VJ

aP
IRRIGATION .ENGINEERING
BY

(. R. SHARMA
P. s. E. Class I ; C. E. tHons). (Roorkee) ; M. I. Eo
P. W. D. Irdgation East PDnjab.

VOLUME I

INDIA PRINTERS,.
BOOKSELLERS $: PUBl.ISHERS.
, '

TRAINING COLLEGE ROAD. JULlUNDUR.

INDIA
 Printed by
Sh: Kahan Chand at the Steao Press,
Putligllar, .A.MRITSAR.
Published by
It. R. Sharma, P. S. E., Cla.ss I; C. E. (Hons). (Roorkee); M. I. E. (India),
Punj~b Irrigation. '
AmrI&IJar Agen1s:-
. Jaswant Rai Gyani & CO.,
Hall Bazar, AMRIrSAR.
 FOREWORD.
By
A.M.B. Montagu,
IS. F •• C.I.E" M.l C.E .• A.C.GJ.
Chief Engineer, Irrigation works, Punjab.
I look upon' it as a great privilege to write the foreword to Mr. Sharma's
comprehensive work on Irrigation. I have known Mr. Sharma and his work for a number
of years and have always been impressed by his keen interest in the physical foundation of
irrigation practice. But he does not confine himself merely to prinf'ip\es of hydraulics lllvolved
in the art and practice of irrigation He has made a prfllonged study of all mc:.tters that can
be said to touch upon irrigation in general. Even the revenue side received his careful and
zealous attention.
In many respects. Funjab irrigation leads the world. Without irrigation, the Punjab
would still consist of a few strips of land contiguous to the rivers, upon which the anxious
cultivator would sow a crop uncertain whether or not it would reach malurity. Today, by
virtue of its canal systems, the Punjab is the granary of India. Valuab e crop are now grown
in areas which were originally arid waste whereon nothing was found t ut thorn and the camel
which throve upon it. During the last few years, no less than 14 milllOn acres of crops have been
matured by irrigation water from the fiftt::en separate canal systems of the Province.
The actual distributinn of the available water tf'tween the canal systems has
necessiated the construction of feeder canals which convey surplus waters of one river to a
river in deficit Distrihution nmong the iJl(livldual channels of a canal system IS a relatively
simple matter. Far more difficult is the o'stributi()n among the Individual cultivators. In the
Punjab, the greatest athntion is concentrated 11pon this di.strihution. Where thi, distribution
Is faulty, cultivators promptly bring facts to notice. The individual irrigation officer is
concerned not only with distributiorJ by existing known methorls, hut in many cases utilizes
his scanty lei~ure in endeavours to secure fresh forms of outlets which will still further improve
and stabilize the distribution to the cultivator's watercourses.
Among the~e energetic and enquiring gentlemen, Mr. Sharma occupies a high place.
His work on outlets is well known. His activities in other directions have not perhaps received
all the attention they deserve. This book is a compendium of all knowled!!e aVd.ilable to him
on the suhject ano will unquestionably prove of the greatest value n'1t only to the practising
irrigation Engineer but al"o to the student who is preparing himself to serve in that great
service, the Irrigation Rranch of the public Works Department, Punjab
It is but natural that many of the views put forwarrl I,y t e ~uthol are coloured bv his
Punjab experience. Reailers foreign to the Punjab may concludt' lha,t unone weight is laid
upon the Punjab practice ard Punjab views. The ans\\er to any ~uch implied criticism lies in
the fact that the Punjab is very confident of its c::tDacity to design. construct
and arlministff irrigation systems of the largest size. In many w"},!;, such as
distrihution of water already mentioned, the des;ign of major works on the
shifting sands of rivers, the economical design and m~tnan(e of relative Iv
minor structurei, the Punjab Engineer may justifiably h(lld that he is to the fore front nf
modren practice and knowledge. In other directions, the Punjab Engineering is first to
recognize that he can l,..arn not only from other countries of the world. but also from other
provinces in India It is probable that the construction of high cams will snon occupv the
attention of the Punjab. E xpf'rien 'e ill this field is negligible and we turn to other Countries
notably the United States of America for instruction and gUluancc! III thi~ field. The Punjab
is behind other countries in the use of machinery, but it is probable that post war development
will impose an advance in this direction also. S
 ii
In many ways, the advance of kn'lNledge in our relatively restricted field, is phenome-
nal. For example. in the mltter of pl ll)iag water ton the su~-soil water resevoir by means
of tUbe-wells, it is p.)ssibla tnat the praGtij;e described in \1r Sharma's book may have advanced
by the time these VOlll:n~S are in th.e huh of the public, but for many years to come the
student of i,rigation engineuing may turn with cJilfirljnGe to the3e pages for a useful and
comprehensi're gtlide to irrigatiJn puctice. with spe~ial reference to the Punjab.

Lahore, ; A.M.R. Montagtt.

15th January 1945.
 PREFACE TO 1st EDIITON.

The science of Irrigatio) Engineering in India has developed by leaps and bounds
during the last two decades. Th:ltrrigation Works of India, especially of the Punjab, oc.upy
a prominent place in the world, both from the point of view of the academic interest as a
science and of the successful results of large irrigation engineering projects. Since the epic
books 'Irri~ation 'vVork5' by Bligh, 1907, and Irrigation Pf'cket Book. by Buekley, 1911. were
written, the development of the Science in India is confined to the technical papers co ntribted
by the eminelllt Engineers available in the proceedings of the various Engineering Sccieties,
such as the Punjab Engineering Congress, Lahore, tht' Institution (If Engineers India, Cenrral
Boaro of Irrigation, Bombay Engineering Congress, the departmental technical papers of
P. W. 0, Punjab and Bombay and the research publications of the Punjab Irrigation Institute
and Central Hydro-Dynlmical Research station Poona, Bombay.
The purpose of this hook is to present the science and the practice of Irrig::ltion
Engineering in a concise form comprising practically all the modern developments. The book
is essentially ment to be used as a Text RO'Qk for the students preparing for the Engin~ering
Degree Examinations of the various universities in India and otRer c0mpetitive examinations
of the Central GOVErnm"nt. Selected examination qU!3stions usually set in Degree
Examinations of the various universities in india are giv=n at the end of each Chapter. Some
examples of the typical design .. of irrigation Works have reen worked out for the guidance of
the students. Since irrigation Engineering (distinct from HJdraulics) is not a suhject taught in
the British or other European unive 'sities, no text booh dealing with the subject are available.
This book is intended, therefore' to meet a lqng standing need of the student comm~nity in Tndia
The book also deals with the actual practIce of science in the field and is meant to
be used as a refel ence book I ha ye, therefore, attempted to cover the require men ts of the
students in the examination que .. ti0ns (covering the syllabus of the v:'lrious universities in India
and the competitive examinations of the Fedral public Services Commission). In th~ Chapters
which are considered beyond the scope of the studen's, no questions are given at the end
The subject of Irrigation EngiKeering has developed mathematically so much that it
g~es beyond the capacity of an average stu<~ent or a practiC::lI Enginee, to cram up all the
formulae. In aI, accompanying volume, the diagrame gen' nlly used in Irrigation practice are
given which can be used by the students and practical Engilleers. for solving the problems
relating to the design of Irriga!ion Works. .
The book i" devided into six parts and forty four cha.pters. The first part, comprising
two chapters, deals with Lift Irrigation, the s~cond comprising 20 chapters deals with Flow
Irrigation, the third ?ompri~jng six ch_apte~s deal; ~ith Ta.nk . !rrjg~tiQn the fourt~ ?omprising
f our chapters deals WIth DraInage Engmeenng-, the fIfth ccmprtsmg SIX chapters dealsrivlth Ground
Water Eugineering [water· logging] and the last part camprising six chapters deals with general
informa.tion usually required in engineering practice. The fourth ::Ind fifth parts comprise the
subject which is usually definer{ as Hydrology in American practice
Obviou>ly the whole of such a comprehensive book cannot be original Detailed
references to Ihe publications cOIisulted have been mentiond in the text. The list of author~
is SO large that it is not possihle to acknowledge gratefully the help and the U8e of their work
by naming to them individu 11ly. Similarly the refences to the proceedings of various Eagineering
Societies, referred to in the text are gratfful 1y acknowledged.
The help rpndered in compiling. editing anti improving this Book by the following
Irrigation Engineers of the Punjab is thankfully acknowledged: [lJ R::Ii Bahadur B. N, Singh'
T. S, E., retired Chief Eng-ineer, punjab Irrigation. [2J Rai Bahadur D. K. Khanna I.S.E.
Superintmding Engineer [now Chh~f Engineer Irrigation Works.] [3J Rai B::Ihadur B 1.. Uppal
I.S E., retirerl. Superinteding Engin Eer Punjab Jrril!ation. [4J Rai Behadur Kanwar Sain I.S.E,
Superintending E~ginee~ [5J Rai Eaha?ur B,~. Kap~r, Director ~rrigati:n Resear.ch, Lahore. [6]
Rai Bahadur Hakim Rat, I.S.E. Snpenntenrlmg Engmeer [7[ ]atIDder Smgh, ASSIstant Engin~er.
~uld [8] K.S. Pdthak Assistant Engineer
 iv

The author is '-'extremely grateful to the Honourable Minister of Public Works

Department (now premier of the Punjab) Malik Khizar Hayat khan Tiwana and the
J
Honourable Member of Revenue in-charge of Irrigation Department, the late Sirdar Bhadur Sir
?undar Singh Majithia, for se lectillg the author to do spade -work of Civil Engineering teaching
In the Punjab as the first professor and the h .. ad of the Civil Engineering Department at the
Maclagan Engineering College, Lahore. The author availed of this opportunity to study the,
subject as a whole and to compile his lectures in the form of this Book. He is also grateful
for the encouragement received from the Honourable Minister of Kevenue, Rao Bahadur
Chaudhri Sir Chhob) Ram.
The author's thanks are also due to A. M. R. Montagu, Esquire., C.LE., ChIef
Engineer, Irrigation, for the trouble of going through the whole Book and then writing the
foreword. Extremely useful and valuable suggestions to improve the Book were kindly supplied
by him.
. Last' but not the least. the Author is specialy indebted to E.S. Crump, I.S.E; C.I.E
Retired Superintending Engineer.under whose guidance, the author carried out experimental
research work for more than 5 years.
Amritsar, K. R Sharma,
27th May, H:l44.
Preface to 2nd Edition.
. Part III of this book dealing with the Storages and Dams [Tank Irrigation] has been
revised and enlarged. It comprises now six chapters instead of four in the 1st edition.
. Design diagrams' are bound 1D volume III. These are intended to be used by the students
1D the class room and could be issued to the students in the Examination Hall for the soluticD
of the problems dealing with practical designs.
. The help rendered by Dr. J. K Malohtra, Ma.thematical officer Irrigation Research
Instttute East Punjab in editing the 2nd edition of this booN is gratefully acknowledged.

Amritsar. K. R. SHARMA.
1st. Decem ':ler, 1948.
 NOTA nONS USE~ :rHE BOOK' - ,
(A) Hydraulics.

Steady flow is tbat state of flo'.v in a stream where the discharge across any defined
section of the stream remains constant in respect of time. .
. (Tni'orm flow. Uniform flow is steady flow in a stream when the depth does not vary
wIth constant d;~ch<lrge.
Non U~litorm How. When depth varit,s in 8teady flow in a stream with constant
discharge it is non-uniform flow.
Normal flow is that shtll of steady flow of a stream, where the fall of water surface
corn'sponos to the consumption of energy by friction. It is stable. The principal symbols are
subscrIbed with the letter 'n'. to indicate this condition (This is also called Neutral How).
Regime flow is that state of stream, flowin a in self-borne alluvium, where there is neither silt
nor scour. Regime flow also rostulates normal flow as a preliminary condition. The principal
symbols are subscrihed with the letter 'r' to indicate this condition.
Critical flow. When water flows according to Bernoule's Theorem there is a particular
condition of the flow, which is of spicial interest and is known as the critical
condition of How. This condition of flow is more simply called Critical HoW. The depth
and velocity are also called critical, Mathematically it is defined to be the condition of the
flow when the change of velocity head is just sufficient to produce the change in elevation
of the water surface. It has certain properties for which the student should refer to l!ar~ VI
Chapter IV paragraph 4. The principal symbols are subscribed with letter 'c' to llldlcate
the condition.' .
Constants.
C Arbitrary or experimental co-efficients.
fL Lacey's "silt factor"
g The gravity constant.
K Theoretical constant in the "free fall" discharge formula K=-vg-(iJ 3 /!=3.0888.
N The co-efficient of rugosity of a channel. .
Na Lacey's co -efficien t of Roughness in a channel.
w The weight of 1 cubic ft. of water=62'S Ibs.
W .~otal weight.
Discharges.
Q The dis~harge in cubic feet per second (cusecsl of a channel or work.
q The discharge in cusecs per foot width of a channel or work.
Energy.
E The total energy expressed in feet head of water above a fixed datum. When
plotted, this depicts the "total energy line".
The 'energy of flow' expressed in feet head of water above the bed. When the total
energy line has been plotted. the "energy of flow" is depicted by the intercept be-
tw.een the bed line and the total energy line. (In hydraulic books, the term specific
energy of flow is llsed instead).
Ec Energy of flow in critical conditions of flow.
h The head or energy required to produce a velocity h ==~
H Depth on crest including velocity of approach head:s:G+ha
Ha Afflux bead
ha The hea? or energy equivalent to a velocity of approach Va
Hw The. aval1abl" working head, i. e. the difference in total energy levels between two
sectlons.
The minimum working head required between two point
 vi

Hl The head lost between two points, due to aU cayses e1l1Cept destruction in a standing
wave.
HL The head or energy distroyed in a Hydraulie jump.
Velocities.
The mean velocity in feet per second of a stream at any section.
The mean "velocity of approach"
The mean velocity of normal [Neutral] flow corresponding to Dn .
The mean velocity of regime flow corrpsponding to Br . .
The velocity of flow, at a section with critical flow.
Kennedy's standard silt-charge velocity Vo =0.84 Dg·64

x = ~ =critical velocity ratio [Kennedy's]

Vo

Dimensions.
L The length of a channel or work measured along the line parallel to the direction of
flow between two points.
Lt Tl.l.e length of parallel sides of the throat of a weh:. flume. etc.
B The length of a channel bed transverse to the direction of flow. i. e. the width.
Bt The width of the throat of a weir, flume etc.
D Th@ depth below the surface of a strpam ber! at a stated point.
Dn The normal depth of a stream corresponding to Vn •
Dr The regime depth of a stream corre!'ponding to Vr.
Dc Critical depth corresponding to Ve.
A The cross sectional area of a stream at a stated point.
Pw The wetted perimeter of a cheRnel.
Bs The top or surface width of a stream.

R The hydraulic mean depth of a stream=-~.

Pw
Miscellaneous.
S The actual slope of the total energy line at any given point.
Sw The aetual slope of the water surface at any given point. .
Sh The actual slope of the bed of a channel at any given point .
.")0 The slope of the total energy line in the case of normal flow.
Sr The slope of the total energy line in the case of ngime flow.
f Co-efficient of friction
s The side of at rapezoidal channel.
G The gauge rpading: Zero of gauge must be specified.
p \.:fV· The total pressure on a cross section of a stream.
p The intensity of pressure at a stated point.
r •t The instantaneous radius of cnrvature at a point in general, aDa lor thtn_
streams ......... of the bed; exectly ......... the mean radius of curvature of all the filamen~. in
a stream,
(B} Hydrology

Areas in square miles:-

A Total area of a catchment,
Ae Effective area of a catchment.
Ao That a.rea for which dispersion is uuity or thlt ateawllldh can be Wholly eovered by
. a storm with unvariable intensity of storm. .
As . Area covered by a particular storm,
Ah Area between two isohyetals.
Ag Area of in1 1uence assigned to a. rain-gange station,
 yii

R?1infall or precipitation in inches:,

P Mean annual total rainfall over a catch ment.
Pa Total annual rainfall over a catchment for any year.
Pm Total monthly rainfall over a catchment fur any month.
Pt Tetal rainfall during an interval 't'.
Ps Total rainfall of a particular storm as recorded at a rain-gauge.
Ph Mean annual precipitation between two isohyetals.
Run-off [for volumetric studies} in inehes:-
R Mean annual total run-off from a catchmeBt.
Ra Total annual 'run off from a catchment for any year. _
Rm Total monthly run-off from a catchment for any month:
'Ra, Total run-off of a particular storm.
Discharge for [intensity stt~dies] in cusecs:-
Q Maximum discharge from a catchmeNt.
Qs Maximum diseharge from a catchment on account of a particular stOl:'m,
Qo Maximum discharge from a catchment Ao.
q JJi5charge per square milt: of catdllnent.
Time in hours:-
T Inlet time.
t An interval of time.
st Duration of storm.
Temperature in d~grees [Fahrenheitj:-
o Temp~rature in degrees Fahrenheit.
Ba Mean annual temperature over a catchment.
Om Mean monthly temperature over a catchment.
M is,ellaneous;·-
L Distance in feet 01 watershed from the stream along the line of flow.
B Absorption into the soil in inches of depth per hour.
E Evaporation in in~ 01 depth per hour.
D Total loss in inches per hour=E+ B.
I Intensity of rainfall.
Im Maximum inten;ity of rainfall.
S Maximum slope of catchment from watershed to the dlainage.
F Reduction in, inches due to rain initially held by trees, crops and undergrowth.
To determine mean annual rainfall of a catchment:-
Exmapels,
(a) Isohyetal method;-
p=~ Ph.Ah =;;S Ph.Ah
;;S Ah A
(h) Weightage method:-
Extra notation. 1
't - ~ .::'".; 1" ;-
, :, < I .'

Pg Precipitation at any rain ..gaug8!~tation~·"

P=~ Pg.Ag _:2 Pg.Ag
~Ag A
('e) The straight average method {ior a large plain area} ~'In '.. , .

p A1Pl+A 2 P2+ .' An~

Al+A 2 + An
 viii
q

Uld Pl ,
Note,-The total area is divided into n sub-divisions,
Pa• Pn
':.i ...
Al , A2 .............. An
represent the average rainfall for eaeh sub-division.
Run off formulae (volume)
(a) Vermeule formula:-R::::P- (11 +O.29P, (0.358--0.65)
. :J:J
(b) Khosla's formula:-R=P- {} +C
2
Extra notation .,
C=A constant which allows for catchment characteristics humidity' glacier contribution
etc, but not for absorption, evaporation and transpiration covered by the
tempera.ture factor 2'
e

MaximHm discharge from a catchment (int;nsif/l): -

(a) Inglis formu1a:-

Q= ?~OOA for fan-shaped catchments

'vA+-i
And Q=7000,/A- -'Z40 (A-lOO) for an etongated catchment.
(b) Khungar and Gulllatj's formula:-

Q=645AoZ;:nax - F C Ae
T Ao
t
m Index of disper~ion,
Zmax Maximum height of a theoretical hydrogra.ph for a rainfall of ma.ximum possible
intensity in the catchment,
(e) Flow of water through sub-soil under weirs and dames.

Symbols:
H Head in feet or difference in water levels upstream and downstream of a work
[Percolation head]. '
P Pressure head in feet in a pressure observation pipe measured above the downstream
water level.
Pc Pressure head at point C.
G Gradient, or rate of change of head.
GE Exit gradient.
t Temprature in-degrees Fahrenheit.
p Density of fluid.

fl Viscosity= '47-1~~~O?k~()f+~000682 t 2

·
K merna t'IC . 't y=
VISCOSl fl-
p
Q The discharge in cusecs of a channel or work.
q// The discharge in cusecs per foot of width of a channel or work.
p
pressure head expressed as a percen tage of the total head = -Ir- X 100.

Determination of exit gradient:-

(a) Floo], with pile at downstream end:-
 JX

Extra notations,
d Depth of pile line below floor surface.
b length of impervious floor.
b
=(i
l+V-I~
2

GE =EXl't gra d'lt'n t =(1

H • '~
... 1 A. -

(h) Floor with pile at downstream end with step:--

Depth of pile line belo,"" upstream impervious flolr.
Depth of pile line below downstream pervious floor.
Length of impervious floor upstream.

--~c----
H
[d1 -d 2hl A. •
K
l-K'
. where K is given by the equation

V l-K2 Cos- 1 K= . ?t~_

K d 1 -d 2
Pressures at different points one pile line or depressed (laor:-:
d1 Depth of pile line below upstream floor.
d2 Depth of pile line below downstream floor.
bl Length of impervious floor upstream.
b2 Length of impervious floor downstream.
b bl+b a
~E q, at junction of upstream floo~ pile line.
,pD q, at bottom of pile line.
+c q, at junction of downstream floor and pile line.
Mutual interference of piles:-- .
C-19 d_2~ d1 +d,
- vb' ' - b -
C the correction to be applied as percentage of head.
b' the distance between the two piles.
d, the depth of pile whose influence has to be determined on the neighbouring pile of
depth d i .
dl depth of pile on which the effect of pile d 2 is sought to be determined
h total floor len~th,
 x

(DJ Dimensional Formulae.

Common Dimensional formull
Name of quantity, Symbol. English in terms of mass,
Unit length and time.

2 3 4

Geome ricai quantities.

~l
Length (any linear dimension)
Depth
Head . H I
Diameter . d }- ft. L
Raf1ius . r I
Hydraulic mean df!pth R I
Wetted Perin.eter . pwJ
Area A (ftoll! Vt
Volume . . V (ft.)1 ~'-i ,~" I.V'
Kinemalical quantities. i,
ime • T second T
!;~f'\ ~~ '."'.
Angular velocity . 1i! Radians/Sec. lIT
Angular acceleration • a Radians/Sec.1 Tlrs
.Velocity • v ft/sec. LIT
Acceleration due to gravity g ft./sec.' LIP
Lacey's silt factor fL ft./sec. 3 LIP
Kinematic Viscosity • • =IL- 1'~ locii
ft.'/sec. L2JT
p
Discharge Q :ool! i"
ft a/sec. vJr
Dynamical quantities.
Mass. . M Mass M
Surface tt n~ion v to./ft. M/T'
Viscosity . • (-l Hi. X ~ec,{ft.2 MILT
Pre'sure (unit stress). p
Modulus of elasticity.
:--.hearinp; morlulu$
e
es
1
J
lb./ft.' M/LT'
Density p Mass/ft. 3 M/L3
, BI
Momentum ill Mass x ft./sec, ML/f
Force F)
Weight Wr 11> ML/P
J
Torque Pq 1 MVII
Energy E l- ft.x lD. -rs
Wo1'.l& EwJ
(E) Greek Alphabets.
0: Alpba ,{ Laulbada 1.1 Upsilon
13 Beta fl Men t Xi
'Y Gamma t!-: Neu 1] Eta
a LI Deft::!. (Jw Omicron ...l. Zeta
E. Epsilon 7\ Pi x ChHki)
/J Theta p Rho (j!) Omega
i iota ~ Sigma ¢ Phi
iT Kappa T Tau V. Psi
 Author's other Technical Papers.

1. Underpinning Satghara Rest House, L.J.C., Paper No. 143. Punjab Engineering
Congress, 1931.
2. Remodelling of' Mithalak Distributary, Paper No. 154 Punjab Engineering Congress, 1932.
3. Effect of outlets on the Regime of a channel. "Indian Engineering" Calcutta October
and November 1932.
4. Silt conduction by Irrigation Outlets. Paper No. 168 punjah Engineering Congress, 1933.
5. Design of an A.P.M. Paper No. 176 PunJab Engineering Congress. 1934,
6. Design of irrigar.ion channels paper read in "Institution of Engineers India" in 1935
and copiE's placed in the Punjab Irrigation Libraries received vide Superintending Engineer
U J.C. No. 962 dated ~8th July 1937.
7. Silt selective Distributary Head Regulators. Paper No. 189 Punjab Engineerin5 Congress
1936.
8. Design of R.C. Extensible Bridges i0r Drains, August 1936. printed copies placed in the
Punjab Irrigation Libraries vide Chief Engineer':> No. 386, dated 29th January 1937.
9. Standing Wdve or Hydrauiic Jump, "Indian Engineering, Calcutta" October 1936.
\0. Determinat.ion of water profIles over Hydraulic Works. Indian Engineenng. Calcutta
19l~.
11. Experiments in Spepage Losses from canals. Paper No. 209 Punjab Engineering Congress,
1938.
12. Automatic Silt extractors without loss of water. Indian Engineering, Calcutta, February
1938.
13. Cross Darin'lge Work (Auto-Suction Weirs) Indian Engineering, Calcutta. Febru:try 1938.
U. Improved adj ustable proportional modules and open flume outlets. Punjab Engineering
Congrflss 1940.
15. Model Experiments to determine suitable length/of skimmig platforms on models of
the Head Regulator at Rasul, Punjab Engineers' jo\unal, May 1939, Lahore. .
16. Theory anti physics of seepage flow from canals. Paper No. 231 punjab Engineering
Congress, 1940.
17. Use ot P0rtable Flumes for water course discharge Measurements. February 1940 Punjab
Engineers' Joumal, Lahore.
18. Minimum M )dular Hea.d 1M. M. H ) for an Adjustable Proportional Module, January ~940
Punjab Engineeri' Journal, Lahore.
19. EXDeriments to determine" frue Basic Sub-soil Pressure at R. D. 299,000 U. C. C
(W. l. R. Bib 16).
2U. Metf'r experiments to determine Losses in Mangtanwala Feeder (W. I. R. Bib 5).
21. Tank Exp.erim"nts to determine laws of seepage losses at R. D. 348,000 U. C. C.·
(W. I.R. Bib 71.
22. Determination of Seepage Los;es in U. J C. from Rashidpur to Shadiwah from meter
discharges. (W. I. R. Bib 14).
23. Trough experiments to determine Seepage Losses by Point Method in U. J. C. (W. I. R.
Bib 44).
24 Sub-soil con(litions in adjoining fields of Thur anG cultivation.
25. Determination of losses in Low.:r Chenab Canal by taking Seepage Discharge Observations in
a closure (W. r. B Bib 21)
,26. Determination of losses in Uoper Cenab Canal by taking Seepage Discharge Observations
in closure. (W. I. R. Bib 21)
27. B. S P. Experiments at Khanki to detllrminp. B. S. p. costours in a complicated case
of sources and sink and to oetermine True Soil Pressures in River. (W. 1. R. Bib 17).
28. Report on Seppa~e Los,es of U. C. C. & L. C. C. (years calculated and actual compared by
months). (W. 1. R. Bib 31.
29. "Hump" Investigations unoer c'lnal.
30. Point Method Apparatus and its use on L.C.C. Distys: to determine their losses.
 X ..
,;

31, Plotting and analysis of 39 Daily B.~.P. Stations (W l.R. Bib 43).
32. Plotting and analysis of B.S.P. PIpes along canals calculations of losses by using
observed S.LC. (W.I.R. Bib 18).
33. Capillary fringe and soil evaporation st1ldies with Hydmdynamic soil Pressure Observations
(W.I R. Bib 42,.
(a, Containing fore ward by Mr. Blench.
(b) Detinitions.
(c) Note dated 5-6-39, 15-12-39 on ,- Hydrodyna:nic Soil Pressure observations at
Lahore.
(d) Note dated 8-6-39 on "Experiments at R,D. 180,000 L.C.C. to measure evapora_tion
from the .bed of a pit.
(e) Note dated 15-9--39 on "Capillary fringe and Rise of water level in Bore Holes at
R O. 1S0,00a L.C.C:,"
(f) Note dated 20-9-39 on "Soil Pressure Observations· to determine evaporation
gradient by Progressing Lowering of ben of pit at RD. !80,OOO L.C.C."
(g) Note dated 16-11-39 on "Hydrodynamic Soil Pressure Observations and Capillary
Fringe studies at RD.' 150,000 LC.C.
(hI Note dated 25-'7-40 on "Hydrodynamic Soil Pressure observati()Ds and capillary'
fringe studies opposite RD. 46,000 Chichokim~llian Distributary." .
(i) Note dated 2-1-40 on "Technique of Hydrodynamic Soil Pressure observations below'
the effective saturation line."
(j) Note dated 25-3·40 on "Capillary Fringe studies at Hudiara Nala."
(kl Note dated 26-340 on "Hydrodynamic Soil Pressure observations at Sheikhupura site.
34. "Division of the Punjab" by an engineer. June 1947. .
 IRRIGA~~ION RNGIN~ERING.
Volume •.
Table of Contents Page
Foreword by A.M.R. Montagu C.LE , I.S.E., M.LC.E., A.C.G.I., Chief Engineer, Irrigation Works Punjab, iv
Preface by the Author IX
List of other Technical Papers by the author iii
Table of contents xi
Introdu;ti0n 1
PART I-LIFT IRRIGATION
Chapter i-Open Well I r r i g a t i o n . , 3
Definition. S:.lUr~e of Water Supply,location of w~!l" d.scharge of Irrigation Wells, suitability of well water
for crops, classes of wells, construction aud sinking of wells, methods of raiSing water, duty alJd delta
of well wat"er. cost of Well Irrigation, Well Irrigation versus Canal !rrigataion. Questions.
Chapter ll-Tube well Irrigation 13
Tube wells and their object, water bearing or non-water bearing formation, water dividing, History of tube
well practice for Irri;.:ac;on, Boring of tube "'ell and different methodS, percussIOn rope. bv<ing, "nchor bolt method,
Rod boring, Boring by hydrauliC wash, Kotary borin", core drills anEi stove pipe method; tube well pIpes; .and
blOwing during boring, Fishing tools; types 01 tube wells; suitability of tract for tube Wf'll irrigation, Tube wells
chaks, selectiGn of pump sets, selection of strainers, ,hoking of strainers, size of tube well pipe and strainer
length. Different op,rations of sinking a tube well, eccentricity in the bore and its correction; Back washing dnd
back blowing of strainer wells, Tube well tests, ExtractIOn of pipes and strainers of failed wells, const! uction of
Kiosk, w. courses, sumps, operator quaters, assessurem and water rates. Discharge from a tube well. condition
ot tube well irrigation in the Punjab and the United Provinces. Financial espect of tube well irrigation, canal
versus tube well irrigation, Examination questions.
PART. II-CANAL IRRIGATION.
Chapter I--Classification of Canals. 42
Irrigation and navigation canals. classes of Irrigation Canals, fin~ classifiication. receipts from
irrigatirn canals, parts of canals, Punjab Canals, distribution of river supplies to the Punjab Canals
Functions of Canals, Questions.
Chapter H--Principal crops and Assessment. 49
Introductory. Soil conditions in the Punjab, rainfall, principal crops in the Punjab, fertilizers and
manures, crop diseases and their cure, aconomy of water. Snb-Irrigation Internal distribution, Units of
Measurement of areas, asses"ment metbods, record of irrigation, special charges, Miscellaneous Reccipts
Duty of Canal Water, Efficiency, Questions.
Chapter III--Hydraulics and Control of Large Rivers. 62
Introductory, Major Divisions of River Channel, torrents and streams, river regime Theories by
Molloy Oldham etc. meandering and avulsions of rivers and their extent, erosion, swirls. silt and velocities,
in Punjab rivers, reclamation works, spurs and groynes (Denhy's). object of guide banks (Bell's Bund) an':!
extent of narrowing, their desigr. and construction, retired embankments. marginal Bunds an 1 ;;purs, pitched
islands, I" yout of river embankments. cross sections of emba'lkments, Question.
Chapter IV-Head Works. 87
Introductory, selection of site, layout, types of weirs, afflux. ponil level, waterway for weir, effect of
weirs on the' regime of river, undersluices. divide wall, fish ladder, head regulator, silt control at head
works, guide banks and marginal bunds. moveable weir crests and shutters. Gate lifting arrangements.
Chapte. V-Design of Weirs on Permeable Foundations 115
Introductory, Theory of design and its development, parctical weir design", determination of uplift
pressures (Khosla's theory), theory of exit gradient, curtain walls, and sheet piles, Khosla's design of
modern weirs, wing wall design, examples of weir design, Questions.
Chapter VI-Design of Irrigation Channels. 151
Introductory, Hydraulic formal<B, Kennedy's sih theory and diagrams, Wood's normal data of Kennedy
..hannels, Lacey's silt theory and calculations, Lacey's diagrams, examples, summary of Lacey's theory
comparison of Lacey's and Kennedy's theories, Author's design of irrigation channels, velocity and silt
distribution in a channel section, diurnal anQ. seasonal variation in silt charge in a channel, silt atrition, silt
movements in a channel, ~ilt sampling and analysis, corelation between silt and hydraulic data, weed growth
and its eradication, lined channel section deSign, super-elevation in water roads, Questions.
Chapter VII - Design and Maintenance of Canal Banks. ... 19'i
Typic~l Distributary main and branch c"nal cross sections, earthwork specifications, puddle, earthwork
repairs, construction of high embankments. sand and puddle CQres, berm formation, strt'ngthening of ba.uks by
silling tanks. acciden~s to canal banks, closing breaehes. Questons.
Chapter VIII--Lining of Channels. 209
Advantages of lining, suitability, types of lining. construction gf brick lining, compaction of back fill and
banks. drainage of storm water. lining in reaches with high spring level, l06ses from lined "hannels, Questions.
Chapter IX-- Cross Drainage Works. 217
Masonary aqueduct, siphon, supp",r passage. level crossing. author's auto-suction wiers. calculations of
an acqueduct design, Questions.
Ghapter X-Falls and Meter Flumes. 232
Definition, Necessity of fa.lIs and their location, fall design practice, classification of falls, destruction
of energy of falls, methods of destructicn of surplus energy D.S. of falls; contracted versus full width falls
falls with cistern or glaCis, cistern design, trapezoidal notch falls and their design calculations, Montagu type
fall and design calculations, Inglis type fall and design calculations, meter flumes, Cramp's design of meter
 xiv
flume, author's design of meter flume and calculations, examination questions.
Chapter XI-Silt Excluders and Ejectors 268.
Intr\9duction, definition, approach channel to a silt extractor, escape discharge above extractor, approach
design of the extractor. tunnels and their design, extractors C',m meters, effe-:t on canal regime, efficiency
of silt extractors, tYl'ical design and its calculations, Author's automatic silt extractor without loss of water,
examinat;on qupstions,
Chapter XH-Fluming and Headless Meters 282
Definition, cla,sification ot flumes, hed profile de~ign, upstream approa"h design, theory of departure
or expansion, Burkitt's method of expao:-;ien, athor's method of downstream splay, examples with calculationr
and desi~n of flumed acqueduct, flumed bridges, contraction with depressed floor, Burkitt's Headless Meter
Flumes, Author's improved and economical headless meter flumes. weir cum spillway, syphon flume, surge
tar,k" <11,d chambers and rigi,i tOP flumes,
Chapter XIII - Syphon Spillways and Hydratomats 303
Introducticn, discharge formula of syphon spillways, types of spillways, priming methods and their
comparison, de"ign vf syphon spillways for High Dam£. graphical method of desIgn. effect of splllw&\'s on
dams, ice trouble in sp.llways Hvdratomats, six example cf design of various, ype.; of ~yphons with calculations.
Chapter XIV-Di,trioLtary Head Regulators and Distributors . ... ... 329
introduction undershot distributary head regulators, over-shot regulator, (Wood's rising cill gates),
Cribb's d:stributary Head Regulator, King's "ilt vanes, head with skImming platform, Cantilpvered skimmer,;.
Author's Slit sE'lective distnbutory Head Regulators, Ex,mple of de·ign of SIlt ~elective Hedd Regulators,
proportional di,tributors, field tPsts of silt excluding devIces, examination questions.
Chapter XV - Outlets and Tail Clusters 340
Tvpe of modules, ctefin,tion, modular heRri, non-morlular outlets, tilted pipe outlets, Kennedy's Gauge
Outlets, Harvey Stoddard improved outlets, Kirkpatnrk module (Jamrao type). Crump's adjustable proportional
module. Author's improved '. P. M (submerged semmJf)duh oTJflce outlet S. S. 0 0.), Crump's open flume
outlet, ~ utnor's weIr flume and n.rrow opE'n flume OUI Jets, Pipc-cum-semi-me:>dule (double module outlets),
rigid modules (Kent Glbb Khanna and GhHffur), tail clusters, watpr-course discharge observat,ons, Cipollette
"elr and Author's detachable and protable flumes, outlet chaks, examiloation qestions.
Chapter XVI-Irrigation Projects 361
Introduction, prellmina. y survpvs, surveys ano alignment, location and layont of Headworks, longitudinal
sections ",nd cross "eeti"ns of mian can. Is," land plans, Inca' ion and alIgnment of distributaries, area, command
and draw-off scatments, chak plans and alignment of water-courses, masonry works, estimating and rules
of preparation of project estimat" and construction, direct and indirect receipts and financial outturn, plans·
required in Ir·rig-ation Projects, examination questions.
Chapter XVII--Remcdelling Channels, 482
IntroductIOn, necesslty of remodelling channels, hench marks and hydraulic surveys, history .of channels
di3.gnosis of troubles, ch'tnnel sectin'J~, outlets, head regulotor, drowned bridges, J,-,wering of channels, eontrol
paints and metprs, strengthening banks, silt clearance and berm cutting, watching channel after remoaelling,
H_ registers and characteristic curves, examinatio.\ questions.
Chapter XITIII--Inundation Canals _ 391
Introduction charcteri'tics, ~ilting canal head re"ch, changes in head reaches, selpction of off-take
subsidiary tu a'is "unns at h"ads, miscellaneou< works. P. cl"xton's theory of Inundation canals, concluslOllS,
examinati 1\ quest om,.
Chapter X,X--Discharge Observations anc. Regulation 400
Intro<luction. liver gauging sites, hill flVer gauging sites temporary and permanent gauges, metho,ls of
stream gauging. v~loclty area method. s"gmentation, varIOus methods of velocity observations, current mt'ttH"
upkepp of nwte,s, .e-ratin~ of current meters, river discharges by stretching ropes, river discharge by pivot,
point methOd. nver dischRrge by current meter3, ~anal discharge observations (tirFt, second and thrrd ClaSS)
spec al prpcantions for cahlbration dischiirges, preparation of discharge tables, Tegul;o.tion, Questions.
Chapter XX--Water Power from Canal Falls. 422
Introduction, power ava;lable from canal falls, P~' to< of hydroelectric installation, design of head raee
tUI bine pit, draft tube, tail race, inlet bpnd etc in a hy<lroalectric installation, selection of eqUipment, suction
pip" design, specific speed and its use in <lE'sign q with examples. genpra 1 principles of hydraulic turbmes, number
of units of hydroelectriC plants layout of H. E. plants, interlinked hydroelectric developements of canal
falls, flour mille; run by turbines, gird wheel flour mills, their design and layouts.
Errata of Volume 1.
Volume II.
PART III-TANK IRRIGATION (STORAGES & DAMS)
Chapter I-Storage Reservoirs or Tanks.
lntroduction, Definition, Catchment discharge from rainfall and snow melt, preliminary investigations and
survey, final investigations. choice of site, capacity of reservoirs, irrigation capacity of tanks, capacity for water
supply to towns, abwrption and avaporation lossps from tanks, silting of reservoirs and their remedies,
auxiliary re,erVOIrs and supply channels, flood absorptiv ... capacity of tanks, reservoir flood routing,
floud control leservoirs, leservoir ope,ations, sptllway di,charge capacity of reservolrs, spillway tuniJels,
emcrgency spillways, biblography, examination questior.s.
Chapter II. Earth Dams.
Introduction, definitions:- Foundation of earth dams, Material of construction of Earth Dam, Practical
Criteria for th ... safety of earth dams. safety agalDst oven, pping, seepage flow net and seepage flow under
dams, Method of dealin~· with excessive seepage through Eanh lams, Prec3.utions to guard against free
passage of weter thlough darns, Remldies to be appled to reduce excessive flow under the dams, Stability
of Hydrauhc fill lams. foundation of earth dams Elastic theory of shearing Stre~ses in earth dams,
:Slip Circle net hod of testing earth dams Taylor's "tability numbers, construction detaIls of ear:h ~ ams
 xv
construction of Hydraulic-fill-dams, observations and studies to watch safety of dams after construction, Waste
weirs, Tower and outlets, Maintenance and repairs of earth dams, failures of earth dams, Examination questions.
Biblography.
Chapter III. Rock Fill dams 59
Introduction, definitions :-T)'pical cross-sections of Rock fill Dams, Foundation charactenstics of a Rock
tiiIl Dam, Design of rock fill dams and its parts. Settlement of rock fiJI dams, Core wall type of Rock Fill dams,
Composite type of rock fill dam, Earth Core type of Rock fill dam, hand packed Rock fill dams, Biblography.
Chapter IV. Gravity Dams (Masonry or Concrete). 69
Introduction. definitions :-Advantages of masonry dams, disarlvantages of masonry dams, Examination of site,
Forces acting on a gravity dam, Silt pressure, ice pre~sure, Earthquake forces Uplift, Wind pressure, Wave pressure,
Requisites for stability of Gravity dams, Design of If)w gravity dams, Profile of a low gravity dam, DeSign of
high dams with Example, method of mf)ments for testing stability, graphical mf'th"d of testing stability, Haes~ler's
graphical method of testing stability, Vertical shear and the Ellipse oistresses in the body of the dam (solved
example). Constructions details of masonry dams, Construction details of concrete dams, Temperature control in
dam~, Failures of masonry dams, Temperature stresses in dams, EXamination questions, Biblography.
Chapter V. Arch and Multiple Arch Dams. 104
Introduction, definition, Arch Dam types, Masonry arch dams, shell theorey of arch dam design, Ela~tic
theory of Arch Dam design, Trial and load analysis of arch dams, Design proceedure of Arch dams, Loads on
arch dams, Investigations of arch dams, Model tests of Arch dams, Failure of Arch dams, Multiple arch dams,
Design of Multiple arch dams, Rib-shortening, Professor Cain's formula, F. H. Fowler's Diagrarus, F. A. Noetzli's
formula, Diagram of working stress intensities, Design of the buttress, Failure of multiple arch dams, Compara-
tive cost of a Multiple arch dam with othtr types of dams, Example ofaxisting Multiple Arch dam, Biblography
Chapter VI. Buttress aDd Reinforced Concrete Dams. ~ 126
Introduction, definition :-Forceb against buttresses, Earthquake loading for buttr<ssed dams, Spacing of
buttresses, Design of buttresses, Upstream face of buttress dams, Shrinkage cracks in buttresses. Masonry buttress,
demand example of design, Shrinkage cracks in buttresses, Masonry buttress dam and example of design, Slab
buttress dams and their various types. columnar buttress dam, Truss buttress dam, Cantilever deck dam, Round
head buttress dam, Diamond head buttress dam, steel buttress dam, Connection of facing & foundation, Buttress
dams on soft foundation, Reinforced concrete dams, Reinforced concrete Dome Dams, columnar and slab design
of Buttress Dams, Biblography.
PART IV-DRAINAGE ENGINEERING
Chapter I. Rain Fall and Run off ... 139
Introduction, hydrology, rainfall phenomenon, definitions, effect of climate, variability of the annual rainfall,
measurements of rain fall, intensity curves, spacing of rain gauge stations, factors affecting run off, IUn off estimate
in a small catchment. hydrograph, combination of hydrographs of various areas Gullati and Khungar's analysis,
run off from large catr.hment. run off from very large catchment areas, examination questions.
Chapter II. Surface Drainage (Open Drains or Ditches) ...... \. 160
Introduction, classification of drains, alignment, capaCity of drains, drain section design, cr<l'Sa.aections,
distance marks, maintenance of drains (weed clearance and slou~hing sides etc.) prevention of silting in open
drains, inlets. discharge observations, extensible and snbmersible bridges on drains, other masonry works,
seepage discharge of drains. history of drains, examination questions.
Chapter Ill. Sub-Surface Drainage 173
Introduction. benefits of sub surface drainage, systems of this drainage, effect of this drainage on
spring level, spacing and their depth below ground level, grades, discharge capacity and size, tile selection
concrete versus clay tiles. laying of the tile drains. outfal!s, freezing and thawing, maintenance of tile drains,
surface inlets. vertical drainage.
Chapter IV. Soil Erosion 183
Introduction, effect of soil erosion, gully erosion, methods of controlling soil erosion, terracing, effect ot
vegetable and plant growth, control by check dams and detension basin<. sailab bunds and wat bandi.
PART V-GROUND WATER ENGINEERING
Chapter I. Soil Physics ... 191
Introduction, soil and its textures, soil cJas8ification, physical properties of soil, soil water, soil characteristics,
other soil constants and equilibrium limits, swelling and shrinkage of soil, stability of 50ils, properties of clay
and collOids, distribution of wat€f in soil crust, soil temperatures, soil atmosphere, soil sampling and 3.nalysis.
Chapter 11. Surface Evaporation, Soil Evaporation, Transpiration 212
Definitions, effect of temprature on surface evaporation, effect of Manometric pressure, effect of Relative
Humidity. effect of wind, effect of depth of water, measurement. of surface evaporation. soil evaporation, factors
affecting soil evaporation, measurements of soil evaporation, transpiration, factors affecting transpiration,
amount of transpiration.
Chapter III. Grouud Water Reservoir Movements. 222
Introduction. general configuration of the ground water reservoir, definitions, experimental evidence of
·B. S. P. conception, well observations, water equivalent of well rise, sub soil flow (Darcy and Hazen). Author's '"
, 8ub soil flow formula, and its use in the field, well Versus B. S. P. observations, water table in a soil crust, capillary
fringe in a soil crust, fluctuations of wells in water logged areas, effect of barometric pr€ssure on B. S. P. level,
yearly cycle of rise and fall of grounds water reservoir, hump in water table under the canals.
Chapter IV. Theory and Physics of Seepage Flow from Canals ... .., 249
Introduction, history of the development of the subject, Wilsdon experiments, Author's experiments,
temperature correction, moisture variation below canal bed in three phases, pressures and gradients below bed
in saturated and unsaturated phase, conditions of losses in three pha~es, Haigh's formula of absorption losses;
reasons for the unsaturated phase not causing any addition to the water table, method of assessing .seepage
 xvi
losses, discharge observation method, flume ob,ervation method, evaluation of percolation intensity co-efficient
from seepage discharges in cloures, point method (point method apparatu5 and trough methGd), statistical
methed, Dr. Yaidianathan's method by observing seepage profile, examination questions.
Chapter V. Water Logging and Antiwater LoggIng Measures ...... 274
Defil'lition, factors responsible, the infertihty ~f water 10liged lands, water requiremenb. for crops in water
logged aleas, causes of water le>:ging, ri,e of water table in various Doabs in the Punjab, the cure of water
logging and anti-w6ter logging measur~;;, lowerinR of water table in Bist D0ab and, its remedies, stable water
table near the "nd of Rechna noab and water inventory in R,chna Doab; examination questions.
;;hapter VI. Reelamatiol'l of Thur and 5"", Land '" ••• 284
Introduction, salt in the soil crust, formation of Rukkar ~oils by base exchange in the salts, soil clalisifica-
tion for purpose of reclamation, Telia Kallar Lanrl5, re~lamation methods, reclamation opelations, permanency ol
reclamation by leaching, Immunity of Kallar trouble iu well irrigated areas, reclamation of Sem lands, adjoining
fields of Kallar and good culturable lands, eXamination que5tions.
PART V[-GENERAL SECTION
Cltapter I. Soll StablJization .,. 2&3
Introduction, properties of ~oi), ideal soil lor bl'st stabilization,clay percentages,<:>ptimum moistures content
density of soil, soil compactiem at optimum moisture in large works, various methods of soil sta.bilization,
detrimental salts in soii, use of soil s'abilization ill road eU/iineering, use of stablilized soil base for aeroplane
landing !!roundo, soil stabilization in water logged areas
Cllapter II Design of Bridges and Culverts 309
Introdu~t!On, I. R, C, Standard loading reinforced C?'1crete bridges, masonry arched bridges timber bnd>:es
amd ublll8 of design.
Cbapter III, Technique of Hydrollynamic Sub-SoiJ Pressure Observation 321
Introduction, design of appalalus I'nri its working, design of pressur6 tapping pOint, theO£y of bellzine.
differential manometer, temperatu"e errors, design of P.T. pipe in stiff soil crust, record of pressure
unaffected by temperature, diaphragm manometer, use of Chattock Micro manometer, use of hydrodynamic
pressure ob~eryation to measure height of capillary fringe, measurt:ment of Darcey's K or a.uthor·s (P,r.C),
soil evaportion, flow into the water table, effect of drains .. nd canal_, ground water flow down the Doal;) and
transpiration.
Chapter IV. Hydraullc Jump and Water Proll.les 336
Notations. general equation of floW' in open Channels, critical conditions of flQW with various prdpertie~,
standing wave, hydrauliC jump on a level flow, jump on a glacis,' jump with lTough drowned, determinatioa
of position of jump, back water f'qu Hions in different cl!Jndltions, backwatet function~, determination of
water profiles, examination qu<'stions
Chapter V. Iiydraulles applied to Iniga tloR 3S()
Introduction, hrdro~tatlc~, Bernhouilli's theorem. flow through small orifices, small an<i large traingular
openings, sluices, weirs, submerged weir-, pipe~ and syphons strength of pipes and cylinders, units and
dimensions, fundamental unitil. derived units, method af solution of problems by dimensional analysiS, laws of
hydraulic similitude transference equations, geametrical similarity, mechanical similarity, kinemetical similarity,
dYnamical simillUity, examination questions.
Chapler VI. Canal Plantation ... ." 375
Introduction, suitability of area for plantation, and the choice of species, shisham plantation in commanded
areas and in uncommanded areas, kikar plantation. plantation in building cOlllpounds, malfttenance cj_
plantation and commercial exploitation of plll.ntation.
Glossorv of terms Associated wi·h Irrigation engineE'ring
(a) English terms arranged .-\ to Z I
(b) Hindustani terms arranged A to Z xx
Errata II! volume II

Volume III
Diagram used in designs of Irrigation works-Plat~ No.1 to XXII
 ixvii

:' ; ....
:. :0 '

..... ,..-.
".,'

.', ::

:::l
.' 7:. ' S
..' - .§
~
. ; . ,;
.' "

: . ~.
-{, .;

>': ',';:
' ;.: -

", -.. :'

• I
xViii

•
 IRRIGATION ENGINEERING

!1nh,oduction
RRIGA TION is the art and practice of causing water to flow and spread over land with the
ll oJject of nourishing the growth of plants.
in gardens, or trees in fortsts.
Plants may be agricultural crops, fr'uit trees
From the point of view of art, it is a branch of
Engineering and fro.n that, of actual practice it is one of Agriculture.
Plants require for their growth proper weather conditions of heat, light and moisture
In irrigating countries situated chielJy in tropical and sUJ-tropical regions, the first two are
a'oundant and do not need to be supplenented. The last na<neJ is frequently deficient there
~n~ is .the only one w~ich can 'ne suppli.d artificially: on ~ large scale,. a~d t~is is done by
IrngatlOn. In hot chmates, even. W,len the total ramfall IS a ~undant, IrngatlOn is required
for certain crops such as rice en account of the uncertainty of rainfall and the need of constant
supply of water. In countries with scanty rainfall, irrigation is an a)solute necessity for the
growth of agricultural crops and for producL'1g water supplies fvr other necessities of life such
as water for drinking and domestic purposes. Irrigation suppknents rainfall to procure all
a~enities of life to the human beings. '
. Irrigation may be natural or artificial. Natural irrigation occurs where rain falls, or
where a river floods over the lands in its vicinity known as "Sailab" in Hindustani. Sometimes
both of these natural systems of irrigation are controlled by the construction of suitaole
bunds around fields to retain rain water and by making em ..)ankments along rivers to
regulate rivers spills.
Artificial irrigation is the artificial application of water to lands for agricultural purposes.
The science of artificial irrigation may take any of the following forms and may involve the
construction and maintenance of works pertaining to them:-
(a) LIFT IRRIGATION -Water is lifteri from open wells, tu"._;e wells, natural drains or
canals and used for irrigation.
(b) CANAL IRRIGA TION.-Artificial channds taking off from the rivers are eonstructed.
Water is tuen used for irrigation by flow froJ_TI the distrLuting channels through
outlets. In the case of permanent canals the nver at the canal off-take is controlled
uy constructing Head Works. In the case of inundation canals there are :no
permanent Head Works.
(c) TANK IRRIGATION -Dams are constructed at suita'Jle sites in the hills, where
rainfall is aDundant, for storage of water in reservoirs and then water is let oft
from them in channels which are used for irrigation of crops in the plains.
(d) DRAINAGE ENGINEERING-Drainage Engineering inclUdes Hydrology and
deals with the science of measuring and analysing the flow of precipitation of
water (rainfall) evaporation and analogous phenomenan to provide the correct basis
for estimation of the supplies availaole for irrigation purposes from the natural
drains such as rivers. This further deals with the artificial channels known as
surface drains or sua-surface drains, which are designed and constructed to drain
 2

off surplus rain water.. ~h~ w.ater from such drains is sometimes used for
irrigation purposes, by hft Ifngatl9~>' , .....
(t) GROUND WATER ENGINEERIl-<"lG+Tbis deals with the science of water occuring
telow the surface of the earth, its distribution and movcment. This further deals
~ith the dynamics of water ta ,Ie mowments as infl.uenced lJy rainfall, irrigation
and other factors. 'Vater percolates to the undergrcund reservoir from the canals,
from other su)siqiary irrigatiol} ,Ehannels, and from the irrigation fields. Irrigation
results in excessive additions of watcr . to the water ta 'Jle which gradually rises
under ground. The lands are said to J e waterlogged when water ta.Je rises
so near the natural surface, as to di~inish the proc:.uce of the crops. The works
required to reduce and prevent water-logging, necessarily pertain to that branch
of engineering which deals with irrigation.
,., .

_, The o'Jject of .thiS book is essentially tc deal with' ;.trtificia.'l irrigation involving works of •
the nve types as enumerated a·cove. This Dook. 4eals with the techniqu of the Cfj)llstruction f
and maintenance of works relating to all of these five branches of irrigation engineeriI4:i with a
view to equip the student with su'ficient practical and theoretical aspect of the projlems relating.
to them. A large number of solved examples of. the. design of irrigation works are given.
Selected examination, questions ar{) given at the end of chapters, important for the ·:sfudents
preparing for the Degree Examinations of the various Engineering ,Univetsities in India .
.' Italy' I~ay wen ce considered as the parent country in the science of distribution of .
water. It was the first country to .take up large schemes involving large irrigation works. The'
Italian storage reservoirs !ln~ aqeducts have served as a guide to irrigation engineers in other
countries. Indian Irrigatian no dou')t followed the foot steps of Italian practice. The first
large canal was constructed more than a century ago in the enited Provinces in India. The
Ganges Canal is still a great engineering success, and was followed by large canals in the
Punjab. The Punjab canals are first-class examples of the development of the science of
irrigation engineering. The .reyenue return from the Punjab canals is the highest and the
efficiency of their working surpasses all irrigation works in the world. It is a well known
fact that the Egypti.arirrigation system is founded on the Indian system. The science of
Irrigation engineering:,independently developed at great strides in. the United States of America.'
Large irrigation schemes both in the forms of tank and canal irrigation have teen successfully
constructed there. . The ,Al!1erican practice undoubtedly now leads the world in the teChnique
o,f .the design and construction of high dams.
, ~ ~ :,,~. , :1 ~ , .
 PART I

LIFT IRRIGATION
Chapter I

iRRIGATlt)N FROM UPEN' WELLS'

''' .. ..
; \

1. Definition.
, Wells are merely holes in the ground t(!{tap·.the underground water. They may be
entirely unprotected, partially protected or entirely protected by means of lining ofbr~c~,milsonary.
They are called open wells as distingui.shed from the tu .. e wells:" ,A~~,~~.;.... ,
2. Source of Water SUllply.
The source of water supply for open wells is the sheet of water found under the gro~nd.
This underground reservoir of water commonly called the water ta·.lle is fed 'by .rainfall which,
soaked into the soil, permeates under the grclUnd. Water is 'held thera by the ;~ncec of the
impervious stratum below this sheet of WaLer. The retent;ve SU0stratum is no~hihg but the
consolidated sodiu n clay hardened by the 'heat in the interior of the earth glaDe which increases
towards the centre.
The laws relating to the flow of the sub scil ground water are dealt with at length in pa~t
V of this volume. There is only slight lateral flow in the sui) soil reservoir towards the mam
drainages, i. e., rivers; and as the country slopes from the foot of hills to the confluence of the
rivers, the main underground flow occurs along the slope of the ~?untry.
The underground water is available in three forms :-
(a) Rock hollows.
(b) Artesian basins. •. '
(c) Ground reservoir.
Rock hollows filled with water are found in the hills, they are generally small and cannot
be used for irrigation. The artesian Dasins are found near the foot of the hills. The rainfall
permeating in the hills reappears on the natural surface on, account of the high pressure gradient
and the presence of the i:nperviou5 soil consisting of coarSe sand'a_I!d gIavel a"ail<l-ole !!~ar the
foot or flieilitl-s-;- In some cases artesian wells and springs-can 'De used for--irrigation. Some
springs'iouu-dtfUhe Himalayas have great medicinal value on account of the dissolved salts in
water. In the plains. -rr-,e ground reservoir is fhe-chieT~~~;:ce of water supply. WenS-are- aug
down into the water f)earing strata and water is lifted .and used for irrigation, - ,,
3. Locat~on of Well.
NAf'URlfl.
Suitable site for an irrigatiQR well is that wht}r~_a:_weli 'can
I::e found~d inda'-y strata (called mota or karl below sub soil
waters level arid'-i:he clay strata can ce pierced to ensure
permanent supply as shown in Fig. In a sandy soil. a bigger
well will te required for the same discharge to ensure no
disturuance of the soil Lehind the lining or steining of a well.
as the rate of supply depends on the velocity and the area of
percolation. The soil movemerit is governed by the critical or
optimum velocity of percolation. ' ,
4. Discharge of Irrigati<Hi WeUs. ~<

An irrigatiGn well shol1ld r.,e sunk about 20 to 25 feet

'1 elow the spring level so that in'years of draught there is a
margin of about 10 feet leaving a Jout 10 to 15' feet
infilteration head. . ,., ., ...~
\'
 4

The- average discharge from an irrigation well, water'r eing lifted by Pe~sian Wl:e~ls, ha5
1een found to be 0'137 cusec (page 253 Farm Accounts 1933-34 Punja1:l). The dIscharge IS more
than this in the case of mot or charas. The discharge require:nent of a well is never going to b(t
more than 0'15 cusec.
The discharge formula to estimate the discharge. from an open well may te used in the
following form:-
Q=4/3 a AH where Q=Discharge in cusecs.
A=Area of the well in sq : ft.
,
H=Infilteration Head or Depression Head (Percolation Head) ..
a=Percolation !ntensity co-efficient in cUsecs per unit area
per units head.
The constant ','3 represents ratio of the actual area A' ?f
the safe cavity fOffiled to the area A of the well as shown In
Fig. 2. The usual value of a for tte Punjab soil of t~e
water i earing strata is .75 X 10- 4 which gives a corrcs~ondig
value of the Transmission constant of about '00023 In the
Darcy formula of viscous flo N throgh the soil.
• : • • • ':, I I": EXAMPLE •
.:.', :.': . ':' ::. '.
/"':.... '..... . Find the diameter of a wall to give a discharge of
0'15 Cs with an infilterati~n head of 12'0 feet.
Let percolation intensity co-effcient per sq : foot='75 X 10- 4
Let the diameter of the well=d feet.
d2
Area of the well = - 7T 4-- sq: ft.

Q=I"330Xa A. H. = 1,330X.75xl0- 4 X
1r
-4~
u2 X12 '15 cusec

..'d 2 --
'15
;75xiG~ '5(12 -163
--
:. d = 12'8 say 13 ft.
For an average sand of 16/1000 inch average diameter in the Punjab, the optimum velocity
~f .005 it./Sec has teen found to move the finest soil particles in the experiments carried out out
the Irrigation Research Institute, Lahore (vide page 228, Paper No: 248 punjab Engineering
III
Congress) . .

The actual velocity of flow in the above quoted exampl{(==

0009 ft jSec. which allows approximately a factor of safety of 5. A factor of safety of at least
3 should always be allowed to provice for the years of draught when the well has to be
\\'orked under the worst conditons.
5. Suitability of we:l water fOf crorS.
All sub ~oil waters are !lot suitable f~r irril$atiori J?ut'poses, ~~ter. of the well shoul.d be"
tested to e.xamlne the amount of salt contamed 1n solutIOn. The lrngatlOn Rllsearch InstItute,
Lahore has cevcloped a technique in evolving an expression to which the name of Salt IndeJi
 5

is giveri. T.his is a measure of the quality of irrigation water in relation to the soils. The
sodiu'll salts nre,ent in tlle Irrigation Water cause l'ase reaction to take place in the soil with
its consequent deterioration. If calciull salts are present in the irrigation water, the reaction is
opposed and delayed some what .
. Salt Index=(Total Na -24'5) - (Total ('a-Ca in Ca C0 3) 4'85. When the e'{nression is
np.gative, the water is suita'-.le for irrigation of the crops, and when Dositive it is unsuitahle.
The brackish water as found in the Western Punjab is unsuita~)le fOf irrigation, while the sweet
water of the Jullundur, Hoslliarpur, Amritsar, Gurdaspur and Sialkot distr:cts is very good fOr
irrigation. .
~ 6. Classes of wells.
ra) ImpervIous lined weIls.
These are the only type., of wells userl for irrigation purvo5es. They are usually of tyoo
shown in Vig. 2. The de3i~n anti the Df0CeSS of sinking such wells is descrihed in detail in the
su"sequent paragraTlhs. When the botto'1l of the well is in san~y soil, it is usual to provide
ipverted filters at its rott'1'1l so that a dangePJJ., cavity is___nill_for:nerl during t"'e years of
draught. TIie"luverfen filter simply consists of alaver of eoarse sand aHheO-oUom, a layer of
fine gravel in the micldle anrl a layer of bajri of a'~out i" to t"SiZe at the top. These layers
are a!~out one foot thick each.
(b) P'JfV:OllS Unei Wells.
A QQl~ is dug in the ground.kthe water table and dry brick or stone lining_is then laid
on.a well £ur~} l!p._!.9- a h~igm0f8 or 10lecl:lhe soIl InsIde IS scooped and ~eH can Fe sulik
easl1Y~ feet }-~]Q"".J:lle spring leveI."WIiile sinking is done, the sandWiIlno dou',t flow out
from the joints: -This will ;"e very mu:cn reduced, if brick ballast of size 1" to 1" is filled around
the lining, so th8.t it may sink along side the lining. Such wells are~conomIcal III the case of
,small devths. And when the discharge required is not much, water percolates from the sides
also. They are quite successful when the arrangement for lifting 'Water is Rati or Denkli. I,
() Kacha Wells.
t This supposes that clay stratum is available from the natural surface down to a few fee~
below the spring J&yel.an.4.!hat the soil is stf()I!K_tlllough to_stand_witho:ui. tlw_ ht:1p of_th,e lining.
Such wells are used for low lifts up to 10 to 12 ft. by means of Rati or Denkli. -.
7. Linei Or Pacca Wells. .- ~. ..---
All open wells used for irrigation are circular. The thickness of the lining or steinins
is usually 2 bricks i. e. 1'5 ft. for depths up to 40 or 50 ft. For depths up to one hundred

t
:.s "...,VII•• ,..A.,."
". C"rr'NG fl>LiI rI. _, f
.,.U. ~u" • . ., ~DI:~ tultl IIt"N CIJR •

Fig. 3c, Fig. 3 a. Fig. 3 h.

,
 6

, feet, the thickness of lining sho~l~ ie2t tricks, i, ~." ;t~;k RPWL:''?:tlN~ for~eR~~s. ~?\q~,e;lifol,~
100 it. are never built, Leca~se hftmg of wa~e.r bewme-J V«x:Y ,~neyp~?mlFaL , 1~~ hnmg ~s PUllt
in lime mortar of proportlOn 1 : 2 (one h!ne ,and 2 ,S,\lT~Jq qr ~fl.11c;l)~ Qr, sei~J,C<llt Irlqr~ar. of
proportion . one cement ,an~ ~ sand). When. a large nU.n _er of. JN~Jfs Js:rl!qll~~c~)C ~ e ~u}h,
it is advisable to get speclal uncks ,iJurnt which copforfi?, \ ,to the CIrcular snape of'. the wen.
Ordinary flat tri,cks need. a lot of .1aLou,r to d.r~ss .the;~ to the. circ~lar;shape, . and also
j

require a~ exceSSIve quantity of mortar to fill ~he ]Oll!ts; . \\ he~ spe~lal Dncks ar~ used, 25 eft.
morar wIll do for 100 eft. masonry, but for tile orilInary bncks '33 ctt: wl11ve,needed .
. The lining is built on a well curJ. The' 'Wen- Cut)' may 'De of wood, iton or reinforced
concrete .. Sections of suitatle designs for WOOGen and iron well curus are sketched in
Fig. 3a, 3D. \
The wooden well cur;) is mace of hard wood as Kika:r, . Sllisham ,or Sal. The iron curb
is made of 6 to 8 triangular frames of angle iron covered Dy' l' to t'
plate all round. There are
usually 6 or 8 tie rcas ~"to £" ,diaincter. The load of lining on the Cllr J is shown in
Fig. 3 (cl.
The well curb' is a continuous tent team in cetween the tie rods supporting the
0
triangular wedge of masonry of isosceles shape with angle 45 when the masonry IS in Ce'l!-ent
,
0
mortar, and with angle 60 when the masonry is in: muLi or lime mortar. fhc lengtas of tne
rods should be at least double the height of the weight triangle of masonry. fne tie rods
are anchored at the top in a circular plate iron f' to i" thick anu about 6" wide.
The section of the wooden well cur,) can l"e designed according to the formula;-

M= ~-- or d2=M2/Z where d=dept~ in inches . . .

2 -V2~ M=cendmg moment lO lOch tons.

It is supposed in deriving this formula that '-

"the sale tensile stress in wood is ! ton/sq. inch and that the section is
triangular with top width 12". R. C. section of a curu can easily D,e calculated'
, considering it to ce a Learn of triangular section carrying a weuge load in ;__etwcen tne
rods as shown in Fig. 3. (c).
B. Construction and sinking of welJ,:;.
(a) A circular pit is dug in the ground up to the water taole or up to the bottom of
the clay str';ita, which ever is higher. Tile pIt snould De sujicient -in circuinference
to allow a space of at least 2 ft. around tile lining. In clay the sides 6t the pit
can usually stand vertical.
(b) The well curb is laid when soft soil or water table is r~a~~cl..'

(cl Masonry of the required thickness is, then, built on the well- cu~h 'to it height of
about 8 to 10ft. a Jove the ground level. ;',

,(d) Then a temporary platform is made on the top of:linjng to ~1.)-pport t~e weight of
gunny bags filled with earth ,to facilitate its sinking. 'The lO,<id' is placed OIL the
outer sides of the platfonn leaving a clear space ;jn the middle ,SU, hejen;t for lifting
the material dug Ly means of a pulley arrangement lai4 pver tll~ ,platfOrm.
(e) The soil within and h~Iow the curD is gradually s~oped by- mean.-s . ot
Kassies to
fill a big bucket which is pulled out,oy the pulley: arrangement. The -loaded curu
" thus sinks gradually. ... .
r.

(I} If is necessary to see th,<!t flte .cu~~ with lining sinks vertically:' 'fo watch this
. 4 plum.) lines are hung at equal intervals on the; b'uter' 'circumference of the
.'; ··steiriing. If the steining gets out of plum;), the workmen ,an_) instructed to cut
the obstruction under the curb on the side which lags lehind in sinking.
 The o1:lstruction may te in the form of roots of a _ tree or pieces of stone.
If tl'ere is no such obstruction, it is then a case of the unequal soil friction against
He lining. The weights on the platform are the1] shifted to the siee which is
sinking slowly. If unequal loading cannot cope with unequal skin friction, then
the soil behind the lining should "te wetted by using water jet under pressure.
(g) The c1igging of the insi£ie soil by men is usually possible up to five or six feet
j elow the spring level by cootinuously bailing out water. When the ingress of
water is more than the removal, then digging by means of kassies cannot proceed.
FurPler digging is done by a jham which is a large kassi. A diver dives and fills
thc jham with soil, which is lifted by the pulley arrangement. This is no doubt a
slow process. Wells have j een successfully sunk from 20 to 25 ft. below the
sDring level by this method. The modern dredgers like the Bull's dredger clench
the soil by weight of tl':e' blades and when the chains are pulled, the blades
autO'natically close with the soil filled in t1ce drcdger. The digging of soil under
water is very much facilitated by this method.
(h) When the curb with steining has thus been sunk to the required level, the space
around the lining is filled with good clay earth consolidating it by ramming and
watering in successive iayers of a'Gout a foot de" tho
(i) Finally the top parapets are built on the top of the steining to suit the water lifting
arrangement which is usually a Persian Wheel.
9. ,Methods of RaiSing Water.
c>'

The following water lifts are more or less in general use in India :__:;,"
1. Rati or pulley block arrangement for small garden plots.
2. Denkli or lever for small gardens and vegetable plots.
3. Charas or mot or leather bag for agricultural crops.
4. Persian Wheel for agricultural crops.
These arrangements are shown in Fig. 4 (a to d).
"

1N6. DIN,," .

hg. 4 a. Fig. 4 b.
 /.1FT IRR~TION - TNI. tNUlt1/1

Fig. 4 c.

Fig. 4 d.
In very few cases, the water lifting arrangements have been modernised in the form of
chain pumps, ordinary suction pumps worked by bullocks. and modern pumps worked by the
electric motors or the oil engines. The e;nciency of different water lifting arrangements was
tested in experiments by Captain Cliooorn given in Roorkee Treatise on Irrigation Works
Vol: 1, reproduced below: -

No. Class of lift. Lift in feet Labour Foot tons per hour
per head

1 Rati 17 Man 32'26

2 Denkli 15'4 21'99
:J Chain Pump 15'0 30'55
320
"
Cattle 588
4 Chams
5 Persian Wheel 30'0 "
57'1

A ton of water raised one foot ton.

10. Duty of well water.
The duty of a well cannot ce defined in terms of acres per cusec discha:tg~ because
there is no fixed and continuous discharge. ... --..
~
Area to which a well gives protection in the case of draught is d.efined as t1M tiuty of
the well. It is thus the average area irrigated from a well per annuIn.
 9'

The experimcnts carriei out in t~Je Unitcd Provinc'~5 by ~IL Anthony gave the
following figures:-

No. Class ot lift Lift Duty in acres per annum

-~~----~-~-------,

I. Charns w.th cattle 25 3'91

2, Charlls WJth men 28 27f)
3. Denkli IS ')1
4, Pati 18 '8

This is a very poor show, but in the Punja~, the usual duty of a well worked by
chams or Fersian Wheel is 25 to 30 acres. Csually there are three shareholders having an
area of eight t) ten acres each. The higher duty for the Punjab wells is due to the
following causes : ~
(i) In the Pun:ab the wells are generally fitted with Persian wheels which are suita~le
for lifts of 20' to 50'.
(ii) The cattle and men in the Punja') are hardy and, therefore, capa~le of putting in
strenuous effort.
(Hi) Holdings in the Punja~ are small and, therefore, water is used economically.
11. Delta of Well Irrigltion.
Delta is defined to be depth of water required in feet to mature a crop. The usual
depth of first watering fro.n a well is three inche3 in a pbugled field and the suosequent
waterings are of 2" depth each. The average depth of watering from statenellt No. 1. Farm
Account 1933-34. Punjao Agriculture Pu.)lication No. 46 works out to be '193 ft. or 2'33
inches. Pro:)aJle deltas for various crops in the Central Punjab with average rainfall of .20"
per annum (IS" in monsoon months of July and August) is given in the table below from
personal 00servations : - '

No. Crop. ~o. of waterings. Velta in feel.

-----_._--
~dO ~'>ll
1. \V"e it 4
')
Cq·to 1 5 to 6 10
3. "Vegei..a.b:e:i ",
" 6 '!'1
4. Pet~t )es 8 l··t) f:v.
S. I ler 'T'H)S Chari.
p
Ma.ze dn I'Senjl "~ ,• 3 -6
6 S;,ahal Ber,elffi G 1'2
7. S'lgarcane 8 to 10 1'8
8. Maize d -...
~ ',.n " 5 to 6 },O
'"
A comparison of weI} delta with the canal irrigation figures in Part II ' Chapter II
will show that a')out i water is used in well irrigation as co:npared with canal irrigation. The
yield of crops from well irrigation is usually somewhat better than that from canal irrigated
crops on account of timely waterings and Jetter manuring facilities.
The reasons for low delta in the casa of well irrigation are ;.-
(i) The cultivator is aware of the cost and the la')our e:nployed by him to lift water
and, therefore, he uses it very economically by resorting to small kiareis
in his fields.
iii) Well water-courses are usually well maintained and there is no wastage from
breaches and overflows.
( iii) Well water-courses are always kept clear of all jungle and grass and are often well
consolidatad and plastered with mud to reduce absorption losses.
(iv) Water is lifted according to the requirements of crops at the will of the zamindar.
He has not to wait for his turn as in the case of canal irrigation.
 16

12. Cost of Well Irrigation.

ilhe cost of well irrigation from Persian Wheel is worked out on page 253, Fan.
Ac-co1&at ]933-34 Publication No. 46 of the Punjab Agriculture Department.
Cost of well Rs.774/-
Area cropped 17'4 acres
A verage number of irriga.tors .(

Time taken to irrigate OIle acre 2'84 days

COST PER AORE

Rs .. a. p.
Overhead charges 11 13 0
Bullock labour 11 4 0
Manual labour ... 11 12 0

Total Rs. ... 34 },3 D per a.cre.

Cost excluding manuallabour=23jl say Rs. 231-
The manual labour does not cost a cultivator anything as he or his children usually
supply the requisite labour. The overhea.d charges for the cost of a well etc., are also very
excessive. Bullocks will also te used for ploughing and cartage and other miscellaneous work.
The well irrigation is definitely many tim(!s more expensive than the canal irrigation.
The Irrigation Research Institute, Lahore, has also worked out the expenditure and
income of one family in Jullundur District possessing ten acres of land on a well with one third
share in the well. The gross receipts in a year work out to Rs. 741/- while his actual
expel'l$itnre is Rs,,383/- for paying land J:evenue, casual labour, bullocks, share of well and <"

Persian Wheel, seed and kamin and nliscellaneous expenditure. His net income is about
Rs. 358/- per annum.
13, Well Irrigation Versus Canal Irrigation.
Well irrigation has got a definite advantage over canal irrigation, I:;ecause a cultivator
. can raise water when it pleases and suits him.
In certain tracts, well water is more beneficial to the crops as in the Bist Doab in
the Punjab.
The well irrigation should te encouraged parallel to the canal irrigation as it is an
effective antiwaterlogging measure and serves to keep the water table low. Well irrigated
tracts should 1:e debarred from canal irrigation when new irrigation is extended to a tract, so
thllt wells may not tecome extinct. Canal water should be given to about 50 Yo intensity of
tke irrigable ar.ea so that the well irrigation should also develop.
The canal irrigation in the Punjab has a definite advantage over the well irrigation in
the manuring value of the canal water ca.rrying fine silt and clay in susr;ension. All canals
take off from alluvial rivers, .
 11

14. Examination Questions.

1. (a) Define duty of a well.
(b) What factors govern the delta irrigation from 6pen wells?
(c) Why is delta for open well irrigation the lowest as compared to the delta f<!tr tube well or .anal
irrigation? (p. lJ. 194t)
2. Describe the various appliance. used in the punjab for lifting water from opeD wells an4 which is
the best in your opinion ?
3. How will you cla5sify the soil in the Punjab with respects to the clay contents? Meutiou the crops
which will flourish in each type of the soil.
4. Describe the ·process of sinking an open well llnd the precautions that you will take 50 that the
steining of the well does not get out of plumb while sinking.
5. (a) How would you judge the suitability of well water for agricultural purposes?
(b) What do you understand by (i) Sa.lt lnde)!: (ii) Salinity of water?
6. (a) What is the average discharge from an open well in the Punjab?
(b) How much below the water table, an open well meant for irrigation should be sunk in the
Punjab conditions?
(c) What factors determine the diameter of an open well meant for irrigation (
 14

irrigate finally 17389 acres. 20 tube wells are at present installed and are wofking.
It is very successfull so far. .
(e) Ganges Canal Hydro-Electric SC;1eme was conc~ived and carried out by Sir William
Staml;e E /::.I.E .. Chief Engineer, United P lOvmces, India. The energy has Leen
developed from the low head falls on tte Ganges Canal. The complete scheme
envisages the generation cf 30,000 Kilowatts of which more than 24,000 are from
canal falk The bulk of tl'e power is useJ to energise 1500 state operated tube
wells wilh an everage c1isc:large of l'S causes each. It is meant to irrigate 1300
square miles area Tubt! wells are e.-nerted to run for about 3000 hours per
annum. This sc\e:ne has proved a financial success and the returns are
very promising. This is the biggest undertaking in india for Irrigation from tube'
wells.
i. Borln~ of tube wells (An extract is given from Bi"J. 28).
A. choice of site for boring.
(A) Wate,: divining.·-Considera·le di'ficulty often arises in determining a suitable si.te
for a boring. Eut for locating springs or flowing water near the surface there are often other
indicatioJ.s which may prove valuable.
Water divining has receJ.tly received much attention' and there is a')undant
proof available to show bat Le art of "doNsing" is a fairly reliable means of detecting
the presence of flowing water near ground level. Many theories have been formulate to
explain the p"henOil1enOn '. ut no prcof is yet availa~le. (,:::i.). 1.)
There are indications which may lead t6 the discovery of springs or water near ground
level, in cases wl~ere nothing would a?pear, to t':lose unaccustomed to observations of natural
phenomena to induce a Lelief in their existence. The following are· some of the
simplest. (Bib. 1.)
In the early part of the year, if the grass assumes a brighter colour in one particular
part of a field than in the remainder, or, when the latter is ploughed, if a part is darker (or
damper) than the rest, it may te suspected that water will l::e found beneath it.
In summer, the gnats hover in a column, an,i re'llain always at a certain height ahove
the ground, over the spots w:le~'e springs are concealed. In all seasons of the year, more dense
vapours aris~ from those portions of t'~e eartYs surface w;lic3, o,'{iag to Le pre ;e~lC'~ of
subterranean springs. are more damp especially in the morning or in the evening.
In selecting sites for wells to tap 1'1e u;)per layer of su')-s')il water it is aha ne{:e'is1.ry
to study the phnt life in t'1c locality. It is CO'11mon knoNleolgc tlat certain tree:> thrive "ell
where the su ,-soLI water is more plentiful or nearer to ground level. .
Tl:ese rules apply generally to springs near the surface. WilCre the source is lONer
these are rarely su ficient. It is clai:ned that with the aid of a "divining rod" or electrical
instruments springs several hundred feet ~elow ground level can i~e located fairly accurately.
For further information on t'iis su ject a reference may te made to "The Physics of the
Divining ~od," by Maby and FrankLn. pu·.)lished by Bell and Sons Ltd., London, S. W. 2.
But expenence with these instru:nents has not Leen very encouraging and the only safe guide
for correctly detcrmining deep sub-soil aquifers is a trial coring.
(B) Boring for W dIs.- The diameter of boring tu':es is measured on the outside as
against ordinary pipes which are measured on the inside, Lecause in the former case we are-
conce.rnd wi.th the size of the bore made. This point should be taken note of when
orden~g J:;ormg-tubes from foreign countries In practice coth the methods of measurement
are Lemg used by the borers almost indiscriminately; and in India generally inside diameter
is. measured, forthe sizes of boring tools strainers, pipes, etc., have to be adapted to the inside
dIameter of the boring-tu1:es A safe course is to mention the outside diameter as also the
thickness of the loring tu':es when placing an order for these.
Another point, about which there is a good deal of confusion is regarding the
nomenclature of casing pipes, blind pipes, blank pipes, plain pipes, solid pipes, etc., and these
words are used indiscriminately and in different senses by different borers.
The generally accepted practice which is adopted in this book is as below.
 15
All pipes and tub(s used for boring have leen described as "boring-pipes" and
"boring-tubes," or 'casing-pipes" and "casing-tules." . ' .
1A..ll pipes used for the tube-well itself (whether between stramer as m a stramer well,
or in c.ontinuat:on 'with a slotted or perforated pipe as in the case of slotted tu':e-wel~s and
perforated pipe-wells, or in cavity wells where "boring pipes" are not used) are descnbed as
"plain pipes" or merely as "tubes"
The small length of plain pipe (generally 4 to 5 feet) used at the bottom f nri. of a
strainer or slotted _tube-well and having a cap or bail-plug fixed to it is described as
·blind-pipe."
In the case of cO'l1pound wells the larger diameter plain pipe on top, in which the
bore-hole pump bowls are housed is sorr,etimes descri~~ed as "housing pipe."
. The use of the words "blank pipes" and "solid pipes" has been done away with as it
is confusing.
Though boring requires skill and care, yet in principle it is extremely simple The
operation consists, as its name implies. in boring a hole of a dia'l1eter varying according to
circumstances, and in a vertical direction. Many syste:ns have teen and are now follO\yed in
carrying out this Jdnd of work.
(C) Ior:ng system3.-Boring systems may te BOA1D7HRSlilJS TIJ.aNINC}/
divided into three' .main clas,es. viz, 'percussion
boring, by which a hole is made into tte ground
by recucing the strata to powder, water jet boring
which makes a bore by washing out the strata, and
rotary boring by which a hole is drilled into the
strata and a solid core obtained. The percussion
system is suitable for ordinary sand or gravel strata, :) Oflt£T JOYIT .
while hard tenacious clay is better penetrated with
a water-jet plant and the rotary system of boring
a
OltlIJTNIl.EaD.s)"O ANINCN
with core drills is adapted in the case of firm and
hard strata. The driven tube is the simplest form
of tore hole. This method is rarely employed for
larger wells.
(i) (a) Pe;'cussion rope boring -H.e old system
of well boring is the original Chinese rope boring
system which is described in the beginning of this
D 5W£I.I.£.DIiND CR£.sSED
8 ()/t1D TH.Il£IWS 70 nil/NCR

o
book and is of great antiquity. The present
method of rope coring is described below:
A pit is dllg at the site where toring is to l-e
done, about 6 to 8 feet in diameter and about 15
to 20 feet in depth, and the boring·tn e is lowered
into it with a cutter shoe fitted to the f:;ottom of FLUSH ,JOINT
it. These cutter shoes can be "slip shoes" or
"screw shces" and are of slightly bigger diameter
than the boring-tube itself so that they may BDIUN6 SHoeS
provide a clear passag.e for borbg-tub~s as hey <r,!
sunk. These shoes are made of tempered steell~ig 1.
The slip shoe as its name implies is slipped on
to the 'toring-tu e when strating loring and left I:" .Ill
at the bottom of the tore when the tn e is with- t •• ; 'I~

drawn. The slip shoe is of a larger diameter than iI 'I \-:'' ~

the screw-shoe for the same diameter of tU:Je. This S S
shoe is useful in in stiff clay, rock soil, etc., and LIP SJlO€ CREW SHOl.
for deep bores as it clears a slightly l.arger (fig 1.)
than .the .scr~w-shoe . an_d. thus the fnet fi.,tJ e surface of the boring-tul:e is reduced
both ill smkmg and In withdrawel for ~lea¥ctnc~e ed by the larger shoe remains open to

\:B\jtj:)1
~~
''-. '/1-'1:'.. .
~I
,<;~ /
 16,

Scm:e extent while the tute is being withdrawn. On the other hand the screw-shoe is. screwed
to the 1 ottom end on tr.e tore tul:e and comes out with the tube when it is withdrawn. In this
case as tte screw-shoe has to be withdrawn with the 'oring-tu':~e there is no extra clearance
left while withdra\~ing. Therefore a screw-shoe is chiefly used for sandy soils.
Screw-shoe are generally good for the average size tuee wells in India and with
experience this practice has been estaDlished here. There should be female threads to the
shoes to receive the screwfd end of the toring-tube which should fit in butt t.J Dutt with the
shoe. In the case of very deep borings, or Dorings in clayey strata, the bore is started with
a larger dia'l~eter Doring-tu1:e and carried down till the friction on its outer surface makes
it difficult to carryon the coring any further without damage to the tu~es. A slightly smaller
size roring tu 'e is then lowered inside the larger siz, tu ,e. The boring' is next continued
with the s:naller size tu JC which is free froTI any frictional resistance to the hottom of the
larger size ture. If necessary a still smaller size of f~oring-tu::e is lowered inside the inner
tu~e when the friction on the latter also recomes excessive aft<:r carrying on the Doring lower
down. Thus the diameter of the tore-hole is successively r :duced as the conditions
demand, All th' tu es should however extend right up to ground level or aJove the bed
level of the pit in ,which the bore is made, so that they may all be ex:tracted with jacks when
required.
This system of lowering boring-tu"es one inside the other to distribute and overcome
frictional resistance in the su )-soil strata is known as "telescopic boring." Care has to be
taken to select the size of the various soring tu',es in such a way that the inside diameter
of the innermost tu'~e, reaching the bottom of the bore, is su"ftciently large to allow the
lowering of tu' e-well pipes and strainers inside the finished bore.
The boring-tu:es are ste 1 tubes with the following types of joints :-
(1) Flush joints.
(2) Swelled and cressed joints.
(3) Socketed joints.
These are shown in Fig. 1. The flush joint is the best for ease in lowering
and with,'rawal "ut t~ese tu' c<; hay!) tf) . n flicker elan socketed pipes or swelled and cressed'
pipes to ha';e su ;';cient strengtil at the joints.
After eeing lm'"f!reri in t:~e pit _a<; ,',:scri'Jed above, the boring-tUDes are clamped in
position Wit'l woo 'C:l cla'llps, a'1'l ,yutly fille·j with water. Water is necessary for Doring
o,:eratiollS ::tnd must :~e poured from Le top of c:1sing pl'Je u,til W:1t'~r i<; stnEk in Ete
t--ore-hol~ it<;~lf, T f! adm.l ,. oring is done with a tool known as a "shll;ger' or "san.l pu;np '.
TI,is imple;nel1t is a steel pi~ce of a',out 1/4" t1-tickn~<;s, varying in lengt'l aa.t dia1l'!tcr
according to the size of well to 'e ·,ore'1. It is aJout 2" IC3s in diamet,:r t lan the casing pipe
and its length varies from 8 feet to 12 feet for ;;orings from b inc';es tu 16 ine~les in diameter.
A cutting shoe of "hard steel" or tungsten steel, is riveted to the sand sludger at its lower end.
The cutting shoe is tapered out slightly at the 'I;ottom to give it a clearance in cutting the
material at the bottom of the 'core. A flap or ball valv~ is fitted inside the sludger pipe just
above Le cutting shee.
The sludger is suspended hy a wire rope from a pully on a tri1)od. The legs of the
tripod are to L e 'curied a,:out two feet under ground to prevent their sliping or tilting. The ___ .
leg opposite the craJ-winch has to : e securely buried and anc.'JOred so that the tripod will
not turn over when su ject to a pull fro'll the winch, The sludger comes centrally over the
boring-tute and can work smoothly inside it. When the sludger is worked vertically up
and down by manuallaJour, or an engine with cranking arrange:nent, etc., a circular motion
is also automatically imparted to it by the torsion in the wire rope. On the downward
stroke of the sludger the flap valve is force j open and the loose material pounded at the
bottom of the bore-hole enters into the sludger pipe As the ul-stroke tegins the flap valve
closes thus retaining the loose material inside the slur:ger pipe, At the same time the up-
stroke creates a momentary vacuu"YI at the l:ottom of tr.e t;ore as the sand sludger is lifted
upwards. Owing to this vacuum or suct;on the pounded loose material in the bore is sucked
up and gets eetter mixed with the water at the bottom of the 'Dare, The loose material thus
pulled up and remaining suspended in the water enters into the sludger pump on the next
 17

downward stroke oHhe sludge~ and is retained by the, <JIve. At the enrl of th downward
stroke the cutting shoe cuts some more of tht matl~rial at the bottom' of the hore which is
pulled up and enters tte sludger in suiJsequcnt strokes. After about 30 to 40 stJ;"okes the
sludger is taken out and the loose material retained inside it is empt;ed out. This material
is carefully examined as it is a sample of the stratum at the cottom of t1:.e Loring where the
sludger has been working at that time. A change in the characteristics of this material is an
indication that tr.e stratu'1l is changing and a careful record of the nature and c'.cpth of these
strata is ke]:lt. The samples of ~he excavated material arc also preserved.
Ttc c0ring-tute is clampci-l with sleepers at some convenient point a'~ove ground level
and over the clamps a pliJ.tform is made which can 1 e loaded with su f;cicnt weight to' over-
come the surface ad;1es:on on the 'Doring pipe and force it down as the sludger excavates a
hole l:elow it. Thus boring progresse, and one casing pipe after another is s.;rewed on and
lowered till the required depth of ;)orc is reached.
(01 An1her~1l(jlt methoil of leading blring-pipes.,-A convenient and more efficient
method of applying load to the boring-pipes is the "Anchor-bolt" methGd.
The anchor-bolts consist of two mild ste.el rods H inches or more in diameter and a:oout
16 feet in length. Both the1e rods are threaded with square threads for aJout 2! feet length at
one end and aJout 1 foot af the othe'r end. If tlwse rods are not availa')le in one length two
rods 8 feet in length each may be used to form one anchor :001t by cou?ling the two pieces
together tightly with a solid coupling. To each of these rods, on the end having thr,~ads for
about one foot length only, are baIted together two rail pieces, R. S. Joists or channels about
6' x 4" in section and about 4 fect in length to form tIre anchor. For this purpose holes about
1i inches iJ;l diameter are drilled in the centres of the,se pieces and the anchor DOltS are passed
through these and tpe two pieces ooIted to the rod.
A pit about 5 feet in diameter and 7 feet in depth is dug at the site where a boring is
to be made, and the boring-pipe with the cutting shoe at its bottom end is placed vertically
in the centre of this pit. The two anchor-'Dolts with the channel anchors fitted at their bottom
ends are placed on either side of the boring-pipe and ~ept about 4 inches away from it. The
channel anchors are placed in a cross fonn. The pit is then filled with earth and well rammed
taking care that the boring-pipe and the anchor-bolts remain in their correct position.
A wood'en clamp is made from two pieces of sleepers 10" X 10" by 5 feet long.
Each of these two sleepers has a semicircular hole in the centre so that wh'~n clamped
together they form a circular hole slightly larger in diameter than the outside diameter of the
boring pipes. On either side of this hole and about 4" away from it the clamp has two
smaller holes about Ii inches in diameter into whichA:he anchor-.bolts are loosly fitted. These
two wooden sleepers ·are clamped together with four mUd steel bolts It inches diameter and
a ')out 21; feet long.
This loose clamp is fitted acr0SS the boring pipe and the tcp of the an::hor
bolts and the two halves are clamped together with the four-tightening bolts
described a'bove. The loose clamp is held in position and prevented from moving up or
slipping down by nuts and washers screwed into the top threads of the anchor-bolts on Doth
sides of clamp. '
A second wooden clamp similar to the loose clamp described above but with the
central hole tightly fitting to the outside of the boring pipe is also required. When clamped
together its central hole must grip the borin:~-pipe tightly. It is fixed to the boring-pipe below
lc()se clamp at such a point that the vertical distance between the two clamps is just sufficient
to put in two jacks with their caps fully screwed in On top of these caps iron plates are
placed to act as bearing surfaces when the jacks are worked.
This second clamp is fixed to the boring-pipe as de,cribed aJove and the jacks (20 to
40 tons capacity each) an~ put in position. The sludger pump is now worked and the jacks
are put into operation by unscrewing them, this exerts a pressure on the loose clamp. But as
it can not move up the pressure reacts on the bottom tight clamp itself. " Tht; bottom, clamp is
thus pushed down and drags the boring pipe with it. When the jacks have compeleted their
travel they are screwed in and taken out, and the verticility of the boring pipe check(>q and
 18

set right if necessary. Tb.e tight clamp is loosoned and raised up, and again fixed celow the
~op loose clamp at a P?int j?,st .low ~nou~h to accommodate the screwed in jacks. The process
IS repeated till the Lonng pIpe IS dnven m. Both the clamps are then taken out and a new
boring pipe screll cd to the bottom pipe and the operations repeated till the required depth of
boring is reached,
Another convenient method of applying load to the boring-tube is to use a
weighting frame mac'e up of duplicate R. S. Joists assemoled around the tu::e and
resting on the ground. A pair of winches are bolted to the joists of the weighting
fram~, and a pair of pulley sheaves to a clamp fastened on the boring-tube at con-
venient point above ground level. A!" dia:netcr rope with one end tied to the
weighting frame passes over these pulleys and is wound over the winch drums. As
the winches arc operated F,e load of the weighting frame and winches comes on to
the clamp on t:1C casing pip'~ whic'1 is thm pulle i d:) va and enters the bore mac.e by the
sludger. But n.e weighting frame is sufficiently loac'ed so that it is never lifted off the ground
and always rests on it, Thus there is no risk of its slipping down and injuring the workmen
operating it. This arrangement is also used for extracting horing tu es. For this purpose
the clamp with pulley shave, is fastened to the boring tul:::e a little below ground level The
rest of the operation remains the salLe as tefore and th tu')e gets pulled up.
The pipe with the weight on it, supported over a clamped platform, is, at times, rotat~
ed with manual labour to help its sinking down. In this case it is necessary to watch the
movement of the pip~ going into the bars, If the pipe goes in all of a sndden, i.e, if it slips, •
it will take along with it the sorrounding earth which will jam the pipe, and it will be difficult
to move it either up or down Hence it is necessary that the pipe should be held back with a
rope and a cra'J, and only allowed to go in slowly.
A careful record of the boring progress and the strata samples met with as also of the
morning and evening spring water levels inside the bore is to be maintained. When a hard
stratum is met with the sludger is replac d with different tools such as chisols underreamers,
auger, spring-rimor, etc., to pierce through it.
(c) "Morninr," "Evening" and "Working" water level.-(Indications they give with a
"Lailor-test" regarding the water contcnts of a stratum.)
The excavated material coming to the surface with the sludger will indicate what type
of stratum is being passed through at the bottom of the. bore. The existence of water-
cearing strata is indicated by tr.c presence o~ fine, medium or coarse sand, gravel, boulder,
or round-edged kankar etc. while clayey strata are generally non-water cearing. An indica-
tion of the pressure at which water is fio~ ing in a particular stratum is given by the level
to which water rises in the toring-tub~ wtcn it is passing through total stratum, It is clear
that when the bore passes through a non-water bearing stratum the water level inside the
boring-tu: e will 1e very low for there is no water coming into it from the non-water bearing
stratum at the bottom of tte bore. But as soon as a water bearing stratum is reached water
will rush into the tu'. e. The pressure of water in the water tearing stratum, as indicated
by the level to which water rises in U.e boring tn e, should ce carefully observed toth in the
morning, I::efore starting work and also Guring the progress of the work in the day time.
The former gives the static pressure of water III the core proviced no boring has been done-
at night and the watet has had tirr.e to attain its natural level; and the latter indicates the
working pressure. n.e working pressure is generally lower than the static pressure, for as
tLe sludgcr is taken out to ttc surfacc to cmpty It of tLe excavated materiaL SOlLe of tae
water also comes with it and thus throughout the day water is f eing tailed out from the tore.
Hence the working ani cycn;,ng water levels are effected by this continious bailing out of
II'ater. The working wat~r level also gives an indication of tte fiow of water in the stratum,
and if it is not appre-:ia'oly lower than the static water level it may be safely assumed that
t Le fiow of water in tLc stratum is satisfactory ann that it will give a good yield. A record
of morning water levels observed before starting work and of evening water levels taken
after stopping \\ ork s'1oukl \:;e carefully maintained. As a matter of fact it is desirable to
\~onr1uct a "Oailer lest' as each individual water-oearing stratum is met with while boring,
so ihat sun,l: infom:ation regan:ing tt.e probable yield hom the stratum may be had by
 19

observing the difference between the static water lcve 1 and the lowered water level in t 11J
stratum after a "bailer test."
In carrying out this test the slucger with its flap valve is used as a bucket to oail
out water. If the level of water insic'e the tuce-well remains steady even after ,Jailing Ollt
water, for say a couple of hours (apprc.ximately 80 'bails of 6" to 8" diameter and 8 feet height
in the case of 8" to 10" diameter torings), the stratum may te considered satisLlctory for
location of strainers, or for further cevelopment and gauging in the case of a cavity well.
(ii) Rod. boring. - Another system of boring is the rod boring system. This is similar
to rope boring except that rods e \ tend from ground level to the boring tool instead of a rope.
The rod boring system is used for very deep borings or borings in very hard strata. The rods
are of ash weoel ~ith the metal male 'and fe:nale screws spliced on to their ends, or steel
rods with screws (in length from 3 feet to 20 feet and fro:n 1 inch to 3 inches tJ_ick). Boring-
rigs are constructed either of wood or wrought iron tubes, with a head-puIley-DIeck over
which a wire rOFe is passed to support the boring rods and tools. The other end of the
rare passes to the wind-fass in the hand-power-rigs or to the winch or winding drum in a steam
driven plant. •
In order to obviate the jarring e 'Cect of the great length of rods, particularly in
very deep borings, a tr~p link is introduced 20 to 30 feet a'.::ove the tool as it is dangerous to
exerci~e a percussive actlon of such power which will expose the lower rods to the danger of
breaking. The method usually adopted is to employ a form of sltding joint, on one of the
rods, 20 or 30 feet above the tool. This is based on an lllvention of Oeuyenhamen. It
consists of a slide joint of two parts which are aole to slide upon one another for a
distance of a')out on foot, and so arranged that during the descent one Leco:nes detached
from the other. The upper part is balanced by a counter-weight suspended to a lever, and
the lower one only allowed to act by percus~ion, Certain adaptations of this joint in the form
of jar-bars and other sliding joints are now generally used.
Square rods with taper joints are sametimes used for shallow borings. In this
method earth auger, cross chisels, flat chisels, Dull-nose auger, etc., are used as boring tools.
The square rods are rotated with drill, and the auger is allowed to penetrate in the earth
about one foot and then withdrawn. The earth is removed from the auger and the clear
tool is again inserted into the bore. The above process is repeated till the boring is made
to the required depth. In Gujrat bores up to 150 feet are sunk with this method and with
manualla'bour. For shallow borings this method is generally more expeditious than rope-
boring.
(iii) B)ring by HydrauliC Wash or Water-Jet-llorlng.-At places where su 1icient water
is availade for working a water-jet plant, the hydraulic washing system is a very convenient
m~thod .of boring t~lrough. hard and tenacious soluble days in w~ich progress .of boring
WIth ordmary perCUSSIOn iJormg sets would be very slow. Hollow oonng rods or pIpeS, with
a s",:ive~-he~d at the top a,nd conn?cted to a. chisel at their bottom end, are used and a rotary
mohon IS glVen to these oy rotatmg the water-tu:Jes from ground level. A sIaN rotary motion
is given to the boring-pipes aha to e {pedite their sinking. For 8 to 12-inch diameter borings
going to depths of 3 to 4 hundred feet, ordinary W. I. pipes a'')out 4 inches in diameter are used
as hallow boring rods. These are sus:_Jeade,j fro:n a pulley in the derrick of the oaring rig with
an adjustable length of wire rope, and pass inside the boring tube. At its bottom end the
water-tu e is fitted with a chisel having two nozzle orifices, one on either side of the chisel,
The nozzles are kept at about 9' aJove the bottom of the bore. A suitable connexion with
an electric or oil-engine power pump is made (at the upper end of the pipe) by means of a
flaxi'Dle hose pipe, and water is pumped ..into the tu:;e. The capacity of the pump silOuld not
ordinarily te less than 15,000 gallons per hour and it should be capable of working up to heads
of 200 feet. The water which is pumped through the tube to the bottom of the bore-hol~
shoots out through the nozzles and rises in the annular space ,etween the water-tui;e and'
the casing pipe carrying with it any comminuted material and overflows into the settling tanks
at ground le,:el. The vel~city of water in this annular space should be su ncient to carry
all tte commmuted materIal with it. This velocity can be regulated by controlling the dis-
charge from the pamp. After settlement and clari'ication in U.e settling tanks this water
ean be used again for boring purposes.
 20

Although this method has proved efficient it is not very widely used owing to the
difficulty of obtainin<Y su'Iicient water to ensure a constant supply at full pressure at the
Lottom 'of the hore. " Where it can be used there is a great economy in time as the tool needs
to te raised less frequently, fn addition a clean working face for the chisel is m~intained
at the hottom of the bore-hole, Care should, however, he taken to record changes m strata
very vigilantly when working with this method.
As recording of morning, evening and working water levels would not be practicable
with water-jet i;oring, a cO'11')ination of water-jet boring and rope boring methods has been
adopted in many cases. The leoring is started with the water-jet system, but as soon as a
water-bearing stratum is reached the water-jet phnt is put aside and the rope boring system
is adopted. This enable accurate records of the variom water levels to be kept as also of
the depth and sa'11vle of the stratu-:n. When the water-bearing stratum has been crossed
the rOj::e l.oring plant is removed and the water~jet plant is used again. A combination of
these two methods gives expeditious results and incorporates the advantages of both the
systems
(iv) Rotary bori'lg.-The rotary system wao, originally devise1 to drill wells through
unconsolidated rock, such as beds of clay, sand, etc. Where tho material to be .drilled through
is soft, the rotary syste:n is used to far greater arlv 1. n tage than the standard percussion
system. Here speed is a great factor, as much' as 1,000 feet having been drilled by this system
in 24 hours (Bi·,). 3).
The practical operation consists of rapi'iily rotating a column of pipe, at the lower
end of which is a cutting ·oit. Iri the bit are hole':> through whic':l water, with clay in
suspension, is sent under pressure The fluid mixes with the cuttings and carries them up to
the surface in the annular space eetween the drill pipe and the bore hole. The clay that is
held in suspension in the water "muds up" the walls of the hole, prevents caving, and causes
the pipe to turn more rapidly as there is practically no friction tetween the boring pipe and
the silles of the bore hole. The muddy fluid, cecause of its greater density, also trings to the
s~rface cuttings which could not te lifted by water alone.
A derrick like that of the standard rig is used, tut machinery and tools are unlike
those of a percussion outfit. A pump with proper pipe connections is provided for feeding a
constant stream of the mud flush to the tools when drilling. Usually a second pump is kept
ready for use as a stand by. Thin mud or slush plays an important part in drilling. A slush
pit, an essential accessory. is usually dug near the derrick a'1out 40 feet long. 15 feet wide, and
3 to 4 feet deep. A ditch where sand may settle out of the mud is cut from the well
circuitously to the slush pit from which hose or pipes lead to the pumps.
A machine called the rotary, is used f~r rotary ho~i~lg. This rotates the drill pipes and
at the same time permits th'~ir b,;ing lo.verc l..in be :)Or0 as re:ruire-i. The chief parts of a
rotary are a rigid base and a turn-ta.)le rotated ny a system of gears actuated by means Qf a
a: chain drive sprocked. The drill pipe is held by a ca )le, which passes over the crown pulley
;(ttached to the Doring rig and around the. drUll shaft <If the draw works. This drill pipe is
clamped by a Dlock to the turn ta lIe and forcej to rotate with it. Modern rotaries are more
elaborate and are fited with dou';le turn ta'Jles, special clamps drive rings, etc., having
interchangea.Jle hushings for holding casing and turning drill pipe. .
A constant stream of mud slush or water must be circulated through the hole while
drilling. This is supplied by pumps which force the water through a hose pipe and a rotary
swivel connection and then cown the drill·pipe.
Wrought iron or stael pipe is used fo,r drilling purposes. This comes in random
lengths of a')out 20 feet, it is a'::out {th to ! an inch thick and from 21 to 6 inches in diameter.
Special mtary tool joints are used to connect two pieces of pipes, in place of ordinary couplings,
for greater strength and safty.
When the bore has 1 een drilled to the required depth casing of l;e proper size is inserted
to support its walls.
(v) Core drills.-The rotary system of toring with -core drills is hardly ever used when
horsing for water' in sand, gravel and clay as suffiient water is usually fouud at depths which
 21

do not justify the use of this machinery. But this syste n is largly used for oil borings when
great depths are atiained for rock boring. In rock sand-stone, trap, slate, granite, etc., the
rotary core drill with shot bits is found most effici~nt.
The ordinary core drill consists of a tool called a ('crown" which is a short
piece of cast steel tute, into one end of which a number of "black diamonds" are fixed
circumferentially. ' Black diamonds are an amorphous variety of diamond usually called
caroons and are only valuable on account of their extreme hardness. Chilled steel shots
are now generally used in ,place of dia'11orids, The upper end of this crown is screwed
to steel pipes of the rotary drill. Machinery on the ground surface causes the pipes
and the crown to rotate aJ explained in the case of rotary boring. and the crown cuts
through the strata causing a "core" to rise up the hollow tu~e into a core barrel. Lead
balls or some gravel a')out !" to i' in dia'1leter are then injected with water in the bore so
that these balls or gravel rr:ay act as a wedge between the core and the crown. If the
tool is now rotated fro;n. top tre core gets sheard 0'1 at the bottom and breaks. The
tool is then raised and take:l out and broken core comes out along with it. The crown,
when it has been taken out, is given strokes with a copper hammer and this makes the
core drop out'from it. Fro:n time to time the core is broken 0'1, the tOClls raised, and the
core extracted. Water is pU'11ped down the hollow rod as in hydraullc washing and rot lry
boring and enters into the bore hole acting and enters into the bore hole acting as a
IUJricaht to the drill and carries off loose deJris when returning to ground surface through
the annular space between the drill and the bore hole.
(vi) Cllifornia or strove-pipe method of boring-This is a method of boring used for
wells of 6 to 30 inches in diameter. The casing consists of short sheet-iron cylinders, forced
down by large hydraulic jacks and perforated in place by a special tool. Materials within
the casing are excavated Dy a sand bucket (sludger pump).
This method is chiefly used far unconsolidated alluvial depoists and is known as Cali-
sornia method, as it was originally developed there.. It is also called stove pipe method on
account of its special casing. The casing is made of lap riveted or welded cylinders of sheet-
iron or steel, usually of 10 to 14 gauge, 24 inches long, and 6 to 30 inches in' diameter. Two
sizes are used one of which just fits within the other, so that the joints of one may be adjusted
to fall midway ;:etween jojnts of. the other. Sections are added one at a time as sinking
proceeds, each 2 feet section adding 1 foot to the length of the casing. Outer and inner
sections are united simply Ly centing with a pick; and the casing is watertight. Casing
may te started from a prorerly recessed steel drive shoe and. drive pipe. Casing is us,l,lally
sunk by two or more hydraulic jacks iJuried in the ground. and pulling at a casing yoke placed
on top of the casing pir::e.
PI. sand pump is used for horing as in the case of rercussion rope boring described
before. A strata chart is carefully kept, as usual and after the well has ceen sunk. to the
required depth a cutting, knife is lowered inside the casing anJ vertical slits cut in the
casing in sclec:ed "ater-', caring strata. Revolving cutter used is such that it punches
five hc1es at eacl reVolution of the wheel. Vertical slits are of a form and
size that do not clog reaqily. For best results in fine material, a' natural strain.J'
is forn ed surrounding the casing hy re;noving the finer materials adjacent to the pipe by
pumping tl e v.ell as low as possi'le for several days.
It is seldom necessary to reperforate a well casing. The yield of an old
well may often be increased by using compressed air for back-blowing or' by heavJ
pumping, thus loosening the surrounding sand and gravel. (Bib. 2)
6. Tube-well pipe.' (An extract is taken from Bib. 28.)
(a) Tnreads on boring-pipe.-from American practice it has found that the ~
threads pipe (i.e. 8 threads to an inch) lacks strength, owing to the deeper cut
required for the thread. It was found that the pipe often broke in the thread when subjected
to strain. BesiC'es, the 10 thread pipe proved more consistently watertight and less liablE
to admit particles of sand and grit, which would bind the thread and destroy it in unscrewing,
The Americans strongly recommend use of 10 thread pipe' (Bib 33).
It is howover eXFerienced that when casing pipes with 10 threads to an inch art
,jacked out, during the process of extraction from a bore, they at times slip from the joints
 22

.pecially if the threads are partly worn out. Casing pipes with 8 threads to an inch a~e
better in this resj::ect tecause their threads are. cut deeper and ~herefore they do not S~IP
;0 easily. But greater depth of threads necessitate~ that the thIckness of pIpes should l!e
correspondingly greater so that ~hey I?ay ~ot lack m strength. and may not part (Dreak)
at the joints. Generally speakmg pIpes WIt~ 10 t~reads .to an ~nch are preferable where the
thickne~s of pipes is less .than 1/4th of an m~h, lJut wIth thIcknesse~ cetween 1/4th and
3/Sth of an inch pires wIth 8 threads to an mch are preferred espeCIally when used for
boring purposes.
(b) Collar-bound pipe.-Boring pires if socketecl sometimes become "collar-bound"
in a hole. In such an event the pipe can be worked for about the length of a single joint,
but will not pull above a certain point. This is caused by mud loading heavily upon the
collars, and by a mud-laden collar, usually the 'r:ottom one, coming in contact with a tight
place in the hole. The condition may te aggr~vated .by the .sand filtering in :'lnd ,findinl? a
resting place upon the mud-laden collar, (That IS why It IS advIsable to hav,~ tonng pIpes wIth
tlush joints, fig. 1, where tte collar at tl'e joints are disrensed with), In such cases, it is necess-
uy to keep the pipe m constant movement until it will pull freely pass the point of o:)struc-
tion and pass easily down through the same place; otlcerwise, such a contiition is sure to cause
the pipe to 'freeze' rapidly, The expre,sion "frozen pipe" is applied to toring tutes rendered
immovable in a bore by muri, sand, limestone cuttings, and other gritty substances, settling
nound the outside of the tube.
When a pipe is collar-';_lound, too powerful an effort to pull it through the tight
place in the hole will often cause it to become lodged so that it will neither pull nor
fall back to the bottom, a common occurrence in the Vijapur area of the B:uoda State. This
condition should, if possible, 1:e obviated by the exercise of patience in keeping the pipe in
~onstant movement. Usually a collar-bound pipe, if worked patiently up and down the ';)ore,
the collar 1eing pulled gently into the tight place and then lowered i'efore the pipe gets lodged,
will clear itself sufEciently of the sand and mud gathered upon the collars to permit free
movement, Unless great care is exercised in controlling the ripp, it may be injured in dropping,
because a pipe thus lodged drops with a tremendous force under H.e impetus of its own weight
when once starteo. Therefore a free fall of even a few inches may result in serious damage
if the pipe is allowed to strike the bottom.
Sometime dynamite is used to give jerks to the pipe and thus to loosen a jammed
pipe from the surrounding earth. The vibrations taking place at the time of exploding the
dynamite shake the whol\.' length of the boring-pipe and also give vibrations to the surroundi.lg
;oil. Due to these vibrations the ~oil round the pipe which has stuck to its sides is likely
to get loose and the boring-pipes tecome free. This process was tried in Padra (Baroda State)
)oring for the extraction of the jammed casing pipes with considerable success.
(e) Bad pipes i,e. pipes in a siRte of partial collapse, or split up, while l;oring.--A pipe
:emaining in a bore for some time will frequently "go j'ad". This expression refers to a pipe
that has become dented from a stone bruise, etc, and is in a state of partial collapse from
outside wall Fessure or is perhaps split and partly flattened, The pulling of such a pipe for
the purpose of replacing the damaged joints is often inadvisable tecause of the danger of
breakiug t.he pipe at the "bad place" and leaving the lower part of the string of pipes in the
bore, Such a pipe is oftenJrozen, and attempts to free it by applying too much force would
~ause the weakened joints to part.
"Bad pipe" is usually first detected ty the tailed or sludger hesitating in its drop as
it is let down the bore. This hesitation if recurring continuously at the same place, leads the
borer to suspect that the pipe is giving way at that place.
One of the remedies for a bad pipe is "swaging". The 'swage' is a heavy piece of steel,
oval in shape, with a small groove cut into the steel as a watercourse. At its greatest diameter,
the swage should be a fraction of an inch smaller than the diameter of the boring pipe through
which it has to pass, Thus a boring pipe with a 10 inch inside diameter should permit the.
entrance of a swage 91 inches in diameter. The common method is to drive a swage of a small
diameter through the bad place first. This swage is then withdrawn and a swage of a larger
iiameter is used. This procedure is repeated until a swage of maximum permissible diameier
 23

is driven through the impared pipe. The swage is rua below the stem with long stroke jan
and tools working Lnmediately a Jove thc swagc. The operation of driving thc swage down
through the bad pipc and jarring it back throug:.1 the sa'1lC place should be repeatcd un till the
swage will pull frccly through the impairc,l pipe, It can then be safely assu l1cd that the pipe
has i cen extendcd to its normal size. If a swage has 'teen driven through a spilt joint of casing,
and cavings are falling into thc pipe through the split portion, great care s:10uld 'J:e taken
to prevent the cav;ngs froTI wcdging the sNage in the pipe, the swage bcins so nearly flush
with the pipe, tJ.at any hard obstacle, may causc it to wcJge.
7. Sand hIowln~ cI.·,<fin; boriJg, i.e., Sanl rushing into t!te b~_'inJ pi,;>e during boring. (An extract
from Bib, 28,)
While boring is in progress there may be encountered a stratum of fine sand, or sand and
clay etc., which rus'les up the bormg pipe to a consUerable length and thus hampers the pro-
gress of further boring. This sand is generally very fine (with or without clay) and dry. It would
appear that this particular sand stratum is sandwiched between two strata and is under
considerable pressurc, It may also be that it is in an arc 1led condition, all the while, betw€en
the two strata aJOVc an,t below it. When t'lis stratu'll i3 struck during boring its stability
is disturbed and thc iJoring pipe provide.> an opening for the release of its pressure. The
sudden release of pressure causes the sand to blow up ins:de the boring pipe to considerable
heights, In one well in Okha, Baroda State, sand was blowing up everyday from a depth 01
680 feet below ground level to 140 feet, i c., to a height of 540 feet.
The di,licultya'lout sand bloNing is that when the hore is very nearly cleared of all
the sand and the '00ttO:n end of the boring tu',le is reached, the sta:jlity of the sand underneath
is again distur'jed and it blows into the pipe up to the old level and during this blowing up
the drillus tools, etc" may get struck up or jammed or damaged in the rushing sand; and i1
provides a di'1icult pro:)lem to deal with.
The simplest mdhod is to release the clamp which is holding the boring pipe so thai
the pipe may sink down under its own weight whenever suticient sand has been cleared from
inside the bore. All the while the sand is teing cleared out, the boring pipe should be kept
full of water right to the top, which will keep th) sand under pressure and prevent it from
rushing up, specially when the pipe is being cleared near its bottom. As the boring pipe is
unclamped it will sink down by its own weight when the sanl is cleared beloN the cutteI
shoe. In some cascs it may becoTIe necessary to e:den,l tIle ;)oring tu')e a'Jove ground lev31
by adding another piece on top of the tU'Je, and filling it wit~ WJ.t~r to increase the head of
water over the rushing sand. This process of watcring cleeming, ani sinking is followed u].til
the stratum has been crossed by the :;oring pipe
If the thickness of the stratum is great and the pressure of sand high it may not
be possiole to pic ~ce tllfough the stratu 11 in one day. In such cases even though the pip.:
is kept full of water during the night it may not be a')le to prevent the sand from rushing
in durin~ the closure hours. The well at Okha, referrei to above. is an exa'1lpl of such
sand blowing. Every mornin~ the driller had to spend a')out 5 hours in cleaning the sand
before he could commence further sinking of the boring pipe.
The principle followcd in the above method is to line the bore while at the sa'1lt
time taking out the sand, whic',l is prevented from rus\ing in iJY keeping it under pressure
of water. It is o'lVious that if this pressure of water on the sand coulrl be maintained
there would be a lesser chancc of the sand pus':ling its way tllfoug'l and blowing in even
during the clOSUie hours. This is achieved by the use of '·aquajel." "Aquajd" is a powder
of the colour of cel1cnt and hcavier than clay, Wh~n mixed with water it forms a jelly
which is very tenacious and whic\ d0e, n'Jt m:x with sand. So w:le.:1 t'le trou')le of sand
blowing is encounte:-ei 'aquajel' is addcd fro n the top of thc boring pipe It, being heavier
than water, sinks b tl.C botto n, W:lere the Joring to)l is wC)rking. By the action of any solid
boring tool (not a sludger w:lich will tend to collect t1)e 'aquajel' inside it and bring it up
with it when the sludger is lifted up) the 'aquajel' will get mixed with water and a jelliy
will be formed whic'1 will stick to the side, of the bore as a plaster and will bind the sand
and keep it in place near the bottom of the boring-tu'Je. The boring operation can then be
stopped, and the layer of 'aquajel' will form a sort of bur) at the botto'll of the boring-tuhf
and prevent sand fro.l1 blo Ning in during closure hours,
 24

If the rushing sand is coarse it indicates good water bearing stratum provided its
iydrostatic pressure is high. When such a stratum is struck water will rush in the boring
cube and along with it may bring in sand also. But "this sand do;;s not rise so high as the
lry fine sand referred to above. It is easier to penetrate through such water bearing stratum
ind a sand pu,:np su"tices for this purpose as the sand docs not rush in so often or as high. '
~. Fishing t~ols. (An extract is taken from Bib. 28.)
Wh')u sludger or other tools re:nain imide a boring pipe due to the snapping
()f a wire rope there are different types of fishing tooh which can be used to fish out the
)ludger. etc. fro:n the bore. For a detciled description of these, "oil gas and water weUs'_'
by Lucy Manufacturing Corporation Ne\v York, or other' Looks on this sui ject may be
referred to Bi') 6.
The handbook on oil wells, published by the Gil Well SUPFly Co., Fittscurg, U. S. A.
ane also the handbook pu lished by the National Tools Supply Corporation, Oho, U. S. A.,
He good refrence books on this sU'Jject.
A borer should keep a log t)ook in which he should keep measurements of length
and diameter of each tool, their thickness at various points, length of threaced parts, numi:er
of threads per inch in the various tools, etc. These constitute 'important information for
a borer~ for in case of an accident he will have full information to prepare fishing tools, of
the correct size easily, to ena~le him to fish out the broken parts of tools left inside a bore.
9. Types of tube-wells. (An extract from Bib. 28.)
Tu'pe-wells generally comprise the following types : -
(1) Strainer wells (including radial wells).
(2) Cavity tube wells. '
(3) Slotted-tubewells (including shrouded wells).
(4) Perforated pipe tupe wells.
(S) Artesian tu',)e wells.
In a tube·well we can draw water not only from the water L-earing stratum nearest to
the surface of earth but also from one or -more of the water tearing strata lower down, and
this makes it possible to ,draw out a larger quantity 6f watar fom a tu'ce-well as compared to
to an open well. The water bparing stratttrr1 nearest the ground suface is not. however, tapped
in the case of larger tube-wells in order to avoid interference with the supplies of the adjacent
open wells and prevent their getting depleted.
Open wells generally get their supplies from the first water-' earing stratum When
we start boring from ground level we reach the' uppermost water' tearing stratum in which
water stands at we call "spring water level" or "static water level" and it is this stratum
which feeds most of the open wells. If we continue boring we pass through imf:ermeable
,trata of clay, conglomerate or stone alternating with other water~'~earing strata lower down.
in the case of tube-wells we tap one or more of these lower water-bearing strata.
Radial wells and ground water collecters.-In some cases it is found that vertical
,trainer wells need replacement very frequently owing to 'rapid incrustation,. of the strainers.
ft is generally 'relieved that the incrustation occurs from the lil eration 'of dissolved car",on
dioxide as a result of the release of pressure or change of temperature due to the tapping of
underground water from aquifers at diferent depths. The loss of CO 2 decrl/ases the solvent
power of water and causes precipitation of the minerals on the vertical well strainers.
As an improvement. radial wells, with radially driven strainers, have i:;~e;n constructed.
These strainers are driven radially in a horizontal plane in the same water-teanng stratum and
are thus at all times under the same water pressure and temperature.
A shaft of suitable diameter (12 feet or more) is sunk down to the water bearing
stratum to be tapI=ed. " '
The shaft is constructed of rein£orced sections (air-applied concerete and sheet steel
sections) each a: out 10 feet long. At the Lottom there are tr.eble-walled steel sections
comprising the cutting shoe, the intermediate and port hole sections. These are all welded
together with the annular spaCE reinforced and filled with concrete.
Through the port hole section the slotted screen pipes about 10 inches in diameter
(with a pointed driving end are driven to about 200 feet length horizontally with specially
 25
constructed sliding hyraulic packer. The screen pipes are made from i" Copper b~aring steel
plate punched with 11" x 1" slots.
There are about 36 screens driven radially from one shaft ~nd kept just above the'
clay ted underneath. (For London Water-works Installation, thl)se screens were kept one
foot above the clay bed.) The open ends of these screen pipes project inside the central shaft
and are fitted with sluice valves and vert:cal back-wash pipes which extend to ground level,
and can be operated from there. All these s:reens are backwashe.d and properly developed.
The sand thus taken out averages about 3 c. ft. per lineal foot of screen pipe projected.
The water from all these screen tu·. es flows under gravity into the central shaft and
ispumped out with deep well vertical motor driven pumps.
The yield from one such radial well is of the order of 7,O;)() g.p.m. with a drawdown
of about 7 feet."
Cavity wells are tube-wells, which, being without strainers, draw their supplies from',
one acquifer or water- oearing stratum only.
10. Suitability of the Tract lor Tu1:e Well IrrIgation.
(a) Geolo~lcal strata.
Below the soil crust in the ground water reservoir, there should be available ~
6
coarse sand stratum from 100" to 150 deep to locate the strainer of a tu')e well. Geological
conditions in the Punja~) are suc':l that unfortunately such a water bearing strata are not
available everywhere. Trial bores are ab;olutely essential to ensure suitable strata for every
individual tube well. Water bearing strata are often separated by clay or kankar bed layers
which are only partially pervious and yield muddy water. The trial bores indicate the
position of such layers.. If ~he strainers have to pass through such layers, the portions should
te closed by putting blmd pipes.
The Geological conditions In Northern India:-
"There are three crests of the crustal warpings or underground rock ridges sweeping
across the Northern India as shown in the map of the Punja~ Plate I, Vol. III. One major
and a number of minor crustal warps, run transversely across the three main crests. These
ridges of rock, varying in dep.ths'fro~, at least 2,000 feet to surface outcrops, have an important
bearing on the flow of SUb~SOll water. > '

"The middle one of the three main crustal warps, is of vital importance to the
Punjab and United Provinces. It runs from Sargodha, through to Delhi and thence to
Allahabad, Bhagalpur, and ShiJong. The major transverse crustal warp appears to run
from the head waters of the JlJ,mna river, via Delhi and Ahmadabad to. BomJay. The
northern portion of this warp follows, very approximately, the boundry line of the Punjab
and the United Provincess The rocky floor, of which these crustal warps are an integral
part. is covered by.alluvial deposites of th~ Indo-Gangetic 'pl~in. .The entire str~tch from
Karachi to Calcutta IS of the same geolog1cal type, but W1thm th1S type there 1S marked
difference of detail. In that area of the United Provinces covered by the upper waters
of the Jumna and the Ganges, the. alluvial sub-soil i<> lenticular in formation. Ltmticles 01
clay are contained in a matrix of sand. East of the line LucknowJ~awnDore, the clay
becomes predominent and the characteristic tends to reverse. In the Punjab, west of the
major transverse warping below the Jumna, already referred to, the characteristic change!
sharply. The clay appears ir: beds or layers, of varying thickness and very varying extent.
In some instances the sub-senl water appears under pressure between two clay bends and
acquires artesian propertie~. In the Unite? Pro,,:"inces the gener~l d~rection of the sub-
soil flow is that of the Ju'nma and Gange3 flvers, m an easterly dlrectlOn roughly parallel
to the main crustal warps. So far as the3e crustal warps are concerned therefore, the flow
is unimpeded in t~1e United Provinces though obstruction is probable in Bengal, between
Jal Paiguri and the Ganges river. But in the Punjab, the direction of the sub-soil flow is
roug'oly normal to the main crustal warps and is impeded by the second on the San;olha
Delhi line." RY GKVK
UAS LIBRA

11\\1\111\\111\\\111\1 \111 1\\1

1135
 26
(b) !oll.
The soil survey should be carried out hy taking samples for every 100 acres of the
proposed area for tube well irrigation. The pH value should be from 7 to 9. The significance
of the so called pH value is explained in Part II Chapter II.
(c) SuitabilIty of the Pumped Water.
It is determined by working out the Salt Index as explained in tho previous chapter
on well irrigation. Tho Salt Index should be negative for suitability of the water for
irrigation purposes.
(d) Source of Water Supply.
The source of water supply should be such that the yield of the tube well does not
deteriorate in course of tirr:e. The yield is effected in two ways; firstly by the choking of the
slrainar and secondly by insufficient inflow to replace the depletion when the soil is not
suffciently rervious. The first factor can generally be controlled by an engineer by a suitable
design of the strainer, but the second one is often beyond the control of the engineer due to
the irregular presence of the clay lenticles which are only partially pervious.
11. Tube Well Chaks.
(a) Size of Chak.
The size of chak depends on the discharge of the tube well and the delta or the depth
of the water required by the crops which will be raised. There is yet very little information
on the delta performance of the tube well irrigation in the Punjab. This cannot be compared
to the United Provinces irrigation from the tube well because the average rainfall of the tracts
~rrigated in that province is from 30" to 40' per aAlnum, while even half of that is not availahle
'In the Punjab plains. In the PunjaD canal irrigation is fully developed and the delta:
~tatistics of this are availahle. The average figures of delta for the canal irrigation are given
III Part II Chapter II. The delta performance of tube well irrigation is likely to be about
three fourth of canal irrigated depths of waterings to mature a crop. The Karol tube well
irrigation scheme has teen based on an average Rabi delta of 1'1 ft. and an average Kharif
delta of 1'7 ft., while the adjoining Shalamar Disty., of the Upper Bari Doab Canal has an
average of 1'57 f1. in Rabi and Z·57 ft. in Kharif. Assuming that the pump will work for 5000'
hours a year, it willI e able to deliver annually acout 420 foot acres per cusec of the pump
supply. Allowing th depths of watering as stated a! ove and assuming tte discharge of a tube
well to be 1'5 cusecs, the Rabi area is likely to ·[·e 190 acres and Kharif area 1.20 acres. With
irrigation intensity of 75% an average chak works out to be 424 acres.
(b) Suitable discharge for a tube well.
Excessive discharge for an individual tube well results in wastage from breacheo on
water-ccurses and excessive depths in the fields. Too little discharge needs a extra number
of wells and results in wastage in the fields by taking relatively longer time to fill a field.
A discharge of 1'5 cusecs is considered to be suitable for a tube-well chak. Depth .0£ first
watering is likely to te 4 inches and that of the subsequent waterings 3", which are very nearly
the same as for the canal flow irrigation, but the major economy of water is likely to result
from lining of the water course, and from division of the fields into small kiaries, because in
this case a cultivator has to pay for the energy or the volume of water used.
(c) Location of Tube Wells and Water-courses.
A tube well should t e located near the centre of the area to be irrigated, so that
the radiating water-courses are not more than one mile in length in any direction. It should
be situated at the highest place in the chak so that the water-courses do not run in embank-
ment. The depressions or drains should form the boundary of the chak. Main water-courses
should be aligned in such a way that no field is more than half a mile from the Government
water-course.
In designing water courses the usual practice is to provide a field command of '2 ft., a
slope of 1 in 3000 in tte subsidiary ( zamindary) water-courses and a slope of 1 in 5000 in main
(Government water-courses}.
 27

. It is de.>irable to line the main water-courses to save absorption or percolation losses.

Tiles one and a half inches thick laid in cement mortar 1 : 4 will serve as a useful non-erodable
lining, while the sodium caroonate, lining is liable to be easily damaged and subject to cracking
on drying, .
12. Selectio:l of Pump Sets.
An extract is given from the Punjab Engineering Congress (1941) paper No. 248 by
Messrs. H. L. Vadhcra and A. R. Talwar.
"Selection of a sllitable type of pump is of primary importance in any tube-well
scheme. Before making the final choice, different kinds of pumps and headgears installed in
the United Provinces and elsewhere were examined."
"Centrifugll pumps with horizental spindles are usually installed at the nor.nal lowest
level of the sub soil water table to ensure constant priming. Due to the occasional rise of
water-tahle during the mOJ.soon season, When the demand is slack, the installation is in
danger of being drow.lerl by be rise of the seepage water in the well. Actually, a numeer of
motors are said to have been ruineJ in the United Provinces in this manner."
"The di nculty has been overcome by using centrifugal pumps with vertical spindles
and driven through vertical shafts from vertical spindle motors installed 8 to 10 feet aJove
the pump. The only drawback with this type of pump as o:£ered by the tendering firms was
that its discharge varied more rapidly with a variation in the total pumping head,than in the
case of other types of pumps, otherwise it is the most suitable type and has been proposed to l:::e
used on the Karol TuDe-Well SchcT.e."
"The propeller type of pumps gives a fairly comtant discharge against all heads, but
its efficiency is comparatively lower than that of the centrifugal pump. The bore-hole type
of pump was also examined, but on account of its lower efficiency it has not been adop cd
for the Karol Scheme. The use of a bore-hole pump is obligatory when the suction
head exceed 22-24 feet and sump wells cannot be constructed to place pumps of other types
at a lower level due to large changes in water table. . '
The vertical spindli)' certrifugal pump which have been used were specially manu-
factured by the Harland Engineering Co. Ltd., and are known as the "Spiroglide pumps."
The following special features of the Spiroglide pump set are noteworthy:-
Pllmp and Motl)r. The pump is provided with cutless rubber bearing instead of the
standard ball and rollar bearings. This arrangement renders the bearings immune against
the effects of dampness or flooding of the well. It also means that the pumps require no
periodical lubrication, one gland being the only part requiring occasional attention.
The headgear is extended downwards and provided at its lower extremity with a
steady bearing. The unsupported length of vertical shafting is thereby considerably reduced,
giving additional rigidity and freedom from vibration.
This feature avoids the necessity of the pruvision of an intermediate steady bearing
for the vertical shafting carried on a girder spanning the pump chamber, such bearing being
prone to get out of alignment in course of time due to the settlement of masonary walls of
the chamber, which alter the position of the girdtlf supporting the cearing, thereby causing
vibration in the vertical s'1afting.
The weight of the vertical shaft and pump motor is carried on a thrust ~earing
located in an accessib~e position in the headgear, where also is the means of adjusting the
position of the shaft and the impeller in ttc pump casing.
The vertical shaft between pump arid motor is of an abnormal strength, teing of 3l"
diameter. The reason for using vertical shafting of such a large diameter is to ensure that
the first critical speed of the shafting is beyond the full load speed of the pump sets, so that
when the pump is bJing started when running and when being shut down, the shaft does not
pass through a critical speed which would cause vibrations.
The motor-supporting headgear is of massive construction and has a large area base,
machined on the under-side for supporting on chamber. After erection, dowe pins area filleted
 28
through the headgear base into the girders SQ that should the neccessity arise for the headgear
to be renewed there is no difficuity in re-erecting the pump set in correct alignment, the gjrders
being levelled when first installed, so that no .packing strips or shins are used between, the
headgear and the girders."
13. Selection of Strainers.
A few types of straines usually used in Northern India are described below :-:--
(a) Cook Strainer.
This is manufactered in America and consists of a solid drawn brass tube slottered with
wedge shaped horizontal slots. The slots are cut with a circular cutting tool from inside the
tube, to variouse gauges to suit the coarseness of the sand, the usual gauges in the ~e!ltral
Punjab varying between 6/1000" and 16/1000". The strainer lengths are generally )Olnted
together by means of screwed collars of brass .
(b) Tej Strainer.
This is manufactured 1 y tr.e Rtr<. ble Water Supply Service of India and
consists of a brass tube consh ucted of a 1 rass shtet bent round to form the tube,
the vertical joint being brazed. The slots are wedge shaped and cut cefore the sheet
is tent to various gauges to suit the coarseness of the sand. The strainer is made in 8' lengths
generally and from 3" dia: upwards, the lengths being joined together by means of screwed
collars of brass, '
. This strainer is similar to the Cook strainer, except that it is not made from a solid
drawn tube. It is niether so robust as the Cook, nor are the slots cut so accurately. It
possesses the advantage of ceing considerably cheaper and is easily obtainable, ceing locally
manufactured. .
(c) Layne and Bowler Strainers.
They are iron slottered strainers. The strainers are made in America and, consist of
wedge shaped st.-el wire wound to a suitajle pitch round a slotted or perforated steel or
wrought iron pipe the lengths teing jointed together by scr !wed collars. These ar,~ heavy
and ronust and can stand rough usage, but having to be importe 1 from from America, it take.>,
a considera~)le time to get delivery.
(d) AsUcrd Strainer .
•
The strainer is generally made in 8 feet lengths, the lengths teing·jointed by half rings
bound with wire and soldered over. This strainer has to be very carefully handled lest the
(wires are broken or displaced.
(e) Brownlie COllvolutei Strainer.
The strainer consists of a polygonal convoluted steel plate round which a copper mes~l
· itrainer consisting of heavy parallel copper wires woven with copper ri::>bon is placed.'
(f) ltggett Strainer. "
· This is new strainer provided with cleaning devices in the shape of cutters which can
•be turned in the slits and it is claimed, by this means, that the clogging of the strainer by the
, deposition of solid matter on the outside of the strainer and in the slits, can be prevented.
'The cutters are operated from the surface. The strainer is said to be somewhat expensive.--
• t
(g) Phoenix strainer.
It is calcium plated and is supposed to be free from the danger of choking and corrosion
caused by the chemical action.
All brass Tej strainers appear to be the best because br(,lss is not readily acted upon by
water and is easily procurable locally Coir and Munj rope strainers have been used in the
Punjab with success for .small supply tube wells meant usually, for domestic purposes.
Hard wood slottei strainers are at pre3eat in ;:.a eX.?i!rimenhlstag,~ hr ilie on tube wells in
the Irrigation Scheme in the Punjab.
14. Choking of strainers.
The strainers get choked up usually in two w~ys; (a) che:midal1y (b) mechanically.
 29
(a) Chemical action.
• 0 , The chemical action may deteriorate a strainer in two ways, firstly by choking and
. ~condly by corrosion . . If calcium 'b"icarbonate be present in water to the extent IS parts per
105 the reduction of. pressure'. due to pumping releases carbon dioxide and causes calcium
carbonate to be precipitated .on the strainer. The .effect is th . c~mulative. In c~ur:,e of time yield
begins to fall on account of choking. The corrOSlOn of the stramer metal results In the complete
collapse and the sub ;eq uent choking of tl1e portions below it. The mil(,i steel and cast iron are
attacked by the sodium salt. Zinc is particularly susceptible to s)diu:n carb.mate and
aluminium is even more so. Copper is attacked by sodium carb)nate and sodium chloride.
Brass is not readily attacked by solts present in the soil while calcium in the form of plating is
non-corrodlb~e. The sub-soil water in the western Punjab is usually saline and unfit for
drinking purposes when salinity is more than 15 parts in 105 , and not suitable for agricultural
puposes when salinity is more than 60 parts in 105 . The chemical choking by deposition of
carb)nates is very much reduced by providing a large slit area or low velocity of inflow -which
means less cepression head and the redu~ed liberation of CO 2 ,
(b) Mechlnical Cho',ing-.
It is simply blocking of slits with the mlterial such as fine sand. This can be guarded
against by providing suitable slits expanding inwards. The surest remedy to remove this
trouble is tbat the vebcity of inflow should b; lower than, the, optimum velocity which can.
te expermently found capLi'?le disturbing the material. Theprqper screening or shrouding of
the strainer with coarse material will remove this trouble to a large extent.
The pulsating action of the centrifugal punps tends to break the adherence of the
sand, particles in the slits by arching action and is also. useful to retard the deposition 01 the
carbonates.
IS. Size of Tube well Pipe and Strainer Length.
The minimum diameter of the suction pipe is fixed after the considration of the maxi-
mum permissible velocity through the pipe. The frictional losses vary as the square of the
velocity and directly as the wetted area, A 10" diameter is considered suitable for suction
pipes for a discharge of 1'5 cuseC5; this gives a velocity of 2 76 feet per second. which is less
than 3'0 feet per second, the st;ll1dard permissible velocity in water supply schemes.
The factors a"fecting the design of a tube well strainer are: (a) Transmission constant
of the soil, (0) Depression head, (e) Length and dianeter of the strainer and (i) Shrouding.
The transmission constant of the soil varies considerably fro!n stratum to stratum.
In wa.ter bearing sand, consisting of particles" say: 16/1000" mean diameter, a mean
velocity of 0 005 foot per se'cond ha3 ~ee_l fOU::l it) n w,~ 0.11y tIi') fi.lf~3t PiLrticles of a ncgltgible
daimeter. This would limit the disclarge to 'OOS cusec per sq. fo::>t ,or the strainer surface.
Large diameter strainers are, however, very expensive and, therefore. to keep down the cost
shrouding is resorted to. From practical experience a 10" dia"lleter strainer of 123 feet . total
length is found suncient to give a disc'large of 1'5 cmecs with a maximum depression head
of 120 feet. The dialleter of the strainer sl.ould, ho Never, be n::>t less than that required
for the optimum velocity., .
The in draw into the strainer is not uniform throughout the length of the strainer and
the gross discharge does not yary directly with the length, of the ,stI;':Liner and, therefore, it
would be uneconomical to have a uniform diameter of the strainer throughout its length.
It is desirable to vary the,dia'1letet of the strainer retaining the optimu;n velocity of three
fe€t per second in the pipe as far as possiole. ..
Suitable lengths of strainer of varying diameters from 4 inches to 10.inches may be
used. A}ypicallocation chart is shown in fig. 1.
The saving in cost <:onsequent on using the smaller diameter strainer is considerable
for instance:-
The cost of strainer and plain pipe for one tulle well of uniform diameter of 10 inches
t.hroughout i~s length. is Rs. 4 575/- compared to .Rs. 3,48~/: when stl'ain~rs and pipes of varying
SIzes from -4 lll, to 10 lll. are used for the same tUje well glvmg a net savmg of about 25 per cent
in the cost of strainers and plain pipes. '
 30

There is, however, one objecti~n to

the use of the small diameter str<l;lllers. LOCATION CHART OF STRAINER .bO PIPE.
The strainers of 4 inches to 6. l?ches TYPE DESIGN
diameters, especiallY' the Tel Drass 2.
SeA.L • • 1401t: ':.''..
"CR:I.40
·
s t ramers Whl'ch have been used . on the
K 1 Tube 'Nell scheme, reqUIre extreme
carearo.In their handling. when lowering the
and also dUrIng the process of
same Great care h as t 0 be exer-
"
shrouding.
cised to see that the bottom of the lowest
T
I
strainer is not allowed to touch the
ground and to ta~e t~e weig~t of !he
strainers and plam pIpes whIle bemg 0 "Ii .. ."Ntl 1',".
lowered into the boring. It is a ll1atter i
r~ ~~~
of simple calculation that a 4 inches
diameter strainer will not tear the super-
imposed weight of strainers and rising
pipes.
- . . +. - - --
I• •

._.+
nltA'H'1f
In view of the extreme care required
in handling the delicate smaller dia- •
- Te. - -

I
meter strainers of 4 in. to 6 in. dia-
meters. their use is discouraged and
only strainers of 7 in. to 10 in. dia-
!
I .41".
meters are used.
16. Different operations of sinking tube
e ,7.e ,.,~,

wells.
Extract from Paper No. 248 PunjaiJ
I
Engineering Congress, 1941.
,!
"Boring.-Boring is done by the
percussion system. To accommodate
10 in. rising pipes and strainers and allow
... ", -
iI

-14 ••
:.: I. -__-- .- .t.4-.
-
~--
:= 'e.• srlfA'H,1f

!"
"..
for 4 in. thick shrouding 18 in. casing
pipe is sunk by means of a sludger. The
sludger with a "teel cutting shoe and a ".. ",.
Q
0 --_.+. ."'N. ~/".

non-returning. valve is either worked by "~ ," --~~- -- 1,-.

