BETZ Handbook of Industrial Water Conditioning

._... ~ .._,__ ... .,..,.. - -....• ,._...,. •• tr., .,.,.--...~----~...,....._...... _ _ _ ~-· -·-~· ~ - ~ .... .....
Jttl~~~'
- . ~ - --- .
https://boilersinfo.com
BETZ HANDBOOK
OF INDUSTRIAL
WATER CONDITIONING
SIXTH EDITION
Published by
TREVOSE, PENNSYLVANIA 19047
SIXTH EDITION-1962
Sixth Printing-March, 1973
Library of Congress Catalog Card Number: 62-21097
Copyright 1962
BETZ LABORATORIES, INC.
TREVOSE, PENNSYLVANIA 19047
All rights reserved. This book, or parts thereof, may not be re-
produced in any form without permission of the publisher, ex-
cept by a reviewer who wishes to quote brief passages in
connection with a review written for inclusion in a magazine or
newspaper.
Printed in U.S. A.
Acknowledgment
This book was prepared by the engineering and technical

staff of Betz Laboratories, Inc. Many individuals contributed
to this edition and previous editions. We particularly wish to
thank the following for their editorial contributions to this
Sixth Edition.
H. L. Boyer A. J. Piluso
A. B. Cherry 0. H. Preis
R. B. Conlan M. U. Priester
J. M. Donohue L. F. Probst
C. M. Frova R. L. Reed
P. J. Gaughan J. H. Richards
H. B. Haldeman C. J. Schafer
R. H. Hayman J. R. Schieber
J. B. Lord B. F. Shema
J. J. Maguire W. A. Tanzola
S. 0. Meyer J. G. Weidman
W. L. Nieland G. A. Wilhelm
C. A. Noll
6 BETZ HANDBOOK
TABLE OF CONTENTS
INDUSTRIAL WATER TREATMENT PROBLEMS AND PROCESSES
Chapter Page
1. Industrial Water Conditioning 8
2. Water and Its Impurities ................................................ . 12
3. Aeration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4. Subsidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
5. Coagulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
6. :Filtration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
7. Precoat Filtration ......................................... , . . . . . . . . . . . . . . 42
8. Chlorination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
9. Hydrogen Sulfide Removal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
10. Iron Removal and Iron Retention . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
11. Lime-Soda Softening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
12. Hot Process Phosphate Softening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
13. Silica Removal By Magnesium Salts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
14. Sodium Zeolite Softening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
15. Cation Exchange by Hydrogen Zeolite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
16. Demineralization ......................................................... 102
17. Dealkalization by Chloride-Anion Exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
18. Hot Lime-Hot Ion Exchange Softening ..................................... 115
19. Oil Ren1oval . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
20. Deaeration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
21. Chemical Feed Systems ................................................... 135
22. Boiler Scale Control ...................................................... 151
23. Boiler Corrosion Control .................................................. 167
24. Carryover ............................................................... 174
25. Silica Turbine Blade Deposits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
26. Measurement of Steam Purity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
27. Condensate Return Line Corrosion ......................................... 194
28. Embrittlement of Boiler Metal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
29. Boiler Blowdown Control .................................................. 210
30. Chemical Cleaning of Boilers .............................................. 221
31. Cooling Water Treatment: Once-Through Scale Control ..................... 229
32. Cooling Water Treatment: Once-Through Corrosion Control .................. 236
33. Cooling Water Treatment: Open Recirculating Scale Control .................. 247
34. Cooling Water Treatment: Open Recirculating Corrosion Control .............. 258
INDUSTRIAL WATER CONDITIONING 7
35. Cooling Water Treatment: Closed Recirculating Systems ...................... 270

36. Pretreatment and On-Stream Cleaning ..................................... 274
37. Acid Feeding to Cooling Water Systems ..................................... 279
38. Slime and Algae Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
39. Cooling Tower Wood Deterioration ........................................ 300
40. Air Conditioning Systems .......................... .' ...................... 311
41. Industrial Waste Treatment ............................................... 319
CONTROL WATER ANALYSES AND THEIR INTERPRETATION

42. Methods and Equipment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 329
43. Expression of Analytical Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
44. Acid (Free Mineral) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 338
45. Alkalinity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
46. Alkalinity (Hydrate) ..................................................... 343
4 7. An1monia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 345
48. Calcium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
49. Carbon Dioxide ......................................................... 349
50. Chloride . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
51. Chlorine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
52. Chron1ate ............................................................... 359
53. Color .................................................................. 363
54. Conductance (Specific) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
55. Hardness ............................................................... 370
56. Hydrazine .............................................................. 374
57. Iron ................................................................... 377
58. Manganese . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 380
59. Nitrate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382
60. Nitrite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
61. Oxygen (Dissolved) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
62. pH .................................................................... 390
63. Phosphate 394
64. Silica . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
65. Solids (Total, Dissolved, Suspended) ....................................... 402
66. Sulfate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404
67. Sulfite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 407
68. Turbidity 409
69. Composition of Prepared Reagents ......................................... 411
Conversion Factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415
Index .................................................................. 417
8 BETZ HANDBOOK
1
Industrial Water Conditioning
What Is Industrial Water Conditioning?. . . . . . . . 9

External Treatment . . . . . . . . . . . . . . . . . . . . . . . 9
Internal Treatment . . . . . . . . . . . . . . . . . . . . . . . . 9
Water Conditioning Control. . . . . . . . . . . . . . . . . 10
Industrial Water Problems. . . . . . . . . . . . . . . . . . 10
used directly or indirectly in an

ATER
I
W indu~trial process, is classed as indus-
trial water. The use of w~ter .in boil:rs for
steam generation is an obvwus mdustnal us~.
Cooling water, either o~ a once-t~roug~ basiS /
or with cooling towers, Is a promment mdus-
trial use. Water is essential to large air con-
ditioning systems. But, water has a multitude
of industrial uses-some as commonplace as
washing of equipment-others specific t~ an
individual industry or plant, such as use m a
high pressure spray for debarking logs. Water
is incorporated in many products and foods.
Water is used as a carrier in industrial proc-
esses and as a convenient means of disposing
of wastes. Figure 1-1 • External Treatment
by Hot Process Softening
WHAT IS
INDUSTRIAL WATER CONDITIONING? use of equipment, which may include hot or
cold process lime-soda softeners, zeolite and
To many people, industrial water condition- other ion exchange systems; deaerators, filters,
ing is shrouded in an air of mystery. Much of clarifiers, etc. Such equipment is used for the
this confusion is due to a lack of understand- purpose of reducing hardness and alkalinity,
ing of what industrial water conditioning is eliminating dissolved oxygen and for the
. . . and what it is supposed to accomplish. removal of suspended solids.
Basically, industrial water conditioning em- The use of external treatment should be as
braces the broad field of fitting water to the complete as can be justified by the severity
job. Its purpose is twofold. One-it involves of the water problem and engineering
removing or minimizing undesirable charac- economies.
teristics of water, such as removing hardness Regardless of the purity of the water pro-
by softening to avoid boiler scale. Two-it vided by the use of such equipment, addi-
involves adding desirable properties to water, tional internal chemical treatments normally
such as adding chromate ion to give cor- are required for complete protection against
rosion inhibitory properties. scale, corrosion, and a host of other potential
Industrial water conditioning may be sources of trouble.
grouped into two major parts or spheres of There are often cases where an investment
operation-external treatment and internal in external treatment equipment is not war-
treatment. ranted because of non-critical requirements,
a good water supply, small size of the plant,
EXTERNAL TREATMENT or a combination of these reasons. In such
cases, internal chemical treatment alone, will
This phase of industrial water conditioning produce successful operating results.
has many names: preliminary treatment,
pre-treatment, indirect treatment, primary
INTERNAL CHEMICAL TREATMENT
purification, etc. Whatever it is called, it
means doing something to a water to make it Like external treatment, this phase of indus-
more suitable for its intended application trial water conditioning is known by several
before it reaches the point of use. names: secondary treatment, after-treatment,
External treatment usually requires the direct treatment, etc. In this case, it means
10 BETZ HANDBOOK
doing something to a water at the point of use ing systems. These cas~s include the use of
to make it suitable for its intended application. chemicals at various points in process appli-
In boiler water systems, the objective of cations to impart desirable properties to
internal treatment is to make possible the a water.
control of scale formation and corrosive Water conditioning is in many respects
action. Steam purity and caustic embrittle- unlike any other field of engineering. It de-
ment must be controlled. It also may be mands an unusual variety of talents including
necessary to use internal chemical treatment · those of the chemical, mechanical, corrosion
to prevent return line corrosion due to dis- and sanitary engineer, microbiologist, physi-
solved gases liberated in the boiler system. cal chemist and bacteriologist.
Internal chemical treatment is used also
in cooling water systems for the prevention WATER CONDITIONING CONTROL
of scale and corrosion, as well as conditions
brought about by biological growths. Proper Control water analyses are an essential part
application of internal chemical treatment of any industrial water conditioning program.
results in improved heat transfer by eliminat- The processes of corrosion, scaling, slime for-
ing these insulating deposits. mation and the like are dynamic, not static.
The treatment measures employed for their
There are cases, too, where internal chemi-
control must be adjusted frequently on the
cal treatment is used in the solution of specific
basis of control water analyses. Such analyses
water problems other than in boiler and cool-
need not determine every ion or substance
present in the water, but only those required
for control of treatment process. For example,
in sodium zeolite softening the important
tests for control purposes are hardness and
chloride. The hardness test is employed to
determine the end of the softening cycle and
the chloride test to determine the end of the
rinse cycle.
INDUSTRIAL WATER PROBLEMS
Following is a list of the more common water

problems that plague industry. Each industry
has its individual problems. This list is not
intended to be complete, but only to indicate
the great variety of problems common to
most industries.
Boiler Water Problems

Scale Deposits
Excessive Sludge
General Corrosion
Pitting
Carryover
Embrittlement
Feedline and Heater Deposits
Figure 1-2 • Internal Treatment High Blowdown Rate
Chemical Tanks and Pumps Carbonized Oil
Steam and Condensate Problems
Carbon Dioxide Grooving

Oxygen Pitting
Turbine Blade Deposits
Superheater Fouling
Oil Contamination
Condenser Leakage
Process Contamination
Cooling and Process Water Problems
Deposition of Scale
General Corrosion
Fungus Attack of Cooling Tower Wood Figure 1-3 • Plant Operator
Product Contamination Conducting Control Ana(yses
Pitting and Tuberculation
Slime and Algae Deposits
Red Water
Preliminary Treatment Problems General Problems

(removal of impurities)
Poor Quality Water Supply
Turbidity, Mud, Silt Waste of Water-Lack of Re-use
Excessive Color Inadequate Pretreatment Equipment
Iron and Manganese Improper Control of Existing Equipment
Soluble Silica Unsatisfactory Chemical Feeding
High Hardness Stream Pollution
Acidity or Alkalinity Waste Disposal
Objectionable Gases-C0 2 , H 2 S, NH 3 High Costs
High Solids Improper Over-all
Microorganisms Water Conditioning Program
BETZ HANDBOOK
12
2
Water and its Impurities
Sources of Water . ......................... 13

Rain Water. . . . . . . . . . . . . . . . . . . . . . . . . . . 14
Sw:face Water. . . . . . . . . . . . . . . . . . . . . . . . . . 14
Ground Water. . . . . . . . . . . . . . . . . . . . . . . . . . 14
Impurities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
F water as it occurs in nature were "pure" of hydrogen and one volume of oxygen, chem-
I water and nothing else there would be no
need for water analyses and water condition-
ically combined.
ing. Whatever the source, water always con- SOURCES

tains impurities in solution or suspension. It
is the determination of these impurities that The various sources of water can be con-
makes water analyses necessary and the conveniently classified as:
trol of these impurities that makes water con- Rain water.
ditioning essential. Surface water (streams, ponds, lakes and
Pure water is tasteless, colorless and odor- impounded reservoirs) .
less and is one of the most universal solvents. Ground water (from springs, shallow wells
The chemical formula for water is H 2 0 (or and deep wells) .
HOH) and it is composed of two volumes Water is evaporated from the surface of the
Figure 2-1 • Calcium Sulfate Crystalline Scale in 5 inch Line
14 BETZ HANDBOOK
Figure 2-2 • Pipe Section Pitted from Dissolved Oxygen
earth by heat furnished from the sun. Water organic matter and may pollute the water
precipitated from the atmosphere, high above with animal waste. Swamps may discharge
the earth's surface, is chemically pure. As their waters during floods and carry decay
water descends through the air oxygen, nitro- and vegetable matter, color and microscopic
gen and carbon dioxide (the normal gases organisms into the stream. In addition, sur-
present in the atmosphere) are dissolved in face waters are exposed to pollution by ani-
the falling water in proportion to their partial mals and humans, the sewage of cities, and
pressure in the atmosphere. Rain water also the wastes of industry.
encounters dust, smoke and fumes which are SuRFACE WATERS. Surface waters obtained
dissolved or retained in suspension. Bacteria from ponds, lakes and impounded reservoirs
and the spores of microscopic organisms may are similar to those from flowing streams.
be picked up. However, within these large bodies of water
RAIN WATER. Rain water in descending considerable self-purification may take place
through the air and in percolating through and the water quality may change to a large
the upper layers of soil absorbs carbon di- extent.
oxide with which it forms carbonic acid. GRoUND WATER. Ground water obtained
This action increases the solvent power of from shallow wells may be hard or soft de-
the water so that it dissolves a certain pending upon the mineral characteristics of
amount of the mineral matter of the soil the surrounding area. Natural filtration
or rock with which it comes in contact. through the sand will usually provide a water
Water obtained from surface streams may be relatively free of turbidity and with low
rendered turbid by the presence of clay and organic matter. The water from shallow
silt. Agricultural land may contribute to the wells is usually softer than deep wells.
The water from deep wells usually contains

relatively high concentrations of dissolved
minerals. These waters are usually clear and
colorless due to the filtration taking place
through the layers of sand.
Spring water is usually similar in character-
istics to water obtained from shallow or deep
wells in the surrounding area. However,
springs are more likely to be affected by the
normal surface pollution due to their rela-
tively shallow origin.
IMPURITIES
The impurities that may be present in a

water supply can be divided into suspended
and dissolved solids. Suspended solids are
those which do not dissolve in water and
which can be removed or separated by filtra-
tion. Examples of suspended solids are mud,
clay or silt. Dissolved solids are those which
naturally dissolve in the water and which
therefore cannot be removed by filtration. The
presence of hardness or chloride in a water
supply is an example of dissolved solids. Gases
may also dissolve in water, but unless they
combine chemically with other impurities also
in solution, these gases will be expelled from
the water on boiling and are not considered
as dissolved solids.
Fig. 2-4 is a listing of the analytical deter-
minations made in the examination of most
natural waters. Illustrated are the chemical
formula, if one can be given, the difficulties
commonly caused by the presence of that con-
stituent and the usual means of treatment to
prevent these difficulties. This listing covers
the usual determinations made in the analysis
of a water prior to treatment. Of course, for
certain special water uses, additional analyses
are necessary. For example, boron is of im-
portance in irrigation waters because of its
effect on the growth of certain crops. Boron
possesses little significance industrially and
usually is not determined in a water analysis.
Similarly, water analyses are made for certain
constituents that are added to water as a
:nethod of treatment. For example, chromate Figure 2-3 • Delignification
IS employed as a corrosion inhibitor in cooling of Cooling Tower Wood
16 BETZ HANDBOOK
water systems, phosphate as a precipitating the constituents that normally are present
agent in boiler water and sulfite is used to naturally in untreated water supplies and is
control oxygen corrosion. Analyses must be not intended as a complete listing of all of
made of such treating agents for control pur- the analytical determinations that should be
poses. Figure 2-4, however, deals only with made where specific problems are involved.
Fig. 2-4-Common Impurities Found in Water
CONSTITUENT CHEMICAL FORMULA DIFFICULTIES CAUSED MEANS OF TREATMENT

TURBIDITY . . . . . . None-expressed in Imparts unsightly appearance to Coagulation, settling and filtra-
analysis as units. water, Deposits in water lines, proc- tion.
ess equipment, boilers, etc. Inter-
feres with most process uses.
COLOR . . . . . . . . None-expressed in May cause foaming in boilers, Hin- Coagulation and filtration. Chlor-
analysis as units. ders precipitation methods such as ination. Adsorption by activated
iron removal, hot phosphate soften- carbon.
ing. Can stain product in process
use.
HARDNESS ..... Calcium and mag- Chief source of scale in heal ex- Softening. Distillation. Internal
nesium salts exchange equipment, boilers, pipe boiler water treatment. Surface
lines, etc. Forms curds with soap, active agents.
pressed as CaC03.
interferes with dyeing, etc.
ALKALINITY ..... Bicarbonate (HC03), Foaming and carryover of solids lime and lime-soda softening.
carbonate (C0 3), and with steam, Embrittlemenl of boiler Acid treatment. Hydrogen zeolite
steel. Bicarbonate and carbonate softening, Demineralization. De-
hydrate (OH), ex-
produce C02 in steam, a source of alkalization by anion exchange.
pressed as CaC03. corrosion, Distillation.
FREE MINERAL H2S04, HCI, etc. ex- Corrosion Neutralization with alkalies.
ACID ........ . pressed as CaC03.
CARSON Corrosion in water lines and par- Aeration. Deaeration. Neutraliza-

DIOXIDE ticularly steam and condensate tion with alkalies. Filming and
lines. neutralizing amines.
pH ........... . Hydrogen ion con- pH varies according to acidic or pH can be increased by alkalies
centration defined as: alkaline solids in water. Most natu- and decreased by acids.
·ral waters have a pH of 6.0-8.0.
1
pH= log ------c:---
(H +)
SULFATE ....... Adds to solids content of water, Demineralization. Distillation.

but, in itself, is not usually signifi·
cant. Combines with calcium to form
calcium sulfate scale.
CHLORIDE.,.,., Cl- Adds to solids content and in- Demineralization. Distillation.
creases corrosive character of
water.
NITRATE ... , .. , Adds to solids content, but is not Demineralization. Distillation.
usually significant industrially. High
concentrations cause methemoglob-
inemia in infants. Useful for con-
trol of boiler metal embrittlement.
FLUORIDE ... , , . F- Cause of mottled enamel in Ieeth. Adsorption with mangnesium hy-
Also used for control of dental de- droxide, calcium phosphate, or
cay, Not usually significant indus- bone black. Alum coagulation.
trially.
Fig. 2-4-Common Impurities Found in Water (cont.>
CONSTITUENT CHEMICAL FORMULA DIFFICULTIES CAUSED MEANS OF TREATMENT
SILICA . . . . . . . . . Si02 Scale in boilers and cooling water Hot process removal with magnes-
systems. Insoluble turbine blade ium salts. Adsorption by highly
deposits due to silica vaporization. basic anion exchange resins, in
conjunction with demineraliza-
tion. Distillation.
IRON ......... Fe ++(ferrous) Discolors water on precipitation. Aeration. Coagulation and filtra-
Fe+++ (ferric) Source of deposits in water lines, tion. Lime softening. Cation ex-
boilers, etc. Interferes with dye- change. Contact filtration. Surface
ing, tanning, paper mfr., etc. active agents for iron retention.
MANGANESE . . . Mn + + Same as iron. Same as iron.
OIL . . . . . . . . . . . Expressed as oil or Scale, sludge and foaming in boil- Baffle separators. Strainers. Co-
chloroform extract- ers. Impedes heat exchange. Un- agulation and filtration. Diatoma-
desirable in most processes. ceous earth filtration.
able matter.
OXYGEN . . . . . . . 02 Corrosion of water lines, heat ex- Deaeration. Sodium sulfite, Cor-
change equipment, boilers, return rosion inhibitors.
lines, etc.
HYDROGEN Cause of "rotten egg" odor, Cor- Aeration. Chlorination. Highly

SULFIDE ...... . rosion. basic anion exchange.
AMMONIA . . . . . NH3 Corrosion of copper and zinc alloys Cation exchange with hydrogen
by formation of complex soluble zeolite. Chlorination. Deaeration.
ion.
CONDUCTIVITY Expressed as mi- Conductivity is the result of ioniz- Any process which decreases dis-
cromhos,specific con- able solids in solution. High con- solved solids content will decrease
ductivity can increase the corrosive conductivity. Examples are de-
ductance.
characteristics of a water. mineralization, lime softening.
DISSOLVED None "Dissolved solids" is measure of Various softening process, such as
SOLIDS ....... total amount of dissolved matter, lime softening and cation ex-
determined by evaporation. High change by hydrogen zeolite, will
concentrations of dissolved solids reduce dissolved solids. Deminer-
are objectionable because of proc- alization. Distillation.
ess interference and as a cause of
foaming in boilers.
SUSPENDED None "Suspended Solids" is the measure Subsidence. Filtration, usually
SOLIDS •...... of undissolved matter, determined preceded by coagulation and
gravimetrically. Suspended solids settling.
plug lines, cause deposits in heat
exchange equipment, boilers, etc.
TOTAL SOLIDS . . None "Total Solids" is the sum of dis- See "Dissolved Solids" and "Sus-
solved and suspended solids, depended Solids."
termined gravimetrically.
BETZ HANDBOOK
18
3
Aeration
Theory Involved. . . . . . . . . . . . . . . . . . . . . . . . . . 19
Basic Aeration Methods. . . . . . . . . . . . . . . . . . . . 19
Water Fall Aerators . .................... 20
Air Diffusion . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Limitations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
AERATION is a process which consists of bon dioxide and no measurable amount of

1"1. intimately admixing water and air in hydrogen sulfide.
some manner. This mixing may be accom- Both carbon dioxide and hydrogen sulfide
plished in the form of a thin film, drops or are relatively soluble in water. At 20 C and
spray. Aeration is based on the establishment sea level, carbon dioxide is soluble to the ex-
of a state of equilibrium between the gases tent of appro~imately 1700 ppm while hy-
present in the water and those present in drogen sulfide solubility is approximately
the atmosphere. 3900 ppm. The partial pressure of each of
Aeration is employed primarily for the these gases in a normal atmosphere, how-
removal of undesirable gases from water. ever, is practically "zero." Consequently, the
Carbon dioxide and hydrogen sulfide are establishment of a state of equilibrium be-
corrosive gases and, when present in appreci- tween water and air by means of aeration
able quantities, must be removed to avoid will result in saturation of the water with
corrosion of lines, pumps and equipment. Al- nitrogen and oxygen and the practically
though aeration results in saturating water complete elimination of carbon dioxide and
with another corrosive gas - oxygen - this hydrogen sulfide from the water.
process is the one most commonly used for The basic requirement in securing removal
the elimination of carbon dioxide and hydro- of gases by aeration is the establishment of
gen sulfide. a state of equilibrium between the gases in
In some cases, the purpose in removing the water and those present in the surround-
carbon dioxide by aeration is simply to re- ing atmosphere. The removal of gases by
duce chemical treatment requirements where aeration is favored by increase in tempera-
the water is subsequently lime softened. Car- ture, increase in aeration time, increase in
bon dioxide can be removed more economi- the volume of air in contact with the water
cally by aeration than by chemical precipita- and increase in the surface area of water
tion with lime. exposed to the air. The efficiency of aeration
In the removal of iron and manganese is greater where the concentration of the
from water, aeration is frequently employed. gas to be removed is high in the water and
The oxygen introduced by aeration encour- low in the atmosphere. For example, aeration
ages the chemical oxidation of iron and efficiency is higher with a water containing
manganese, thus aiding in their precipitation 100 ppm carbon dioxide, than with one con-
as the insoluble hydroxides. taining only 10 ppm.
Aeration may be employed in potable
water treatment for the removal of certain BASIC AERATION METHODS
tastes and odors, such as those resulting
from the decomposition of algae. In sewage - Aeration is primarily a mechanical process
treatment, aeration i~ used to promote the for which two general methods are em-
biochemical oxidation of organic matter. ployed. The first method is that where the
water falls through the air as with the use of
THEORY INVOLVED spray nozzles, cascade devices and forced
draft type aerators. This method of aeration
The removal of gases from water by aera- is the most common in industrial water con-
tion follows Henry's Law which, in brief, ditioning.
states that the solubility of a gas in water is The second general method of aeration is
?irectly proportional to its partial pressure where air is bubbled up through a solid body
m the surrounding atmosphere. At 20 C and of water, as in a tank, and is known as the
at sea level, water in equilibrium with the air diffusion method. In industrial water
atmosphere will contain approximately 15.8 conditioning this method is usually confined
ppm nitrogen, 9.4 ppm oxygen, 0.5 ppm car- to the treatment of relatively low flows.
20 BETZ HANDBOOK
WATER FALL AERATORS. Many variations of the advantage of smaller space require-
the water fall principle are in use. The sill].- ments. Coke tray aerator~. are widely used in
plest is the use of vertical risers to discharge iron and manganese removal since a catalytic
the water by free fall into a basin. The effect is secured by contact of the iron-
risers can operate on the available head of bearing water with freshly precipitated iron
water and efficiency of aeration is increased sludge. As shown by Figure 3-1, these units
by making the fall as great as practical. consist of a series of coke filled trays through
The addition of steps or shelves to break up which the water percolates with additional
the fall and to spread the water into thin aeration supplied during the free fall from
sheets or films increases contact time and acts one tray to the next. Wood slat tray aerators
as a further refinement. Spray nozzles are are similar to small atmospheric cooling
another variation of this principle and can towers. The tray slats are staggered so as to
provide efficient removal of dissolved gases. break up the free fall of water and to spread
However, space requirements frequently limit out the water surface before finally dropping
the usefulness of this type aeration in indus- into a basin.
trial water conditioning. Forced draft aerators, as illustrated by
Coke tray aerators and wood slat aerators Figure 3-2, are favored for many industrial
are relatively similar in design and possess water conditioning purposes. These units
Courtesy lnfilco, Incorporated