--
-- ·_--t
~
,rltAIN'"
power plant or manual labour depndin;
upon the hardness of the strata.
Strata Samples.- Soil samples are
taken at every 10 fiet depth and also at
0 I
&4 ••
."N" ~I~.

every change of strata. They are pre- "e

served in a compartmented l,OX and are
also sent to the Irrigation Rescarca
Institute for the determination of trans.-
... - :-B-:~
-
-
--.
- -_
.$ 'TItA'NIl~

~~~
I I
mission constants and measureiLent of ae·e
percentages of clay and /?ankar. --- ----
- - - --
I
Sr/lfAINtl"
/
.~.

Water Samples. -Water sa'11ples of

each water tearing stratum are taken in a
Ut
-_··t
~~::~~
properly cleaned and steamed Winchester
tluactz bottle and sent to the Irrigation
=~
, I

1. .
l-<esearch Institute for the determination :: 7 - 4,.; ,rltA/H""
of the pH value, conductivity, salt index, --- --
quantity of salt present in water. as also
the total s01ids, in order to find out .-----
whether water is suitable for irrigation ot'
not and its probahle effect on the 1fe of • a'e =-=1i_:;;'~'
::.: ,.",u• ..... If.oll·
the strainer.
Fig. I.
 31

Water Jevels.-During the process of boring, observations of water levels in the bore are
. recorded regularly twice a day.eefore starting the work and after finishing it for the day. ~he
difference in the water levels observeo on the close of the day's work and tte nex.t mornmg
before starting the work gives an indication of the recuperation of water levels m the core
and the nature of the strata at the bottom of the casing pipes.
Location of strainer and shrouding. After the completion of a bore, a location. chart of
strainers and rising pipes is designed, guided by (a) the transmission constants d the SOlI strata,
(b) the analysis of samples of water, and (c) the water levels in the bore.
After the strainer has been lowered in the bore according to the location chart, !hc
process of shrouding and extraction of the casing pipes is started simultaneously. Shroudmg
consists of i in. Pathankot bajri retained on 1'/10 in. mesh. An accurate record of the
material used in shrouding is maintained showing its calculated quantity as also. that .a~tually
used. In certain cases, specially at the junction of sandy and clayey strata. bIg cavItIes ~re
formed which require considerable quantities of tr.e shrouding material to fill t.h~m. . Dunng
the location of the strainer, great care has to be taken so that the strainer and nsmg pIpes are
always suspended and are not allowed to rest on the bottom of the bore as otherwise the
strainer at the bottom is subjected to unsafe loads, and is liable to te crushed.
To ensure that the strainer has been located and the shrouding has been done without
any da'uage to the strainer, it is necessary to take soundings inside the tube well ~t frequ~nt
intervals. A history sheet for each tube well is maintained at site, giving complete mformatlOn
regarding each bore. '
17. Eccentricity in the bore.
In all cases where borehole pumps are installed, whether the wells are
compound wells or otherwise, it is of great importance that the finished well and the bore
itself should be as nearly vertical (i.e. in plumb), as possible. Borehole pumps ge~erally
revolve with a speed ]::etween 1,500 and 3,000 revolutions per minute, and it can easIly be
imagined how damaging the effects of eccentricity can be in such cases on all the moving parts
of the pump and motor. The setress and strain on all bearings, shafts, bushings, etc., are. so
greatly increased due to the eccentricity that all the care and patience exercised in securmg
a truly vertical bore is amply rewarded' by the ease and smoothness attained in the installation
and working of the pump later on.

In spite of exercising reasonable care in boring the finished well is likely to ge~ s~ight~y
out of plumb partly due to some eccentricity in the bore itself and partl) due to van.abon. III
the verticality of the top pipe when earth is filled round it after extraction of the bonng tute,
e.g., when bell-sockets are used for a compound well.
1he ins~de diameter of the top casing pipe is always kept slightly larger than the
outside diameter of the pump bowls, so that there is a certain gap letween the outer face of
the pump- bowls and the inner surface of the casing pipe. So long as the eccentricity does
not exceed the clearance allowed in this gap, advantage can be taken of this in reducing the
eccentricity. The eccentricity will be neutralised if the pump is installed not centrally in the
casing pipe, but in such a way that one edge of the lowest pump bowl is almost touching
the ca5ing pipe on the side to which it is inclined. But if the eccentricity exceeds the clearance
allowed Py th;s annular gap (i e., the difference between the inside diameter of the casing pipe
and the outside diameter of the pump bowls) it cannot be eliminated without setting right the
eccentric casing pipe.
Me3.surement of eccentri~ity-A simple method for measuring eccentricity in a bore is
by the use of a reel-like instrument which can easily be made by joining together two circular
discs of the same diam~ter by means of a central hollow tube. The tu,;e
should have threaded ends with nuts to fit on to these so that discs of diEerent diameters,
according to the size of the bore. may be fitted on to the ends of this tube as required. The
discs should have holes punched in them so that they may be immersed in water without any
obstruction. The discs and tubes should be sufficiently heavy to enable them to be lowered
in bores full of water without difficulty.
 32
,.'</'"
The tube with discs can te suspended by means of a thin strong steel wire or copper
wire so that it can be moved vertically up or down in the bore. A tripod having a pulley'
rigidly fixed to its apex is placed over the tube-well and the disc instrument, fitte.j with discs
about 1/8th inch smaller in size than the inside diameter of the tuhe-well, is sus?cnded frOill
it with t1;e wire passing over the pulley. TLe instrumcnt can now be lowered or raised as
desired. The tripod is so adjusted that the ,disc instrument will ·come centrally over the top
of the tu' e-well when freely suspended. Let us assum~ that the ~isc instrument is so hung
that its top disc is centrally situated and flush Wit'l the top of the b)re tube. Let the distanc~
tetween the pcint of sus;:;ension (i e. the pulley of tripod) and the top of the disc instrument,
when its top is flush with the top of the tube-well, be 10 fect. The instrument is then lowered
in the bore, say, by 10 feet. This distance can be found out y taking measurements of the,
length of the suspension wire, Let us assume that the wire deviates from its central position
and the deviation as me\lsured at the top of tte tube-well is l/lOth of an inch. Therefore
at the top of the upper disc (whic'l is 10 fcet inside the tubl~-well) the deviation will be' twice
this much i.e. 2/lOth (l/Sth) of an inch from the vertical line throug~l the pulley. Hence in
going down 10 feet it has become eccentric by 1/5th of an inch or thc eccentricity is 1/5th
of an inch in 10 fcet. (It will be noted that the eccentricity is not l(lOth of an inch in
10 feet, but 1/5th of an inch). The e,:centricity at 20 feet, 30 feet etc., can similarly be
observed.
Setting right eccentric wells-If the eccentricity of the larger diameter casing pipe (in
which the borehole pump is to \-e lowercd) of a co:npound well is found to exceed the
:: clearance allowed by the annular gap 1 etwecn the inside dia'neter of t'1e casing pipe and the
" outside diameter of the pump bowls, the casing pipe will' have to be Drought into plumb before
the borehole pump is installed.
A simple method of setting right an e~centric casing pipe is to loosen the earth on the
sice t6 which the casing pipe is inclined and te;cn to force t;10 casing pipe b:lck by applying
jacks on the other side (opposite to that to whic'l it is inclined). Wacl1 t~e casing pipes have·
been forced into plumb, eCLrth and gravel are filled on the side opposite to that in which the
pipe was inclined, thu'> p 'eventing Le casing pipe fro.n resu ning its cccentric position.
To set right an eccentric casing pipe, a new bore, say acout.t\\O thirds the diameter of
the casing pipe, is sunk close to the casing pipe on the side toward5 which the pipe is inclined.
Tlm bore is carried down to about 10 feet a,-;ove the b)ttom end ,of the .. casing pipe (which
as explained earlier is generally at a'~out 80 feet to 100 feet b.~low ground level in the case of
compound wells). The point where this new bJrc-is made is kept su!'ficiently away from the
eccentric casin~ pife so that on completing the bYre its b)ttom end does not come into clash
with the bJttom end of the ca';ing pipe. Theboring:tube is then withdrawn with tte borehole
left intact. Powerful jasks are next appli!2d on the side'opposite to that in which the casing
lJipe is inclined and at a point a out 8 feet ! clow tre surface of tlc.c ground. This enables c

the jacks to be fixen pro~eJy and theY,get a g()od :se;:}t bClind for applying force. As the
a
b:lttom end of the casing Dil=C is fixc r !. and the~e is hollow space made by t1e boring on
the siee opposite to that fro:n ,,'h~re j;Jck" arc applled. i\e ca5ing' pipe can easily te pushed
till it is ab301utely vertical. Whia this has be';n accomplished the:e will be a boUow space left
J-etween the neN and th.e old position of the ca5ing pipe Earth and gravel are then filled in.
this open space and also in al1Y orcn space Idt on the opposite si,'e where the new boring
was made. The ca.sing pipe is thus fixe 1 in a vertical position and all e:centricity is rerriQVe<i... '
.
' /

The jacks are tl--,ell re novd after making sure that the casing pipe does not tend
to spring back to fce eccentric position. The well is n:>.v rC:l.ly for installation of a 1:)) [e'1,ole,
pump.
18. Back washiag and back-blowing of strai!lcr wells.
In thc case of strainer wells 'it sometimes happens that the strainers get choked with
fine particles of sandancl clay or due to bridging (descri' ecl before) and thus have to be
cleaned; or it is found neccessary to draw ollUargcr quantities of sand from thc surrounding
strata through the. strainers, to inlprove the yield of the well.' This is done by resorting
to "back-blowing". Tl:e process of fo'rcing out water or air under pressure from inside the
tube-welt through the strainer slots, into the surrounding strata is' called "back washing" or
'back blowing" respectively. In some cases satisfactory results are obtained by these methods
md it is advisable to try these for improving the yield from a tube-well whenever it is found
l~cessary.

If su T:cient quantity of water is available from a high level storage tank or from a
pump which can pump water under pressure into a tube-well, "back-washing" can easily be
done with water. This water entering the tube-well under pressure passes out
through th strainer slots into the surrounding strata and thus pushes out and carries along
with it all the fine particles of sand, clay or dirt and cleanses the strainers substantially. But
if the strainers are only partly clogged, the water may pass out of the strainers through the
unclogged -rcrtions and tLe whole of tte strainer may not get cleansed properly. In such
ca~es a. water jet is carried (lown to a point directly opposite each portion of the clogged
strainers by using a small pipe, which can be lowered freely inside it. At the cottom end of this
water-pire, tLere is a foot piece alJout 4 feet long with rerforations on the sides and with
it5 lower end sealed with a can. The top end of this water-pipe is connected to a high level
storage tank or the pump. The water from the storage ,tank or the pump emerging under
pressure hom t1 ese perfora tioll5 in t1.e foot-piece passes through the strainer slots thus
washing tLem thoroughly and cleansing them. If any of the was~ed material falls inside
the tube-well it can be pumped out from the tube-well along with the water from it.
The other rr:ethod is to use air unc!er pressure, from an air compressor. The plant consists
(if an air compressor and air line, foot-piece, eduction tube, etc. The air line
is a G.!. (galvanized iron) tu1)e from 1 inch to 2 inches in dameter, through which air
uncer pressure is passed from the air compressor to the foot-piece. A foot-piece is a per-
forated tube ab0ut 4 feet long and of the ,same diameter as the line with its lower end sealed
or plugged with a cap. An eduction pipe is used for discharging water along with
any sand, etc" which may have come into the tube-well due to i_lack-blowing. The well
pipe itself can te used as an education pipe, but as the washed material from the strainers,
etc., has also to be pumped out along with the yratl)r it is preferaole to have a separate
eduction pipe about 4 inches in diameter, so that tue ~lelocity of water pumped out may be
sufficiently gleat to carry with it all the clogged matrtrial, etc" from the well.
The tube-well is sealed with a sealing cap at the top and the eduction pipe and
line pass through it There is a sluice valve on the delivery side of the eduction
pipe which can, c closed when air is to be passed out through the strainers and can be
opened when it is dsiord to pump out watr and sand, etc. from the well.
The eduction pipe is lowered into the tU2e-well till its t:OttO'll is al~out 2 feet ahove the
top of the strainer pipe The air line IS then lowered inside the eduction pipe till its foot-
piece comes dirctly opposite the strainer pipe, The sealing cap cetween the tu')e-well and
eduction pipe is then closed as also the sluice valve on the eduction pipe. The air c )ming
from the air compressor and passing out of the foot-piece has now no passage for escape
except through the slots of the strainer pipes, and through etese slots it pa~ses out into the
surrounding strata. But in doing so, as it passes under pressure throug'l the slots, it cleanses
them and removes the material which was clogging them. When one portion of the strainers
has teen cleanerl, the eduction pipe and air line can i)e lowered to the strainers lower down
and graduallY all the strainers are cleansed. After cleaning each length of the strainer the
sluice valve on the delivery side of the eduction pipe is opened and the water containing
the washed cloggings, etc" is Ipumped out. Thus gradually all the clogge,j strainers are cleansed
and there is a proportiom.te increase in the discharge from the well.
In the case of compound wells an eduction pipe of a suitable size, which can easily
be lowered inside the main strainer tube, should generally be used for back-washing and
back- 'Dlowing purposes. But if the diameter of such an eduction pipe is found to be too
small for satisfactory operation the main strainer pipe itself can i:e utilized as an eduction
P!pe, In such cases a loose joint is prepared for extending the length of the main strainer
pIpe h ground level For this purpose plain pipes of t'1e same diameter as the
strainer tube have to be lowered from ground level inside the compound well. A rubber
ri.ng fGrming the loose joint is screwed on to the bottom end of these plain
pIpes and when lowered inside the well rests on top of the main strainer pipe, The bottom
 34

por~ion of this. loose.joint is tappered so that the main strainer pipe slide3 into it and abuts
agamst the ru?ber rIng, thus forming a water and airtight joint due to the weight of the pipe
on top of It. This pipe is then used as an eduction pipe and the air line
can be lowered inside it and worked as described a bove. When the straners have cl~aned,
the upper plain p~pes with the loose joint can easily be lifted out of the compound welll~aving
the well clear for Installation of a deep well pump as usuaL
19. Te3ting the T'ube ..Well.
After the ex.traction of the casing pipes is finished, the tube well is tested for its
discharge. If th.e Yield of the tube well is satisfactory i.e .• it yields a discharge of abottt 1'3
cusecs 101'.2 lmiXlmw~ aepresslo!7 heCla ot l2'') teet, the C0!75tmctf<:1ll of the ma,Smzary p~rt of
the work is taken in hand. The arrange~ent of the stilling c'1amb er and IT.eter flUme to
measure the tube well discharge is shown in Fig. 3.
The most int~rcst~ng PlAN" MlASIIIIIJII; DlYI&E FDR 1.6 &USEe D/SUIAR(;£
feature of the. constructIOn Sc A J. E. • 1/
of a tube well IS the put ting /'00
in of a water-tight joint Sl'cr/ON ON A.II.
between the top of the nsing
pipe and the suction sluice-
valve. This joint is 7 fef~t
below spring level and it is
not possible to fit the suction
valve when water is rushing
out of the rising pip~ under
this much head. T() over·
come this difficulty a mecha-
nical rubber stoPPer has ...:frr--=;::::::o'i--":=:""

Montagu,
been in Fig.
devised 4. A.M.R.
by Mr. ""i'~'d'~~~111~~~~~!!~~j:;;tijr-;;;;;;;>i3~
Th~ stopper COIlJ.pletely
staunches the flow at water
in the risin.g pipe. It is then Fig. 3.
9uite .e~sy to put in the rump founea tion after ttc suction sluice-valve has be(!n put
m positlon.
20. ExtraetioJ. of strafilers an 1 pipes in the case of umuccessfu} wells.
If, on testing, a tube-well it is found to be u::lsatisfaciory and it becames necessary to extract
t~e strainers and pi~(lS it can be done i)y lowering a wire rop~ with a ho 'k . inside the strainer
pIpe so as to catch tnc "eye" on the bail plug and then pullmg the stralller tube up. As
d~LriOed. 'oelor<;. .this can casily be acco 71 plished provided tne strainers a.nd plain pipes ha ve
not been m poslho~ for a long time and have thus not been jammed b;: the surroundlUg soil.
If these have been Jammed in position the wire rope method of. pullIng .them ~)Ut may not
be successful. In SU(~h a case a simple device is to use a tapenng, comcal, pIece of wood
like a fr.ustruJ?- of a cone, loosely fitting inside the strainer tube, whic:1 c~n be lowe~ed. inside
the stramer ;t)lp~ bY.Jneans of steel rods. This conical piece of w?od IS l~wer~d I.ISlde the
tube-well. wlth .ItS wlder edge at the lower end. to some point OppOSIte a plam plp~. It it is
lowered to a pomt OPposite a strainer tube the latter may get damaged when pullmg l( out.
Gravel or stone chip, are thea poured inside the tU;Je-well and these collect round the conical
piece ?f wood and form a sort of a wedge between the wood and the pipe. .When th· wooden
piece .IS now pulled up by means of the rods it gets jammed to tte ~:)lpe [;ecause ()f the
",:edgmg effect of ~he gravel between the pipe and the wooden cone an~ It tends t~ P1l11 the
pIpe also along WIth it. A clamp is tightly fastened to the steel rods lust ~DOV~ tne 1:op of
the ~ube-well. an~ re.sts on it. Another clamp is fasteped .on the tu :e-~el~ pIpe Itself on t~e
outs1de: holdmg It tightly, and jacks are place:l below th1s clamp .to hit It up .. The pull IS
transmItted from the jacks, through the clamp~ to the tube-well pIpe, and th~ pIP~S trcmsmit
it to the cla!llp. on the steel rods resting on top of the pipes. Thus the ro? .Itsel~ IS pulled up
and transmIts the pull to the canical wooden piece inside. The tu ;e well IS m thiS way pulled
 35

-
IIrtNAIIIQJ. SrfJl'l'ER'DR ID INC_ _ ~
up by the jacks directly from outside and by
~he. wooden piece and the wedge of gravel from
, ScAl£4~: '''I InsIde and is gradually extracted out. This is a
v~ry convenient method of extracting strainer
pIpes and can be used where the strainers have
~:een lowered for a long time and have got
_ _ _ _ _ _ _ _ _ _ _ ~_ Jammed. The advantage of this method over
~1'IItr.,."K,,"'_ the hook method is that instead of the tensile
'OUJzt..~~b~OI7:r--~ force teing applied to the hook on the bail plug
,./.
~1Ij\':1'l
l only, which may get Lent or sheared, it is
a:)plie 1 to a large surface of the tube-well pipe
1===iFl .3' .~ 9 itself and is therefore more effective.
J. '.I.,,_~.. .A special collapsible steel tool is also
~i=;:;::::!~~~:3~'/4~--:;t:} (, btal.nable which, can ~e low.ered inside the tub~-
~ ~t, :well mstead of We LOmcal pIece of wood. ThIS
. . ta I tool can be expanned out after being lowered
_-:f'~:;=.: _J __ • insi ,e the s~rainer pipe.s thus forming a so'rt of
. wedge to gnp tte pIpe Just as the gravel wedge
holds the pipe in the method described above.
9. Miscellaneous We rks.
After the tube well is installed, the following
works are carried out :-
(a) Construction and lining of water course.
The earthen water-courses should at least
be run for one crop for consolidation before the

.. .
fa
lining is done.
(b) Pump Sumps.
A suitcble design is shown in Fig. 5.
'

(c) Measuring devices.

It should be in the form of long crested
weir to require low working head of the type of
meter on canals, Part II Chapter IX. A stilling
cevice in the fornl of grids will generally be
required after the water leaves the delivery pipe.
(d) Operator's Quarters.
Two rooms with a verandah. kitchen ana
court yard will give enough accommodation.
22. Assessment and Water Rates.
The esta~lishment required for each well
consists of· one operator and a beldar. The
operator's duties are as follows; while beldar
acts as his assistant : -
(a) to start "and stop the pump,
F:ig. ~. (b) to record the time and the meter reading
at the beginning and end of each cultivator's turn,
(c) to record tIle area irr:gated and the crops sown.
(d) to issue cultivators' recei?t~.
(c) to prepare tie daily routine return.
(f) to look after the Government channels,
A ziiladar is required for about 20 wells to check the revenue work of an operator ann
to prepare the demand statement. Besides a supervisor is needed for about 20 wtllls' who is
supposed to we rk as a mechanic and an electrician to carry out the necessary repairs.
 36

The Operator maintains a printed and

b}und log book containing tfle following
information : -
(a)~Datc. (b) Time of starting and stopping
the motor. (c) Meter reading at starting -and
stopping (d) lCnits used by ca h cuI ivator.
(e) Value of electricity used. (f) Field numbers
irrigdted. (g) Name of crop. (h) Area irrigated.'
(i) Gauge in measuring tank. (j) Total volume
pumped in cubic feet.
In addition to the log took the operator
maintains a cultivators' receipt i:ook. This is
in triplicate. When a cultivator has finished
irrigation, he signs the receipt which shows ttc
time of starting and stopping the pump, the
number of units consumed and the amount
assessed. The oFerator keeps the counterfoil
with himself, sends the duplicate to the
zilladar and hands over the triplicate to the
cultivator to enable him to check his water dB
at the end of the crop.
On receipt of the cultivator's receipt, the
zilladar posts the units used in a ledger in
which he leaves a page tOf each cultivator,
This forms the b)s's of the demand statement.
The ledger is also kept in dupl'cate. The
" ., cup'icate copies arc di3tributed to tte
cultivators at the end of each crop and take
the place of the pare '~a (demand notice)
v~macular form No.8 A used in canal irrigated
areas.
'Water charges have ceen fixed per
1,000 eft. of water. Actually, however, the
rate is converted in to electrical units by actual
measuremert ts.
-~i;-fj~~I~~~~~~~:£J~~ At the beginning of each crop, a complete
te3t of each tube well is carried out with
re ;pect to its discharge, electrical consumrtion
delivery and suction head, pump efficienc:e5
etc. Electrical consumption per 1,000 eft. of
water and the cost of an electrical uuit for
each, tube well is work(~d out and intimated
to the nmindar concerned by posting notices.
The rates per thousand eft. of water are
uniform for aU the tube ~'ells, iJut the rate .Fer
unit of electricity varies from "ell to "eU,
depending on the number of units required to
raise the standard volume. The advantage of
Fig 5. this system is that the cultivator can read the
meter before and after his turn starts and kn ws exactly how much is to be deb:ted to
hi, account.
Water rates r er 1,000 eft. of water are the same as charge~ in the United Provinc~s.
These are three annas s~x pies per thousand eft. of water for khan! and tW? annas fo~r pIes
per thousand eft. of water for f,rabi, These correspond t.o Rs. 2.12 per watenng of four lllches
.in rabi and Rs. 3.18 per watering of four inches in khan.f.
 37
Assuming deltas of 1'8 and 1'1 for kharif and Rabi (vide paragraph 5 [ the cost 01
maturing one acre of khan] crops works out at Rs. 16.7 and of rabi crops at Rs. 6.81.
23. Discharge from a tube well.
The mathe,natles to determine the discharge from a tub~ well has not yet been fully
deVl'loped. The approximate methods to judge the discharge of a tube weU can be applied
in two ways. Firstly, very nearly correct estimate of discharge can be made from the actual
performance of the existing tU'le wells. In the Punjab Engineering Congress paper No. 129 ot
1929, M.1:. Ho Net Sll?erinten(ling Engin(~er has given actual performance of a.
large num'Jer of tube wells varying in discharge from 8 to 36 gallons jJ~r
squar~ foot of the strainer surface per fcot depression head. The safe figure
for an estimate of a tu')e-Ncll disc'large may be taken as ten gallons per hour per square
IFATUIW. foot of the strainer area per foot depression head. Secondly an
estimate of the discharge in a tube-well can 're fanned by
experimentally determining the safe optimum velocity of say,
005 ft./sec. as ment~oned before in this chapter which will disturb
the soil particles irrespective of the depression head, The soil
samples used in such experiments should be representative
samples of the soil crust in which the strainer is to be put, which
is not generally practica'0le.
The formula of discharge was first developed by Dupuit
{I863) and then developed by Kozeny (1927). Reference is invited
to pages 255 to 260 of book 'Public Water Supplies' by Turueaure
and Russel, New York (1924). Fig. 6.
Fig. 6.
Let Q=total discharge of the tube.
H=Deoression head =D-d
D=Depth of water table above the impermeable boundary.
d=Depth of water in the tube above impermeable boundary.
r=Radius of the tube.
R=Radius of the influence of the streams radially entering the strainer.
k= Tra.nsmissi.on consta.nt of the soil. '
The formula of discharge in cusecs
_1fk (D l_d 2) :. "'

Q--R-
log. - -
r
In this formula k depends on the porosity P and a soil constant S'By X' so th3.t k=k'P
The a'3su'llption in the derivation of this formula are:-
(i) Water table is at rest and is horizontal.
(ii) Bottom of the strainer rests on a horizontal impervio'ls layer.
(iii) The flow is radial and total diseharge Q remains constant acctoss the surfaet-
of a series of concentric cylinders.
The formula is not correct because the assumptions are not justified. There are ne
concentric cylinders Gf equal discharge entering the strainer. There is nG theore tical limit
to the value of R, radius of influence. The assumption of the impervious lower boundary h
absolutelv unjustifierl.
However tables are given in the a'Jove quoted reference to predict. the value 01
R and the likely values of k in different soils based on the observations of the authors. :
24. C niitions in the Punjab and the Unitej Provinces compared.
The conditions of tube-well irrigation in the adjoining provinces of the Punjab and tht
United Provinces, in India are CQ apared below:-
(a) In the tube-well areas of the United Provinces, the geological. conditions an
such that an immense sub-soil stream flows parallel to the crustal warpings. In the Punjab
the flow is transerve to the warpings Consequently in the United Provinces there is (f
uniformity of conditions in respect of depth to water table belcnv natural surface. whic~.
 38

does not exist in the Punjab. Such uniform conditions lend themselves to standardisation of
plant and n:ethods which make for maximum econO:T y in sinking and equiping of tube-wells.
In the funja b, the variety of geological condifions necessitates every single site being carefully
examined and tested before the tube-well is completed and equipI=ed. Apart from eXI=ense,
delay in construction is certain to occur. .
(b) The nature of tte sub-soil, in the tube-well areas of tb.e United Provinces is
l:enticles of clay in sand. In the Punjai), sheets of Leds of clay occur irregularly in addition
to numerous lenticles both Jarge and small. There are no geological maps which would enable
the engineers to avoid such deposits when siting t;~e new wells. Conser;uently in the United
Provinces the entire water table is interconnected and the whole of the stream cantiL utes to
tI-.e discharge of the tubc-well system. In the Pun: ab, the irregular 'geology of the sub-soil
results in pockets of water cut off from the main stream. If a pump should accidentally be
sited in one of these pockets, the discharge may cease abruptly after a few months' working or
may te greatly reduced.
(() In the hbe-v,ell area" of ttc Unitec Prcvircc,_ He soil is l'gJ t in texture, uniform
in quality, free from salt with a pH value aproximating to the neutral. Conditions could
hardly be more favoura')le. In the Punja b, t 1,e widest variations of soil occur. Many areas
are impregnated with various salts the pH value varies from 8 to 9'2 and over. Unques-
tionably ther,: ar,~ areas suitable for well irrigation. Such areas must be sought for, and
carefully tested. Otherwise Government may fined itself saddled with tube-well irrigation
;ystem in areas, which may deteriorate in course c.f time.
(d) In the Punjab the water of the su~)-soil reservoir is not sweet every where as in
the United Provinces. The salinity and the Salt Index of water for every strata where the-
strainer is laid, must be tested, otherwise the tute-well iu:gation: will fail in the Punjab by
deteriorating the yield of the crops, matured with saline water.
(e) In the Punjab. the areas which are available for tube-well irrigation are very
much scattered unlike the areas available for the United Provinces Tube-Well Irrigation.
because the major portions of the Punjab plains are under canal irrigation. It is only the
patches, uncommanded by flow irrigation or those adjoining the Dhayas along th' rivers, which
needs artificial irrigation.
(j) In the plains of the Punjab, average rainfall is 5 to 20 inches per annum, while
the rainfall of the areas irrigated from tube-wells in the United Provinces i~ more than 30
inches per annum every where. The number of waterings from tube-wells to supplirr:ent the
effect of rainfall in the Punjab will be at least double of those requ:red in the United Provinces.
The Punjab cultivator shall have to pay abJut double the price of water. The average cost of
irrigation per acre from tube-II-ell wells is Rs. 2-4-0 per watering in rabi and Rs. 3-4-0 in
,~hanf. If the number of waterlngs is 4 in raM and 6 in khari/, the cultivator will pay double
the amount as compared to the canal irrigation.
25. Financial aspects of the tube well schemes In t1:e Pun;ab.
The financial aSI=ect of the tube well irrigation schemes is not at all bright in the
province of the Punjab. The type of expensive irrigatian cannot be popular because large
areas in the Pnnjab are already being irrigated by flow irrigation from the canal at very
cheap rates.
The electric energy has been prepared in the United Provinees at a cost of less than
three pice per unit, while the cost of energy per unite from the Mandi Scheme in the Punjab
is more than two ann as.
The cost of sinking tube-wells must be more in the Punjab than that in the
United Provinces, because the water bearing strata are irregular. The lifts are also relatively
nigh.
The life of the tube-wells in the saline waters of the Punjab can hardly te taken as
12 years, while the depreciation charges in the United Provinces are worked on the basis of
17 years life.
In spite of these difficulties, the tube-well irrigation is likely to pay a very important
part in the future irrigation schemes of the Punjab Province, because there are lots of areas
 39
where water can only be supplied by tube-wells, and because it is likely that cheap energy
may be available from canal falls by small hydroelectric sehemes or from crude oil engines
using country fuel. Moreover, the tube-well irrigation may have to be extended as an
anti-waterlogging measure in the waterlogged tracts or as a famine relief measure in certain
arid tracts.
26. Clnal Versus Tube-TNell Irrigatim.
A few of the disadvantages of the tube-well irrigation are stated below:-
(i) Working expenses are extremely high, as wells depend on mechanical means oi
raising the water.
(ii) The wells are dependent upon a source of energy. Apart from the question of
cost referred to in (i) above, a failure of the energy supply is accompanied by a
stoppage of all pumps dependent thereon.
(iii) A tube-well is liable to progressive deterioration. The strainer is l'able to
choke and is difficult to clean. Replacement of the well may be necessary
after comparatively short period of operation.
(iv) Maintenance of a delicate mechanical installation will always present its OWfJ
difficulties.
\v} The tU"be "ell water is clear i.e., free from silt. Cousequentlv weed growth an<l
algae must ce expected and their remoyal catered for. Clear water tas nol
the manuring value which silt In canal water provides.
The advantages of tube-well irrigation are stated below:-
(i) Wells may be sunk and equipped as required.
(ii) Water can be turned off at any moment, to take advantage of rainfall.
(iii) The supplies are likely to be fairly constant excluding always the deterioratiol1
of the well.
(iv) The loss in transit is much reduced.. If water is sufficiently valuable, the lining
of water courses is a practical proposition.
\v) Wells may be sited to command any desired land, subject always to the technica.1
limitations for this means of irrigation. .
\vi) Volu'lletric assessment is possible, in fact obligatory.
(vii) Water is likely to be used with rna "{imum e nciency accompanied by benefit tt,
the cops and to the su )-soil water ta')le.
The canal irrigation is not an un nixed ble3sing. The supply availa')le for irrigatioL
is limited to the disch.arges of river.> at th.e time of sowing and maturing of crops which havt
already been used up in the Punja'). Flow irrigation from the canals is essencially accompanied
by a great loss of water in the distri lUting channels and the wasteful use in tht
fields. This wastage brings in turn the curse of water-loO"ging. The disadvantages of canal irriga.
tion are dealt with in detail in part III Chapter II. 0

27. Exa.mination Questiom.

(i) Discuss the merits of irrigation by means of tube w~Jls against irrigation by grav:ty canals is
the western dis:ricts of the United Provinces.
(ii) Describe with sketches an electrically worked tube-well installation used for irrigation. ExplaiL
&!l air lift pump for a well lift 20 fcot. (T.C.E. 1935;
2, (a) Name the different types of wells.
(6) Why woull you prefer wells over other sources of supply of drinking water?
(c) Describe the various me hods of boring deep wells, P.U.l9~
3. (If) Yo?r are required,to place all order for pumping machinery. Please lay down specification,
which you, as a CIVil Engineer, will communicate to a firm of suppliers, P.U.194:2
(li) Describe the various type.; of pumps suitable for use in the tube well irrigation. Which tyl1'
do you consider best and Why?
 40
4. (a) Why is t.he cost of the tube well irrigation more in the Punjab than in the United Provin~es ?
(b) Why is delta in the tube wen irrigation lower as compared to the flow irrigation from canals?
5. (a} What is the life of a tube well machinery in the Punjab and how will it compare with its life
in the United Provinces conditions?
6. Describe the various types of tube well strainers used in the Punjab. Which type in your opinion
is the best and why?
7. (a) Describe the various factors which con tribute to the reduction of yield from tube wells in the
Punjab in course of time.
(b) What precautions and remedies would yeu suggest to minimise choking of strainers?
8. (a) What is the suitable discharge of a tube well meant for irrigation?
(b) "'hat points will you keep in view in designing the water-course sy,tem for irrigation from a
tube well?
28. BihliJgraphy.
1. "The Construction of Wells and Boreholes for water Supply," by J. E. Dumhleton. The Technical
Press Ltd, L')ndon.
2. "\Vater Works Handbook" by Flinn, \Veston and Bogert. Megraw-Hill Book Company, Inc.,
New York anb London. '
3. "Oil-Well Drilling Methods," by Victor Ziegler. John Wiley & Sons, Inc. New York,
4. "Hydrology and Ground \Vater" by J.M. Lacey. The Technical Press Ltd., L')ndon.
5. "Tube-wells," by T.A. Miller Browrr\ie. Thacker Spink & Co, Calcutta.
6, "Oil Gas and Water \Vells," by Luecy Manufacturing Corporation, New York.
7. "Casing Troubles and Fishing Methods in Oil Wells," by Thomas Curtin. Bulletin No. 182, of
the Department of the Interior. Government Printing Office, Washington.
8. "Johnson National Driller's Journal," St, Paul. Minn.
" A Study of the Yield of Water Wells. Bulletin No. 1238:'
" Testing Water Wells for Yield and Drawdown No.638."
.. The Chemistry of water. Bulletin No 1237."
9. Brochure issued by the American Well-Works, Galveston, Texas, U. S. A.
10. "Water Analysis," by J, Alfred Wanklyn and E.T. Cnapml.n. Trubner &. Co., Llldgate Hill,
London.
11. "Baroda Blue Book of Water Analysis for Agricultural purposes," by Dr. C. C. Shah, Agricul-
tural Chemist, Baroda. Printed 1938.
12. "Hydraulics and its Applications:' by A. H. Gibson. C:>nstable amI C:)ill,nny Ltd., London.
13. "Minutes of the Tllbe-wf'lls Conference, United Provinces, July 1934." prjnted at the Advertiser
Press, Saharanpur, United Provinces.
. 14. Notes by Mr. F. H. Hutchim;on, LS.E., Superintending Engineer, Development Circle, United
ProvIl:ces, on Pipe Lines, Shajra Sheets etc. and instructions for the gUidance of Tube- well Operators, Section
MlstrzS and Supervisors.
. . 15 "The Ganges Canal Hydro-electric Scheme with its associated State Tube-well Project," by Sir
Wilham Stampe, Superintendent, Printing and Stationery, United P,·ovinces, Allahabad.
16. Irrigation Administration Reports of the United Provinces and their supplements from 1929-30
onwards.
17. "Report of the Royal Commission on Agriculture in India." Printe:.l at the Government Central
Press, Bombay, 1928 .
. 18. "Report of the Indian Irrigation Commission, 1901-03, Part I, General." Office of the
Supenntendent of Government Printing. India, Calcutta.
19. .. Compressed Air Plant," by Robert Peels. John \Vilay &. Sons, New York. Chapman &. Hall
Ltd., London. Season Edition.
20. "Brochure issned by Messrs. Ingeroll Rand," Co. of U. S. A. reganiing Air Compressors.
21. "The Physics of the Divining Rod." by Naby and Frank is. Publi~hed by Bell and Sons Ltd.,
London, W.C. 2.
22. "Notes on Tube Well" by John Ashford Paper No. 40, Punjab Engineering Congress,
Lahore 1918.
23. Article on Unreinforced Cement Ccncrete pipes and pipe-lines by Mr. KG. Smith,I.S.E., Indian
Concrete Journal, November, 1942.
 41

24. Report OD the State tube-well project estim~te, 1934-35 to 1942-43 by Mr. F.H. Hutchinson LS.E.
25. Note by Mr.R. S. Chaturvedi, C.E., A.M.I.E. (INDIA), Assistaut Irrigation Research Officer, United
Provinces, on "The investigation of groundwater resources in the western districts of United Provinces."
26. Report on "Well construction and pumping from wells for drainage and irrigation in California
and Aizona, U.S.A." by Mr. A.R.B. Edgecombe, I S;E., United Provinces, Public Works Department, Irriga-
tion Branch. "
27. "Water supply from tube wells by T .A. Miller Brownlie paper No.39, Punjab Engineering Congress,
Lahore, 1918.
28. "Hand Book on tube wells" by: K:D. Sanwal GovemmenJ::,Prjtiting Press U. P. Allahabad 1944.
29. "Tube well in Borstal Central Jail by W.S.Dorman The Punjab Engineering Congress, Lahore 1920·
30. "Tube well practices of the Public Healt Circle P.W.D. Punjab by D. A. Howelh The Punjab
Engineering Congress, Lahore, 1 9 2 9 . : !. \

31. "Tube wells on N.W.Railway by J. Varden The Punjab Engineering.Congress 1930 papers No. 13t
and 152.
32. The Karol Tube-well scheme by H;L. Vadehm and A.R. Talwar, ;Fhe'Punjah Engineering CongfetJs
Lahore ISH.
33. Bulletin 182 of the Department of Interior published by the Washington Government PreiS .. 11.S,A.

--: 0 :---'
 PART II
CANAL IRRIGATION
Chapter I
CLASSIFICATION OF CANALS
I. Irrigation and Navigation Canals.
Canals are divisible into two main classes viz .• irrigation canals and navigation canals.
Examples of irrigation canals of appreciable magnitude are available in India starting
with the Ganges canal which was designed and built by Sir Proby Cautley in 1845 and the
Madras -canals by Sir. A. Cotton at about the same time. The finest ex:amples of th.e irrigation
canals are now available in the Punjab as shGwn in Plate No. l. Navigation cannot be
expected to be successful on irrigation canals. because they have to follow the main water
$lleds or ridges so as to provide a sufficient head of water to flow over the adjoining land. for
irrigation. Moreover. they have to be designed with velocities sufficient to guard against their
silting up. consistent with the limitation imposed by the permissible scouring velocity for
materials forming the bed and the sides.
An ideal navigation canal on the contrary. should have a very nearly still water
channel, so that navigation may be possible in both directions. rt should generally follow
low country for economical construction. Moreover, a navigation canal should approach
conveniently large centres of traffic. Examples of large navigation canals exist in the
western countries such as Suez, Panama and Kiel canals.
An attempt was made in the beginning to combine both the functions in the case of
.. Ganges, Western Jumna and Sirhind Canals. In these canals, small country boats were used.
o •• ,; The bridges had to be designed high enough to pass the boats underneath and at faUs, lock gate
arrangement was provided. The velocities in these canals are 3 to 4 ft. per second, and loaded
.. boats cannot be pulled for the up traffic by men or animals. The income from traffic by boats
has been too low to justify the cost of the additional works for this purpose.
It is mostly the first class of canals i.e. Irrigation Canals. which will be dealt with in
this volume.
2. Classes of Irrigation Canals.
The irrigation canals are divided into two classes.
(a) Permanent Canali.
A canal is said to le pel manent when its source of supply is SUfficiently well aSSUred to
warrant the construction of a regular graded channel supplied with masonry works for regulation
and distribution. The canals which are provided with permanent canal head works fall in
this category.
The permanant canals may be perennial which receive assured supplies from the rivers
througout the year, or may te seasonal such as kharij channels in which regular supply is
available only in the kharij season.
(b) Inundation canals.
Inundation canals are those which depend for their supply on the periodical rises in
water level of the ri.ver .from whi~h they are taken 0'1. They are not provided with permanent
headworks. Water IS Simply let mto thea. when the river rises through the marginal flood
embankments and the~ are pr?vided with a regulator 3 or 4 miles away from the river. The
proolems connected WIth theIr working are complex and peculiar and are dealt with in
Olapter XVIII Part II.
 43
3. Financial classificatIon of canals.
The financial classification of Irrigation Canals is two-fold-
(a) Productivc works.
When the forecast of the income usually prepared at the time of the preparation
of the project of an irrigation scheme shows that the income from the proposed canal will
exceed the yearly charges of maintenance by a sum equal to at least 4 per cent of the
capital invested, the work will rank as a productive work. The precentage of the net
return is fixed by the Government from time to time according to the prevalent rate of
interest in the market. A certain number of years is allowed for the development of
irrigation after construction of the canal, during which the interest charges accumulate
as a simple debt to be paid off as the revenue gradually increase3. On the expiry of
the term usually ten years, H.e net revenue receij: ts ~Lculd have cleal ed off all the
interest accumulated during the construction and developement, after paying for the running
charges.
(b) Protective works.
Protective pu~lic works are defined to be those, which although not directly
remunerative to an extent .which would justify their retention in the class of productive works
are calculated to guard against a probable future expenditure for relief of the population.
They usually take the form of famine relief works. The construction of an otherwise expensive
irrigation scheme may be started tJ employ the population during famine3. The lining of the
canals though not directly a. remunerative work is a very useful protective work to reduce the
future expenditure on the antiwaterlogging measure and to save the deterioration of lands by
water-logging.
4. Receipt from Irriga.tion Canals.
The receipt which make up the income of a canal fall in two categories (II) Direct
receipts and (b) Indirect receipts.
Direct receipts comprise, firstly, the income of the water rates fixed for the different
crops irrigated; secondly, the receipts from plantations; thirdly, the income from water-power
such as mills or hydroelectric power; and fourthly. the mi'>cellane:>l1s receipts fro:n water used
in bulk for filling tanks, for building houses, for watering road-side trees or for consolidation
of roads.
Indirect receipts are the receipts creditable to the canal department due to the
increase in land revenue by canal irrigation to barani or chahi lands. The difference cetween
the usual barani or chahi land revenue rate and the revenue rate of the irrigated lan<is is the
indirect income due to the construction of the canals. In the case of crown waste land lying
uncultivated, whole of the land-revenue comprises the indirect income of the canals. The
interest on the sale of crown waste land put under irrigation is also credited under this head.
5. Parts of a cana.l.
An irrigation canal is divided into the following parts:--
(a) Headworks.
The headworks comprise all the works necessary to dam, to control the river and to
regulate supply into the canal.
(b) Main Canal.
Usually in the head reaches, the canal is in cutting and below the natural
surface and is not required to do any irrigation. In thi;; reach it is called the main canal.
(c) Branch Canals.
When a main canal leaves the high ground and must, therefore, bifurcate into
branches covering the whole tract meant to be irrigated, the canals in such portions are
called Branch Canals. Very little direct irrigation is done from the Branch Canals.
(d) DUributaries.
Small channels which take off the branch canals alid distribute their supply through
outlets into the water-courses are called distributaries.
(e) Minors.
Sometimes when the country is such that the water-courses will have to be longer
than two miles to reach the fields. it is usual to take off small Government channels from the
distributaries, which are called minors.
(f) Water-courses.
This is not a Government channel and constructed and maintained by the cur
tivators according to the' alignment sanctioned by the canal engineers from an outlet
down to the fields of the cultivators. Their design is dealt with in detail in Chapter No. XVrr
Part II. '
6. Punjab canals.
A list cif the Punjab Canals is given in Appendix 1 at the end of, this chapter. The
earliest canal constructed was Western Jumna Canal about the year 1817 from the Jumna
river. then came Upper Bari Doab Canal in 1850- \859 from the Ravi river and then Sirhind
Canal (1872) from Sutlej river. The Lower Chenab Canal was completed in 1902 from Chenab
river and the Lower Jhelurr Canal in 1905 from the Jhelum River. The Triple Canal Project for
the construction of the Upper Chenab Canal. the Upfer Jhelum Canal and the Lower Bari Doab
Canal was commenced in 1915., In .this project. the Upper Jhehlum and the Upper Chenab
canals are mostly feeder canals connecting the Jhelum river with the Chenab river upstream of
Khanki, and the Chenab river with the Ravi river up stream of Balloki respectively. The last·
named 5 canals (serial Nos. 4 .to 8 in Appendix 1) are called the five linked canal because the
waters of the three rivers mentioned above can be jointly utilised in these five canals according
to the seasons requirements. , , ' .
The canals at serial Nos. 9 to 19. take off the Sutlej-Beas system and were constructed
from 1922 to 1928. and are called the Sutlej Valley Project canals. Most of them are khari/.
channels The Haveli canal was completed in 1939. This is also a feeder canal from
Trimmu (Chenab river) to Sidhnai (Ravi river) and its water has been utilised to 'give perennial
irrigation to the old Sidhnai Inundation canal. .
, The old Inundation canals in Muzaffargarh and Dera Ghazi Khan districts taking off.
the Indus river were taken over by the irrigationDepartment,P.W.D. from the civil administration
in 1880 and the Shahpur canals taking off the Jhehnn river in 1894. The· lower Sohag Canal
and the Para Inundation Canals were built in 1882 from the Sutlej river and have subsequently
been absorbed in the Sulej Valley Canals .. The Sidhnai inundation· canal was opened in 1886
taking off that the Ravi river at Sidhnai and has subsequently teen aosorved in Haveli.Canal
Project. They took water from the rivers in floods, without any permanent head-works.
. The p~ )ject to irrigate the Sind Sagar Doab is under construction' by taking off that
canal with a. disc~arge of about 6.000 cusecs from the Indus river at Kalabagh.
Lots of areas of Hissar and Gurgaon districts in the Punjab and in Bikaner state need
protection by canal irrigation and tank irrigation. Projects for constructing dams in the hiUs are
under contemplation for this purpose.
7. Distribution ()f river supplies to d.I:IIerent canals in the Punjab.
Some canal systems are inter-linked and tt.e supply available in the rivers is not
always sufficient to meet the indents of all the canals. Consequently. the supplies of. various' ...
rivers are distributed among canals according to orders framed from time to time by the Punjab
Gevernment. The following may te taken to te the usual procedure:-
(a.) Western Jumna CanaI.-The Supply at the river Jumna at Tajewala is distributed.
between the Western Jumna Canal in the Punjab and the Eastern Jarnna canal
of the United Provinces according to regulation rules for the canals. .
(b) Sirhilld Clnal.--Th~ Sirh~d cana~ is entitl~d to take up to its m?-xim,um capacity,
all the supply avaIlable m the flver SutleJ at Rupar. . .
(c) Upper Bari Doab Cloal.-The Upper Bari Doab Canal is entitled to take up to its
authorized capacity, all the supply that reaches Madhopur with the sole
exception that duriJl~ kharif from 1st April to 31st !:eptemcer the Kashmrj
 ,(5

(Basantpur) Canal on the right bank has a prior claim up to a maximum of

120 cusccs.
(d) The Northern (Linked) Canals.-The Upper Jhelum and Lower Jhelum Canals taking
off the river Jhelum, the Upper Chenab and Lower Chenab Canals taking off the
river Chenab and the Lower Bari Doa)) taking off the river Ravi are inter-linked
and they are entitled to all the water in the river Chenal:> and Jhelum which they
can take up to theirauthorized capacities. When the supplies of the Chenab and
the .1helum rivers drop helow the combined capacity of the linked canals, a
destribution programme CO'illes into force.
(e) The Sutlej Valley Clnals.-The Sutbj Valley Canals are entitled to all the water that
'.. COUles down the river Beas and a'1y surplus from the riv.;r Sutl~j ov:r and above
the requirements of the Sirhind Canal. The actual distribution of supplies among
. the' three partners. (The Pu:<jab Government, the Bikaner Sta~e. and the
Bahawalpur State) of the Sutlej Valley Canals is carried out in accordaTlce with
the orders issued by the Government of India, based OIl the Anderson Report
of 1935.
(1) T.!e Montgomery Pakpatan li1.k.·~rs entitled up to its authorised capacity to all
the water availa:)le in the river Ravi at Balloki.
(g) The Haveli Canal.-The Rangpur canal, the ThaI canal and the Panjnand canal
are regulated according to the orders issued by the Government of India, based
upon the recommendations of the Anderson Committee.
(h) The Inundation Cluls.-Draw up to their authoriz~d capacities all the water that
they can tap from the rivers both in rabi and kharif.
i. Fnuctions of Canals and other Irfigatim Works.
(a) Proteeti::m.--The most important function of the irrigation works iq to protect the area
dealt with against serious loss during seasons naturally unfavourable for agricultural operationS..
They provide protection against famines. There cannot be any famine in canal irrigated areas.
The surest remedy against the recurring famines in the Hissar district of the Punjab is the
introduction of the canal irrigation.
(b) Improvemen\s o{ erops.-Another function is the substitution of superior for inferior
classes of crops. The natural result of the introduction of a permanent supply of water to a
tract formerly dependent on a fluctuating rainfall. Thus. in some Indian districts, we find wheat
replacing barley; sugarcane and indigo replacing light millet crops and, as a general rule, the
cultivation of mixed crops practically ceasing. _.
(c) Manurial value of canal water,·~The rivers of the Punjab are all alluvial. The water
of the Punjab canals contains a sunciently large percentage of fin~ clay and silt in suspension.
The silt in the canal water has a great manurial value. The yield of crops raised from the canal
irrigated fields is more than that of the crops irrigated from wells without other manuring.
(d) Addition to the wealth of t.he country and the Government Revenue.-The irrigation
works add to the wealth of the country both directly by the enhanced value and quantity of
products, and indirectly by an increase in the valve of the land. The increase in wealth of the
Punjab province can well be imagined from the figures of the irrigated areas as given in the
following table ;--
Irriga~ed area
Year. In Source:
acres.
1867-68 1,025,156 Wells and river spills.
;!
1877-78 t .~, ~
2,341.103 U.B.D. & WJ. Canals added.
, ,.'4;
1907-08 fit I 6,039,944 L.C.C. & Sirhind Canal ap.ded.
1917-18 9,063,901
" Jhelum Canals, U.C.C. & L.B.D.C. addeQ..
1937-38 12.800,()OO Sutlej Valley Canals added.
1942-43 14,000,000 Haveli and Rangpur Canals added.
 The government revenue in the Punjab, as credited to the canal department, from ,the
realisation of the owners and water-rates, was about 8 crares of rupees tefore the World War II
out of the provincial revenue of about 1I crores of rupees. The expenditure on the maintenance
of the canals in the Punjab was only about 1! crares of rupees annually,
The profits of, some of canals r efore 1920 after paying for the interest charges on the
capital outlay are shown in the following tables : - .
Profits expressed as
Province. Canal percentage of
expenditure,
Pwljab. Western Jurnna Canal 12·3%
Upper Bari-Doab Canal 1.6'5%
Lower Chenab Canal 43'6%
Sidhnai Canal 31'S to 51'8%
Sirhind Canal ,., 21'7%
Lower Jhelum Canal 24'0% (now 35%)
United Provinces. Upper Ganges Canal 10'4%
Bijnor Canal 11'0 %
Bombay. Bombay Canals 10·0%
(e} Population,-The absence of famines by the introduction of irrigation results in an
increase of manpower. Good and sufficient feeding results in good health of the population. There
is also an increase in popUlation. This increase in popUlation is absorbed in additional labour
required for agricultural operations and for the maintenance of the irrigation works. The
robust manhood of the Punjab has always been the pride of the Indian Army,
(f) Effect on climate and health.-The Irrigation Works such as canals and storage
reservoirs are feeders of the sub-soil supply. With the rise of water table, the climate becomes
damp. Irrigation on large scale affects the climate by making it cooler and damper particularly
at night. The average temperatures in the Sargodha and Lyallpur districts of the Punjab
have dropped by about 10 degrees by the development of irrigation, on account of the increase
in the cropped area and the other vegetation in the form of gardens. The dust storms have
been reduced by 75 per cent. Damp climate and vegetation naturally increase some diseases
such as malaria; but this is more than counterbalanced by the increased vitality due to better
standard of livmg_
(g) Phntation.-The banks of large canals are generally planted with trees, This is a great
advantage to the tracts through which they pass on account of the shade, timber, fuel, and fruit
that they provide. Trees and shade are a real necessity in very hot climatES, and as constant
extensions of irrigation tend to break village plantations up into tillage, the niaintenance of
permanent and well-cared-for plantations on the irrigation works themselves become desirable.
Moreover, these plantations, if well managed, become large sources of revenue.
(h) Navigation,--Another important function of a large canal is the facility it offers for
navigation. This function is often neglected probably because the engineer is Irore attractEd by·
the manifest advantag.;s of the result of the attention he pays to extensions and improvements
in irrigati n. It would be bett(r, however, to take a broader vi6wof the situation and endeavour
to develop all possible sources of advantage to the people and State, and the utilisation of the fine
waterways formed by great canals should not be neglected as it has so far been,
(i) Bathing.-The domestic advantages of irrigation works should not te overlooked.
The facilities given for bathing and watering cattle must be greatly appreciated by the inhabit<J.nts
of the towns and villages situated near the banks, In the colony areas of the Punjab, can~l wafer
is USed for drinking purposes as the ground water is brackish. ' ! (l 1
(j) water-Power.-Water-power is frequently made available for Use by the constntclion
of canals and tanks, In India up to the present time this source of revenue, has to a large, extent
been neglected, owing mainly to the fact that the sites suitable for power generation (falls) are
frequently far removed from the manufacturing 'centres. Electrical generation and its
transmission o~fers such a simp~e solution of this difficulty. tha~ ~e may look forward ~o seeing a
great advance m a few years hme and probably many great CIties may soon owe thelr lighting.
ventilation. and commercial prosperity to the same beneficient work that supplies them with food.
The Ganges grid system in the United Provinces in India has proved a great success in the
electrification of a large tract of that province.
Considering the above-mentioned advantages of canals. the irrigation is the best cottage
industry for the predominently industrial Province of the Punjab.
7. Examination Questions.
(i) Explain the follOWing terms : -
(a) Water-courses, (b) Branch Canal, (t) Minors, (d) Indirect r~eipt8,
(ii) What are productive and prot'llCti\>e canal works P
(iii) What are the difficulties in having combined channels for irrigation a.nd navigation t
(iv) What do you understand by permanent and Inundati'On canals?
(v) Explain briefly the Triple Canal Project and the five linked canals of the Punjab.
(vi) Elucidate: - "Irrigati'OD is the best cottage indus try for a predominently agricultural Province of the
Punjab." (P. U. 1942 )
 Chapter I
AJeP-,,-",DI,{ 1
CANAL IRRIGATION IN THE PUNJAB
PUNJAB CANALS
i ..,
I ,
. ro
~
. :.0
<Il'O ro Delta
0 .; Gross -"
.0 '0 .. :0 .~ ~ ~
roN
Z r ...ro
bO
~::l ~ ~... 'iii ....
! Name of Canal. area. ~
~'E~ ~.S
-;;; I ..cu ;:;6ro ·s......~ ::l .... - cO ....
·C
ro :.0ro
..,
·C I on ::l 0
UU Q.l ro
..... ro
g~ g
~
.;; C_O'
~
(J)
I i p., I

I! 2 I 3 4 5 _, 6 I· 7 1 8
\91 10

i
~- \~es~-~~-n-'.]-]-~-~-n-;~e--~a-2-·~-~-aa-I--'·---8-00-0~-:-2-7-2-5-88-1--'-1--2-2-78-0-0-8 ~~ 8905~:~~I~:6~r]~~-e
2 I! Sirhind
British
Canal (S .. C.)
Area 5257 ~
9040K
2424776
I
I' 2026205 773072 1395622 1: 1 2'0
'1
1'2
{Rupar)
3 I1 UppC'r Bari Doab Canal
I [U.B.D.C.] Madhopur.
I6900K
~5770R 1564135
'
1445707
I 1081465 1268971 I: I 2'E
4 Lower Bari Doab Canal 7000K
[Balloki] L.B.D.C. 1801410 1438166 I 939415 1249779 4: 5 3'3 1'5
tis;>\).\.
5 Vpper Chenab Canal 13084F
[Marala] U C.C.
Used in Irrigation 5180 1536852 1445000 566860 704643 I : 11 3'4 1'3
6 Lower Chenab canal 11 032K
[KhankiJ L.C.C. 995SR 3772715 28732691 1873867 2472500 I : 212'7 I"
7 Upper Jhelum Canal
l\1argla Head (D.] C.) 87 83 F 1 : 1 2.9 1·3
I 8 47K 538Il7 502735 I 303666 33951;7
Used in Irrigation 1242R
8 Lower JheJum Canal 4099 1340645 1239597\ 756390 936934 I: 22'33 1'34
[Pasul] L.J .C.
9 Pakpattan Canal
[Sulemanki]
6094+ 735 1392766 1271122 57% 729000 t 1.513 3 1'6
[M. P.
Link]
]0 Dipalpur Canal
[Ferozepur] 6950K 1044060 978823 50% 486622 1: 1 3'91 '3
11 E8stern Canal
[FerozepurJ 3320K 395039 347965 46% 160000 1: I 5'4 '3
12 ! Gang [Bikaner Canal] 2720 Irrigates Bikaner State.
1 Eastern Sadiqia Canal 4917 Irngates Bahawalpur State.
13
[Sulemanke]
14 Fordwah Canal 3366 do
[Sulemanke]
15 Mailsi Canal
[Islam] 4883K 752964 688770 5,(% 373400 1: 1 '('S
16

17
Bahawal Canal
[Islam]
Qaimpur Canal
3000 Irrigat" Bah.w.,pu' ".,..
do.
I
[Islam]
18 Abasia Canal
[Islam]
230 i do,

19
20
Panjnad Canal [PanjnadJ
Haveli Canal [1939]
[Trimmu]
7770
5249K
2750R
III 16096 10044631 736276
do
692000
(1941)
21 Ran~pur Canal (Trimmu) i 2710K 318425 306124 183674 107303
(1941)
22 Shahpur Inundation 12010 Max:
Canals (5 Nos.) ,----- 128000 108000 66170 81864
(]helnm river) 1
1295 ord: \'
%3 Muzaflargarh i9 .
Inundation canals ' ~~5 Max'l 660367 595000 358165 359455
(Indus river) 7 Nos, 6270 On!: I
24 Indus Inundation
1
I I Ij ...
25
canals. Dera Ghazi Khan
(9 Nos.) Indus River.
Ghaggar Canals 3 No.
17350 ord:
I 1530
I
,I 0955max: 702294

131228
M£9461 291123 J 331917

L 64143 44407
I
107596
I'"
26
Inundation.
Swarasti Canals 2 Nos.
(Inundation.)
i
597
I
i 188761
r
154880 J 77440 I 33020
I

I. ·
Note.-K=Kharif Max: = Maxirnum F. = Fall supply
R=Rabi Ord:=Ordinary
 PART IT
CANAL IRRI GAl ION
Chapter II
PRINCIPAL CROPS AND:ASSESSMENl'
1, Introductory.
The primary object of all types of irrigation whether lift irrigation, cand irrigation
or tank irrigation is to supply water to mature agricultural crops. It is, therefore, considered
desirahle to acquaint the students with the principal crops sown and matured in the Punjab.
The procedure of their assessment is also described briefly. The major portion of the revenue
of the Punjab is derived from the assessment of the irrigated crops. The economical design.
construction and the maintenance of an irrigation project is a means to an end which can
be jurlg'2d in its efficiency from the quality and the quantity of the agricultural crops produced
and the revenue which they bring to the state. The subject matter of this chapter is no doubt
the domain of an agricultural engineer, but as the efhciency of an irrigation engineer can
best be jurlged from the revenue return from an irrigation scheme, it is, therefore, essential
that he is incharge of the assessment of the revenues accrucing from irrigation enterprize anti
is also equipt:ed with the requisite knowledge about the agricultural crops.
2. Soil conditions in the Punjab.
(a) The quality of crops grown depends upon the texture of the soil. The texture
of soil is determined by its clay contents. An average soil in the Punjab contains clay from
12 to 15 percent. The soil containing less than 2 per cent clay is useless for crops other thn
Barani grams.
Classillcatlon of solIs.
Class. Clay content% Suitable crops,
Heavy soil. '.; -I 40 Sugarcane, rice cotton and wheat
(Produce of the last two below normal).
Normal soil 10-20 Cotton, wheat: maize, vegetable, oil
seeds, fodder crops. All give the best
product.
Light soil ~ -8 Wheat, Gram, fodder crops etc. Yield
of wheat celow normal.
(b) The ~hemic~l charac.teristic of the s<?il is very im~or~ant to determine the suitability
.f t~e crop sown 1ll a SOIL The Important chemIcal charactenstIcs are : -
(i) Salt content.
(ii) pH value which defines the degree of alkalinity or acidity.
The principal salts found in the Punjab soils are sodium chloride, sodium sulphate and
sodium carbonate. The yeild of the major crops such as wheat, rice etc., is not afLected up to
'18 percent of the total salts. After which the yield drops and at about '25 percent of the salt.
soil becomes absolutely infertile. The average salt contant in the Punjab soil is '05 to '15
percent increasing towards the westernly direction.
The pH figures give an indication of the alkalinity or the acidity of soil on a
logarithmic scale of the degree of ionisation. The scale ranges from 0 to 14 ani the figure 7
indicates neutral soil in the sense of chemical reaction. Below 7 the soil is acidic and above 7
alkaline. In the Punjab the pH value of soil is everywhere more than 7. It veries from 7
to 11 rising in westrenly direction. The average value in the cultivated areas of jullundur
Hoshiarpur, Gwdaspur and Sialkote districts is 8'5. It is now very nearly established by tll;
 50

researches in the Irrigation Research Institute at Lahore that soil with pH value from 7 to 8'5
gives normal yield of all the principal crops sown in the Punjab. The yield declines with
pH value from 8'5 to 9 and with pH value equal to 11 the soil is infertile. The soil of
Gangetic plain in U. P. is very nearly neutral and there is no trace of sodium salts. The pH
value is no where higher than 7.
3. Rainfall.
The crops require fairly well defined quantities of water at definite stages of their
growth. Mr. Wilson, the Scientific Research Officer, Lahore, estimated the following total
requirements of water for some of the principal crops:--
Crops. Irrigation in field.
Rice 44"
Sugarcane 40"
Cotton 20"
Wheat . 12"','
Maize 18"
Fodder 9"
The rainfall will reduc~ the depth of irrigation required by artificial methods.
(b) In the Punjab, the rainfall is very variable. The contours of equal rainfall called
Isohy'ets are shown in the map of the Punjab appended as plate 1.
It is evident from this map that in a very narrow strip near the foot of the hills b::>rder-
ing north-east, the rainfall is 35" to 40" on the average in a year. In the plains the rainfall is
, ; . everywhere less than the normal requirements of the depth of waterings required for the princi-
1;. ,pal crops sown in the Punjab. The rainfall cannot be turned on and off as required by the
crops at the time of sowing and maturity. Moreover in the Punjab, the rainfall is concentrated
up to 7S percent in the rainy months of July and A.ugust and for the remaining part of the year,
the rainfall is very scanty and is very precarious.
It is therefore, that the Punjab needs very badly the help of artincial irrigation which
has lead to the development of canal irrigation system and which surpasses in its efficiency all
sueh works in the world.
4. Prin(ipal crops in the P'lnjao.
. The principal crops sown in the Punjab are given in the Appendix. No. I appen~ed. to
thIS chapter. The English and the Hindustani names of the crops are gIven. The tune of
sowing and harvesting the crops is given in columns Nos. 4 and 5. The column No.6 gives value
of depth of the canal irrigation required in feet known as Delta. Column No.7 gives the average
produce of the crop in maunds. Column No.8 gives the approximate water rate charged for
the irrigation water. The value of delta varies from canal to canal according to rainfall and the
efficiency of the cultivators. Similarly water rate changes from canal to canal according to the
soil condition and the anticipated produce. The average produce is very variable according to
the nature of the soil from one district to another. The average produce is fixed in the settlement
reports of a district according to the nature of the soil. The figures of the columns Nos. 6 to .~ __
should be taken merely as a guide to acquaint the student with approximate usual valueS.-·-"
5. Fertilizers and Manures used for agricultural crops.
Introduction.
Of the principal food materials required for growth of plants by far the most important
one in which the Punjab soils are dencient, is nitrogen. The deficiency of nitrogen is the main
problem of manuring. Much of the farm-yard ma::mre is burnt as fu'.)!, while a larg~ quantity of
combined nitrogen is exported in the form of oil seeds, food and other grains, and animal products
such as hides and bones. The only way of stopping the wasteful practice of making manure into
cakes and burning it, is to provide an alternative supply of fuel hy planting quick growing trees
'near the villages. .
 51
Farm-yard Manure.
Farm-yard manure increases the retenhve power of the soil for dissolved substance~:
It causes the soil to be puffed up and increases the pore space which improves the tilth and
general conJition of the soil. It increases the water holding capacity of the soil and helps the
biological activity going on in the soil.
Farm-yard manure contains about 75-80% water,' 0'6-0'7% nitrogen, 0'1-0'3%
phosphorus (as P 2 0 S) and 04-0'5 :/0 potash (as K 20). The urine is much richer in nitrogen and
potash but contains only traces of phosphorus. A to~ ?f man~re c<;mtains about 15 lbs ..nitrogen.
5 los. phosphorus and [0 hs. of potash. The composItion vanes wlth the type of the ammal, the
quality an~ the quantity of food, proportion and nature of litter and the stage of decomposition
that has taken place in the manure itself. 5illce the liquid part of the manure is much more
valuable as compared with the dung, far more car~ should pe taken to preserve this than is
usually the case. The 'Lest method of collecting urine is the use of litter. Any waste material·-
wheat straw. w'z'Jat b'ZU:>l, tJ/ia or SZ'SO,'1, p.tlllr," sug ~rcane trash, grain bhussa and any other
vegetable waste can be successfully used as an absorbant of urine. The litter should be spread
under the catth in the evening and may be carried next morning along with the dung into pits,
which may be of any suitable length and breadth and only two feet deep. When
the pit has been filled, water may be occasionally sprinkled over the stuff so that it remains moist
but no free water stands. A turning is given to the material after every two sprinklings of water.
'Ihe ~tuti becomes ready tor carting in. a;}out g mont'ns' time.
As stated in the introduction, quantity. of f~Hm-:yard manure available for use for crops
is inadequate. Something can be done in thls dlrec hon by the manufacture of composite
from any rwbbish material on the lines suggested a )ove. Compositing of waste material is now
a recognise? part of the activity both of the ag!i.cultural exrerim~nt stations and village improve-
ment assoclations and we may expect some addltIon to the manunal resources of the country as a
result.
Green Manuring.
Another method by which the deficiecy of farm-yard manure can be made good is by
green manuring. It provides all plant food ingredients in the soil and supplies a large amount of
humus which improves the texture and water holding capacity of the soil. Experiments show
thatGowara is the best crop for burying in as a green manure under condition in the canal
colonies while (san) hemp appears to be the most suitable crop in the submountainous tracts.
Green manuring with D'~ainch(l has proved exceedingly advantageous under waterlogged
and alkaline conditions prevailing in the l(allar tracts of the Punjab where rice is mostly
grown.
Artificial fertilizers.
Of all the essential elements required hy the plants only three viz: nitrogen, phosphorus
and potassium are important froal manurial point of view, because they are found generally in
small quantities in the soil.
(a) Nitrogenous fcrtiliZers.
They tend primarily to encourage above ground vegetative grGwth and impart a green
colour to the leaves. They delay maturity of the crops. In cereals plumy condition of the
grain is increased while straw gets weakened with exce3sive applications. Examples are
ammonium sulphate, sodium nitrate.
(b) Phosp.llatie fertilizers.
They are a? .essential part in t~e formation of grain, hasten the maturity of crops,
counteract the condItIOns produced by mtrogen, encourage root development especially of the
lateral and ~brous rootlets, ~trengthen the s.traw cer~als and increase the weight of grain
comI;>ared WIth ~hat of straw, mcrease the quahty of gram of cereals and of grass in pastures
and mcrease reSIstance of plants to dieases. Superphosphate is the common material used as a
phosphatic manure.
(c) Potassi3 fertilizer.
The presence of available potash h;;\s much to do with the general tone and vigour of
the plant. By increasing resistance to certain dieases it counteracts \he ill effect of too mucR
 52
nitrogen, while delaying maturity it works against the early ripening effect of phosphorus. It is
essential in formation plumb and heavy grains. Putassium sulphate is the material used as a
manure.
The results of manurial experiments in the various parts of the province with regard t.
th~ use of manures have shown ;-
. (i) Farm- yard manure has almost always given the best results. It is the cheap-
est means of adding organic matter and nitrogen to the soil. Average quantity required per
Ol"Op is about one to two cart loads pC! kanal (tacre).

(ii) The application of nitrogenous fertilizers such as sodium nitrate and ammonium
sulphate has given good results. The quantity of manure required per acre per
crop is al;out one maund. These manures increase the yield of wheat, but the
increase hardly covers the cost of manuring. In the case of cotton and sugarcane,
these manures have given profitable returns.
(iii) The application of phosphatic and potassic fertilizers has almost always resulted in
a financial loss. This is due to the fact that the market prices of such fertilizers
being high, it is not economical to use them for agriculture.
•. Plant Diseases of thc Principal Crops.
(i) Cotton. Cotton is mostly damaged by ;-
(a) Boll worms.
(b) Jassids.
(c) Root rot.
(a) Boll worms.
In the beginning of the cotton season the cater-pillar bores into the top tender portion
of the shoots. When fiowers, buds and bolls appear, the larvee turn their attention t) them and
flowers add buds of bolls of the plant are attacked; observations have shown that up to
75 percent of the flower buds and 60 percent of tne bolls may be damaged by this pest. .
Remedies. There are parasites which feed on the boll worms. This parasite should
be encouraged. .
In the beginning of the cotton season the attacked shoots should be collected
alld burnt.
. The plants should be shaken thoroughly by dragging a rope over them. The boll worms
w(U drvp on the ground. The ground should be immediataly watered to drown the boll worms.
After cotton crop is over, cotton sticks should be removed from the fields, cutting them
2" below U e ground.
(b) Jassids.
The adults are reddish in winter and greenish-yellow in summer. The pest attackS
€6tton in Juneand remains on the plant up to Novembtr. It sucks the sap from under side of
the leaves with the result that the fruit capacity of the cotton plant is very seriously affected.
No effective control against this pest is known. Rough hairy leaved types are slightly resistant
tt) the attacks of the jassids. .
(c) Root Rot.
H appears in patches. No effective remedy is known. It is suggested that cotton should
Bot be sown for a year or so in the field in which root rot has appeared.
(ii) Sugarcane.
There are two main diseases; red rot and stem borer.
Red rot is an insect disease. The insect bores into tht stem. Resistent varieties should
be -sown. The caterpillar in the case of the stem-borer attacks the central shoot and kills it.
This is called the 'dead heart'. These dead hearts should be collected and burnt. The insect
also hybernates in stubbles which may also be removed and burnt
 53

Interculture and irrigation recommended.

Crude oil emulsion placed in irrigation channel is also useful.
(iii) Rice.
The caterdillar is the most destructive pest of rice. This kills fhe central growins
leaves which dry up. but when plant is attacked at t\e fhwering stage. the ear-h~ad stands beeD
and is devoid of grains. To control this pest the best remedy is not to provide facilities for the
borer to breed. This can be prevented by the a')sence of the early crop of the rice. Rice stu"jbles
should be ploughed up. collected anj burnt. .
(iv) Gram.
Gram suffers from wilt in the month of February. After flowering, the plants dry uP'
Recent work indicates that salt in the soil accelerates the intensity of the disea'ie. In reclaimeo
land the attack is less.
The gram cut-worms.
The caterpillars get into the pods and des(roy the seed. The damage never is very seriollS,
(v) Weelt.
The most common diseases are:-
(a) Ear-cockles or Mttmni.-This is caused by worms which produce black bolls in
place of grain. Seeds free from disease should be sown.
(b) Rust or Kttngi.-This is a fungus disease which is serious only in cloudy wintel's.
The leaves show yellow or black spots. Rust nHistent varieties of the seed should
be sown.
(e) SUint or Kangiari.-When this fungus attacks the wheat plant black powdery suB-
stance is produced in place of grain. Th'} seed before sowing should. be treated
with hot water.
(d) Bunt.-This larglyoccurs in hills.·~ 'Kanctls of grains are destroyed and replaced b~
ill-smelling blackish powder. Seed from healthy crop should be sown.
7. Economy of Water.
Water is the fiinished product of the Irrigation Department and has been broug~t to tht
. fields by incurring a great expenditure in the construction of head works and the canal system.
It is very essential that it should be used very ~cono nically. It is both in the inter~st of the
Government and the cultivator to use it carefully and economically and thus o')tain more and
better crops.
(a) The cultivators should be encouraged to line the water courses. They can savt
about 20 percent water and thereby increase th~ area irrigated by 20 percent.
(b) Wastage should be avoided by adopting khat kia'fl
system of irrigation. In this system a field is divided into
AU kiaris of one kanal and su1)sidiary watercourses are construted
so that water does not flow through the already irrigate 1 por-
" tions of the field. When one ki((ri is full. wateris closed and
,! led into another one. The arrangement shown in Fig. 1.
There used to be legal binding to enforce this very econo-
mical system under the rules of the canal Act No. VlII 01
:51, 1873 but this has been stopped since 1928. The enfring~­
ment of these rules used to be punished by levying additional
water-rates.
(c) The wastage should be avoided by proper alignment
of the water courses so that there is no heading up in thf
Fig. 1 water course to command any field.
(d) The water cours=s should De properly maintained to avoid wastage from breaches in them
 54
(e) Excessive depth of watering should b~ 'l.voided. It is no u.se flooding a field with
one foot depth of water when death of 3 inches will do.
(f) Economical use of canal water is an effective anti-water-logging measure in the
water-logged areas to reduce the menace water-logging. Excessive watering is also injurious
to the plants.
8. Sub Irrigation.

ll, ,Ir-

There should not be any bend up or down

as the pipe line is not closed, that is,
water tight. If the ground is very uneven
the arrangement should be as shown in Fig. 3
(iii IS" deep trench slightly wider than
the external diameter of the pipe be dug
and pi pe laid a little water to flow after laying 8 to 10
feet length.
(iii) The lateral spacing of pipe line differs for different crops. Lawn 8 ft. intervat
Lucerine 6 ft. interval. and vegetables 3 to 4 it. apart.
•
(iv) There should te a vent at the end of each pipe line for inspection and to prevent
air lock.
(v) The crops need irrigation as below in dry weather. Grass and cereals 3 to 4
days. Vegetables and Lucrines 2 to 3 days.
(l) This system of irrigation is only possible when water is available every two or three
days but in smaller qnantity. The initial expenses for laying pipe lines are very heavy.
This system can only suit small gardens under' strict supervision. Water saving is no dOUGt
large.
9. Internal distribution of water.
Water is supplied from a permanent masonry structure called outlet or moga, the
capacity of which is fixed according to a water allowance for a definite area as explained in
detail, in Chapter No. (iv) Part II. The chaks are designed in such a way that ehe discharge
of the outlet is not less than one cUSeC and more than 2'75 cusecs. Water is conveyed to the
fields through channels which are called water courses or "khals". Mainwater courses are de~igned
by the Irrigation Department and the work is carried out, the cost being recoverable from the
cultivator with the land revenne in small installments called acreage rate. .
All water courses are essentially zamindari channels. The cultivators are responsible
for their maintenance under provisions of the Canal Act No. VIII of 1873. Usually only one
 55
naka is given from the main sanctioned water course to every square or holding of a cultivator.
The water courses as shown in Fig, 9 are constructed alJd maintained by the individual owner
of the square or a holding,
(/;) The distribution of the water by the cultivators amongst themselves is essentially
the concern of the zamindars. [hey can take the help of the Canal Departrr:ent under provisions
of Section 68 ot the Canal Act No. VIn of 1873 to get share or wari of individual sharetlOlder
fixed acccrding to the area owned Ly him in a chak. Only Divisional Canal Officers are
authorised to investigate and pass orders on warabandi cases. The turn is usually tix d in 'hours'
and sometimes in 'pehars' in villages whue the cultivators are illiterate. l~sually the cultivators
maintain a common clock and a gong.
(c) The cultivators then follow the turn and the time allotted to each shareholder of
a water course. Lots :of disputes arise in the internal distri'oution of the canal water but a
canal Officer does not interfere except as provided under the Canal Act Section 68. The
revenue staff, the zilladar and the p:ltwari, are supposed to give the necessary advice when
required by the zamindars. In the cases of warashikai (taking water out of turn) only
r"medy available to the canal department is that the person who disregards an authorised
warabandi promulgated by the Divisional Ca:1al O.hcer u'l.ier Section 68 of the Ca:J.al Act,
can be charged special rates under Section 31 and 33 for using water in an unauLlor sed manner.
Such cases are instituted by the canal officer on the written application from an agrieved
person. It is generally possible to set right irregular working of the wari by rendering
sympatfJetic advice to the cultivators.
(d) Nikal is often the cause of great trouJle. The last man receiving wa.ti on a
water course ben~fits by the amount of the water contained in the water course when the
turn changes to the lands in the begining of a water course. Sometimes the cultivators who
get water at night have a great grievance. The night wari can be changed to day wui in
alternate year.
(e) The changes in chak bandi, ndas and supply of water through the intervening
water courses are made under Section 20 of the Canal Act No. VIII of 1873 by the Superin-
tending Engineer and the Divisional Officer. The Divisonal Officers posses final power in the case
of water courses and the Superintending Engineers in the case of chak bandi changes.
10. Un~Zs of measurements of arelS.
The land measures are of two kinds, in the Punjab the bigha measure and the kanal
or ghumao measure. They are shown in the following table:-

1. (a) Ordinary Bigha Measure,

-Description. Indian equivalent. Area in sft" Area in acres. Rpmarks.
1 sq. karam 1 B swansi 22'6875 '000521
20 Biswansis 1 Biswah 543'75 '010416
20 Biswahs 1 Bigha 9075'0 '2083=5/24
I. (b) Shahjahani Bigha Measure.
Descriptl~·o:=n.:_._ _ _ _Iu-cdian equlUalent. Area in sft Area in acres. Remarks,
~'Gatha 1 Biswansi 6806 '00156 Note-This IS
20 Biswansis 1 Biswah 1361'25 '03125 used in southern
20 Biswah 1 Bigha 27225'0 '624=5/8 Punjab and in
States.
II. (a) Ghumao measure; Karam 5.0.
Description. Indian equivalent. Area in sft. Area in En.((lish acres. Remarks.
1 sq. karam 1 sarsahi 250 'UUU574 (a) Th'iSTsliSed in
Central Punjab.
9 Sarsahis (kan.) 1 marla 225'0 '00516 (b) 1·21 ghumao
20 marlas 1 kanal 4500·() '1033 equal to 1 acre.
8 kanals 1 ghum ao 36000·0 '8264 {c} One karam is
two paces of a
man or 3 trots.
II. (b) Acre measure; Karam=5f (colony areal
""D""-es-c-ri"'p-t-iO-u_-._-::::: ----iI"'n"""dian eq u i valent. Sq uare feet. Acre.
1 Square karam 1 sarsahi 30'25 -000694
9 Sarsahi 1 marla 272'25 '00625
20 Marias 1 kanal 54'45. '125
8 kanals 1 acre 43560. 1'0
 (ii) rutting a Government Irrigaion Channel.
(iii) Enlarging or damaging the outlets.
(iv) To oj::en the outlets when closed under proper authority in tatils.
(v)Warashikni (taking water out of turn), the powers have now generally teeJl
transferred to panchayat.
(iv) Waste of water" Abzai."
(b) Water lost from breaches from water courses is. charged at special rates as
it is the duty of the cultivator to maintain the water courses in good condition.
The special charge levied in each case can be upto six times occupier's rate at the
diieretion of the Divisional Officer.
14. M.iscellaneous receipts.
Miscellaneous revenue falls under the main heads and is charged and collected by the
Slib-Di visional Offic-er according to the rates sanctioned hv the Gover!1ment w hen the authority
to 'IlSe water has previonsly been o'Jtained from the Divisional O'flcer.
1. Mill rents.
2. Fixed contracts for water supplies to Municipalities and other local todics.
3. Minor items. Water for construction of hou'>es, burning bricks etc.
The usual rates for miscellaneous use of canal water are as below : -
Rs. Unit
(a) Brick making and pise we 0 3 0 100 eft or 100 No.
(b) Concrete & masonry 0 2 0 100 eft
(c1 Metalling of roads 10 0 0 per mile
(d) Consolidation of kacka roads 30 0 0 D~r mile
(e) Water "upplied in bulk 1 0 0 2500 eft
(j) Local bodies 1 0 0 6000 eft
(not commercial purposes)
(g) Watering avenues (kharif) 2 8 0 per mile
(rabi) 5 0 0 per mile
15. Duty of canal water.
Duty is the measure of the working of a channel. It is defined to be area irrigated
a crop, per cusec of the mean discharge. It is got by dividing the irrigated area by the mean
discharge utilized in the period in question.
The mean discharge in any month is the aggregate of daily discharges during the
month divided by the number of days in a month.· The mean discharge of a channel durin~
a orop is the aggregate of the daily discharges throughout the crop divided by the mtmber 9f
days in the crop.
Kkarif crop=April to September=183 days.
Rabi crop==October to March= 182 days.
Delta is the average depth poured on land in feet in a crop or per annu"!Wl.
DeltaL.=2-~-N (1)
Where L. =Depth of water in feet
Q=Mean daily discharge in a crop
A=Area irrigated in acres
N =Numcer of days in a crop
Delta in feekoAcre foot.
Acre foot is the term llse:! in America. representing one foot depth of water ever 6ll1t
acre, i, ,. 43560 eft of water.
Full supply duty of achal'lRel is th area irrigated divided by the full supply capacify
of tke channeL
 It
Full supply factor is the authorised or the stipulated area per crop or per annum for
(:)ne cusoc discharge at the outbt head. This is used to work out the proposed discharge @f
the outlets.
Kharif rabi ratio is the ratio of the proposed areas to be irrigated in kharif crop and
in rabi crop. The mual ratio is 1:2. i.e. kharif area is one half of rabi area.
Intensity of irrigation is the percentage ,)f the cultura'Jle commanded area which is
proposed to ?c irrig~te~l. .Son:e land should. be allowed to have rest ?uring 0l:le year.. The
usual intensIty cf IrngatIOn In colony area IS SO to 100 percent and In propnetory vllla~es
about 40 to 60 percent depending on the existing wells in the area.
16. Eillcien~y of ca!1al L"rlgation.
In consequence of the. devehpment of th_e c~nal irrigation throughout the Punjab
province, the wa~er of all the ~Ivers have to be dIstnbuted .mo~e ~arefully than i!1 the pa'it.
I t is very essentIal that the engmeer sho~ld watch that the dIstnbutIOn of supply IS eqvitab1e
to the individual watercoursos. The efficIency of the actual performance of the water used from
individual outlets and the channels is watched by an engineer by adoptihg the following
methods:-
(a) Actual discharge observations of the water course by the use of a\lthor's
pJrtable tin flumes or ciwllete wires.
(b) By watching monthly delta nlculations for every channel in a delta statemont
and then final delta for the crop.
(c) By ploting the outlet efficiency dia.grams. The form and the outlet efficiency
diagram is given in Appendix II appended to :thiS chapter. No outlets .are reduced simply on
account of excess in the area irrigated by a water course because It may to due to good
husbandary or timely rainfall at the sowing period.
17 Examination Questions.
1. A distributary is designed foralternate runnin~ to irrigate 4800 acres in ,.abi with discharge of 44.-
cusecs. In a certain year it runs for 60 dayti in rabi with an average discharge of 30 cusecs and irrigates 3000
acreS. Work out the channel's (il Full Supply Factor, (ii) Duty [iii) Depth of water (Delta) and explain what
you understand by each.
2. What do you undentand by fluctuating assessment ? How doell it diff~r frora permanent
asseSiiment ?
3. What measures can an engineer take to ensure economy in the Uile of canal water?
4. Exp:ain the follwing term~:-

(a) Nikal, (b) Khata Khasm (c) B::mdo1ast, (d) De:ta., (e) Efficiency diagram, (fl WJrashikni,
(g) Tatiiing, (h) Shajra, (i) Shudkar, (j) Khatauni.
5. Describe briefly a few of the common diseases of the following crops and their remediei.
(a) Rice, ,b) Sugarcane, (c) Wheat, (d) Cotton.
6. Give the times of sowing and maturing and the likely produce in maundi per acre f!lC ttl, following
in the &anal irrigated tracts :-
(a) Wheat, (b) Rice, (c) Cotton, (d) Maile, (e) Tobacco, (f) Barley.
7. What factors govern the suitability of the soil for various crops?
S. How doe .. all. Engine6r watch the efficiency of the distributing channel. P
 60
Chapter II Appendix No. t.
Principal crops grown on the canals in the Punjab.
1 2 3 4 5 6 7 8
<5
Z
Name of Grop. Sowing time. Harvestiag time. '"M8oof:!. ",;::Q)
...
OJ ...cd
a;
"'::I'd "
~ 'S:;s ~
~ ._ Pol •
~ • ~
·c English Hindustani From To From To aj
Q)
Ifl
0 .;; o...S 8. ~~~~
---------,-.---"~----

Kharif Crops.
Rice Jhona, Dhan June July Sep. Oct. 2·5 8 6/1
Chawal Munji
2 Maize Makki June Sep. Sep. Nov. 1'5 14 ,,-,' 41·
3 Great Millet Jowar June July Sep. Oct. 1'0 218
... Spiked Millet Bajra July Aug. Oct. Nov. 10 8 3/4
5 Millets Charri May july Aug. Oct. 1'0 32 2/8
6 ~oth June July Oct. Nov. }",) 8 2/8
7 Pulses Mung, Mash June July Oct. Nov. 10 8 3J4.
8 Vetch field Guara April June June Aug. 1'0 32 ',1;) 3/8
9 Cotton Rui,. Bari, KaJ?~ I
April July Sep. Jan. 15 61 51 4
10 False Hemp SaUl July Oct. 125 8 4/12
II Hemp San July Oct. 1'0 8 . 4/12
12 Italian Milet Kangni March April July Aug. 1'0 32 2[8
13 Indigo Neel ~ ~ ·ti ,_-._'~!
March June Aug. Dec. 1'5 8 ~l }
6/4
14 Henna Mehndi June Aug. Oct. Nov. 1'5 8 :ri 6f'
15 Gingell)' Til June July 8 4/4
or sesame ;t ,.;
16 Sugarcane Ganna,ponda March Feb. 3·0.':!· 30' ll!~
Naishkar., kamad
,/ ""'l~'
~. --
'is'
Minor Millet
Chillies
Madal
Mircl:l
May
Feb.
June
March Sep.
. Nov.
2'0 if, 8
1."1
\ :~
,?t
2/8
6/4
19 Melons Kachra (phut) July ;';1'
Oct -75 8- 4/12
Water nuts Singhara June July Oct. Nov 8 .to 7/8
20 ,(11 t--·
21 Turmeric Haldi March '\
April Nov. Dec. 8 $ ,;
6/4
22 Vegetables Tarkari Feb. May May 8 5/8
23 Lucerne Lusan Oct. Dec. J';~e uP ;:D
2f8
24 Grass Ghas May Aug. May Sep. l' 2/ 8
Rawan June
)if)
July Oct. Nov. ·,d 3
25 Cowpea 2/8
26 water melon Tarbuz -. July 8 4/12
.',"'. May
27 Raghior June Nov. 2/8
Mandhwa or'
'''{ ,I" ~dD1arif L '0'" (,f;:,

1 Indian Repe Toria

~l:·f
Sep.
Nov.
Jan.
Nov.
S
8
" 414
2 Turnips Shalgam1," fr"J ' Feb. 1'75 i 5/8
3 Potatoes Alu Jan. 21l 8, 5/8
4 Vegetable Tarkari Oct. Nov. Feb. 1-75 8 SjS

1 Wheat Gehun, Gandam

Rabi Crops.
Oct.
Oct.
. Dec.
Nov.
April
March April
125
1-5
14 4/4
2
3
4
Barley
:Mixed grain
Oats
Jao
Berra
Jawi
Sep.
O~t.
..
Dec. Feb.
1'25
1'25
8
12j
11
6/4
414
414
5 Gram €hana, Chhola Sep. Oct. April 1'0 11 3/4
6
7
Lentil
peas
Nakhud
Masur
Matar
Oct. Nov.
Nov.
. April
1'0
1'75
8
8
3/4
S/8
8 Poppy Post April 20 8 6/4
SauDf ,~U Nov.--::) March
9 Kiseen April 1',5 8 6/4
10 Coriander Dhania S~p_ April Mav 1-75 8 614
11 Safflower Kasumba Oct. Jan. April 1 75 8 6/4
12 Rape Sarson March Ie" 8 4/4
13 Linseed Flax Alsi April 1'5 8 4/4
14 Linseed Alsi 1 '5 8 4{4
15 Rocket Taramira 8 4/4
16 Carrots Gajar S~p. NJ~v. Feb. 8 5/8
17 Vegetables Tarkari Aug. Oct. April 1'5 5/8 8
18 Lucerne Grass Lusan Ghas Oct. Dec. June 2/SF
N.B.-Crops gra7.ed are charged follider rate (F) and crops used as green manure are not ckarged at all.
 61

t 2 3 4 5 6 7 S
iU IU • 0)'
S. Name _of crop. Sowing time, Harvesting time. td
.....
b.OuOOl-oi ~ .S t q)
t -g~ ~ ..... "'-' ...<J
t1l::;"tlU
No. 0:; (OJ Q) •
From T()
English Hilldustani F.rom To Q > ....",-,.~~""'-' ~e~"
<
. ----~---~-------. ----~- . .-----~.----

19 Maina Oct. Nov. Feb. April 2/8F

20
21
Millet
Person Clover /
Senji
Shaftal
"_ F~b.
March
May
2/8F
2j8F
22 Radish Mllii S::;p. T N~v.~:) Nov. FeIJru.ac, 8 5/8
23
24
25
Mustard
Fodder
Egyptian clover
Rai
.0' 1-)
Methra. /J
Berseem
f.
Oct.
Sep. ..
Oct.
March
Feb.
Nov.
April
May
3 4/4
2/8
.2/8
26 Garlic Lehsaa &t April 8 ;6/4
27 Cumin Zira Sep. Nov. April May 8 -6/4
28 Chico.ry.
.. t,"
Kasni
,',
Zaid Rabi Crops.
-8 0/4

(
, ' I
I Maize ,'/-
Makki :7 f- ?-
,I
Ti
....
March April May Jnne 1'0 '2/SF
2 Millet Chalri July 2/8F
3 Yo,) Swank ... \. l/.».;: ....
j~ly Aug. 2/8'F
4 Common t ~1 China :.J. ' ::' May June '2/8F
Millet ~JL J ,: i"1
Tambak~'
;~

, ~l!ld
I

5 Tobacco
,;"
" nT Feb. r March June 2'0 8 0/4
6 Melons Kharbuza May June l'G 3 4/12
7 Water MelollS:' Tarbuz M~y June july I'G 8 4/12
8 Onions ,) Piaz : I
Jan Feb. May juae 8 5/8
9 Potatoes )'; Alu ,j
L • Feb. 2'0 8 5/$
10 Vegetables
ii.1 Tarkari ,e! :J:!J 1 ' ~;< i March Aug. 8 5/$
II Cucumb~r Khira June 8
I';'; ~f !(,t{J! ~ : 'I~
lIt
N.B.-Crops g~ are charge~ f!»4!l~,tI,te {FJ ud crops used as green manure are not charged at all.
 PART II
CANAL IRRIGATION
Chapter III
HYDRAULICS AND CONTROL OF LARGE RIVERS
1. Introductory.
India posses a high mounain range called the Himalayas forming the north boundry ,
throughout its vast Sub-coIltinent separating the North Indian plains from Tibet (China"
Russia, Caucassia and Afghanistan. Himalayas is a Sanskrit word meaning Home of Snow. In
the Himalayas, there is pcrpttual snow which Leds the great rivers of Northern India by
melting. The rivers get their peak discharge due to snow melting in the months of April and
May, which period synchronizes with the time of maturing of the raM (winten crops and the I

so~ing of kharif (summer) crops. The rai.ctfall in Northern India is concentrated in its intensi-
ty and magnitude in the months of July and Augu st to the extent of about three fourth of
the annual average rainfall. The rest of the rainfall occurs in the months of Deceml:er and
January when the river supplieS fall down due to decrease of the Snow melting in the low
temperatures during these cold winter month. On account d the heavy rainfall in the month of
July and August, the rivers of Northern India rise in high floods which are many times more
than the normal supplies in them throughout the year. The ratio of flood discharge is about
SO to 100 times the normal winter supplies of the rivers. All the rivers of Northern India
emanating from the Himalayas carry in their waters lot of suspended staff in the form of fine
silt and clay, when flowing through the vast plains after leaving the hills. All of them form
n~w lands and erode some whil.~ passing through the plains. They are, therefcre, called
alluvial rivers. The hydraulics of these rivers is very complex and peculiar due to the flood
discharges being very high as compared with the normal discharges requiring flood scctiom
vastly out of proportion to the sections req uired for winter supplies.
2. Major divisions of river channel.
The river channels of the large rivers of U~f! Indo-Gangetic plainf are divided inhl
four major divisions ;-
(a) Mountain stage.

The river flows through the hilly gorges with _rocky bed. Water is clear. The bed
slopes are very sttep from 1 in 100 to 1 in 500 wi:h occasional rapids and adrupt falls. In
hills it is mostly snow water.
(b) Bo.tlder stage.

When the river leaves th~ mountains it flows through the sub-mountainous tracts with-
boulders beds having steed gradient of the order of 1 in SOO to 1 in 1000. Water is generally
clear except in the rainy season. [he section is usually well defined. The boulders of the
round splunial shape vary in size from adout 1 foot diameter in the' hills to about t to I" siz€
H

shingle or pajri as the river entcr:> the plains.

(e) Trough stage or rlYers in plians.

Th~ r~ver sectio-'_ls are well de~n~d chan~els with high banks. Water ()arries large
amount of sIlt III suspenslO~. The ~elocltles are hIgh enough to erode the bed causing d€cp
~cours an.d to damage the SIdes ca~sl-'_lg serious erosions. The proportion of suspended mathn
III water m the form. of clay and sIlt It usually as high as 1/300 to 1/1000. The bed slope is
pretty steep from 1 III 1000 to 1 in 2000.
 (d\ Dcltaic Stage.
. . The ri:,er in .this stag~. is near its .out!all into the ;;ea. The river fans out in many
tnburanes and IS su bJect to spIlls. VelocIty IS low. Slope IS generally very flat from 1 in 5,000
to 1 in 10,000. On account of the low velocitv the river red is generally silting and rising.
The water surface is also rising, eventually spilling over and forming new channels. The process
of land formations is shown in Fig. 1
lnL7A'r'C ""'''fAYtDN The cross section of tht' country would show
tte river running at the highest point and the land
falling on each siee of it at a very gentle S~O;1f. Fig. 1.
In course of time the river would break t~rou:(l one of
its banks and assume a new course, when the sarr:6
process of the gradual rise in its banks would take
place. A valley would thus be for'lled retween the
Fig. 1. two courses, and this itself would ev.!ntually be
obliterated by the river breaking into it :md similarly silting up the area. Thi" action of
gradual rise in the country by successive d !posits of silt along different courses of the ri vcr
occurs chiefly in its deltaic portions. but if may also continue higher u') when the river breaks
through one of its banks and take, an entirely ne ,v course, The silts is of a very fine. unsta')le
oharacter, and thus the river flows in a c:lannel. the bed anr\ sides of whic'l are quite una'lie
to resist any violent attack by t1.e r:ver. Where the river discharge is fairly constant, a rrore
or less permanant regime might eventually be establis"ed, l'ut this is far fro'll being the
case, for during the flood season it may 1::e fifty to one hundred times the minimum winter
capacity.
In a Maximum flood the width may be 10 to 20 times and the velocity and silt in
sl1spension several times what they are during minimum low water. The great fluctuati( n;
in discharge are thus continually changing the conditions, and it 18 the fioods which chiefly
affect the regime of a river. .
3. Torrents and Streams. : ,;
(a) The presence of a delta close to its mountain source may I::e held as a feature
distinguishing a torrent from a river or stream. The course of a torrent may, like that of a
river, be separated into four main divisions, viz., the hill portion dehouching in a ravine width,
for a short distance below, a deeply cut head: this rapidly merges into a delta. or broad
spreading fan of sand and hill de:)ris, tailing as a rule into a swamp or broarl marsh formed at
the finer clay particles brought down by the torrent. From this marsh small drainage~
break out in places to unite further down into a defined channel or stream, Fig. 2. ,,',
These physical features are marked by on~ peculiarity, \.; . the complete a')sence of

..
. visible water in all the divisions except for a short tim6 during heavy rains.

......
Even the final
nallah rarely contains any wate}
unless it derives it from S')mf
spring unconnected on the sur-
face with the torrent. The
floods which pass down these
torrents are short lived, though
extremely violent, and the slope
on the fan generally varies from
25 to 15 feet per mile. Two
seccessive floods rarely or neveJ
occupy exactly tte same posi·
tion on the fan. for coming
down loaded with silt the~~
spread out in dif'erent direc·
tions, reing constantly forced
asife by their own deposites. 11
will l:e easily understood tlea'
Fig. 2.
 64
the surface of the water in a torrent is not smooth like a river flowing with an established
section, but a series of waves one following the other at short intervals. It is now generally
admitted that down to a certain limit, the bed in the fan moves with the torrent. and that
the snrface of tte fan after a flood does not by any means represent the actual bed during
the flood.
Ib) Streams.
Streams derive their supplies from hill springs, sometimes from lakes which are simply
reservoirs filled by rainfall~-in the last case the supply must be considered as extremely
precarious. In the passage from the foot of the hill to the low lands, a considerable proportion
of the supply is absor;:ed in the soil and many of the smaller streams entirely disappear a
short distance below their sources. In the larger streams the water thus lost by percolaticn
probably reappears to a certain extent lower down a channel, but the water from the small
streams simply goes to feed the great su':J-soil reservior. Canal can be marle most profitable
from small hill streams and springs with their headworks in the gorges, but the channels have
either to ce lined with masonry to pr2vent excessive loss from percolation and erosion of the
bed from the great velocity engendered by the high slop2s or to be given comparatively .flat
slopes and numerous falls, including depJ1>its which help to minimise the loss from percolatlOn.
!.Of spring streams there are two distinct varieties viz. the small river with distinctly marked
channels emanating from hills, dry or nearly so near the hills, but gaining steady increments
: ,to its supply, as it passes through the heavy clay lands below; and the local stream rising from
a swamp or nest of dry channels, which gradually accumulates more and more water in its
-down-ward course. A river of the first variety has generally a tortuous course, a sandy bed,
and is subject to severe floods. Its channel sometimes occupies a local watershed, and may
indeed be said to be slightly deltaic in its nature, and it differs from the large rivers in not poss-
essing-a valley-this renders tce floods more violent in their results and causes them frequently to
spread over the surface of the country_ and to eventually find their way into the adjoining rivers
on either side, thereby often cutting up the doabs by cross drainages.
4. Molley's River Regime Theory.
- Fig. 3 on plan and longitUdinal section illustrates the theory of R.A. Molloy, late
Exp.ctive Engineer, Punjab Irrigation Department, published in Technical Paper No: 1 t8 of
Government of India, Central Printing' Office. Simla.
- (a) Taking the plain first it will be seen that the river bends alternately from one
bank to another. At the crest of each bend is deep water and in the lengths where river changes
over from one side to the other, shallow 'crossings' or 'bars' are fonned by deposition of silt in the
bed.
(b) The bars are produced by the action of the side channels. The side channel has the
"off-take" and the outfall as shown in the plan. The su btraction of discharge into side channel
at the off-takes results in the reduction of velocity in the main stream which results in the
deposit of silt forming 'bars.'
(c) Eetween the side channel and the main river are "islands." These islands are
formed when fl)Od subsides and the river divides itself into main and side channels.
(d) Levels of the river.
The longitudinal section shows the conditions of surface levels. Mean gradients are
shown by dotted lines and actual ones are shown by firm lilles.
(1) Taking low water L. Section, it will te seen that owing to obstruction offered by,
'bars' actual gradient compared to the mean raised above the bars by an amount indicated -by
"e" and is depressed by "d" to gain sufficient velocity to pass the bars.
(ii) In floods, the conditions shown in the top L. Section are brought about when
flood is rising, the bars are silting up and the discharge of the side channel at the off-takes
tends to inc~se. This tendency is resisted in two ways as the floods continue rising.
 65

(a) Rise of + k and drop of -h at the off-takes resulting in reduced gradient of the side
channel.
(b) Back flow at the outfall.
" This process will continue till 'he flood covering the full section IS reached and the
e onditions sketched in tte L. Section Fig. 3 are bronght arJout.
I . ~

MOLLOY~ RIVER REGIME THEORY

~
Q
t:
"
lot
III CI:;
..;,
"'If
~
...
...
~
0
fie
.., C"l
';(
i! ':!
~
b.O
1L:
I
~ Q
::a
_,
~
ts
;r:
""
~
~
).
t-

(e) Defilnations.
- h=the lowering of surface d the main strea"u due to discharge having been drawn
oft by a side chan!1el.
 66

+k=the corresponding rise in surface level at the outfall of a side channel where' it
rejoins main stream.
+e=the heading up above the off-take (and the adjoining bar below it) of a side
channel the discharge of which is reduced or stopped altogether.
-d=the lowering of water surface at the outfall of a side channel which becomes dry
or is closed.
-l=the general lowering in levels of the surface and ted caused by a cut-off.
zero=the condition between a general lowering and a general raising.
+l=the general raising of level of the surface and the bed due to an increase of the
tortuosity of the river.
General remarl,s on Molloy's Theory.
Molloy in his theory explained all variations in the behaviour of rivers by change in
discharge, but there may be other factors such as change of curvature and change in gradients
also contributing to changes in river regime. He states in para 7 Paper 118, Governmen
of India
"Such inequalities in the range of rise and faU are only to be accounted for by grea t
variation and disturbance of momentum" Usually the implied postulate of all river formulae
is absolutely uniform in condition, section, gradient and.velocity. But the actual characteristic
of a reach of such a river "(an alluvial one)" is irregular alterations of all the conditions and
consequent changes of momentum. The term channel does not apply to such a river; it is
rather a series of pools, or compartments, of which the sides are the high-level islands, and the
bar crossings are moveable, transverse panels very difficult to move".
This description of river by an Engineer who studied the rivers as life time problems
clearly shows that rivers are not channels like the irrigation channels as Mr. Lacey considers
them in his theory of channels ~n alluvium as described in 'Chapter VI of this part.
;). Other River Regime Theories.
(a) Oldham's River Regime Theory.
Mr. Oldharq following the theory of':M. Dausse. French Engineer, enunciates his theory
hy four principles. .
1. Every stream tends to a condition of equilibf.illm in which the velocity developed
s just sufficient to enable the stream to transport its s0lid butden. It the velocity is in excess,
the stream will cut down its bed thus reducing its gradient, the veloCity and the silt transporting
power. If the velocity is in deficit, the silt will be deposited, thus reducing the section, and
;llcreasing the velocity required to transport silt.
2. Every stream is alternately collected into a single well defined channel called a
. 'reach" or spread forming a "fan." Fig 4. These fans are seen in river bed when it dries
(I p in winter.
3. The gradient of a
stream is not a uniform one.
The bed will be flatter in the
"reaches" and will be steeper
in "fans" relative to the mean
gradient. Fans are shallow and
th us with reducing velocity.
4. Both "reach" and "fan"
work gradually upstream.
"Reach" encroaches the upper
'fan" by erosion and is encro-
ached upon by the lower one
by silting.
Fig. 4 (b) Mr. 'Ellis' Tributary
Theory (Vol.CXII The Engineer December 15,1911, uvtlines that a rivers bends to the infalls 01
the tributaries owi~g t~ its b.ank being there blank, and thus weak, and the tributary forming
a channel on that slde. fo bnng tae flver under control he says some points should be firmly fixed,
and no tributary stream should be allowed to enter nor irrigation canals to take oft from
the river except at fixed points, He observes that the conditions usually oI)tained are:-
.(i) No stream coming in on either si'ae-a straight, well defined channel.
(ii) Streams coming in on one side-a well defined channel on the other side only.
(iii) Streams coming in on both sides an ill-defined channel and a disposition to form
a shoal in midstream,
(c) Another theory is that rivers running north or south attack their west bank
owing to the rotation of the earth Dringi:lg those banks against the current. This assumes that
the water of the' rivers does not acquire the same rotatory motion as does the lands, which
assumption does not appear te:la·)le. The destruction of Dera Ghazi Khan on the west bank
of the Indus was said by some to be due to this cause; Mr. Molloy held with much greater reason
that the westerly trend of that river there was owing to its deHectioL by small reclamation banks
on the eastern side which were maintained year after year in the interests of cultivation.
6. Meandering and Avulslons of Rivers.
(a) An extract from River Training and Control by Sir Francis ]. E. Spring C. 1. E.
Technical Paper No. 153 of 1923 of the Governmeut of India is given 'relow which descrites
aptly the process of meandering of large rivers of the Indo-Gangetic plains.
"The manner in which rivers of Northern India fined it necessary to meander more
and more nearer they get towards the sea-in other words the lighter and less coherent becomes
the sand composing their beds, may be illustrated by a rough measurement of the Indus between
Kalabagh, where it first becomes alluvial. and the sea. In the following table, column (A) gives
rough measurements along'the centre line of the river, while column B gives more accurate
measurements round the 'cends, Both columns have been scaled off a 32.miles to an inch map and
doubtless if column B had been scaled off a one mile to an inch map,the proportionate increase
of length due to the river's windings would have been considerably greater than shown here :-
A B Percentage
Successive 100 mile lengths. Lenath measured Length measured of
fairly direct. round bends. meandering.

1st 100 miles, beginning at the sea 72 100 39

2nd do 75 100 33
3rd do 72 100 39
4th do [ ! 69 100 45
5th do 82 100 22
6th do 82 100 22
7th do 93 100 7
8th do 98 100 2
9th 100 miles to near Kalabagh 97 100 3
--------~----------------------------------------------------------
The meandering of the Ganges are less regular, as may be seen from the following statement, which,
however, as showing irregular distances, is perhaps not very convincing : -

A B Percentage
Successive Sections, Length measured Length measured of
fairly direct, round bends. meandering.

From the sea to Sara 185 200 8

" ~'Sara Ben ares 425 540 27
" . . Benares Allahabad 85 130 51
" ,. Allahabad Cawnpore 110 135 23
" .. Cawnpore Rajghat 175 195 11
" .JRajgbat " Garhmuktesar 37 42 14
., Garhmuktesar .. BaJawali 65 73 12
 If lines be drawn conneting tt.e limits of the bends of such rivers, it will be seen
that while the meanderings of some of them do not extend outside the breath of one mile, those
of others, e, g. the Indus at 700 and the Ganges at 200 miles from the sea, extend to a breadth
of 10 or 12 miies. outside these limits are slightly higher lands, but still for t1:le most parts
alluvial, into which the river has not as a fact cut. But the fact of the river not having cut
further to right and left may generally be ascrie:ed less to the mat3rial of th~ untoched land
being of a more holding nature than that in the river valley, than to the fact that for each
part of each river ther2 exists some relationship between Clhesive properties of soil and velocity
as qualified by cross sectional area, such that no more than a certain extra percentage of length
of channel is needed by the river to prevent its cutting still further into the up lands.
But even when such rivers are being described have meander cd as for they ever will
within the limits of the shallow vallevs which they have scored out for themselves in the
alluvium, they are no nearer than ever to the attainment of a stata of permanency or
equilibrium within those limits. Indeed in the course of years, there is scarcely an acre whithin
the valley limits which will not sooner or later be eroded quite away and in turn reformed.
Action of this sort goes on for the most part during the flood tim%, and much of it beneath
the .surface of the silt-laden water. What happens is somewhat as follows:-When the flood
~eg.ms to rise with the melting of the Himalayan snows in June, it finds (a) ready pr2pared for
~t smce the previous year. a more or less deep channel. As it rises higher tb) it spreads'
Itself over the low sandy s[Jits and fills short-cut channels. At its half fioods stage (c) it finds itself
topping extensive areas covered with he3.vy reeds, grass and brushwood. Later again' (d) if it shou-
~d happen to rise somew:1at hig'1er t'1a:t that f.":e'1t years, it t:ns cultivd.ted area, and drive3 the
mhabitants of the more or less temporary villages, to whom the cultivation is due, to take shelter
on trees, boats or house tops. This last state of things (d) is seldom of mlre than a few days,
duration, but the canditions descri'oed at (b) may te of six months and the intermediat stage.
(e) of perhaps three or four weeks' duration.
Now, owing to the comparatively brief time during which the flood tops the higher levels
where cultivation is being carried on, it seldom has time to do any surface erosion at such places
on the contrary, the tendency is to raise such places by the deposition of silt, owing to the
great reduction of velocity due to the lessened depth of water and the restrining influence of
the vegetation. Moreover, a skin of vegetation. though of no avail against the edge, caving
Jr erosion, is a very effective preventive of surfac," erosion. Therefore so far as mere superficial
a~tion is concerned, the tendency i:.-, for the higher plac3s, below highest flood !ev21 to grow
hlgher still to the great profit of the cultivators, to whom a f2w day of exposure and semi-
starvation are of slight iinportance in comparison with the incre" sed fertility of their latelv
flooded fields. '
But, flowing as it does over a great breath of the valley, swiftly in the deeper parts
where ~ast season's chal!gel is, and more slowly in the shalbwer parts, the river finds itself tak-
ing shallow short-cuts across gre3.t be;11s, at a C) n)1rativ0iy l).v velo:!ty a Nm~ t) the shallow-
ne:iS, and then cataract~ng down into the m3.in stream at a very high local velocity at the down-
'stream end of <each short-cut. It ofte:1 happ::ns, that if the bend is a very big one, the s\ort-cut
channel has not time enough to erode its bed and banks to any gre3.t ext:nt; ani 1:1 this case 01
the fall of the river, the results of the action will be exhioited in the form of an a')ortive chan l"ls,
which, had the flood only lasted a few days or hours longer, might have esta )lished itself right
across the bend. But should the fiood stand up bng enongh fJr the short-cut channel to erod~
itself adequately, we find the folbwing stat~ of things, viz (a) the main river mnning round a
great, say 10 miles long, with velocity due to a fall, say, one f,)ot per mile measured reund the
bend, and with a favouraDle stream cross section; and (D) a comparatively small side shoot with
the same fall, say 10 feet, in perhaps 3 miles inst::ad of ten, but handicaDped by its compara-
tively unfavouraDle cross section, If the net result of the favourable fall and unfavoura'olE!
section should be that the velocity in the narrow she>rt-cut is e fective in cutting its way so as
to rapidly widen a!1d eeerer. tte small siee channel or chute, with the conseqUEnce that, in a few
days or hours-the author has seen such a change occuring on a very large scale in 2.1, hours-
the small side channel will itself have constituted the main river, widening itself by caving, and
the 1'ng bend will have got silted up. A summary of Mr. R. A. Molloy's aCCOU.1t of the causes
of cut-offs, in rivers of the Indus class, has already been given in this chapter.
 69
Now this short-cutting actio::! is constantly going on at one or the other part of the
river. Sometimes it fails and sometil1es it succeeds; but whenever it succeeds; its effect on the
river for many miles up and dowlJ-stream may easily be imagined. For miles above and below
the recently established short-cut, there follows a period of utter lawlessness; banks caving,
channel silting and new channels forming. u,ltil at length. proba',)ly with the advent of the cold
weather 10"1 warer conditi.ons. tb.<>re IS a truc~ for a few mO.:ltbs. The next rise of flood finds
things still a bit chaotic, a:1d velocity in places quite in excess of what the soil can stand upto; and
the result is perhaps two or three more months of lawlessness, untill at length, by increasing
its tortuosi.ty, the stretch of 20 to 30 mil2s of river eiiected by the one short-cut settles d:Jwn
into a st;; te are armed neutrillity, ready to exite itself afresh. as badly as ever, when the next
short-cut happens to succeed in estaolishing itself. The action which the auther has attempted
to describe is that which Northern Indian river Engineers have Leen in the habit of calling
avulsion; but the author prefers to use the term CUT-OFF, as expressing better than avulsion
the action in question."
7. Extent of meanu.t>ring 01 Rivers.
An idea of the extent of meandering or the alluvial rivers can be had from tables
in the last paragraph An attempt has been made in Publication No.4 (1939-40) of the Central
Irrigation and Hydro-dynamic Res?arch Station Poona, India, to work out the formulae con-
necting the dimensions in a meander belt of a river as sketched below in Fig. 5.
W =width of river. ·.n~_ .~": <l':n:! . ·; '''.1 ..... " . q ~~l"tr,.~ ..... t; ... S
~~1Jj. ,',\!i .'.'':~' '~·rY'. ,~,j'J, v,}tli [' :; '~J
Q=Actual discharge of the river.
r~~~f1"" ~1i~_:
,q~~Jz:):' pi
Mb =Meander belt. ;.
oj; MI =Distance between successive meadtR~ on the axis of the river.
R=Radius of the meandering land.
The incised river data gave ;JC -.lj/'" d:lfn~; it'" (, l'-.U.I,,',l ~ :,!
Lt~i':~~:r~;-i-r1 :}d~ 1<) ~fIHJr.H :td; ~ : ! : '
W=2'48 Q! ",lil; ;.;vtllJ:J ~t1J hril; :~'JU;d :,iJj .. !,

Compared with the Lacey relationship .\ 'C" ;:~,'",·r !J;',(:,,",'

.1 .
P=2'67 Q2 for regIme channels.
EXTENT fJF .ANIJER1Nti 0' /lIVER The data appear to comfirm
that Mb MI and W (i.e P. wetted
f---- -- --/If, -.-.--.----j
perimeter) all vary roughly as Qt
(in other words, are scalar) and tLat
the meander length is of the
order of lOP or ronghly 25
1. 1.
Q2 t030 Q2
Radius of curvature was worked
. 'out from the ~ame data as shown
below~-

.j Fig 'N'o. 5. I ~

(1) For incised rivers: (High flood section cut below N.S. and not liable to
overflOWS)
 MI =25-4 V Q; Mb =56-4 vQ; and W =2'48 V Q from which R= 14·23 viT,' say
14 v
Q~
(2) For rivers in flood in plains: (River likely to flood the country on both sides in .
high floods).
Mb =29'6 vQ Mb =84'7 vQ; and W=4'88 vQ~ from which R=20'64 vQ. say
20v'g.
8. River Erosion.
Erosion is defined to be the destructive eFfect caused by the river's lateral action on
riparian lands and works, while scour is that due to the river's vertical action on its bed.
Erosion may be said to be of three kinds :-
(i) Normal. This is caused by straight, unobstructed currents parallel to the river's
bank. Such erosion proceeds gently and regularly, and may deepen the river
channel next the bank from 30 to 40 feet below low water level.
(ii) Abnormal.-It is caused by straight-flowing currents impinging at an angle
against the River bank. Such erosion produces caving, or embayment, of the
bank, and may I:;e rapid; it may deepen the river channel next the bank from
60 to 100 feet below low water level.
(iii) Extraordinary.-This is caused by rotatory swirls set up where the onflowing
currents are bounded i:1y dead or non-flowing water. Such erosion is extremely
destructive locally, and may deepen the river channel where it occurs from 80 to 120
feet I:;elow low water.
The principal factors on which erosion depends are; the velocity of the river, the
duration of its flow, the nature of the material of the bed and banks, the angle which the
current makes with the bank and the curvature and general alignment of the bank,
Erosion is due to the local lowering of the surface and the bed. A rise of surface
level often reduces erosion, and, when this is combined with the reduction of velocity, erosion
must stop.
9. River Scour.
Scour is defined to be the destructive effect due to the river's vertical action deepen-
ing its bed. Scour depends diefly upon velocity but also greatly on the amount of silt carried
by the river, if that is in excess, there will be silting; if that is in deic'cit, there will be scour;
and if there i~ only just the amount of matter suspended in the water which its velocity will
enable it to carry, there will be neither raising nor lowering of the l::ed. Thus scour is
worst during a falling river, because the water then contains a small amount of silt; during
a rising river the water is charged with silt, and this tends to fill up the hollow in the bed.
DEep channels are likely to become deeper owing to the concentration of flow down
them, especially when the water is not fully charged with silt.
When a flood is at its maximum it has already deposit~d som~ of its silt 'upstream
and will therefore tend to scour; thus a river may be deepened, although both the flood level
and the velocity have decreased.
An unerodable covering protets the bed from scour, thus a rl!ined masonry structure
dumped stones and refuse, or matted vegetation, will tend to prevent scour under them.
10. River Swirls.
A swirl in a channel is formed by a formed moving current by the side of still water
which thus gives the flow a circular motion. An obstruction to the flow in river, such
as an artifical spur, will produce dead water and a fast river stream passing near it shall
cause formation of swirls both near the upstream and the downstream ends. The
 71

obstruction may be in the form of an ol~mas(mry bar in the bed, or stem of a tree sticking
in bed and causing dead water and the c~nsequent swirls : -
(i) The reason why swirls cause deep scour is that they throw off to their outer
circumference the silt that was suspended in the water or scoured from the bed and
fresh silt can enter the swirl area from caving banks.
Ui) Swirls have dead water at their centres which may cause silt to deposite there
when they occur near a caving bank. silt falls in on the land side, thus raising the
bed there; the deepest water is always en the outside of a swirl.
(iii) Tne destructiveness of a s~irl depends upon the velocity of the current, the
area of the still warer, the size and nature of the sand of the river bed, and the
duration of the flow.
(iv) Swirls are the most destructive form of erosion, and protective works should

11. Silt in River Water. -

therefore be designed so as to avoid them and certainly not to produce them.

'There is a vast different cetween the sands of various rivers and even between the sanci,
of various localities on the same river. People speak of the sand of the river bed as if all river
sands were alike, whereas as a fact, there is as much different tetween sand and sand. as there is
between, say, pumpkin and a potato, or between road metal and powdered wmdow glass. The
shape of the grains of sand bas a good deal more to do with their transportability than their
size or even their specific gravity.
- In the boulder stage of river, the rolling mat,rial near the bed may te size 6" to 12"
diameter, and the suspended sUJstance from 1" diameter to the road metal size. The suspended
SU9stance called bajri or grit does actually travel always above the bed like the silt of the canals,
or the suspended sand in the trough stage of the river, but it strikes against the ted at some angle
and is again shot up for some distance.
Further down along the great rivers of Northern India the rolling substance is grit
or bajri and the suspended substance is the coarsest sand. In the plain in the· trough stage
th rolling substance is coarse sand while the suspended substance it silt and clay.
It is only in the deltaic stage of a river that the ted materiel and the suspended
substance are of the same nature and consist of silt and clay, It is only in this stage that a river
could be said to re flowing in self-silted aluvium but in this stage it has no definite channel. It
is always discarding an old course, forming new channels.
It is usual to classify both the bed silt and the suspended silt by mechanical analysis
using standard sieves. The classification goes according to the mean diameter in millimeters of
the particles. The C~ntral Board of Irrigation, India (C. B. 1.) and the American practice is
given below:-
Classification 01 Mean diameter in M. M..
Class of Silt. Central Board of Irrigation India. American,

Colloids Below '001

Clay () to 11256 0'001 to 'OOS
Very fine silt 1/256 to 11128
Fine silt 1/128 to 1/64
Medium silt 1;64 to 1132 (),(lOS to 05 (silt.!
Coarse 1/32 to 1/16
Verv fine sand 1116 to Ij8 0"05 to '1
Fin~ sand 118 to 1;6 '1 to '25
Medium fine sand 1/6 to 1/4
Madium sand 1/4 to 1/2 ·25 to '5 (sandl
Coarse sand 1/2 to 1 '5 to 1 '0
Verv coarse sand 'I to 2 1 .to 2
Gravel 2 to 4
 72

The specific gravity of sand ar;d grit in a river varies .very little. from 2'63 to 2'74, but
it seems to have no bearing on the subject of sand transportatlOn III the nver waters. The angle
of repose of the bed sand of the Northern India rivers does not vary much. It is about 31 0 to 37 0 ,
and it does not appEar to have any law connecting it with the fineness and the coarseness of the
sand. There is also very little variation in the pore space of the silt and sand available in rivers.
It is about 40 to 45% of the volume irrespective of the diameter of the sucstance However the
coarser and heavier the sand, the smaller the amount of scour and the comequent erosion.
12. Vellocities in the Punjab Kvres.
Velocity of a river varies with the gradient, the nature of the cross sectional area
(H.M.D.) and the character of the ced and banks (N, the coelficient of rugosity). Thp velocity
depends more on the depth than on the gradient. The prevailing velocities of the winter
supplies in rivers may be from 0'1 to 5 feet/second and in floods from 15 to 30 feet per second.
As greater scour follows inc~eased velocity:. th~ tenden~y is for th: deeper parts of the section.
to tecome deeper still. A nver has to adjust ItS' velocIty to what ItS bed and banks can stand
by changing its section.
The velocity of a falling river is greater than that of a rising river. In the former case
the channel is draining away, while in the latter case the flood as it rises has to fill the channel
pools and its res~rvi?r area. When t.he v~loci~y is retarded, ~ilt ~s drope?, and :vhen the velocity
is increased, silt IS pIcked up. A fallmg nver IS most destructive m scounng actwn.
A deep channels scoured out during the flood season may silt up ·in the succeeding low-
water season, as its section will be ..too great for the reduced velocity then. In the following
flood season, the silting will at first reduce the velocity below what it was in the previous one;
but if the floods continue sufficiently long and are sufficiently instance, they will scour away the
deposite and re-establish the former deep channel. If, however, for any reason the floods are
small and last only for a short time, they may not have sufficient power to scour, and the silting
up may remain permanently and thus affect the flood regime of the river.. "iv.c' I:.' 'J Hi ,
13. Permanent ReClamation or Protection Works. ~ "t,

(a) Reclamation Works.

The object of reclamation and protection works is three fold: : .OrH 10 ']1
fl ~)L_'hj l.i
(i) To remove sinuosity of a river to ensure a well defined cha_·!(;,i"i
(ii) To.reclaim land subject to floods for cultivation. ~ r ;r.;:

(iii) 1'0 protect valuable land property on the side of a river which is subject to
erosive action.
The river main stream wand~rs from side to side, and therefore the most important
principles for reclamations are:-
,
"Never throttle; always close and work with a full head" (Molloy's).
In other words, the river should not have its main channel narrowed.
The closing of scour side channel and the prevention of spill over fore shore should
be affected by what has been usually called a "reclamation bund." Reclamation bund unlike
the Guide Bunds are temporary works for local use. They also require more repairs.
,b) Design and construction of a Reclamation Bund.
The highest point. of a mi~stream island and one protected by vegetation shonld te
selected for an. abutment lilt? whIch the bund should "te well keyed. The bank of a bund
should be contmued some dIstance beyond the length to be protected. Where ridges and cross
channels are met, cross spurs 300' to 500' long should be constructed on the ridges both upstream
and downstream as shown in Fig 6.
 , ...
The cross spurs serve t.
'RECLANAT.1DN .UNO produce still water especially

~.. ,."".
C. CNAIINIL
1'1 ...."". 1""
."". .~N£1t4L PLAIf
upstream to induce silting
and to protect the main
bund. The best material to
It. 1110.£
construct the bund is
S ••- ..... pure sand and failing that
the soil of the locality. The
material should te obtained
some distance away from
the bund on the upstream
side.
The axis of the bund
should be preferably at
right angles to the riveLand
not less than 60 in any case.
0

A good key trench should

$~CT'DN AT •. he dug 6' to 8' wide down
f.,' .of to spring level. The bank
should te raised 4 ft. above
;" L) the flood level allowing
~:JrlO! .1<" (' :(L'.:!
about 2 ft. free board and
H' to 2' afflux head. The
section should 1::e raised
according to the required
strength. If necessary, an
escape channel should be
constructed .

.,~ ...~.~~~," To protect the bund

~
from wave wash pilchi ma-
ttress or grass fascine should
1"" ,•.1>,;<:.._ be provided. A roll of 6'
.ii"----~., . diameter will do.
Fig. 6 (a). "'
The reclamation should be recorded every month and 'contours plotted to watch effect
of Reclamation Bund.'
I :.~

(c) Spurs.
The earliest form of river control was the construction of spurs projecting from the
Lank into the current so as to deflect it from the places threatened by the attack.
Where the soil of the bed is firm, this form of protection answers very well. I.
loose soil there is danger of outflanking.
A spur is essentially a fixed obstruction. It is bound to cause swirls both upstream
and downstream near the ends. The swirls will cause deep scour and are very destructive.
The spurs on account of the formation of swirls near their nose are sometimes used
to induce a deep river channel on the side provided with spurs at regular intervals.
The protection of banks 'by spurs is only successful when the extreme upsteam
spur is well-founded in unerodable soil. The spurs yield only temporary relief and it is
preferable to resort to the construction of guide banks.
(d) Form of Spurs.
(i) An ordinary type of spur consists of branches of trees Or jungle and stone ia
alternate layers so that the spur is previous as well as capable of sinking
unevenly without breaking up. The stone being the permanent part, the more
the better. '
 74
(ii) If ston~ is not available at the site selected, and abutt:nent is rormed by digging
a trench which is staked and lined with brush wood mats all round. This is
fille i with bags of earth. The bags are well trodden or founded. The stakes
are hraced. The construction is extended into the current. The materials are
perishable and the life is not long
, (i:i) Planks are some times driven to form a continuous sheeting, instead of stakes or
brush wood.
(iv) In Egypt and the Punja'J spurs are so:ne times made entirely of stone. Permeable
spurs are considered to induce silting up by the slow flow through them. The
stone spurs are permeable.
The spnrs are usually constructed sloping downstream at an angle of 120·. A spur
of le~gth L normal to the bank protects a length of 3L upstream and 4L downstream•.'.
~e) Gro~nes.

Spur8 of permanent nature are called Groynes: Denehy Groyne as used at Norora
weir is a type of permanent spur,
The led of the spurs is
'uQ(iJ; in the form of T, Fig. 7
a ,out 400 feet long and
is connected to a guide
bank, the main ft.ood
embankment (Guide
• Bank) by means of an
earthen bund. The
Groyne is heavily faced
OE.N£H Y GROYNE wi th stone and is pro-
tected by reserve stone
piled on the spur at the
'I T end. The Groyne
most upstream should
be well protected or
finished with an impre-
gna'Dle head,
14. Temporary Recla-
mation Works
(a) Brownlow's
M .... ""'''.11 III,. ... ..;1
Weeds.
These consist of
!! ;" r series of lines, each
$'CTI(JN DN 6.F. c. fastened to an anchor
i;'r--$'...-t&l~1--'-' ,--.,,--,--,-.-;-..:.t-+.6+~ t lock upstream in the
i~de¥b : ,~w ~"1' river 1 ed. and suppro-
ted downstream by a
bum'. Trees and brush
Fig 7. wood are attached to
the lines, which by deadening the curre,',lt couse silt to be deposited. A spur of any d~sired
,length can be formed by increasing the num~er of the lines. The success of the weeds depends
upon the holding of the anchorages, but when once silting commences, they hold better.
(b) The Bengal "B.m:1'tal". This is a primitive ccmtrivance for causing minor and side
channels insilt carrying riv~rs to be silt~d up, so that the main channels may get a concentrated
'flow down them which will kee? the:n clear and cause them to scour This results in the
:,reclamation .f their foreshotes for cultivation. A bandhfll comists of a line Qf bam000s upto
50 or 60 feet in length and from 1 to 3 inches in diameter, worked during low water into
and across the bed of the channel to be closed and fastened together with a string at the top
to form a temporary fence. which is rendered sufficiently impermeabl~ by fixing mat screens to it.
 75
(c) Hurd\e D~kes. .'
\
Description of Hurdle Dykes. is available III ~ on Missis~ippi river )'-"Y'i':t';. F._
l}aw!!O~,. 1900. The arrangement is shown in Fig. 8. . t. ::/;. . . ,: i i f

"'.~,). 1U:Jf:'}) ,.:),

>f
,
. 1
I, I I
-rI !
,j'_L II I

~X· I

t- ~
f<

.. Di .. \U 11 t-'I""
~, ,! ~ It J

sqI~.e

lx:
'" II
~
'i " ~
~. ,
'l c ~ 'I :

>If .v.- or: I (~ ~ ~J

h_ ~'~.:, _, '"'_'_~ _ _'~_ It'',,~.-rl-~

, To p :event e~osi'an of the river ted from undermining the dyke, a flxibl~.· mattress is
formed on sloping ways, brought on to platform carried on special barges. and is lowered on·
to the bed, weighted with stone. It is manufactured fre m wattling brush uron poles, spaced
about 5 feet apart, and these are lapped together and fastened by spikes and wire to give the
required ",idth. The l:ursh wood matteress is spiked to the po!es at the -edge and elsewhere at
intervals to every third pole. The width of the mattress vades from 60 to 120 feet. according
to the depth of the water and the liability to scour. The hurdle consists of. a row or parallel
rows of pil1:e!rdriven thtogh the mattress at about trd of its width from its upstream edge, either
singly or in clumps, each consisting of three or four piles, the tops of which are drawn together
 76
i>y wire rope to form. a 50rt of pyramidal structure, the piles at river bed level being from
I to 10 feet apart. In the im[>roved from, the'curtain, or wattling row, is braced at an angle
of aOOllt 45" by vertical diagonal braces. which are heeled against a row of clumps and securely

--~7~; •••~~a"----
- ---------------
--~---------------

-----------~-~---

'Fig,9.

GlIIDt. BAIIIt pRo.T£eTION

/
ItD"~!"
" D. /ME,.." Ie...... seoll..
To THItlIf"••• 0' .,,,,.• • T.,.I..
A/U,f V . .~. nQ'" • 2·... T(If+fl').

.I".. A 0' A,..""" STOIf! AS LAIl'."~ D To

MUll T/f/C.,,,.,, 0' .""0If A. ""/0. "'.T.
/JIIl1I1I TIIICIo,.." ., ",."." _ .",..H.,.:
 77
fastened at top and bottom to the pil~s against which they abut. The piles vary in length
from 25 to 60 feet, and they are driven into the bed for about 15 feet, witt the largor end down,
by means of a hydraulic jet and weight of a:)out one ton. They are spaced so as to come to
about one pile to the lineal fo;)t of the dyke. The tops of the pil.os are kept about 20 feet
above low water, except that of one in e 1ch clump of wattling row, which is 25 feet above th at
level, in order to catch drift and prevent it from' passing the hurdle. The piles of that row
are connected along its upstream fdce by wattling of fine brush or by curtains mude of brush:
this hrms a permeable barrier, through which the silt-laden current passes, at a velocity so
much reduced that silt is deposited on the river bed upstream and downstream of it. The
great arlvantage is that they can reclaim a fore shore on .vhich permanent reclamation 'L unds
.. can ce designed,
(d) Cross Fixed Spurs.
They consist of double lines of stakes 3 to 5 feet apart with ropes passing r mnd.
Brush wood and twigs are filled between the stakes and are usualy loaded with .sand bags. They
are quite successful in the case of ~mall streams and torren :s.
15. Extent of narrowing by constracthn of Guide Banks.
The great rivers of Northern India, have' very high maximum flood discharges. It
often happens that in the course of a great river in the plains with surface width in maximum
floods from one to 10 miles an Engineer is confronted with the problem of constructing a bridge
fot a road or a weir for a canal. It is. not generally possi~;}e to fined near about the site a
rockey gorge section or aq increased section in plains, It will not only te exp<:nsive to span
the whole width, but also nnengineering on account of the uncertaip.ty of the deep scour in .a
very wide section. The section can safely be curtailed by, constructing artificial gorge \Yalls
known as guide banks. when t le river is sandy-cedded and capable of enlarging its section by
scouring its bed without any appreciable afflux. \
In the case of bridges such as railway bridges founded on well foundations, the factorS
governing the extent of narrowing have been summarised by (late Sir) Spring in 35 Technical
Paper No. 153 on River Training and control as gi ren bel>w :-, ..,..
"The experienced engineer will without any dout x:ecognise the wisdom of selecting
the place on the riv,er's cross section where the bridge shall be located, so that the area of
scouring out of bed section that must be done in the first rise of flood shall be minimum-other
things, of course, being equal. Also he will not fail to observe that, within lirr its, the greater
the narrowing, the deeder may be the scour, and therefore the mere expensive will Le tJte
stone protection of his guide banks. Thl:ls a practical limit for narrowing is soon reached. The
author is inclined to the opinion that in an ordinary sandy bedded river of the class under notice
with fall of 18" per mile and less, narrowing may in practice l:e limited to what will cause an
aU:over mean scour of from 8 to 16 feet 1 et~ ecn a:.;utments. The comparison between the
area before and the area improved by scour after contraction should be made with regard not
to area alone, but to discharge throug,h the old and the new areas, as gOVerned 1y the velocit~,
got at the varying depths of the successive com?artments of their CfQSS s2ctions."
The construction of the river ~ection is also done in the case of canal we'f3 bV the
of guide ban~s, but the problem is much .m?re co.mplicate~ tha~ that in the case of. bridges' with
open foundatIOns on account _of the p,erm,lsslble dlscnarge Intensity, the allowable ajiux and the
required crest level of the weIr to mallltalll a fixed pond level to feed the canals. The mathemat-
ics and calculations are given in Chapter XIII of this part to arrive at the permissible contra~­
tion at canal weirs.
16. Design of the Guide Banks (Bell') Bands).
A type plan and cross section of a guide I::aak IS shown m F'ig, 9, and 10
respectively.
.J~L ~~to~t and alignment of Guide Banks.
(i) Guide Banks are constructed in pairs symmetrical in plal'l.
 78
'(ii) It is essential that the design should be such t:lat no swirls are prodhted; The-
" pair of the guide banks should be ac; smooth as possible. 'There should te
no spurs projecting from the guide bank as the s~mrs produce swirls.
In order to exercise a gentle entrance for the river between the guid~' b(l_nks,
the embankments s~ould be curved inland near the ends. The widening shou,ld
not be excessive, otherwise it will lead to the formation of islands.
(iv) Care should be taken not to select a site for the bund where the rivH has a
tributaryor a sirfe channel flowing paralbl to the bund. These four factors
, .~ .
ensure proper alignment of the Lund.
(v) Sufficient still water area should be provided as in reclamation bund. This" can
be exercised by providing separate and additional bunds 'of reclarnation type
. upstream.
The essentials for a -guide bank are :-A proper alignment as symmetrical as
possible, a sufEcient still water area, an ample apron, impregnable head, and -a
large reserve of stones for r;epairs during its early years., '_ FiK. 9." is a typical plan
and Fig 10. is a typical section of a guide bank. fhe ~ uide banks are permane'nt
works unlike the reclamation bunds.
(b) Section of guide bank.
Top of the bank should not be less - than 10', and in the Railway practice 20 to 30
~eet to allow the working of the material trains, Side slopts should le 2 to 1 and free' board
. to 5 ft. In calculating the height, allowance should _be made for heading up, and also for
~e~tlement of the banks which may be 10,% of the height, The inside slope should Le protect;}d
with stone pitching. and the outside one by good pacca earth; and floating ropes with jungle
~~ <n.attress may be used to withstand w~ve wash. ' The earth. for the construction of the bund
(}ttld always be taken from the apron SIde and no borrOW-pIts shoUld be allowed on the back
water side as these may induce percolation.
(c) Slope of a guide bank.
" b (1) It is essential that the slope shold neither be undermined nor slipped. If it fails,
\ teach would occur as the earthen bank behind cannot resist the river action. Ample stone
~ \)ttld be provided in pitching on the slopes and reserve stone should be stacked on the back
: 0 replenish the weak points.
'-,

(2) Pitching should be atleast 3' higher than High Flood Level.
",.' (3) The pitching stone should be selected having a high Specific Gravity and weighing
t?ln. 100 to·l60 ths. In water it loosE'S its weight equal to the wight of the volume of water
'; l~Placed. The effective weight is, therefJre, reduced.,
th (4) The stone should l::e laid on 6" deep spawls to pr-otect the bank from wave wash
rOugh the intrestices in the stones,
I
vet,. (5) As, regards thic~ness of the stone req~ired, it w mId ~aturaIly incr, ase with the
inclcity of the flver. Mr. Sprtng used the followmg ta')le of thlckn~sses which apparently
ttdes the layer of spawls :-
---......
p'lI.I] "'Pe-r-m-il-e-o-f-r-iv-e-r-in-in-c-h-es-'' - - - - - - : - 3- 9 12 18_ 24
~ -----------,--------------------------~--~---------
:o>aQ'i...classification, Thickness of stone pitching in inches.
VeI'}"
------------------
Co
<\""e
coarse
-
16
22
19
25
22
28
25
31, ."
28
34
-F:Vleq i tun 28 31 :14 37 •. 40
m~
34 37 40 43 46
Very fine 40 43 ,;uO 46 49:,;:J 52
 79
(d) Apron of the Guide Bank.
Apron is the name given to the p:ut Df the pitc'1cl slo?~ onth~ river side of the Guide
Bank which is laid horizontal at the bN water levd. In ord ~r b ma,l{e the apron safe and
to arrange it efficiently, it i" n~ccess~ry to asc~rtain th~ d~pth of the d~epe.>t scour negl~ct~d
.swirls. The breadth is not to 'oe less than It/time.:; the deepest scourin Fig 10. The av~rag~ depth
.should be 11 times in gentle arid It tim~s in turDulwt river of the sbpe pitching thickness when
launched.
The apron is laid at low wa ter ltwd and it launches by it; own wdgh: when the scour
starts in floods as shown in Fig. 10. The average slope which a properly laid apron a5sume
when launched, is 2 to 1.
(e) Impregnable he!ld of the Guide Banks.
J. Mr. Spring's design of Impregnable head, .q~,: • ~.! I.,' l' 'tt) ~t ;;.;q
':..... j;:'?~l J ~~
1. The entry should be gentle.
2. It slould be curved hto reduce embayment.
The proper radlus to be alloweG. depends on the velocity ofthe river, Mr. Spring gives Ul.!
following ta:,le to design the radii:- ' . '
Fall p'·r mile of river in inches,

Probable ill'l.ximUill,
Sand classification.
abnormal scour,
... --~---~-~~-
I J~~~12
1
3 "6 9 I
".._..... .L......:
.L_~ __ _ l
--'---'j-_._._-,---------_.__._.
RadIUS of u.pstream Curved End in feet

Very coa::se , l)nder 20,feet, " : 200 ' ,2,50',! \300 II 350 1 40<}
,Over . . . . 250 310'1,375, 440·; 500
"
Ccarse 'eiider 30 feet " 300 '. 360' i 425 f 490 ' 550

Med;~m
Over ..
Under 40 feet
350
400
430
425
I 510 i
I 550 I 625 I
590 I 670
700

Fine
. Over ..
Under 50 feet
450
500 I550
590
I 650 '\ ',' 750
675 760 1
! 850
8S0
..
Very line
Over ..
Under 60 feet
6 0 0 , 725
sOO :,',,' 700
825
8i)0:
I ,,925
900 \ ':1-
1020
1000
Over 800 900 1000 1100 1200

l. The upstream curv~ should be carried 120 0 to 14 0. foun,d hI': "b~Gk to "CCUI<'
protection of the inl~nd side.
2. Pitched apron should be carried throughout the curve.
3. The curved portions should De of full height of the bank.
4. The pitching at the impregnable head s1.J.ou 1.d be 25% extra.
5. The width of the apron should be double fw deepest scour instead of II times.
17. Constru~tlon of the guide banks (B~ll's Bunds)
The construction of the guide bank should De commenced from the a'}utment, that is tht
weir, and should proceed regularly upstre'lIn~ If its conpletion cannot be achieved il1 one: season,
the reduced length c0mmenced should be fbished off and should have a tenporary imorcgna'Jlt
head formed at its upstream end, so that it may hold the river, and n0t De O,lt l1an~~ed oy aI,
e nbayment. In the su'.)sequent s 3aso.1. this head should be removed and the guide hank similarl~
continued or completed, To ensure rapid progr3s, proper aDd early supplies of stones should bt
al'ranged for, and it will facilitate w;)rk if these can be broug;1t to sit~ by trains.
The slope and the apron should be built between tc nplac~s 100 feet apart. In par'" 16(d:
the construction of the apron in the dry condition has been descriued, but this may not alway Ot
'possi'Jle, and part of it may unavoidably have to be formed in deep water. When thi~ is tht
 80

ca,e, the quantity of stone required for the sU'Jaqueous part or

,the apron should be' calculated;
this amount should be carefully and regularly dropped iI?-to the water from, boats and pr:ecautions
sho~ld be taken ,at fir,st not to extend the apron too far lest it be ,undermined and the material
wasted.
In the most difficult case, a length of the' entire guide bank, including the
em',ankment, may have to be formed across a deep-water area. If possible, that area should be
temporarily reclaimed by silting it up, 'DUt if this is not feasi'<,le, the apron should bt: formed in
advance and gradually built up auove the ,water level" after which .the superstructure can be
constructed in still water but this involves avery large expenditure of stone,
~he cost of guide banks for a Railway Bridge has varied from Rs. 7S to Rs. 184 per foot
run. Those for the Sara bridge across the lower Ganges, wher.e the deRth ,was ;excessive and the
sand of the river very fine, were estimated to cost Rs. 475 (£32) per foot run. The cost depends
greatly on the price of the stone, and that i" due principally to the l'~ngth of its lead and nature
of its haulage. The extent of the sectional area of the protection also affects tne cost.
18. (a) Retired Embanklllents.
, The conditions of :unembanked rivers are unsuited to the cultivltion of the adjoining cul-
turable lands on account of the fact that the areas remain under water throughout the flond season
i.e., July to Septem>er. The river Jhelum floods the lands on the left bank from Rasul to Shahpur
upto a distance of 30 miles in high floods exceeding 6 lac' cHsecs,' tlie maximum being about 9
Tacs. [his is a typical case wh ~re protective embankments would go a great length to save the
residential property and the destruction of crops in floods. But it canalised river with closely
constructed embankmenrs is prejudicial to the certain repatian' tights of the popUlation who
derive relatively greater benefits; as compared with the loss in floods, from the bumper crops in
the spilled areas for a couple of years following the floods. A system of' retired embankments
provide5 a solution of the problem. The embankment should enclose the area which has no other
means of irrigation except the river spills and cut off the area which is ensured of irrigation from
canals or wells. '

(b) Advantages of the retirement of embankments.

(i) The minimum interference with the natural operation of the river in raising the
country by slit deposit.
, (ii) Tt e maintenance of a large river reservoir capacity. ." .... ,
(iii) The filling of this reservoir reduces the' peak' of the high flood.
(iv) The emptying of this reservoir will maintain the river at good irrigating level for a
longer time.
(v) It may be practicable to utilize such a reservoir so as to give temporary storage of
water for canals when the, river: level falls. •
(vi) The, redw::tion of the high-flood level will preserve -the regime of the river, an i thus
it will diminish the erosion of good land and the throwing up of sterile
sand bal}ks. '
(vii) The longer life of the embankments will permit of their being constructed carefully
and slowly, well in advance )f the period when the river will approach them, and
thus will give time for their self-consolidation by settlement.
(c) Disadvantages of the retirement.
(i) In'.lreased cost.--Against this lUust be set the saving of replacement of embankments
'oy retired 106ps and the loss cif crops on eroded and inundated areas.
(ii) Increasei of open canal head.·_·As the embankment is retired from the river, so will
, th~ length of the open canal head between the two be iIJ.crea:;ed. Not being
protected from floods, it is particularly exposed to silting.
 81

(iii) Reduction of the spi1lcd area. Ge'1erally it will he pos~ible to grow on the riverine
area post- irtunriation crops, such as wheat, instead of rice, and where this is not practicable,
'iorests might be planted.
(iv) They interfere with the river's natural operations in raising the cou'1try generally.
~v) They deprive tllt' protected land of much fertilising silt.

RIVER TRIJNINC WORKS AT KHANKI HEAD-WORK.5

Fig. 11.
(vi) By shutting Dff inundation water they cause the protected land to deteriorate by the
rise of 5U )50i1 salts,
(vii) They raise the river high Bood bvd.
(viii) They cause the river, by changing its course, to erocIe its cultiva Ie banks and to
throw up sterile sand banks.
Of these disadvantages ~os: (iv) , (v) and (vi) refer to damage to surface conditions
Le" cultivation etc., and:\' os: (vii) and (viii) affect the natural regime of the river.
(d) The aclvantages of the retirement outweigh the disadvantages, and indicate the
desirability of constructing embankments well away frem the ec"ge of the river. By adcting
reclamation bunds to the retired em'uankments, the latte' can I:e made safer by the silting thu~
induced, which will tend to produce along them foreshores tree from scour channels and more
suitable for post~inundation cultivation or forest plantations, If work is carried out on a proper
system of reclamation, the emhankments can be provided with spurs projecting to, the points OJ
the river main channel where it is likely to attack the cultura ,h; land. If necessary, the spun-
should 'ce ended with Deneny's Groynes.
19 Marginal Bunds and Spurs,
(a) The marginal (unds are aligned and deSIgned iike the retired enba'lktnent
upstream of a canal weir on both sides of the river. Their chief o')ject is to protect
 '82

the' countryside' from the river spi.lls which will be caused by the afflux head of 3' to 5'.
usually provided in case of canal \\elTS and to guard against the outianking of the weirs.
These works are required in addition to the usual guice banks or Bell's Bunds in case of the
canal ~'eirs, unlike the road and railway bridges where river bed is capable of teing eroded to en-
large the section witl:out any afBux head.
The marginal l-unds are usually provided with earthen spurs protected at their ends
like the Denehy's Groyne out to the main stream of the river to protect the land enclosed by ~he
marginal bunds and to have a defined and straight flood discharge section upstream of the welTS.
Arrange'T'ent cf the marginal bunds and spurs as provided at Khanki Head works of the Lower
Chanab Canal are shown in Fig. 11.
The marginal bunds have a free board of at least 3'0 feet in addition to the afflux he~d
provided in the weir design above the highest flood level. They are projected upstream of we1rs
upto lengths wh~re the top levels thus calculated coincides with the ground level.
(b) Spurs. In India. and especially in the Punjab, besides guide ~a~ks several o~her
guide works have teen constructed in the rivres above barrages. The trammg works chlefly
0nsist of armoured spurs at suitable points from the river banks or marginal bunds. fhe types
cif armoured spurs adopted in the past are:---
1. Hockey spurs.
2. Inverted Hockey spurs.
3. Point spurs.
A. "T" Head spurs.
The type found to be most satisfactory is the "T" Head spur. The spurs are constructed
in stone with heavy aprons. Various arrangements of/-these spur~;' h~ve bee:n adopted in
the past.
Spurs constructed along one bank of ~h~ river. .In so~ cases, s~)lirs have ;o~ly been con-
structed along one bank. Arlexample of th1S lS the nver Chanab aDove KhankI Headworks.
:\t this place more than a dozen spurs exist along the right ,pank.
Spurs (onstructe~ on both banks of the River. In '~ther cases, spurs on both banks of
the river have been constructed. On the river Ravi above Balloki, armoured spurs exist on
both the banks thus tightening the river.
" , , ~ I: I j ; ....

On the old weirs of the Punjab, large natural betas exist in, frpnt of the central
bays of the weir. These' betas extend sometim~s for a length of two ta three mileS, as at Khanki,
Rasul, etc. In these cases, the central bays of the weir are masked and' their capacity to
escape the flood disch;trge is reduced. From a persual of the a'oove, it will be seen that methods
adopted lti the past for traIning the river were not ver)' satisfactory.'
, '
' ; .'. ' ' :, . .
\;

(c) The Pitllhed Islands. The e (fect of the pitched island on river training was examin-
ed by means of models by Uppal and Mushtaq Ahmad in 1936. In addition to, this . oth~r
method, of river training i e., armoured spurs coustructed at different places on both the, riv~r
banks anJ on one bank, extension of the guide banks and the sandy bela, along the guide banks
wen.! eX<,lmined Results obtained.showed that the conditions obtain~d with the pitche'l(:island
were the most satisfactory. In 1943, A.M.R. Montagu again revived the idea of the pitched
islands, The construction of.a pitched island in the river-~ed effects a change in the velocity
fluctuation and the tractive force. The tractive force near the island considerably increases and
causes scour. It draws sIt wly the main river channel on to itself and holds the river thew.
The working of the island is satisfactory, especially in the vicinity of control pJints i.e., weirs and
railway bridges. It has proved very useful in actual practice upstream of the Islam weir. :," ~
The use of the pitched islands helps in : -( Paper No, 276. Th~ Punjab 'Engineerihk
Congress, Lahore.)
(i) Training the river ~e~t.ralal--ove the 'r'eir.
 (ii) Relieving the the intensity of flow on the marginal 'b~nds, 'the guide bank~ and,
outside of river curves.
(iii) Training the river in the reach away from control points.
(iv) Protecting ,the valley land agaist devastation as well as rec~aiming i~.
(v) Improving channels for navigation.
20. Layout of river embankments.
(i) When any new line of flood embankment is under consi('eration the inhabitants shouk\
always be consulted. An embankment may shut off the floods from' land whi;ch has hitherto
benefitted from them and the people may prefer the old ,Hrangements to the new one. A single
ra.bi crop in the year (HIe ra.bi is generally the more valuable crop) with freedom from canal
assessment, may suit them the best. Their villages or homesteads are usually placed on high
ground or protected by local ring exem0ankments. Thue ma y thus be a temptation' for the
people to cut embankment, an extremely easy operation because the men who watch it can be
evaced. For the above reasons the location of a flood embankment should be very carefully,
considered, and there should be no delay in providing means of irrigation. If no alternative
means of irrigation are provided, there is a great temptation in floods to the population
to cut them to spill their lands for ra.bi irrigatioa..
(ii) A flood embankment should not be so near the river as to be in much danger from
erosion, but the ground, as already stated, generally falls, in going away from the river, so that
when an embankment is set well back it is in lower grould, mare eX[lensiye and mQre
liable tv breach. The most suitable alignment is a matter of judgement and depends
largely on where the main stream of the river is, at the moment" and' on wheather it seems'
likly to shift.
If the main,stream of the river has lately made aij. inroad and cU,taw:ay ~n em'iJankment,
but has'not shifted its course, great caution is ne~ded in fixing the line of the n,ewembank)nent.
In 1882 there was an erosion on the Indus a fe,,, miles north Of Dera Ghazi Khan, and a flood
embankment was cut away. A new embankment was constructed a mile inland. ThA erosion
recurred and the new embankmeht was cmt away within a fAw months of its completion.
Embankments are, where po'Osible, made in -,straight or properly curved rtaches.
A flood embankment, at Wast at its upstream ~d, termin<l:t~se_ in .gr~:)U.nd whi"~h is above
fl ood l e v e l . . , -, _::-
(iii) The ,top of an ecn:Ib~nkment is generally 2 or'~';r~et",al>ove the -High Flood
Level of the rWet, It should, -of course, ,c' be gntded c pa'f'a-lle~ to- ilie~Iteral'-'f4... F. Level.
but neither the gradient nor the height of the fbod is known with accuarcy,
There is generally a record dr-'mark' of some high flood, anri this is taken provisonaUy
as the flood levd or the level is calculated approximately from the flood readings on the, nearest,
river gauge. If experience shows that the- embankmen1:'is tooilow; it IS tai~ed. " t' ; ,;
I ' ; , ' , ': ;~ ,I ,: I 1; ,
(iv) NaturallY,a flood, embankment shall ce e~onol}lka~ if the, alignment follows the
cobt0urs of high land bu.t this will usually set the bUJ;ld V~J,y mij,ch., retired from· the" river .and
shall also increase the length. Often the embankme,nt .shaU lJ.qve to 'be lopped inw<;1.rd to protect
some village or <;lther, valua'ole property, r;,,, ' ;'. '1' I ' ,
" , ('1) In the neigb'bourhood of Deta Ghazl Khan it' n~s IO:'1g been" the, ctisto'lIl ,to ha,ve a
double line of flood emhaDkments in places wh.;te the fir'sf line is in any!sort bf danger, eith ,r from
river erosion or from ordinary breaching owing to lo~ ground or bad soil. Tlli§,shoul,d be I accept-
ed as a principal for all localities in which the breaching or destruction of an embankment will
cause widespread damage. Even if erosion' i.s foreseen: it may be very rapid '~nd there t? ly not,
be,time.to mako a second line. It should ,be there before hand. 'Sometlme3 the lInes tit"
connected by cross embankments whose function is to'localise damage in the event of a breach'
occuring in the main embankment; Sometimes there is local embankment the func!ion of which
is to proteoLa local- area itt casetlle maitl'embankinent is bteached htgl1e,r ~?; 'for mstance.. the'
j h
1 ~ 1-\ .! " '! jf! ) i.' f i 1 'I r, \).... ,i! .... '11 d ,1 ~ l ; i '
 / /1
I~!!&/
• I /
I' /,
.f;-/
~ '"
"

/
/
/
I
I

./,
~

-
~

II
\ \1"

\ l~ I
IIif
!t
If I
f
I
,~
,t I
II I
'I I
 84
cantonment emtankment. The flood em'ja',kments at Dera .Ghazi Khan are .shown in Fig. 12
';';hich gives the g .nerallay"ut of such works. .
21. Cross seetim or margin, I bun;! Of retirei embankments.
(a) G( neral re~uireme'lts. A river emban·kment should te high enough to prevent itts
o::ing overtopped by a:l extreme flood, shouL! have a thoroughly consoliclated section sulici~nt
to prevent percolatio!, watel from causing it to slip, should have a width of hase a1)le to suppor
the SUptTstructurc:, and to l'revent creep of wat r under it with a key trench where necessary'
to ass~st in resisting this. It s':lOuhl have its slopes protected {rom guttering by rain and from
erosion by wind and wan-lap. To retluee. tte cost and to increase safl ty the highest land
ptacticab1.; should 1e selected, l ut care shoul,[ i e take 1 not to trav, rse, if possible, ground
which is sandy, fr.a le, ntUc:\ cracke,i. or i'npregnate,l wit'1 S3.{ts, or intersected hy numerous
channels, or irregular in longitw;inal or CTOSS se,ction, as his wiH lead to di, erenccs in s(-'ttle~
ment. The ne;g,],_ol1rlJood of villat;es sIwuhl be aV<)iJed, as their preSe;1C<3 Lacreas,"s trespass
Ly traffic.

When an embankmp':lt is likely to be breachedby ero";on of the rh'er, and a retired

alignment is decided ori, which it may, e nee 'ssa "V '0 exte.:ld her oa.fter, it will be necessary to
U!/Rq 6MIl(J"""'''''r !I. join the end of t .e n'2W i ank AB to .tie Ql:d line CD by.~
/~-------- cl·ossLank. Ca' e sL,ouhl i.e taken IV ,"e hg.13) that. thIS
Si.-// ~ . cross-I. ank I as a do\\ nstr am f'ireCi ;(Ill, BD, and not an
--~~ "j,b~l ---:-::"1:1. ul,q;eam one, ): C for tl:e latter .will form "vIrat is known
.. -lIftvt!tt/" e as a 'pocket'. WI' n the 01(1 j ank s Lreached and flood
i " 'wat r is aUIlitted to tile new )'2nk it will in the latter case
Fig. 13 . ':," ~;: pond up at B to tLe level of the river opposite the cut C,
but in the former one the water will rise at B only to the level of the river oppo5ite it.
tb) Section.
The proper section depends on the height of the bank, the nature of its material and
foundation and the way in which it is to be formed. A typical cross section is given in

•
~~~~~ Fig. 14
LlClty .,."I.HCIf

Fig. 14, If the river water is likelv to be on b'Jth sides, the bank slopes should re
1 in 3 ell I-oth siees, The earth work without clods should be laid in lavers of not
more than 6" thickness and ccpsolidated before the next layer is laid. '-"'here tile g;ound is l'ke
to give way unc'er the bank: the base should Ie widened so as to gain the necesEary amount of
"upport. Where wave wash may be expectEd, the riVc'T side slope should be flattened and wl~ ~:
thl.~ material is uUielia\ Ie hoth the slores ~bould be 111creased. The top of the bank shDuld be
given a slight slore for drainage to the land side, so as to diminish t~'ndency of the river s'ope
to gutter during rainfall, as rain scoreS will aid wave wash in damaging that slope.
(e) Key Trench.
Inspection of river banks will show tl- at owing to the absence of pressure and the eff2cN
of U:e w(ather, a surface soil is gemraUy more porcus and fissured than a subsoil of similar'
matErial. Foreover when· watn meets two disshnilar materials, ·such as mate earth and
natnral soiJ, it has a tf'w"ency to collect a.t their junction and to endeavour to P1SS tetwfen
them. The key trfnch which is a small rrjddle trench provides the junction which unites the'
body of the emtankment to the Stl :soil, destroys the continuity of the base plan ani secures
the thorough examination of the part of the base it occupies, and the removal from it of
 ... ~--.--.---~- -
n·,·
.- ••..
:(
" .

........... _._

, .-~

..;.:._:::....-.::;......::: ',':..- --,

'

\
.
.
... /
/"

......
.; .~ ~

.
~un." "' ........ ~\.\\\.\\;\

~"'~ ,
~ .... - . ......4 5\.Slf .

(cl Key Trencn. .I

Insptction of river banks will show tl- at owing to the abser1c~ 01 l-'l<:;""Uc'~ ---
of tte weather, a surface soil is gem rally more poreus and fissured than a subsoil of similar;
material. Fore over when, watn meets two dissimilar materials, -such as mace earth and
natnral soil, it has a tE'n(~ency to collect at their junction and to .endeavour to pass tetwEert
them. The key trench which is a sman rrjddle trench provides the junction which unites the'
body of the embankment to the su :soi1, destroys the continuity of the base plan an:i secures
the thorough examination of the part of the base it occupies, and the removal from it 01
 85
vecretation, roots and ru',1,ish. A key trench ls most imnorhnt where thegrounrI is porous sandy,
or b is cracke~ a'1d tran,svcrscd by fissl~r::s or bv wate~-C0U~S"S etc. Its section will depend
upon the 11lgh-f1o"d (epth.agaqst tile el~b\,'kl1'llt, till~ nature ?f the sub-soil and
the importance of tile em ank IF'nt. OiUiH:J,,,ly It, Lottom wHlth may vary from
4to 6 it. its de lth ~FHn 3 h 5 fe"t an,l i~, s i!e slo')f's J} to 1 or 1 to 1. A water-tight
foundation is essentially necessary for aa orcl1uary hYllrau~lc work.
~b) &lip Trench.
Wherever there is a~ ahrupt c1:lange in the level of tJ...e ground, or of the top of the
em l)ankme.lt un,:er CO,lstructIOD, one or m');-e ~:,:) t ,;.ac.;,es Fig. 15 s\ould be marie so as to form
SLIP TU./If£H£$ jU,JC~:U;l (;0\\ ~:i tp f''' \-enc ti,e cr.,(:;p of water al(wg the jun-
ClillNNfM. ell"S$'~ BANi( JUNCTION ct:0_1 (J :e of (t:e ('aHh work, They mily have a bottom
"ie; tt of :3 or 4 [eet, a depth of 1 ~ to Z f; :t, an,t s~'l." slopes
1 to 1. .1, e'r SI([,S and t1Jose of tlle iryegu:,u:ly bl'ing
dea.t with s ;oull j>e sIo:)ed so that He eminnk:nent during
sc:tLl ,nellt may seL:e 1 !,(ltly a~,,'il,t Lem. TjH~ l~ss the
ca!)aC~LY of be so t to \\it:istand p"Tc::;lation, tl,e more
numerous and the larg.cr s;lOuld te Le the slip of trenches.
(e) Banquette (Bel m) or Pushta.
The advantages of berms as shown III Fig. 16. are
Fig. 15. that they :-

Fig. 16.
(i) Cover the hydraulic gradient of the internal percolation line and thus render the
embankment less liable to slip.
(ii) Provide space for a roadway which is b2tter located there than on the top of the
embankment.
(iii) Provide space for stacking material, or an area from which soil may readily be
olltained for the emergent raising of the emLankment when necessary during·floods.
(iv) Furnish a better cross section for subsequent enlargement or raising of the embank-
ment.
22. E:){amination Qnestions.
1. Explain with slreches the (1) Dimehy's groyne (2) Bell's Bund. What considerations deterlJ1ine
the tength~ 0\ both I (T C.li, 1933)
2. (a) Sketch the most suitable type of a pacca Oroyne.
(b) What points will you consider in designing groyne training wo 'ks ? (p,D. 1942)

3, Describe various typeB of river training and pro\.ection 'Narks, (P,U.1942)

4. (a) What are the main featufes distinguishing a torf€'nts from a fivet or a stream?
(b) What ate the main divisions in the course of a river and a torrent I (P.D.1941)
5, Sketch a suitable cro58 section of a guide b,lllk as used in river training works. Explain the
process of launching of aprons in such works? (P. U. 1943.)
G. (a) \Vhy don't the Railway bridges need a system of protection w rks embankment in addition
to guide b3.nks like the canal head works?
 86
(b) What points will you keep in view in determining a suitable layout of a system of marginal
,bunds with spurs ?
7.; (a) What is the function of retired embankment and what are their advantages and dis-
advantages?
{b) Sketch a suitable cross section of a:r\ earthen embankment and explain the f0110wing : -(i) Key.
" Trench (ii) Slip Trench (iii) Berm or Pushta.
8. Describe with sketches 'a suitable layout for thc design of guide banks and the function of th
curved ends up and downstream.
iI. (a) Explain the following terms in the case of rivers:-
(i) Erosion (ii) Scour (iii) Swirls.
('D) Why are swirls most destructive in their action?
 PART II
CA-NAL IRRIGATION'
Chapter IV' ,) ,.
HEAD WORKS
t j
Introductory.
Permanent canals are prcvided with permanent works at the site from where they
take off. These works known as head-works comprise the construction of a perm<}nent W~f
or a dam across the river along with other subsidiary works which will be dealt with in thi$
chapter. The aims and objects of the canal head-works are summarised below : -
(a) To prevent water required for canals passing the. head-works at a level below
that at which it can be utilized an,d when total supply is wanted in the canal to prevent any flow
downstream.
(b) To raise the level of -the supply so that it can commaJld by flow the area to _ be
irrigated.
(c) To gain command economically when the canal hasto pass in expensive cuttin~.
(d) To reduce the fluctuations of the level of the river.
(e) To cont~ol the silt entry into the canal. )
" (f) To render the headworks permanent thus ensuring the require.d supply ~nt~,i f~~
C'anal and to prevent them from being affected by the vagaries onhe river. ' "
2. Selection of site.
(a) The decision whether the headworks of any canal should be situated in the hil
boulder tract, or in the sandy trough. must be determined by the{ position and level of the
country which require's irrigation.· There have been warm controversies at various times between
experienced canal engineers. regarding the relative advantages of both positions, but the
pr~cticability of both classes of site having been demonstrated by actual co'nstruction, it is quite
eVIdent that at the present time the area to be commanded must determine the selection, unless
the excess of cost renders one of the sites a financial impossibility.. f
(b) There are, however on general grounds, four strong arguments in favour of the
trough site. Firstly, the well known fact, that a dam in the boulder tract, even when it is to all
appearances perfectly water tight, does not hold up all the water in the river, for there is alw~ys
a strong sub-soil flow which appears lower down, and which combined with the percolatIon
fro~ the country side along a river, and possibly with the supply from some affiuents, will give
an Increased discharge at the trough site secondly, a canal from boulder head-work,s is certain
~o lose a lafgp proportion of its supply from percolation, and this loss may be greatly increased
l~ the channel passes through iights and also, which is mostly probaiJle, before it reaches the
firm solI of the dob. Thirdly, the tracts lying at a short distance from the hills rarely require
wat~r with the same urgency as those lower down, and the quantity they need can generally
be gIven by petty works from minor hill or spring streams. Fourthly, the certainty of getting
all supply in the river at any moment when required, is ensured by a masonry·· weir fitted with
proper regulating arrangements better than by the boulder weirs in b0\11der reaches which are
not absolutely water- tight. '
. (c) The trough weirs are more expen~iv~, tflan the boulder weir as the stone has to be
Imported from long distances. Trough weirs require very expensive training works and marginal
 88

bunds. A heavy annual expenditure on the training works must be considered a necessity for
all big trough weirs.
(d) The points to be kept in view, whib selecting the site of a weir in the boulder reach
of a river are as follows:-
(i) A side channel should be so12ct('d which is not dir2ctly SU')j 'ct to Hood action and
into which the cold weather supply can easily be diverted froTI t'.le \lui 1 challel.
(ii) The side or supply channel should lr.ad dir'Cctly to the fixed o'f-taking site for thr.
excavated canal. If the canal was to take-off from the main channel, it shall n2ed very
massive and expensive works. .
(iii) The control over the suoply channel is mJ.intained by m:lkin'S its be i slope after the
first thousand feet or so something less tban that of the main river. Tho site shoq.ld be such
that a rise of 5 to 8 feet is attained at the canal head.
(iv) The sitesbould be such that sUlht':lle Sit2S wit'1 firn_'}cd a'1j "'ld~s are availaJle for,
'constructing a bund at A~ a weir at B and a waste weir at C shown in Fig. 1. as
. ..t. ' .. '. .' '. . '. .. .
.~~ c. 'U~ (e) The rLain requi-
" ~._~;- n1. ~ t .. t .
--.~
! t ~ .... _. r 11>-.
, . ..
,sites
fJ~t:
of a suita:~le site for
" '. _ ~. -. ... ~ 1 1 ! 1 l '" 4Jl-::.!>"" h' ad-work in the trouge
._____-::;::,..?J__ ":J'--~~' -----------":----v.. stage of rivers, may De
~ /-~ ~--:t-1"-~"r~-:;-~ <' t / ; / / \, briefly stated a's--;-firstly,
:,_______...--- , "'-.e"", 1j'~ ~' ~;> ~ ./:,., .6- C· I / a. na:row, straig~t, well:
_:;: ----... --.:.._-~.5:.")/ _. ' d .fined ~hannel WIth banks
-,', ;'-.Il')(;'I." ni ;-? .------~ __::__,,~_CM~ .!..".!!!~J::__ not sUi)merged by the
H".Uo. highest Hood; secondly
Fig. 1 a canal line capa1lle of
attaining command of the irrigable area, with,necessary slope by moderate dig;:sing; an~ thirdly,
the presence of_suffiClent material for construction near the site. The a]:)senceof mat~n~l~ such
~i; clay for bricks,' sio~e fo~' building and lime, timber, etc., can be got over by the provlslOn ot

-t;i t~h;IJ.tI?: ~,: '1,

.;,; .) \i> -~'IIT
" Ili • '11' /" •
. ~"/:)1 hnt; t' ,.,,1.),
. 7-~J·
!~\ff:" n=..:<_
hIUJ:;

Fig~ 2
 89
railwaY line-the (xpence of long transport is. t:owever, a serious item. Inferior sites for the
weir can be improv, d and rendered safe by training works. but no expedient will thoroughly
corr~ct a radically bad canal channel line if carried far along the Kha.dir; it will alwa~ s be open
to attacks from th fiver, rendering "a solution of continuity" possible, and most expensive and
training works over lrng distances a necessity: moreover, th~ evils of over-saturation and deter-
ioration of soil will 1e severP.ly ttlt and found difficult to cure. If aligned again in deep digging
for a grEat 1 ngt1-J.. the initial cost 0' (0 lstruction will be enOrmous and repairs h<;avy. as the
alignment must b: merely 1 S5 straight the expend ture in the adjustment of drainage, waich
it cannot cure tv ayoid. will be no light item. Of the two causes the channel in digging is
pro ably the better in the long run.
:~. Layout of HeadwJrks.
A typical lay )-It of tl:e canal hcadw,",rks is showning Fig. 2 and in aerial view of Ylerala
F esd works in Fig. 3.
It consists of th' fo1l0.ving works:-
(i) Weir proper, di\'i(~e'l into bays 1)\' divide pier5.
(ii) Under-sluices.
(iii) Canal head regulator.
(iv) Fish ladder.
(v) Divide wall.
(vi) RiVET control works.
. Thp weir should be situated at right angle s to the main stream of He ri\'er as far as
possible. Fig. 2 shows U:e layout cf the teadworks at Rasul for tte Lower Jtelum Canal. 1 hf'
Rasul weir is 4090 feet long. It is divided into 8 bays of 500 feet width each ty the divide
pins to avoid cross fiow in floods. The fish ladc1er is situated at the left end of the weir. The
divice wall Ee~ erates the weir from the undersluices.
The undersluicfs arc constructed on the side of t1;e river where the canal takes off
through the stl ucture known as He tead regulator of the canal. If there are canals taking oft
from both sic'es of the river upstream. Hen ttere shall be two unc'ersluices, one of each sides.
There shall sirr.ilarly te two divic:e walls. The river control works consist of guic'e tanks in
contiJuation d tte a utrrents supplemented by the marginal bunds and spurs to control tht'
river spilling the country-~ic'e upstnam of the weir on account of the a-FEux calJser! by tl;e weir".
T11e headworks site is verv suita Iv situated betwen high ground on both sic:es and tI-.erefore no
extensive embankments ar' sh )\~'n in Fig, 2. The extensive river control works at Khanki
are shown in Fig. II i 1 l'r,;vi 'Ui C'l'l?t!r ano th')se at Sule:nnki in Fig. IS of this Chapter.
4 T) p~~ of Weirs.
(a) Weirs are classified into t .\'0 classes according to the design of their 1~()Or3:-

(i) Gravi.ty weirs.

(ii) Non-gravity weirs.
When the we:ght of the masonry and tl~e concrete of tLe floor of tl:e weir oV.rcotl1e~
t1:e upward pressures under the floor caused ly the head of water against the weir. it is called a
gravity w;ir. In tte caSf of the modern weirs. SUC'1 as, the Emerson Barrage at Haveli the
HOt)r consists of t1::e reinforced concrete slaD and its weight is less than the upward pr.:sstrre.
The R. C. floor sla') is conti'luou:; under the divide weirs, the weight of which keeps the structure
safe as a whole against tht~ uplift pressures. Such weirs are call d n()r:$,~ravity weirs.
(b) Weirs are clas,ified into two classes acc.crding to the conJrol ,.f tr.e snrface flow OVer
r en.
(i) Op'n weirs or simple weirs.
.1
 91
(ii) Barrage.
In the first case, the crest levd of the weir will J e determined by the permissi~,lt
afflux during the maximum floods, the discharge per foot run and the pond level. The plmrj
level can be maintained by a permanent masonry crest with its top at pond level as at Rasu!
or one at a lower level supplemented oy the fallinf shutters as at Khanki or Marala or counter-
balanced gates as at Sulemanki and Islam. Permanent raised crest is unsuitable, 'r;ecause i:
shall cause excessive afflux head in maximum flood and there will re no control
of the river in low flooes. The choice cetwi en a, weir with shutters <Jnri one witL
counter-balanced gates is largely a matter of cost and convenience in working. A shuttereri
weir will be relatively cheaper, hut witI lack the speed and the effective control possihle in
the case of a gated weir- \
A barrage is a gate-controlled weir right across the r:ver with crest at one uniforn;
level. The Sukkur weir in Sind and the Ferozepore Panjnad and Haveli weirs in the Punjab
are examples of this type. A barrage provides a perfect control of tht river channel upstream
in low floods and affords better facilities of inspection and repairs. A barraged weir should
provide efficient control as the channel leading to the under-sluices and there'D), help in the
control of the silt entering the canal. .
(c) Th· weirs are classified as below according to_ the function. th<-y perform"
(i) Storage weirs, (ii) Intake or Diversion weirs, (iii) Wastf ~eirs. The storatge weirs will [l
.fealt with in Chapter HI. Part III along, with the storage reservoirs. It is the second
~lass of weirs which form tke subject matter of this chapter as med for diverting the supplies
mto the canals. Waste weirs are escape channels for floods which cannot be stored in the
reservoirs. These are constructed to ensure the safety of dams and their design is dealt with
in Chapter III Part IIr. .
\tl) Part of a weir.
A typical qoss s~ction of a weir (Bligh type) is given in Fig. 4.

.... .. -
t-"-L,--I------.-- I..

Fig. 4
A ~'elr consists of the following:-~

(i) Upstream curtain waU,

{iiJ Fore apron.
(iii) Crest.
(iv) Downstream Apron or floor.
(v) Downstream curtain wall
(vi) Riprap or Talus 01' pervio11s protection
(vii) Wells under the crest.
(i) The upstream curtain wall used to h~ of shallow depth 6 to 8 feet feep. unlike
the modem sheet piles which can be sunk to any depth, -
'(ii) The length of the lore apron is determined a~cording t~ the empirical formula
evolved by Bligh on page 169 of Design of Irrigation Worb.
 92

W 0:=4/ '-"ri where W ~ leng th in fed of fO:'e apron

(=creep coefficient
H=head against the weir
The floor lev·J upstream of the weir is determined from considerations of the penni-
,;sible scour depth below the maximum flood discharge level at tre site of weir plus the afflux
head allowed, The scour depth can be worked according to Kenr.edy's or Lacey's Theory of
design of Trrigation Channels as shown in calculations i.l chapter VI. The thickness of the
fore apron is determined from considerations of water tightness because the uplift pressure
below is balanced by the weight of water above it. About 1 ft. mlsonry over 1'0 to 1'5 ft.
of concrete will do,
(iii) The masonry Cfest is designed as a wall retaining the water pressure as shown in
L. Fig, 5. The worst couditions for the weir floor are
_"_ when the water is headed up to the top of the shutters
upstream and is at the Talus level downstream.
, The rfsultant pressure P is equal to the shaded area
of the p: e sure diagram acting horizontally at the ce~tre
of gravity of the shaLed trapezoid. Similarly wel~ht,
W of the crest acts v~rtically at the centre of gravl1Y
Fig, 5 of the crest section. The resultant of P and W must
pass within the micdle third of the base. This determines the width at th .. base of the crest wall
and the top width is fixed from the practical considerations to allow enough spaces for the
anchoring and the working of the shutters. The height of the weir crest is determined from
considerations of the permissible afflux to pass the maximum flood discharge as shown in t~e
calculations in Chapter V. The required depth on crest for tte maximum flood discharge WIll
fix the crest level.
(iv) The length of the downstream apron the total length of the impervious floor a~
determined by the allowable creep coefficient for the particular soil of the river as given in
para 2 Chapt-er V .of this part. The method to draw the creep gradient line shown in dotted
is explained therein, The thickness of the downstream apr.)n depends upon the adual pressure
acting below it according to tr.e creep gradient line. It is calc hted according
to the formula.
H~h
:::r;
_.- ..---~~-
p-l
Where t=thickness of the floof in feet.
H=Total head as shown in Fig. 5,
h=head lost in creep up to the point where thickness is to ~ determined,
p=Specific gravity of th" floor material
4
-represents
3 the factor 0'1 ",afetv
. /~~--'-

P is 2 25 for stone masonfY and COllcrete in stone ballast and 2'0 for brick work and
concrete with brick ballast,
The level downstream of the weir crest used to be kept the same as the upstream floor
level (top of the fore apron). It should be kept the lowest as worked from the following three
considerations :_--
(a) Depth reqUired for the flood discharge with the actual surface slope of the river
before the weir construction for the intensity O'f discharge per foot run over the river.
(b} From consideration of the depth required to form a hydraulic jump downstream
of the weir on the impervious protection.
 93
(b) The floor level downstream should be ;tepressed su"ficiently below the upstrealh
floor level to allow hr the retrogression of the ri' ar bed levels which follows the construction
of the weirs across alluvial riveT as explained later.
(v) The downstrp.am curtain walls too used to be very shallow, usually 8 to 10 feet
deep, because it was not p::>ssiblc to c()nst~u~t them as the spr~ng level in tne river-be~ could
not be lowered very consIderably by pumpmg. In the modern weirs the sheet pIles are drIven to
form a very deep cut off at ttle end of the impervious protection.
{vi) The total length of the pervious protection is calculated according to the tormuia

for free over-falls, L 1 =10C r~-q-

vi lOx75

and for sloping downstreem aprons, Ll = lie I Hh ~ q; jPage 16401 Dams and weirs by W.G
. vi lOx j Bligh of 1918).
Where L) =length of the talus
c=Creep co-efhcient for the river-bed soil
q=dischaJ~e per foot run
Hb = Higb. of the crest wall above bed
The wells under the crest simply serve as nails in the river bed. Their usefulness as
ilfstroyers of creep head had neVEr teen attahed \\ith SUCC€Ss bHauEfI the sraces in bet\\een
them could not be made water-tight. The usual attempt to drive wooden piles betweea the
wells has not bef'n much successful. It will be usual to consider these shallow wells as an ac:ldi-
tiona} factor of safety and the creep length required was provided, without taking them into
consideration.
5. AtIlux;
Afflux is the rise in the maimum flcod level of tIle river upstream of the weir caused
as a result of its construction across the channel. In the beginning, the effect of 1he afflux
in raising the river levAls is only felt up to a short distance upstream according to the length
of the back water curve but in course of time the river bed rises due to the silting up caused
By the additional waterway added to the section upstream of the weir by the afflux height
and the effAct travels upstream till the river slope upstream of the weir is the same as before
its construction, In the design of weirs founded on allu vial sands, the afflux is limited to 3 to
5 feet.
The amount of a fflux will detemine the top levels of guide banks and their If'nghs
and top levels and sections of flood protections. It will govern the dynamic action as the greater
the afflux of fall ot levels from upstream to dowDstream, the greater \\ill be the action. It
will also control the depth and location of the s tanding wave. By providing a high afflux
the width of weir can be narrowed but the cost of the training works will ~o up and the risk
of failure by outflanking will increase. The discharge per foot run, tbe depth of the scour with
the action on the loose protections upstream and downstream, as well as the depth of piles at
either end will increase with the afflux.
6. pond level
The pond level is the water level required in the undersluicfs pocket upstream of the
head reg~lator of the canal to feed the canal with full supply The full supply level in the
canal at Its heads nepends on the levels of the country which it has to irrigate, and the
peqpissible ~lope in the canal. The working head at the canal head regulator should be allowed
to be about 3 to 4 feet. while the waterway for the head regulator may be designed with less
than ha~f ~he available head. Enough margin should be left in fixing the pond level for the
f U1 ?re sIltmg up of the canal bed and the silting up pf the river bed downstream of the well
WhICh will cause sluicing difficulties for occasional washing down of the silt in the pocket.
 94
, , Pond level dete.r.mi~ the hmght ()f the undersloice 'gates and the height of the shutters
above the permanent mwM'Y ,crest of the' wQir.'· ,; , ,
7. Waterway for weirs,
The waterway for the weirs has in the D::J.st been kept arbitarily limited by the
permissible scour depth wkkh, would fix,the floor level upstream of the weirs, Tt the water":ay
is restricted very much., the float' level., will be relatively low necessitating heavy and expenswe
pumping to lay the upstream and downstr,eam impervious aprons.
The scour depth for a discharge per foot run be worked from Kennedy's silt formula
Va ='84 D'64
which applies to the Punjab conditions.
It can be'modified to suit othe" i?ra.des df silt by introducing another factor known as
C. V, R, and denoting it by x. ", ' :"
Va ='8.\. x, D-M
. . \ .
Let the discharge per foot run be q cusecs . I :

Tlten q = Va 1>: ='84 x 1)1-« ,

,61

: :'0=
..
(it .(~.)
, .• ' X
!.: '," •

,and wheri~ x is HIiity

D=l:il q.&l ''
,~) Anotaer method to determine the sconr depth is that given by lacey in his papel .
on .,Stable Channels in Alluvium" N~ 4736, (19:JO} Institute €If Civil En~ineer6; London.
" ' "7' ,'·('QiL.~I):1;6' "," ", , ' '
\0;=
,I
--as", ,
' ,

',;'
,Vo',, iw\
and,R=
. :7.JvS
. -
fL'!'
..•
Henee ~=j305 (3~fL)1
> " , i, ".\b~t P .. 2'67 Ql, ,
, ! ~. !. : ' , ~." . , - ,. '. I (t j. ~'" " 1. :., "~I • •

,.'. "h~(di~~harge. per, f~ run or :(i==~= ~~=yQ-

. ,..., . '! ' , P
w 2'67 vQ ?;ftJ7
Substitui!lg i.n terms of q
q2
R=.9 ( -f~' ,
)1
,; ., (c)' In t.he table given below, the, actual waterway provided' between the abutttJent§
is compared with t)Jat worked out according to La~ey's formula F -:-2'67 Qi .in the case of ,
some of the Punjab weirs, Waterway for weirs, for the avp.rage conAitions in thp Punjab shoulrl
be 1'5 times Pw worked out according to tb,e ALacey's formula. The eXperif'nce has shown that'
even the restricted weirs like ~hose' at Ferozepur. and Islam sliMer from the satne tendencY'
of forming islands upstream of the weirs as in the case of the reJatively WIder weirs. The
 o o
I~' -:> o

<:~ :~:Jj:\ ~
000
o II) 0
000
N'" ~... If)"

o
lI?
"" ....
00

o o
o o
o O.
o·
'"o
:--____:_
'"'" 00

: •• I
o o o o
o o o o
o o o o
.... ....<D
!Y)' ",' a; <D'
00
'" '"'"
c 'o=' 0 o o
'? o
o o
0 g co
o co
E o o
.... ,,,' O.
..... lI?' lI?
~
o
~
Q)
" '" '"'"
00 '"00
:-:I
.... o o o ,,,
....o .... o o o o

-o
~
~
o
o
.... o
l~~

.
o
00'
00
""'" __N----_.----------
'"
o
'f)"
o
....
....

4> o o o
o o o
~ oC" '"
~
If".;
r1')
C'I
'C'l
C'I+ .... .,,' ::.
co o
",,"
0:>
.....", "" ....
'd
'"
-------------------------------------------------1
o .0 o o o
~ 0 o o o
""_.
1('; <D
. lOt ,,,
- !"':.
0>
o
O·
o o
.... .... ....
0' O'
,.\.~
~.
lI?
....
~ '"
~ c c
bb co lIS ,,,'" o
<D.
""
.~ '"'"00 C'< ....
lI?
o ""
~
'"

:c'f!),
tn. .~
I
j
1 •
 96
restricted waterway increase the scour depth and consequently lower the, upon levels. This
entails relatively higher expenses of the pumping to lay th~m.
8. Effect of weirs on the Regime of a river.
(a) The river regime is affected by the construction of a weir across its channels in tlw
following ways.
(i) The slope of the river upstream of the weir flattens due to the pondhg up of sup pEes
(ii) An increase in tortuosity, as a result of ponding up, as the bulk of silt charge of th"
river water deposits in the pond, leading to the formation of irregular shoals.
(iii) A progressive degradation or retrogression of bed levels downstream, due to the
picking up of bed silt by the relatively silt free water escaping over the weir.
(iv) These ef:ecb continue lor the first few years. but l~ter due to ~ontinom siltlng up of
oond and increasing tortuoslty, th:! bed levels of the stream wIll t2nd to fIse as the b~d levels at
the weir are fixed. This can be explained by t:1e fact that the water will need great~r head, to
overcome tte increased distance which it ha., to travel. An increase in tortusity will necessarih'
enhance the rise will be felt higher U? the river than would otherwise be the case As a result
of this progressive rise of bed level, there will be tendency on the part of the river to regain it,.;
original slope,
(v) A stage will come when u;Jstream pond ab30r~s no further silt burden. Owing to
the off taking canals drawing cOTnparatively silt free water, the exeess of silt will g) downstream
of the weir wl i~e the amount of water passing over the weir will be below the normal due to
canal withdrawls. The river celow the weir, will, thus, have to carry an excessive silt charg ~
with a lower discharge, This will result in progressive silting up downstream, an increase in
tortuosity and, therefore, a recovery of bed levels downstream.
(b) The changes in the regime of a river caused by the coustruction of a weir have an
important bearing on the design of ihe weirs as outlined bdow;-
(i) Retrogression of Levels.
In the nrst f~w years following the construction of a weir. the retrogression of bed levels
downstream is rapid and progressibe. In the case of the Punjab weirs this has ranged cetween
4- to 7 feet. This lowering of the bed levels in the early stages if not duly allo Ned for in design
may result in a failure like that of the Islam weir in 1929. The retorgre5sion may. undermine
the stability of a work by an increase in the exit gradient beyond the safe limits. It will in-
crease the destructive action of the standing wave as with the increas:d fall and decrC'ased
depth of downstream water due to the lowering of the water lewIs at that eLd, the wave will
tend to travel down to the Dlock and loose pr ,tection area.
As a result of the retrogression in bed levels, while the low water levels have been
found to drop from 4 10 7 fe~t, tIle maximu-n fio:)d levels have not bee \ knoNn to have dropped
by more than 1 to 1'5 ft. The dow lstrea n rlood level should I)e depressed below the upstream
river bed by about 4 feet.
(ii) Restoration of the original slopa upstream of the weir.
In the course of time the river upstream will regain its original slope wh:ch implies that
the effect of afflux due to the construction of the weir will not be confintd in magnitu 11.: anll
length to the usually accepted distance as determined by the back water curn (apPfJximatel:v
:l.Ha IS It will travel very far up and will be felt in full all that distance. In other words,
the entire bed of the river will ultimately rise uniformally throughout the zone of protection and
training works. As, however, the full affect will not be felt until after lapse of many years, the
free toard, may, in the first insta 1C::-, De reckoned a JOve t'le H. F. L. a, determined by the back
water curves
(iii) Recovery of downstream bed levels.
The process'ofrecovery of downstream bed levels after the initial retrogression, is SlON
but steady. It may take 20 to 30 :vears but the bed levels may rise higher than those before
 f17

the constlUction of the weir, At Khanki the rise above the original bed level has blen of the
costruction of the weir. At Khanki the rise atove tl~e original 1 ed level has i een of the order
of 2 feet. A rise in the downstream [-ed levels may lead to the loss of control of the river in
respect of silt regulation, making it necessary to raise the weir crest, Khanki and Marala are ins-
tances of ,>uch raising. The Khanki weir crest was rassed hy two fe3t in 1910-11 in days numbered
4 5 ann 6, and in 12[7 in th1 rella i ning bays It was again raised hy a further two feet in
, 1920-22, Marala was raised by two feet in
.1925-26 in order to o:Jtain control of the
river and to improve silt conditions in the
main line. Tht'se facts point out the
nec .. ssaty of cesign for a pond level
I suffici -ntly ahove the full supply level in

i
+
the canal so as to lea\'e ample margin even
after He rise in the downstream ced has
taken place .
••
.# 9. Undersluices
(a) Object of the underslulces. The
•
t
.~
nndersluices ara required to keep the river
U:l.Cer control aiming at the following
1'oi[;ts :--
: (il To scour the silt deposited on the
river ted in the pocket upstream of the
canal head regulator.
(ii) To preserve a clear and defined
river channel approaching the regulator.
(iii) To facilitate the working of the
weir crest when movable in the form of
shutters or counterbalanced gates.
(iv) To lower the highest flood level.
(v) To pass the low floods without
dropping the shutters, th~ raising of which
entails good deal 'f latour and time.
(iv) To control the silt entry into the
ca'1al
I (\'ii) fn the case of sudden fkods, the
undersluices can te opened to fill the river
l eds downstream to protect the downstream

+i Talus from the H.rdraulic jump action.

(1 ) Tte capacity of the undersluices.
The capacity of tte undersluices is
.
<.

't
i fixed on tte following consi(~eratins :-

'I (i) To fUsure scouring capacity, it should

1 e at least dou'ole the canal discharge.
! 1
i (ii) The undersluices shouid be capable
of passing the low flood diseharge except-
I ing the three month of the monsoons
. ___l_
without over-topping the shutters.
(iii) The uncersluices should l-e capable
d ra~ing the winter fresl:ets under thi):
Fig. 6 Fig. 7 gates with canal in flow.
 98

(iv) At high floods,. tllc undersluices should he capaule of passing about 10 to 15 Yo

of the maximuHl flood disCharge. Jhis aims at rerIucing the length, of the .weirs became long
weirs will cause idand formation' upstream and will also serve to maintain deep water chann~l
towards the canal regulator.
If there are two off-takes, one on dther sil~e of the river, there shoul(j be' two under-
sluices of such capacities in each case as ddermillerl from tl!es~ consicerations.
(c) Design of the undersLi:es.
The following points should be kept in, view in dtsig1ling the llndersluices : -
(i) The crest and the floor level in tte pocht :upstn am should 1 e at the . lowest
pr~ctica'Jle l~vel of the rivtr.
It is usually kept at, tte Je!l;el of the apron upstream 'of the
weIr crest.
(ii) Large spans are preferable. In t);6 case of old '~eirs the lllH'ersluices span~ used to
Le not longer than 20 feet tecause wic:er gates were not availa' Ie and were diificult to work. Now
the gate design has teen very much improved by the use of counter r alanced weight as in the
ca,e of stonyc's gate arqmgement cescri: cd later in this chapter. .Kow spans of :30' or more
are used.
(iii) The floor thickness and its legth are (;etermined according to the weir design as
outlined ir. the next chapter.
(i\,) The discharge intensity leing maximum in the undersluices. the foundations, the
floor and the Talus should te maue extra strong.
(v) '1 he floor downstream of the crest should not I:e keDt level with the crest level
because the formation of a proper hydraulic jump as destroyer of energy is very uncertain on a
level floor. ~breover from the considerations of retrogression, it we, uld I e advisable to dpress
the downstream floor up to the likely retrogression b~d l:v,'i downstream which may te at
least four, fec·t.· ,
(vi) A' bridge downstream of the unrlersluice s)lOuld always I'e ployided to lift the gate
hy cranes if Stoney's arrln.g~m~nt fniL., on a:ty account.'
Typ~cal sections of the u:ldersluices are shown Fig. 6 and 7.

The length of 11npet~-ions apron \lpstream~of the sluice gat!; is got from the empirical
iormula.

C=crer,p'ccefficien{
H = Height of the shuttered water Lv< 1 allove Huot level.
The length of the Talus below the undersluices is got from the formula on png~ 216
• Design of Irrigation works" by W. G. Bligh.

, I ,--- 1-"
;, \ j' Hb
10 .Xv
,) 75
q l ' fee t ;
',\" nere [' ~',~ 1:ngtlll1

C= creap coe:11cient
Hb c,~height of tht p,:rdtallcn t welr cresC
'. and q = Discharge intensity per fC run,
The thickness of the floor IS calculat..,d fr<Jm the creep gradient the~ty iik'~ the welf:;,
The depth of the Talus is 4 to 6 Jed, usudlly one and half times the depth of floor downstream
of the; weir. . .
 99
10. DivIde Wall.
The divide wall is s.imply a long divide groyne built betwecd the weir and the ilnder-
sluices. It separates the turbulent river in maximu"m floods from the pocket in front of the
canal head regulator. The wall extend, upstrean to a little distance Jeyond the 'oeginning
of the head regnlator and dowustr'am to the eur: of the Talm of the undersluices. This NaIl
plays a very im~ortant part in controlling t~e entry of silt into the c mal Jy enclosing a pocket
of very nearly stIll wat ~r awl. by se::n.ratmg It fro n the tur ml ~nc~ and vagaries of the alluvial
rivers in floods.
OIVIDE WALL In the case of old headworks such as at Rasnl.
the divide wall is simply in earthen bund protec-
ted with stone pitching on both sides and provided
with stone apron as in the case of guide banks.
The upstream end is provided with extra strong
protectio'l in the form of aprons. An earthen
ctivi('e groyne requires a lot of space for its section
and shuts 1ff a c'lllsidera' Ie efective water-way
of the weir and it is, therefor!:', now usual to buiI:i
the divi,'e '.vall in ~olid masonry as Shown in
Fig 8.
Normally thE Ie will be water on both side,.;.
_,,"CH';;": ~.~ I• ........-·.--j but di erence of pressure on 'loth siclEs of the
wall is Lk ly, due to still water I,eing in the pocket
Fig. 8. and a strong river current (~estroying its velocity
head by striking against it. Similarly, if flood is being pa?sed thr~)Ugh the undersluices, there
will i:e rdatively less de?th on the weir side. It is usual to deSign tht wall for a pressure
difference of 3'0 f et depth of water <)n either side. In the case of masonry divide wall, it is
necessary to provide Nell foundations for as least 100 ft. lE'ngth fro:n the extreme end taking
them well below the deepest possible scour.
11. Fish ladder.
Irr the large rivers, the ftsh are always moving from one part to another. In the
beginning of winter they leave the cold water in tbe hills and move down to tl e relatively wann
water in the plains. Before the monsoon in the months of m.'tY and June they move up again
in search of clear water. In the mOllths of July and August in Northern India, the female
fish lays eggs in the water which meeting the juice created by the I'1ale fish in clear water
fertilizes th~ eggs. The meat of the fe-nab fiish is usually piosonous in the months of July and
on ac(ount of the presence of eggs in the 'Dody, The fish t!terefore travel up and down hundreds
of miles along the large rivers and it is, th~rd')re, essential that as )ace should DE' prov:(:eJ in tLe
construction of the modern weir for the unintt:rrupted movements of fish.
Fish Ganeasily travel against current of water with velocities up to 100r 12 fect rer
second. The design of the fish ladc:er should 1 e such that'the velocity of the current against
which the fish haw to nlove s~lall not I e higher tha!l this limit. There is usually a head of
a'~out 16 to 20 feet from the upstream of a we'ir to the downstream water level 011 the river
in the winter m.mt!lS In an op:n gap in the weir just like the O:le in the Ganges at Hardwar.
the velocity of th" current will l e very high, and therefore even the strong flS'l will not Le a;.Je
to travel upstream. It was nothing short of <.:ruelty to these poor animals to see swarms of
large fish 2 to 4 feet long collecting downstream of fish gap in the we'ir at Hardwar an i
dying of sheer exhaustio_1 in attl!;npts to move upstn:am. A prop:,r fish ladd'cr ha~ no",' Leen
provJ(:led there.
In the modern weirs the fall of water is broke:l ill small strips. Typical \;e3igns of fish
ladders are shown in and Fig. 10.
Tiie suita ,Ie site for tlH! fish ladder is near th'~ divide 'vvall because there is wat'!l
throughout the year in river bed downstream uf the undersluicts (lUly. It ¥; usually localed
Letwe.:; the wtirand the c:ivide wall'and in SOme cases it ha been !uilt vithiil 1h, divide lAal..
The walls are huilt oblique and holes are also staggered as shov"n on tLe plan so that the fish can
 100

take rest after passing O:lC hole before they move on to the other. The;e walls also serve to
break the fall in small steps as shown by water line in the section.
12. Read Regulator.
The object of a canal head regulator is three fold:-
(i) To regulate the supply entering the canal.
(ii) To control the amount of silt entry into the canal.
(iii) To shut out river floods entering the canal.

(2)

-----
1Tt&k]
------
:~ -
i

'"" .,~.~ ..

:~~§~ ~

S'_f.JFf':O tJ£Nfl. DESN:N , " " Wlot. flANGE IN /IPS'f11EAM wmR!nn

Fig. 9.
 JOI

r
t
I

Fish ladder long Trimmu Weir Wall under construction Fig. lOla).
I

j
A view of the fish ladder long t he right divi,:e wall Trimmu Weir Fig. lO(b).
 10'2

Upper Chanab Canal Head Regulator as remodelled in 1937

Canal Head Regulator Downstream view.

 The water-way 'Of the head regulator is fixed so that the full snpIJly dio>charg-e 01 th~
2anal couIc! easily pas3 over the crest of th~ heitd regulator with th'~ d_;signed pond levd, with
lmple· factor of safety, to allow for any silting up of the canal. The regulation is arranged In-
provicing gates son-:e times in two set~, O:1e r·sing on the other dropping. as shown in Fig n
is in the case of the Sirhind Canal '-egulator and Upper ChaCla-) ('anal hC:ld regulator ifI
rig. 12 a) and 12 (b).

5JlfHINO CANAL RECiUl.ATOFt. In tllf~ cas;; of old Head regub.

.ATfS CLOSED GATlJ 0l'£1f
Fig. 11
the spans of 20 to 2S f. et are now quite common.
aking off at Haveli is given b:~low in Fig. 13.
)rs, th~ spans were small
t
usually 5 to 7 feet, by providing
dummy piers dividing U e sp"n
of the downstream bridge.
tecause the gate liWng arrange·
ments ~ere crude in the form of
travelli" win'h mounted or,
rails. 1 h(> original Sir hind. canal
head rfgulator har! 39 spans 01 5
f '( t each. In the modern canal
head regulators, the modern
arrangement of radial arm gate,
[ift<,d by the rack ane{ pinion
methJd has reduced the numeeT
of the head regulator bays
A s :;ction of the canal head regulatO!

~fCTJON THROUGH HAVlLI CANAL HEAD RE.:

1-'-. - •• --~.-- •• -.- "1 (
,.." .........
. ... ".
, .
i- ,·4-4 4 ,;..L..:...I
Fi~. la
The silt control in the head regulator l(esign is provided by ha.ving a suitable raiSe(1
:rest. In the Sirhind Canal a crest wall of 7'0 feet height was put in dU3 to a serious troulile in
he Canal. A rennanent high crest shall need a high pond level to feed it in winter with higtJ
osses from th\) pond and it is, therefore, lIsual to provide a rising cill g ,te as shown in Fig. 11.
The head regulators at Khanki and Panjnad have teen provided with silt excluders in
ront of them in the pocket. A separate Chapt~r.No. XI of this part. has been devated to the
iesign of silt excluders and silt ejectors. As tIle riwr conditions in the pccket are usually ver~'
nuch di'>turbed, the silt excluders in front of t~le head regulators in the pocket have not proved
;0 efficient as a silt ejector ruilt in the canal WIth steady flow af-Jout a thousand Ieet downstream
)f tht head regUlator as in the case of the Haveti Canat:
III the case of the maximnm floods, tbe head regulator of a canal is to be completely
:losed lest heavily silt-laden water of the river silts up the canal. There are now daily silt
)~servations being carried nearly at all head'tlorks measuring the sil.t content of water passing
nto the canal and that of water escaped under the sluice-gate. 'VVhen the percentage of silt
;harge in water entering the canal exceeds a C6rtain figure fixed in ea(:h case, the canal is closed
 l04

This function of the canal head

WATER fJlt£$SlIRl OfAeRAM5. regulator has a very important effect
• on its des;gn. In maximum floods,
the worst hydraulic conditions are
created. Hydraulic head causing
flow under the regulator is of the
IIfllOtlt:
order of 25 to 30 feet with the canal
closed which may be about dou.Jle
the head for which the weir floor is
designed. Such a high head shall
require a very IOil~ impervious floor
with the modern introduction of sheet
pil's, and excessive floor thickness.
The gravity sections will ce impossi-
ble and even the invertetl arch or
reinforced concrete sections will te
wry expensive. Tbe economica
solution of this trou'Jle in the cas:o of
large rivers would be to make the
whole floor of the pocket impervious
by providing a water-tight concrete
of say a'...:out 1'5 to 2'0- feet depth
ensuring that the joint of this floor
\\ ith the crest of the Canal head regu-
lator is abso~utely \vater tight. Such
a floor is required when it is desired
.;; to provide a silt excluder and shall
also lessen the cost of the undersluice
Fig. 14 aprons,
The stability of the head regulator piers should be tested to withstand the overturning
momenis caused by the high pressure head in floods, The designs of the crest wall of head
regulator, the gates and the dreast wall will he worked out for the water pressure diagrams
shown in Fig. 14.
It is now usual to provide a solid masonary crest wall and a brellst wall in reinforced
concrete The horizental arched breast walts ar,; now o:)solete. The reinforced, concrete breast
wall can also te reinforced at its bottom to act as a beam to support its own weignt of the gate
lifting arrangements instead of providing additional as'in the case of old hearl regulators. Tce
layout of the head regulator is usually at right angles to the tndersluices, parallel to the divide
groyne wall. Sometimes oblique entry i.nto the can II is considered desire able and the head re-
gulator is splayed out 1 in 4 from the uncle rsluice, taking OJ the canal at right angl;}s to the
splay. This has hen tried at Sulemanki, This will provide additional still-pond \\ etC'r-way
in the pocket and sllall also insure relatively ~mooth ently into tl:e canal. There is no
mpirical formula to determine the length of the Talus dONnstr2a'11 of the head regulator. It
wouH ce ample to provide a Talu~ equal to 4 to 5 times the de?th of cam.! ani 2 to 3 feet thick
in concrete blocks of stones of 110re than one foot diameter.
i3. Silt control at Head Works.
There are three n ethods of reguLttioll ll'iually acloptc(', at the canal hcadworb to
control the silt entry into He canal.
(a) Still Pond System.
In this system the river water in excess of the cd!lal requirements is not allowod to
escap; under the sluice gates. Only the supply required for the canal enters the pocket. Thl~
. velocity of water in the pocket is very much redUCer{ on account of its excessiv'J \\'ater-way. The
 105

silt deposits in the pocket and clear water enters tho canal. This syste'1l presupposes that
, the canal regulator is provided with a hi.gh crest wall as shown in Fig. 7. Wilen the silt in the
pocket has accu:nulated to say beioN 24 hours a_1d the stui~e gates are o;->ele1 to sc)ur out the
silt deposits in the pocket. This process is rcpeated when egain the silt accumulates in the
pocket.
(b) To take supply from the curved river ap~roaeh cJ.ennel.
In this system, the undersluices remain closed and the river is allowed to develope a
curv~d approach channel by e3caping su??ly i.l e:'{ceS3 of c:nal re'i.uir~m,ts over the weir by
opemng shuttered portion or by co Istructing an)ther set of Ul ierslu!ce, in the midJI~ of the
weir as at Khanki. The wat~r is the:1 taken fro11 the outside of th3 curve wh~n t:le chann~l
section is deep as shown in Fig 11 Part l[ C Ul)ter II 1. Then i, cross flO'N of silt in. a channel
sectIOn on a curve which re~ults in bw silt charg" on the outside of a curve as co:npared with
that on the inside. After thus selecting relatively clear water froen th~ river channel. this
system can be cem'Jined with the still pond m~thod for further silt selection in the pocket itself,
(c) To usc under SlUice Gates to escape ex~ess river Suppl~.
In this system, the river supply in excess ef the canal requirem:mts is escaped under
the sluice gates. In this system reliance i., placed on the fact that the coarse silt shalf conce~­
trate near the po~ket floor and it can safe~y be escaped below the gates. This system though ~t
looks sound in theory is very treacherous on account of the uncertain approach chann21 ~ondl­
tions in the river. When this system is used, there mUst be provided a fairly high crest of head
regulator with suitable desigh, Even then the daily silt obServations mUst be teken to find out
the silt content of water entering the canal. When it exceeds a certain figllre for the particular
canal, ;t should be closed and tne pocket cl~ared.
The still pond system has ceen very successfull at Rupar and Marala. This is by far
the most effie ient form of regulation far canal head works which are not )rovided with silt
excluders with the only def"ct that the water has to be wasted in scouring the pocket for aboul
a day in a month.
14, Guide Ba.nks & Marginal Bln1s.
The chief idea intlle design of these works is to concentrate the river into a narrO\"
channel by a pair of long guidd banks. It is claimed for this arrangement that is not only
iuduces a clear and direct flow to the weir but it also greatly reduces the nUll'Jer of spurs
needed along the margmal bunds. Being forced into one restricted c':lann~l, the river...-has no
room to swing abont, and there is no possioility ·of cross flow just upstream of the weir:' 1£ the
river does swing a"Jout, it must be a )OV2 the entrance to the guid~ ba 1ks w:1ic~1 is SO far from
the weir that little or no harm can occur do the protective works behind the guide banks at the
weir itself,
The fullowing points should be kept in view in the design and layout of the guide
t anks:-
(i) The choice bet\veen parallelism convergence and diverg3nce must be dictatd bv
the concition of the ted of the river,_tor constrction purposes, during the working jieason, For
it is better, if possible, not to be oJliged to lay the apron in deep water.
(ii) But, if practicable, it is Letter that the guide banks should approach each other
r.ear their udpEf'cnds, tefore their uppe: curves begin. The urnount of construction may be
say, anything up to dou·)le the CO:1l )inej thic~me,s of the bridge piers helow low water.
(iii) The length of the upstream part of the guide banks may be made equal to, et say
upto to a tenth longer than, the bridge. But attenti))1 should be paid to the possi0ility of the
river tending round above one guide bank into the still water area at the back of it, and erod-
ing the main approach bank. In specially wide Khadirs this may involve the me of very long
guice banks.
 _ ---
_-- -- ...
'.

;;. i "

Fig. 15,
 107

(iv) The lenf th of the downstream part of a gUIde may be a tength to a fifth of the
length of the bridge, ~ccordin~ to the jugdment that. may be formed as to the ~ctivity of. the
swirl or disturbance lIkely to te caused l y tbe splaymg out of the water on leavmg the hndge,
for the swirl, if there is one-must be kept sufficiently far away so as not to endanger tnc
approach bank ..
(v) The radius of the curve of the downstream end of the guide bank may be such as
the material trains can run on, say 200 to 600 feet; because it is convenient to take ttc stone
service line by this route.
(vi) The radius ef the upstre~ 11 CUlVeo part of the guide .bar,k maybe any thing from
SOO to 101_0 fe3t, according to the estImate of the PrJ >a )le v J )Clty of the current past it. fhe
curve should be carried well round to the back fully 1200 to 140°.
Free board for guide ba.'1k should be 3'0 f3et to allow fuV afflux head above the highest
flood leVel. The thickness of the sto_1e in slop~s and aprom should b~ as given in the last chapter
paragraph 17.
In the case of the Sulemanki head works the guide banks curve inwards forming
bottle neck, the width l:etween the noses being 1600 fe:~t compared with 2223 feet between
the weir aimtments. The 0 \j~ct was to induce central flow b floods and to prevent the formation
of shoals. This 0 )ject has not been attained, because the island formation on this heabworks
is as acute as on the others, but on the other hand, this has resulted in a strong flow in floods
round the guide wall towards the unriersluices with the likely effect of damaging the ted
protectIon upstream in the pocket. The marginal bunds with the protective spurs are required
at all heads in addition to the guide banks. A typical layout is shown in Fig 15 as used in
the case of Sulemanki weir.
The underlying idea in the design of these works is that firstly, a partially controlled
and restricted flood by the marginal bunds or retired eml>ankments should approach the river
section between the guide banks, and secondly, they should 1?rotect the country upstream of the
weir from river spills due to afflux caus~d by the construchc~ of. the weir. The design and
layout of such pfJtection works has already reen dealt WIth m Chapter III. The free
board is kept on similar considerations as for the guide banks.
]5. Movable Weir Crests or Shutters :
(a) Objeet:·-
The object of placing a movable crest on tht pmnanent crest of a weir is ·to secure
the maximum fair weather supply level required for the canal, and the maximum high-ftoods-
level, so as to lessen flooding, the height of flood .embankments and the danger outflanking.
When these two level; .:orrespond, the greatest eillclency of movable crest is oDtained.
Another advantage of a movable crest is that it lessens the silting up of the backwattr
pool, which if it occur, excessively, may interfere with the supply to the canal. Movable
FlASIIlIIIA_ - crests are~ ho.wevEr, very t:xpensive.
ItfVAYIOH nCTIOH and may I e d~ffcul~ to. work; as their
. c~st and J?1amp~latlOn mcrease greatly
-:- WIth theIr heIght. It is generally
advisaUe to make nem low, and to
depend upon the solid weir to
olltain the I alance of the elevation
necessary.
Movable crests may te divided .into
following categories:-
(b) Flash Boards.
l.,..... ., •• t."" ,...
They are usually of wo;,d as shown
Fig. 16. lU Fig. 16.
 lOR

The ordinary flash board is suitahle for all cases where the head held up does not
excee~ ,5 feet, although, if. the h',ight is more ollan 4 fe~t s1lUtters will prolJa' )ly be more efficic:>:t.
This tYfJe is simple in construction anrl can be used in all cases. Even although boulclers may
damage the flash, boards and tlJeir spports, th'se are ch~ap and can be easily replaced. Flash'
boards are, hO\\ever, di:ficult to repltce until the water has fallen to the leve of their [Jase, and
are therefcrc best ada pte 1 to high (;ams or esc-are v,·cirs over which the water flows"
but rarely.
(c) Hinged Shu,tters.
Th shutters are iron gatts hinged :1t tLe rase ?n th~ weir crest. The hing~
ing or otJ-:er guiding arrangerr,ent ~t tl~e base of the shutters IS a .dlsadvantge, i eeause many
stoT}es of first siz~, or great~r, c rned b~ ~ost of t'le stream" are ha)b to lodge in the spaces'
in which the hinges and gUlde or spportmg rods, work. .,
HING£D SHU-rTERS Fig. 17 shows a typical
shutter used in lncia, wh:ch is
dro ?ped Iy releas:ng the curved
lLlVATION SfCTION honzental lever. Ttc raising
of the shutter is erected by a
!;."-'!.•..e'!..!~!..!~'.!:
•
f~!.~_I:~:..L __-
hand crane which hooks on to
the smaller (hanging)' loop on'
..,tv fW ./.'" 'AfIII the upstream face of the shutter
Trained men can raise and set
these shutter wite ease,S, 6 or
even 7 feet of water, provided
, that the backwater below the'
I,,, '''ON dams is not so high as to inter-
." wtQ' • ., •• "TO'"
fere with the workieg of the
crane.
The dropping of these shut-
ers is perf, ctly easy, provided
that they are not over-topped.
If once over-topped, it is almost
impossitl~ to get them down>
Fig. 17. although this has been effected
with some risk When the river
did not rise rapidly after over-topping the shutters. If the shutters are over-topped. and tl~e
rise of the water continues> the dam will probably re d~stroyt d, not so much by the impact of
the water falHpg over the shutters~ as by cross-currents mduced alo;lg the dam from the portion
where the shutters are up, towards that wtere they are down.
" .
(d) Automatically Droppin5 Se~tt,ers.
They are hinged Type Shutters up to 5
or 6 ft, height. As a rule the oesign. consis--"
ts of a projecting bevelled arm which is
forced by tbe fall of one shutter and pulls

~
.
OIA I H _ ,I.tt
,
'"
~; lilA, ","
out the key and sets the next one free.
The arrangement is shown in Fig. 18
They are raised by men with about 2 to
3 ft. depth of water, or raised by using
traveling crane on the crest or from boats.
The shock produced by thc simultan-
eous fall of a long line of shutters is a
seJvere trial for a \veir. :;
~ig. l8..
.. ,
:~
 IIJ~

, HINGED HYDRAULIC SHU'TTE.RS (e) Hinged Hydraulic Shutter'.

An expensive but reliable {OIllJ
of hinged shutter (Fig 19) is work-
ed tJy mears of a hyciraul:c ram,
of \~hich the plunger has a niller
at its top J earing against . tht
shutter, and is actuated t.y ~ ater
under hydraulic pressure conveyed
in a pipe in a water-tight tunnel
formed he low, the weir crest. The
shutter can thus be raisell or
lowerc d when drsired, and is under
IfUS.~.
complete control; it can 'r:e made
so as to st0re up to 8 feet in
depth.
1--.- ,:.-.__, (/) Trestle Weir Crests.
Fig 19.
This form is of. I,'rench ongm. a.oJd has l een made to store as much as
20 feet in de pth :mt a)() CIt o:le-lalf of this IS a usual limit. The trestle; (Fig, 20) are made
of iran and are built up as lattices; they
rUST/I.E WEIR Clf66T arc hinged at the base to the we:r crest, to
~vhich they are normal, and are preferably
spaced apart at a distance equal to th depth
" ; 'of water they store, so thac they no not
ovcrlal) each other when lying flat on the
cr .st. When 'fl1ey have to be fn:cted they
are hauled up by tackle, and then net-dIes
are plaCfd ,against them; the lenght and
onscquent weight of these determi:C1e the
height of the trestle. Long needles .ma\'
have to 'be Uted bv machinerv ; their section
and thus their w~ight, can· be re,luced uy
supporting them by removable walls hung
,on chains between the trestle.s. Some higt,
trestles have wooden needLs 2 to 3 feet wide
stiffened by steel angle irons; and are worked
by machinary. Som ;ti TICS the trestles sup-
ports horizontal shutters, a'ld then, on
aCCOll11t of the weight of these and th
a'1lOnnt of water pr~ssur" on them, tb·
trestles l:ave to 1e spaced cbscr. Thi~
form has not i een used in In(lia. 1t
is adapted to the case 'Where there is a
consice~a ble depth of \yater in the rivN
" downstream.
"Fif; 20 (g) Mac'1anical Weir ShuttH3. "
In America a favuorite type 'vi machanical weir is th 'Bear Trap" of which t) ere
"'tA It TItA ,. IfI!EI1f . are seyeral" varieties. In its simp1 5t
from this consists (Fig. 21) of two Ieawr"
pivoted at t]-,e base and working in a
closed cham"er; thes:. when raised for
sCClri1g storag", make '1 triangle \\ith
the weir crest. To rai.se thG leaves,
water is admitted 'under the dovvnstream
one which is thus lifted and carries the
upstream leaf with it; to lovver 'them
a downstream valve in' the chamb~r is
o'Jcnecl and as the water level decreases,
the leaves fan and lie flat in a recess on
Fig,21.
the ted. As the upstream leaf has no upward pressure on it, it is sometimes made in two
parts with a loose' idler" covering the upper one to keep out floating debris; in other forms
the upstream leaf is joined. Forllerly the leaves wer~ made of w"od, 'out recently they have
l~een built up of iron plates and fram<'s in length up to 70 to 80 feet, and are braced to secure
:igidy. One objection to this design is its cost, at the total cross-sectional length of the shutters
1<; about three times the storage depth; another is that there may not enough natural head
available to raise the trap-artificial pressure is then supplIed to overcome this deffect. But
traps have however teen successfull in use for many years and have not given trou·Jle.

_ _ _.....-F5~AT ~ W~C.H
16. Gated Weirs.
(a) Gates moving vertically in grooves. These
Weir Gates are usually made of wood for small spans
GATE.D say 10 - 12 ft. or these are fa'lricated from steel for
large spans, and raised or lowered by means of winch
WlE..R5 or a 1 ravelling cranes. Tf ese gates usually operate
in cast iron grooves embeded in the masonry of the
Piers. The general arrangement of this type of gate
is shown in fig. 22.
()) Counter-bah need Gate. Th is type of gate
consists of a gate hinged at the bottom of the channel
and counter-balanced as shown in fig. 23. This
typo of arrangement has hen used in the case of
Hydro-electric schemes when excess had to be passed
over. When the prssU'fe on the upstream side of the
gate increases due to increase in the upstream water-
level, the gate is made to tilt with its fulcrum at the
hinge. and it permits excess to pass over when it is
automatically back to its normal position by virtue ot
the counter-weight.
(c) Walton Gate. These are provided with a hinge
at the bottom about which the gate revolves as its
fulcrum. At the top of the gate a projecting lid
about 2 - 3 ft. with a slope usually 1 to 5 is pro-
vided. This slope helps in smoothing out the path
of the fluid in as much as it prevents
presenting a sharp edge. This design has been
found to give a high co-efficient of discharge
which is about 3'6 as in the Weir formula. These
gates are counter-"Jalanced and are operated through
a winch fig. 24.
(d) Sluice Gates with Rollers. Fig. 25. In the
case of large gates or gates subjected to high fluid
pressure, the frictional forces between the gate and
the grooves in which it slides. are of sufficiant magni-
tude to prevent the movement of the gate. In ordeI
to facilitate the working of the gates. the rollers are
interposed between the groove and the gate in ordeI
to avoid sliding friction. The usual method is to
Fig. 2:1. mount the roller on the gate by means of a rolle bm.
and the roller rotates through a pin which passe~
through the roller bO_li.

Ie) Stoney's Arrangement When, however, the pressure is very considerable due to"
great head d water or the exceptional dimensions of the opening. anot~er arrangemen~, viz:, thal
of free rollers, is made use of. These rollers run on axles, but there IS no pressure mduced 01
the laHar: the axles are only used to keep the rollers at the proper distance apart, and are fiKec
 111
n a frame. This roller cradle is suspenied in t'l:l grrJ:)ve un'\ttac"t~d to the gate. As the side
of the gate bears a~ainst the rollers, wh~n the former h moved, the r.)ll ~rs r ~vovle and the whole
frame of rollers rises and falls with the gate. The friction induc ~d is pur,~ roller frict,on. which"
between smooth surfac~~, is so nettling very snlll in be 1. s} m,e'} S1 t ut iti c~-e licient can be
entirely ignored, the lifting pON~r b~ing only t'le weig'lt of the gate plus the friction of the lift-
ing apparatus. Again unlier sui~a,)le conditioll3, the gate can 'oe cO~lbrp:)ised by ",eights hang-
ing from overhead pulleys. and in such cases gates of the larg:!st size can be manipulated bp o.ae
man. These free rollers are termed "StoneYs Patent Antifnction Rollers".

1.;f~.~..."
Fig,24. Fig. 25.
The dirgram Fig. 26 represents a sluice gate fitbd with. free rollers ill sectional phIt allti
elevation. It will 'te seen that the fro me of the gate on plan ju:;t clears the sluice wall. and
proj~cts beyond the greove. Thes~ spac;s are cla~ed hy staunching rods. These are simple
round rods of about 2 inches diameter, which are fast~ned to the gate, but have l.. teral play.
so that when the gate is under pressure thes;: are hrced into the COLler, e:fectually closing the
•
, '~. ItDLLOf ~'" (.

RADIAL ARM GATt'

.... ....
~.aou ....

.................
~7_." . "..
'_.A • • AIIt ...
(.«tt
en•• ,

i (
J
Fig. 27.
aperture. On the gate being raised they are carried
1 p with it. still bearing agaiFlo:;t the side of the gate
; nd the sluice wall, the friction thus iuduced is,
IO'h"Cver, quite trifling. Owing to the susperrding
J ulleys the roller frame moves haU as fast as
the gate,
Fig. 26.
 rrz
~;" .' ff) Radiai Ar~ Gaff s. This type of gat~ is illustrated m fig. 27, and IS hinged at 3I
~a' Ie pomt as s!town in the drawing. It is rai.sed Of l.)wered by mBans of a winch, and
this arrangemeat is. used where the un;tler-shot method of regulation is. desired. This,

Fig. 2S'r
type. of gat-e', therefore; finds general applintio-;:} in the regulation of Hydro·electric channeh.
The usual devices to reduce friction against be end of the g ltes usually fail due to rusting
or de}}ris eatching im the hll1t rs. The modern practiee
\1f5CA.,." W'IiIt:N is to· provi(e Ranial Arm Gates. In tl:e case oi
can'aI regulators, the gate is used behind the opening.
orr nSf 111'\1'1'(,:""""" It has 110 side grooves or roller bearing Lut . is·
4r:T'ACN4.
pivoted above downstrea.m wa.it r level as shown
in fig. 27.
17. G~te' Lifting Arrangements,
(a) Screw Gear. A simple arrangement of screw
". "4, gear is shown in fil!. 28. This arrangements is·
used for lifting small gates with small hfts. The
gate is sllspended from a rod which is screvved at
its U'pper end, and is raised or lowered by means of.
a nut to which is attached a: handle or a hand- .vheel.
The s~rew is raised or lowered bv rotation of the
hand-wneel 01' the handle. "
(b) Ec:reW Wh'eh. The usual winch used fm lifTlrii--
minor and distri;:,utary gates is sketched in Fig. 29. In
this the power is applied to the male screw, which can
be. quite short, a little over the lift in length. As sho\\ n
in the illustration. the thrust collar and plates are
situated at the rod head and consist of three plates.
the upper, the distance plate. and the base plate. The
screw is threaded through .th ~ base plate, while the
u1?per plates are superimposed. ann the who]!'

Fig. 29.
 113
jolted througb as s'hown on plan (Fig. 51: The upoer a.nil lower plates snoulli
Ie of brass, the middle distallce plHe, W;llC\ is su jeet to· no strain of friction.
lx:ing of iron. ,The solid screw rod paises
into a pipe in fhe head d which is be
female scre\v. This shoull 1e of htass'
and in some ca"es, the whole dpe is ma<'e
of ,brass but t:lis seellS a needless e:dra-
vaga'1ce. Th'~ pipe hear! is held rigid ')'Y
t" 0 arms running in guide 51 .is cut in
the frame of tile stanr!ard. Thus tLe,
tor.,iona Is tr, ss is at on, e absoy', lcd and
the position of the ends o'f e,e arms ono
r.f which can "c ma('e ~o pr0j.~ct . eyo.ld
the frame, inr'iraterl Iv I]'cans of a
pointer ('n a graduate,l scale, showing t~,e
e~ act height at' "vI ic 'j t.e gate stands'
...a~:ove its sill Tre pile is rigi<liy connect-
ed with tLe gate, and from it; ring s etion
is clearlv much Letter "uited to withstand
compress;on than a solid rod of greater
\I eig h t a f metal. It can also ' e mad c of
)rdi:Jary gas or water pI; 11!lg. Tlle
introduction of ball bearings a'love and
Lelow the thmst discs would be a further
improvement.
(c) Travellers. In some canals tho
gates are not fitterl with Ii fting arra-
ngemement individually, ',ut t::e regulator
is proviLed with a traveller wmen, which
Fig. 30.
__----------- -------------
;;;;r
f; PINiOH ,.c"RHANGE.ME.NT
RACK-::=-~----~-~.< '.'---=-=-:e~

UPSTREAJ'.1

Fig.3J.
 114
is use'd for raising or lowering- the gates when necessary. This traveIIing win~ consist;;;
of a win,~h fitted over a trolley, moving on rails ov~r the regulation gates, and this trolley
is brought over the gate to be operated upon. The winch consists of a simple reduction gear
and a drum carrying a chain or wire rope for raising or lowering the gates.
(d) Gear ~inch. In the case 01 large a.nd heavy regulation gates, it is desireable-
to provide each gate with its own lifting arrangemer. t. In this case the gate is suspended
from rope-drums, winch, in turn, are operated through a series of reductIon gears as showf
in fig. 30. •
,e) Rack and PiIlioD Atrangement. This arrangement is illustrated in Fig. 31, and in
generally used for DislfJ butary Head R€guJator Gat~s The lifting arrangement coes~!s of
a worm and worm wheel reduction gear, to which is attached a pinion which works against
a rack attaehed to the gate. The rotation of the handle causes th pinion to rotate. and
these raise or IoweI' the gate depending on tho directi()D of rotation of the handle.
 PARTU
CANAL IRRIGATION
Chapter V
DESIGN, OF: WEIRS ON PERMEABLE FOUNDATIONS
1.. Introluctory.
Practically all the big canals of Northern India take off from rivers after they have
left the hills. Rocky foundations are not available for the weirs constructed across them to
pond up supplies for the canals. The weirs are usually of small height, 10 to 15 feet (as
distinguished from the dams) and are constructed in the permeable river-ced. The irrigation
engineering in India occupies a very conspicuous and poineer possition in the experience of the
design and construction of such works. The science of the designs of weirs on permeable founda-
tions have developed by leaps and bounds in India as the student will find in the subsequent
paragraphs in this chapter.
:!. Development of the Theory of Design.
(a) The law of flow of water through permeable soils was enunciated for the first
time in 1856 by H. Darcy who, as a result of experiments, found that the velocity of flow·
varied directly as the head and inversely as the length of the path of flow. [his law is
expressed hy the equation. . .

H
v=k-
L

Where v= Velocity;
H=Head
L=Length of the path of flow
k=A constant called the "transmission constant".
:Darcy's law is of the same form as the law for the movement of water in capillaries
eIlunciated by Poiseuille in 1841-42. .
The validity of this law in relation to weir design was tested by Col: Clibborn in 1896
in c<'Imection with the proposals for repairs to the damage of Khanki weir on the Chenab ri ver
in 1895. This weir which feeds the Lower Chenab lanaI, was completed in February 1892,
In January 1895, 100 ft. of the weir crest in Bay No; 1 subsided by about 2 it. This was
the first major weir to be construc,ed ~n the alluvial ted of a Punjab river, aJ?-d this damages
occurring so soon after the constructIOn, gave food for thought to the engmeers respensiule
for its construction and maintenance. But this inve,tigation, though affording the fist rational
basis for design, embracing as it did the conception of failure by undermining and by uplift "
due to the flow of water through the sub-soil (of the weir, did not matenally add to the know-
ledge for the p~rposes .of pr<l:ctlc~l design. It gave only qualitative indications which, however
. were generally In the nght dIrectIOn.
(b) Hydraulic gradient tht ory.
. TheIiydraulic Gradient Theory for Weir Design, apparently orginatod between Sir
John Ottley and Thomas Higham and was developed as a result of experiments by Col. Glibborn
\1895-97). • . .
 116

With the pu'Jlication of the results of Col. Glibborn's experiments in 1902, the
Hydraulic Gradient Theory came to be generally accept ~rl in India. The following passage
from Buckley's Book "Irrigation Works Tn India" (1905) Page 175 will bear te:;timny to
this: -
"rt has been mai1Jtained that, in those cases where th-o chief danger to a weir is from
under scour and not from parallel currents, the true measure of security of a weir in a per:ne'able
hed is the distance through the soil which a current of wat'or would have to travel before it
could rise up below the weir, and that it is of little conse1uence whether masonry is laid
horizontallY on the weir be,l or sunk vertically helow it, so long as the curre::J.ts passing thlOUgh
the soil below the structllre are exposod to the friction of the same length of passage. This view
appears to be sound, but it is ess~ntial to attach to the application of this principle, the
condition that the weir must be protected from horizontal scour on the face of the toe, that it
: must have sUTIcient weight to resi~t the horizontal pressure of the head it supports, and also
'sufficient weight to oppose the up'Nard pressure in the base of the foundations which that head
may produce and, fui-ther, that the surface of it which is exposed tJ erosion should be of material
sufficiently hard, and sufficiently heavy to resist that erosion."
(c) Creap Theory of Bligh.
In 1910, W. G. Bligh enunciated his creep theory in his book "Practical Design of
Irrigation Works" and elucidated further in his book "Dams and Weirs" 1918 He did no
~xperimental works, out aTlalysed tIle failure of the Narora and the da'llage to the Khanki weirs.
According to this theory, the cr~ap was marc important and destructive to undermine a weir
than' th~ percolation through the soil b~low it. The tnginneer:; haVing the experience of closing
hreaches must have felt the truth of the statement when they find that it is most difficult
Or rather impossible to close a leakage by the side of a log of' wood or . roots in the soil. A
oank of a distri~)uting channels, oth~T..vise strong. b~com~s unsafe if out-let walls be constructed
right across it and it neej;; a speci3.1 eJort to IT13.kB the joint between soil and masonry effective
a.gainst creep. Bligh assu ned t3e percolation wJ.ter to "creep" along the contact of the
I)ase profile of the weir with the sub-soil, losing head enroute, proportional to the lentgh of
its travel.
The length of travel in a weir profile as given in Fig. 1, would be : - I·'
_=_-L_e;;.""u' ./ue,'£Nr ,,,,It "1I0!t '" L=b1 +d1 +dl +b2+d2+~+b,
---_ "'(a_ULIC "'AD •• 1tIT LI",. .. A ~

_: --1_ m:P' ~.~! =b1 +2d t +b2 +2d 2 +bs

If H=total head over the weir the
loss of head per unit length of creep
would be :-
C= __ . _ _ _ _H____"_~.
..
.' 1

~ '
,ir, "
. b l +2d l +b2+2d2+~ .
"; Fig. 1. He called this loss of, head per unit length
as Creep Co-e'Iicient· and assigned a safe
v~lue of C for different classes of soil. T~us, if in a given weir design, the value of C· were less
than the safe value assigned to it for the given class of soil, the design would be conside~ed
safe. The following were the values of C recommended Page 155 "Dams & -\Veirs"
bY Bligh.
Class I: River beds of light silt and sand, of which 60% passes a lOO-mesh sieve, as
those of the Nil~ or the Mississippi, percolation factcr C= 18.
Class II: Fine micaceous sand of which 80 per cent of the grains pass a 75-mesh
sieve, as in the Himalayan rivers and the Colorado; C=15.
Class III. Coarse-grained sands, as in the Central and South India; C= 12.
Class TV : Bonlders or shingle and gravel mixed with sand; C varies 9to5. ,
 117

Because of its simplicity this theory found general acceptance Some works desjgned 011
this theory failed whil() othe_rs· stood, dependi!lg nn the extent to. which the then engineers
ignored or took note of the Importance of vertIcal cut-offs at the upstream and downstream ends.
ld) Electricity Flow Analogy.
Pavlovsky (1920) approached the problem of the flow of water through sub-soils of
hydraulic structures from the analogy of flow of ebctricty th[.)u~h a conriuctor, According to.
Ohm's law: ---
Current C=-- Potential difference
--0·----
E A
-=B --.
- _ . . __ .-
E
ReSIstance L' p .L
Where E=Potential difference
A=Area of section of conductor
L=Length of conductor. ",!

and p=Specific Resistance of the material of the conductor. This IS itlentical'-

Darcy's equation for flow of water through sand viz:
H
v=k--
L
The work was pUJEshed in Russian. Pav]ov,;ky ac1lieved success in solving a number
of proble:n:>, :'ut as the laJoratory csults could not be shown to agree with the field results,
this m ?thod d d not i 1'l?ire C).1 i. I :nce a 'uClng the engineers and remained more or less of
academic interest. The remits 0 )tained on the Panjnad mod:l in conjunction with tht'
researches of Dr. Vaidhianat:1an onthe electric models have shown ~onclusively in Vol. No.4
Irri~ation R~search Institute, Lahore, for the first time in the history of research on the
subject that :--
(i) The distribution of pressures und~r works on sand foundations caD be exac Uy
reproduced on hydraulic or electric models.
Iii) All seasonal and other variations from the normal conditions can be reproduced
on Hydraulic models by superimposition of silt, temperature, or hoth, and
by simulating the stratification.
(iii) The pro~)lem is suscepti )b of ~athematical treatment.
The conclusive proof about the reliajility of model and mathematical results in
applicatio:l to field ccndition; marks a gre it advance and de:luite lancl-:nark i.'1 the development
of this branch of engi:1eering which has led to a final solution of the pro;)bm of design of: weirs'
0:1 permeable foundations.

(e) Khosla's Investigation in the Punjab

In 1926-27 trou-)le at the syp!1ons under the Upper Chenab Canal (Fuuja~)) became
acute. Cracks app~ar~d at the u?stream and downstream ends due to, the undermining of thj;,
sutJ-soil. Repairs wer~ carried out on the acce?ted Bligh theory, but the trou_!le persisted. A set
of pressure pipes with. weH points were inserted in the floors of th.:se syphons and th~
observations disclosed that the· pressures indicated by these pipes had a':lsolutely no rela tionship
with those calculated from the BI gil Tleory. The3e re3earches were ca.rried out by Khosla
and are embodied in the Punjlb E:lgineering Congress Papers Nos. 13~ and 142 of 1930. The
simple creep theory of Bligll was repudiated aild some provisional and important conclusions.
were arrived at, nataLIe among which were :-- .
(i) The outer faces of the end sheet piles were much more effective than the inner ones
and the horizontal length of the floor.
(ii) The intermediate piles if smaller in length tha,n the· outer ones were int!ffective
except for local redistribution of pressures. .
 11S

(iii) Undermining of the floors started from the tail-end. If the hydraulic gradient at
exit was more than the critical gradient for the particular sub-soil, purticles would
move with the flow ot water thus caus'ng progressive degradation of the su'o-soil,
resulting in cavities and ultimate failure.
(iv) It was absolutely essential to have a r~asona()le deep vertical cut-off at the
downstream end to prevent undermining.
(v) There was an urgent nece,sity for research work in the la')oratory with regard to
pressures under th.e existing and the new structures. The former could Ice done
by ins~rting suitable located pressure pipes in these structures and by maintaining a
continuous and comprehensive record of the observations of pressur~s form those
pipes.
In 1929, it was decided to extend the Panjnad Weir \Punjah). This afforded the
opportuJjity of putting in a comprehensive set of pressure pipes and of conducting full sca~e
experiments as suggested in the Punjab Engineering Congress Papers 138 and 142 of 1930. ThIS
was the first full-size e~p;>riment in the world, and the results o:)tained from it (1932) paved the
way the finC'll solution of the pro5lem. The main conclusions of 1921-29 derived from reseaches
at the Upper Chenab Syphons were confirmed and the following facts were established ;--
(i) The flow of water through the sub-soil is in stream lines and therefore susceptible of
mathematical treatment.
(ii) The ratio (Q) of uplift prefsure (P), at any point along the base of a particular weir
founded 'on premeable soil to the total head (H) is constant and independent of: - .
(1) Head (H)
(2) Class of sub-soil so long as it is homogeneous.
(3) Upstream and downstream water levels.
(4). Temperature, provided it is uniform throughout the s,_tb-soil; but it vades

(i) Silt deposit or scour upstream or downstream of the impervious floor.

I, '/ (ii) Temperature which varies from point to point in the sub-s'}il and in different
seasons of the year.
(iii) Law of loss under the floor was nearly a straight line and far the shpet piles
something like logarithmic.
Sir R. P. Hadow suggested that the field results be tested on a scale modle in the Lahore
Research Laboratory. It was not tilllatft in 1934 that a model of the Panjnad weir was set up.
This investigation was carried out by Dr. Har:0ans Lal U pp:lI following. Hele Shaw method and
the findings were presented at the Punjab Engineering Congress in Fe';)ruary, 1935 (Paper No.
ISS). The results clearly vindicated the correctness of the field observations made by Kho~la
at Panjnad.
The results of the field investigations made by R. B., A. N. Khosla, I. S. E .. Punjab,
supplemented by the work of Dr. Vaidanathan by electric methods and the work of tracing
stream lines by Hele Shaw method by Dr. H. L. Uppal of the Irrigation Research Institute, were
merely qualitative. Professor Warren Weaver, Head of the Department of Mathematics at the
UniveIsity of Wisconsin and at the time working with the Rockfeller Foundation, developed his
mathematical treatment of the flow of water through the permeable sub-soils under dams.
Weaver's work, as will be seen later, provided the inspiration for the complete solution of the
problem, which has been summarised in this chapter from Central Board of Irrigation, India
Publi.cation No. 12.
(f) Weighted Creep Theory.
. ~ In September, 1932: La~e analyse~ over 200 dams, all 0v:er the worl~, and evolved his
WeIghted Creep Theory, whtch III effect mlght be called the BlIgh Creep 1heory corrected for
 JI9

vertical cut-offs and slopping faces. Lan dropped .I weight of three for vertical r:e p and one!>
horizontal creep. While this theory was an improvement of the original Bligh Theory, it W 15
,~mpirical and lacked the background for a rational blSis for desig1.. His m~morandum on
the su~ject app3ared later, as Pap~r No. 1919, in the 193; Transactiol1s of the A'11erican Society
of Civil Engineers. This pap~r is valuable for the wealth of information aO:o11t the numerolls
weirs it deals with. .
(g) Exit Gradient Theory.
In 1929 a notable contribution came from Kifl Terzaghi (V. S. A. and Vienna). He
stated and proved by laboratory txperiments Hat failurC!s occured by undermining, if the hydra-
ulic gradi~nt at exit was in exc~ss of what he called the "tL> atation gradient". This Nas t~e
same as the critical gradient enullciated by Khosla, but it was more explicit: in so far that It
implied a state of floatation of the soil mass at the to~ of the work if the exit gradient there
eXCl~eded th3 limit 1 : 1 a1 which limit the upward force due to th~ flow of' water was almost
exactly counter-balanced by the weight of the soil.
In 1935 Haigh (Paper No. 182 Punjab Engineering Congress, L:;.hore) and Harza (PapeJ
. ~o. 1920 American Society of civil Engineers) independently pr lduced two very useful papers
on almcst similar lines. Fundamentals were once again introduced and a definite attempt was
made to break away fro',n empiricis~n and to lay a rational basi" for design. They took .!fote oj
·the exit gradiends as a controlling factor in the stability of weirs and discusse~ the distrihut_ion
of pressures which could be considered a" safe. Harza got agreement between th<=or~hcal
values of uplift pre:isure'i and those o'btained fro~n the electric models for some of the sImple
cases dealt with by Weaver in his mathematical sections.
(h) Failure by the uplift pressure due to the formation of the jump of the downstrel.tn apron
The worst condition for the sub-soil flow occurs when the downstream. bed of the riveJ
is dry and water is he lded up to the pond lev 1. This recieved attention oi'51lmost. all th.e in-
vestigatox:s as shown above. The failure due to the uplift ptessure also· occurs m m~xlmum
floods in the region of the trough of the hydraulic jump form:!d 0:1 the downstrea;m :apro~: Thf
shaded area in Fig. 2 shows the amount of uplift pressure which is of the o~er of 10 to 15
feet in the case of large rivers of the Punjab. This causes da:nage in tW0 way;, ;firsHy by rak-

. Fig.2 .' . I
,
...
. i
;
.
ing out the joints of the masonry floor and th~n lifting the klvlra'lja a1'ld secOl'!.dly" by \lif.ting th~
floor from b.;low the bOttom profile of the Nelt. The damage due to the first metho4 IS usual,
but due the second it has never been experienced, because before the full effect'· ,ofl the uplift
pressure is felt, when the flood subsides. This factor as influenc:ng the design 'otmVf;em weirs i!i-
dealt with in detail in the author's Publication No.9 (Indian Engineering Odo'oor 19~).
(i) The pressure dllIerences due to the river acting as a sin'{. ,
Yet th~re is anothe~ v~r~ impo!tant fact,or de~~rmining the stability pi a wei~ which ha~
e~caped the ~lOtIce of the various mvestIgators, rh~ flver and seepage. dr.,ins act· as strong
smks coll~ctmg the seepage flo~. .The p,:"essure dIfferences causing the flow produce uplift
pressure If the oed be cov~red With Imp~rvlOU~ floor and also cause strong floatation gr<ld1ents
in the open bed. If a pipe provided with a filter point be sunk progressively in a river or
seepage drain bed, the water level in the pipe rises, ]Jut after a sinking of 15 to 20 fet t there
is usually no rise in the water level on the further sinking, The pr,:ssure diagram become~
 120

..,
'0

, ; I
i-'
I

", I
I
I
1

, ......1-
'0
"
I j.:
,.,l
 121

,
".
'.
•
~ ')t

1i
'.' ~Ti
,.
..
0:
'II.
~
,
'
1
l
I
l-r1•
... 1

\,

....
~
;>
iE
~
:!l )
~
j~
. !'
Ii I
~
L~ :.",
I
~ ~,
~ ~
'j~
1 :
,!:
v ( ~ ::
i I Po;

~ ',.. :

1
:oj
, l-<

,
..c
~
j

-·n
1 :1
• ;1

1 ii
})

~ 1 !
I ~. f-I
'" I,
'"
~
,i

.,'1
!'
I:
il
II
1
I
~ ~ . . _j I
. . . . . L.L
II
II I;
,iI I:
II ;
r Il
 ]22

. .
1·······
,..
' '.
",::
;." . ' I0
'T.
...
.
: '.;;. 1
;. '. .
0 I
~ .
I' 1 •
+
'....! I

I ..
I
I.. -.;.- - 0

I
t -:
-
I •\I

f _.i.
~
..
,
' .. \I

~~ . T
..~
;

• i
.. .:.
."
... 4 I
r. ~ I -,. >i:
-t-;:-- . !
... • O! I ....
:1::

"
"
;.,4:
".,:

'~
"

.. I
.. 'I
..,
:;
I
"Cj
2: j
f
~ :Jl
~l!
-+•
"
<l)
~ 't:; tm
:;... '0 c::
:c rn
! .
-1
"..
(.) 0 ~

en if)
<I,)
~ "0
rn
I
~

!:: I
...... e<:> ...
.~
i
~

~ 0
ell
o
...
-,
i en
::: I !>O
ti: s .,
..:!_. t•.

':l!J
~
...
.. Qt;
ti;

-c¢ :E
I=Q

r I ....
::t

.,.I
c¢
0
.... .<1,) 0
...:::
..... ~. \::
~
t .S
.s..... i :~ .....
.
(.)
<l)

!l , (f)

,I ,t { ~,
\
'/j

I
-i "
fJ~'

~ ,I
"_'.'

.1~1 <; ~.
I';
i .,.

". J'.
I' ~

'.
r ''t
\.. ..
•.
i
~ 11 ! '
\. r f Ii
, -;-'~
t .L Il

't.
11 !~
ii t· \: •t
,:
:i
:,}
q
I,
I:
II
II
f'

I II =.
it Ie
!
,
j1 . i;'~(~ ';,
:",:,': .,'
.: .. ::,_.;;y
.. ..
~ ~.
I
..L
 123'

tf "
,' ..
~ '. ~

c:
.8
.....Q
V
if,
~
.~

,.,
>-
:Q
~
oj

.8
~
{F. ,
I
.1
:j

j .. .
, ':. . i.

I
L_·
 124
~symptotic to the highest water level recor?ed in the pipe. (This level has been defined to ~
the Basic Sub-Soil Pressure level in Cha Jter Irf P ..lrt .;) T 1! pr,nsure diJ ~r !nc~ bus r'~c )rl~lIti
above the free wate~ hvd i.1 the riv~r be~ is the pre'>sure det~r:mn.i.lg t'1e minilll~rn ttlic~n'J1S
of the gravity floor m the fiver bed, and 15 a measure to dete:mme the floatatIOn grariJe:,ts
Ttle author o-_)served 200 feet away from the downstream end of the wLir at h.hanki. ~
pressure difference of the order of 3 to 4 feet. Thes~ forces call ~ into play irres ective of the
hydrauli~ pressure by pondin~ up water against the weir . rest causing flow undi,r \\eirs and
should, therefore, I-e separately determin~d and allowed for in the wdr cesign in river led
and the design of the floors and bridges and the meterflumes on the drains. '
8. Practical Weir Designs.
(i) Rock fill Weirs.-In the original rock fill weirs of the type at Okhla in fig. 3. (a)
there was no curtain wall proj.~cting i..·elow the 'Led_ The stability d the sand foundations is
tntirely dependtl.:1t on its weight and effective base h:ngth These weirs were not wat r-ti~ht
II.lld in floods the downstr;;am aprons were very much damaged and had to i_e repaired l.very
year. Clay puddle lining of the fore apron very soon fLlllo\\ed as in bg. 3 C.) and the top
'Skin of the downstream apron was abo laid in mortar as in fig. 3 (c) to protect the loost! stone frOl1l
Leing ripped up in floods. The neccessity of deep curtain walls under the crest and otter
point lower down was soon felt and walls wt:re introduced as in }jg. 3 (c) and 3 (d l . Insr;i.e
of these improvements th. early type d weirs could not be called rennanent structures, tccaUSe
thQ exce&sive leakage. c{Hl.(inued and enormous am"ual repairs \\cre necessary after the flcod$.
(ii) PermaDent masonry weirs.-The first attempt with vertical crest and pacca fio('
. downstream was made in the case of the Narora weir. fig. 3 (c). Tl.e downs ream fiOLf
·and the previous apron were soon damaged which pointed out the necessity of an impervious
rear' apron and the insuJiciency of the oase length Vertical drop of 13 fe.:t at the crest. waS
·considered to Le excessive causlll~ river action in the downstream talus. Hydrautc gradieI,t
theory was then generally accepted.
Bligh, analysing the failure of Narora, brought out his creap theory stating that a
·creep gradient of 1 in 15 was necessary. The wtirs which followed were designed accoroing
to Bligh:s conception. Tpe su~sequent weirs had sloping downstream aprons "ieh very small
drop 7 feet and 6 feet at Rasul and Marala provided in the sloping aprons 1 in 15 in both cases
instead of vertical drop. The weirs were provided with shutters, Lut no raised crest. These
weirs shown in Fig. 3 (f) and 3 (g) were provided with deep upstream curtain walls tut Loth
of them suffered from tne undermining of floor due to insu:ficient Lase length. The apron at
Khuki was only 108 feet long, while It should have teen 195 feet, allowing creep coe:ficient
of 15. The Marala weir has 3 well lines which are pretty deep but it is doubtful if they are
water-tight in between the wells i~ecause this weir has Deen constantly giving trouJle due to
,tIIldermming. Khanki and Rasul Weirs were re ,uilt in 1934 and 1931 respectively. The failure
was nearly complete by undermining due to the well lines not hing water tight and no
provision of sound end curtain walls for cOlltrolling the graaient. They were hollowed out in
-course of time and the collapse came in flood protally due to the uplift pres~ures caused
in the trough of the jump. Marala weirs actually did not give way, but in 1930 floods, the •
Kh.ranja of the downstream floor was ripped in the region of the hydraulic jump. The pile
lines upstream and downstream have now been provided and the ~ctions rebuilt.
A beautiful example of Bligh weir is provided by the Ganges weir at Hardwar shown
IIIFig. 3 (i). There are shallow curtain walls both upstream and dowmtream and the gravity
doors arc provided according to Bligh's conception. This weir has stood very effi~jently the
te,t of time. The unprecedented floods in the Ganges in 1935 left the weir very nearly
IlDscanned.
(iii) HaveU Project Weirs in the Punjab.
Two ?f the Sutlej Valley Project weir sections are given in Fig. :3 (i) and 3 (k). The
era of sheet plIes started. At sulemanki, there was a clay layer below the weir and it was
considered enough to put only one line in the begining. Panjnad weir Fig. 3 (kl shows thr••
 125

ery deep sheet pile lines. They were gravlty weirs and in some eases top KhaTanj(l, Wilt
;:'placed by lightly reinlorce.d concrete ~laJes to withll~and effectively the hydraulic jump
action on tnt Hoor. SutleJ VaHey Project weirs. 4 In Dum,;cr, stood extremely WtlU
except the one at Islam which collapsed due to the retrogression of river levels downstream.
A comparison of the last Bligh's weir in Fig, 3. (iJ and those which followed as in Fig. 3 (k)
J (j) and 3 (1). sbows that tte introduction 01 deep shH~t pile line introduced new factors whieh
did not exist in Bligh's weirs. Bligh's weir had shallow curtain walls, 6 to 10 feet ceep, upstream
~ ... d downstream and even the inside corners were filled with concrete. Bligh's cCfap theory could
;;ry well apply to shallow curtain walls. He could never imagine that his successors would put
in sheet pile lnes 40' to 60' d~ep. It is grave injustice to say that Bligh's theory had reen repudf'
a,tefl, or that his conception had reen proved absolutely '\\rong. Naturally with deep cut-ofts
c:reep 'Would re v ry nearly impossible, the loss of head shall have to follow some other law. It
b.as already re(>n stated that R. B.. A. N. Khosla, I. S. E., Punjab did the nscessary work in this.
direction. His theilfY to calculate pressure is explained later in this chapter. Even now
the law of loss under straight floor is very nearly a straight line just like Bligh. and it is only
in the case of deep cut-ous that the change has come in.
(iv) Khosla's weill.
The Panjnad and the Emerson Barrage Fig. 3 (k) and 3 (1) are eeantiiuI examples ot
Khosla's weir's. Emerson Barrage is tho cheapest weir which has so far been constructed.
in India. Tt is not a gravity weir. Th. floors are of reinforced concrete slabs to resist thp·
uplift pressures and are weighed down by the Berrage piers.
Khosla. has brought out clearly the importance of exit gradients which had beeR
the great source .f trou ,Ie in the past weirs and hilS supplied the engineering world with a
correct ~ethod of calculating the uplift pre&Surilli with d~ep cu.t-olb, and of deslgning weii' floor~
conomlcally.
e
4. Math,metl. te .etermine uplift prMSures.
A very mnch abridged summary of the
formule is taken from Chapter VII, Central
Board of rrrigation, India, pu~lication No. 12
by R. B., A. N. Khosla, 1. S. E., Dr. N. K
Bose. Ph. D. and Dr. E. Mckenzie Taylor,
O. B E. B. D. Sc. The student should refer to
the original book for the detailed knowlerlge
of the sU'Jj'ct and the proof.; of the formulae
stated het.in.
It is generally known that the stream lines
of flow under a floor AB as given in Fig. 4 art'
confocal ellip,es with the centre at O. ~b6
Fig. 4. middle of the floor AB (=b) the major
axis and with foci at A and B. One of tt:e
stream lines will be AB, which is the limiting form of this family of elli~es.
The ellipses are given by the equation
X2
-(-~-c-O~Sh~U r-- + ( ~-
y2
sinh u r =1.. .......... (A)

..hen 11 is stream line ftmction.

From this equation any particular stream line call t;e detffcmill~d hy giving II. iait..b~
Yall1e to u, So that equation (A) will take the form : - _.
Xl yl
A2-+-SI-=1
 126

The stream line along the door AB is given by putting

u=O
from which cosh u= 1
sinh u=O
and equation (A) reduces to y=O, that is the line A B,
The line of equal pressure are at nght angles to the lines of flow and are in the above·
casE' given by

(~
x2

cos v r =1 (B)

where v is the pressure function.

From this equation any particular equi-pressure line can be obtained by giving suitable
value to v, so that the equation (B) will take the form : - .'
_!-~ __ _y2. =1
AI2 Bl2

,\t the upstream end A, V = 1T, the total head or pressure,

Substituting this value in (B) we get y=O, or that the full pressure line is AX' which
is horizontal and in plane of the floor at the upstream end,
The entire head or pressure V=7i' is gndually lost along the direction of flow throug~
the sand medium till it is finally reduced to zero at its exit along BX, The ra:e of loss 19
given by the spacing of the equi-pressure lines shown in Fig. 4,
If P is any point in the mediu-n, alld the pressure and stream line functions at this'
point are given by u & v, then the relationship (A) and (8) can be expressed as below;

Z= -.!:>_ cosh- 'if) (C)

2
where Z=x+iy and w=u+iv
The position of a p:>int P in a phle (5ay Z-platl~) is usually denoted by the two
i::artesian co-ordinates x and y. It is convenient to denote this numryer pair (x y) by the
compound symbol (x+iy) , and this number pair is conveniently called (after Gauss) a complex
number. In the fundamental operations of arithmetic the complex number pair (x+i'O) may
')e replaced oy the real numJer x defining i to lnean (O-i'!) we have i2=(0,1) X (0,1)=(-1,0)
and so i2 may be replaced by-I.
:. i:l=~l

5u')stitnting these values in te) we get

x-Hy= ..~ -cosh (u+iv)

c= ~ ~h u cos v+i sinh u sin ~

~nquating the reals with the reals and the imaginaries with the'imaginaries we get :~

b . hu' cos v.
x= z-cos (E)

and -z-
y= b sm
'h u sm
. v (F)
 127
or cos v- x
_b2 -cosh
... u

sin v=-,.;--,---,,-Y_--
i) • h
-2--sm u

Squaring and adding we get:-

__X_2_ _~+ ,'f·· y2
cos2v+sin2v=l=

which is the same as (A)

( -}-C9Sh u r( +-sinh u r
..

'\
Similarly we can get equation (H) as ;-
cosh 2 u -sinh2 u= 1 x2
( _lJ-cus v ) 2

In the case of a si.nple floor the pressure at any point along: it (whete Y'=O) is given
by equation (E) by substituting:-
u=o
b
or x=-cos V
2
or 'cosh 0== 1
orl, ".,:";"", . f '
,_:,~'., -, v=eos -1

or as V='ncP. cP= _I_cos -1 (2x) G -

7r x
General Form.-If under a floor we introduce a vertical obstruction like a line of sheet
pile or wells, the configuration of stream and pre;sure lines of Fig. (4) will be distorted. But
it is possible to bring [1ack this distortion to the normal configuration of Fig. (4) by means of
a transformation known as Schwarz ChristoHel transformation.
The fundamental general equations for pressure distrIbution under th;, foundation
profile as sketched in Fig. 5 with the sheet pile at any point under the floor may be summarised
as below :.-

Fig. 5,
 128
c..... .....,...
L .+. = d,-.,. ,()..coS'Y-l'<~-I_d.. -dl ..."f(ACO'Y~Aa)
ty trK V fr 'It L
+ j I,{coo v- A .),-'1 } +;ld,-d,)

O. Prtl5.r••t E :-

m.
F. =H-
Preuure at C : -
.. c:09 _ i (->")1..1_1)

Pc = =cos _1 (A~+l)
.... ,r_are at D :-
PI) = ~ COl _I (A~It)
Where J{ =«=05 e i. ghen by
taD '-9= .. &

ad )1..,= Lt-Ls
2
. . . . "- 'RIM of ~ aDd L. are given by equatbt.1t~, ~~,~ L.-~~~"\1
GE at
,
c= _ip~
dy
(exit)
,,-,
tI '" \

" ,
" r", '. •

'-,

y~, U(_!_),
d1-d. d,=d.. =d
==_1
'I7d
W-: ••all now deal witb particular C'ases 01 this gellttal lorm. , .M ".. _" Iftaift~
ClDDCel'Ded WIth the values of ,the pressures· at the key points, ED and C aa~U,* ,...... .. tIae nt«
Ifadients we shall obtain. onJy the values of these quantities. ' . '
CJase 1.
Floor with pile line not a:t end Fig. 6.
f-.- - - , -----r 'this is Wtaver's genetai c~.
' ...... - J , - 4'
Pt '
L ......" .. __ _'-(tt
d
tI
I

fi~. 50.
• •• /
I
 ·129

Pl'essures are;,as below

, PE == _II ~os _1 ('\'C:L)
',. ~ , A ,r

Pc - H cos-
7T
1

c
(~1_:+-'I)
A
H _1
P D =- -
7t
COS ({..L)
These values of P E, Pc and Pn.re given ~.,_,~J, -_,_-.~ .
in plate II and Table I '. , . ...:.-.!!p . , t QIL-
Case 2 ,,' '. . A, , . f:-:-
Floor with pile at end ~a'i tn Fl~7.
Here dt=d:=d. b 2 =(J il\d Ql-b I
t • ti
p H-
=,,,w cos - ( A
- -2) I
E ", "A ',~ .-
H _1 A-I, )
J., ,0
and PD = -. c')s (. - -.----
w, X Fig. 7
,; Pc =
!\
'R
'-. cos
W
_1
C·,X) _"_()
)(

.,. where A=
1+v'I+u,2"b
- - - - ecause
"
2 '
, >

11 2 =b 2
=0' Values oIr ~nd PD a:re give~"ln Plate, No. 't'I A. Vo~. ~I.
at ' .', ' .
,i, Exit grapdient at C.
H I
.) GEXIT = ~7\'d . \!.\: .:.
\1alu6s ~f GE are, given in plate No. III Vol. nI. When the flooris absen:~
~
: A=
v' 1+4,2 +1 +v~~~
-~~-' '2
.
---~-- = I,as· !Xl ==a 2 =0

I H
r":X1T '::!;:' lI't<'I

'" -

~ ~:~PI: fiOO!,._k~O pile " . --$_f1"wmm&uzwny¥Z1-~"'''.~...:.::;'''

'h6'pressures along tie ba.Se are glven by ~-' ':,' I - '.
P::!!II: ·.Hw---cos
_1 ( 2) .
-D~
t.;
l<.' I
I

\.~t ,.

Exit Gradie1.t ;- ,. r t r

H _I . J; Y.
8XIT """ 7i'XO x \l A "",x : , Fir. 8
In a simple floor the exit gradient a.t. the immerliate toe is infInity. which as will h
:aplainp-rl in the next paragraph. contians a condition of deiillite inst~bility., 'fhl:! puticles a
e toe must mov~. . ..
 130

TABLE I'
Khosla', Table for Pile at end.

l."

d' ..
'-~---
,a b

'02 6,92 50 2S'545 0;063

8'<] 12'7
'04 12'6 17'9 9,56 45 l3'O06 0'066
'06 Ill:,} 21'S
..;" 11'34 40 20'S06 0-070
'08 ri't; 25-1 __ 1'2-S0 ,0' as IS-007 o'07S
'l() J~-4 28'& ' 14 12 30 15'508 0-0807
')2 21'2 30'S IS'I6 25 13010 0-088
"--/'1140' 22'7 32'9 16'12 20 10,S!2 o {}9~
-f& ,24'} ,aij'o., 1697 18 9514 0103
'>1 t+ 25'3 37-0 17'85 }6 8 SI6 0-109
'2~ 268 38'9 18'61 14 7-518 0116
'22 27-e 40'6 19'27 12 6-S20 o 1:!5
'24 287 42'2 19'80 11 11020 0130
-28- .30~ 45'2 20'91 10 5'S2 , ,.0-136
"&? 32-) 47'9 21'S4 9 503 0'142
"&6 8 ;' 453 O-ISO
336 ;;';
50'4 "_,-: .. ' 22'63 "
-40 34-g, :;2'7 23'41 7 4 04 " '0'1580'169
-44- 36'0 54'7 24'07 6 3'54
-4S 37'} 56'7 24'71 '5 ,305 ~l
0'182
'52 37'!) 58-5 2S-30 4 2'561 0'199
"liS :lS'1t 00'1 25'81 3 2'08 0'225
-60 39"tJ 61'7 26-30 ! 1'62 ---;--0'252
'64' 40'S 63'2 26-72 1~ 1'401 0269
'68- 40'9 64:5 27 1 ,8 I 1'207 ' 0'289
"7Z .I't; 65'S 27'01 lilt 1'125 0800
'7r;. 42'1 - ... ,.- ' 66:9 , 28'00 1/2 1'057 0309
....oo . ~ ~

4-2'0
" ~-....

6~N
"';" 28'3S. lJ4 1'015 O'H6
'90
~

43'7 70-7 r 2(.r21~

"':' , \) , "
)'000 ., .'" O'31S
:too 446 728 3000
IJ() 45'2 74'7 30'6S .1;,

1'2() 45 S 7tH 31'21 .1

1'30 40 0lj 77-1} 31'75
}'4(), 46-7 79'~ 32'23 • • . ,.1..

NO< 47'1 80 .. 32'70 ."

I-GO< 47'3 81'4 33')0
1'70< 47'7 82'S ;l3'5~ , :- -.I

1'8&
19t
47'9
4S'): ' '.
: ' .: 83'6 33'90 '. ~" .. ,- ,: .~
"
',' , ~!.i
84'1 ""'-' 34'21 ' 't'", .'

Z:OO' f! . .. ,',
..__ I'")-r
48"9 C'
8Ht .~~ 34'30 ,-! 4- ;~ ".4- : .::-1~·; "
~ ':
 13t

TABLE 2
Khosla's Table for pile not at eDel
p
t/>E -~xlOO
H

=100 X
1 -1 (A A 1)
7t--cos
bl b h
Y : u 2 = 'do- «=--d
2
u 1- I

h, 1>1 or base rates

-b b

0 0'1 0'2 0'3 0'4 0'5 0'6 0'7 0'8 0'9 1'0

a value of <h
0'1 3'16 282 255 2'22 1'91 1'59 1,28 0'96 O'6~l 0'35 0
()'2 n 22 567 5'06 4'45 3'SI 317 2'.';2 1'87 126 0'63 I)
025 786 7'10 6'31 5'53 4'75 398 3'17 2-:lS 1'58 0'78 ()
1/3 995 929 827 7'26 6'24 539 4'15 :l II 2'OS 0'99 ()
1)5 14'93 13'74 1230 10'8:1 9'37 7'82 6'22 4'66 :1 11 1'56 ()
() 75 21'90 2000 1800 1580 1370 11'50 900 6'50 4'50 2'00' u
1'0 27 '18 25'40 22'81 20'08 17'38 1493 11'87 8'92 5'S7 292 0
15 35'80 33'70 30'90 27'40 24')0 20'50 16'H) 12'50 S'30 4'20 0
2'0 42'48 :19'93 36'79 3H'29 29'18 2491 20'32 1544 10'24 :;'03 0
2-5 47'40 4480 41'50 3760 33'20 2850 23'50 1790 12'20 6'10 0
3'0 51 22 48'67 45'21 4' '03 36'42 31'32 25'89 19'84 13'46 6'85 ()
4'0 5698 54'37 5048 45!86 40'74 35'25 2935 2292 158:1 8'08 0
S'O tH'07 58'17 54 OS 49'01 43'62 37'90 31'76 2509 ]7'63 9'27 (I
60 0
tH' !2 61 3( 56'64 51'32 45'79 39'77 3:HS 26'76 19'06 11'09 0
7'0 tl6'88 6365 5S'&4 $'292 47'11 51 15 34'48 27'98 19'62 10'90 0
8'0 IlS'S4 65'43 ,,000 54'1:1 48'22 42'28 359:l 28'91, 21'11 1160 0
9'0 704S 1l7'OO 6!'L; 5541 49'15 43'04 36'67 297ii 21'88 12'(10 (I
100 72'01 68 12 (,ZOO9 55'96 50'13 43'72 37'fl9 :10'::1::1 2257 1Z'74 0
11'0 73-14 69-11 1)2'89 56'G() 51H7 44 28 37'fI'l :3{)' 96 23 '(\l) 13'22 1.\
12'0 74'37 69'96 63'44 57'1(1 :-10'9(' 44'76 38'3:-1 :H '43 2358 13'68 ()
14'0 76'23 71,38 6454 58'01 51'74 45'48 39'00 :12'15 24'32 14'37 0
160 7772 72'27 65'28 58'64 52'oS 46'24 :i9'60 :-1274 24'92 H'93 0
18-0 ~8'93 73'20 65'86 59'17 5300 4650 39'S4 :H~41 25';17 15 HI tI
20'0 8')'00 73'84 66'37 59'43 5:i'14 46'83 40'41 33'71 25'87 1588 0
25'0 1'210 75'10 (17'()4 6030 53'81 4746 4102 :14'06 26'45 16-68 O.
:~O-O :-14'21 757Fl.7 (,7'77 ()()'75 iH'2:l 47 88 0
41'H :l4'6:l 2(1'95 17'32 ()
 132

TABLE 3
Khoala's Table for pile not at ~nd

Floor with pile not at end

rpD ==
PD
~HX 100=~ ~Ocos-l(~l )

0'5 0'6 0'7 0'8 0'9 1'0

bl 0 0'1 0'2 0'3 0·4
1)

a value of 90
50'00 4998 49-97 49'96 49'93 49'92
50-08 50'07 50'04 50'03 50'02 49'68
0'1 50'07 50'00 4992 49'87 49'81 49'75
50'32 50'25 50'19 50'13 49'60 49'52
0'2 50'10 50'00 49'90 49'80 49'70
0'25 5048 50'40 50'30 50'20 49'36 49'17
50'17 50'00 49'83 4966 49'50
50-83 50'65 50'50 50'34 48'56 48'28
1/3 50'3S 50'00 69-62 4928 48'87
51'72 51-44 51'13 50'72 47'06 46'30
0'5 51'62 5080 5000 49'20 48'38 47'66
0-75 5370 52'94 52'34 47'57 46'39 4550 44'54
53'61 52'43 5122 50'00 4878
1'0 55'!l6 54'50 47'71 45'56 43-63 42'00 40'80
58-00 56'37 54'44 52'29 50'00
l'S 59'20 50'00 46'83 43'89 41'23 39'01 37-50
62'50 6099 58'77 56'll 53'17
2'0 5382 50'00 46-18 42'50 39'30 36'50 34'82
65'18 63'50 60'70 57-50 3454 32'71
2'5 54'37 50'00 45'63 41-39 37'55
6729 6546 62'44 58'61 31'39 28-94
3'0 6002 5507 5000 44'93 39'98 35'25
4'0 71'06 68-61 64'75 39'06 3374 29'13 26'44
66'26 60'94 55'50 5000 44'50
5'0 73'56 70-87 44'17 38'45 32'93 27'47 24'53
72'53 67'07 01'55 5583 50'00
60 75'47 50'00 44'11 38'07 31'62 26'19 22'92
73'81 68'38 0193 55'89
7'0 77'08 5602 50'00 43'98 37'83 31'50 25'32 21'57
78'43 74-68 6850 6217 2456 20'43
8'0 02,33 56·07 50'00 4393 37'67 31'14
9'0 7957 75'44 6'l86 37,61 30'89 23'99 19'47
69'n 62'39 56'15 50'00 43'85
10'0 80'53 76'01 43'79 37'43 3066 23'48 1867
76'52 69'34 62'57 56'21 50'00
11'0 81'33 56'22 50'00 '3'78 37'35 30-53 23-08 17'84
82'16 76'92 6947 62'65 30'23 22'15 16'61
120 62'79 56'28 50'00 43'72 37-21
14'0 83'39 77-25 69-77 37'1S 30'07 2207 1556
69'93 (1!'H;; ;;628 :ill 00 43'72
10'0 84-44 7793 43'71 37'11 2996 21'78 14'86
78'22 70'04 62'89 56'29 50'00
18'0 R5'14 56'34 50'00 43'66 3707 29'88 21'60 13'99
00,0 86'01 78'40 70-12 62'93 21'22 12'57
56-37 SO'OO 43'63 36'98 ,2974
87'43 78'78 70'26 63'02 20'98 11'26
25'0 6303 56'40 5)-00 43'60 36'97 29'68
300 88'74 7902 70'32
 133

5. ThllOry of Exit Gradient. ~ . -c' .

A brief summary is.given below. from '~hapter VIII of Central B~af. . d ~f Irrjgation
publication No. 12 as mentioned tefore.·
"Weir failure from seepage flow can occur by;-o-
I

(a) Undermining of su'Jsoil. ;

I

(b) Uplift due to pressure under the floor being in excess of the weight . on the/floor.
Tte failure due to undermining.is most common, so that a knowledge of it!; causes and
of the mtasure:; to prevent it, is of utmost imp()rtance Doth for design of new .works and for
ascertaining th~ safety of existing ones. Even in the Second case if the floor be burst due to
excessive pressure the fmal failure is due to the reduction of tne e~fective bngth with the conse-
quent increase in the exit gradient. '
The unc'ermining of the sub-soil starts from the tail-end of the :Work. It begins at the
surface due to th ~ residual force of seepage water at this end being in exc{ss of the restraining
force" of the subs)il which tend to hold ,the latter in position. Once the surface is disturbed
tl,e dislocation of sub-soil particles works furtner down;and, if:prog<J;.}ssive,lea.d';; to ,the forma-
tion of cavities below the lioor into which the latt~r may collapse, According to the commonly
accepted ideas. this undermining is supposed to result from what is knownas "piping" that is, the
er.)sion of su i-soil by the high velocities of flow of water tllrough it, when' such velocities exceed
a certain limit. But as will be shortly explained, this conception of undermining is incomplete
Water has a certain residual force at each- point along its flow throught ,the su')-soil which: aclSj
in the direction of flow, and is proportional to the pre:Bure gradieIllt at that point. '. :Au:the tailt
end this force is obviously upwards, and will tend to lift up the soil particles, if it is more than
the su' merged weight of the lattpT. The frictional resistance, cohesion etc., of the, aoi'!-Cent soil
will have to be considered in certain Cases. Once the surface particles are'distutb'id.1he resistance
against upward pressure of water will be further reduced, tending to progressive disruption cf the
su>-soil. 1 he flow gathers into a series of pipes in the latter ca,e and dislocation of particle~ is
acC: lerated. The su·. ,-soil is thus progressively undermined. Soil erosin can also occur thrOUgh
natural pipes, or faults in the sub-soil. '
As water flows through the sandy sub-soil, it is welt known that the velocity of flow at
JiJ .... H
p,oint any in them~dium is given ,by Darcy's Law :- v=k-
L < ;

, Tris watpr'exerts a, force (F) on the sandy medium along its line of flow. Besides this
force, the particies of the mediu:n are subject tq two other forces.:;-. "
(I) the force·-of grawty, ot weight {W)which acts downwards,
,j:
'anti' (2):.the :i?rce of buoya;ncy (B) which acts opposite to gravity. These two latter forces
;ac 1;e comomedillto one and If: : - : ' ;.'!" ..

'rw ,= Weight of unit volume 1twater' ) ," I': ::o,:_.

p= Specifk gravity ~f ~and pa~ticJe3 lit l ; I!; " ,

; " " - : .1 - f f" ': .. ' • " ,\.".~. •

'l;;; = Pore space in unit volume;thell the ~ight of sail p~rtieleS' per unie. volume
=W(I- ~ )pW ~'" . ' , .. ' " ( : " "
.,' , Theweightofdisplaced~ater~V(17~)=:=B·. ( ' ..
Hence the resultant of ,the forces, that is, the weight '~f tk('> I~Oa" particles less I buoyaney~
, . Ws=w(l- ~ )(p--'- l:) [ i : I • .' . . . " ' " I: . ,

, ',' This ma~.be, ~alled, t4~; I)uqmerged, 'Yeight \Y~whlch win obvlol\-Jly\a:lw~ys ~~ qo~nw1l~'1
LS P IS greater than umty. ( 1 _ • ..... I' ~ ~. ,, ' ..' '.,.<" ,:: !) I: .: :': '
 134

. .,. . The next step is to determine Hie value of i .

I Let us assume a cvlinrler of soil along one of the
L ; stream lines ,as sho'wn Fig. 9.
~ Let rla=sectional area of cylinder and
-~ --! dl=length of cylinder. .
, ", The pressure at face A = p da and ae~ along the'
\ '
\, P alonge stream line from left to right.
\.
, .... The pressure at face B= (p+dp) da and acts
along the stream from right to left.
Neglecting the curvature of the stream line we
. "'... l get the force acting on the cylinder along the stream
line as
Fig. 9.
pda-(p tdp)da= -dp. da

.. The force per unit volume dl. da=_~p .

c£ the cylinder=- dp.da
. ill

Hence F=-~£""",Ptessure gradient at that point. So that the action of water on any
,dl '
pomt in !he sandy medium through which it percolates can be given by a force F, which, per nnit
volume, IS equal to the gradient of pressure at that point and which acts in the direction
of flow.
Thus the force acting on any particle in the sub-soil consists of as shown in Fig. 10
•,
_
-!~ r---
r-~ ---l~ If'

_1ii:_11 '=iZi::::!l22,::::t:z:::,=::%I,
Z:::Z
, :, Ii
-.1 ____ _
(a) a force F = -<!p acting in the direc-
dl
tion of flow, and (b) the submerged weight
I W.=w(l- \il) (p-l) acting downwards.
I
, /
I
For stability of the particles, the resultant
..., R of F aud W. must have no upward
component.
Now the vertkel component of R is
Jgiven by

Fig. 6. From Fig, 10 Rv In the region

A to C must necessarily be downward as F has no upward vertical component.
From C onwards, F has an upward vertical component, which goes on increasing
rapidly as the stream line approaches E the exit end. At E, F is vertically upwards; at -
X' it is infinite, beyond which point it rapidl falls off as shown in Fig. 10, \ Thus for some length
beyond X' at The surface, where Facts verticalh upwards and Ws verticallY downward, the
resnltant Rv= F-Ws will be vertically upwaros, 'so that, the soil particl,s in this region will be
1if ted up and 'be in a state of unstable equilibrium. This region of unstal,le equilibrium will
extend to the limt where R=O or F = Ws. At thi5 point the soil partiCles will be just stable,
and F acting upwards will have its critical value whlch will be just· resisted by the submerged
weight of the particles acting downwards. The slightest increase in this value of F will lead to
instability. The soil particles will start to be lifted UPior "float" so to speak. The gradient of
pressure at which this occurs has been aptly. called by Terzghi the "floatation gradient." Haigh
calls it the bursting gradiE!nt. Another auther has called it the critical gradient. Thi5 is
equal to the critical value of the force Fe and is given by'.
 ... ualauon gramento:: l'critical =\V. =w(l-E)(p-l)
If.the pore space=40 per cent and p=2;65 and:w=J Gf ={I-'4}(2'65-t}= ex 1'65=
'99, say umty,
, Th?s for th~ class ?f soil me.ntioned above the floatation gra~ient Gf =unity or In..
Th!s floatatIOn gradIent Will vary wlth the pore space of the sub-soil and .the density of th6-.
soil particles. The density of the '!joil particles though generally in the neighbourhood of 2'6 £;)'
271 for the Punjab sands may vary in" extreme cases from )'g to 28, Similarly. the p::>re space
whir.h for the Punjab sands ranges between 37 and 42%(may he as low as 20 per cent and as
high as 45 per cent) Working on these limits. the extreme value of GI may range from 044 to
l U ' ~
Table for Gf

Densities
Pore space --- --.- --_ ..
2.8 2'6 2,. 2'2
., .,'

. ,2-0,.' . : 1.8.

",
,',
0'20 1'4( 1'28 1'12 ' 0'96 0'80 . ,_ o !,.1; t:
0'25 1'3; 1 '20 1'05 0'90 0'75 061} "
0'30 "26 1 12 098 0'84 070 ' ,
0'56-
0'35 1 '17 1'04 O'9( < 078 1)'65 0','52
0'40 1'08 O'tl6 ;" '0'84 0-72 0'60 0'41'4
0"5 0'99 I}'S8 0'77 0'66 0'55 0'44

, The theoretical value of the floatation gudients is given by the formula at this
page and the table above, but' in actual weir de.iign there are of her fa'ct.ors which cannot
be allowed in theory such as the formation of scour holes and retrogresc:ion etc. All the
under mentioned factors increase the value of critical gradi~nt:-
(i) Scour holes can be caused by the formation of swirls by a tree sticking in the bed har
the downstream end 'of the weir.
(ii) The exit gradient at the lower end can be increased by the formation of the local
surges or waves which resemble the action of the intermittent pump. The rising and
the ebbing of the waves produces suction on the soil.
(iii) The su.-lden application !1nd reducti<?n of head also tends to increase ,the' ~lue 01
" the exit gradienf. " "
hn<' (iy) The 'pr~sure diff~rence. betwee.n th.e fr~e water level in a drain Qr a Si~K and the
! 1' Basic Sub-Soil Pressure level proouce. flow into, the river bed causing floating
., , gradient and critical flow disturbing the soit praticles.
(v) Retrogression results in th'e lowering of the downstream bed level, with the conse- .
quent increase' in the value -of the exit gr'adient. ( , . , ' " • , .:' I

, . ' To allow for such unc~tain river.conditions; a -factor of safe~y is a\10 ~ed' as given-:-
i>elQw'ifi fixing the. value of the exite gradien1:-, \~~~ .;.>

~~, tr.; Nature of soil :<.Ii Fa.ctor of ~afety' Gf

Shingle. 4 to 5 -.. "25 to '2
,., Coarse sand .H'T'" ''tf!l'lV-C; 5 to 6 ·,li(1 ,. " .' fh " '2 to . 17
Fine sand ..." ' ~i)l.~ 6 to 1 . : ni -17 to '14
6 f Sheet .P;)~. .' ",,~, .. ') ~ ,

f ,~ .• ~ j , :- .

Sheet Pile Section; "

In the modern weir the sheet piles replace the curta.in walls of ma~Qnary well lines or
)1 B~i.gh's w~ir.,_be~au~e they can· be driven to a~y Qepth ~ith~t ·the; t:edious -pfQcesli of well
 .~~iK:;>>mi~t~;>\?,tr{i"'.n ~;~lZ' >":),~;~i,:., ~t~~t®C~',!\,; (';1-,'
, ~' lIt. e., 134, t f " "" t'tl~ Tr.:1"" ,;::f;
, .,. '- ,:1 ~ "t! J

.$iilkirig, or the.very'e:lCpen~ive unwatering required in case;o( curtain wall. ,..A typical section of
sheet pile is shown iil:.fi~.:t... Th~:fliles are interlocked as <~hown therein.,. (
To aU intents and purposes they are supposed to be water-tight,. M

,. ,._ '''~'.' ............ .-

"

I
.~.:; "~.ilJ.'.,.w,,''\i,.
- -""tb} Depth of 'sheet pUes: . >, -_ •., -.,.. " .. , .,_.,_,.... - .... - • - 1 • •
::1 'Tn"'pa:ragf:lph No: j'-ehapter V formulae have been given to d~termine the depth of \
normal s~ur; Thi'l sCour' d~\)ttrus'uafty 'c'a1c'iUated froi\n::-acey-"sToriiiuTi' aTe somewhat higher
than those from Kennedy's. It is on the saf~ side to acc~p~ Lacey's figlHes. Accord\ng to
Spring scours are classified ~s: - '
(i) Normal.scour, 30 to .40 !ft. caused by the unobst(004ed currents. i

: "" (ii} Abnormal 80 to S:,flt,' ca.u.sed bf 'the defJecte(f~·'·bu.t nevertheless onwards ~oving
- 'Currents."" ' ., ",' , .~. . _ "
" ,
(iii) ExjraonHnary. 80 to 100 fLC"tl~!>Cd b¥, swirls set up in forwanhnoving water.
Lacey has given a clearer definitjol} of clllsses of scour.
, . . I' .''1 •
Let R be t!:le depth of scour in a channerIlowing straight with level oed. .
class A Straighi reach ' ': ) ;:2;;- ~ ,'.
i: i tClass
·
B . , , ' Moderate
'1(' Ii, ~;t .'\fiJ I·.$';~}i ' "
Bend" <'
"L' ;"w,.;' 1,.,
;'..,.,:,.~,'x.
_,
J:,5, R i "

, .lI ·!lItN~ ,)Class,C ", la~;,r·~:;fil.W Severe ,Senti t· f l _ . , ; .' ;tlif"iw 1'75'R ;.;
Class D ,J;if;llt ;Right-an'gle(l ~~: 1!gRgi - ~JL):0I»2:0 "If t
., ~ ~ .~ t
• I '),
,::' t ,.~ ~ "t.', ;:!!~ii,:ijf"" ,~r"f< :1'.!, t H ... • .. '

Class A is likelV to ocC'ur just below or upstream of loose aprons and· Class B along"
the aprons of guide banks in a straight reach and C & D at the noses of guide banks, The
top level oHhe apt;on!i shoul. be u'fed allowing SC0ur; -depth ,'aceording to', I<enneliy's and the
b\,Uom;,of :the sheet, f>iles loollid go. below R as calculated from Lacey's ,' formula Plate III.'
Ccl F~netton of upstream ahd ~oiriistr9im 'pile li~e. " ' . ,.
, . - -~dour hol~~ c~in occur Doth upstream and donwstream so that ~a pile line is: req'uired at the
up!'\tream end of the tJoor just as at the downstream end in order to prevent failtlre by slipping of
the sub·sQ.i] into the sC!)ur _holes by simple earth pressure. In the case of <I; sco~r hole at the Jlpst,r~am
end, the flow of seepage water wlli tend to keep the soil in ;posltion; as Its force will Oppose
slipping due to earth pressure. 1\s the downstream end this force will assist and accelerate earth
slipping du~ to earth pressure as the twCi fQrc~s, will act together.
': rh~ depth of the upstream pile will be governed by the depth of SCour only and to
the rlPWQstteam one by the depth of scour and the exit gradient. l " ' ,~
Under normal conditions the bottom of the scour upstream will be higher than the
bott)m olthe scou~ dow!lstea'1l hy the amoun~ of afflux or differefic}i );l~tw~ep. upstream and
downstream levels In maxImum flood. The sheet piles at the upstream enG can therefore. be higher
t~,an.~hqs\y,a.t\th~d<;>~p~t~~a~;end)?y"tpat,.am~~t: ", , .. ,: .. ' "
, , " • ~ t ~ ,.

\ ""The ne't'ef~t)~'forpikstatltM!'~s't1'tlatn'aitd ddWr.!tre'aI'll endgli~1)"bvjoUiJ. '1:hey~are to protect the

work against undermif,ing of the foundation soH. The pressures under the dOVv(!strearri door lllcrea~e ,
as the de?tA of ~he downstream pile ,l'nt1 increa~es. The upstream pile line is relaHvety more
effective to destroy rre~sure read and 'reduce~ uplift pres ures under the weir while the pile line
at the downstream ell d· is most effective to reduce the exit gradient, '
 1:.l1
(d) Function of illtermedlate piles.
The intermediate sheet pile rJ\es are r=quired neitrer to prevent un ierm:ning of ti.t:
fl r at the upstream nor at thedownstr:am end. not do t:lCY materIally alter he pressure
d?~ribution to give l,~ss upl ft pre3sures un er t1e do vnstr 'a n floo. But tf>ey act as
i:portant seco!ldary lines of defence SO that eVen if t~e yacCIl (im,?eT ioa?) tJ. o')r, lS damaged
at either the up,tream or the d )w"'lstream end iy ,{ulure of t l end ple5 Un er a n r al
ure the rest of the fbor a.ld th.~ superstructure will e sav"!d fro n c )ILl.p'a by the 'n r-
~~dia.te piles, ~he ~Jlift pre,mro.!s, will ~n,crea5e ,mt the, uarIer niilitl~ ~m le arre ;ted as the
exit gradie~ts Will s,tlll be ,eto.v ta"! cntlc.ll ,value, . r.1'~' structur: ,will t~u> be sa\'ed fr)'n
conplete faIlure whIch on a m,l) lr w?r~ w011: li be dtsastrous, OplllI,ons dlfer ~~ to ~he:' er
there should b~ .0:1e or m 1re mter:Oldlate plIes. The best vclue wIll be oAamed wIth om
deep pile line diro::ctly under the CLost, . " ",r lj!} ~I . }t

1. KhOS}l's method of Design of ml) lern we'rs ' ,,-: h { J i (;

. ,f 'co, i fiT wino to
(a) _The two essentials to ~! considered in weitdesign, therefore, are:-
- 't :.

r! (i) Residual or uplift pres~ure under weir, and (ii) Exit gradients. ,': J je, " ' , " , " ,~
These two essentials are inter-c~nnected_ For any given fcundation profile of a wei!
in a given class of soil, there will re a dpunite distriuution of pressure eXit gradient. To safe-
guard against underrninin!5. the exit grarlhnt mllst not l:e allowe'l to exceed a certam safe
limit. generally 1/5 t01/7. The upstream pressu,res must be b:pt as low as possibb. 'consistent
with safdy at the exit so as to keep the floor thickness at the minimum. " .
(b) In paragraph (4), the mathematical sO,lutiohS, have been given for simple
cases to determine tht upltft pressures 'and the eXIt gradIents, A complete methematical
solution of a weir section with sloping or level aprons with 3 to 4 she~t pile lines is not vet
availa'~Jle_ Khosla evolved a: method of design lrn'Jwn as "rhe Method of fndipe,1dent Varia')ie's
the justification of which haS 'Deen methematically given by Dr. J. K. Malhotra of the Irri-
gation Reserch Institute, Llhore.
In this method, a complex weir section is split up into its elementary standard
forms-the mtire length of the fiu;)r wit'1 anyone of the pile line" etc, ma:-dng up on: such
form. Each eknentary form is t":~n tre3.te,i a, iurle')~nrIent of th~ others. The pres,>utes
at the key points are then rea,d o~ th~ floor of ~late n, These k o ) , points are junetioa
points of the floor and the plb lme of ~hat' partIcular e'lementry form, the bottom point
of that pile line and the bottom corners 10 the case of a depressed fioor.
The exit gradients can be obtained from Pla'te III.
The readings at tl)e juction points are, then corrected for:-
(a) the mutual interference of piles, ,,
.. c~

(b) the floor thickness

1;~d (c) the slope of the floor. "1"(I1CB-+~-" ' )-: . ' Lj<i;) 1 ).~});ii tll;'~
.-<"':' The method of spliting up a complex w!ii section is show~ 'in the catcul~tf~ns' \' ~ ·'''};)oft
(c) Mutual interference of the Piles. ;i '.', ' ;

1---- - " -_._ ---t The lputual interference of piles is w<?!,-ked-6ttt

, ; from Khosla s formula Fig. 12. . ,_," -

;_-,t:-r.F----Il--:·-·'-~-ir-t"1-
J . 0 __ '
C~ I~JV· ,~D -
! ' I " where (' =the correction to be applied ,
1 i j_ percentage of head .
. Fig. 12.
 138
b ' = distant bet ween the two piles.
D=the depth of pil:s tb~ influence of which has to be det.,rmined on the neighbouring
pile of depeh (d)
d=deptb of pile on whlch the effect of pile CD] is sought to be determinep,
and b=total floor length.
This correction (C) is additi ve for point.; in the rear or hack-water and subtrative for
points forward in the direction of flow. This equation gives results within about 2~ per cent
of those obtained by experiments and almost exact.y o·)tained from theory uy Dr. Malhotra.
But this equation does not apply to the e"Cect of an oliter pile on any intermediate plie
if latter is equal to or small~r t\an the hrmer and is at a distance 1 ~ss than twice the lengtn
of the out"r pile. Su ject to these li'1litations t~is equ ttian can 'oe a')plied to find the influence
of outer pilcs on int~rn c-dia,e ones and vice versa irresp~ ;tive of their depth or spacing.
The mutual influene of the piles is locel. It m1.in1~ ex:ends u? to a distance equal to
the depth of the pil> beyond which it gradually falls 0.1 till '.he residual effect at twice that
distance is n°l!l'gible in most cas's. Where the spaicng IS close and th ~ depth of the piles great,
the residual e'1ect may 'Je cO.:1siflera )le. Tee a love el_uatien givts t:11s inluence for all ca~e.s.
It takes note of the depth of either pile, of t:le distance between the two and the floor length,
(d) Correction for the floor thi~kngss.

The pressure percentages so o~tained refer to the join of sheet pile line with the surface
at E, C, Fig. l3. To o·.)tain the corr '.sponding pressure at E ani C it is necessary to determine
the pressure along eac':l faco at L1~ sheet pile line. L1~ pressure dro} alo.lg the face of a sheet

1-·_·-*·-·_·....-.....11--·_.,- J,. --_.-1

"'1----.----ll-.~.
.'i 1 i . ,.f
- . _ .._J
~ "" i ~/

"-
Fig. 18.
. :.)~ •.Jf{). );.1 . ) t:. If'

pile line, facing ~irection of seepage flow, at any point' disten\,bl fr?Dl ~~~ ~p{'\~tte<1:ll1; e~~e of the
floor =(70-:,~O.,ti-) per cent o.~ ,th~ totf}' .,d,rqp ,;llo_lg bath the i<we5 of the shft~t pile, the
pressure drop being assnmed to be uniform alog each face.
(e)' CO~Iection for slop.
. The prssure percentages o~Jtained from plate II Vol. II are to be correcte-d in. case -of
Sl.Oppl~g floors. The pressur~ percentages under a floor s)opir{g down or sloping up in tht'
dlrectlOu of flow ~re respectlvly greater or less than those runder a horizontal tIoor for
t!:e sa_:ne base raho~. ~s suc~ the per~ent~ge correcticns would !be positive fot the <lown and
mtnus for the up slopes foHowmg th~ dlrectlOll of fiON. The correction curve if given in_Fig. 14,

The correction applied iS~ xc whcre I is the hotizontallengfh of the sloP!IlP floor. c is a factor as
rad from Fig 14 and h' is the horizontal distance apart of the pile~.
 De.ign of wing walls.
8.
(a) In the case of Flank walls and return walls design is complicated by the fact titJfI
the soil behind. getssaturated,by the rise of spring leve~ when the river I~ses in floods. W~eIi 'thu
'ver subsides both w:\ter and earth pressures act behmd the w<\Us. It 1S usual to put m filter
floints or in verted filters conIlect~d to weep holes to drain away water arid to relieve water
·~ff."sure behind them. The provisionpf.>jwep.p holes should be con5id,er~d as additional factor of
safety and thp. wall section shoulri be . designed both for the earth and the water the press8.res'
behind the map shlJwn hereafter.
14

:t
()
;::
IJ
U
"
~ IS
II:
o
IJ

~ 10
~
;[ J-f
'"~ 5 'J
a: t-H- I"'R ',1-1ITrl 'n
I
HT
(l
Q

"
. 2 4 , , ,(j

Fig. 14.
Slope (honzontal divided by vertical)
lb) The foundation of abutment and fla.nk wall between any pair of pile lines shonld go
down to the level of the bottom of those pile lines. [f these latter are at different levels the flanks
founda.tion can step down from one .end to the other to suit these levels. This. in fact, meanS of
complete boxing in one oithe foundation of the entire weir by means of pile lines of suitable depth
upstn:am downstream, and on the two flanks, With such an arrangment of levels for the flank
fouflrl::ltions. the uplifts under the weir due to flank flow will not be more than those dlle to
direct flow under the weir. The foundation of the upstream and downstream return walls of
the Ha.nks should simi.larly go down to the level~ of the bottom of the upstream and downstream
pile lines respe _tl vely
A,. -,- . _.,Ie) A wall for dry and saturated soil. Fig 15
(il dry earth pressure
ll"., )o,lo
.\0"'\ "t_
1' .
. 6,

.~
~
U-"":""ATIa P1=w1b 1 ( ~-~
I-sin rb )
,.,~
.. I+sln't'
~.,_.,......~-n/!. U_l- when Wl =weight of dry soil,
hI =beight of dry soil,
~""''-'';'''~~;!1 y "';-zi ' C
rf =angle of repose of th'af soil.
Fig. No 15
(ii) Pressure due to saturated soil

p.=w'!
~
hz(11-S!n :)
+sln r. .
where w~=submarged weight of the soil
=(p-l) (1- e) xw fils if p=2'65 for sand 2 =40% . ,,' "_ ~

=0251Os.
4> =angle of repose of saturated soil (50 to 10° less than ~ usua.lly)
h2= Height of saturated soil
(iii) Water pressure height h2
. P a = wh s where w is weight of one eft. of water.
 - -----------------------,L.
l ___________ ---...-~ ...

. , ~,~ -,

'.r
Fig. 17 ~'. <

First the calculations will be made according to Crumps's method Fig. 17 (negltctil'lg
Ctictlon) Plate No. VI and VII, and then the complete watpr profile will be worked out
.ccordin, to auther's method as per Plate VIII to x taking friction into account.
K=19'S7
L_4-0
F =Drop in feet to the point where jump takes place
it.=upstream depth ar,d )T=Downstream depth
C=Ctitical depth==-.:3/ q~~=
v g 'V
3/ _:U_l~'
322
13'15

For ' Le' ::::'305 from Plate vr, K+F =2'1. I

. C .

~'35 and J__ =1·3$

c c
:.K+F=2 1 x 13 15=21'~ ft.
x='565 x 13'15~'7:4 ft. and Y= 1'605 x 13-15=2(,2 ft,
,

'''te VI the same values of depth 'upstream and downstream

Fig. I for level floor.
of the jump are obtained from

Drop fr.~m crest=27'6~ 19''7=1'9

Distance from the begining of the glacis=4x?'9=31' 6 say S2 feet.
Actual length is 38 it. which i's O.&. for H,F, cO:Qditions.
Downstream floor level=468
Iii) Test for half discharge' ;
211
K = - ~ X -
I )2/:1. ,"
= I~,g it-
( 2 3-1 '

Oepth on downstream floor in H,F, condition=491-46:S=~3 it.

q,=271 =a x 231>/1 . -." " - ," ,. " "," '-,
qa-=1355=kxD5' lJ/
:.D=·64 X 23=14-12. lt"f:"
l

135'5 :
,-V.eIoc:ity=- _ - - =9"20 It. per second.
1'4'72
 141

9'21)2
ha= - - - = 1'3 feet.
2x32'Z

The total Energy level downstr ..am =468+14 72+1~3 =481'02

L=(477.S+ 12.6) -484'02=490'1-48.'02 =6'08 feet.
C=critical depth=i X 12.6=8'4 feet.
L 6'PH . ,'. ,
c- = l:T-l =.744 from plate VI.

'fOp from crest level=23-12.6=IOA ft,

hich means 41'0 ft: from beginniD~ of crest,
x='4137 x 8'4=3.92 feet
y= I'S3 X 8'4= 15'4
The Downstream depth is only 14'72 feet and the length of a glacis only 38 and
.ere.fore there will he no hydraulic jump on the glacis b.ut it will take place so newhere on t~e
vel floor when depth upstream shall crop to the corresponding available depth of 14'72 thIS
m not be calcu la ted neglecting friction.
~viiiJ Author's method:- takingfriction into account.
Upstream flpw
> • '-' ~ 'III,.

It is ~ssl1med that the loose apron shall lunch down in flo;>d§ toa s!ope';of 1 in 3 and
he maximum SCOUT depth below floor level is 15 feet. . : . .~
T,.I£.• level upstream=497'2
' h': 40 f eet depth= -271 =6'S feet per second.
Ve IOCI't Y WIt· .
. 40 .
'. 6'8 2
.~--='617 feet say '6 feet.
ha ,,;
64'4 I .
Actual water level=497'2-'6=496'6 I
Flo N is above critical conJitions against a negative slope ~f f hi' 'N.
. ( 271 271 .) t "
Average ve!ocIty= 40+ 25- =9'7 feet per seconn,
C

From Plate X, the corresponding value of £='011

C in Chezy's formula= I 2g ==76'4

v' f

Yn=neutraldepth= I q2'.a f 27_!_2--~=3 35 ftOet.

~ c2s ~ 76'4 2 X i

Yc=Critical depth= I q2 ' === 7 271 2-= 13'15 ft,

~ g ~ 32'2
the equation of the water profile

L= yn_ z+Y (.1 +~). F(/'I

g s g
 144

4Q
I j let y L =40' in the begining L
yn =ZI = 3 ',:-' = 120

For 21 = 12, the value (}f F (Zl) from Plate X =0.91

L,=3'3?__ X 12+3'35 (3 +~!6~4)2) XO'91

t g
=40x 3+3-35 l3+180) x 0'91 =170+612XO-91=120 +-556=677 ft. ,.
., ,

(ii) Let Y2=33,5 feet z9=L= 33'S = 10 & F(Z2l ,91 !

- yn 3'35
",,
:, L2 =3xS3'S+612xO'9l1=100'S+S57=657'S
T 2-Ll=677-657'S= 19'5 f('Pt.
[iii] Let Z3=8; Y3 =Sx 3'35=2680 and F (Za) from plateX=0.91~7 '
. , L3=3 x 2S-S+612 xO'9137=80'4+5S9=639'4
L a,-La =657'5-689'4= 18'1 ft
.. . Plotting these three points in Fig 18 we get the last point at B from thd plate with
y=2.S'O ft.

--
:_ --
~~-~---~~~~---------------
~ ~==-t.:: ~ g;_~~ ::_.: _~BAl!_:;_r~~-: ~~J"~4
._-_- - --- ---- --- --
- ._:_~
-- -~-----
_}

F------------------------ --
~ \
4{ II.; f
r--'- 40:. _ _ '"-1 l 4'·. '..I.S
Fig
----r~----..::.• <"".~lO~'...!.._·_,_---I'--~'il IJ
.
.:.!.

. . . . . . --...... . . - _, f.
_- '-
~
- ......................~
. :l~

...; . . :
:: ......... ....~
f ~-
-_ - __
- -.......
' .,........_. -L
l~i ,- Jo:.-.---t--,- --"'- .f'.-"-'---'-1
lbJ Flow over upstream approacll.
Negative slope=i aud cement plastered

. Average velocity=t( _22!_ + .~Z_l__) = 15'6 ft, per second From Plate X, f=O 0095:22
25 13'15 .~

/ Zf! I 64'4
C in Chezy's formula= vi T--= v' 0'0095 =83 I

Neutral depth=y =
n
l_cr_-~= (
{I c2 X S
2712
832 X i
) 1=3'5 feet.';

The equation for negative slope and flow above critical

y -_ +--
L=--ll_.
sSg
Z+Yn (1 C2) F .(Z)
 ~45

(il Let Y1;= 25 It in the beginning :. ZI = Y -':""" • 2~. =7' 14

Yo 3.:)'
& F (Z1)='9148 from plate X

Ll=Y1X4+13.t-( 4+ 3;2 YO'9148) <., t

)
=25 X 4+3'5 (4+214) >t 9148= 100+ 763 ~9148= 798'0' f; '
, .~ .:

(ii) Let ZI=5'0, Y2=3'5x5~17'5 and F (Z2) ='927

L!=17'5 x4+763 X'927=70+707'3=777'3
L1-L:a=798-777'3=20'7 feet
(iii) Let Z3=4 and Y3= 14'0 :lnd F (Z3)='939
"

~3=14x4 +763 X '939=716 3+56=771'2

~1- L,= 777'3 -7il'2=6'1 teet.
Plotting these points, extend the curve to get the depth at the beginning of the crest, i1
comes to l::e 1:.'. ti feet while the critical depth is 13'15 feet. The flow against a negitive slope wit h
depth a1)ove critical cannot drop below critical depth--this indicates that the heading up upstream
will he :bigh£r thap 496'6 by aDout a foot and the total energy line cannot be 97'2 but about a
foot higher' .
(C) Flow on level crest 6 ft. wice.
This length IS too small f@r E= 19'7. Double this length shall ensure coptrol section 8:t;lQ
parallelism at the middle of crest (Auther's Puhlication No. 10) and crest length equal to EsllaD
ensure it to te at the end of the crest.
The insufficie>lt length of crest and the energy losses in fric~ion upstream of crest are likl v
to drown the weir long before it could be expected r;ecause 4 It. drop on a pnperly deesiglleq wei,
should ce su'1icient lfor depth crest of the order of 30 f')et.
, The actual position of the control section, cannot be calculated in this case according to th.
e'lluation for level floor.

" L= f(Y--{~- ) -:! ,

, " C , '. i I
, (d) Fiow on the down stream glacis 1 in 4. ; f ' ,
, " I
, Assuming that the levels upstr:)a:n shall adj~'l! themselves as ,ktated ab"ve andJhat th,
tcritical depth shall be attained at the end of th ~ crest;, '
Let depth 'in the beginning be 13-12 feet.
Value of £='01 from Plate X
c= ,12gf _= ,..,,I~'014
""
=800 nearly

Yo :% {If C2~S =(~~!:1 )1 ;"'36 ,f';" . .i , !','

The equation for water profile with positite stu,.",", dept1itWlowwi~

y
L=_o-Z-yo
s
(1
----,_
's g
2
C ) F (Z)
 1M

A\ Y= 13'12 ZI= __!_= _!_3'12 =:J'65 and F (Z..) from Plate X is '943
'"' y" 3'6 .

L1=.r_-Yn
ssg
(_1___
c~ .) 1" (ZI)= 13'1 4-3'6 (4-2001 x 0'948 X
=52'4S+-3BX J96x '918=52'48+ 706'OxO 948=72]'7
(ii) Let Yz=3 x3'6= to'8. Z2=3'0 and F (Za) ='96~
Lz= lO.Sx 4+ 706 X0'963= 722·j
(iii) Let Ya=2'5x:i6=9'0, Z3=25. f (Z3)='989
Ls=9 X 4+ 706 X '989= 734'3
L z-L.=722'2-721·7=&5

. L,,-La=762'7-734 3=28 4 , .
"l ~''1 (l":' .:. -
(a) Water pro11e on the gla,1is d~wnSt~e1,::n of the jump.
Water on the·level floor is fi;s;ed as 491
supposing glacie floor =468
Dp.pth =23 0 feet
(:1 ~:, ::f ~ ,
J)cpth above critical, positive slopl'! of 1 f.:J ,. :"<itul,) .I
Ya =3'6 ft:et as before for the slope

(i)y.=23'O feet Z.=J.2__ = 2?_ :::;6''l

Yn 36 . " v; .1:) "i'·~t~~{(, c·'~'t :f(' Jr~1
'j Jii1n:"it-:~ l() Ij:i l "<' ~1
:. F (Zt\='91S j

L=~ ~S' Z-y

II
(1- -
S
Z
C--. )
g
F (Z)
')
Ll=23X4+7C6x'918=92+648'l=740 I
(ii) ;'::5 & Y.=5x;~ 6= 18'0 & F (zz) ='927
L~lSx4+706X 927=72+654'4=726'4
LJ-!.a=740·!-7264=137 fMC
(iii) Let Z3=4, Ya=4x:-t'6=14'4' & F (Zs) ='939
I ,
La=14 4X4+706x'939=7Z0 5 '1. :L

If, ,
Lz-La=726'4·-720'5=59 feet.
Now plot both these prn!lbs a .. sl,o\f'n in Fig, 19 and find out the position of the ju np
using ettrves of plate VIII with slo;Je 1 in. 4 a~d Dl on the first curve giVeS Da on ~he
otbf>r. Fig. 19 shows that· no hydraulic Jump 15 formed in the highest flood condit)n
wit!} floor l~vel 453 jut it c lUll J ~ f Jr ne 1 at thl! p'u,i of the glads if th" downstream floor
were lowered to RL. 625, Thp. w ~ir shall work drowned with discharge intensity
!7l with upstream lp.vels shooting up frorn ~'(J t~ 4,0 feet. , ple uplift pressure due to
Junp w.)uH be of the order of 21.),0 fc!et but KuranJa dl'1lage IS not ltkely asthe flo or is
RC. slab doubly reinforced.
 (ix) Determination of uplift pressures. (Khosla's method)
The worst conrlition will be wh~n the water upstream is ponded up to the pond level Le'
RoL. 493 anrl downstream water level is at tioor level 468.
Hpad across weir=493-468=25 feet.
(a) Exit gradient.
To determine exit gradient, the weir is considered to be of an elementary form,
Depth of downstream vel tical cut~:off. d=468-444=24 feet. Total length of impervious
floor. b= 140 feet.
. 140 H 25
• , « = - =5'83, -_ =--=1'04
24 d:.!4·

From Plate III Acorresponding to ~ =1'04 and «=5'83

,
Value of GE =0'184
The critical value of exi~ gradient for TrimlllU sandy soil being ",1·,0., tb4t "factor 05
safety=5 which IS ample.
(b) Uplift pressure 1st p:pe line. Elementary form Fig 20,
Percentage pres'sures •. Percen_tage corrections.
Upstream sheet pile line. ' Thkknes~.
'. Total drop along th~ sheet pile;' line";'100-67·f
:~,
" l, : ~ :
.=.3.2'5. .. " " L

b .
_l_=O
b
Drop towards C=32'5Xl~=9'8. 'I

...
:.q,c"'F.67·S " Proportionate drop over floor ~.• thickness ·0

1'5
}'S feet=-··x9 8=0'8
19

; '.: '.~
Iatprference of the intermediate sheet pile
,I line ;-
-. ---f ,~. -""-- ""'\ D 225 .
" 'I'"
.' -=-··~=O·312
b' 72
r'"
~+t1 == _E_5.t_lJ'!" 0'286.L
."-),
1:/ 140 ..;.t
Fig 20. :. pc=30 (froW plate VI) 1:4 !Jope :-
, -" 1,11,-. .
cpc (corrected)=67'S+O'8+3 0.'
~12=70 1
~c=3 4 X _!_~::u 1'2
7'1.
. PerCl·ntage corrections due to thickness of floor as well as interference of intermediate
~heet pile Ime are additive: the eieet of the intermediate sheet pile line is to ponL! up tho
seepage flow so that tl1e pressure at C would rise.
The percerit<}.ge correction dn~ to the upslope in the direction of flow is negative.
" .
. . (c) Interme~iatp. sheet pile line ;- '
Section reduces to elementary form Fig. 21.
 ':i48
Thickness
Total drop along the pile line =20 %
b 140 , Drop along DC= 1--(0'7--0'53 xO 4)
(lC= d= 468 __-446=636 Dropafong-ED=O'70--O'S3 X 0'40
0'512
~l =.24_ =0'53 =-0'488
b 140
:. Drop along D'-:='20 xO'S12= 10'24
~-=0'47 along ED=20xO'488=9'76
b
:. CPG =38 0 ~'5
1>£ = 100-42=58 =-'
J.. c -xlO'24=1'6
'P 22
Interference of sheet piles : -
Corrected values :-- t"
<k =38-0+ 1'6+3=426 1135 32'0
.foE =58-1'6-1'9+ 1'7=56'2 CPE =19 v -7'l. x--
14()
=1'9

/'/.0'5
~1'l·"j .... ,pc=19y'-64- x
39
l4O =3'O
J.~,. __ .-I.D'r.--,- ---t . 1 : 4 slope :-
36
Fig. 21 J..£
'P
=34x 72 =}'7

The effect of down slope in the direction of flow would be to increase the pressure and a7;
such the correction is additive.
(dl For the last pile, the elementary form of the weir is sketched in Fig, 22 and the
same is used to work out the exit gradient.
Thickness correction :-
Downstream sheet pile line :-

a= 140 =5S3;~t=O Total drop=36%

24 b
:, Drop along ED=36xO'3=10'8
.' 4> = 100-64=36
., 3'5
:. cf>£ = - X 10,8=1'6
,', !
24
, ; Interference of intermediate sheet pile line:-
Correder! value: -
fE=36-1'6-2'8=31'!f!"
cf>E = I9J~64'5-446 X 28'5+(24~3'S)
64 140
1---
=19y'~8~~ X~~ =28
64 140
Total Head PCfOSS weir=2S 0 ft.

I Percentage Head,
Head in ft. of water
above R. L. 468'0
-----~,---'-""--"
Lotaticn.
I---~-- C E C
J

100 70'1 25 17'5

Upstream pile line 14'1 10'7

-
56-2 426
Intermediate pile' line 7-9 0
31'6 o
Downstream pile line , .
II
 I 149

9;' f')oor tbiekne~s. '

(aj Floor up to crest.
The downward water pressure is 25 ft. to 15'5 ft. The' maximum uplift pressure h
17'5 ft. The downward water load is more, the thickness may be 2 to 3 ftct from practical
considerations.
(b) Glacis downstream_of crest.
, Uplift pressure--: 14'1 ft.
4 H-h
t=~-­
j P
(as the pressure calculated bdow the floor. the dividend is p not (p-l)
t =thickness for gravity section.
H-h=uplift presmre
p=Specific gravity of. material = 2',25
'14'1 '
t=--- X --~=8·3 it.
3 2'25

(c) Thicknes5 of the downstream level floor.

4 H-h 4 7'9 '
t=- X - - =_ X _'-- =4'7 it..
a p a 225
(d) The gravity sections are very thick. It is a typical case whf're reinforced coneret!)
sections will be not only cheaper but also desirabl~, becaustt ma~onry kharanja has not proved
safe for the back pressure in the jump. Reinforced concrete sectinn can t~ designed as a
continuou~ slab under ,the piers which are 7 ft, thick and 60 ft. apart. It will serve as the
foundation slah for the piers and will also take up the uplift pressur", In the midrile,
theoretically, steel at top only will be required but at least half ofthe steel shonld be provided
at the bottom So that it is capable of taking up downwards loads due to the formation of
probable cavities.
10. Design of Talus or Pervious FI'lor. "
(a) The length and thickness are now no longer fixed according to Bligh's formula as
~,tated in Chapter V.
The upstream stone apron should contain enough stone so that whf'n it settles down in
floods it should take a slope of 1 in 3 with a minimum thickness of 2'5 ft. The scour bi'low the
floor level= 15ft. in thi:; case. 40 ft. length of stone with 4 ft. thick~ess will do.
(b) Downstream TalUS.
"'. This has to serve a triple function.
(i) To serve as an inverted filti~r.

, . (ii) To settle d,own to maximum SCout with a slope' of 1 in 3 with minimum t-hicknp.ss
Qf 3 0 feet. • ",C

(iii) To withstand very high velocities doWngtream of the !l)'rlraulic jump. The 511,C of
the blocks should be heavy.
In this case the scout hlow floot level is only 15 It. !llH] thrrrfore R. 1ength 30ft to 50 ft.
with 5 ft. depth will do. In addition a If'ngth 30 it to 50 ft. at least Ehonld 1 e provided just
downstream of last pile to serve as inverted filter. First a layer of shingle anti gpawls about a foot
thick over the river bed sand. then boulders about 20ft. thick of 1 It si;;<;e and then concrete
 150

blocks 5' X 5' X 5' size. In the second portion of the T::tlus which is mp.ant to sink, the blocks may
be of small sizp. i.e, 2' X 2' x 2' bFc::tl1se the velocities will be relatively low and also for uniform
and even sinking, a small .size is desirable.
to. Examination questions.
1. Sketch a de'li!!"Tl for a weir wit't shutten aeroS'1 a T,~~ai ,treltn to divert W'lter into a small
canal. \Vidth of stream SO', sirle ~Jr>n" (/2: I, bel !pvd 1)5,)'0' H'P'L' 1'60'0 Flood discharge 6800 cusecs
H. Gradient I in 12. top of banki 665.0. Bed level of canal 655.0 F.S.L. 658.0
(T. C. E. 1933)
2. (a) Skctch a design for the head of the canal in question No ; 11) givl'n above. Bed·width=10.0'
(T.C.E. 1933)
. (b) l:'esign ann give a dimensioned ~ketch of the cro~s section of a weir to be built on fine
r:aiait.cous sand IC= 15) R.L. of F. SupDlv in c'ln1.1 6.J'J.O. R. L. of bei rinr=G20.') Height of dr',? shutter~
=3 ft. m~ximl1m, discharge per foot of weir=75 cs. \Vinter di~charge of river is just sufficient for canal
·requirements and the weir is just subm'rged: when 4 fe;,t of water is passing over the crest. _.T.C.E . 1935
1
3. vVhat are the functiol"s of upstream and downstream piles line in a weir on saud foundation and
What is the objection to a downstream pile line.? (P. I B 1936)
4. 'Describe briefly Bligh's creep theory. (P. 1. B. 1936)
5. (a) You are TPquired to construct a canal Headwork.>. What points will you consider in
seJectjng the most suitable site?
(b) What observations are necessary to makp in connection with such a site?
(c) Why are trough hea1-works preferable to head works in a boulder reach?
6, State why you cannot do awcty with un1ers[uices in the design of headwork~ ? (P U. 1942'
7. 'Vould you prefer a hi 11. boulder tract for the heidwork, of a canal or the trough site? Discllss in
detail. 'Vhat points would you cou-lider while selecting site at e:;.crl of the three places? (T.C E. 1934)
8. (i) Sketch a typical lav-out fo1' a canal h"ai w()rk~ in the sandy trough stage of a Himalayan
.lver showing the position of (a) weir and under,luices (b) Canal Head Regulat'>r, (e) afflux bunds
~

Id) Divide wall (e) River training works and (fl And other works that are usuallv required.
(ii) What' pOints would you consider in sele;ting a sutiahle site for hea:! works?
(T. C. E. 1935)
9. Explain the siginificance of exit grac!ient in the de,ign of a weir? 'What do you understand by
the term unity gradient? What factors of ~afety are allowed in fixing exit gradients and why?
(P. U. 1943)
10. De,cribe ·the various methods of control for silt entering the canal usually adopt"'d at
canal Headworks ? (P. U. 1943)
11. Design a lhnk W3.\l lw,tr~am ()f a w~ir in m1.,::m"V' re~aining dry san1 up to 10 feet from top
Ilnd saturated sand for the ne·xt 12 feet to tha bottom of wall.
12. Why is a fi,b, la1 br rqu:re:! ia a wair? S:ot~h a suitl.ble d~lign for 12 feet differenc~ of level
In the pond and that downstream of the Weir. (P. U.1944)
13. D~scribe the effect of constucting a weir acos, an allUVial .river on its regime with spec:<11
reference to retrogression of levels.
14. E:xp1ain th~ w:)rkin~ of St'Jney's gd"l9 a~ U'l"d for the understuicl'ls. Sk3tch the roller bearing,
lind staunching bar arrangements and explain their working? _"ccC ~--

. '.5. E~plain the following terms as used in canal head works, (a) Divide wall, (b) Barrage. (e) Intake
or diverSIon weIrs (d) Talus (e) Pond level (f) Breast wall (g) still pond method.
 PART II
CANAL IRRIGATION
Chapter VI

Design of' Irrigation "Channels'

; . ., ,

• Introduetion.
In hydraulics the students have already learnt to work out the sections of open channel!>
using Chez)" Bazill or Manning slope formula. Tn"re was no reference to the sf)luhle or th6
suspended material carried in the flowing water. In irrig-ation earthen channels or in natural
ttreams in erodable hed, water carries a certain cargo of suspended or soluble stuff known as silt.
~

Silt is defined to be the solid material carried by flowing water whether in suspension
or solution. .
Sand is defined to be the coarse stuff generally rolled .by flowing water and deplsited in
·the bed of a channel.
The problem of silt transport in irrigation channels bafficM the early irrig~tion pngineen
and is ~tiil a subject which has not yet rearLed tLe final stage of research. Th.ere are numerou.
controversial theuries put forth !,y various authors. T11i" proLkm is very important to a canal
engineer who designs irrigation cham els which· oftm silt up and sorr.t t:mes scour. In the Cltse
of challnels which are silllDl2'. the water levels change and tte di~triL utlpn of water is disturr cd.
Thp.~ is discontent and agltatipn among the cuhivators There is shortage of ::-upply at the tails.
The efficiency of the canal liylih:m is lOO much iOlpilir, d that the frt'qUf'Dt :>Jlt c1l:'aranct's ruP. found
necfssrry wUch entail a lot d exrenCilure and lablur ~lIld tLe cem, qU(Ilt bc}tbt'ratlon. Similarly
the scouring of the cbann'Pls UL se's GistIl)' ution cDO en1:1angul'o Hie sah ty of \\orks I y ulluerllllDlng.
This troul-le has led to various inve~tigations in orcer to anive at (.csigns of irr:gation channe:l.
which should neither silt up nor scour.
2. BydJaUlie formulae.
Bffore the developtr.ent of various silt theories is uescriben, it is considered advisable to
summarize tl.e formulae fm n:acy rtlennce v.Lich an ly to the design ot sect lOllS jn hydraulit"
flow.
-~{~
\a) Chezy 1854. . , . \" . .
.u·-'r:" ;' .: .
Ch"~v, a yrench En~ineet, produced hts I\mpiricat fonfUila of
below by fittmg hIS observatIOns of some channels '
.fur,:, . .. .
~ ~~ channels ....

<t,\, V=CyR5 where 'V=mean wlocity flowy.t,:; c,

R=Hydraultc mean deP,th:,

S=~tope
C=a co-efficient
Subsequeuettt hydtautcians have been ttying tl:) work: out tbt'J math~maHcal basi~ ttl
d~rive tHs formula as below. in an or-en dianne}, tlia-> pt ~Sure is atmo~pheric ahd may, therefore
l::e nez1ect"d. heHd du~ to the slope in the channel is assu J.ed to he lost in friction. H~n:ca th"
hydraulic gradIent is equal to the slope of the caanud if t.e laHtr is uniform.
 152

Let S=slope of channel

A=Area of the cross section of the channel
pw= Wetted perimeter ".
T =mean velocity of flow
W=weight of one dto! Water in lbs ~ II Y,!,,' ,~, "'")

F = Frictional resista.ncp in lbs per second ,>

l=absolut(l coefficient of friction between water and sides of the channel pet
square foot per foot veloclty.
fl = Coefficient of friction in lbs' (per square ft. at one ioot velocity.)
Consider a section of water of length l moving along the channel as in Fig. 1.

"I'ig. 1

ASSllme slope of the channel Ulliform and equal to S.

Frictional resistance of the section for length I.
, F =wetted area X frictional resistance pel"1lilit are'a in Ibs~

F=PI.( f.~, ;: ) in lbs.

The value of 11 is taken as 2 as the average, Actually it varies with tile lu;tllre o(
abe surface and the velocita from l.7 tl) 2'3
work done in overcnming friction
Ew=friction resistance X dj!lt~nce moved
'=Fxv foot Ibs.

loss of potential energy per second . ',,' ,

= wej~htx change of altitude in f~,et '
=wi\lxvS it. I b s . ' (8'
Equatin~ ,equations tA) and (B)
, VZ
,PI fw---=wA 1 v S
Zg
" 2g
v-=f--' A
p ....
., ~:'~f~ ks.
 153

NOMOGRAM FOR KUTTER'S FORMULA

KEY Dl.4GRAM·

Fig. 2

"~I .;
-,;j ',I itt Shf' ".l

:.:'t'}l ~ . ~ '. '.' -~ ,

•~'; ~ >

, :_-;.r ",.J ~'

:':: }~,i ',\ ., ~, IJ'Jl< f:w,-~;,.

'~"::~~ Jlj't lh (,',} t. rd lo ~.uur. ,,.

"!co,:: ',,') .... f
,.' -~0

Note :..;..In friction experiments to detptmine the value of frictional resistance in 'I't;s
~er squ.are foot, {' is used to express the unit of frictional resistance. Its value in terms of f is
igiven by the relation.
i'= fw
2g- ~
 1M

/-W-
In terms of f'. the value of C=y -f'-

The Chezy's equation of the fundamentRl energv equation of fiow of witter

lind is the time-honoured practice. Solution 01 Chezy's equation can l:e got from Nomogram
Fig. 2
(b) Bazin, 1897.
From He results of his {xpeyirr.ents of flow of water though channels, Bazin decuced
tlle foll )wing formula

1575
C=--- where R=H.M.D. & k=a con;tant
1+ k
yR··
The value of k de ~ended on the nature of the channel and has the follo\\ing values

Clellr smooth sides of wood, bricks, stones etc. k='2

Dirty sides of wood, bricks, stones etc.

:Side!r of natura I earth
(c) Manning. 1886, Transactions of the Imtitute of Civil Engineers, Ireland.

Manning analyserl SOTJ'1e of the channels in his charge and produced his empirical
formula in C.G.S units which when expressed in British units takes the forms .

.,here N is the coefficient of rugosity.

This formuh is signifil'd furthff in practice ann is u~ed as :!In {'xpotntional formula
.,f discharge of a channel. Q=kd 6 ! a the solution of which is given in monogram in Fig, 3.
(d) Kutter, 187Q

Kutter, ,a~ American Engin~er, from, hi~ an~lysis of ~he Mlssissipi :River sectionsl
produ('e.d an emplflcal formula for C 10 Chez}' s eqllatlOn wherem he made C to depend on
fa) nature of bed (b) HydraulIc mean depth and \G) slape.

41·66+~81_!. +_JJgz~
N S
c- ~~--~~-~~--~--.~-~- tn ft. ffj, unit!!!:
1+'41'66+_~~~28) t&"
.in <m!tric ml.iCs

2~+_~_+j}~155
C== - - - - - - - .
J+~3+ 'O~15S V'~""
 155
(DISTRIBUTARY DISCHARGE DIAGRAM
For Q=kd5 / 3

Clisic, ••
""
" ..•
.J
""PTH d.
''1

I•

.':1.

Fig. :1
..., I" '
r _C -

..
~,
.11

If}}:
t

"0
,£XPLANA110N:-
.'t./HE THROI.ICH r:./'VEH VAL,IJU OF Q AIfO
,~ "'VES VAI.U,. 01' I< ON "HIII. IUCAI.'.
'II
'LINr. IUl)IArr,trl ,-If.OM k rHUS ':OIlNO
'SIV. R';4.¥,R". V" ..... IO 0/1 DIll'TH .. DI.CHAIN.

The gl'nerallv acCpoled values Ot the co-efficient of roughness 'N)are:

'tl'009 Weil planed timber in perfect order and alignment.
0'010 Plaster in pure cement and other glazed stlrfaces. ,
'0 'OIl Plaster in sand and cement. Iron and other smooth pipes in goad «de:!!'.
0'012 Un planed timber.
0- 013 Ashlar and well laid brickwork.
0

0"015 Rough Brickwork. Good stone-work ill fair order.

0'017 The same in inferior condition.
0'020 Rubble ma~{)nry. Coarse btir-kwork and regilneD.
0'0225' Canals in earth in tolerably good order and 'regimen.
0'025 Canals and ri-:ers in earth in tolerably good order and regimen.
0'030 Canals and riVETS in bad order and regimen.
0'035 Canals and rivers obstructed by d etri tus and in bad order and regilDeD.
~'050 Torrents encumbered with detritus.
 156

In actual practire ihe Manning's coefficient of rugosity aDd the Kutter's coefficient of
roughness are taken to have the same values as glven above. In fact they ~ the same value
()uly for one meter depth. ' .
(e) Barse-s 1916.
A.A. Barnes produceo empirical formula in his book. "Hydraulic flow reviewed"
London for earthen channels in average conditions v=58'4 R·694 S496
This wa"',the .first attempt to find out a formula where the variation of the roughness 0
the channel sPCtioll was ign'Jred and '-'. fon:nula covering I3() observations was found without N)
;j, (a) Rectang ll hr channel for maximum di~charge.

Using Ch';zY':dormula.
',_ rA~'
Discharge=Av=A.C. ,/J.{-;-=A.C.V, -. S
l'w
Let B be the ber!. width aud D as depth.
Discharge is Maximum for the sa TItJ slope when PIV IS minimum.
B
The condition is when D=:C (vide Page 163 of Lewitt Hydrulic.)

{b) frapezoidal channel for economical section.

The most economical section for a trapezoida
channel will he when discharge is maximum for a
given excavaton, This will be the case when P w
is minimum.
Neces!:ary condition (Lewitts Hydraulices
Page 165.)
(i) If a semicircle is drawn with centre at 0 (mid
point of surface width) and depth 'd' as radiuS
three sides of the section shall be tangential to it, Fig. 4.
FigA,

(ii)The hydraulic mean depth R shall be equal to'Q-where D is the depth of the channal
. 2
(c) Circular section f?r 2aximum velocity.
. . .. ' . A
Depth for maximum vp1pdty is given by the maximum value ofPwas velocity is proportional t.

VH:'_ when S is constant

D = 1'62 r; where r is the radius of circular section.
(vide Page 167 of Lewitt Hydraulics.)
(d) Circular spction for maximum discharge.
,'~, A'J
Depth for maximum discharge is got Whenl'w
"
/~ '.'
=A x CVR'S =Cyp;: S, the value of S remaining constant,

D=I'~O T, where D=depth

t=radius of the section (vide Page 168 of Lewitt HydrauUes.)
 151
4. Ker.:aedy's Silt Theory.
The above mentioned formula apply to design of (lpfn ch:'lnneT ~ection') without any
refereno'! to the silt which the W;)tPT carries in thp (,::!Tthen irri~ation eh;mnet sections. It was
pointed out in the introduction to this chapter that the research work was Vf>TV essential to
arrive at the correct section of canals transporting siIt so that they. fe r thl>' snccessful wor~ing
of the irrigation system, do neither silt nor srour. The pioneer rese::lrch work on the~e Ime~
was none by Mr. R G. Kennedy, Expcutive Engineer, Punjab Irrigation. Upper B::lrt Doab
C:tn::lI. This is one of the oIde[,.t can<lls in the Punjab constructpd in about 1850 Mr. Kennedv
selected :tbout 20 sites on the channel of this canal system in about the mindle reach He
selected channels which Itad not been silt-cleared for the past 30 years and harl becom~ stable.
These ch:tnnels were neither silting nor scouring. His observations were publishpd m "Paper
No. : 2~26, 1894·95 in the Institution of Civil Engineers, LondoJl, Er,glanol.
Ib) Sediment in the flowing canal is kept in ~uspension soIeTy hy the vertical
component of the constant eddies. which can always te ocsf'n'ed over the fun width in any
stream boiling up gently to the surface. These eddits rise on acconnt of the roughnrss
of the ted ann work UP ag::linst the depth of the channel. From iht sii1f's also, some such
eddips may occur to a much smaller degree but anv S11Ch eonies must be for the
greater part horizontal and of nosiIt supporting- power~ The SlIt supportfng powpr in a stream
is, therefore proportional to the width of the stream and not the total wet!erl
perimeter. According to Kennedy, a Regime channel is that whkh neither !JIlts
up nor scours.
Ccl Mean or the critical velocity.
The critical velocity i .. defined to be the mean velocity which will just k~p the
. channel free from either sitting or scouring. This is denoted by the letter v., •
The expression to flnd out critic:o.l velocity was found by Mr. Kennedy t.y pff>ting
observations with velocity as axis of X and depth as axis of Y. From this the fQllGwing efluation
was deduced : -
Vo =CDm ='84 D'64
This express;on gave the uen-silting veJrcities for channels ",bieh h2.d silt flF the
same. character as in the observr:d channels of Kennedy cn the l'pper Bari Dcab Can'!,l Table 2.
~ut If thl: channels had cilt of different size and gr<:.de, thty Iun nen-silting with a ve10(ity
different from Vo {mean velocity).
C. V. R. !critical velocity ratio)-CriticaI velocity ratio is the Iatio of the a~tual
velocity in a channel to the critical velocity calculated according to the .Kennedy's
Vo fonnufa :- -
Vo =x cdm
, C. V. R represents a facter wl:ich cres not wfl"ct dq)alJ..e frem rq!ime. bttt
a factor ~ h ch is a measne of the variation in the silt cenditions on Upper Bari Doab {'anal
in the Kcnnecy's ob~(;rv2.ti(ns. It is a1;;0 a factor which takes into account the scale effect
o~ t~e diir ers:o;n.s of tIle clanr.eJs. Kennedy consiceTed the question of bank erosion quite
dISt r.ct f] rm th fi1t eu! ~hT'., tlngh cf IT.c'cllhtcd irrpcrtance· in fixing the limiting
depth and the Eafe lirr.itirg .,;e1ccity of a channel. He HJDsidered tIle safe velocity in fhe·
Funj~.h sri} tr Fvard against sice erosicn as :;·S ft. Fer scccnd. This WQuld mean a depth not
more tLan 10·0 ft.
(d) Slit tr:l.nspnrtive power of a channel.
Kennedy found expression for silt transportive power in a channel as below:-
let it vary as Vo .. A=a factor
Total silt canied=A.B.v.. It B=width
'Vg =not sirting V61ocity.
 158

Let p=a percentage of silt by weight in water

Q=Discharge
:. A. B. VO II =pd and Q=B. D. v",
A. B.v!}" =BDv"
1 I
P ..-I;;::r
Vo =(1\ ) .D
This is of the same form as Kennedy's
Vo =cdar

equating the values of index of D

'64=~1_
. n-l

Therefore n=2'S6 or Say 2'5 .

.:. the silt transportive power v~ ~ v",lf' '.
5. Kennedy's Diagrams. 'ftT> '" ;:. j'
Kennedy carried out no investiga.tions to find out the cor~ect slope {orm'u1a, which
applied correctly to the design of irrigatlOn channels. He simply took the Kutter's formula.
and gaw the value of N (coefficient of roughness) equal to '0225 as the average value for all
irrigatiOn channels. He designed the hydraulic d agrams printed at the Thompson College
press at Roorkee, India. He gave diagrams for differo:nt slopes showing discharge, depth,
v~Iocity a.nd C. V. R. Kennedy's diagram applicable to varying slopes is given in Plate
~o. XI.
(b) It WaS soon found out that the value of N in Kutter's fomuIa Was not constant.
Kennedy 'himself suggested the value of '02 for large canals. Similarly in the case of small
irrigation channels, the value of N has been found to be '025. Evidently arbitrary selection
of the Kutter's formula and the value of N = 0225 was not satisfactory.
(e) The author had the occasion to work out the value of N in the Manning's formula
for the irrigation channels and found N to vary from .018 to .028 as published in author's paper
on "Design of lrrigatian Channels" in the proceedings of the Institut:on of Engineers India 1936
The variation of N in Manning's formula for irrigation channels in regime was much more than
that in the case of Kutter's. Evidently exhaustive and sound research work u'Ocessary to find out
the correct slope fom.ula. applicable to the irrigation channels in r~gim:;. This deficiency was
made good by Gerald Lacey as explained later.
6. Woods'normal data of design of Kennedy's channels.
From the Kt'nnedy's diagrams for the same value of C. V. R. lot of designs can be got
out for the given discharge as shown in an example below. A channel of 60 cusecs may be
designed as below with unity C. V. R.
Bed ,f; Depth Slape Bed with ratio to depth
215 2'0 1 in 5000 JO'n~
11 29 1 in 4000 3'8
4-0 4'3 1 in 2000 '93
Evidently all thpse designs are not suitable., Kennedy 1:t1 his note dated 25th August;
}004 admitted the necessity of fixing proportion of ced to depth. Mr. F. W. Woods, Chief
Engineer, analysed a lot of data of channels of L. C. C. and issued a table of bed ratio of
width to depth in a not'~ of 1917. Table No.3 shuws the normal.data of design for Kennedy's
cbannels as given by -Mr. Woods.
 159

The Woods permissible bed wid~n depth ratio is plotted in Plate No. XI in chain
dotted line. The suita'ole channel design will be near about this line •. If the ~ilt c'JUditi >ll5
sIlt gradE' and silt charge). are such that ::\ channel would r?n non-srItlOg wIth a C. V. R
~wer than unity say '9, relatlvely narrower sections will be rtquued and vtce versa, f)r chanD",b
requiring C. V. K mor" than unity.
Mr. E. S. Lindley. Executivp. EI\gineer, punjab Irrigation, took so~e o~servation~ in
]h::lng Division ann some III Lyallpu: D,vision and wrote ~ap.;r N'J. 4'd. m 1919. P,ll1jJ.')
EnJine:;riug c'ongres3, Lah:)rc:, on "R~gime Channels." He ;,.orkc;d OClt th)re"n the fol1ow'!l5
empirical relations.
B=3'8DloSl ( A) wnere B=bed width in feet
(: ,'r(l
V ci = .95 D·57 (B) D=depth in feet If,;) t
Vb=·59B·355 (c ) Lq = velocity in terms of D in ft; ,
;1 i per second.
Lq=velocity in terms ofB ft. pot:
".~
'.
,.

second
DM rmklimr l~lf \' rf " I.xflnfiiIMf L'ihe dellC( (ira. ( reL e[a of 6c(( wrut'ir ane( a.3pCir cUrt
p{a5
an important part in determinirg regim~ seCtions. The diJfect in h,s obscvaLons was thai: tht:;y
did not relate to channels known to be m regime. The author was in charg,; of these ct'ianneb
soon after and found that none of them was stable Howev0r, it was a valuable mass of data
and has been used by Lacey in his analysis I

7. Lacey's Silt Theory.

(a) Gerald Lacey, B. Sc., M: I. C. E., T. S. E., p. W. D. Irrigation, United Provincc~,
India, considerably advanced the subject Qf silt transportation in channels and their dcsigll
worked out in his publications given below :._ '-'
(i) Proceedings inst: C. E. London Volume 223 1928.2~
(J) do 'volume 229 193\.}-31
(iii) do Volume 237 1934
(iv) Technical Paper No.1. P.W.!), United Provinces India 1932
(v) CentnI Board Irrigation Publi<;ation No. 20 Simla, India 1939:
He has be·~n developing the subjeCt in all these publications by stages. The~stu4'ent h
advised to refer to h's htc~t publicat~on No, (V) mentioncrl above as the} a\ltho~itati:c v~r:.son
of his th<}ory eVen though It g.ves . V.EWS abd formulae very much at var ance with. hIS ongmal
conception of the problem as publIshed eadier. A brief summary is given here to mtrodnce thE
stuient to the subject.
(h) Lacey's conception of Regime.

According to lacey, the regime COnditions shall be estabUshed when the folloWin~
fundamental requirements are fulftlled.
(i) Dl~charge should be constant.
Iii) Channel flowing uniformly in unlimited incoherent allUVium of the same charade!
as that transported. Incoherent alluvium h supposed to be . the loose granular material 'which
can be sco,l!~.:d O,ll: as readily a~ it can b.: d'!posited
(111) SIlt grade (and StU charge\ C(mstant.
Ev:dently all the~e nquirements 'lre difficult or rather impossible to be fulfilled it,
nature. Lacey, U,erefore, quallfied the regime conditions as b~!ow ;_.
True Regime.--The channel should flow in a. unlimited alluvium piain of the samt
 t60
TABLE NO. t
Exponential formula of ftow of water in open channels in use
Q=kd5 / 3; Table ofS/3 power : _D5f3, i.e. (deptb)/O/l

_, .!~
. ~o,.Q~ '0 , '1 -2 '3 . '4 '5 . '6 '7 .-9
Z

0 0'000 0'023 0'069 0135 0'217 0'839

I 0'315 0'427 0'552 0'690
1000 ] '172 1'355 1'548 1'756 1'966 2'915
2 2'189 2'42'1 2'663
3'175 3'444 3'721 4'008 4'302 4'605 4'916 5'897
8 5'235 5.562
6,240 6'591 6'949
•
5
6
10'079
14-620
10'503
15'111
10'933
15'608
7'315
1I '371
16'Ill
1'688
11 815
16'621
8'06S
12-266
17'137
8'456
12273
17'660
S'S51
13'IS3
18188
< 9'254
13659
18-723
9'663
14'1::6
19'264
19'812 20365 20'924 21'490 22061 22'639 25'008
7 23'222 23812 24'407
25'615 26'228 26'846 27'471 28'101 28'737 31'336
8 29'378 30'025 30'678
32'000 32'669 33'344 34025 34-711 35'402 36-099 38'222
9 3~-941
36'801 37'509
39'664 40394 41'128 41'868 42'613 43363 44 J18 44'879 45'645
. 10 46'416 47'[92 47'973 48'760 49'552 50'349 51']51 51-957 52'768 53'585 .
11 54'407 55'234 56'064 56'905
la 57'745 58591 50'443 60'300 61'161 62'027
62'898

TABLE No.2
Table giving values of vo=O'S! D·64 (Kennedy's vol for all values
of D (the depth) from 1 to II·9.