Figure 3-1 • Coke Tray Aerator
WATER INLET
INLET CHAMBER
STAGGERED
SLAT TRAYS
AIR li'LET
Courtesy Pfaudler Permutit Inc.
Figure 3-2 • Forced Draft Aerator
provide horizontal wood slat trays which by pumping air into water through perfo-
divide the falling water into small streams. rated pipes, strainers, porous plates or tubes.
Air is forced up through the unit by a blower Aeration by diffusion is theoretically superior
with uniform air distribution across the en- to water fall aeration inasmuch as a fine
tire cross section. Air flow is counter-current bubble of air rising through water will con-
to the fall of water and positive air circula- tinually be exposed to fresh liquid surfaces,
tion prevents any stagnant areas. Because of providing maximum water surface per unit
these features, forced draft aerators are more of air. Also, the velocity of bubbles ascending
efficient for gas removal and require less through the water is much slower than the
space for a given capacity. velocity of free falling drops of water, pro-
AIR DIFFUSION. Air diffusion is accomplished viding longer contact. However, for gas re-
22 BETZ HANDBOOK
moval, air requirements by the diffusion Aeration has also been applied as an aid in
method are higher than for odor removal rapid mixing of chemicals, and in backwash-
and the cost of the air supply is a factor ing filters as well as for the esthetic effect in
restricting application of this type of aera- water works plants.
tion in industrial water conditioning.
LIMITATIONS
APPLICATIONS Theoretically, at 20 C it is possible to re-

duce the carbon dioxide content of the water
In industrial water conditioning one of the to 0.5 ppm by means of aeration to equilib.
major applications of aeration is in the re- rium conditions. Practically, it is not eco.
moval of the corrosive characteristics im- nomical to attempt such complete gas re-
parted to water by the presence of carbon moval and reduction of carbon dioxide to 10
dioxide and hydrogen sulfide. On many oc- ppm is normally considered satisfactory.
casions aeration is used to remove the carbon Although removal of free carbon dioxide
dioxide liberated by a treatment process. For increases the pH of a water and renders it
example, in boiler feedwater conditioning, less corrosive from this standpoint, aeration
it is common practice to acid treat the ef- also results in the saturation of the water
fluent from a sodium zeolite softener in order with dissolved oxygen. In some cases, where
to reduce the alkalinity of the boiler feed- the original oxygen content is already quite
water. Carbon dioxide is produced as a re- high, no undesirable effect is thus produced,
sult of the acid treatment and aeration is However, in the case of a well supply, high
employed to rid the water of this corrosive in carbon dioxide but devoid of oxygen,
gas. Similarly, when the effluents of hydrogen aeration may simply exchange one corrosive
and sodium zeolite units are blended, the gas for another.
carbon dioxide formed is eliminated by aera- The efficiency of aeration is greatest where
tion. In anion-cation exchange, aeration is the initial concentration of the gas to be re-
similarly used. Forced draft aerators are moved_ is considerably above its equilibrium
usually employed in these applications. value. Therefore, with waters containing only
Aeration is also applied for the removal of a small amount of carbon dioxide, lime treat-
methane from certain ground waters. Since ment is usually more efficient.
methane constitutes- an explosion hazard on Aeration alone will not secure the com·
release in confined spaces, aeration is em- plete removal of hydrogen sulfide, but must
ployed for its removal. Electric motors in be combined with pH reduction or chlorin·
the vicinity are undesirable and forced draft ation.
aerators are not employed in this appli-
Where odors are not due to the gases of
cation.
decomposition or to volatile oils, aeration
Iron and manganese removal can be se- will not remove these odors. Tastes and
cured with a variety of methods. In all but odors due to industrial wastes, phenols and
a few of the processes used, aeration plays a cresols are unaffected by aeration. The me·
part since it is a convenient method of sup- dicinal taste imparted to water by the chlo-
plying the oxygen required for the oxidation rination of water containing phenols and ere-
of iron and manganese. However, aeration sols is not removed by aeration.
alone should not be considered a universally
effective method for the removal of iron and
REFERENCES
manganese, but rather only as a part of the
treatment steps required. "Aeration of Water", Journal, Am. Water Works
In municipal water treatment practice, Assoc., Vol. 47, pp. 873-885 (1955)
M. L. Riehl, "Hoover's Water Supply and Treat-
aeration has been employed primarily for ment", 8th Ed., pp. 18-19, National Lime Assoc.,
the removal of odors due to volatile materials. Washington, D.C. (1957)
4
Subsidence
Theory Involved. . . . . . . . . . . . . . . . . . . . . . . . . . 24
Equipment Employed. . . . . . . . . . . . . . . . . . . . . . . 25
Applications and Limitations . . . . . . . . . . . . . . . . 25
24 BETZ HANDBOOK
UBSIDENCE is the process which permits The viscosity of the water is also a major
S the settling of suspended substances from
water.
factor influencing settling yelocity and the
water viscosity is, in turn, dependent on the
The main purpose of subsidence is to ef- temperature. For example, the rate of set-
fect clarification of a water by permitting tling at 0 C is only 43% of the settling rate at
the settling of suspended matter and conse- 30 C. It is for this reason that greater effi-
quent reduction in turbidity. Subsidence is ciency in turbidity removal by subsidence
of value chiefly in the removal of coarse may be obtained during the summer months
particles which settle rapidly and which can when water temperatures are higher. During
be more economically removed in this fashion the winter months, with normally lower
than by coagulation. A secondary result of water temperatures, the rate of settling will
su~sidence is removal of bacteria. Percentage be reduced and lower efficiency results or
bacteria removal generally closely parallels else it may be necessary to increase the time
turbidity removal. While dependent on many permitted for subsidence.
factors, turbidity and bacteria removal by Settling velocities may be calculated from
subsidence may vary from 30% to 80%. Stoke's Law.
y = 18.5 D2 (51- S2)
THEORY INVOLVED
z
All particles heavier than water possess a
tendency to settle due to the influence of Where:
gravity. The time required for settling is V =Velocity of fall in feet per second.
dependent on a number of factors such as D = Diameter in inches of particle.
the weight, shape and size of the particle to-
S1 = Density of particle in pounds per
gether with the viscosity and frictional re-
cubic foot.
sistance of the water.
S2 = Density of fluid in pounds per cubic
To illustrate the effect of particle size, the
foot.
following tabulation is presented:
Z = Viscosity in centipoises.
VELOCITIES AT WHICH PARTICLES OF
In employing this equation it is assumed
SAND AND SILT WILL SUBSIDE IN STILL WATER
that the particles are spherical, falling under
lOG (50F) viscous resistance in an infinitely large body
Specific Gravity = 2.65 of fluid.
The use of Stoke's Law and tabulation of
Diameter of
Particle, Order of Time Required settling velocity is based on still water. Theo-
lnmm Magnitude to Settle 1 Foot retically, velocity of flow in a horizontal
plane will not affect settling velocity of the
10.0 Gravel 0.3 seconds
1.0 Coarse Sand 3 seconds
suspended particles because their downward
0.1 Fine Sand 38 seconds movement is caused by gravity. However,
0.01 Silt 33 minutes velocity of flow in a vertical plane and con-
0.001 Bacteria 35 hours
0.0001 Clay Particles 230 days vection currents directly tend to prevent
0.00001 Colloidal Particles 63 years settling. Design of subsidence basins is there-
fore such as to keep such interferences to a
The specific gravity of suspended particles minimum and to assure continued quiescence
is one of the main factors governing the rate of the water.
of settling. Particles of high specific gravity A study of the tabulation of settling veloc-
will naturally settle at a faster rate than ities will indicate that for industrial pur-
those of a specific gravity only slightly poses the practical limit on solids removal by
greater than that of water. subsidence is for particles of a diameter of
0.01 mm or more. Smaller particles have It is necessary in basin design to make

such a low rate of settling that the time re- allowance for the sludge accumulated be-
quired is greater than can be practically al- tween cleanings. Twenty-five per cent of the
lowed. In large storage reservoirs having a tank volume is usually provided for this
retention period of weeks and months, it is purpose. When proper settling is no longer
possible to effect removals of particles of obtained, due, to sludge accumulation, the
less than 0.01 mm diameter. tank is drained and flushed.
In the larger tanks and particularly where
EQUIPMENT EMPLOYED there is a considerable amount of sludge,
mechanical means are provided frequently
Subsidence may be employed by using a for continuous sludge removal. Mechanically
simple tank which is filled with the turbid operated rakes and scrapers travel just above
water and allowed to stand. After the sedi- the floor and collect sludge to the draw off
ment has settled, the clarified water may be point for continuous removal. The design
decanted off and the settled solids removed and speed of operation of the scrapers must
from the bottom. The tank is then ready to be such that a minimum of disturbance is
be refilled for another cycle. This method is presented to the desired settling of the sus-
designated as a "fill and draw" system, and pended solids.
the settling rates will correspond to those
previously tabulated. APPLICATIONS AND LIMITATIONS
Continuous methods of subsidence are in
greater use in plants of recent design. This Waters containing turbidity due to relatively
is accomplished by passing the water through large size or high density suspended solids
a basin or tank of such size, shape and de- may be clarified to a suostantial degree by
sign that the velocity of the water is reduced subsidence. This applies to natural waters,
to a point sufficiently low to permit settling sewage and industrial trade wastes. Subsi-
of the readily settleable solids. Clarified dence is an economical method of obtaining
water is removed from the tank continuously clarification where it is applicable and in
from an overflow, the settled solids remain- such cases is employed prior to coagulation.
ing in the bottom of the tank. Suspended solids of low specific gravity or
In smaller plants of less than 1 million gal- those of fine particle size cannot be removed
lons per day, usually only one settling basin by subsidence within a reasonable period of
is employed. In larger plants, two or three time. Increase in basin size or retention
basins in series are used. The use of three period may not be feasible. In such cases,
basins is favored to permit the removal of chemical coagulation is necessary to increase
one basin from service for cleaning, while the size of the particles and render them more
two basins remain on line. Where continuous easily settleable.
sludge removal is used, one basin is adequate
even in plants of greater capacity. REFERENCES
Basins are usually rectangular and if only
one basin is employed, the use of a square H. E. Babbitt and J. J. Doland, "Water Supply
Engineering", 3rd Ed., pp. 499-509, McGraw-Hill
layout provides the most economical con- Book Co., Inc., New York, N.Y. (1939)
struction. However, in many cases the basin M. N. Baker, "The Quest for Pure Water", pp.
dimensions must be made to fit the space 286-298, Am. Water Works Assoc., New York,
N.Y. (1948)
available. Sloping bottoms are provided to M. L. Riehl, "Hoover's Water Supply and Treat-
facilitate sludge removal by flushing with ment", 8th Ed., pp. 37-39, National Lime Assoc.,
high pressure jets. Depth of settling basins Washington, D.C. (1957)
will vary usually over the range of 12 to 20 "Water Quality and Treatment", 2nd Ed., pp. 131-
162, Am. Water Works Assoc., New York, N.Y.
feet with an average of 16 feet. ( 1950)
BETZ HANDBOOK
26
5
Coagulation
Theory of Coagulation . .................... 27