0 '0 '1 '2 '3 '4 '5 ·6

1.' ,~ '7 '8 '9

0 a·oo!} 0'192 . 0'300 0'389 0'467 0'539 0605 0'668 0'728 0'735
I 0'840 0'693 0'954 0'994 ]'042 1'088 Jo132 }'223 }'266
1 179
2' 1'310 1'352 1'392 1433 l'472 1'511 1'549 1587 }'661
I 1'624
}'696 1-732 1'767 }'803 1837 1'873 1'9'17 2-007
·1'941 1'975
5
6'
" 2'040
2352
2'645
2'073
2382
2673
2'105
,,2'412
. 2'70[
2'137
2441
2'729
2'168
2'470
2'756
2'200
2'500
2'785
2'231
2'529
2'813
2261
!?'S58
2840
2'~92
2'586
2866
2'323
2'615
2893
7 2'920 2'947 2'973 3000 3'026 3'050 3076 3'10t 3'127 3'153
8 3'178 3'203 3228 3'254 3'279 3305 3'330
9 3'355 3380 3404
3'4'l8' :N53 3478' 3'502 3'526 3'55 ) 3'574
I() 3598 3622 3645
3,667 3.690 3.713 3,736 3.759 3,783 3,806 3.829 3.852 3.875
If 3,897 3.920 .. .3,943 3,9'66 3.988 4,010 4,033 4,055 4077 4099

.,- ~ ~-\
 TABLE NO. 3.
NORMAL DATA OF DESIGN FqR"'KENNEDY" CHANNELS •
._. # . -

1\ 2 3 4 5 6 7 S
Ratio Bed Critieal Mean
,Di5charge, • bed, . ,B width pepth Gradient Kuttelf'sN VeJocity Velocit~
\
depth = D (D) I in- . Ratio
\ (HI

2 2'() 20 10. 2500. '(}225 1'00 O-.sG

250.0. 1.0.3 1.00
, 4
6
"0 1 ,2.3
2.6
2.7
·,·3.5
1.2
1.4 2857
.0.225
.O2~5 1.00 .' :U12
8
10
12
2.75
,2.9
30
40
4.75
5{25
1.5
1.6
1'75
2857
3333
3333
3333
.' ,i0225
'0225
-0225
"()225
I.OJ
0'99
1-00
1'01
il.07
1.13
l'IS
1'22
14 3'1 5'5 1'80
16 3'15 6'0 l'93 3636 "()225 "00 :1'25
IS 3'20 6'25 1-95 3636 -Ot55 1'00 128
20 3-3 6'6 2-0 3636 'O2~5 1'00. 11.'32
25 34 7'25 2-15 3636 '0225 1-01 1'39
SO 3'5 8'0. 2'25 3636 '0225 1'02 il''46
35" 3'55 8'75 2'45 4000 "()'!2ii ~ (j'9S 1'48
40> 3'6 9'25 'Z'~5 4000 «)225 0'99 a'50
45' 3'65 9'75 2'65 4000' "()225 1'00 1'53
37 10'25 275 4000 "()225 1'00 J·.:;6
50 1'6'1
6() " ,3'S ' 11'0 2'9 4000 -0225 I-OG
70> 3'9 {2'G 3'0. 40.00 il2'25 'l-Ol .1']3

~o. 4'0. 13-0 3'20 4000 0() 225 0.'91 1'16

90., 41 13'5 335 4000 ·0.225 «>-98 1~79
4-2, 14-5 340 4444 -0225 1-()Q .1'82
10~ ,1-91
125 4'3 16-0. 3'65 4444 '0225 l'W
l51) 4'4 17-0. 3,69 4444 -0225 11 00 :21)0
175- 4'5 18'5 405 4444 -0225 I'Gl 2'07
4'6 19-5 4-30 4444 '0225 1'01 ;2' Hi
ZOO
4~7 22'0. 4'60 4444 ~225 ,l-G1 2'23
225
250 4'8 22-0 4'70 4444 -0!25 to.. :.2'30
5'0. 2~'O 4:8 4H4 -()2Z5 • , t-Oa ,:2-33
300
3S!) 5'15 26~5 5'15 5000 -(l225 . .()-9.7 :2'33
40() S'3 ,285 5'3 5000 o()225 l'O~ ;2-44
450 5'5 3 1)'5 5'5 5000 -0225 I'Gft :2'49
500 5'7 32'0 5'65 5000 O()225 • !I-OO '.2'54
6-0 355 6,() 5000 ",()225 1-00 26Q
GOO . f·Ql 2-70
700 6'4 :39-0 6'1 5000 -()Z2S
800,,- 6'S 42-0 6'3 . 5000' ,"()225 1'0l :2'81
900 7'2 460 6-4 5000 -0225 'l,()2 2'85
·1000 7'6 50'0 6'5 5000 -0225 'lOS :2'89
54'0' 5714 '0225 C398 ~82
noo 80. 6'S
<)-99 2'85
1200 8'33 58' , 6'9 5714 , '0.225
130() - 8'67 61:0. 7'0 5714 -0225 1'00 :2'88
140() 9'0' 640 7-1 5714 .0225' -{,OO 2-89
1500 9'4 675 7'15 5714 -()225 ()-9S -2'90
1600.
1700
9'8
10·t . 71'0.
740 .
,72
7'3
5714
5714
o()225
~25
1'00
100
2'98
.3'00
180.0
2000.
1.0'5
H'3
:77'0
83'5
7-35
7'4
5714
5714
-0225
-0225 . 1 "()l
101
..3'04
:3'10
3-13.'
220() 12'0' l;l0'0 '7'5 5714 il225 1'04
12'8 970() 7'55 - 5714 '0225' i'03 3-15
.2-400
!2600 i3'5' 103'0 7'65 ·571-4 "()225 1,02 .3ilS
-()Z25 o()'96 3-~6.'t-
'2800
3000
142
15-0- .
·16'1'
ll2-5
120'0
130'0
7-9
S'O
8'0
6666 .
666S
6666
-0225
-(}225
. ()'9S
.0 97
::3'()2,
3'08 •
3300
17'2 139'0 S-l 6666 -0225 0() 97 ~'3'h!
3600
1S'3 {4S-0· S'l 6666 -0225 '(}'9i) 3-17
3900
4300 19'5 .158'0 S'I5 6666 -()225 O():9S ,S'I8
4500. 205 168'0 S'2 6666 '0225 '()'98 '3'19
5000 22'5· 185'0 8;2 6666 '0225 0() '98 :3'20
260. 2150() 8'4 ,·6666 '''0225 o()-99 ,3'25
6000.
7000 29-8 ,250'() 8'4 6666 '0225 1'00 ..3'27
.
~

'8000 -- ~. '33'5· 281'0 8'4 6666 0225 '1'01 ..3·3S

9000 37-2 315'0 - 8'4 .16666 '0225 - 1'01 ,3'35
10000 41'0 345'0 8'5 6666 '0225 IO()Z 3'38
,H'O 360'0 ,S& 8000 - '0225 .(r93 :3'13
10000
15000 60'0 530,() 8'S5 8000 -0225 0() 94 .319
20000 7S.0 696'0. 89 BOOO ~~225 '()-94, .3:20
 162

grade ::IS the IT a 'erial transported. ThHe should in thc()ry he complete freeilom for latera}
movement. Sandy rivprs in alluvium plains achieve to some extent this fleedom and by
meandering adjust their length and slope.
A constant rli:::charge transporting silt of a given gr::lde and flowing in a. self trans~
p()rted alluvium pl::llll of the same grade, tends even tuaU)' to assume a gradlen t solely
determined by thp riis :hargp and silt grade, the mp<ln ve,cci'Y, hydnlu~ic mean depth and
wettf'd perimeter will tend also t() tr::lnsport a fixed regime silt grade.
Artificial channels as irrigation channels can never be in their trne regime. unless
they are neglected like rivers in plains. They may acq !lire initi::l1 regime generally or
final regime in SOllIe ca~es.

[nitlal Re'!ime: -Channels excavated in the first instancq with defectiv;l sl{)p~s and to
somewhat Ilct:-r"w dimensinDs an~ free fly immecliatPly throwing down the incoherent silt on !he
bed to increase thpir slopes anel by ~Aneration of the in~reaspd velocitv to aChieve a non-s\lt!ng
'Initial Regime. Such channpls if ber,lls are gras<;pd.\\hich is usually the rase of irrigatlOn
('hannel;;. \lill be sUllject to a '·()nsif.eraUe lateral restraint. The channels of this type in the
Tniti<l I Rfgime attain a working stability and do neither silt nor scour, but they are not in
tinal regiTT'e."
. Fi:J.al Re:;ll1e :·-It represents the c()noition~ set up eventually in the()ry when a!l the
varIables are equallv free to vary. In the final regime, the wetted perimeter, far from belllg a
variable, is constant
(c) Lacey's mean velocity relation.
I_af'ey like Kennl~dv recngnizes that the silt l!'l suspenoefi by the vertical component of
rddies, awl anvocates that thev are generated in a channel spction at all points by forces normal
to the welterl perimeter. For this reason he arlopts the hydraulic m~an depth as a variable
rather than the vertical oepth. He rliscarded his original re'lsoning in his latest publications
that R being the bed rock of hydraulicians must replace D (the ~epth).
(D In wiele channels. there is hardly any differpnce hetweeri R&D. In elliptic or
scmi-ciTcubr chann r;1 s::ction th ;r, is no side in th" tru,] s;']se and therefore R is the correct
ba,is, but Kr;nncdy c::ms'dered that the :rrigafon channel se:tions weTO trap"~z::Jirlal in shape in
which case th,'_! eddi':;s prod l.c;d fro:n the sides do not appreciably hdo the silt transportation.
Kennedy rightly neglected the side in trap ,zoidal channei sections and thcrefo"'e. he evolved
the critical velocity formula in terms of d and took the total silt transported proportional to,
bed-width. .

(ii) There is a greet d:vergence of opinion ;>,bout the shape of the irrigation
channel,. Th0 author conside~'s the irrigaton elY.nnels are generally trapezoidal
in shape wh'~n thw tra:l,port silt or sand; but when they carry
fine silt or clay, they tend to be elipLcal 0,' sem-ci~cular in shape near th'3
tail. reachp.s of the c1istributar:es and in the CaQ0 of Kh",rif cha':ln',ls. Rivers in alluvium in
plains are not ch,mnels like the irr'gation Gnes and the sha::n of their section in floods may be
el ptical even though form:ng many tributaries <"nd encroached by uJ!yie'd:ng islands and
vegetation. The term cha:,mel can hardly b1 applied to large riv.'rs of the Punj<"b subject to
2,vulsions and me:mdp,r;ngs as cxplainc:d in Chaptl;r If of tll (,)1rt. AutJ.o:, oes not consider
that one s')t of f07'mulae eouid be evolv,]d which could be app ic~.ble to large rivers of the
Punjab and th" small LTigation channels kno\vn as m:nJrs. La~c)y elaims that his formulae are
applicable to channels of all sizes.
('ii) Lacey ev'):vcd the forowing formu~a for the mean velocity (non silting velocity to
regime conditionsl for the Pllniab data of Kennedy and Lindley:-
L=Ky' P-Lk = 1.1547 ,/FI,K ( A)
v = mea 1 regime velocity
K =:l ror. ... tant
 162"

fL=silt factor
R=Hydraulic mean depth
It will he not~cced that Lacey has taken his silt factor "fL" under the root sign. lr.
lis own words 'it is pref.~ra,)le, to denote slIt gfJ.de hy a llll':~ar ratio rath~r than by a velocity
atio and for this purpDS~ th~ silt fastor "f ,. has b~en introduced by the writer.
In t~e Punj:,b c.ondi:ions, taking the value of f . . equal to urlity, he evolved the formula
,y logarithnuc plottmgs III the form
v=1'138 R·1995=1.138 R5 nearly. (B)
Critical velocity ratio as us ~d by Kennedy is equal to the squ,~re root of the sil factor fL.
Lacey's equation for silt fac :or

fL =X2='75-;_:
(C)
where fL=Lac;y's silt factor.
X=,c. V.R. (Kennedy)
v=mean velocitv.
R=Hydraulic mean depth

The silt factor is nearly three quarters of ;2 which is the turbulent acceleration and
'J
1 4
the silt fast~r when .unity is al"o ~onsistent wIth the rugosity coffici3'nt N.=O·02 ~5 f t / •
ThIs equatlOn can b~ utlLz~d to calculate the depth of mJ.xim'ln1 s:::our:; for the bridge
foundation in large rivers
2 Q 1/3
R=O'75 ~_~ ='4725 (__ ) (D)
. fL fL
(d) Lacey's wetted perimeter Discharge relationship.
From the logaithm!c plottings of large amount of data, . L3.cey derived the e~piri9!~l
relation in the form
(E)

and-P ",: =7'11l (F)

R
\
v = O'141~w~ where 1." (G)
R
Pw=wett"d perimeter
\
Q=discharge in cusecs.
R=H.M D.
v=me;:1li velocity.
The constant 2.67 i~; not a true constant as it varies with different localities. In
the Punjab, the Rese?.rch InstitJte. Lahore, found its value to vary from ,2.2 to 3.12 fro 11
24 rebcrime sites keDt under observations for a Ion"b time • Lacev'
j_ ..., rematkecl in the India:1
Engineering. Calcutta, AUg(l"t, 1937 as helow : -
" It must hI} rem<;m bered that the expression P w =2.67 Q 112 gives the valne of th ~
m;nimum stabL; pc;rimeter in channe1s in incoherent alluvium with considerable accuracy. At'
the dat-a so fa,' ava'lable with considerable as curacy tends to confirm the expression. Th:
value of the comtant in individual chann;l-: vari~s with:n limits ronghly from 2'20; to 3-::W.
If the banks are tenac:ous. the w:dth may be less, if the b~d tenacious, the wid::'l
may be greatel", the~eforl~, there is no. absDlute rigidity in the constant 2'61 when applied ':)
small channels with even a sm'lll admixture of clay in the bel or banks. -
In lal"Yc ch;l.llnels it is to our inLrest to make then ?s d'~~p and n'l.rrow 'BS i,
consistent witt] ~t'it iHy In sma.ll channels ap )roaching the limiting velocity there is Il'
ohjection to m lking them wid~r if the clay present provide.> a snall additional measure
stability."
 164

The upper limit of the wetted perimeter constant is about 3'\6 and the lower one 2'2
(or the Punjab Canals.
(e) Lacey's general r ~gime equation.
(i) It has been said that 1he Kennerly's acceptance of Kutter's formula with value,
of N='0225 was noc HI)' safsfa' tOly j acey'~ a'tcn pt tu evolve Regime flow equatinn i<: a
great :ldvance to our knowledge. Ignoring the attemp1 s of Mr Lacf'Y to giw' mathematical
proof of his general. regime eq~ation using the dimensionless numrer ,he analysed a ~arl2e amo';!nt
of data and his regnne equatIOn fits the data very well. Lacey s general reglme equ:ltton
entirely fitted with the author's observations puhlished in paptr rear! in the Institute of
Engineers, India, 1936, giving variation of N in Manning's formula. La~ey rightlv claims on
page 12 Central Board of Irrigation Publication No, 20 "it is very improba )le th'l.t any improve-
ment can be made in this rel<ltion in the near future,"
Lacey's formula for regime flow
v=16 R2/ 3 Sl/3 ......... (H)
This equation applies to perfect regime and has the d~ciden advantage that it applies
to all earthen channels in regime irrespective of a.ny rugosity consideration.
(;i) In the case of c~annels not in perfect regime but postulating cetween the initial
and the final regime, Lacey denves his flow equation in the form of Manning type in which
Rl14 1'3458 Rl/4
Chezy's coefficient is represented byN-in metric units and bY~~N-~- in foot units.
Lacey denotes his rllgositv co-efficient, as Na and terms it an absolute
rugosity co-efficient, that is, a co-efficient determined solely by. the average siz6
and density of the incohflrent bed mat"rial of the eha.nnel. The standard grade
of silt is that which connotes a rugosity co-efficient of -0225 when the hydraulic mean depth
is one meter. It is, therefore, permissible to write, in terms of the Lac~y's definition.
N='0225md 1/2='0225 fL 1/4 where md =mean dlameter ........ (I)
(iii) Vertical Exaggeration, Lacey pidnres the hydraulic channels as the
wetted surface spread out on the horizontal
Perspective view of wetted surface flattened plane, with the hydraulic mean depth erectpd
Out by a Dimensioned HydraUliC Mean therf'on. Prototype and model may therefore be
Radius erected thereon depicted as in Fig. 5.
Let the horizontal scale of the model be

~ and the vertical scale be+-then the ratio ~ is

tf'rmed the "exaggeration" E (Le., the ratio betwepn

the vertical and horizontal scales). Then, in the
Fig. 5
horizontal.

~=pw =X
Rl Pwl

in the vertical, ~
R}
= y

. E==--=~
x p
•. y P wl
But ,from equation (F)
P w =7'111 v
R
 16E-

A~ain let HL and HL1 • be the rliffer nees in levels oetween corre~p,:mrling points on prototype
Hu--s
and model, at the end of lengths Land Ll respectively then ~L =5 at).d -
Ll
1

. _____
S 1
=-H...--- . ---
L, _
Y.-·-·1 =y
.. SI L Hu X
. S _ 1 _ VI (J)
.. -s~- -E---v- .. j

This relationship is of vital imporbmce in model experiments.

Returning now to the Chezv's fomuia
v C Rl1251 12
-V;-'-=~' -R,11 2 S1112

C v R 1. S/12
or--c;- =-v~-' R:I~ . -SI,J
Substitut~-f"r S from equation (J)
I, C v312 R/12
C.. = v1 312 . RIll
But Lacey postulates t.hat streams flowing ill envelopes of the saml'! silt (Le., whose fL value
is the s:lme) must havtJ the sarne co-efficient of absolute rugosity N a , hence for such channels,
from equation (A)
v R1fZ
-V-;-=R I 12
I

~ubstitute value of v_
v,
r. Rl/4
. ~-= R8 4.......... ··(K)
therefore C in the Chezy's formula when applieJ to fegj'ue channels various as R1I4
Hpnct the Chezy's formula may be written
v=KR31451/ 2

or V=_~I . R 3 14 SlJ2 ...... 'L) i

a i
where N. is a true co-pfficient (a m"asure of the absolute rugosity of the silt envelope.}
Datermination 01 the va.lue of Kl
Now the values of Kutter's Nand Manning's N coincide at a. depth. of one motre,
Lacey <11!cided to retain this coincidence, familiar to all hydraulici<l:ns.
V= 1'4858
_ _ R213 S1I2 .
(Manmng)
N

= l'485R (3-208,2/3 S112

N

== [Lacey) E:
N. .
(3'208)314 S172

whence Rl= 1'3458

and v== 13458 RSJ4 S112 (M)
Na.
 166

the above formula is strictly applicahle to regime channels, but may be llserJ.
fretly as a, substit1lte for ei~her Kutter'~ or Manning's in non-regime alluvial channeh or
channels with rigid boundaries etc. with improved accuracy and greater facility, by
adju"ting the valup. of Na .
(iv) Regime Slope.
Laceyevoh"ed his regime sl()re fOTll,ula in 1932 in Technical Paper No.1, Vnitei

Proviuces as S=_1__ f L5 i 3,/Ql/6 but in 1932 in Central Board Irrigation Publication No. 20, it
'1788 ,
has been modified by him to
S= __1_ fL 5 /3/Qli6= 000542 fL 5 / 3/Ql/6
18443
(v) La ;ey's Shock theory.
1 he use of all such equations presents great difficulty in assigning values to the
rugosity co-efficient. The"'e is no Ii 11it to K utte['s "over-all" co-efficient N in obstructed
non-regimp. channels, ami for such chanoels no rati-onal equation can be fra111ed. One
source of error is in measuring the slop3. This if measClr,:d locally over the discharge
run, may be too small to be accura~ely ass?ssed, If mea,und a, gC03l sbp~ OVJr a large
slope base, the hydraulic Iw;an d :pth at the di~charge site, owiug b th ~ natural tend'~ncy
to select narrow reach% for measurements, wIll often be in exces,; of th:J averag ~ hydrau-
lic niean d';pth throughout th~ slDpe bas~ Th; slope a;; mr;3.sured wir, thus b~ in excess
of th0 slope applicab13 to the hydraulic meal d'~pth a; mea,uf.Jd. The sbpJ may also
include "sheck" du;~ to bend, 0' irregulari :ie; in th ~ channeL an 1 al30 "shock due to
channel condition" as opposed to c'lann~l mal;rial In SUC:1 cir-.:umstances it is prderablc
to assign to Na a value \\ hich the bed material warrant3, and to account s')pd.ratr:ly for the
energy destroyed in sheck by making an a-;:>pr.:>priat,; d·~duction from the gross slope as
measurel. The equatioo can then b,_; wntte 1 i l th·} fmm
V = I"~i-"~ R3J! (5-S)1/2 (N)
J.'II.

The following tahle of thp rugosity co-efficient N recommendeil for different channel
conditions demonstrates the futIlity of ~eeking otller than an approximate empirical solution
and the arivantage of employing a simple exponenti.al equation of the Manning type. Value of "3"
the fraction of the slope "S·, lust in shock due to channel conditions

Channel cundition N 4'S" Dcs~rlptlon of channel

~-~-~-~

'0250
--------

'000 S
----_------~----
Natural stream channels.
---
Perfect
Good '0275 '174 S Straight hank. full stase,
Fair 'O~~OO '323 S no ri'ft:; or deep pook
. Bad 0330 -_--- '426 S "EnQ;ineprin'! N pws" V ,I 7~ 1916 Pac:e 371
Very goad· ----~-

'u~"5
---_----

'0 Iv ;:,
-
Ea-theril--(_llaH~~~iS an<f -a-n~tS---- ---------
Good '02fO '190 S under urdinary [ondit,ons.
Indifferent '0275 '::31 S BuCkley's In - a ion ?
l::l<.d ·\;~l.V '437 S l"ocket Book

The t ,bl,~ ,h ws that mor : than iJrty per cent of th~ energy dest 'oJed in the channel
may be d s.sipated by ~hannel i.regulariti;s.. It is- important. to ~ote that if a ~ain canal is
excavaLd m the first mst; nc~ to an eX,-" SSiVO slup.; 1t will adjust ll.elf by the cr~atlon of canal
irregul2;r.ties. Such a canal would eventually achiev0 some kind of balan~e and remain stablt!
the slope, how.:vc:, would be m exc,-,ss 01 that wkch the grade of SIlt Lanspo[ted would
demand in a r,,;!Sime channc;1.
S. La:cy's Dhgrams.
I accy pl l li;hed his elia.grams in TechnIcal Paper No.1 P. W. D. United Provinces;
In'lia. 'i hes-~ d azra:ns WJre ba ,r;d 0,1 his- original quatious.
 167
1/8
V = (JfL2 ) ( B)
o 3'8

P w =2.67 Qll2 ( C) R= '7305

.....V2
~ ( D)
fL
fL 513 ~; "

S=·· .- - (E)
17l'8 Q 6

D= ~R~ - [ (~R~Y
A
B= ----sD
o
K=2(r52+1)112_~ .':;; "W p::' r'lilt _(H) , ,
, ,The channel sectIOn was assumea'to be t~pez01dal wIth sl(:;es ! to 1 though he believed
that it would become cup-shaped aft~r,il.Jii'f· year~,r:unning.
Side ~lopcs ",,'. S :!I!~ K
Vertical 0 2'0
One half to one . . , 0'5 1'736
60 b slope . i ' 0'5774 I", 1'7321
One to one 1,0 18284
l·~to L 1'5 2'105
The diagn',ms are given in Plate XV bJt they need to be revised tor the equations
finally arrived at by him as summ3.rised in paragrdph (7) above. Th'.l origillal diagrams can b~.
,used till new ones are produced.
9. Examples.
(a) Des:gn a regime channel section discharge 500 CUSilCS and

V 0= ( ._ _ )
3'S
--~_
QfL 2 113=( 500x l'212)l.!G
38
.. --
.
=2,4 feet per secilud

A= ~~O =208 sit.

R ='~~~Vo2_=~73~S?<~_42 =3'47 ft.

iL 1'21
F'brside s'ope t to 1, K=l'736
D=,A _/(~ )2_A\t =17'26 '-(298-120)lj2=3'93 ft.
:GRK l lRK KJ '

B=DA-sD=~08 - ~ x3'93==SI'04 it.

'" J'93
The channel section should have bed width=51'O ft. and depth=3'9~
. . fr}j3 1'21 5 / 3 •
S=Reglme slope= J788QI/6 = I788X500 j6='00027
'
say= '27 ft, rer thousand.
 168

It is t~ save these calculations that the diagrams are usen.

(b) Design a channel section for 200 cusecs and'V =·3

vo

Lacey's silt factor= fL = (~) 2= 92 = '81

o
From diagram XV (h' f0r fL -'SI and Q=200 cusecs,
H=:10 ft. awl D 3'-45 feet
From diagram XV(c) for fL= 81 and Q=-=200 cusp.cs.
Regime slope S='16 ft. rer thousand feet,

(c) Sectio'1 for SO ('usecs with~ = 'S

o
2

Lacey's si:t factor fL=(}) ='8 2 ='64

o
From diagram XII (A) f'r fL='64 and Q=80 cusecs.
n'lpth D=3 5 ft.ann bed width =B= 16'5 feet.
From Diagram X[[ (c) for fL='64 and Q=SO cmecs.
Regime slope S='l ft. %

(d) Design Kennedy's section for ,:ischarge 200 cusecs and 'V_= 1'0 and giveh slope of
vo
.~ ft. per thousand.
From Plate XII for slope 2 ft. per thousand and Q=200
B=26ft. and D=3'S ft.
From the same rliagram, according to Wood's normal data of d~i&1) for~-~;J; cU~t'!cs
B=21 ft. aoo D=A'3 feet. ;. '""" ''',(iu lI, .
The ~lope rpquired='216 ft. ~er thomand. 'i':t :-r .(~
10" A surrmlUY of Lacey's CcnclwfcllS.
"In all regime ch:mnpl" in incohprent alluvium the primary funda'1lental variahles
are the mean velocity. the hYdraul'c mean depth, the water ~urfac~ slope and the bp,d siH grade
Of the four variab~es V, R S, and fL, two only are necess:uy in order to obtain the

third anrl fourth. The first equ1 tion is V = (4/3) t VfL R

In this equation when thl) valu3 of the silt Lcto~ is u,.dy th~ grade of sand hal a ruso,ity
co-efficient Na of '0225 and standard sand is thm defimd, When the hvdraul'c m ~an deoth b
equ8.l to one mf'tre (3'2808') th(' value of .'Ja is identical numerically with the value of N as;igne 1
by the Kutter, Manning or Forchheimer equations for flow,
The second equation is v=16'05 R213 S1}3
In this general regime equation the rugosity is implicit, and the equation is a?pEca.~l ~
to every type of regimr; alluvial channel from very fine silt to boulders. FL-J'l1 th3se tW)
tundamental equations a variety of other equations can be derived. If a flow equation of the

\Ofanning or Forchheimer typ~ is sought it must take the from v=~458R! Sf

Na
The valup. of Na is given by the equation Na=·0225 fI}14 !
[f it is desirerl to apply the flow equation to non-regim) channels in which shock and athar

ch anne I can d 'ttlOns

'1 p ay a part. t h e flow equatIOn
. ta1{es t h e f rom v=-N
1·34&::SR.a
~, S ~
<I, ( - s) 2
a
in which s represents the fraction of the gross slope dastroyed by shock. .
In regime channels, free from shock, aDd of which the silt is of standard incoherence,
thp silt factor ~.nd mean bed s]t grade are correlated by the equation fL = 1'76m 1/11 the
diameter mr being me~.sured in millim0tres.
 169

If shock is prec:ent in a regime channel as a variable. the silt fa 'tor becomes an 'nversfl
t11I1ction of the silt grade and m.,l/2 f.. = rr. I / 2 fL=m-'(2 fL etc. In the moulQed E03e relation, where
q =R v='375 Ql,2: Srxm5f6JqU l
the slope i~ a functi n 01 ? d silt g> ade and ~he d~scharl2'e intensity 'q' the shock bj 1~ implicit,
In the EqUIvalent expressIOn of Lacey Soc fr»3 (fr/iL) 5,3/ q l;3 ; the s~lOck is e"plcit. When there
is no shock the expres3io;' s are id,;ntical. The modified Bose equation applies to silt of a
stal1dard coherence. The equation of ( asey
Sxf 5,3/ql j 3
when shock is absent is applicabb irle3pective of the precise coherence, cohenmce being impl'cit.
The s'I?plest dime~s oned ~xpression for. thp. general regime equa i n is the Lacey
_Malhotra equat~on v/vo'X p·~)/m 1I2; m wh;ch the kmematic vi~cosily is implcit.
Under Ideal CJndltlons when all the varIables ar~ equally fre~ to vary, the wetted
perimeter ~nd dis:harg,~ are correlated by the ~acey. expression Poc IQ12; from-this equation
coupl6!d wIth the t"'o hrst furdamental equatlOns lU terms of V,R,S and fL, all other regim<l
. ralations foHow vidp. page 38 of C.B.I. PUblication N. :W.
11. Comparison of I{ennedy and Lacey Theories.
la) The theoretical concepticn of silt transportation is the same in both cases. The silt
is carried by vertical eddies caused by the friction =lgainst perimeter. Kennedy considers
irrigation chann"l section generally trapezoidal :lnCl, therefore, .neglects the sides and derives
his critical velocity formula in terms of d{ depth). Lacey believes that the regime sections in
alluvium are cup-shaped :lnd therefore, derives his regime velocity formula in terms 01
R 'Hydraulic mean depth). L:lcey considers that the irrigation chflnnel cannot be in perff'C1
regime on account of tJ;e imposed sections in the beg:nning and the subsequent berm growth.
They cannot therefore be ngime channels flowing in self-silted alluv;um and hence they can
only be called in initial regime. Kennedy deals with irrigation channels oilly and calls them
in regime whf'n they do neitLer silt ncr scour.
The author considers the change of D to R as a change of fmm only. In wide channeb
they have very nearly the same value. Both claim to satisfy the !"ame data. Tt e empirical
relations derived by both arc essentially correct, The superiority of Lacey's work over that 01
Kennedy is due to other reasons mentioncd hereafter. .

(b) In thA regime velocity equation V=I'13v'fL R Lacey used f L - as a silt factor insteao

of Kennedy's C.V.R. Lace,y's fL= (~) 2 of Kennedy. Kennedy simply stated C. V. R. (~)
varied according to the silt conditions (silt charge and silt grace) Lacev did not Ip:1ve fL as II
guess work but COfrplafed it with mean diameter of ~he hed silt in his formuJa fL= 1'76m 1 / 2 fOJ
regime channels and in nl n-regime channels N a =022S fL 1/4. This is a very distinct improveme n1
made by Lacey. This omission in Kennerly's work h:1s, however, ceen made good in author's paper
read in the Institution of Engineers. India, 1936, the summary of which is given in Paragrapb
No. 13 of this Chapter,
(c) The selection of Kutter's formula and giving N an arhitrary value of '0225 by
Kennedy in his work was not correct, Lacey has produced a general regime flow equation iD
formula V = 16 R2/a 51 / 3 after analysing a very large IT'ass of data of regime channels, R. K.
Khanna, Assistant Engineer, Punjab Irrigation also worked on similar lines and evolved a
regime flow equation in the form V=13'07 R58 S,3 which is of the same form as Lacey'~.
Lacey's work in workng out Slope equation and in investigations of shock ill irrigation channeh
stands out conspicuously in th~ domain cf resfarch.
(d) Kennedy's wcrk sufiered from the defect that he did not notice the importance 01
bed width depth ratio Thi, dP.ficiency was made gcod by the work of Woods and Lindley a5
mentioned before. Lacey produced his formula to this effec~ in terms of wetted perimeter i.e ..
PooQ'f2 and the value of K in the formula P=K Q!lJ, is got out by Lacey as K=2'67 for averag~
regime conditiclDs. He admits that there is no absolute rigidity of the constant value and
accepts its valiati( n from 2'2 to 3'2 for regime irr:gation channels. Similarly in. the case oj
Wood's bed width rat'os there is no absoh;L~ r:gidity because they are for unity C. iI.R. and tht
ratio must change for channels ca,rying silt requiring C.V.R. other than unity. It is, theref~rt'
 170
appare.".t that the regime chan ',el <.;p.ctirm ran have wettecl perimet.;rs 20% more or le)s than the
L',~ey's fo 'mIla P w='· 37 Q112 and s' 'llilarly 'r11o ving \VO lds' normal data of de3'gn of Kennedy's
channel a var:at'on of 20% is q.lite likely. We do need a guide to limit our selection of bed
width and rlepth rc~tios whether it i~ in the fOem of Lacey's wetted perimeter formula or in the
form of Woods' ned width and dtpth raEos.
(e) Kennedy did not fix regime slopes for hi" cha lnel'), However, his diagrams clearly
show that steeper stapf's are re1uired f,Jr small channel> and the flatter ones for large channel;,.
f 5/3
In h's original wary. Lacey produced a regime slope formula in the form S= 178ffQl/6 and in his

final work, Central Board of Publication No. 20, he put this in the form Sex: where fL is the
fL5i/3
q3
silt factor and q disch;jrge inten<;itv per foot width of the channel. The final formula shows that
there is no rigidity ahout the c< nst tnt::;. Tre or ginal tormula of Lacey allows very excessive
slopes. There is not a single channel An the Lower ]helum Canal, Lauer Chenab Canal and
Lower Ba~ i Doab Canal where La( ey's regime slopes are available. These canals would have
nl'ver been constructed, jf the f!esigr.e s han cared for the Lacey's regime slopes The author
worked out in Paper No. 154, Punjab Engiw:ering Congress 193! in reply to Lacey's criticisms
that if Lacey's regime slopes were allowed in the Mithalak distri h ut2.ry, the head supply shall
have to be raised by 4'5 ft. and if the same pro:ess were extended tc head of Lower ]helum
Canal, it would hav._; to b,: raised ahout 15 feet ~ven after consuming all existing falls. The
author cons'ders that a further l;mitation in thC! form of r,:;gime slop'~s is not required. It is
enough to design the regimfl channel sections with the availa')le slope b'lsed on the general
regime equation and cr:tical ve1cc:ty fvrrnula limited by the wetted perimeter or bed width depth
ratio consider&tiJns.
If a Tegime slope formula is to he used ::IS a guide, then the Pnnjah Trrlgation Research
fnstitute formula (1937) S='00209 md 85/Q.2 1 is bett:r ap?l cabl; to the Pclnj ~1 c0nditions (1937.)
Annual Renort) where md=mean diameter in millimetri s. Tn designing irrigation channels
according to Kemwdy's theory. the author's regime slope formula (paragraph 13 below;
2

S=~(I+.!_'_?~) is the be,t guide as it is di n~niom.ll\T marl} corn~Gt than Lacey's regime
7500 p+'5 0

slope formulae mentioned above, for all the quantities in this are dimensionless.
12. Defects in Kennedy's and Lacey's Theories.
(a) Both of them aimeil to find out the average regime conditions. None of them
considered the effec t of the varied silt conductive powers of the outlets and the off-takes on the
regime of chann€! Mr. lacey stated that the silt f<,.dor was constant for a canal system which
is not a bct, while Kennedy said that C,V.R. decreased towards the tail of a channel,
(b) They also took no account of silt left in the channel by wattr that is lost in absorp-
tion which is as much as 12 to 15% of the total discharge of a channel,
(c) The effect of silt attrition was also ignored in both these theories. The silt size does
actually go on decreasing by the process of attrition among the rolling silt particles dragged along
the bcd.
(d) They also took no notice of the scale effect. The Ravi river at Sidhnai, Indus at
Sllkkur and a small irrigation minor may have Lacey's silt factor unity or Kennedy's C.Vft. unity,
but actuall)" th~y carry silt charg'_: and silt grad; very many tim;:; different.
(e) N(;ither la:ey I:or Kennedy ha5 been able to define the silt grade and silt charge and
the size of the chann,;l for unity silt factor or unity C. V.R.
13. Author's design of Irrigation Channels.
. (a) The author puhlished his paper on "D,osign of frrigation Channels" in the proceed-
Ings of the Institution of Engineers, India, 1966. The first 4 factors as p.numerated in Paragraph
12 above were dealt therein as influCllcing tue change of C.Y. R. h Kenn" y's theory. The selection
of C.V.R is no longer a guess work which was the major defect in Kennedy's theory. A brid
ummary is given here and the student should refer to the original publication for a detail, d
study and for the proofs of the formulae.
 171

(b) Notations.
a1 = Silt charge in water at the beginning of a reach and reptesent~ the ratio of the
amount of the silt carried to the volume of water containing it.; a 2 =Silt chargp. in water at
the end of a reach; f1 =Grade of silt charge in the beginning of a reach and represents the
average diameter of all the silt particles ahove 0·04 milimeter; f 2 =Grade of silt charge at
the end of a reach; Pl' P2 etc., = The silt conducting power of outlets in the reach expres~d
as a fraction with resjJp.ct to unitv in the reach bv weight; ql' q2 etc .• =The corresponding
discharge in cusecs of outlets having silt conducting power as PI' pz etc.; A=Absorption
in cusecs in a reach; Q1 =Discharge in cusecs at the beginning of a reach; Q2=Discharge at the
end of a reach.

rl=~2_=Ratio of silt charge at the end of a reach to that at its beginning.

a1

rl=~ =Ratio of silt grade at the end of a rea~h to that its at begining.
il

fS=· 01 =Ratio of discharge at the beginning of a reach to that at its end.

Q2

A = Dl = Ratio of depth at the beginnin~ of a reach to that at its end.

Dz . ..

R c =..3 2 Ratio of C.V.R. at the end Of a reach to thlt at its beginning.

Xl
B=Bed width in feet in trapezoidal seC'tion· with side slope '1/2 to 1.
D""":Depth in feet trapezoidal section witli sidt(ll2 to 1.

P=-~-=Ratio of bed to depth.

X=C.V.R. (Kennedy) in formula. v=x.vo=x.c Dm=:icS2D·68
(cl Formulae.
Change in silt charge by varied silt conduction by off-takes and absorption losses.
D C ={_g1-(1'lql±i'2q2±PSq3+ ...... ) }, <;\-(Plql+P2Q2: aQ3+···) " {A)
1 Qt-(A+ql+q2+qa+······ .. ) Qa
Change in silt grade by varied silt condllction by off-takes and absorptioB losses.
r~= { Ql-(P13ql+P2~~2+P33Q3+ . . ,.)} i . . ~)
x! (C)
Ch~e
. .......
in , C. V. R for r1 and r 2 ; R C=-=r
Xl 1033 . r.· 3• r.· 1=r1·S3 . r.· 3."- 11'
t
S "= ~~- (1 +__!_Z!_)
7500 p+"5
3't ...
D= __Q._.__ tfl&.r:-I.',;m.:oltI £~()ll ,8\t ;.r,:'1 .)'[7'
·t',,,·,,,.,, " " ! , , , .." ., . . ,\ (E)
{ .6:.!x'1>+.5\ }
i' \. ' i '.,
_.... ,. ' "" :-. ~ f ~~ ,,'
'the ratio of C. V. R at the end of a Mlaeh to that at Its beginning can be calculated
from equation (C) or takPn from rlate )otI. (A)
Actual non·silting velocity- of a channel section is a mUltiple of Vo by some factor which
dt'.peEds upon the silt-charge and it's grade. ·Kennedy called it Crftical Velocity Ratio (C.V.R.)
and jn Lacey' 5 theory it appel:lrs as !liilt factor. From the formula given above for the change in
cy.R. it is clear that it not anI):' depend.;; on the Change of silt grade but al~o on the chl,lnge of
sIlt-charge and the channel sectlOn mverseJy as DI/8. It can only be a factor, as uspd by ltennedy
and may conveniently be called Critical Velocity Ratio ,C.V.R.) and not a silt falZtor. A unity
 172

C.V.R. or unity silt factor (Lacey) have not the same significance for a river, canal distributary
and a minor. If a canal was running non-silting with C. V. R.=·85(for example Lower
Chenab Canal at head) it docs not follow that its distributaries, drawing the same silt charge
and grade, would run non-silting with C.V.R=·Sf\ They would require C.V.R. as influenced by
the fa.ctor A in eqllation( C) above. It is therlliore, that the off-taking channels from the Lower
Chenab Canal are actually running with C.V. R. morc than unity, even after considerable side
exclusion at their head regulators. It is nothing bi.it misnomer to call C.V.R. or any power
Elf it as silt factor.
(d) The design diagrams XIII (B) and (C) for equations (D) and (E) are based on
Kennedy's critical vp.locity formula and Lacey's gfmeral regime equation V = 16 RaJ3 SI/3. With the
worked out value of C. V R. ir m X[II (A) as eXIllainp.d above, and the available slope, p (the
ratio of bed ",idth to dtpth) is got from plate XIII \B).
It is just watched that the value of p is not off by more than 20% from Wood's bed
width· depth ratio or Lacey's relation~; =7·2V. If p is out by more than permissil>le, the
tentative longitudinal section of the channel should te changed. If it is a case of excess slope
giving relatively low value of 1'. it should be flattened by designing suitahle falls. If the slope
is fiat giving value of p more than the permissible and also if it could not be steepened, then
maximum permissible valup. of p should he selected and design completed from equation (0) and
(El. In the latter case, the non-silting design is not possible and the section designed should be
aedaTed to need annual silt clearance just before the period of keen demand for equitable
distribution of supplies to the outlets. With this known value of p, depth is got from Plate XIII
{C) and bed width i~ then equal to pD.
[e] It will be interesting to mention here that the regime slope formula as worked out in
the said publication is in the form of eql1ation (D).
Slope is a function of X ano p only. Both are ratios and have no oirnensions. The
slope, which is rlimensionless is uniquely determined by these two oimensionless numOOrs. The
part played by X is much mora predominent than that by p, as explained in the discussion of
the pape(.
Slope is eS8entially independent of discharge. The channels could certainly be run
non· silting even with flat slopes, if permitt~d by dimentionless numbers X and p, The bed
width-depth ratio is deteTminf'd independent of discharge for a given slope and silt conditions.
14. Examples illustrating the use of Author's 'Nomogra.ms.
(1); Head reach from Head to R. D. 16000. Mithalak Disty. Parent channel section
U. S of the head, discnarge= 171'0 cusecs, Bed=74 ft .• Depth=7'6 and side slopes ! to 1•
. Area=7'6(74+ 7;6 )=591'3 sq. ft.
_.
710
Velocit v=s19 1 a =2'892 ft. per second and C.V.R.=2·892=-94
. ~~

. Silt conductive power of the head regulator by weight as deteritimM 'by actual'
experiments =90%
• 0 9 '\ Depth in parent channel _7'(; -2 '05 d I t -'9
- .rl == . ;1'. Depth in oft take -3'7 - an e r 2 - )e

Using formula, Rc=rllp. r· 23.A 1/", from Nomogram Plate XIII [A]; R c =1'05
·,'C.v.R. required in the head reach of distributary='94 X 1'05='985.
Discharge of Mithalak Distributary at head= 14;J'5 cusecs
Slope=l in 4444- C.V.R =091:15, from Plate No. XII{ (C) using the n~ocram. , ...5'0\
Similarly from Nomogram No. X111 (B) when
Q= 143' 5 cusec;, p=5'O anj C. V.R.=·985,
deptn=3'68 feet
Now Bed width=pxD=3'68x5=18'4 feet
Use D=S'7 and Bed width= 18'25 feet.
 (2) Reach from 16000 to 24500
Discharge = 1::J3 CUsecs
Slope= 1 in 4'lOO
Discharge of outlets in Uppf!f reach=7's cusecs
Absorption in outlets in upper reach=~'8 CU'iecs
Silt conductive power of outlets in uPPer reaeh= 122 9/()= 1'22
_ Ql- (Plql+P. 2qZ+·········)
f]- - ----------
Ql"-(A 1 +ql+qZ+ .... ·····) ':,
, 143':::-7'5x 1'22
: = 143"·5;;-----;0;\2;--;0,8cc+
co7'-;'5;-:)- =1'010

{ 01-(PI3 q c l--Plq,-l- .......... } }i

". 1~~~~;;:~~~:'. }"!"'~~~)
,L 133.7
\ fa= ,Q, = 143'5= {tlW
12, 13;;5'7 ,
:. Rc = Ratio of C.V.R. at 160tlO to that at its hea.d.
Ro =fl1/3.ra·3.rlllO '
From Nomogram No, XI1I (A) Rc =1'011,
:.X=·985x l'U11=I'O
Let p be the ratio of hl"d width to dept:h.
From Nomogram XIII (Bj. p=4 66
From Nomogram XIII (CI, !)=3'57
:. B=pxD=3'57x4'6=166
Keep D=3'6 and B= 16 5 feet
(3) Section Down Stream R.D. 39275
Discharge = 107 cusecs
Slope= I in 4000
Absorption up to 39275='2'8 + 1"{ + l'S +O'8=tl6 cu,ecs. . __ -
Discharge of A.P.M. outlts set a.t bed level with silt conductive power 12210=60
Cusecs,
Discharge of A.P.M. outlets set at S/lOth with silt conductive power 11()~'.=6'36t9·JS
+6'31 =2'l'03 cusecs.
_ 143 5-(1'22x7'5+1'l x 22) =1'027
f 1- 107

f2= L
( 143·5-(1·2211 x7·5+I·l'X22)
107 .
\-1 _. .•
J - 97,
--.07 == 1 34 , r -
143'5 • .

From Nomogram No. XIII (A) for the 'value of rl.r. and r.. Rc "",1'035;
X='935x 1'035=1'02
From Nomo~ram XIII (B) p=4'5
From Nomogram xrn (C) D=3·37. ft.
B=pxD=3 37 x4 5=15'17
Kedp D=3'~ ; B= ]5'25 feet f
\4) Section at R.D. 58,000 Ii
Discharge=S8 cusecs
Slope =1 in 4000 ;~,'\ "- :,
Ahsorption=S'6x 1'3+1'0+1-1=10'0 cusecs ,
Di~charge of A P.M. olJtlets at bed level with silt conductive power 122%=7'5 cusecs. ,
Discharge of A.P.M. outlets at 8/lOth :'vith silt conductive powet 110%=22 Cll$6C.q,
Discharge of A.P.M, outlets at 9/10 Wl\h silt conductive power 115%=16 clUOIOJ.
 174

143·S-(I'22X7·5+1·10X22+1·15X16·0+ ... l =1'030

r 1 == 88

r.= {
143'5-(1'223 X7'5+1']3 x25+1'15 3 X 16'0+ ...... ) "')
88 Jt= (~~_Q) i 88
='9~;
..._. J -' ~ .1::;'
143'5 .
ra= ---=163
88
For these values of r l , r 2 and ra from Nomogram XIII (A) Rc =1'025
:. X= 1'025 X '985=101
C.V.R. 12'0es on decreasing up to the tail of the channel. the maximum being at
R.D.39275.
From Nomogram XIII (B) p is found; "' .... ,;,. :
I'
I \ I , ','

1 I ............ f ,t : t"}--'JV
When S=--- X=l'Ol , r 4 ,

4000' 1': '('p . . " ~. • ' ._

e'h", ) t··", J":", c. ,\1--·,:, t,~ ~ =',
:. p:=:j4'6 i \','.(:1 i i
From Nomogram XIII (CI D=3'lS .... • r!
:, B=3'15x4'6=14'3, Say 14'5
88 ,
5;)1. :; r,.t.{.
\"U
~? 'f~ ,
."
.
V=- 3'15 =1'75£eetperseconiM~'I" "".,.,)- ,'+'C/_ <:r:,
. , . ' .. :-l! -or_~\" . " j,fl. ,;J. V ',tf' ItIJ£l'v.o-:- "n. "
y
3'15 (14'5+ ___)
2 ".' .0'" 11 "'-: ,.!t ,~
IH ,.. ? lid H p I '(l;';', -{~1!) ~ UW'l I "-.::
10. Velocity and sUt distribution in a channel section. t ??",,- ?:
... 1," l ,
(a) Variation of velocity over cross section of a channel.
The velocity of flow vari"g at ditferent p()ints of the cross section of a, chanQet The
frictional resistance of the sides cause" the water to 5low down towards the sides of the channel;
I" I" 19 and the frictiol1al resistance between the water and
I I the atmosphere causes a slight reriuction of velocity
" at thA free surface. The maximum 'velocity will
be on the vertical centre lihe of the channel at a
~ point a little below the free. !lurfacc.
. The variation of velocity over the cros!'! !'ection •.
of a rectangular channel is shown in Fig., (j labove.
The cUrves shown are lines of equal velocitYi
they have the greatest value at the centre, just
lelow the water surface, and dEcrease towards
Fig" the !'ides and base, In Fig. '1 are shown
the variatiJns .of, velocity on horitontal section
lines taken at different depth!'1. The velocities at
different points of the section hnes a, b, c, & d
are platted on a base representing 'the width of
the channel.
Fig, 8 ghaw~ the -variation of velocity on the
vertical seetion lines, 0, 1, and 2. The horizontal ci
ordinate repnsents the velocity and the vertical
ordinate the depth.
The mean -velocity on any ~ettica.l section
occurs- . at approximatelY"6 of the depth; it .varies
with the type of channel and with. the nature 01 Fig. 7 .
the &ides~., The discharge '. of the", whole channel
 175
may be obtaineo by oividing the section into vertical rectan 1
ano finding the mean velocity of pach rectangle. Using this rn~::
velocity, the nischarge through each rectanglA may Ie obtained.
• The SlJm of all these discharges will te the total discharge of the
channel.
The mean velocity of each rectangular strip m::!y be taken
approximatlev, as the velocity at a depth of 0'6 or the total
deptb. In a branch canal of over 50 ft. bed width trle mean
velocity in fr~e nntral !'cgnwnt he.s teEn found to be jnst douDle
the mean velocity of tl.e ~k're Stgrnent. 1'1:-.e ratio reduces to 1 ~ on
small channels.

Fig. 8 sm distrIbution in vertical pla.ne.

(b) The author carried out observations to nnd out silt
distrihution in vertical plane. The observations we;'e taken in the middle of tlle Bhek
distributary, Lower Jhel ll m canal, Pllnj:tb from a platform by means of silt sampling eottles
and compared witti He total silt charge by weight passing over th'~ cre~t of a fall Tbe
result wtre published tn Fig. II, PJ.ge 4, Author's paper on Design of Irrigation Channels in
the Institutlcn cf Engineers India, 1986. They ~re reproduceo in Fig, 9.
If rhe silt clnrge intensity over the "ectil)n be 10;), 6f) to 70 p~r cent intensity of silt
by weight will be near tbe surface and about 130 per cent
near the bed. The cent per c.~nt silt charge iiltensity
will be at about '6 D from the surface. The portion
DE represented the silt dragged along the hed. It was 01
observed that coarse ~ilt was also available near the il
surface as near the bed though in a very small amount. g.
Tbis showed that the vertical eddies do work from bed ~ 0.$0
right up to the surfa.Ge in flll depth5. If some
disturiJance In the form of ohstruction I e placed causing
vertical eddiel", the silt charge intensity at the surface
may be cent pF;r cent anti even more.

(c) Silt lifting eddIes In vertical plane. 100

Kennelly stated that the silt was carried by vertical Percentage of silt charge hy weight
eddi~s produced by the roughness of the bed The Fig. 9
eddies ~re defiuitf'iy produced on this account as proved in Reynold experiment.. , and they
com,ert the st;eam line motion into turbulent motion. Lacey also states that silt is suspended
by tile verticl-ll compon~nt of eddies, but urges that they are gcn~rated at ::lll pnjnts by forces
normal to th~ wp.tted p::rimeter. Can such eddies be strong enrmgh to work up ~gainst depth ot
10 to 12 teet in large canals? The silt cany:ng capacity would thus be more in channels with
shallOW depths alld less in case of those w.th large depths.
Ihe author cC:Ds;cets that very much stronger silt lifting eddies Ate produced hy oti1er
factors than by ftJ"ccioCl against the perimeter. One factor is velocity distribution as explained in
para 1St::!} above. The velocity distribution in thA vertical phme gives::t rollmg motion Top
water moving with rel3.tively higher velocity topples over and becomes the bed water, and the
bed wattr lagging behind rises up to take the place vacated by top water. This explains the
EtIong boilmg up to the surface which is usuaLy ~et,n-in the ca,e of large canals. In addition .
to the forward motion, there is also the rolling motion. Suc~ silt lifting eddies are many times
stronger tban those caused by the roughness of the perimeter. it is therefore that the large
canals are carah!.:: of ~arrying rplatively higher silt charge by we;ght ane grade.
K.B.KhLs'llani in hi_<; article in Indian Engineering. Calcutta, September, 1937 explains
the rolling the:'ty {"If waler c('nsidering th'lt the forces acting are in the f,lrm of a couple prodllcpd
by the fr.cii.Qnal rt sistance and the forward motion of the mass of water and works out the lever

aIm 'of the couple Z= f and his regime velocity formula in terms of Z.
 176

nii would be the case if the moving mass was a solid. but the ::Iuthor considers that the
explanation of the rolling water due to vplocity variation is relatively more sound. The simUar
lateral effect has already been found in his observations by A.S. Gi~b ai explained on page
177 in (d)
(d) Lateral silt distribution in a channel section.
A. S. Gibb, Executive Enginerr Punjab Irrigation, carried out observati')ns on Gugera
Branch, Lower Chenab Canal published in hi., paper No, 28, 1916, punjab Engineering Congress,
SeCTION Lahore. He made observations by floats, "The
/' path of surface· Hoats gave surface lines dire~tly.
and rod floats were aS3Umed to follow, lines givmg
a meln between snrface and bottom stream. The
line of bottom flow were dp.duced from thtJ observed
PLAN
surface and mean lines. Fig. 10 (opposite) merely
/_J __ -~-I--~-;;-}_.... I illustrates what is wen known. namelv that in an
.... _ _•.!.o__ ~ __ I f·n ordinary straight canal the bottom water flows from
---'-X:=-':R;A; ".'" the middle towarrls the margin, the surface water

f-"
(' - -.,;.nCi t ~'-.! flows b::.ck towards the middle, and there is an
l .~l)~_
-,
- .!uz.--__
----... ---- upwar d current III
. t he margma
. I strIp.
.
--,-;;;~AC'_K:-_
~ ~t"'"; _~ <," ~
--.Iti' _ _ ~, ~__ __J
Gravitation is the only other known force acting
-----,------ on the silt particlts. Consequently from these cross
I.
circulating currents, allowing for gravity, it is possible
Fig. 10 to get an idea of the a:tiv!tie5 of the silt particle~ in a
channel. Sand suspended near the bottnm of the canal,
as it is carried downstream, will re gradua Uy pushed along on or near the bottom (travelling by·
saltation as an American experimenter terms it) into the strip of water near the margin, there it
will come uncler the influence of the rising current, and a'l gravity is acting- against the current,
the ascent will he considerably retarded with th~ res:llt that water there will be more rlensely
charged with silt than elsewhere. At last on arriving near the surface the sand is carried away
towards the middle of the stream, fallin~ throJgh the wa'er :IS it goes, its fall being ass:sted
by the ultimate downward tendency of the current. Eventually the sand grains again come
under the influence of th,~ bottom current and start off on their rounds again. Heavy particles
of silt, it may be assnmed always remain near the hottom rather than [ght one:;, anr{ they are
thus more under the intlup.nce of the bottom current towards the bank. The d'~nsity of tte
silt charge in the marginal strips of water will be still further increased, oHing to the fact
that the fo[warri velocity L; le5s there than it is in the middle, and a given quaQtitv of silt,
whilp going the round above describpd, must be contained in a smaller volume of water when
it is near the margin than when it is near the middle of the steam. It app·..:ars, then th ..t
the marginal strip of the parent canal. which is to be drawn off into a distributary, normally
contains the most hi!!hly silt charged water in the whole crOl;S section, and tend., contin . lally
to be fed with. a specially selected snpply of the heaviest saur{ availabb. The c.ross currents
in q11estion result from the presence of th~ boundaries of the channel or in other w,)~·d., due
to the relatively reduced velo~ity ncar the sides and higher mean velocity in the mid stream. If
the vp.locitv near the side could be increased by providing pitching the difference of vebcity
rdative to the mid mean velocity shall be reduced and the amplitude of the cross waves
shall increase."
(e) Effe~t or convergence.
The effect of convergence has been studied at length by C.C. Inglis, Director Research
Institute. Pl)ona, India. If a bell mouth approach is construct"d as is US1B.l;y done upstream of
contr::.cterl meter flumes, the silt tenli to concentrate in the mid-stream. If water be abstracted
from tht! bell mouth approach, the silt charge on the off-tak~ is rela,ively less.
(f) SOt distribution at bends.
The cross section of a c.hannel at bends is as shown in Fig. 1 t
There is shallow depth on the inside of the curve and greater dept'" toward3 the
outside of a curve at a b~nr) in a channel. The inside tends to grow berm, while the outside
is eroded. There is flow of silt along the bed as shown by arrows in the Fig. 11.
 177

This results in ~he. conCtntration of silt

charge on the InsIde of a curve and
there is reduced silt charge towarfls the
outside of a curve .\\ hert the channel is
deeper. An off-take taking off water from
the out~ide of a curve cont~ins relative Iv
low silt charge by weight and grade .
Fig. 11
16. Diurnal and seasonal variations in the silt charge in a Channel.
(a) There is considerable variation in the silt charge and silt gr ade of the silt c:lrried
daily in an irrigation channel. A few observations from the author's paper on lJ"sign of
irrigation channels are given below showing variation of the silt charge card·d at R. D. 1351)0 of
the Bhek distributary, Lower Jhelum Canal. These observations were taken by the author while
engaged in experiments for determining the silt conducting power of different outlets.
Date Si;t charge in Disty: expressed as
ptrcentage by we.ght of W3.tr
16-5-1931 ,'102
55-1931 ,171
20-6-1931 .132
9-7-1931

13-8-1931
21-8·19~1
27-8-193-1
-- .275

,331
,350
.165
3-9-1931 .14
13-9-1931 .182

18·9·1931 .150
1-10-1931 .:203
2-10-HI31 .120
5-10-1931 .I!-O
7-10-1931 .150
8-10-1931 .160
-There
-- -
is a free fall R.D. laSOO. The longitudinal section of the reach upstream ot
at this fall was observed nc~asionany to see if it was silting or scouring during the course
experiments. The observation'l showed that the bed of the channel did neither silt UJ) nor
0'
scour during the period in '1uestion, in spite of the fact that therl'l was a ,~onsidf'r'lble variatiop in
the silt ch~rge carried in water from time to time. The Bhek distributary is II. rt'gime channel. n
has not changed its levels during 43 years of its running. The variation in the silt charge by weight
was from '12 to '35, about i. e .• 3 times. Two points are explicitly c{lnr from these observations.
(i) In the formula for the change in C V.R .• (Rc=T 11 / 3 r2'3. A 11&)
if the depth dops not change, and if r1 the silt charge changes. there must be corres-
I onding reduction in ri silt grade in order that Rc does not change causing neither
silting nor scouring.
(ii) Silting or scouring does not take place immEdiately as the stlt charge or silt grad.
is changed. TheTe is some natural resistance offered by bed bdore it is ero "led
(b) The seasonal variatio'l of silt charge is well known. In my observations on the
Lower Jhelum Canal I found it to vary from·5 per cent by weight in summer to a little ~elow
1 per cpnt in winter. 1he silt charge by wFight is maximum in the rainy s.-·ason when the pa.ren'
river is in floods. The summer silt is muddy and winter silt is fine a.nd coarSe sand. The
irrigation channels pa-;" tile high silt charge in the months of Jl1ne, J11ly and August "ithrmt
silting up. The channels berm up most in the months of July and August. when the fertilizing'
fine s'lt charge is very high. Deterioration of channels in the rr.onsoons is mostly due to
.berming up. The irrigation distributing ch:mnels silt up most in the months of October and
November mostly due to the falling p~rent river picking up coarc:e silt in digging the winte)!
channel and partly due to the water picking up of the coarse silt in the canal itself due to th~
large reduction in the silt charge by weight.
 178

t 7. Silt attrition.
(a) The subject of silt attrition in canals has received little recognition so far and it is
generally argued that on account of the low velocities in the canal, there is no possibility of
silt attrition. MoreoVfr. it is sometimes argued that the progressive reduction in the silt grade
is not Clue to the silt being worn down by attrition, but due to thp. silt selective behaviour of
flow in dropping- coaT~e stuff and picking fine sh,ff. This wculd be the case if the velocity were
reducing but in Ion,; lengths of canals with fl.) change in velocity, as the lined Bikaner canal,
the redllctinn in silt grade mmt mean nothing but silt attfi tion. In rivers the part played by
attrition is evident. The pebbles (grit) are ground flown fiIst to coarse sand and then to Lne sand
by the process of rolling and impact against the bed and among themselves, as they are
transported from the hills to the plains. The silt attrition is therefore an establIshed phenomenon,
but the limits of its operation remain to be determinea. rhe velocity at which tu!"bulence
start;; is known to be al-,out '80 foot per second, Thoore mmt be a sfc::md critic::ll velocity beyond
which the silt wou'd drag along the bed and the condit:om n .. cesoary for silt attrition would
come into play. D, Owen determined such a vdoc:ty to be equ:l1 to about 2'5 feet per second.
Observ:ltions and experiments could alone determine sllr.h a velocity in actual canal practice.
The author saw from below plate glass bed in the C'lll!"se of some other experiments that the silt
was rolling and dragging arong the hed of the Mith1.lak Di3tributary, Lowpr .Thelum Canal,
when the velocity was onl:? 2'0 feet per secon n. It is pro'Jable that the second critiGal velocity
necessary frlr silt attrition might be va.ria:ll0 with the q,untity of the silt to be d:-agged along
the befl. There can he n') denying th'~ fact that the drag~ing of the silt along the bed always
ta'<0s IJlare at least in the case of main and bra1l6 canals. The effect 0 f the silt ~ttritioD
mu"t :It least come in th+'! design of the1'~ channels, The author in his paper on "Design of
lrrigation channels,' worker! out the effect on the value of C,Y R. if the silt grade changed
from fl to fa by attrition There was then no authoritative record of experiments to determine
the actual change in silt grace. .
(b) Experimerits have been carried out by C.C. Inglis C.I.E, Director, Central1rrigation
and Hydrodynamic Research Station. Poona.
"The object of this exp'~riment with silt abrader is to see whether abrasion can be an
explanation of the feature of large rivers th::lt the bed sand becomes progressively finer from the
hills to the ~ea As described in the annuall'eport for 1937-33, the Silt abrader c()nsists of a
circular tank of 2 ft. radius. 4 feet high, with a central shaft to which three tiers of hlades are
fixed, each tier being set at 40 degree to the other two, The shaft was revolved 30 times per
minute. so the velocity of the blades at the outer edg~ was 6'9 feet per second.
Sukkur silt was placer! in the tank to a depth ot 6" and water maintained at a constant
level, 3'-6" over the bottom of the tank.
The abrader was work"d for 1637 hours (68 days) since last year's report, the total
working hours up-to-date being 8787 hours (366 days). At the end 01 this time, the silt on the
bottom was thoroughlv mixed and analvsed by the Puri siltometE'r.
It will be spen, that hetween 7150 and 8787 hours abrasion of silt particles of 0'25 to
035 mm grade took place, resulting in the increase of percent silt between 0'11 to 0'25 mm,
diameter. Thtlre wa~ verv little perrentage increase <0'08 mm., which shows tbe limit of abrasion
becau<e helow that limit, the silt goec; in suspension, "Table II shows the progres~ive
reduction in the graile of silt at various periods during the experiment."

Table II.
fd=J.76y'dmm ba3ed on
Total hours of working dmm. particles 0.075
.... mm. diameter
Original sample 0229 ,_ 0'84
3775 hours (t57 days) 0216 0'82
5S8l hours (224 days) 0196 0'78
6220 hours (259 days) 0'193 0'77
7150 hours (298 days) 0'181 0'75
8787 heues (366 days) 0'169 0'72

[c] The Poona experiments dearly show the cff!C!ct of sJlt attritirn. or abration, _ nut
aClual obst:fvations in canals are m:eded where the discharges and the velCcitiCs do not change
 179

in long lengths as main line Upper Bari Doab Canat Bikaner lhed canal or Haveli lined canal
to determine the act.ual reduction in silt grade.
18. Silt movements.
The phenomenon of silt movements is peculiar to the large distributing channels in
the tail divisions of a large canal. The channels of the Jhang Division Lower Chenah (anal
in the Punjab (]hang Brancrl lower in its tail reach), Bhango Branch Dhaular Distributary
and C.ultan Pakha distributary are especially notorious in this respect rbe water leveL;
suddenly rjs~ fr,)m'5 ft to 1'0 ft. depending on the size of the channel in ilS middle reaches even
though no excess supply has r fen admittrd at the hf'ad, This results in suious leakage and
brraches The authnr had a personal experience of this trouhle in the channe1~ menticnf>d above .
. It was at first argued that there was actual increase in the di:;chargc of the channel
at he'l.d due to relatively muddy water coming in the canal. This is quite likely when the
regulation is done with referoo.r:e to a gauge fixed downstream of a bead in an earthen channel.
There is about 10 to 12 % increase in discharge, if muddy water follows a clear water flow
turn in earthen ch;mncls without any change in the section due to reduction in the co-efficient
of friction. J. P. Gunn Executive Engineer, Puniab frrigation carried out detailed observations
on meter flumes and found that an mcrease in discharge dlle to muddy water was 3 to 4 per-
cent. The remedy would therefore be to construct meter flumes near the head of snch.
channels for precise reguhtion. The Dhauler oistributary was provided with one at R. D.
500 in 1925. This has not. however, cured the silt movement troub~e which is now (1943) the
worst ever experienced.
There is one peculiar point about these silt movements that they are the worst in
the beginning of November and about th", end of Fehruary every year, synchronising with
the change of 5easons. TbA Authe, considers that tbe silt movements are primarily due to
temperature ~hanges and the r.omequent viscosity changes in the channel baving fine b~d
silt. The bed silt is lifted up in water and the piling up of water is due to the unerona')le
bars in bed or the drownrd bridges and the increased fiction on acconnt of relatively coarse
silt in suspension.
The silt movemtnts are no doubt intensified in the lower reaches of a canal. if there
are a lot of rais~d crested falls in the upper reaches suhject to long periods of low supply an·t
also if 11)t of silt exclusion has been resorted to in the case of distributary h~ad regulators in
the upper reaches of the canal by constructing skimming platforms.
19 Berming and Siltillg-up of Channels.
The distributary channels silt up most in the head rea~he_; and b:mn in the tail reaches.
The mindle reaches are generally free from sUt trou)le.
Tile rea:IO:1S for sihirrg up in the heli relches
Y' I
[iJ Non-regime section. I~
.-•.f-. 1""'; ., t"t "t 1l:1 r.,,_r •
I
rf the spction given is not suit;l.ble, tte cbannel water would drop its cargo of silt which
it cannot carry in the head reach and the lower reaches will thu; d0al with silt in water, out of
which objectionable silt had already been dropped in the head reach _5.
[iii Defee tive Bead Regulator.
If the head regulator design is such tha t excessive silt charge enters the distributary,
the coarse silt would raturally drop in the head reach.
[iii~ Insufficient slope.
If the channel~as been given insufficient slope, it would naturally tend to incrPil.',e the
slope by silting up in the beginning of every reach in the channel downstream of the control
points.
[ivJ Defective outlets.
If the outlets do not draw their due share of silt, the channel would silt up in. the
head relch mostly a.nd iu other reaches to a less extent.
 180

[v] Fluctuations in the supply.

If the channel runs long pf>riods of low supply it would silt up most in head reach to
adjust its cargo of silt from thp. reduced depth and velocity.
After the adjustmflnts of the silt C::lrgo in the head reach of a channel,· the middle
reaches pass off the supply safely without silt trouble. The lower or the tail reaches of a channel
are again the source of great trouble dUll to berming up. The reasons are enunciated below :-
[i] LOW velocity.
On account of small depth and small velocity water is capable of carrying a reduced
cargo and low gratie of silt. The side velocity being very low, fine silt deposits there to being with
[ii] Growth of grass on the berms.
The growth of berms flourishes most in the tail reaches due to low velocity and the
fertilising fine silt deposited on 1he sides. The deterioration of the channel section start!; from
the sides. Water is headed up on account of the rflduced section, and spreads over the berms
with the consequent rise of the bed. The berm rises up by grass growth and silting in turn
in bed encourages further rise in levels. It is, therefore, that berm cutting in the tail reaches
is "ery necessary to feed the tails in the months of July and August.
~O. Silt Sampling an:l Analysis.
[AJ Silt is transported by a stream either in suspension or by rolling on its bed. Silt
sampels are,analysed as below:-
[iJ Coarse sar.d, i e. particlfls above .2 m. m diameter.
[iii Medium sand, i.e. particles between '2 and '08 m. m. diameter.
[iii] Fine silt, i.e. partfc1es h( tWeen 'OS and ·02 m. m. diameter.
[iv] Dissolved material and clay below '02 m. m. diameter.
Item (v) i'l ignored and the summation curves are got for the first three items to work
out the mean diameter. Silt intensity can be calculated from the mean value of coarse sand,
medium sand and fine silt observations taken at a prticular site by the formllla:-
Silt intensity in eft. per% eft. =grams per litre X 5/8 and silt intensity in eft. per
cusec-day=grams per litre X 54.
[B] sampling, of suspended silt.
[a] Bottle -:SQ)pler.
It consists 01 a brass frame bolding a one-litre bottle fitted with a rubber stopper.
The stopper is operated by a lever at the top of the suspension pipe. When the lever is pres5eo
down, the bottle is openc(l. The length of the suspen!'ion pipe of the sampler is varied to suit
the depth of water in the chanr.el at the sampling site. Tte two essentials regardinl'!' the working
of the sampler are:-
[a] The roouth of the bottle slJOuld be opened only when it reaches the >requiren depth.
[b] Th~ mouth of the boWe should be kept open for the minimum time required to fill
the bottle .
If the bottle is kept cpen for a longpr time than that requirpd for actually filling the
bottle, coarse silt particles keep on falling into the bottle f>vpn after it is full. The time required
to fill the bottle should be ascertained experimentally by the observer. The iliap<lm and detailed
description is given in the Punjab Irrigation Reseach Institute Publication Volume II No. 15
of 1936.
[bJ Tait Binckley Sampler.
The Tait Binckley S::lmpler ccnsists of a ((ntral pipe with rubher extensions on both
sides. The f'xtensions are twisted by an arrangement fitted at the top of the suspension pipe to
enciose a sample of water.
[c] Uppal Samplc:
The. Uppal Sampler, as developed by Dr. 1I. t. Uppal of the lrrigatiotl Research
Institutt>, Lahore, consists of a brass barrel with guides fitted at each end. Brass diaphragms
 181

move in the guides, vertically, like the shutters of a photographic camera. Both the dia.phragms
and guides are ginn a slight taper to secure a leak tight contact. The top end of the diaphraa-ms
are conn~cted together bV an iron rod which is worked by.a lever at the end of the 5 uspen;on
pipe. Detailed descripto) with photograph" is g;ven in the Report hr the year ending April.
1940, of the Punjab Irrigation Research Institute. Lahore.
{r.) B:lr} Silt Sa;npling.
The old method of filling cigarette tins from the oed of a canal in a closure is !'lOW
discontinued as silt thus sa'npled out is not considef(~d to be a repre~entative sample of bed silt
carried in w14ter. cecause wh~n the canal supply is dropping before a closure, some fine silt in
suspension also drops down which i, again picked up in full supply conditions.
The samples of bed silt are taken by m~ans of an app1.ratus which is described in
detail, in the Punjab Irrigation Research Institute publication Volume 11 No IS. This
consists essentially of an eccentrically mounted scoop which dig5 into the bed when revo~veri
and a cowl which protects the sample from being washed away when the apparatus is brou ght
to the surface.
To take a sample, the apparatus is lowererl to the bed and the top rotated which in
turn rotates the scoop by m ~ans of wire. It should be noted that the pip~ forming the handl!1
is kept clenr of the cowl so as to avoid the water in the pipe interfering with tho sample. The
sample is air dried and approximately one lb. kept for examination. The following hydraulic oatas
are obtained at th~ time of sampEng:-
Q (Discharge), A (Area of cross Rection,) S (slope of water surface X I02).R
(Hydraulic mean raclius) , D(Mean dcpth).Pw (Wetted perimeter). B(Becl width), N(Kutter's
co-efficient for rJughne;; and G (mean ga1Jge of the ch'lllnel during discharge observations).
The temperatu:-e of water is alsl") recorded at the sarr.e time.
(DI Total Silt Charge Sampling ill a Canal.
The me!hods describ~d in paragraph (B\ above omit the silt rolling nlong the bed and
the method described in (C) above omits the silt in suspension. Th~ author used a scoop
in a trough intercepting the jet at the downstream end of a free fall long crested weir to
collect II total silt charge sample Thi" method is described in detail in the author's paper No.
168, punjab Irrigation Congre:is, Lahore, 1933.
(E) Silt Analysis.
(a) Ke:medy's Siltometer.
De ailed description is a availahle in pap~r No.9 Technical Publica' ion cl:tss A. P. W. D ..
Irrigation Branch, Punjab. It is now obsolete as thi" method gave qualitative results on l ". Thtt
dropping silt in a tu~)e was collected in a gradu'Ited bottle and the times of filling the variolB
divisions were noted.
(b) PUri's Siltometer.
Th's is an improvement over the former. A detailed description is availa}lle in the
Punjab Irrigation Research Publication Volume II Nos. 7 l\ud 9. nlC bottom of the sijtomp.ter
tube opens into a circular trough proviaed along its circumference with detachable cups which
are just large enough to b~ covered by the bottom of the tube, e.g., after 2 !,26.::iO seconds, so
that each cup comes in turn unner the tube and receives the particles which reach the bottom in
the interval between two movements of the trough. When all the silt has passed down, the,
cups are taken out and the silt in each is washed into a separate glass tube with a narroN long
bottom graduated in cubic cp.ntimeters. The valume of the silt which settles down io; rea.d
assuming that the density of particles of all sizes is the same. we gd a proportion to the
weight of the particles received in each cup.
The sizes are read from a table prppared and the volumes for the cup,;; are Stlccp.ssive
totalled up to give the total volume below each size. This is then plotted against the siz': to
give a summation curve similar to the pressure size curve obt4lined from the optical siltometer
described below.
(c) Vaidianathan's Optical Lever Siltometer.
This is the most correct method of silt. aMlysis. 1t is described bv Dr.
Va:dianathan in paper No. 167, Punjab Engineering Congress, 1933. A prief description
is given below;-
 182
.Optical Lever Siltometer
(i) When a given sample of silt is released from the top of the siltometer
tube the arriv,Ll of various partiCles at the bottom of the water column is
indicated by a change in the pressure as shown by the fall of th~ mercury
meniscus (M\ in the m~nometer Fig. 12. This is magnifi~d by the optical lever
arrangement and the movements recorded on photogr lphic paper. The light
is intercepted at known intf>rvals {usually one second} so that a spiral curve
broken at the corresponding intervals is obtained A base lin'; from which
measurements have to ce taken, is obtained when the slIt has completely
M passed down.
The onJiTlatesof the spiral curve are measured from the base line
at the varicus breaks in the curve. They give numbers proportional to the
pressures of various in"tants,
The theory connecting the time of fall of the particles at a given

,. temperature of the water column with their average size is described

at some length in paper No. 167 Punjab Engineering Congress.
1933 Here it may only be stated that from a knowledge of the time and
Fig. 12 tbe temperature i~ is pos3ible to read the average ~ize of the corresponding
partcles from a previously prepared table. This t::tble can, however, serve
only for a given length of water column and has to be sl'parately prepared for each silto meter
of a differf'nt If>ngth,
The rate of change of pressure at any given instant on the other hand has been sh()wn
to he diredtly proportional to the rate ot change of the weight of particles deposited at the
bottom of the siltometer tube and consequently the rate of change of weights is directly known
fr Jm the pressure readings. '
Pressure-Summa tion -Curve Thus we get the corresponding valnes at
any instant of a number rroportional to the
.... total weight of particles deposited up to that
" instant and of the aver"ge diameter of the
partides arriverl. at the bottom since the previous
measuremt'nt of time, These values are entered
on a sheet and are plotted on a graph. A free
hand curve pas3ing very closely through the
plotted p~int is now drawn and the curve and
the values are passed on to the statistical section.
Form of curve is shown in Fig. 13.
Iii) The percentages.
The plotting anrl. the curve are checked
first and then the values of the presslJres read
from the curve for integral values of the diameter
Particll' size in mm. of the particles. If the value of the pressure
Fig. 13 for the datum line is not zero it is then subtracted
from each value We consequently get a series
of numbers proprtinnal to the weights of all the, particles . whose mem diameter. is less than
the correc;ponding intgeral value cnterer! in the prevlou5 column, The las~ number 10 the column
(savL! is proportiunal to the total weight of the silt !"ample. . , .
'Eash number is then divided bV the number Land multlpltpd by 100 to give the
percentages by weight of the particles below. pach size, The last are w:lat are known at .tht
"summation" pncentages, ie, they re Jrl'spnt 1h~ mms, of the ~epar;Jle per,centages by weIght
of all tIll' particles whose sizes are less than the glv0n sIze. If the snmmat.lOn percentages are
plotted against the values of the mean diameter we get the so called SUillmatton curve.
(iii) The si'le distribution curve.
Each value in the summation percentag" cohmn is next suhtracted from the value
just su('cee(:ing it. Th~ difference.s form a new ~olumn giv~ng the "dish.i mton ~ercpntages:'
lhese give the proprrtlOns by wClght of the fractIons of particles whose dIameter hps bf>tween
 183

the value .of the diameter .entered in the p~evious column in the same row and its next lower
vallIe. It 1S thought suffiCIent, for all practIcal purposes to take the mean of these two values as
the mean diameter of the fractions.
The distribution percentages a"e next plotted as ordin'ltes against the diameters of the
fractions as abscissa. A free hand curve passing closely through the individual points is next
drawn. It shows how the particles are distributed acording to the m~an sizes and is known as
the size di,tribution curve. A typic'!.l siltogram or the paotographic record of the analysis of
the silt sampl\! is shown in Fig. 14, and a typical sllmmati)ll curve in Fig. 15.
SILTO::;'RAM OF BED SILT

.--".

£--"tii~ -----8 AS..£.

'L ~}J;L&'
. r J

Eig. 14
SIzE DISTRIBUTION CURVE

Fig. 15
(d) Uppal's Air Sntom,ter. ~

Siltometer for quick and accurate analysis oj: silt samples has btltlD developed by Dr.
'H L. U ppat A jet of air bas been utihzed to separate the par tid-es ,belonging to diffenmt
grades.
The apparatus consists of the following parts ;-
N A blower.
(iij A stiHmg cbamJer.
(iiJ) An air tank.
(iv) A collection chambet.
A standard blower supplied by Messrs A. Gallenkemp has been 'Used. The speed of the
motor is aojusted by means of a tachometer. In the stilling chamber, which consists of a
woorlen box 2' x 2', the air is stilled befote it begins to flow in the air tank. the air tank is
m~de of glass plates 4'x3 i and oW apart. fhe top two feet pJrtion of the tank is used for th'"
passage of a stream of air, whiie the bottom one toot i" occupied hy th0 coilection chamber.
A sample of the silt is introduced at the top left hand comer of the air tank. The air blast
separates it into different grades which collect in the collection chamber bel<Jw the air tank
The collection chamber is divided into one hundred compartments by mt'ans of celluloid
 184

strips. At the top-ends of the s t 6p" fin" needles are fi Ued so as to a void any turb~lence which
might be caused by the jumping of thp particles. Each compartment of the collectlOn Chamber
has been calibrated for a known speed of this bhwer. A direct distribution curve is obtained by
the use of this siltometer. The bottom of the collection chamber is removable so that after
analysis the sample can be readily tal,cen out.
A large number of samples have been examinei in thii apparatm and it has bl'l3n shown
that duplic:ltes :lgree well and that the method is rapid since it Saves a considera">le a.mOUI:lt of
labor:ous calculations.
21. C.>rrelation between Silt and hydrauli'.3 Data
Size Distri:mtion Curves tflongh characteristic of variou" silh are not con venient for
defining them for purposes of correlation with the other hydraulic data of the channel. ~or
tUs purpose it is neCeSS!'lTy to define the silt by m~ans of" smgle values" capabb of be~ng
derived from the summation of distribution curve. The single values now being used are:-
(i) Weighted Mean Size "m".
The summation curve of a silt sample gives the percentage of paricles belo'" any gi'len
size. Bv taking the reaclings for two ginn siz's and subtracfing we can a~certain the pe:centage
of particles who~ e diameters lie between the two sizes and w~ic 1 m ly very nearly be assumed
as having a diameter lying mldway hetween the two siz~s. If this last value is multipLed by the
corresoonding percentage, and the sum of all sur.h products is divided by the S,lm of p~rcelltages
(usually 100) we get a value for the mean diameter of all the particles contained in the s tmplc.
This is the wp.ighted mean size or"m" and furnish83 a useful measure of the degree of coarseness
of the sample.
(ii) Standard Dcvia1ion "a".
Every sample does not consist of particles distributed in exactly the same way, and it is
quite possible that two samples with the same mean size may differ, one hiving a preponderance
of parti.~l~s with diam~ters near the mean size. and the other with diameter va.rying much more
widely. Hence it is necessary to know how the variolls sizes are distribuh'd a:7out th~ mean
size, and consequently the "standard D.:;viation" is calculated as a measure of their" dispersion."
To o~)tain this the deviation of each size from the weighted mean size is squared and multiplieci
by the correspondiug percentage and then the sum of such products is rlivided by th\) sum of the
percentage. The square root of the quotient gives "a"and the smaller it is the more uniform
can the sample assumed be.
(iii) Schoklitsch Number or uk".
The maximum diameter of a particle that can be determined bv a siltometer is 0·6
m m., so that in practice the mass curve is bounded by the 0 and 1.1'6 m.m , ordinates and the
line (Distribution Curve) may be taken to repre,en~ a normal mas .. curve. If then the area A
lying above and to the lp.ft of the mas, cnrve, be divided by the arf13, Riving below the mass
curve :md to its right the fraction AlB remains constant as lor g as the I mits 0 and 0·0 remain
unaltered.
This was pointed by Professor Schoklitsch, who found that this fract:'on worked quite well
for specifying shingle~, lthough open to many theoretical objActions). This number is referred
to as the Schoklihch numher or "K" and when not otherwi;;e l".pecifined is taken to refer to
diameters lying between 0 and 0'6 m.m. Should it be desired to specify these limits more
particularly it may be written as 0'< 0·6 in th~ same way 0.51, 2·0 would refer to mixtures whose
mass curves lay betwpen 0-5 2·0 m.m.
It has been often asserted that the presence of fine suspended matter in the watpr will
alter the viscosity. so a series of rough te,ts were made on water containing from I oz. to R oz.
of fine susFcnded matter per cubic foot. A plot of these values is "hown in Fig. 4
of the Punjah Irrigation Rpsearch Publication No. ] 5 Volume If and with the
exception of one point they lie on a fairly good line, examination of thA
wide point by statistical mf'thorls showed that Chauvnnets criterion ju<;tified its reduced
(Brunt D. "He Combination of Observa.tion" 1933 pag!) ]30). Th~ v!sm>ity in~reases from
852to 914 from pure \\ater to oz. of ~ilt per cubic foot, which may be se'n Gn a reference to the
approximate tables to be that produced by about 20 0 difference of the temperatum. It m'ly
 185

be "Concluded, therefore, that the normal amount of suspended matter which rarely exceeds
2 oz. per cubic foot. in canals makes no practical difference to the viscosity. The temperature,
however, is recorred at the time of observation so that suitable correction may be applied at
any time, if found necessary.
It might te argued that though the fin} suspended matter bas no effect on viscosity, it
would exert some direct influence on the movempnt of bigger pa'ticles. Silt was, therefore,
analysed in the Puj Siltometer using water w th increa5ing quantities of suspended matter.
The f<'sults l,elow show that silt constants a e unaffected 1 y the presen~e of fir.e particles,
TABLE
Effect of Suspended Matter in the Water on Silt constants.
(Sample from Lower Gugera Branch R.D. lEOOO).
:::'usj.>elll,ed matter Ill. a K
Ounces per eft.
~------------

0 '3U 11 '892 1'008

'12 '2894 890 0'932
16 '2952 '904 0'969
'39 '2993 916 O'~99
62 '30(:2 '946 1-001
I'll '3022 '988 1'015
4'09 '2958 '89(1 0'972
S'U '2950 '862 U'367

It has already been shown that L:>cey's silt f<letor fL = I'76m where m is weighted
mean diameter of bed silt exclnding m<ltter below '07mm diameter, and Kennedy's critical
velocity ratio X varies as below between two sections Nos 1 and 2 : -
,," .33 ·3 ,2

-~:- = ( ~~~). ( -~~- ), (-§~-)

whHe X=C.v,R,
W = Vv'eight of total silt charge carried and rolled.
m=Weighted mean diameter of silt carried and rolled excluding soluble matter
and clay below '67 diameter,
Q=Discharge in cusets.
22, Wee:! Growtb.
The weed trouble in the Punjab canals does not exist because they are fed
directly from the riVf'fs. The weed trouble in the seepag';: drains in the Punjab is very acute.
The Deccan canals fed from the storage reserVO;f3 usually suffer from serious weed trouble,
A very authoritative account of this trou')le has been describen by c.e. IngLs C.I.E. Director
Poona Research fnsti:ute and V.K. Gokhale. in Bombav P.W.D. Technical Pape" 1937-38
on "The I~radication of water weeds from the Dec:an ('ana1s." /\ brief summery is given here
and the student should refer to the original publication for a detailed account.
(a) During the mOll"ocm, the Deccan Canals obtain most of their supply from rivet
flow; but for the rest of tl e year they are dependent on suppl es from IDrgE' artificial lakes
formeo by the construction of ma,>onry dams across river valleys in the foot· hills on the eastern
side of the W.,stern Ghats. These d'lms are ahout a mile long and vary in h~ight from 100
to 270 feet. The Canals, eXcf'pt the Mutha canals, take off from above pick. up~weirs constructed
acros~ the Rivers som~'20 to 53 mill'S downstr~am of the storages. These weirs vary from
10 feet to 40 feet in height; and from 890 feet in len~ th in the ca.se of the Paravara pick-up~weir
At Ojhar, to 3618 feet in the case of the Godavari pick-Up wp.ir at Nandut Madhameswar, Tn the
caSA of the Pravara left, the canal when first opened in 19]9 was capaJ.le of giving a discharge
of 500 cusecs with only 6 feet of water, whereas by 1922 only 272 C1!secs could be obtained
with that depth and only 73 ~usecs or 15 per cent of the 19]9 fignre after the weeds had grown,
within 2t months of a full weed clearance. Weed growth is not uniform throughout the yeaf.
It generally starts in October, and by Deceml:er it is heavy and under the old method of
 186

running canals, another clearance was required early in March and another ahout the end o(
April,
(b) The following water weeds are found in the Deccan Canals : -
1. V::lllisneria spiralis,
2. Potamogeton pectinatus,
3. Potamogeton perfoliatus,
4. Potamol'eton indicus,
5. Ceratophyllum demersum,
6. Hydrilla verticillata,
but for simplicity, water weeds may be dividen into only 2 groups :- .
1. Vellisneria spiralis-a gr3ss like weed, up to 14" in length, whIch grows in tufts,
and
2. Potamogeton varieties and associates-the worst of which is Potamogeton
perfoliatus which gr()ws to a length of 12 feet or even more.
No.1 is the first ,"epd to appear, and is comparatively, unimportant except in small channels-.
No. 2 indicate~ a much more serious condition of wee).
,c) Factors which may be expected to affect the growth of water weeds.
[il Infection,
[ii] Temperature of water,
[iii} Chemic<t.l C Imposition of water,
(ivJ Velocity-effect on weeds due to silting velocity and mechanical, or
drageffect,
[vJ Depth of water, . _._.~~I___._____
. [iv] Deposit of silt, ..
" [vii) Rpduction of light due to turbidity and depth. " :/
Several inherent difficulties make the proof of relative effect very difficult.
fi] Infection
Under Deccan conditions, water weeds grow profusely in all nalas and p
infpction is universal and this enquiry showed that there is no relation between weed growth in
canals ani! intensity of infection.
[iiJ Temperature of water.
In the Deccan water the temperature varies between 200 and 300 C, and he nce it is
highly favourabl<t for weed growth.
[iii] ChemicnJ composition of water.
There is no traceable relation between intensity of growth in various canals and
composition of clay has a secondry efect; because it increases turbidity and reduces the
amount of light at the bed.
[ivJ Velocity has little effect except where it exceeds 2 feet per second; but it ha.s
an inilirect effect. in that when the velocity is in excess of regime velocity, deposition of seeds
and cuttings dces not occur.
[vj Depth of water. Water weeds can grow in clear water at depths rlown to 18
feet.
[vi] DeposiHon of silt. It is undoubtedly favourable to weed growth. but deposition
only occur where the velocitJ"is cons'derabl~ below regime velocity a.nd s11ch low velocities
do not occur in propedy tJesigned channels. Whpn canals are newly opened, they are generaJly
run with low discharges and water has to be headed up at rpgulators'. This reduces the velocity
and causts l~eposjtion of silt especially along the banks, rendering the water clear;
so that all conditions arc tr.en favourable for weed growth. Thereafter the w£eds twd t)
persist.
[v;iJ Reduction of light due to turbidity. There is r'a'iOn to anticipate that exclusion
of light will reduce \\e~d growth, in that light is required for carbon d.Ssirnilation and tissue
building (photo-synthesis) and there are no cases in whir.h "eeds (Trow wr.ere the water is turbid
throughout thl:; year. \Vepds generally grow luxl1riantlv where l'iIt is depositing, sevpral faetors
being favourable then; but weeds do not grow wherr, silt is depositing, unless the water is also
relatively clear.
It will be seen th,lt turbidity was greatest at the head of the Nira Left Bank ('anal
(except in a fair season of 1928-29) when. there was a 11l.arked drop !due to discharge 1.;t out
 187

from Bhatgar being more constant tha.n in other year~). This canal has alwdo}s been more free
from weeds than any other. The Godawari Right Bank was the only other canal in which
weeds were, at this time, decreasing; and the disc rpading were also small. In 1930 there.was
a marked increase of disc reading in the Pravara Left Bank Canal and thereafter weed . growth
became serious. The worst canal for all weed-growth Was the Mutha Right Bank Canal in which
the disc reading generally exceeded 4 f~et.
(d) Conclusion:-
[i] In every case examined, weed growth and turbidity bore a close inverse
relationship;
[ii] No othM single factor inhibited weed-growth unner normal canal conditions.
[iii] Where D>3d, no weeds grew, where d=Disc reading and D=depth.
[iv] Where D>oz/2. only Vallisneria grew. .
[v] Turbidity, which le:Jds to exclusion of light was the dominant factor controlltn~
weed growth in Deccan Canals.
The growth of water weeds in the Deccan Canah h:ts hitherto interfered to a serious
extent with Canal administra.tion, due to the fact that closures for weed clearance were
necessary in the hot weather at intervals of 40 to 50 days. These closures, which took place
when the discharge had fallen to minimum supply, necessitated an increa"e in the period between
waterings and considerable dislocation of normal supplies, with the danger of mistakes and
the certainly of water being taken out of turn by SOill'} irrigators. The consequent dislocation
of normal conditions did not end when the canal was reopened, but persisted for about 2
rotations after each closure. These closures inevitably le<fto serious loss to irrigators, a severe
strain on staff ano a considerable amount of unavoidable friction. ..
Temporary labOlJr had to bj collected, often a difficult matter,· and· w)rk had to be
carrif'd out against time, because no mechanical method 01 weed clearance has been found which
i5 as satisfactory as doing the work by hand.
The rush rotation system has been so successful (nn most of the De,;can canals) that
th~re is danger of its being looked upon as a })enacea. Where, however, the slope of a canal
is very small, it may be very difficult to drain off the stagnant water completely, and where this
occurs, the wp.eds will not die off and may even increase. There is also a danger that some
of the weeds will aceo nmodate themselves to changed closure conditirms, and persist, because
they have a marked capacity for adaptation, and arl~ descended from lane} plants.
In certain cases it may not bG desirable or even possible, to adopt the 'rush rotation'
~ystem. Weeds will, 1 ow<!ver. be inhibited where there is adequate turbidity. .
Und"f Deccan condition no weeds grou,. where the depth exceeds three times the depth
at which a 3" white disc just rlisappellfs from view; and where the depth excepds d 2 /2 (rl=Disc
reading) the Potamogetion group does not persist.
·Where natural turbidity in the parent river is inadequate, it can be created by
providing a pick-up· weir designed to hold about three days' ~llpply, for range of 2 to 5, fed
in water level.
Water can then be let out from storage. intermittently. flow being run for say I day in 3.
Wavewash will then ervde the exposed face of silt, deposited during the flood season. and will
keep this silt in susp~nsion until it reaches the head of the cana.l. Where. howevc:r, the pick-up-
weir basin is to'1 large (Lake i'ife) or silt~d berms with fairly steep faces have not hact time to
form (Nandur Madhmcshwarl or where the pick-up-weir basin is too small (Ojhar weir at h~ ad of
Pravara canal~) turbidity will be in defect. Tn such cases, weed growth can be reduced by
running velocities considerably in excess of regime velocities throughout the yea.r. .
2j. Lined Channei Sections.
K. B. S, I. Mahbub Executive EngineElf, Published a very cpmprehensiv0 paper No 260'
on "LiniI:g of Channels" in the Punjab Engineering Congress Proceedings. 1943. A brief summary
on design of sections is given here
(.-\) Design of Sections,
The best form of lining section would S(~em to be an arc having sloping sides, more or
less at t.he same slopeas the angle ~f repose of the soil. The arc in the bed should b(! ta.ngential
to the Side slopes. It can be eastly shown tha.t we get the mOst econoIIlical section, i.e., the
 188

maximum area for the minimum wetted perimeter if the centre of the arc is at the F, S, Line
with radius equal to depth,
This section is also useful as it has a higher silt carrying capacity than a wide shallow
one, During low supplies, heading up has to be done at off-take sites which would cause some
silting. This in turn reduces the velocity and also the value of 'fL.' In the sp.ction proposed,
however, with no level bed, the silt can deposit on and effect. the rugosity co-efficIent of a
relatively small portion of the perimeter and hence cannot greatly affect the velocitits, The
effect on 'fL' also would thus be corn.spondingly less.
The side slopes may be kept as 1 : 1 for radii less than 1:2' and Ii: 1 for radii over 12'
provided that the angle of repose of the soil is not flatter than I! : 1, as shown in. Fig, 16•
~~~~~ ______________~~~~__________~__-,~L;~8-~.
. .
,
"

Fig. 16
The sectional data would be given by the following ~ormulae:~
Distributaries C:lnals
Side slope, 's' 1:1 It : 1
Section area 1'78 S2 1'925 52
I • '
WcUed perimeter 3-565 3'85 s
Hydraulic mean radius (R) 0'5 S 0-5 s
The velocity according to Manning's formula
III
=!:_48() R2;a 51/2:;:: !-~8~R2/3 (~_)
N ' N' lU,OUO
where 51 is the fall in 10,000'
with N=0-018 we have
v=0'S26 R2j3 Sli 2 and its solution is given in nomogram Fig. 17,
S1 ~~ f or
2 3
' harge= A V= 0. 9 2s 2'-3 •5).
:. D ISC 12 f or SI'd e s 1opes I' : 1-and 0 '995 2/ Sl'd e s 1 prS
v =0'40s t SI
Ii : 1 and Lacey's fL=0'7Sjf
2

Graphs connecting Q, 5, s, fL and v have been plotted for side slopas 1 ': 1 with
N =0018 (Plate XIV -A) and these can be conveniently used for design.'
The free board may vary from I' for D=4 and :l' for L·= 16, which would be e::J.l,I.iY'!olent
toO'S DO-5 ' /

(B) The limiting depth,

The limiting depth of this section for all practical purposes may be taken a<; IS' and so fot'
discharges exceeding 2.000 cusecs, an alternative type as shown in Fig, 18 may be adopted. This
was adopted on the Haveli and is also proposed to be used on the Thal Main line'
, 'In this the sectional data is given by the following formulae. '

When P=~.!C=~ aud N (Manning)-O'018

D D
Side slopes's' == 1 : , ' '
Section area
Wetted perimeter
=D2(p+ 1'7854)
-D~p+3'5807)
.I
 189

LINED CHANNELS

HANNING's FORMULA

.~"
.coo ""'~
f.:l"
(.I;
C)
~
0
z.....
Soo \l
10·1) rn
.:::::
u
.,..
1&1
,...
;Jl

~I
100
::::
....... 0
..e
.... 100
fC
IU
l.
p:;

~I
I ••0 z
!it .... ~
c., f.:l
'Do 1&1 :;?l
0 ....
....l reoo
Q.. ....
c...:>
tn f.r "- ....l
.... ~ :Jl
<
4'0 III
IU : 'til, fL 'I et:
').. A
I&.. ....
.... >-
\.I ::t:
3·"
......
.. ,
0
-.I
\iJ
~

'·0.
'000

.11""
SOOCI
,_., <1>.

~ooo

\'oao

eooo
0".

F~. 17,
 190.

;....
ii-
i:"
~.

;f..
.::::1 Hg.)8

,,..
..
,>'
._.-It ·f'
~
y.,locitv =0'826 R2Pl Sllf2
With 51 =1
Discharge={p+ 1'7854) 0'822501 3 D8!3 (P+l'9248) 0'825501/3 D818

Lacey ,s fL= (0'75 V2) = 0'511

'---R-- ell3 DI/S 0'511 (1/ 3 Dl/3

. . P +3'5708 p+3'8496
c= - ---
p+17854 1'+1'9248
Graphs with N ='018 for side slopes 1: 1 have been plotted correlating the various data
(PlatE-s XvII (b) Vol. III which would be found useful in designing any section,
This section is better than a trapezoidal section, as it is more stable and economical
and also does away with the np-ed of having a toe wall to support the sides.
The slope may he kept as the steepest practicable subject to a maximum velocity of
6 per second and a minimum value of fL = 1'2.
Iron or concrete rings, flush with masnnry may be provided in the sides every
1,000' apart, fixed vertically to enable any person who falls in the channel to come out.
(C) Kennedy, C. V.R,or Lacey's silt factor in lined section.
Kennedv's C,V.R.(X~-,=l 0) "nil Lacey's silt factor (fL:c-~1 0) connote regime channel
conditions in the earthen irr:gatJon channels normally maintamed with self-silted berms and
silted bed, The roaghn"'ss of the perimeter of the earthen channel section is greater than that
of the lined channels. The silt supporting erldies in earthpn chanBel sections are relatively stronger
and more efficaciou<; to support and to roll the !'ilt charge in channel as compared with the
lined channel. In actual practice 20 percent higher value of Ktnnedy's C.V.R. or Lacey's silt
factor i.. allowed in the brick lined sections and 25 per cent more in the cement concrete or
cement plastered sections.
23. ~uper-ele"ation .in Water Road5.
(A) It is a well known fact that when a body moves in a circular path. it gets deflected
from its tangent to the circle by a force acting towards the centre of the circle. In the case of
canal sectLOns un circular curves, the centrifugal effect of the mass of water moviDg round the
curve causes damage and slips to the banks on the concave side loutside of a curve), The curve,
thllS Sdon developes into a hnrse shoe bend impeding regular regime flow and inducing silt deposits
on the inside of a curve. To maintain equilibrium and to counteract this tendency of the water
moving round a curve to attack its conC'l.Ve bank under the action of centrifugal force,
super-elevation or raising of the outer !l;ide is most essential .
The centrifugal force developed on a curve in the case of canals with velocities les3
than 3·Oft. per sec, (which is the li:niting velosity for side e~osion lor the soil conditions in the
Punjab) is too small to have any effect on the concave side of the channel section. The centri-
fugal for ~c results in the eccelemtion of the velocity of water near the concave side of the
 191:

cut\'e and retardation of velocity near the,convf'x Sifh~ of the curVp.. The dischargp.' intensity
er foot width of the channel section dops not remain constant as in a normal section in a
rtraight reach. Both, the increase in the discharge intensity on the concave side and the
~cceleration of the velocity may result in producing velocity near the concave side higher than
the sa.fe scouring velosity of 30ft per sec. The maximum safe-scouring velocity has been taken
to be 3'Sft per sec. in extreme cases in the case of tenacious clayey soil crust. The silt distri-
blltion at bends is describp.d in paragraph lS(f} of this Chapter. To design the non-silting
benns, the bed of the channel should be super-Elevated a'1 shown in Fig. 19 w that the discharge
intensity is the same per foot width of channel. The depth of the channel on the concave side
should be reduced according to the increase in the velocity If the developed maximum velocity
~exceeds the safe-scouring velocity of the soil of the bed and the sides, the channel section should
be protected by pitching or by paving. .
The minimum radii of curves in irrigation channel,; are fixed in practice based on the
experience of earthcrn channe' sections and their valu&s are given below : -
Capacity of channel. Minimum radii to be given.
above 3000 Cusecs ,j JiJ;tJno ocb ;ll.::w\W:'l ~\.dj k 3000'
3000 to 1000 " 3000' ,
1000 to 500 " 2%0' "
500 to 100 " 1000'
tOu to 10 ., So()'
Below )() " (, ' 300' l-
As far as possihle larger radii shoUld be glven. -Hi· l.I·~;U o. .i
(B) Calculations.
, It water be moving on a curve, the presiure varies front one stream tube to another.
The centrifugal force on astream tube is balanced by a ~ifference of pressureq on the two end;,
Let 'r' be the radius of a small strt'am tube of width dr and let 'v' be the velocity'in fep.t pet
second. It can easily be proved that the change of pressure caused by the Genttifugal forc"
shall be:-

dp. =\\1!~ Where w==weight of unit eft. of watet.

dr gr.
v- Velocity ft. per sec. '
-wr(wherew is the angular velocity in citcular units).'

G~ ~ ~=
... S
r:::[:,. 4
r=radius of the stream tube in ft.

~.:ti~f~1,;""ue. diffeeec," in(:;

In a channel of WIdth B, the lUlllts of rare, Rand (R+Bl,
Integrating equation (A) within limits Rand (R+B).

~
~
.J"....
11\
p -p
_l_~_!:=:
W 2a-'1
w 2 (,
~R+B/2 ..... R2
} .' '

"'~'::l '"
~ \..) ill 2
i:; "Q '" F'= 2g
- (B2+2RB)
\oI~
" ~ Q
4{ ...
,-»'-'
~t'
-~4"
\i <:)
...
Fig. 19
 192

In the straight reach. let Q=total discharge in cusec~

q = ~ discharge per foot.

v= bVelocity in feet per sec.

D=Depth in feet.
Let VI aild Dl be velocity and depth at the middle of the curve on the concave side
(outside of tbe curve) and V 2 ond Da be velocity and depth at the middle of the curve on the
convex side (insine of the curve). F represents the total pressure differences on the outside of a
curve relative to the inside and let us assume for the sake of simplicity that the water surfaCe

is depp'ssed to the extent of~ on the inside and elevated on the outside by ~. The depth D

reduces to Dl Fig. 19 near the outside of the curve in the outfall beyond the middle of the curve,
. (V 1-V)2 F
"-~g--=~r

Vl=V+v'~ IC)
similarly depth D expands to D z on the inside of the cur l1e

:. v2 =v-v'-g-F (D)
On the inside of a curve, the average velocity in a channel section, is retarded by the
"ritical velocity due to net pressure difference caused by the centrifugal force on a curve and
on the outside of a curve, vice versa the velocity is accelerated hy the same amount .
.~.
(E)

a.nd D z= __<L_ (F)

va
super--elevation e=(Dz-D1 ) (G)
In rhanne15 with sloping sides, the super-elevation can be calculated neglecting the
slopes. In a large canal, the error thus introduced is extremely small and insignificant.
Ie) Limitations.
It is evident from the above treatment that the design of non-silting and non-scourine
bends with non-changing discharge intensity per foot width to avoid swirls downstream of the
curve is governed by the following canditior1s:~ . .
(i) Due to retardation, the velocity on the inside of the curve should not drop below
the non-silting velocity for the depth there.
(ii) The maximum velocity developed on the outside should not exceed thp. safe.scouriIlg
velo~ity of the mat"rial whir.h the section is made of. .
(iii) The full supply levels are not changing.
'i he first condition cannot at all be fulfillen in earthen channel section becanse, if the
depth he increa.s"d more than the d~pth in the straight reach, actual velocitv must
drop bf>low the n ·n-silting velocify. It is only in lined channel sections that this condition
can be fulfiJ1ed. BI}.CaU5e these can be flumed with the available slope to about two-third of
ben·width on account of the reduced co·efficient of rngosity of the lining m::tterial (brick wor1<,. In
r.emeJlt plaster aT cement concret, even greater reduction of the section is possible.
It is aJso impossible to fulfil the second condition in the sandy soH or silted bed
conditiO'Il ber<'.u~e the velocity developed on the outside of the curve is certainly going to be
more than the scouring velocity of the earthen channel s ction in the straight reach of the canal.
If soil of the bed is day which is not liable to be eroded up to certain maximum velocity, say'
 193

o ft. sec. and roughly assuming 20 percent increase infveTo:;ity on the outside of a curve,the
~(I)dmum veloeity in the normal st.raight reach 0 a channel could at the maximum

j)e
3 '0 X.!
1'2
=2·Eft. per sec, which is Kennedy's Vo for 55ft, de)th It is evident that it is only
' .'
pOssible to design non-silting super elevated bends in earthen channels with depth less than
5ft
5 . The t h'Ird cand't' , f u Ilf'lIed only In
1 Ion IS . the regIme
, channels. In non-regime chann~ 1s t h e
design of non-silting bends on the inside and non-scouring bends on the outside is imprssible in
arthen channel sections.
e
f I
Let us assume that the bed is unerodable clay and will not be depressed on the inside
of the curve to satisfy the condition (i) and ~ii) above, then water

I
o
surface on a Cllrve takes the shape as sketched in Fig 20. Depth Dl
reduces to depth D on the outside and the depth D does not
, 8 ", change on the inside. . 1! ~
(vc.-:YL =F 1
2g
,
).

or VI =v+v'2gF ','n (H) f' ~- '1

"H: .
D=.9.:..
v
!l'
Super-elevation =e::::t(D~Dl)
(D) EXample 1.
Design a super-elevated curve in a lined section for the
following data OLl channel:-
Discharge ,= 1000 Cusecs,
Bed width = 60'0 ft
Depth =6'0 it
Slope ='175' per thousand.
Side slope == ito 1.
v = 2'64 ft, per sec.
N =0'0225.

C.V.R.

Brick pitched sections or rough m~sonry- ~ =0'015 ,

neglecting side slopes for lined section, llsmg Mannmg,s Formula: ....
Effeetive bed width:=: 62~_~'2.!.5=40·U ft.
OU225
Usual side slope for economy is kept I to 1 in lined channels fot side soil compaction,
Averacre velocity=. _1000 ___ =_~0~=__!Qo..~_3'65 ft' per Sec.
o (K+D)D 46x6 276 ..
Let the radius of the curve=2000 ft.
F=~~. 2RB±~~
2g R

3'65 2
2x 2000x4~+46a .... 0·00955.·, ft~ say 0'01 h
= '64T X (2000)2
 194'

vl=v+~=3'65 +V32'2x'Ol

.I :'
. 1000
DIscharge peT' foot run=~' 46 =21'8 cusecs.

21' 8
, D 1 = ----=5·18 ft.
i 421

21'S
DS=3:06=7'03 st.
Super-elvaJion =e=D 2 -D t =7·03-S·18= Ij85"
1'85 .
Slope of super-elevated bed::: -_ =1 in 21·6
4Q
Since v2 =3.{)f) feet per sec is grea.ter than the Kennedy's Vo far 7.01 feet depth, thlt
is, 2'92 feet per se.:.. the nesigned curve section is nonsilting; and the discharge inttDsity shall
remain constant throu~hout the width. The abo"e calculations' give the meximum
super-elevation in the middle of the curVe. It !:hall be zero at the beginning and at the end of
the curve.
The contraction in the upstream of the curve and expansion on the downstream
respective Iv shonld be 1 in 5 generally and 1 in 3 in the extreme ca.'ie. The exp~nsion sho',dd
preferably be designed as per paragrph S(Fl Chapter XU part II.
Example II.
Design a super-elevated non-silting curve m an earthen channel sectian with the
following data.
Discharge = 265 Cusecs.
Bed width = 20 ft.
Depth J: = 5'00 ft.
Sides .h~I'.iI::j; ';~r-'f,"\i' = -! to 1.
./
Slope 1 = i in 4444.
Velocity - 2'35 ft per sec. and Kennedy'5 v,,=Z'3;
Kennedy's C.V.R. 1'0

Effective bed width=B+ D =20+~-=22'5'

2 2

Discharge intensity per foot wiljth= 2~~ = 11 '6 Cusecs~

2",·;, '"
Let radius of the curve=lOOO it.
y2 2RB+B2
F= 2g X. i{:F .

From equation (HI V1,=V+v' 2gF . ~t35+ v'zgx (0 0041 .

=2'35+ O·SOs.c2·S55 'ft. per IJec-ond
D1= }~ =4'1 it. say 4'0 It. But D=S'O It,
Super-elevation =5'::'4 1"",0'9 ft. s-ay }'O"'it
 19~

Thp. s':'ctinn designed is non-silting but the discharge intensity may not be exactly the
same throughout. the width. It i;; not so injurious as the swirls (vertical rollers) which sh.all die
out downstream of the bend of the curve in the straight reach.
~uper-elevation in rivers.
The super-elevation in the water surface as calculated in the foregoing examples of
irrigation channel sections is r::Jther trifling but in the rivers very hkh velor.ities of the. order of
:to tc 25 feet per sec. are pos-;ihle especially in the boulder reach with steep slope of the order of
1
500 The Ganges weir at Bhimgoda, Hardwar, in the United Provinces, India, is situatd on a
curve. Tn the flood" of 19~3 the differencfl of w::Jter level on the left side Flank was recorded to
be 3,0 feet higher than that on the right side Fl::Jnk. In ordflr to counter-act the effect of the cross
flow and the varying disch'l.rge intensity, the crest levd of the weir bay on the left side had to he
rctised by 3'0 ft. the nc;xt bay by 2 0 fed and the third one by 1 () ft. There are 8 bays of about
500 feet each. fhis .. hawed that the velocities, which developed on the outside of the curve
(left Flank), were of the o[d~r of 35 to 40 feet per sec. in this boulder-bedded river at Hardwar.
24. Exa.minati()n QUestions.
1. Describe Kennedy's silt theory. A channel i!> silting badly in the head reach How would you
proceed to determtne its C.luse and what remedies would you sugge<t ? (T.C.E. 19::14)
!<, Quote Bnin's revise,l, KlJtter and i\hain~ formula for flow of water in open earthen channels
Which of these is employed in India and Egypt? Derive the Chef_Y's formula from the first principle. (T.C,E. 192d)
3. Dcoign an earthen channel in ordinary soil to carry 500 cusecs.

Kennedy's_V_ =0'9. Use any method orformula you like that gives reasonable results?
Vg
(T.C.E. 1928)
4. What do you unierstand by (a) regime channels, (b) initial and permanent regime of channele?
Determine the hy lraulic mean depth and water surface slope of a regime channel to carry 8 ') cusecs.

2-=0.8
Vo (T.C.E. 1935).
5. Complete the design of a channel on the enclosed L. Section form which gives N. S. level etc.
[)raw type erO,lj sections of this channel at R D. 5000 and R D, 121)00. Inspection road will be at tlle top
pi tlW bank on the left unto the fall and on the natural surface below that. "Work out the land-width reqUIred
m the left from theccntre line from head up to the fall. (P.I.B.1939) .
6. A minor of 30 cusecs discharge and a bed slope of 1 in 5000 gives constant silt trouble. Give pOSSible
reasons for the silt trouble and state what measures you would take to stop it. (P.I.B.1935)
7. (al What is critical velocity ratio?
(b) Wnat are the velosities for various type of soils ordinarily met with in the Punjab?
(c) VVhat points will you bear in mind in aligning and designing an earthen irrigation channel?
(P.U. 19!2)
8. Describe a brief es,ay on the design and grading of irrigation channels.
(P.U. 1942)
9. (a) 'Vhy d ) the irrigation channels silt most in the heacl reaches?
(b) Wh~t do YOll understand by silt movements in irrigation channels? Explain their causes.
10. Compare Kennej'y and Lacey's silt theories, 'Vhy is Lacey's conception supenor to that
()f Kennedy?
It Hr)\v do the diurnal and seasonal variations of silt charge in the irri ation channels affect
their working?
12. Skdch the best form of a lined channel section. Design a lined channel section for 200
cusecs discharge, slope 0 3 per thousand, N =O'OHl in Manning's formula. _
. \ 3, A channel is to be designed to carry a full suppy of 600 cusecs with a depth of 4·:,feet.
A velOCity of 3'0 fe. per second ;s considered suit.'lble. Side slope 1 to J, sketch a dimensioned channel and
calculate the slope required for this channel. The co-efficient of velocity can be ~elected frOID the following table.
H M D. 2'5 3 3'5 4 4'5
C 67 70 73 76 78
(F.S.C. 1937)
15. How wou ld you design a stable channel in coherent material? Describe briefly the factors
jnvolved,
t An earthpn channel of trapeZOidal section wi th sides,"\- to 1 has a bed width of 20 feet and dppth 3'5 feet
ThE, slope is 000025. What will be the uischarge? Will this channel be stable one? (F.S C. 19H)
18. QllOt" :Yfanning's formula, explaining the various terms and say to what it .is ap~licable., A
draiuage calvert ander a can'tl has a level barrel 200' Ion" faced with brick rnasonary. 4 ft. wide and 6 feet high
The Jl~or is extended horiZontally beyond the end and the wings are splayed at an angle of 30 with the axis'
There IS a vertical breast· wall at each end. Using Manning's or any other formula, derive an expressllu}
 HI6

connecting the discharge with the upstream and downstream water levels. Also obtain an expression for the
pressure urder th" roof of the culvert at a point 50 ft. upstream, of the downstream and, assuming the upstream
water surface to be 10 ft. above the floor level. when the discharge is 140 cusecs. (F.S €. 1940)
]7, A canal lined with concrete has a section consisting of 45 degree side slope£ joined by a quadrant of
a circle of 10 feet radius centred in the water course. Calculate its discharge when running full with a slope of
I in 5000.
18. Derive a formnla for the depth of water in a circular pipe which will give the maximum discharge
for a CQnstant slope.
19. State Bernouli's theorey. How would ycu calculate the velocity head ill a chano,el kpowing the
velocity distribltion?
 PART II

CANAL IRRIGATION
CHAPTER VII
Design an d Maintenance of Banks.
1. The; b~nks of irrigation channels should be strong enough to withstand the water
pressure due to the depth of water in th.e channel. The earthen bank; are porous and,
therefore, water can percolate through. them The normal pore space in the Punjab soil
i,:, a\)Qut 40 ),~ by volu'lle. The bank; as originally comtructed are rather loo,e, hut whfn
consolidated In course of tim~ in a couple of years, are compacteri better than the normal soil
cm-;t. Tile width of the bank is cletermined from the considp.rations of the hydraulic gradient
which is usually kept I in 5 ;IS shown in Fig. 1
. The actual gradient of flow is 1 in 2'5 to 1 in 3 as shown dottfld, and the permissible
hydraulic gradient presupposes a factor of
__,..,.....
-....\
. ,._F •...
safety equal to about 2. The height of the
bank above full supply level in the channel
L. is called free board. It is usually allowed
1 25 feet for the minors, 1'5 ft. for the
distributaries and 2'0 ft. for the branch and
main canals.
2. Typical distributary cross sections.
Fig. 1 Typical distributary and minor cross
/' ';'in;
'1 . sections are sketched in Fig 2 to 4.
0(;
(a) Fi~ 2 giv~s the cross section for a dhtributary when bed is level below natural
ground surface i e., . N. S.

PARTLY IN DIGGING "'-.

1-'-4
Fig, 2
., t ,"·1

oWtlhl':'J a:):)/.~l
:fli £1'1[011;: 2i; ,0
The bank width=Z+D+ -~_ )1 wit. Hi ~)~?,.
tv
where Z=beight of bed above N. S.
D =depth in feet i1bs:'
B=bed width in feet

Free board=F=I+ ~ in feet and the berm width=D

The digg-ing is 1: I up to N. S. and the side slopes of the banks inside and outside
,lie It to1. It i> assumed that the germs shall silt up to the dotted lines to a slope of 1/2 to L
thus le'l.ving a clear berm of length equal to D at full supply level in the distributary. The
inner edge of the berm is kept at wat-r level by cutting the lip. In course of time the other
e1ige i" "loped up 1 in 10 by the earth made available by cutting the lip. The earthwork dug
from tile hed is used on the banks and the additional earthwork required for them shall be
obtain·~rI fro n :)Jrr,)wpits UJually pld.'2ed in the flelds at least 10 feet away from the canal
land limits.
 198

The compensation for the earth got from the cultivators' land is paid according to the
rules in force framed by the Local Government. The aim of the ideal design should be that the
earth obtained from digging in the bed is eqlliil to the earth required ill the hanks. The chan nel
is then said to be designed with th~ bJlancing depth. [f the digging is more than the earth
required on the banks, then "ither th~ bank dimensions are increased or it is dres,ec!
half a foot lower than the bank level as a spoil behind the bank on the side other than t lle
bonndan or service road .
. (b) Fig. 3 shows the cross section of a distributary when it is in filling with bed leve I
above ground level. The height of the free board is kept the same. The bank width is kept
to suit the permissible hydraulic gradient line of 1 in S.
In filling with bed above N. S.

: --:=::~~~~~-~:~~-..:_-=-o:.;_-:::
-8 .• , ..,,--~...p-tf~~(l..i
'&
Fig. 3

. The berm width is kept 2D to 3D according to the requirements depending on the

. embankment height.
The side slopes are l! to 1 both inside and outside. All earthwork for banks is to be
obtained from the borrowpits sufficiently away from the banks.
In the case of large channels the borrowpits can also be put in the bed leaving five feet
berm from the inner toe of the banks on either side and 10 feet wide barriers across the bed
after every 90 feet.
(c) Fig. 4 shows a cross section when the channel is iu digging with full supply level
below the groul1d level i. e. N. S.
~! In Digging F. S. below N. S.

Fig~ 4

Tn this case the earth obtained from digging is more than required. The digging is
1 to 1 on sides and usually a berm 2 to 3 feet is lpft between the inner toe of the spoil bank~
ani! the digging limits. The excess earthwork can be used to raisl! the boundary service road so
that it is not flooded in rains as shown in Fig. 4 and the earthwork still in excess of these
requirements has to be dressed in the form of a spoil bank on the right side as Shown
therein.
3. Main Branch and Canal cross section.
A typical cross section is sketched in Fig. 5. The free board is 2'0 feet. The bank
widths are 17' to 22' on the patrol road side and 12 fpet on other side exclusive of the dowel
which is 1 foot high ami 4 feet wide at the hase with 1 foot top width, sides 1'5 to 1. Usually
thpre are two plantation lines on eithpr side with bounrlary road far cart traffic in between them
as shown therp.in. Side slopes of b:Juks are It to 1 and digging 1 to 1.
; The inside berm dimensions shall change according ta the considerations outlined for
the distributary sections in paragraph (2) above.
4, (A) Earthwork specifications. "~f

(a) Setting out for earihwortlt. ;\' ');li f(';;' .' .' ,

On canals and large channels, preVious to the edlt1ni(!t1c~riietit ~ \Torkl the ~Dtre Hrie 1§
 199

marked by pegs at every chain, curves are pr"perly laid "ut, all half-breadths are carefully set
\)ut, the top and bottoy? erlges of the excavation and the toe of all the embankments and spoil
bank are clearly lockspltted. •
(b) Profiles for eartllwork.
Before construction, a complete profile should be set up on eVEry 500 feet of distance
'<, \ and at every chanFe of section. This profile should be a 10 feet length
;?-~~ -.. .i. of thl' artual completed channel or embankm"ot, the excavation being
*(11 n '
dug to the fJrOPI'T lewl, banks thrown up to the correct height and width-
, .' ,f. -. ... and all slopes dre~s~d to true form. C::Ire shol110 be taken that the ends
;' I of the profile banks are stepped so that they be picked over at the time
± ~ of construction of the banks adjoining them.
e !
~~~ . -+ (c) Stripping soil prior to earthwork construction.
~ /

,.
U) Before begir,ning work. the surface area of gr()und to b" occupied
bv all banks "nd spoil shall have all its jungle and roots grubbed and be

f .-

):
...(- plnughed over. so as to completely eradicate roots and grass, and other
~ jungle, from it. Where the earth below the top layer is not suited for
: making substantial covering to roadways and embankments, the top layer
... of good soil should be set aside for this purp(Jse~
(B) Excavation of channel.
(a) The excavation to be dug in lifts of from 2 feet to 5 feet, as
may be ordered by the officer. in chargp., and-in each chain, each
lift to bp. completen, as far as possible, beiore the one below is commenced.
Care should be taken that the finally completed width of the channel is in
no place exceeded. All gangwavs, roads and stepping should be left
within the channel and not cut into the slope. The 11n<11 dressing of the
slope will then consist of diggi.ng only and no filling or making up will
be necessary.
0'"
~
(b) Deadmen.
Deadmen or such other marks as the engineer-i n-charge may
0 direct sball be left at points indicated by him. These should remain in
...
I
tact till measurempnts are completed, but final payments shl)111d be

..
-.
~.l- deferred till all the marks are removed. Where natural surface is re~ular,
.; deadmen or benches shall be left at equal distant intervals .
.I- Note.-This paragraph applit's to borrowpits as well as to channel
...
t. .. -t'4
.. 0
excavatIon.
(C> Slopes for Ear~hwork.
, l
,~
...
~

-IQ
The stannald:slopes, best suited for the ordinary avcrge soil met
with in the Punjab plains, are I: 1 in excavation and It ~ 1 in
~
v- embankment, but where inferior soil is met with and in the case of flood
~ ..... ~
t.. embankments, special section shall be proposed by the local officers in the
,~ ~ estimate~ for the work 01: reported for orders of sanctioning- authority wLen
I the work is put in hand if the estimate was made out and sanctioned for
Fig. 5 ordinary conditions.
(D) Borrowpits (outside borrowplts.)
. (a) Borrowpits should be dug only wh"re unsvoidable spoil for the £onnatio~ or the
banks being led along the chaDn"l, if possible, in preference to taking it form borrowprts. No
borrowpits shoulrl be within ten feet of the toe of the bank and if its depth exceeds 2 feet, the
distance to top edge of pit should not be less thau 15 feet. Where bortowpits exte~d for a
considerable distance, a bar, separating them. 10 feet wide at the top, shall be IP[t l~ every
chain so as to preven t drainage from running al mg the back of the canal bank. ~orrowplts sha.U
beas shallow as possible, and they will therefore extend over the whole arua avallable for their
 200

formation. No borrowpits sh,>ulrl b~ dug in th'l c1.111.1 b",d b",hN bed levp.I except under special
sanction. All borrowpits should be properly laid out bV the subordinate-in· charge before
digging is begun. •
No borrowpits sh'JUtd at aU b~ m1.de in the berms of channels. Approximately 2t feet
width of tile lio of the berm near tf-te wa.ter ed~e ma.v nOWdver be rlug away not in the form of
borrowpit but it should be continuous without "ta-ti;s" for at lea;;t a chain. The lip c:m be
dug to about half the full supply d~pth but not lower.
(b) Table of quantities of earthworks.
In distributary estimates the table of quantities show, for ear.h 500 or 1000 feet length,
the quantity of earth that NiH be obtained from channel excava.tion, and the quantity that must
come from borrowpits.
(c) Widening of distributary bed t() oMain earth.
In new construction extra earth reqllired for bank;; shall preferably be obtained by
widening the hed of the cannel itself. The bed may be widened to thrpe times the normal
width without causinf1 defects in future workin~ and m'l.intpn::tnce, such widening to be of the
same amount through each length of low gnund. and not to vary frequently.
(E) Emban~ments Laying.
(a) Alll'mbankmenh should be thrown up in layers which should never exceed half a
foot in height, and to the full completed wirlth. Each layer should be commenced from the edge
f~Tthest from the excavation so that all earth is thrown into the sloDe and not tipped over it.
CaTe should be taken that the top of ea~h layer is level or slightly hollow in the cen :re ; any
rounding should be dressed down level before the next layer isb egun.
(b) Breaking of clods
All large clods should be broken up in the borrowpits and care should be taken that
no clods larger than a man'~ list. nor any roots, grass, jungle or other rubbish are brought in
baskets to be buried in the banks.
(c) Allowance for settlement of banks.
The height of the bank should be in all caSfles one-tflnth greater than that shown on
the drawings, to allow for settlement.
(d) Dressing earthwork.
When thrown up to full height, the bank should be dressed to the slopes and dimensions
ordered.
(e) Watering earthwork.
Whenever water Can be lfOt down in a channel, the embankments of which are under.
formation, the area reserved for borrowpits to be moistened to obviate aU risk of clods being
introduced into the banks.
(F) Spoil Banks.
(a) Spoil shoul~ be tipped as directed by the engineer-in charge, and spread evenly
over the whole area avallable In layers of not more than 1 foot thickness. It should be dress0d
to the slopes and form shown in the type ~ection of the work. Bad soil and sand where possible
should be faced and topped with good earth.
(b) Spoil banks for plantations.
Spoil banks intended for plantations to be provided with long and cross dowels,
forming compartments SO' X 501 so that no rain water can flow off.
(G) Ramps (Earthwork).
The plan;: of canal ma"onary works shall inclurle such a s:te pl:tn to a s11itable scale to
show the ramps. Such site plans should sho Ti the ramps to any subsidiary work> in th
 201

neighbourhood, such as bridges on nltch rlistributaries, mill channels etc. The plans of
distribllt~ry m~sonry works should a Iso show how the ramps are to be arranged. Before
construction is started, the ramps shall De marked on the ground bv lockspits.
{a Foundation pit (Excavation).
The foundation pit should be dug truly to the level shown on the drawin~s. W erC
the soil is strong and the pit is above spring level the whole area, may, at the discretion
of the enginter-in-charge be opened out prior to the commenctment of c~nstructi')n. In
such soil the bottom of the pit should be dug truly to the plan of the bottom of the
foundations, and the side slupes should bl': as steep as is compatible with safety for sLps,
under the shock of ramming.
(b) Inspection of pits prior to laying of foundations.
No cOIL<;truction to be commenced without the order of the engineer-in·charge and
such order in case of all works in the cbannel of tbe canal or branch must not be given ulltil
the pit has been inspected by him and passed as true in depth, form and len~th.
(c) Foundation plans.
For large works a foundation plan should be prepared by the Engineer-in-charge
showing reduced levels of each part of the pit.
(d) Watering of pit.
prior to the commencement of work, the bottom of the pit should be thoroughly
watered, but all ramming of dry ballast or laying of bricks or stone over its area ahould be
absolutely prohibited except III the case of slushy foundations,
(e) i. Soft foundations.
In laying concrete in slushy clay foundations it will be found that the lowest laver
of 3 inches depth will on ramming sink into the Sl il and mix up with earth and be rendered
useless as concrete. In such cases, to ensure the full thickness of concrete being good,
excavation should be carried out to such extra depth as may be ordered by the officer-in-charge
and dry ballast should be rammed down to form a satisfactory bed for the concrete to rest
on. In the case of foundations in bad soil or below spring level, only as much of the oit
should be dug down to final level as can be rendered safe during the day by the laying of the
bottom courses of concrete or masonry. Extreme accuracy of form or level cannot be looked
for, but care shall be taken that tne level is nowhere higher than that on the drawings, and
that the edges of the lower course of concrete or masonry as laid are nowherE within outer footi~gs
as designed.
(e) ii Foundation Plt (filling).
The backing of walls and filing of eXcess excavation should be carried up in level
layers of w>t more than t foot in thickness. The top of the filling should be, as a rule, about
2 feet be low the top of the masonry. Before commencing any layer of the fllling, the surfa:e
of the layer below should be cleared of all spawls, bats iLnd other debris, and each layer should
be floodtd to a depth of 2 inches over. night so as to be consolidated and to prevent
absorption of moisture from the new masonry or concrete. Care should be taken t.hat no clods.
roots, grass or other rubbish, is buried in the earth filling which must be laId gently to
prevent dust getting on to the top of the completed course of masonry. Where the filing consists
of shingle, care should be taken that enough sand or finely broken up loam is mixed with it
to ensure complete filling of all interstices. No silt or soil having angle of repose when
wet, less than 2 to 1, should be used for filling.
(H) Turfing earthwork.
Turfing should be carried out over the areas in the manner directed by the Engineer-in •
•charge. It should be done in all cases immediately after the setting in of the monsoon~.
and should be kept well watered until the seeds or roots have sprouted. It should be
protected from cattle.
 202
Watering by spray being preferable to ~atering lJy flow, should be adopted, where
possible.
5. Earthwork specifications for repairs.
(a) Repairs to banks (Holes and ravines).
All holes (gharars) and ravines should be, wherever possibl~, first fully opend out
to the bottom, aU lumps of fallen e1.rth dug away. and the sides dug down in steps of not
more than Ii ftet depth. All iungle. grass, roots, or oth0r rubbl5h, should be thorougly
cleared and the work when ready for ftlling should be inspected and passed by the engineer
or subordinate. deputed for the purp:ne, before fWing is h~guu. Filling shnulrl be done in
accord'1.nce with the spocification<: given for fOilndation pits except that constant wetting-
during the work may take the pll-::e of fil)odin:s elch 13.yer. Flooding the top layer of
the ilay's work should. however. bO) done every ni~ht. ramming with wooden rammers
while the wnrk is in progress, should be done when directed by the engmecr-in-charge.
Ib) Repairing of banks by earth from berms.
Where a silt berm exisls, earth for filling and for repairs generally should be obtained
~s far as possible, by cutting away such berm, care being taken that a layer of at )past six
inchps 1hick of silt next the bank is left untouched, except under special orders of the
Ingineer-in-charge and that cross dowels are left at close int~rval" in the silt berm, to permit·
!,uch borrows silting up quickly. Any ba.nk which i5 to be widened, should be ploughed
or cut into steps.
(c) Raising of driving banks.
Raising of driving bank shall not be done with sandy earth from silt berms.
(cl) Rep)'iring of banks by eareh from sp:>il banks.
If there be no berm, or if the s:>il o~)tainable from the berm be insufficient, earth
shall be obtained from the spoil bank, if such exists, or from outside excavation in getting
earth from the spoil bank b::>rrowplts on top shall be rigidly, prohibited, as in wet weather
they form tanks and lead to damage by breaching. Earth, therefore, i5 to be obtained from
the back of the spoil, or by widening the drainage gaps in the spoil banks, where such exist.
r J Repairing
p of banks by earth from borrowpits.
Where there i" no spoil. earth sh mld be 0 )tained by levelling do ~n any high lump~,
if there be such, and last of all from borrowpits. Where borrowpits are unavoidable, tlley
mu;t be dug as far from the toe of the bank as possible (the minimum distance to be 10)
feet), must not exceed one foot in depth and must be neatly set out parallel to the bapks.
The long slope which forms at the toe of all banks by washing down of soil from the top
and slope shall in no case be dug away.
(0 Silt Clearance
The spoil from silt clear :!.TIces of cha.nnels should be spread out evpnly in the
n')i~hbo'lrjug b::n-rowpit.s. if s11ch exist. [f thev do not, thl! spoil s]lou:d be spread eveuly a.long
the back of the bank thus widening and strengthening it. Care should be taken tha.t the spoil
is n )t hoaped up on the top of the bank, or thrown in lumps on the outside so that it may no~ be
blo Nn in by the winds. '
6. Spe3ificltions for puddle
. . The c.lay should be dug and exposed t? the air for two or thrle days. The clay
contamlDl?' sodIUm c~rbonate (called sodium clay) IS the best for punelle.
If nec-cssary. sand ShOllH be aoded until the mixture is suitable. and thfl whole should be
wetted and" e1 \I 0 ked up in a rug mill or Ly nen's fect into .a. smooth plastic IT ass. The working
wit;} a pug mill or kne'lding with fe~t tends to COvflrt !lodium carbonate into a colloidal
s'Jlution and the cor ods tenrl to choke the fine pore spaces to make the puddle impervious ..
 ~;
203

7. Construction oi hi~h embankment.

Special precautions have to be taken in the construction of high embankments, such
as approaches to aquf'ducts in the case of large canals for example Solani aqueduct, of the
Ganges canal in the LJnitpd Provinces.
[::IJ Speci lcations of earthwork in genf'ral as descrihed in paragraph 5 above apply
to the earthwork in high embankments concerning the layout, removal of jungle, the size
of the clods, location of borrow pits etc.
[hJ The u~ual section adopted in the good old days was as shown in Fig. 6 with a
p1lddle core. The earth was laid in 6 inches layers a1.d c'JrHJlidat-:: i by w)odqn rammers ::Ind
the puddle core abo'clt 5 feet thickness rose as the bank progresspd. The main reliance was pl<lcf>d
on the efficiency of the puddle core as a water tight substance. . This efficiency of the puddle
d"pends on the retenti()n of the moisture in it because the densification of the purldle is due
to the chocking of the pore space by the sodium clay being pugged into colloidal state
to fill the pores. The dried puddle 15, therefore, even worse than ordinary sdl banks becau~e
it cracks. The dry soil arouud the puddle core extracts the moisture in course of time.
reducing therehyits efficiency. It is therefore. th'l.t high embankrnents in the past have always
he en a source of great trouble. .
rc] The modren practice is to stabilit' the whole. mass of the soil in the bank.
Various methods of soil stabilisation have been described in Chapter III part VI such as :-
'lij 1: 0 manipUlate the component paris in such a way 't'nat the mixture will. proGuce
compact soil. .
[iiJ Use of vario,ls admixtures such as electrolytes, chemicals, binders and
adhesives.
[iii] By the p.lectrochemi-al process such as the application of heat.
[ivJ Campaction and df'nsincation at the optimnm moisture content.
Only the last method is of practical utdity when earthwork is to be carried out
on a larg-e 8cale. In this method the soil is made impervious not by chocking the pores but
bV compaction. Compaction results in reduction of the film of water around the soil particlf>s.
When the soil is compacted at the optimum moistllre content the dry density of the soil is
maximum. Therp. are vari()us methods of compaction at optimum moisture content. The one
described below can be used in the field in a la.rge scale.
[iJ Measure the sand content in the soil by drying a soil specimen, then pulverising
it and passing it throngh ~;ieve No. 270, calculate the percentage by weight of sand left on the
sieve.
[iiJ Calculate the optimum moisture content· in each case according to K.B.S.I..
Mahbub formula. Punjab Engineering Congress, Lahore Paper No. 257. ,.
. W=25--14S
where \V =percentage optimum moisture content
S=percf'ntage of sand in the soil.
[iii] Determine the hygroscoPi: moisture content of the soil_
The difference of the optimum moisture content and the hygroscopic moisture content
will give the amountof water to be adrleu while compacting.
[ivJ Lay th., soil in 6 inches layers and then add water as worked in [iii] aboVf~.
With a little experience, thp. variations due to the prevailing temperatures can be easily
allowed.
[v] The soil should then be rolled by means of 1'3 ton dentated roller. The roller
is .m~de of concrete having staggered teeth projectinl? three inches ::In~ d~iven by b~lllocks. The:
sOlI IS supposed to be consolidated When the irnpres"lOn by the proJectmg teeth IS not more
than iN deep or when the surface has been rolleci by 16 to 20 times.
[viJ After consolidation of eaeh layer, the density of the soil is tested. It should be
about 1~8 .
. [vii] Each layer should be cover ted with' a couple of inches of sand. after consolirlation
whi:h should be removed when the next one is laid. This is just to stop soil evaporation during
the interval.
 206

awl parallel to them, with cross banks at intervals of 500 feet to 1,000 feet apart. The series of
compartments, which are thus formed, have inlets from the canal at the upstream ends and out-
lets to the canal at the downstrp::Irn ends of the compartments. A portion {)f the canal tiL5charge
is passer! through each compartment by these inlets a:ld uutlets, and it is thus gradually filled
up with silt.
(b) The 'Long Reach' system.
Under this system (Fig 8) external parallel banks are constructed, ::IS in the first
case, with similar cross banks, but these are at 4000 to 5uOO feet intervals. Head inlets and
tail outlets are constructed,. and not only a portion of the canal supply but the whole of it is

Fig. 8

,.
diverted into one silting reach at a time. The canal channel contiguous to th~· siltini~r~;f~::;tn
operation, is closed at its head and tail to prevent the deposit of silt in the canal. WHenr, thiS
system is employed, spurs also are constucted to encourage deposits. :' ';"1 ~" ~': ."::;';

(c) Thc system of 'Internal Silting.'

Under this system (Fig 9) the canal banks are set back at a, little distance' frnm the
normal section of the canal channel. Inducements are laid down to encour::lge the deposit ,..,f
silt internally on the herrns This system is the simplest, the best, and the most economical
of the three, but it can only be applied to a new canal which is so constructed.
The papers give full details of the operations, with illustrations. It was found in the.
operations of the Ra.kh Branch of the Lower Chenab Canal in the Punjab that in 22 months
16,000.000 cubic fpet of silt was deposited in a length of about seven miles. It W::IS found best
to commence with the "long reach" system and complete the nperation with the 'In and out.'
system. The proportion of silt deposited to Vi'ater passlng was ahout 1/2200. Tbe silted portions
were eventually ploughed and sown wit h shisham very successfully.

, Fig. ;10
_··';-'·1" . '_

H
"'H"..'HSAL SII..TING $Y$TtM

,.If Q. k a-A
To obtain the most speedy and satisfactory results with a long series of reache~, /a:;:-h
reach should be closed and made reasonably secure by olJtaining a sufficiently heavy deposit.
Other reaches should then be taken in hand, and the completion of the warping in the reaches
first opened, which being a.51?w process. should not be undertaken until all the r~aches have
received their first heavy slltmg. The mner bank should, therefore, be made suffiCiently strong
to withstand the he'ld of pressure when the canal is full and there is no water in the reach. One
other point to be noted is that silting reaches should not be made too narrow. A. G. Reid
says that the Width of the tank will of course he the who Ie of the Nidth between the canal bank
and the boundary road, but it may oftpn be desirable to increase the usual width between the
outer boundaries when passing through low ground in which silting reaches are contemplated.
A.M.R. Montagu, Superintending Engineer Western J .lmna Canal ~now Chief Engineer]
evolved a system of silting tanks on the Western Jumna Canal with a view to rtffiOVe 'he sik-
 207

trouble of the canal. Clear wa.ter. after ~1ropping silt in the silting tanks, picked up silt from
the reaches of the canals and dIstnbutanes bwer down and thus caused scouring of channels.
This method gives a great relief to the silt troubl~ in a canal where proper silt Extractors and
Ejectors cannot be constructed on account of the supplies being not available for escapage
II. Accidents to canal banks.
Breach~s in the ~an.al banks are .f~ir.ly common. They may be due to raJ Intentional
cuts by the cultIvators to Irngate the ad]OImng lands or [bJ the mual causes for normal
breaches .
. . The ::Iccidents of the former case, can be avoided by patrolling banks at night, by levv
of addItIonal charges for the water thus wasted, and by strict judicial action against th~ persons
at fault.
The breaches can occur due to the followIn!! causes:-·
(a.) wEak Bank. Insufficient bank width should be nnde good by period;ca1 strengthening
of banks which is normally required after every fifth year.
(b) Overflows. The overflow can also be avoided bv periodical raising of the banks
to the designed sect ions.
(c) Leakages. The leakae-es occur through insect or r::lt holes. Tl~ev take a considerable
time usually more than 24 hours before they develop into a breach. If patrolling by beldars
is efficient, no breach sbould occur on this account.
(d) piping due to excess supply. Even though
the bank be strong. the exeess supply
over-topping the berms results in failure at
t he banks by piping as explained below in
Fig. 10. -
Let CD be the surface of canal wa.ter,
where it normally run~ full supply. CG
Fig. 10 rt-presents the saturation line in the bank.
The bank as sketched above is extra safe. Lpt there be an excess in the canal so that watpr
ltvel rises to EF. What actually happens is that the line of saturation is changed to EH. A
breach should not occnr even if the point H be higher than the toe of the bank. At the most
water should seep out from the canal. The velocity of the seepage cannot be high enough to
dislocate the soil particles of the bank. A failure never occurs by percolation through th ..
banks.
Under normal conditions the soilof the Dank al;tQve CG is dry and has probably been
never wetted. The Tise of the saturation line froIll C~ to EH wds for the· first time the soil
of the trapezium CGHE of the bank section, Dry soil on tirst wetting contracts and the bank
stands at places by arching action. Thus an open pipe is forme,i between EH and E'H' ai
shown in the sketch ll.bove. Water flows out of it as a leakage "'S if an open connection exists
from water in the canal to that outside. The water cornming out starts washing out the soiJ
and eventually develops into a breach. If in the beginning the mouth E of the pipe EH coulq
be located and closed. the breach would never occur.
12. Closing breaches.
(i) Closing a breach in a small distributary or minor.
Water of the breach spreads on the adjoining land~ and usual~y there is no place to
take earth for closing the breach. Th~ earth has to be obUllned by cuttmg the outer slope of
the existing bank. Enough earth should he collected on both sides· of. the breach on the . existing
bank. The earth baskets should never be thrown in the water. All Jungle should be removed
from the breach site by men in running water. The process starts from both ends by slipping
the earth from the heap and protecting the channels sidE. by grass~ clods usually ava~lable .1rom
the berms. No grassy clod should be allowed to be washed down mto t~e breach SIte. WIt~ a
rush of earthwork at the end, the breach can be closed straIght away progressmg
from the bank. f
 208
(Ii) ClosIng a brelch in a mllor distributary or sma.ll Branch canal.
• fn this C:lse it is necessaTy to reduce the flow through the breach, otherwise a ht of
earth will be washed away before the breach is closed. This is usually done by, driving a
double line of stakes as shown in Fig II and then putting planks (,r mattrt'sses against them
if available :l.nd if not then filling jungle in between the stake.. pressing it down with
bags filled with sand and by men walking over them. No earthwork should progrr·si ,before
C4NAl. the flow through the breach has been arrested to
---- ----.----.--- some extent in this way.
Mt'anwhile earth is piled up en bothe sides.
The closing is starte-d from both sides by slipping
earth from the heap in form of a ring bund as shown
in Fig. 11. All jungle from the ring bund site s30uld
be removp.d before earth work progresses. No earth
Sl.Ol'e basket should be thrl)wn in water it must always be
slippp.d from a heap. The last gap of about 10 feet shonld
be closed with a ru~h when enough earth h:ls teen
collected on both s·des. Straight closure in l .. rge
Fig. t 1 channels is not possi bIe,
(iii) Closing a breach in a canal
The closing of a breach in a canal follows the same method of the ring bund but the
jungle or planks do not servp. the purpose of arrpsting the flow through thc breach. The
double line of stakes should he driven as before if depth permits and a double line of gunny
bags filled with sand is put in. The inter-spac~s are plue-ged with berm earth, A temporary
i)ank of gunny bags is raisecl in the position of stakes and bushing as shown in Fig. 11. Th6
closing of the breach is then done by constructing a ring bund behind.
t3. Examination Questions.
1. How would you consolidate a new canal bank in filling if it has clods and is full of
ilaUar 1 ( P. I. B. 1941 )
2. How would you in~rease the amount of silt entry into a silting tank inlet? (P. I. B. 1941 )
3. Draw to scale the cross section you would adopt for building a two mile long embankm~nt
(each from a new distributary of the fo.lowing dimen~ians; the soil i;, light and contains a small amount
,f kallar.
Bed width 25 feet F. S. depth 3 feet
Slope 1 in 4000. F. S. L. 5 feet above N. S. L.
. Stace what precautions you would take on opening the channel and running it to ensure that the
!mbankment reach would become safe and water-tight as quickly as possible. ( P. I. B. 1940 J
4 Descrihe What method you would adopt for each of the following operations:-
(i) To form berms on a distributary.
(ii) To check side ero,ion on a large channel. ( P. I. B. 1939 )
5. Give Specifications fOt" :-
(i) Repairing rat holes and rav:nes in a canal banlt.
(ii) Deadmen for borrowp ts. . ( P. 1. B. 1936 )
6. Draw cross sections of a channel of dimensions given below for the following cases:-
Bed width 20 feet, F. S. Depth 4 feet, F. S. Discharge 160 cs.
(a) When natural surface is 3 feet lower than the bed level.
(b) When natural surface is 2 feet higher than the bed level but soil is very bad. (P I. B. 1935) /
7. How much labour should be arranged in each day for the construction of a fall on a distributary
In a ten days closurf'. having the following quantities of work to be done?
Earthwork in foundations. 200 eft, Ceme:':'t concrete 400 cft.
Brick masonry 1500 eft. Dry brick pitching 300 cft. . (P. I. B. 1935 )
S. Give sp"cificatiolls fvT Glzarabandi and describe how you would carry out and check this
class of work to prevent fraud. (P. I. B. 1936)
9. You are required to make embankments of a channel leading to an aqueduct. Lay down
,pecifica.tions for earthwork, (P. U. 1942 )
 PART II
CANAL IRRIGATION
Chapter VII I
Lining of Channels
t. The lining of channels has progressp.d only recently in India. It was brought up for the
ntst time in 1917 by T A. Curry in paper No. 32 Punjab Engineering Congress. Messrs. R. S.
Dnncan and Som Nath Kapur contributed their paDer Nos. 221 and 25::l in the Punjab
Engineering Congress on the "Haveli Lining." F. F. Haigh Superintending Engineer, inchargr
nf Haveli lining wrote various notes on the subject. A comprehensive paper No.
260, Punjab Engineering Congress, Lahore 1913 by K.B.S.r. Mahbub on "Lining of channels.'
'pu published.forAabri'cf
wasblicati)Ds summary
detailed study.is givcn here and the student should rcfer to the originai
2. Advantages of Lining.
The main advantages of lining ate:-
(a) To save water for extension of irrigation or increasing the water
already served. , supply in area~
(b) To prevent water from reaching the water-table ;lnd raising it. thus avoiding
water-logging.
(c) fo improve command owing to flatter possible slopes.
It may thus be worth while in some cases to line only the head reach of a distributary.
So as tothe
where command
commandthe adjoining high areas and leavd the channel unlined in the lower reaches
is ample.
(d) The
and alteration of stability
outlets of section, which in tbe case of distributaries should reduce remodeIlin~
(e) Rcduction in maintenance costs.
3, Suitability of lining,

The principal factors which have to be considered in deciding the suitability ot

design are:- an~
(a~ Permeability, i,e.) reduction in absorption losses. Generally speaking, bitumen linin~
or a suitablp. form of Concrete or bri;k l'ning would reduce these losses appreciably. Th~
presence of any cracks, however) would also have a material effect on thfsp. losses.
(b\ Co efficient of Rugosity. which will determine the carrying capacity of the channel.
It is preferable to use M"anning's formula in designing the section as it is simple anf]
fits closely
deduced fromto several
Kutter's,observations.
for which values of N have been determined by various experiments and
The values of N generally accepted are :~
Svrface Perfect
(p,ment mortar surfaces 'OIl Good Fair
'012
Bad
Brickwork in cement mortar ,'012 '013 '015
Concrete lined channel '012 '013 '01$
'014
'017
Refer to article 23 chapter VI of this part for finding '016 'U18
Plate XIV of Vol. III. out lined channel section al1(j
(c) Durability.

This may be described as the resistance to the following dIsintegrating force!!: .....
(i) Weathering;_ This is caused by the disruptive action of alternative freeting aDO
hawing and by expansion and contraction rpsulting from temperature variations and alternatt
wetting and drying. The resistance of any surface to Weathering is thus a fUEction of itt
 210

water-tightness and volume-change characteristics, It may be remarked in this connection

tha.t the co efficient of exoansion and contractin of brickwork is about t that of cement
concrete or cement mortar.
(ii) Chemical Attaek: - Thp. mostimDortant destructive agencies under this he:td are
sulphate of sodium ann ma!5nesiurn, commonly encountered in so called alkali soils and water.
Corrosion of concrete by alkali W:lteor can be m:J.terially reduced by the use of sulphate resisting
cement. Surface coatin~ treatments are also gen<>rally effective in prolonging the life of concrete
or bricks though this mav not be a permanent remedy,
Alkali solutions inc:rease in strength in dry seasons when dilution is at the minimum.
The United St:ltes Bureau of Stanrlareds specifies that concentrations greater than 0'1 of 1 percent
enclanger concrete.
(iii) Wearing: -In the case of concrete brick-lining this wearing action is negligible
for all velocitits under ordinary conditions.
(d) Initial eost and subsequent maintenance.
This would vary with 10cal1tv an,:! the availability of various materials. Generally
brick ma'>onry lining wi)uld be cheap~r than stone eoncrete lining,
(e) Structural Stability,
The following factors may be considered under this head: -
(i) Reinforcement: -In America. it is a common practice to ninforce the lining concrete,
Reinforcemeo t i<; dC3igned to reduce th"l siz~ of c::>ntraction cracks and to assist in preventing.
failure of the lining due to settling of the sub-grade ar to back pressure from a saturated
subgrade.
On the other hann, this may delay relief beiag obtained by local failure in small patches,
in thp. case of heaning up of water pressure. thus causing extensive oamage. flis was found to
be the ca"e on the Haveli Main Line,
(ii) Thickness of Lining: _-The side s of a channel to be lined should preferably be
[,£t nearly of th'~ S'l.me slope as the angle of repose of the natural surface, whicn is of the
following order: -
Material Angle of repose Slope
Firm clay well jrain<>d 45° 1:1
Clay loam, awrage sandy loam 36°-33° Ii-It: 1
Sandy or gravel soil 33°-26° It- 2 : 1
Where the slopes are left steep~r, the side" have to b~ designed as sloping retainiilg
wall~, Etchevery ([rrigation Practice anrl Engine~ring Vol: 11 by B. -\. Etcheveryl has work.... d
out the following thickness of concret'l lining corresponding t) varioils depths of canals, based on
Coulomb's lormula:-
. Maximum depth of canal in fect.

Side Angle ~o surcharge and thickness Maximum surcharge and

slope of of repose of lining. thickness of lining of
canal. of earth. 1" 2" 3" 6" 1" 2" 3" 6"
-'--~~--.~.----- . ' - .---.,~-----------~---

1/2: 1 1 : 1 53 106 160 32'0 ]'6 33 5'0 10'0

1}2 : 1 1/2 : 1 16 3'2 4-8 9'6 0'6 1-:!. 18 36
1/2: 1 2 : 1 10 2'0 3'3 60 04 18 12 2'4
1/2 : 1 3 :1 0'5 1'1 1'6 3'2 0'3 06 09 l'S
1 :1 15 : 1 158 31'6 473 9* 8 4'8 97 U'S 29'U
I : 1 2 : I 3'8 77 U'S 23'0 19 3'8 5'7 11'4
: I 3 : 1 }'9 3'8 57 11'4 08 1'7 2'5 5'0
1'5 : I 2 :1 370 740 III °0 2220 11'3 226 3tO 68'0
1'5 : 1 3 :1 6'2 124 8'6 17'2 6'S 5'1 7'6 15'2
 Thec:e results may be considered as useful for practical application only within:
reasonable limits.
The thickness required N()uld be much m'lre. if provision has to be made for any
h ydrostatc press 11 re across the IininQ'.
In the worst conditions, the lining is subject to pressure due to saturated soil and the
,. differential water head across it. Assuming angle
(Ifrepose of earth to be 30° and angle of slope of
1 ning as 45°, the differential head that a lining of
6" and !::In thickness would stand, as obtained from
\ calculations, is indicated in the graph in Fig. 1.
I
It would be seen that with greater depths the
1\-+ .'
.~f\W-
II lining would only stand a slight difference of
, hycirostatjc pressure on the two sides. This clearly
indicates the importance of proper drainage so as to
I I \ keep the difference of hydrosta tic pressure across the
a I\IN Ening as 10'IV as possible.
i " -1 I IT,' ~"'CIC (iii) Strength of lining.
!----- f-- ~ r-- J 4,",,,
'"~
..7.
1 eli TH'CM ,'If' • Any reasonable strength should be good enough
~ ~~
-- ~ f---- 1---- if the materials used are such a~ do not deteriorate.
in course of time. A piece of Bik~ner lining shows on
•
..
~. .......-

I
.,ffllE'HIAi
rl 4 S
MIA6 A(HSI "NIN.
, 1 • ,
test a compressive strength of only 690 Ibs/sq. inch.
The compression strength of l;~ cement. sand
mortar after three months is conc;idered to be over 3,000
bsfsq. inch. Both cement concrete and cement
Fig 2 brick masonry would thus give more strength than
the Bikanet lining.
(i v) Eat th Backing.
In the U.S.A. Bureau 01 Reclamation speci.ftcations. it 15 laid down that lining should
only ce placed on an undist1lrbed material or thoroughly compacted back fill. It is also
neces~ary to provide a.dequa.te facilities for drain"lge of rain water tJ ke~p the backing in proper
form, The American practice is to lay open joint pipec; in gravel-filled trenches to serve as.
longitudinal drains. I'eeder lines are brought from the sid~ slopes where n~cessuy and outlets
are provided as required to discharge the acctimml1lated water.
<C. Various type; of LiRing.
The type generally u<ed for lining so far are:~
(a) Cement or lime concrete .. ~" to 6"thick.
(0) Cement mortar-!" to If thick. ,
(c) Stone \1asonry set in cement or Hme mortar -6~ to 12" t~k.
(d Road oils.
(e) Sodium carb"nate plastt>c.
(tl Clay purldle-3" tn 6" thick.
(g) Brick lining combined with cement, mortar-3" to 6~ thick.
(h) Preca,t cement concret{' block's.
(i) Tar or bitumen imp:egnated cloth c.)vered by masonry
(a) Clnerete Lining.
. This is a very useful type and is generally arlopted. Tt is dtlcable if lctid properly, and
reduces the ahsorption 10sses by nearly 95 %. The co efficient of rng()sity is very low, and in:
view of the high wlocities possible, tho section is considerably reduced. The All Am~rican
Canal is a crmcrete lined channel '
On the Gang canal ~Concrete lining of Gang Canal by C. E. Jefferie5, Punjab Engineering
Congress, Paper No. 102) whIch consisted of 6" of 1: 1 : 6 mixture of lime, grit and k.snkar
balla.st, a value of Kutter's N =0'013 was adopted. The value actually obta.ined wa'l, ho~ver,
 212

reported to be 0-0145 in 1935 and O'OJ f-3 in 1939. As all the ingredients were obtained from
local kattkar the cost was quite low, viz., Rs. 22/8/- per cent sq: feet.
The joints were made Vor Y shap;d and hter filled with bitumen i'n the ca~e of the
Gang Canal. A Jap type is not desinble as it breaks at the shoulder under tension. A butt
Joint 1" .wide und(~rlaid by a 1:4:8 concrete sleep:r 1'0' '.vide. and of the same thickn~ss liS the
lining is conc:iderEd to be better thaa a V or Y shapad joint, as this would o'wiate the necessitv
of filling with bitumen afterwards. and would simply mean leaving plain constuction joints at
specific places. .'
To avoid cracks, it is desireahle in sn..:.h .cases to lim't the siz~ of blocks to saY 20 I X 20'
or 15'x IS', Alternate blocks should be laid and preferah 1 v an int(>rv"ll of seven davs allowed· so
that the setting, contraction may take place. ,Alternatively, strips 51. wide may be left all round
the blocks, and these strips rpay be con~reted' -after a week
To avoid the subgradef ahsc>rbing moisture from the hottom portion of the concn'te and
thus making' it spongy and pernicable, it shoulrl be thoroughly moistenp.d. According to the
Bliearu of Re~lamati'm. ·U.S A. th~ proper moisture penetration is 12" in sanny soil and 6" in
olher slils; except when this much moisture causes the subgrade to become muddy.
Other alternatives are:--
(i) Use of oil paper. This has been found to give very good results. The cost works
to about Rs. 2 per c(>nt. square feet. '
, (ii) . Spreading of any crude oil on the subgrane so that the whole of the surface is fully
covered. ThTs'may be usefully adopted whf're oil paper i, not available. The cost of using
linseed oil at Rs, 3/- per gallon works out to Rs. 1/10/- per cpnt square feet.
, (iii) 1:6 cement pla<;ter on the subgt:lde. The plastering need not be very accurate,
and half tlre~labour tate~r'normallv paid for plastering should be good enough for this type of
work. The cost ihdurling materials wou'd work out to ahout Rs.3i8/~'per cent ;:quarc feet.
(iv) Spreading t" of 1:4 cement mortars and slurry a couple of hours before plaCing the
concrete. The slurry can be poyred direct fr'lm the cans. This has teen found to be quite
successful in' actual practice, and the cost wdrks'out to about Rs. 1/8/- per cent square feet.
The consistency of concrete used in lil'ting is a verY important factor. The United States
Bureau specifications for: canal lining lay down that this consistency is critical. . "The concrete
must be fluid enough {<i'compact well, especially under the reinforcement, and yet stiff enough
to stay in place or the slopes. The slump must be carefully regulated. If it is too low, honeycomb
is difficult to avoid,' and if it is high, a bulging or wavy concrete surface may be produced. A
properly designed mix, using well ~rad~d aggregate containing enough of the fine sizes to allow
atlequate finishing hut not so much as to cause crazing of the finished surface is necessary for
satisfactory canal lining. .
(b) Cement Mortar.
This type is naturally not very durable u~less suitably' prot~cted, and as
such can only
be used in conjuction with some other protective mat~rial. According to Etchevery, 1" thick
cement mortar stops 75 percent seepage. Actually. however, it is mue». more effective .
.,,' . "J.::.. ,
(c) stone Masonry.
This has a limited application mainly on account. of tpc cost imd thus should/ be used
only where stone is locally available. Th;e co-efficient of rugosity is, comparatively higher in this'
case, as stone cannot be very finely dressed. ". ",
(d) Road OilS.
For using thpse the soil is rolled or burrowd so as to seCure a peti~ttation of 1/ to 3". It is estima-
ted, that Ifl6". thjck.l~yer 9f erude ~~l (at about l/B,lfoi agaUop per hundred square feet) stop~ 15
p~ cent.se~page,.1Jl.)S IS thus not ,as 1mpermeabl1! ~s the other type~ The ~o-efficien~ of tugosity is
falr,Iy ~lg1i: and Its effect is no~ durable. t~lS does nO,t, stop bliIrowlBi by a~lmals and wil1
prevent weed growth for only a few spasofis, '
.'" (e) '~odlum:'eaTboDat( llnbig.
"Th~, h,.!~~~~ u_se,~~P1'1 :V\/atereotl~ses ai?-9. small chat1llds and consi~tS of i<piastet inadt!
 213

up of clay, bhooso and s Idium carbonate in the proportion of 100 eft. of cl;!y, 6 mds of bhoosa and
9 seer::. of sodium carbonate. The cost to the zamlndar who provides bis own labour and
bhoosa according to the figure given by G. R. Sawhney, in his paper (lin,ng of water-courses by
G. R. Sawhney, p, E. C Paper No 24l;) on the subjPc t is Rs. 25/- to Rs. 40/- per mile.
The author had an occasion to see some of this type of lining on the Lower Bari Doab
Canal. The plaster was found to he damagfd in most cases ano could not he considered to have
a mefullife for more than 2 to 3 'ieasons. The saving of water effected bv its arloption could be
attributed more to the proper alignment and grading of water courses than to the plaster.
The method was also tried in the United Provinces bnt was not found to be a s"ccess.
Thev had founr1 that lining with u~ar (local salty SOLI containing lOq!~ clay and6% sodium
carbonate) was more effective. A bitumen aud mastic plaster would, however, be much more
durable than even Usar.
If) Clay Puddle.
This ha~ been in use fOT a long time as the only material required is suitable clay.
Th'! selected SOli IS allowed to weather and then pugged throughly after saturating it with
w'"lter, by men marchwg up and down. This pugged clay is then put in position and covered
with about l' layer of Silt. This was considered to reduce seepage by about 80 per cent.
Recent researches have d~arly shown that there is an optimum moisture content for
each soil at which the dry bu.lk densitY': obtained 1S the maximum: We cannot, therefore hope
to get the denSity by saturatmg the SOlI, as has be~n the practice m the past. as by cont rolling
the moistu[p contf'nt and keeping it. as near the optimum as practical,le. This moisture content
can be easily determined in the laboratory by us:ng a compaction apparatus An approximation
can only be obtained by using the formula (paper 257 P.E,C. by K.B.S'I Mahbub).
W=25-'14S
where W = "/oobtimum moisture content
and s= % of sand in the soil i.e. particles owr 0·05 m.m. diameter
The percentage of sanri can be determined either by a Chain Hydrometer after taktng
into account th"1 time taken by various particle5 to settle after dispersal, in accordance with
Stoke'~ law, or by pas5ing the disptlfsed soil through a 270 mesh sieve, with water froni a
wash bottle playmg over it. .
Rolling can then be done in 6" layers at the optimum moisture content with toothed
roller, as de,cribed in netail in Punjab Engineering Congress Paper No. 257, as these ensure
a higher intensity of pressure anrl uniform kneading from botton upwards.
It may .be noted that the deslrable placements ra?ge of moisture content i'l very
much reduced WIth hlgher clay contents and a more ngld control over moisture is thus
necessary in such ca!"es· .
(g) Brick Lin :ng.
This was used on a large scale for the first time in America in 1933 on the Southf}rn
Texa"l Irrigation Canals. It was also adopted with smtable modifications for the Haveli Main
Line in 1937.
The Haveli Lining c"nsisted of two layers of tiles 12" X Srx 2k" laid in cement witli t"
of 1 : 3 cement plaster sandwiched in between. [he botto.n layer of tiles was bedded on a 1"
layer of 1 : 6 cement plaster. 1he masonry was reinforced with 1" M. S. bars laid in the
1 : 3 plaster, forn ing a grid 12:1:' X 12t' in the sides and 24!' X 24i' in the bed.
This lining failed in several reaches within a year of its construction due to water
getting behind the lining from external s~)Urces and. fro.m wave ~c~ion oVt'rtoppi.ng the lining.
In one case the water cau"ed a ~reach ot the bank behmd the .hnmg. ~f.ter whIch t.he lining
sLibsiced. In the other cases WhICh were many, the pressure behmd the hmng caused It to fail
and fall into the canaL No lining of practicable thickness could have stood under 6ither of
'hese circumstances. The ddects urought about by the failure were :- .
(a) Inacequate compilction of the back fill.
(b) Lack of proper provision for the drainage of the banks •
. (c) Insufficient free board.
Nothing was thus found wrong with the type and no similar trouble could be
anticipated ift~e above defects were effectively r6ffivved in future dt signs.
 214

5. Construction of Brick Lining.

(a) The following are the main precautions to be observed :-
(i) The earth to be used in the manufacture of bricks should not have a salt content
above 0 3 pew'nt and the quantity, of calcium carbonate should not exceed 2 percent. The
clay content should range from 10 to 20 percent.
(ii) Great care should be ~xercised in the moulding and burning of bricks No pilla
bricks should be aHowt'd to be used under any condition.
lili) The bricks ::;hould be thoroughly soaked before use. This is important as bricks
which absorb water quickly lose it as quickly when taken out of water. Each mason should
thus be provided with a kerosene tin of water in which his immediate requirements of bricks can
be kept.
(iv) Sand should have a fineness moduless of preferably not less than 1 2, and
should be free hom (rganic impurities. The percentage volume of silt in it shoulri not excefd
6 per cent.
\v) The consistency of mortar should be properly regulated by having slump tests.
the masons should not be allowed to mix water in the mortar pans at site, under any
conditions. '
(vi) The olaster should be allowed to sft properly for about two days before laying
the masonry pn top.
. (vli) 1 he 5ubgrarle should be properly moistened or oiled to avoid absorption of water
from the bottom layer of masonry or plaster.
Thl} bats obtained as a bye pre(Iuct while burning bricks can bl'! broken into ballast
passing thro 19h f' ring and used in concre1e lining, which can, for ease in execut.ion, ce restricted
unly to p~rti JDs'n ted .. . .
Ihis concrete can be lald III compartments, as ~entlOned already and levelled and
compacted by means of heavy screed~ or tampers fitted with handles weighing not less than 7 lOs.
per linear foot. These should not b; les~ than 3" wide and should have !;" thick flat iron fixed
on the underside. The surface can then be floated with a flat board IS' long and about gw
wide, after which it can be rolled. if necessary, by light rollers, weighing 10 to 12 lOs per ft.
run.
(b) Brick lining in general has the following acvantagcs over concrete lining laid in
situ :-
(i) Comparatively low cost, as bricks can be manufactured at site, thus avoiding long
carriagps.
(ii) Highly skilled or specialised lar-our is not required, as the work is of a simple
straightforward nature,
(iii) No elaborate or expensive equipment is nectssary as would be needed for concrete
linings.
(iv) Due to the large number of joints, cross cracks due to contraction are greatly
reduced, and buckling caused by expansion is pli_ninated.
tv) No expansion joints are neCessary.
(VI) Brick lining can be more easily adopted to construction in circular sections.
(vii) There is no risk of any undertected thin areas due to poor workmanship, as the
thickness is controlled in this case by the thicknes~ of the bricks. __
(viii) Repairs when necessary can be canied out easily and expeditously. "
6. Compactions of the back fill and banks.
(a) The importance of having a stable back-fill which would not settle behind the
lining cannot be over-emphasized. This should be obtaiop.d by compaction of the soil at~ the
optimum moisture content. This can be easily attained by rolling tht! earth sufficiently in 6"
layers, with a 1'3 ton toothed roller 4' wide haVing knobs 6 x4" and 3" high, at the optimum
1

moi~ture content, Generally, 20 t? 24 rollings would be good enough for the purpose It
may be remarked here that the maXImUm dry bulk density would be obtained if the soil contain!:! I
70 per cent ~a~d and 10 ~o 20 per cent Clay I
As it IS not pOSSIble to roll the ends properly. all extra ",idth of Ii' on the inner s:de
may also be compacted and then removed before lining,

I
,--{
 215

(b) Banks.
If all the voids in a s~il are filled with water, the volume of water for a given pressure
is called the natural volume of voids. The greater the pressure in the soil, the smaller is the
natural volume of voids in it and vice versa. Settling will thus take place, if the actual volume
of voids in the soil is greater than the n'ltura1 voillme when the soil gets wet and apparent
co esion ceases. Consequently' the more permeable the soil, the sooner will it settle. The
so 1 in a bank should thus be rolled to the natural volume of voids corresponding to the pressures
wh eh will be in the bank to avoid future settlement.
The density of the finished bank should be ehecked by measuring it ;n a s:lmple fOJ
every 500 eft. or so of the bank. This can be conveniently done by taking out the sample in tho
shape of V, so as to accomadate a wedge, 6" square base alld 8" height, in the hole. The wed~
is then placed in this hole and a measured quantity of sand poured outside it from a graduated
cylinder, so as to fill all hollows. The volume of the soil would then be the volume of the
wedge plus that of the sand and its dr~ bulk demity can thus be cal'~ulated. In the case of smail
channels, it may he possibl~ to run the earthen channel for six months to ensure connection of
the back fill and the consolidation of the banks before the 1inning of the channel is done.
7. Drainage of Storm Water.
As alrf'ady mentioned, inadequate drainage facilities for the disposal of storm water were
in a great meaSlure responsible for damage to the Haveli Main Line.
The following precautions in this respect therefore seem desirable ; -
(i) No berm should be provided in the channel as it simply forms a receptacle fot wateJ
which can work its way behind the lining. . .
(ii) The dowel and bank should slope away from the channel. A sl:ope of 1 m SO for the
bank and 1 in 20 for the dowel seems desirable.
(iii) A suitable drain should be provided at the toe of the bank to drain away tm
storm water.
(iv) In cases, where there is a spoil, a berm at least 5' wide should be left on the spot)
side of the drain.
(v) 10' wide gaps should be left in the spoil 250' apart to dispose of the water in
the drains.
A typical cross section for a lined canal is shown in Fig. 2.

Fig. 2
8. Conversion of existing channels into lined channels.
For canals and branches it wil.l generally be advisabJ~ to construct a new lined channel
along with the old one. as the lining of the old channel cannot be carried out without intt>rruptint!
the irrigation supplies,
In the out case of small channels, the question as to whether the construction could b.
carried out in rotational closures may. however, be investigated in detail, in each case. In view 01
the saving which could be effected by this m6thod, as compared with a neW channel, considerable
expenditure would be justified on temporary arrangements for fpeding the channels durin~
construction. For instance, banks might be strengthened to permit channels to run with largro
supplies for shorter periods and the rotational programme might be modified to meet construction
requirements.
9. Lining in Reaches with High Spting level.
The linning of tbe head reach of a cana.l having high spring level will not generally bt
 216

a: velY economical proposition due to uplift pressures under the lining in closures.
There may be circumstances, however, in which it may be desirable to line' a reach in
an area of a high spring level. To enable this lining to be properly maintained and inspected
it would be nece"sity to arrange a system of pressure relief under the lining, baving a gravity
or pumping outfall, depending on the condi~ions at site
This press'ure relief can be arranged by vne of the following methods as described' in a
note on Confocal Conics applied to sub-soil to and from open channels, by F.F. Haigh:-
(a) ;\ contiI:uous inverted f,lter under the lining.
(b) A system of drains or porous galleries.
(c) A system of vertical reI ef pipt'~,
In the case of (a) there \\ould be great difficulty in avoiding any leakage between the
canal and the filter.
(b) Thi" method also suffers from this defect to some degree, since all the length of the
drain must be adjacent to the bed.
(c) This method suffers the least from this disadvantage and is proba\>ly the cheapest
to construct.
The spepage dischaJ ge and thus the sp::lcing and size of relief pipes or drains can be
easily calculated from the formula derived by F. F. Haigh.
Whatever system is used, unless the outfall is through the lining, connection must 'be
provided in the shape of drains or ducts from the relief elements to the outfall.
10. Losses from Lined Channels.
F. F. Haigh bas suggested thA following formula for the brick-lined channels :-
K=l'25 Q.056 where K=lo!:s in cusecs per million sq. ft. of wetted perimeter~
Q=Discharge of the channel in cuseC5.
Th:s formula gives results generally on the safe side (if there is no drainage). This means
a saving of about 75 p.c. in losse, by absorption and percolation as compared with tte earthen
channel section. .
The loss in the Bikaner Gang Canal has been estimated to be about 15 to 2 cusocs per'
million sq. feet. The tests of the losses from the Havtli Canal so far show similar figures.
11. Examation Questions.
]. Describe th e advantages of brick-lining over the concrete-lining.
2. Sketch a suitable section fer a lined channel in fillin~ and describe the precautions which should bll
taken so that lhere is no settlement in the back fill and the banks.
3. (a) What are the advantages of lining channels?
(b) What saviug in 101lSfS from-canals by absorption and percolation can be expected by
lining them?
4. Describe the various types of lining usually used and say which type in your opinion il! suitable for
(a) distributaries and (b) main canals.
S. What additional precautions are nece~5ary when the lining of the canal reaches with spring level
bigher than the bed is to be done?
 PART II

CANAL IRRIGATION
CHAPTER IX

Cross-Drainage Works.
1. Thfl alignment of all canals and distributing channels are selected in such a way
that they run along the ridge anrl no drains would thus be intersepted by them. However,
sometimes they have to cross the drains wh n the country is irregular and uneven. Works
necessary tn dispose of the drains are called Cross Drainage Works. fheyare usually classified
as follows:-
(il By lateral diversion. i.e., by excavating a channel parrellel to the canal the
stream can be thrown into another drainage line, for the disposal of WhICh provision
has been marie.
(ii) By passing it unrlerneath the canal either crossing the stTf'lam on a raised
; aqueduct, ()r, if the headway is insufficient for a clear passage, the bed of the stream is depresst'd
below normal level, anrl the water passing in a tunnel underneath rising again on the further
side. The latter is termed a syphon or a syphon aqueduct.
(iii) The nrainage water can be admitted into the canal itself. This is termed an inlet.
Uv) The, drainage can be taken into and cross the canal at the level of the bed of the
latter, the inlet on one side and the exit on the opposite side. This involves one regulator across
the canal and one at the further bank across the exit of the drainage. This is termed as a level
crossing.
(v) The drainage can be taken over the canal by an aqueduct. This is called
a super~passage, to distinguish it from an aqueduct proper.
(vi) The canal can be taken under the drainage
~ C
L
line by a depressed syphon or syphon-superpassage.
2 Masonry Aqueduct
(a) In Fig: 1 (a) we have an example of
the usual design of a masonry aqueduct. In this
work the canal bed is 25 fp.et above that of the
drainage, giving sufficient headway to pass the highest
flood, which is 18 feet deep.
Like all masonry aque(lucts, the construction
mainly consists of an arched bridge with platform at
canal bed level. and provided with two solid parapets
which retain the water flowing through.
Fig. 1 \aj
T~ reduce expense, the waterway of an aqueduct is made narrower than the average width
10 .... -1 of the canal in earthen banks, Owing to the smoothness of the
sides, the co-efficient of rugosity (N) !is much less thaI) t~at
applIcable to channels with earthen "ides and bed. bemg
'013 in. the former against '025 or '0225 in the latter. This
alone greatly increases the velocity, so thclt a. considerable
reduction in section can be effected, even it the original slope
of the current were retained. As however. that velocity can
safely be incleased to 5 feet per second, if sufficient bed slope
be given, a still further reduction in the width of the wateIwiY
in the aquf'duct can be effected. In the exan:ple we are
considering, the reduction in bed width is shown on the genera I
plan in Fig 1. (c)
(b) the thickness of the arch throughout is 2l feet, the
Fig I (b) radius being 20 feet.
 218

The value of the co-efficient in the formula, thick-

ness=nv rcan then he approximately increased in proportioI)
to the depth of watn carned ; in ordinary cases :1=0'4 ~
The following rule tor deducing the increase to the value 1 - - - - - - \
of n will suit in most caSES (Bligh's rule). Let d=df'pth of
,water, then n=C·4+V02 (d-2). When the d"pth of water is 1:=::==c::J::::::-:::jrl
2 feet, 'n' will remain -4 ; with S feet depth n=U'46, with 6 feet """'---'-A-'-1"::J.4lI....-......
n=O-48, with 7 fept n=O'S, with 10 feet n=0·56. This is for
large spans of over 25 feet. For smaller spans 'n' should be
taken as 0'5, r.o matter what the depth of water is.
In the example the correct crown thickness would then
be, wita 7 feet water carried=O'S X '\I for 0'5 x 4'5=21 feet
This would increase to 2t feet at the springing,
The parapets, as is usual in larg aqueducts, are widened
p

out to carry a roadway, as communica tion for cart traffiic must

be kept up along canal banks. The parapets here are 7 fett
wide at base and 6 feet at top. corbelled out to provide a 10
teet roadway, with an iron rail fence on either side. The Fig. 1 [c]
, thickness of the p:papet in this ca!'e is excessive. It can be made half, but should not
exceed two-third of the depth of water. In the next example we shall see that it is
made about i f) in width, D being the depth of water in channel.

(c) Tht> piers are {-- or :0167 S in thickness. They widen out by offsets to 7 feet,

o.or U'23;;, at the base, S being the span.

There is no definite rule regarding the ratio of the thickness of piers pIoportionq
to the span in the casp of large span bridges. It may be taken to vary from

~ to~,. For heavy works of this description the proportion; as in this case,

,would not be excessive.

(d) The floor is composed of inverted arches with a wersed sine (If 5 feet, the thickness
in the centre of span being 4 feet, and that at the !'pIing line of the inverts 9 feet. The
object of this invert is evidentlY t( distribute the weight on the piers evenly over somewhat
,shallow foundation. It is very doubtful whether the inverted arch does really act in this way
'the use of an invert is more to prp,vent a floor from blowing upwards nue to water pressure
, underneath, and it i~ used with advantage for this purpose in work SUbjected to a head of
water. In case:. of bridges. howaver, there is DO appreciable, head of water against the work
"consequently in'lerts are. apparently not necessary. The objection to their use is the great
obstructiou they offer to the free passage of water, by decreasing the effe<:tive depth to the
extent of nearly 3 feet. .

: ,(e) Reference to Fig. I-c will show the disflllsition of the wings always a most
. important point. In almost all aqueducts and super-passages double sets of wings are required,
viz,:, Two long curvf!d land wings to form the connection between the masonry aqueduct and
the earthen banks of the approach channel, and two water Wings connecting the face of the
'abutments with the river banks on either side. The land wings form really a continuation ot
.the parapet walls, and are of the same section at the top.· Being subjected to hardly
. any earth pressure, they can be built with verticalsides of the same width~lhr~)\:~~out as the
top.
(f) The length, a!! shown in Fig, 1 (c) is determined from
considerations (" percolation
:0' efficient. The land wing!! and water wings should be long enough to permiJ the distance
 219

Type Section

111
Fig ...
'UD/H.• III
• "4IL .... L .. '!L .
\_' /1' $j;
•. 71')'·...... r, ~ 'F
.: 'ts1 J :: ,I A0' 1.6 .tll~L ' .. :'1
,31'1 \0 ::Inil '" :>j:t i',lt1! '}. )T .• l!'.t"jJ&oi:~b a'li.,;;
'J' . .
t __
/'

.
.
, :T
\.J -,.; ----. - U5UAL
T
S,.,Al.tOW·"SYPHfJH -~:-
~ .j~ :~n
. DEEPER SY~UOH
---_._------
X to be more than the gradient 1 in 5 for the head frl)m the water level in the canal down
to the bed of the drain which may be considered dry, the distance y in the bed of the canal
is paved with concrete or masonry (impervious floor) to destroy the pressure of depth of water
in crt'\ep. Solani aqueduct, of Ganges Canal is shown in Fig. 2.
3. Masonry Syphon (Syphon Aqueduct).
An in~tructive example of a syphon passing dra;nage underneath a canal is given in Fig 3.
In the transverse section it will be seen that there is a fall of 3.5 feet in the syphon
This will enable a large discharge to be passed thorugh the work. at an increased velocity. the
waterway being considerably curtailf'd than what would otherwise be required. The spans are •.9
feet, the pier~ 2.5 fe"t thick, the headway 6 feet, and the floor 3 5 feet thick. The abut!llent IS
of reasonable dimensions. The disposition of the wings is apparently good. The earth lines, so
necessary to form a proper idea of the suitability of the win~s, are not given.
VERTICAL WELL TYPE OF SYPHON

-------_--_ .. _-- ------

_.
i.ii.:W-lr.::U~'j,i",..nil·~~A;;,.~ .;.~" --_ . . _---",..;!.-- -~.-------
• ,.,:?:Ithtt•• •• '<:ft fH'I6Nb".If- . .tat-o~.
.
••
, ..... ,.,.
..
1' .~2.r-·n.A", tEItID$ to U~~RL"M. Iii
..... ".,.-It~A.., H £."'1)$ ~F""AS"/'f t' "'...~

I
..
" ,.,
.L~
I ~i~~~~~"~~~"§"~"(~~'~"~'w~c~=
's·
Fig ...

j_
I
'~.~'~'~L-
-,' ...........
___________
6 ..... '

"~'_"
..._____________________
The canal is 7 feet deep and the parapets are 3 feet 6 inches wide at base and 2 feet 6 inchs at
top. The roadway is formed by an outer arch-way of 6 feet in width, the piers being lengthened
to carry it. This provides a road way of 7 feet. The exit of the work has a sloping rise to the
bed lev~l of the stream, which is as it should be. The head producing upward pre~sure is 4 feet.
i he arch, with concrt'te covering at the crown, is 2 feet thick, and with a specific gravity of 2'0
will just balance this pressure. No exception can be taken to any of the details, except possibly
that :t thickpr parapet corbelled out on both sides woulri provide an equaUy wide roadway, and
thus save the lengthening of the piers and the external high level arch. A cross section of a
l-ertical well type of syphon is shown in Fig 4.
 220
5. Super passage.
Cases where the levels are such that rivers have to be taken over a canal, are compara-
tively rare, as thfy involve a verv heavy work, if the stream or torrent is large, and so they
arc avoirled so far as possible. 'They" gmerally cccur in the upper st!e~-.a...,canal, the
headwork of wh:ch is situahid among the lower hills bigh up in the course of a river. There are
some very lan~e works of this descr pUen in the olel Ganges Canal. An pxeellent modern
example is given in Fig ] (16 of the Budki Su)::eJpassage en the Sirhind Canal in the Punjllb.
The length of the lVork is seven spans of 30 fett each, the width tet~'eeri parapets
being 4()O feet The arches ha\e to l:e raised well acove tho (anal bed on account of the
exigenCIes of navh!ation. and al,o of the level of the bed of the torrent. The 30 feetwirle spans
are therefore suitable. The thickness of thl' arches at cro\' n is '45y'X::- TIle thickness of the piers'
at springing is 6 feet or .2.S. ,W.e have already seen that a proportion of· \Is giving a
thickn~s of 55 feft is sufficient. The piers are therefore, too thick. The' bottom width could
be retained at 8 feet the piers ha.ving either a stra'ght or a eurved baUer as has >b~en designed,
The foundation;, which are probabh' on good soii, are admirable.
The abutment, at the Inok of thp. section, is dearly much too heavY, a very common
fault among designeTs. To prove this the actual incidence of the resultant line of pressures on

Fig. 5(a)
" .. of
if.u:u
'~~"\l'

.f; )i :

the ba.si has bee., gr!l\phically found; the wethod of working, consists in first fi~ding the ce~tre',
of gravi y of the ~alf arch and its load of water, the latter reduced in depth. as shown by the.,.
horizoatal dotte<J; 1in~. to an fquivalent ma~s of masonry, This process is shown in Fig 5 (a). and,
the reciprocal fUnicular polygon above. The forces 1 and 2 are the areas of the two halves in.to
which the half arch has been divir1ed. Having found the centre. of gravity ot the half arch, a
horizontal rne is drawn through the centre of the arch crown to intersect the vertical through
this centre of ~ravity, and from..{t.e point thus found the resultant tne R is drawn through ~he
centre of the arch at its springing, till it intersects a vertical line through the cen.t.r~~ of gr~vlty
of the abutment artt'i its water had In the force· polygon the load hnfl, composed of the areas
and 2, is continued down to,meawre 3( 0 square feet, the lIne, R is then drawn from "the
tern ination of 1. and 2 parallel to its reciprocal cutting the horizontal P at a point. From,this

.; ~u_ ~__;_ 1"~ ..

f
. '_",_,~" h~".~"·

'j! 1 ~ ~::.
L ... _:.~.;

point another line, RI joining the termination of th~ verticall(ad;lfn{T,;~~d;a"j1isi":obtallied·

gives the final resulta.nt RI. This projected on the profile of the abutment in Fig. 5 (a) from it.
 -_-, ..:::..._

,221

proper starting point, viz" the last intersection found. cuts the base of th", abutment at a point
~ome S feet within its heel. As no credit h~s been given to the weight of the earth. ba.ckin~ with
.. water above it. the resultant line Rl need only just fall within thp. b'lse. This proves that the
' abutment is l1nnecf'ssarily thic.~ The line P represents the horizontal thrust of the arch, and if
measured, will be found to closely ~orrespond with rt or 8 X 30=240. i.'ft the cakulated ho.riz'Jn.
tal thrust of the arch, If the abutment were made 8 feet wide at the springing and 13 teet· at
the base, it would probably be of sufficient section . As built. it is haH as widf! again ,as these
, dimensions, and beside5, is provideri with a l~rge buttress of which no account ha<; been taken
. Thi~ particular analysis is due to W. G. Bligh
. It might be mentjoned here that the calculations for the effect of buttresses in an
abutment or retaining wall are affected as follows:- The- wall should' b~ con.c;idered a, havjng a
base equal to its normal thickne:;s plus the length of the buttresses, but formed of two materi~\s
of different specific gravities. the solid portion being of the proper specific gravity of the
material and the part behind of a lighter sp~dfic gravity, equivalent to that nf'a material spread
. over the space of the same weigM a'i the solid buttrp.s, only. Thus, supp:)sing a walt 6 feat thick.
is provided with bl1ttress~s projecting 4 feet thick and 6 feet apart. ie' at to fap.t intervals. and
let the specific gravity of the \yaU be 2. then the spscific gravity of the 4 feet wide space be~ind

, will be 2 X..!.._='8 and the effective base width of the wall would be 10 feet, not 6 feet .
10
The disposition of the wings is generally similar to that usually adopled In aqueouets
'!""""_",.._ _ _ _ _ _ _ _ _ _- - . . " . . _ _ _ _ _ consisting of water -wings as cnrvtrl
continuations of the faces of the abutments
and splayed land ... iogs which. carry
t======-==== parapets in continuation of those 1U' the
aqueduct proper. These wings are shawn
in the plan over all Fir S(c). As a further
precautiun. the ends of the land-wings are
connected by a cross wall at bed level.
which a:pparently goes clown to the full
depth of the foundations The hnrl-w'ng.;
being in solid ~round ,are stepped lip
Fis. 5 (e)
in foundation, which is shown in the
levation Fig. 6 ;a) .
Dhanal1re Level crossing upstrea-n v:e ',v, Gang~s Canal

" '

- t :-
"
"
,6. Level Crossing.
Dhanaure level crossing in the Upper Ganges Canal is a nice example of .this type
of work which is cODstructed when the bed levels of a drain and a canal coincide at the
junction. The usual layout is shown in Fig. 6 and Fig. 7.
7. Author's Auto-suction Weirs.
The author carried out a few experiments on the subject of cross drainage works to
deal with small·seepagp. even-surface drains across canals. The results were published in the
Indian Engineering, Calcutta, Febluary, 1~38, in an article. Three methods were suggested
therein---
(i) Auto- suction Weirs. (ii) Syphon Spillways, (iii, Venturi Flume.
A brief summary of the said publication is given here.
(a) Our knowledge of disposal of drains across canals is limited to aqueducts, level
crossings and Syphons. In thp. aqueducts, the drain bed is required to be well above the full supply
level in the canal. Level crossings are only feasible, when the drain bed and canal bed coincide
Syphon requires that the barrels are not completelY choked up in order to permit initial
waterway for the floods in the drain If the bed of a drain happens to be between the bed level and
full supply level in the canal, none of these devices are applicable. A Syphon would be
completely choked up and would have no initail waterway In a level crossing, the drain would
tend to silt np the canal. Aqueduets and level crossings. are out of question when flood levels in
the dra.ins are lower than the full supply levels in the canals. Even a super passage is not
possible when 8. drain has got a small depth and bed level near about the bed
level of the', canal. The sllbject of drainage works across canals perhaps has not progressed
beyond where the primitiva hydraulicians had left it.
(b) Remarkable advance has no doubt been made in the design of sucH on hydrautomats
which could lift up small drains into the canals. A suction hydrautomat requires a minimum head
of at least 50 feet. Similarly with high heads available in the canal, the small drain could by
pumped into them hv the e~ection, of compression hydrauto~ats or hydro-electric pla,nts, Big
falls are not generally available lO the canals mar the drams and all of these deVIces are
anapplicable,
(c) This subiect is of very great importance to the province of the Punjab where seepage-
cum-surface drains are a pressing necessity to relieve waterlogging. The contiguration of the
, country is often such that these drains are inttlfcepted by the canals. It often happens that such
drains have their bed levels above the bed of the cana.ls ann their flood levels lower than the
supply levels in th: canal~ P1!mping into the canal by oil engines. has generally been tried
pumping deals sahsfacton1y wlth seepage wat,er. but flood wa.ter remams acc1J~ulated for a long
time till it is cleared by ~he pumps, The gravity outfalls are not generally avatla!:lle except by
running the drain long dIstances parallel to the canal where seepage from the canal itself is
difficult to co~tr,Ql.
{d) fh~.,disposal of sma.ll seepage· cum surface drain .across the c.anals with bed levels
above the canal bed and flood levels below the supply level m the canalIS not a subject which
could be said to be of impossible solution. Such drains could easily be drawn into by the canals
In all the above mentioned three conrlitions , the drain should not be disturbed but the canals
should be t;e\lted ?n the line~ as described hereaftpr.
(i) Auto-Suction Weirs.
Model experiments were carried out by the writer on numerous forms of this device
for drawing a natural dr']'in into a canal. The form of the weir in Fig 8 was found to have thfJ
maximum effici~ncy and c?mparatively. low. working heads. The results .of two experiments
were published In AppendIX I of the said article. The model was 2 feet wlde. The maximum
dep1h on crest H was 1 foot. The length of crest was 2 H, i.e, 2 feet, out of which 1 5 feet was
level and .5 foot was curved with radius of 2 H There waS a drop of 0'12 foot from the crest to
the point where the glacis in 5 started The sl~pe below th~ orific~ for the drai~ was ta.ngential
to the curvature of the crest. The control sectlOn of the WOlr. whu;h wa.s at a dIstance of H feet
{rom the belZinning of the crest, was not affected.
 ~Lr/J.'(\,
l.~?\
(~ ::."//l !.J r,~ I~
.. \ t.;
0···· .... ..... I

\.\ ",
'l,~
"t
\ ." /
.............. _ . .. I •

1... 1 6R i\ \-~.
"'-.~---.~
 'l24

~
=
'"
U
II)

'"
b.D
~
- ': CIS

'"
~
(.)
p
-g
P
• 0"
<:
·2
'"
"0
rJ)