Types of Equipment . . . . . . . . . . . . . . . . . . . . . . . 28
Coagulants. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Alum ........ ·........................ 30
Sodium Aluminate . . . . . . . . . . . . . . . . . . . . . . . 30
Copperas. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
Ferric Coagulants. . . . . . . . . . . . . . . . . . . . . . . 31
Magnesium Oxide . ...................... 31
Coagulant Aids . . . . . . . . . . . . . . . . . . . . . . . . . . . 33
OAGULATION is that process whereby finely particles into larger masses, the suspended
C divided particles of turbidity and color,
capable of remaining in suspension indefi-
turbidity particles, bacteria, microorganisms,
etc., become enmeshed by the floc. Any sus-
nitely are combined by chemical means into pended particles carrying a negative charge
mass:s sufficiently large to effect rapid are more easily enmeshed by the floc and it is
settling. believed that color is removed in this fashion.
Ordinary subsidence usually will not pro- For each coagulant there is an optimum
duce a water sufficiently low in turbidity to pH zone for best coagulation. It is desirable
be suitable for domestic purposes or for most to control the pH to the point of maximum
industrial requirements. Additional settling precipitation of the coagulant, equivalent to
time is not usually of benefit because of the the minimum solubility of the floc. The best
small size of the suspended particles and their pH for this precipitation is dependent not
slow settling rate. Filtration, without coagu- only on the coagulant, however, but also upon
lation, will not remove the fine particles of the mineral characteristics of the water. For
turbidity if a filter medium sufficiently coarse example, while alum will coagulate best at a
for modern filtration rates is employed. Coagu- pH of approximately 5.5 for pure distilled
lation is required to agglomerate the sus- water, for most natural waters the ideal pH
pended particles, thus making the water more is higher and may range from pH 5.5 to 8.0.
readily filtered and also causing much of the The temperature of the water also affects
coagulated matter to settle out prior to filtra- the process of coagulation and at tempera-
tion. tures near the freezing point increased mixing
time as well as increased coagulant doses may
THEORY OF COAGULATION be required.
In general, successful coagulation involves
Coagulation reactions can be written very three main factors:
simply showing the use of aluminum or iron
1. The presence of a minimum quantity of
salts precipitating to form a floc of aluminum
aluminum or iron ions to form an in-
hydroxide or iron hydroxide. While these re-
soluble floc.
actions will be shown for the sake of simplic-
2. The presence of a strong anion such as
ity, coagulation is a much more complex
sulfate or chloride.
phenomenon than thus illustrated.
3. The pH of the water must be controlled
It is thought that the precipitation of an within a definite range.
aluminum or iron coagulant produces finely
It is not possible to predict from the analy-
divided precipitates of the hydrous oxide.
sis of a water which will be the best coagulant
These precipitates are positively charged and
to use, the quantities that will be required
thus repel each other. Neutralization of the
or the optimum pH for control of the process.
positive charges by negative ions such as sul-
For most efficient and economical coagula-
fate and -chloride in the water causes coales-
tion it is necessary that laboratory tests be
cence of the fine particles, thus forming a
made with different coagulants at varied con-
gelatinous porous precipitate of large volume.
centrations and varied pH levels. Laboratory
Flocculation, as practiced in water treat- stirring devices should be used in these tests
ment, is the further increase in size and volume to provide uniform test conditions. While
of the precipitate caused by bringing together previous practical experience with a given
by gentle agitation of the water a number of water will aid in selecting coagulant dosage
t~e individual precipitated hydrous oxide par- and optimum pH, in order to consistently
ticles together with enmeshed turbidity to secure the desired results it is necessary to
form larger and more readily settleable change coagulant dosage and pH control to
masses.
compensate for varying raw water character-
In the flocculation of the hydrous oxide istics.
28 BETZ HANDBOOK
Coagulation problems vary greatly not only water of the most desirable characteristics.
because of varying raw water characteristics, The presence of iron and manganese in the
but also because of the ultimate use of the raw supply introduces additional factors af-
water. For potable systems the removar of fecting coagulant choice and dosage. Indus-
bacteria and microorganisms is quite impor- trial wastes present in the raw water also may
tant and these requirements may affect coag- complicate the coagulation problem.
ulation control. For example, a slight excess of
coagulant may be used over and above that
necessary to produce a relatively clear water TYPES OF EQUIPMENT
in order that bacteria removal may be in-
creased in the sedimentation basin, thus re- Three essential steps in the proper control of
ducing the residual chlorine concentration re- coagulation processes prior to filtration are:
quired in the finished water and minimizing 1-Mixing
taste and odor problems. These considera- 2-Flocculation
tions would not be important in the prepara- 3-Sedimentation
tion of boiler feedwater where clarity of the It is important that the coagulant be
filter effluent at minimum coagulant dosage mixed rapidly with the water to be treated in
would be the controlling factor. In boiler feed- order to secure efficient coagulation at mini-
water conditioning, however, softening is usu- mum chemical dosage. By rapid or "flash"
ally important and coagulation may be com- mixing the coagulant is uniformly distributed
bined with softening to achieve a finished throughout the water before precipitation is
Courtesy Graver Water Conditioning Co.
Figure 5-1 • Sludge Contact Clarifier
CHEMICAL
SLURRY POOL INDICATED BY SHADED AREAS
Courtesy lnfilco, Incorporated
Figure 5-2 • Upjfow Clarifier
completed. Mixing is secured usually by six hours are usually provided. It is important
motor-driven propellers placed in small basins to so design the basins as to avoid short-
where the coagulant is introduced or by add- circuiting and to keep the velocity of flow
ing the coagulant at some point of turbulence. quite low. Modern basins are usually designed
Following mixing, the next step is to secure for continuous mechanical desludging rather
flocculation by bringing the small particles than the periodic draining and manual de-
together under the influence of gentle mix- sludging employed in older designs.
ing, thus permitting them to increase in size Upflow clarifiers which combine in one unit
and therefore to settle more rapidly in the the functions of mixing, flocculation and sedi-
sedimentation basin. Flocculation may be mentation have been installed in large num-
achieved by the use of baffied basins where bers in recent years. Particularly in industrial
change in the direction of flow will produce practice, these units are favored because of
the desired agitation. Lack of flexibility is in- the marked reduction in space requirements
herent in this method since the agitation se- and installation cost. These units are designed
cured will vary with the rate of flow. Mechani- for only one or two hours retention in com-
cal drive paddles or mixers are generally pre- parison with two to six hours for conventional
ferred for this purpose since their speed can sedimentation basins.
be controlled to conform with the rate of flow Figure 5-2 illustrates one design of an up-
and floc characteristics. With proper floccula- flow clarifier. In this unit the coagulant pre-
tion a uniform dense floc can usually be pro- cipitates in the presence of a recirculated
duced. slurry of previously formed precipitates. Floc
After mixing and floc formation, the treated formation is encouraged by this slurry re-
water is passed to a sedimentation basin which circulation and finely divided precipitates are
provides the retention time for the floc par- avoided. In the secondary mixing and reac-
ticles to settle. Retention periods of two to tion zone the raw water is circulated in con-
30 BETZ HANDBOOK
tact with three to five volumes of recirculated The use of sodium aluminate is advantageous
slurry, aiding in bringing treatment reactions under certain conditions because there is no
to equilibrium. Clarified water separates from need for additional alkali ..
the top of the slurry pool and sludge is auto- Sodium aluminate may be used in conjunc-
matically blown from the concentrator com- tion with alum, particularly for color removal.
partments. Units of this type provide efficient Alum alone, or in conjunction with acid, may
clarification of water in a single compact be used to coagulate color at a relatively low
unit. pH with sodium aluminate added after prim-
ary color removal for pH adjustment and
COAGULANTS secondary coagulation.
Sodium aluminate is employed frequently
ALUM Al2(S0 4 )a.l8H 2 0. The most widely
as a coagulant in conjunction with both hot
used coagulant is aluminum sulfate or filter
and cold lime-soda softening. In addition to
alum. Commercial filter alum contains
providing a lower turbidity of the softener
slightly less water of crystallization than in-
effluent, sodium aluminate assists in more
dicated by the above formula and corres-
complete precipitation of magnesium, thus
ponds to Al 2 (S04 ) 3 .14H20. This material
reducing the magnesium content of the sof-
is specified on the basis of 17% Al 20 3 con-
tened water. This effect on magnesium re-
tent. Alum is applicable as a coagulant over
duction is obtained to a greater extent in cold
a pH range of approximately 5.5 to 8.0
process softening, but is also of value in hot
and reacts with either natural or added al-
process softening.
kalinity to produce a floc of aluminum hy-
droxide as illustrated below. Ammonium alum is not widely used in co-
agulation processes, but is useful in some spe-
Alz(S04h + 3Ca(HC0 3h 2AI(0Hh + cial applications. In lump or crystal form am-
aluminum + calcium = aluminum + monium alum dissolves slowly and is employed
sulfate bicarbonate hydroxide
where alum pots are used for coagulant feed
3CaS0 4 + 6C0 2
in small installations as in swimming pool
calcium + carbon
sulfate dioxide practice. Truly proportional feeding is not
achieved. On the basis of aluminum content,
Alz(S04h + 3NazCOl + 3Hz0
ammonium alum costs approximately three
aluminum + sodium + water
times as much as filter alum. The ammonia
sulfate carbonate
content may be of value where chlorine-am-
2AI(0Hh + 3NazS04 + 3COz
monia treatment is used by eliminating the
aluminum + sodium + carbon
hydroxide sulfate dioxide need of a separate source of ammonia.
A(z(S04h + 3Ca(0Hh 2AI(0Hh + Activated alum is a commercial brand of
aluminum + calcium = aluminum + alum containing insoluble silicates to aid in
sulfate hydroxide hydroxide the coagulation and to produce a tougher floc.
3CaS04 This material contains less water of crystalli-
calcium zation than the usual filter alum. Black alum
sulfate is a commercial filter alum containing pow-
SomuM ALUMINATE Na2Al 2 0 4 • Sodium alu- dered activated carbon. Black alum is em-
minate is actually aluminum oxide stabilized ployed where it is desired to obtain both co-
with caustic soda. A good commercial grade agulation and a degree of taste and odor
of sodium aluminate contains approximately removal with a single material.
55% aluminum oxide, 35% combined so- CoPPERAS (Ferrous Sulfate-FeS04 .7H2 0).
dium hydroxide and 6% excess sodium hy- The natural alkalinity of most waters is
droxide. This material may be used in place not sufficient to react with copperas to form
of an alkali in water which must be coagu- the desired ferric hydroxide floc. Lime or an-
lated with aluminum in a higher pH zone. other alkal.i must be added to produce this
result. The first reaction with lime produces as ferric chloride. Reactions with ferric sul-
ferrous hydroxide which in turn is oxidized fate can be shown as follows:
by the dissolved oxygen in the water to form
ferric hydroxide. This oxidation occurs only Fez(S04h + 3Ca(HC0llz = 2Fe(0Hh +
at higher pH values and copperas is a suitable ferric + calcium = ferric +
sulfate bicarbonate hydroxide
coagulant only in the alkaline pH range of
8.5 to 11.0.
3Ca504 + 6COz
calcium + carbon
FeS04 + Ca(OH)z = Fe(OH)z + sulfate dioxide
ferrous + calcium ferrous + Fez(S04h + 3Ca(OH)z = 2Fe(OHh +
sulfate hydroxide hydroxide
CaS04
ferric + calcium = ferric +
sulfate hydroxide hydroxide
calcium
sulfate 3Ca504
calcium
4Fe(0Hlz + Oz + 2Hz0 = 4Fe(0Hh sulfate
ferrous + oxygen + water = ferric
hydroxide hydroxide In general, the ferric coagulants will co-
An advantage of both ferrous and ferric agulate at much lower pH values than the
coagulants in comparison with aluminum co- aluminum salts and can be employed over
agulants is that the floc produced will not the wide pH range of 4.0-11.0. Ferric coagu-
redissolve at higher alkalinities. However, the lants have been employed successfully in color
need for coagulating at a higher pH and the removal at low pH values. At high pH values,
consequent use of larger amounts of alkalies ferric coagulants are also useful in combining
restricts the general use of this material for color removal with oxidation and precipita-
coagulation purposes except in conjunction tion of iron and manganese. The ferric hy-
with lime softening. droxide floc is heavier than alum and settles
FERRIC CoAGULANTS. The ferric coagulants, more rapidly. Chlorinated copperas is also
ferric sulfate-Fe2 (S0 4 ) 3 and ferric chloride classed as a ferric coagulant. Chlorine will
-FeCl3 , act to precipitate ferric hydroxide oxidize copperas and form a mixture of ferric
as is the case with copperas. However, since chloride and ferric sulfate.
the iron is already in the ferric state, dis- 6Fe(S04) + 3C(z = 2Fez(S04h +
solved oxygen is not required for oxidation. ferrous + chlorine ferric +
Ferric chloride is available either in the sulfate sulfate
anhydrous form, or as FeCl3 .6H2 0 or as a 2FeCh
water solution containing 35-40% FeCl3 • Fer- ferric
ric chloride is quite corrosive in the presence chloride
of moisture. Reactions with natural or added In practice, the copperas solution is mixed
alkalinity can be illustrated as follows: with the chlorinator discharge prior to the
2FeCh + 3Ca(HC03)z = 2Fe(OHh + entry of the mixed solution into the coagula-
ferric + calcium ferric + tion system. Theoretical chlorine require-
chloride bicarbonate hydroxide ments are 1 part of chlorine for each 7.8 parts
3CaCI 2 + 6C0 2 of copperas. Since this reaction produces both
calcium + carbon ferric chloride and ferric sulfate, chlorinated
chloride dioxide copperas has the same field of usefulness as
2FeCI3 + 3Ca(0Hh = 2Fe(0Hh + 3CaC(z these materials. Because of the comparative
ferric + calcium = ferric + calcium ease of feeding copperas as compared with
chloride hydroxide hydroxide chloride the ferric coagulants, chlorinated copperas is
Ferric sulfate is available as the anhydrous sometimes preferred.
material and as Fe 2 (S0 4 )a.9H 2 0. Ferric sul- MAGNESIUM OxiDE (MgO). Activated mag-
fate is corrosive, but not to the same extent nesium oxide was developed as an efficient
32 BETZ HANDBOOK
Fig. 5-3-Data on Coagulants and Coagulant Aids
Suitable
Commercial Grades Weight, Handling
Name Formula Strength Available lb/cu ft Materials Miscellaneous
Aluminum Lump Powder, 38-45 Lead pH (1 "/o
Sulfate Powder Other, 57-67 Rubber solution)
Granules Silicon Iron = 3.4
Sodium Crystals 60 Iron Stabilized