.>
 Auto-Suction weir. The drain discharged into the
canal against a reduced prf'ssure 01
the jet of water down the glacis on
i account of the pronounced effect of
j'
..+C04'YA~.
the centrifugal force around the
curve of the crest. The design of
....___ ._i, the weir was the same in both
experim~nts, hut discharges of the
drains were varied. It is clear {rom
the results tha.t in the first
experiment the efficiendy of liftio8
dr<t.in water into the canal is high
but the working head of the wei,
is comparatively high with low depths
on crests_ In the second experiment
the efficiency is about 52 per cent
~l; "} : of the depth on crest of weir which it
(n,
Fig. 8· ,c, '. gcod enough and the working head
with 1 foot depth on crest is as low all
18 per cent. If scale of the the prototype be 5 times that of the model th~ depth on crest will bt
5 feet. The working head requirf'd would ue 1 foot for the weir. The floods in the drain would bt
drawing into the canal with level..i.n the drain 2-5 feet lower than the full supply h;vels in the
canals downstream of the '''!eir. The discharge of the drain taken up bv tht' weir would be on.
cusec per foot run if the height of the opening for drain in the prototope is 0-15 foot.
(ii) SplUway Syphon.
The canal could be passed over the drain by designing a low head spillway syphon 01
the form sketched at the next page(Fig_ 9) which is suitable for efficient prilLing.
Low head spillway syphons are very suita.ble to pa"s the drains undisturbed undCl
the canal. The spillway syphon can casil)
be designed with the low heads so thai
o _ '·I_C, the pressures at the summit are equal to
or gr(;ater than the vapour pressure oj
water. Average temperature of flowin~
wator in the Upper Jhelum Canal has.
been found to be about 25 degrees C' iD
condection with absorption experiments.
The water vapour pressure for th~
temprature is about 13 feet head 01
~~~~:z;?z~:'Z~~~~~~~zt~?Z~~"~"'~
..
water. The pressure a.t the summit of the
syphon may be calculated acccrdidg t.
the following formula.:,-
Fig. 9

va
P=(A-H- _I
2g
-HE)
..! " ~ H t: n ~'1'It9(J !

P=Pressure in feet head of water. .

A-Atmospheric pressure-34 feet. '." .
H=Elevation in feet of the centre of the summit .etion, above wa.ter level upltreo.m 0'
the syphon. .
v.=Velocity at the summit of the syphon.
H!= Losses in friction.
(iii) VenturI Flumes.
The simplest solution to pass the drains over tho carta.l is to construct a Verttltfi
 226

"'~t''!l'_''''''
,
--'-. -.. _. .-.
~U . . LI.~U. ~
~- - =::"r. . "~ flume in tho e canal as Fig, 10, Let the
section of the canal be constructed' to
I '
\ s,
~
I
,.. ~ e.A'H , . ,
f '
~
one half so that velocity of the cana 15
pvzazzzuzzzz';);';zzzz ZZ) lIZ ZZZ~ ',__ tinc7refased, say. froffi 3'5 feet per. :eco~d
--., 0 eet per seCOlld , y ...... r .
Fig, 110
2 7" 3 - 2 2
'.s,

_.
,::--::~_
• • -, • ,.
/
,
.
The loss of head~ _:!!_ = __ 2
2g 2g
=0'57 feet
2g 2g
Tho bed in the canal may not be depre5sed for the sake of cton.veltieat, .
...:,V?!"-.....
j''';~: r' inspection~'
.
--.k
' • -.'

The uplift pressures under the top slab are likely to be ·tt~ubreicj~e but t_~ft" can
..nainly be managed at no additional cost, . T ---

(c) Conclusions,
The loss of head roquired in no case is more than a foot if these designs are adopted in
, canal of, say 4,000 cusecs discharge. Let the depth in the canal be 9 feet and the bed width 150
(_t. A drain of 150 cusecs diSCharge could be easily disposerl of hy adopting any of thp. al-)ove
,n.mtioned designs, A fall of about a foot coulrl generally be created because if the drain has got
'Md level above the bed level of the can'll. the canal would be in deep digging. ThA main dra:n'l
~ve hardly been constructed in this province yet. When the I'lrains are extended to fields, there
.m be numerous cases where an ., Auto suction Weir," will provide a very cheap device to dravy
tmall seepage drains into the ~rrigation c h a n n e l s . , ,. ,.
t. Example R. C, Trough aqeuduct.
A distributary is to cross the main channel.
Design an aqueduct with the following d,Ua.
Main Canal :-
Full Supply discharge ti7CO cuse~s.
:Bed width t=60'
Bed T,eveI fiC: 544'5

Full Supply Level ,,=549'6

Slope "'" 1 in 6666
Main Velocity _2' 1 tt,/Sec.
DIstributary :-
Fnll Supply discharg. : tr:::409 cusec., ",;
Bed width . ==4Ift.
Bed Level f=550'7
Full supply levd =550,1 ~.~ I
Slope ..... *=1 in 6666
"~
Angle of intersection _~ • .', t::~~~" ot>.......,,~ ==61 0 30' 13#
I~
Natural surface =556,12
Branch Canal Flume.
Let thfl' seclioo be contracted to ~ It. having :J spans ()f 12 ft. with tWo-'pTets (If
"3 tt, each.
Distance between abutments=36+2X 'Z S=40 (; ft.

N'r It.
Discharge r~ 'Nfdth=,~
. ~
= J~'~ CDsees.
, The section of the aq-ue'drtct is ress than 60 Ct, as it IS' n'ta~otfry a.-n& the ~o-effid:ent 01
lliction is Ips!! and so- the velocity is more. '. '. _...,
By doing this, we have saved the cast IJ)' 1 .
f~ f
&Jj,iieaFdep'fh=tDc=~ q2_ ==~195X19'5 -=2.1tt
8 32'2 '
 227

The depth cannot be lower than this. Let depth be 5'6 ft.
Then C in the discharge formula

Q==CB t H3/: ; or C=Bt~3)2 = HJ2 where q = discharge per foot width and B is 1 f,t. widt)

v2 2'lx2'1
H=Head in ft =56+- =5'6+ 5'669
" 2g 2x3~'2

. C=___!_~-- =1'44
.. I X (5'6)3':
From curve in paper No 12S-Puniab Engineering Congr~~s·. (Fane's curves 'Piite IV.
C=I'44 ",hen drowning ratio =0'99 :.Loss of head=5:669 X '01='06 ft. \0
, DP.pth has been tncrea.~ed by 0'5 ft. at the flume sectlon 5'1 ft. to 5'6 ft. Therefon
the fioor will be depre;sed by O'S ft. The upstream approach in bed=I in 30. ,btl.• _
Tbe Hoar level is transitioned b the floor level of the downstre,am bed.
Discharge Flumc.
Lpt the floor levd of the distrihutary flUme be 552'1 to ensure clearance for the brand.
can,al The be<i is r:tised s) that clear:tnce of at least 1 ft. is ensured. between the bottom of t114'
slab of distributary and full supply level of the canal as it is an aqueduct. ' .
nepth of the water=3'0 ft. Let C=2'1
Discharge per ft. width=CH3/2=2.1 X 33 / 2 = 10'91 cusecs.

Width !'equ'red= ~=37'5

ft.
10'91
Ag~in from Fallc s Curve3. paper No. , 125 of the Punjabi~, Congreii. taw
drowning ratio is '97 for C=2.1
Los~ of head-'03x3='09 ft. Say 0'1 ft.
D'esign of R. C. Trough of the Aqueduct shown In Fig 11.':
Calcula tion for the bottom sla':> :-
Let it bA I ft. thick .
The load on the slab is that of the slab and water.
Wt. of sla.b 1 ft. thick= 150 Ibs. '
Wt of water 4' depth=4x62'S= 2501bs., .
. (for wOl;t condition excess supply upto the top c)f tter~~. X!
NP.t 1.lad=400 Ibs /sq. ft. ,
The slab is continuous over 3 spans of 12' clear .
angle of intet!'ection =61°,30',15" ' , ~'
. The reinforcement. is o~lique and parallel to the sides. '
.dt, '.: .,"',' , " ,
"aW

,R,~. TROySIt,
L. ISfC'fg,"" ,

tt.1J it; ~:;:;"Efiedi,*~ ~pan:'*, Sin ~~O~OI15' 13.1+bearing qfslab;"';!&~

' ..=750 Ibs/tJ' ;=1.=18000 Ibs/O":i n_ i' _Ii

c
 B M = Wlz 400 x 15 x 15 9000 X 12 Ibs inches
•• 10 lu

Shear = WI _ 400 ~ 13'7 =2740 Ibs

2
M
tit" Let d be the depth.

d=c"MIO ='485X.t 9000 X 12=8'1-:

V I V 12
Let the depth be 12'/ with 1'5" covering.. 10'S'" is th,4fffective depth.
Design of Reinforcement. H r "" a
,:. £/~J'
M 9"
Sq.;~b.~'~of16 4~'
108000 . '. \
.V! Steel=ISjd =i800X'S78x--To-:-s-= 65 dia:
c to c.
Bond.
::;; O=3x 1'77531.
U= 2_?~0_ _ =68 tbs which is safe
5'31 X'S7l:Jx 10.5
Temperature Sieel=As =p. b, d='0025x 12x 12='36 sq. inch" 4 64($
pacing 6" c to c, both on top and at the bottom.
of'-l16 " dia .
Side Walls.
fhey are to be cesigned for the water depth of 4' including 1'5' free board.
Overturnmg WW = ~~x~
' movement due to water pressure=~6_ 6 ==665 ft. Ibs.

WH =62'5x4x4
Share= ___ - - ---=500 tb-s.
2 5,\

let d be the depth then d=C_!f_ = '085 X 665:>( ~~ = 4'2'"

b 12
From practical consideration it will be kept 12" with effective width==lO 5'"
~s=~=
f, Jd _______
18000 ~~~_X_1~ __ .. ='41;
X 'SSx 10'5
use 1/4'/ bars 6'/ c to c (for safety)
femperature Steel.
A. =p b d='0025X 12X 12='3G sq. inch
Use 5/16" dia bars 6" C to C both inside and outside.
The side slab of the aqw~du:t is acting as a beam as the sido is monolithic with the
bottom !lila':> capable of taking load so we put 2 bars at the top and two at the bottom to tab
the load of 1he bottom slah np to a length of 6 ft.
The bari are are 5/8 11 dia, ; 2 number.
EXpansion and contraction in Fluming.
(a) Let creep ro-efficient be I in 7
Creep len,goth required=7x 4'4=30'S or 30'
(b) We contracted the upper ch:mnel to the flume width==37'5' /1
Actual bpd=41' ; and the difference=41-37'5=3'S' I
The contraction on the upstream side and f'.xpansion D,S. side be \ "~J:j:~th 01 t.bI
!><,cca protection= 3'5x~ =__!_~=S'75
2 2
The flaring is designed from vertical to 1 to 1
The Jf'ngth of the flaring wall for depth 4'4'-4-4 X5.. 22'
 229

Total length required=22+S'75 =30'75' say=30' ~""

Difference on oneside=1.75'; Flaring on each side=4'4 It .~
Total expansion and contraction=S'15' .•
Allowing it 1 in 5 length of pacca pro~ection on s.ide=5xS·15=30·75 or say-30 1
Sections of the walls change from vertical to sloplllg 1: 1.
Percolation through banks.
The percolation head=Distributary water level-bed of canal.
=555'1-544'5=10'6 ft
The safe percolation coefficient through banks is 1 in S.
The length req11ired=tO'6 X 5=53'0 or say 55'
Questions :- Cross-Drainage ·Works.
I. Sketch a design for a brick arch syphon to carry a discharge of 40 cusecs uncter a disty of bed width
50' wa.ter depth ~ '5', height and width of banks 4'0' R.L. of bed disty 560'0'. RL. of bed of dtain 556'0'
T.e.E. 1933,
2. (a) What are the diff~rent methods of disposing of crOS3 drainage intercepted by canals?
(b) Sketch a design for a syphon to carry a discharge of 100 cusecs under a disty of bed wiath of IS'
discharge 100 cusecs, beds slope 1'2 ft. per mile, RL. of bed of disty 610 0, R. L, of drain 60S'0 maximum depth of
W;l,te1 in draIn =4ft. T.e.E. 1934,
3. Briefly describe the various canal wcrks required for crossing a natural drainage. F. U. 1942.
4, An Irrigation Channel carrying 700 cnsecs, with 60 ft. bed width. 5'1 ft. depth. side slope 1/2 to 1 and
bed slope in 6666 passes over a drain in a ma.sonry acqueduct. Design the follOWing if the archway ha3 a span
ef 20'0 ft.
(a) Flumed rectangular section for the channel with no loss of hea.d assuming co-efficient of rugosity fat
brick masonry as 0'013.
(b) Archway for the drain. •
(c) Masonry Sides, waH and approaches for the aqueduct.
P. U JQ.U_
 230

l OfSCHARC;£ or WEIRS
...
oe • eeH~·
Q . . . OlSCHAIlCU: 'If CU6IC.
8. e WIO'rH 'If FL!.T ~
;:,
~. ~ eo- 'F"'C~"'T. ~

II. = "."'., OVI!" C,,.ST ,,,el,.,,,,. 1 ,_

.,) Ol",v..".CITY HI.O. i)
()
() 'J-!I ...:
......
IIJ
"-
.0.. ~

'0"
~ 2:

X.
'.-41 t
()
,..
U
0
\J z:

,... III
II)
cr:
~

....
...G:: ••
oJ
;)

"-<,..
Z

..,
~ '(!
~
.
,a: ' ••0
...
iii ...ct0
.... et
~ U
) ...
",.
"I
,
.~
.... 4 .• CD
;::)
v
".~"o
;lo. .
0
"-
~,e Q

~
••• ...
~ "C

'"J:.
\.
'0
... $.0

.f ••
~
III
e"
II::
"-
()

~
"
~
(I
.t••
'" \I
...
~
1&1
~
~.r "(

'"
Q
~
a..
'·tI
"-
0

'"~
'lilt

.... "
V\

','0
 FLOW OF. WATER OV£R A NOTCH

) .,.,.

v ••
lei
III
'70
~

-J
,.,
,,/ ,~
....
I.. ·c ~

..
VI ~

~
;:,
It' .s~
I ·s
,Q"
... ,,"0
:t;
'-J
t)1 "-
0
·4

••

,Of
..,

.., .,

·1'

~o
A. 8.
Fig. 2
 PART II
CANAL IRRIGATION
CHAPTER X
Falls and Meter Flumes
I. Definition.
A fall may be defined as a work de.signed to secure the lowering of the water surface of
a canal and the safe destruction ot the energy so liberated. In Northern Tndia such works are
commonly named as falls but in South India and America, they are called ;anal drops.
2. Necessity of falls in canals .
. Canals arc earthen irrigation channels. They require to be designed with the velocities
nep.end IDg on th.e nature of the soil of the bed and the sides so tha.t they neither silt nor scour
thelr beds ';lnd sl des. Su~h a ve~ocity is known as the critical velocity. Lacey fixes the regime
slope p.llrbcular for a glvpn. dlscharge. In every discharge formula, the bed slope is ::t
predommant factor to deternllne the velocity of flow. In order that the canals Iun non-silting

Fig. 3

they can only be designed with the requirf>d amount of slope in the bed. If the slope of the
country exceeds the slope which can be I!iven in a canal as shown in Fig. 3 below, vertical
falls or drops have to be given in the canal be4 and the water surface.
3. Location of Falls.
The foll )wing factors control the selection of the suitable site for a canal fan.
Ca ) In the case of main and branch canals which do no direct irrigation, the site for
the 1all should be determined from considerations of economy in the, digging of the canal and
the making up of its banks. The most economical section of a canal is that in which the earth
1ug out from the bed is enough to make up the required eal thwork for the ba.nks. The depth
of digging is known as the balancing depth of excavation. A suitable site for a fall would be at
a place where the oepth of digging drops below the balancing depth, such that the reach with
digging more than the balancing depth downstream of a fall is not longer than the reach upstream
of a fall with digging less than the balancing depth.
(b) In the case of the distributing channels, such as distributaries and minors, the falls
are located from considerations of co nmand of the are I to be irrigated. Suitable location of falls
with outlets taking upstream of them results in command of extra areas.
(c) The combination of a bridge with a fall usually results in economy, because the
floor and walls of a fall can be utilized to serve as foundations and abutments of a bridge. If
there is going to be a fall near a road crossing, either the road can be diverted to the fall site
or the fall may be shifted to the road site.
(d) Where a canal has to be bifurcated, a regulator will be required. A fall should be
combined with the regulators to save the cost of the masonry works.
 233

4. Development of the practice of Fall Design.

. . When t~le grtat irrigation projects came to be constructed by the British Governmerit
III h dIa there eXIsted no theory of the falls to guide the desio-ner and there were no records of
the successful pr::Jctic~ to be followed. These projects were t> the Western Jamuna Canal 1817,
thl'! Eastern Jamuna (anal 1823, the Ganges Canal 1842 (now Upper) and the Kittna Canal
(South India) 1852.
(a) Ogee Fdll of Sir Probby Cantley.
. In the design of an Ogee Fall an attempt was made to avoid the destructive effect of
verbcal. drop on the downstream floor by providing a curve from upstream bed to meet
tangentially the oownstream ted. The direct impact w~.s r 0 doubt avoided but it rf'~ulted in draw
oown with excessive velocity lind ,cour upstream, which \\ as soon rectified by constructing the
crest as shown dotted in, Fig. 4 All falls on the Ganges canal are Ogee Falls. 1he forward
velocity of water remained excessive and nestructive in causil)g
" the bed and ~ide era: ion downstream which necessitated
the recurring addition of oose stune pitctling on the. bed
and on the sides. Every f 11 possesses such protection more
than a thousand feet in length.
Fig. 4 (b) Rapids of Liert : Croftm R. E.
Ogee falls were followeil by Rapids where the drop was comiderable. Rapids following
the weir design are still sllccessflll falls. because they utillsed the part played by the hydraulic
jump (though imperfect in destroying the mergy of the fall without the designers knowing
;this inherent improvement in the Rapid Design). The rapids are very expensive. The western
Jamuna Canltl Falls were made rapids with very flat glacis sloping 1/10 to 1/20 d<)wnstream to
admit of Timber Traffic over them.

.. (c) Vertical dl'op with cistern •

Rapids were folloNed by falls having vertical drop with cisterns as shown in Fig. 5 .
.The dimensions of the cist~rns were put in arbitarily in light of the experience of the designer.
Another device in,the form of grid was usually used in the cistern intercepting the dropping jet
of water as slnwn in Fig. 5. Grid consisted of baulks of timcer either horizor:1al or inclined,
spaced some inches apart. The grid became clogged with jungle carried by the stream and its
clearance was not gen~rally practicable except in a closure. 1 he grid timber rotted and had to
be replaced. This device has now been abandoned.

L-,L-_ _ __
"Is 8£:> ~·J.LE"U. Fig, 116

.
.
~ ~ \
"

(0) Trapezoidal Notch Fall (Ried 1~94).

This design consisted of one or more trapezoidal notches in a high breast w~ll ~cctOss ~
channel with smooth entranCf! !lnd a flat lip proj~cting downstre::tm to sPlay the drOpplOg let. The
notches were calculated to give the required discharge at a Half Supp~y ~epth and Full Supply
bepth, the error in between J-.ein.g neghgible. The dis.charge d~pth reJabon 10 the chan~el abo'Ve ~he
work was maintained. This relatIOn means that the dlscharge III an earth~n char)J~el vanes ~ccordmg
to mean exponential formula which is expressed as Q=k D513 were Q IS the dlscharge m cus~s.
 ~34

1) the depth in feet and k a consotant. A typical de,ign is worked in paragraph No. 11 of this
chapter and is sketched in Fig 124. fhe object of the scientific des:gn m keeping the crest at the
upstream beds level was to stop silting up of the upst~a!ll c,",annel in Llw supplies,
The dr owned falls give a lot of trouble dnwllstream inspite of the provision of well
de~igned cisterns while in the case of free falb there was no troublt! at alL Never-the-Iess the
trapezoidal notch fall w~s so suc :essf .. f that it held the field in India for many year/? and was
copier! ~ll over the world. There was one seriou;; def~:t in these falls that they could not be
'.!sed as regulators in addition. In some cases groove" upstream of the notches were prov!ded
(or inserting horizontal karries but they gase a lot of trou~)le by getting choked with debns •.
(e) Free OverfaU of the weir type.
With the general improvement of irrigation practice, the more economical and accur~te
distribution of water became pss~ntial Engineers foun i them ;elves obliged to measure the
disch:uge at frequent intervals in irrigation channels. fd.lls had to b~ designed as meters of
supply. Nutch tall~ did not admit correct cilibration to g VA a di3clurge table depending on
11. gauge reading. Broad cre;;ted-weir., with fre~ overfalls supplied the necessary solution.
Further research work on broad-crested weirs p·ov..!d soon that tile ire:! fall formula
could be used even up to a drowning ratio of 85% to 90% when a hydraulic jump was formed
downstream of the weir crest.
The efficiency of the hydraulic jump as a very pJtent means of rle5troying the energy
of canal falls was brought out clearly by the research work of the M"jamy Conservancy Board of
Engineer;; in a publicatlOn by Sherman M Woodward on "Tl1eory of Hydraulic Jump and back
water curves" 1918.
This brings us to the modern designs of falls which are de3cribed in detail in this
c.hapter.
(f) Contracted Falls.
The falls of the above mentiond types were gelerally c!)n,trltcted of the full channel
width. In the case of trapezoirlal ~otch falls, the maximl~ n c )llstruction of the width at the
fall was never more than lth of th~ bed width. But the Slltlej Valley canals in the Punjab were
constructed soon after the last great war, the rates of construction were very high, and th~
~olution for ecnomical Cl)ll3tructioll was found in contractin~ '1r tlu'ning tlu fall, and comhining
them with bridges. The fluming wa3 done in some ca'ies mlre tl1'ltl ha f the bed wirlth of the
channel. The resultiug action downstream was generally ser;o 1" and in mlny ca'3es disastrous.
(Paper No. 147 Punjab Engineering Coug-ress, Lahore by E.L. Pro.heroe on Dam:lge to Falls
on Khadir Branch, VJry Pakpattan Canal).
This opellP.d up an interesting and a new field of Res,arch work indinting that there
was still surplus energy even after the formation of the hydraulic j Unlp on a glacis or the impact
in a c:stern which had to be destroyed further down. The developmpnt of th ~ subject took place
in two directions, firstly to invent ways and means to arre:;t the destfl1c~iv(' PJtentiality of the
high velocity jet leaving th~ fall near the bed in the for.n of hafflt! walls and block~ of different
kinds finally devolping in the InglisJaU design. and secondly attempts were made to arrange
horizontal impact in the hvdraulic jump instead of a jU'np OQ a ghcis leading up to latest
:vlontagu's Fan dpsigns. Both these designs are des:ribed in the S~')5 'q:l~nt paragraphs. The
I atter has proved to be the most efficient. /
~. Classification of Falls.
Falls may be divided into four principal classes in accord:tnce with the purposes of the
approach.
Cllss I.·-Fal15 designed to maintain the depth diScharge relation.
It was believed at one time that a high-crested fall w;)uld ca'He Silting in the channel
upstream of it. This is no doubt the case when a channel is run far long p !riod. at less than the
designed full supply during the silting 53ascm A, the(~ is g'e.l.~ Vlriltlo l of the silt ontent OV..~r
the year in the head reach of a channel, the pan lin~ effe2t of ;). hig\cr'::3te 1 fa.ll h the head relch
mey have the effect of creating silt in the canal sY3tem below.
 " 235

If the channels are run £1111 or nil supplies as is 11suaUy dJne in the casp. of distributaries
and minoIs, and if the period of low supplies are not protractf;d, m()d~rn opinion holds that the
silting effect occurs only for tte limited distance of the brl.ck water curve and that any such silt
deposited in bw supply turns will be radidly swept out during periods of supply and eqn:ltJrium
restored ..
The principal types of the class are the trapezoidal notch falls and the low-crested
rectangular notch falls. The trapez'Jidal notch falls are described in detail later in this chapter.
The shape of the trapezoidal notch is rles;gned to give conect discharge at the corresponding
channRI depth at say, half S:.lpp:y and full supply. Thp. error at intermediate discharg ~s is
negligible.
Class II-Falls designei t) maint tin a fixed supply level in the channel abave fIe wor:.
There are usually two reasons for cons'ructing a work of this type ; -
la) When a hydrauLc power station is combined with a fall, tke de5:gn of such
machinery often requires a m)re or I S5 fixed Intake level.
(b) When a sub.,idi:Jrv channel takes 01 t'le parent some distance above work.
Such works do suffer from the defect of silting caused in the channel above the work
by a raised crRst fall, and the constant level is only near the work just upstream. while Hle
levels drop off in the chann"l according to the back water curve in low supplies.
The principal types of the falls of this category are syphon spillways and high cre,ted
weir falls. The syphon spillways are described in Chapter XIV of Part II. High crested
weirs may be used when abs,)iute constant level is not requiIed, but it is only necessary to
maintain a high level above the work.
. Class 111- Falls designed to admit of variations of the surface level atove the work
at the will of the operator. Such falls have also to serve as regulators.
There are three m~thods of regulation usually resortEd to
(a) Sluice gates (b) Horizontal stop-logs or karries (c) Vertical needles. In prac:ice
falls of this class are confined exclusively to rectangular [!otches.
The width of the notches depends on the size of the available gates especially when
they are made of steel. Considerable economy is possible by the manufacture of such gatp.s in
stanrlard sizes. The reduction in number of notches will increase the cost of gatt's. [he mcrea;t
in the number of notches with the corresponding reduction in the discharge intensil y per foot
run will result in reduced cost of the distribution of energy of the fall in the cistern.
Horizontal karries are usually of deodar wood of 5" x5" section and quite strong for
spans upto IOta 12 feet. The ewls are protected by iron straps. There is provided a groove
. at each end having metallic rod across it. The r'lgulation staff is provided with poles with
metallic hooks at one end which can fit in the grooves of the karries to lift them up. fhe water
level is regulated bv removing or addiug kaj·ries according to the requirements,
This is a very defective arrangement, bf'came it is very difficult to remove the karries in watJr.
Moreover, they are often chokeri with jungle which has in many cases resulted in complete
blockade of the falls accompanied by many bff~aches in the channel upstream and the outflanking
of the fall itself.
Regulation by means of vertical needles is very much superior to that by the horizontal
karries. Their lower ends rest against an angle iron sunk in the crest of the fall and the upper
end, bulking against the regulating foot bridge, and projects a foot or so above it. The needle3
can be easily pulled in and removed by a man standing on the foot br:dge. Moreover, they a-e
put near the ends of the span witli free openin~ in the middle and. therefllre, they do not choke
with jungle. Where gates cannot be procured, the needle regulators should be preferred to the
horizntal kari rpgulators.
Class IV.-Falls under this category are designed from special consideration without
any referencc to the <l.pproach conditions.
The principal types are : - " ..
(a) Cylinder Falls usually called well-falls. 1"

They are quite suitable and economical for low discharge and high dr )ps. Water il
thrown into a well over a crest from where it escapes near its l>ottOIJl fhe destructi m oj
eLergy is usually complf'te in the well.
 236

(b) Chutes or rapJds which conduct the stream in an open inclined trough.
(c) Pipe Falls. Where a pipe replaces the chute. ~. ;

. i
Each of the first 3 classes of falls may be designed as.
A. Meters. P. Non-meters.
. The trapezoidal notches can be calibrated to work as meters provided the
channel upstream does neither silt up nor scour. On the whole, it may be said that the
trapezoidal notch fall is not a satisfactory meter.
The rectangular notch falls can satisfactorily be designed as meters. In the case of
sharp-crested rectangular falls the metering is limited by th~ free fall conditions available.
Water level in the chanuel below must be 10 <ver than the crest level of sharp-crested rectangular
notch.

It i'l only the broad-crested rectangubr notches viz weirs which prove as su :cessful
meters of supply. Tte design is described and discussed in datail later on in this chapter.
Th ~y can be designed as meters with a drowning ratio of 85 to 90 per ce;:lt of the depth on crest.
6. Destruction ot the energy of the tall.
The mass of water dropping from a certain height at a fall in a canal represents the
f'nergy ~hich mUst be destroyed in the fall design. The following methods are usuaJly employed
for effiCIent destruction of the energy thus liberated by the fall.
(iJ Impact.
. (~) Impact t)f falling jet over a vertical crest .against the dow~str~am Bo?r is ver):' efficient
In destroymg the energy liberated by a fall. Sometimes the droppmg Jet strIkes agatn,t the
floor at downstream bed level and often in falls with a vertical drop the downstream floor is
deprpsspd to form a cistern, which serves to add to the depth of water acting as a cU5hion and
reduces impact against the floor.
. (0 Impact can also be arranged between two strflams striking against each other. In
a slUice regulator there may be rising and droping gates. Water pa~sing through the lower gates
can be made to rise up meeting the dropping jet from the upper gates.
(c) Impact can also be arranged between the fast moving stream leaving the faIt and
the slow moving stream in 1he channel downstream as in a hydraulic jump. H.' draulic jump is
now recongnized to be the most efficient means of destroying energy liDefd.ted at falls.
(d) Similarly impact of hypercritical stream OJ) a glaci, downstream of a weir can be
arranged against a permanent waH across the glacis known as the baffi~ as introducdd by Inglis.
Impact.against
been tntld. staggered blocks on the gIaci'i bdore formation of the jump has also

Colyer's Biff wall.impact of reverse flow in a. cistern has been tried with success as in the
(e) The , case of
. (f) Lateral impact has not yet been trierl- The supply of the channel may be divided
Into equal parts and passed downstream from sides. High velocit.y jet issuing from opposite side
walls should be allowed to strike against each other in a cistern or a wah.
(ii) Aeration.

The uspiulness of· aeration under certain' conditions appears to be generally recognized by
engineers in practice, but there appears to he complete absence 01 theoretical treatment of this
subject. This has been successfully tried in America by spraYIng tht high veiocity jet <lown-
stream of the dams. A tnoroughly aerated water is believed to beha\'e as au emuls.oil III which
stream linps cannot axist. If thi~ be accepted, aeration provides a means of ureaking of the
stream line in high ,peed jets. The consequent destruction of energy would appear to de_!)end on
the internal impact and the work done On the wetted envelope.
(iii) Super-tnbulence.

It is usually the name given to the turbulence in a hydraulic jump ot that in a welJ
designed-cistern. Formation of lot of standing waves dONostream of a drowned falll" a case of
 IS:
...
II>
II>
pr
~
..,~,
::0 "
~

S
0
0-
~
'"
Q,
en
::T
0
~,
::l
0T:l
en
M-
III
U~
at!
,.,'"...,
-
0.
~
I' c
~
fJl

::v
c.h
~.
ciQ'
 '.

l
I
-f'---

1
f .·
 .239
.. "' ~-

super-turbulence though inefficient in its objPci of energy destruction. It cannot be considered

a separate means of energy destruction because it is nothing but amplific'l.tion of the impact.
7. Methods of destruction of surplus energy downstream of falls.
The various forms of im :>act and other methods of destruction of energy described
before, aim at the destruction of the horizontal component of flow of water. Water flowing
especially on an inclined floor has vertical compon"nt of flow which generates dangerous
pl:e,."ialitie<; in ero:i;ng th~ b~d and the sid~; of the earthen channel downstreom of the falls.
Energy of the fall, d11e to this rea,on. is n:>t compl~tely de3trojed in either hydraulic jump or
impact in a cistern. There is al~ays some surplus emrgy which mU5 t be destroyed before water
passes on to the earthen channel section. The following methods have been tried with success
in practice.
(i) Bltne Wall.
A wall built trans"ersely across the line of thw of a strP.am firstly to head up water
above it to such a denth that·a hydraulic jump shall form ann secondly to withstand the
actual impact of a high speed jet of water and to dissipate energy. The height of the wall is
merely a matter of jud~me:lt and so:netimes a seriou; trouhle results from keeping the wall high
by creating a second falL
(ii) Fllation blleks or arr3WS.
They are U:;U1.lly b'~ilt on downstream floor of the falls below the glads or the cistern.
Their object is to divin~ the hottom high speed water laterally~ They are spaced and staggered
REINFORCED CON: "FRICnO~ BLOCK'- so that there is impact between water so
(A) The Block (B) The Reinf. divided and deflected. They just serve to
Scale I" = 1 foot. reduce the' bottom velocity of water
I

SIDE ELEVATION lea:ving the pacca downstream floor of the

fall. Height oithe friction blocks may be
up to t depth. They are spaced 1·5 to ~·O
times the height of the hlock. The deSIgn
is shown in Fig. 6. The distance ~tween
the successsive lines is equal to tWice the
height, Several lines may he staggered
in·relation to one another. The distance
between the successive lines is equal to
twice the height. Several lines may be
staggered in relation to onE another. The
distance up to which r01,lghening of bed
is required by arrows or friction blocks
is given by the following empirical for-
mula given by Montagu.
.·5. D 2312 H 1/2
+, L=C_ _.. _L_
Dl

..._ _-+
..,
I •

\
+.4---...
.......
•• IT

i--.;._j
+
....,
1.
(~) .
where Dl =depth of the cistern .
. D2=depth do,",:nstream•
HL=drop at the fall .
e :::a co-efficient.
Fig 6 . (.

The co-efficient taken is unity for vertical impact. three for horizont<li impact, four to
aix for inclined impact and eight for no impact.
(iii) Dentated Sill.
A sketch of tbe Rehbock's dentated sill is given in Fig. 7 below.
 240

,!,he object to this sill is to deflect up the high

velocIty jet from near the bed. This is built at the end
of pacca floor downstream of a fall.
\i v ) Deflc~tor is built on the ,arne lines as a dentated
sill. It is of uniform height unlike the dentated sill as
shown in Fig. 8. Its obJect is just to deflect up the
h:gh velocity jet near the bed. '
Fig, 7 (v) Staggered blocks.
The staggered blocks are usually rectangular, blocks as shown in

:'~'_$1i Fig. 9 in plan Thy are of t1:e height not more than 1/5th of the
depth They are staggered in pTan. The energy of hottom water is lost
roth in impact against them and in intercolLsion by lateral deflections.
Fig. 119 Ihey alfo deflect up the Ugh velocity jet. Taey are the most efficient form
PLAN of devise usuallY!I,e.d downstream of the ,cistern
~E "" CII/f""l~""iIAl4 or the hydraulIc Jump to destroy thp. surplus
c:::Jr---"T"-....,c:::J--"T"'"-__,L-.J..L.:..;;..::r-~~I:.:.::.:.=:::;) energy which escapes destruction in the cistern
or in the jump.
(vi) Cellular or RIbbed Pitching.
( Internal frktion of water in a channel
is a negtgibl" quantity hut roughening of the
Fig. 120 perimeter can ~ucce~~fully be developed to destroy
• surplus energy downstream of the f~ll, This has. b~p.n t(ed in Cthe form
- ••,c.~:: . ~4.. of C~nular to Ribbed pitching. ThIS type of pltchmg IS constructed by
, •• u •• _ putting alternate bricks on end mstead of on edge as shown
. cit.. .... in Fig. 12.
,i .

(vii) Bur wall.

Fig. 1~

,;F. Colyer evolved the design of a Biff wall in the 'Punjab as

~~own in Fig 13. This is put in at the end of the cistern. The object
._ Js~~<r~ftec::,t back. the water from cistern. to create super-turbulence
(i(,,;,,~fif?11.a~ tun qUlte successful wherever It has been used. '
:,:- • >

8. Contractei Versus Fall-wIdt] fall. Fig. 13

Advantages of falls with full-width of tbe ohannel.
(~) Discharge intensity is low. The energy desiruction is efficient in a cistern or in a
hydraulic jump energy. Thf'! cost of energy de3trllction device~ is very low. Free falls with hill
width can successfully bo constructed without cistern and othf'r energy destroying devices, as has
been done in the c~s~ of falls constructed on the Sarda canal in the United Provinces
~b) The action in the earthen channel section is very much reduced necessitating very/
little elastic protectioIl in the form of pitching _
(c) Such falls are most suitahle for corr~ct gauging of the depth on crest of the felll.
The error due to los~ of head in entry from the earthen channel upstream to pacca channel
between the side wa.ll:; upstream of the meter is negligibly reduced .
. lq) Wide faUs with no complications of contractions and expansions serve as free faU
meters up to drowning ratio as high as 90 per cent when provided with a glaciS.
,. Disadvanhges of full channel width falls.
(a) Haisrd h:gh crest i~ likely to cause silting up in the upstream channel if it runs long
periods of low supplies. The silt when piCked up by water in full supply or When the silt content
is low as in the month! of October a.nd November, will likely cause silt movements in the channel
dOWDst '~'1lD.
 24l

(b) when these falls are combined with bridges, they prove to be relatively more
expensive thm the contl"acted falls. Contracted falls can now safely he con~tTucted using the
modern enel"gy-d,stroying devices described hitherto. fheyare compautivdy ec()nomical. when
combined with bridges. .
Advantages of contracted falls are shted below:-
Cal They serve as ('foportional distributors with low-set crest, causing neither afflux nor
draw down.
(b) They prevent excess silt entering the off-taking distributaries due to dividing the
flow at the sides induced by the coutracted entrance which tends to concen ,rate the bed silt in
midstream. .
(c) 5"fety of floors is increased as £luming increa"es the creep length.
(d) The cost is very much reduced especially when they have to be com!Jined with
Toad bridges.
9 Falls with cistern or glacis.
The selection of the cistern or the glacis downstream of the crast of a fall depends
primarily on the prev'liling levels. Theoretically there is no need of a glacis wha'(l the drop
available is equal to or greater than Ij3rd of the total energy deptb on enst of a fa!!. tut in
actual practice the energy destruction is very imperfect in the case of falls with a cistern when

the drop is about: as shown in Fig 14 below :- ,\ "" ",

~:, alOjai:!

" s.,.'.I..

Fig. 14

The critical jet leaving the weir crl\st simply skims over the cist"rn fnrming standing
"'aves. There is neither nestruction of the energy of the f'lll in impact in the cistern nor a
hydraulic jump is farmed because the critical jet is n,)n-adherent and un,table. The cistern
design should be used when tbe downstream water level is up to or below the crest level.
(b) In thf! case of drowned falls a; shown in Fig. 123 above, the glacis design is very
desirabl~ to drop the critical depth leaving the crest to the hypercritical condition upstream
of a jump.
(c\ Glacis df.s·gn is essential when the drop is not enough and it is de'lired to use the
raIl as a meter. The h~'draulic jump is formed on a glacis with a fall of full channel width up to
a drowning ratio of ~o per cent, and in the case of contracted falls lip to a drowning ratio of 80
~r cent. The provision of the glacis en~ures the hvpercritical flow downstream of the cre~t
which does not allow the buffers of the jump to foul t~e c{)ntrol section on the weir crest.
m. Cistern Design.
Th€'oretic~l treatment to work out the dirLensions of a cisterIl h<l~ escaped notice so fll.
lfut a. lot of empirical formulae propo~ed by the variou'! engineers based on thei~ experience of
stich works are mentioned billow The object of the cistern cou1d be thtee~f61d, firstly to reduce the
intensity of impact of the dropping Jet against the downstream floor, secendly to provide cushion
 242

l
, ,'i\~:'; .

I ".

"
:~.,
\
rig., 15

to d~stroy the energy of the drops, and thirdly to prodllce reverse flow by providing a suitahle
, enrl~wall to ,ensure an imp::Ict in the cistern.
, 'The impact against the fio)r depJnds upon the vertical drop. In Fig. 15 a fall with
'. :.v,erticl\.l droP.is shown ,
Let _E___ be the depth of the cistern
4
The drop without cistern =(I1L-Dc+D)
Drop with cistern =(~L+l'25D-Dd
Discharge =gl/~ Dc3/Z=309 H3(2 cusees per {ootrll.rt
Impact or force against th~ floor is equal to the mJrnentum degtroyed per second
M wav 2
perpendicular to the surface minus the pressure due to the depth = ------'-v-wDa=:~···~.~-wDa
. g g
.",~ ...... ",,~Jo

'\
wv 2 v2
Impact per unit area=~---wD=w(_-D)
g g ~.

Impact wit~ no cislem= w)l 2g1!lL-g D,+ DLp) ~

. ,,' ,
( ,') } .

=w ~ 2lHL·-Dc)+D ~
l J ., .
;I mpllct . h' clstern=w~
WIt
. ( 2g(HL---D
~--.-___ c+1'25D)
~~. I
'250

=w 1 (,
, L g
{2HL-Dc)+1·25D ... '
} .

Thc.re;is increase in imp::lct aganst the floor of the clst6ti by an iittlotifit' eq dal fa
'wD h .
i' '-4 when t e depth of the cistern }is,D/4.

!be cushion is just the wpjght of volume of wa~er displatecl by the jet as shown itt
Fig. 15 mmus the weight of the water dropping per sp.cond. This will naturally increase if W6
increas depth .. If we put a bucket below a tap, water strikes against the bottom but as th ~
eepth of wat~r In th~ bu:ket increases, thp. cu~hion increases andeventually no effpet i~ felt at-
. the bottolll:. 'Th.~ rdal a-:iv<!.lt'l;gP. in a d;:;tern hr the destructil)ll of energy is from the cllshion
. effect and lOthe 1l?Pacqrom1 91e reverse flow while' action ~-ga'inst the floor increaseS.
 243
(bi Dimensions of a cisttrn. 1
(i) Sir Proby Cautley put the depth of the cistern as one half the height of the fall wiF:o
any reference to the discharge intensity.
(ii) Captain Dy~s givec the formula
x=yHL x~D-where x=depth of cistern in feet.
HL=drop in ffet.
D=depth of water downstream in a channel.
(iii) Bombay Pra::tice P. W. D. Hand Book.
x+D=E+EJl3XHLllj ; length of cistern " ~: EY(HL+D+x).,-E

where E=total energy at crest.

(iv) Following is in use "in Bihar and Orissa, India, known as Glass formula as u:>ed by
E.L. Glass Chief Engineer,
x+D=2'25 £l/Z'HU/3.
Length=5(x+D) ,
,,_,,"{v) Etcheverry- .
formul:t as evolved in America.
..... .. .
,' ~ '

Depth=one sixth of its Length

,,' (vi) Montagu's formula, Punjab, India.
Depth of cistern=t Efz
when Efz is the energy of flow downstream. for the discharge intensity q over tb
'Weir and the fall HL from Plate VII.
Length=l=4'Ef2
The formula gives the minimum length used in his faU design, but in the case of fallr
'with: vertical drop, it may be of the order of 6 Eft. In latf'st practice, in large falls the cistens
bed is roughened by two rows of friction blocks one at '4l and the other'at '7l from the beginning
'of the cistern. '
11. Trapezoidal notch faUs.
(a) Notched F~lls used to be generally constructed on canals and distributaries in the
Punjab. They have many advantages. The notches are so designed th;~t, whatever the dischargo
palsing down the channel, the aperture of the notch' is sufficient to pass tha t discharge at tho
water level of the channel at the time. Tha following rules were drawn up by A. G. Reid for
'use. (Punjab Irrigation Branch Paper No.2, 1894.)'
Notations.
l=length of cill of notch in ft.et;
n=2 tan a where a=angle made by the sides of the notch with the vertical;
l+nx=width of notch at any point x above' base;
c=co-efficient of discharge
=0.78 for canal notches and 0.70 for distributary notches;
Q=discharge in cubic feet per second
H=depth of water in feet above cill of notch. and measured to the normal surface a. few
feet upstream of the notch. .
D=deptn of water in feet for discharge D in the reach downstream.
Hd """Height to which notch is drowned in feet.
Hl =height of fall in feet.
ka =-Head i1\ feet due to velocity of approach= v!
. ' . 2g
Note:-In the following formulae Q1 and QI are the sptcial values of Q for which the
i1()tch is calculated. Similarly HI and HI are the special values of Hand DI and D2 are th
specialavllles of D corn sponding to the discharges Ql and Q2 respectively. Als') e1 and e, are
the daphs of drowning of the eill for upstream depths of water d1 and de.
 244
C&nai Noteh.s with Complete Fall.
General equation is :- D=icy 2g (!H 3/2 +.!_n H5/2)
5
=5.35c H3/2(1+0.4 n:fI)
This expression contains two unkown qnantities l aD.d n, and for its solution two values
of H with corresponding values of Q must be assumed.
Then for calculat1ng the dimensions of canal notches, with complete or free fall, the
equations are
QaH 1312-Q H 3/2 0)
n=2.14cH;3/2 Hlj2(~2-HI) tai
> ;, 'fp:m.~ \2)
1 . 'H .,(.,'
TABLE NO.1 ," .11

Table of Oo-emcicnts of Formulre for :Discharges of Notehes with complete fall.

~
K
, L
~.-[~-
K
l
L
-1--

H 6'0562 X '7H1 512 15'1405 X '7H13/2 H 6'0562 X '7fI 1 5/2 15-1405 X '7H 1 3 11

0'5 dty 1-33441 0'26688 3'3 0'01193 0'01574

'6 0' 8459 1 0-2030Z '4 001107 001505
'7 0'57538 0.16111 '5 0'01029 0'01441
·s ,It 0'41207 0'13U19 '6 0'00959 0-01381
.~
]'()
"I
1
. '01:
.
;
0'30697
0'23589
0'18588
0'11051
0'09435
0'08178
'1
'9
·s
0'00896
0' 0083 6
0'00785
0-01326
o 0127-t
0'01225
'2 0'14954 0'07178 4'0 0'00737 001179
-3 0'12242 0'06300 'I 0'00693 0'0113t
h: ~ b,~:..
·4
·s tt
0'10171
0'08560
0'05683
0'05136
'2
'3
0'006:;2
0006 15
0'00581
.. ,4,., . 0'0109"
0010(10
0.0\022
'15 0'07285 0'04727 '4
-7 ,Oj~" 0'06260 0'04256 '5 000549 'jJ'zd '-f 000988
"8 • JJ 0'05426 0'03907 '6 0'00520 rb 5rii" 0'0095 6
'9 .(., 0'04740 0-03600 '7 0'00492 J:I eM" 0' 0092 6
2i) 0'04]70 003336 '8 0. 004 67 0'00897
£~!nl
.J 0'03691 0'03100 -9 0'00444 0'00868
-z 0'03286 0'02S9! 5'0 0'00422
0'00402 :n liJW, 0'00844
0'00820
"3 (H)2940 0'02705 '1
-4 ()1)2'64of " 1'1:, 0'02533 '~rlf '2 0'00382 ..l:' !; 0'00796
-t;
"&
0'02387
0'02164
(l'02387
0-02251
,IS, '3
'4
0'00364
0'00348
".
,.,1 0'00773
0'00752
-7 0'01965 0'02127 '5 0'00313 0'00732
-g. O.OI79S ;!!~rl:J' 0'02014 '6 0'on313 0'00712
'9' 0'01647 0'01910 '7 0'00304 0-00693
3~ 0'01513 0'01816 'S 0'00'l91 0-00677
"l!! "" j'''''.;._"
'1 0'01394' 0'017Z} '9 0'00219 0'00658
'2 f)'1j)1288 001648 6'0 0-00261 0- _0642

Distributary Notebes with Complete Falls.

Take HI any convenient depth not 1es9 than one·third or more than one half of th;J
full supply depth in channel downstrerm of the notch. and let Ql and Qz be the discharges
conesponding to' the depths H t and H2 in the channel downstream of the notcb.
Then the fonowing equations mu<;t be ll<;ed ;-
n = K (Qz-'l.83 Ql) {3 )
l=L (5.66 Qr-Qz) (4 )
where K and L are co··efficients tM [Ju1nerical v~lueg. of which of for dill~d'lt V.ltli6S' of &. <l.C"
given in the table.
 245
TABLE No.2
. Table of Coefficients of fortnulae for Discharges of Notches with Incomplete fall.
M ,M
0
-~~l----

------
l °1
H 6'15 X7H1 5j2 15'38 X '7H1 /3Z H 6;15 x'7H/'/2 15'38 X '7Hif3'
0'5
. , 1'31401 0'26272 '3 001175 O'() 154.9
'6
'7
0'83309
0'56661
019985
0'15860
."
'5
0'01090
001013
0'01482
() 01418
'8 0'39887 0'12983 '6 000944 (NH360
'9 0'30229 0'10879 '7 0'00882 0'01305
1·0 023228 0'09288 '8 000825 ()·el.254
'I 0'18304 (;'08051 '9 0.00773 ()'0120G
'2 014725 0'07066 40 000726 ()'Oll~.J.
'3 0'12055 0'06267 1 000682 ()'01l19
'4
'5
0'1001.6
008430
• 0'05607
9'05056
'2
'3
000643
000601
O'OlO79
001041
's 007173 0'04590 '4 000573 ()'oHlOG
'7 006455 0'04190 '5 0'00541 -o0097:J
·s 005344 0'03846 '6 000512 (j'1}0941
'9 0'04668 0'03547 '7 000485 0.0091%
2'0 0'04106 0'03284 '8 000460 "()'0088:J
'I 0'03635 0'03052 '9 0'00437 "()'OO855
'2 003311 002846 5'0 0'00415 ()'OO83()
'3 0'02895 0'02663 .} 900395 0'00806
'2 ... _.. ()'007-83
'4 002603 0'02504 0'00377
'5 0'02351 002350 '3 0'00359 0' 00761
'6 0'02131 0'02216 '4 () 00343 o '0074 <l
'7 0'01939 0'02094 '5 0'00327 () 01)7:N
'8 001771 0'01982 '6 0'00313 ()'0070l
.~
'9 0'01622 0'01881 '7 0'00299 ()'00682
3'0 o ~1490 001788 '8 0'00287 - ---,' 0'00665
'1 0'01373 001702 '9 0'00275 ,000648
'2 0'01269 0'01626 6'0 0'00263 0-00632,

Distributary Notches with Incomplete Falls.

, The following simplified formulae for incomplete falls on distributaries are app licable
only when the following conditions are obtained.
(a) The descent of the fall must not be less than one- third or more than one~balf of the
full supply depth of the up~tream reach.
(b) The depth for a givt'n discharge in the downstream reach must not be less than the
depth for corresponding discharge in upstream reach.
Failing either or both these eonditions, the formulae for canal notches with incompleto
falls must be used.
Let Ql be the discharge due to depth in the downstream channel equal to the des::an\
()f the fall and let Q2 be the discharge due to depth of twice the descent of th~ fall •. (The not-eb
will thus be just free with depth=HL in the downstream reach and drowned to a depth = Hl.
when the depth of water down-stream of it is 2HL .)
n=M (Q2-2.5Ql) (5)
1=O\5,37-l QI-Q2)
in which M and
given in table 2.
° (6)
are co-efficienb, tbe numerical values of which fOf different values of H are
 246

(c) Fig. 16 shows th", form of notch uSed in the Punjab. It will be seen that the plane
of the profile of the notch is set back 1'5' from the downstream face of the notch wall. The
profile itself is formed by a horizontal eill and by twr') equally inclined straight lines which arp.
the traces of the crowns of two inclined surfaces of varying curvature The limiting radii of
curvature are shown in the sketch. Ali arcs of C'Clrvat, re are drcular, and all centre~ lie in the
plane of the profile. The same form maj be used for large distributalY notches, but in their
case the plane of the profile should be set back 0·75' ollly from the face wall. and the unit of
radius may by 0'33' in place of O.S! as shown in the sketch. Cn aU ca1!es the upstream splay
should be 45° and the dowDsteam splay 22°
Thp. height of the notch?s on distributaries should be in every case that of the estimated
depth of full supply. and on canals it sh(mld, as a general rule, be aho of that depth. Thrre
will, however, arise cases pn canals in which at times the volume to 1-e pa,sed i~ considlp
greater than that dne to the full supply depth .
.ElEVATioN
which has been as';umpd for the purpose of
caleulating the profile of the notch. and in such
ca~es the notch ShOUld be carried up to the
. surface level due to the maximum depth
of water which will be carried by the upstream
reach. .
The thickness of the notch wall must be
sufficient to withstand the pressure to which.
it is subjected. Where two notches are
separated by an isolateil wall, it may be
laid down as a rule that hoth faces of·
<

the wall being vertical, the thickness should

not be less than half th~ height on canals,
PU.N and 0'4 hmes the height on distributaries.
The lip of the notch must be
corbelled out beyond the face of the crest
wall of the fall The edge of the lip ~hould'
be an arc of the circle of radius such' that it
passes through the following thrpe points :-
(a) On canal falls-At the end of the lip
the two points in which the downstream
splay of the notch meets the fare of the
nothch wall; and in the middle, a point two
feet outside the face of the crest wan.
',tit " (h) On distributary falls-At the end ef
jb.&~n i the lip the t.o points in which the
·.1 [~dj ", downstream splay of the notch meets the
surface of the n 1tch wall; and in the middle,
a point one foot outside the face of the crest wall.
(c) EXlmples--Reid gives in his paper the following examples ot thf! method of using
these f()irrl\lla~_
Frae FlIlo, Cm\I-Ctl·~uht~ thl} dimension1 oi thl notche; for an 8-0 ft fall on a
canal under the following conditions :-
Upstream Reach
Ped wirlth 100 feet
Inclination I in 66666
Full supply discharge 1869 Cusecs
Three rajbahas, with an aggr"gate full supply of 42~ CU>6CS to take off trom this reach.
Downstream .Rc-ach.
Bed width 87 {fet
Inclination ". I in 5,000
Full supply (norm:! 1) 1,445 cuscct
Minimum supply oYer the faU-770 cusecs.
 247

Assuming side slopes at t to 1 the dp.pths of full supply in the upstream and downstream
re;lches will be found by Hi~ham's table to be 7'0 and 6'0 feet respectively.
We have first to consider the depth over cills of notches (Hm) at which the normal full
supply of the downstream reach (1 ,445 ~usecs) should be passed.
The depth at whicb a supply of 1,445 cusecs wOClld be passed in the upstr(}am reach is
found, by Higham's tables, to ce 6'0 feet, or exactly the same as in the downstream reach:
The limiting value of Hm will therefore be :-
(il Depth of full supply at head of up stream reach=6'O feet.
(ii) Depth in upstrt'am reach due to a supply of 1,445 causes =6'0 ft.
It we take Hm = 7'0 feet, the working of the upper reach will be perfect as long as all
the rnjbahas are running full supplies (424 cuse~s\, but if one or more of the rajbahas are closed,
the notches will be very tight, and will head up the snpply. If, on the other hand, we make
Fm =60 feet or lO foot less than at the head of the reach, there will be a great draw under
normal conditions, i.e. with all rajbahas open, which might cause a serious bed scour, and would
lower by one foot the head availa'1)e for the rnjbahas immediately above the fall.
The value to~e ascoumed for Hm will lie l::etween these limits. Generally it may be
said that the nearer a falLis to the head of the canal or branch, the nearer should the value
assumed approach the maximum limit. In the present caSe we may take an exact mean and
assume Hm=6'5'
The depth due to a supply of 1,445 CUSECS in downstream reach (Dm) has been shown
above to be 6'0 ft.and ratio of full supply depth above cills of notches to full supply depth
. .H 6'5
III downstream reach -~= -
Dm 6'0
The minimum working supply is 770'cusecs, corresponding to a depth (Do) of 4'15 feet in
downstream reach, as will be found Ly the tables.
The supplies Q1 and Q2 to be taken in calculating the notch will be those due to depths
D1 and D2 in downstream reach, and these depths may be determined thus :- -
6°0-4'15
D1 =DO+HD m -Do)=4'15+ 4 =4.'61

_ 3 • 3 (6'0-4'1
D2-Do+4\Dm-Do)=415+ 'l
The value of QI and Qz correc;ponding to these depths in downstream reach ~a}' be
calculated (from Higham's Tables) thus:- ,
Red width 87 feet Inclination = 1 in 5,000
Depth D1=4'61 Q1=924 cusecs
D-D 2 -5'54 Q2= 1264 cusecs
The corresponding depths over the notch cills HI and H2 will be
H t =D1Urn =4'61 xAoS=5'0 feet; H 2 =D 2!f_lll=5'54 X 6'S =6'0. ieet
Om 60 Dm 60
We have therefore to calculate the dimensiom of a notch which will pa.ss 924 and 1284
cusecs, at depths over the cill 5'0 ft and 6'0 ft. respectively.
H 1=S'0; Q1=924; c=0'7:1; H 2 =6'0; Q2=1264.
Q2 H 13,2_Qt. H 23/2 _ (1264xl1'181-(924x14'7')
n 2' L4(;h~.3/2tr~3{2lH2--=--H~) -2 14 X 0'78 X 14'7XfF18 X (6:0::"'5'0)
0

= 548'72 -2'0 feet

27447
'~I
'L___Ql_ -04nH = 924 -O'4x2xS'0
5';J,jcH13,'2 1 5.a5xO'78xlloU;
= 19.81-4.0=15.81.
 248

The above valu~s of 11 and l are for a single \. otch, if we design the fall with, say SIX
notcl.es the vulues for each notch will be
n=,~_:O =1; l= 15'61 =2'63.
6 6
The width of the notch 1'0 foot above full supply or 7'5 above cill, will therefore 1:e ! of
L ..' I 7'5 +-2'63=5'13 feet and the profile of the notch will be as
! ' - s ..I._': in Fig. 17.
"""._.... It will be found by applying the gen~r::ll equation Q=S'35 c.
H3 12 X (l+0'4 nH) that the full supply of 1,869 cubic fpet of the
upper reach. which may be anticipated when all uppl·r raj bah as
are closed, will be passed with a depth over the notrh d=7'5
feet so that the water will bl> just level with the top d the
notch, as shown in the sketch, and there will under these extre-
me conditions be a heading up of 0'5 feet. If it is desirable
to avoid this, or to leave a gnater margin, the height of the
notch walls might be reduced to that required for the normal
full supply passing i)ver them. namely 6'S feet, and exctssive
supplies could pass over the whole length of the wall, hut there
are few circumstances in which this could he recommended.
(d) Calibration of Notch falls.
Fig. 17 Nethers-ol's Tables.
The late sir Michael Nethersole pre Dared a series of tables to facilitate the calculations of
the correct discharges of notches for distributry falls. Tables give the dischages through the two
side triangular portions of the notch as distinct from the discharge of th .. rectangular portion in the
centre. These tables were published by him in 1903 at the Thomason Press, Roorkee, India. The
velocity of approach cannot be correctly allowed for in the calculations. SometimQs the value of
the coefficient is increased to allow for this. The calibration is nothing but approximate.
(e) Trapezoidal notch fall is still the best type of fall in channel, in which variation
of supply is very much and no metering and regUlation of supply is required. However, this
type shonld be avoided when it is not a free tall.
12. Montagu Type fall.
The design of the glaCis prnfile does not seem to have received the attention it deserves.
The ... hole theory of the formation of a hydraulic jump postulates an horizontal velocity
and it is during the change of this horizontal velocity from the hypercritical to the subcritical
stage that the dest rurtion of t nergy takes place.
2. The fact that all water m lYing down a glacis haq a vertical component of velocity
seems tel have been lost sight of. in evp.ry pUblication on the ~ubject.
It appears that the high speed jet which so often persIsts below a Hydraulic jump is
the outcome of this vertical component of velocity, the energy of which is unaffected by the
occurrence of the hyilranlic jump.
3. It will be cle::lr that a reverse slope on a glacis such that the hyper-critical stream
is moving horizontally. is the best solution (Fig. 18)
The drawback is that the Hyrlraulic jump is extremely sensitive on an horizontal floor, and that
the level 0: ~uch a floor is only correct for one
--_L
--
I
ideal set of conditions Any departure from the
conditions postulated, leads to trouble which may
clttSr be serious.
But for a varying discharge, the profile of
the glacis should be so designed that the
maximum horizontal acceleration is imparted to
the stream at all states of discharge. By imparting
Fig. 18 maximum horizontal acceleration, the maximum
horizontal velocity is attained in any given length
of work; such a design is thert fore conductive to economy in construction.
 249

.4. It will te clear that a stream flowing on an horizontal bpd has no horizontal
acceleratlOn because the reactien is norm~l to the bed.
. .. I~ will also b~' cle;r that a Slr~am flowing- over a parabolic hed appropriate to the
1m hal hOrIzontal velocity \ , will have no horizontal acc" lpration because there is no rea~tion.
So"'ewhere between thpse two there will be a path upon which the horizontal component
of the reaction imparts a maximum horizontal aceeleration. It is this
path which is the most efficient and ecnomical glacis profile. (Fig. 19).
The equation of the ideal glacis profile derived by A.M R. Montagu
is stated bplow and the student should refer to the original publication
No. 10 Central Board of Irrigation bv Mo.otagu for its proof.

;/4
x=vv·_-- . (v +y where
. g y
v=the initial velocity of water leaving the creest.
X= Horizontal distance along the ordinate. .m'&91iaawob !,'
y= vertical distance below the horizontal. ,.abl.e tb~Qc'6'm:l:lf!'
:It I p\ 1 e'"

Example. Design and sketeh a fall of M mtagu type for the foll~~= lUi
U'lt

Discharge=207'S cnsecs ; Bed width=3f-0 feet, F'S'D'=3'6 ft. .' .J

783'94. 78034
F'S'L'= '--'---, Bed level= .~--' Fall=20; N'S'=784'72
7~ 1'84 778'34 '
Calcu :ations.
2.0 fall at R.D. 2.000 Kokri Distributary Sirhind CaOlll
Data. Q=207'6 cusecs; B=31'O feet; D=3'6 feet.
207'6
Mean velocitY=(3-f+--:-326136 -=1.761 ft. per second. 'f''';
~j,:

~ 783'94. 780'34 . .. '

F .S.L.= 781'84' B.L·=77tF3T' Fall 2 0 feet> N,S',=784 2. 7
. v2 1'76 2
(1) Flume:-h a =-2g =-64-:~f =·OSft.
8" S. D Downstream=3'6 feet; Ef.,=3.6+·OS=3·65 fpet.
From plate \II for HL =~·O and-Ef 2 =4'65, read q=U'6Cs
2076
B; ='--~6-
=17.89 Say 18.0 feet.
11. "
(ii) Crest.
Q =207.6; De = 16ft (critical depth); B = 18.0 and q = 11'4 es.
Efl = 1.6 X 3/2=24 feet,
R.L. of crest=783'99-2'4=781'59 Length of crest=2'S
H=..l·Sx2·4=6·0 feet. i
(iii) Approach.
Side expansion each sine upstream=(31-18)·i=6'S feet.
Expansion 2 to 1 length=2 x6 5~~ l~ 0 feet.
~. Cistern.
Dppth~~*Ef2=3. 6 5= l'S3 feet: R.t. of cistern bed=t778'34 -1'8~j=776'Sl
• 2
Length of cistern =4 xdepth=-t X 1'83=7'3 ft.
S Glacis.
X=V (4 /y- +y (I)
ygy
 250

q = 207'57 = 11'53 cusecs; Dc = l'6ft,; q 11'53 7' 2ft,/3ec.

v= -=----=-
IS Dc. 1'6 ,)' , .'
Substituting value of v in formula (I) above

x=7'2x'3524xy'y + y =2'54vy +y
RL of crest =781.59 and RL. of cistern floor=776'Sl
. Ditlerence=S'08 feet.

y y'y 254y'y x
" 1 1 1'0 2'54 ~'54
;1q 2t 1'4142 3'59 8'59
3 1 '7321 4'4 7'4
4 2'0 5'OS 9'08
5'08 2'252 5'72 10'S
Horizontal length of glacis is IO'S feet
Departure downstream.
.. 7.
Departure on each side =i (31-18)=S'Sft,; Splay 3 to 1 ;
Length =6'5 X 3= 19'5ft.
8, Exit Gradient.
For minor works founded on clay soil, the exit gradient m~y be 1 in 3 to 1 in 5
Exit Gradient GE =_I:!_ I vide Plate III ; where,
d 7T'y'~
d=depth of cistern wall at the end of pacca work =2'15 feet. "
b 56
a=--=--=26'5
d 2'15 ,',,'
Head=H=worst when water level is up to crest and downstream bed
dry='jSl'59-778'34=3'25 feet

1
-
='086; From plate III, GE == 3'25
:us
X 'OS6='13 which is safe'
/
i't
"'YA
Upstream pressure.
(1) Upstream curtain wall. Corrections of thickness; total drop =15%
.J.. _15X30_ ,- 01
d= 1'4 feet, b==56'O' 't'c~ -~.- - 4 " 10
a='0225 lUU
. , 4'5x'9
,pc =85 pc, from Plate II ProportlOnal to thIckness= _ - 2'S6
1'4
o~
II

In terference of Glacis toe wall

¢c corrected=85+2 86+ '6==S8'4G 0/19 !~'32=19 /-l'33=~~~ ::::,6°/ .

/(J y' 29'8 V 29 8 56 /0

(2] Glacis toe wall, d= 1'4 ; b,..:56 0% Correction for slope of Glacis;
b 29'3 56,
_j._ ---' =.53;a==- =40, For slope in 2 read from Fig. 68; .
b 56 1'4

;C=48 %; 9D=48 % ; .LE=48 == ]0'8 X6'3=2'3Q~

't' :l.9'1 I"

Corrected 9~=4S%
(3) Coerected ¢E =.pD=48+2·3=50·3%
Downstream curtain wall. Corrections
0=2'15 feet Thickness of floor ; drop = 1%8
~=o
b
30 =5'4%
rpE=18,-< __
100

!_= 2.15 ='038-\ Proportional to thickness

a 56 .

Corrected ¢E=18-2'27-'37=lS:360/0
Thlckness of fioor.
Soil Mlow the floor is sandy .mixed with kal1kar and'theretore pervious. Worst
H=3 25 feet. '1
Table 3.

Full seepage :'ei: . . .jJ:l· Thickness Thickn'ss

Description. Percentage. :1 flow , ,irhn of floor provided
pressure calculated
. , iT 1 j'
:
·'::~.:fi:;i~:i
Unner cistern 50'3% 1'64 1'64 1'9
10'25 from toe Ii J 1'4 ;"'<VIili 1'4 14
of glacls tc;·!i
19'8 do '9 '9 1'15

Th;ckness of the glacis.

The thickness of the downstream ghcis has been designerl for the worst conditions of the
trough of the hydraulic jump and it oc;urs when the supply in the distributary is minimum,
Area of trough at the jump=8~4=16 sft, per foot width.
Area of pacca floor under trough of the jump=S'S X average thickness=*S·5 X l'f)== 16'15
sft pe) foot width. Hence thickness is sufficient,
Curtain walls.
Upstream curtain waIl=1/3 F.S.D,=1/3x3·6""'1·2~ft. thickness=l'4 ft.
Downstream curtain wall.

t of downstream depth= ¥=L8 it,

':.tt -I • __ ', ":'_:_,1

thickness provided at the axis of the channel==2'15 feet as the bed is bowed, as per Centra)
Board of Irrigation Publication No : 6 Fluming Page lO~
~ ~ ,1 . ~ !>

Bowing of Floor downstream of cistern.

It is raised up 1 in 10 to RL, 771,57 petmitting a bowingt>f_!_to .~belowthe designed,

30 40

bed. End curtain walt is also bowed, Let. it bt.1!..in

(:$(J
this Case=l,O ft.
 252
Downstream bed protection.
2.ft thick loose bats far a length of 4 to
5 D 3'6 X 4= 14 4 say IS feet bed and 10 fflet
on sides in bed.
Roughening of the cistern bed.Cistern bed is
roughened by two rows of friction blocks.
Side walls. Side walls are flared from the
vertical at the end of the crest to a slope 1
in 1 at the end of the Cistern Fig. 20.
LONGITlJDINAL SECTION.
13. Inglis type Fan.
SCALE {HOR: The inventor Jf this type of contracted fall
~£R:
iC.C. Inglis M.I C.E. Director Research
Station Bombay) described the essential
features of his design in paper No: 44
Technical Publication P.\V.D. Bombay and
paper Mo: 170 punj ab En~ ineering
Congress 1933. A lucid description of tne
nece~sary points is gi ven in Appen iix V of
Fig. 20 I the Central Board of Irrigation Publica'ion
No : IO.A sununary is presented here.
(a) Notatj( ns
Q=di ;charge in eft. per second [cusecsJ
q = }ischarge per foot width of a channel or flume.
qo = Discharge per foot run of overall throat width of flume i.e., including piers.
\' =velocity in ft, per second at any selected point.
d1=Depth of water in upstream sertion of achann~l [e g., ahove a fall or contraction~.
d 2 = Depth of water b'l.ffle pavement at toe of fall [assuming no standing wave has formed
d 3 = Depth of water in the downstream section of a channel or flume [e. g .• below
standing wave.]
dx=Depth of water downstream of standing wave above baffle pavement level in a parallel
sided channel,
dy=Depth of water downstream of standing wave above baffle pavement level in
expanding flume. j,;j.>'(d
d z = Water depth in cistern. downstream of the baffle pavement
D1 = Depth of water upstream, above sill level. ,;}I i:

D=Effective .depth of water a00ve sill upstrGam =D1+ch x

hv = Head causing velocity
H= Fall in water level i e. difference of water levels of upstream and downstream of fall
(below standing wave)
Hx= Fall in water levels with parallel sides.
Hy=Fall in water levels with expanriing side".
h2 =Difference of level of water upstream of and on baffle pa.vement at the toe of
the fall. assuming no stancling wave i.e. the effective head for downstream velocity v 2
hb = Height of baffle
B=Bed width of a channel B 2 =Bed width of flume at the contracted section
rexcluding pien).
Bo =Overall teil width of flume at the contracted section including piers. ,
Lh =Dist:mce of the upstream face of baffle from toe of falL
(b) Description of the Design. ' 'I
A section and a plan of the fall is given in Fig 21
Des'gn consists of ;
(i) A standard long -throated weir flume followed by a glacis slope and a pavement on
which a baffle is fixed to dissipate eneTgy A cistern downstream of the baffle with a deflector
at the downstream pnd of the cistern is provi~ed.
pacca pavement ends at the deflector There is a se{;ond cistern downstream of the deflector
the section of" hich need only be pitched,
 253
(ii) At the toe of th.~ glacis,

-.-.,t-~ very high velocities aTP generated,

and the energy in excess of th.<lt
required for normal flow has to be
dissipated. Under such conditions
energy is usually dissipated l,y
means of a hydraulic jump which
forms at the point where there
is a balance between the total
j_
energy due to pressu' e pIns the
..r- rate of change of momentum of
T 'Inn tap. high velocity flow and that of
..... I ,... ..
~
the low velocity flow in the
do'Vnst;eam channel
. II
~ () , ;:
~
:ce ..:: (iii) Though a hvdrau1'c jnmp
I
I
,.~III
r- -t
oo - "

r-
rlissipates a consiilerable amount
of excess energy in the form of
I "

f ..
"
I ...'"~ lIo.
~
r-
-c:
t)
- -
,.."b
heat, yet residual eddies and
turbulen<>e persist :lnd
rlistribution of velocities down-
the

..
Jo.
I ,..i <,.. ~
stream of a fall does not
I M approximate to the distribution
I ....'" n....-t ::}
in a normal chan nel. It follows
I a that when the materials of the
<: nr channel bed are erodable, heavy
I \ £"
I scour results, and this is
a('centuateil where the sides
diverge sharply below a fall,
because cork screw eddies are
then generated. Where the height
of fall exceeds l/:;rd the depth
of the channel, conditions are
UJ:
markedly improved by adding a
properly designed baffle.
(c) The maximum dissipa.tinn
'Il1 of energ~' by a hydraulic jump
,~!lh occurs when the jump forms at
;{! the toe of a glacis. In practice,
if there is a sloping glacis, the
hydraulic jump may form:
[i] on the glacis, [iiJ at the
toe, or [iii] downstream of thp.
toe. ~
Fig. 21
(i) If the hydraulic jump vouId form on the glacis, where there is no baffle; i.e. if
da-the normal depth of water in the channel downstream would b~ greater than dx (or d y)-
the optimum depth, the baffle should be fixed on a platform at a higher level than the canal bed
level, so that the natural wave would then form at the toe on the platform.
(ii) If th .. hydraulic jump would form at the toe of the fall i.e., when da=dx (or dy], the
baffie should be fixed at pavement level.
[iii] If the hydraulic jump would form downstream of the toe, i.e., if d 3 would be less
than d x or [dyJthe glacis should be extended and a ci,tern provided of such depth as to bring
the wave to the toe and the baffle fixed on the bed of the cistern.
Plate XV enables us to detem1ine the level of the platform on which the baffle should
be fixed for an overall discharge of qo per foot run bet Neen abutments for a fall of Hx feet
when the channel has parallel sides and when the sides diverge" The height of fall Hy ca.n be
turned into equivalent height for parallel sides Hx -vide Plate XV and kilO wing Hx we can fine
dx from Plate :XV and as Hy+dy=Hx+d x , we thus get d y , the depth of the baffle pavement
belOW the water level downstream of the fall.
 254
The height of the baffle should be equal to l'3d 2 ; d 2 being giwn for any qo a.nd Hx on
graph of Plate XV and the distance of the baffle from the toe of the fall=S'25 d 2•
It is important and interesting to note that as the position of the
standing w:;).ve depends only on the discharge per foot run and the head available.
it is immaterial whether there are piers or gates in the throat. provided the piers do not
cause afflux and the coefficient of gates is unity; but if the coefficient is les. than udty.
then the head available is the difference of head upstream and d)wnstream (Hg ) less the energy
head destroyed by the gates or piers; which can be calculated.
The baffle ceases to be effective when Da/Dl exceeds about 0.60. because then it
begins to become ' drowned" .
The graphs shown in Phte XV allow for nornmal (IO%i losses due to friction and the
graphs in plate XV allow for the part of the prc:snre of the standing wave which is balanced by
the side walls.
(d) Even though the energy is effectively dissipated by baffle near the toe of a fall, the
distribution of velocities is not normal, hence a chtern and deHector are provided. At the point
wh~re ~he downstreem flume width is equal to 3/! bed w;dth of the channel, thl'! nepth of wate
WhICh IS necessary to give the mean velocity of v=cXO'8! dO 64 is calculated. The difference
?ctw2en this depth and the norm~l depth do vnstrcam giVdS the amount by which the pavement
IS su:-!k below the downstream bed leVel. Plate XV show;> the value.;; of L in terms of Bo for'
various values of BalBo and divergence of 1 in 4 to 1 in 10. A deflector is fixed at the end of
the chtern i.e. at a distance L from where diverge'lce ~tarts, the height of which is 1112 of
[F.S.D.+depth of the cistern below the diagonal canal bed Level downstream.]
[e] Side divergenee.
The side divergences should not be sharper than those shown in Plate XV otherwise a
return flow occurs. A cistern with a semi·elliptical cross section gives a much better distributi n
of velocities than a cistern with a flat bed. The side slope deflectors are bel1eficial by caus.ng
the bed roller to form over the fclll width. thus preventillg the formation of lock screw eddies •
[fJ Seour downstream of bed ddlector.
Experiments have shown that where a deflector Fig 22 was fixed, the material tended to
bank up behind the deflector to a slope of about 1 in 3 to I in 4 and for the deflector to be
fully effective this scour should be allowed to occur because flat pitching prevents the form'ltion
of the beneficial borjzontal bed roller. Protection should be laid to the natural slope of 1 in J
[gJ Examples
Design an Inglis Type Bumed fall in
a channel with the following data:-
Discharge=700 C1lsecs; Drop=6'63 It; Bed
width = 60 ft.
Depth 5'1 feet; Fun supply level= upstream
downstream

t 4R-~,,'
·;':·~':-·':/:;~I'i~,;;·;~;,!;· ,~ .........., . _.... ::. ".0 ~ .1.\
_. ,.::'-.: ; . : : :....

secono:
622·S9.
Bed levels= --_,
615::16
Mean velocity=2"19 it per

Fig. 22 N.S.=624·O; Side slope t to 1.

(i) Design of flume:-

Area of the channel section-S·l(60+--1, =319 sit. 5-'

~
 255

fL{from V= l'15v' fLR=O'81

Height of the hurrp in throat=_<!3~=_5'L='64 ft.
8 l:l
Dl=7/S d a=7/8 xS'I=4'46 feet.
d)=height of the hump+D 1 = 64+4'46=5'1 ft.
v l =veloclty of approach=2'19 feet per second.

hv=head due to velocity. of approach = ~0~ .086 ft.

},'ote:-Instead of 2g, c.e, Inglis adopted 50 to calculate hv.
D=Dd-hv=446 (-'09 =4'55 it.
Q=3'09 BDl,0=700 :.B=23 4 ft. say 24'0 ft.
lt is desirable in l'ractic p to design the throat in an exact multiple of feet s() that the
;pan width is about 1'5 D 1 ; let there be 3 S?arlS with two piers of :2. 5 ft. thick each, the di"dn;g,)
co-efficient reduces in that case by l.S P c. .
. '
.. The new D (.
700 2/3
_. ) =4.52 feet.
24x3.0-l
This will give Dl =4.43 and the height of the jump='67 •
[iiJ ~en mouth ups tream approach,
Radius=2Dl-5~, 2 X 4'52 1 '°= 19'2 feet,
The curve shou'd su;)tend 'an angle of 600 and then continue tangentially till the wall
penetrated the side slopes above F.S.L.
. The sill of thrO;:Jt may be formed to upstream bed level by a sloping pavement.
(iii 1 p05iti:m of ga'lge cha ill cer,
The posltion of the upstream gauge chamber from the end of the thi'oat=4D15=3tVi
feet say 38'0 feet.
iv. Throat Length=2'SD=Z'SxS'l=1I'3 feet.
Throat width=3x8+2x2'5=29 feet.
qo =--~OQ=2!'14 cusecs, The glacis slope will be 2'5 to 1 and will be joined to the thro'lt .
2~

and the baffle platform by radius 2D",,9 04 feeL

Side divergence B3 = _~Q_=1l'76 from plate XV for-:§___=ll'16, the side diwrgenc6

. da 5'1, d " ,

;, should be 1 in 9'2. In the design, it may be adopted as 1 in 8.

(vi) B::l!fte anI Baffle platform
Hy ='0 rs ttP.:l u wcl.ter-level downstrean1=ctest ievel+d.,l-, 6.21·oa=o~:a2·59 r511-:--3~1 '0(1
'="663 ft. Fr,):n pIa!',· ;{ V fur q 1 =24'14 and Hy=6'63
01' from 'qU'lti'>n <1 2 =' J I q) Ih:l 5 (.-\)
and x= 7:3 go .5! : Ix ,',1 (B)
fL ()II
an d sram·,' • =2'07= ,'CK'U2=l [2
Btl 'N ' .
 •

dx form equation (B) above=5'S6 feet.

h6=height of the hottle = 1'2 (Dc-D 2 )-1'96 feet.
/ -----;;;- / 24 '14 2
D ={I-_={I--=2 63 ft.
o g g
The baffle should be 1'96 ft. high in i length and should have its top sloping down at
the ends. The dbtance of the baffle from the toe of the fall=5h b =5=1'S6=9'8 feet ;,Hy+dy=
Hx+dx; dy=4 87.
R.L. of the baffle platform=downstream water level--dY
=621'06-4'87=616'19
(vii) Cistern downstream of baffle.
Minimnm length of the cistern is given by L=6'3 Q.3=45 feet.
The bed width at the end of the cistern expanding in 8=48'78 ~..et q per .loot widt\}
at the end of the cistern = 14'36 cusecs.
Depth of the cistern is given v=-Q-='84 cd 2 'u
q2
of fL ='81 and then C='9 :. d. =6'02 feet.
Actual length of the cistern is usually kept rather liberal. Let it be 76 feet in this case,
41 fee,t with pacca sides and bed and 35 feet downstream of the deflector with pitched bed.
The cross-section of the bed of the cistern is bowed. The amount of bowing in
;nidstream equal to 25% R wnere R is the hydraulic mean depth. The actual bo\\ing of cistern
ted can best be done by letting a string held across the section touch the three points, two ends
and the middle. The curve obtained is catenary
The depth of watel in midstream at the end of cistern=621·06-614.82=6 24 feet.
The bed deflector will be=6.24 =.52 feet say 6' high. Side deflector =~264 =1'04 feet
. 12
say 1.0 feet.
(iii) Vanes.
It is usual to provide vanes in the continuation of piers to stabilize the hypercritical flow
The vanes may be 6' R.C. walls with 6" free board,
Sometimes in addition to the deflectors, staggered rows of nctilinear blocks say 4 in
nnmber are provided to stabilize flow downstream of the pavement and to reduce scour. In this
case the second cistern can be omitted and only one cistern of about 41ft. length at the worked
out level will do.
h This type of fall is considered to be v'ry efficient, for distribution of energy
hut the cistern element is unnecessarih' elaborate and very expensive. fhe position of the
gauging site is defectivf because it will be opposite an earthen bed, which is liable to silt in lows
upplies. In this the correction due to velocity of approach cannot be correctly applied as will be
!>xplained later.

e4. Meter flume.

The correct design of meter flume received thp. first impetus fn m thp. experiments
carried out hy E.S.Crump I.S.E.punjab Irrigation, described in papTr Nos: 26 and 30-A of Punjab
Irrigation Branch PUblications, Class A 1923.25 Th,,"e experiments brought out three points
very clearly·
\i) The discharge formula Q=CB~H3f2 was applicable to the long- cCfst(d weirs
and that the value of the discharge coefficient C was 3'O~ whidl is its theortical value
(ii) The length of the crest sh"'uld be 2H.
(iii) The value of the coefficient was constant and indepenced of H in a long-throated
weir flume upto a drowning ratio of 91% (model L of (tumpe contracterl flume with?
length of crest 2H and Expansion 1 in 10 downstream on both sides}. In ca,e of flume
with vertical drop it was constaut upto 66% drowning ratio
 257
(b) Discharge formula.
Let q be the discharfe per foot run: D be
the depth upstream.

Initial Enugy E=D+~~

2g
Depth on crest=Initial energy -Rise of crest
H=E-x
Let y be the depth of water on the crest and
the motion in a frictionless channel. Fig. 22
The discharge rer unit width of the wier q=A.v

wit h respect to y

for the maximum value of q for the values of y; ~~ =0

1
(H-y)1/2_IrY-H--iT2 =O, or H-y=ty:. H=3J2y or Y=iH
\. -Vi
This is ju~t the necessary condition which gives the maximum discharge for the kn·)wn
value of H. Therefore discharge.

q=iH.V2g~=iv/l. H3/2 A

J:r:"'l' llf3
=3'088 H3 1 2=3'09I{3'3

we obt::~ :itil~~~ep: jfm~r 8

27 H3)lJ3=iHor H==iYc
{I gi =(
< g
. .
SubStitutillg Yc in terms of H in equation [A]

q= v/~g~'H~-Yc J3/3=gll2 yc 3'2= v~ (B)

It is therefore that the critical depth is the de)th of water ff·quired at the crest to give
the maximum discharge or the coeficient 3'09, provitied the flow is frictionless in a rectangular
channel. It has been shown that the water surhce profile in an open channd, is parallel to the
b~d profile in the condition of critical ffow. This is the section of a weir, where parallpl:sm of flow
is attained. It is the section which gives the maximum discharge so long as the parallelism is not
fouled by the conditions of flow upstream and downstream. Any ilisturbance of this control
section results in the reducti >n of d15charge of weir. It is to attain this condition of parallel flow
that the crest is kept levd
(c) Length of crest
Advant~grs of long length of crfst.
(i) A long length of cr{ljt a~ usuall) ad opted in weir flumes ensures a constant cQO)fficient
of discharge with all depths on crt}~t.
(ii) A lcmg throat provides a considerable increase in the modular limits. Free fall formula
is applicable to as high a drtlwning ratio as 90%.
(iii) In long-throated' weirs. the ce~sation of modularity and the disappearanc~ of the
hydraulic jump occur simultaneously, so that tll'~Y o:1er this !!reat advan tage over their short-
necked brothers that one can tell at a slance whether a flllm~ or a weir is dE-livering Its full
m<)dular discharge or nl)t,
. The water surface profile shown in Fig. 23 i, convex in the beginning. then attains
 pnalbl!sm in the critical C);} iitialS of flON (uSHlly c1Hel flu point of Inflexhnl and lastly
drops belo", the critical con iitions in a c!)n~aV~ curve. fhe convexity connotes a CUflrature of
fila'l1'~nt.; wi}ich is <tttendd by an hcrea;;e in til:J v~Iflcity of successiv~ Wam ;nts from thfl surface
to the be i. This will m~a'1 that t~e act 1l3.l presmr~ ag1tn'lt the b~d will be less than tue actual
depth in a convex flow. Su~h a flow is ea'lily Iia')Ie to pressure inflations by a high pressure
water available below it as in a i't TIp. The crest, ShO!lld, therefore, be hng e.:Jough so that the
mnvex flow flies out on the crest fJl[oNed by the critical conditions of fhw wh"re· parallelism
is attained. The oarallelis:n c,):1notes that the velcity is the same from the surface to the ted and
that the pressure against the bed will be equal to the actual de.Jth· For constancy of the dischar~
ge even this is not enough but the level creit must be longer stilt to ensure hypercritical flow for
suficient l'>ngth downstream of the critical section, so that the buffer of the hydraul c j Imp in
the case of meters with a gla~is and the convexity of the dropping jet in the case of falls with
a vertical drop do not foul the critiC'll section. Hypercritical jet is adherent and not liable to
prpssure inflations. It has been ohserved that the control section or the point of inflpxion occurs
in the IT id<ile of the crest when tht length of crest is 2H. There is thus a level
lr:ngth of H feet dowDstream of control section which is enough in the ca3e of £ImrIes
with a glaci;;, lecame the height of th~ jump at the beginning of a glacis shaH never be more
th"'n l/a H ano therefore it~ spi:l5h shall alw'lys be less than H. Ill. the caie of meters with
.vprtical drop. the cl)nv~xit.Y introdll~ed by the drop mly be serious req1irins a length of crest
grf>ater than 2H, The author wo~ked out the length of level crest from c)U)iderations of a stable
ann an ~l(lherent hyp2f(~ritical jet flropping in depth by 5% from the critical depth in his Article
in Indian Engineering, Calcatta, December 1936. The length of level crest worked out to be
(~quaI to H downstream of the control section with f='0066 and if the crest was of cement
. concrete or pla"tereil masonry it shoul<i be about 2 H with f= '0044. The I"ngth should. therefore
be 2H in the case of brick masonry flumes with a glaciS and ~H in the eas" of falls ""ith
vertical drop or with cre3t surface cement.plastp-red.
[n the above-mentioned publication of the author, effect of the circular appro a ;h
cUrve t.n the crest was a1s'"l mathematic Illy worked out. This will tend to increa3e the
(onvexitv of surface flow upstream of 1he control section and, therefore, a relatively greater
length of level crest will be required upstream of the control section or the point of inflexion.
This effect increases with the discharge and therefore the length of crest were suggested therein
as given below :-
Discharge in cusecs per foot run. Length of crest.
Bdow 15 2H
15 to SO 2'5 H
80 to 50 30 H
15. Requisites of a meter flume.
The requisite,. of a cn rrect1y d~sjgned meter flume are :-
(i) The ]f'rgth of crest from 2H to ~'H.
(ii) Available working head or drop at the fall more than the minimum modJ.lar
head.
The minimum modular head is the difference between the water evel upstream at tht
Cauging s:te and the downstream water level, when the buffer of the jump just starts fouling
the control sl~ction and the discharge drops by not more than one per cent. .
Tvpe ta' M.M H.in terNS gf H
Free overfaJl 33 per cent
Contracten flumes (Crump's L type) 9 per cer,t
C0ntrar.tecl me 'cr flu'11c (t in 5 d,')wnstream expansion) 20 per cent
Full chann.,l width meter w:th 1 in 5 g J a c i s ' 10 pre cent
A safety margin of half a foot at least should be allowed for the likely silting up of H.C'
channel.
(iii) Strean line approach to the crest both on sid~9 and in the bed. The obejct is to
mmimises the Io~se3 d 11e to abrupt entry and shoc~. A. bell mouth approach a~" already
described in case of the Inglis Fall design ig the best in the case of contracted falls. In th," case.
of meters with full wirlth of channel rlownstream, the side approach will be straight with vertical
walls and it jg mual to provide a circu1ar approach curve in bed with radius 2H (Crump's original
design:: The author avoided the corner at the beg:nning of the curve by giving an S curve,'
ThA corner was always found to be filled up with irregular debris and was not, therefore;
~ esirable. (Fig, 24.)
 259
(iv) Correct Gauging of the depth on creSt.

~
f!lO.' ~ It would be very accurate if the depth at
the control section where parallelism i'l attained
___...-....:'=-R=t.;;;:$~T CJuld be accurately measurf'G, because there
~
~.
(' 3
~ _ .. _ .. ___ -l the discharge will then re simply y'gDc and no
U~.8tfl -'4-..... - correction for the velocity of approach will be

·~R-;FI
rpqnired. The control !"ection is simply a point
of inflexion on the surface curve and it is very
di ffi :ult to loccte it correctly. The de pth on crest
has, therefore, to be measured UDstream of the
Fig. 24 n:st ann the correcti m for the velocity vf approah
has to be applied. The es'e:1ti 11 fen nre:; f1r 'he correct Gauging of meter Gauge would be :-
(a) The surface pro'lle sh rm 1d l'e very nearly horizontal for an appreciable distance
upstream and downstn am of the Gauge hlle. This mean'> that there should be no change in
tl:e section in this length The change in section shall produce a slope in the water surface. A
Gauge well in the bell month approach to c -est is, therefore, Ollt of question.
(b) The velocity of approch should be correcty calculated for the section oppos:te th3
G~uge ho:e. This is only possible when the section (!oposite Gauge hole would neither silt nor
",cour for the range of the v triahon of the discharg~ OVd the meter.
In the ca"c of contracted meter flumes with bell mouth approach up to the crest, the
gauging has to be done in the earthen channel upstream where the bed level changes by silting
in low supply and the velocity of a'.lproach cannot be calculated correctly. Such falls serve a;
meters only in the full supply conditions and even thpn the results are approximat~, as they
neglect the energy losses from gauge well to the control section which is a consider abe distance.
In the standard meter name design de~;cribed by the author in his paper No. 154 Punjab
Engineeing Congn~ss, 1932 the floor level opposite gauge well wa." designed in such a way that it
rt~hlained free of silt even with half supply. fhe method is illustrated by an example in this
chapter. The type is shown in Fig. 25.
(c) The design of gauge wel! and gauge hole shottld be such that the surface pulsatioll3
in the channel are not cc nveyed to the water snrface in the gauge well. This suggests the use
{)f a sill§ile g'1uge h\)le locatt'd below the crf'st level. In the standard meter flume desgin
lmntioned above, it was prov:ded to be a circular hole in a metallic plate fixed flush with the
wall outside. A 3" X 3' hole in the wtll as shown in Fig 26, and the area of the hole in the
If. iddle of the iron plate from 1/2000 to 1/5000 of the gauge well area.
The position of t.he gauge ho!e from the beginning of the crest in the standard meter
flume design is kept 3H while there is provided a straigt reach 2H least upstream of it.
Thus horizuntal !'urface pr0file o.)po,ite gauge well is ensured, The
pUlsation on the watFr surface of the chamid are produced by action of the wind on the
sllrface. It is advisable to d:vidd the large flumes into compartments by constructing thin
R B. Walls 9" thick projecting '5 ft. above F. ~.L. and to construct gauge wells on both sides
to take the average gauge reading for accunte gauging.
Plan of Meter Flume

.~"t "'" ;
. "'1 " ~

r-t ' l:\

- _.L',
~ig. 25

. ___ :.r-:-' '.I$/",t:.=.

,.. \::'::'LOOA O • .<fZ{fN 114 I*10RTAIt

,,.
. • ~.... 0.• ("4.'.;
elM: ('<1ft ..
 260

'dJ The water s trface in the g:t~ge -w:eII should be accurately measured. It should be
M"easured by neans of a. Hook Gauge whIch IS n ~t usually practic:tblp. bl'cause f'ducated. gauge
readers capable of handlIng sl1ch gauges are not available. The accurate record of gauge IS now
made possible by installing automatic wat~r
level records. Legget's water level recorder IS
usually used in the Punj!l b. This instrument
I which giws a chart of ihe water surface level
1>10 .. ,- . . _ in the gau~e well is manufactured i~ th~
JI~7077zr.~--pt~Tli~~ Central Irrigation Workshops at Amntsar
F up-jab, tut it is a costly instrum"nt and
cannot be inst<>lIeu everywhere. It is U!;~~l to
put in sloping gauges in the gauge we Us <;hvlded
accurately to th.. second place of decimal of
a foot or showing cusecs. A sloping gange
Fig. 26 provides enough length for small Suh.divisions
l6. Calibration of meter flumes.
The formllia of discharf e as worked out for a longo-crested weir tiume in paragr" ph 14.
of this chapter is b"sed on the assumption that the channel was frictionless. Let hi be the loss
()f heae" from the gaug~ site to the c,mtrol s 'ct;on on the weir crest.
rn the case of raised crested meter flumes, the approach curve in bed is usually circular
as de~c~ihed bpiore, Even if it L omitted or if it is a short slope, the stream lines are curvred
irom thtl bed upstream to the cTe,t The curvclture in the stream lines cannot be avoided exc:pt
by giving <> very flat slope, say \. in 10, which is nClt po,sibk 'in the practical designs. The
effe~t of circular approach in bending the stream lines alld therebl producing centrifugal force,
wa~ worked out by the author in his article Ddcem")er, 1936 Page 199 Tndian Engineerings
eulcutt... Let it he dt'noted by F, so that F =5 H. The discharge formula q =3.09H3/2, can be
writton in the form q=C (G+h.-h+F)3/ 2 per foot run of the weir, where G is the gauge reading

and 11. is velocity of approach head= ~2 and C is the coefficient of dischargE'. If hf = F, C will be
- :lg
equal .0 its +heore~ical value 3.09 but if hr is greater than F ac; in th~ case of small channch,
the value of C is loss than 3.09 and in case of large channels, the effect of the cen ~rifugal force
is very pronounced, the value of C being more than 3·09.
..".
~ .., I ,_
J ~ I-
!-,_,...
'- ....
J..,..
:;;;

._ V
V
~
l;'

C/. C(G+ho.) ~= C H " WHERE G .. GAUGE READtNG.

/

-
.. ~'-
••
--r-I-
.. ...
'J
20.

OtSt!HA/fC' PER
U. -'0. n.
;t\,";:'

FOOt RUN II' Iif£TER nUNE.

40.
.,
-I'
,

. ....
The author had to carry out large scales observations on Lowe.l Jh~ lum Canal 0
mpter flume! with the Standard Crump-Sharma approach e." with silt-free floor levels which were
opposite gallge hole site up to half supply conditions. One interesting pCJint in these
 261

observations was that discharges were obsprved oIPosite the gaug'e hole by running of
traveller on two ropes stretched across the side walls in the section which waS always free
of silt and Jpvel. The vel lcities were taken with current meter. The value of actual cO"fficient
was sensibly constant for every m~ter flume with respect to H but varied with the size of
the channel, viz., the discharge for the different: meter flumes as plotted in Fig 27.
(b) After the value of C has been actually observed for a meter flume or taken from
the graph opposite, the discharge table for the meter flume can be work eo out with the help of
the table No.3 of thi!'> chapter or the graph in plate IV tB).
The o;)ject of the table is merely to provide a readv means of performing the proce"ss
usually referred to as "c()rrecting for velocity cf approach," in other words. of applying tt e ahove
f.ormulaQ=Cb (G+h a )3/2 in view of the difficuhy arising from the fact that the term ha is itself
· h arge.
pende:a t on th e dISC Q We have V 2 =
h = -"_ 1 Q2 _
----.~
a 2g 2g AS
50 that the formula for di::charge may be written
. t(G+ 1 Q2) 2/3
Q=CBI zg-'_A2- where Bt =width of meter at crest (2]
The values of C.Bt., G and A being known. the value of Q can. of ~ourse. be d~termine.d by
trial and error. It IS, however, a tedious and lengthy pro :es~ from ~hlCh the pracbcal engmeer
would be glad to be saved. In the attached table No.3 and curve, 10 Plate IV B, correspondmg
CI and CBt G
values of ---c ~ are given which enable the dis~harge Q to be calculated directly and
rapidly from the formula. Q=C 'Bt Ga'2 [3J
in which the value of C / is obtained by mUltiplying the known va.lue of C by the value 01
C'
[ C ] taken from the table or curve.
The calculations of corresponding values of _~ CBt G have been made as
follows:- C and A
From equations I and 3 above we have Q=CB t [G+ha ]3/2=C'Bt H3 /3
whenc CI G+ha ]3'2 [ £ 2/3 ha. (4)
~ C=( ~ : er C] =1+ G
But h a = _I. Q2 ______
~ I (C'Bt G3/2)2
._
2g A2 - 2g A2

"'~=21 (IBtGt=(C/)2_1_(~)
G g A C2g A
whence from (4)
,Cl)i =1+(C')2_1 _(CBtGi) _(5)
C C 2g A

putting~=y and CBtG =x=__£Q_ if Z=Height of crest above U.S, flOQl.

C A G+Z

Eqnation (5) becomes :_yf =1+y2 ~2 ; or x=_!_y2g(y2/3-1)

2g y

which enables us to calculate values of X= C~G for various values of y== ~' •
Corresp mding values of these two quantities are exhibited in table :3 and cur~
attached in Plate IV (B).
By way of example we will take the case of weir where :~
G= 16'0 feet; C"",3'40; Bt=240 feet, A-4500 square feet.
We have x .. CBtG _3'40X240X 16 0=2'90
A 45UU
 262
From the table or curve Plate IV (B) we find the corresponding value of
~I

y =~- to be 1'42.

whence C'=1'42 C; =1'42 x3 40=4'83

and Q=C'BG3j2=4'S3x240x (I6'0)3/2 ; =74'200 cusecs,
TABLE No, 3
CB,G
Table of corTtsponding values of Y= ~I and x=---:t\ =_1 -yl2g\y2/2_1),
y

C' CRG C' CB,G C' CBtG "..,

\..- CBtG
C
,-----..-,-~.- .. ~
-
-A-
C
-~.~-----------.-.~-~--~--
--x--- -
C
--- ---~----
-A-
--- - ..
C
-.-~---- .. -~---~...__-.---
---x-
1020 0'907 )'240 2'541 1460 29H I '6~0 3'071
1'030 1'\)99 1'250 2'571 1'470 ~ 9;; t 1'690 3073
1'040 1-251} 1'260 2-6()0 1'481) 2963 1'700 3 ()75
1050 1'390 1 270 2'626 1'·190 2'972 1'710 3'077
l'oeC) 1 507 )'280 2'8':;2 1 500 2'981 1,720 3079
1'070 1 6! t ] '290 2'676 1 510 2'98~ 1 730 3081
1'080 1705 1,3('0 2699 1 520 2'996 1'740 3'082
lr09c} 1790 1'310 2721 1,53 ) 3·()O.3 1 750 3084
l'wa 1'8(;9 j-320 2-741 1 540 30!0 1 760 3'085
]'110 1941 1 330 2761 1'550 .3 016 1770 3'OBC)
}'lZO 2007 1'340 2'780 1'560 3'022 1'780 3'087
1'130 2069 1'350 Z 798 1'570 3'028 1,790 3087
l'14() ~'127 1'360 2·815 1 581) 3013 I'SOO 3 (J8R
1'150 2181 ll70 2831 1'590 3038 1810 3088
l'16() 2231 1'380 28!6 1600 3'013 1820 3'Oil9
1'170 2278 1'390 2861 1610 3047 1'830 3'089
1180 2'323 l'~OO 2874 1620 30;:;1
1'lSO 2'365 1 '410 28S8 1,630 3055
1 200 Z 404 1'420 2900 1 6~0 3'0';9
1'210 2,441 1'430 2'9'2 l'S50 :3 062
I'220 1. 477 1440 2923 1'6GO 30(15
1'230 2510 1'450 2934 1'670 3068

17, Meter Flume Design Examples (Author's Dovelopment of Crump's DesigIJ ,)

[AJ Design a ro, ter flume in a channel with the following data:- (
Discharge = 150 cusecs, Depth=38 feet; side slope=l to 1; Bed1Vidth=-20 feet
Drop= loS feet, Range from full to half supply,
[a] Crest level :-Width of the flume Bt=:lO 'feet. f;.
' h arge per f oot run= 150_
D ISC -- = 1'5 CUiecs
20
The value of C from Fig, :~7 = 3'045,
Let total energy depth above crest=H
:,H= [t1iiI3=[/~h ]2 1
3=1825 ; Depth in challnel=3'S feet. I.

7,5
velocity of approach = 3-8= 1'97 feet per second f
,

v 2' t 972
ha = -- = ---='0652 feet.
2g 644
The depth on c est=- }'825-'065= 1'76 _ . ow:
Let urst earn water level=500 0'
The crest 1'wl=500'= 1'76=498'24
(b] FloOI level oppo,ite gauge hole
half supply discharge=15 CllSCCS.
75 ,. ,
Discharge per foot run=2j)'=3'75 cusecs' per foot rna,
 Scour depih or non-silting depth
D= I'll q 61= 1'1) X3'75'61=2'S feet.
Total energy depth on crest for half supply
3'75 )2/3 =1'15 feet
H={ ___
3'O-!:i
3'75
Velocity of approach= - - =1'5 feet per MCond.
2'5
t 52
Approach head=~='035 ft. . _.:.l.:..~,:?
Water Ievp.I in half supply=49S'2t+l'I5-'035=499'3S--- - .
TLe floor Ie-vel opposite gauge well=499 36-:ol'S =496'88 : .
(c) Length of crest=2H=2 X 1'83=3'66 say 36 ft.
Radius of upstream aprroach curvp.=2H=3'6 ft.
Distance of gauge hole=3Hh:3 X 1'83=5'5 feet.
Straight distance beyond gauge hole=2H=3'6 ft. ., ~,:~"
Radius of the curve of the side" aUs is 2H, Keep 5'0 ft. in this cue.
[dJ Glaeis, The glacis slope I in 5
Critical depth=i X 1'83= 1'22 ft.

Crump's L=I'5 ft· and_!:_=J.:.?-=I'24

C 122

From plate V for -~-=1'24, K+F =3'4;~='42 and-.!-2'Oa

CCc c .)
F=3'4X 1'22-183=4'14-1'83=2'31
Jump will take place at a di,tance from the beginni~g of gla.cis b
5 X 2'31 = 11'55 say ll'6 ft,
Depth upstream of jump='42x 1'22=0'51 ft.
Depth downstream of jump=2'02 X 1'22= 2 46 ft,
Downstream water level=SUO-15=498'S
The bed level downstream=498'S-;"8=494'7
Drop from crf'st to downstream floor level 498,24-4947=3'54
Length of glacis=3'S4 x5= 17'70 feet
The jump is at a di3tance of 1]'6 feet and, therefore, well within the glacis.
(a) Length of d"'wnstream protedion=5D=SxS'8=19'O say 20 feet.
This may be partly pacca if staggered blocks or friction blocks are added, but in tho case
v mlaU channel this may by only '65 feet thick dry pitching 10 ft, long,
The level floor u,ill be followed by loose bats protection I· 5 feet deep in bed in a slopt
of 1 in 10 for a length of 10 feet and sides bats-filling 1'5' slope I to I.
(b) Design a meter flump. in a channel with the following data:-
Discharge 700 cuspcs. Drop 6'63ft
B~d width 6U,O,Depth 5'1, full supply levels

Upstream 627'69
= downstream =612 U6
Upstream 622'59, I 2' ~ft·~'·
Bed levds=~~-- '=6-- -;sldes-to.I.N.S =6 4,0 •
downstream lS'96 2 .
Calculation, The d,op in this c,se is more than the dopth, rhe meter num': s!)all t,e with
fl vertical drop
Bed width downstream =60 ft.
The width between the side walls downstream of the crest shall bo 60 tt.
The width at the weir cresb.. 60-1 "'" 59 it. .
 264

When water drops over the weir crest, the drop causes convexity in the stream jinf's.
The convexity results in the distribution of the prefsure, The jet Jeaves the crest with preSSUTf'
at the bottom below atmospheric pressure and at the surface at atmospherk p:essure. Ttie actual
water level below the jtt is at C'D I instead of ABCD. Fig, 28 the differen Je has been observed to be

F,S.L

Q S p·e.L Fig. 28
,1

about O' 5 foot in thi falls of the s;ze as in this example. As the jet is being discharged against a
pressure varying from about 0'5 foot below atmospheric to at .nosph(> ric , the discharge co-efficient
of meter flume varies and increases more than the theoretical. This introduces another uncertain
factor and the remedy is to keep the cistern width 6 inches more on either S101;: than the crest
width so that free air can get below the jet.
The discharge per foot run of the crest width; q= 7_QO = 11'9 cusecs
59
The value of C ir c.>m Fig. 27 is 3'065 for q = 11'9
1]'9 2.
Toial energy depth above crest H={3'Oo:t ='2'47 feet
Velocity of approach (approximate as the floor level is not yet fixed}'
"'H-:;;n
II'. . v2 2'4 2
=~---=2" feet per second; :. ha=c-. = -2~ ='09 feet
5'0 ~g g
The Gauge Reading=G=2'47-'09=2'38 feet. . ;jii~'d:"'"
Crest level=F".S.L.-G=627·69 -238=62531
J i
(ii} Upstream floor level.
Half supply= 119
-2 = 5 '95 cusecs per foot run.
.

According to Kennedy's formula, non-silting depth=::"

D= 1 U'q·Sl= 1'1l=5 95. 111 =3 3 feet

HI = ( :'::5 )i- = 1'585 feet

5'95 1 ~o
Velocity of approach=~3'3 =1'8 !t./sec :. h= ~~ '='JS!
Gauge roadin~=Gl=1'585-1 05=1 535 feet
Water level III half supply:ocrest level r-G1 =S2531+1'54=*62685
Tho floor level opposite gaug.3 well =626'85 -3 3=628 5S
tUi) Length of crCl~t=2H=Z X 241 =4'fH say 5'0 feet ~"'L .0 •

The distance ()f cauge well hole from beginning of crest=~H=3x2·.~=7·5 feet
The straight portion llpstream Gauge hole =2H =.) feet .
. Tho rise in bed to floor";622'59~-621'06=1'53 feet aUowing a slope of 1 in S=l'S3X;S
=7·6 feet . .
 265

The radius of the side wall ~t upstream=6'5ft.

[iv] Cistern. Etcheverry formula:-

Length=3y1tCL H=3x,,/t:F6:fx2'47=12'15 it say 12'2 leet

Depth=one sixth=1/6 x 122=2'0 feet
MOHtagu formula : -
From plate vIr for HL=6 63 and q=l1'9, the value of
Ef2=4 25 ; Depth=4 25 x !=2 12 say 2'0 feet
Length=4Ef2 =4 x 4'425= 170ft.
(v) The protection downstream of cistern:-
The total length 4D or 5D=5 X 5'1 =25'5 say 26
This consists of 10 ft, level floor with a curtain w;ll1 at the end and then followed by
15 ft. of loose bats fill:ng 2 fept deep in slo;>e 1 in 10 bed and sides 1'5 feet deep in slope 1 to 1.
. (vi) Floor design (Bligh's method) Fig, 29 :-
The design is simple, using Bligh's theory, A creep co-efficient of 1 in 7 is enough for
the usual Punjab clay soil, and in some cases it has been kepi 1 in 6 wiih success. The worst

---- - - - - - - - -17 - - - ---,

III

NDN SII.TIN(; fLOOR TOR HAL'SIJ'I'l'f:

'. GONDlf/ON

.,-. -.f
_ . _ _ 27.0'_,
'40 .-n.':"__.-t
Fig. 29
condition for the creep is when water level is headed upto crest level a.nd the downstream
bed is dry,
Creep Head =crest level--downstream bed level=6Z5'3l-6IS'9(l)_ 8'35 fee t,
Creep length required=7x9'35=66.
Actual creep leogth=89'9 which is enough, Fig, 30,
The upstream floor opposite ~auge could be :heapened by making it in dry brick pitching 6 feet
deep and need net be changed in this casco The actual creep up to begining of dstern=54 feet.
Head lost=54j7=7'7 feet.

The thickness of floor=!_ X ~~--=-_7_~=~ X -_!-~-~--- ='Z'20'

3 p-l 3 2'0-1'0
Keep 2-75 for half length and 1'75 feet consisting of '75 feet masonry· of brick on on~ foot 1)(
concrete for the remaining length of cistern.
(vii) Cistern floor Des'gn Khosla's method, ,
Depth of downstream cut off=3'5 feet,
Total length of impervious floor = 52 feet.

0.= _52 = 14 Sand H =~ 35 =2'67

3-5 ' d 3'S

1 II 1
--~ -'115:.GE==-.--_ =2-67x'l15='318
?t VA d ?tVA ~ ." ,

increase the masonry depth to 3-25 and cOD(;rete"",'75 for the end 'Cnrtaln wall then
 266

52
a=--IS' 1
4- " ,. -nv',\. from plate 111='12

GE=9'35 X'12=28lthe safe value is 1/6)

4

This is safe mough for the Punjab clay soil. Note that
on account of clay in the soil the pressures calculated from
Khosla's theory are reduced to 70%
Pressure at the begining of the cistern.
b=27 +25=52 d=12':;S feet
p 27
-!. =~='52 and a=5Z/IZ'38=4'Z
b 52
Fig 31) ,pE=64%, ,pc==35%, ,pD=49%
The correction for depth=drop (49-35) = it X 11'38/12 '38=: 12'85%
The other corrections are np.gligible.
pressure=47'S5 X 9'35=4'46%
The soil being a mixture of clay and sand, the effective press'U"e
='7x4'46=3'122 feet
Actual depth of floor=3 7,;it. is therefore safe.

QuestIons,
What do you know about the Standing wave meter flume, Fxplain the ase of It. (T.C.E,1933),
'2 Discuss the advantages and disadvantages of a V notch'versus plain crest in the case 01 canal falis
Show by a sketch tbe effects of extension of impermeable platform in a weir (a) upstream of the drop wall (b)
downstream of the drop wall, (T.C.E,1934)"
3 (i) DiscU!,s the advantages and dls~dvantages of a plain crest versus trapetoidal notches in the case
.1 falls in (I) Main Canal (ii) Distributaries (3) Drains and escape channel.
(ii) Calculate the height of a crest required for a '2 ft. fall in a OistY,with full supply discharge of 150
~usec'. bed Width 25 ft. bed slope 1/5000, lengh of crest being 20 ft (T.C.1935)
4 At a certain point OD. a Disty, the channel dimensions are as given below:-
Discharge = 1l0cs. Bed width=20 feet.
F.S.depth=3ft. Top width of bank 5 feet.
Top level of bank 1 ft. 6 inch above F.S.L. Full supply level at natural surface,
It is desired to raise the full Supply Level upstream of thIS pOint by 1 ft. 6 inch. Design a broad crested
meter dume fall allowing for the possibility of somt slight silting occurring downstream,
Give dimensi ned sketches not necessarily to scale from which a Draftsman could prepare working
c1rawiDgs. The thicknes~ of wingwalls will be fixed by the Draftsman. Giye reasons for the particular ratio of
width of flume to bed width of channel ;ond of depth on crest to depth of channel. (P.LB. 1 HO)
5 Describe a method to check beavy bed scour immediately blow a fall in main canal, (P.LB. 1 ~39)
6 Design a broad created meter for a distributary carrying a discharge of 91'S cusecs with full supply
depth 3'3 ft. and bed wid th 13'75ft, Difference in upstream and dowmtream level may be taken as .75 ft velocity
.f approach sbould also be taken into account. also, draw its dimensioned sketch (which need not be to scale)
thowing plan of the meter with position of the gauge well and its L. section with downstream protection etc. All
dimensions relating to the d'sign sbould be clearly shown. (P.LB. 1938).
, 7 Prepare the pr~liminary free hand dimensioned sketches and calculations for designing a broad __ _
created meter flume for thebead reach for channel with A. F.S. 300 cusecs iD sun dry loam soil. (P.I.o.1937)
·'>;Ii· (. 8 Design and sketch a broad crested measuring weir for a channel with the following data:-
(I) Discharge = i 00 cusecs.
(2) C=3'1
(3) Difference between upstream and downstream wa.ter levels=O.8 ft.
(41 Depth of water in channel 3'1 ft.
(5) width of channel = 17 0
(6) sides=l to 1. (P. U 1942).
9 Describe briefly with help of sketches the various types of the canal falls. (P.U 1942)
10 De~c!lbe the vuious methods of destroying surplus energy downstrea.m of the canal falls. with sketches
.04 give reasons why you consider a particular device superior to others.
11 Describe the signilicance of long crest in a long crested meter flume and explain the essential
nquillites in a good meter flume design;
 267
12 Describe the advantages and disadvantages of contracted falls. Explain with sketches the chie.,
characteristic of Montagu type fall as used in the Punjab.
13 Deduce an expression for determining the discharge over a weir with a clear overfaU. How wow.
the above be modified if the weir is partially submerged during floods? (F.S.c. 1935).
14 The' upper and lower suTtace water are 6'0 ft. and 2'0 ft. respectively above the crest of submerget
weir 70 ft. long. Calculate the discharge; take C=O'S for the drowned portion and C= '557 for undrown~
portion . (F S c. 1937)
15 Explain the phenomenon of standing wave and state the condition necessary for its formatioa
Draw a sketch of s,tanding wave .fll!me carryiq.~ out its essential parts ~nd expl.ain the functio~ . of eae)
Deduce an expreSSIon for determining the discharge of a channel on which a standmg wave flume IS Install.
State assumptions m a d e ' . (F.S.C. 1939)
 PART II

CANAL IRRIGATION
CHAPTER XI
Silt Excluders And Ejectors
1 Introduction.

The idea of silt excluders originated with the late H.V. Elsdon, Executive Engineer
Punjab in his paper No. 25, P. W.D. Irrigation Branch Publication L922. Three ejectors or
extractors and two silt excludf!rs were designed and cl)nstru ~ted in 193t -35 on the Upper Jhelum
Canal by E.S. Crump, Sup'rmte!lcling Engineer. ~alampl1r Extrator on the Upper Doab Canal
followed soon after. Khanki and Darlnpur Silt excluders were constructed recently, and the
H<l:veli main line has been provided with a silt ejector. A very co:nprehensive paper No. 211 was
Wrttten by F. F. Haigh, Superintending Engineer on silt excluders in the Punjab Engineerig Congress :
1938. A brief summary is given here. '
Definition.
The silt excluding device built in conjuction with the head regulator of a canal on the up~
stream side in the river is called a silt excluder.
When silt excluding structure is constructed across a canal, it is usually called a silt
extractor or ejector.
The basic principle on which both are designed lips in the fact that in a flowing stream
carrying silt in suspension, the concentration of the silt charge in the 10 Ner layers is greater than
in. the upper ones Fig. 89 Consequently, if we can escape the lower supply without interfering
With the silt distribution. the remaining water will have less silt in it per unit volume than the
water upstream of escape,
[i) The essential features of the design should, therpfore, be to provide concentration
of silt charge near bed by some such devices as reduction of friction by paving or by plastering to.e
bed and the sides. or hy increasing the section and redlcing velocity, thereby lowering the bed just
upstream of the excluding device. This can easily be arranged in the casOl of a clnal sp.ctio1l flowing
straight with uniform flow. In the case of a river the conditions approach to head are usually
curved and disturbed. Silt ejectors in the canal are, therefore, more efficij>nt than the silt
excluders at the head of a ('anal.
[iiJ The other most important feature of design should be to provi ie saperati(m of the bottom
water charged with c mcentrated silt from the uP?er watj>r wi~hout anv dist'lrbance, This
requires that water should enter the approach to the chamber below the Diaphragm slab with
the same velocity as it is flowing in the canal aDproaching the work without disturb.l.nce,
(iii) Lastly the high silt-laden water thus stealthily taken away from the canal should
be es-ap d through the tunnels connected with approach with a velocity which does not disturb
the entry of water a.t the approach.
3. Approach Channel to a Silt Ext ·actor.
Undoubtedly the o' ject should be to provide a long straight approach chanl'leI. in which
the silt can settl"! into the lower lavers of the water, Ula thu~ increase the efficiency of the
extraction. Anything affer the nature of curve, whir.h will dispbce the silt ('oncentptiotl t,) the
s;de of the channel as well as reduc':) it, or any obstruction on th,~ sides or had wnich will set up
turbulence and hence destrov the b@d conc",ntf'ltion. should be avoided.
The great arlvantage that an extractor has over an eXc!11der i'.l, that the approach
channel can generally be securpd without difficulty in the cas~ of the former, while in case of the
latter it is usually necessary to turn the water through a right ~nf I.. hend before separation
is dfected awl in a o.y ca~e, the approach channfll, which cf'nsi~ ts nf a natural river bed, will
probab;y haye curves iD it and will certainly have very inegular boundaries.
 269

(i) The approach channel shQuld be designed with the flattest slope which will carry
the l,eaviest grade of silt likely to approach the work.
(ii) It should be pitched or lined up to 3 to 4 D (depth in the channel) upstream of the
work.
(iii) If possible, the bed may I::e depressed upto 1-0 ft. by a gradual transition curve in
the bed. There should be 110 change of section on the sides.
(iv) The minimum length of approach channel below a head rpgulator to ensure the
stilling of the turbulence down-stream of the head and then the restoration of the
normal silt distribution, may be taken about 1000 ft, for a canal with about 6000
cusecs discharge. .
.4 Escape DisehaI'ge above Ex~racter.

TIl''' efficiency of an excluder may be defineil as the reduction per unit of the silt
intensity in the caml supply when comp:ued with that of the water approaching the work.
This, though the onlV practical standard, is a fal"e criterion, Th~ true measure of efficiency of an
excluder is unity minus tht ratio of the silt entering the caBal to that whi ·h would· enter it if the
excluder were not working. The pc int about this dIstinction is that the addition of the escapage to
discharge the canaIrdischarge increases the silt aflprnaching in the canal and increas··s it in a propor-
tion greater than that of the discharge, because in Kennedy's theory, silt transporteilv aries as D ol3
and discharge varies as D5/3. We mmt not, therefore, bl ndly a~cept the idea that the great2r the
escapage the greater the efficiency. Research is neces;ary to determine ",hat the optimum propor-
tion is 20 % or below is suggested a3 r<!asonable.
5. Appr.:>ach design of the Extractor.
The St paration of the escape water from the canal supply at the edge of the diaphragm
should obviously be arranged without disturbing the silt distriJution. It is easy enough to arrange
this for fixed canal and escapage discharges by placing the diaphragm at a height such that it
divide; the normal stream into the correct proportion. In practice, however, it is always necE'ssary
to vary both the canal supply and the escapage and if the height of the diaphragm is fixe i it will
generally nlt suit the varying prop::>rtion of the two. In the ideal case, water near the bed should
en.tar the ejector with the same velocity with which it 'approaches it and in the full supply condi-
hOD, the stream lines will be horizontal for the water passing OVAr the diaphragm. If. however.
it is required to unr the canal sllPply full, while sufficient wat-~r is not avilable for the discharge
escapage or if theescapage is run full, when canal supply is low, we shall have stream hnes
what some as Fig. 31- ,

Flg, :3

~~------------------
"71." ,»,»; > » ?/ ) ,.>/) .'/J / »); .'))';;,/7,',' //) ))~,.
A. B.

In the region AB there may be a certain amollnt of turbulence set 'Up which
will cause a less favourable silt distribution. The question is, is this disturbance serioUi ~
According to EIsden it would be as shown by his proposal to have the height of the edge of the
diaphragm variable. In praetice this device bas not been made use of and little attention has b.e,~
paid,to the point in the design. This point needs investigation in models.
 270

'0 (a) TUDnel Entrances.

For f'conomy aud to avert any danger of silting. the velocity in the tunnels must be
Iligh. We have, therefore, to transform the comparatively low velocity of the escapage at the
approach entry to the high tunnel velocity. This must be done without allowing the draw of the
tunnels to affect the velocity distribution upstream of separation and in certain cases when
escapage head is valuable, a gentle transformation is necessary to avoid loss. The former point
can always be secured by. placin~ the entrances a s?fficient ?istance downstream from the edge
'f the diaphragm, but thIS may mvolve an expenSIve cantlleverd slab. The method by which
this is usually arranged is shown in Fig. 4.
(b) Tunnel DC Ilgn.
The tunnels them"elves must be arranged to evacuate escapage at high velocity. say not
leu than 10 ft. per second. They must aho provide contr01 of the discharge sO that the sam'} velo-

SALAHPUR EJECTOR ~ ~
U.S.D. CANAL , ~
~

'''&Ll.-&, •

Pig. 4

-
t

my is secured at the entrance to each tu~nel. This mlybe done either by keepingthf saille tbtu'UI
iimeasions and varyiog the width of clnal served by each tunnel. o.r by keeping the widths fi(fTv4fd
271
 273
and hence the discharge the same, and varying the tunnel sectional dimensions to secure the
!:ame discharg.. with the varying tunnel lengths, The loss of head in ~ach tunnel may be
calculated hy Manning's formula. At Jaba, th~ widths served and the tunnel dlln~nsions are
the same and the discharges are equalized by a varying degree of contraction at the tunnel exits.
At Dadupur, the nischarge of each tunnel is the same and the tunnel dimensions are varied.
There is a control weir at th., end of each tunnel, the crest llwels of which are varied to suit
variations in width. At Khanki and Madhopur there appear to be no devices for equalizing the
tunnel discharges. The tunnel roof ~hould be designed to take the fnll water pressure above it
with the minimum pressure which may occur inside it, assuming the entrance tn be blocked,
If the tunnels act as weir for the canal supply, the possibility of uplift occuring with the tunnels
closed at the downstream end, and a velocity <:iepression oyer the roof should be studied. The
oocapage, if less than full supply, is regulated by gates on the down·stream end of the tunnel.
At Madhopur , a surge chamber is provided at the down-stream end (If the tnnneis which serves
to obviate any danger of water hammer. Thh arrangement also pnrnits the regulating gates
to be placed with the chamber and enables gr()OVei. with only onl! face machined, to be useo.
At Dadupur. the gates are placed on the outeI' ends of the tunnels and the grooves are machined
on both faces. the gates, against the upstream face. At Khanki, Bhong and Jaba the exillting
undersluice gates were used to control the tunnel discharge.
Cc) Tail Race.
The channel from the ejector to the river for escaping its dhchar~ should be designed
with Somewhat steeper slope and relatively higher value of C. V.R. as comp'lred with the canal.
7. Extra.ctors combined with a Meter.
At th~ h~ \d of a c<\nal a silt ejector can' conveniently be combined with a meter
fiu""e wit h crest higher than the diaphragm level. For accuratA gauging it would still be better
if the fioor level opposite gauge wells be kept higher than the diaphragm leveL For idea.l
conclitinns, it should btl th~ top level of the diaphragm slab. The meter flume shoud be with fn II
width of the channel without bringing in coml)lication of the contracted approaches.
There has not been any serious trouble with the existing silt extractors resulting in
thpir choking due to jungl p getting into them. There is usually no arrangement in the form of
trash-racks to remove jungle. laba had a slight trouble to begin with.
1. Effect of ex~luder or extra.~tor 0:1 the Ca.nal regime.
So far the silt excluilers have been hu.ilt where the silt entry has bep.n affeding the
regime of the Canal to a dangerous extent. In the case of the Upper Jhelum Canal, the canal
capacity had been reduced conSiderably and was still diminishing, as a result of silt entf\'. In th.,
case of the Lower Chetlab Can~tl. excessive silt en try at the head had caused the hed uf the main
line to rise. and was giving great trouUe in th(~ branches and distrihutaries. In all these ca,es,
therebre, it was obviously neces!-ary to exclude or extad silt from the canal, as much as possible,
and the excluders were constructed as a methOd of affecting this; but in cloing so, no attempt was
made to consider what grade or (,}uantity of silt was suited to thp. canal regime, and to regulate
the silt entry a ;cordingly The excluder has alwa"s been designed to be as 6fficient as local
conditions would permit at reasonable cost. It is evident, of course, that if the excluder was too
effective and resulted in too rapid retrogressi:>n of the canal, it could be dimsod or worked
intermittently and thi~ had actually to be don~ in the case of tbe Bhong excluder, where the
retrogression caused endangered masonry works in the canal.
On the Upper ]helum Caual the efied of the excluders has bf'en so far to stop the
progressive silting up of the canal which was in progress, and to restore its capacity to that
desi!;ned. On this canal the capacity is ruled by a minimum gauge of 11'3 at JlggU. The
discharges given by this gauge in different years are as given below:-
1934-7695. 1935-7875. 1936-8160.
Tn 1937 the gauge of 11'3 was never attained. In January of that year a gauge of 11'25 gave a
disc harg~ of 8::lS5 cUlLecs which appears to have been the maximum run.
On the Lower Chenab Canal. the progress of silt erosion is about l'S8 ft. in the head~
reach upto R.D. 40,000 and '93 ft. upto 140,000 The. effect of silt excluders in the case of the
Western Jamna Canal and the Upper Doab Cabal has not yet been felt.
 274
The amount of silt exclusion in illl cases wag arbitrarv. No attempt hag, howover. so
far.been made to reguld.te the silt <tt entry. to a gr;:J.de and quantlty sutted to the slopes for
whIch the canal ha.q been des;gned, but sucft regulation must be the aim in the case of an established
c~nal system. How th\s ai n has to be achieved, however, is by no means clear. To r~gulate
SIlt we must have sone standard for mea'iuring it which can be rearlilv observed. The most
generally accepted stand'ud for silt gra.1ing is La.cey's 'fL' which cannot be den.ned more clearly
than to say that it is a ch'1.rl.cteristic which relates silt grade! to disharge and slope in a regime
cha.m.el. The difficulty of the proble:n is increase i by the f::tct that' fL' vaTies throughout a
canal system, under the inFfucflce of such factors as attrition and the unequa.l silt extrac.tion of
off-ta.kes, and before we could regulate to a given'fL' at the head we wouln have t() relate that
'fL' to those of tile canal as a wIlole. There does not. therefore, seem to be much prospect of silt
regulation in the immedi'lte future, of 'fL'. .
However, the author's formula Re=rl 1/3 r3·3.~ 1/6 giveing the change of C. V. R. with
respect to the silt charge and silt grade can successfully be used even in the ca.se of silt excluders
as has already been done in the case of the silt-selective distributary h'}ad-regulato r
Chapter XV. ,

9. Efficiency of silt excluders or Ilxtractors.

There has been some doubt in the past as to the best method of calculating the efficiency
of excluders. The one in general use gives only the reduction of silt intensity in the canal water
as compared with that of the approach flume. On the Upper .lhelum Canal, howev€r, it was
formerly customary to take the ratio of the total silt escaped to that of the approach flame as
the efficiency, and tbis method is employed at Madhopur as well.
If Q, I and S are the discharge, intensity and total silt contents the SUffiX P5, f, c and
x denote the approa~h flume, the canal and the escape respectively, the efficiency is given
'by Haigh's formula.

E= IL-Ie = I - l (A)
I, I
1£ observations of the approach flu,ne are not available, the efficiency may be obtained
from the canal and escape observations by the formula. :-

E=- Q_,JIx- Te) (B)

Ie Qc+1xQx
According to the second method referred to above efficiency is given by :-
E'= Sx=" _Q-,,~= QJ, _Q1Tf-QJe
Sf Qrlt QcI.+QxI" QII,

whence E' =E+ QxIc (B')

QfI f
: It is clear from this that the latter method gives a greater value of efficiency then that
calculated by the former. It may also be noted that if the excluded efficiency is nil, i.,e., if the~-­
intensities in the two downstream channel are the same as that upstream the latter formula"

would still show high efficiency gx while the former would correctly ~ ive a nil value •
. Ql,
In comparing the efficiencies of excluders it must be remembered that the great"r the
proportion of the supply escaped, the greater efficiency. As has been pointed out abo 'Ie, the,
efficiency will not vary directly with the e3capage. Since the intensity de~n:asf)s rapidly with·
depth, additional escapage will increase the efficiency, but !lowly. .
Anoth.r point to be borne in mind in this connection is that the efficiency mmt re:
affected by the grade of material carried by water. Since that grade has been found to cause .:.
trou.ble .in practice, effor~s are .ma?e to exclude everything la.rger than ~'2 mm. diameter. T?e i
proportwn of the total SlIt whIch IS great"r than 0 2mm. and the ralatlOn of th1l coarscst s}'lt;.
carried to this grade will, however, vary at different sites. ...J
 275
JABA S'L.T £XTRACTOR
SqLE,j'so
',.
'111 '
tl(l: PLAN
MANHOLE. (OVER BOLTEO OOWN
T~J~PI"T (J:PlIFT IlY W~TER --..

6·0

Fi

.,2·o.,....~-----

This same .excluder may be expected to work more efficiently where the proportion of
KH A111(1 (XC, UOER
PLAN OF TlINNElS
s"M"'V~' .
• ~- .....4'"

Fig {1
 276
Th:! same excluder may be expected to work more efficiently whtre the proportinn nf
coarser silt is gr..-ater than wbere it is small. but on the other hand, the coarser the grade of silt
carried. the greater the slope and velocity ann consequently the les:) the concentration of <ilt in
the lower layers. It also seems prot'able that the coarser the silt present, the greatf'r the
efficiency when ba;e~ nn the same grade, but this is by no means proved by our present
knowledge of the sutJ]ect. ... - ..
10, TypicaIDesigns.

The typiral design of a silt ejector is shown in Fig. 5 alf usel4~t, Jaba. tle Upper
Jhelul1l Canal, and a different design of a silt excluder is given::f~::Fig'~H~!! !i~ti

IL Automatic Silt Extractor Without loss of Water.

The author described a design of such extractor in the Indian Engineering, Calcutta,
January 1938. A brief summary is given here.
The silt movements take place after the Monsonn in the months October ::Ind November
,a~d the silt ejectors would be most effective then, but on accout of the Rabi sowing penod,
they have to b ~ kept closed as nnt a drop can t hen be wasted
'. [aJ Different types of silt extra tors or ejectors have already been tried. Tbe silt ejector
at R.D. 1,00,(00 of Jhang Branch has simply got slits a couple of feet below the silted bed leveL
The depth of the channel is about 10 feet. The depth of silt above the slits stands by mutual
cohesion 0 t particles and no silt or water flows out. If the silt above the si Its is disturbed, thl!
canal 10 Sf'S more than one-third supply under pressure head of 12 feet. Such ejectors wor ked
by tt:e pre~sure of thl' depth are useless ou account of loss of a great amount of water.
Improved ty; e of silt ejectors have been constructed on the Upper Jhelum Canal where the
discharge lost throngh them is controlled by means of gates on sluiCf~ valves. Chak-Sikandar
ejector is a nice example of a very efficient type. The maximum discharge is reetricted to the
dIscharge of the canal up to 1 f00t depth. Even 1he controlled discharge is as much as 250
cusecs whiCh cannot be spared for escapage thrtlughout the year.
[bJ The principles of silt exclusion for distrIbutary head regulators as described by the
author in his Paper "A Silt Selective Distributary Head Regulators" published by the Punjab
Engineering Congnss 1936, are equally applicable to the silt extraction from the canals 1 his
paper describes clearly the detailed experiments wbich were carried out on full size modf'Is made
of glass to determine the dfe:t of streaml;nes in each case. About twice the normal silt charge
in a cubic foot of water ~oulO be extracted if tbe red water of the canal be stealthily extracted
without the slightest disturban~e. It is very essential that no disturbance should be generated
in the act of stealing the b,.d silt-laden water. \Vater should be taken away with the s:'-Ime
velocity with which it is approaching the silt extracting device, which ShOUld not in itself be
an obstruction or a disturbing factor,
(c) If the supply which e~capes in such a case is, say, 10 per cent, the silt extracted
by weight will be. one-fifth of .the. total silt charg~ of the 'channel This is. an appreciable.exclusion
as it mostly consish of the obJect·onahlt COHse sIlt. In order to further lDcrease the efficIency of a
silt evtractor, the eftorts should be ilin~cted to concentrate the silt charge near the bed of a
canal. This could 1 e attained by reducin~ the roughness of the approach channel sectlon.---
The reducing in rough/lPss means eliminatIon of vertical silt carrying eddies. Cement plastered
sides and bed will do upstream of silt ejectiug aperture in bed of a canal' up tn a lengt h of.
say,S times the dl'pth of the channtl. The concentrated silt charge near the bed will thus be led
to' the apertures "'ithout disturbance. The degree of concentration of silt charge near the bed
depend, upon the reducti"n in ronghness in the approach channel. Glazing or even covering the
perimeter with mica 8r celluloid !'hf'ets will be more efficient than cement plaster. The author
thinks that ~uch expfllsive measures are not in fact necessary. Even cement plaster will reduce
the roughness of the channel section to oDe half in comparison to an earthen section and conse-
quently is lkely to double th.. etflcier cy of a silt ejector which is sure to ml'et the requirements
in all cases. 1he scientists having res"urses of Hydraulie Research Institutes should help the
engineers by evolving some cheap methods of hydrogen ironisation of water reaching a silt
extractor to concentrate all silt charge near its bed.
 277

(0) After providing suitable approaches for efficipnt silt ejection, the design of the
ejector should inv@Ive a.. little los;; of water as possible. The aim should be that there is no loss
of water as there is harrlly enough water in winter for the crops and no water could be spared
from any canal for purposes of silt.ejp-ction. Tbere is bound to be some loss of
water if the silt-ejecting device is worked by the hydraulic pressure
uf ' depth. Unfortunately. all existing ejectors are worked by pressure. They
invariably entail a considerable loss of water. We cannot afford to waste a single drop of water in
winter and they have to be closed down for six months. We should, therefore. think of SOme
other property of silt or water to work oat silt ejectors. Laboratory experiments could be usefully
employed ,(' dis~over Some dev ce to eject silt from canals adoptbg some simple principle such as
variation of viscosity due to difi(?rent sitt charges in water. As described bdow, the writer has
worked out a design for silt ejection from a canal, utilizing the difference of specific gravity of
silt and water. This device essentially aims at little or DO loss of water in the proce')s of silt
ejection.

[e] Suitable Design.

A suitable design of an automatic silt extractor without loss of water is: given in
Fig. 7 [aJ. It is drawn to scale for construction at R. D. "0 of Mancher Distributary. Lower
AUTOMATIC 51'-1 EXTRACTOR
Chenab Canal, the discharge of distributary
WITHOUT LO'S OF WATER
'. or
." being 70 causecs. Regime slope cannot be

.......
given and the available slope is 1 in 5,714.
....
~~
~ The designed depth is 2'5 feet. Tbis distributary
Os
takes off from one bay of Chenawan escape
which has got a very low bed level as shown
.,
in the drawing and which can easily take
eway the silt extracted from ebe Man-
...v cher Distributary. Smooth approach in
.......""'-'NCHER"
- ... ______..
IIY~
.
<II

cement plaster is provided upstream of

«I
: 1,.0 F.
'"
:ul silt extracting apertures.
the channel is not
The width of
restricted. The
I i.' RERM
,.~
reduced roughness of the approaoh chan-
:'I!L.<~~~~'-ll nel will thus concentrate the. silt charge
-<
;!I near the bed. -

Fig. 7 ., Water will leave itsi eargo

above the aperture as a suitable
O&TAIL. OF CH ... "'aE~
expansion is provided for reduc-
tion in velocity. A guaze deflector
is provided as an additional factor
to catch the silt. The automatic
device consists of a swing gate
which is permanently loaded on
the other side for the head of
water above the gate which is
3'5 feet in this case. Additional
loads are provided for the weight
of silt on the gate. The additio.lla
Fig. 7 (b) . .
weights could be adjusted in such a way that the swing ga.te should open when the SlIt, 15
say 9 inches over it. Provision is made that the gate opens just enough to drop. awa.~ the sllt-
The gate will then automatically close the aperture till it is again loaded w1tb. sIlt upto a.
stipulated height. Very little watpr will flow out when the silt is dropped. InspectIon ehamber
is provided as showh in sec tion. A very steep slope of 1 in 20 is given in the tunned :so that th.e
 278

silt moves out into the escape with the least amount of water. There are no sensitive moving parts
needing special attention. The ball- hearings of the swing gate shall require to be oiled occasionally
This device provides a suitable design which could be installed on canals which have got syphons
with good outfaIls. The working of this silt ejecting contrivance is automatic and without much
loss of water.
12. Example (f Desine:- (silt Extractor at R,D. 600 That main line)
{i} Tunnels.
Each tunnel serve~ 80 feEt of the bed width. To ensure correct ilistribution over this
width there is proposed a diaphragm separating it into two orifices, spread 6 feet apart, each
taking 1Z cusecs under a head of 0'5 foot. The ar"a required for this is given by q = 7 Ay'H ;
whence A=2 4 sq. ft. i.e. 2'4 X I': the orifices have been stream lined. The tunnel on the left of'
'
the diaphragm is 4'x4 ior which a=16, f=1O'6 and m=l. The dis~harge in 1his channel a~
'
compared with th"l t in the right channel loses additional head in a length of 30 '. This is ca1culat
ed an below. From Manning taking h=O,0165; v=O'9m2/3 SlJ2 where s is the drop in 10,000'

301=.~
2
s- 0 v ; now v=3'75, m=l; whence 5=0174%. Loss of head is xO 174=00522'
81 m 2 '3 . 100
This may be compensated by raising the velocity in the right channel to VI so that
(v '-:_-_~~~ -0'0522
64 -

v= 1'83+3'75=5·58 ft. Area of tunnel at the narrowest section= 60

5'S;; ,
= 10'75 sq. ft.=2·69 ft, X 4 ft.
The length of tunnel downstream of thA end of ,diaphragm is 76+170=246'. For the
tunnel, a=32; Pw=24, m=133 ft. Loss of head per % ,foot=O'1l5 ft . . ', loss of head in 740.'
length=O'29 ft. Between each tunnel there is a difference of about 60' in length. Loss of head in

60 ft.=0·118 X 60__ =0'0708 ft.

100

Maximum velocity
Difference in length, Difference in required in tunnel Area required Depth. Width,
loss of head. at the paint of
sudden expansion.

2nd tunnel 61). ft. 0'0708 5 ~9 20'4 4

3rd tunnel J20 ft. O'141G 6'71 17,1 4
4th tunnel 160 ft, 0'2124 7-4S 16 1 4

rotallo~s of head in any tUl1I1~l. therdore, is

Loss at entry =0'500
Loss in 4' tunnel """OOSZ
Loss in 8 1 tunnel =1) 290
Loss a.t exit "",0219
1'061 ft,
(ii} Out talt,
(a] Surface fiow:-J
The minimum river level with which it will be deCessary to ,work the' extiactdfs may
be taken as 675'0
The normal discharge per foot of tUn'u-el IS 15 cusecs which is reduced to 12 cl.1secs in
the cistern. Assllmin§ the final pond at R.L, 69! 0 and a T.E.L at the tunnel exit of 693'OJ
H-693~675=18 fept.
Hf~~..:: 52. Hence it wiU suffice to' put the cistern floot at R.L. 670'0
 279

With a pond level of 690 and tunnel soffit level 582 of at the downstream end, the
head will be 8 ft. and the discharge 1400 cus"cs
:.Discharge pp.r foot in cistern=34 cusecs. Downstream T.E.L.=683·9
The minimum downstream level required to keep the standing wave on the cistern
floor is 677 4. Keep cistern roof at R.L. 678.0 to provide ior this depth.
With a pond level of 692, the tunnel intensity will be 49'5 and the maximum down-
stream level consistent with the discharge about 682'0
With a rlownstream level of 685 corresponding with about l'5lakhs in the river, the
extractor will take 1300 Cs. .
[iii] Sub-soil flow and pressures.
With the extractor closed and the mllllmum downstream water level of 6'i3'0 it is
estimated that spring level will not exceed R. L. 682'0, Pressure at the back of pile line will be
governed by spring level a.nd may be taken as 3/4 of the head from spring level to outfall i. e.
;)/4 X 9 or say 7'. The depth of pile line required for an exit gradient of 1/4 is given

H 1
by;-GE=!= 7\d ·VA. - whence d=9'3 say a 10' pipe line,

As reg:nds pressure, the alternative, of an open cistern involving heavy wing walJs and
roofing the cistern has been worked out and the latter is found to be considerably cheaper.
[he maximum pressuri' under the cistern floor will bil abont 9'. It is necessay, however, to
provide for silt depositing on top of the roof for which a level of 590'0 is suggested. The
preSS'lres resulting from this are greater than the sub-soil pressures and hence rule out the
design. The roof will be loaded to the level ab initio to take the sub-soil pressure.
The thickness of the barrels, walls and roofs will be approximately as shown in
the drawing Fig. 8.
(iv) An eeonomieai thickness of a slab.

Thic~ness ?f a slab does not appreciably affect the cost of form work, hence in workivg
out-an ecnomlcal thIckness of a slab the cost of shuttering and centring is not considered.
Let M=Bending moment.
t=effective thickness of sla.b.
c=cost of concrete per eft.
s=cost of steel per cwt.
1'S" =Concrete covering for steel.
Futther assume that distribution steel is 25 per cenl of the main ste!l;and the area
. 5 tee l'm square mc
af mam . h es=-----_
M
, 140UOt
Cost of slab per sq. It :-
,!
Steel= 12'5~ X _.i90s ___ ~s ; 0 C ... (t+15)c
14000t 144X 112 ;;68000t' C n rete --u---
Total cost per sq. ft. =K=~~~ + i!_+l'S)~
368000t 12

From K to be minimum dk=O; i.t'._l_ X Ms+ !.C::l:O

dt 368000 t2 12

or tt =_l_xMs; 01' t=O'OOS1 /}'1s-,

30670 eVe
 2M

Q
it

Fi~~ 8

I!a
..

. 1.
 281
.
At Kalabagh Headworks s= 15'25 and C=O'4l
. 1'1.<;-'2' Xl jM=Ol0822x y'M-
••
t==O·C057X./ .-------
"U4_; v.1

Que;tlons.
I. Descril-e briefly and explain the working of any existin, work. which should be named, if p,",ssible, for
c:<:cludingsi]t at,
(i) Canal Head Works; •
(iiI Branch R~gulator;
(iii) A Distributary Rf'g lalor. (Pb.I.B 1941).
2 Show by ilimensional ~k .. tch~s a tie~ign for ~iJt coptrol at a Distributary Regulator taking off 300
.. use~l\ at right llnglf's above a fall discharging 200 cmecs. Explain how your design can be altered to TeduG~
proportion of silt taken by the Disty; offtake. (P.I. B.1941).
3. ExpJain any device for reducing the in-draft of silt into the head of a canal or disty; Also explain a.nv
form of dltometeo. IP.I.B.1938 ,
4 A channel is found to be silting badly in its head reach. Describe how the defect would be studied
al:d remedied if due to a defective head. _ (p. I. 8.1935
 CANAL IRRIGATION ~ .,..i' "

- CHAPTER XII

,~·e;.,~·f> . '"
Fluming and Headless Meters
~.~ .....
1. DefInItIOn. ,')u'm,.,<! ,'. '.
\.' . ~y! 1'1";" " ; ' . : .~_~ ._;.. ; .<,~:_~~, _. :~'
A flumed work IS one, huilt in a stre;Jm, the waterway of .which . is reduced below the
nor.mal. , Tt is necessarily accompanied by an increase in 'velocity. ,." ,,... ,;,'
This result may he obtained in a variety of ways.; , .
(al A reduction of width.
The discharge per fOr)t of the width nece;;sarilv in :n'ases. This may be accompanied
by an increase or even recluction of the depth. Examplf's of this type of flum{s are lined channe I
section, contraction at the aqueducts and the flumed bringes.
(b) A reduction of depth.
This may bp acc)mpanied by a cnmtant Width. Works falling in this category are weirs,
flumed or contracted falls and heaclJess meter flumes.
A fe"" examplf's of ihe clesign are given here and the 5tudent should refer to A M.R.
Montagu'g paper No 6 Central Board of Irrigation, Simla and F.H Burkett's Paper No. 125, Punjah
Engineerin5' Congress 1929,for a detailed study
2 Classification of flumes
Flufl1es my be dividerl primarily in to two main c1asses:-
Class 1. Flumes with' frep,"water surfacej.e,a water surface open to the atmosphere. In
such cases the pressure on the free surface remains constant.
Class J[ FJume!: with' sealed"water surface, i.e. of ~ hirh a roof forms part of the desigen
The eS~l'ntjal hydrauLc difference between these two types
rOTH I'tim' we is expillined l~elow :_.
In class I thp, surfilce is open to the atmo~phere and the
pressurp, on the surface remains cpnslant. A],o the area of the cross
o section i~ free to alter with t he change in velocity. Fig. 1
I Consequently.
BEO

J
Bernouilli'~ equation in the special fOJm Z+ D+
Fig. 1 Tf'mains sensibJy true if the effects of curvature are. nea;lected.
ENtRy'" But in fpe case of Class 1I the surface IS not free.
-----rj--.:.::.:..;~~:.::.l~=~_ Provided the surface pressure does not fall below zero
"'''TAL LINE

l'
- - - ._~/rY ..€.!!..£.f!.rj;l:_J,Ib.L_ (when cavitation takrs place) the above f'quation has to
be modified. (Fig 21 ."
~4 ",%)
r ~ y._"" R I I D !
:;n z
Considtr a part cle 0 funIt
' mass a t a d'/'
cp th D below'
~
l

,~I
the surface moving with velocity v.
\;,,,,1;7
'j. . The potential energy=Z-j- D-d
SID v2
t
z.
Kinetic energy=-.
2g
Let Pr be the pressure on the roof. Then the
Fig. 2 pressure on the particle=pr+d:. the total energy'
per lb.of the particle, measured in "f{et head of
2' v2
watet"=Z+D-d+ v_ +Pr+d=Z+D+ +pr
2g . 2g
 283

Since total energy remains constant (save for the reduction by friction, etc, represented
by the slope of total energy line), it follows that :- _
, If the velocity energy in 'feet he,ad of water" be plotted above the roof, the pressure
thereon is represent"'d by the intercept betNeen ,th~ ,veloci~y en~rgy line and the total energy
line. .

3. Bed profile.

(A)
.
General
.. . Conditions of flow.:
. .
; Water has Inertia. It follows tbat any sudden appli'cation of any extet'n:\hforce,?;lilVill
bririg the laws of impact into p l a y _ ' ,J l >_ f:'"

All {orce5 tending to a)ter'the, direction of flow must therefore be applied 'gradually if
mpact and .the consequent loss of energy are to be avoided. The gradual,application of It force is
followed by a change of velocity or dirtction-the stream lines are curvt d. - .
'.'. It w0:tIld appear tha,~ thi"i curvature may be increased by in<:r~a~ing the' external
'Impressed 'force upto a limit at present unknown. ...'' . .
~.' ,If the lImit be exceeded, thffniture of flowi~g ~ater exbibits the property of t~ying to
.c'prr:ect ,the impropC!r c 1lrvature, by tne' interposition of a layer of minute eddies, in the form of
'hprizontal or vertical rollers, The 'eddies consump, energy and the total el}ergv of the stream is
reduced. ' .
If the curvature be still further increased, the eddies are. no ~onger effective in correcting
the cu~vature of the boundary, tlW stro::am IllES b'e'ale down and pure impact results,
, . A vertical o':>struction across a channp.l.if! -t.~e, fo~m ~of a wall, will prolitlce horizontal
rollers both up"trea,m and downstream of the' obstruction", If the o\;lstlUctioq. is excessive, there
will be pl1reimpJ.ct, fl,gainst ,the wall and also between water pil:s;;ing_, oVl':r it and that moving'
slOWlY ahead of it, It implies that all changp.s iIi the profile must he gra.dual.
(B) The forces actIng upon a stream of flowing water ,are:-
(a) Gravity, ",hich is c:mstant in v~lqe< direction and sense. ,. .
, (bp<riction, the value of which varies with the area of the wette~.e_l\velope>. the velocity
of the stream always acts against the . dire~tion of flow:, ' " , , , , . , '
. (c) 'The reaction of the envelope" which (wheI1 sep:lrated from friction) acts normally to
the surface 'of the' en velope at the point. Its value depends 6n anum )er of factors. ,.
It is the variation in the reaction of tne en ~l<;>pe that causes .curvature.

(C) Ideal Bed ,profile,

The chan~e in ~ection should be such that the jet remains adherent so that vertical
compon<;;nt of the reacti,m at any. point IS t:qUa,l to the pressure due to the depth at tha~ point.
Although the qualitative eftects of curvature In the stream lin~s are well known, our knowledge
of the qu lntit~tive effects; t l1t is the '~qlld.tlon5 defining thp. fundamental laws, is practically nil.
Consider a r\ct ang l'ar , n chaunt:! of which the hed risp.s and falls while the width
_,~____________• .u4W~I~.~.·.7$.aLr~,w~·_______ remains con"tant, Ideal chang-e in section
would bp, in the fo:m sketched below:~
Fig 3, It is comb'nation of double S
curve to avoid all corners, 'The stream
-_ _ - - - -......_. " line upstream of the crest shall be curved
upwards connoting vertical component
~ of flow due to cl'ntrifugal effect. The reac-,
1 ion 51 all be l's5 tt an that due to the depth
of water at any point The ;;vater sur[aCfl
Fig, 3 profile sha:I J::e convex. The convexity
COU'iotes that the jet is not adherp,l1t and is liable to pressure inflation,., A very t asy. approach
upstream of the crest IS, theref, re, reqnired. On the down-stream Side the stream lines ar"
lonGave and, therefore. adherent.
 284

The convexity upstream of the crest, can be removetl (or rather renuc(':d if friction be
taken into acc"11nt) by providing an easy uniform negative
C'''l'sr
_--==__f
-" i
slope. J'rom experience one wo tld suggest a suitable slope
fi}'
from I in 3 for low velocities to 1 in 5 for high velocitif's.
MD _ $.C_ _ Tn thp case of the uniform slope. the corners need to be
transi tioned by introducing short circular curv"l tangential
Fig. 4 to both plane3 as shown in Fig. 4.
The upstream curve to crest with radius 2H as introduceo by Cr,.mp in meter flume
nesign looks, no doubt graceful, but the e.ffec~ of c~rva!ure i~ prcdu~ing Cfntrifuea) efff'ct is very
prononncei as calculated by the author In hiS artlcal In Tndlan Engllleering (alcutta December
1935, On the downstream side S curve shall t-e ideal and the iDclease in d~pth may le arranged
even in a relatively shorter distance liub-critical stream and hypercritical stream for type
Critical stream, the Montagu's design of bedprofile i! ideal. (Para 12 Chaptu X).
4. Approach or contraction upstream on sides
In this case, a curvamre on the side, will no doul,t intronucp. the Cf~ntrifugal effect which
usually result!' in heaving up of the surface about the middle of thf' stream and which also
accounts for the concentration of the silt charge about its middle, but the jet is perfectly adherent
to the sides due to the inertia or the momentum of the approaching water, There is no chance of
the reaction against the side-wall being less than that due to the depth against it. There is r.o
possibility of the forming of vertical rollers which usually occur in the expansion downstream of
the crest.
The design of the approach should aim at the following requirements of a suitable
approach:-
(a) Thue should be no sudden change in section to produ.:e the losl of head in f'ntry.
(bl The change in section shoulcl bP. gradual to avoid clirect impa~t against the wall.
There is no mathematical solution yet possible to work out the conditions necessary to

. ~ ~' " fullfill the above-mentioned requirements. One

I-~ ..J /"
-........... ··........... ··..... "1
- ' , - , __
. '
. .. I
could only say that to avoid impact. the forward
. component 0 f the incommg. water shouId be equaI
.~ to the centrifugal effect so that the rea ~tion
I.~
•

. /
~
J
--....
-.. .
---'
-:1'
!(If.... _t_
_
tit·
I
agaInst the wa II is not cbanged H owever, .rom
f
expt-rience it can be ~uggested that a splay of
---, 1:2 for low \elocities and 1:3 for high velocities
• .. " will do for the approach on sides. In plan it
. --' C,"N~._ • _.- shall be ot tne form as sketched in Fig. 5

~n ideal cur!i~;uld be with infinitp. radius in the beginning "Ind the end as shown in

Fig. 6. However, it is not possible to

design such a curve which may ce lllid
,..,.".
.
out with precision hy 8n average
Engineering Subordinate. It will be
enough to design a drcular curve as ~ / ! "
shown in Fig.6. ~ ~ ~~
The radius of a (ircular curve • \ • I II I~.
can ea"i1y be worked out assuming it 10
be tangential at the point where the
..~ ______ . _____C~.~IN'
•
#-..--._--
contraction is complete. The curve Fig. 6
needs to be transitioned in the beginning
for a short distance, as shown above. to avoid loss of hE ad in entry.
5. Departure or Expansion on sides downstream of the contraction.
(A) Expansion and dispertions of a stream of water is another Hydraulic Phenom~noD
which has recei'led the scantiest consideration of mathematicians. Even experimentally lltt16
work appears to have been done on the subjed.
 285

An ideal expansion wOl1ld be when the stream is just arlherent on the sirles. If it is l'ls<;
than the ideal, the jet will no doubt be arlherf'pt on the sides; but thp. work will be expensive. If it
is. more rapid than needer!, the jet will not }Je adherent and low presiure pockc[s near the side
wIll cau~e the formation of vertical rolh-rs with the consequent loss of energy and with the
eddy-mJtion causing scour on sides in the earthen section as the stream leaves the pacca work.
The reverse flow on sides is the clearest inqi~tiol1 of a too rapid expansion.
[B[ Forces acting.
The forces acting on a stream flowing in a section are as given bclow:-
[iJ Reaction.
Thp reaction of the 'led and the sirles is opposed by the internal pressure of the water
which can easily be romputed at any point of the perimeter. [n a rectangular section, the reaction
aRainst the bf'd will be equ,1} to the-pressure due to depth, and against the sides it will be variablll
according to the depth. In:tn kleal expansior1 the jet should remain adhere!lt to the side. Any
factor tending to increase the bed reaction will help the lateral dispersion.
[iiJ Friction.
. The effect of friction can be depicted by giving a downward slope to the total energy
lme, The effect of friction on total energy line elm easily bp. calculated. Friction always acts
aglinst the direction of flow. It may bp. considered as a component [resol"ed along the bed] of
the reaction. Its effect is to ileflect the reaction against the direction of flow and the total value
of reaction. The net effect of friction is to ittcrease th<' rate of lateral expansion. In a frictionless
channel. lateral expansion of the jet or oispefsil:m will be nil on this account.
[iii] Gravity.
A mass of water [M1 moving with a velocity v has a momentum Th h
gMv e c ange Of
momentum per second is a measure of the force which a stream is capable of exerting. Any struc.
ural change in a channel section tending to retard the inertia shall produce a change in the inter-
nal prpssures and the consequent disper~ion of the jet. The in tertia acts in the line of the direct·
ion of the motion of the mass. Velocity on the line of floN has no bearing on the inertia transverse
to the line of flow unless there is something to change the line of flow with the consequent chan-
ge in the internal pressures
Tn a hypercritial jet leaving a crest of a free fall weir. the forward momentum is prerlom-
inent, and there is no lateral exp'lnSiOll, even though the side be suddeuly removed. There is no
lateral expansion of the jet, if it follows its natural parabolic path, but if the bed profile is flatter
than the parabola of the dropping jet, there is lateral depression because the line of the velocity
has been changed which results in redistribut:ion of intt:.rnal pressures. The dispersion in hyper-
critical j .. t will be very slow.
If the stream Ipa'ling the crest is subcri tical. the parabola of drop will be steeper still
with this low velocity. The departure introduced by a slope will be more pronounced resulting in
the relatiwly greater de pres ,ion. MoreovPr, the forward force due to momentum is in this cass
'k:ss than the side pressure, the jet will exp:n1d adherent to the side. Ml1Ch more rapid expansion
or dispersion of a subcritical stream is, thl'refore, possible.
rhe water surface profiles are governerl by tha conditions of flow as described in Chapter
IV Part VI. The effect of friction and the bed slope flatter than the theoretical parabola of drop-
ing water is essentially to disperse the jet laterally. The condition of flow may be subcritical or
hypercritical with level. downwa.rd (positive) slope and upward (negative) slope; but the expan-
sion against negative slope cannot in practi~ he arranged unless the surface is curved. The disper-
sion necessarily means the reduction of tbe discharge intensity with the consequent reduction in
the critical depth.
[C) Conclusions.
(a) The greater the discharge intensity, the higher the \'elodty artd the less the dispersion
in both subcritical and hypercritical coI)ditions. A stream flowing with lower initial velocity
 28~
will expand its cross-sectional area by a given amount in a shorter length of the work than the
stream flowing with higher velocity, other things being equal. . ...
.. [bJ fhe flatter the bid slope in the outfall, the greater the dIspersIOn In both the
subcntlcal and hypercritical flow. . .
(cl With. hypercritical stream leaving the crest, the radim of curvature .In the, !)utiull
should be infinite at the beginning of the lateral expansion. The radius .of curvature may
decrease as the expansion proceeds dep"nning on the first two conclmions.. .
A straight line expansion, howe \'er gentle, i~ not satisfactory, because a stream will not
adhere both in the beginninO' anrl at the end. In the caie)1 su'lcritic'll flow, a ·circular splay
tangential at the contracti~n is likely to serve the plrpJSe, if a decreasing. radius cannot be
araranged. .
(d) Il'l the cas. of flumed bll" the expan5ion should bnil or very slow upto the jump
of the ~lacis, but may be rapid after the jump in the subcritical flow. The rOjJghening of
the bed by artificial. device such as· friction blocks, will help thA rapid di<;persion.
. (D) Length of Expansion.
. . (a) The f::lctor that remains unkown is the lene-th. Owing to the existing· ignoraJilce
of the manner in wh;ch the varyinJ pressu es act on the inerria of .1he m.S3, the accel~ratinn of
the lateral expansion cannot be predict([d at present. Experience ahne must be the 'ginide in
determining the important factor of length. The departure splay is of t~e. order, varying
fram 1 : 5 to 1 : 10 as usually allowed in practic'll dcsigns. .
(b) F.R Burkett dealt with the question of expansionfrom considcration of pressures
a lno in hispaper No. 125 Punj ab 'Engineering Congrpss, Lahore, on headless meters as given
btlow;- ., .
':Now it appears obvious that, a ch'lnge il~ the direction of fhw along the side wall.is
used hy water pro,sure at right angles to lh~ lattt~r. It should, therefore, be our aim to keep this
pressure constant and as the ve10cllY of .vater drops, the radius of curvature of the side wall
sho~ld decrease. As shown in diagrarn 'Fig 'j! RA2 should remain constant, ,where R is the
radIUS of curvature of the sine wall~ ana A is the area. of waterway.
"If the floor downs tre'llll folhwed a cons t an t lila pc' a du'fcren tial eq nation could be
obtained connecting dista.nces nonwstream of the crest with off-sets to the !Dide- walls, but in
practIce, this i<: not po.;sible, a!> tp.e qlhstioIl, Qf.critical velocity upsets the uniformity of the Jl90r
'gridH;nt' .;.. "''0 1 ""',',.. . . . ..•• .. ..

)( -"
SIN ¢,=_ R, 11 ,

51N tfJ. 2)(- (If, - R.) 5/~

:& 'I. "
- Rz

SIN

y,. •

Yz .. l( TAN~
z

. . Fig,7
Fig. 7 5hO\, s how these walls may be la'd out for any initial radius."
IE) The Aut~l()r'S lVIet!1oi . ' .. , _ .
(I) The rate (if expan ion is to be d~temined from the graph give;) in Fig, 8, Th;s will 11'(
1he length of the outfal',
Tn t'lt~ Ca.,e of Ilbcritical flow downsteam of the contraction, as u;ually available ill. thfl
ra'!~ of flamed I I :dgos or hearlldSS meters tlw expa,!lsion may be circular, tangen ti.ll to the
eo[,traction •.
 287
'$4
_r r- -r , 11 , I 1
...
(l
H-I--t +tt
,. . . I J~
~
i -~-q-l-'r-;
I I ~"-+-~I-+--'
;;;
:t .25
~-+-I- ---i+
I

i '
I
1
,-+- .-rt---~--'':;'~-t-r-T
1
t-."
~
-t""H I N r
I '

I I,
; ..S " ,

('i'...r.
i

.
\I I.... (
...
"'t

-ri---- 3±1tfi~f<~~'
ttrn
~
h.:':l'"

;W+r'
(l

It
I-f--f-- -
r- 1 I l_j__ ;'~11'rl<'r.,.iy I r:~1f'(' ,
\) .21t

. _' I ''d+ .
;;:; I I ! J 'Ifl ,_I_L-, I I~~
HI - --" - -- I 1
'1 t-LJ .__r·..c~ i (' ," l--r, f

+-L.;~JCMTr]::p;j
~ _I_-l T
~
~
IT H "
-I ; t: I·
">---1-+' rt·_ /YI;;. I
'~
I i
'1',
,. ! Fig,'8 w.

.,5
,
I i ,-ti-(I- - "r;--'- i I"f-:'I_rlr---~
I I ,I I , ,(

I
...J
"'(

.........
Q::

,
. ~'~
r
+-l--
:"k
'm:
~tt-'!"k.:=r- :Q=I_' . ,
~
I r,
11',
i
,I,
I
1
"{
LL
c i -n
I 1,1
~ I'\.
~iEEE_l ~
.(0
"-
~
, -- -- )',
~ )
!If,;!' •' . t- , 4."; t i·e.1
"l-i·=R--+=:--+ '\ ,",
.{;I,~':;l
~
I .,
I i I II
I i I
"\I
1\ .t-
Jt$

BED
" '4
:m.:)1
j1

In the ~as~ onhe falls wh,en water leaves the crest with hypercriti0al flow, is then
foll()w~d by a jump, on the ~lacis a!ld listly, t.he subcritical-flow, an ideal shape of an expansion
is Bernauillis Lflmmiscade as
used In the d p slgn of the roof block )f the A. P. M. outlets. . It
has lot ,\-I-n infinite rad ius in the. bf!ginning and then it decreases towards end of the Expansion
as shown in Fig. 9 below: '
The curve recommended is the one f()lmed by bending a uniform flexible strip of some
homogeneous material (such as (~teel or celluloid ~
in the manner indicated above, Knowing the
length of expansion as determined by its rate
of change as fixed beforehand, the curve~cJUld
,'"
14,:'0 J td it,' l~e drawn in a plan to 3. reduced scale ,nd th,
percentage df pr-rture read at suitable .distanc~~s
\ .. .. from the beginning lind then accordmgly laId
. - - - - - . _ _ _£'.1J.!i£_. _ _ . _ _ _ • _____ at site
Fig,9 (ii) Flaring. .
Tn flumps, long vertical walls are Df)t only very expensi ve, ,but also ,they cause
consi<1erable reduction in w"terway pven w'hen t.hey are 't uilt' at -the toe of the side slope.
The usu;:;,l sid@ ~1()9" allowed in p;tchi ng is t to 1 in distributaries and minors and 1 to 1
in Branch and Main Canal~. The chAnge of vertical section to the normal means considerable
contraction upstre::tm and expansion dnwnstre<\.m. J t is. therefore, usual to fine the sid€ wall from
vertical at the crest to the permissible side Slot e at tt·,e he~inniT'fi of the contractio-l'l upst'eam
a;q,-d that at the end of e:>ipans;on dowDstrearrl, The typical sections of flarid walls are ;shown
~ p~~ XIX_ :
The sections (h), lC) (d) as comparocl with section fa' show considerable saving in the
sectio. of masonry and concrpte. The saving of cost in a long wall is about 40%.
7. Example No: 1 Flumed -\queduc;t.
A channel with the followin~ data has to pass in a masonry aqueduct .flume with
v€rtical sides with N ='013 with no loss of head (with the slope of the channel), De~jgn the
fluming. . . .'
F. S. Dis?harge 70~ cusecs: Full SUPhly level 549'6 Bed.width 60 feet; Sl()pe 1 ~D 6666 ;
:8ea level 594'S ; Slde lope! to 1 ; Depth 5'1 ft. ' . ' ... .
[:tJ Assume width at the contracted section as 42 feet: Sectional area=A=S'l X 42=
214'2 sit.; Perimeter=42+2 X 51 =52'2 feet.
 288

A
R=H . M• D . =_~_ 214'2 . f--.
---:::041 ~.
P 52 ~
Using Manning's formula:-
104~: X4'12/3X(_1_)l{ll=-3'3 ft./sec,
v= 1'48;; RlI/3 51 /1 ;=
N I~ 6006
iat Dis~harge=42x~'1 x3<~=706 l::auseC5 which is nearly the same as required. 700.
of contraction up<;trealb allowing 1 ; 3
tb) Lf'ngth
FLUMED BRlDGE-Srale IllO()

=-3X 6~-42 =27 feet; Radius of circula.r C!;)ntraction (2R-9)9=27l :. R =394 feet.
2

(c) The departure downstream I in 5=-,,5 X ~~2=45'Radius of circular expansion'

=(2R-9;9=452 :. R= 117 ft.

Jq> The walls shall be flared both lJpsirt'am 31}0 00 v.ostnam to } to J sjde sJl>j)I'A9 J»
Fig.IO.
S. Example n
Flamed Bridget (Montagu)
It is required to flume a ch~nn~l with the d 1ta given below withou t txceeding the
critical velocity to a width of 10 ft. whIch is fixed fOT an existing railway bridge. Find
the fevel of lhe bottom d girders allowing a freeboard of 15 ft. with 100 as bed level of the
earthen channel.
Di~charge 250 cau.ecs; Slope 1 in 1000; Bed 25 feet; Depth 4'1 feet; Velocity 2'3 it per
seconci: 1 to 1.
CalCUlationS.
LbJ Throat under the bridge with vertical sides. the discharge intensity=250/10=25
cuse.c~: Critical depth=t1-i;r- =2'89. Th~ actual depth must remain well above the critical by
at least 10% say not les, than 3'0 feet.
[c] Up~tre'lm side contraction 3 : t ft. Let it be 1 in 3 at bed level; The length of
< 25- to
upstream approar:h =3 X _~_-=22 5 t
. f

Rarlius=f2R-75] 7'5=2l5 2
:. R=37 33 feet.
The wall shall be flared t to I slo~ with I foot free board.

The radius of top of well; contraction each ~de=7'5+ ~d-=10'05 feet;

. . a. _ 22 ..:;2+ 110 '4 615

R=«%R-·-IO 05) 1005=22 5 .. R - ------:rot--- =-2[--30'0 feet nearly,
(dl Downstream Deptarture.
Let the !'ide splay be I in 5 at bed level;
25 --_
Leneth of oJ,ltfall =05 x -- -10 =.37-f
.;) t.;
~ ~

Harlius at bed 1~vel=.2R---7·5 7·S,,,,3752 :.R=97·13 feet.

The wall shall bl" narc!c t~ h:.l.H !:::: or.;:; slope with 1'0 it. freeb:mrd. The expansion at top
 289

level=7'5+ - 4'1+1 .
- - =1005.
.....
2
The radius at top leTel= ;2R -1&1'05) 10 05
_37'52 .' ,R=75'2 feet.
-
(e) Throat length,
This depends upon tqe straight portion .
required for the bridge,let it be 1(1 fpet
Three mp.thods ,)f solution are po~sible
!
'!
in this example, only one of which will be
illustrated here'
(0 The bed may be sketched in a.nd
wing wal's designed tlaerefrom,
(ii) ThA wing walls may be fixed and
bEd designed therefrom.
iiii) Sketch in the water surface and oesign
the bed and the sides.
o It is prol osed to illustrate the 2nd
method here.
(f) From Efs Diagram of Montagu
Plate Hr.
Discharge 25 cusces; Depth of 30; Ef2 4'08 ft,
Total energy level in the open channel
v'
=100+4'1+ - - = 10"'1
2g
2
, +23 = 104'18 Of--
2g
Bed level ill the throat=lOt'18-t'08 =100'10;
F'S'L'= 100'1 +3 = l031
The hed need to be raised by 0'1 ft. oaly which
is tri !ling
Now work out the design with no chang!> in
b~d level.
Read depth for Ef 2 =4'18; Depth=3'28; Water
Ievd= 100+3'28= lO'j 28 :
The botton'\ of gilder= 103'28+ 1'5= I04'7~
(g) Friction bss.
...
0. J

---WJ-:-·,1 _-
Velocity in contracted section= 25 =76.J It
3'~ •
per second; '-..-r--f-- ~~

Velocity in the bf'ginning=2'3; Average veloc:ty

... - ..
f

764+2-3
110:::-------
. 2 =50 ft, nearly .
~

1
H.Y.D. in the beginning; Uea= (2'; +4"l) .'!== 110'9 say 111'0
l'erime.ter .... 25+4·1 x 2 yll+5'j -3·H1; R=!_!LQ- -3 25 ft.
3~ 17 °
 290

10;< 3 28 3'l:~8
H.!t!. D, in the contracted 5 'ction= - -' - -, = --, = \.97 ft.
JO+2x3'2816 56

A
nverage H 'MD
I' ' =3'25
- - -\-1
297 52'2
- =-2-- = ""61
/I
ft,,' For N='OI5 J'n Manning's furmula w;th
'l

R= 2'61 and v=2 0 from Nomogram PIette XIV S= _1 __

]400

" 'IiI
' approac h =
Loss m

37'5
--=' 0 I 6 f t.
225
14UO
-
U1 11

Lo~s in outfall = ~ ='Ol.7; Rate of loss in throat for R= 1'97 and v= 7'6~
1400

1
S= 6C.o X 10,-'016 ; ~et 1055='06 ft, 5ay 1 ft.

The downstream water level may be shown 1 ft. lower. The Totahinergy liDe will have
slope of 1/1400 in approach and outfall and of 1/000 in the throat,
This slope is f\\ther very small and there will be no change in the design.
9 Example III. Contraction with depressed floor, (MontagH)
In the above example, contract to 5 ft, width with vertical side walls (A, ~f, R
~
.Montagu's Solution C. B. I. publication No, 6), Then mean velocity enfrgy Ev= "2g =0 OS2;
~fean value:,f Et =4'182, Assuming the win~ wa.l{s of thp. approach and oeparture to be vertic,d,
I.e, the sectlon rectangular. then on the section at entrance to apPloach tran!'.it)on, Fig, \ ~ ,
Qo=250 cusecs: Bo=25 feet: :,discharge per foot width q=lO cusets;
E t =4'182; :,D o =4'09 fect; and Vo=2 44 f P.t pef second,
The Throat Assume a flurning ratio of one-fifth width of throat, Bt=5 fect; Discharge
per foot of width, q = SO cusecs,
The velocity is n)t to exceed the critical ve~ocity, Exrerience dictatfs t'Jat the designer
5hould k-eep well cltar of the critical point.
Examine the curve appflpriate to the value of q=50 cU!'ccs Sp.!ect a point thereon,
'...·ell cl,~ar of the locus of the criti :a\ puint. Let ~he point selecte(\ have the co-orclin"tes.
Dt=7'O; Eft=' 7'8
Then tht line of the ted may be set off downwards from the Total-energy linl, 7-8 ft.

qt
anj tl:e line of the surface u~t up froUl the bed, 7 feet; Vt= -_ =7'14. This is satisfactory
Dt
The TraD.;iti.DJ ;-Begin by assuming an approach transiticn of ];2 splay in pia-no
Its length will be ; - L a = Bo-:-_~t X ~"",20 teet.

..-
'2 1
:. The slope of th~ bed from level of normal channel to le vel of the throat wi}} be
Eft-E{c
- --. ___ = 12 %; Let the :length
' of tbroa t L t= 10 f eet.
La
I ' , "7"1:"- ". ". , ~,. c

The design (resumed) ..Refer now to FIg 11. Assume any nUin1:er of sections X Y Z etc.
on the plan and elevation and proceed as follows, t~ construct the tabular statem<:n t shown' at
t ne bottom of the plate, The total discharge Q remams constant.
 291
Measure the succeisive bed-widths B on seection XYZ etc. ano record them in the
table.
The di5char~e per {not width q at the successive sections equals Q divided by B at each
section. Calculate and -record. ",
The value of E f appropriate to each s ction is determined by the intersection of the
0

appropriate q curve with the "curve of transition characteristics". Record this.

The depth at the point is read off tile' Hydraulic niagrams at the same time as Ef and
r~corc!~d
The R. L. ()f the bed is determ;~ed by subtracting the value of Ef kJm the R.L. of the
total-enugy line at the section. Fig. it.

FLU,.,,£D 8/tIO&£
.,
(1 W

'''- . .:
If
;
'{.""
I
Z
,, ,tT
.',
,
T,
I
:,
,,
,~<t'
I t-~ .. , l
I

I PI-AN ,
, "-
~•• ;

.. , '}- : SCALI:::
,
c. :L, .. ~--t-~,,: ,
, ' J I

,
:
,
, ,,
,
1
!
I
~ IEL£VAil01'4
1:sc.. lEr..... :.
'ifOo :
, I I . l"e"'::- ~ y#~ I l l ,...

..J"!:M!·=!/!9f=FT~·!:!1fL='=N!'~:'::;:::;:~'4~=EE~::::=;:=:;:::~=.=£'~'::~='~H~.'C:'==F:;!I::!;";;4~.~~.r..,
- I i 1 P ,... " ; ::
'I l, I ~ ~ 1
, I: I
! :: : I f
,t :~,Ofl,
I
I
,
l
I
I'
I 1
:! -ii; r--- : ; :~
I L I ~i" Fig II
t
: : .WD r : : ~J1
I , , II J

t OATA • ,
OISCH:Q'I-_ __ - - - - - 25" r;:;;;;'£C5 ---- - ---.---1
.Bllel .,
=~ lto
... .......
_/Ot'll

.. 0
; ~
......
,• ;... ..... •• ..."
'It
...
...;...
.. " ...... • .,•.
.. . ! : !! ! !
... .. ,; : ....
... ..•... .........
"!
t
to. Example :'1. Hel:lle3s meter fiun1'3 (Burkitt's paper No. 125 Punja) Engine-ering
Congrfss)
(a) The theory is sim[lle. ff at two sections in a strca:u dose t06'cther, where ~here
are no sudden changes to upset a str{am t"ne flow, the breadths are b o ,md bl and the ce:}ths
Yo. a.n? ~ l' the dtpressi~n in water surface between the tWI) sections b':ling h, then tile djscl1~rge
Q IS given by the eq_uatlOD:- -

Q=8-025 ho Yo Yl 't' ~'bo2~·~~2-'.

. ~2 • Yo
21 t -,h.
 · ,'- 29%

For:- v I_V 1=2gh' ° Q! _ Qf =2gh' ° 0Q! 2i!h AoJAl'

1 G ' 0 ° All! A02 ' ° Ao2_A1 1

:.Q=8·025 b. Yo b l YI { -b 2 2 b 2 11
o Yo - 1 Yl
h }t
... 8.025 b o Yo Yl 2
h l.t
Ibb~2o 'Y0 -Y1 2 2
\

(b) Hearlless meter at R.D. 103076 Mithalok Distributary

F.S. Discharge=3~'1 cusecs; Bed level=6~2'ol~

62 ' ;)
624'85
F.S L'=624 75' Allow a loss of 1 ft·

Bed width= 10 ft. ; Sides t to 1 ; N S. 622'74

Velocity 1'3 ft. per second; Depth 2'7 ft.
(c) Calculations ;-
Velocity of approach= 1'3 ft. per second
v2 1'3'
Velocity of approach head=ha=~= _ ='03 ft.
2g 84'4
, ietthe crest be level with the downstream bed; Depth upstream=Z'S It.
Depth downstrl'arn=2'7 ft. ; Drowninng ratio=~_:_~=='964
2'8
The co-efficient C for a drowned weir (rom Fane'~ curve Plate tv for this dtovmi1lf
ta.tio==2'S2. Let B be the width at the maximum contraction

~: ,/ .; ~~-

fd) Width opposite upstream gauge for nonsilting conditt. .....-'·-

Depth*28 ft.; vo=.84 d "='S4x2·S· '4 =l'63 say 1'1 ft/see, . ~c_
Width=Q-= 38'1 ==8'0 ft,
Dv2't; ~~ 1'1
There will be one gauge where widt.h IS 8'0 {to and one where width is ~'5~.
(e} Draw·oow:n to upstream gauge.
. - - ft, =H; Let depth at the 6rst ga-uge->wen~pH; The drdjY &;
Total energy aepthz:2 8 +'03=2·83
-i - (l-p}H ; .; Velocity=v' 2gh1; DIscharge ptt foot
- /-"-' .

_pHxy!2g -v
(1-pJ H==p. Vi-po
/

.
Pv
.Jl-~==3S·r
s Xv'2g -x2'83
1 3 /Z
='1265. The solution olthis' equation is g:ot (rom ffiowli

curve Plate IV in Punjab Eng;ineer:ng Congresg Paper 125 1929; but it can eotsily be If){ tty the
use of slide rure.
"Set the re-v-ersf'fr sIi'de to '1265 ot'! the b:
scale anrl then read under the cUrs~1' 6ft the'
II- and C scale 80 that total of p and I-p is one,"
'.' I-V='011 Draw down=h1 - 017x2'S3'=.0476 ft.
(f} DrawootlowB to s{cond gauge.
 · 1.. arge In
D ISC1J • • as'1
t en 51 t" = ·--cusecs per ft. Draw ·down =hs={l-p)H
J 3'52

- 38'1 1 'd .
'Where H=d+haa::2'83 feet; py'l-p= 3 S2 X v2g -tI- X 2'833/3 with sli e rna,
solving
I-p-·I03.·. nraw·down+ha=2·83+· 103=29 ••
(g) -Depre<:sion or difference between the two gauge readings
h=h 1-h 2; Depres<:ion='294-'047a= 247
This will be the difference in the full supply conditions the discharge for other . obs~rved
values of depression 'h' being according to formula in raJ above. The design is shown In FIg. 12
11. Example V. Ra~ed crest headless meter Hume. (Authn(s]
In the last example, tne width at both gauge· wells should be the same: but the crest
be raised at the second gauge·well
. (a] Let, width=8ft. o?posite'upstream gauge,
[b] HeIght of crest
The depth upstream approach=2'SO'
Lpt CTfst height be R
Depth on crest= H-R=2'SO-R
Depth downstream 27-R
Drowning ratio= 27-R . Let R-t'''.
28-R ' ..

Drowning ratio=2'~-I'!' '928

2' -1'
e from Fane's curve=2'9
. < ~eb\~c~~r;::t2'91~48~'1'~~:~:'3 c~
ft. O. K.
11an(~~l~ 38'1 '~W~'~:'\lien
,.m: " ,.
R... l't .
I" , •

-
-4-'--

.j
/
 294
(c) Draw-down to first gauge as bdore='0-l76, Let depth opposite the second gauge
1
be pH where H,-.:}'43 ft. :,PV - p = 8.~~';X8Xl'43~/2
Solving with a ~lider u1(~ I-p='2 \ , : f
Draw-down to s;c0nd gallge= 2x I 43=':?86 ft.
,dl D"pression in thp. design. of full supply h='2R6-'0476='238! ft. 12. .i:;{H' ...,
12' E J ample VI. Authors's Improved Headless meter Flume. ;:j)
(-\) DeLcts in Burkitt's Headless Meter t~ lumes.
Ii) BUlk tt's Headless Meter Flume app:ies to vt'ftical secti ms and therefore, entails
expenditure of heavy wall sections.
(ii) The correction for vdocity of approash is rather very cumbernme, only p03sible by
trial and error from th~ gauge rearllng, of the first gauge s;ection.
(iiil Expensive gauge-wel's and ina curat~ ganging.
(8) 1mprovcmen t in author's improved Head les<; I-Iume.
A flurr.e of this typ@ was designed and constructed by the author at R. D. 1206:~5 of
8hudinala Drain, UpPff Jhelum Canal comprising the followin.g speci'll fCatures:-
(a.: The hydrauLcs is developed to suit the trapezoidal ~ections so that the sides need not
')e vertic:!!.
(b) The op.c:ign is v{'ry che'lp with only pitching for the side walls.
(c) Thf'! gauge wells are suitably locdted in the central pier divided ill t .vo parts. 0116
,:-ecorrts the gauge at the contracted s~ction, and the other upstream onl'! is s,o designed rlS t,_)
teeOf@ the tieDth and velocity of approach head serving as an open Pitot gauge LJr the
'lDstrcam Sec:iOIl.
(d Tte correction for velocity of approach is automatic.
(e) A difierentirll gauge was installed in the fom. of inverted U tll'oe ::lnd water
~'as suc:ked ur from both gallge wells to glVtt a direct reading of the differences tl-!at is,' H'
nea;! causing floW b"tl"een two section, .\ vernier ~cale lias provided to give thl'! rea,lino
correct to thinl pl::lce of decimal. 0

1/) T1.e formula of Discharge is very simple. If H is the refad:ng of d;fferentiaJ gaugp.,
ri.e dischargp Q=Ay''lg I-{ where A is tHe sectional area at the c()ntra'~ted section.
(g) Data. ./
l' S Channel Data.
B= 19'0 ft. ; d=7.0 ft.: s= }'5 to 1 ; Q = 700 cu<;ecs.

A=/ 0(19 0+lC·3)=7·Ox29'5=20S'5 sq,ft; v=2 ~~~ =3'39 ft/sec,

<~-. r':- =t='--~~
L=fF-

"'. -
-.~/~.{.'
- .
' -

~---. ';'r- .,,~:, .... "1

v2 :i':i9~ ~
h =~ =---- =0'178:> it
d 'tg 6'1'4 .
Flumed seetion:- Design £Ol F,S. afflux of 0 3' Try F = I 0 (R.ise in 1ed)

A=x(B+x) . and w=B·t2x,' ',h= A_=_x~*x)

, '2w 2i:3+4x

v=.J"2gh- =- ."")'}:") y'h ; H=x t-h; H+l·O=7.3

TO,178 . J'fi8; x+h"=6478 .:
Fig. 13

) .,..h·=S·478 ; i.e., x+x__:_~+x =6'478; x(l +_~±;<-)~-o'4;8

, 2o+4x 2B+4x
:-18+5:<, ),,8/l(B+x)3/l 7110
!('-fi:f+ tX G'478 (1) : ---~V2~:; 4x - 8'025- (2)
Fro:r-; d) 3Bxt5x2 = :'05: ' ,.:o'912x ; B(3x-12 956)=ix (25912 -5;,;)
 295

2S'912-5x
B =x ,------
3x-12'956
-
(i) Let x=4'6; Left hand expresslOn=-- -
, 4'A~n2-9S6-2X4'6,312
-________ =182'5
31<46-12956

(U) x=4'7=Left hand ex-m"sslon= 4'7'112'PS~-2 '>(4?__,!_lz =129'4

. JX47-u,t!:6 _

(iii) x=4'S; Left hand f'xpression = 4~~~(1_3__~_5~=3.~_i_:~~= 98'(

3.-<.48-12'9:6

liv) x=4'716; Left hand expression=4· 7tR2(1'1956-2x4716r_: =123'4

__ .. • J ,,4'716-12'956 , •
25912-5x 2332
B=x - - - - - =4716~ -=924 ft,
3x-12'956 1'192
(C) Solution x=4'716; B=9'Z4
A=x,B+x) =4'716 (S' :4U+4 7 \6)=4'716x 13 956=65'816

w=B + 2x=9 420+9432= 18'672; h = ~

2yv
= ~16_= 1'762
31'344

V=V2gh =8 0'15 x 0-762= 10'65 ; Q==va=IO'65x6S'82:-;-701

,.
CU$eC1l.
H=x +h=4'716+ 1'702=6'478 correct.
Variation of Working Head.
Assume Manning for C1_ annel. .
i.e .. v= ~~~_R21~S1l2 ',i e VXR2/3. v=K.R213
N '
Q=700; A=206'S; v=3 39; P=19'O+3'OOBX 7 'C=19'OX2S'242=44'24
R= A = 206'5 =4'67 R2J3=2'794 ; K= 3'39 =1213; v=}'213 R2J3
P 44':B ~'794
'~t
,,,-I
TABLE I
d. A P R Rt v Q
2 3 4 5 I; 7
:,w'_)U U 93b6 23;3 --~--
l'u 'l'l6l 0907 1'136
20 4400 2621 1'679 1'413 1'714 754
30 7050 29'82 %-164 1 775 2 ];;0 1516
-40 100,0 3342 2992 2079 2'250 252,0
50 132'5 3703 3580 2'313 2'840 3764
50 168-0 40'64 4 )30 2575 3,124 525'0
70 206'5 44-24 4,670 2794 3390 7000

(D) Section opposite U, S. Gauge-Well.

The section is arbitrarily selected to Ie IS feet wide with slope 1 to 1 and the floOJ
level, viz" one foot raised above earthen channel section upstream with the object: that the floor
. remains clear of silt undp.r all conditions of flow Actually the discharge is obtai!.l_e,d independent
of the measurements of the sectional area opposite upstream gauge-well unlike Burkitt'f
-headless Meter flume Fig, 14.
(El Flume Conditions.

A-x\9 24 +x); W=9 24+2x; h==+w; v-S'025v'i1'; Q=vA

 296

TABLE II
A v=v' 2gH Q
w P'24+x A h=-H=x+H
2w
.,
~ d 4 5 6 7 8

'5 10'24 974 487 0'238 0'728 3'914 19'0

1'0 Il''ll 10'24 10'24 0'456 1 456 5 42 55'S
I5 12'2t 10'74 16'(1 0658 2I5S 651 104'9
2'0 1324 11 24 22'48 0'849 2849 739 166'3
2'5 14'24 Jl'74 2935 1'032 frt 3'532 H 8'15 2392
3'0 15'24 12'24 3372 1'2U5 4'205 881 3232
35 1624 12'74 4459 1'373 4873 9'41 419
4'0 1724 13'~ 4 5296 1'536 5'536 9'95 5n
45 18'24 13'74 6181 1'696 6196 10'45 646 ,.
.. 71fl rR fl7 I~ q" tiS 82 1'762 fi ~7S 1065 701

Working head=floor level+Column 6 of table 11

-(D.S.bed leveI+Colum 1 of fable I)
The flume will work in free fall conditions upto a discharge of 323 giving a working head
.t 16%.
Galibration :-
Read H from th~ differential gauge installerl in open pitot GaugeWell which registers the
",ater level at the first s. ction plus the velodty of approach head and also the Depth "x" from the
(ange rf~ding in the Gauge Wtll B. Dischar€e=(9 26+x) XV2>;H CUStlcs.
A difference of gauge reading C and B is denoted by G; the formula for discharge
?ecomps with H=G+h
Q=R 1 "-2V':!gH
y' A 1'l.-Aa 2
13. Example Vl[. Weir spillway sYp'hon flumes.-Fig. J5.
Cl)nversion of R. Culvret @ R. D. 54,600 of P.R.K. drain into a weir spilway syphon
10 carry 200 cusecs.
Data.
Section of the existing culvert is as shown in the sketch.
8ed width of drain= lS'J it; F. S. depth=4'Z ft ; Bed level=73Z'07 F.S. L.=:::l733.'Z7; Length of
:ulvert is 47.4 ft. ; floor Jevel=7~'7.23 Depth above floor level=736.27-73.t23~4.04 say 4 ft,
Upstream $ide.
Q~3.0x6xH3/!; 154=lBHM; H=4.IS ft .
..t the lowest point of the roof p= 12.5; assu:ne depth=x

A=125 X'
154 = 12.32
v """ ............ _- ._v' - =
12'322
I 2''l55
- x~--~-
,. 12'Sx x ' 2g x 2~ x2

But for a free surface x=deqth=4.1S _~~~5

X· .
.

Let x=3.9: R.H.=4.025 x=404; R.H.=4·091 Position 01 wave with )5~ CJ!eCl, .~.
. Q=]54 cusecs L=4.JS+.16 -359 =4.34-a.59=,75 ft.
K+F =4'18+'16= 4.34

1st try W=11.0

ActuaIK+f'_ -2.378
. 1826
't'hitl> ~. ~ore • 2nd try W-I~
 297

.,
I'
I
I
L '7~
-_ = - - ='43b
l. 1 7 L~{
_'-\jhJ
':
. ,.
r~'~
1'"~:t ...:;
.
K--1-F
C ·=2303
""'_.....""/ \I'"
"' t: :
.
,'". .
I· •
I Q 0

1J !
.J
3rd try w = '0 feet
(Ii' ,ft
154 ~
q= --- = 1;,'4 4:r' ,
10 ,I
If ,

I
:,~
,. It) I
I • ~~
I

", • "'t I
,
1

';z: I
. .... <:>
~
I
I
I
K+F ··='2228
'.' ...
.
~
~
,_"
,f ....
<I')

L J \)
I
Qj;;
"-l 1.i;
against 434 ='22'2g
1'946
.. \ WI
\t\

which is equal.
Wa"e shill! iorm wheA ,.~.

eli J
W=lO' x ='545 In'
'J "

c':
r
~G
m.
·}:H :)',

x=106'
Farrel depth='l:8' "':'
Reinforeem(nt. (\" I,,; ' . ' . . • •
Maximum vacuum pJtcfJtial=5'9- r&: ...· j
2 S,,-2 S
4'54= '·36lt.; v=' ,.~ = ". .~}:.\ iefJl.IT'gF
:Lg tg 1 86
Pressure dn~ to vacuum
=1 S6x6~'5= 116 Ills,
Add v;eigh t of cov; 1= II) 1\)5,
ero wd luad = I W 1[IS,
Total " 296 !lB.
say 3no lhs.
Max, span=7'O feet;

72
300x .. =18101Ls.
~
 298

R='555% = 1840 Ibs.; d= .1

v74
M =)25=5" max. 6" max.
.
. 12x6x'55
and use thIs throughout; Iron=--lOU--- =.4

4 bars of,r dia: max: '44 which is ample. "I'

Neglecting the effects of curvature we may arram;e for the upstream lip at R. L.
= 736:23 i.e. 4'0 feet a JOve H:e fic or of the culvert. Regarding the lattf'r as an open flume wl:ir
we have for the priming cischarge; Q=3x 60x 40312 = 144 eus.cs. •
R. L. of D. S. Up.
Depth of flow in the earthen channel for 144 cusecs'
(144)3/5
d=42---- =345 ft.
200
The downstrp.am lip of syphon must dip into tht level to ensure sealing; surface 'level in
drain for depth=3'45=732'07 +3'45=73552.
We will therefore keep 'the downstream lip at R' L.=735 SO
Position of wave at point of pri::ning with discharge 144 Cs
We have L=drop=36'23-35 52='71 ft.
1st assume widfh= 180ft. q = -!~ =8 cusecs.
18

, [q2 / 82 . _ ,L ·71'._~.
C={I g= {I g=1 2,,8 Ie =)258= ,,6;),
from Crump's diagram Kt F ==2 488 ; K+F=2'4S8X ,·258-3·lS tt,
Giving a floor levd [limiting]=736'27-313=733'14
The designed bed= 733'03
This shows that the wave shall form U. S. of 18'0 feet width of the flume 1the end 01
the syphon).
No speci"l precaution to form a ·"ave is necessary as thp. floor is already low enough.
WP. could on the' contrary reduce th width cO some extent in order to bring the wave nearer to
p

the lowest point of the cowl and then extenrl the cowl at an upward !tlope to downstream F.S,
expanding at the same time. This expedient would give us a higher coefficient for the syphon,
2nd Try a width of 15 ft,
144
q=~_-=9'6cusecs: 19 '6 2 L 7
C={I--- =1421 ;L='71-=_':__!_='50
K+F
.... --C---"-2,·395
b g C 1'421 -
.'. K+F=2'395x1421=3'4 It. ", l ,'~ ,
This gives a limiting floor level =73623-34=732'83 against 73207 i.e. R:"'L. of bed of drain,
3~d. Try a width of 12 feet.
144
q= -,- = 12'0 cusecs.
J~
c= v~/(--
122
g
""" 1'641
·71. 'from Crump •S d'lltgram
L =--F6!7"
C

K+F =2'295; K+F=3 78

C ,
Giving limiting floor level=736'23-3'78-732"45
 299
We might go further in this direction, bu t to do so would involve a risk of wave
forming too far downstream. By adoDtins- a width of 12
feet ot the l'wesr point of the cowl we shall be on
the safe side with a nief'ly b::llanced design. Dotted
water line shows conditions of flow for a discharge
of 144 cusecs when the priming just starts
D'>wn<:tr"am priming conditions.
Down"tream end will become sealed when the
water surfac p tou 'hes the low st point of the roof 31ab at
R. L 7395 @ 16 itet from rlown ~tream face wall,
fhe cross section at this p0int= (735 5-732'07)
X II =:3'43 X 11 ~37,73
Q Q c . Xl lC I l
v="-A' ='~"F73 . {l"l,i·.'/, J

v 2
ha=_l~=
0. 2 X
1
2g [;;7 73j~ :lg

A t crest, the width = 17 f~et

\ , = II~d ; ve 1OCl't y= ..O·

se;:tton ::,'.,
17d

,. V 2 V 2
lace level 1
2 .. Q
:Lg 2g ~17UU

,.,..Q2( .. J ...___' _ ) in depth at

~h 7 UU 18000<.11.

, _" 1 [2J
lboUOd3 '

For \ he .!rain in f III <:llDiJly we h \IT"

Q=200 cusecs; d=4'Z ft.

a.ssume
.
_SL =[.....~.!
20U
15 [3
J
[3]

,Try d=r 3'5 feet; from L3J Q-l·47-6

R. H. from equation !'l:
Ii'

d=343.l.H3'6 2 [ .. ' . • __ ~.. 1 1

91700 J 8600 X 3 StJ , ~

=3572 ; d '-0 =: 'Oi2 1

Fig 15
Try n= 345; Q = t44'S; c>'=e.;35S3; d'-d-'1I3.Try ,'=3'59 Q=154
d l ... 3'i5+154 2 ( 1
..
91100
'.,
l~tiul) x:J jVli .. ) which is the same as aS5umed.
 3()O

14. Example 8. Surge Tanks and Chambers.

(A} Surge Tanks in w, ter supply.

To cushion an parts of the water conduit against the full eff.'ct of forces dlie to VI attr
it is variously equ:pped with ff~lief valves, bursting plates, and surgl! tanks. A surge tank is
" Types of ~urge Tanks
doubt!y valuable to -e penstcck hecause it
wili not 0111y ahsorb energy during deceleration
but it will also provide a ready reservior from
wh'ch th~ turbines can draw temporarily, as
when they are starterl, or, during normal opera
tion, when a sudden heavy demand causes rapid
oprning of the turbine gates, Surge tank
t) pes are illustrated in Fig. 16 "',
The simple surge tank often lives upto its
name when a resonant (aDdition canses succeeding
surges to become ever greatu until the tank
spills over.

(B) Surge Chamber Lr controlled Hydraul c

Jump (Crump).

At Dubbiwala culvert. Upper ]he1um Canal

a silt ejector was t uilt and the escape supply
of 200 cusecs had to be dropped in the
arljoining Jhelum RiVEr in which H. F. L. was
75 feet lower than canal F. S. L. Water was
carried in a 6 feet wide flume with 1 in 10 slope
and the hydraulic jump was allowed to occur
Fig 16 .
in a chamber covered by a reinforced concrete slab which was ancoored and weighted down
as shown in Fig 17.
The idea of the chamber was to destroy the energy of fall of 7() feet with a depth of
4'00 fed in a chamber downstream of the hypercritical jet.

15. Example IX. Rigid Top Flumes.

(Hypercritical to sub-critical without a jumn) Fig. 18 Suppose the origin:!l velocity
is v l and D 1 . If the channel is open at the top, the static head at the point K is D1 and the
velocity head is V12/'lg.
Suppose the channel to be of uniform width and the bottom, level. In order to secure
the gradual smooth enlargem~nt of cross section, supposp a rigid top to be added 10 the
channel as shown from E to H, rising along s straight line. We know hy experience that if
the angle is not made too big the water will cling to the inclined plane anli as the cross spction
increases the vl'locity will be smoothly and gradually reduced, as exemplified in the expanding
tube of the Ventnri-meter. '
At point B, where the depth has increased to D and the velocity has decreased to v.
the velocity head will have diminished to v2j2g.
By Bernoulli's theorem the static pr"'ssure at B then equale

. >i

In general, this pressure at R would not equal D exa.ctly, but would be something
greater, as BF. Bv means of the above expresion, the static pressure corresponding to eacHl
depth can be easily calculated for given conditions.
 301

.......
•" .... ,...e:- ~'.

~ frretotc. MJI:lt:;NAI':'I'
I,e J1/-4.'fCI(

,G(:f!"'il$ ........ , ... " ~

...
Fig. 17
The resulting static pres~ure, as shown in Fig. tR gives the r.urve EFGH. This curve
which crosse, the surface (d the water at E and H indicates relations of gn'at fundamental,
importance which should be ~arefully noted. At E the static pressure on the bottom is exactly
that due t~ the depth of water AE. As soon as the velocity is rpduc~d and part
of the velr)city head is· thus converted into pre'isure, the pr'l5>ure on the
bottom is the greater than that due to the depth of the water at"ng. Thl1s at P thl.! pressure is
that caus?d by a head GP. Tliis means that at the PJint N there IS a pressure afainst the upper
bounding surface of the water eqllal to th::lt caused by a head GN. If a small piezometer tuhe
shoulrl be inserted through the upper surface at N. the water would ri:;~ ill i.t to the hwd of G.
This pressure tend .. to burst the cover of the conduit. The curve EI'GH is really the hydraulic
grade lim through the expanding section.
The interior bursting pressure against the upper surface begins with Z'3fO at F and
increa<;e!", very rapidlY at first. and then more sl )wly, until it reaches a maximum value such
as GN at N. From this point onward it gradually decreases utniU it a~ain reaches z:~ro at H.
The pressure on the bottom at M is ltgain exactly that. cau~ed by the depth of the water.
As the velocity decrpases betwpen E and N, the corresponding dimunition of velocity
head is more than enough to raisa the watH ::llong the rigid. ~lanting. uprp.r houndry surface, and
hence there is an accumulation of excess static pressure. As tt'e velocity decre::ls.es from N to H,
the gradual conversation of velocity head into the pressure is not sufficient to raise thl'l water, and
hence the previons accl1mulation of pressure is gradually drawn upon in raising the surface to
higher and higher elevations.
If the whole apparatus is open to air, the water at H would nl longer follow the uppt'r
• I. -- slanting SUJ face but would break loo!e
t--.,r-......;_-::;::::::::==========-=::;'liP1~- and flow away with level surface from
this point. If the air were excluded by
suitable means 1 eyond the point H, it
,
I
would be uwler a vacuum increasing in
I 02 intensity with the coutinued expallsion
: -Vz of the ·:ross section.

,,:~ ~
Tbe cross section at which the
upward pressure on the surface E R is
........... _ m •••*_ •• M • • " ......m ...... the maximum at NP is of a particular
interest. At this place a tangent to the
Fie'. 18 curve at G is parallel to the water
surface at N, and the change of velocity head is just sufficient to produce the change in elevation
of water surface so that there is neither accumulation nor reduction of pressure, On account of
this particular balance between velocity, head and depth, this particular section may be said to
mark a critical point or a condition of critical flow.
vc 2 Q1
It can prove that; Vc =gDc =g Q.; and Dc = _ = --
g g
 302

The curve EFGH, it extended according to its mathematical equation, has as asymptotes
the straight lines OL and OK. The curve EH need not follow a straight line. Any smooth curve
would do and Bernaul', Lemnscate provides the ideal surface expansion. fhe curve EfGH would
change correspondingly in shape, but would have the same properties as pnumerated above, and
WOuld, for each depth of flow, be at the same distan~e above the top
fhe df'pths at E and H always have a definite relation to each other because lilt these
two points the sum of the depth and the velocity head is the same. Depths so related are called
alternate depths.
Let Dl=depth at E; D2=depth at H; Q'-"-vIDl=v 2D 2
V1 2 V 2
--- +Dl= 2. +D2
2g 2g
.' $ubstituting for velocitie3 their values in terms of Q and transposing.

2~~li -2g~~i-=D2-Dl; %. g::--:%~:i=D2--Dl

Dividing thrcugh b:y D 2 - Dl and substitutin~ Dc3 for Q2Jg Dc3= _~Iy .
Dl+D 2
If the values of any two of the depths are given the value of the thi.rd can rea.di:y ce
oh~ain,d. .1 his device in th~ simple form was uspd by the author to reduce M.M.~, of the
Onfice semI module as published in paper No, 2::S7. punjab Engineering Congress 1940 and
described in paragraph 5 (e) chapter XII part II.

, .
 PAr<T II

CANAL IRRIGATION
CHAPTER XIII
, Siphon Spillways and Hydratamats
1. IDtroduction.
Instead of alfowing wat::!r to spill over the crest of a dam or weir the surplus water may
be dtalt with by a siphon spillway. This m::J,y comprise one or more siphon units. :\ siphon, {I,ot
syphon, :IS it is sometimes incorrectly spelt) is a pipe or tube bent to form two legs of unequal
effective length, by which a liquid can be transferred to a lower level. over an intprmediate
elevation, by the pressure of the atmosphp.re in forcing the liquid up the shorter leg of the pipe
immersed in it, wh;le th~ excess weight of the wattr in the longer lpg [when once nlledJ causes a
continuous flow. The flow takes place only when the discharging extremity is lower than the
liquid surface, and when no part of the pipe is high"r above that surface then the same liquid will
rise by atmospheric pressure [about 34 feet for water near sea-level]. As applied to a dam
the siphon is usually formed monolithically with the dam. Such a siphon is illustratt'd
disgrammatically in Fig. 1 wherein are shown the names usually given to the various part;;. As the
re"ervoir overflows, water flows over the crest of the sipt on as in a plain spillway, bnt the design
is such that the air which is contained in the upper bend is shut off from the outer atmospheTe
either by a water seal b<>low the downstream leg, or by a curtain of falling water. If now
t he air is exhausted from the upper bend the siphon will run full and is said to have "primed"
The discharge now very much greatpr than could be obtained with the same reservoir level and
;to phin spillway of eq\lallength to the siphon spillway.
The air may be drawn from the upptr bend by an air pump, but it is usual to arrange Hat
priming shall be automatic, the siphon being so designed that the water flowing over the cre,t
itself carries away the air from the upper bend.

In all siphons the inlet is bell-mouthed

or funnel-sha ped so as to fec uce loss at
entry and minimize surface ,iraw-down
above the !'iphon mouth. Usually the
upper l'p of the mouth of the siphon
Olt is carried down below crest level and in
this case it is necessary to provide
air-vents in the hoon of the siphon just
below the level at which it is de:;ired that
siphonic action shall cftase, Fig. 1. If
these were not provided. a siphon would
DAM continue to function until the reservoir
RE5ERVOJII
level was reduced to the level of the
siphon lip and mu~h storage would
be lost.

Fig, 1
2.: Discharge formula.
The umal syphon formula is:- .' . ,
.'..
Q=Cd A y'ZgH where Cd=Co-efficient.Head; A=Throat area,
The value of Cd has bren taken to be'S in some of the low head designs ,adopted in
Jhe Punjab for platn siphons used to keep a constant level in the channel. ',1,
 304
A. H.Naylor, who built the Laggan and Dunalastair Dams in Scotland. considers in
his book on "Siphon Spillways, 1935" that the head should not be introouced into the expression
for the charactenstics of a siphon. His opinion is that in view of the lack of
knowledge of the priming level before construction of the actual siphon, and in
view of its variability according as external conditions vary, the performance of a
siphon should be assesed a<; the ratIO of the discharge to the discharge of a "perfect" siphon
of the same throat area. By a "perfect" siphon is me;:jnt tl-}e purely theoretical conception of a
siphon with a perfect vacuum across the throat. This ratio, it i, suggested, should be called the
"Efficiency" of a siphon.

Efficiency = '1 -y'Q=- - where Q is the discharge of the siphon; A is the throat area; a is
A :Lga;
the :ttmospheric pres"ure in feet head of wa ter units; feet and second.

At sea level a=34' and the expression then becomes: 11= Q

47 A
~
Q can be calculated with a fair degne of accuracy or may be measuren by means of a
s~ale model. In a few cases it is pos"ible to check its value against the p:rf?rmance of the actua I
slphon, though this is usually rather difficult to measure accurately. ThiS IS a true efficleny and
can never attain a value of IOU per cent. It is a measure of the intensity of di,char6e ,.rver the
crest. The only shortcoming is that, like the coefficient of discharg" it can take no account of
priming level. But it has the advanta:;e of defIniteness for this very reason. A further dis.advant-
age i" thp, disappearance from consideration of the somellVhat indefinitu H. The .,"oefficient.of
discharge" diverts attaintion from the true desiderata of :t siphon. It snggest?, that the dischuge
is proportional to the square root of the head availablp. in any particula.r design, wherea, It
depend.;; rather upon the vacuum attainable at the throat action.
Type of ::liphon Spillway.
Siphon, may b" classitipd according to Head as High, Medium or Low Head siphons.
. High Head. Consider a ::.iphon whose area of cross section is constant from the
throat oownwards. If the lower limb be imagined to be lengthened, tt en, as the operating head is
increased, the degree of vacuum at the throat will ·increase. When the head becomes greater than
00 feet (the actual figure depends upon the design. barometric height and the ameuat of air in
solution iN the water) the lower part of tae siphon will not flnw full, or altern;:Jtively violent
"{:ulsations of flow will begin to take place. It is desirabl~ to avoid this, and therefore for greater
operating heads the outlet area should be reduced by a nflzzle, or by tapering ttw lower
~part of the siphon, or in some other manner incre::tsing the resistance
to flow. Such convergent siphons are conveniently designated as High Head ~iphons.
Medium Held, With somewhat smaller heads a ~iphon of constant cross sectional "rea is
suit able. Such siphons are often used where a greater discharge could be obtained with a divergpnt
lower I'-g. !'hey have the advantage that priming occurs at a lower level and t hat from the
con"tr"ction"l point of view a unifrom cross sectIOn is the cheape; t. Siphuns in which a lower Lmb
is of const:mt are'!. may be designted as Medium Head siphnns.
Lo·.v Head. fhere remain the siphons witt! divergent lower lim::'s. These are
classed as Low Head siphons.
The clasaification into high, medium, or low hea d siphon s accord ing a~ the outlet
<1rea is is les~, equal. or greater th:tn the area at the throat, does not allow definite ranges
of head being assigned to each type. There is a :ertain amount of overhpt>ing. It is <:uggested
that the classification acccrding to head sho1l1d be, over 20 feet. 2() to iO feet, npto 10 feet. This is
purey arbitrary. However, for heads les"> than ahout 20 fr.et \c"epcndent upon the design) the
dis,--;harge may, in general, be ;ncreased by the u;e of a low~~head type of siphon. In all
siphons the maximum discharge is primarily limited by the de:iign of the upper bend. Unlp."s
the available head is very low, it is usually pns~ib:e to design t~,e rema:nder of the siphon so as
to utilize the full capacity of the upper bend. With a llW head" thp,re are tWJ m'l.in reasons why
it is difficult t) obtain the full discharge of which the thrr'at design j.; capable. These are ti.6
difficulties of ensuring priming at the df:cirefl watf>r level and the loss of b~ad:n a divergent (,utlet:
4. Priming methods.
No two syphons are alik~. It IS usual to elassify them accor ing to thp methods
adopted t'l en<iure priming.
 305

Fig. 2 Laggan ~iphon ~howing the nutlets and end siphon 1fi action.

, . r'

fig 3 La~gan Siphon sh?wing t4e effec.LoI the Jet disperseri;n

. thil ?Utltts . ·
..'

.' I
 raj AIr Pump.
The air pump or Pjector is the suitable .ievice by means of wh'ch priming mav be initiated
:'It will with the water level below the crest. Provided that th,l mouth of the siphon is situated
s·lfficier.tly JOI)l1 to remain immersed, the coiphon can be primed with He water-level many feet
beluw the crest, thp criterion being that thE' vaCuum at crtst must n<Jt ex:eed about :t4 feet of
wa1er fa complete vacuum is unattainable owing to Lbcration of dis.;olved air at low pressures].
Th" air pump can be applied to any shape "f siphon provided that there is through-out the
priming a water seal to the lower Ilmh_ The great disadv;Jntage of this type is that, ~hould
anything happen to impair ihe prorer functioning of the air pump or ej Clor, the dam wOllld be
endango::red_ fhe air passages bping sm<>U, the po_;sibility 0: choking by rlebns or freezing
cannot be rulpd out.
An intfresting application of the ejo>ctor principle is to a battery of siphons To avo'd
shock ann undue vibration it is usually arranged th<>t th ... siphon units shou!d prirne at different
times. This is easily ('nsured b; the t-jector principle In the pal Uion I ttWtt'n a( j acent siphons a
small inclined passage is formed above the Clest. After the first siphon is primed, the vacuum
above the crest causes air to be drawn from the next siphon untiH this p;i1nes an,1 so on succes-
sively throughout the battery. The time lag between successive primi ngs can be incre<>sed IIV
renucing the size of air pa~sages ; whj ... h must be small, as otherwise the priming of the tirst
siphon will be hindered. This system cannot be recommf'ndpj, as it makes the priming of the
whole battery and the safety of the dam dependent on the priming of a single siptlOn.
This wethod was used at Mittersheim and also for the Oswego lock on the New York
State Barge Canal U.S.A. In the latter case it is controlled and priming depth on crest is
d
3 where d is the depth at the throat.

(b) Auxiliary (Baby) siphons.

In some siphons a small auxillarv or "baby" siphon is formed just helow the crest
This will, of course-; flow ·full by the time the water reaches crest level The sheet of water issuing
froll'! the baby siphon is arranged to shoot ohliquely across the lower limb of the siphon so as io
seal it and prevent ai r entering from below. The enclosed air is grad ually carrie i a wa~ \. y the
surface of the falling water until priming take" place.
A good example is the Maramsilli siphon insta.llation Tht'l baby siphon is one foot high
and 8 feet wide' at the throat against t! feet by 8 feet in ca;;e of the main siphon. Priming I)f
t ... baby siphon occun betm~ thp. water level rises to the crest of the ml.i.n siphon. With the
?etter design of the two types of baby siphon u ,ed, Fig 4 priming of the rna-in siphon followed
In 15 seconds ..
While the bab" siphon obv-ates th~ ne.;d hr a scaling basin b I',,. the lower limb, it
h'ts many disadvantages and it i~ pro'nble that equally good results can b~ att'l;ne'l by one of
the methods descrihed later. The
NAItAMS!Ll.l effective tvp water-bvel for the con-
siderat ion of storage of water is not
the crest of the main siph on, but
the CTPst of the auxiliary siphon.
So -hat fLOm the point of view of
ar-
t
maxim1Jm discharge the auxillary
syphon must l:e looked upon as an
ob,truction placed imide the
! siphon. The minimum size for the
auxill ar\- syphon is axerl hy the
considerati"n that througliont
priming the sealing' jet must
i
.:r__ _ _ __ function as such. As the flow over
the main siphon crest increa<;es, the
jet from the baby siphon is increas-
ingly depressed and the combined jet
will strike even lower down on the
outer cover, The baby siphon must
Fig. 4
 307

re sufficiently large to ensure that the combined jet never falls below the lip of the lo·....er leg. Thill
!<ize can he eas.ly calCllhted from consideration of the momenta of the two streams of water.
From the Maram'lilli tests it w"uld apppar that with good de"ign priming can be relied upon tv
take J?lace with a rise of watn level of less than

d . where d is the di'pth at the throat. The babby siphon

8 •

is expensive and difficult to construct. as it must extend unbroken, the full

width of the siphon. Any p'l_rtition would cause a break in the issuing jet of water. It is a
source of weakness. Its cowl must, if necessary. be thin at the crest section and yet the outside
wil,! be suJ.jec;t to a partial vacuum with possibly cavitation and destructive vibrati0n. Th3r,!
is a dang~r of choking of the rel:l,tively sm'lll air passages.
~c 1 Priming weir.
The priming \\eir W"S, introduecd to ensure priming where for some rUSon ·it WllS
desirab'e to form the lower limb of a siphon at a flat
sl'Jp" Fig 5. Tn this devictl a weir is constructid at
crest level from the crp.st hack wards along the side
an d acro,s the hack of the si ph(ln so that when
overflow C'JllltEences, the falling nappfs complptelv
seal the air in the crown of The siphon. AIr 1; th'!:,
gradually carried away by the fall'ng water until
pr:ming ensuf's.
fig. 5. rests on the Bear River Siphons of the MokelumoJ.
Project indicated that the primin~ weir h, itself e0uld not he rdied upon to pronuce priming
with the perm::i"ibl~ hed a'J,w ~ the I:rcst of O'3d The outhll of this siphon wa<; theJl
modif ed so as to ensure priming illdeppnrlently of t:1~ pr:m'ng- weir. It was then
fO'lnri • that by its e'f ~ct of d m')ling the lengtfJ. of \V'ir the prirniag weir (ansed th:
siphon to prim'! 'uith a little mHe t7a,1 half the f.se of water-levelnece5sa. r y. Nhen the priming
weir wai boarded off. fhe tarbulence ::l.od re,jll~ttOn of di ic!11.fge cau';ed l)y th is d-:!vice probably
nentral ZH its arIv,mta{e i . l'or t~le sane fduon t'le pjm ns wei' Cionot be rega,ded as a
sati~factory device

(d) Heyn's Flexible TJngue P,i nin5 D'llvi3e.

An ingf'nious app1ication is the system suggested hy H~ y"Tl
in which a flexible steel tongue directs the over, n"",
across the siphon until priming i1egins when the we ight of
watpr prcsses down the stefl tongue ag"ins~ thl" inner wall
so that an unobstructed flow ensues Fig 6. bU,t it is well
to avoid moving parts.
(to) Priming Stf'P or Tooth.
If the sloping lower limb is not too Stf'fP then,
Fig 6 by surldenly turning it dcwn' near the lower end,

rr,mlng may he bro 19ht about without auxiliary oevice. When over'flow commpnces
the ,h~et of running W<iter. owing to its momentum. is unable to follow t"e surface of the lower
limb an I shoots f, rward a:ross the siphon, thus forming a seal.
rhe primi!~g e'fIc:cDcy of a sheet of wat~r shooting across and sealin'! tIl} I )·.\'e· [m')
depends npon the amount of air which C'l.n he carried avyay. This will b~ prop()rt;oll'll to tllc
velocity of the f;"lling nappe rather than the amQunt of over-flo:", fhe lUQrp acute the angle
at which the nappe strik, s tne outer cover, the more flifectively will :lir Le c3rrifd away by thi,
water. fhe step or c~;j,nge" of slopt sOold,i. Jtierefore. 'be situated sufficiently far down the
lowpr l'mb to aliow the water to attain its maximum possible velocity. yet it lnu~t be solfflclentlv
high up to e'lsure strkirlg t he outer cover wea above the lip. There must be no dangff of tll,'
 30R

napT e c1ingine to the lower surf~e. This can be avoided by a definite disconfinui,ty such as a
small step "g. (7)

This device is known tn ensure priming with depth diG as used in Germany.
The lower limb is often terminated.in a sealing: basin, the cill of which is
' I vel with the bottom edge of tae siphon cover. Examples of this type have-'
already heen mentioned Ihe lower p)r'iofl of the siphon should be steep, but
.1 lnmt not in thi .. case be quite vertical, so that any air carl ied down into the
sealing pool will bu) ,Ie up cltar of the siphon, Thus the spajing pool acts as a
non·return valve for the air and ensures the steady evacuation of t~
siphon. Fig 8
An evil of a submerged ouqet may be noted. Consider the upstream
water level to b~ ~t the level of the mouth of the siph!)n ur, if air-vents are
provided, the level of the top of the air venti. An} ri,e of the upstream level
will now compress the endosed air sJthat the water surface inside the mouth'
will rise by a smalL r amount and therefore, priming will be delay~d, A rise in
the downstream water-level wilt llkewise compress the air and may even depress
J the upstream water-level inside th~ siphon sufficiently to prt'vent priming, In
It such cases an air relief valve should he providpd iLl the Grown of the siphon.
1 ig 7. A simple flap vahe is all that is TequifEd.- _' -, '

Sealing Bas.in . (J) Vertical or overhanging crest to ensure J}rimiDJ~.

Water flowing over a weir the vertical profi~e
of which. is an 8.rc of a circI~ will spring cleat- when It
reaches a point about GOO from the vertical ~xcept in the
case of a ~ery thin adherent napp~. Fig. 9, .
In the Verona siphons constructed by Greg()tb,
the crest is an <lverhanging sharp-edger{ weir, This has,
'Of course, exceJl.~nt prining qualities, a risf of water l~vel
ca'x>ve. the crest of 2 inches to. ~. inches, 1. e.
>

_.__~--
12
to ~~', pr~dudng priming in
'8
1 '0 2 minutes. ' .Theje
. . '.
, are h~avy hydr;:uIVc JoSSI'S .in. sllfh a sipho~ an~ the
.. >coefficient of discharge is only 0'40, ..
'.
. . Other instances are the Canberra West Lake
Siphon, which primed with 2 inches over the crest,

or_d_ ,
27
so~e of the siphon., at Huntington Lake and the
. !

leaburg siphons. The, last'Gamed were. ~f various siif's, b~t all primed )"1th head of from .-;-_

to-~
.. above the crest level ' . ;i

The g }od priming (juafitit1~ of SJph,6tiS with ~yerhatJglng crest ate obtaIned at the
expense of the disch:trgei which i~ somewhat reduc,ed by the large changes of dir~ction of ~.w .
(g) Comparison at Ptiming Devicesl
. The air P11~ti is .bY) l:y~s1titable wher,e .tt1af;iu:'ti (!()htrol i_9 desJred. 'Pfiming by ~his ftiMn~
takes place at a steady tate. dependent npO'!l fhe capacity of the .{lit pUitIp (except whE!n a yacmtn1
;tank is provid~d), So that it is' suitable wp.erEi the rise 01 water level may be very rapid. .
, A weU des.i€ned a\1xili~ry siphon can ;nd:~lce 'priming .~ith a l~eB:d jlbo_ve crest level Qf
J.' ••
 309
c1
S· The priming weir is not very relia ble as a priming
I

device, but in conj unction with other priming devices

allow!> of priming taking place with a reduced size of
water· level. The capacity of the siphon. is. however,
reduced ry the increase of resistance due to the presence
"" of ttl' weir. A water seal reduced the discharge. but
I used in conjunction with a priming device it makes prim·
I ~,
ing more certain. A few inches oVf~rflow usually ,suffices
ttl bring ahout priming. If. however. the lip ttrminating
the lower limb of the siphon is deeply immersfd in the
water seal. priming will be retarded and mly even be
prevpnted; Provision mllst be made to pr~:vei1t
corr,pression of the air enclosed in the siphon or overflow
Fig. 9 at the crest may b(- reduced 0r even stopped.
A joggle or step in the 10 IVer limb of a siphon can be arranged to cause priming with a
height of the water surface abov~ crest level of d/3. In conjunction with a water seal to the
lower limb this m:ty be reduced to diS or less
With the lower limb steep, vertical or with ovp-r hanging crest. provided that the design
is sound, priming can be relied upon with a head of water above crest·level dl3 and will generally
take place \\ith a head of d/4 or less. . '
F )r good self priming qualities there must be sharp curves. large. changes. of direction or
projections in the siphon The requirements for a high discharge are easy curves :and small
changes of direction, so as to minimize hydraulic losses. A compromise must be struck between
these opposites for good design. Priming must fint be ensured with the permitted rise- of
water above crest level and then the shape improved as Imuch as is compatible with this
c<)ns;.deration.
4. Design of sIphon S ?iIIwaJ tor High Dams.
(a) Throat depth. ,
The crest of a siphon' may be sharp as in the Verona siphons designed b} _;rogott
or it may be wide. In the form~r case, the ?riming depth would be,lessened and tlle crel;t,
heing sharp. would not take mnch space, a large outer radius could be adopted, at the l!>ame tIme
containing· the whole siphol1 top' inclUding it's inlet within the limited top width of the' darn.
But it has been found that a sharp crest reduces efficiency and ,therefore ,discharge. and also
causes cavitation and vibration to the structure and is therehre undp.sirable, though welcome
for a reduced priming depth. . ,:
Experience has shown that .in a siphon design,. the minimum permissible -vacuum should
not exceerl 24 ft of. water, because wIth lower .pre,ssures too .much air in sohltion is released from the
water passing through, which ca1J~es vibration, cavitation and parting of the water from thfl crest.
If we tuu, limit the vacuum to 24 feet, 3:nd this at the crest of the siphon, evidently if v is the
" velocity ~t the stream lines'at the crest, then; v~==v'2gX'l4='8X2v6=36 It per second
. '. No~ the flow atthe throat ~f a siphon. as proved by professor Gibson .. Davies and
others ;in the papers contributed by Ylem ,to ,the, Institute of Engineers, London, approximates
to a free vortex flow, which has the iwell-known property, viz, v x r.or the product of velocitv and
- the radiu!l at any point in the stream equals a constant. Fig. IV ,
Therefore at any section. where the. radius,,::;::[ and \Telcicoty=v;
vxr=voxro (at crest)=39Xro

:.v=39~
r
:. the discharge through the sli:t of height dr.. and width· t foot==dq
" rOJ
=V Xd A;";'~rox dr Xl' , ~ \ "

r
 310

i R
Q= Total discharge through the throat R= I -~~~Odr.=39 ro log B_
Lro r ero

.', the discharge q per foot width of siphon throat is=39 To logp.}{__ '.
r.

Efficieney '1= -- ._Q_~ =~Q_­

A V 2ga 47A
R
A =( R -To)x·1 per f 00 t WI'dth"" • _39
'1- Tolog;r 0
-~------
47 (R-ro)
Discharge per foot width of siphon tube q=nx47(R-roJ

Negative pressure at the crest==24 feet; Velocity at the crO~Il~39!?

:) ;·'·/C •.. , II'>
" ;-~fT'~<1!:~.< 1

(39[<:.J
R
Pressure at the clown={R-ro)-~igft,

(b) Width of siphon.

The throat section must be rectangular in oropr to get a maximum area if! a given
height. The width of each unit is determined chiefly from strnctural considerations. Then~ is
llsllally a high vacuum pressure acting on the outer cover. However, if the cowl is curvcd,a
fairly wide siphon is permis:,ible without making this unduly high. Thp. effect of vibration must
be borne in mind. During priming there is a likelihood of violent vibration of the water column
and jf the sbhon cover is thin :md too wide, it may vibrate in sympathy. No rule can t e laid
down, but width of about 2d gives good proportions. If the width is 'much less, the hydraulic
mean radius is !'lensibly reduced and friction lossps ~nd th,' cost of construction are increa~ed.
The discharge p~r foot width is worked out and now the discharge of each siph(JU is
got by multiplying it with the throat width
The number of siphon tubes can thus be worked out by dividing total discharge cap~city
by the capacity of each siphon.
tc) Level of the outlet.
High head siphons are ohviously indic"lted Whffi the drop exceeds 34 feet over the flam,
Water should be shot clear of the toe of the dam in the fvrm of jets. Since the discharge is
limited by the vacuum at the crest :lnd since the lower the orifIce the greater t hp, jet 'velocity,
the orifice must be reduced in ;lrea all the mm p • A low orifice with its greater Vf~locity and its
proximity to the toe of the dam can shoot a jet further downstream, so that the jJo-,sihility of its
affecting the dam is reduced, A low orifice ha~ many disadvantages, h"wever, Th.e extra length
of siphons me'! ns extra c( st. Th ~ gre'lt"r construction of the passa{!e inrreasps the posi bility of
obstruction. The down;tream toe is weakened in the region of greatest c)mpre~sive stress. fhe
hig h velocity about 700 feet per second with 120 feet head might damage the surface of the
C'Dcrete. In this connection it should be pointed 011t that there is no fwidence of deterioration
of smooth, dpn<e concrete ullc~er velocities as high as 60 feet p(r second, if there are no
irregularities to U?:;~t tlte Sm nthness of flaw and if the water is fr,~e from grit.
Let it be limited to 50 feet per second in the case of con :rete and assu:ning co-efficient
discharge.= 8; v=8V~gH'" H=60 feet.
The level vf the outlet c011ld be 00 feet lower than crest.
In ord .. r that the jet should strike the ri ver bed at a maximllllJ. distance frum the dan,
the jet must be inclined upwards.
Let 8= inclination of nozzle: x=height above river bed; l=horizontal travel of jet;
v=initial jet velocity.
 ,----
Neglecting air friction, which will not be important fot' a large jet. i~ is easily shown

that for 1 to be a maximum: Cosec e=V'~!~- + 2 and

v

then l=-~cos e~ ,in e+y'/~n2B+2~~---\

g l ~ J
The area of the orifice at the outld ; Ao= Q-Sq. ft.
. 50
\
(d) Verlicallimb
. d) Th~ cross section aru at throat is /I. sq. ft. and that of the outlet is Au sq ft. It is
desIrable to r~duce A to Ao a; quic::kly as possibl ~ from cl)nsiderations of the strength to give the
maVJmum thickness between tht siphon units.
If pressure is kept constant, the ratc of inc-ease of Kinetic energy will equal the rate

of loss of potentiai ~nergy ; d(--~~ )""'--dh

~0 ,:~ ~
2
Integrating within limits hl and h't.; hl-h't.=9 ( _
2g .n.

This gives the length of the transition for the cho.nge of the sectional area.
(iiI Shape of vertical limb.
. From the hydraulic poin.t of view 3 circular . shape of the vertical limb's
lIi~al as having the least wetted perimeter. From the point of view
of stress in the dam the concrete between the :;iphons should be as thick as possible, and free
from sharp corners. This would suggest an ow.! section or a rectangular section with rounded
corners. But the extra expense is not justified 3nd a circular section is usually adopted.
(iii) Bend on the vertical limll.
0' As in ve:itigated by Professor GIbson aJ1.rl Davies. there would be sone loss of he'l.d a~
the. lower bend and it has been found that, in a pip(', the minimum loss of head occurs when thl}
ratIO of twice the carlius of the bend to that of the pipe is=5."
(e) Inlet Cowl.
We shall now consider the design of the inlet. In the early siph'lns, the lip of the inlet
was kept at the same level as the crest of the ~iphon, so that, as soon
as the wate~ level w"nt down. t) the Cf<:lst, sip honlc acti m ceased, air rushmg
in as soon as the lip was uncovered. But this arrangement has the obviously great disa,:ivant;:.ge
that the surging of tile water Cue to waves would prime and deprime the siphon, caming
intermittent wnrkillg. Also, even when the siph n primed. the shallow screen of water ""ver the
lip would t~nd towards the formation of a vorteX. sometimes e"en a hollow vortex, resulting in
, air being SlIcked into the siphon and prevent-ing its priming even when the water level was far
above the lip levd. Thi" wa". therl'fortl, remedied by carryir!g the siphon lip weli lJ~low the
crest level and by providing air-vents at a suitable height to deprime the siphon. N( w the
usual practice is to give an area for the widest cross section of the inlet, twice that provided
for the throat of the s.phon.
[fJ Air ven ts.
As already stated above when the l:p of a s;phon is catded 50 deep below the c~est ot
,jthe siphon, ior purposes of depriming, an air-volt of ;:. ;lui table size mnst be provided in th~ hood
with its top at a level where it is de~:llfed to stop the siphonic action. The air-v,nt is lIst}BLy
rectangular with its wicHl equall to the \\idth (,f the siphon and its height not very large) b,C>I,use
the gradual uncovering of the vent by the gradual fall of the flood level would result in partial op-
eration of the siphon over a long period, causing noise and vibration, With a high vent thpre is
 " . .- \ . 312 . ','; _. . -.- .' ,.'
a.lso te_ndency of vortex fortn:J.tion with the ri.;ing water levp.I a.nd this would either prevent or delay
~he pnming. Information available regarding the size of an air-vent ll'ee3SHY to deprime a siphon
· IS .very mellgre. We h.ave got some experiments conducted by Inglis at Poona hut better than
,thIS we have t.he exp~nen:e of the siphons actually working in different part; of the world. The
are~ of the aIr vmt IS generally expressed in terms of the throat-area. Stickney, the grandf~ther
of SIphons and w.ho lOI.g back designed and ~onstructed siphons in different pans of the world,
says that th~ raho should be 1(24. The ratio in case of the M~rmasi1li siphons has deen kept at
l/t'6 by Davies. In the Lake Fife siphon, however, it is a, small as IllS and this has been found to
reduce the flow: The larg~st ratio is in the C'lse of the Wood Creek siphon and this is 1/~·3. The
HetchHetchy.slphons have a Jatio of I/IO,the Dunalascair larg siphons. have a ratio o,f 1/16 and
the Laggari .slI?hons recently constructed in England have a ratio 0 f 1/ 11.5.Except in the cas~ of
tu:o .the rabo In case of mOst d the siphons of the world ranges from maximum of 1/4'3. to a
tIl1mmum of 1/28. The gener"l practice is, ho\\ ever to err on the safer side. Scalemodel expenments
may also be co~duc~ed but the results though cannot he correctly representative in this respect,
on account of VISC0Srty surface temi m etc. We normally migut keep the ratio at about 1/1 'l. the
,throat area. -
. 'The shape of the lip is also important. If it has a sharp edge, then, under a strong
draught of. aIr, a~ th~ ri,in,! w,lter rea::h ~i th"! lip. the draught will depre,s th} water surface
b~low the ltp and It would r,nuire several in:;hes of ri3! of wlt~r leV'el a~.)ove the lip bflfore the
alr-venH.. llc~ually sealeil and thus priming would b€' delayed, though, of course,for ,the very
· ~easons explaIned above,ddpriming would be fairlv quick If, on the other hand, th€' ltp surface
IS broarl. and rounded, then. '1.<; soon a, the rising water reaches the lip; air will be cut off, as'
there Will be no water depre,sion n'lwand the priming will be satisfactory. But as the wate~-
I.. vel falls below. the lip th:! air rushing under will be. at sub-atmospher~c
pressure llnr~ thIS. would su~k up the water, which w,muld throttle the ~ur
flow and ~h~ slp~on Will not stop working until the water lweI fa.lh some distanc~ below t~e lip.
The depnmmg 15 thus delayed and the water in the r~servoir would bo. wasted. Thus, WIth. no
draught of air, a sharp-erlged air lip is ideal; but where the rise of water level is to be stri(;tly
I ' . limiteel a broad lip is better and safei:'; but
·J.I1~ P.t.LHSl. we m~st.. on the other hand. be prepared
fl.r..p to lose some storage,: due to the delayed
depriming. The actual slope to be useel,
therefore, depends upon the co?ditio.n of l~e
reservoir working.·A compromIse IS alNaYs
desirable. To pnwent the interference in
priming and depriming, due t~ the sur,ge
of waves it would be convelllent to pro"lde
. an exterio'r air lip as shown in Fig.It. though
tbis would involve an extra cost.

"
Fig. 11
[g] Jet Dispersers.
The rock in the' hed of the river downstream of the dam may not be
,., "P-\, ha ~(l awl m::tybe intersected with a numter of cleavage planes The iets of water. ~ith a.
initial.-ye,locity of about 50 feet per spcond are likely to, ¢i~inte~rate or re~we the~ !~C~ tn th~ bed
'. by theIr 1Illpact fhe present day designerg, who have t') deal WIth such hIgh vclO",1tle, and s~ea
large ma:~'ses of water; use, what are called, "Jet Dispersers." These have been used for th~:sluIr.es
in the M~ttur dam al1d also, in the new L10yd darn Su:h ~i9persers h:we been us~d' also 10 the
~agg~n s1phons, recently con~tructed in England. These (itspet;;ers ~ave been patent.ed by Mes?rs,
(jlentield and K~nnedv and they "con,ist essentially of a set of rarhal v<l;nes of ra.dIally varYIng
pitch. so that, ths core of the issuing jet receives a higher speed d rotatwn than thij surface, and
thertfore, owing to the resulting centrifugal f"rce being greater at .the c~ntre. t,han at the
'circumference. the jet on 1~aving the di;per:,ers litrrally explorles, .b~eaktng up l~tO. a. ftn'j spray
the ener~y of which i" soon dissipated by friction v;ith t.he ai:-'" Uue to t.he frIctIOn, the path :01
,the partIcle". d water will no longer l:e parabolic and 1t Will spread 1tsf!lf ov~r_ a hrgoe area,
· thus redncmg _ 'iJrl_[Jad. These' di!tpersers.' since they· offer .. some r~slstance. to .. ,the
free flow from the siphon outlet, naturally cause a certam amount of bac1~-press'lre,
 313

Hence, it would be better to design for a slightly larger .outlet vent than usually necessat y ;by
about half :I. foot ciameter.
6. Graphical method of design.
"According to usual hyrlrndynamic theory for two dimensional flow of an irrotational
fluid, such flow c:m be expressed by the equation: '
¢ +i :J_, = f (x+iyl ; when rP is the fluxe ; <p. is the veiocity'potentiai
This is llsua!ly written; w=f (z)
If in the plane of tt), q, is takPn :IS the abscissa and f a~ the ordinate, and straight line~
4>=1, 2. 3 .......... and 1>=1.2.3, are drawn. the Q) plane will be divided into a netNork of
eg1lal sqaarps This represents a s:ra;ght, uniform. pu;tllel flow from left to right.
Now suppose w=¢-f i 1.=1" (t)=F (l+i /J,
where F is any function. The l ann 11 lines in the t plane corresponding to the 4> an(L,,:J_, lin~s in·
the w plan~ wlll,inter,;, ct orJhog.)nally to form squares,but in general, the sides will be curwd and
the squares unpqual III size. To every p'lint in the w plane there is a corresponding P(jlflt in tte
t n1ant and 'I lines repre' en t a possible distribut on. of str~amlin"s. By successive in termf'diate
, transformatiolis "f this nature the relationshi p cI> +i:J_, = f(x + iYI is finally reached and it is
sometimes pos;ible to oiJtain a distrirution of stream lines and velocity-potential lint'S which
will fit the boundary s rfaces of the How under comideration, Sch \1\ aTZ anrl lhristof1Po[ have
• shown how such transform'ltiol1s may be mathematically performNi. b"t a solution is only po~sible
in a few <:imple caseS Nhere the solid boundariPos are stnght lines or circnlar arc,. . ' "
Hy virtue of thp properties ot the 4> aud ~O: lrnes of intelsecrin~ orthogonally and of
dividing rp.gion of flow into cnrvilinear squ'l.re~. jf it is always po'slole to 0 "tain a graphIc" 1 solution
provided that the boundary conditions are known.
Consider the case ~f a siphon the cross-section of which is throughout rectangular and
of cons.tant breadth so that the flow is in two dimensions onlY. Tn orrler that the flow into the
mouth of the siphonmay be two-dimensivnal also. it' will
he considered to be a unit in a battery of siphoIls all in
; operation at the same time. The inner boundary forms the
"(;.i,stream line ';'=0 and the outer boundary . the stream line
'i =n, "here n is an arbitrary number of stream tutes, The
:;; difficulty lies in finding two known lines of constant velccity-
, J:1( potential, or what amounts to the same thing. the directIOn
o~ oi the stream lines over the remaining two bounnaries. Tf the
reservoir is deep as compaTed ",ith the diameter of the
siphon. then at a distance of several diameters the IjJ lines
will be radial to the mouth of the siphon and the 1> lines will
be concentric circles, as BB. [t is necessary to fix the
remaining boundary before attempting to trace the stream
lines. If the siphon includf's a lon£;. straight. parallel leg
Fig.12 there is little error in as,uming that the tiow is para.llel to
the leg at the centre of its leIlgth i. e, that the 9. lines at the centre are normal to the sides.
as AA. Fig. 12.
Now divide the area ABBA into stream tubes of equal flow. Four is a convenient
number. These are sketched in by guesswork. At AA the lines will be equally spaced. At the
cre:.t section, where the flow will approximate to a free vortex, the line, will 1e more closely
spaced at the crest then at the crown, Finally at RB the spacing will again he approximately
uniform. Sketch in 1he 1> lines so as to intersect the 9 lines at right angles. It save!'; time if
these are spaced at interva1s along the strea. m tllbe. whic!l is near~st to tile centre of tne siphon.
equal to thE: th:ckness of the tune. ft wlll prolnoly be f )lind that mnst of the elernental area~
are by no mean' square', the mear} distance between the 4> sides being unequal to the mean
distlnr~e befween the 'y side~.: rhe I" lines should now' be adju<>ted so as to make the elements
between adjacet t ¢ line" a, nearly square as pos~iole. 1'h"n new 9 lines call be "ketched in
crossing these ¢ lines orthogonally. and the process should be continued until the accuracy
desired has be0n attained.
It ic; a 1vi,able tt") sketch in the whole flow netw.)rk in the firs': place. however
approxiwately, as the terminal conditions affect every y and ¢ lioe. The flow in the upper
 314

bend never quite attains a simple, free vortex: flow, but unless the angle of the bend is small,
the approx:im'ltion is very close at the centre of the bend and the cen tral rp line may with little
error he assumed to be a straight radiallme.
7. Etlect 01 siphons on the dams.
1 be flowing water ext-rts a centrifugal force at all bends, a longitudinal force where.
ever an acceleration of the strt'am takes place, and in addition tangential frictiun forces. The
effect of these ann the lightening of the dam by the replacemf'nt of masonry or concrete by air
!lr water throughout the !:iph'ln shollid be calculatf d whm considering the stability of the dam,
The stabilitv of e:lch sectIOn of the darn between contraction joints should be considered
inrlependently. The general effect of th"se forces is to incTfoase the overturning moment on
the dom •
The centrifug .. l force at a bend of angle 0 is in a nirection bisecting the angle of the bend

and ha<; a magnitune of 2 "-'--Qv

, .. ' g
sin~-
2
where v is the mean velocity. The accelerating force

excrte<l by the siphon walls are given by

" 2g
wg
.ov, where OV is the change of velocity. Note that

,the total accelerating force acting on the water is wQ_.

g ov, but only one half of this is exerted
by the siphf.n walls. The longitudinal friction is given ':>y wAh! , whf're hf is the head lost in
~riction in the part under consideration
In Clmsirlering the stre~s in the siphon cowl the centrifugal force cannot be used, but
rather the actual distrihution of pressure on the underside of the cowl.
Where the siphon is carried throu~h the dam, the possibility of the maximum shear or
compressive strf'sses heing increa;ed must be remembered.
Where the water from siphons flow down the face of a dam. there will be an increased
pressure on the rlownward curved toe. Where the impact is below the dam, the nature of the
rock must be considered.
Not lfast important is the possibility of vibration ann its dfect on the dam, Thin
reinforced concrete stTUctures are particularly susr.entible. Vibration only rfsults wheu a ~iphon
is not flowing full ann free from air. so that its elimination lies in depigning to avoid sharp radii
with consequenent vacua and vflrtic~s and in arranging for priming and
beaking to bl! compli>ted in a short time. The an anging of the siphon units to prime ;l t
different levels and the me of mech:lDit:ally-opcrated valves are of partkular advantage
in tnis respect. A typical pre~sure diagram got in experiments by R. M Freeman is showr. in
Fig. 13.
4

Fig 13.

~
" P",SSllll'
 S15

8. Ice trouble in siphons.

One of the grp.'l.test fellrs ill connection with siphon; has bep.n the dan~er of freezing.
Ttere should be no horizontal projection und!'r which an ice sheet might catch.
A spell of hard fro<:t is always accompanied hy reducp. run-off. which. in thp. case of a
hydro-flJectr c scheme. means a fall in wat"'r level Before a l'se of water level can take plac:e a
t~aw must ha~e set in, and long before there is any apprec'able overflow. any ice will have
dlsapp~ared .. fhe pipes supplying the still-water VI ells would likewise thaw out befJre any
apprecIable nse taken placo. :,
It is highly imorobable that the interior of the dam win ever reach a t lmperatnre
~pproachi~g freezihg. but the horizontal limbs of the siphon, are drained as a. pre~aution again,t
Ice fcrmatlon.

9. Hydratomats.
The Hydratomat in its many form; converts the pncrgy from low falls of wllter in canal
systems and rivers to air power eithe" compr.essed air or vacunm and such air pown is transmitted
. to lifting units some distance apart. principally for the purpose of raising water.
. From a cursory study of the followi11g, it will bp. appreciatp.d that the system from
beg-inning to end is free from any working parts and thtrdore. superinrlende lce, othe·i. ~~an
that required to start and stop. is eIim nated. \{oreovpr it Cin be 5afdy aBullt"ci that th~ mltlal
cost. construction being mainly simple reinforcfd r.onr.ret p • is low. .
In the inste!!atiou of the Upper Chenab Canal Punj abo the system may be considered In
two parts, where it bllS been uSl'!d to lift a seepage drain into the canal. .
(a) Rarefier or vacuum power unit which is situated at the fall (b) Lifting or pumping
unit which in this case is raising water from a drain about 1 500 feet away.
The Rarefier as illustratpd in Fig- 14 comprises a ;iphon which is raised to a hf.i~ht
slightly aho"e the VdCuum required at the lifter, In this instance the siphon Rarefier;s built ifltO
existing maisonry pi.!r:> an 1 is ~ f-!ec 6 inches wide, while the entp.ranc6 water.vay is 6 feet
deep.
In the cen.tral or p'lrtitio:l wl!l "B" are fixei thl'p.e P(illin~ Ports "F'" whic1.. whe'l
starting up, are u'> ~d to pri m ~ the sipho:1. When SWuUE~ open by lever~ from the out-;id~, t h?s~
ports E!ive a free pas<:age of water from up~trealll to downstrea n, Taking the lowest port, which
IS alwa\s submerged, water pa,;ses from the upstream leg and on Spli"hin~ on t!le down .. trea'n.
traps much air frorn the in'>icie of the sipao:1 fhe air pa'>sc!s out of the siphon a 1.rl i5 freed to
atmosphere. This has the e'1ect of creating a putlal va~nUll inside the ,iphon and the water
level rises until the next highH port comes into action and so on until the siphon commenr(,s
running when all priming gates are closed. For quite a large siphon of this description :tnd to. a
vecuun:of 16 fept '-\ith a fall of aeout 5 feet, the priming is accomphshed 1Il a penod
of :to mmuh~~.
Flow can take place like an ordinary siphon npwards in pa~sage "A" over the putition
wan "B" and thence dowawards in passag~ "C" an" so to downstr.. alll. Aspiratin€ Fins "D"
which aTe perf )rated, are arrmged in the throat of the siphon at intervals ac:ross tile width.
The~e fins communieatf' with an air d ICt "E" huilt into the top of the partition wall or crl!st < B"
which"com'lluilic:ttes with the air main outside, connecting- the Lifting unit.
Now the water in pa<:s'ng these perforated fins draw:> in air from the air main which
enters the water in the f'lrm of small air bu lbles. This emulsion of air an<i water passes down-
wards in passage 'C' and out of the ~iphon where the air is freed to atmosphere.
Thtls a partial vacuum is created in any r.hamber connected t) the air main.
It is. when air is admitted to the siphon in increasing volume, that tlu discharge slows
down until a point is rp.ached when th"! siphon becomes unstable and finally break,; down. To
ensure that the siphon runs steadily with the maximum air capacity anti. at the same tirne i~ free
from any possibility of breaking down, a special gO'lernor-gear is arranged which effectively and
simplv controls the voillme of air admitted un fer all conditions. It will be understood that an air-
mam connF>cts the Rarefier to tht lifter an.t thus the Rarefier is continuously extracting air from
the Lifter and maintaining a partial vacuum.
 316

'!nUli ?'J' H
"':,JfiJi~, v: 1e:;',:1 1: ". '.:J
,lfHi ).11;(1'; "",p;;;"it
, f'll f to ,~j;l,;.te y'1C
5; mOll ft:dl
1701;> : :rr: .4

Fip'.14
 317

HY~ratomat Lifter.

,.
t'A(.(VM AIR MAIN
To ~AR£-F'F.R

OUTLEr W. L.

Fig. 15
 318

Referring now to Fi~. 15 it is observed that units consists of a concrete chamber

having an inlet at the lower level "A" communicating with three lift pipes "B" and an
outlet from chamber "D" to the higher level "C"
lnsin .. each lift pipes "B" is pla"ed an air inlet "E" communicating with atmosphere
by air pip·s "r''' rising through the roof of the Lifter.
Assuming that the air is shut off from the air inlets and the main air valve on thp.
vacuum main is opened, the water level inside chamder "D" will rise up aDd similarly inside
thp. lift pipes ";3" The air pipes "F" are now opened to atmosphere with the result that air
enters the water through the air inlets t'E" which are placed Slightly a JOVe the lower wa.ter

'.. ,.,•..1 _L .
~.'t- ,_,
1-, -I. _,...,Io!--,.,-'-r
rI1 m~'
M: I : I~·t
~. -,-i: ~~-~-~ ~~J_-1't"
I
. \
I
\
b~.-II,
~
" .1;:

4-~
, ..
I"
,

II<
....
I
I
,1
I. I1 I
I
'.L_.-.J
4·.,
II II l'
,I.

1 '1-4..- ,.lII 1.
,
I

'2

...
, 01.:"1, t
~
I"ol.t!
I'
10,.,"1
I
! I I I, I I I I
I

c .•. AT THROAT or THt:

HA~ SATTIRY OF ""ROH

,,,..,r.:".L61'AT',,,,..'" '" ''': A..... 0. •. $I:crloNAI.

.,"",. U.&L. : .". ~.Ht AT .1._-. .'.VATI." .•.".
1,""011

Fig, )6

$/(~rc" $I(OWiNS 1''11&

5n'HDN IN REI..ATrON rill /

TkE H61GHT OF D ... ,."

leve Y. Aerated columns are thus formed inc:ide pipes "B" which having a loWer specific gravity tun
the Mlid water column inside chamber "D," rise and overflow into the cbambe·r "D" and 50 tf1
the higher level "C",
 319

The air, which was contained in the water, is liberated in chamber "D"and is continuously
fextractf'd by the Ralefier.
Flow in the lifter is continuous and wide variations in head and quantity can be given.
[n this instance. pumping is flom drains and the Lift is variable, in fact,
the plant i!: arranged for pumping from O' to 10 feet and in order that reason ably efficient
conditions will prevail over most of the range the air inlets can be raised or lowl'red to follow
the lower water level although it is not necessa.ry to attend strictly to this adjust m en1.
The type of Hydratomat may be used in irrigation for pumping water into higher level
distributaries, pumping water from drains etc., and in works connected with watl'r supply.
On the other hand ther~.are sy.. tems which compress air in similarly simple mamer
and the air pow!'r is c ln~yed t) batteri. s of tube wells whele pumping is car1'ied out by aiT
lift pumps. Hydratomat Limited, of Victoria Staticn House, London. S.W.!., are responsible for
these developmtnts.
10. Efficiency of Hydratomats.
The efficiency of Hyrlratom ats is rather extremely poor usually from]5 to 20 %.
11. E)"ample II (Laggan Dam Sip.l0D\Uy A H. Naylflr).
(a. Data. Flood dischal ge 380:) cusecs with priming head 1.0 ft. Dam section given in
Fig. 17; Height 120' Base=51.0· rop width = 15 ft.
(b) Throat Depth.
In an actual mndel test it was found that the priming depth rquired was dJ4 to dIS
Keeping it d/3, the throat depth will be 3'0 ft. Fig. 16.
(c) Radius of crest.
This depends on the space a'lailable. Keep it 6 feet for the crest and 9 feet for the crOwn.
(Fig. 16).
(d) Discharg" velo :ity at the cre~t for permissible 24 ft. negative pressure=v'2g~ 24=39 it
per sec.
R
_ Q _ ~~~r 0 log r0 _ 0 _ _
Efficiency '1 - 47I- 47(l<-ro) -6810 where ro -6 and R-9
Discharge per ft. width=.68 X 47x (9-6) X 1=96 cusecs; Velocity at the crown=39x~/8=26 ft.
per second. .

Pressure at the crown=-3-v~ =_~_~62 = - 13 feet head of water

2g 2g
(e) WU.th and number of units.
The width of ~iphon tubes at crest is usually 2d. Tn this case it is kept 6 ft. to incheS
Discharge of each unit-6'S3 X 96=650 cusecs. To give 3600 cusecs 6 units are required.
(f) Outlet.
[i] Level of outlet.
Permissible velocity in concrete=50 ft per sec. ;

... Head =r: ( ~ 50 _=~y-oo feet,

'~Y2g
The outlet shoul'i be at 60 ft. below crest level.
[iiJ Inclination of the outlat,.
r-~"'
cosec ()=y .lg~ +2 where x=height above r""er bed ... 64 it. say.
v
 320

v=50 ft· per second; 8=3160 say 30 0

r r~
l=~!.cosO~ sinO+v sinzOX 2gx
"'---'-,
~ = 128 ft.
g . L 2
v J .
[iii] Area of oulet.

~
Q =so-=I3
Area=,; sq. ft.; Keep it 4'-6" dfa (16 sq. ft. area) on account of Be
,nd immediately

behind.
(g) Vertical limb of the outlet and for fixing jet dispersers.
(i) Transition.
The area at crest=20 sft.. Area. of the outlet = 16 sft.
Length oftransition=h1-h 2= Q2( 1 _~)= 650.2 ( ! _ 1 2 )=S H.
2g ,<\"a A2 644 162 20 1:.1;,
It is kept 10 feet in original design. '() 1
(it) Inlet cowL
The area of m luth equals twice the area of the throat. The velocity ot entry
650 = 16 ft. per second This is rather high and has caused vortex in the actual.
= __ The area
4u
and eotry mOLY be increas ..c from 2' 5 to 3 times the area of throat..
(ij Loss .. of head.
The most impjrtant los"e;; of head are due to bends. There is no satisfactory information
on the subject and it is hrgelv for thi" reas 'n tha.t scale mfJdel flxperiments are Sf) advisable. The
loss is not a property of the bend itself but of the whole assemblage of bends and str::t.ight pipes.
Tilers is a certain amoult 01 loss due to the redistrhution of velocity on entering a bend until an
approximation to free vortex flow is pstablished, Then there will be eddy losses on the inside of
the bend, additional to ordinary friction loss, especially when the bend is of a small radius' and
finally there will he loss due to the readj ustment of velocities in the straight piDe succeeding the
bead. It would not be pxpecteri. therefore thl- the los;; would be simply proportil)o-l.l t·J the
t
•

angle of the b~nd. Bonchayar and Vlallet found the relationship between lo~:s alid angle of bend
to b. very irregular, the less being less for a 6\.)0 than that for a 30~ bend.
After a careful considerat'lon of avatla')le information the following values have.
been ch0seIi :
2R v2 . 2~
Upper bend angle about 135°, ~d- =5, loss=0'35~; Lower bend, angle 90 c' - ·:-=5, loss
4g U

v2 2R v2
"...• Outlet bend , angle 30 0, d..... =6, 1088=0'15
=0'30· 2g g
.< . •

The distribution of velocity across a pipe is not uni.form even ;n the case of a straiaht
pipe. Therefore. the kinetic energy of the is.ming jet wJl be greater than the kinetic ene~gy
corresp:mding to th" m~an vdoci~y. O.ving t) the rligh value of c. the kinetic energy is
VZ v2
taken as 1'05 9 instead of 1'12 as recommended by Professor F.C Lea for normal velocities.
~g ~g

It is now possible to proceed to the calculations, we have;

Total Head H=Los:s at entrance, he +-]053 at bel ds, hh+frictiol1 loss, hrrR. E. of jet. ho
 321

raj H=t6' [h} H';"66'

VI
=0'02-'- v'
=O'02~
2g 2g

IT >!\lmq) ! ' -0"35 va'-

'" 2"
~ fP,011· ~l_-' ~, 2&'
Lower Bend ,0'30 \,5_, L;l)oz~ n(i. ~JrlT ,8(,~
" 2g ..

=0'30 X I 62 v'
__2 _ .1
2g

I
0 va v t
30 Bend 0 15 L, -0',15 X 1'62
-0'24 .!~2_
_2,_
2g " 2g "
)~!J{) J S-X j 0016'. 2~
h f (Q being 125)
Mouth 0 to 1 : 1= 12
,1 !'~<)~ 1€Q, _.,9,
v'=O'1 JV 2 2
m=I'94
\,1'
,/; iiil= i'l':':' }JTt~- - An-
p. '
'.
(,.' '

\ V l
,. '.-.n,o",
-- _l_
:Jg

'l~q -ftl_!!_ •
I'viI1 2g

~:;)=1~5~!-1~1~ X64~" ~~s";':'{: '·_.~i

i
[b] tr.87

hv- 1 05~82 = l'05x 2 63vas ,a " >.i

'=~'39 'v 2'
2, 5
~
~g 28 ib2 76~ ."
',tiy'l I "'::2"'6 'vi
3
2'" 2g
. '3< . "
0

2
2g
raj Total H=S6 ... 4'26
[bJ tola llt=66=4'31 V~2 V2
:: ,/,' e,,!j ,: •

whHe H is t11easured from 'the upstt~aIl1 water level at priming 2,

Hence " " , ,; ,,,- E. :, ' to the centre of the Doade
[a] vi=29·L.·~ Qlld:a 2v.=203x291_59t cusecsJ' ,J
[bJ vll ==3j'3 :, Q=a.vi=20·3x31'a"",640 €usees .,'
 322

The ewClenCles
.IE' ,. r] 590 62~ [ 640
are La 20'3x47 = % t)J 20'JX47 67 %
These compare with a permissible efficiency for this design of upper bend of 68 %
[i] Effect of the ciphon nn the dam.
The effect is shown in Fig 17,
12 Exampel H, MaIikanve Dam, Inlia. (Prof. K,D Josli Mysore)
Design a spillway siphon batttry from a flood discharge of 2n,,000 cusecs for the
Ifarikanve Dam of the section given Fig, lB. fhe priming head not to exceed 15 ft.
(a) Throat Depth.
d=-3X 1-5=4'5; Top width is only 16 ft.; The radius of crest=4'S ft.
Radius of the Crown=90 ft,
The crest shall be a full semi-circle.
(b) Discharge.
For 24 ft, negative pressure, the velo~ty at crest= 39 ft. Sec, '- {?J: t ";II·
~
Q 39r loge r-R \ ~~vJ ti-t==i
'°0=.,,,
\
~

PJIO:#.fficieD.cy -'ilA"" 47({-rf =.aB?~ w]Jere r=I-S Ji.; R-9'/)lt. ( ..e r"':;m
Di~charge =·58 X 47 X 4' 5 X 1 = 12~ cusecs,
Velocity at the crown = v = 39
X 4~ = 195ft· per second. . pressure
9

v· 19-52
at the crown--4 5--=-4'5--
tg 2g
=-105'
(c) Width and number of units,

WC"HTJ
0". COIYCJI£TE

-W6/tiNT " ,

WA.-ell .,,~ 1
1 .2

Ij f ..
Let width=2d=!!x4'S=9 ft.; Discharge
for uni t=9x 123=107 cs. Number of
units f Ir flood discharge of 20.UOO cusecs
is 19_
(d) Outlet.
(i) Outlet cowl.o."r _',
The c)-efficient of discharge as given
by eftlciency=-58; permissible velocity
in concrete st'ction =50 ft.
per second, Velocity with 100 ft.
head,
='58 X v' :tg X 100 =46-4 it. per lie cond or
say 50 ft/sec.
Fig 17
(iii Angle of exit, 'n, ~,

Cos 8 == .y / ~~~+2
v2
; x=32 height ot outld above bed
-8=37° ; v=50 ft. per second.
'iii) Horizontal distance from Olitlet, where jet sheet meets the bed
f= v· cos 0 {sin ()
g L
+ ./sinZ8+
v
-~~~
ylJ
1; V=50 and ,,==32:.1=132 fl.,
 323

1
[iv] Area of the outlet A.= -~ = ~f =2,hq. ft.

(~)Vertlcal leg. I
V·ngth ot trllnsition for :rea change from 40.55ft. at etast to 225ft' in outlet.
b~_hl=g2( __!_:.... 1 )_~I07J ( _!_ '- -~)=28ft.
Lg Az AI 64 4 222 40 5'~

T----
The dia.meter of the outlet: -D=v~ x2=S'3
7\

Fig. t8

:L

tf] Inlct cowl

The an'a at entry =twice th~ area at the crown=2 x9x 4'5=81 sft.

Velocity of entry = 1100 .

---= 125 f t. pet second. ne head req 11ired for this velo"ity
tH

v2 12'2
over the top = -- =.----- =2 5 ft. nearly
2g :.!g
[g] :\Ir vent.
; I 1 · "
Afea of air vent=-·of area at throat=<- X9X4 5=3484 ft
12 12 • I
31; ,

3'4
height of vent =-> ='4 ft. say=5 inches
9
Note:
The crest nep,d not be a e )a1plete semicirde but only upto 60& are on the tcservoit
side. This will give )'6 ft. atiditional thickness behind the tube instead of 1'5 ft. only. The
entry area at the entry should be increased to 2'5 times to avoid vortex tonnation..
13. Example III. Plain Siphon~.
Fig. 19. represents a sipholl in nature. the velocity of the wator in the siphon in
 .
meters per seeond IS '1= V /- - -2gH
--r
(l+t)+>.. -(t-
in which H is the difference in wlter levels in mebrs, ,( ( is equal to 0'4 fnr a plain sharp-edged
circular orifice and is equal to 0 1 ,for a, well-rounded mouth-pif ce," A is a frictifln factor,
dependent on the nature of the matenal, L IS til. length of the siphon in meters. d its diameter in
IT;eters, Also ps, (Ule pressure at the summit) is

Ps=,,{hCl-H,- ~,I
- :E w) in wI ich ho= 10'3
g .
meters, the pressure head due to the atmosphere;
HI is the height of the summit of the siphon
above the upper water sm face. measured in
rr eters; v. is the velocity of the water at the
summit; and :Eill is the total losl in pressure
head up to the summit due to the rf'si"tance of
the pipe. Ps must be equal to or ~reater than e,
the pressure of the wattr vapour if the siphon i.
to onerate.
This is,
Fig, 19 PS,5e for the siphon ,to opeTate
or _!'-,5!._ .5 Ze for the siphoR to operate
" "
Values of e alid for Ze for water are to be found in the followinr tabJe.
PreSitare Tempera1 ure T"ble for water Vapour.
TABLE I
Temperature in
,/
degrees Celsiu9 (T) 00
e in atmosphere9 20n 400
O'OO~ 0'023 600 80'
Ze in met". 0'072 1000
1'.",P'06 ~ _ , 0'24 0196 0'462
-~. ps'~
..
.
0'74 2'02 4'81
1000
1'0:1

Thus, if water at 400 is used,.-

y must be equal to or greater than O'!."'-,neters if the siphon
is to operate.
1=--2i H;:---'
In the monel,
Vm= 1(l+O+Am~ll1
V dill

-; Lm= t__ ; Om=.-d

Where Hm= H
n n D ,(
C as. before, lies between ()'4 and (J 1 and $houtd the sarrie as in nat tire. ~ .x. ~-w.jff
probably not be the same as X since the 1rictional tesistalu:fi to fio" for the modf}l .lid
for nature will not necessarily be' ihe ~ame, x

2g- _ ( ~(»)nt
5
(1 ) III for the model •
IS" 11 ) m - . (vs)1m
(ps) m = y (hO'-nl ~ .

(Ps) m mnst be eqnal to or greater t.han-~' if the modei siphon' is tcY opftatf/(
n ,..
, ~.

tps.)m:> -» for the model t()' operde'"

..
 325
f
or (pg)m 5~for the model to operate.
i' n
,
Example. If the temperature of the water in nature is 41)1) and n i; 10, will the mo~cl op~ra' ...
. For water at 40°, Z·=O 74 meters (see table I).

Hence. (Z .. ) m = ()'7~ = 0'074 meter~ which would indicate a temperat.ure fOT the wafe'
lO
less than 5°~ (see table) which is i\l1practicable.
Example. Assume the fo lowin~ data fnr the m)(el of a siphon:-
n=10, Hm=06m, d m =008m, L".=80m, (H 1 )m=0'4m, T=T m=200C:=O.t and A=Am

=C'025, and ~O>=( :6O»m=Am Lm IV<'m .

2g , Dettrmine if. the mode1 Slp
. 1lon, as well as, tht
dm
nOli'll"
1 d; ,,;;
; 1a~n:)
, li;; \0 ~3·. :, • 1,

. lSi:)

f" ( .•

Now(ps)m = ho-(Hl}m- (vs )2m -(~ w)m; or(p~'m = 10'3_0'4_I'i4~ -O'025X~. = 1--',[,)--
'Y 2g y l!:)'6~ 0'U8 19'62
=9'36m. From the table, page 345. for T=20o C, Z1=O·7·h. Since the value for

~p_s)mof
')'
g'36m if greater than the value of Ze of 0 74 m the model siphon will operate.
.

~
Now -y-~. ~h.-Hl- -- . ~(lJ
vst
'( 2g
,__ ~ ,ij1£() ~IH f)' .. !~l' I~ "

·lhen J~~ == lO'3-4'O~ S·5Q.2__:_o-\)2f'~· ~~:() X 159~~-2!- -10'3 -4' r-.. .. ~3 9=0 9m from"'~~e
\: 19 62 d~1 il ~tl 0 t! 0 , ~;
?n ;:(",; k i·I'th
~ " . ' '," ,
:tablel page 345; bt r"",zoClC: Ze ~ O"-Im. SinceLhas a value of O'9m. ,which is greater than
. A 'D....
the value for ZI!:=of 0 'i 4 tn. the siphon in nature" ill operate. ".
14. Example IV, Siphon to maintain a const::mt ll?vel in the parent channel (Crump):
. 1 hese are plain siphons as described above. Tltey are constructed to mainta;.n a constant
level in the parent channel either to feed a high level off· take with fluctu tting supplie; in the
parent a'S a siphon at RD. 61,000 Gnj rat Brinch, Upper Jh~lum Canal Punjah, or to fe~d th~
tuthil e plants constructed for Hydro~dectr'c tnstallation at a const~nt hea.i with varying supply
~o!lditions In the canal such as at Renala Khul'd. Lower Bar( Doa~ Canal, Punjab,
f • •,
 (5. Example V Gulping Siphon (ern'1lp).
Rasul Siphon of UpperJhelum Canal.

~ C The Upper Jhelulll Canal TUn') in the reach along the steep slopes 01 'abbi hills on the
I)ne side an-i the river Jheluru on the other, about 75 fPoet beIllw. There are a lot of barrel-type
,iphon.. under the canal to pa.ss the storm watp.r frllm these hill si<i.cs. The storm water com~s
down the hillside which is sun dry aud umally silts uo the siphlln barrels when the flood IS
draping inspite of &teep sloppes of the orderl/300 available on the outfall. If the
outfall and barrel are not closed before the next rain. the masonry barreh are blown on account
of the blockade of vater-way as the silt in tb.e barrels gets consolidated and strengthened with
~rass growth in a month ( r so.
E ;. Crump, I.S E.. cureri the troublp. by con~tructing a Gu1ring siphon (spillway
iiphon). The intermittent flow in low sop plies served to keep a defined ch I'nnel upstream and a.
dear channel downstream in the outf.\Il a, sh) .vn in plan and section in Fig. 20
Data.
Siphon under canal 12 ft. wide barrel: 320 ft. long with segmental 'Invert arch in flow
lIld semicircular arch above top. Flood disch~rge= 1400 cusecs; Flood level 'lpstream of siphon
=786'25; Crest level of 31 ft. wide w(!ir=7S1.7S; Siphon width=6'O ft.: Sipbon depth =6.0;
Crest level of siphon=781'0 .
A. Calculatbns.
Discharge=3'09 x34 x4 53 / 2 .... ~ XV2g x6 X6y'S'25 = 1000+420=1420 cusecs.
CalCUlations (or sip 10n reinior ced concret•.
B. Roof of siphC'n.
This is at all times subject to pressures less than atmospheric. The greatest DPgative
"ressure occur when the upstream water~level is low,Ht' namely at R. L. 776'50. With this level
the maximum working head is about S.O fpet. Assuming the siphon formula see Fig. 183.

Q=5Ay'h; the velocity v=5Vh; and the kinetic energy VI =~5h =25 X 5 =2 feet say; the
2g tH.4 64 .4
itatic pres"ure at the crown is that clue to a heihgt of 7870 -776.5= 105ft. of water
The total negative presure is. ther~fore, that due to 10'5+2.0=12.5 feet of water or
1'Z'5x 62'5=782 lbs!~q ft. Add for weigQ.t of slab= 118 l~s/sq ft. Total=900 s:q ft. lbs ft .
. ,C. Sides of siphons

The average pr.,sstlre on th~ sid-s at the crown of the siphon is that due to 12.5- 6~
"d}·Oft. of water i. e. W .=9.5 XSl5=593 Ibs p6r foot.
D.
These will be su'-jected to earth pressure fron the earth below the existing floor. The
prpssure intensity may be taken a, 30 d Ibs1sq. ft. at depth d. TM resultant per ft. run for
'hft depth d will, therefo:·e. be 15 d l . Instea.d of blllding a retaining wall it is proposed to
reinforce the sides an'! support tlv~ vertical sla':>s by R C. beams placed along
,he upper edgp", of the pit and normal to the siphon roof at the siphon exit. The
tloo,. beam will be 15 ft. long and loaded at intervals of one foot measured from the downstream
~d I, 2,5. 10, 25 50. 90, 15(1, ISO 220, 250 280. 320 and ::n0 lJj, rp.spectively.

Total m!)ment= 10885 ft. Ibs.

. The redul'tion R at the upstream end is thus given by R X 15'0=20885:. R1392 Ibl
At 5'0 ft. from the upstream enc tLe bending morr;ent is 1392 xS.O - (320 X4+320X3+280
x2+:lS0x 1) =69(i0-~050=.3·)1O It. lbs.
At 6'0 ft, from thc end it is 3910+ 1392-1390=3912 ft.lbs' These heing practieal1yl
equal indicate that they are near the maximum value which may be hken as 341 ~ ft. Ids.
 827

Fig. 20

'~ . /hICIl IN f:£NttI'T

... /os

~
" " /.,
S70;YS
1hl;~'X'c

For a beam of 9;' in width this give~ an intensity

.
of ~9~~=5.l13 ft.
QI~
llls lYet foot 5213 ia It.

per incb.
 ~28

.. r;/5213
GIVlng:-h=v'74=V~74-=8 4
"

The beam "ill be9"X9" with 0.555 p~rcent reinfo~C3ment i.p.. a= .00555 X 9XS 4=0'42
sq in. \llow 3 No J /2" dia bars. For the sh:::rt beams of O'U span. we may assume these
uniformly loaded" ith a load of 3(10 lbs per ft.

Fig. 21

"-s tUs Is less than the loading on the sides of the siphon at the ctO\l n, the
reinforcement already DTlJvirled is a.mple to take this load, provided the vertical leinforcemtnts
alrearlv provirled ar,> also ample to take thi5 load and are ,nade to olerlap the last three of
the incliLPd roads. The lat ter roads ~re placerl on 5" centre", so that Wp may assume that the
terminal re ctions from the t'lO heuns Rre spread over the low&r one ft leng:h of the 8 roof,
slab rhe total compres~ion on th's strut :ln1ounts 1392+900=229~ Lbs. The ct,.ss ~ection
of the strut is 8x12=96sq, inches so that neglecting the assistance given by reinforcement
229l
the compr6'ssive stress is only ~-'2-S-- ,.... 241bs sq ft. which is very small and calls for

no further rtinforceml'nt tl'an that already proviced for tpe roof slab.
16. Exampl~ Vl. Use of siphon as atomatic Mak€'.and.-brdk arrangement,
Plain siphon designed ,I:; shown in the sk>tch Fig 21 wa5 us~d by the all thor to
eliminate the errors in tank '1Jeasurements of the outlet d:scharge dUll to varying levels in the supply
channel as descriLed in the Anthor's paper No,237, Punjab Engineering Congre.s. lahore, .940
The relea'ic valve arlmitting air at atmos;>hpric presmre is opp.ntd an(1 water from
the channel let i7jto the tank and the stop-watch started, . /' '
 PART II
CANAL IRRIGATIO~
CHAPTER XIV

Distributary
..
Head Regulators
. and Distributors
1. Introductbn
A distributary head regalator is an important feature of the pro')le:ns of distributioR
of supp~ies. ~t is. m )reover, a n ~.>~",ary lillk hetween all p.arth'!n parent chafl1ul and an Off-taking
distributary .. rli,;tr.blltary head is a r,·guiator. a meter of supply and a silt s ~l~ctive st~ucture.
A ilistridutary he:ld 'Na.'> designed in the beinning slmpll1a> a r~gulator [t llsed to have
a floor or a. Cfe3t at beel I ,vd of a p He,lt .;ilannel with ga'e" to r"gula~e supply . The regulator
was undershot. I he dist.rihutries were boostf'd wirh E'xcessive silt charge and their di:icharge
tabl,ts, based (In gaug~s tixed in their he4d r"ach~s, had to b..l re"i;;ed occasionaHy.

To begin with, silt exc1u,i'lU wa" the first to receive ~ttention. A silt wa.lI in the parent
channel away fr·Jm t"e gates was the fir_;t d<!vL:e triec! in p:actice. It v(:;ry soon proved to be an
ine;ficient remedy
Gibh (111 pi,," er work in sue ·essful excl'lsion of silt (as described in Paper No. 28 of
the Punjab Engineering Con~ress, 1916). Comtruction of a Gibb wlll did meet with
conside a;)le success in som~ cas·~s. (Work nn Similar lines by Els len was also described in Paper
No. 30 of the t'U[ljl') En~im.ering Congress, 19161.
Wo ,d~ d0s::ri Jed his wl)rk in plper No. 29 of the Punja') Engineering Congress in
)916. It w,J.~ pro v d,~ I to t!xcl,de .;ilt by making thi'! regulttion overshot, us ... g riiiing cill gates
Metering (If supply was arranged by ma'<:ing a head r~gulat')r t·') work a'l -1 rte fall over the
gates with ovprsh It [' gulation. It was the first attempt to correlate the silt selective function
of a rlistrihutarv h ~ad regulator with the working of a'1 off· taking channel. The device mE',
with a little HC'C:3S only in a feN cases Later on Wdlton designed a useflll rising gate on
similar lines.
King d··v ·hped a "ilt excluding devicp.. in the form of silt vanes (as dp.scribed bv
him in Pa.)en N H. 41, 6~ and 13:) of rhe Pu Ij lb £ngine"ring Congr0Si). Gibb wall'! as used
hv their author on the Low~r· \':11 'na') C"nal Ea.'>t Circle NI-re so;)n provided with
R.C. slab~ with a tunnel IJellw tlle;n. fhis devdope.i tIle design of the skimming
platfom a<; a d<!vice t() e'( 'lu,le 'lIlt. ;,ki'l1m g platform, and silt yanes met with
consirlera ~lp. SUCC'S5 in som I ca ies, . ut they fail:d lJarlly in cases ... hert! supplY in the parent
c'lannel was di ;turbed behrd thu pl .... tforllls or the vanes c ·uld function. Gibbs, silt wall
Wood's risin~ g ItM. Kil'l.~'s s It van S <1[v1 Routh's sk;m ning platf.)nns were useful silt
exclud;ng devices fhey se vud their puq)os~ vury well, Lut they flid not approach the
ideal i.e. wh.;re t,) cxdllde s:!t, and ho V lTIUCi1 of it fhe \lllhor In his paper No.
189 on A silt selectiv~ Di,tri1)utorv Hea(i R 'g,llator, Punj \b Enginee~ing Congress, 1936
made up this rlefici ... ncy by p 0 lucing a design which would pass as much ,;iIt as the off-taking
di<;tri:mtar I could carry acc )rding to t ne C. V R. allowed in it with the a va.lablt slope.
Pi: Metering of supplies at the distributary head regulat.)rs was the last to rece've
attf'ntion. I'his was arranged in the form of open flumes as de~cribed by Crump in frrigati(JD
Branch Paper No. 26 CI.ss A and as improved upon vide paragraph 15 Chapter X of this p.ut.
2. Old und6r-3hot dlstribatuy heacl-~e:6ulat)r.

A typical cross s~etion is shown in Fig I The reguhtion wa.s done in this case on tne
crest which used to be at the btd·lev.. ) of rhe parent channel and in some CaSQ3 a foot hig<lcr
 330

The crest wa') foil wed by a cistern t to

'SI.C·TION 2 ft. defp and about 10 'fl'et long.
There used to be a ory brick (n.B.}
pitch;ng downstrf'am of the cistern.
The upstream wings were sloping 1 in 1
and splayeo'at 4",,0. In some cases D.B .
• jlitching in I~ed in par!'nt channel was
.1 .put in. as shown in plan awl section.
• Thf' Iron ~ates med. were Hted by
~windes rising in groovps about 6' deep
I
.a'l shown in the plan The regulation
was eione on a gauge uSll:l11y fixed some
50 ft. eiown-stream of the head and in
'§: !lome cae~s fixed against the down-
__ ~~::==~~)-_::!:= • - , Ii •• Fig
strpam1. return \\'al: at G in the plan
'1;' There are no experiments on Ttcord
which would give the silt conductive
'oC power of t lJe head but it t:an easily be
f imagined that it wouH be of the ()rder
1.1. of 125 to 1~0 per cent by weight of the

Fi~. 1 ; i silt charge in the parent r.hannel. The

silt trouble was most acute in the off-
taking channels fitted with such regulators,
These head regulators were provided with silt wall; as shown dotte,l in the croso:: sl'ction,
but serveri nJ u ,cful pur,)p5e bjcause the draw towards the head was so great that
the silt cnuld jemp over these walls.
3. Wood's rising eill gate (overshot rezulators),

A regulator of the type which existp.d at the head of th~ !l'a1am disributary Lo\\cr
Jhelum Canal is shown in Fig. 2. The distributary takes off at R D, 21700L of N Jr h~rn ·'Branch.
It had a raised crest with a rising gate behind it after the desigll of W.)o is. The Insic principle
in this d~sign is, that top water in the parent channel contains relatively low silt charge aDd
5hould be takfn into the distributary. .
Results of an experiment on this type of head were pnblished by the author in pilper
~;.; .. ,;,'1 ,.No. 189 P.KC. Lahore. Silt takvn
,,, by 1he Head Jegulator i<; 110'4. % by
._,.,...,.,r;:;:;:;;;~~=~~~~m~~
\.
.>..\i~D ~JL i
; I..
11 !cH; I weight on the
,!,~
averagp. and si.eve
analy;isshows that the dhtriLutary
to k at leas t as much coarse. silt as
that in the rarent channel.
Alth:mgh it took very nf'arly
surface water of the parent channel,
the !>ilt conductive pc,wer of the )lead
was very high. It was due to
" ! disturbance and eddifs upstream of
the gates cau'led by unsuitable
approaches. Even coarse silt easily
oj jumped over the vntical crest and
the ga~es The Salam distributary
was, theref· re, a bad silting channel.
.It was 180ft. ",ide with 10 foot
depth for a ( ischarge of 30 cusecs
and had a slope of 1 in 2080. This
expereriment showed that· in oreet
Fis,2 to exclude silt it was not 1Iufficien1
 3~1

to take simply surface water from a parent the channt'l. but also 'imperative to provide suitabie
approaches up-stream, so thlt there was no disturbance u..)stream of the crest.
4 .. Gibb', drstributary Head Regulator.
The existing design of the Fatehpur D:strihutary Lower Jhelum Canal head regulator is a
Gibb design given in Fig. 3.
The Hearl is an open flum'} and i~ provided with Gibb's groyne wall 011 lines of Elsdon's
slIggestions described in paper No 30 of the Punjab.

Head regulator of Fatehpur Distributary off· taking R D. 83,800 of Northern Branch.

L. SZCTIC,.,

Fig. 3
Engineering Congress 1916. The basic principle underlying this design is
that the cross Wolves of silt srrould not bo allow;ed to enter the distributa,y by constructing a
Gibb groyne ~ all and the amphtude of these waves &'1ould be incr"ased by providing pitchIng
ou the side np;tream of the head in tft,} parent channel. [he wa.t~r"'ay a.t the entrance to the
chamber, enclOsed by the groynf', is alloNed comidering the effectiv depth as two-third. A hOIL:
of three feet dia.meter is provided in the G,b'j wall to (HCap.! the additional supply from the
approach cba~nber This hole: is at the hod-Iev,}l of th~ parmt cha.nnel fhere is a dry brick
Side-pitching 30 ft, loug upstream of the head in the pUClJt ch'l.nnpl.
The velocity is increased on the sides nn account of the presence of the side pitching
This reduces thll dittcrence between the side velocity and the meiln vplocity of the parent channel..
as water approacht's the hoad ; iJut incrt"ases the amplitude of the waVt;S of cross flow of silt near
tho bed, from the middle of the channel towards the sides, Cross Waves of silt, thus, miss the
entrance of the head and a~e intercepted hy the gwyne wall (Para 15 (d) of Chapter VI Part II),
Silt observations wew carried out by the author as pu lishcd in paper No. I ~9
Punjab Enginne~ring Congres. SLIt taken by the head regulator.; is about 8S'7~%
of the silt passing through clownstream of the fall at R. D. 641)00 Northt'rn EnDch,
but the silt taken is some.vha~ tiner than that in the Northern Brandl
Mlthalok Distributary is situated betvteen the Fatehpur Distributary head anrl the fall ~ilt
exclusion at tt.e head regulator of this distrir,utary influence to some extent the results of this
experimrnt. It is concluded that the design of the Fatehpur d'stributary head is just a suitable
des:gn to give a proportiona.te S11are of silt to "n off-take.
5. Klng'e sUt vanes.
1h se have ler>D used with !'UcCess : in some cases in conjnnction with design of the
distritulalY head legulator as described in paragraph 2 above. They are shown in plan in Fig 4
Tte basic principle on which they are designed is that the water near the bed of the paren
 332

channel conhdns relatively a high silt cn:lrge, Fig 9 Chapter vr Part VT, ano shoulcJ, ther~'!ore,
be deflpcted away without disturhance. Watl'r entering the distritmtary would thus contam a
a relatively low silt charge lind grade. The prindple is very souno, but it failed in some cases
where there was a violent approach with a strong draw towards the head picking silt· over the
v~a .

,
~: 30

A Fig. 4 B j . •
rhp si!t vanes are usually thin R· B. or R. C. -wall,; 3" thick constructerl on the paCca
platform as shown in Fig 4. They are sometimes metallic :1$ in Fig. 4 (B). The radius of van':s is
kept nlore than ~5 (lr 30 ft. usually they are cut short when they are inclined to the straIght
dow at an angle of 300.
The re'ght is 1 to! depth. The spac'TII! htween them is Ii time the lleight. The
number of the vanes ne! ends on the di!lcharg~ of the channel to off-take relativ,~ to that in the
parent cl:anrJel "he proportional bed should be covoed by vanfS with a minimum number
of ':t.
Th",re are nt) &ctual observatior,s on record to give the silt connuctive power of heads
provided ~ ith this device. .
6. Head witb a ~kimning platform.
H~ad Regulator of the Satehara rlistr butary
off-takingR D. 760()(I of Northern Branch.
. ..... An extract from experiments described by
He author on this type in rapfr num!-er 189
Punjab En~in~:;rinl! Congress, i~ giVen lJplow:-
(a) This type is constnlctf'd ~ t the Satghara
DistriLntary hf'ad If'gulator t;>king ofI. at ·R D.
760(10 L Northern Branch The ne~ign of the
hf'::Jd is giYen ;n Fig 5 It is an open tlurre head
~ regulator provined with a Giilh wall ann a "kim-
QJ min£' phtform. A circular groyne wall is

",j. ~
providfd with w1jnst::Jhle regulating shutters
ahove the reinf(lfl'ed I'oncret~ platform. The
depth on the pliltform is I !- ff'et ag'linst normal
:)£1 ., 0
.
supDly depth in the p:lT.'nt channel of 70 feet .
~

-.:c< . Shuttf'TS are adjusted in snch a way that .i5 no·

,_ ,-" dra w down at the Ilose of the groyne wall.
A surface float in the parE'nt channel put in
line with the nose of the groyne wall comps
straight to the nose. lhere is no fl:sturbanc6·
in the supply appro'lching the a:'proach
chamber abwe the platffllm. There is a fnrther
clarifiration ot silt on the platform by_ escaping
the supplY "traight bE-low the shutters; In.
this case the hf'ad, has the maximum efficiency
Fig 5 as a silt-excluding nevice and the tunnel below
the platform is nev.'r choked.
the experiment was carried out in two parts. The first part consists of three
 333
observations with shutters w':,rking, and the second part with th~ shutters down; ::IS, if the
groyne wall above the platforlll was a sailo one. rn th'se experim~'lt" the silt of the, ~orthern
Branch was measured at the fall at RD. 64000 Northern Branch and th~ silt in the distri,butary
was measured from th~ hydraulic jump downstream of the 'hpari of the distributary. '
, Result of the fir,t pal t M the experime:1t a, pu')lished in JHper ~o. 18-) f'. E. C, show
that the silt conduction of a lwad regulat,)r with a skimming platform wr;; a" low as 729% by
wpi~ht and that. it took distinctly fiin'~ silt in comparsicn with silt in the parent channel upstream
of it Result's of the second part show that even a skimming platform Ill1gtlt fall to achieve
high silt exclusion by its defective working.
I he waterway at the entmnce ahove the pll.tform is an imporhnt factor controlling
the entry of silt into the chamber enclo,ed by the groyne. From the following practical
eonsidcrations it has genff 11y to be more than that required.
1. to suit the ave -g: supply conditions in the parf'nt channf I,
2, to allow for ri1hi a"rI Kharif supp~y ltwels in the parent r.hannd.
3, to allow surface width between the pier and the toe of the sides so that the tunnel
is not choked.
Supply of I-he distributary is limited to the authorised discharge at the meter flume
El()wn~trearn of it, and thp excess supply turns round the nose of the groyne wall, Th;s causes
a considerah!e draW-down and d stUi'hance near the nose of the groyn". wall.
Snrface fioats from near the berm or pare,nt channd do not f'vpn enter the approach
chamb~r Water appro'lching the head gets a tlVi~t. and there is pro iuce,j a horizoatal roll~r
with axh parallel to th~ central line of the pJ.rent channel. Muddy watpr at the bottom of the-
parent ch;jnne[ tnu; enters the hfad and surface, water ,escape" into the parent channel. The
conduction of the head was, thprefore~ 89'7% in comp"rison with tbat in tIle parent cll1.nnel
Round eddies are produced jlolst at the entrance of th':l tunnel, and work, rotatiDg with their axil
perpendicular tf) the bed, btlhw the platforn RJund eddles cannot carry dt thr,Jugh th 1 tun~el.
When the tunnel gets choked to an appreciaM~ extent, the efficiency of the head a~ a ~l,t­
excluding de".'ice falls, . The remedy is to provide shutters as aCtually dom in the case of tile
head regulator,
(b) cantilevered skimming plai form.
This type is ctnstruct~d at tbe head regUlator of the Khunan Distributary taking oft
at R D. 207,000 of the Northern Branch, Lower Jhdum Canal. Tho:: d. sign of the hea,l is given
in Fig 6. It is -;imply a flume regulator provided with a centi-Itvertd skimming platform.

'-:;

Fig. 6
The head ~ook silt 72'4% by weight on th~ average in comparison with tliat in the
parent channel. It was evident from the ~ilt analysis by sieves that tat f'ilt taken by the head
was distinctlv finer than that passing in the parf'nt channel.
7. Auth.nr's silt-selective distridutary head-regulator.
The. hasic ~riciple on ,which the hpad is .d('signerl lies in th~ fact that in a fJowing~
stream carrymg SlIt 10 suspen~10Jl, the concentraclOn of the SIlt charge in the lower layers is g:oe_ate,
than that. m te u~per on~s. Fig: 9 Gh. _VI ~ar~ II. Cons(:quently, If we can escape the Lowe)
s!-Ipply Wlt~lOut hmterferlDg WIth the SlIt distrIbutIOn. the water remaining will have less silt in
1t per Ull1t of volume than the water upstream of the hearl,
 33t

(i) The essential features of the design shaull, thHefore, be to provide concentration
of 'silt charge mar the bed hy some such devices as reductioll of friction by pitchinf! or
plas 'ering the bed and the sides.
(iii Thp other and the most important feature of design s hou ld be to provide sf'parati lD
of the bottom water charged wIth concentrated silt from the top water without any disturbance.
~his requires that water should enter the approach of the head with the same velocity as it is
flowing in the canal, approachIng the work without disturbance.

(b) Design descriptfon.

A typical design of a "Silt selective Hp:ld Regu!ator" is given in Fig. 17!. The structure
cal'l be divided into three parts-approach chamb~r. regulator in the form of l{arries, and weir
fiume. The approach chamber selects the silt. The supply is regulated upstream of the weir
flume, which meters the supply. The student should refer to paper No.IS6. Punjab Engiueermg
Congress. Lahore for a detailed description.
Pitching upstream of (he Head in the parent channel on the sid~ a~celerat\!s the side
velocity, and it is further increased by the reduction of depth pr'Jvided by raising th~ bed of the
puent channel in. front of the head. The difference in the mean velocity in the Mindle and the
side velocity in the parent channel, is reduced. Water on the side, simply swe~ps straight along
the head without at;ly disturhanca. Cro3s waves of silt near be'd of the channel Lorn the middle
towards the sides are cut off by this q'lick-moving W'lter in front of the h)ad r~gulator. Th~
floor of the approach is kept higher than the bed of the p lrllnt ch~nn"l ann e)(clude~ bed silt of
the parent channel according to requirem .. nts. The pri)file of the side t; I is kept the same from
bed of the parent channel to the floor of the approach chamber as the vdtical wall is likely t'J
cause disturbance A sTTooth entry is providfd into the chamber b)th upstream and downstream
on sides. Th~ watArway at the entrance to the :lPP;O'lC'l c!1amber i:> th"! c3ief factor which
controls the silt-selective power of the head.
It is convenient to maintain the required supply level in the parent ch'lnnel hy
regulation at the olontrcil point. A needle reE!ulator ii the most su~tabll devic:! wh;ch interferes
the least with the sIt clfnage in the puent chunel a ld i1 ea3Y and c mve lient to manipulate.
However, if the distributary h'~ i-regul<l.tor i, de ;igned according tro the av-!rag~ supply level in
the parent ch'lnnel for want of a control po nt in it, ri'gulation is provid~d as shown in Fig 171
in the form of .2 foot Klrries secured a 10 bcke i in th~ gro >ve,. [n thii c \se a. straight length of
at least 'ZH is nquired from the regulation K ,Tries to the gaug·~ hole of the meter flume. This
distance will be increased aGcording to disturbance caused by uguiation by means of Kanies.
The regulator i" follo.ved by a IU}ter flllm~. fhe depth on crest of the flume j.; kept'
about i the depth in the approach cha.mber. It has g)t a !engrh of crest equal to 2H and the
distance of the gauge hole fron the beginning of the crest is 3f-{ The a;>proach curve in the bed
is hid with radim 2H. The crest is f )llowed by a conv~nient gla :is, Watp,r to the gauge well is
admitted throngh a single h Ie. the area of whfch is about 1/ L.OOO of the area of the gauge well. The
hall': is provide.l in an iron plate secured flush with the wall and lo:ated an inch or two below
crest level at a distance of 3H from the beginning of the crest
[c] EXPJrimentaI test of an existing head of t)il type.
The head Tegul-.ter was huilt at the head of Melay distributary at RD 83332 R
Southern BranCh, Lower Jnelum r:anal. fh;) d~sigtl was similar to the typ~ design given in Fig
7 .. The s:lt of)servations pUl)lishp.d in p:lp3r No 189 Punjab Engin'lering Congre3s, s'lOwed that
tillS type of h ....d regulabr took 702 % silt by weight in. co npuis m with 1 hat in the parent
~bant1el dowllstream of it. It took distinctly a finer gradr at silt in comparsion with that in the
p:lrent channel. fhi'> proved as e:ncient as a h.lad with a skimnlng p'atform.
(d) EXperimental rq ult of models.
Very detailed experiments were carried out on this tYP:1 of head nn full-size glaz d
models to trace the silt laden stream lines entering head. fhe fol!oNing conciu,j.Jns were
arrived at : -
. (i) The silt conductive power of a silt-selective h 'ad regulatol w,)rking under ideal
. conditions does not vary with the discha: ge of an o:I-tak'1, so long ai the depth in the approach
 335
~hambeT is not. changed or. in other words, so. long as th~ ~atio of the depth in the approa.~h
hamber of a sllt-selertlve head to the depth 10 front of It In the par€'nt channel is not varied.
Silt Selective Head Regulator.

t.

Seetion on A. B, C. I~" ~ ;

Fig. 7
Ideal conditions of working mean that the wi ith of the entrance to the apprl)ach
chamt)er has been suitably selected, so that the bottom wa 'er of tre parent channtJ' flows
straight along the parent channel along the sloping t to I crest and hao; 'no tenrlency to ri5e into
the appr 'ach chamber. Thpse conditions are available, where no rolling bed flouts rise up into
the approach chamber an d the surface floats indicate neither drawdowri at t1',e entrance, nor
eddies in front of tIle bed. .
(ii) Under ideal connitions of working, the silt conouction varied with depth jn the
approach chamher. In other words silt conduction vdrien :lccording to some power of the ratio
depth in the approach cha.mber to the d ~pth in the parent channel in front of the head on the
pitched floor. Results of the experiments satisfy the following rdathn : - ,
., '

[A)

where Pl = Silt conductive power of the head (e:xpres,ed a'! percentage) in ideal conditions
H.=Depth in the app~oach chamber; D=Depth in the parent channd o?posite to the
head.
The value of index x in these observations is nearly t.
(e) Advantages of this type.
li) This is a very efficient excluder of silt as compared witQ other types of head
regulators de,cribed before.
(\i) This can be designerl with any required silt conductive power to suit th~ silt
conditions which could be safe~y allf)wed in the oft-take with the aVl\ilable slope and C.V.R.
(iii I The design is very simple and free Irem the encnmbrances of groyne walls anl\
platform which often cause disturbance and reduce the efficiency as a silt excluding device.
 336

(8) Example of design of a silt selective Head Regulator,

NotatIon for Fig. 7.
D;scharge in canal U5 of th~ orf take=Q : Depth in c:tnal=D u ; Be 1 width in canal
=B ; Discharge of the offtake q; Bed width of the offtake=Bt) ; Depth in the offtake Do ;
Depth on pitChed floor in front of head = D =0'9 D"
Permissible silt conduction of the head=P )/oJ From Nomogram
Df'pth in approach clJamber = H. Hate No. X [[ I
l'r< j,>ction of th~ dowmtreem wing S=.q/Q (B+DuJ:l) ; Width of approach=W,,=K

.S ~ D where K = I 5 to 2; Radius of DiS :;ide wall of appro ach= Rl =3H.i ; Width of flume
•
= Bt (not Jess than Wa/2); Depth on cre,t' of tltlme = H = (q/ ~Bt)2/1 (from i H. to H.. )
Straight pnrti '11 up "tff'am he",mr] Gauge holc='L 5H : DistallcH of gauge well h lie from beginning
of crest= 3 5H, Length of crest=2 SH.
(b) Des gn of s lr-sel ctive Distributary head Regulator. fur the parent channel
conrlitiom :-
Slore=1 in 5715 F.S.L.=61S''! CV,R .98 BCG=6W'2
B pc"
V= OJ =(.
'D ept h5
= '0
Discharge: 520 cusecs ; B=8~6)<5=43'O ft. and v=2'29 rt. per second.
Off-taking Channel,

Discharge =63 cusecs13.=p=45; Slope=l in 4000; Depth=2'6' B=2'7x4 5...,120';

D " ," .
\'0= 1'0 v = 184 ; 1'.S.L.=61I'4 ; Bed= 11 4
lci 1'0 determine the required silt se~ective power of the head. Silt index of par"nt
channel=C,V,R,=.98; Silt index of off-take=C.V.R.=l'O
Rc= ~'\T:_~:_i__l1 off tak~_ =!_.OO =1'01
C.V,«. III parent ctlannel 9,s

A=.r!!P~~;ih-~a~('ti~a~~an~~l=.: ~O = 1925
ft=silt selective power of h 'ad: R c ,,=r 1 ·3 . r 2 3 . .\.1/6 awl asstl'ning fl =fa=r, R:=<r· 61 " A 1/6 ;
l'OI=r·&3 1.9251/6 ; .', r= 85 and 1hus rl=r ~5°'
". . , 2 /0 'I
."ote : _ott E assnmec that the s It gpde shall al"o vary as the 51 t cbarg'; bl weight.
The observations _plotted in plate 7 of the pap~r No. IStI Punjab Engineering Congre:H. LarlOre
>how that the ratio of the average diameter wa.~ very nearly the same as the silt conductive
po\\er by w~ight .
(i) Depth in :lp;)foach r:ha'nber for 85% silt selection, reI",tive to silt in the p.Hent,
1/' 1/3

p~ 100 (~''-- 'but D=.9 x D ,=45. '85= (-~..-)

-~ =.6 or H .. =,6x4'5=2·7 it .', Rl"",3H.~~8 1ft,

4e) Set} ack of the npstream wing .

s= $;-.{B+-_D..!'_l=_.?_3 (43+5 12)=55 £t

~ :l ;.;:"u .
(f) Widtb of etpproach=W ' W ~K_i_xO =1'5X5·S·1~Q=15{t.
a' a- Ha :l.,.
6g} RadillS of upstream lpproach curve=3 H.=3 X 3=9 0' say .10
 337
(b) The lVidth of tiume= 15/2=7-5 ft_ say S-O ; C from Fig_ 27 =3-03
2/3
63 ,. 63 2-91 2 I
H=(-------) =1'89 it. ; v.=-- ~-=2-91 ft:/Eec and h = - =13
8 x 3-03 . 2-9 x 8 • ti4"4
_',G = I ~9-'13=1'76 It "~ Length of crfst=2 5 H =25 xl'9 .,,4'75 ft.
Gauge hole f.)rm the beginning of crest=3-S H =6-7 ft. and Sl raight portion upstreaQl~2'5" i,f
H=4-75 ft. say S·O ft.
11. Practical lieU t(lsi of all silt-ax:chding device~
(a) Tt;s e:lsy to work out the pr )portional surface width of the parent channel for the
discharge of the offtakf'_ A surfar.e float shnuld be allowed to flo:lt in the parent channel from 15
feet upstrea'll of the hear! of the ofi-takf' at a proportional distance from the watar edge_ If it
jll<>t enters the off t:lkt', then it is taking its due share of bed and surface water. The silt conduc-
tiv(! pc:>wer of the head will be about cent ppr cent. [f the surface float dropped from even
beyond the prop )rtional share of the surfacp. width, gets into the offtlke head. it is then working
~s :l silt excluder. a'ld silt-hden water :It bed is entering th... head less than the projJortional.
This applies to all silt excluding devices di.,cribed in this chapter for distributary head r"gulator.;
and also to distributors.
PropOl;tional Discharge and silt Distributor,

...

I.ONGTUDINAL SeCTION

Fig. 8
 338: "
[h] In the c~se' cia si1t~selective distiibutary head regulator,; the sat' seledive power
Can easily be predicted by 'simply running the surface floats, DetermiJte' the distance
from the water edge upto whkJ:l the floats euter the off-fake head. It should be the same as
the width of approach allowed in thedesigo ,for .the stipulated silt co~ductiv~ power for whi~h
the head is dEsigneg .. If it is less, the !'ilt selective, flower, will ,he correspondingly reduced m
direct proDortion 'rhe fact was established in Hie experiments des'crihed by 'the author in Paper
No, 189 Punjab Enl!ineering Congress, Lahore i ,-

12, ProporUenal Distr i butors,

[a] (n dist;ibuting ch~nne i~" proportional di~Jribution of 'sllp,.pli~ja: is ~lesira'bi{ a~d ,il!
yery much appreciated by the qlltivators It i, a goon. practice to combine 1:hepf,f-takhg tI)inor~
an,d outlets into one distributor. lhe proportional distribution of supplies is' at once: arrang"q
by, m~king all off. tak"s includin!{ the parent channel into open flume wein with their 'crests ~t
~he ~arr,e level, There are two arra~gements 'usually adopted a<; shown in F~g.. 8. and another
ln. Flg. 9. When the off-taking mlDQrs and outlets are sm~ U as compared wIth the parent
;channel, ; tl:te arrangt')ment shqwn in Fig, 9. is suitable and. economical.. Wb(m the
off-taking channel carries a discharge more than 1/4 of thp. la.rent channel,' the
arrangemeut shown in Fig; ILis dEl~ire:l ble thongh usually somewhat expel sh e. .
ib] The proportional distribution of the silt charge of the parent channel can eaSily be
att:!ined in th~ arrangement a.5 shown in Fig:8 '6y dividing proportional water-way of the
ch;Jnnel approaching each flume by extending the partition walls upstream of the cre!'t. In f'let
!he partition walls can continue as silt varies with height equal to t to' t depth tQ ..r!l~trict or
lDcrease the entry- of bed silt iato any off..take according to the requirement.
When the off-taking minor is a small one as. compared with the parent channel, the
arrane-ement shown in Fig:9 is convenient and economical There is enough p.xpoimepJ<.ll work
on record (Khan Bahadur Minhaj·url-Din I.S.I:. experiments on silt di;tribution Indian
EngiAeering, Calcutta . 1928) to show that the right-angled off takes a~ shown in
Fig. 9 take relatively high silt charge a'> compared with the parent channel due to.
the curwd entry of water, This can, howf>ver, be set right at liltle cost by constructing there
a silt vane as shown so that only proportional bed silt en trrs the off-ta'ke. . "
ProportioFlal Distributors with silt vanes. Scale 1/200
,

~.
.-.-...., .. ~.--'.----.. .......... ~.-.....,

Fig, 9

.-.-~-- ).I':~:.·'·:-,~:~,~
-.,"'~"~;""::-::f:::;;:i
=;::::~:::=i:::=:j~_3ro=Fk':~;;:BiiI!i1:"~13~!~..$i'~""", <;, :.""

'- 'u., . c,!,'A-'31f--

..
 339
,
The calculations of design of distributors are given in Fig, 9, The structural details
are as per standard meter flumes design as given in para 1S Chapter X.
13. Examination Questions,
Gi'!e a sketch of an escape and a regulator f9r ;a.distributary having a discharge of 150 Cusecs a
bed width of 20 feet and F. S, depth of 4'15 ft. Thecescape lias 'to take a discharge of 100 cusecs and will hav; a
bed width of 14 ft, and wa.ter depth oHJ 54t, R, T•• of bed of,distrib,ut~ry 560'0. ~ L. of bed of escape 557'0
Explain a Walton gate:·fo,r ~pin~l head regulator (:P. B, 1:. 1937) ,,'
2, Work out and s;{etch.the design for a proportional silt and discharge distributor for a trifurcation
with the following data:-

, Righli off-take Left off-take

Parent ch~nnel '. ;'
, at 60 0 at right angle

,Discharg~ " 520 3UO 100 !120

Depth .,' 5'65 4-6 3-5 ',·a·s
Bel width '. 340'0 ,28 14(3) , lUI
Slope per thousand '2 '2 '23 -235
C. V, R. 1'0 1-0 1'0 10
Sides ii to I ~to I ii to 1 ; to,1

(a) D~scr'ibe the ba.-ic principles 0n which the problem of silt exclusillil,. j .." base!l,
(b)' Wha,t are the advantages of silt selective heaq-regulators over other types? .,
(a) Explain how you would determine by funning sur'fa.ce fl.odts in the patent c,ba;llnel ,'r.ll~i.h,er ,a
head is working as a silt selective and silt excludil'l~"'device. ,'" " "
(b) Explain the basic prjnciple on· ,which the desigll. 'of King's sile vl'nes' is based" ' "
5. (a), Why,did.\Vood's d<>sig" of riSing cill gate. fail even th'Jugh it aimed at taking, surfa,ce ,~ater
hy oV,ershot regulation? ".', " ' , , "
(b) When do the ttlnupls below the skimming platforms c11~ke and why? Sketch and explain the
conditions when th .. y will not choke. '
6, Design a silt-selective distributary head-regulator for the following data : -

Left off-take at '

Parent channel , righ t,~,ngl e,

Discharge 520 63 '

Slope I in 5715 1 ill 4000
C V. R. '98 1-0 '
Bed 43'0 12-0
Bed width depth ratio "{!'S 4'5
Velocity. 2-29 l'S4
1 . , ;:, y ..., ' _ ! , : . , _I '. ~ ~.' '." " " ~.,

. ,. '7. Whal are thiltbree iu'nctions ora d isfributatyhe'adAregulatot ? $ij>l;l.in their devel<;>pr;nelit lil the
modern distributary head-regulator designs? " ' . , .' ",' , '\' .' ',: .
 PART II
CANAL IRRIGATION
CHAPTER XV

Outlets and tail clusters

. . .1. The masonry structure throlJgh which water is admitteil from a government
dlstnbutmg channel into a water course (cultivator's channel) is known as an outlet ur mogha.
In America it is called a turn-out. An outltt may be a module or a non-module as defined
below:-
Definition.
The term module, in hydrauI:c~, was originally applied to a contri \ ance deviserl to
pass a fixd supply of water independent of \\ atel surface levels both in the supply and delivery
channels. With the invention of the gauge-outlet and similar devices it would, however, sp.em
more- convenient to broailen the term "moilule" to include the latter in \'fntions also. and to
defin~ it as meaning "a device arranged to pass a supply of water incerenden c of water surface
level m .the channel into which the supply i~ deliverEd." With this definit rn. modulEs fall into
two maIO classes:-
(a) Rigid ModUles-Passing a fixed supply.
(h) Flexible modulfs-or (Semi· module") Passin{=! a supply which varies in some
characteJistic manner with surface ltvel in the supply channel but whirh is
ind.e.pendent of the variation of the water level in the dtl,very channel (water
cour~e).
(a) Ameng rigid moilules are GiLb's Vortpx Modules, and the Kent "0 " (a plessure-
float device) ; Glllfur and KharJna's rigid modules.
(b) Flexible modules m::!y be sub divided into :-
_ (i) Orifice type-The K .G.O. (Kenner:y Gauge Outlet) and the original Harvey
Stoddard Stmning wave outlet. Crump's A.P,M. (Adjusta1lc Proportional Module). Sharma's
jmproved A.P.M. (".S.O.D)
(ii) Weir (or fiume) type. Various forms of oren weirs e g that used by Gibb for tail
clusters, short-throated flumes as used by Harvey and Lindlt-'y fnr tail clusters ; Crump's open
flume outlet., minor-he::!ds, proportional distributors, meters, and IODf -throated flumes: Sharma's
narrow open flume outlets.
(iii) Combined type - The Harv, v Stoddard Improved (proportional) Outlet, which
is essentially a combination of su I merged orifice ann weir.
2. It follows from the definition of a flexible me dule t11at a vertical gauge, or scale,
fixed rela~ive to the module can be calibrated to show the disch::!rge oi the module corresponding
to aT, y mark on the gauge at which surface level in the supply channel may ~tand. The height
in feet, measurtd to this surfact level from the zero of n,f' g"luge. will be designated by the :etit:f
G, it being uncerstood that the zeTO mark repnsents zero discharge as calculated from the formula
used to express module discharge in terms of gauge. In fleXible mC'dul(s of orifice type, for instance,
G is measured either from the centre or the upper surface of the jf-t, ace >rding
as the jet is rated all round Of only on the upper surface lind discharge varies as Gl12. In the
weir type, G is mQa~ured from the weir crest, and discharge varies G3 i2. In this combined type,
I 2,3

the zero mark is at wt:ir crest leyel, and t1.e di~charge is given by(.9 ) = (_g_) = G where K
k K
and k are COllstants for anyone module.
 341

For rigiil modules, the gauge G has no meaning. "inee discharge is inoependent of gauge.
3. With varying field levels, and with the zamindar at literty to silt clellr his wafer
course, whenever and as often a~ he wishes, thp ~u Jply orawll by a non-moclular outlet is fur
ev"r ...:hanging independently of surfac~ \evel in the s'~pply channel, and thereby affecting the
general distri ution of supply in a ma''llttr entirely l~eyonrl the control and manllgement of those
responsible for distribution. On a moduled channel, on the oth ... r hand, d:strihution IS render~d
entirely indepenrient of the arbitrary ch:>nges in water C()urs conrlitlOllS, and is dependent only
D
,

upon conilitions in the supply channel5 under G'JVernment control. Th;s gfl'!at llilvantal e of the
monule is by O'o,w generallv r~cognizQd : the old non,n0dular outlet, except a,; a pu~el'y te ..i~)Qrarv
expp.dient is noomed to disappear, and in what follows, it will be tac: tly assumed that we are
c mcerned only with monult'd channels.
4. Following the defin'tiop of the term TT'odule of paragrllph I, all mGdules-whether
rigid or flexible-may be more precisely classified. in relation to the supply channel, in term.;
of a single characteristc ratio :-

r= ~q-/ ; '!9..
representing the ratio between tte fractional deviation oq in the normal supply
q Q q
q of the modu'e, and the correspondia.g fractronal devia.tion

-~Q- in the normal supply Q of the parent, passing below the module. This ratio 'r' will be calkd

the FLEXIB [U TV of the modllie. For a rigid module, the FLEXIBTLITY is zerO. For
orifice types lik'! the K.G 0 it is usually IllS, than unity; while for weir [or flume] types it
tends to l;e greater than unity
(b) When the "r" of a module is just unity, thp. module is proportional, i.e it shares,
proportionaUv with the parent, in any slTlall deviation in the normal surply of the parent; so that
on a distributary fitt,d throll~hout-including the heads of its minors--with S1.1 h modules diurnal
fluctuations of supply would affect all parts of the distributary to the same extent i e
by the same percentage inc' ease or dt crease of supply. so that if these fluctuatiolls
were the only difficulty to be overcome, the pmportional module would offer a simple hut ptrf, cb
s')Jution of the problem of distribution. fhe proportionai s 'tting of the orifice type semi-modulel i"
when G='3 D and for the wei! type outJe·s is when G='9D where G is the gauee and 0 the dep1 h
in the chann!'!l The proportional setting of the crest of the outlet, therp.fore works out to b3 6/10 D
and 9/100 in the case of the orifice type and weir type outlets reop ctively. The mlthemltiCJl proof
of this was workerl out oy ES. Crump in lois pap~r No. 26, 1. B Publication Punjab.
Ie) The problem is, ho\\ever, more c(,mplicated, as changes in channel regime, coupled
with the usual rlls ..-ridion of hearl supply to within a prescnbflrl ma dmum lim;t. intru tuc~ a
secone" difficulty. with which the pnporti onal module is poorly adapte! to cop" [he rigi t
mooules, on the other band, behavrs in a directly 0ppo~ite manner ; while completely
imm'llle from effect of regime cIllnges it take.; no share ...,hatever in
fluctuatuatio(ls of supply. In lact. the proHen 0f distribution presents two
independent difficulties the requirements of which are in direct opposition: the first is met by
proportional modulrs, and the sec<'nd calls lor rigid modules Whether the K. G. O. or the A. P.
M. set at bed or belr,w or modules of simil,r f1exiblity, which occupy an intermediate position
betweell these two extremes' offer, for this reason, tJ e Lest solution of the pro'.Jlem, ,remains to
be considered.
5. Every rf'ach of a channel pnns in what rnav be called a "COlttrO} point ," that is to
say in a masoD>lTY work usually so desigued a;; to maintain (a) a permanent relation between
I

upstream surface and di<;charge, and (b) parallelism of ~urf,.jce and bed lines with varying
discharge. Fluctuations in supply may, thfrefore. be regarded a" causin5 a general raising or
lowering of the smface line without affecting its slope. Regime changes. on the other hand,
result in a steepening or flattening of the surface line, and their effect on disHbution is rearlily
appreciated by visual sing the surface line; of llnV r~ach as swinging slowly about it'! downstream
end i. e th~ "control-point" as pivot. All off takes in the reach (l,re affected in the same WaY ;
but to an extent increasing with their distance from tl e 'control point." Th~ 1"1 ach as a whole
benefites or loses, as the case may b accordjn~ to the los~ or gain of the reaches below it. The
A
,

advan ta€e of frl:qu~nt control points in reducing the effect of regime chang~s is at once evidt nt;
the effect is nduced inver~ely as the n11mber of sub-reaches into ~hich a silting reach is divided,
 342
.- .~;

\ 6. The sensitivtness of a module to regime changes 6h~iously' varies nirectly 'with its
FLEXIBILITY as defined in paragraph 4 above. It will, however; lYe useful and cond:uciva,
to clearness, to distinguish between the two term> and to oefine the sensitiveness S of a modul.e.r
as' the fr'l.ction increase(or decrease) of module supply per his,;a ris" (or fall) in channel surface: :
With th'is definition as per ct:ump's papor No. 26 Class A,P.W.D Technical Irrigation Brandl.'

Publications S = _L_ applies to all modules, and enables the precise effect of a, given regime,
6D .' . ,
change to be calculated in terms' of the FLEXIBILITY of the modules concerned, and of the,
normal depth D of the c h a n n e L . , , ': ' '
For e(ample : applying tl,!e above relation to the case of d head reach of 2·S feet average
depth, in permanent regime, and subject to an averag~ surface swing of±O·t foot; the percep.tage.
season~l variation in the total draw-off of the re::lch would be abollt±6 per cent f01: pr()portiona.~
modules, ±'l per cent for K. G, O. Or other .orifice modules having a setting (G/D) of 0'9 (a falr
average for such modules) and nil for rigid modules: Tn other words, Ie G 0s were used in th~
h~ad reach, no re-setting nf them would be called for; ;vhereas if proportional modules were subs-,
htuted, they would have to he readjusted every three months or :sO, to maintain a well-balanced

7. The idea of adjustable modules is not neW. wifh nrm-mod ular outlets •. :r~e~ad}llst:ments
cistribution throughout tl;te year:. . ,
,
meant reconstructton. In the cae, of K. G. 0 .. it t;lkes the form of resetting on a masonry \'Jr
c?ncrete foundation. By adjustbility is meant the provision of some suit'fhle means of altering :the
s ze of the orifice'say, of a monule, with a minimum of trouble to the se empowered to make
f'e-adjnstme'nts, but with a degree of difficulty suffi~ient to prevent re-adju,tment being· made
il} :>n ·ilIegal or unauthor:zed manner, by the zamindar or petty official. In Crump's opinion, the
nght mean between facility and difficulty is to b" ol;>tained by means of a key of solid masonry
or concrete, substantial pncugh to defy tampering, hut at the Samfl time' small enough to be
reAsonably cheap in renewal. Such a key, WOUld, of course, have to be dismantled and rebuilt at
each adjustment of the module. It is assumed and regarJed as feasiblp. that re-adj ustments
wo~l(l be carried out under the sanction of higher authority, by the Sub-Division?-I Offi~ef
concerned, and that he would be.held responsible for the correctness of the re-::IdJustments
: made anti recorded by him. As regards the rreq'lencv of the re-arljllstm~nts, it must bl.'!
• remembered that all regime changes:"'an'd it h tht'se aJJne that call for atiju;;tahility-take place
slowly, so slowly tha,J:., even if propo.rtional mcidtlles were used in head-reach~5, ,{eadi,!!stment
would ordinarily be neces'ary not mort' frequentlY-than once a crop. ClUmp brought ,', forward
, tne idea of adjustabiiity alld proportionality and therefore his outlet is ;called .'\djustable
, Fropo tional Module (><\.P.M). ' , , ,
! 8, Modul:u limits and minimum modular held (M:M H )
AU modulES, whether r:gid or flexible, require a cNtain minimum head to ensure
, h~odulatity ..In th~. ca~e ?f rigid module~. there is alio an UppH limit beyond which const~~cy of
,dlscharge falls. ThIS ltmlt must obviou~ly be high enough to avoid trouble, within the range of
running conditions. In all modules the advant:ige of low M. M, H is self-evi"'ent, since it means
, le~s. evpense .in earthwork both in modulil1g ~xisting channels add in const-ructmg Lew one~ with
";a VieW to modul<\r equipment, and.in the fl)rm~r cas dess distilrbance of. existing regime. In
~ m od,.l~ng existing. channels it is usually found n<c·:s,ary to raise channtl,~udac~ levels iI_1 order
, to obtam madulanty for the worst-commanded out13ts. This is du'! to the fact that wI1:,h. "the
, old non-morl,t11ar outl"t, the importance of insisting upon an ample m ugin of command was 'not
~enf'~ally ap~reciated by irrigation offi::er:;. l'heir ind Heren ~e h'l5 commonly resulte l in the
Ze1mndar .bem:{ a'l~wed -and even encouraged.-:-t() hring 'Under Irrigation, l;l.nd which, by
,rea,;on of ItS elwatJOn, should. iJl the intersts of e:TIcient distribution. never ha.ve been adm'Ued
~oirrigati.~n by. f.low. N~e(lles~ to ~ay, a continuence of. thi.s. easy-going attitude ill quite
lI1c'lmpatwle "Yitn the modular eqUIpment of channel on sClentthc Iml:ls. '.;'
9 Essential conditions to be satisfied in outlet design. "
, ', .. (a) A~ .ou~let must be strong an,d ShbUld DJt have moving parts liable to derangement
, (.r reqUlrmgjper,lodlc attention.' , , , , ' "
, ::'." i(bL lxgerf~ence by the cultivator m;1st be difficult, and if made; should be rea,~ily
,defec'l'ble;,.;,; ,;:', ,~. ; J, ~ ' •• ..: .. c, "; i ' '

(c' The outlet should draw its fair 'share of the silh:arried l y,the:. piu::en;t. (}'9&pJ:!.eJ.;,'
(d) It should be p ssi'1le for the outlet to work efficiently with a. SIT a 1 working head. 'rhe
 343

outlets

Fi~.1
. l>15tHARGE OF CAi'JAL OU rLETS
Q=KAy' h

5
I:

~.
,..
., i "t
.
., .
.i
I' . ,
. •.
., .f ~

•.,. ,
.
. ".'
EXPLANATION
, 10"'" A~It.U ~C.U. " ••..
4
'1.;

.,c •• ~ .,.... ~.,/IU "!I

.4 J
Ii;. SA.re..

... ,:, ,.(',IfT FM ,.014 ,.,,.E '?' ewllf

...
.....
.
"" '"'"
e-= .......
.,AH.r,,, All. LEN.'II Wlrtf ~

..... .. ' NUN .If $CA~' ,elt t..

~
...
::! ......
!l
''I
.,
.~

'"-= • ., Q
"

~
.&
~
~ ., .y ::t
,." .,
~ .,1

"~

j4

. :,I··
-If '{

: ~ '.... .1' .'

1,;
;; ......

•1 .

...
.. ..
.
 344
distrihutary. They may be circular OT rectangl11ar in section .. The
discharg
form 111a is Q= K A v h - where Q is the diich'uge in cu>ecs. A is the sectional area in sqnare fPoet
ano h is the working hea i in f~et (Difference between the water level ill the parent channel ami
the wateT ievel in the watf-r c mrse) and K is a constant" hich :is taken as 5 in the case of masonry
orifice" and as 6 in thPo case of wooden and iron orifices. There j<; umally providpd a face wall on
the distributary side as shown in Fig. I and drop wall at the end of the bank in the case of
pacca outlets.
The wooden sho )ts are usually put in during the con<;truction of a ch'lnnel Thpy are
simply put in the bank with puddled earth around them. They are meant to be temporary and
kacha out lets and are replac~d Ly permanent ou tlets aftH rnnning the channel 3 to 4 years when
the chak boundar fS are established by usage and some idea of the actual water levels in the
water c mr,e" is availabh·.
It is apparent form fg I that th" discharge of t~e outld can eaoil. be increased by digging
the water course ~nrl thereby JO\~ering the water level in it. Thi~ Will increase the value of h
and. therefore. the discharge Equitable di,tribution is not p'1ssible with this type of outlet
when the cultivators can increase the discharge by simply digging the water course which is a
::aminddri channel The solu'ion of the formula Q = 5 Av'h is giv.-n in Fig. 2.
1 [.Tilt~ pipe Outlets. (Free fali pipe out1ets)

The masonry ot pipe drownf>d orifices as desceibed above were re,>laced hy tilted pipe
outlets as sketched in Hg 3 The pIpe is supposed to discharge f 'ee fall above the ~ ater lev-el in
the wat~r course. The p;pe may be tilted as shown in Fig. 3 or horizontal
Thf! discharge furmula of this outlet = Cd. A V2gli- where (.1=06 co-efiicient of the
Aischarge: A= sf'ctional area of the pipe in sq ft.: h=head from F. S. L in distl : to the centre of
the pipe at the outfall.
fhis type of outlet is no coubt a ~erni ·module and the discharge

TI1_T &:.0 PlltE. C)U'T LI.T / c:Jnnot be it creased by digging the

water-cour~es. This succfeded
very well for some time. but the.
(:\-~.:=-~I!.!itJ~'" cultivators Soon invented the method
of constructing a ramp in the water
conrse by heading up water to the top
level of the pipe. The discharge
Itoo, Ot• .,:
increased by 15 to 20%
Fig. 3
The long pipe outlets do not usually rlln full at the exit and the r1is~ha.rge eo·efficient
is mostly a measure of the co efficient lit the contraction of the free fall Jet at . the v~na
./'1
contracta By drowning the pir e. it star ts running very nearly full ~ore. I he contwl,sectJO.n shIfts
frorn the vena contracta of a falliIl~ jet to the s. ctional area of the plpP. a~ t.he eXit. 1he dlschar~e
c,,-efficient inc. eases to L'S and even more in some cases, I he d;s~narge IS lllcreased e \Ten th:mgh---
the head is lost tqual to half the diamtter of the pipe by drowning.

12. Kenreiy's Gau~e Outlet

Th is was the earliest type of semi-module used in India. I~ is ~ketched _in Fig_. 4.
I t was invented hy R G. Kennerly. Chief Engineer, Irngatl?n. Pun~ab. I he outl~t
is PJo\'ided "ith the bell·mouthed approach into an exp:mding cellvery p.l~e as shown m
hg 4 .) he jt t in the air chamber is a. rated all round. 1 he discharge formula
q=cd y2g A v' G- where G=the head mea!'ured from F, S. L. i~ disty: to the centre of the
PI ific' ; cd= '97 = Discharge co efficient; A =St ctional area of the onfice at the end of the month,
 Fig. 4.

This outlet required in. practice a Minimum Modular Head of about '22 to '25G and its
discharge was independent of water level in the water course when the working head exceeded
M. M. H.
This m'Jdnfe ha~, however. bep.n supflfo;eded became its di5charge could easily be
increased by closing the air holf's feeding the air cha.m ber. After the air holes are closerl, the air
of the cham 'Jer is sucked away by the jet and it..; prc·ssnre in the chamber drops below the
atmosphere The pres·;ure head cauging tlow increase, to (G + '\ -A') where A is the atmospheric
pressure and A' actual air pressure in the chamba below <J ·.mospheric pressure.
13 Harvey Stoddard imprJvei outlet.
fhe Har \ley Stoddard improved oatlet (Fig. f) was an early design to attain proportion.
ality. It consi"ted of an orifice outlet coml)ined with a raised standing wave flume iri
L. SECTlO:-'; continu::ltion. The loss of head was
ToP 01' wALL relatively more than in a standing
wave flume. M. M H. was about
22 to ~:; 'j',.
A standing wave flume gives
proportional discharges with a
small loss of head, so long as thp.
;;3 depth of the parent channel is small .
•• _.: 0.'50 j._-- /.77 - - - + - - - J"7--~LL___ Whtn however, the distributary is
Fig. S. a deep one, the flume, which would
have to be approximately the same depth Ito he on the right side of proportionality) would have
to be made so Il"rrow that practical cOnsidf'rations would rule it out.
]4. Kirkpatrick Outlets. ,
Jamrao tl/pe open flume is shown in Fig. 6 atd the Jamrao orifice semi-module is shown
in Fig. 7. Th" open flume of Fig. 6 has no level crest and has a short throated section about
2" or 3" wide, where an angle iron frame IS nxed. The up,trearn approach is 2 ft. long and has
a splay of 1 in 4 1 he downstrt'am apprna"h f1urne is 10 tr lung ::Ind the width of the flume
at it~ downstream end is ~~ H. hi, kp<11J L k f ULd that the shapes of
the upstream and tho downstream !lames were n)t cap::tble of further improvt>ment
and that any further increa~e in their length~ or
L. SECTlO2'< I .
iates of diverg"nce cou rI not Improve upon this
w_.
o
tyr>e of nesign. Th" values of eo· efficient K in
- the formula q =:= KBGl.5 was not found to be
_ constant for f1iffc'rc'nt values of G but if the
inde~: 1'5 were changed to j'6 the formul::t
agreed within a per cent of the actual observa-
tions. which c;)vc'red a wide range of
both Band G. fhe discharge formula as
pvolved by Kirkpatrick for this tyPfl of open
flume outlet is· q=3'2 BGI 5: The M. M. R.

found to bp necessary is= ~

It will be seen that the M. M. H. requil"ed
Fig. 6 for this outlet· is more than that required for
the Crump type of open flume outlet,· also the
 246

narro '< as the open flume outlet suffers from t,:"o ot?e~ defec:s : -
[a] The control section does not remam wIthIn the angle Iron frame depth on crest
i ncreasf'S.
lbJ rhe'e outlets are more sU'iceptible L. SErION
t() gettIng choked with jungle than the long
thro3'ed open flumes. W.L

Orifice type outlet in Fig. 7 is an

attempt to reproduce a K. G. O. in
Dl<lSonry. The upstream approach is only
:l ft. long and is of the shape of a truncateti
Cone with a convergence of 1 in 4. The
contr"l section is a square orifice in an angle
Ir nn frame. The downstream flume 10 ft.
Ion!?, has a horizontal floor with the wide
\\ alls at a splay of R/lO. The most novel
fe" tUft- of the Jamrao semi· module IS the
introduction of baffies as a means of r('cover
Ill.: head. The baffies work as a "roof" slop- ~--~
iag gradually upward, in the downstream Fig. 7
l1ume and by tlleir addition "considerable extra heari is recovered" At the same time the outJPt
tii charges under free atmospheric conditions. "Even if an aperture is cloied, the effect is not
cunsiccr<lble as air will enter from the next aperture and so on." .' rh~ optimum l!ope for the
r<)(Jf i,e lower edge of baffiEs} is 1 in 15. fhe optim'lm number of baffles i,; 9 of which the first
(~ are equally spa.ced and the last 3 somtwhat spread out" [he dis:::harge formula
ilppl'cabie to this orifice semi-module was determined experimentally, by
h.irpatrick and 15 q=7'2 B Y V'f(;- As the or.fice
is sqnare and B=Y the formula may be re'<ritten as q=7'2 8 2 y'He -dS will be apJUe.lt from
the drawing. This type i; not at all adjustable. Tne minmum working- head required is taken
from the table given below:
TABLE NO.1
Siz~ of Module. M.M.H. with 6 Baffles. M.M. H. with 9 Baffles
He He
8"
4'1 48
7"
He He
4'1.) 4'7
He He
39 4'0
H He
~:F7 4'5
4"
He He
~'5 4'4
He He
3'3 4'3
15 C.. ~mp's Adjustable Propotional Module.
The setting of the crest for proportionate discharge is 6/10 D. The structural details
aTP expHned in Drawing iri Fig. 8. The immunity from tampering is provided in the form
of cast iron base ~nd cheek plate 1 0 foot wide with a cast iron roof block. The adjustability
of the outl·t IS arrangeti by providing bolts passing through the roof block and the cheek. Tnese
lJOlts can be removed after dismantling the rnas:>nry on top of the roof block and it can then be
lCl1,ed dud Ic.,wered according to reqnirements. The masonry above the roof block can again be
jlll' i'L l'he ~st of adjustment of the size of the outlet is very small. It is an orifice type outlet
"Ifh tlie additional advantage of a<ijust;J.)ility and proportionality. Thc discharge formula used
I~ ~iven below ;-
 247

q =edy' 2g B. vv'~ Where q= Di~charge'in cuseC5; cd=O 91 =discharg~ co·~fficie It:

B=Width ; y=Height of opemng; h=depresslon of roof block. H=h+y=depth on crest.
The head" h "is measured upto top of the jet leaving h.e orifice under the roof block
T~"re is levell~ngth H provided in be'tl after the roof.to ensure that thf jet of water Ieavin;
with y depth should exert pre,sur~ on the floor. It IS, therefore a submergp.1i orificp.. It I~
semi-morlule becauso the dischuge is supposed to be indipendent of th", condtions in the wat .. r
course and tile m ,dularity IS indiclted by the forllution of the hydraulic jump downstrearn ot
the orifice A.P M. is, tberefore essentially a submerged semi-module .
Crump in his experiments suggested ;:t coefficient Cd=cdy'2£ =7'3, He al<:o gave a table
(f th~ minimum modular hmits based on his experiments in paper No 26 P.W 0 Irrigation
Rranch, Punja=> pUJlicatio[ls c ass !\ His e'(p~riments werp carried out with width 0.5 ft, ann
depth on crest '2'0 feet It was soon found th'lt the co-effitient C and M.M.H. as given by him
did not apply to all cases. :::rumo's A. P. \'Is. have been manufacture i and used in eight standard
width '2,25, '32, '4, '5'63, '8, and 1'0 ft.
This outlet is used all over the Punjab This has proved to be a very succes<:full
outlet on account of the immunity to tampering and tf-te adju3tability. The cultivators have been
attempting to increase the discharge by raising and tilting the roof block [his defect is ven'
easil~' (etected and set right. They sometimes construct a covered drain behin~d the side wall
below water level in the parent channel and then drop it In the out-fall downstreo m of the ro) f
block by piercing thf' side wall which is only It brick thick. Though difficult to detect tbis
defect in the supply turn;;. it is ea'iy to ciscovtr it in closures.
TABLE No.2
Sharma's Table of silt conductive po :ver of outlets.
~....
o·
Z....
Q.)o..
_
~E
,_::=
....
·U
v CO -
I-< ~
Vo ~~
o..-:_
mJ)
cd ....
co

i-·-jo----oo--- - - wti 111 l~o 1,j~ 114 t!4 lH lh

1'00 92 100 llO 124 130 112 112 112 112

'9S 93 100 109 122 123 111 111 III III
'96 9~ 100 lOS 120 125 llO 110 110 110
'94 95 100 107 liS 123 109 109 109 109
'9! 96 109 107 116 121 lOS lOS 108 108
'90 97 100 lOG 114 119 lOS 108 108 IUti
-S8 9S 100 106 112 117 107 107 107 107
'86 t}S 100 105 111 115 107 107 107 107
-84 89 lOll 105 110 114 106 106 106 B6
'82 99 100 104 109 113 103 106 lOS lOS
'80 100 100 11)4 lOS 112 106 106 106 lOS
'78 100 100 103 107 III 105 105 105 )U5
'76 100 100 103 107 110 10:; 105 1()" 10:';;

Note [iJ The silt conductive power of the outlets is expressed as percentage rtdat ve
to the silt charge in the parent channel determinerl in Paper No. 168 read ia the Punja~
Engineering Congress 1933, Its variation with respect to the quality of the silt carried in th ~
channel wa~ noticed iu the ~xperi~ents described in paper No. 168 and was varified by analysing
distn butanes knowo to be III regime,
 3-18
, "-'" ' . ' . '
.' -
[iiJ -\.P.l\L~Ailjustable Proportional Module.; I.B.=Irrigation Branch; S,S.O;O.=Sub_
merged Semimociul Orifice Outlet, J
[iiij In the case of pipe outlets set at bed level the sil_!_ conductive power drops to

about 90% with all valuf's v __ , if vertex i$ formee upstream of the outlet in the rar n t Channel.
V<l
16. The Author'; improved A.P~M. [Submergp.d Se'Tlimodule Orifice Outlets].
The author carried on very exhau,tive r,search work supplementing
the work 'a\ready done hy crump fo' all width" and depths on
crest hom 1'0 to 4'0 feet Tbe rt~s11lts were pUblished in papers No. 168 and No. 176. Punjah
Engineering Congrpss, Lahore, The first paper gave the meaSUlements of the silt conductive
L. SEcnO"l"

I:. <j

PLAN Fig. 8

.<.,

:.

power of the irrig"tion ()1]tlf,ts as given in table No 2 and the second one investigated the
hydraulics of the A p, M.e outlp.ts. Whilp. discussing tbe paper No. 168 and 176 the following
conclusiops were aect'ptfd unanimously hy a committee consistillg- of E. S.' Crump and fl. W. M.
Jesson ~uperintendi[lg Engineers. and tbe author, pubhl ed in Appendix I o~ithe author's paper
N'Q P. E, C, LahOJe ,
(i) The channel fitv,d wjth A. p, M. out Ids set at 6/ J Oth had It tendency to silt up
because A. P. M. outlets as constructed did not take their due Sfl'lTe of silt. The propnrtional
silt conduction wall not enough becau~e 10 to 11% water waR lost in absorption arid the due
share of silt to be drawn by an outlet was considered to be 110 to 112 percent. The improve-
mpnts made by the author in the approaches of the A. P. l\1, outlets as sho""n in Fig. 9 were
sati;factory.
(ii The rli~charge co·efficient C in the formula of an A P. M. Ip=CByJ-h) was· not"
cnJ)stant. '·S, bllt showed considerable variation from 6'2 to 8.4. In the case of narrow outlets
'2 It. wiilth With H=4 0 and ),=20, the co-efficint droPPt'd as low as '6'~, "·In Crump's outlet
the CO-t;fPcient decreased with 1he dec"'ea,e in width 'and with the 'increase"in depth on crest:' .
. [iIi> The Crump's A. P. M. r€:ised to be a submerged· semim'ojule .hen the hydraulic
jump took place on the glacis b,yond the leve IfJ.oor of length H downstream of the roof block. In
such cases, the co' efiicien t :is high as 8 4 was recorded, In this case the mere formation of jump
dId not indicate modul:'rity.
 ¢I .., I~0
~,

,"
~
~
"~ "'l
::Q
..... :a:"' ...0
II
lo.
.
t; ,: .~.
~ ,, ....
,--',"

>~
~ ~
.~

, Q» ()
"
.~
PI 1"\
., ~
~;
c:;,
>:
15 '" l:iI,
~
(;)

-,
t ..
0
-"'~

~~
."
.'
~i
~
i:;.~
~
C'-
~J~

(\'
- ~-"I,

~ r.:
-..,.'.
>')

,~:"
...
,: -(:_"

!5 <:

)0.
c:
5-
..,
0

-
(I)'

3
"0
"1
0
<
ell
~

>
------:--... _----
, .
PI
It
:r
'i:!
~
if)

'Jl

~:! ,.
0
=:~ ,......-
o .. g
"," I
.lI-
I • ...
T
..
'. ,

~-4.. ••

;;-1
T~
;.
.A.
~
n
(/)

P>
~
"t. o.
N
II
-~
en
0

\
., <: : I

.....1
"'j
 T
. I
~.-2$
~
~·.I,.

---+ . ., I..
~ -..o.....!--+-t

OJ

"'"
~
I
_J_

-~-::.L.,;"
t', iflciillllIf '¢\Ii"t *f-

/'
/

. . .
II II II II II II II II II
~17 ttl ::r:: "'.Q 0-10 Cl- 0-
-1 II Il Ao -
~i~ "'I
(I)
0.
q;;1.Q
+t-:l ~I
V)
 [ivJ The minimum modular head required for the Cmmp's A. P. M. was founc1 to he
considerably more than that given by him in paper No. 16. When the outles were narrower
than 5 ft. and also when the depth on crest was more than 2'0' M. M. H. was r"ughly :-i3% of H.
(b) Fur.ther exppriments Were carried out by the author to att::lifl perfecti m in thp, points
Iii) to (iv) enumerated above nnder the supervjsion of E. S. Crump Superintending Engin'.er,
Upper Jhelum Cana.l, at Shadiwala and the' results were published in Paper No. 237, Punjab
Eng-ineering Congress, 1910
In the case of the A p. M. outlets to ensure modularity by visual examination of thp
hydraulic jump, the following conditions must be fulfilled at the control section upstream of
the jump: -
(it Depth not mo ethan 2/3 H (or allowing for frictionallGsses in the aproach not more
than 2'3 KI total energy
{iii Filaments bey"nd the c >ntrol section either convergent, :tdherant or rectilinear.
(iii) let leaving the orifi ~p aciherant. chat is, in cont::lct "ith the bed.
The first one W1S fulfillec1 in Cru ,jJ's design by limiting the valve of y not greater than
'SH and tho last two conditions were not attaimd as indicated by the low pres,ure pockets I eyonri
the control st'ction under the roof bock at it" end. The jet leaving the orifice was therefor.. i I hIe
to pressure inflations and capable of ::Ittaining a coefficient of 8'4 even beyond the theoretiral value
of S'O The'ip. deftcts were rectifie,i in the author's improved A. P M as !'.hown in Fig. 9 ly
making the following changes with a view to rpduce the coefficient variation and also to make it
work as a perf~d submArged semi-m'ldllies.
(i) The curve of the roof block (Bprnaulli's Lemniscate) was giTen a tilt of 1 in 7~­
greater than sbpe of the bed to ensure an adherant jet and to reduce the variation in C by giving
a bell-mouth appro::lch upto the control section.
fiiJ The level floor was removetl and a slope of 1 in 15 was given in hed to ensu e
rpctilinear or convergent adheran~ filaments in the flow beyond the con trol section indicatmg
pressure on the ben equal to y or slightly more than y and-in no case less than y.
Ciii] The value of y sholllc1 be between 0'2H and O·SH.
The upper limit of the value of y wa~ fixed by Crump to ensure that the outlet should
run full bore with no vertical starvation of the control section and the lowpr limit of Y='3H was
introduced by the author on the Lower Jhelum canal in 1930, because when y is less than this,
the hypercutical jet becomes non-adherent and the discharge increases In the author's improved
A. P. M. this limit is y=·2H.
The defects as given in [i J and [iiJ were completelv removed. The co-efficient variation
was very much reduced as given in plate No; XVI A. The variation of C dne to H has now
totally disappeared and the co-efficien t now varies only with the. size of the orifice i. e. ratio
B
Whatever be the position of the jump, the discha:-rge does not now vary so long as it ]s
Y formed
Coefficient Variation.
B '2 ·8 1 2 3 4 ·5 6 8 10
Y
C 7'0 7'28 7'43 7 54 7'6 7-76 7'82 7'86 7'87 7'S8 789 7 90
[cJ Ihe minimnm modular he~d is comprised of the followiug losses in the A. P. M.'
outlets ;- ,.
(al Loss in the approach upto the control section indicated by the coefficit nt C.
(b) Losi! in the hydraulic j u m p . , . .
(c) Loss in thQ outfall jn expansion. . .~ .
(dl The frictonalloss from the COn trol s~ct ion to the water course.
Crump simply stated that the math_ematical formula for the losses was very
cumbersome. The author worked out the mathematics a<; published in the; I;'unjab Engineer; Th-e
,~, C2 :W(k)-2k;3'
Mall, Lahore, January, 1940. It takes the form hm=h(l- -) +.054· - . kJ h t3.6 f}
' . ... : 2g F (k) . ( .)

~2~:
~¥~2
.h+f2h where hm =M. M.'II.: C==Co efficient in the f{,rmula of discharge; k=v/h=Height
-
 350

of orifice/Depression of roof block; £1 =a factor for outfall losses=5 ; £2=factor of friction loss
in outfall='15 tn ,25:
F (k) = i Y i{iq:. :'::1>1'1 C"k - k) An approximate formula as sometimes suggested by mere guess
work in the form hm = 75h is used in actual practice.
The author analysed a lot of observations of M.M.A. as published in Paper No. 237 and
found that the outfall losses were more than 50% of the total M.M.H. It was thought desirable
to have a standard design of the outfall. The bed prof Ie has already heen fixed and tbe side
walls were carried straight for 1.0 ft. length be~ ond the roof block and then expanded with a
radius of 25 ft. on both sides to the width of the water course in all cases. M M.H. observations
are not comparable unless the outhll design is the sa'1l~ anri <:tand"l.rdised This formula cannot
be used III practice but actual observations Were plotted in Plate ~('. 4 of paper No. 237 and are
noW drawn III plate No; xvn tAl for use of the students, M M.H. is roughly 20 to 25 per cent of
H
(d) Though the M. M H. re~ults as given in Plate XVII (A) show improvement over the
Crump's A. P. M. outlets, they are still very high. In some cas~s it is necessary to put an A. P.M.
in cases of water courses with. poor c .mmand. It I.\'''S investigated in paper No: 2a7 tn r"duce the
M.M.H. further by covering the outfall beynnd the straight sides with a flat roof with cluved
edge in the begining as shown in i<ig 2. The f!:duction in M.M.H. was recorded by about 30%
and the results are given in plate No:XVII B for me of tM'students This aovantage is gained
by tne addition of the surge chamber in which the jump takes Place under controlled conditions
Better re,ults arfl pos",ble by improving the design of roof covering into a gradual f'xpansinn. ,~
(e) De3ign of an A.P.\1.-Crump designed a nomogram with C=7.3 giVing the solution
of the equation q =C I3Yv/h" but C is variable. fhe author designed a nomogram for the
varia.ble C ano published it in paper 'lio 176 Punjab Engineeing Congress 193. and it is produced
in Plate X \; I B tor use of the S1 udunts. The method of its use is explained therein by means
of a key diagram.
With a slide rule. the following alternative method for the design of A. P M. may bc
of interest. Fix H according to the required setting of the outlet. [0 de1t:rmine the size of A. P.I\i.

keep h=y = H/2 and calculate B = C?Vh A. P. M size. having width of the flume equal to B

or just higher than B (:alculated) to suit the standard width. will give the required width of the

A' P. M. Then calculate. a= C~

yyl1 =a
h+y=H
i

'( .
,11
"

CAl
(BI
,,(. ')

1'1.
"I; ;<!).J1{;'
'.

<
d;.
:.j(;:"
;.{;
.!
The solution for equation (A) and (B) is given by one setting of the side rule. This
method is not only more accurate hut also quicker than the Nomogram method.
Rule. "::)et the reversed slide to 'a' on the 0 scale. [he values of hand yare given
under the cursor of the A and C scales respectively.
Proportionality versus rigidity. .
,f) hoportionality in an A. P. M. is a great scientific achievement. and is an asset of
incalcul&ble value in outlet, of regime channels.
It is the considered opinion of Crump. who inventeri this outlet and the author. who
developed it. tha.t proportionality shonld not bl'! s"crificed on regime channeis. which could be run --
nonsiiting by suitablt grading and silt selection at distributary head regulators: (fhe author's
paper No. 189 Punjdb Engineering Congress' 1936).
If a channel cannot run in ree-ime, proportionality;n its outlets become, a disadvantage.
Once it is discarded, there is no limit to tile lowering of the setting and to make outlets
suh-proportional., i.e, rigid. The author's S.S.O O. may be set even below bed level, if the
available working head permits. [he Uge of an S.S.O.O. approaching a rigid module is far superior
to the known dev-ices of rigid modules with mo\oing parts such as Kent's,
(g) Practilliof cultivators to foul the control Section.
I he 4I1O·eificient of discharge Cd in the formula of the discharge of a submerged semi-
module is , he product of Ca and Cv as in the case of Vena contracta of a free fall orifice. If
the co-efficient of sectional area Ca and the co-efficient of velocity Cvare changed. there shaH
 351

be a correspon iing change in the discharge even if the jump be for n~d and the control section
be convergent adherent Ca can he changed in two ways.
Ca) Vertical starvation fhis means that the outlet does not run full bore. This happens
in Crump's A P.M. when the depth on crest is more than 2 5'. The discharge can be increased by
having a wider roof block or by adding a lip downstream of thp. roof block. In such outlets
zamindars U5U'>Uy put in a wood an strip of metallic flap downstre-un of the roof block. Thev
just make u? their shortage although the dischar~e cannot be Increased beyond the theoretical.
Thp. chances of vertical starvation have been I:urecl by a fit in the lemniscate curve of the roof
block in the author's improvement of the A P.M.
(b) H)fi~on'al Starvation. It i~ either due to exesdv~ width of an approach resllting in
back now or due to rough sid~s The Zamindars usually cure the former by putting a Iloriz >ntrl.l
woorlen plate up3trea'11 of the roof !'hck and the latter by ruobing the sides approaching the
A.P.M. bl )ck, similarly HlP, value of Cv can be varied by tWI) methods:-
(i' Excessive length of approach. The 10SS'H in fr:ction are increased before the control
section resulting in draw down. It has been observed to be as much a'S three inches in some
cases,
Unsuitl,le apprr)acles The head hst in entry causes serious flrror in the discharge.
We have Sf) h· explained the hvdraulic, of the Eagineers and the cultivatosr but the
subord nates executing the construction of the outlets introduc~ sometimes changes to benefit
the cultivators fne Engineer in-chuge should he c:l.refull to watch them These are practised in
three ways resulting in 12 to 15% increas~ in di<;charg~. (i) Bac'{ tilt in the ro:>f block, resulting
in concavity of stream lines with consequent increase in r1ischar~e (ii) Groove downstream of
the ro')f block in bed causing partial aeration below the jet. (iii) Upward tilt in down<;tream
level floor causing a low pre,sure pf)cket. All these factors have been effectively cured in the
author's improvements of A P M It is enough to show that the hydraulics of the sublLerged
semi- modules (A.P. \1: ) is much morE' complicated than it is u<;ually supposed to be. It does not
mean that the control section of tl)ese types of outlets can only be affe;:ted but that other
types a t outlets K.G.O. Harvey Stoddard or Jamrao, Bend outlets are worse still in this respect.
17. crump's open flUme outlets.
The design of the open flume outlet is given in Fig. 10. The length of crest is 2H. It is
eX3.ctly the same design as originally invented by Crum.) but the author's approaches and the
sta ndaT d outfall design as accepted by Crump have been added.
Open flume outlet with i s proportional setting of 9jlOD takes its due share of silt, that
is, about 112%.
Minimum Modular Head is 10 to 15% of the depth on crest.
The coefficient of discharge is steady, 30 in the formula q =CBH3/2 for all values of H
and Width from 0'3 to 1'0 foot so long as the length of crest is 2H. for TI!in ow open flume.
18. Author's narrow open flume outlets.
(al Crum! did not investigate the behaviour of narrow "pen flume .outlets. In Crum p's
- design the length of "lpproach is very long ann the length of crest is 2H. In the case of narrow
open flume outlets, the coefficient drops very much due to very high frictional losses and
M. M. H. rises. The author carried out detailf'd investigations on outlets wi th widths from 0'3 ft.
to '05 ft. The experiments were dividfd into three secti~ns.
(i) Length of crest 2'H.
(ii) Length of crest 1·25H.
(iii) Length of crest '5 ft. in all cases. .
In the first two cases, the coefficient variation in the formula q_CBH3/2 as found out,
is giveR in Plate XVIII (A) and the Modular minimum Rea is, as obsorved, were found to satisfy

the relation h m =0'18 ( 3 ~9 . )2 R which can be obbPined from Plate XVIII (A).

In the third case it was sl1ggestf'd by Crump to reduce the friction losses further when
discussing the results of the first two cases and the ohs~rvations in this case were taken uader his
suppervision and results were published in paper No. 237 Punjab Engineering Congress Lahore.
The design of the outlets js given in F;g II and the results of the value of C and hm are given jn
Plate No. XVl[f (B 2) and XVIII (B l ) respectivdy. Th I ngth of crest being constant and small.
the control section shifts beyond the contracted section and under oertain conditions tho
 852

co-~ffi~ient of disch~rge, increaSfS eVf n wore tha,n, 3'09 The visl,lal examin?-tion of the jump does
not indicate modulanty 1U :-.11 case'> hut the provlSlon of the reqUlred h m as per Plate XV[ll "
essential to en~ure modular condition. Is
. (b: .Open flume outlets ?oth narrON and wi~e are ven' sensitive ~lltlets .becausq the
discharge vanes as H3/2. The dIscharge of these slmple outlets can easIly be lllcreased by
adopting undermentionerl devices fOIl ling the control se~tion
Ii) 'he discharge is increased by heading up water in the small channels by cattle and
men sitting accross the section rlown~tream of the outlet.
\iil Then have been r."sp.s where two men sat the whole night in water with a cloth
strp.tehed across near the tail rear.hes of distribut~ries. This deVIce passes water below while
heading up 3 / / to 4/} and cannot be detected.
(iii) The cultivators usually throw bricks and rubbish accross the channel downstream
of the outlet which is usuallv known as Daft )Bund),
(ivJ rh .. discharge of the open flume outlets can '>e increa'ied by inserting horizontal
lamina as shown in fiig l2.
The upper portion runs as a weir and the lower nne as orifi~"'. The sum total of these
rlischarges is more than the weir formnla
discharge The greater the nnmber of
lamina the larger the increa,;e. By lU-
sertmg even one plate, increase of lO
to 12% is Lkely,
i v J The discharge of the open flume
mit ;et can also be increased by simply
putting a pumpkin nr a pla.te in front
Fig. 12 of it as shown in Fig. 13. Ihe control section
shifts from middle of crest at G to C i in a widf~.r
section and more dischargd is passed even thougf-t ,t may act as a drowned weir [he increase
is { bout 20%,.
19. Doabie moiule outlet, (Orifice cum A P. M. or Open Flume3).
SECTIO~
The first known outlet of thiS
t\'PP was called· after the Scrachley'S
name. It used to be a masonry
orif.ce outlet discharging into a well
and then followed by an orifice in
the well. The modern practice {Paper
No. 146 P. E. C. Lahore, '1931 .by
K. R Sharma, is to conscruct a pIpe
outlet a.t bed lev~l of the parent channel
and then to take out an open flume or the fig. 13
A,P.M. frilm the well as shown in· Fig 14 ' "
These typ~s of ouUets are e,~onorriical and ensur~ a good ~ilt entry and ::easy adjusta ':lility where
Pipf'~Cllm F. °
,..
J;{ ;

Fig. 14 (a}

Pipe-Cum O. S. M. Fig. 14 (b)

 ,-
~
'-:
'"0
('!)

d
I'D
'fl
ciQ.
-
~
::I
0

Z
..,..,P>

.~
If v
~
0
()) "'Cl

..
(t)

<>
~
~
:It
('!)

...,::s
...
Z
;::

.....
...... I
J
;- ~i:
iOlt
:. ...
:3n
':)
p
.-+
n>
M-

;S4! ~t~
'tit
",'i
~
it
_...:fi..
"

..
'!o
~""~~----: .....
-Ui
tI o

' __ L ______________________________________
~

r
~"
-------------------------------------------------------,

....o , : -
o
0
p
.....
Ii'
r+

II
b:I
II
(Ji
~I..a

..
~i
'
I
 353
they have to be constructed across wide patrol-roa.d banks in the case of Branch at.d main canals.
The design presupposes that the available working head is enough and to spare for dJuble loss
first on the drowned pipe outlet and then in the module outlet. The loss of hearl in the pipe or
tn'lsonary may be kept 0'1 ft, in all cas~s and the rest of the available working head to be used
to set O. F. or O. S. M. (A. P. M.) as low as possible.
20. Rigid modules.
(A) Kent's module
Its design was based On a pressure float device as originally used by Italians in the
design of their modules. These outlets failed because the floats could be raised by inserting
sticks through the air-vents and the discharge could be increased.
(B) Gibb's module.
(i) This outlet was invented by A.3. Gibb, Executive-Engineer, Punjab Irrigation
~LSpublished in Paper No: 12-A Class A.P.W.D Irrigation Branch Publications. The module is
named after it" inventor and gives an almost constant discharge over a considerahle range of
water levels upstream and downstnam. It is only a rigid module without any
moving part. fhe. module consists of a chamber, semicircular m plan, called
pLAN

=-"=- -- -
-_;._ ==----
-:.....~--~-:_

Fig. 15
the eddy chamber, round which the water flows. Water enters this chamber through an inlet
pipe, which delivers weter to a 1800 rising pipe in which. a free vortex flow is developed. It then
enters the eddy chamber-vide Fig. 15.
 355

the range of a module of one Sl'rni·dccle. This was not borne out hy experiments. the range f
one cusee spiral module with 3 s"mi·cirdes wa.. only 1.42 tim..s that of "a single semicircl;'
module.
(g) Maximum pennissible tlownsheam depth above the module flooT was found to be
()'Sl G agamst O.5t D accnrliof! t,) Gjbh. When a • 'n 10 diverging flume was added to the spout
the maXilllllTl downstream water level increaSl'd to n·] [) This modification is. therefore desirabl'
th)\Yhen the Jellgth of exit flume or -;poul: i~ =2D.the Tange is maximum. Gibo's ide:
th at the length of spout shc.uld 00= B was incorrect.
(C)Other ri~id mooufI;Js
KIIANHlsRIC10 MDDUU OUTL£T Rt{. )\ llan'l3 Assistant EJ);;inpu
,.jPunjab Irrigation and ChafluT EXt'cu'iv~
::; EngiDec~, Punjab 1!J.iga1ion, have brDu~ht
':1.: out deslgns of RJgJd l\irnlules without
,c', moving parts . The Junjts of their working
l nl'ed to he deteJrninc? in actual practjc..
. ID bot h cases. fhe de'l]gns cue shown in Fig
"","'~'f"I'7..,."~,,,";;;:~~~f6...!!~ 6 and 17 Jespectively.
~.~ 21. Tail elus 16rs.
FIg ,b [.,;, Toe tail ma50nTY work on a channel is
caned tail duster. When the fhscharD'e of a
di ... tri',utary or minoT TH'lllf:e3 bPIl)w 5 cu·ecs. it is desirable for p.quitable disrribUl~)n to
construe: ;11 the 1a.i~ outlt:t, in the LHn of a cluster. The length of the water coone from the tail
dus.t~r s'wull! 10 noca"e be mnr" HY'n 2 mile-., ,be most sujtable oUIlet for use on the tail cluster
is an optm flum~ ..~il outlets at the t~ il aTt.~ open ilumes> with. CJests at the ~e ~evel.. Usually
liJt' dep,h on crc'~t IS kept one foot. I fie type de:>lgns of tile tall clusters are gIVen In FIg. IS (0
FIg :-cO.
Design of oemi-moduJe outrets ill ebanneis with varying eli c!larglb.
(il The supply avaihble in the Sutlej Valley canals is so short In t!"'e ~riti".al sowing and
maturing penods that it has been found necesisary to nm the,e channels WIth wilat is call.d
normal supply for c~ rtain important period,; in the agl'icu! tura! rotatioll. fhis normal sup::>lv is
55 per cent of the iull supply. The design of ontlets on these channels should, therdore. be sudl
that they take their proper share of the discharge both at full supply anti at normal slJPplv. III
other words,! t is necessary to rl~sign t he out lets on the e s y"tems as proportional nudets:
(il) fhc following are tl.e three types of proportional outlets which are suitable for us,
on these canab;-
tal An open flume outlet having its crest at 0,90 and having an available working head
of nct le,;s than OAD wm draw a prop.>rtional disc:harse at both fun and norrna l. supply conditlons
in a channel.
~b) An O.~ VI. with H,==t143D and an avaih.ble working head of not less than 0.4D win
draw a pruportional diSCharge at both full supply ant:: normal supply conditions in a Channl I.
provided G does not excet'd O.69D.
tC) All. 0" M. wit h G = 0.7,.,.D and Y =H~ will draw its due share of discharge al full
and normal supply levels. in the latttf case it works as an open flume. fhe workmg head
reqUIred in thIS ca'le is also O.4D.
,iii) It will be seen that a working head of 0'4D i'i requiJ'Pd for the design of outlets
with the reQllired conditions. An ex~min<ltil)n of the aVdllal>le working heaf\s of outlet~ On th~
Sutlej Valh'Y Canals shows that whereas connitinns vary consirltrablv, a substantial number of
outlets does Dot pO.~SeS;' a working hean of 0.40. It will thus be recognized th.u w]ltreas it is
possible to design a large numb"r of outlets such that they would work proporliona}]y under both
lnll aul n)flUl.l s'lp;:Jly C )llditiollS ttl~ nu.noor 1)f those whici\ cannot he so designed IS OClsidpra:.>!e.
{iv) For pnrpnses of design, therefore, the outlets on ti~ a)[} -llece:mia distril>utalY
Oil the Sur lej V"l.lly canals may be classIfied as follows:-
(a) Outlets which can be designed to work pmp0nionaiJy under both full aud nonnal
supply conditions·-Such outlets must have an available working head of not less than O.4D
,b, Outlets which cannot be designed as class A but have a sufficient werking head to
IJt" modular at full supply conditions, Sue!:! olltJets are those Vlhicll bav~ a working head of 02D
to O.4D.
(e) Outlets other than tnose included in class A and B. Such outlets have a very poor
com mand i.e. less than O'2D
 354
The exit at the downstream end of the m"<iule rJisdl'l'g''S tbroueh a SPO'lt OT flume in~()
the outlet channel. In a free vortex now veloctiy x ra<iius=com:tant. In this condition,
the w_ater at the out.. [ circmnferellce of th<! stream i:!ows at :t comp~r3.tively high level d,ne to
centnfugal force and the surface slopes dovvn tow:\rds the inner cir(';!lmfer('nce. A l"enes o[
b~mes is rixed in the eddy chamber with their lown edges sloping a the required heigh.t ahove
s~!I ·)f the module. If the head causing Row is In excess. the water bailk~ up at the outer
crrcur_nference of the eddy chamher and impinges against the baffles. imparting an' urw~r~
Totac!Onaldin'ction of flow to the water <l,'hich spin'" rnund in the Comp::J.Ttrn.·nt f--etwe~n tw....
c:uccpssive baftl~s and rinaUy drops on tIll' oncoming str~am of wa.ter. thus <iissipa.ting excpss
enf'Tgy. nIP. action of the haffi~ is not unif"rm and. except at maximum dissipation, some baffles
are out of action. fn g~nerat one baffle is far more effective than the others. .
(ill Discharge form'la.
,~ibh'" formda fot:' iis<:huge was
o _ 1·5 r
m:!-i I
~=roV2g (dl+ho~ -{ ----l<sg"m+--Iogem------
m:!-t
?-
1
l
rn 3
m ~m:i J
where li=Radi11S of inner s~ml-drde of the (~ddy cha.miw:r; rl)=Radiuscf the Ol1ter seml,cirde
d,=Depth of water at inner cinnnfO'!renre.
dn=!}poth of water at outer circumference: h ==Head ..au<:ing flow, v~rying from 110 to hI :
h,,+ 0= Total diff(~rence of level measu.red fmm the minimum wati"f level in the parent channel
0(' dist-ributarv to til" lion, of the eddy chamber (i.e indu.ting hea.d lost in inlet pi.pe=h" )

m=rofry=Ratio of outer radius to inner radiu,,; Q=Dlscharge.

B=Width of eddy chamber=ro-rt. ' . .
_ The [<}rmula is based on free vortex fiow. in which the velocity at any pOint van~s
mver<>elyas the ~adius and by BemouHli's theorem that the total energv nf'a.d of all filaments IS
con!'tant : - - ,.
t fatal energy 11ead due to velocity pbs static head measured fmm the same datum.)
Gibo's formula held only f(lT his standard design in which
. h
m=rofrt =2, and D =1/7 ._cA.

(iii) Poona ExpeIiments of Gibb's Modules.

The resnlts of the pooua Experimet1t..~ on Gibo's modl~le have be(>n Published in Reoearch
Publication No.3 by CC. Tnglis and Rao Sahib n. Y. Jo!!lekar of CeniT"l Irrigation and
Hydroclynaruic Rc!<earch Station Po lOa. Bombav, 1940. The conclusions :lrrived therein ::Ife
summarised :- -
(a/ Of the 6. bamt>~ in the eddy charnbH of the standard 3 cusecs modulE" in wl1idl
rad~us of outer ~c:ir('~umference of eod_y_."_h~_I11~':leT: =2-0
radIUS of inner drcumferenc· d eddy ctu.mlJer

Md the r~tio of llll-:_~~~~~sing 'I7«locj~_at_ °llt.er _ c~rcnl11fere~c~ = 1/7. the first

D=depth of water at outer ctrcumference+h o .
4 baffles weremo'-" effective in killing the hea.d viz. 0'12 ft. in 3 cusecs module than b~ffies 5
~nd 6 which increased tt",· range by only. 0 18 ft.
0

(bi When a hell-mouth was added to the inlet pipe. conisdeTably less he ld was
requirftd 1;l'~ a~tain modula.rity and the range was increased.
(c) lhe range was a minimum with m=r,,!r1 =2 as adopted by Gibb;
but he wa"· wrong ill assnming that the r31'lge of his moflule could be
inCTl'ased hv increasing the ratio BID; and the values of BfD within which a high Tange is obtained
are 0"8 to .'5.
(d) [ncreasing the number of baffles in semicircular mocble had a negligibJe effect on
range. though fluctuation.. of di~charge within the ran~e. werf'o sl!~htly n:ducedo
(e) [he range was not increased by"increasing the I.. ngh of the arc of curvature through
which water flows before entering the eddy ch:lmber.
(fl The Tange of a module can be increased by incn<lsing the number of spirals of the
eddy cham~r. Gibb held that the range of a m()dl11(' wlih 3 H'.mic;Tcles would be three times
 356

After the outlets of a distributary have been classified as above and before proceed'
with the design of outlets, an attempt should be made as explained in paragraphs 10-25 to lDg
if any of the outlets clas£ed D cannot be converted into cla~s A or those classed C to class B orse;
For successful dj~t~ibution it is essential that the number of outlets of classes Band C shOUld m;
reduced to the mInImum.
(v) The type of outlets that should be adopted for each class is as follows:-
1. Class [AJ. Fer this class of outlets which can be designed to draw a correct discharg
at both normal and full suppl V conditions, the following type should be adopted:- e
(a) An ollen flume outlet with crest set at 0.9D provided B t does not work out to less
than 0.2'.
(b) If in [aJ above B works out to less than 0.2' an O.S.M. Set at O.75D may be designed
such that the value of Hs ranges from 0.375 to 0.43D.
(c) If an O.S.M. outlet as above cannot be designed for a particular discharge, then
O.S.M. set at 0'69D may be designed satisfying the conditiom Hs =0'43D
I[ Class [BJ. For this class of outlet, it is not possible at present to arrange for
proportional working at both full and normal supply conditions. The best type of outlet for
this class would be either :- "
(a) An open flume with H equal to five times the working head availabe or.
(b) An O.S.M. outlet ~et at 0'75D with Hs from 0'375D to 0.4SD. whichever of the
two would give a lower setting. so as to enable the outlet to draw some discharge at low
supplies.
Ic) In case the width Bt of the open flume outlet works out to less than 0.2' or it is
not possible to design on O.S.M. accordnig to [b] above for the parhcular discharge then the
outlet should be designed as an O.S.M. with crest at O.69D, provided it could work modularly
under full supply conditions.
Ill. Class (C) For this class of outlet which'has a working hea.d of less than 0.2 D it
is not possible to design an outlet of the semi-module type. The only outlet possible is of the pipe
or orifice type and the best type is Scratchley outlet.
The working head Hw to be adopted for purpose of design of the outlet should be the
average of working heads observed during the time of keen de.nand (1st to 15th September) for a
period of 10 days. A fair proportion of these working heads should be personally checked by the
sub-Divisional Officer (hy surprise visits if possible).
(vi) when water level in a distrbutary rises above the designed full supply level on
account of chan~es in regime, the open flume outlet draws a high percentage of excess. To
guard against this, all open flume outlets, whether of class A or B. should be fitted with
roof blocks.
23. Water course discharge observations,
The discharge of water courses downstream of the outlets are observed by any of
the following rr,ethods:-
(i) Float obstrvations
(ii) Trapezoid.1 notch cip ,H.-tte Weir.
It
dErAIl OF
1--. -L. ..
'''ES.,
L+f -- --t
(iii) Wooden flumes or Portable Detachable
Tin flumes.
(i v) Calculations from the discharge formula

r
,
L ·1 of the outlet
In the first case, the surface wirlth is measured
say w feet and depth D in the mipdle is taken. The

l'
o j
1 .5",_1' Ell .
I .
surface velocity is takfD by running a float or a cow-
dung piece for 20 feet length in a straight reach a~d
! 'l: liiliu-s " /:'-0 elf'• .,. 7i']i- by taking the time by a stop-watch. Tile velocltYd
i , ED D WA"1.1f etU/lf'SI. per second is then worked out say v ft. per sec?n
-t. CU""'''GIDG£ The discharge= 06 wDv cusecs. This IS a faIrly
approximate method.
Fig,21 In the second case a trapezoidal sharp·crested
notch is used as shown in Fig 21. It is called
cipollette weir. Its us~ ~s desc~ibe? by A.S. Gibb, in pap~r No: 14,. Punjab P.W.~.
Irrig:ttion Branch, PublICatIOn. It IS fixed across a water course WIthout headmg up w:ater 10
it or into a field by diverting the supply from the water course. The gauge readtng IS ta~en
and the discharge can be got from tables published in the said publication. This method glves
 ,... ; ... -t-~ ~., ....
-- ----t---:-~--

,..._.
.:" ,I I
_.~,'
/'.

. .,.. /
')
~

I
I

..: I
.(/
( ....
... I

"

"
T
.. ..." ~. ):

~
.. 0-
&I)

~
p •
~
::z:.
,4
~ ,.
II)
eO

J:
~
"
IcO

.
<to
d:
. ,... .,i,
f

rt .j(
Z tr ~ 4,
c:a J-. ~ , '.
&tI % '",
~ ,
c
...-I· d.

ta.d -!
,~

"
l
i'jI
±, •
II
.ft
'f 10
 foZ .~[.I
$J>'''.'' ~
t·, fo" I" ,.,
NJ
,.,
8~.1J YH:1SIQ
,.., .., ,.. !I'" ... I ••
• ..
If-}
4,7" i'
"
-1.

. ,_'-:<.:. ,..
- ~-.' :.' "!.. .....:
'"
I

..
aumu-uu dLqlqlOd S,l3CU1EqS JOJ qc.lel~ a~l't!tlOS!a

~~.
"""....~...... -.i"i. ...~._ ':"'-1' " ';>...

."f
-
-'1 0

-
~
(fJ
~
~
....:l
10-
~
0
a:::
.."
.-
-
:;:
>#

t
.".:.
0
C'I

~
b.O

~ .L
0
IX<
~
0
IX<

'.
Y:
...III
a- ._
 357
accurate results only when free fall conditions for the weir are available without headin
water in the water course. The working head is required equal to the depth on crest Whi~h ui~
not generally available.
24. The Author's portable tin Hume fOT water course discharge.
Experiments were carri;d out by the author to design suitable
flumes for this purpose and the results were published in the Punjab
Engineer, Lahore. February. 1940. A brief summary is given below for ready reference;
(a) An accurate determination of watercourse discharges is a very essential factor for
successful irri~ation and equitable distribution. Water is supplied to the cultivators from an
outlet which is supposed to discharge the permissible discharge of the water course. The
design of the outlet is based on certain coefficients which are liable to vary with different
conditions of the design of the outlet and its imperfect construction. The public opinion is
getting enlightened dav by day and the cultivators are now keen to get theu permissible
discharge correct to the second !Jlace of decimal.
(bl A cultivator is not satistied to know that the Irrigation figures are more thad the
permissi bIe,
(c) The discharge observations of water courses by means of floats generally result in great
dissatisfaction among thfl cultivators. [he overseers and zilladars always report different discharg,~s
taken by means of tloats, rhe discharge observations of a water course are, thus, very unreliable.
The Use of the cippollette weir, as described in Irrigation Branch Paper No; 14, t::lass A by
AS. Glbb, gives fairly accurate results. fhe free fall conditions for a cippollette weir are not
generally available and the observations in the drowned conditions by an average subordinate
generally give wrong results. It will, therefore. be a great improvement to observe the discharges
by means of an open flume under modular, conditions, when the hydraulic jump is formed
downstream of it. In practice, it \Vas observed that it needed special efforts to make the
cippollette weir water-tight because water falling down the crest generallv scours the downstream
bed and the weir is creeped out. Moreover, the velocity of approach under different conditions
of the upstream water course section. affects the discharge considerably.
(d) The idea of disci arge observations of water course by means of a portable flume,
was introduced by Mr. Abdul Gafoor P.S.E (Sub Divisional Officer. TandalianwaJa) in his
note dated 2-8-1926 circulated by the Chief Engineer. A coefficient of discharge 3'0 was
adopted for the wooden flume for all readings which is not correct. A gauge of the flume was
provided upstream in the return waH which is liable to be affected by the I )SS of head in entry
and velocity of approach under different conditions of the water course upstream. Moreover,
the design of the dume was not suitabl!'! as it required as much as 25 per cent of the head fat
free fall conditions. The experiments were taken in hand by the writer in summer 1931 to
improve the design of the flume and to ca.librate it for different heads and under different
drowning ratios by means of actual discharge observations. The discharges were observed by
melns of tank measurements.
(e) The drawing for the type design of two sizes of portah le flume is given in Fig. 22.
The flume No. 1 of one foot width is capable of measuring discharge upto 1'00 cusecs. The
flume No. 2 of 1 5 feet width is capahle of measuring discharge upto 3'00 cusecs. The flumes
were made of galvanisea iron sheet G.W.No. 22. They are quite light and can easily be carried
by a beidar or on the carriflr of a motor car. One beidar can fix them in position in ten minutes.
Each of the fluRles is in two parts. One part comprises the glacis and the other consists of the
upstream approaches and the crests. One part can be folded on to the other.
(f) The flumes are made according to the standard design of a meter flume described
by the writer in his pader No. 154, Punjab Engimering Congress, 1932 and sketched in Plate
No.2 of the said paper. They have a uniform width throughout their length. The length of the
crest is 2H. The upstream approach curve in bed is laid with radius 2H. The gauge is providpd
on one side in feet reading upto '01 foot and on the other side in cusecs at a distance of 3H
from the beginniug 01 the crest. The flume is straight up to a length of 2H upstream of the
gauge and is curved out with a radius 2H from the upstream return wall. The floor is kept
level opposite the gauges. The dowDstream part consists of glacis 1 in 5. This design ensures
a standing wave with a working head of 10 per cent of the depth on crest. The proper location of
the gauges and the silt·free level flow opposite thl'm removes the error generally introduced
by the velocity of approach, which would otherwise vary with the shape of the waC':L
course upstream.
(g) The graph for the discharges of both the flumes is given in Fig. 23 for different
heads in modular conditions, so long as the hydraulic jump is formed on the glacis. It will not
 358
Design of Portable Flu:nc>

(It 1~j~' :'1('"

;;iJJilJ:) j., It
S£CTION
~
~'.IRtlH ,.c.,
• .. , -to •.,
$/. !Sfi( ,,,..'. al~N if...,
_.. _-- ---- ----, --- - ------- -,,-~ -- - --- ---- ---,--
e..{I7i(11¥ w.

-------
;If

I
I
!,
:~
~
vAUGE
U
n ~.
\r:~
:1::
. - ::-' .' :•
." t J~oh
:Ii t,
If
- -t--f"s-
t
-, CIl!ST;;t·a td.
I'"
ii-:.:1t4(;(ln-(f
• '''-1! /
Y.PJ
•
t
I!:
l~ G-Auf;$
:1 ~
- ,_ \,._, "
I II
i!!~===~~==~~~:!!'~V
~ !JeOM .tr. c: 3i4'" "
C;.' SHUTII

,R:'+~
. ;!.UlON ~
,".,

",,1 ,rr 'nbf'l'

generally he neces ary to lun these (tumes in rlruwnec conditi,ms, bec::tuse in no case the requi:'e,f
wNking Iwad is mf)re than 0·1 foot for modular condition; of the flume, Such a small ht)ad is
generally available by pntting a flume in a hranch water course or in an open field. !11e
discharge can De read if(,m the graph ')1' a cu~ecgauge may be provid~i in the flumt: itseH or by
n~ing table NO,3, Tll!! main and thfl most impl)rtant precaution in obsPTvation d the dischar'gt'
by means of a portable flume is th'lt its crest should be absolutely levd. This can ea"ily be arrang,'d
hy means of a mason's spirit lev... t The spirit level may even be rigidly fixed with tne side walls
and on the rigid cross bar across the ~re,t. fhls shoulrl 1ll practice be te~ted invariably by
letting in a little water befoTE' the water course is cut on to the meter flume so that the
depth on crest is 0'005 foot as read from the gauge. This should clearly show unevenUfSS in
the cr(:~t.
. The upstream ends should be well puddled aJJd the water course bflrm and bed should
be s,haped upstream of the flume so that the entry is smooth There should be no leakagt'
oolow the bed or on the sides. This can be very easily arranged on account of the ed f e
provided upstream and the curved return walls, Care should be taken tha.t the tb')T opposite
gauge is clear of any deposit at the time of observation,
TABl,E No, 3
Sharm's Table for Portable Flume.
Discbarge Oischarge
Gauge Gauge
Flume width 1'0' Flume width 1'5' Flume width 1'0' fiumfl widthl'S'
----0-16----- -'-o,~z -- - -,- o~io..-------=~~·-~O'~r_-------14ir-----

fIlS 0-25-· tPI ,.iCf'~-;::'ttS$:':>, ,t c ' ()'46 1'06 1':7

{}ZO O"lS ";'.i I\"~" 048' :';I'la,':, j" }'67
0'22'
024
{} 26
!'} 28;
tI 80
03il.;'; :.;;;.,;;"; ,; .....~;.t"
087
042
u'47
0 ti3
"rn
--'bnfll
IS)
(Yi

.
E.

.
0_
,1':&56':' :·';,052

m.s$O.lil (')
l)·SO .
054
056
0'58
i\' 1;:·(OJ50; "j"2O '': IJ1S~I'i"ll]
1'28·I~nWjiil'l·'
1-36
I 44
1 53
: 1'78
1'90
:.!·03
216
2'29
f} 32 O'5~ 0-89' 060 1>32 242
t)'34 0'13S- 0'98 U 62 :l'S6
E)' 36 o'n 1'07 0'64 2'70
0'38 0'78 1"17 0'66 2'84
0'40' 085 1·27 0-68 2'98
042 0.g2 1'37 0'70 3'12
•
 ~hl When th~ discharge is to be ()b~erved for ~n out>t whic 1, i.~ work:ng lr,'e [,jll cr
·,·jth a hydrauli~ jump, it noes nnt matter If water levd i~ sowlI.vhat he::tded np or depr -s,~d hy
in<;ertion of the f}nme. The gauge of the m")ter wiilllecome steariy after a few minute.i ill such
a c~"e. Bllt ",h<':n the discharge of an outlet is to be ob3erved when it is not working "dularly.
it i" imDp.rativ> th'1.~ the ~....ater level In water course should not be changed lJy insprtiol1 of tlw
flume in water-course or on the sides. rhe discharge of the watercourse is generally approvimatpl.y
known from the conriitions of the wnrking of the Olltlet. fhe flume should be fiXld so th'lt the
gange H',iding on the CllS c gilllge is in level with the wat'"r lwvel in the watercOllrsp. -\ pIg shoukl
he fixp(l in the watercourse with its top Bush with the water level in watercourse befor(; \ .. , er is
onen .. o into the flume. In case water ("vel in the watercourse overtops the ]wg or is dc, ressed
tleio".v thf top of the p"g, the obsp.rv:ltion should he repp.ated by adjusting the crest a CCI [ding
to appf'oximately )bserved gauge.
. rhe flume should he fixed neu the outlets as much as possible so that the loss in ()bser-
vatlOn on the wat~reour~e <iDes not ;.l.!i"ct the observation. A distance of about 100 fee,
downstream of the outlet "'ould be su cipnt.
(i) The wri t~r thin ks t i l t th ~ introductior, of this accurate method of ob5ervation of di:-
charges will lead to great sd.ti"fa " ion a nang the Zamindars. Every suhoHlinate shouin be provii.kfl
"".jth these portahle flumes an,! t dis 1 1 ge of the watercours~ should alway., be observed j.> nt 1 •
hy the liHadar ann the Overse~r with thesfl flumes on compldints of shortage from t he cult; vat or, .
ft is very difficult 10 satisfv the Zamindars afterwards when a wrong dIscharge is once inhmated
to them by a Zilladar who generaIiy bases it on inaccurate float observations.

25. 01ltl"t. Chaks.

The 'lrea irrigatpd from one individual outlet is called cllak of the
outlet. The discharge of the outlet is u aTken from the culturable commanded area in the s~j(l
chak based en the permissihlp jnt~nsit y of tte irrigatir·ri and the water allow~nc'C. The maximum
(-li,,~harge of an outlet is 7:75 cusecs ann thp minimum 1'0 cusec. The length of watncot1rse irrig'lt*
ing a chak should not be more than 2'0 miles. The procedure to design f he watercourse and to fix
the chaks and the method to work out the area, commanrl ann draw-off statement~ is explainE.d
in ChaplH xvn of this part 0'1 irrigation projects. I he A. P. M. outlet set helow 8i 10 D or even
upto 12/1Of) or rigid modu les He suitablt' t:ipes of outlpts in the head reach of a channel so tha~ the
increase of sup)ly at head is conveyed to the tail re::lches where it is most wanted. The proportion" 1
outlets slIch as O. F @)'jIOth, and A. P.M. @6/10th, ;ue suitable for the middle reach.
The open flumps are suitahle for the tail reaches of channels an(l prefelable in the form
",f distributJrs and dusters. They are sensitive and can absorb the exce~s wt ich usually floods the
tails ;n rains Or when the upper outlets arc closed by the cultivators for silt clearances of the
watercourses.
Examination Que,tions.
I. \Vhat do you kn')w about Kennedy's gauge outlet and explain the use of it. (T.C E. 1933)
2 Give the formu la us.d for calculating the discharge of three types of outlets used in the 1rrigation
Branch. vVhat is the normal seUIl>.£( fo~ each type?
3. Give the formula for mea-snring the discharge of water-course with fair accuracy, exp!;.ining the ,arne
and any co-efficient used. How would you mea~ure the discharge with great accuracy? (P I.B. 1141)
4. How would you calculate the minimum water level required at. an outlet from a contour plan of the
outlet chak?
\Vhat is the minimum size of outlt't chak you wruld normally permit on a canal in which the water
allowance is 3'25 c"secs per 1000 acres? what are the objections to outlets of smaller s;ze? (P LB. 1940)
5. V\'hat difficulties. if anv, would ycu expect to find in the irr;gation of ... clzak which has an cutlet
of 3'5 cusecs? ~ (P LB. 1940)
6. Make a dimensioned sketch of a clistributary tail cluster of two outlets, one 'Iront' and the other on
the left s;de (at right angles), the paT1iculars for which are as fnllows:-
Full suppJy discharge front 20 cusecs,
Left I'S cusecs
Working hear! frant '5 ft. left ·75 ft, (P.I B. 1939 )
7 The enclosed blue print plan shows the area which is proposeD 10 be ;Tngated from a channel
m~rked 0:": it. Propose in it a suitable chak (without marking l1alws of watercomses, and design suitable outlets
tor these. Each rectangle may be taken to contain 25 acre". Permissible annual intensity is 75 pc. Khanj Rabi
ratio 1: l a.nd full supply factor at outlet head is 88 cusecs of I{/zal'ij permis'ible. !P.I.B. 1939)
8. Give the standard formula for calculating the discharge af e"ch of the following outlets.
(a) A.P.M. outlet, (h) O.F outlet, (e) Pipe outlet and state the relative advantages and disadvantage
of each of the~e types of outlets? (P I B 1938)
9. On the enc'osed blue print, mark the alignment of thp principal water couree from tne two tail
outlets allowing one Naka to each square. (P.LB. 1938)
10. A certain minor distributary would he improved if the water level wen', lowered as much as
practicable. A tail cluster for this minor has to be constructed for the firEt time. The R. L.'~. of wat .. r surface
 360
n the water course at the site of the cluster are 512.40, 512.0 and 512.1. These levels are sufficient for command.
The reduced level of the water surface in tbe minOT as nvw measured is 513.2. The discharge of the water courses
are to be 1.3, 1.4 and 1.7 cusecs. The positions of the approaching water-courses are shown. Design the tail of
cluster. (P.LB. 1938)
11. On the Blue Print supplied mark the alignment of the principal water-courses from the outlets.
allowing one naka to each square. (P I.B. 1938)
12. Describe the two m@thods of measuring the discharge in a water-course. POlDt out what precautions
are necessary and how the formulal are derived. (P-I. B. 1937)
13. Design the water-courses for 2L miaor on the enclosed blue print. Describe briefly the essential
working principle of an A.P.M. How do you tell on inspection when an A.P.M. or an O.F. outlet is working
properly?
What are the respective advantages of an A.P.M. and an O.F. and in what situations would yon prefer
one to the other? (P.LB. 1937)
14. Give a dimensioned sketch of a two-way tail cluster to distribute 5 cusecs supply at the tail of a
minor. The front watpr-course is aligned on the same centre line as the minor. It is to be given 3 cusecs with a
working head of 5 ft. The outlet on the right takeli vff at right angles and is to be given 2 cusecs with a working
head of 1.5 ft. (P.I.B. lS36)
15. What governs the size of an uUet chak? What IS th .. optimum size for it? (P.I B. 1941)
16. Whatis the difference between a semimodule and a rigid outlet? WhaJ; is the main advantage and
disa.dvantage of these types of outlets as compared with those of iron pipe outlets? (P I.B. 1936)
17. What is the maximum length suitable for a water-cour~e? (P,I.B. 1935)
18. Desctibe the hydraulic features of the following types of outlets and state under what conditions
you would use each type.
(1) A.P.M. and (2) O.F.
19. What are the hydraulic shortcomings in the working of:\ Crump's Adjustable Proportional Modul( ?
How can they be remedied to make it work as a perfe"t submerged semimodule? (P. U. 1943)
 PART II
CANAL IRRIGATION
CHAPTER XVI
Irrigation Projects
Introduction.
Irrigation Projects usually fall in various categories such as :-
(i) Gravity system canals i rriglting the traci. hy flow. The rivers usually have a
relatively steeper slope than the adjoining country due to their tortuou~ length. If the main
canal bE' taken out from a suitable site in a river, it is designed with a relatively flattened slope.
The main ca.nal at its tail with full supply abovJ the natural grouni surface is splct into distribut-
in<: branch canals depending up,m the config ur ttion of lhe ground.
Ii) Punlping system. In this ca,e the supply is pumped flOm the sub-soil reservoir or
from a natural stream into the distributing channels
(iii Drainage Scheme The drains rna" tJe surfac a, seepage-cumsurface orains seepage
drains ano underground covered drains to remove! storm water or lower the water table.
(iv) ... tOT;lgll reservoir schem s as de3cribed in P;trt II [ of thi~ book.
lVj Reclan,ation Schemes. They are nece_isary to reclaim the thur and long~d-
lands.
It is not c0nsider~d neces~ary to descri 'e all such schemes ;lnd projects in this book. In
what foU()ws, the gravity system project is d'scribe at length. The principal req1lirements are
generally the same; although the detail and design differ a lot ill other cases.
2. Preliminary Investig2iions.
(a) Reconnai~sance Survey:- Before a p:oject is sel 'cted f)r detalled investigation,
it is essential that the whole of the country in its n~ighbourhond shoulrl be pxamined by means
of a reconnaissance. The principal obj .. cts of this prelimina y surve.\' are to b:} oota ned.
at comparatively srr:all expense of time and mor'ey, gpneral information a.s to the
nature of the tract examined. the facilities offered by it for irrigation, ::lnd the relative merits
of all projects practicable in it. Such a reconnaissance en lDles a general plan to be drawn un sa
as to utilize those facilities in t!1e best and the most comp ehensivd manner possible s() that each
individual project prnposed will work in, and will not clash with other schemes feas~ble .
To enable a proper comparison to be made of the relative advant"ges of competing
schemfs, approximate geuel-al surveys of the 'Porks and irr:gable areas should be undertaken, and
rough plans and esjmates of th~ pfl)p),ed work.; Should be rna Ie. fhe de3 i gns for these shoukl
not be considered as final ones, nor the estimatf's for them as txact. bu t sutiicient care should be
taken in th"'ir prepartion to obviate extensive a:terations thereafter, as those may greatly lesspn
the value of the prelimin,lry work. C:lre ehollld also be t:lken to prepare the designs and
estimates of competing pmjec s a'i far a; . pos'iible on the same general Lnes, so that a fair
comparison as to tl1eir costs all(i advantages may be obtained. .
(b) Hydrographic Survey. As soon as it has been a~certained that a favour a scheme
is practicable III the area und r reconnaissnce, all hynrograph c investigations lltcessary should
Le started, ano thty should be continued for as long as possible.
Rainfall data must be studied. High mean annual rainfall does not pre:lude the
necessity of canal irrigation of the rainfall is not available at the sowing time of the principal
crops.
(c) Soil· Survpy. Soil survey should be taken to know the alkalinity and aciditv of
the soil and its ~uitability for the irrigated crops. Agricultural Depa.rtment shou~d be consulted
about the crops which would be suitable for the tract.
(d Com'l1unicati()n c • The Roads and Riilways departments should 1;e cor,sulted to
extend mans of lommunications, to cheapen the cost of construction and to develop the
rapid:ty.
(e) Geological survey. The Geological Department shonld advise on the general
geological conditions. and in licate what special precautions are necessuy to ensure the safdty and
success of the works.
 362
(f) Drainage. me draiage of lands should be invt'stigated and jJrovided for in the
schemes.
3. Detailed Surveys.
After the f~asjbilty of thf' project has ;hen established with (lue consideraton of the
preliminary investigations, detailed survey;; are requ;r~d toinve~tigate the f lllowing aspect!; :-
d) The !rcation of Headworks. The dptaiJeo riVI'T ~urveys are required. The usual
plane table or Techometer survey methods II.re not applicable in river beds of large riv!lTs
special methods not usually dec;criced in survey book~ are required: The student is advised to
study the book wr tten 1y Rai Rahadur A. N. I( hosla •. LeveIiing of Precision Across Rivers
19:4 (Disc. Method)" which deals with thi .. suhjectiQ detail
di) Alignment of Canals and Distibutaries. Cant' ur surveys usually to a. scale of 2
inches to a mile are requind. These contours may be at 5 ft. intervals. They are generally
available from the Survey of India Department, and if not, they h::lve to be preparen.
(iii) Al"gnment of water-courses and Irrigation chaks. The contoured. survey plans
showing contours at one foot intervals on a scale of M inches to a mile are required. The areas
are divided into rectangles a nd Squares and levels every 500 ft. distance observed along long and
cross lines. These i ~ not genprally availabe with the SlJrvey of· India Departm3nt and have
to be prepared by a special staff of the Irrig::ltion Dep:trtment.
4. Alignments.
~. ',Having determined the source of supply and its relation to irrigate lands, the third
questio.a\i$tile alignment of the canal. This should be so made that the canal should reach the
highest part of the irrigable lands with the least length of line and a minimum expAnse in
construction. the line of the canal should follow the highest line of the irrigable lands,
preferably, skirting the surrou T1 ding foot hills. and passing down the summit of the watershed
4i:viding the various streams.
.: To get at the best alignment. preliminary surveys are necessary. On a large scale
contour plan the C. L. (Centre line) of canal, may be marktd. The flnallocaiion may be made
on the ground with the aid of a few trials.
()bstAcles to Alignment.
(,;,,'.' (b) The best meJhod qf avoidiDg the cross drainage wodes should be considered.
Estimates of ,cost of materials required for consbuction and subsequent maintenance should be
made. . 'c '.,. " !
):, .'
I.
Curvature •
. ';;~l ;••. (c) ~ ,A dire(ft or straight course is the most economical, as it
gives the .gre~test f~edom
of flow and CilU$CS tke least erosion of banks. The inst::rtion of sharp bends mevItably results
in the destruction of the canal banks.

.!
MINIMUM CURVES IN IRRIGATION CHANNtLS.'
[.;.
'-.L. ,The minimum radii for curves are given below ;---
Capacity 01 channel Minimum Radius of the curve
..-.:
Over 3,000 cusecs 5000 it.
3000 to 1000 ., 3000 .,
1000 to SOO •.. 2000
500 to 100 1000
106 to 10. " 500 "
I_ess than 10 " 300 "
As far aspOS~ible.longer radii should be given.
('or
Bench Marks.
• ...., [d] Permanent Bench Marks should be left at regula~ intervals •. These should gi~e
,inscribed redu,ced levels for cross reftrence purposes.
5. Slope and cross Section. ~ ..
According to Lacey's silt theory, there;j~Lonly one section of· a channel and only one
slope at whidlihe' canal carryhlg a given dis~harge will carry a particular grade of silt (silt
factor). The wajer in. the, main ,canal carties. 9ilt with a high silt factor, . This is gradually
reduced in Branch Canals and Minors. The channels section slope i1\ hence fixed by silt analy'i:!is.
Any excess: in the slope of the country.. has to be destroyed
.:.;...,' , 4:~ .
by'suitable
r
dT(;)ps (Falls).
. .
•.
 " The cross section of a canal may be so designed that the channel may ,be' wholly m
excavation, wholly in embankment or partly in excavation and partly in embankment. The typical
ross. sections are given in chapter VII of part II. .~~.: ~
,
~~" :,--:

Fall and Drainage Works-L. Section. ,/ f . . ,

. As the natutal fall of the country through which a canal runs is usually greater tMn
the slope of the canal, falls are provided. The location of these is usually fixed at thel site whete
the canal come3 too high above the surface of the ground, while their distance apart is So
arranged that they shall not have an excessive height or fall. If a canal can be so located and
aligned that it will skirt the. slope of the country on a grade contour it becomes possibie to give
it tile most desirable slope throughout its length without the introduction of falls, but where it
runs down the slop~ of the country, compensation must be made for the difference between the
excp.ssivo ground slope over that of the cand If the alignment of the caRal cuts the natural
drainage lines, either the drainage should be diverted or cro.;sings provided. In no case are the
drainage lines to be obstructed.
6. Location ot H€adworks.
The aim of the canal Engineer is to supply water to the irrigable lands by rHrect flow
from the distributing channels. Hence the headworks have to be so located that this condition
is brought about. It may either result in tapping water upstream or rai;;ing the pond level of
the river by aptificial obstructions. A headwork comprises the following :-
(1) Diversion weir or barrage; (2) Undersluices ; (3) Head regulator: (4) Silt Excluder;
(5) Silt Ejector, (6) Escape.
fhe desL;n of the fint three works has already been describ~d in Chapter IV of Part
11 and items Nos, "and 5 have beeD ex;>lained in Chapter X[ of Part II.
Canal Escape.
Escapell are int.,ded to escape surplus water from a canal into the nearest river or some
other natural drainages such as river ereeks. They serve as safety "alves for the canals. Their
function is three-f.,Id:-
(il The excesses upto ten percent of the authorized discharge can generally be utilized
without endangering the canal or distributary banks. The abnormal excesses which sometime;
enter a canal due to sudden changes in riv<!r supply.leTels ill freshets have t) be escaped at some
suitable site below the Hearl Regulatar of the Canal.
(ii) Escepes are also necessary at tail of Branch canals and up3tream of high em':)!I.nk-
ments in a distributary to escape the excess~s in such channels due to sudden rainfan.
(iii) Escapes are very desirable at convenient phces to permit reduction of disch~rge
in the distributaries and Branch Cana.ls to closs normal breaches and intentional cuts by the
cultiTdors for llnauthonzed irrigation.
The eft~ct of reduction from the canal head !:I.kes usually a long tim~ It is expedi,tnt
. to locate reasonable sized escapes at eTelY 50· miles in a long canal.
The capacity of the escapes depends upon the effectivcnes'l of the fac~ors contributing
excesses in the channels. As the escapes also serve as safety valves to t!xpediate repairs to the
damage done to canal banks by (luts, it is the rough rule to fiX the capacity of an escape equal
to about one half the discharge of the channels from which the escap~ takes off.
The head regulator design of ali escapFl is just similar to the design of a canal heati
regulator described in Chapter IV Part n UnaUy the escape head regulators have low crest
levels with undershot regulatiun so tJaat the escape head may als) work as scouring sluic{'s when
water can he spared for escapage.
The channel section design of an escape is jU3t similar to the design of surface drains
as described in Chapter II Part IV. The slope should b! a" steep as the levels can permit.
The design of other masonry works such as bridges and falla if unavoidable, is slffiiJar
to on. used in the c:oase of canals.,
7. DI5&rlbutries and minors.
No direct o~tlets should be given from the Main Canal and Branche!. as th~ regUlation
. of supplies is very dIfficult. All outlets should take off from Distributaries and Minors. .
The water is drawn a.t proper intervals from the main line into moderate. sized branches
"~ich ara so arranged a, to cOlllmand the greatesf area of land and to s~pply the laterals
, (distributari•• and water courses) in the most direct manner ..
 364
Location of Distributries.
Distribution from a distributing canal is m0st economically effected when it runs along
a ridge, so that it can supply water to its branches and to channds on either side. In ~he c~se
of main canals this location can be made only in occasional instances, but the distnbutanes
taken from these, mains should be made to conform to the dividing lines between water courses.
The capa :ity of the distributries, which traverse the sepal ate dra.inage divides, are prop()rtional
to the duties they have to perform, (the natural bounrling streams limiting the area they have
to irrigate).

'.'''--

Fig~ 1
.
'-_ ...... ,

,
...

Fig. 2

Careful serveys should be marle and grea te3t care taken to h aJance cuts and fills and
to so locate the distributries that the least loss of water shall occllr from P'1; alation.

Go.d
Fig. 3 Fig. 4 ,
The figures 1, 2 and 3 above show an ideal distributary system. Such an arrangement
enables the least mileage of channels to command the greatest area of country by furnishing
wat~r t~ both sides of its line, At the same time perfect drainage is obtained by the water
fl? W1!lg In ?oth directions into the natural drains. Fig. 4 shows the faulty alignment of
dlstnbutanes and drainage channels.. .
 365
Design of Distributaries. I ,?
\
The capacity is proportional to the the duty performed and water surface kept at a
level where flo N irrigation is possible. Th~ cross sectional area is diminished with steady draw-
off from the outlets The capacity of the different parts of the system must. be based on the full
supply capacity. .
In order to avoid high banks and to ensure the surface of water being above that of t~e
countr~, the slope of the disty : should be made as nearly parallel as possible to that of the land It
traversps. Hence alignment falls must be introduced.

Dimensions of Distributar:es .
. The' greater the ,amount of water discharged by a distributary, the smaller will be the
proportIOn of cost of malUtenance. Hence hrga sized distributaries are better than small ones.
The minimum bed usually kept is 3'0 ft. i:' . .
8. Intensity of Irdgation and water allowance.
The relation of water supply to the lanrl depends on the raiafall and the q 11ality of land,
as the depth of each watering is fixed by the composition of soil. The soil survey should give
both the chemical composition and the fea<;ible texture. The soil must contain chemical plant
f?od and humous (decayed organic material). It should be free from Alkaline salt. The periodi-
clty of watering is fixed by the climate, the crop period and the number of "aterings required
for the maturity of a crop. The losses in transport have also to be considercd.
. Generally speaking the intensity of irrigation in the Punjab may be kept about 50%
In the case of proprietary villages, where means of irrigation such as wells exist. In the case
o~ ~ro'vn Waste land of the colony areas. the intensity of irrigation from 75 to 80% will do
gl,:mg about 20% land for rest during the year. Near the Cities where the vegetable crops are
raIsed, the intensity of irrigation shoull be cent percent.
Generally speaking a water allowance of about 3 cusecs per thousand of the C. C. Area
is enough in the Punjab and in the case of the sugar, rice and garden area, about double the
quantity will do.

9. Capacity 01 Channels.
The record of areas taken from the Settlement records of the villages to be irrigated by
a distributary is used to work out the capacity of the channpl. The area is abstractl:.d in a
statpment called the chak-bandi form given on page 366 in Form A. Usually chaM, salab, abi and
banjar shamlat induding ghair mumkin areas are not allotted any water.
The uncommanded area is excluded, The remaing area is called C. C. Area to which water
allowance is applied and wari is allotterl.
The command statement is worked out as shown in Form B to work out the levels
required downstream of the outlet to command the chak served by it. It is a bad practice
to include the uncommanded area in the C. C. Arpa, because it rasults in temptation to the
cultivators to head up supply in tm. parent channel. This upsets the equitable distribution of
supply and results in silting of the channel. Then the Capacity Statement of the distributaries
and minors is worked out as per statement form C attached. The columns of the Capacity
Statement are self-explanatory and need no comments as detailed information in designing the
sections has already been given in Chapter V of this part.
 Sl{1'ml3}I 1
~ tIC
' '-At ) ..
til
...o
. QC.9
"c 1:
;il'"
....
""""
....O~\ ..... .~ t,}
;

z' ~
<. ~~Ol
~ ':...,.,"':'.
,~. ct-
t: > 0
.;,-<·ti(UU
c:I .........
Q,)""" 0
.
U.
~ -
..... .<I

110
" ._,g ....
.; ~
.0 .9"0 'i
.s;Q:! ~
.... ., 0 ....
Q

....o a-.. .. o..c: Po

gCl- c .. ,. .. .
::: 8
Po
~

:.~8u "' ..
kIlO
<
,.
ui.9 '
\
....• I ,- ~"O :.:
'O.~ .~
.; ~Cl
0)

-..
'"'
.; ~!, I--Pa P..

= ·r
1~1 (!):t
ui.9.:.:
~"08~:ri
13.; >.o.:i
.., .~

<OC
I pa1)U'BWWO:l-U[l
..j&:;!"':i
~2:A
00

'EC'
p<lpll'BWWO:)
..."'~.
.. ..,
~
I

~
..a .._
..... .1 ~ papU1lW
'
UIO:l-lI!l
p<lpU'IIWWO:) -
~ 'ii'R° u
~ .. "tI •
~'5 ~:::
it"

...
R ~
'tr. O"..c:
~
.....Ol
..:::
IS /l)<lPU'Bwm03-U[l
'"'
z·
~
..::: ::s~
• \.) papU'BwwoJ
R~o
._ .0
.....I>->
= ~ i~I P<lPU'BWWO:l-U[l
Eo<
:;u~

...
-<
15'"
0- .!_ <D
os

...
Ol

....
_~0
"'~
~ ~! P'BPU'BUlW,,:)
. -
0 'Cl"'
(fJ

Z
-<
~:::;::;-

"0
... R .
-...
~.~
0 '", P<lPU1!WWO:l-Uil ~ o!:s;t
;:-
p3pUllWWO:)
(J)
=s SOl lIj
(i .
u
. t! ~::;;
~

J;
0
u
e'l':
I ....
I 0'"
~
.;: u"
I
\:! popa"ww03-U fl' i .,
~'" JI"P"1IlUWO~~'1
00 .0
,1 - .... 0.0
, .~ c: ~.!
i ~.~[ . ·1
\:! ~ p<lpU1lwm 03-U fl to-
~~-g ~
tJ~ g'S
~

!
i d~ . papUllWUIo:) I ~'.'"

... .;
tlO

,
I. ~
.,'"
..:::
l (C :!
k

°
.....
..c:110
'ON \ pt.~(.!l ."
uaUl91nas ~

E>~m'BN
l<lUMO II ... •..
R
-
t:i~
0 ....

- ::;!
~
}lJ

u ~ 0

...... ~
:V1 f1 '1f}1 \
JO IlUl'llN ! I I')

0
cO
tl1 v d I I" Z
'iii
.... JO IlW1IN
'oN l11!Je S I j
'0::
"
III
 ra87
C":)
q~dap 'S'iI Sl[llnU3U C'f

-W'g:
qlP!M pag: iju!lJaIaJ. 10 -'1 C'I

-
> C'I
)l pal! IIoqd!lOsaG

adolS
S C 'f
....cd
0 ....
• <ll
;.-4 ... '" 0
~
·M'A'J.lO Iii U
.... Ii
;l en
::s :ri
q:>ea, {pea
~3nqos!p lelo~
~e

!
- 0
0
i ~
,
...
00

satpea l
Aq 1l0ndlosqV ...
C'I
iii
Pi
.. -.
r-.

0-
>.
!-<
...
fD

saq~eal

Aq a3leqos!o
-... P1!3H ZIl!l[.l0M
II)

qovaU
-
o

....0
''J'M II! 'TS'.!l

"']0.0
...
'<to

lGnno
~ I·S C":)
JO alhvq:ls!O ~i-
-., 0.0'"
-C;;i-~
C/l-_
~Q
"II<
Qt"-l
'!fs'lJ

'l!alV 'd 'V

- 0
1:1>'
0 ....
;; .~
~Q
I
'"
,Q
--...
C'I '"'
Z
1-1
f-
1-1
....
If)

><
A
~
en
0
~
I CIl ~
0
~
l1l3:1
lad Al!Su3lll1 AlS!O JO -<f M I 0 ... ~

'AlS!G en
. JO -qoS!G 'S .!l
llaA!li 3q
ol st .lal'llM I
q:l!qA\ lOJ 'I'aJ~1 ...'5" "'I
<ll
00
., 0.0
.~
Zl'~
~
CO
.... I-~
,.c:: I<
."... -- Q.
.~
~I~
r-.
~ 'V3l V
"
U
pald')o:lV
'sJe3A;
U U)
S lSVI !l0!l-e8!llI
-. .; /aliV.I')AV

"...
0> I·
. alq!Ss!uxlad IV~lJll~
II)

.,"
'"
--~-

'<to
0 'V":)''J
c3 .
'3li1!iI!A Cf:)

lanno JO
ap!s pllV 'o'u
'lenno
JO ap!s pile 'cru N

'oN IvpaS ... , \

 868

pgq:>ledsQP
WJOJ q:>!qM
qH~ JgHQf JO
;}:rep PU'Il ·oN
'-~J~O~l~d-!~~:>~dO~J~\~---------------------------------------------------
JO ::Ilt!a

i
----------------------------~-----~--I
·ON I"e!JJS
 369
10 Watercourses aad 0 utlets.
The watercourses are designed on contour phns, scale 4 inches to a mile, showing
contours at one foot intervals. The watercourse usually follow3 the ridge to avoid embankments.

FigS

The length of watercourse is generally limited to 2 miles. The watercourse is usually branched
to reduce the length and to gain command, lowering thfTeby the highest levrls nquired down ..
stream of outlet when working a!> calculated in form (B) given in para 9 above. The usua
watercourse slope is 1 in 5000 and the command at the fiel. 6 inches. The workin~ hearl fo
'he outlet is kept not less than 1'0 foot. A typical layout plan of a watercourse is given in Fig. o.
 370
The desigl'l of the Qutlet is worked iIi a: statement called the Outlet ~tatement. A. copy
of the usual form used is given in Form D. The columns are .$elf-e~planatory, anc;l, need no
"luddation. The sl!b-proportional outlets are used in the head re:lches of' a Distributary, such
as A. P. M. set b~low .6 D down,to 1. 2 D. Proportional outlets are desirable in the middle reaches
: set at .6D up to .8 D because the setting )f the ~. pM .. to give due share of supply is ;8D~. In
the tailreaches of a channel where the command IS poor. open flume outlets are desirable, as
they neer! comparatively low WQT king heads and are sensitive to absorb the exees,> reaching the
ta'l in the case of slack demand. Theory of ontlet design has been described in dptail in
Chapter XV.
11 Rules governing the sUbmis'sion of Estimates for and the con",auction of IrrigatLn work"
, ,l. Ge~eral.

I The following rules define the proced'hre to be adopted in connection with irrigat fon
projects. Throughout the rules. the term "Irrigation projects" or "Irrigation work" SllOUld be
held to include the Navigation, Embankment Drainage, u ater-storage projects or work,.
The expenditure on works only should be the criterion for determining the authority,
whose sanction is necessary to an estimate (original or revlst'd).
II Classification of Works.
2 Irrigation works for which capital accounts are kept are classified as either" Productive"
or"unproductive". The criteria to be adopter! in deciding questions of classificattotl: are given in
paragraph 12 of this Chapter.
Rules regarding the prepar3tion of Irrigation Projeets.
3 ·T]ie. papel s to bP. suhmitted with the projPct for an irrigation work consist of a report,
such pl"n~, measurements, quantities, rates and specific~tions as may be necessarv to enablp the
suitability of the designs and the <ldequacy of the estimated cost of the project to be adjudic<lted
upon, and an abstract classified under the heads pre:crited in rule n, bringing the various
component items together. Thes~ documents togethel form what is called the' Estimate".
4 Every project for an irrigation work should contain a full report explanatory of the
project. It i1'; impossible to lay down headings which will be suitable in tIl cases, but the
following points should, in general. be dealt with.
5 The report should give a general description of the. proposed works, including the
sources from which the supply of water is to be drawn, the maximum floods likely to be
experienced, the 'ltlantity of water available at different. periods of the year and the quantity
proposed to be util:zed, Reference should also be rrade to any rights in the water already
existing. The reason for th~ adoption of the particular scheme recom mended in pref~rence 10
any other shQuld be ex.plained, if necessary. the rainfaH and depth of s?ring level in the
tract affected, the !';ourses of existing irrigatirm, and the means of orainage proposed. if slIch
are required, should be dealt with. The lengths of the main channels and distributaries
should be given. These Jengths heing compared with those of channels 6f similar capacity
actually constructed elsewhere, :lnd a full account of the construction governing the designs of
all important works, apptmded, with a careful \lualysis of any engineering questions involved.
The cost of additional communication (other than the cost ot construction communications)
and other extaneous works involed in the project should be regarded as parts of its cos~, but the
report ~hould include an estimate of expenses under this head so far as data for such an estimate
are available.
6. Figures shouli be given showing the area of land commanded, the average area
l1su:Jlly cultivatf'd and the area probably irrigable. Tlie opinion of the local revenue officers as
to the desirability or necessity of the projected work, the fairness of the water rates propo~ed and
the probability of t,he an.ticipa.t~d financial n~sults being realized sho'!lld inva'riably ~e statecl. The
returns e xpeded from the work~, with a note a~ to the basis on which thy are circulated. should
be exhibited in form No. III appended.
7. The lepbIt should also deal with the .qnt'stion of labour .and the sources whence
it is obtainable, the probable effects of the operations on the existing rates, the localities
whence materials can. be ,obtaineci and th1l facilities for manufacture. The manner in which
the rates adopted' 'forestimate purposes h;J.ve been arrived at should also be con'tmented on.
8. The methor! propose for carrying out the work should be described ann in particular
such matters as the extent to which it is intended to use mechanical appliances, to employ
manual Iahollr to give out work to large manual labour contractors or to resort to the agency of
departmental la?ourof j)e.tty contractors should be dealtwith; T,he ,No. o~ cirCles and divisions into
 " 371

which the work is to be divided and the' time likely to b<l 6ccupied on construction should also
b€ stated. Any permanent increase in' the cadres of Eagineers which will be necessary to provide
for eventual maintenance of the works should b~ calculated and ex:plained, and sanction shoUld
be applied for any posts, temporary and permanent, which the authorities subordinate to the
Provincial Government are not themselves competent to create. "
/9. Special attention should be given to the following matters, which have been the
cause of excesses in the past: - , .
(a) The preliminary operations, including the surveys, , both of alignments and soil,
with triaL borings, where necessary, should be completed under the orders of competent
authori~ies. It shoUld t e clearly stated, in each case. to what ~xtent this has been done and
what still remains.
. The cost of such operations should in the first instance be charged to the Major Head
" 18 -Other Revenue Expenditure financed from Oroinary Revenups (2) Misc: Expenditure".
. (b) In addition to the usual charge of 5 percent fJr unforeseen contingencies
which percentagp. may, for reasons recorded be increased if circumstances render it desirable. all
incidental expenditure which can be foreseen, such a s sheds for workmen and stores, etc., should
be separatl~IY provided for in the estimate, The utilization of "over alI" rates, intended to include,
such expenditure, is to be deprecated. '
(c) The estimate of the cost of acquisition of land should receive special attention"
and a valuition should be obtained in every case from the District Officer concerned. '
(d! The Distri t Officer should also be consulted as to the' number of bridges which
will be required, and :::are shoull be taken that an adequate number of c"capes is provided.
. (e) Special attention should be given to the probable cost of foundations. the,
measures to be adopted in connection with them, and the amount of pumping likely to be
necessary.
, (f) The allowance. if any for special tools anrl plant should he carefully considered.
The inclusion of a Jump sum for the purpose i.;; gpnerally to be deprecated and. if this method,
is adopted, the provision allowed should be tested with reference to the probable cost of the
actual plant required. The amountof credit anticipated under this head and the Source whence
it will be obtained should be fully explained.
(g) Where the e:;timates of several structures of the same kind are derived from a single
type d~sign, the most unfavourable conditions likely to occur should not be overlooked.
(b) Estimates framerl on the hasis of analogif's from existing works are usually un-
reliable, and, before this method is adopted, the correGtness or otherwise of the analogv should
be very carefully tested for selected portions of the work. Tn panicul:Jr, analogies drawn from
snaIl works should never be relied upon for the prepration of larger projects.
(i) All calculations of qUflntities and costs should be independently checkEd to obviate
clerical error~.
(j) When the circumstances of a project are "uch that tl:ere is reason to suspect that
expenditure. other lh'ln that which Cin be dtfinitely foreseen at the time of framing the estimate
may become necessary during the course of construction provision to meet it, under the heart
• unforeseen expenditure' should be made in the estimate. When this is d(}ne. the circumstances
should \:;e fully explainen in the report
(k) In estimating the f<?venue likely to be ohtained from a project, consideration sJlould
be ~iven to the fact. that. in some .cases~ the f,ull discharee or .storage Will not invariably be
avaIlable and that, m others even If avaIlable, It may not be reqUIred The average discharge
or storage likely to be utilized, arrived at by analogy from similar works elsewhere, should be the
basis of the revenue estimates.
\1) In calculating what price will he secured fot Government land, the sale proceeds of
which from part of the estimated revenue from the project, regard should be paid to the probable
relati?n of supply and .demand, speci~lly when 3: project or different 1?fojects, are ll~ely to bring
land mto the market Simultaneously In the PunJab Of any other provmce or an IndIan state.
10 fhe expenditure proposed to be incurred upon a project should in every case be
restricted to the minimum amount necessary to ensure the succe"s of the undertaking. Estimates
~, should be framed keeping in view the rule that eVNY . public officer should exercise the same
L'Vi?ilance j1\ respt:~t of expenditu~ ,incurred from Go\'ernment revenue as a person of ordinary
 372
prudence would exercise fn respect of the exp~nditure of his own money.
11. fhe co:np\ete estimat~ for a project besides incluriing all anticipated direct charges
.. houlrl furth~r in:lude a:; iniirect ch'lrges the amount required to cover the capltalizltion of
abatement of land revenue on the area o:cup:ed by the works, cost of audit and accounts
establishment and simple interest on Capital Outlay prior to the work being brought into
operation. The direct charges should be classified. under the aopro .Jriate sll~-hea.rls of Account.
the main headings being (1) Works, (2, E,talJli3hment. (3) fools a'1d Plan t and (4) Leave and
Pensionary charges. The items incluied under the head "Work," should be cia,sifi.dd under the
prescribed "head Service" and 'D~taileci heads' enu n~ratpd in F0rm l[ The c03t of surveys'
including expenditure incurred prior to the ~u) nissil)n of the project, should also be included in
the estimate An abstract framed on th~se line;, in the Form No. Il appended, m'lst acc')mpany
every proje<;t submitted for sanct'on. In the cas~ of a large project the subNorks mlY, if de3ired,
be further sub·divided sc> a<; tf) show individual large works.
12 If it is known that the project will constitute the sole charge of one Of more
Superintending Engineers the estimate of th'! Cl)st of th} Sup~rint~ndin~ Engioeers and other
establishment should be framed in d~tail, a suitalJle amount hein~ added to cover an ddequate
shine of the cost of the Chief Engineer's ec:tahlishment. Even where a S'IDerintencJing Engineer's
charge i<l not in question, be cost of the esta'JIish'11ent required for the supervision of the
construction of the project should ordinarily b<! estimated in detail. hut Government may, at
their discretion, calculate the provision to be made for establishment in such a case on a
percen tage ba..is, provid~d that th~ percentages are j u:;tit'ted by c'Jmpanson with past actuals.
In the ca<;t} of ordin 1 ry fools and Piaot, either th~ actull anticip 1tej c )st, or a rea<;onable
percentag,~ allONance, bi,ed on past >'.ctu1.ls may be entered in the estimate.
Unless Government directs ot'1erWlse the provision to be made br es~ablishment should
include 25 per cent on the estimate of Works Outlay for dep'lrtnental establishment and 1 per
cent for Audit and \ccOllnts, while the provision for Tooh and Plant will be I ~ percent on ~he
Works Outlay Estin ates for large surveys for new irrigation projects should. hONever, provIde
only for a charge of 5 por c"nt on the cost of sp':)ci~.l e,tablis'lrrv~nt to cwer th~ supervision
charges thereon. In the case of irrigation project. hr which neith<>r Capital nor Revenue :\CCJunts
are kept, it is unneces<;ary except in the case of large surveys for lieN irrigation projects, to enter
provision for Established and fools and Plant in the estimate unless, for any reason, it may he
deemed desirahle to do so in order to forecast the ultim(l.t~ fE.su1t of th" project.
13 Charges for capitalization of abatement of land revenue should be calculated. at
twanty times the amount of land revenue remitted. while provision for Leave and PenslOn
allowances should he at the rate of 21 per cent on the gross charges for Establishment.
14 Simple interest on Capital will be calculated at the rate sanctioned by Governments
from time to time. on the probable annual outlay.
15- No provision should ordinarily be marie for the minor head "Suspense" as this
head in the accounts represents services of a general character not necessari ly pertaining to a
particular project. If howtver, one or more Divisions are expected to be maintained
exclusively in connection with stock required for the proj(ct, provision for "suspense" may be
inclurled, but only to the extent of the balances likely to be outstanding under Suspense on
the date of closing the constructiln
IV. Storage Projeets.
16' The report should, in addition to the information specified in rules 3-15. give thearea of
!ank and contents when full, the area to be irrigated per unit of storage the length of the dam,
Its maximum height awl shape, materials proposed to be used in its construction and the mode in
which the Wa ter is to be let off for irrigation, The questions. of the available water supply, number
of times the reservoir will probably fill. rainfall and proportion flowing off the catchment, lo~s by
evaporation and ahsorption etc: should De fully dealt with as well as the quantity of floodwater
for which provision must be made. the flood absorption cap~chy of the reservoir, and the water~
way of the escap' weirs or sluices. The re~ults of any experiments bearing upon the strength
flf the ml,terials proposed for use in the dam should be dea.lt with, as also the silt content of
the water and the probable effective life of the reservoir.
V. Projects Affecting any other Province or an Indian State.
17. Where any other province ot' an India State is also concerned, the report should
detail the arrangements mutually agreed upon fOf financing the works, ihe terms upon which
 373
the water is to be sharerl, the agency by which the work; will be con,tructed and wherl' an
Indian State is conCflrned, the ag~ncy by whom th! aC~f)unts are to be audited on behalf of the
Stat" . When a project or different projects are likely to brin a land into the market
simultaneously in differen~ provinces or Indian States, and the sale-p~oceeds of such land form
part of the estinnted revenue from the project, the report should state what arrangement the
Governments concerned have made to meet the contingency. A draft of any formal agreement
into which it. is propo ;ed to €ilter to regulate these and any other matters in respect of which
agreement is deem"d is necessary should a :company the project.
18. In all projects wJ.ich may affect riparian or other interests in Indian States,
Government will ascertain the views of the Durbar or Durbars through th~ political authorities
concerned.
VI E mba.nkment5.
19. In the case of new lines of ri ver embankmen ts, the report should show cl >arly the
financial responsibilities of Government in connection there with, and the manner in whi0.h it is
proposed that the outlay shall bp. recovered.
vn Nature of sanction.
20 The sanction accorded by the Provincial Government to a project for an irrigation
work shall be regarded as in th ~ na.ture ot an a·Lninlstrative approval to the project,and not as
the final technical sanction to the detailed estimC1tes of the works. Such technical sanction will '0~
a~cordej by tbos:! officers of the Public Wurks Department. Irrigation Branch, to whom
pJwers hav<! be:m d ~legated by G)V rll 11,!,lt. Odtail)d working estimates for individual works
in exceS3 of Rs. Ilakb must further be approv<!d by Government.
YIlt. com.mencement of Works.
21. It is a fLlnd tmental rule that no work to which these rules apply shall be commence 1
upon an Irrigation project, until the folbwing conditions have been fultitled.-
(a) rh~ approval of Govarnm :nt to the project has been obtained.
(b) There is a slnctLOned d~s;gil and djtailed e,tima.te for the portion to be commenced
(cl Funds have been allotted for the work.
When these c >nditioni have been fulfilled, Public Work's Department. Irrigation
Branch. are co:npetent to authorise the commencement of construction.
22. Government in the Finance Dapartment. in consultation with their \udit Officers
may prescribe rules to regulate expenditure de~itable to a sanctioned project upon such survey
and preliminary operations as may be necessary in order to ena')le the detailed ejtimates to be
drawn up.
IX. Rules Governing the Accord of Technicll Sanction.
2~. When a project has been sanctioned by Government, an officer of the Public works
Department, Irri~ation Jjranch, to whom power has hean ilelegated by GO'lernment, may sanction
detailed, estimates for component parts of the project against the amount provided for the
'Sevice' or "Detailed" head in the abstract estimate (Form No. [I).
24. (a) Detailed estimates, subsidiary to a project estimate' may be for a single work
such as a bridJe, a certalO num'Jer of miles of excavation, etc; or for a distributary in which
a numb"'r of small works are incllHied. In the first case, the cost of the work should be specified
by detailed heads in the abstract while in the second case the abstract should be prepared
to show the component items and liabilities separately in detail under the various service heads
of clas,ification But if a work of exceptional magnitude costing more than Rs. 10.000 becomes
necessary, a separat p estimate shou'd be prepared for it.
(hi A separate estimate should be prepared for eaeh distributary or, where thp.re are
minor channtls, tWG estimates one for the main distributary and its important branches and one
for the minor channel" but the whole e'<pendlture under "A. Preliminary Expenses" should be
provided for in the estimate for the main dIstributary.
. (C) General charges for establishment, audit and account, and tools and plant should
be entered in a project estimate, but will not appear in any estimate ~ubsidiary to it, but. after
. a proiect estimate has been closed, estimates chargeable to the open ca;>ital account shoul d
contain provision for these charge 0; except I per cent on account· of audit and accounts
establishment, at the rate given in rule 12.
 374
- 25. When it uecomes evident that 'the amount provided for a "Sub-work" or a
• detailed" head in the project estimate will be exceeded, the following rules must be obsevred
. (a) Subject to any orders which Government may pass in the matter, the officer-in-
charge of the project may transfer provision to meet such excess from another detailed head of
the same sub-wor« on which a saving is anticipated.
(b) Should it become evident that an excess over the amount pnvided in the abstract
estimate for any sub-work will be exceeded, the officer-in-charge of the project must report the
fact immediately to the Finance Department. He should, at the same time. intimate what
savings, if any, are anticipated upon other sub works of the project. .
(c) rhe Finance Department may -transfer provision from Ot,e sub· work; on which a:
saving is anticipated, to meet a probable excess on other, or it may permit an excess over the
provision made in the abstract estimate for any sub-work 11pto an amount to be stated by it.
(Note.-The sub-works into which the minor head "Works" is divided are enumerated
in J.or~n_No. II appended to these Rules.
X. Modification after accord of Sanctions,
26. After the Government's approval of the Project for an irrigation work, the Chief,
Engineer m'ly, if necessary, and suhject to the provisions of rule 25, modify the details of the
works; providt'd that if any such morlitication is in the opinion of the Chief Engineer substantial.
a report of such modificatIOn shollid he made to Governmpnt.
Note.-Modification will include abandonment of iterns included in the original estimate
or provision of items not included therein, and an increase or reduction in the area to be irrigated
by ~ the Proj ect.
XI. Rpports of Probable Excesses.
27, Whenever it is ascertained that the expenniture upon any project is likply to exceed
the amount sanctioned by the Provincial Government by an amount' higher than that which
the Public Works Department. Irrigation Btanch are empowered to pass, Finance Department
should be immediately advised of the anticipated exess without waiting for a revised estimate.
The revised estimate. if npcessary should he prepared in due course. and sudmitted to
Government with a full explanation of the causes of the excess ann of the probable effect on
the financial results of the work. The Finance Department should also be imrnediatly informed, _ .,
if at any time during the course of construction, it b"comes probable that a work, sanctioned as .' -.'
productive, will fa.il, in operation, to satisfy the criteria which must be satisfied before a. work
can be regarded as productive. - '. ~ . _~ '._
_ ~ " ~
.... _ l , ' h • .""i·;"
X 11 Supplementary and Revised Estimates.
28_ Any development of a prJject thought necessary while the work is in progress
which is not fairly contingent on the pr,)p~r execution of th;! work at first sanctioned, must be
covered by a supplementary project e,timate. accompanied by a full report of the circumstances"
which render it necessary. The abstract mmt show the arnount of the original estimates and the
total of the sanction required including the supplementary am')unt.
29. A revised estimate must be submitted when an original san~tionerl estimate is
likely to be exceeded oy more thln 5 p ~r cent and a second revised estimatd when any excess
is ~mticipat ... d over a sanctioned revisci estimlte. The revised e3t;mate should be accompanied
by a comparative statement in Form No. ) Il.appende<i, comparing the revised estimate with
the latest exist:ng sanction of competent authority. and by a rep'lrt showing the progress made
to date.
30. When a revised or supplementary estimate is su~mitted under Rule 28 or 29, it
should be accompanied by revised financial forecast statements as required b the case of an
original estimate (vide Rule 6).
XIII. Utilization of completion report as revised estimate.
31. WheD excesses occur at such an advanced period in tbe construction of a work, as
to render the submission of a revised estimate purposeless, the excesses may with the concurrence
of tbe Finance Department be explained in a completion report prepared as prescribed in rule
33. The adoption 01 this procedure in this way .dispenses with the necessity -for the immediate
report of the excess required under Rule 27. : :~ .
 XIV: Completion Reports.
C

. 32. , The estima.te for const(~~tion aLan irrIgation work should be clo<;e~ a.c; s'o;~ ~c;.
the project is praCtically in full oper<I!tion. although there may be c~rtain works provided for in the.
estimate. either unfinished or which 'it is not d<:sirable to construct at once. .
33. As soon as the constCllOtV'o estim te has been closed, the public W lrks Depart_
ment. Irrigation Branch will prepare ,for sU'lm'ssion tJ Govprnment through the Finance Depar:.
ment, a completion rp.port of the project: 'co Ilprising'the following documents in Form No. IV
appended. Completion repor.t in th}> form s~()uld. be..prpareci only in resp~ct of p£(')jects ianctio '! d
by Government. For sm'l.ller proje~ts, and other open capital works which are sanctioned by top.
Pu'llic Works Department, lrrigation Branch, co npleti ,n reports s.hGuld be prepared in 0 e of
the forms p'''escribed. . ..'" . '" ." ....
: Schedule A. A, statem mt showing by wor'~~; and sub-works, clas,>Hied ullder the relevent
aetailed heads. the actua.l expenditure on works completed upto tile' da.te of the closing of the
'Construction estimate. . . .:' ~ . ' . ,. ,.
Schedule B. A statement of works which are within tJ.. e scope 01 the sanctioned