Aluminate Steel with approx.
Rubber 6"/o excess
Plastics NaOH
Ammonium AI2(S04h· Lump 60-68 Lead pH (1%
Alum (NH 4 hS04 .24H 20 Powder Rubber solution I
Silicon Iron = 3.5
Stoneware
Copperas 55"/o FeS04 Crystals 63-68 Lead Efflorescent
Granules Tin
Wood
Ferric 90% Fe 2(S0 4 h Powder 60-70 Lead Hygroscopic
Sulfate Granules Rubber
Stainless Steel
Plastics
Ferric FeCI 3 .6H 20 60"/o FeCI 3 Crystals 45-55 Rubber Hygroscopic
Chloride Glass
Stoneware
Magnesium MgO 95"/o MgO Powder 25-35 Iron Essentially
Oxide Steel insoluble,
fed in slurry
form
Bentonite Powder 60 Iron Essentially
Steel insoluble,
fed in slurry
form
Sodium Na 20.3.22Si0 2 41° Be Solution 87 Iron Solid grades
Silicate Steel are available
Rubber with varied
Na 20-Si0 2
ratios
means for the removal of soluble silica from unique and useful advantage 1s that no
water. Magnesium oxide is also an efficient soluble salts are added to the water with its
coagulant in the alkaline pH range of 9.5 use and no lime or soda ash is required
and above. Magnesium oxide provides ex- for its precipitation. The additional advantage
cellent coagulation properties in conjunction of silica removal is secured with this coagul-
with hot and cold process lime-soda soften- lant and unlike aluminum coagulants, no ions
ing and can be considered as a coagulant are introduced into the water which may re-
specific to the lime-soda softening process. A sult in complex boiler scales.
Figure 5-4 • Jar Test Coagulation Studies
COAGULANT AIDS ionic, cationic or nonionic. Polymers whose

functional groups in water solution give posi-
Polyelectrolyte coagulant aids are being used tively charged particles are cationic. Polymers
in increasing quantities in water clarification that dissociate to form negatively charged
processes. The increase in use far exceeds the ions are called anionic. Polymers in the non-
increasing demands for water as a raw ma- ionic group provide both positive and nega-
terial indicating that the synthetic polymers tive charges in solution.
have definitely proven themselves superior in The mechanisms of the aids are not com-
many applications. pletely understood but a great deal of re-
Polyelectrolytes are high molecular weight search is being done in order to develop fur-
water soluble polymers that contain groups ther knowledge of this function. Such factors
capable of undergoing electrolytic dissocia- relating to colloidal behavior as Zeta poten-
tion to give a highly charged, large molecular tial, Brownian movement, van der Waals
weight ion. Originally, the term was applied forces and electrophoretic mobility are being
only to synthetic polymers or copolymers but studied to determine a more scientific ap-
has become more inclusive through general proach to the selection of coagulants and co-
use. The term now includes naturally occur- agulant aids. At the present time the most de-
ring organic flocculents ;many of which func- pendable tool for checking out a system of co-
tion through hydration alone rather than agulants is the well-known jar test as illus-
through electrolytic activity. In some cases, trated in Figure 5-4.
the term is applied to combination of natural The pH of a water to be clarified is usually
organic flocculents and inorganic weighting a very critical factor. In actual practice, there
agents. is generally a very narrow pH range where the
Polyelectrolytes are further classified as an- quickest and best results are obtained with
34 BETZ HANDBOOK
any particular coagulant. In many cases, the verely tax the effectiveness of normal clarifica-
application of a polyelectrolyte will not only tion processes and it is here that the aids can
broaden the pH range over which satisfactory often be of most value.
flocculation will occur but also reduce the
quantity of primary coagulant required.
In many cases, the use of polyelectrolyte REFERENCES
aids has been justified on the basis of im- A. P. Black, "Basic Mechanisms of Coagulation",
Journal, Am. Water Works Assoc., Vol. 52, pp. 492-
proved quality water rather than any cost sav- 504 (1960)
ing. These polymers may also prove advan- J. M. Cohen, "Improved Jar Test Procedure",
tageous in permitting higher flow rates through Journal, Am. Water Works Assoc., Vol. 49, pp.
existing clarification equipment. This factor 1425-1431 (1957)
J. M. Cohen, G. A. Rourke and R. L. Woodward,
can be very desirable to a company which "Natural and Synthetic Polyelectrolytes as Coagu-
would otherwise be faced with a capital invest- lant Aids", Journal, Am. Water Works Assoc., Vol.
ment for new clarification equipment because 50, pp. 463-478 (1958)
of an increased water demand. Increases in C. R. Cox, "Laboratory Control of Water Purifica-
tion", pp. 163-190, Case-Shepperd-Mann Publishing
throughput up to 100% have been attained Corp., New York, N.Y. (1946)
without sacrificing the quality of the effiuent. W. A. Hardenbergh, "Water Supply and Purifica-
tion", 3rd Ed., pp. 331-367, International Textbook
In many clarification installations, the aids Co., Scranton, Pa. (1952)
are employed when influent conditions are A. A. Hirsch, "Manual for Water Plant Operators",
such that the heaviest demand is placed on pp. 56-76, Chemical Publishing Co., Brooklyn, N.Y.
the efficiency of the unit. During the run-off (1945)
"Water Quality and Treatment", 2nd Ed., pp. 131-
period following storms and during cold 162, Am. Water Works Assoc., New York, N.Y.
weather operations, surface supplies can se- (1950)
6
Filtration
Theory Involved. . . . . . . . . . . . . . . . . . . . . . . . . . 36
Types of Filters . ......................... 38
Gravity Filters . ........................ 38
Pressure Filters . . . . . . . . . . . . . . . . . . . . . . . . . 39
Filter Operation . ......................... 40
Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Limitations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
36 BETZ HANDBOOK
is the process of passing a liq- lated particle~ normally does not extend
F ILTRATION
uid containing suspended matter through
a suitable porous material in such a manner
deeper than ~-4 inches and most of the fil-
tration is secured at the surface or in the
as to effectively remove the suspended matter first one or two inches of the bed. This coagu-
from the liquid. lated mat acts as a fine filter for smaller
Filtration is employed in the treatment of particles. It is necessary in selecting the size
industrial water in order to remove or re- of filter medium to have sufficient coarseness
duce suspended solids and turbidity. The tur- so that some penetration of the top few
bidity or suspended matter may be present inches of the bed takes place. With no pene-
initially in the raw water undergoing treat- tration of the bed by coagulated material,
ment or may be the result of chemical coagu- the loss of head would increase rapidly and
lation or a precipitation process of water filter runs would be short.
treatment. For example, in lime-soda soften- It is desirable from the standpoint of filtra-
ing, calcium and magnesium salts are precip- tion to provide a size of filter medium which
itated in the softening process and final clari- will prevent floc from passing through the
fication by filtration must be done prior to filters, hold the floc as loosely as possible in
use. In general, most water treatment proc- order to permit easy washing and hold as
esses involving coagulation and precipitation large a volume of floc as possible without
are followed by filtration in order to complete clogging. The removal of turbidity is affected
the removal of undesirable impurities from not only by the sand grain size, but by the
the finished water. particle shape. Sharp, angular particles pro-
Filtration as such does not provide removal duce larger voids and do not remove as much
of dissolved solids although .it may be used fine material as rounded particles of the same
in conjunction with a softening process that equivalent diameter. From the standpoint of
does reduce the dissolved solids content of backwashing the filter medium should be of
the water treated. such size that backwashing will free it of ad-
hering floc without it being lost during back-
THEORY INVOLVED
wash.
Sand and anthracite coal are the materials
With proper sedimentation following chemi- usually used as filter media. Bed depths of
cal treatment, the heavier coagulated parti- 15-30 inches are employed depending on the
cles will have been removed prior to filtra- type of filter used.
tion. Only the smaller and lighter particles Quartz sand, silica sand and anthracite
of floc reach the filters. When a freshly back- coal used in most gravity and pressure filters
washed filter is first placed in operation, must meet definite standards, namely effec-
many of the finely coagulated particles pene- tive size and uniformity coefficient. The ef-
trate into the filter bed through the many fective size is that size such that 10% of the
voids in the bed surface. As the particles lodge sand grains by weight are smaller and 90%
between grains of filter medium, flow is re- coarser. Thus, the effective size is the mini-
stricted. Coagulated particles then build up mum size of the bulk of the sand particles.
on the surface of the filter bed. With con- However, this does not indicate either the
tinued filtration, the rate of flow increases coarseness limits nor the degree of variation.
through the larger unclogged passages and To assure that the variation is not too great
decreases through the smaller and partially the second measurement must be made. This
clogged passages. As the flow penetrates the is the size of the_ sand grains of which 60%
bed further, the water spreads out and veloc- by weight of the sand is finer and 40%
ity decreases with lodging of coagulated par- coarser. This size divided by the effective size
ticles again at points of low velocity. is the uniformity coefficient. For example: if
Penetration of the filter medium by coagu- a screen analysis showed 10% of a sand was
finer than 0.40 mm, and 60% was finer than size of from 0.35 to 0.50 mm, as conditions
0.64 mm, the effective size would be 0.40 mm warrant, and a uniformity coefficient of not
and the uniformity coefficient would be 0.64 over 1. 75. Pressure filters are generally pro-
divided by 0.40 or 1.6. vided with filter sand with effective size of
The most desirable effective size and uni- 0.50 to 0.60 mm and the uniformity coeffi-
formity coefficient depends on conditions of cient of 1. 7 maximum. Experiments made at
operation and effluent quality requirements. the Chicago Water Department concluded
Modern practice usually calls for two grades good results can be obtained with coarser
of filter sand; for gravity filters the effective sand, with an effective size range of 0.60 to
Courtesy Roberts Filter Mfg. Co.

Figure 6-1 • Cutaway of Gravity Filter
38 BETZ HANDBOOK
0. 72 mm, with a uniformity coefficient of moval of accumulated suspended matter is

1.3 to 1.5. This work indicated a 26 inch bed secured by backwashing up through the filter
depth optimum. bed, expanding the bed ··and washing the
Anthracite coal prepared in similar sizes is solids to waste.
equally effective and may be used in place of The essential parts of a gravity filter, in
sand and gravel in filters. Extended filter addition to the filter medium, can be briefly
runs, less backwash water requirements and outlined as follows:
higher filtration rates have been experienced 1. The filter shell itself may be of concrete,
with anthracite coal filters. In many indus- steel or wood and may be square, rectan-
trial filtration applications, the anthracite gular or circular in shape. However, rec-
medium is preferred because no silica is added tangular units of reinforced concrete are
to the water due to high alkalinities and heat. most widely used.
The lower specific gravity permits a lesser 2. The filter medium is supported by a gravel
backwash rate. bed which serves the purpose of prevent-
Other materials such as quartz, calcite and ing the fine sand or anthracite from pass-
magnetite are sometimes used as filter media ing down into the underdrain system and
for special purposes. also acts to distribute the backwash water.
The supporting bed may vary in depth
TYPES OF FILTERS usually from 12 to 24 inches and consists
of several layers varying from 1 ~ inches
Filters are divided into two types-gravjty to Ys inch in diameter.
filters and pressure filters. 3. The underdrain system is provided for two
GRAVITY FILTERS. As the name implies, the purposes-the collection of the filtered
flow of water through these filters is obtained water after passing down through the filter
by gravity. medium and supporting gravel and also
"Slow sand filters" or "English" filters are for the distribution of backwash water to
mentioned in passing because of their his- the underside of the filter medium. It is
torical significance. These were the first form essential that the underdrain system act
of filtration employed: These filters, now es- uniformly in both operations otherwise
sentially obsolete, use beds of fine sand. filtering and backwashing rates in one
The water is filtered at very low rates, usually part of a filter would exceed those in an-
in the range of 2-10 million gallons per day other and result in inefficient operation.
per acre (0.032-0.160 gpm per sq ft). The The underdrain system may consist of
removal of suspended solids is secured by header and laterals with perforations or
straining through the pores in the surface of strainers suitably spaced. False bottom type
the sand layer and by adherence of the solid underdrain systems are also popular.
particles to the sand grains. Biological action 4. Wash water troughs are provided to col-
in these beds accounts for the good results lect backwash water. These troughs may
obtained. As backwashing of such large areas be of steel, cast iron or concrete with the
is impractical, at intervals the accumulated newer plants employing aluminum, asbes-
growth and suspended matter with some of tos-cement and fiberglass for greater re-
the sand from the top of the bed are removed sistance to corrosion. They must be of
by rakes. adequate size to handle maximum back-
In rapid sand or coal filters, the water is wash rates without flooding. Troughs are
passed downward by gravity at a relatively spaced so that the horizontal travel of
high velocity, usually at rates between 2-4 backwash water will not exceed 3 to 3~
gpm per square foot. Pretreatment by coagu- feet. Wash troughs are placed approxi-
lation and sedimentation is necessary in order mately 24 inches above the top of the filter
to permit these higher rates of filtration. Re- surface although where high backwash
Courtesy, Infilco, Incorporated
Figure 6-2 • Battery of Pressure Filters with Pilot Operated Diaphragm Valves
rates are employed, the freeboard above and manual operations previously required
the filter surface should equal the inches on standard gravity filters. Their application
of vertical rise per minute. to industrial water filtration problems is
5. Certain control devices are required to mcreasing.
insure maximum efficiency in filter opera- PRESSURE FILTERS. Pressure filters are some-
tion. Rate-of-flow controllers automatically what more widely used than gravity filters in
maintain uniform delivery of filtered industrial water conditioning. Pressure filters
water, operating from Venturi tubes in possess the advantage that they may be placed
the effluent line. Backwash rate-of-flow in the line under pressure and thus eliminate
controllers are used similarly to provide double pumping. The use of pressure filters is
proper wash conditions. Rate-of-flow and required in conjunction with hot process sof-
loss-of-head gages are also considered nec- tening to permit high temperature -operation
essary for most efficient operation. It is and to avoid thermal loss. The general de-
possible to group most gages, together sign of pressure filters is essentially the same
with the operating controls, on operating as gravity filters with respect to filter medi-
tables for convenience for either manual um, supporting bed, underdrain system and
or automatic operation. control devices. The filter shell, of course,
New concepts in automatic gravity filtra- differs from a gravity filter and wash water
tion are offered by several manufacturers troughs are not employed.
that reduce the need for various controllers Pressure filters may be of the vertical or
40 BETZ HANDBOOK
horizontal type. The filter shells are steel, and also permits a small amount of suspended
cylindrical in shape, with dished heads. Ver- matter to accumulate on the bed. As soon as
tical filters range in diameter from 1 to 10 feet the filter produces clear water, the unit is
with capacities from 2.4 gpm to 235 gpm returned to service.
at a filter rate of 3 gal per sq ft per minute. During operation, the suspended matter re-
Horizontal filters, usually 8 feet in diameter, moved by the filter accumulates on the surface
may be 10 to 25 feet long with capacities from and the increased frictional resistance is shown
218 gpm to 570 gpm. The general trend is by the loss-of-head gage. When loss-of-head
away from horizontal pressure filters because reaches 5 psi, it is advisable to backwash the
certain areas of the filter bed adjacent to the filter to remove the mat on the bed surface.
shell are inactive during filtration and back- While industrial filtration rates are usually 3
washing. One vendor has redesigned the hori- gal per sq ft of filter area per minute, backwash
zontal unit with cells thereby improving rates are much higher in order to lift the sus-
operations. pended matter from the surface and expand
Figure 6-3 the filter bed. Backwash rates of 12-15 gals
per sq ft per minute are employed for sand
Filter and Backwash Rates for Vertical
and rates of 8-12 gals per sq ft per minute for
Pressure Filters
anthracite. The backwashing is continued for
Diameter, Area Filter Rate• Backwash Ratet 5 to 10 minutes, then the filter is returned to
ft Sq Ft gpm gpm
service, following filtering to waste until the
3 7.1 21 106 water is clear.
4 12.6 38 189 Surface washers are of assistance in main-
5 19.6 59 295
6 28.3 85 425 taining maximum filter efficiency. These units
7 38.5 116 578 consist of horizontal arms, equipped with noz-
8 50.3 150 755
9 63.6 190 954
zles, supported just above the surface of the
10 78.5 235 1180 filter bed. High pressure backwash water sup-
plied to the washer causes rotation of the
* Filter rate of 3 gal per sq fl per minute arms and water leaving the nozzles impinges
t Backwash rate of 15 gal per sq fl per minute
on the media, and loosens the accumulated
Figure 6-4
suspended matter.
Typical Filter Bed layers and Sizes For
Pressure Filters APPLICATIONS
SAND ..... 12" layer-sand, 0.45-0.50 mm effective size Filtration is a requirement in the clarification
10" layer-sand, 0.80-1.20 mm effective size of turbid waters for most industrial uses. In
4" layer-gravel, lf.I"·Ys"
4" layer-gravel, %"·1!.1"
the preparation of domestic supplies, process
8" layer-gravel, 1"·%" water, boiler feedwater and, in fact, for almost
4" layer-gravel, 1%"·1" every use it is necessary that the water be
ANTHRAFILT 18" layer-No. 1 Anthrafllt, 0.6-0.8 mm freed of objectionable suspended solids. While
9" layer-No. 2 Anthrafllf, 3/32"-3/16"
9" layer-No. 4 Anthrafllt, 5/16"-9/16" both coagulation and sedimentation may be
4" layer-No. 6 Anthraflll, 13/16"·1 %" employed for the removal of a portion of the
suspended matter, filtration normally also is
FILTER OPERATION
required to complete the production of a water
of satisfactory quality.
After backwashing, and placing a filter in Filtration is also employed as a part of other
operation, the water is filtered to waste for the water treatment processes. For example, all
first few minutes. This procedure rids the of the precipitation methods of softening em-
system of possible suspended solids remaining ploy filtration as a final step in the process.
in the underdrain system after backwashing Filtration is employed following hot phosphate
softening as well as both hot and cold lime solids as in gravity filters. Pressure filters are
and lime-soda softening, except in some special more difficult to inspect and clean. The re-
cases. placement of filter medium, gravel and under-
Filtration ahead of exchange softeners and drain systems presents more of a problem.
demineralizers is standard practice. Filtration However, the most pertinent criticism of
is also a final step in certain iron removal pressure filters in comparison with gravity
methods and in the removal of oil from con- filters is the inability of the operator to ob-
densate. serve the effectiveness of the filtering and
backwashing operations. While pressure filters
are capable of providing an effluent of as
LIMITATIONS
high quality as gravity filters, this inability
Filtration, without coagulation and sedimen- to maintain constant observation of the proc-
tation, is usually an unsatisfactory method for ess may result in the production of an effluent
completely removing turbidity and suspended of poorer quality than is obtained with gravity
solids. In such an application, the filter pro- filters. Consequently, gravity filters are pre-
vides only the removal of the larger particles. ferred where the conditions of the installation
Pressure filters are favored where double permit their use and where a water of highest
pumping can be avoided by placing the filter quality is desired. Any malfunctioning of a
under line pressure. Pressure filters also usually gravity filter may be quickly observed and
require less space for installation. Pressure corrected. The selection between the two types
filters are required with hot process operation must be made after a study of the conditions
and prevent thermal loss in· filtration and and problems involved in the particular instal-
backwashing. Oxygen pickup is also prevented lation.
by the use of pressure filters. REFERENCES
Disadvantages in pressure filtration are that
the appearance of the filter effluent and the G. D. Dickey, "Filtration", Reinhold Publishing
Corp., New York, N.Y. (1961)
filter medium are not under observation, back- "Filtration", The Public Works Manual and Cata-
washing effectiveness cannot be observed and log File, Public Works Publications, Ridgewood,
loss of filter medium cannot be noted. Wash N.J. (1961)
water troughs cannot be provided in pressure E. Nordell, "Water Treatment for Industrial and
Other Uses", 2nd Ed., pp. 363-387, Reinhold
filters for as effective removal of backwash Publishing Corp., New York, N.Y. (1961)
42 BETZ HANDBOOK
7
Precoat Filtration
Theory Involved . ......................... 43
Filter Operation . . . . . . . . . . . . . . . . . . . . . . . . . . 44
Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
IATOMA_CEous earth filtration is a proc:ss

D in whrch a filter cake or precoat of dia-
tomaceous earth is used as a filter medium.
The precoat of diatomaceous earth is formed
over a permeable base or septum, by adding
a slurry of diatomite to the filter shell and
either filtering to waste or recirculating the
slurry until the precoat is formed. The unit
is then ready for the filtering operation.
Diatomaceous earth filtration is a special
form of filtration that is employed basically
where the utmost clarity of the effluent is re-
quired or where space and weight limitations
are quite stringent.
Diatomite filtration achieved prominence
during World War II, primarily because of
the need of the armed forces for filter units for
potable water that would conserve space and
weight, which would be portable and rela-
tively simple to operate. During the war, these
filter units were also applied to the removal
of oil from condensate on shipboard, where
space and weight requirements must also be
kept at a minimum.
EQUIPMENT EMPLOYED
Courtesy Johns Manville Products Corp.
The principle of diatomaceous earth filtration Figure 7-1 • Photomicrographs (X100) of Fine,
is relatively simple. A permeable base or sep- Medium and Coarse Particle Size Filter Aids
tum of some type is required to support the
diatomite filter cake. The septum must be of THEORY INVOLVED
such a nature that it will prevent the passage
of the filter aid, must be permeable to the Chemically, diatomaceous earth is silica
flow of filtered water and must be of sufficient (Si0 2 ). It is produced from mineral deposits
strength to withstand the pressure drop. formed by diatoms (a form of plant life re-
Among the materials used as a supporting lated to algae) . Diatomaceous earth is really
base are filter cloths, porous stone tubes, wire composed of billions of tiny fossils, resulting
screens, wire-wound tubes, and porous paper from the silica "skeletons" left by diatoms in
filter pads. This supporting medium is first prehistoric times. Diatomaceous earth deposits
coated with a slurry of diatomaceous earth have been found as much as 1400 feet in
which then acts as the filter medium. The thickness.
water is filtered in passing through the diato- The structure of diatomaceous earth is due
mite coating. During the filter run, additional to the shape of the tiny plants from which it
diatomite slurry is usually added. When a originated. These plants, or diatoms, are so
high back pressure drop develops, due to the small that they must be examined under a
accumulation of suspended matter removed in microscope. Individual figures are of many
filtration, the filter coating is sloughed off by shapes such as flowers, discs and needles. Over
backwash, and the filter returned to service 10,000 types have been identified. Commer-
after again precoating. Coagulants, with at- cial filter aids are produced from the crude
tendant pH control, are not required. material by a gentle milling process which
44 BETZ HANDBOOK
separates the diatoms from one another. The shell is usually of steel. Diatomite for the pre-
finished product is a fine powder, consisting coat is added as a slurry and the raw water
of rigid particles so tiny that hundreds could pumped through the filter to waste or recircu-
· be placed on a pinpoint. lated until the diatomite evenly precoats the
. When a precoat of diatomaceous earth is filter elements. Only a few minutes are re-
formed on the septum, the irregular, but sym- quired for this operation and the filter is then
metrical, shape of each diatom forms an in- ready for service.
compressible filter medium consisting of ap- As suspended matter is removed during the
proximately 10% solids and 90% voids. In filtering cycle, the pressure drop across the
effect, a screen is formed with extremely tiny filter element increases and rate of flow de-
openings-so fine, in fact, that even most creases. The filter is then backwashed by re-
types of bacteria are strained out. The object versing the flow through the unit, removing
·of the septum is to provide support for the the filter cake from the elements by the re-
precoat. The openings in the septum are not verse flow of water. Spent diatomite is drained
sufficiently small in size to prevent the passage from the filter shell.
of individual diatomite particles. Instead, the
One system uses a variation of this proce-
septum supports the mat or filter cake formed
dure in that the elements are backwashed
· by the interlacing of thousands of diatomite
· particles over each opening. Some diatomite from filtered water storage to waste, ending
particles pass through the septum during the with the unit filled with filtered water.
precoat operation, but once formation of the Other designs may employ compressed air
precoat is complete, the interlocked mass of to remove the filter cake from the elements
diatomite particles prevents further passage. at conclusion of a filter run. In another
Commercial diatomite filter-aids are pro- method, the air trapped and compressed in the
duced in several different grades offering a filter dome is utilized to remove the filter cake
range of filtration rates and clarity of filter from the elements, in conjunction with back-
effluents. In addition to applications for water wash by filtered water.
filtration, diatomaceous earth filter-aids are Leaf type filters, which are employed fre-
employed widely in the filtration of process quently in the removal of oil from condensate,
liquids. employ no backwash water. At the conclusion
While diatomite is the filter aid primarily of the filter run, the unit is drained and the
employed with filters of this type, there are filter cake is manually peeled off the filter
also available various grades of purified wood leaves.
cellulose which find application in the filtra- In some operations, the precoat is the only
tion of high pH solutions where silica pickup filter aid employed. More frequently, however,
is to be avoided. Other media used include additional diatomite is continuously added
perlite, asbestos and carbon. to the raw water during the entire filter run.
It has been found that where a precoat only
FILTER OPERATION is employed relatively short filter runs will
be experienced. Suspended matter will tend
The primary clifference among the several to plug the openings in the precoat, acting to
different types of precoat filters lies in the conform an impervious mat. Pressure drop in-
struction of the filter element. These con- creases rapidly resulting in short filter runs. By
struction differences affect the manner of filter adding additional diatomite continuously dur-
operation. ing filtration, the filter cake is maintained in
In filters where wire-wound tubes are em- a more porous condition. Lesser pressure drop
ployed, these are usually mounted vertically and longer filter runs result. The diatomite
on a removable plate and inserted as a unit added during filtration is referred to as body
in the filter shell. In the larger sizes, the filter feed to distinguish it from the initial precoat.
Because diatomaceous earth is an abrasive