estimate, and of which detailed estimate~ h!lve been pre'paten ;"nd sanctioned by competent'
authority, which were incomplete!ot had not been begun on the date of the closing of
the construction e s t i m a t e . . ' .
Schedule C. A staternE'nt 0.£ works, whether included in the construction estimate or
riot, which have been sanctioned by competent authority, between the date of closing the
construction estimate and the date of the submission of ·the completion report. • .
Sehedule D. A statement of works for which no estimates have been sanctioned upto
the date of the submission olthe completion report. but the probable exp~nditure on which can
be foreseen and which is necessary to complete the Project. (. .'.
Schedule E. A statement compiled as a combination of statements A,B,C & D. This.
statemflnt should also show, for purposesof comparison, the sanctioned estimate by works an i
sub-works, classified under the relevarlt tletailed heads of account.
A repQrt, on' the works executed lipto. the time of the ,closing of the construction
~stimateand an index map or maps showing the Project as completed, will accompany the~e.
documents, fhe report wilL discuss the - financial result already attained and expected in the
fuiure and will be accompanied by' forecast fin:lncial statements in Form No. 1 based on
Schedule E ,above 1. e. on the total antidpatell ultimate ~x:pendit\1re on the PrQj~ct. Part III
of this form will be signed by the Chief Revenue Authority of the province.
34. l'he schedules A to E ,accotnp:J.nying completion reports should i!litially be signed
by the ~ ?:ncer. in- ~har!2'A of the Project (who is particularly responsible _. f6i-.figures in co~u;m~~
5-9 of '-'cnedule D and consequently column 10 of Schedule E.l and ,coumer-sIgned as "venfIed
by the Accoun~ant Geaeral, in token of his verification of actuals and classification.
35. The finaIlcial statement submitted with (';)mpletion reports should similarly be SIgned
and cQlmtersigned>c but 'in this case t4~ Audit Omcer sho.uld do so under the words "Actuals
and calculations checked". ,
36. These documfnts should ordinarily be prepared ani submitted to Government within
'6 nonths of the closing oj the cons~ruction estimate, o:r 12 month~ in the case of.. an exceptionally
large work. If this. is nat found possible within the . per~od specified, the Finance Department
·should be advised of the reasons for delay _and t,he probable date when the documents may be
expected to he ready. . ' .
. 37,. Schedule E.will be treated as a revised forecast of expenditure against the'sanctioned'
Project .. All important works whi~h had not been commenced and which were within the scope
'of the sanctioned estimate should be included in schedules B, C or D as the case may be.
38. Subject to the restriction tha~t thetotal expenditure against thq project shall not
exceed 'the amount sanctioned for the Project by ltn amQunt ,gr~ater ,th~n th'lt which t~e
Departments are empowered to, pass, the Public. Works D(jpartment,. Irr.lga!ion Branch, 15,.
competent ,to pass expenditure between the (late of closing the Gonstru~tIon estimate, and that
of. tDe' approval' of the completion report by' c()mp~tent authority on- (a) 'Works entered
in schedules B. & C. (b) Works, ~n~ered .in Schedule D, within the limits and subject to the
'conditions 'specifif!d. .' .
39. On receipt of approv,al of the Government to the completion yeport. works included
in Schedules B. & C. ma.y be carried to completion by the Public Works Department, Irrigation
Branch, within their powers of sanctioning excess over estimated amounts, approval of Govern-
ment being obtained to any higher exces<;, Public Works Department, Irrigation Branch may
also. on receipt of such a )proval, sanction further outlay on other works aga inst the open capital
account of the Project subject to the conditions usually laid down.

FINANCIAL STATEMENTS.
FORM No. I. Part I.
Summa.ryof the estimated direct charges to capital account.

Establishment including Tools and Less receipts

Suspense. Total. on capital Net
Years. Works. leave salary and Plant.
pension charges. account. total.

1 2 3 4 5 6 7 8
---------------------------------------------------------------------------*
Form No, 1 Put [l
Summary of the estimated indirect charges to Capital.

Cbarges on account of
year. Capitalized abatement of land revenue. audit and accounts Total
establishment

I 2 3 4

FORM No. I PART III.

Estimate of growth of Irrigation and revenue receipts and charges.

Revenue receipts and charges.

Gross revenue due to work. Net revenue due to

work.
Charges
y Irrigated area Enhanced both
ear at end of year. land direct and Including Excluding
Direct
revenue or Total. indirect enhanced enhanced ~
receipts. indirect against land land :;
revenue. revenue revenue, revenue. E
QJ

account. P::i

1 2 3 4 5 6 7 8 9

FORM No. I. PART IV.

Estimate of net financial results of-years after the probable date of completion of the work.

Simple interest at Net revenue

% on capital outlay includin~ Net
Direct capital Direct capital to end of previous enhanced Simple revenne
Year outlay during outlay to end year plus half land reVenue intere~t less less simple
the year. Qf year. outlay during tbe column 7 of net revenue. interest.
year. part III.
1 2 3 4 567
 377
Form 1I. Abstract Estimate of cost~_~-:-;-----;-___ Project.
Minor head. Sub-work. Detailed head. Amount.
Works. Direct charges
1. Headworks. A. Preliminary expenses.
B. Land.
C. Works.
K. Buildings.
O. Miscellaneous.
p, Maintenance.
Contingencies.
2. Main Canal. A. Preliminary expenses.
B. Land.
D. Regulators.
E. Falls.
F. River and Hill torrent works.
F. (I) Other cross drainage works.
G. Bridges.
H. Escdpes.
I. Navigation.
.T. Mills .
K. Buildings.
L. Earthwork.
M. Plantations.
N. Tanks. and rivers.· •
O. Miscellaneous.
P. Maintenance.
Con tingencies.
2. (a) Branch No. 1 As for main canal.
Z. (b) 1?ranch N o.2 do
.. i.
3, (a) Distributa,ries
~ .
group No.1.
A. Preliminary expe ses.
B. Land.
C. Works.
·K. Buildings.
O. Miscellaneous.
P. Maintenance.
COBtingencies.
3: (5) Distys : gtoup As for group No I
No.2. .
3. (c) Distys; gnlUp No.3. do
4. Drainage and pro-
. tective works, .
5. Watercourses.
6. Special Tools and plant.
Unforeseen expenditure.
Establishment includ-
ing leave saJary &
J=ension charges.
T. & p,
Smren,e.
De'luct receipts on
Carj'al Account
TotaLDirect charges. Indirect charg~.
Capitalized abta.. .
ment of land revenue.
Audit ano. Accounts .
. Total indirect charges
Grand total.
 FORM IlL REVISED ESTIMA TE.
Comparison between original and revised estimate,

o" I Modifications Total (/)

Minor Sub- Detailed flgtl~a sanctioned by sanc- revised ' E ~
work h ead . es t1- ('om PI' t ent t'lOne d e~ t'Ima t e Sa.vmgs. xcess. ~8
head.
rna e. authority. estimate ~

FORM IV. CO:VIPLE nON REPORT.

Schedule A.
Schedule of works showing actual expenditure on works completed upto the date of the
closure of the construction estimate,

( lassification.
•
Serial No. Su ':>-w,)fk Na!1e of C03t of
Minor heacl Det:tiled head work.
of item wJrk.

FORM IV. COMPLETIO~ REPORr.

Schedule B.
Schedule of works of which detailed estimates had bp.en sanctioned prior to the date
of the closure of the construction estimate but whIch were incomplete or had
not been begun on that date.

Classification. Outlay to
date of the ~I-o
,.C<l.I>,
~o>.
<a
.o .... ~
_.... §
v
.0 0· ...
Serial No Minor Sub- Detailed Name of closure of ~,..c::: <1:!
.0....,- "'1-0-
'Oll.l .... <'dv~
of item. head. work. head. work. the con- 01-0 .....
1-0 ::l ::l
c,..c::: ::l .g~15.
struction ~""o
1-0 .... 0
1-0 "0 El
~;; ~
estimate. .... 0
(,)

FORM IV. C0-'1PLE nON REPORT.

Schedula C.
Estimates sanctioned against the' Open Capital account suhsiquent to the date of the
closure of the construction estimate.

Amoun t Sanctioned.
o Date on
Z Minor Sub- Detailed Name of
<a head. work. head. I
worK.
Works. Estatlish- Tooh and
ment. Plant.
r ota1. which
'i:: sanction
""
(f) lapses

FOR~t IV. COMPLETIO~ REPOrt r.

Schedu~e D.

Schedule of works of which no detailed estimates have.yet b~en sanctioned but th~
construction of which is necessary for the completi)n of the Project.

Clasification. Probable cost of work.

o
Z·
-----
_8 Name of
·c"'~
.,..., Minor Sub.work. Detailed work, W k Establish- Tools and Totat
4.)-
mo head. head. or s, ment. plant.
 379
FORM
COMPLETION REPORT. IV.
Schedule E,
Comparison of expenditure by main and s'lb-head;; with the provision in the
estimate sanctioned by Government.

~ Expenditure up to Difference between

Classification. .§ -E the date of closing Probable further probable total out-
--------c~-7n :>, the construction expenditure lay and provision
'0 (1)..0. estimate. sanctioned
('d .S '0 i:i..:. .... ,.....----"'-~. -----;"'-,;-;-,-:....
;_~--:-.- - - - ---.,---=-=-=-:":-=-:":-=-:"::__-.
1:: Q,)<I) <I) ...... -.!<iXl ~
... "t:~o:l "00 <I) ..:.
"0 s::: t:: S "0.] ~ §~ a; co..... U (l) ~ <I) '5 ;:::l
<I)
.......
.S!.S
.~ t) ~
t:: S"'::S
0 0 "0
;:;
....
s: -;:::__. s:: 0
;.;
...
0
~ -:::
d.... :;:: .g(l)
0 ...... ~.n
C '"CJ
c:; <I)
:0
('d _
0
•
<Ii
Ul
'@ :> c:> u s: <I) ::l "0 1l "; 8 ~ -g U ~ 0) ~
Cd ..0 ('d:>, <I)

~ 8~~ t::"O-G t::~..c: "0 S:::C"Oi:5..c: O,.d o§::l"O 8b~ ~

~ Po; 0(1)'" o~iil f-; o::l(l)uiil U5iil ::>:;::"0 f-; 0... .... ~

1 2 3 4 5 6 7 8 9 10 11 12 13 14

12. Classi!iCltinn of Irrigation. Navigation, Embankment and Drainage Works into

"Productive" and "UnProductive".
1. Prnjects of Irrigation, Navigation, Embankment and Drainage, are of two classes :-
(I) Productive and (2' Unproductive.
2. To admit of a new work b~ing classed as Productive Public Work, the following
conditions must be satisfied :-
(a) There must be good reas~:m to believe that the revenue deri 'len from it, will within ten
years after the probable date of its completion. repay the annual interest on tile capital invested
calculat~d at 6 per cent but in preparing a project: for sanction no deduction is to be made from
the total capital outlay on account of. antiCipated excess of revpnue over simple interest.
Note :-Capital invested includes (I) direct charges, (2) indirect charges and (3) all
arrears of simple interest, if any i.e; balance of total interest over total net revenue.
(1) It must be susceptible nf having clear Capifal and Revenue Accounts kept for it.
(c) Its cllssification as a Productive Public Work must be authorized by competent
authority.
3. The rules for determining (l) whether a work which has been classed as productive
shall continue t~ be so classed, and (2) whether an unproductive work may be reclassed as prodtcc-
tiV6 are as follows, the percentage rates referred to being those prescribed for the time being,
and being subject to alteration at the discretion of the competent authority :-
I. Every irrigation, navigation. embankment or drainag, work for which capital
accounts are k"pt ShOllld, until ten years after the date of the closure of it, construction estimate,
be classed as productive, if the net revenue anticipated from it appears likely to repay, on the
expiry of that period, the annual interest charges on the capital invested (including direct and in-
dird charges and arrears of sirrple interest). calculated at 4 % in the case of works sanctionerl
before the 1st April, 1919, at 5% in the case of those sanctionedbetween the 1st April 1919 and
the 1st August 1921 and at 6 o/"in thecaSfl of those anctioned after ]st August 1921. Conversely,
if it is not expected to yield tt e relevant return, it should be classed as unproductive. If moreover,
at any time during the period of comtruction or within ten years of the date of the closure of its
construction estimate, it b'!comes apParent that a wor:< originJIly classed as productive will not
a~tUllly be remunerative according to tht cri tedon prescribed ahove, it shon ld be transferred
from the productiv to the unproduct:ve works. and similarly if it becomes obvious, riming the
same period, that a work sanctioned ::IS unproductive will actually prove remunerative, the ttansfer
of the work from the unproductive to the productiv~ class may be effected.
II. Every work c1C!ssified in accordance with clause 1 above will retain its classification
unchanged during the eleventh, twelfth and thirteenth years after the closing of its construction
estimate. . ,
III. If any irrigation, navigation. embankment or drainage work for which a capital
account is kept and which is classed as productive fails, at any time after the expiry of itn years
from the date of the closing of its construction estimate, in three 5uccessive years, to yield the
 380
releva.nt return prescribed in dause 1 above, it should be transferred to the unproductive class.
A work classed as unproductive which succeeds in yielding. in three succes,ive years, the relevant
return prescribed for a productive work may. on the same principII', be transferred to the
producti ve
IV.class.
If an existing irrigation. navigation, embankment or drainage
work be extended or improved, the creterion of productivity prescribed in
clauses I to 111 above or improvement, as whole system, including such
extension or improvement had been executed simultaneously with the
original work and· the date of sanction referred to in those clauses for
the purpose of determining the percentage to be returnpd by the system as
a whole. shall be that of the accord of sanction to the original project. As
an exception to this rule, if any extension he, owing either to its nature or
magnitude, such as may reasonably be considered to be a separate project and
if it be susceptible of having clear capital and revenue accounts kept of it, as
distinct from those of the project as ::l whole. it should be treated as a
separate project and in that case the conriitions relating to original
projPcts and not those relating to extensions and improvements shall be
applicable. In all such cases separ ate capital and revenue accounts should
be main tained for the extension in order to enable the productivity test to be
periodically applied.I, III & IV are, however, subject to the proviso that
V. Clauses
the compet€nt authority may postpone the transfer of a work from one
class to the other in cases in ~ hicl:l. it is satisfied that its success or failure is
purely due to transitory causes. .
4. Forthe purpose of the determinin~ the productivity of an old
work which has been developed by the Government only the capital
expenditUle expended by that Governmer.t should be regarded as the r
capital at charge on which. interest is chargeable. !:' ~ __.L...j.-I.-"__
5 .. The traQsfer of a work from the productive to the un productive ~ ! ~
category, or vice versa, will affect the recording of all future transactions in :II ~
~.~ ~
;:.
connection with it. No adjustment will be made in the general accounts in
respect of past transactions, but the necessary transfers will be .effected by;; ~ ....
the Accountant General in the Pro Forma Accounts of the work in question. j ~ o
o

1~.
o
Plans.
The fonowing list of plans comprises all those generally required;-
!':Il
"4

(a\ INDEX pLAN :-(Art. 58 (a) Usual scale 1 mile to 1 inch. This
should be a tracing of foolscap size, or folded to it, so that it may be bound
'tYith the report: a duplicMe copy of the plan should be placed with other
~/ ~ ~
::t
~
plans. (b) Storage reservoir :- (i) General Plan of the Catchment : - o,
Usual scale 1 mile to 1 inch. If topographical sheets exist, this plan should ...
~

be prepared from the~, the boundi~g watershed line should be clearly ~

marked, and the area 10 square mIles of the catchment printed on the l~
_.f::-,_.
plan.
. (ii) Contoured. plan of the reservoir;--Usual scale 660 feet to 1 inch ...
T~e plan should .shOW ~ll the main c~ntours and also all the works connected lilt

WIth the reserVOIr partIcularly the tall channel of the waste weir and it
outfall; a table of the areas and capacities of the main contours should b!
""
printed on it. ... :'" ~.' .

. (iii) Land .plan: - Usual· scale is ~2

1 . The
,000
plan should be

prlepared from th~ ,.Government land plans, if there are any, and s ••
. .. Fig .. 6
 381
should show all land to be acquired and the offsets from existing fields to
where Its boundary changes in direction. They should show the roadcrossings and village
boundaries. A typical land plan for a distributary is shown in Fig. 6
(iv) Dam :-Plan longitudinal and cross-sections; details of foundations and closure
arrangements.
(v) Waste Weir :-Plan, longitudinal and cross-sections details of sluices and temporary
crt:s t.
(vi) OutIe: --Plan, longitudinal and cross-sections details of tail and fore-bay:>
approach bridge, valves. lifting Hlds and capstans.
(c) Weir or Headworks:- (i) Contoured plan of the backwater :- As for restrvoir.
(ii) L':llld plan :- As for reservoir
(iii) Flood Embankments ;-Lmgitudioal aBd cross-sections.
(iv) HeadwOJ'ks:-General plan showing all works longitudinal and cross-sections;
details of all sluices.
(d) Main Canal and Branches:·-(i) Longitudinal Sections:- To show canal bed and fnll
supply line, ground line, and top of embankment; the location of all proposed works to be
;ndicated. Notes to be given of the bed fall, full supply discharge and the irrigated area under
command of different sections of the canal.
(ii) Cross-sections: -To be given of all typos and unusual sections.
(iii) Cross -drainage and other Works:-Plan, elevation, and cross-section of each main
work to be drawn. For minor works type designs to be prepared.
(iv) Plan of frrigable Area :-On this to be shown all villages, roads, etc., the main,
canal ani! its branches and sites of the principal works, drainage lines and channels and
the boundaries of ir) igable and unirrigable land.
14. Examination Question.
Mark a suitable alignment for extension of a channel for Irrigation of the area shown on the blue
print plan supplied. Divide the area into suitable chaks and work out the minimum F. S. L. required in the.
channel at each outlet head, discharge for each outlet and total discharge required in the channel at the beginning
of the extension. Design an A. P. M. outlet for one chak and a standing wave flume outlet for another.
Draw free-hand of dimensioned sketches for both. Intensity 70%. Full supply factor at outlet head 88 for Kha"lj
permissible. Khanj Rabi ratio 1 : 2. . (P. I. B. 1941)
2. On the attached contoured plan, the alignment of a minor is marked. Discuss for the advantages and
disadvantgaes of this alignment and propose a better one if you consider this feasible. (P. I. B. 1941)
3 On the attached contoured blue print plan mark the alignment of a minor taking off from the most
suitable point on the distributary to.) irrigate the area inside the irrigation boundary of the proposed minor.
Divide the area into suitable chaks keeping to the village boundaries as far as possible and mark these
chaks in the plan finally showing the positio];l of outlets and water courses to irrigate each square in each
chak (P. I. B. 1936).
4 Describe briefly what information you would collect and the survey you would conduct for the
preparatIOn of a complete irrigation projoct. (Mysore 1939).
5 \\'hat are the main pOints to be kept in view, in aligning a canal for irrigation purposes?
(b) \Vhat preliminary surveys would you conduct to enable you to fix up final alignment?
Mention the important deta to be collected for properly designing the channel.
6 What data would you collect and what are the plans you would prepare for an irrigation proiect?
7 Work out the detailed flow irrigtion project for the Karol Tube Well Scheme area from the Lahore
Branch. The following alternatives should be adopeted :-
(a) The eXisting Shalamar Distributary taking off from R. D. 218500 R. Lahore Branch to be
enlarged and remodelled to irrigate this area.
(b) A new channel to be designed taking off from R. D. 23000 R,crossing the Shalamar
Distributary. '
(c) F. S. Facter=250 Rabi Kharif ratio 1:1. Intensity of irrigation 100%. .': .
Complete draw off, Command and arta statements and alteration forms of outlets to be worked out
The alignmellt should be marked on the attached contoured pIau and Longitudinal Sections to be prepared with
cross sections at every thousand feet. The land plan should be prepared showlDg the detail of the hlnd to be
acquired. The typical design of the fo~lowing works should be given:- .
, (i) Distributary Head Regulator (ii) Meter Flumes. (iii) Falls (iv) Crossing of the Shalaipar ,Pistribut-
ary. (,,)Outlets. (vi) Syphons over the drainage a'ld tail clustero. ' , '; .
. " The cost of the scheme will be evaluated under the following sub head:- ','
(i) Preliminary surveys. (ii) Land. (iii) Earthwork. (iv) Service roads. (v) Masonry' works. (vi),
Outlets etc. (0. Miscellaneous.) I' (~" U. 1942)
 PART II
CANAL IRRIGATION
CHAPTER XVIII
Remodelling Channels
1. Introduction.
In the lac;t chapter, the student was acquainted with the design of new channels and
the preparation of their projects. This chapter deals with the troubles and the working vf
the existing channels. It is intended to describe here briefly how to di<.tgnose the trouble and
to apply the remedies. The object of all remodelling of the existing channels is to aim at
equitable distribution of supplies with a reduced cost of maintfDance. As the holdings of the
cultivators in the Punjab are generally small, the question of equ it able distribution of
supplies is very important. In the colony areas, the educated cultivators clamour to have,
the discharge of their outlets correct even to the second place of decimal.
The work of remodelling channels requires a great experience of this type of work,
because a good deal of complaints and dissatisfaction starts when on~e the outlets are changed.
Really honest work alone can make a remodplJing successfui. A brief summary of the author's
experience of this type of work is givpn here and the stud en t shou Id study the followi!!g paper,
available in the proceedings of thll punjab Engineering ((lngress, Lahorc, for a detailed study.
(a) The Author's paper No. 154. 1931, (b) S. L. Malhotra's Paper No. 172, 1934 describin<g
remodelling by R. A. Routh, (c) A. W. M. Jesson's Paper No. 230, 1940.
2. Necc'sity 01 remodelling channels.
The channels having the following troubles need to be remodelled:-
1. Chronic shurtage at the tails
2. Widespread comlapints of individual outlP.ts.
3. Drowned bridges.
4. Excessive silting up and frequent silt clearances,
5. Excessive supply over the authorised full supply discharge at head necessary to
feed the tail.
6. poor command at head of a distri butary or at heads of its minors.
3. Bench Mar:{s and HydraulIc Survey.
When the necessity of remorlelrng has been deCided upnn, then just put reliable bench
marks along the channP.! to he renodelled, by douhle levelling, along closed circuits. The
masonry works are li:lble to be demolished, the bench marks should be fIxed on the R.D.
(Rf>duced Distance) marks or by constructing masonry blocks 1.25' X 1·25' X 1'5' neer the land
boundary limits at every mile with top one inch above ground level. An arrow head marked
in cement plaster should indicate the point actually o1:served in levelling,
A detailed hvdraul'c survey of the channel should then be carried out and the longitudinal
s~ction and the cross-sections p}.)tterl. Typical remorlelling. L. section and Cross ~edions, are
given in Plata XXII. D. II and Fig. 1 to 4 respectivel,!, Tre hydraulic survey should comprise
all the information shown therein,
When preparing a longitudinal section, the points to obsen'e a!'e :-
(a) The horizontal scale should te 2*=1 n,ile and Vf·rtic:lI !'cale 1/50.
(b) Last designed and existing f' II supply and ted Jeyels shC'uld be shown in different
colours on the paper section. (When a tracing is made, do not use blue colour, as it produces
very faint lines on the print especially on glazed prints.)
(c) Crest levels and working heads both existing and proposed, for all outlets and also
the w::tter levels in the water courses should be shown. The water level in a water course
should he that prevailing when a high field is being irrigatea.
Record plans of all masonry works should be checked at site, while carrying out
hydraulic survey to See that they are correct.
 383

~ 70"',
.
I
I T I
I
I
"•I I

i I I. I
.I
I I
I I
I
I

I :·"11' ...'" ,
In·,.,
I
I
I

I
I
•
I•
•I . -~
., I +.
I \
,
I
I• + t \

...I
+ ~'
tv., I
4;
Io.

t
t I
I
I
I
I
f
.¢'
..
t
.I
I

r (jl
;'"

~
I
I
I
I ~
!
r H
!
to,
t:' I'

~~
I... CJ'I
I'"
It
Ia.
~
<:)
r-
+I I!:
I~
I• I C I'"
I'"
~ C
I
I
I
4-
~~
NE'l
" I
I ..
i;~
I
• I
•., I
I

I
t-. , I 1 .... I
:~
u·,., I I I I
t.,.,. ,/
I
I
I
I
I
r~

I I
I

!-. r
/
I
I
I
\
I
I +
'"
c,'
I I
'- I .,
t

.. +N '::::,') ..
I
+i
I \
II \

I ,,
I
I
.
"\ I

"l I
..... '
+~,
T
,
I

••
.I
...L
I
I \
i.
·1 •
t'
f

,•
\
I•
j_
\ ~ t
I
II I n'

-'I,
I
J I I
I
t '\
,I
I
ILl' •• '
f'
I

o.I"'n,
I I

.-"
,I
~.~., :• I
.
t.

,
I
I 1
I
,
t
 384
4. History of a Channel.
The history of a channel should be read if one is available. If not available, one
should be prepared after reading the old files to give the following information. The reasons
for aU changes in the original channel design with respect to slopes, sections, outlets, ialls,
di;;:tribution miner head and distributary head regulator should be recorded as given, out by the
engineers then. The record of all silt clearanees should also be traced if available.
5. Diagnosis of the trouble.
In the existing. channels, the supply levels of the parent channel cannot be chal1ged
appreciable and the supply levels of tails are also fixed irom considerations of the command
of the area to be irrigated fhere is usually little scope of steepening the slopes. The channel
has to be run with the available slope. It can generally be run non-silting with the available
slope by controlling the silt conditions and the bt'd width depth ratio (Para 6 Chapter VI Part II).
A reference- to appendix 1 of Author's Papar No. !54 P. E. C. 1932, will show that the
channels do often suffer from excess slopes steeper th:tn the regime slopes. The excess sl~De
cause silting up of the channels by eroding· sides and widening, by meandering and shoalmg
aad above all by sweeping velocities killing. the silt carrying vertical and rolling eddies. The
causes of trouble should be investigdted under thd following su~·heads;-
la) Defective slopes (b) Wide Channel sections (cl Defectivij outlets (d) Defective Head
rlJgulator (e) Drowned bridges (f) Lack of control points and distributors.
6. Channel Seetions.
The channl'll sections should be within 20% of those obtained from L::Icey's formula
P .. 267 Ql '2,or Wood's bed-width depth ratio If channel is too wide, as is usually the case,
the remodelling should provide its contra~ti()n to the designed sections. Very wide channel
sBCtions are general! attained in course of time by cattle tress pass, and collapsing ef berILs
a.d pedastrian crossings. A relatively wide section silts up more than a narrower section on
account of the decrease of the vertical silt lifting and rolling eddies (para 15 C Chapter VI).
The contraction should be attained by hanging spurs if sCQuriug is also to be attained along
with the herm formation which is, a slow proce~s. rf it is desired to form
berms quickly and also not to disturb the equitable distri1ution of the supplies
to the outlets, the channel should be silt·cleard to the designed bed
tavels with bed width equal to designed bed width plus depth and- then to
construct longitudinal bushing. Both processes are explained in detail in the author's paper No.
154 and are giveR in paragraph No 9 Chapter Vll of Part rr. Grass should Qe sown on new
oorm formation. If bU'lhing be not available the contra~tion is attained by silt filling on sides
protected by gachi pitching (berm clods) or by compacetd earth protected by dry brick pitching.
7. Outlets.
The outlets should be, designed to take their due share of silt, to give the' corred
discharges based on correct coefficients and to work modularly under full supply conditions
in the parent channel. [n t he head reach of a: chimnel, A. P. M. outlets should be set as, low as
p()ssible 'f'ven '%D lower than the bed ) according to available working head, as permitted by the
reqnired M.M.H. Tn the middle reaches they should at least take there due share of silt which
is about H2 per cent. the author's A, P.M. with '80 setting permits this much silt conductive
power and in the tail reaches there should be open flumes set at .90 or bed level. The author's
modifications in approaches increases the silt conductive power of these outlets by about 6 to 8
pereent. .
8. Head Itegulator.
The -rrobable defects 'n head regulators of different types have already been given in Chapter
XlV Part II. Blind silt exclusion by constructing skimming platforms should be avoided. as it is
likely to give 1rouk.le in the parent channel. With the known C. V. R. in the parent channel
and that which can be permitted in the off-take with the available slope. the required
silt conductive power of the head can be calculated and a silt selective distribntary head
regttlator can be designed with this silt conductive power.
(Para~raph 8 Chapter XIV. Part II.)
9. Drowned bridges.
The drowned bridges cause unnecessary headirtg up and upset the regime. The heading up re-
sults in siting up of the reach upstream of this. They should be raised with 1'5 ft. clearance by R.C.
 385

Slab or T. Ream bridges in the distributaries and with 2 ft. clearance in the case of Branch
and Main Canals.
10. Chakbandi, Command and Draw off.
Having decided upon the causes of trouble, the next thing should be the preparation of correct
chakbandis and the area statements. Tne classification has been given and the preparation of the
area statements is explained in Chapter No. XVII Paragraph No.9. Similarly Command and Draw-
ott statements should Le prepared a$ explamed in pargraph No.9. Chapter No. XVII. Part 1[.
11. Lowering of channels.
No outlet should have a working head less than 1'0 ft. If the working heads are high or if there
is no irrigation from a reach, it should be lowered as much as' economically possible. From
consideration of seepage losses. it is desirable to keep the water levels in the channels as
low as pessible.
12. Control pOints and mete! s.
(a) There should te control points at about every 5 miles with a
minimum working head of 9". [he suitable design of control point is shown in Fig. 29.
The outlets should be groupl d to take off from upstream of the control points and
combined win in the form of distrIbutors. <\ 111f~ter flume at about R. D. 1,000 of the channel
should be provided so that th: supp y passing the distrbutary is correctly gauged, if correct
metering is not possible in the head design.
Two control points in a major distributary are absolutely essential, one about 2 miles
upstream of thq tai~, say wlJere the dIscharge is 15 to 20 cusecs, so that the silting and b(;rming
up of tail, reach does not spoil the regime ot the whole channel, and the other about 2 miles
downstream of the head regulator to localise the effect of any mistakp. in the design of the head
regillator or in selection of silt factor or C. V. R. in the beginning. These should be provided even
at a risk of slightly flattening the slope. The middle reaches shall then remain immune from
silt trouble even with relatively flat slopes.
(b) DL;tributors
There should be distrihutors at the heads of all minors in the form as given in Fig. 8.
aed Fig. 9.
13. Raising and Strengthening Banks.
Wh<ln raising and strengthening banks do not ::I 11 ow horrow-pit measurements, adopt
bank measurements. You will get much better value for your money. Below are given suggestions

or calculating bank widths etc. Bank width =D+ F + v/~- (take the result to the nearest foot)

where F=height of de~igllE(l bed above ~o. S.

D=designed full supply Depth: B=designed bed width.
Height of bank above full supplv lev61 up to 10 cusecs discharge=I'O' 11 to 25 cusecs
clischarge = 125'; 26 to 1Of) cusecs discbaTge=l 5'; 101 to 250 cusecs discharge = 1'75' ;
above 250 cusecs discharge = 2'0' Berm width at full supplv level up to 10 cusecs
discharge=2'+d/2' 11 to 25 cusecs dis:;harge=I'S'+d/2 ; :'6 to 50 cusecs c1ischar ge=2'0'+d/2;
51 to 100 cusecs discharge=2'S' +d l ~; 101 to ::!SO cusecs discharge =3'0' +dj2 ; above 25U
cusecs discharge =3 5/+d/~.
The formulae ~ivcn above provide strong banks and liberal berms. If funds are scarce a
lighter section may be adopted.
14. Remodelling operations.
In the case of small channel,:lll Works Tf'quired, such as silt clearances, remodelling
outlets, raising bridge-, constructing meter flumes and control points and COl traction of channel
sections should be done in one and the same closure. fhe stregthening and raising banks should
follow as soon as possible.
In the case of major distributories the reaches 3 to 4 miles, starting from the tail reach
sid~ should be tackled. All works required as mentioned above, should be finshed in the reach in
one and the same closure. It is preferable to take reaches as defined by control points at the ends.
Some engi neers are in favour of taking up remodelliug reaches in major distriburaries,
starting from the head. This may be possible only when the distributaries are not taking ~xcess
 386

a' haad and only outIes to and masonry works are to be remodelled to the modern designs
In the case of channels drawing excess at the head, the remodelling reaches must be selected
bl'ginning from the tail.
. All remodelliJllg of outlets must ne~cssarily be carried out in the beginning of a crop.
kharij or rabi, say in April or October so that the changes in outlets do not adversely affect the
crops when they have been sown to avoid discontent among the cultivators cr.d excessive
remissions.
] 5. Silt clearance and berm cutting of a Remodelled channel.
After a channel ba!' been contracted to the designed sections, the protruding bra:rtches
of bushing should te cut off and chanllf~l should be fina;ly cleared to the designed bed levels.
The tail reaches need annual berm cutting in the month of 5eptember generally, due
'0 berm growth throughout summer caused tv very fine silt in the wa.ter near tail reach which
fertilises tht berm growth especially lD the rainy season. Ihe belm growth heads up supply
with the consequent rise in the bed levels.
It may be that in the remodelling, the channels ClDnot be designed with regime
sections due to the existing fiat slopes in certain reaches. Such reaches should be given sections
as limited by the wetted pt::rim~ter discharge rtlation or hy the hed width ~epth ratio as far as
possible and should he declared that they shall nped silt clearances say every 2 or 3 years
as the experience will show. (Page ?56 'of this book).
The remOdelled channels with correct values of X. Sand p in Kennedy's designs and
with regime slopes, correct values of p,~ and silt factor in LacE'y's theory, with discharges over
about LO cusecs, do run nonsilting and no not need silt clearances However, this does not
mean that they could be neglected for all times to come. The damage caused to them by cattle
trt'sspass bv way ~f widening should be repaired by hushing the berimir,g up s~ould be
straIghtened by cuttmg and the bed interference by the cultivators by way of dajJs, bed Clods
and branches sticking in bed shoul.1 be removed by bed clearanre whenever the necessity arises
preferably before the month:> of keen demand in the Punjan (June and October).
16. Watching cnannels after remodelling.
The remooellen chann ... ls need to ba specially watched for at least a few years. The
following observations should be taken ;_
(a) H. Re~isters (Modularity Register)
The H. Registers are now maintained as normal routines glvmg observations by the
subordinates once a months of depth on crest of all outlets and the actual working head.
The cloumns of the register may ct' as per specimen below ;~
, !
I R.D·I Designed April May June July Similarly
I of Data for
No ,outlet! Type August
i and I !
---~--

to
I Side I ;B y H hm H hm H hm H hm H I hm March.
i I II
1
i 2
i
3 I· 4 5 6 7 I 8 9 10 11 12 13 14 15 16
I I I I
"
(olmn 9 shall show the fluctuations of the VI ater above the designeds water levels and
column No. 10 shall gi ve the indication of non-moduL r outlets where actual working hea4 is
les-.; than hm requirr;d for the modular working of tha outlet as got from plate X VIII (A.&.B)
column 8.
(b) SpecIal gauge slips.
Special gauge slip should be introduced for at least 3 years after remodelliBg
a channel giving upstream mettr gauge and downstream channel gauges at all
control points including the hp.ad of the dlstrihutary and the heads of the minors daily. All tail
gauges should be read daily. After the channel has stabilized. it will be necessary to have just
the head gauges of distributaries and minors and the tail gauges. The overseers, the zilladars
and the S.D.O. inspecting a tail ptrsonally should report the actual gauge reading. This serves
a useful check on the work of gauge readers.
Ic) "Characteristic curves," 01 distrIbutaries.
(i) There has been a progressive development in recent years in the m· thods of
 387

distribution of supplies form the remodelled Distributarie". With the universal :lpplication "
modules for outlets and Minor heads, the distribntion of supplie~ has now become automatics
and interference by the Gauge Readers has almsot ceased, The sy"tC'm has yet by no means
attained perfection and freqlltnts silting of channels introduce conditions, which are inmical
to the proper working of the nAwly evolved automatic distribution of remodelled channels
Moreover Distributaries are being required to run with different suppliRS in Summer and Winter
to meet the require nents of land reclamation, and of additional land given out for temporary
cultivation. All th.is senously upsets the distribut;on to various outlets and minors. Alttlough
the minor heads c:oulrl be adjusted every crop to make them take their authorised supply and
no more, it is neither practicable nor Hesirable to adjust outJets in th!' same way It WOUld,
therefore, happen that when pxcess supplies are run at the head. the intervening outlets wonlj
also take a certain share of this excess It is of utmost importance tn be able to calculate
readily, evtn though approximately, the excess discharge that would be needed at the head of
any channel to deliver specific suppl as at different points of a channel
(ii) To meet this requiYerrent. charactHistic curves have been evolved by R.B. Hakim
Rai, Superintending Engineer. These curves afforn a f'ady means of finding l'ln anSWf'r to variou,
questions. which otherwh~ would need lengthy cal ;ulations each time it is desired to carry out
such investigations. SOITJe of the problems for which a ready solution could be obtained by the
applicat.on of these curves are:-
(1) Excess di!Oharge requirerl at head to give say 1-0 cusecs at different points of a
Distri butary.
(2) The effect on tail discharge of lowering the Full Suppl Y' Level by silt clearance
in a certain reach of the Distributary.
(3) The effect of tatiling vf outlets in a paiticular reach on tail suppl:e<;.
(4) If any minor is over drawing the effect of redudng the supply in the Minnor
without reducing the supply downstream of it on the head discharge of the dis ·ributary.
There would be many other problems for which these cnrves WQul·i prove useful.
(iii) For the present the follwing characteristic curves have been drawn ;-
(1) a-Curve. This gives the excess discharge reqUired at e;J.ch point of toe channel to
give one cusec extra oischarge at the tail. This is the basic curve. as other curves are derived
from it.
(2) fi-Curve. This gives the exce3S discharges at head required to give out CllSPC extr a
discharge at any point of the chane!.
(3) 'Y-Curve. This gives the percentage of the extra discharge at any point that
would reach the tail
(4) A -Curve. This gives the increase in depth in 1 he variou;; reaches when excess
supply to give one cusec extra at tail is running. This also gives the e:iect of stlt clearance of
certain reach, s on the taj I supplies.
(5) ::;g -Curve. Below the A Curve is)given the excess withdra·val by all the outlets in
a reach when extra sllp:)iy tf) give One cusec at tail is running in the channel.
No~e : -For disci1arge other than that for whkh the curves are drawn. values can be
obtained by direct propflrtion (except in the case of 'Y curve which remains unchanged).
(iv) Method of plotting the characteristic curves;-
a Curve is called the basic curve, as calculations are made to determine values for
plotting this and other curves, are then derived from this. The tabular form used for this,
purposd is S1nwn in' tlble l. The channel is divided into small rear.hes beginning from the
tail uowar is The discharge at the taU is to be one cusec in excess of that
sancti lued This results in inCfj~ase in the depth of the channel, which is found by
using the formula Q=QD5/3; Percentage increase in depth can easily be .d~termined as it is
approximately 60% of the percentage increase in discharge. After h;lvir,1Z det~rmined the
increase in depth in the reach, the excess di!Ochargp. drawn by eac 1 outlet is worked out and
the total withdrawal plus the extr::1 supply reaching the tail plus the absorption lossps give the
proposed discharge at the tail end of the next reach. The same process is repealed till the
discharge originally required at the head. and that now detemined is worked out. For purpolles
of th6slt caiculations. it is a"sumed that all the Minor heads would not draw any share of
excess 2.<; it would be Dossible tn adjust them for the increased depth.
Figures tbu, obtained are then plotted as ordinates and reduced distances of the channel as
abcissa to get a curve. Calculations for elC~ Minor are also made separately from which and
 388

......
o
(7Q
~.
 3S9

with the help of these curves figures can be easily worked OLlt, if the min'lr h3ad.; are not adjusted.
The excess discharge drawn by each Minor per hissa rise in depth is worked out and the .. xcess
discharge required at thf'l head to give this much excess in the Minor is aloo determmed
as would be clear hom the exarr.ples worked out. ..
The figures for the f3 curves are obt::lined by dividmg the ordinate at the head as ppr
a curve by the ordinate at the haad of the reach. The points thllS obtained when plotted give
the f3 curve.
The figures of the y curves are obtained by aividing WO by the ordinate at the head
of the reach as per a curve.
The figures for A curves are obtained from column 6 of table No. 1 of this Chapter.
The ~ figures are obtained from column 14. These curves are shown for Maduana Distributary
of Lower Chenab Canal in fig. 5
(v)Examples of the use of curves.
(i) What extra discharge is needed at the hea i of Maduana Distributary to give I.S
cusecs at tail and 0'75 Cusecs at 350'00 if.
(al Outlets are not adjusted. and minors not allowed to take excess.
(bl Both outlets and minors are not adjusted.
(ii) If one cusec is saved as a result of silt clearance in the reach, head to 35,000 how
much of it whould reach the tail?
(a) From a curve excess required at head to give 1'0 at tail =2'4 Cs.
do do do at R. D. :~50()O=I'4 Cs:
:. Excess required at head to give 1 5 Cs. at tail = I'S x 2 4:.:3·6Cs,
do do do 0 75 at R. D. 35000=0'75 x 1'4= 1'0 Total 4'5 Cs.
(b) If Minors also draw excess, addition for this fo be made as follows:-
Rewazabad minor:-for 1'5 at tail, rise in depth at, head of minor=1'5xO'1=O'15 (sees
curve); From the curve for every hissa rise in depth @ head of Rewazabad minor, the excess
discharge neede.l at Maduana head = 1'2 Cs .', Excess discharge nef'ded at head if Rewazabad
minor is allowed to take excess='lSj'l X 1'2= l'SCs (A); Similarly for Karimpur minor,
Rise in depth='09; Exc~ss discharge needed at Maduana head = 09/0'1 X 1'5x'S9-'S
CS (R) and in th~ same way for ::iehtewala minor cxcess discharge required at head would be=
'051'IX I'Sx 1'9='14 cs
For supplying 0 75 Cs at R.D. 35000 eXCess required at head =0'75 X 1'19 (From a
curve', = 1'4 Cs (C)
This is p.quivalent to sending= 1'4/2'4=0'0 Cs at tail.
The excess discharge needed at head in c:onsequence of the two minors above 35'000 i.a
Qaimpur and Sehtewala. drawing excess is worked out below:-
For Karimpur='09j'l X 0'6X ·59=0·32 ............... (D).
For "ehtewala ='05/'1 xO'o . :: J·J=0·57 ............ (E)
Tota I excess required at the head on account of minor heads not being adjusted = (A
+(B)+(C) +lD) -T-(E)=l'S+ SO+I'4+'32+'57-4'89
lotal excess at head in the case of 1{b)=4·6+4·89=9·49Cs.
(III) If one cusee is saved in the reach, head to R.D. 35,000, 60% as &een from
curve, will reach the tail. Thus if Rewazabad Millor in the way does not draw a[;,y e.hess
discharge reachin§1 the tail=0'60 ;'<1=0'6
If the minor also draw excess, decuct for this as follows. 0'0 cusecs at tail would raise
supply at Minor head by '1 X 6 ='06 and the millor will get the discharge of '6 X '65='39 Cs,
and thus tail would get '6-'39=21Cs
 390

TABLE 1
Calculations for increase in discharge at he~d of Madduana Distributary
for an increase of one cusecs at tail.
._ , ..c
0 "0
v
...
..c:: I~ .... I.>
Size of ou tIet.
I,
• <'a
.; v
'"0..
Ig .....
Q.;
+0>
q,)
I-<
0 ..... = ._

\~
+0> (IS 0
q,) 0
I-< ..c:: ..c::
.... ,::C 0
..c::
..... 0.. .... 0 1-0 "0
c:
i:it 10. u- .0. 0.. v
"0
"0 Iq,) q,)

I~
a;>..d q,) q,) V
+0>
(IS 0 "1; "'C...; "CI .iii s:: CI)

s:: ttl ::s .8 ..d

'"s::
<l.l v e:! 0 .S !
'Cl
s::<l< r H B Y h
....
u Ig
+0>
.... .:a.....
~ b.O v s::

Ii
"0 (IS
.... a;>- "0"0
..c:: (!j~ (IS c: 0 I!.) I!.)
'"v A
:<j rlJ 01
(IS

I"::~
(1;
u ..c::u
u (IS ..c::
u .~"O '" '"
V '!i
<to c: u
~~
(.)
(IS
Q'" .v.
i:5'" OV
I-<
U
s:: ~ '"
(IS
IX
i:it (d
~
ill
':g I- 01 101 !~ Z U
~--- -~---- --~--~--~.-~---------~----

2 3 4 5 6 7 8 9 10 11 _~:J13 14 15 IH
63<!10Ti 477 1 , ,Excess! I
1'0 I I 5-77
" 63210/4'77 5'77 1-4 1'57 '17 61043L, 1'6 i '25 '45 ['15, '89: '94 '051 577
to I 57548R 1'6' '25
1
'3311'27 '68: '73 '05 3'29
55936L i 1'6! '4 '49;1'11151'1-62 'Il '33
4950°1 "58
I To tal i 3-08 329 '21 9'97
I
49500: 8 64 9'97 FALLl'lOl 'I 49465L \I-41 '32 '58 1'82,1'231-30 '07 19'97
to 1 ' 1'0 149425L ,Rewazab:ld minor_ 4'32.4-82 i 708
49000\ 49100R' 1'4 032 0'68' 072: I' ~R 1'46: '10 I '71
I, I To 'tal I \ti'I1' 7'U8 l--:W, 1776 I
49000! 16'24
to
!

i
17-76 1'90 i 20 '1 46982L, 2'0 i '9 , '55 1-45 1 94 2 '00
455, 42R 24 '4 I i
'43 1'97 1, 76 ,1'81 '051
1 1-'061--i
I 1
17'76
6'46
34625 1
40()77L 228 '32 I -51 1'77 I-59 \1'62 031 I '65
I I~S994R 12
7l.J '25[
/36!Z'35 1,ozI1'031 '01
1

(SO ON AND SO FORTH) Total 6'3) 6'46 'IS 24'S7

,
16. Examination QUestions, .. '.' '

1. An existing distributary of 100 cusecs capacity designed f~~ an ~Ilnual inten<ity of 80% and a crop
ratio of 1: 1 is to he remodelle<i to thp. same intenSity but to a smaller 'Kharif Rabi ratio 1,2 Work out the old
and new outlet al'owance and ~tate fuli.., wbatd"ta you would collect a;; basis for remodelling the channel and how
vou would determine the new slope to be given to it? (P I.B,(I936)
. 2, How will you determine the nece!!sity of remodelling a channel and how will you diagno<e the trouble?
3, In the remodelling of channel!. the sl"ptls are fixed according to the existing conditions and cannot
Qe changed. how will you determine whether a parti:lllar re"cn sh"ll need !lormal silt clearance or not? Why do
"\:e tail reaches need annual berm cutting and silt clearance in Octuber?
4, (a) In remodelling channels, how will you fix the SEtting ef the outiets?
Ib) What is the effe;t of drowned IiIridges on the regime of a channel? ,<
5. 'What observations and care are necessary after a channel has been remolletllld ?
 PART II
CANAL IRRIGATION
CHAPTER XVIII

1. Introduct.lon
.. Inundation Canals
The canals which usually flow only in summer, when the rivers are high, are called
'In~ndation Canals'. They have no permanent headworks and no weirs across the rivers.
Owmg to the changes in the river cou~se the off-takes. of the ca.nals have often to be changed and
fresh channels dug or creek~ cleared mIt. The supplies of tnese canals aTe not laways as desired.
In most of the districts in which inundoltion canals ex.ist, tlle rainfall is very scanty,
and the crops chiefly depend on canals, flver inundation 'lnd wel!~. In order that the water of
a. can~l may ~e m::tde to flow on to the surface of the country, the canal must generally have a
dIrectIOn maktng an acute angle with that of the river aUfl a general slope flatter than that of
the river. but when the country falls away from the river, a c::tnal may off-take at right angles
to the river. Sometimas a canal. after crnssing the river valley, runs nearly parallel to the river
along the slope of the watershed and irrigates land many feet higher than the flood level at a
point immediately opposite. The bed of a canal is below the surface of the ground, hut th.,
water level is generally above it, and the irrigation is by" flo N ". The land near the head of a
c~nal is generally too high to recei,:,e flow i.lri~ati Jll except in floods: It J?ay be irrigated. by lift
eIther from the canal or from the flver. Owmg to the smaller lift thIS IS much eaSIer than
irrigating form wells, and the water contains ;fertilizing' silt. Land irrigateri by canals yields
two crops, the KhariJ or summer crop and the Rabi or winter crop. The land for the
latter is soaked with water in August or September and afterwards ploughed. The.
Rabi crop is sown in November and December, and is generally matured by water railled from
wells The principal summer crops are rice, indigo, cotton and millet, but the largest of all crops
is the winter crop of wheat and barley. The inundation canals of the Punjab used to irrigate
about Ff4 millions of acres yearly They were more then fifty number. Nearly half of them
have been converted into perennial canals by the construction of the Sutlej Valley and the Haveli
Canals. The Dera Ghazi Khan and Muzaffargarh Divisions have inundation' canals even DOW.
There are private and Government inUladation canals in the Shahp,ur district. ... ."
2. General Description of Canals.
The bed level at the head is generally about the same as the winter sub-soil water level,
and thisis about the same as the low water level of the river. The depth of ~atftr in the canal
at any time is thus about the same as the height of the river above its low water level. Iri
most canals it averages about S feet, but it may sometimes be 10 to 12 feet A large canal
usually gives off branches or distributaries. These again give off water-courses which are.
maintained by the people and not by Government. The bed slope of a canal is se Idem steeper
then 1 in 4,000 or flatter then I in 10.000. The side-slopes generally become, by silting, about
half to one. The canals were in most cases dug originally by the c1lltivator without any
Engineering knowledge and are sinuous. The width are also somewhat irregular.
The banks of the canals are often somewhat weak arid liable to breach. In such ca5es
the banks are patrolled and watched. The strengthening of canal bamks. \\here liable to breaches.
is a work which should constantly be seen to,
A canal generally has a masonry flood regUlator, a few miles from the off-take, to
enable excessive supplies to be shut off. The ragulator cannot be placed near the off-take be-
cause this is often chang eo and there would also be the fear of its destruction by the river
erosion. At the flood reguhtor there is generally the off-take of an escape leading back to the
river and there is often also the off-take of a branch or distributary. .
If the canal crOS5es a flood embankment the flood regulator is at the point of the cross-
ling. Further down the canal there are other regulators generally at the off-takes of branches
IDI' distributaries, but sometimes at other places in order to head up the water during low supplies
 392

The Head Reach (often called, for shortness, the "Head" , while the actual point where
the canal begins is called the "off-take" )is the uppermost five miles or so of the canal, and this
often corresponds roughly with the reach upstream of the flood regulator.
The banks of the canals generally have trees and jungle growing on them. Fresh trees
are also planted in lines on the banks and slopes, bnt- not generally near the channel because
they are somewhat apt to fall into it.
A canal generally has a bridl~ road, and often a proper road fit for wheeled traffic, along
one or both banks. The road on one bank is genera' Iy called the Insl'ection Road an i i5 reserved
for the use of the Canal Officers The other bank is open to the public. Where watercourses
take off, rough bringes made of branches of tree3, grass and earth are constructed across them.
For the purpose of fixing the dates of opening the canals they are divided into three
classes. An "ordinary" canal is opened about the 6th May. An "eady" canal is one which serves
that tract of a country where certain crops whic!! require early watering. are grown in sufficient
quantities to renner an early open in!? necessary. Such a canal is opeiled about the 26th April.
A "Late" canal is one where there are no considerable areas requiring watn till about 15th May
or (sometimes) 1st June. All the ahove dates are averages and approximations. The actual
dates depend upon demand for water The height to which the river has risen, the tendency
of silt to dfposit in the canal and the probability of damage occurring from breaches. The
canals generally go dry a bout ::50th September and thus the period of flow is generally about 5
months.
3. Silting of C:lr..als.
The slope in the canal is usually flatter than that in the river. The velocity in a
canal being less than that in the river a deposit of silt generally takes place in the canal. fhe
silt consists of saRd and clay. The sand, with a little clay, is deposited in the head reach of
the canaL At the close of irrigating season. the deposLt at the off-take is gpnerally from 1 to 5
feet deep and it extends downstream. gT'adually decreas'ng in depth, for a di"tance whkh may be lfss
than a mile or may be 6 miles. The deposit is generally the greatest when the canal takes off
from the main stream, pspecially if erosion of the river edge upstream of the off· take of the
canal is going on, and If.ast when it takes off from a creek There may be some oepo!':its,
due to berming up, near the tail of the canal.
There is generally no deposit in thp mirldle portion. There is nearly always a silt
deposit at the insirie of a bfnd, but there is a corrfspontiing hollow at thp outSIde, so that the
gpneral level of bed is about the same as it is just above or bel;;w the bend. Large sums
of money have been wasted by removing the silt deposits which are known as 'side-silt'.
Their removal makes the cross section greater th'l.n elsewhere and they quickly form
again on bends.
A grea~ amount of silt is deposited in the water-courses. The silt when removed
from the canal is gpnerally placed equally on both sides behind the banks and levelled down to
the height of the banks.
The deposit of silt in the head reach of a canal is the greatest evil with which the
Canal Engineer has to dea1. Deposit in reaches further down the canal, if it occurs; IS not of
Sl) much consequence. In many cases a Su',siniary Head is provided :lnd kept clo"p.d. Deposit
in the head reach not only reciuces the supply below the demand, eXl"ept in high floods, but
it also cuts off the supph when the river finally falls and causes damage to, or failure of,
>:tal'lding crops. The subsidiary head is then opened. '
Erosion of banks may occur, especially in large canals and at hends. It is likely to be
very serious in sandy soil, and this is most cornu on in the l,ead reaches. The upst ann most
common method of ~topping it is "Bushing" The various methods of bmhing have already
been described in the case of perennial canals. Sometimes fascines and mattresses have
also been tried where river action is high.
4. Canal Head Reach.
(a)i The general lavout of the head reach of a canal is as shown in Fig. 1
The main problems of inundation canals is to secure adequate supplies for these canals
in the middle of April or very early in May, so that the cultivators can sow valuable crops
such as Indigo, sugarcane and cotton instead of cheap ones as rice, fodder crops. hajra and chari.
 393
If flood waoor breaks into a canal, it freque:tlly breaks out again on the opposite

Fig. 1

side and it may form a deep hole in the bed of the canal. This hole may flnlarge in all directions,
fOIming a pond whose diameter may be much gre:l.ter than the width of the canal. When the
banks an~ made up they are carried round the pond forming 'ring banks'. In course of time
the pond silts up and the banks can then be brought into proper linc.
H the damaged place is in higa eround, little harm may result, except the expense of
repairs, but if the gronnd is low, the canal supply will continu@ to escape and be wasted even
aftert he subsidence of the flood and the closure of the breach may be difficult. The rush and
distrubaance may also cal1se falling in of the banks for a considerable distance upstream and a
short distance downstream of the breach.
It the water merely breaks into the canal head, and does not breaks out again, the
heading up, at the point of afflux, is likely to be consideraple. It may be dP!'irahle to cut the
other bank if the ground is not low. It is nearly alway:> diffic1Ilt, and generally almost impossihle
to close bad breaches (i e. breaches not in high grount) in head reaches until the river finally
falls and money should never be spent in attempts b9 close them unless there is a real hope
of succe;5 awl also something appreciable to be gained by success.
(a) ii Anything which reduces the silt duposit in the head reach of a eanal is of the
utmost value, anrl a small change may make all the difference. Any sharp bends should be
removed and loops cut off so as to shorten the channel and increase the gradient. The width,
supposing the gradient and discharge) to be the same, should nnt be greater than that of the
next reach downstream of regulator It may with advantage be slightly less so as to give a
"draw". Anything which causes heading up of the water, such as the continued use of the flood
regulator or an influx of water into the head reach at a point downstream of the actual off-
take is a source of danger and may cause serious deposit of silt. The in-flowing water may be
simply flood water or escapage from another canal or flood water impounded in a "pocket"
For the reasons just given, a flood regulator should newr be built within, say three miles of
the off-take of a canal if the gradient is steep and six miles if it is flat. The actual site is
generally so fixed that the regulator can be combined with a bridge for a main road or so that
it may suit the off-take of a branch or distributary. It may be many miles from the canal
off· take. Unless it is so, it should not be used more than is absolutely necessary to reduce
supply and cause the escape to work. If there is a flood embankment which has to croSs the
canal it shoud be altered, if necessary, so as to bring the regulator site lower down.
(b) Escape in a head reach.
At a flood regulator there should always be an escape, and it should be large enough to
carryall the surplus water. Otherwise there will be heading up in the head reach of the canal.
The size of the escape con be calculated by taking the difference of the discharge of the reaches
ab and be (Fig. 2). Escapes, usually lead back. to the r.iver, .but .occasi~nally an escape
leads into the canal next below, In thIS cast} the JunctIOn wlth lower canal
is probably not far from off-take of the latter and the heading up caused
by the inflow is likely to be detrimental. Such an arrangment should not be permitted unless
 394
the escapllge is likely to be slight or unless the water can again be let out of the lower canal
ose to where it enters.

Fig. 2

An escape, of course, extends only as far as the nearest creek of the river. It should
be inclined downstream at, say, 45° to the general direction of the river. Sometimes, in order to
get a short line, it is run out at right anF!les. Such a channel is likely to hwe a poor gradient. It
is almost peculiarly liable to have its banks breached in floods. This may not matter, much while
the floods last, but the breac0.es have to be repaired afterwards. In the absence of an escape.
cuts are sometimes made in that bank of the head reach which is nearest the river, at points
where the ground is high and the escaping water finds its way to the river. Such cuts close of
themselves when the floods suhside.
(c) Changes in Bead reaches.
Owing to the changes which take place in the rivers. it is frequently necessary to dig
new heads. If the heads could be dug rapidly, and at any time of the years, matters would
be greatly simplified, but the excavation requires time and it can only be carned on in the
winter and the spring when sub-soil water level is low. Moreover, labour can only be obtained
in sufficient quantities in the winter and the early spring. As soon as one irrigation sea"on is
over, the heads for the next seaSOR must be arranged for. The old head may be retained or
a new one may be necessary. In arranging for heads regard must be had chiefly to the state
of the river as it is at the time. Some little idea may be formed of the changes which will
take place in the immediate future, but beyond that all is mere conjecture. A canal head,
whether new or old. may begin to work badly soon after being opened. When serious erosion
of the river edge takes place just upstream of a canal head the latter may become heavily silted
in a few weeks. or the creek supplying the head may silt up. or the stream may move away and
the watpr level be in consequence lowered. Subsidiary heads are also dug as shown in Fig. No.1
which are opened if the main head has silted up.
(d) Selection ot off-take.
TbP' foU-owing are general rules for off-takes :-
ta} A site where erosion is occuring, or is likely to occur, is a bad cme. Sites in the
main stream., or close to it He thus objectionable.
(h) A site in a small or silted creek is open to the risk of filting up of the creek.
(c) A site in a large creek is generally goods. The site should not be at a bend which
(d) to
seems liable erosion.
A site near the tail of a creek, but not so low down as to be close to the main
stream, is generall)' an excellent one. If the head of the creek silts up, the supply will be drawn
in from the tail.
(e) A back water, such as that marked in Fig. 1 is also a good site.
The principle of placing the off take as far upstream as possible is . followed when other
considerations do not interfere with it. This ensures a long feeding cbannel which is desirable.
A good site lor an off-take can generally be found by going far enough, but both funds and
labour are limited and practically the length of a new head seldom exceeds two or three miles.
A head (Haad reach) should not run for a long distance near the river edge. Such a head offers
facilitits for new off-takes in case they are needed, but it is open to the danger of being cut in
half if erosion occurs. If an off-take is carried too far upstream, it may result in the supply
becoming so great in floods as to be unmanageable.
 395
(e) Subsidiary heads.
(i) When a canal has two heads, the one first opened is called the Main Head and
the other the Subsidiary Head. The following points may be kept in view in deciding which
is to be the Main Head:-
(a) The main head should be the one which seems likely to give the better imm~diate
supply. It is not much me having a good supply for maturing crops if thete is not a good
supply for sowing them.
(b) The main head should be the longer of the two, because there is then less chance
of the silt extending down below the junction and so ob3tructing the flow of the subsidiary head
when opened.
(c) The subsidiary head should have good banks so that, being closed by dams at
both ends, it may be completely' boxed up," and no flood water allowed to break into it. A
channel to which flood water has a access is quite unsuitable as a sllbsidiary head.
Generally. the longest head, i e. the head whose off-take is farthest upstream, will
give the best supply unJess erosion seems likely to occur at the off-take.
The main head usually giv!1s sufficient supply for the greater part of the season. At
first, the deposit of silt in the canal or in the feeding creek is small. By the time it has
become considerable, the river has risen high and the silt has little effect.
It is when the hig\:l floods are over and the river is falling, that the silt begins to have a serious
effect, and this is the time for opening the subsid;ary head. A Slliable date generally for the
Indus is, about 10th or 12th of August, and for the other rivers a)out the ~Oth of August.
Premature opening of a subsidiary head may be disastrous, causing e«cessive supplies or silting
one of the heans if it has a flat surface slope.
(ii) When the subsidiary head is opened it is sometimes de"ireable tQ close the main
head. It may happen, however, that both heads bring in appreciable supplies (this is likely
to occur if both heads are long ones or even if the upper one alone is a long one, and is the
more silted of the two) and both should remain open. No doubt some additional silting is
caused. In one head or the other owing to both being open, but there is also an additional supply,
The question whether the additional supply is worth the additional silting (the C(Jst and difficulty
of closing a flowing stream being also considered) is one of judgement. The quetion whether it
is desire able to close a head, is also affected by the information received from station higher
up the river. Possibly the river may be falling so fast that the head will soon close of itself.
(iii) Special cases may now and then arise and give rise to special measures. For
instance, if erosion occurs near the off-take of the main hea.d and it begins to silt up, and
the erosion seems likely to go on, it may be desireable to close the Main Head, even quite
early in the season and to open the subsidiary head It may even: happen that what was
intended to be the main head is not opened and the subsidiary hea<'l is opened instead.
Sometimes if there is spill water sufficient, cuts may be made in the banks to feed the canal
thereby.
In large number of cases the subsidiary head is not used. Either the main head
continues to give a good supply all the season or erosion sets in at the subsidiary head or its
feeding creek fails or the ri ver recedes from it.
5. Bunds (Dams) for callal heads in winter.
The canals having gone dry in September or Octorber, earthen d~ms are constructed at
th.e off-takes in November. The object is to prevent the rainy river water from accidentally
entering the canals, before they are ready for opening. Any such accident may cause immense
trouble. Contractor's earthwork may be unfinished or unmeasured or mat~rials' for masonry
works nay be lying about. The danger may occur during winter freshets or in ApriL or May
when the river rises permanently.
In order to allow for a slight falling in of the river edge, the dam in the ma,in head
must be set back, say 20 to 50 feet, from the actual off-take. To prevent this space from
silting up when the river risesl a minor dam is necessary at the actual off-take or as near it as
possible. Setting bank the main dam several chains is most objectionahle. The whole lengt4';,
may silt up if the minor dam is carried away. , ,
The dams must be parallel to the river edge. If made square to the canal Wq.~ the :-
off-take is skew, a space is left for the collection of fine tenacious silt. Such silt has at 'tilnes ,
 396
absolutely prevented a subsidiary head being opened aBd has interfered with the opening of a
main head.
The dimeosions of the dam may be :-
Top width Height above water level
on 15th N 0vember
Main Dam 10 ft. 8 ft.
Minor Dam M ft. 5 ft.
If the off-take is from a dry creek, the height can be mea<;ured from its bed. The
material should be got, if possible. from the bed of the canal Sand does very wen. The side
slop"s of the dams should be protected by fascining Long twigs are made into bundles .and
tied up so as to form fascines. and these are laid on the slopes and secured by pegs driven mto
the slopes. at short intervals, between the fascines. .
If a creek is to be cleared, the dam will be ~t the off-take of the creek (from the flver
or from a~l.Other creek-which is n<:.t to be cleared) and not at that of the canal from the creek,
unless the dam at th~ creek off-take is considered to be insufficient
If there is any creek or channel or depression which is not to be cleared bu! by yv~ich
water can enter the canal or can enter a creak which is to be clund it must be closed In a SImIlar
manner, but the material need not be taken from its bed and there need not be a minor dam. On
the top of the smaU dam a grass hut should be made and a watch-man should live in it. For a
subsidiary head the dam should, unless erosion seem likely, be close to the off-take with 15 feet
top width, height 3 fept above H. F. L. and sides protected by fascining. If prosion seems likely,
the dam may be silghtly set back, and if erosion actually sets in a second dam (No.2) should
be constructed. say 200 feet downstream.
A dam can be built on the top of sandy deposi t, but it should never be built on the
top of tenacious silt or on grass or dibh growth. The silt should be dug out to bed Ipvel and
then the dam made.
In opening a canal it is impossible to remove the whole dam. Some of it must
necessarily be left and be allowed to be swept away by the water. If time and labour are
short, the whole dam may be left to be swept away a mere cut being made in it. It is better
to do this than to delay the opening of the canal. The material in the dam when spread over
a great length of canal, causes no appreciable raising of the bed.
6. Miscellanous Works.
(a) Bank erosioll,
Erosion of the banks may oc:ur espeiaUy in large cal'l.als and at bends. It is likely to
be very serious in sandy soil, and this is most common in the head reaches The btst and the
most common method of stopping it is "Bushing." The various methods of bushing have
already bpen rlescribbed in the case )f perennial canals. Sometimes fascines and mattresses
have also been tried where river action is high.
(b) Bridges and water course crossings.
In the tracts served by Inundation Canals, roads fit for wheeled traffic :Ire somewhat
scarce. Masonry bridges are somewhat few. Many bridges suitable for c;une!s, horses and foot
passengers are made of rough tree trunk;, placed on end to act as piers, with rough branches
1aid across them and cnvered with coarse grass are earth. Such bridges require repair, or
reconstruction yearly and not infrequenty portions of them fall iato the channels. They are
gradually being replaced by masonry structures. Bride-es of sawn tim')er are cheap, but do
not last. There are very numerous ghats or places where the banks are sloped off so that animals
or men can get water from the stream or wade across it,
~ . • . . In \vinter the people put up numerom temporary acqueducts, rough wooden troughs
. supportEid on stakes, in order to rarry Nater raised from wells on one side of a canal to fields on
tre other,side. The canal banks, being generally above the level of the fields, have to be cut
.through at such places. Before the canal are openen in the spring, the people remove the
acquedeucts and repair the banks. This procedure is troublesome and objectionable, and is only
permitted at places fixed by usage. 1£ a crossing for water becomes necessary at a new place'
a maso.nry syphon has to be constructed or a new well sunk, the cost being borne by Govern-
ment, If a new Government channel has severed the connection bet'veen well and field, otherwise
by the people concerned.
(c) Drainage Crossings.
, In the Dera Ghazi Khan district there are numerous streams which issue from th~
Suleman mountains (these lie twenty or thir ty miles to the west of the Indus and flow towards
the Indus). The channels are generally dry, but after rain streams come down and break into
the canals at numerous points, generally toward~ the tails or into the western branches and
break out again on the opposite bank. Much damage is done to the banks and some silt is
brollght into the channels, but the water brought in is, on the whole, useful in supplementing
the regular supplies which are not generally very good tow~rds the tails of the canals or in the
western branches, as these have necessarily somewhat flat gradients owing to the land rising
towards the mountains.,
(d) Duty of Inundation Canals.
The "Duty" of water on inudation canal is low, being generally about 70 acres per
cubic foot of discharge per second, (measured at the canal head) as against 200 or 300 acres on
perennial canals. The duty is low because the canals flow only for about five months of the
year, because the soil is sanciy porous and because the rainfall is very light. Long water courses
also, to some extent, cause low duty.
In matters sucn as the observations of discharges, the keeping up of Irrigation
Registers, tbe Distribution of the Supply, the fixing of the sizes and sites of outlets for water
courses, the remodelling of water' courses and the maintenance and extension of plantations,
the principle and practice are the same as for perennial canals, except that trees are not planted
where they are likely to be cut away by the river.
The chainage of an Inundation Canal usually starts from the flood regulator and runs
in both directions. The mile posts, etc" in the upstream reach should be of a kind which can be
shifted. The same remark applies for the whole canal until all necessary straightenings have
been effected.
(a) River Behaviour.
(7) P. Claxton produced on this subject paper No. 119 the Punjab Engineering
Congress 1938. His practice is shown in Figs. 3 & 4. In this paper the writer deals with the
Iaundation Cl\nal practice in all its bearings. He t~kes as an example the whole of the
Dera Ghazi Khan Division in which he had served for many years, from 1910, off and on,
till 1927. He has been able to watch problems as they has arisen and developed.
"It is well known that alluvail rivers, be'lring large loads of silt, carry them on, not
in one sustained effort, but by a series of jumps, or by saltation. The loads have not travelled
down continuously from the hills, but are removed along by the processes of SCOl1r and erosion.
Of these tWI), erosion, the action at banks, in contradiction to scour, that is over the ceci, is
the more important. It gives the heavy excess loads at various points and causes the river
to meand.. r. In doing so it presents the problems with which the inundation canal practice
has to deal".
One of the first principles lays down that silt eroded from a bank very soon deposits
as a shoal on that side. Many engineers argue that it may be, and is, carried over to the other
side. The argument is based on experiments on very narrow streams in which the whole surface
slope is affected; This brings about a cross current which has the power of laterally deflecting
particles. This is due to the surface slope and in rivers is confined b the bank. The silt for
the remaining width is entirely controlled by the strong forward current, across whicn nothing
may pass. The writer does not appeal to laboratory experiments for his proofs, but has shown
by a few examples in other papers that the principle is supported by large dver movements,
sometimes involving square miles of shoal. These shoals appear and melt away in step with
erosions above them, and have the power of deflecting even the main stream. They all obey
the same law which the writer has enunciated, He has, therefore, considered it fully established.
The theory is illustrated in Fig. 3 which is believed to be drawn strictly to principle.
 398
. In this diagram there is erosion from A to B on the right side. At A the stream comes
Into the bank, ba.ving deposited its load of silt on that side in the shoal above. The water A
e. "LgM £MIIAH!!I1I!IIT CI r;: .

Fig 3
is, therefore, clear. From A te B erosion is active. and the loan of silt is consequent Iv increas-
ing. At B it has grown so great that the river is no longer able to carry it forward, and a
shoal begins to form, 'diverting the stream a.cross to C. From C the process. as at A, is repeated,
only this time on the left side. The erosion from A to B is responsible for t1e shoal BCDE
and the erosion from C to D for the next shoal on the left.
It will now be shown how, on this theory the favourable points for inundation canal
heads are d.eter-mined. The favourable n()n-silting points are these just above A,e ann E. The
canal off-takes may be placed a little higher up in the creek which forms the out fall above
one of these points. Creeks are favoured since they are usually more constant than
the main stream. being subject to milder attacks. They al"o allow of the building of groynes
or dams and thereby the off-take may be carried higher up towards the head of the creek itself.
Inundation canals invariably work upstream for better comm::md and such canalizing of creeks
is very useful,
(b) River Control.
In controlling rivers for inundation canals, there being no stone available, u~e of the
still-water pock~t principle must be adopted to the greatest extent. The principle is illustrated
at every Bell's Bund built for protection and training at important works of the province. It
need not, therefore, be described here. Water to be held by earth embankments mrlst be
brought to a complete stand still, and this is effectf'd by the iormatinn 01 the stilI-water-
pocket. Such a plcket is formed when a creek is completely closed by a dam. ThiQ, however,
is not always, possible, and a compromise is got hy means of a groyne. The groyne is a dam
which has the canal bank or the high main land for one flank while the other flank Is in the
o~n river bed. Whenever possible. the nose of the floating fl:lnk is placed on an island, or is
carried up to a high shoal. across the main cr€'ek, w()rking upstream at the same time so as to
create a pocket. Some times it cannot be carrieci across the stream but is finished off to from ::t
pocket on onC side of it. T.he nose is protected by stakes and Prush-wood, and spurs may he
added to throw off longitudinal flow along the groyne. These groynes are not only helpfnl in
securiug the early supplies. bl1t are strong factors in keeping the canal heads clear of silt. With
experience they may often be maintained even through the floods. Whether or not throughout
the season, they form undoubtedly a feature of inundation canals which should be recognized.
A head with some sort of groyne may nearly always ce fDund and by it much silt clearance
may be avoided. Another very important advantage is the control of the river which groynes
afford. Developments of down side creeks are prevented, and the main stream is kept away.
Tbese are some of the leading features of control.
Ic) Policy Combining river behaViour and control.
The usual form of river is shown by the meandering course A, B, C, D, E, G, within
the normal flood embankments which, of course, will not be so regular, out will be more Of
less comtant, marking the limits within which the river may wander. Beyond these limits
the flood emLankment and main lines of canal should be bnilt as at a e and c g. One of the
features of the policy was the combination of the canal and flood embankment by 'iligning the
 399
canal along t~e borrowpits of the embankment on the river sirle, Thus the cost would be that of
the emban~m"11t only. lhe,canal wo~ld have on river bank but will spill over the whole river
front, and Itself, would remam clear, slnc~ It wonld be the deepest channel which would form
naturally along ,the emb~nkfl,en~. eve!! wIthout the help of borrowpits, On the river eage it
wou,ld have a h.lgh marg1n of sI?lll which would keep it from returning to the river. The cross
sectIOn,at the t11;'1e of constr~ctlOn would be as ~hown by the diagram Fig. 4 and in time with
succeSSIve deposIts of sIlt, th1S would tend to grow into that shown by the dotted line.
~" This trunk canal would run the waole
f
:.J t :,
~
lengt~ of the district, more or less constant in
capacIty. It would feod other branches off taking
f .:::...,. at the points A, E, C, G, etc. rhe trnnk canal
, ;t~ ~ " itself would he ltd by Inks, run in from each
.., ----." of the favourable points A. C. E. as describeri
,. above. They would be at angle so as to form
pockets with the trunk canal, at the junction
w:th which regulators and escapes would be built.
These points would also be the heads of the
Fig. 4 branches. Each link would be under entire
control at the junctions A, E. C & G, a regulator
into the next compartment, and a head of a branch canal, being built at each point. By means
of any of these regulators, the flow in the compartment above a link would be brought under
entire control, even to the extent of causing the link to work back to the river by heading
up. This control would regulate silt deposits in the upper compartment to some extent.
These compartments bv arresting silt, would grow veluable for Rabi Irrigation, while behind
the embankment, kharif crops would be sown, The system would be unique in making all
the heads on the river available for service. The trunk canal also would be impregnable, i. e. it
would be impossible for ri Ilef erosion to sever it, Such action frequently ruins an inundation
canal. Hear, however, should erosion threaten to eat into the canal, it would also threaten
the embankment; and as this must at all times be maintained for the preservation of the district,
it would have to be built further back as a loop, and th6 canal would automatioally follow it.
Not only would existing favourable point::; of the river be made available, but short cuts to
others, might at any time be made along with other links, The canal, as a trunk canal, could
make up shortage at one off-take wi eh excess at another, thus makivg the supply
mQre ewn. The large com;:H.rtments w·.)uld also act as dampers ani!. would neutralize temporary
shortage. Thus, this plan provides for every difficulty with whien an inundation canal is
faced. It aims at affording a steady, even ann assured summer supply.
It may) be noted that one of the objects of a trunk canal, such as that decribed
above, is to olDtain, jf possible, not only 8. summer supply, but also a smail winter one. Tbis
is" often possible even with the present system of canals which are not in a position to tap all
the availaMe hads. By having a. continuous trunk canal, all the availa~le heads will be
greatly enhanced.
8. Examination Questions.
1. Define Inundation Canals. Describe the function. location and the working of subsidiary heads in
Inundation Canals.
2. Describe briefiy the Layout and the points you will keep in view fOr determininig the suitable
main head for the head reach of an inundation canal.
3, Why are escapes neessary at the end of a head reach of a canal? \Vhy canno they be done away
with by heading up supply at the regulator to top excess supply?
4. Describe Claxton's systm of aligning head reach of an Inundetion Canal.
 PART II
CANAL IRRIGATION
Chapter XIX
Discharge Observations and Regulation
1. IntHdl1ctIon.
Discharge denotes the volume of water passing a section of a stream in a unit of
time, It is measured in cubic feet per scond, commonly cated Cllsecs in India.
For successful irrigation operations the correct knowledge of the discharge in the
rivers feeding the canals is very eSSf'ntial. The supplies available will be distributed into
the canals taking off from the river, The discharge sites along the great rivers of the Punjab
are fi xed, where daily discharges are taken.
Similarly carred discharge observations arC necessary along the canals and tranc~..
to ensure equitable distribution of supplies in the distributing cbannels called distributaries
and minors,
The methods commonly adopted for the river and the canal discharge observations
are described in this cbapter.
2. Objects and applications of discharge observations.
The objects of discharge observations and uses to which the data 50 called are
pllt, are as below:-
(i) Stati:stical. By collecting daily discharge data for a sufficient numbu of years
and scrutinizing it,it is possi91e to forecast the normal supplies for basing the irrigation
projects on and also the highest supplies for fixing the magnitude of works and storage projects.
Accuracy in observation of discharges at all stages is therefore needed. The Sind-Punjab dispute
shows the importance of collecting discharge data for inter-provincial distribution of river
supptes. The importa(lce of collecting, in the first instance, correct data for highest discharges
is illustraed by the subsequent adrlitions at a comp:lfatively higher cost required at the
Panjnad Headworks and Kalabagh Railway Bridge,
, (ii) RegUlation and. distribution. Accurate measurement of river supplies for distrbuti(ln
among partners of an irrigaion s~ stem and of canals, branches distributaries and outlets fur
equitable distribution' is an essential feature of the sucessful running of callal system.
(iii) Scientific and Hydraulic. investigations. For this. accurate observations of discharges
both in the fied and laboratory are necessary, Of the numerous invesigations in connections
with the advance of hydraulic and irrigation engineering some are' estimation of absorption
losses in channels, determination of co-efficients of roughness or various types of channels both
lined and unlined, sluice gate co-efikient studies of the conditions of flow in streams and
,investigations of hydraulic laws.
·3 (a) River Gaaging Site in Plains.
The site chosen must so far as possible comply with the following specifications:-
(i) It must not be located where the river is too wide and shallow, nor too tight and
~eep. In the former case it results in shoals and slack water, while in the latter during high
floods it may result in excessive and dangerous velocities.
(ii) Widths may vary from a few hundred feet or even'less during the Winter season,
to several thousand fe~t in the SUmmfT months, but the site is to be such that at all times the
section is reasonably uniform. A site where the sectbn is very deep on one side and very shallow
on the other, or shows some such deparature from normality at .different times of the year must
be avoided.
 401

(iiiJ The discharge section line as for as possible is to be located in a straight reach
of the river.
[ivJ It is preferable to locate thp. site for obviom; reasons where ,the rive,r is flowing
in one channel. But if this cannot be avoided, it means that the multIple sectlon line will
require observations to be taken on each arm of the river in flow, On the other hand, a very
large rivpr may more conveniently be treated at more than one arm of flow when in all probabi-
Ii ty each arm would b') treaterl. as an independent si te wi th its own seperate a tlservational
staff and equipment. .
[v] The section line must be so orientated as to be as nearly as possible at a ngh t
angle to the direction of flow.
[VI] A site mu,t never be located upstream of a confluence and sufficiently near to it,
as to be affected by heading up at the discharge site due to increased flow in the channel into
which the flow through the disch:uge site falls resulting in higher and false levels. Similarly,
a discharge site should never be located above a Weir or Barrage, within the effect of the
Pond formed there by the heading up of the river.
[b] River Gauging Site in Hills. 1 ~. "

In this case the same arguments apply as for phins gauging sites, but generally speak-
ing, silts in the hills are more difficult to select ann all the desirahle features cannot be found;
but it is es,ential n~verthele ss, to select a site which is:-
fi] Free from proj ections from ei ther the bed or, the side~.
[iil The site should not be too close to any bridge or fall, Any incoming torrent
upstream or dONnstream of the discharge section should be as for away as possible and not
nearer than 600 feet,
[iii] Straight for at least 500 feet, above the section line.
[iv] Accessible ~n all se~sons.
[c} Artificial channels(Canals).
Such sites gonerally are easy of selection, but nevertheless must be :-
[i] In a straight re'lch of length not less than 10 times the mean width of the channel.
[Ii] Not in close proximity to any fall, bridge or work of any kino likely to cause
obstruction to smooth flow anrl cosequent eddy action.
[iii] If the channel is not pitched, the existing channel must be kept to a clean
section f~ep fom fallen berm~. For a perm'lnent discharge SIte, Hie side slopes and preferably the
bed also should be pitched over a length equal to at least 200 fcet.
4. Permanent and temporary gauges.
[a] permanent ganges may take the form of:-
[i] Graduations directly engravpd into rock or masonry.
Ii] F. 1. gau;es fixed to an angle iron, old rail, or masonry-pillar and g'ven
secu:e foundations.
LiiiJ E. I. Gauges fixed to masonry structure such as Railway Bridge, Piers etc,
rt is most important that the zero R. L. of a permanent gauge. when once-fixed
is on no account changed,
[b] A temporary gauge does not reqt1ifl~ such secure foundations and may be used as
an extention gauge to a permanent gauge which has been temporaril~ left high and dry or in
connection with a disch'lrge site the section line of which is not in itself a permanent fixture.
When temporary gauges are required to be fixed not les,; than 500 feet upstream and 500 feet
downstream of a discharge site section line. for the purpose of recording water surface slope
a.t that discharge section line, they, must be graduated to read to 0'01 foot in order that surface
slope and ultimately Kutter's constant may be sufficiently accurately determined,
[c] All gauges, whether perman~nt or tempora y. mmt have their zeros fixed relative
to a permanent bench mark which, if not existing near enough to the gauge or gauges to
be set from it, must separately be located and provided.
S. Methods of stream gauging.
The measurement of discharge is made by thc following methods: ~

/
 402
[iJ Velocity area method [ii] Gauge discharge curve method [discharge tableJ
[iii] Weir method [Meters chapter X.]
The above three methods are the most {'onmonly employed for discharge observations
and s~all be described in detail hereafter. In addition, the following methods are also used,
thougn they ;.vill, intp.rest engineers only :-
. [ivJ Pilot tubes. This method is based on the the principle that the dynamic or
Impa~t pressure of a current moving with velocity v, and striking a r,lass tube with a nozzle,
pomtIng upstream, makes the water rise inside the tube by an amount 'h' above the sllrface
and that the velocity is represented by the equation v=Gy 2gh. The value of the co-efficient
G can be determined by calibration of the tube.
[v] By chemical means. This method is based en the principle that if a w~ight w 10
of a chemical is added each 5e~ond to a stream discharging Q cusecs and after a thoraugh

mixture if a lb of the chemical i:: contained in nIbs of water thp.n _W_ = _1 orQ=--~.
6~'5 Q n 62'5
The J?etbod reqaires special apparatus and arrangements and is not easily applicable under
conditions prevailing in the field.
[vi]. By venturime!er. This is a?plicable only to discharges c::lpable of being passed
through a pipe and its theory is described in all te· t books on hydraulics.
. [vii] Bya travelling screen. In this method a light varnishp.d canvas screp.n fixed
J~ a. rigid frame is hung from a wheel-ed carriage and is made ~o move with the water in a
spe.mlly prepared rectangular channel. The velocity with which it moves. after being corrected
for leakages from sides is takpn as the mean velocity of the stream.
6. Velocity area method.
This method involves measurements of area and velocity.- It being impossible to
detfrmine ~he mean velocity for the entire section of a channel in one single measurement,
the area IS divided into vertical strips of suitable size and mean velocity for each
compo~ent strip is observed, the total of the discharges for all strips bein.g
the dIscharge of the channel. In other words the cross section of a river IS
observed by soundings and the mean local velocity perpendicular to the cross-section is
observed at as many points as::possible. The discharges at tl:.ese points are denoted by the
p~oduct of the mean velocity and depth. If now a curve is plotted with the water surface
wldt~ as the base and the discharge at each point represented by product of depth ~nd
veluclty as the ordinate. The total discharge is represented by the area of the enclosmg
polygon. The area of the polygon may be determined eitoher by taking straight averages or
by Weddles and ~impson's Rules.
raJ Measurement of area. The area of cross section of a stream involves twe elements
viz, the horizontal width and vertical depth which are both measured as below :-
[iJ Width. If thp. discharge site is located at a bridge or other masonry work, the
measurement of width is a simple affair. In the case of an open channel of width upto
about 1,000' a wire is stretched accross the channel and segments are marked on it tw
means of pendants. At Rupar on the Sutlej a weir has been stretched at a width of 1,700
fe:t of the river. The necessary precaution in such cases is to make a suitable allowance
for the sag. Alternatively, the pendants snould be fixed by a theodolite. In bigger channels
accross which wires cannot be stretched, the segments are marked by pivot-pomt method
which is baised on the principle of similar triangles to be described hereafter.
[iiJ Depth. The points where depths have to be measured are located as above and
in deep and fast stream in which wading is not possible, boats have to te employed for reaching
such sounding poin ts. Depth is measured either by direct reading on a sounding rod or log
line or by indirect methods.
raj Direct Method. (i) A sounding rod consists of an oval wooden rod with E. I. gauge
on it, flat iron of 2"x t" size or bamboos of 2" to 3" diameter all graduateo in tenthS of feet
and with iron .discs of 4" to 6" diameter at bottom to prevent sinking in bed. Bamboo rods
have been used even up to 30 feet depths in low velocities. The depths are measurer! at
downstream ends thus omitting effects of afflux due to velocity.
 403

(ii) Lead Line. In higher velocities and depths, observations are made with a weighted
log line. Log lines consist generally of copper cores covered over with hemp. Such a line does
not shrink"" hen. wet. nor stretches u~der weight and remains free from knots. Alternatively,
when such metallIC lInES are not avaIlable as IS the case at present due to war, weighted
manilla rope can be used, tut with due precautions regarding stretching and wetting it before
lIse. Constant checking during use is essential.
The lead or sounding weight is generally of the shape of frustum of a cone. The weigth
varies from 10 lbs to 56 lbs depending upon depth ann velocity encountered. At Kalabagh even
a weight of 56 lhs has to be used with care and experience to measure depths accurately.
It requires an experienced hand to observe depths properly with a logline. The depth of
water surface below a reference point is first measured and marked on the line. The weight is then
swung and released a little upstream of the observer. After tCJuching the bed it trails down
and the rope or line is pulled, till the weight is vertically under the observer when the
length of the rope is marked against the reference point. The dist:tnce hetween the two
points gives the depth at the point A dif ere:1ce of 6" between two consecutive readings is ignored.
for large differences observations have to be repeated
(b) Indirect method. In deep and rapid rivers the observations of depth with log
Haigh's Depth Meter. Scale 3/8"=1"

C(1a STIMW it/Nil

~~V~~~~-;--I--- 8'10~£0 TO
COIL "NO CON7'AlNq

;;:0"'1 HOLES
1." fJ·S1'AC.'£D
I a JIIouNQ CAse
/
Fig. I
 404
line requires skill, experience and prf'EeVerancc and thp accurar)' of results can to some extent
be questioned. The problem of measuring depths with ease, rapidity ::Ind accuracy has not yet
been satisfacturily solved and attempts are being made to apply some indirect type of sounders
for the purpose of which the following ale worth mentioning.
Ii) Haigh Depth meter. This is an ingeniouc, device recen tly invented by F. F.
Haigh C.LE., Chief Engineer, Punj'lb Irrigation and is based on. Boyle's law viz.
PV is constant and thp, observations so far m'loe give remakably consistent rcsults.
It con~ists of a tube coil in a cylinder containel', in which watfr under pressure is entrapped
by compressing the volume of air in it. The volume of water entrapped is proportional to the
depth upto which the meter is immersed :lnd the dppth is read from a calibration curve
Fig. 2 prepared by actnal obf,ervations tefore it is brought into use. The details of the meter are
as shown in Fig. 1. The ir1stru·nen~ b~sid ,s b~in~ U3ed i.\ rlischarg \ o·)servatio 1~ wJl be oi immense
value in observing scour depths at salient points of Headworks, training works and railway
bridgf's.
(ii) Kalvin tUbes. These have been tried in Sind lind have not
proved to give consistent and accurate results. . .c.:
(iii) Echoe Sound9r. Sugge;tion has beer} mvie by Montagll, to explore the possibility
of using sllr:h sounders as are used on ships hr ob3erving ciepths. This needs further
investigation.
<hI Measurements o~ velrl~iiy The flow being furLulent in opon channels, three to
Calibration Curve for Depth \1"ter four observations of vel,xity should give
J' the correct average velocity at the points
til of section of a channel.
(J
.S Even though the flow in open channels
v i l d lU0US (turbulent. and subject to
1-III
I-'
continuous pulsations, the distribution of
velocities in a cross-section has sorre
> general characteristics.. Attempts at
, V
/' determining location and magnitude of
one single value for mean velOCity for the
III "" entire cross-: ection of a channel have so
....
~
I'M far not been successful and therefore, for
...
c:
80 1/
rlischarge measurement mean velocities
for component strips of area of cross-
lOa
sedioTls ale observed. The vertical
velf'city curve showing velocities at
44 I various depths is accepted to be a parabola
., with a horizontal axis. Actual observations
by various investigators in America and
'"-. 4 India have shown that the p')sition of
IZ. to I!.... a- t I1ID .4() ... ... mean Vflocity in a vertical occurs, under
normal conditions of flow, at 6/IOD. It
Depth in feet Fig. '2 has been proved mathematically also by
Lindell from the paraboLc stape of the vertical velocity curve that the mean veIccity occyrs at
0'6D and the maximum velocity a.t 0·15D. Further, it can be proved that the mean velocity
equals the average of the velocities at 2' 21 D and 0 79D generally accepted as velocities at
2jlOth and S/lOth depeh which is the basis nf the "two and eight tenths method". When
measurements of velocities at known depth are not possible, Surface in segments or even central
surface velocities in a cross-section are measured and corrected into mean velocities by
application of a "reduction co-efflcient," Velocities are measured by the following meth@ds :_
(1) Surfa!le float!':. The,e consist generally of wooden discs from 3" to 6/1 in diameter.
Observations with such floats are liaLle to be influenced by wind, ripples and eddies on the
surface of water and bE-nds in the stream. The shape and weight shOUld, therefore, be Such as
to mnimise the effect of the disturbing forces. To be safe against effects of wind, globular
floats are better than flat discs. Simpler floats consisting of corked bottles, oranges, blocks of
WOod, or even floating debris are used when improved types are not available.
 405

The floats register the surface vd )city and a reduction coefficient has to be applied to
get the mean vdocity. The exact value of this co-effir,ient is still a controversial rr:atter. The
value IS i!>tated to be greater for sandy bed rivers and minimum in bedo; of gravel. Experiments
on different rivers have shown the value to be between 079 to 0'91. In Holland the value has
been found to be 0'87 in s:lndy beds. In the Punjab Goodman collected a large number of
olservations and found the value to be 0'89, which is alreildy in use. The method i" not of an
unquestioned accuracy and is applled in high floods only when o~servatiom hy more definite
and accuraie methods are physically limpo~siblp. ,
Sometimes subsurface or double floats, consisting of metallic hollow, cylinflrical
weighted subme~ed floats attached by means of a thin cord to surface floats were used. They
have, however, become obsolete with the development of better methods.
(ii) Rod or tube Boats. A rod float con<;ists of a wooden rorl, square or round, in section
of width or d'ilmeter from I" to 2" proportional to length for strength against rough field use.
It is weightt'd at bottom by means of a lead weight for immersion to required depth. Hollow
tubes of tin r galvan:sed iron were also used but have been discarded being unsuitable due to
cracking and developing leaks.
That a rod float extending from -;-urface to bottom of a channel shall represent very
closely the mean velocity in the vertical, is an obvious inference. Cunningham in his Roorkee
exper:ments has, bv assuming that the force acting on the rod at any point is proportion&.l to
the square of the diffprence between the velocities of the rod and water in contact with it and
that th'l vertical velocity curve is a parabola, mathematically shown that Vr the velocity of the
rod is equal to Vm the mean vf>locity when the immersed length of the rod is from 0'95 to 0'97
the depth of water, the exad value oCJending upcm p:>sition of maximum veloc Iy in the vertical.
He takes mean value when length of immersion is 0'94;). Parker found that Vr equals Vrn if length
of immersion is 0'9520 with maximum. and minimum valu'ls of '97D and 0'91 ). In the Punjab
Irrigation Department, rods with length of immersion of 0'940 are used and they are mace of
full length and weighted so as to have '06 of length out of water.
Parker is of opinion that rod flgat give too high a result as compared with weir or current
mater.
The method of rod-floats is applied only for caMI observations and is rarely, it ever,
applied to rivers, The rods form a simple inexpensive and direct method of measuring velocity
easily followed by untrained men and are not afJected by the nearness of boundaTies of channel.
section. They have, howevpr, defects like requiring a la'l'gf'r party for observation, difficulty to
control their course, motion along a pulsation without neutralising effects of changes in velocity
observation of more than one cross section and poc,sibilityof fouling bed by '94D rods and a conse-
quent tendency for uoe of smaller roeg leading to over-measurement of discharge.
This over~estiomation of observed discharges by the rod-floats is an admitted fact
and in order to solve this difficulty the rods described below have been introduced as alternatives
(iii\ Lacey Tabular rods. This type of rod consists of a hollow closed tube working in
another hollow tube, closed and weighted at the bottom so as to float in water with only the
top sticking out about two inches in length. The length can be adjusted by pulling out the
inner tube and the depth of immersion remains upto the same point from top. the increased
length of the outer hollow tube getting filled up with water. These rods are avaiable in sets
which can serve for depths from 101 to 10.0'.
Lacey has mathematically proved the relation Vrn =2v 2- Vi whf're Vrn ~ =the mean
velocity or velocity at O'6D and v 2 for a rod of O.8D immersion and Vi = the velocity for a rod
of AD immersion. He has, th, refore, reommended the sinlUltaneou$ use of two rods immersed to
depth .40 and .8D and observations and comparison with current meter velocities at '6D has
proved this method to be quite accurate. [t, however, in-volves too elaborate a system and
excessive time and suggestions to substitute a single '80 rod have been made as described
hereafter.
(iv) Current meters:- Current meters are of two kinds viz. Cup meters, consisting
of a wheel with conical cups revolving on a vertical axis a.nd screw or propeller meters,
consisting of vanes revolving on a horizontal axis. To the first kind belong the price (Fig. 3)
Ellis and Watt meters and to the second, Fteley and Stearns. Haskell. Amsler (Fig. 4) Stobroni.
Kichard and Ott Price meters known as Gurley meters after the manufactureres, Wand L.E,
 406
Gurley, are now almost universally employed in America, Egypt and India.
7. Price or Gurley meter :-The equipment
consists of the following five principal parts
fig. 33
(1) The head- As shown in the sketch the
h~ad is yokeshaped, earring a wheel of of 5/1
dIaI?eter with six conical cups '2' attached to a
honz.ontal frame. This wheel revoh'es in a count-
er clock-wise dirtction on a shaft '4' resting on a
pivot point '5, 6' at the lowEr end, and pas~iDg in-
to cUmml tatoT box 'II' at the upper end in wll.ich
arrangements' 0 indicate electrically the numl:er
of revol11tions hy single or penta strokes are
provided. In Gllrlp.y No. 622 single and penta
arrangements are provided in the sa.me box and
in mder No. 623 separate boxes are pre video.
Meter No. 623 is in use in the Pnnjab generally.
(21 The tail : - The tail· 8~ consists of a
. ,
stem on which there are two vanes, one bf'ing
rigidly fixed and the other capable c f being plllIed
out by sliding in a groove. Art adjustable weight
in a slot is provided on one of the vanes. The
functions of the tail are to balance the head and
keep the meter parallel to the current.
(:3) Ths hanger:-The hanger '23' consist of
a thin steel bar, passing through the frame of the
meter with torpedo-shaped weight at the
bottom, the function of which is to keep the
meter in pItimb when hung by a cable.
(4) The recording or indicating devise-This
consists of a telephone recieved '27' with necessary
battary '26' equipment and ticks are conveyed
to the observer, on closing of circuit after the
i~~~~~!!!~~!!i[iijtll
" single or penta number of revolutions according
It to the commutator box in u£e.
(~) Suspending device- This consists of a
rod, cable or rack-and-pinion and provides for
lowering the meter and weight (when used) in
water. The rod is used in shallow str~ams of
wadeable depth. The cable is used from boats
and bridg<~s, but it has the defect of bowing an@!
bdng deflpcted from the point of obseroation ill
h;gh velocities. In order to enSUlf the location
of the meter being truly at the point and <ilepth
of observation, a rack-and-pinion d~vice is uSfd
in qractice. This consists of a rigid rack rocI
Fig 3 with teeth which works dy means of OiL toothEd
~ear.and can bp. made to stand at any depth in. the water. Details of such an applian€e a~e. shown
10 FIg S. fhe rack rod is graduated by reducing each foot to 6/loths as the velOCitieS are
measur.e~ at 0.6D. At present velocities uptu depths of 2~~ can he easily measure.d by rack-
and'oprmon arrangement and dtsigns for treasuring velOCItIes at greater depths are In hand.
. . T~p. Gu!)ey ~eter~ ~ossess reliability in service ,!~d6r most t.Tving <:onditions in the
fIeld, simplicIty In desIgn, ngloity in construction, adaptability for use III all sIzes of streams,
compact size and are now used to the exclusion of all other meters in the Punjab.
 407
In using a meter revolutions are recorded and the relation between the revolutions
and the velocity is determined by rating the meter. The process of rdting

-AH5UR CURRENT ME7£R

WITH 511SPENDING ARRANGEMENT.
Fig. 4

consists in towing a meter through still wat-'r at known uniform speeds and noting time, number
RACK-AND· pr~ION
Suspension Arrangement for current meter.

Fig. 5
 408
of revolutions and distance. Revolution .. per srcond and velocity in feet per second are
calculated. For each meter under rating. eight such observations are taken at different sPeeds,
and the relation between velo:ity and revolutions per second is determined by either plotting the
curve or by the method of least squares. The relation ha<; a str;li>:ht line equation of the form
v=aN +/3 where v is velocity in feet per secono, N is revobltions per second and /3 is the intercept
of the straight line on vehcity axis representing minimum velocity at which the wheel sta.rts
revolving. A rating ta',le for each meter in use is prepared form this. For Gurley meters
in use in the Punjab, a lies between 2'18 b 'l'38 and f3 between O'(J to 0'05. An up-to date
arrangemtnt for ra ting meters consisting of a masonry tank 400 ft. long. 8' width and 7' deep
with :ln electric trolly fitted with distance, time and revolutions-recording devices, exists
at Lahore.
Rating is carried out in still water and at the temrature of the season and will not it be
out of place to mention briefly the effect of turbulent flow ano chanEU's in temperature on the
perfor;nance of meter. Tnrbulance p )stulates variations in both magnitude and direction of velo-
city in pulsations, durations of which depend upon degree of turbul::Jnce. Investigation in America
have shown that cup meters over-register and screw-meters under-register velocitifs and some
authorities have recommended the mean of observations by b(;th kinds of rreters to .be taken as
the correct discharge. Regarding effect of temperature, Blench found by analysing the dis-
charge data at the head of the L.J. Canal that a Price lY,eter over-re!!isters by I per cent for
every 6° C fall in temprllture. This may be attrihuted to changes in velocity observatir'ns.
The mean velocity in a vt:rtical is obserVEd by a current meter in one of the io]]owjng
waY3:-
(i) Six tenths method. The meter is held at 6/iOth depth from surface and the oJserv-
ed velocity is taken as the mean velocity: This method is inv:lriably employed in the Pnnjab,
except in conditions of turbulent flow. .
(ii) Vertical velocity curve method. This consists in measuring velocities at a point O'S
ft. below surfa.ce every tenth of depth and a point nearest possible to bed and plotting a curve
with velocities as aJscissat: and corresponding nepths as ordinates. The area of this curve
divided by the depth gives the m'2an velocity. This method is employed whtre, dne to turbu-
lance, the position of mean veiocity does not occur at 0'6 D as can be the case in sume hill sits
only.
(iiil Surface velocity metind. The velocity is measured at 0'5' or I'Of below surf::tce
depending upon velocity and turbulance at the surface. M"an velocity is obtained by applying
a. reduction factor. This method is employed only when Ii) and (ii) are rhysically impossible
to apply, say in floods in hill torHmts.
(iv) Two point or tN;~and-eighi tenth mei'lOd. Velocities are obtained at two tenth and
eight-tenth dpeth and average of the two gives the mean velocity of the verdcal.
(V) Stlmmllti ~n lnt- gr:>Uon method. In this method the meter is lowered to the bed add
then raised to surface at a uniform rate. The readin~ of the meter, which represents the
mechanical av:::rage of the r:ltes at which it turns during its journeys up and down IS assumed
to correspond to the m~an velocity. Cup type meters are not suited for this method which
is never used in the Punjab.
8. Precautions for upkeep of eUtrent meter.;.
(a Although the current me'er will stand considerably hani u<;age, it needs care full
handling and atetntion to ensure its proper working. In this connection the following instruc-
tions shoukl be carefully o~'servcd:-
1. Be. sure that the set screws are all tightened up nefore putting the meter An the
water otherwise some of the parts may be lost.
2. Before beginig a measurement loosen the raising nut snd see that the meter runs
freely. Spin the meter cups occasionally during a measurement to see that they are running
freely only by blowing and nnt by hand.
3. See that the meter is swivelling fully and frt'p.ly on the suspension ror. or rack. Tf
cable suspens:on is used see that the weights are free to follow the direction of the current.
4. If any apparent inconststencv in th~ results of an observation throws doubt on its
accuracy, investigate the cause at once. - Grass may be wound round the cup shaft; the cups
may he tilted by tension the moter contact wire the channel may be o'Jstructed immediately above
the meter may be in a hole; or the cups may be b2nt so as to come in contact With the yoke.
 409
S. After a measurem~nt, it is absolutely necesslry to pour out any water that m:l.y have
collEcted in the commutator box, to clean and oil th'" bearing (in order to prevent rust) and to
inspect the pivot-point. The overseer must do this personally, using only the oil supplied with
the meter by the Dhcharge Division.
6. When the meter is not in use, the cups must never be permitted to ride on the
pivot-point.
7. Always see that the Lock Nut, on the Pivot-point is screwed firmly against the
Frame Nut, so that it will stay in place and carry the cups prop~rly.
8 In measuring hw velocities, be sure that the meter is in a horizontal position. If
it has a ter.dency to tip. the Balance Weight on the Tail is to be moved forward or backward
as necessary to give a hOTizontal balance.
9. Avoid tak:ng measuremmts in velocities of less than (l'S foot per second, because
the accuracy ofth~ meter diminishes as zero velocity is approached.
10 Should it be necessary to take measurements in high velocit;es by cable s1:lspension
inste:l.d of rack-and-pinion suspension, it is essential to use sufficient weight and aho a stay-line
in very high velocities, so that the rn ~ttr is vertically suspended below the water surface as
nearly as possible.
11. When the meter is not in use, disconnect the meter line from the battery, so that
it will not become exhausted,
12. Do not strike the telephr.ne receiver, as a heavy jar will injure the receiver.
13. Care must be taken not to short-circuit the dry battery when the meter is not in
use. To avoid this, the poles may be wound Nith insulating tape.
tb) Current Mete L Out:lt.
1. Meter itself, with its rating table.
2. Telephone ccnnected up with insulated wire in circuit, with dry c:!ll an j r;onnEcting
plugs.
3. Oil--can filled with spindle oil.
4. Small screw-driver.
S. Cable for supporting the meter.
6. Hanger.
7. Hangp.r screw.
8. Small tin-box containing 2 spare set-screws.
9. Lead weight pin· screw.
10. Commutator box screw-driver.
11. Penta Commutator box (Model 6~3 only).
12. Cotton sqaare for cleaning.
When sending meters for re-rating all the equipmeat listed above must be sent along
with it.
9. River disehargcs.
The following procedure is adopted for observation of river disch:l.Tges
, :-
(a) Selection of discharge site.
The essential feature for the accurate measurement of a discharge at a site is that the
flow should be streamline and regular, devoid of errors caused by irregularities in the motion of
water. This postulates the following deciderata in selecting the site :-
(i) It should be in a straight reach of the river with regular streamline flow,
undisturbed by bends leading to unequal distribution of velocities, cross flow and eddies,
which affect accuracy of measurement.
Iii) The section should be regular and deep with least tendency for wide spills even at
high discharges.
(iii) The section should have a reasonably sta ble bed and sides and be amenable to a
regular disLharge relationship as far as pos!'ible.
(iv) It should be far removed from confluences and structures causing vitiation br
backwater heading up. .
(v) fhe site should be easily accessible and free from cracks.
 410

(vi) In case of hill sites, the stream should be straight at least for 500', site should be
kee from projections in bed <lind from sides and away from disturbing in~ences ~f rapids, falls,
and cQufiuenc of torrents.
(b, Segmentation or spacing of sounding pvints.
The dist"nce between sounding points depends upon the width of the stream, profile
of the bed and the accuracy desired. In short, the greater the numher of sounding points, the more
accurate the area measured and discharge 0bserved, as there are greater chances of fluctuations
in velocities being evened out. The following rules are used in practice for segmentation 0f
river discharge section lines:-
Ii) The number of spgments should Lot ce less than fiite"n •
(ii) For ri'ler width in excess of 2500 ft. seglnentation is to ce fixed -by the
~xecutive Engineer. keeping in view the pr~vailing cross ~ection, distribution of
dIscharge across the cross section, equipment availahle and time involved in completing an
observation.
. (iii, For river width fran 1500' to 2500' segments shonld be 100' apart over the
portIon of section passsing 75 per cent of total discharge and 200 I apart for the remainder.
(iv) For river widths 7S0' to lSOO/segments are to be 50' :lpart.
(v) For smaller widths segments to be 40',30',20' or le53 subject to condition Ii).
. The principle to be followea is that in the portion of the section where the discharge
IS concentrated. the segments mU5t be nearer and farther apart in the slack portion so as to
have segments of more or less equal discharges.
(c) Pro::edure and equipImnt.
At each discharge site the following equipment is requirpd : -
(i) Boat-This is required for reaching the points for observation of depths and
velocities. A 3S' long Sukkar type boat is in use in the Punj abo
(ii) Suitahle anc bar with about 250/ long 1" diamether manilla ro pe.
(iii} Sounding rods of line and weight according to condition~ of site.
(iv) Current meter with accessories comphte in all respects inclulling a swivel. At
eaoh site a check meter is also kept and the local meter in us~ is checked with this once a
foctnight to make sure that it works s"ltisfactorily under conditions in the field. A bracket for
handing check meter is also provicled.
(v) Rack-and-pinion suspension arrangement, described above.
(vi) Stop-watch for use with meter. An extra-check stop-watch is also kept at each
site.
(vii) A torpedo float with cotton cQrd for indicating direction of current.
(viii) Pocket sextant, with which the angle made by the current with the section line
as indicated bv the float is read. In order to work out the component of the velocity. normal
tll) tlae seetion -line. the obsevefl velocities are multiplied by the sine of this angle and a table for
such corrections is provided at each site.
(ix) A levelling instrument for checking reouced levels of gauges and a 1heodolite for
reading angles.
(x) The cross section line is marked hy three flags, ?OD I apart on fach bank. In case
pf swift currents, the boat drifts down and drift flags apart are fixed for a requisite distance ~ ..
dowDstream of the X section line. ,(
(xi) Tn case of narrow strpams. say upto a width of 1000' as is the case with all the
Punjab rivers in winter, the sectie>n line is marked across the stream by a wire on which

Fig. 6

pendants are hung at sounding points Fig 6. _ ~he boat is held by another wire. stretched a
little upstream so as to have the ta'(;k, and-pu:uon under the pendant.
 411

(xi]) When streams are too wide for stretching a wire rope, the sounding points arb

Fig 7

DI$CH,fRtOl! :SECTION LINE.

located by pivot methnd which is illustrated in Fig 7.

From point A on the cross section line on the bank, a line AP at ri~ht angles is drwn
and from a point D on it, a line parallel to the section line is drawn. The ratio pn:PA is
~enerally kept 1:5 and the length (If AP is 1000 1 or upto .half the ~idth of t~e river. On line
DD, point say 20' apart are marked ~nd rays jr~m ~ passlDg through. tl;ese POII~ts.intersect the
section line at B1 B2 B a etc., each. 100 apart as IS eVlde.nt from the pnnc~p.le of sImIlar triangles.
The point p is called the pIvot pomt For accurate dally work the posItIon of the flags along
the section line, piYot point and direction lines are marked by cement concrete blocks with holes
for flags and the arrangement is thus marked semi-permanently. The layout can be checkPd
by the theodolite, as angles subtended by B1 P.B 2P etc. with AP should have tangents increasing
in rat, io of segments e. g. with IOO' sf'gment and 500' length of Ap; tangent,(j=l~
500 '
200 300
2
tangent 8 =-50
o ' tangent (ja=---etc.
500 For rivers wider than 2,000' pivot point layou~3 are

made on both banks, for work by two different patties from the two banks. Methods of pivot-
point layout are shown in Fig 8.
In case of physical difficulties like uneVen ground, the Jength and angle of the pivot-
point line can be varied and the observer can easily loute the sounding poin ts with theodolite
by his knowledge of tri{:onom"try. The section line is marked by R. Ds and the water edo-es
go on changing with season. 0

Warious Pivot Point Methods of Layout,

· > t~
B, A

" "1
' t 1 "

p 0 P

Fig. 8

<9 P/rar POfl'fT ON lor/l BANK3

~ rJtlIl S"T$ 01' PI"OT 1"01117'/$ ON Olle.
\,.!/. SANK "'r7'N'A FI.W POfN7'S 0" 71ft 4711'"
@ 7N~
DII'UcrIQN I'-IG$ 6tNIltO P'ut Pi'"
p. P'''O'7' PO,1£. ,
A.S. $Cc.rlllivl.'N'E.
A.P" S.P, PI'VO.,l.iIl4,
Z, :. l.,u.,C 01' D'IC£d"'~." 1'10$,
 412

In slack velocities and Jow de,Pths the boat

'( ME.THOD OF TYINQ ~HClOft )
is kept in position by a boatman standing in
water and in high vehcities or greater depths thi!
'" boat is kept in position by an anchor. Upto about
5' pp.r second velocities, the b )at can be rowed
up, from station to station. In higher velocities
while pulling out the anchor it loses
headway and has to be brought to
bank or in slacker wa ter for being
rowed up. This detouring involves a
lot of time. To reduce the . number
ROPI. 5£1'%£0 'TO of detours the anchor IS th· own
RIWC:WI'TH 2.0~3 about 300' upstrFam of the section
.......JT~ ..r Tt:lIlM5 OF Y,."'''
line by means of a subsidiary flag
line or pivot-point arrangement, and
at every station a little headway is
ANCHOR ~F"TE.R SE.I%lNG lost, enabling about 3 or 4 stations to
be done at a time without detour III
&RO"-£.N
velocities as high as 8 to 10 ft. per
second. The difficulty and risk
involved in pulling out the anchor
under such conditions are appreciable,
and the latest method of securing
the anc'lOr shown in Fig. 9 is
the safest, though it involves introduction
,
/
of some drift. Trained crew can,
however, c"pe with even the most difficult
conditions met with at some sites during
floods. Undet extreme conditions met with at
some sites provision of power launches may be
necessary which shall facilitate the procfss of
Fig. 9 observat on.
(xiii) Having divided the section into segments, the first procedure of discharge
observations consists in taking the boat tn a station by wire rope or pivot-point method as is
in use, at site and observing the depth. The current meter is then lowered, by rack-and-pinion
to 0.6 depth and velocity is recorded. The angle of the current with the section line is then
read by means of the torpedo float and pocket sextent. These obsHvations are Jloted in a
discharge note-book, a specimen copv of which is shown in Ap pendix 1 which shows full
procedure for calculating the discharge. The velocity at each station i; modified for it, angle
ani! the product d X Vrn gives the ordinate of the discharge curve at each station. The total
discharge is fQund by straight averaging. Further, the wetted perimeter is also calculated and
the slope ob!ierved for working out the value of Kutter's N. This is taken as the mean of the
slopes on the tW(} banks. Details of meters and other equipment u~ed, along with calculations
for discharge and Kutter's N as prepared by observers at each site daily are shown in the form, ~­
a specimen of which is given in Apperlrlix 1. The metho i of recording is clear, self evident and
foolproof and the data collected can be used to check the accuracy of field-work from day
to day.
In the absence of boats, improvised arrangements consisting of bull skims and
charpoys, called khatnaus are prepared for discharge observation in hills.
Another approximlte method of estimating a floo 1 discharge in the absence of suitable
equipment, is to observe the slope and flood level at the time. A cross section is observed at
lower supplies and from it the CrQ,s Section 3. t the highest flood derluced. Having determined
the slopes and hydraulic mean depth R, the mean velocity is determined by one of the
following emoerical formulae:-
{a\ Chezy's formula v =Gy'RS where the value of G is det{;rmined either by Kutter's
or Bazin's formula given in all books on hydraulics.
 413

(b) Manning's formula v= 1'4858 R2/3 51 / 2

, N

(c) Lacey's formula v=1.3~~8 R 3 14 S1/2

10. River Di<;ebarge with Current Meter from a Bridge.

. When there is a hridge near or at a disc.harge site, that bridge m:ty be made use of
if necessity accol'dingly dictates, either hy holding a current meter directly from the bridge,
or, anr! preferably. by holding a discharge boat at a sufficient distnnce downstream nf the bridge
by. mearJs of a rope traversed along the bridge. The distance marks hr segments should be
pa.lI~ted on the up;tream and downstream railing or parapet. One flag should be fixed on each
raIling and the boat brought in line with the section line and these flags. the sp.ction line being
marked by three large pole3 on each bank.
The section line must be far enough from the bridge so that the effect of the piers on
the flow of water is reduced to a minimum. A distance of 500 feet should, how;}ver, be
~uffi?ient. From a bridge where there are railings, the hanclling of a boat by means of a rope
IS fairly easy, but from large 'N' Truss Railway Bridges the handling of a boat is difficult. The
:ope used is in thrpe pieces jointed in the form of the letter '!Y", the tail piece should be 400'
10 length and t~e two arms about 160 feet each. and all three jointpd to a ~teel flng of about 2"
diameter. In the Fig: 10 when the boat is below point A, and an observation is in progress, the
postion of the three ropes will be as shown by the firm lines. When the observation is over, the
Flagman on the bridge is given a signal from the boat by means of a green flag, to move to
the next station. The two sets of boatmen holding the two ropes at A and D, so that the final
position of the ropes becomes as i3 shown by the dotted lines. When the boat has finally
reached the new position and the r ope DE has fully taken the strain, the boatman holding
the rope at A which is now entirely slack and free of any load, now travel to the third statio.n
G, crossing the rope over the rope DE at E, so that it takes up the position GE. By HHS
method which is repeated for each succeeding station, the boat is securely held all the time
and automatically comes to each succeeding station.

1/(, •

«10 •

Fig. 10

------------- -- ---- -'~----

.. I
R_
--- -----.-
- - __

~
•• 0.. -

( .
In cases where a discharge is observed by directly lowering a current meter from a
bridge; each bay should be divided into a suitable number of segments, so that on obs.ervation
point comes too near a pier or within the zone of disturbed water. The meter IS to be
held just below water surface so that the surface velocities only are recorded;
if attempts to hold the meter at 0'6D are made there will be no guarantee that
the mater is really at O'6D owing to excessive deflections of tbe supporting cable,
Great care must be taken in measuring depths which in this case will be observed
by means of a log liua held from the bridge, and since the point of suspension
 414
of the log line will be so:ne distance from the water .surface, excessive de~ection from the
vertical will result due to the pressure of the water 10 flow, and there IS. therefore. the
possibility of a considerable error being inrtoduced in the observation of depth .. in this manner.
11. Canal Discharge Observation.
(a) The observations are divided in three classes of accuracy. •
Class 1. Dppths measured at the upper and lower ends of the float run and at the
centre section line and meaned,
(lass II. Depths measured at the upper and h1)wer ends of the float run only
and meaned.
Class II [. Depths measured at the central section only.
For all the three classes of oJservations. three satisfactory observation .. through each velocity
station over the whole float run are to be taken, divergence not being permitted to exceen-half
a segment width. The boat holding wire for velocity rod o~servation is to be s;hifted irnmedi.ately
above the upper float run.-wit:e, so that veloc:ty roes may be let into w~ter exactly oppsite the
desired pendants. Placing a velocity rod in water is a job requiring previous practice as it is
essential that it is released in suc'l a way that it does not bob and spin. The best way to do
this is to hold it tightly by the upper end between the fore-finger and thumb and lower if gently
so that it is in its floating position tending to trail downstream; then the rod should be ,relea sed
with a slight fonnml push so that it immediatt-ly gets into the vertical and floats along steadily
before it passes under the upDe~ float run wire_ fhe tim -: taken for the run is t hen to be timed
by an observer on the ha"k. He times the rod over the whole run, using a stop watch, watching
it carefully aU the \\-hile becau>e if it suddenly fods the bed or proceeds in jerks the observation
has to be r"j~cted even if the roj passes perfectly below all the desired pendants. If
after Tecording thre~ results one of them differs by more than 5% from the mean of the
other two. it is to be rejected and the run rt~peater1.
(b) The standard velccity rod is I.High e 1 10 float at 0'94 of the depth i. e., if the depth
of water is 3'0 feet the rod \0 be use" will be 3'0 feet long and it will float with 30 X O· 4=2'82
feet suhmerged and the balance of 0 18 it above water !Ourface. As, however, a velocity rod
can only record the mean velocity between water surface and the bottom of the rod. it follows
that no velocity rod can possibly reAd the mean velocity b~twepn water surface and the bed
of the channel, and it is that mean velocity which is rc quired. It may al~o be found that
even if f Jr example a 3'0 feet rod is used submerged by '282 feet, it will foul the bed at some
point or other when the use of a shorter rod cannot be avoided. The application ef a
corr@ction factor, therefore, becomes essential and correction may be made from the
Francis relationship:-
!-5=r
Vrr,ean= Vrod (1'012-'1l6v'--~) ; where l=length of the submerged portion of the
rod used.
For different ratios of velocity rod length to the mean depth (If water along the path
of a velocity rod, the correction factors to be applied in order to correct a rod velocity to mean
velocity. by the above formula are as follows:-

~ =0'75, O·SO. 0'85. 0'90, 0'93, 0'95, 0'96, 0'97, 0'98, & 099

y rnean=0'954, 0'961, 0'968, 0975, 0981, 0'986, 0'989, 0'992. 0'996 & 1'00
V
rod
In the example given above if it was necessary to use a 2 5 feet rod submerged to

2'5 X 094=2'35 feet while D remained as 3'0 feet then_l_='783 and the corr(ction would be
D
=~·957. meaning, the error?y u~ing the s~aller rod and not correcting for it would be 4'5%.
ThIS error would be on the tIght SIde, meanmg that there was really 4'5% less water flowing in
that particular segment than the observation showed.
The fundamental differp1'lse between a current meter observations at O'6D and a
 415
velocity rod observations is that the current meter measures milan velocity between water
surfacE' and the b~d over a period of time at a fixed point, while a velocity rod measures the
mean velocity between water surface and the bottom of the rod only, not over a period of time
but over a distan~A, and as the latter introduces two errors, a current meter is preferably to
be used when the highest degree of accuracy is required, and necessarily to be used with the
greatest df'gree of care when observation for the calibration of a meter flume or for the
preparation of a discharge table are to be made.
Ic) Segmentation.
The cross-section will he divided into 5 main segments, as foUows:-
1 Central segment, wa; 2 Side segments, W 2 & W4; 2 Slope segments, WI & w5 •
The width of the central segments 6w 3 , shoulrl be taken approximately at the full
surface width, less 3 or 4 times the general depth, and should, if possible, be made a multiple
of 6, S0 that W3 may be a whole number. If this cannot be arranged, it should be a multiple of 3.
Th~ widths of the side segments w2 and W 4 should be equal tf) each other, exten<1ing
as near as possible to the foot of the slapes. The balance of full surface width to be divided
equally bdween the slope iegments WI amI w5 •
The typical section given bela v will show the general arrangement, the notation to be
adopted, and the pnin' s at which velocities and soundings are to be taken:-
\Vidths of segmpnts W1 W2 6w 3 · ..... • •• • .............. w 4 Ws
Distances between WI 1'12 Wa Wa Wa Wa Wa Wa W 4 Ws
velocity station.

·
I nterme d late soun d'mgs ....... rI ~ d f< f1"
' " d 5 rI"
5 "d 6 d'
6 ,{ d 7 d"
7 "d 8 d"
8 9

Soundings ("1 and d ll ) should be taken in the c entre of the slope segments and velocltl6
(Vl and observed when practicable. When velocitips VI and vi! cannot be taken they may
Vll)

be assumed at~ and Vlo respectively.

2 2
Streamers will be atta~hed to the section ropes at all the points of the principal
soundings which are also velocity stations. Intermediate soundings will be taken only )il .he

central segment. at intervals of ~va. Streamers are not necessary at these points.

The method of recording observations and

•• ~I,~ ~ I,...I. I,....

U, (1/1 ~ ~.C- [tJ,a= 1#8 ,--1.-.. "+ GJs I working out discharges is shown in Fig. 11 and
appendix If.
(8 18 $ 4- 12. Preparation of Disoharge Tables. [Gauge Discharge
. ,~
~. 'Ii
III
q ;,~ Methed]
'"
.. .'i [aJ A series of discharges are to be observed
'i~, .;~ ~ ~~ 't ,.. "i ''''II:: at round about steady full supply, three quarters,
half and one quarter full supply, against a perma-
,\ I nently fixed gauge. For each series of observations
one meaned value I!)f 'Q' the Gauge reading 'G',
- - I\nd the Area of waterway 'A' is to be taken out
3 bv rejecting any obviously erroneous observation.
i\ i
'D' or mean water depth as area divided by Mean
~ V Width is then evaluated, and from the four values
~ .... r... V of 'Q' and '0' so arrived at, the connection between
.~
"r- / 'Q' and 'D' is determined by evaluating for
'K' and 'n' in the equation Q=K.Dn uiing the
• ,

1<i" 11
.. ... .'"."-
method of least squares. The £il'lal stage consists
of preparing the discharge table for values of rD'
corresponding to gauge reading a<; read from
 416

a plotting of 'D' against 'G' in order to determine intermediate value of 'D' cornsponding to
intermediate values of 'G'.
(b) For a channel section.
When framing a discharge ta ble using the relation for a chanr.el secti
Q=K.Dn .it is insufficient to assume n=5/3 ~~
a.r.tually It must vary for each individual
SIte; on the other hand, where the highest
degree of arcuracy is not required, it is
sufficient to evaluate 'K' and 'n', graphically
by plotting 10gQ against log D as shown in
fig. 12
'TAN c( .. n As Q=K.Dn expressed in the form
of :-
A.N71 LOG )'= #( LogQ =n 10gD + 10gK is the
equation. of a straig~t line, the four points
i representu,g the senes of observations, at

"
such all'ay that :-
LOG D.

Fig. 12
round about steady full supply, thrp.e quarters,
half and one quarter supply, will fall in

1. The tangent of the angle, the line through the poini'> makes with the log D
axis=n.
2. The ordinate. the line through the point makes with the log Q axis=log K.
The solution to the values of' K' and 'n' is thus completed, but this graphical method
depends on the corr~ct visual appreciat.iJU of th~ divEfge~ce o~ the points from ~he straight
line and the fitting 1D of the appropnately fittmg straIght hne; the mathematIcal mathod
appreciates the minutest diver~ence and pro,;,ides a solution which mathematically fits the
entire range of supply over whIch the observatIOns have been taken.
13. Special precautions for Calibration Discharges.
The followit,g precautions are to be taken before and during a calibration class of
observation :-
1. The guage, whether in a guage-well or outside, against which the discharges are
to be correlated mu:;t be absolutely firm and secure. To frame a discharl1" table againit a
guage attached to a loose and shaking post is a complete waste of time. and in amy case no
such guage should ever be used in connection with any recording of discharge.
2. The guage well, if any, must be functioning properly and it should be seen that
it is not choked with silt, and that the holes or silts are perfectly clear.
3. Each observation of dbcharge should be in duplicate>, using two newly re-ratd
check meters on the same suspension rod or rack-and-pinion, and operated by two independent
observers. 1his serves as a check on the work during its prcgress as, if one observer gets
at a stlltion an entirely different velncity from an other 01 server's, the o' servation can
immediately be repeated after checking over both the meters. Finally the two results, if differing
by not more than 1% may be accf'pted and meamd.
4. A complete observation must take the form of a double traverse, 1 hat is,
one set of observations will te from left to right and immediately followed by another set
from right to left; with two Observers there will in this way be four sets of results and it is
essential that they are all worked out finally at site awl cross-cheked.
5. A regular program aLe for taking the whole series of observations of suitable
suppLes is to be prepared in consultation with the local raguiating officers so that the series
may not drag on indefinitely and yet cover as much as pos~ibR> of the whole range of
supply from full supply down to not less than 40'/'0 of full supply.
6. Special efforts must be made by the 1<egulatiu~ 6[afl to kEep the supply steady
and if the variation in gauge during a caiibration discharge observation is greater than
0'01 ft. the ob~ervathn is to bf' rEjf'cted.
7.. The discharge site mmt be as near the gauge of the flume under calculation as pos-
sible in order to avoid the effect of time-lag. If the distance, however, is unavoidablv apprecia-
ble then time-lag between the discharge site and the gauge or flume, has to be allowed tal. If an
 APPENDIX 1 FORM D 1.
DAIL Y DISCHARGE DATA.
River Sutlej Site Harike Date 10-9-1944
Meter No. and make Gurley No. 171-22,304-S Time from 7/30 hrs. To 12/0 hrs.
Equation of Meter Permanent. i Temporary
Gauges.
I R R. L.B.
Date of last rating 6-4 -1944- -- - -----~

Description of surface floats __Zero_ ~.L~I __~~_~6_1670·56 j 670'56

. Length of float run Beginning. j, ; 9' Ip , 8'9S 1~_~9_~_
Float run marked- with
Section line marked with
Sounqing taken ",ith
Method of suspending meter
Pivot Points
Sounding ROd,$
Rack and Pinion
~::~F~~- ~:~ ~~:.
Weather Clear
- .
~.erysligh:'
Timf'piece used Stop Watch Wind driction ~slight.
_"'strong.
__"""eT)' strona,
Weight used Ibs. and strength
r Fairly Clear
Condition
I
of water~ Ordinarily Silty Yes Current
I
l Intensely "

SURFACE SLOPE ObSERVED.

V = Mean Velocity= (? =3'05
Right Bank. Left Bank. A

Back
~ - I ~i~~r
l'ore. , ellce.
B
~c
k I F- -.- r Differ-
. ·ure.· ence.
- __ -_. . -- ---- .------ 1_--____
i" - yRS=O 040
500' D./S. 7'25 j
- _e _______ ~I_~~ C= - "'_ =76'25
yRS
o
--~-- ..-------- -:").•-
~
,
..
X=41
. 1)0281
6+~---=56
.
40
500' 'O./S. 5
i 6'58 I
i
I- y= CX . =1,463
Fall 1,000 ft.
I 018
YH.
Z=C-X=Ht'90
'":_"~,,"
Mt'an=O'185
~,
N"", y(Z~+7;-244Y)-=~_ =0'028
2Y

Character of river bl'd Sandy

Class of roughness under which it falls II
Passed for check by Initials S.D.O. II daten ......... ..
. ~_ it j ; '
Check~ d by
.~~--- Collator dated .......... ..
R(" ugh-volumed by Collator dated .......... .
Compared & recorded by Head Collator dated ........... .

ON HIS MAJESTY'S SERVICE.

THE SUB-DIVISIONAL OFFICER,

DISCHARGE SliB·DIVISION.
STAMP
 'p;uoq:>uV

-------------~---------------

08·9'"
09·6f:=Zr..l X Z/09(~J)
OZ·l =SI.OXZ/08 (!} pa:mpaa

0'0
UU
· .
- -
'" 00
,:,
--
Ib
C':)
--
'"rj,

-"" -
o
;...

-----------~------_------_ --

iC':)

UN

UN

--- -'"
o
,~

----~-:__:_-'--'-----':':_--- ------ ------------------ ---------

-
--

-
'?
~
o <?
------------------------~---------

o
- - ""-
0) -.to
o 00 8 o o
~------- ------ ---- -- --_.__ --_ ---- --------_----------

t1) .L~ ---~-- .

-
o
~
o
(0
b
c_> ._

'" u
't:..9
~~
o
o
~
--- -
o
CD
g o
CD

§~ o o o
.0
~'';::
.~

~(/)
u
o
'"
~ -
o
~
oC"I
-~

Note :-1. Col. 16=(common width of segments-!the sum of segments on either side of the R. D.) X Col. 15
2. When floats are used, or more than one meter observations are taken at the same section, each obsarva-
ticn of time and revolutions must be rocorded in a separate line in Cols. 5 and 6 respectively. In Cols.
1 and 2 all the lines relating to one station will be bracketed and R. D. on section and water depth
will be recorded once.
3. Co-efficient employed converting surface velocity into mean velocity should be noted in the remarlts
column. Unless specially directed it is to be taken as 0'89.
4. If DO drift occurs. the same is to be shown as nil; the column is never to be left blank.
 411

observation, for example. is from 10.0 A. M. to It.30 n. M. and the distance is 5000 ft. a.nd

the mean velocit ... is 2'S feet per secood the lag will be = 5000 =33 rr.inutes, when the com-
J 2·S.l( 60
parable gange, if it is upstream of the iite, will be that prevailing from 9-27 to 10-57 A. M.
8. W)Jen observi:sg velocities, times must be recorded to the nearest IJSth second
and the products of "D xv" mtlst be taken out to three places of decimal, and when finany
prepe.ring the equation not less than seven accepted series of observations a.re to be taken
mto mat:R.omatical analysis.
14. RegttlatioD.
(a) Use of Discharge 'fa.ble.
The discharge tables are calculated to give the discharges for every hissa, that is, up
to first decimal place of a foot. The discharges for t1<ie intermediate Gauges for tliie ~econd
decimal place are worke::i out by interpolation. U!Sually enamelled Gaugoo with divisions
upto second decimal of a foot are fixed. The discharge tables are rnaintaintld f0r all dis-
tributary head!! and far the indenting site at the beginning and at the end of a canal
sub division on the branch or main canals. Usually there is a fall availatle which is d~igned
to wt!Jrk as meter at the indenting site. If no fa II is available, a discharge !3ite is constructed
in the open channel by pitching the bed and sides for a length upto ISO feet so that a straight
run of 100 ft. length is available for the discharge observation. The Gauge is loca~d in a
gauge well con5tructed about the middle of the pitched reach. The discharges for S'uch sites
are worked out as explained in para 12 (b) above.
(9) Indents,
Every Sub-Di visional Officer works out the requirements of 5uP9ly for the sub-
division by adding authorised. Full Supply discharges of all off-taking Jistributaries and
the seepage losses in the main or branch canal in the !lub-division which are usually
sanctioneci by the Superintending Engineer in the Capacity Statement of the eanal. In
the case of rain-fa.1l or slack demand in certail'l portions of the sub-divi"ion, he focestalls fhe
reqUirement!! of the distributaries ann then adds up the required discharges to work out the
indp.nts. The 'fail S. D. O. wires to the Upper S. D.O. and so on to S. D. O. Head works of
the Canal. The rndents should specify both the Gauge and the discharge at the indenting
gauge. Twenty four hours notice is usually required to meet the indent. The
sub- divisional officers on the tail Peaches of a canal should calculate the time whelil the effeet
will reach considering velocities in Main Canals 3 miles an hOUT and in Branch Canals 2 miles
an hour. Regulation is a very important part: of the duties of an Irrigation Engineer. 'fhere
are usually printed instructioJas hung up at the important regulators on the main and ocanches,
'canals sperifying the duties of the regulation staff. The regulation staff is on duty tor 24 hour
on the important reg,Ilators. They atonce attend tG the orders issued by the officer-in-ch<lll'~ of
,regulation ordering r.hanges in indents.
(c) Immergent Indents.
In case of accidents to canals such as breaches or cuts, immergent indents are issued
'for reductions The foremost duty of the officers and the regulation staff is to attend to them.
The regulation staff and the signallers are supposed to attend to the Telephone and Te~egrapl)y
in.5'tmment~ even at night. .
Bibliography.
1. Paper No. 272, Punjab Engineering Congress, 1945 "Methods of Discharge Observ-
atioI'ls" by S. L. Malhotra 1. S. E. Executive Engineer, Punjab Irriga.tion; 2. Discharge
Observat i on5 by Hayat and Grover.
3. Lewett's Wydraulics.
15. ExaminatioB questions.
1. (al De9cTibe different methods of measuring the discharges oJ fiVers and canal~.
(b) Describe a -'discharge rod." What is the best ratio of the length of the discharge rod to
.edeptho!water? (T.C.E.19B)
 418

2, Describe in detail how you would proceed to measare the discharge of a river about 400 yards wide,
Max: depth about 20 ft, max: sud9.ce v~locity about 2 ft. per seeoud. (T. C. E, 1935)
3, The observed discharge of a canal with 7'5' gauge was 600 cusecz and with 6'5' £t gaur:e 450 €usecs,
Frame a diseharge table showing discharges with 6'0' gauge and 8'0' gauge. (P.1. B. 1941)
4. Descpibe the mellhod for taking a first class discharge of a call.al with, current meter. A branch
canal has a discharge of 1500 cusecs with gauge 9'5' ft. at site and 1000 cusecs with gaugCi 8'5' a4: site.
How wiil you pr(')ceed t·) frame a discharge table for this site?
What will be the discharge with a gauge of 6'5' ft. ? (P. I B. 1939)
5, FrQm your experience describe an accurate method of measurj,ng the actual disQharge of a channel
having 10 to 15 Clasecs capacity, you have a stop watch, tape, measuring rod, Give specifications of the measuring
rod,
6, Deseribe the method of observing the discharge:-
(a) of a nver where it is not possible to stretch a rope across it;
(b) of a canal with 1st class observations;
"J. (a) Describe anyone type of Current Meters used for discharge observations;
(b) Give briefly the percauti~ns you will take for proper upkeep of Current Meters. (P.U.1942)
8. (a) W:ow wtll you work out the indent of supplies in the canal for your Sub-Division?
(b) Explain the process of segmentation in a first class canal discharge.
9, (a) How are \Veddle's and Simpson's rules used for working out canal di-schrges ?
(b) How will you make correction in the dischariJ€ observed by velocity rods. when a stroog wind is
blowing ?
10. (a) How are the velocity observations eorrected I:ry' velocity rods to get the correct mean velocity?
(b) \Vhat points will you keep in view while selecting discharge si te in an artifieial channel?
II, What are the different methods adopted for measuring the velocity of flow of water in a stream?
Mention the eondition~ to be satisfied in selecting the site for aocur I1te gauging of flow in a stream, (Mysore 1>941)
12. How would YOll observe the discharge of a channel using;
(a) velocity rods.
(b) Current meter? Deseribe any type of current meter with which you are best acquainted,
, t, ,~...
{F.Sc .. 1941)
 419

Appendix II.
FiM Cla!!ls Observations.
Discharge taken on ........................ at R. D. tO,Oeo Main Line
Between 10 A. M and 1-0 P. M, By Executive Engi·neer... : ............................... ..
Assistated by Assis.tant Engineer, Reading of gauge at R. D. 1,000 below ff·gulator-.
Before observation 60 }
Mean 605
Aft< r ,f i 1
Surfac-e slope-R. L. 500 f~et above centre of run ... 83470
R. L 500 feet bel~w ditto ... 83t 52
Bifference or fall in 1,000 feet O'IS
Fall in 1,000 feet 0 18, or S = 0'OW1S
Slight br€eze > Strong bref'Z6 >
Much Wind - ~'-~ > Direction >
Le14gtk iif r1tn-l 00 feet Timing doni-with chronograph.
Floats used-loaded tin tUbes Surface width= 126. C6ntraZ segment= 108
w1 =ws=4o'.0. wa=w 4 =5'.G. w.-=IS'.
TABLE OF SOUMDIMGS.

Diiltance I
frorM
right
i Nota·
tiol1. I(j"pp".i Middle.1 Lower. Total. Mean. IAverage.
}J'ota.
t40n,
bank.
~ I I I
o -dl-~
2
2'0 1--;3 -1·-2'~1- -~6-'4-- ~~-·13-1--2-·1-3-
4 3'5 I 4'2 I' 4'0 11'7 8'90 l"90
4'87 "1

t
3~ 4'7 ! 5'1 48 14'6
15 5'1 I 5'3 59 16'3 543 >- 524
21 5'0 5'4 (50 HN 5'47
27 5'4 5'6 81 17'1 g'70
33
39
5'5
5'6
60
5'9
6'0
6'_'
6'4
6'5
6'4
17'8
18'1
18'8
5'93
6'03
6'2,)
t
tlI
45
51 81 63 6'2 18'6 6'20 ' '17
x 57 6~ 64 60 IS () 600
~3 60 56 5'7 17'3 i 77 I
.; 69 60 5'8 5'S 173 5:77 5'8S
75 66 6'0 5'3 J7'9 6'91 j
X
81 6'2 65 5'2 179' 5'97 ,I 5'91
87 6'0 62 5'2 17'4 5 SO >-1
.;
~'g~ ~>-!.I
93 5'9 6'3 54 176
X ~9 5·7 6'6 5'6 17'9 D II
5:"
X
~'
X,
105
III
J 11
122
5'7
56
5'5
4'0
6'5
6'0
5'7
49
8
<!
59
5'4
4'3
18'0
175
16'6
13'2
6'00
i'83
5'53
4'40
ti
:
I
'73
4'40
JI
12(
126
22 ,},'S'
250 I 250
 420
VELOCITY OBSEIRW A TrONi),

d Distance Length Time rf

0
from right Timeo.!
:;l of rod passing first ~ndiDg Veloc-ity
;g bank, used. Seconds. Total. Mean.
0 rope. run, per
Z sellOnd.
~

1'03 ~ 2'15 72 "1

VI 2 2'0 2'31 3'40 (;9

V-2 4 3-75
4'02
700
5'09
8'04
07
64 "1
J 208 69-8 I 44
8'15 9-20 65 -{
930 193 64-3 1'.'16
10'34 64 )
12',,0
Va 9 4'5 13'17
12'59
14'15 58 ~
~9 "1
179
i
14'40 15'42 59'7 1'67
62 )
20-05 2050 45
21'10
v4 '1.7 5-25 2 '30
2['52
23'!7
42
47 )
1
:.15-02 1 !55 45'0 2'22
2548 46 I
28'10
Vo; . 45 5.75
0'14
1'20
2'31
28'55
054
1'56
3'07
45 )
40 "1
36 I-
3(!l
112 37'" 2'GS
)
5') Z 5'54 42
V6
68
5'5
640
8'60
7'21
844
41
4A I-
1 210
I 42'0
9'10
10'50
13'0 I
953
II 30 I 43 I
40 )
2'38

81 13'45 44 I
V7 5'0 14'00 1441 41 ~ 124 41'3
14'5-9 15'38 . 39 ) 2'42
0''08 037 3'7'
1'12 "1
Vs 99 I-52 40 I
5'5 2'15 2'55 40 }- 204
3'02 S't5 43 40'8 245

V9 117
4'12
6'00
456
6'49
II 44
49
)
I
5'25 7'00 "1
8'16
752
9'08
52 r 153 51'0 )'96
I
I V10 122 3'75
10'49 11'42
52
62
J
"1
I I
1208 13·11 63 }-
15'00 189 63'0 1'59
16'1)4 64 )
Vn 124 2'0
18'00
19'30
19'10 I 70 "1
20'42 72 }- 215
21'12 22'25 I 73 J
71 7 1'39
I
CALCULA [roN OF DISCHARGE,

, Va = 524X!'67='
Q) 8i'S
'3
... 5D4 v4=5x5'70x!'22=
63'27

6DG v6 =6 X 5'S5x 2'38=

83'54
D7 v7 = 5'91 x 2'42= .
....... 14'30
...
5Ds vB =5x5'~~X2'45=

D9 v9 = 5'73 x 1'96=
.! J: titl 'n) t=O· .. \J \)3 ~U ·.i :f=..1..; __;j
Discharge=3w3_or 5'4 X
1~ 270'52 I 1,460'81
 421

CALCULA nON OF DISCHARGE.-(c0ntiJ'lUed)

D 2v 2 =3'9xl'S6 == ." 6 08 I
D3V3 =as above ." 8 75 i
D9V9 =as above 11'23 i
D10v lO =4',Ox l'~O - ... 7'00 I
--33-061 82'65
I

"'_'..
ffi~ 4D 1 v 1 =4x2'13xl'44= 12'27
S ~ D 2v 2 =as above = o'ns
~I:: D1QV10 =as above _ ,., 7'00
i ~ 4D ll V l1=4X2'5x 1'39 = ,,' 13'90

~~
__.__~_ _ _ _ _ _ _ _ _ _6 _ _ _ _ _ ! . . - -_ _-
Dis€harge=~or O'67x ... i 39'35 26'30 .
_ __
Total Discharge D=1,5GS'76

OTHER CALClTLA TIONS.

AREAS.
Wetted Perimeter.
Central SegmQnt. Slope and Side 5egments,

... I .
Da 5'24 D2 .. 3'90 -yI~Wl)2+(D2)2 ... I 5'6
5D4 ,..
...
28'50 D3
,
.., 5'24
\

100
D5 6'17 Dg .., 5'73 W 2 +W4 ... 1

6D s ... 35'1 DIO ... 4'40 iI

D7 ... 5'91 i W2x ... 19'27 =48'17 6wa 1080
5Ds ... 29'75 t D2 ... 1'95
· .. 1
I

Dg ... 5'73 Dl ".

t D,o ...
2'13
2'20
V(ws,2+lD 1O )2 ..,I 60
Total ' .. 1
116'40 Du ... 2'50 ' ." ~-,

X 3w 3 = 5'4 ~
I Total 1296
'"
10 ! WIX 8'78 =17'56
'" I
Area= ... 628'5$ 64'73 j
Area of central segment 628'56
.. other segments 6.5 73

, A=Total area ;~., ,

.1 P=Wetted perimeter

Hydraulic mean depth=~ =R =

P

A-ylR'5 =694'29-y1S'31 xU'~0018 == 21'558

Co-efficient= D 1569'76 == 73-S
AySR 21'888
 PAR r II -'" i'

CANAL IRRIGATION
Chapter XX
Water Power from Canal Falls
Introduction.
The necessit y of creating canal falls (drops) has Ceen explained in Chapter X of this
part. When the slope of the country is steeper than the permissible (regime)' slopes all(")wed
for thft irrigation channel, th~ additional head available in a canal is given the form of a canal
fall. Such falls ca.ft be utilized fur the development of motive power asd for other useful
purpOies as under:-
(1) Generation of Hydro·Electrie power which can be utilized for lighting, heating
and various inclustrial purposes.
(2) Dirt'ct drives, lIuch as flour mills and lift wheels The latter can be used fe,r
llfting water to commanc! the high ~djoi.nin!! areas.
~31 :,Hydromats (Saction or compressisn type). (Already described in chapter XIII
of tms part).
(4) AutosucNon weirs.
(Dc£cribed in Chapter IX of tBis part).
2. Power available from canal falls.
An approximate idea of the energy available from 8. canal fall can be obtained "(11m
the foltl,,)wiJII€ calculation.s : -
Let Q = fhe ql1antity of water in cubic feet-per SfMnd flowing throug. the canal.
H=Head in fed through which the water drops at the canal fall.
w=Weight of water in Ib~ per Cubic foot=624 lb. The theoretical water power

available fr om the canal f<ill i8= Q. H. '!:_ 62' 4 ~. H. horE9 power. The actual useful
550 55
t'neqy available depends upon the efficiency of the various parts of tAe insta.llation. The usual
pffide.. ci.es may be taken as under ~ - "
[::I) Pipe lines, intake channel etc.,,-e 1 =!H to 98% i;< .. i
[bJ Hydraulic turbine= e 2 =80 to 90% "
[c] Electric gp nerator=e 3 =90%
Thus the overall efficiency from water to electrical side will be e=ei X e 2 )( ea. This is
u;u:oJlly about 'i5 to 80%.
Note :-In mor1ren practice tre turbines are directly couplee to the generators on a eommon
shaft. The old praCbl!fl was to use a gear arrangement but as this involves a loss ~f power of
anout 5%, direct coupling is preferable.
Taking the (werall efficiency at 75% the electrical power available from a fall
62'4QH x 75 QH_ horse power approximately. 1he ccmrr.cn unit of electrical power is a
S50x 100 12
/--
kilowatt which is eqmal to 1'341 horse power.

. 12 1'34] 16
=
Expressed in kilowatts tke available electrical energy=_2!!__x _1__ QH_Kilowatts

(approximately). Taking an aetual example, suppose the quantity of water flowing through
a canol i~ flO CUS€Cs and the height of the canal fall at a given point is 10 ft. Then the /'
f leotriGa 1 el\ergy available from this fall would be : - -.t

5001~ 10 =312'5 killowathl

 423

Placing 500 Ton Rotor in one of the 82,500 Killovatt amperesigenerating units
Boulder Dam, U.S.A.
/ .

__/ /
 424

Installation of four 82.500 Kilowatt ampere generators, Boulder Dam Power Plant, U.S.A.
 us
, To have ltD approximate Mea as to what this energy is clp~~)h~ of doing. the normal
consumptions of various electric~l appliances are given below: - ;: HIt. ~,il',
[il Ordinary electri~ lamp=40 watts. ' +:'!i'F
[iiJ Electric table fan. =40 to 50 watts. ", l',L'lI
riii] Electric ceiling fan =60 to 80 watts. ....
[Iv} Electric heater for room=500 watts. <::'1 "r di~lt'N ft, •
[v] Flour mill 2£ ft. dia: stone=lO Kilowatts. " /;
[vi] Tube-well with a discharge of 1'5 Cusecs and lift of 20 ft.= 12 Kilo Natts.
One Kilol'latt=1000 Watts. ," '!~ ,-",J n ,
3. Hydro-electri, Installa.tions on canal falls. ..' .iJ

A Hydro-electric installation on a canal fall will consist mainly of the folloNing

parts:-
[a] Headworks for diverting and confrolling the flow of water from the main canal
to the head race or intake channel leading to tho power house.
[bJ Headrace or intake channel.
H , '

~~:;~·:~/~.'~~+'·'.,,{.r~-r~""y~ttft 1)?£ 31
[cl Water turbines together with the control apparatus.
Cd] Electric generators together with switch gear and control appartu~.
[e] Housing for the dydraulic and electrical apparatus, items [c] and [dJ above. \r~\~;
if] Tail race leading the water back into the main canal.
4. Design 01 Hydro·electrie Installation,
In nesigning the various parts of a hydro-electric installation the engineer 'bas' to
consider \'ery carefully the most suitablp. velocities of water at each part so as to determine
the cross section and design. High velocities mean the sacrifiCing of a large percentage of the
total head available for developing power at the turbines. On the other hand, low velocities
require a largtlr section of intake and out· flow channels tkereby involving a higher capital
cost. The engineer has to determine the economical mean I etween the two. In low head platns
it is m~st important that the head lost in friction and eddies be as small as possible and,
therefort>, velocities should be as low as possible. Suitable veloclties in various paTh of a
hydro-electric install~tion are given below :-
[iJ Headrace or intake channel
The maximum velocity in open flumes should Qot exceerl about three feet pAr second
as a higher velocity is likely to cause whirlpools ano. eddies which will enter the runner and
cause cjisturbances ill the smooth flow of water. In steel pipes used on hiE!'h head installations
higher vehcities can be userl. Usual velocities for heads up to 200 ft. are 8 to ]0 ft per second
ann for higher heads the economical limits for velocities are still higner. To maintain a
constant head and constant supply of water to the intake channp.I art overflow weir is necessa:-y
which in modprn practice is designed as a spillway syphon. ,See chapter VIII of this part)
The intake channel should be designed non silting. [Chapter VI, part II}. ':1;: '.,;;" . .-
[ii] Turbine pit. .. , ,
In open flume plants, the intake channel leads direct into the turbine pit. The
velocity of watfT on entrance to the runner on this type of installation should b~ about 2 ft.
per seCClnd to avoid eddies in the runner, Wh~n turbines are prOVided with spiral casings so
as to guide the water in smooth passages, higher velocities can be used. When spiral casings
are of concrete, velocities of 5 to 6ft. per sec. are recommended. F()r plate sted spiral casings
velocities as high as about 20 ft. per sec. can be used.
[iii] Draft tube or suction pipe.
The velocity at the exit end of the draft tube should bel about the same a~ in the
tail race. In ION head plants this shOUld be about 3 ft. per sec.
[iv} Tail-r:l.ce.
In low head plants the velocity of water in tail-race should not exceed about 3 ft. per·
sec So that the out-going water leaves the turbines with as little kinetic energy left in it as
pos ible. Besides, higher velocities cause eddies and d~stroy the suction action in the draft
 4~
.'
tube which leads to lower efficiencies. Slightly hil!'her velocities upto' 5 or 6 ft. per sec. m~y be
allowed as the head incre;}.ses. In the case of Pelton wheel instiallations the tail water has no
effect on the turbirle and its velocity is determined by such conditions as the available area of
channel with respect to station foundations etc, " .
(v) Inlet bend to Pelton wheel nozzles. =..:t ~";i';~; ~,hl'j'7f]: 'i:
. The velocity in the inlet pipe of a Pelton. wheel is generally kept at 18% of the
sp~uttng velocity of wa.ter i. e. 10 Yo of Y2gH where H is the effective head of water und~r
whIch the turbines operate. For he<l.r;is exceeding about 600 ft. it is preferable to reduce ~t
to 7t% of the spouting velocity and for very high heads of the order of 'l000 ft. and above, thIS
percentage may be reduced still further to about 5%. The velocity should In no case exceed
aboMt 30 ft. per second. , ;'l.'." .
5. Selection of e~uipment . ' .... ,

. Hydraulic turlilines may be classified unuer two main headings:-

(i) Reaction turbines which work by me;}.ns ofthe p"tentia-l and pressure energies of
he, wate:. Francis turbine runner is the chief example of this type in the present day use.
It lS a IJll>.'_ed flow tuabine being partly inwarn flow and partly axial flow. ,
(11) Impulse turbine work unner the kinetic energy of water, and the POWlilf IS abstract.ed
from the water by allowing the jet of water to act on a number of buckets fixed to the nm
of a disc O£ whe~l. The principal example of this type is a Pelton wheel. . .
The c~aracteristics of thesQ. two types of turbine runners are described in deta~l .lD a
!ater part of thiS Chapter. Hydro-electric pJ.ants usieg reaction turbines are further subdivided
lllto tll.e following principal categories:-
(a) ·Open flume turbines.
(b) Concrete spiral cased turbines .
.(e) Plate steel spiral cased turbines.
(d) Cast iron or cast steel s~iral cased turines.
~a, Open Hump. turbine.
"_, .".
, In this type cif construction the intake water channel leads direct into the tnrl~ine
runner. FOT low heads not exceeding about 30 ft. this is the simplest and most economIcal
c?nstruction as it eliminates altogether the necessity of pipe lines, valves, turbine casi~g~ etc.
Stngle runner vertical shaft type construction is usually adopted as this gives better effiCIency.
The w~ter after doing' it~.work in the rnoner discharges through a suction tube into a tail water
c~nal J116t beneath the' fo'o~ of the turbine pit. Ihe turbines should be so arrang?d that the
distance from the .snrface of. the incoming water to the highest pomt of the
turbine runner, or gui<;le vanes when these are provided. is at least equal to the diame~er of ~he
;l,lnner, as otherwise ai,r IS likely to enter through eddie5, thereby reducing the suchon actIon
m the draft tube and din'iiilishing the output of th~ turbine.
When rocky foundations are available, this type of construction may be economical
even fOT heads upto SO ft.
(b) Concrete spiral cased turbines.
These are wled 'prIncipally for capacities above 400 H.P. at 10 feet head and above
50f)O. H.P. at 100 ft. head. They are seldom used for heads above 100ft. due to the necessity
of remforcement in the cO'llcrete which makes them more expensive and less reliable. . They
are almost invariably constructed in the single runner vertical shaft type.
(c) Plate steel spiral caood turbines.
These turbines are used for heads exceeding 40 ft. upto 375 ft. Above this head a
<>ast type casing is used, These turbines are usually constructed in vertical shaft single runner
type as this ha!, a greater efficiency~ .
Cd] Cast iron and cast steel spl'ial cased turbines.
Cast iron casings are used only for moderate size units working tinder medium heads
as cast iron is not a reliable material for .high stresse5. For large units and higa headg upt_o
1000. ft. cast steel castngs are used. When the head exceeds 1000 ft. the Francis type of fl:lDDer S
ives too high aspeed. a~d implllse turbines are used.
 (e) Impulse turbines.
These are generally used for heads above 850 ft. They are alsO 'used in smaller uIl'its
dowJli to 2f1l0 H.P. at 100 ft. head. These turbines are usually of the horizontal shaft type
either one or two sets of buckets being used to drive each generator. Turbines of this type,
are in use for heads upto 5000 ft. in Switzerland and for heads varying from '2000 to 3000 ft.
in other parts of the world. .
6. Suction pipe or draft tube.
>. The suction pipe or draft tube is simply an air-tight tube fitted to all reaction type
turbines on the dlscharge,side It extends from the discharge end of the turbine runner to
about 18 inches below the surface of the tail water level. The suction action of the water in this
tube has the Eame efi( ct on the runner as an equivalent head so that the turbine developes the
same power as if it were placed at the surface of thl!) tail water. The action is similar to that
of a syphon, the water exerting a suction proportional to the height of the column. Theoreti~
cally the centre of the turbine shaft can be placed about 34 ft. (height of the water barometer)
above tl1fl tail water at sea level, but full advantage of this cannot be taken owing to the
dissolved air in water and due to srr.all air leaks. In practice the maximum head that should
be used in suction is about 25 ft. at sea lev@t and less according to the altitude of the site.
Straight draft tubes have generally a flare of from 4 to H degrees, depending on the
length, so as to rec uce gra(Jually the velocity of water, which then discharges quietly into the
tail water and with as little energy left in it as possible due to residual velocity.
The draft tube should be supported securely otherwise severe vibrations are
set up which may disturb the whole station. Large draft tubes are usually moulded
iIi concrete as it is easier to obtain grandua curves in concrete than in steel plate. At
sites subject to high floods the electrical equipment has to be placed above the high flooa level
and in such cases the maximum length of the draft tube has to be used so as t9 avail of the
total head upto tail water level. SueQ contingencies should be carefully studied from the data
of rainfall and flood levels extending over a period of 40 years or over before designiog the
installation. If data ava~lable js_ for alesser period, a safe margin should be added.
lU.i 'jf~_"I,"::;'''''''''''''' IV....... '
7. Specific speed.
The specific speed of a turbine may be defined as the r.P.M. at which the runDer would
run if it were so reduced in size, without in any wa.y changing the design, that it would develop
one hon;e power under one meter head. Meter has been selected as the unit of head as the per-
formance of hydraulic turbines is more often expressed in metric unNs. Runners having the
SQme specific speed have similar characteristics of performance, so t?at specific speed is a very
ccr..venient term indicating the type of a turbine runner. High speCific speEd runners are suitable ..
for low heads and low specific speed runners are suitable for high heads. The specific speed is
a ocmplete measure of the possible performance of a given ruDner under any head, both as regards
power and speed, and thus it gives an indication of the suitability of.a given design of runner for
any given set of conditions of head, speed and power. When considermg a turbine which bas more
than one runnp.r OIl its shaft, the specific speed is based on the capacity or. out put of one runner;
hence the total capacity of the turbine must be divided by the number of runners on the shaft
for embodying in the specific speed formula given hereafter.
Let D=Dia : 01 runner in meters.
Q=quanityof water passing through the ruoner in cub. meters per minute;
N = B.H. P. developed; n=Speed in revolutions per minute; and H=net or effective
head on the runner in rr.eters.
Then :-(i) ncq/fI (ii) Qocv'H; (iii) NocH3/2 for a given; design of runner;
(iv) NocD2; [v] QOCD2; [vi] nocl/D [vii] Doc VN-. The formula for specific speed is
n vN
ns=-~-
H514

\\ hen invesitigating any givEh scheme the head; quahtity of w{lter, and the total
powtr output are readily available and the first problem is to find a suitable type
 428
soo of turbine and its speed. The curve in
Fig. I shows the maximum specific speed
suitable fnr various heads, and enables the
••, maximum specific speed being dt:1ermined
at a glance. This curve has been obtained
.!ND/ClTes TURBINES BU/~T AND IN by collecting the data of a number of
j
f- SUCCESSrUJ. OprffIJTION, ,(ND existing plants, and -is not based on any
INDICATES TilE f)EPAR"I/I1f£,~R(JM theoretical calculat ions. The application
t\.r VAL IItS INOICA.,'&J 8" THE. 'CURV",
WilleN MAY 8£ MAllE WrrNSAr£r,(. of the specific speed will now be
300
::a:· \ demonstrated by the following example.
Example 1. Suppose it IS desired to
tri 1\
I devflop 6400 H. P. at IO meter head by a
)I( \ single machine. The electric ~ellerator is
~ \~ . to generate alternating current at frequency
~ ~
of 50 cycles. To detetmine the speed and
the type of turbine suitable for this develop-
§ I ;.
.......
..........
--._ -
ment. It will be noticed that the head is
5uitdblp for an open flume turbine and this
fO() ~ -- should be adopted. From the curve in Fig. I
it will be served that the maximum specific
~
II) • !'peed for 10 meters head is 360. Assum-
II ing one runner, the actual spfed of the
.&:: turbine may be determined from the
specific speed formula as below :-
SO 100 Z()O
HEAD IN M£iRES _ n YN~.. _ n, X H 51 4 _ 360x lOS!<l
n - - - - -, .. n----=- - ---.-~-
, H5/4 yN • /S40()
!Uj ;) iu,; Fig. 1 v
h. =79·2 r. P.M.
This is not a suitable speed for the altprnatoJ, hence we will try two runner~ on one
shaft. The output per runner will then be 3200 B H.P. and sub~titutiIlg this in the above
equation. we get :-
360 X 105/4
blIwN n= ~.J-HOO =112 r.P."'.
'1f;:~'i;;'rhe nearest speed mitabJe for an alternating current at 50 cycies is 125 r.P.M. and this
125.,< y3200
will Le adop. ted. The specific speed corresponding to this speed will be
.
ns =-'---,------
lU5~
=401.

This is higher than the maximum specific speed derived from the curve, r.ut as the head is very
low, some deviation from the curVP is possible. This spEed "ill, therefore, be adoptHt
Example 2. A turbint' is requireo to develop 3000 B.H.l'. at 80 meters hEao. Tbe
generator is to gener:lte altnnating current at 50 cycles frequency. To determine the suitable
type and speed of turbine.
Froln the notes given under selpction of I'quipment it "ill l::e noticed that plate-steel
spiral cased turbine will be suitable for this head. From the curve in Fig 1 the maximum
specific speed corresponding to 80 meters head )S 150. Hence the speed of the turbine will be:-
n= n, x H6/4 150 X 8051~_ =654 r P II,f
yN' \,,3000 •.
The nearest synchronous speeds of the generator corresponning to 50 cycle fJequency
are either 600 RP.M. or 750 r P.M. If we adopt 750 r.P.M. the generator no doubt will be less
costly, as the higher the speed, the less bulky will be the machine. and hence the cost will be
comparatively less. The specific speed corresronding to 750 r. P. M is:-

n = 750 X y__]_Q_~ = 172, which is J5% higher than the maximum specific speed obtained
;",,: s 80514
;"-"'
 429
~';!Ofi ~d 1~;<Jl !l.Z :~H ni flwoila ' ," <.rtt to'll i
from the curvl'. This deviation is not desirable for a head of this magnitude, hence the next
lower speed of 60l) LP.NI. will be arJopted. The specific speed corresponding to this spet:d is
138 and turbine runner of this specif,c "peed will be selected. 1,). ! II " :') 1'1. 'l 'J;; l ~ !,
. .: ,~~j ~.:)~1:1j .. r~'~,. ';:"'P ~ ;~,.~;:v:,
8. General Principles of Hydraulle t ubines. ·. }i:~_~Hb·I' ecLt v~-! . ~f""';\ r': i J'\:i'J'l ~,;o
" t :)~~ rc!n'~
It has been stated under para 5 that hydraulic turbines are of two principal types
ta) Reaction turbines of the Francis type and (b) Impulse turbines of the Pelton wheel type.
Below are given a general descripti .n and the principles of working of each type.
[a] Franch turbines, The runner of Francis turbine consists of a number of vanes
spaced round the circumference of a wheel which revolves on a shaft. The water is guided into
the wheel by means of guide vanes, and n.e reaction of the water on the vanes of the runner
proouces a torque causing motion of the turbine. The guide vanes also regulate the' flow of
water into the runnp.r, being mOl1nted on a shaft about which they ca"! turn through the governor
action. thereby reducing or increasing the water input according to load. In Fig 2 two circles
~' represent the inner and outer periphery of the runner
, and one vane. is shown with tips Band C; LetCl
~_ _~.. represent the velocity of the jet before entering the
runner. _
U1=periplwral speed of outer casing. U2 =peripherel
speed of inner casing. Ul and U II are tangential at B C
. n where nand r are t h e radii of the
Vl =--
respecbvely;-- 2
U2 r 2
I outer and innl'r casings respectively.
, The triangle BOP is the entrance velocity triangle.
L: VA,,! Til' P/tDPue~D in which BP is the velocity of jet at entrance and BO
is the peripheral speed. aDd OP repftsents the relative
Fig. 2 .'
~ ...... -.......
'. -'
~ ~.

velocity between the jet and the casing. <

The wattr should enter the runner smoothly without sho('k and for th:s end OP
should be parallel to AH,i.e. the direction of the vane tip at entrance ..\ similar triangle can
bp. constructed for the discharge end of the vane tip. such as CMK, in which CM represents
the peripherHI velocity. Ck is the discharge velocity and Mk is the relative velocity between
the jet and the vane at discharge. The m'lgnitude of Ck, it will be noticed,' will depend on
thp. vane tip angle and this velocity should in practice be as small as po,sible.
When the water enters the wheel, it sets up a pressure tangential to the periphery

equal to ~-Q_ X CI cos. a in which "w"is the weight of water pp.r unn volum~ a~d HQ"is the
g .'
quantity of water entering the vane in unit time. Similarly. if the discharge velocity CK be,

denoted by Cllthe pressure set up at discharge end is ~g X C~ cos f3 These two pressures set ;; t

leverage of f1 and T2 respectivly. where r 1 and r 2 represt>nt the ou ter and inr.er<radi'i of the_

. I
runnpr rp,spectIve T
y. hus. '
the turnmg momen t a h outer vane en d 15_r1
t te . X wQ X C1 Cos a,' -g-
and that at the discharge end is r 2 x ~g- X C% Cos f3. The effective turning moment on the
g
shaft is the differenre bp.tween the above two moments, and if we muhiply each moment by
the linear velocity of the whelll at the respective 'points, the difference wilt give the work dOlle
or the horse power of the mnner.
 430

A typical efficiency curve of a Frand; Turbine is shown in Fig. 3. It will be noticed

that the cune droops rapidly on both sides of the mlximum efficiency point and it is, therefore,
necessary to run it at or near the load of maximum efficiency.
. It, Pelton wheel. As previously stated, a Pelto~ wheel consists (lEa number of buckets
fixed to the rim of a wheel and a water jet issuing from a nozzle under the pressure head of
water impinge3 en these buckets, thereby setting the wheel in motion. The el1letgy is imparted to
the whep.l solely by the kinetic energy of thp. jet of water. The speed of the Pelton wheel. has
a definite relation to the theoretical velocity of water issuing from the jet and these relations
are given below:-
'-~_"''''''''''''''!'''T'''T"'T''''''''''''''I'''T-r-:-rTlT1''''t.Q Let H=the effective head of water
~+-1H-H+t+.J-+~~l.H-+-t-1I,+t-t-H'+i 9b in meters, then the theoretical velocity
~'D ~ t:±:q:t:tittjj"~,iGIf~,~t;~-~-~~~ of water, usually called the spouting
"~;' ~ L!tlr.~(GU#RA~f;iot~ 80 .. velocity; will be V2gH ~ ...
I
~ 7D.)ba f-++-+-+-t.H~p~'-fl,I!:;CTtr-r-H-++-f-l-+.:l+H 10 ~ (i) The velocity of jet v=0.97V ~H
~ '" .tI ~ lie' meters per iDecond.
1 iD ~ ~ 1"-1>' ~ Iii) Wheel velocity in meters per second
~ 50 200 ~ 1-+-+-:iFt--H-+-1H-rt....,.._'",.~~r'<.:~-4-iI-+-+4-150)... is usually taken between 0'43V 2.gH
..... S f~r.~~ ~~ l'f!,pvF++-+-{4D~ and o 47XV2gH. This gives the best
~.,fO 'a ",,, °.k1._1 ~,.t:~~i-+-+-+-l-I-l ~ efticiency. The lower figure should 're
~ ,~) I-Qu~"'('Xl.l "J,"I!"'_;:;"~4-H-+++HJII ~ used for high specific speed and the higher
'AO"';/ /-::: ~ "LfI"I!E-~~+i++++-!+H-H ...
1D ~ figure for low specific spe@ds.
l:)
<;). I,., '"
,~20.... " T h e term wheel velocity for a Pelton
~l '(J wheel means the velocity of a point on
~·o " "" 8 , 0 11 I t " an imaginary circle, to which the cenhe
o I , S. H P IN iHOUSANIn }ina of the jet forms a t ang.en t. The
diameter of a Pelton wbeel is the
Fig. 3 diameter of this imaginary circle.
lliii} The diameter of the Pelton wheel is det1trmined as follows :-
let wheel v~locity =v=0'45vZgh--meter per sec. Knowing the value of H this can
readily be worked out.

let n=r.P.M. of the wheel; then v=7\"D n/60 ; or 0= vx_60.

. 7\"'n
(vi) Jet diameter. To work out the jet dimeter the total quantity of wate r required
by the Pelton wheel must firi>t be determined. The effiCiency of a Pelton wheel is usually about
90 Per cent and the friction losses in pipes, etc, must be taken another 5 to 10 percent
Knowing the ov~raH efficiency, the quantity of water required for a given capac·ity can easily.
bp. worked out. Let this be "Q" cub, meters per second' thus:-

d==d a of jet== / Q -=-meters.

V VX\)'7854
where velocity v=velocity of jet in meters per sec. The ratio Old varies from 8'5 to about 15
. or higher effiCiencies.
(v) Number of buckets.
The number of buckets has a pronounced effect on the efficiency of a Pelton whee 1.
The efficiency increases as the number of buckets increases upto a certain poin t anf, it then
drops if the number of buckets 15 further increased. The graph shown in Fig. 4 is based on the
data collected from a number of existing Pelton wheels giving good efficiency. The number of
buckets is plotted against the ratio Djd of wheel to jet diamp.ter. This graph will give an
a.pproximate idea of the !uitable number of buckets for the design of a Pelton wheel.
(vi) Size of ltuckets.
The width of buckets is usually frotn 3'7 to 4 times the jet diameter; height of buckets
from 0'7 to 0'8 time the bucket width, and depth of bucket is 0'3 to 0'4 time bucket width.
 4.31 .v, "

CIJRV£-IIE.PRE.S£N18 T1I£ It is quite evident from the above relations that if the
1'RACf? CI. II$I/AI. 111' TIJ ABDI/T 131~ number of jets is increas~d, the bucket dimension will becOiTle
CURY£--R£P,,£S£N1S flltA&'rlt:L smal!er, and as buckets form a very expensive part of the
1It"%(I MANY PlANTS CMlSrlfllCTR turbme, the cost of the turbfine will be less. On the otber
DN "HUE. L,NES !lAVING eWE N
hand, high~r stresses in tbe revolving parts due to higher 8peeu
I
lIERY GOfJD£r:-nCI£IIey CUIWt:.
_. of the turbme produced thereby will involve higher construction
cost and these two factors have to be balanced against each
,
0
r~

--. other to arrive at a suitable arrangement.

8
r The c_urve in Fig. 5 is a typical efficiency curve of a Pelton
I wheel. It wlll be noticed t}lat the Pelton wheel gives practicaUy
,11
5 '"
~ I
-.i

I
I
the same efficiency at part loads as at full load, and in this
respect it has a considerable advantage over a Francis Turbine,
'3 tQ r 9. Number ., units.
- J 1.~
, I~ 1'"-
_I
The greater the number of units in a Hydro-electric plant
,
10
the more ealiiily can they be run at their maximum efficieRcy, and
this has thus an advantage from the operation point of view. On
~-- the other hand, the cost of foundations and installation increases
o S .:II'
as the number of units increases. Usually in an independent
"'IO~~£~J
.IS" installation, four to six units should be installed depending fln
Fig. 4 the capacity of the installation. Six units is an ideal installation,
four units being designed to take the maximum load, the fifth
being kept as a stand bye for emergencies, and the sixth unit can always be opened for overhauls
by turn. As each unit can take about 25 per cent overload, there are in fact two units available
'TYP1CA.1. E~F"'C'E.NCV CURVE
OF'A P£l.TOR WHE.E.L.
100

S
U
~Q

So
I(
~
70
Q.
)"
60
u SO Fig. 5
Z
&.oJ
~
\)
~ 30
\&.
W 20
10

1000 3000 41110 .5000 ,00. '0_

QU:T PU" .... ttQP.SE POW£P,

as standbyes for emergencies and the factor of security of supply is, tnerefore, very great. Five
units may be installed for slightly less security, and in installations of less importance, four
units may be installed. Less than four units are Dot desirable unless the in!tallation is
inter-connected with another so as to depend for inter-change of load in emergencies.
10. Inter-linked canal Hydro-electrie Developments.
The main disadvantage 'Of Hydro-electric developments on canal falls iii the
d'iscontinuity of supply as the canals have to be closed periodically :- (a) For repairs (b) When
water is )1ot needed for cr'OPs, and (c) f'Or rotational running of inter-linked canals due to
insufficif:nt supplies in the rivers t'O feed all canals simultaneously. It is necessary to have
steam or Gil stand bye plants to run during canal closures to have continuous supply of electri ..
;city Such stand bye plant involves large additional capital cost, besides high running expenses.
Si{]gle vertical Hydro-Electric uuit in simple ope.u SEcrrON
fll'lffie for low head;; up to 20' head
"i,~ ~;~};~T ~~: ~I
PL ... N

[====-__ "l~~~~~~Z?2~~~~~L_ Pj; l,!'e,.

Fi 6
Fortunately most of the adjoining canals in the l>un1al, ar~ illterlinked and if one is closed, an
adja€ent one will be running. It is, therefore, possible to havtl interlink~d hydro~lectric
developments on adjoining canals connected. with the same net-work of electric supply system,
plan H(H..IJI-. ;.. '-I..... Section

,f ,
••
I
I
I
,,
I
I
'f

•
f
r 1
I

Fig 7
 433

and this will eliminate the necessity of providing and l:laintaining the standbyt' fuel plants. In
the present canal systems in the Punjab there ar~ at soma places several small falls withm a
f~w miles length of the canal. From the hydro-electric-development point of view, falls with
bIgger drops are preferable, and it is an important point to be kept in view in the future design vf
canal systems. Typical layouts of low-head plants with open Burne and spiral cased turbines
are shown in Fig 6 and 7.
. 11. The foregoing notes give a general idea of hydroelectric installations and their
eqUIpment. The ilesign and selection of hydro-electric turbine and their equipment is a
specialized subject on which the advice of hydraulic firms manufacturing such equipment
must always be oDtained. Names of a ft''H important firms are given btlow:-
1. Boving & Co Hydraulic Engineers, Kingsway W.C.2. ":1\ E f!t}t)wbG
2. AILs Chalmers Manufacturing Co. Milwankee, Wis, U. S.A.
3. Escher Wyss and Co., Zur.ch (Switzerland). The electrical part of a hydro-
electric installation falls withm the province of an Electrical Engineer and is not vealt
with in this t.>ook.
; :;:)Jh\r; ~!,;:,
12. Flour Mills run by turbines. I1b'" I.r.l
~
In the Punjab. the water power at the 'canal falls has been extensively used
to run flour mills. They have been commonly run by the use of country made wooden gird
wheel carrying a pair of stones described fully in the succeeding paragraphs. In a few cases
the low head Francis turbines (suitable for falls. from 4 to 10 feet) have also been used.
Two instances of this type are on the Lower Jhelum Canal at R.D. 25,500 and 64,000 of
Northern Branch. Turbine shaft is coupled to a cam shaft rotating two to four pairs of
stones. Generally speaking turbines flour mills are expensive in the initial outlay and
expensive to maintain as compared with the gird wheel flour mills, .,
13. Gird wheel or tlour mills of t!te old Indian PattUB. ".
The Flour Mills in the Punjab Canals bringiiB a considerable revenue to the Government.
The income from Mills in the Upper Bari Doab Circle for the year 1944 and 45 is given below :-
Name of the Mill. H lld Number of stones. ·,·jl (! Daily Income Amount.
. Rupees.
1 Aliwal 1 & 2. 9 64 /-
2 Aliwal No.3 5 31/- ...
3 Ahwa14 & 5
4 Parowal.
5 RaneVYaJi.
12
11
6
75[- .
. '1/.1 '.
' ):i.tlt/"": ~ i .
-...,..,,_
r:.~

6 Kohali. 8 . iI_no: .'4

7 Kotla.
8 Pakhoki.
i
S ' !4i~:" ~
9 Jaura. 5 , 37f- U
-.
10 Athwal. I 35/- •
;."
II Raya.
12 Nagoki.
'12
5
6616
·r t ...,_ •
'1:1
tI.O
13 Alladinpur. fi 4,7/8 A
14 SUjanpur. 5 331 '51
15 Sarna.
16 Tugial.
6 41V
'·M·
1:11
!.
./
., '1,

17 Nanunangal.
6
5
12
- '2931-
1,,, ., 0
."
18 Dhariwal. et
19 Bhuchar. 6 36/8 x
C<I
20 Bedian. 6 36/8 ;;;-
8 Sl/6
21 Lulliani.
-
t'\I
0
22 Bhanba. 6 33/-
23 Bhanba. 3 16/10

Total 156 1033/2

The above fi~UHS give the inc.ome for one cana.l syste.m. There are fifteen such
systems in the Punjab. and vee" nearlv similar figures are obtained on other canals as we~l.
The quan tity of the flour proluced T@r pair va.fies. from 20 .to .30 maunds per diem
and approximately the num ter of pairs of stones workmg m the PUlllab IS about 2000.
 434
, The niill wheel or "gird" varies very little in form or dimensions on the Bari Doab
a.nd Sirhind Canals. being universally made about 2 feet in diameter, on the Western Jumna
Canal 2 feet :1 inches is the size adopted, while on the Swat Canal wheels 3 feet 6 inches in
dIameter are used. .
The mill stones vary In di~meter from 2t to 3 feet and al'e about 6 inches thick, when
new, and 2t inches thick when rejected as worn out. They are made of Agra sandstone, Ifxcept
on the Swat Canal, where a very hard conglomerate stone obtained from near Michni Fort is
employed.
The "head" under which the mills work varies fr.om 22 feet to 6 feet, being generally
between 3 feet and 5 feet, qut in few cases qpto 9.0 ft. I.

14. Desgin of Gird wheel. ' . . ~. ., ..,

. The water is applied to the vanes of the "gird" through an inclined shoot of a
rectangular section, the sides of which converge towards its lower end or mouth. The angle
at which the shoots are inclined to the horizontal varies from 30 to 60, an inclination of.35
being a common practice as shown in Fig. 8. The dimensions of the mouth of the shoot very
considerably, according to no apparent rule with reference to the height of f~ll or size of
~tones (a common size is SO to 90 square inches, that is 7 to 9 inches wide, and 9 to 12
Incbes 'deeP). ' .
. On the Sirhinu Canal It bas been found that, with working falls of about 5 feet. the
dlscharge rtecessary for a pair of 3 feet stones ( the size in common use) is given by the formula

Q= _4~_, '\'t>he!-e Q=the discharge in cubic feet per second and h=the working fall in feet so

tha\. a 5 feet fall would require 9 cubic feet per second per pair of stones, and a 4 feet fall about
11 cubic feet per second, It has been pointed out by Higham, that Qh is not necessarily
constant for all falls, and he, bas q~oted experiments on the Sirhind Canal showing that a
t • •. . ' . •

discharge of !~_ cubic' fe~~ ~uffices: to give good results on a 4 feet fall in ,the case of some new

~i~is ~;-~~~~;~:·~"h~~~;·~:;';h~~~~~ at Kh:;~r the discharge was ~the working fall vary-
c h
ing from 5 to 8 feet, but in the latter case more water was probably being admitted than the
mills could utilize. It will suffice, however, for purposes of roughly investigating the subject,
to assume that Qh has a constant value=45 for all heights of fall considered. It may also be
assumed that the co-efficient of discharge at the mouth of the shoot is 0'81 and that the axis of
the shoot is incli ned to the horizontal at an angle of 30°.
Let AB represent the velocity and direction of motion 0 fa
particle of water issuing from the shoot and striking the vane pq,
and let the velocity and direction of motion of the vane at the time
of impact be represented by the line CB then the velocity and
direction of impact of the praticle relatively to the vane is
represented by AC. The particle of water will do work in driving
the vane, provided that the velocity of the vane is less.than;C}B ;
where AC 1 is parallel to the vane pq. If the velocIty of the
F' vane=CB 1 the particle of water will merely slide down the vane
Ig. 8 ,
with a ve10city relatively to, it.:__AC] and do no work, and if the velocity of the vane exceeds C}B
the particle of water will strike the back of the vane, and will tend to retard the wheel.
Let AB=v and C1B=U. We may assume the plane of the face of the vane pq to be
inclined at an angle of 60 0 with the horizontal, that being in accordance with parctice.
Then C1B=ABsec. 300 :.U=V=1·l5v
. ,", .-~ _. "'. , ,I:'" ., , .:',{
But v=O'Sl X l'1lv'2gh =7'48 v' h . ' . ",
The table shows the values of U for verying heads from 3 feet to 10 feet:-'-
 435

Head in feet 3 4 5 6 7 8 9 10
U in ft. per sc. 12'9 15'0 IS'3 19'5 21'1 22'5

a.nd evidently no portion of the wheel should have a greater velocity than that as,signed to
lJ in the above table.
Th~ usual number of revolutions made by the native mills when working effi.ciently is
150 revolutions per m i n u t e . ' ~
The f0110wing table shows the circumferential velocity in feet per second of a point at
8. distance R inches from the axis of a wheel making 150 revolutions per minute:-

R Velocity in ft. per sc R Velocity in ft, per sc,

~- .. ---~----

4 5'2 15 197
5 6'S 16 21'0
: 6, , 78 17 2~'3
7 91 18 236
8 ., :,.,) 10'5 19 249 ! bns
9 .,,,,,_,
118 20 262 ~"

]0 13'1 21 27'5 l't·J

11 14'3 '22 28'S
12 15'7 23 30'1
13 17'1 24 31'4" r .f '1,( ~,:,,' '1' <' j 1 :~ ~\., :.'
14 18'4

It will be seen from these tables tbat with a 5 feet fall (a cornman working fall in the
canals) the limitins distance from the axis at which a particle of water striki.ng ,a vane does
any work in driving the wheel at 150 revolutions times per minute is about 12t inches so that any
addition made to the size of the wheel beyond a diameter of 2' 1" will at only be of no use
but any water striking near the ends of the vanes would in such a case actually retard the
wheel. It is a noticeable fact that 2' 0" to 21 1" is exactly the size which the native millers have
universally adopted on the Bari Doab and Sirhinrl Canals. ,
Similarly it follows that the following diameters should not be exceeded for the falls
named below:-
Height of fall in feet Dia: of wheel in inches height of fall Dia: of wheel in inches

ii, 19
~ ,. 30
''4 2~ 32

6
5 25
28
9
10
-'J;}t' -;.. 34
36
The work is done in these principally by the impact of water, and it is therefore
probable that they work most efficiently when the vanes are moving at about half the velocity
with which the water over takes them, 'This being true of other water motors which are propelled
by impact.
,. The component of the velocity of ineffluent water resolved in the direction of motion of
r,; I

1- /~
the vane is v cos 300='\1'; v and the best velocity for the vane would be, therefore, be vV-T
and the shO'lt showed be so placed that its center is situated as nearly as possible in the
position where this would occur.
The following table shows the distance in inches from the axis of a wheel making 150
revolutions per minute at which the velocity is vi and, therefore, the distance from the
axis at w}:ijch the centre o~ the. sh~otshould. if pO!tSible, be placed,
 Fl
436.

Head in feet. Vy! in feet per Distance from axis in

second, isches.
3 6'0
4 7'0
S 7'8
6 8'5
7 9'2
8 98
9 10'4
10 11'0
Dimensions of the shoot:-

The shoot is to give a discharge Q= :5 cubic feet per second. The values of Q and v

and the necessary areas of cross section of the shoot in square inches are given below : -
Working fall in feet 3. 4 5 6 7 8 9 10

Q=Dis : required in
cubic ft. per second.... 15·0 11'2 9'0 7'5 6'4 5'6 5'0 4'5
v=0·8 I y2gh ............ ll·2 13'0 14'6 IS'9 17'2 18'4 19'5 20'S
Area of shoot in square
. hes -Q- X 144 ...... 191
Inc
v
124 88 68 53 44 37 31

The results obtained are summarized in a tabular form below:-

Working head in feet 3 4 5 6 7 8 9 10
Dia : of wheel in inches which
should not be exceeded 19, 22 25 28 30 32 34 36
Distance in inches from the axis
at which the water does most
work. 4'~ 5'3 6'0 6'S 7'0 7'5 8'0 8'5
Area of ori gce of shoot nece-
ssary in square inches. 1911 124 88 68 53 44 37 31
Data assumed :-

Number of revolutions made by the wheel= 150 p~r minute. In~lination of shoot to
horizontal 30°. Indination of vanes to the horizontal 60°. Velocity of vanes when the water
impinging on them·a maximum amount of work in driving the wheel=half the horizontal·
cemponent of the velocity of effluent water. Co-~fficient of disc:harge of the ilhoot=0·81.
Now if we attempt to design a suitable arran~e nent for pairs of stones for various falls
from 3 to 10 feet in accordance with the ab)ve results. we shall notice the following. First
with 3 feet of fall. the wheel should be only 19 inches in diameter, and the best place for the
water to impiJilge is 4f' from the axis, but the boss in 8 inches diameter, and the shoot
ough.t to measure 191 square inches alld we find it quite imp:>5sible to bring anythi.ng like
enough w:tter on to the shoot vanes which ara oaly 5! iaches in length, anrl scarcely :tny of what
reaches them and strikes them near enough to the axis to work in the m')st ecnomical manner.
Witll a 4 feet fall we meet with the same difficnlties, aTld the best point to appJy the water is
51 inches from the axis or 11 inches from tne rl) It of the vanes, which are 7 inche; long, and
as the orfice need only measure 124: sqllare inches, we can apply a large prop)rtion of the
discharge from a shoot say 7 inches wide and 14 inches high.
 431

plan
Fig. 9

Section

SECTION

/
of Shoot
c. L.-orWater
Wheel .

PLAN'
Fig. 10
 438

With 5 feet or more of fall we have no diff!culty but when we have to deal with a
high fall of say 10 feet, we find that the shoot need only measure about Si inches square and
that the best point of application is 8t inches from the axis. There is no objection to the wheel
being as much as 3 feet in diameter, but with such a small shoot having its centre only Ri
inches from the axis nothing is gained by making the dia: of the wheel greater than (say) 2i
feet diameter, as we can apply all the water necessary to a wheel of that size in the most
advantageous way. The tails of shoots are shown in Figs 9 and 10.
or
15 Efficiency Girdwheels
.
Light stones do not give such good results as full sized ones, as well as the quality of the out.
turn -falls off when the stones are driven too slow. With the ordinary pattern of mill now in use
both these causes evidently operate in reducing the efficiency of mills working on falls of less than
(say) , feet, and it would appear that at sites where the fall is as low as this, and the supply
of water abundant, better re,;ults would be obtained by designing a shoot which should deliver
the water on to a larger two segment of the Wheel, or substituting two shoot for one. Also
the sUbstitution of a )Vheel with a thinner spindle, 'perhaps a metal wheel, would apparently
allow of more watf'f being arplii>'d at a part of the wheel where it would work economically.
On the other band, at sites where the fall is considerable and the supply limited, it
would be worth whi~e to make experiments with a view to find out whether the :;hoots in use
are not extravagently large for the stones.
It may also be noticed that where there is a high fall it would be quite feasible to use
much heavier stones. In England stones 4 feet in diameter are commonly used and these are
run at 140 revolutions per minute, and grind 5 bnshels [about 3 maunds] of wheat per hour. It
does not seem customary in the Punjab to attempt to run such large stones. but one reason
may be that the mill houses are so bnilt they could not be set up. In the Peshawar DIstrict,
however, large stones are preferred, and are said to produce better floor than smaller ones.
References.
l' "Hydro-Electric Handbook" by William P. Creager and Joel D. Justin.
2. "Water Power Engineerng" by Danien W. Mead.
3. "Hydro-Electric power Developments" by Rushmore" and Lof•
• , "Funnamental Princinples of water Power Engineering"
by F. F. Ferque.:"son.
5. teN ote o~ Gird Wb.eel Flour Mills" by A. Reid, Under Secretary,
Irrigation. Punjab, (1895)
 ERllATA"

Page 67 Line 21 rea<i found for fined

74 45 course couse
80 3 From Bottom insert cost "
after in,~reased
87 8 read Doab for Dob ,i'..d
"
.'
II
96 I Apron upon
116 "
\_ [
'l'~:;
.. testimony ..
" testimny \<J;
;.;;' :
127 14( t~ r ",' " 2x 2x l:t;,;
tj",t"'b " x
136 7 delete along
"
, :.( F dl,da
"
136'~ 14' read .dp.da for
dl.da-- dp.da

2 From Bottom I b
138 f)f'C for ---xC
b1
142 7 x ,. K
2712 2'712
142 8 ., ~-=
32'2 " 32-:-2
148 6 From Bottom 'I 5 II s
154 17 "simplified ., signified
158 16 Add Known as Kennedy Diagrams after Hydraulic Diagrams
159 II rtlad Vd for Lq
159 13 Vb I, ., Lq
160 table 2 line 2 J, '893 ., '693
162 line 3 From below J, v= II

164 6 L L
.. -t.:1- 'J Rl
173 4 '# + ,. X
1713 3 " below •• on page 177 (d)
203 19 " I "III
210 12 l)Jsert and between concrete and brick lining.
214 12 • r~ad modulous for modules:>.
13 From below " undetected •• undertected.
195
2?:l 5 I HI'S
I'x (5'69l3/~ ,~
1 X (5'6)3{Z
228 4 '085
v- '485
I,
7 "
234 2
36
I,
'.
18000
Fig 16
d~lete very
<

..
/"

Fig. 124
before Pakpattan
1800

last line add waves afte:- silt

235 4 add full before supply
236 "
12 From bottom r~ad in a wall for a wall
241 12 Fig. 14 Fig. 123
242 last line " II
"
insert impart "
before action
4 From bottom '1 of the cistern after depth
10 r~ad - 1'25 D J for 1'25 :0
 440

Pace 248 Line 4 read I fer 1

256 2 ., baffle bottle
2(:; 11; "
_ ,. h
260
..
-
2S8 17 n Fig. 9 chapter VI Fig. 89
'JB1 11 From cottom insert Vol. III after XIX
3 H= after co-efficie. t
304 20 "
read attention for attaintiotl
·831 8 issert Engineering
Congress 1916 after Punjab
332 1 read II fClr VI
334 9 III '1 ,. 17
340 4 From bottom aerated II rated
"
340 last Ene read (t-)i+(+) i for ( ~ )3 _~_ = ( ) i
ACC<3S';luH ~,;:) .. \\3.i'.....
!
Ddte ............. .. .... , ..... ' .
__ ..I
JNIYERSITYOF AGRICULTURAL
SCIENCES
UNIVERSITY LIBRARY
BANGALORE-24

-, 2 AUG 2004-
\ \Pl~(\ l
 6i31· 5~
Cl. No .....~.l::t:A............

Author ..Sl:t.A.RlYl.~

1r ~1...~ . ---.-----
;;'~P..J:i~
( "
\ b~.