material, some special designs of feeding ap-
paratus have been _de~elofed. The specific
gravity of the matenal IS high and the mate-
rial will settle out of suspension and tend to
pack at points of low velocity. Feeding equip-
ment must be designed with these considera-
tions in mind.
The amount of diatomite required to form
the precoat and also the amount necessary for
body feed will vary widely. Factors influeJ;lC-
ing the diatomite requirements include the
grade of media employed, the characteristics
of the influent water and the design of the
filter. However, experience has indicated pre-
. coat requirements in the ;range of 1.0-2.0
ounces per square foot of filter area. Body
dosages are directly influenced by influent
turbidity and filtration rates and consequently
are subject to wide variation. Values varying
from 5-50 ounces per thousand gallons of
filter effluent have been reported. Courtesy Infilco, Incorporated
Figure 7-2 • Industrial Type of Diatomaceous

APPLICATIONS Earth Filter Unit
Because of the savings in both space and Health Service Standards, even without chlo-
weight of diatomite filters compared with con- rination. Diatomite filters will remove cysts of
ventional sand filters, diatomite filters are amoebic dysentery and the cercariae of schis-
obviously favored for mobile use by the armed tosomia (blood flukes) , prevalent in Philip-
forces. Figure 7-3 illustrates the comparison pine waters. Consequently, the armed forces
of the U. S. Army pack and mobile units with favor diatomite filters for mobile purification
conventional pressure and sand filters. units.
Figure 7-3 Diatomite filtration is also applied to cer-
U. S. Army Pack and Mobile Units tain process industries where rigid water re-
Compared with Pressure Sand Filters quirements exist with respect to clarity and
bacteriological quality. In such applications,
Pack
Unit
Pressure
Sand Filter
Mobile Pressure
Unit Sand Filter
the diatomite filter is frequently applied as a
"polishing" filter after conventional coagula-
Rated Capacity, gpm 15 15 50 50 tion, sedimentation and sand filtration.
Filter area, sq fl 3.6 7.1 10 19.6
Weight, lbs 30 2860 350 8050 In the removal of oil .from contaminated
Diameter, in 8 36 18 60 condensate, diatomite filters possess an ad-
Over-all Height, in 22 73 30 82 vantage over conventional coagulation and
Floor Space, in lOx 10 37 x49 24 X 24 6] X 78
filtration methods. No control of coagulant
and alkali feed is required since neither co-
Due to the extremely fine openings in a agulant nor pH adjustment is employed with
diatomite filter cake, a greater clarity of ef- diatomite filters. In addition, no increase re-
fluent can be obtained. Army tests showed sults in the solids content of the filtered con-
that diatomite filters can be expected to pro- densate, due to the coagulant. ·
vide filtered water approximating U. S. Public Diatomite filtration is being widely applied
46 BETZ HANDBOOK
to swimming pools where advantages are ex- Continuity of operation is generally con-
perienced in the somewhat simpler operating sidered a requirement for diatomite filters.
control and lesser need for coagulating chem- If the filters are operated intermittently, the
icals. filter cake will slough off when flow is stopped.
To avoid a turbid effluent, when flow is re-
LIMITATIONS sumed, it is necessary to again precoat the
filter. Intermittent operation is unsuitable to
The major factors limiting the use of dia-
diatomaceous earth filtration. In boiler feed-
tomite filtration are the need for a relatively
water conditioning particularly, it is necessary
low turbidity influent water, and the high
to guard against careless operation which per-
operating costs compared with conventional
sand filtration. mits high turbidity in the filter effluent from
disturbance of the precoat. The silica thus in-
In general, it is not advisable to employ
troduced would tend to be solubilized under
diatomite filters where raw water turbidities
exceed 200 ppm. Uneconomically short filter boiler temperature and pH conditions, pos-
runs result from high turbidity influent waters. sibly resulting in the formation of siliceous
boiler scales. For these applications, ,nonsili-
The purchase price of diatomite filters may
ceous media must be employed.
exceed the cost of pressure sand filters al-
though installation cost may be less because Some types of filter elements have shown
of the smaller space requirements. The operat- susceptibility to clogging with iron, manga-
ing cost of diatomite filters will generally ex- nese and other materials and require removal
ceed the cost of coagulation and pressure sand of the elements for chemical cleaning.
filtration, due primarily to the cost of the dia-
tomaceous earth.
REFERENCES
Diatomite filters operate with a high pres-
sure drop across the unit, as much as 25-50 "Diatomite Filter Aids", Technical Bulletin D10,
psi. Pumping costs are increased and the in- Great Lakes Carbon Corp., Los Angeles, Calif.
(1961)
stallation of diatomite filters may be limited "The Filtration of Water", Johns-Manville Corp.,
by the available pressure. NewYork,N.Y. (1961)
INDUSTRIAL WATER CONDITIONING
47
8
Chlorination
Chemistry of Chlorination . . . . . . . . . . . . . . . . . . . 48
Theory of Disinfecting Action . . . . . . . . . . . . . . 48
Chlorine Requirements. . . . . . . . . . . . . . . . . . . . 49
Chlorine-Ammonia Process. . . . . . . . . . . . . . . . 50
Breakpoint Process . . . . . . . . . . . . . . . . . . . . . . 50
Chlorine Dioxide. . . . . . . . . . . . . . . . . . . . . . . . . . 51
Effects of Chlorine . ........................ 51
Hypochlorination . ......................... 52
I
Dechlorination . ........................... 52
Feeding Equipment . . . . . . . . . . . . . . . . . . . . . . . . 53
Laboratory Control. . . . . . . . . . . . . . . . . . . . . . . . 53
48 BETZ HANDBOOK
HLORINATION is the process in which chlo- tion and oxidation.

C rine gas or chlorine compounds such as
hypochlorites are added to water or sewage,
The addition of chlorine to treated sewage
is widely used for bacterial control and sterili-
usually for the purpose of disinfection. zation. Further, substantial reduction in ap-
The principal use of chlorination is to safe- parent BOD has been experienced where rea-
guard water supplies by killing disease pro- sonably high dosages of chlorine were applied
ducing organisms. For this purpose chlorina- to sewage, either in the partially treated or
tion is standard practice and is one of the completely treated stage.
major reasons for the bacterial safety of our In the industrial waste field, chlorine has
public water supplies. been generally adopted for cyanide destruc-
In addition to its bactericidal qualities, chlo- tion in the presence of an alkaline salt. Other
rine and chlorine compounds are used in the uses in industrial wastes have included the
reduction and removal of objectionable tastes destruction of hydrogen sulfide and phenol.
and odors from drinking water, the oxidation Here, approximately 8.5 parts of chlorine are
of iron, manganese and hydrogen sulfide pres- required for each part of hydrogen sulfide
ent in waters thereby aiding in their removal, present and approximately 20 ppm chlorine
and the oxidation, decolorization and destruc- for each ppm of phenol. The pH of reaction
tion of organic matter. It has been stated that is important to obtain maximum efficiency.
chlorine is probably the most powerful and
flexible tool that chemical science has pro- CHEMISTRY OF CHLORINATION
vided the sanitary engineer. Further uses to
which it has become essential include its in- THEORY oF DISINFECTING AcTION. Chlorine
jection into sewage and industrial wastes prior hydrolyzes in dilute water solution as follows:
to their discharge to streams not only for the Ch + H20 = HOCI
bacterial reduction in these waste waters, but chlorine + water = hypochlorous acid
for the destruction of chemicals that may react + HCI
directly with oxygen, such as sulfides, sulfites, + hydrochloric acid
ferrous iron and others. Chlorine is widely It is the hypochlorous acid produced in this
used for the control of slime and algae in cool- hydrolysis that provides the disinfecting and
ing waters in both once-through and recircu- oxidizing properties of chlorine solutions.
lating systems. Early theories that the sterilizing properties
In the clarification process of water treat- of chlorine were due to nascent oxygen liber-
ment, chlorine is often applied to the raw ated from hypochlorous acid or to complete
water along with the coagulant. Included in oxidation of organic matter have been dis-
the benefits derived from prechlorination proved. Later work has led to the theory that
practices are improved coagulation probably the destruction of organisms results from the
due to the action of chlorine on the organic chemical reaction between hypochlorous acid
material in the water; reduction in taste, odor formed when chlorine is mixed with water
and color producing materials by oxidation; and an enzyme in the organism cell. The
reduction of biological loading by longer con- amount of HOC! is dependent on the pH of
tact with the water thereby providing a safety the solution. At pH below 6.5, the HOC! con-
factor; and oxidation of iron and manganese stitutes almost 100% of the free available
for removal by settling and filtration. chlorine. At a pH of 9.0 and above, the ionized
Chlorine has been used successfully as the hypochlorite ion accounts for practically all
activating agent for sodium silicate in the of the free available chlorine. Since it is hypo-
preparation of the coagulant aid, activated chlorous acid and not the hypochlorite ion
silica. The advantage claimed by this process that is the principal disinfectant in chlorine
is that the chlorine used for the activating solutions, the efficiency of disinfection will be
process continues to be available for steriliza- substantially greater at low pH values where
the hypochlorous acid content is greater. The bination and the resulting retarded action of
free available residual chlorine of a treated chlorine with organic matter, the maintenance
water is considered to be that portion of the of a stable residual in outlying water distri-
total residual chlorine existing in the water as bution systems is facilitated. Although this
the hypochlorous acid. combined residual is ordinarily more stable
than a free chlorine residual, due to its de-
creased bactericidal qualities it is essential to
maintain higher residuals.
The combined available residual chlorine
w of a treated water is considered to be that por-
z
~ tion of the total residual existing in chemical
3 combination with ammonia or organic nitro-
5
..J
<t
gen compounds. Studies on the bactericidal
:::>
0 properties of free and combined available chlo-
ill rine residuals have indicated that for complete
"' kill with similar contact periods, approximately
BREAKPOINT
25 times as much combined available residual
chlorine is required as is necessary with free
available residual chlorine. By the same studies
CHLORINE ADDEO
it was apparent that to obtain complete bac-
teria kill when using combined available resid-
Figure 8-7 • Breakpoint Chlorination
ual equal to the free residual, approximately
100 times as long contact period was required.
CHLORINE REQUIREMENTS. In addition to
The disinfecting ability of chlorine is re-
its bactericidal action chlorine displays its
duced when hypochlorous acid reacts with
strong oxidizing properties in other ways.
ammonia or amines to form chloramines.
Chlorine will react with and oxidize fer-
}?ormation of chloramines may be shown as
rous iron (Fe++) , manganous ·manganese
follows:
(Mn++), nitrite (N0 2 - ) and hydrogen
HOCI + NHl sulfide. Reaction with such inorganic constit-
hypochlorous acid + ammonia uents present in the water being chlorinated
NH 2CI + H20 increases the quantity of chlorine required
= monochloramine + water and adds to the chlorine demand of the water.
2HOCI + NHl In this oxidation process chlorine is reduced
hypochlorous acid + ammonia to the chloride ion which possesses no further
NHCI2 + 2H20 oxidizing properties or disinfecting power. The
= dichloramine + water reaction between hydrogen sulfide and chlo-
The pH of the treated water determines rine illustrates this point.
which product will be formed. At pH 8.5 H2S + 4Ch + 4Hz0
monochloramine is the major product. At hydrogen sulfide + chlorine + water
about pH 4.5 dichloramine is the major prod- H2S04 + SHCI
uct. Mixtures of the two prevail between these = sulfuric acid + hydrochloric acid
two values. Below pH 4.5 nitrogen trichloride Chlorine will also react with and be con-
predominates. sumed by organic matter present in the water.
Where chloramines exist, it is necessary to In some cases chlorine is consumed in direct
provide longer contact time or increased chlo- chlorination of organic compounds present in
rine residual for comparable bacterial reduc- sewage or industrial waste pollution with the
tion. With the oxidizing properties of chlorine formation of chlorinated compounds such as
being substantially reduced when in such com- chlorphenols. Products of this type may possess
50 BETZ HANDBOOK
undesirable tastes and odors. Chlorine may water is generally described as marginal.
also oxidize and bleach organic matter with Breakpoint chlorination is the application of
the extent of oxidation dependent on time of chlorine to water to maintain free available
contact and the residual chlorine present. In chlorine residuals. The addition of chlorine to
either case, this oxidation of inorganic and water containing ammonia or organic matter
organic compounds requires the use of higher will first produce an increased combined chlo-
amounts of chlorine to satisfy the chlorine de- rine residual. After reaching a maximum
mand before a chlorine residual for disinfec- value, increased chlorine doses result in de-
tion can be established. creased residual values. This decrease is
Chlorine demand may be described as the caused by the decomposition of the com-
difference between the amount of chlorine bined available chlorine residuals, formed
applied to a water and the amount of free, when chlorine is first added. This decrease
combined or total residual chlorine remaining in chlorine residual will not start until
at the end of a specified contact period. The the chlorine applied amounts to at least 8
<:;hlorine demand will vary on any water with times or more the amount of free ammonia
contact time, water quality, temperature and present. If this chlorine-ammonia ratio is ma-
pH. The chlorine requirement is the amount terially less than 8 to 1, say 4 to 1, chloramines
of chlorine that must be added to accomplish will be formed and the breakpoint reactions
a specific objective, i.e., bacterial kill, hydro- will not take place. The ammonia will have
gen sulfide removal or oxidation of iron and been essentially removed when the chlorine
manganese. dosage amounts to 10 times the amount of
The residual chlorine requirement in treated ammonia present. When the instantaneous
water is usually established by bacteriological chlorine demand is high, the chlorine require-
testing to determine the amount necessary for ments to obtain free available chlorine resi-
effective disinfection. Normally, a residual of duals may be 20 or more times the amount of
0.2 to 0.5 ppm is satisfactory after a minimum ammonia present.
10 minute contact period at or above 20 C. After the minimum chlorine residual or
Higher residual values may be required if the breakpoint is reached, the chlorine is present
chlorine exists as combined residual and the as free available residual and increases in di-
pH of the water exceeds 7 .0. The need for rect proportion to the chlorine applied. In
higher combined available chlorine residuals developing the free available residual, destruc-
is markedly increased in colder water. tion of taste and odors is usually achieved;
practically all of the bacteria have been killed,
CHLORINE-AMMONIA PRocEss. Ammonia or organic matter oxidized and the free residual
ammonium salts are employed in conjunction is quite stable and not readily dissipated.
with chlorine to produce a combined available Breakpoint chlorination may also be used
chlorine residual that is useful for certain to control slime and algal growths, to aid
specific problems. This process was originally coagulation, oxidize iron and manganese, to
developed for taste and odor control where remove ammonia from boiler makeup water
free chlorine reacted with organic matter in and to generally improve water quality in the
the water. Further, some minimizing of phe- treatment cycles or in the distribution system.
nolic and other type tastes by chloramines is The object of breakpoint chlorination is to
sometimes accomplished over conditions reproduce and maintain a free available chlo-
sulting from normal marginal chlorination. rine residual. Since a stable free chlorine resi-
The lower bacterial efficiency of this combined dual can exist only beyond the breakpoint,
residual limits the usefulness of the process. the process is controlled by making certain
Actually, this process has been superseded to that a definite free chlorine residual has been
a large extent by the more positive breakpoint produced and that it is adequate to carry
process for taste as well as bacterial control. throughout a plant or into the distribution
BREAKPOINT PROCEss. Normal chlorination of system as required.
6
r-.- __AMMONIA-NITROGEN
l_l_L_l __ -- 1.8
- ...
...... AMMONIA-NITROGEN AND
I
1/
5 ......·~ CHLORINE RESIDUALS OF 1.5
CALUMET RIVER WATER J
::::e 4
....,.,' 5 MI,N. CONTACT 7 1.2
::::e
""""
1
""
""
..;
en
0
Q
.....
z
iii:
3
CHLORINS,.. ~
~
/
1/
"\ \ i\
\
pH 7.6
7
J 0.9
z
.....
CD
0
=
1-
~
z0
9
::c / 1\ IJ ::::e
c.,) 2
v v \ \ v 0.6 ::::e
<
v '~ v
1
0
0
I
I
2 4 6 8 10 12
\
14
"
'\ - ---- -
16
-- - i--
18 20
0.3
0.0
CHLORINE RESIDUAL, PPM
Figure 8-2 • Ammonia Removal by Chlorination
The breakpoint process is of specific appli- and reportedly does not form reaction prod-
cation in the removal of ammonia from boiler ucts that cause taste intensification. The costs
makeup water. Figure 8-2 illustrates results for feeding chlorine dioxide are higher than
obtained in applying this process to Calumet for straight chlorine with the result that in
River water. In these experiments it was found order to justify its use there must be defi-
that greatest ammonia removal efficiency was nite improvements obtained. Fortunately, the
obtained with a chlorine to ammonia ratio of method for determining residual chlorine is
8: 1 and it was possible to reduce the undesir- applicable for chlorine dioxide. Chlorine di-
able ammonia concentration of the steam oxide reacts with acid orthotolidine rapidly
from the boilers to a negligible value. and therefore can be measured by the flash
test and the OTA modification.
CHLORINE DIOXIDE
EFFECTS OF CHLORINE
Another satisfactory and proven method for Many industrial waters are chlorinated for
using chlorine in the treatment of water, the control of bacteria and slime producing
sewage and industrial wastes is in combina- organisms. However, in food processing in-
tion with sodium chlorite to form chlorine dustries, the chlorine may cause tastes in
dioxide. This process was introduced in 1944 canned and frozen foods or may cause cor-
and is now used in more than 150 water rosion of the metal containers. Also, it has
works plants. It has also found wide indus- been determined that excessive chlorine
trial application for various applications of affects smoothness and brightness of metals
oxidation, bleaching and sterilization. in the plating industry. Paper mills producing
Chlorine dioxide is used in public water fine paper usually establish very critical limits
supplies for taste and odor control as well as for chlorine residuals.
for sterilization. It has high oxidizing powers In normal concentrations, chlorine does
52 BETZ HANDBOOK
not render water abnormally corrosive to zinc The hypochlorite ions estaplish equilib-
nor does it increase the solubility of lead salts. rium with hydrogen ions, depending on pH,
Reports on irrigation water and subsequent as does gaseous chlorine. The significant dif-
studies indicate no injury to land plants when ference is in the effect on pH. Solutions of
the irrigation water contains less than 50 ppm hypochlorite contain an excess of alkali which
combined residual chlorine. Aquatic plants tends to raise the pH. In some hard waters,
are reportedly harmed by concentrations of scale problems have been experienced due to
3 ppm or more. the precipitation of calcium carbonate by
The concentration of chlorine required to hypochlorites. Usually a complex phosphate
control plankton in reservoirs varies from 0.2 can be added to overcome this difficulty. In
·to 3.0 ppm depending on the sensitivity and other cases, hypochlorite treatment, due to the
type of organisms present. There are some high alkalinity, improves the quality of soft
types that may require as high as 50 ppm for and highly corrosive waters.
ciestruction. Normally, free available chlorine Calcium hypochlorite is soluble in water al-
in water may be toxic to fish and other aquatic though when dissolved in hard water, a pre-
life at low concentrations, and in combination cipitate often results. For liquid feeding of
with cyanide,. phenols and with other sub- this material, the solutions should be prepared
stances may be even more detrimental. at approximately 1% to 2% available chlo-
There are no specific data concerning the nne.
limiting physiological tolerance for chlorine in Sodium hypochlorite, NaOCl, is available
potable water, but it is generally agreed that in several forms ranging from the familiar
the small amounts of chlorine present in chlo- grocery store bleaches to commercial laundry
rinated waters are dissipated by the reaction bleach solution. These products vary in avail-
with saliva and gastric juices as soon as the able chlorine from 3% to as high as 20%.
water is swallowed. It is reported that chlo- Most of them are quite unstable. Sodium
rinated drinking water may, in rare cases, hypochlorite may be satisfactorily fed in the
cause asthma, colitis and eczema. When af- normal 15% available chlorine strength or it
flicted patients drank distilled water in place may be diluted with water for feeding.
of chlorinated water, their conditions im-
The available chlorine, on a pound for
proved and when they returned to chlorinated
pound basis, is equally effective for disinfec-
water, the symptoms reappeared.
tion and other purposes whether obtained
from hypochlorites or from gaseous chlorine.
HYPOCHLORINATION
All hypochlorites are corrosive to some de-
In addition to gaseous chlorine, there are com- gree and must be carefully handled and cor-
mercially available calcium and sodium hypo- rosion resistant materials should be used for
chlorites. The most common forms include storage and dispensing. Satisfactory materials
sodium hypochlorite having approximately include wood, ceramic, glass, plastic or
15% available chlorine and high test calcium rubber.
hypochlorite with 70o/o available chlorine. The
hypochlorite salt most widely used for water
DECHLORINATION
treatment IS the calcium hypochlorite,
Ca(OCl) 2 • Dechlorination may be accomplished by the
When this material is dissolved in water, use of activated carbon or by chemical re-
ionization occurs:
ducing agents such as sulfur dioxide, sodium
Ca(OCih = Ca++ sulfite, sodium bisulfite or sodium thiosulfate.
calcium hypochlorite = calcium ions Prolonged storage in open reservoirs is also
+ 2(0CI)- an effective means for dechlorination. The
+ hypochlorite ions practice of dechlorination may be desirable in
CHLORINE SOLO CHLORINE SOLD AS CHLORINE

AS LBS OF LBS OF DRY CALCIUM SOLD AS GALLONS OF
CHLORINE GAS HYPOCHLORITE OR SODIUM HYPOCHLORITE SOLUTION
CHLORIDE OF LIME
POWDER
® @ R
L£:j
RG
0 L::J
GAS POWDER POWDER 20% 15% 10% 5% 1°/o
70% AVAILABLE 35% AVAILABLE SOLUTION SOLUTION SOLUTION SOLUTION SOLUTION
CHLORINE CHLORINE
I# OF CHLORINE POUNDS OF DRY GALLONS OF SODIUM HYPOCHLORITE

GAS FOR EACH POWDER REQUIRED SOLUTION REQUIRED FOR EACH I#OF
I#OF AVAILABLE FOR EACH I# OF AVAILABLE CHLORINE
CHLORINE AVAILABLE CHLORINE
Courtesy Chlorination Topics, Spring 1956

Figure 8-3 • Comparative Chart Showing Amounts of Material Required
to Provide One Pound of Available Chlorine
either public supplies or for industrial uses. In The use of chlorine gas requires careful
public water supplies, the reduction or re- handling and the employment of special safety
moval of chlorinous taste may be necessary. measures since chlorine is highly irritating to
Many industrial process uses of water are ad- the eyes and to the membranes of the respira-
versely affected by high chlorine residuals and tory system. The physiological effect of various
consequently they must be removed or re- concentrations of chlorine gas are shown here-
duced. Also, it has been concluded that some with.
resins used in exchange softeners and de-.
minera,lizers are adversely affected by the ppm chlorine
presence of a residual chlorine in the water to by volume In air
be processed and consequently dechlorination Least detectable odor 3.5

ahead of these units is required. Cause throat irritation 15
Cause coughing 30
Dangerous for 30 minutes exposure 50
FEEDING EQUIPMENT Quickly fatal 1000
(from U.S. Bureau of Mines Paper 248)

Chlorine gas or solutions or hypochlorites are
in common use and are readily and accu- LABORATORY CONTROL
rately applied to public and industrial water
or sewage and industrial wastes. These ma- The success of treating water or wastes with
terials are applied either at a constant rate or chlorine is dependent upon the accuracy and
proportional to flow, dependent upon the adequacy of the control measures practiced.
specific requirements. The form in which the A number of chemical methods have been
chlorine is to be fed should be properly eval- developed for the control by measuring both
uated and the equipment supplied appropriate free and combined residual chlorine concen-
for the service. The tabulation (Fig. 8-3) trations. The iodometric titration was the
showing comparative amounts of the various initial method used. Modifications of this test
forms of chlorine should be helpful. Reliable have been developed which have improved
equipment for feeding all types of chlorine and the accuracy. The orthotolidine test has been
chlorine compounds is available from repu- very good for the purpose intended. How-
table vendors serving this specialized field. ever, the limitations of this test as well as
54 BETZ HANDBOOK
PRESSURE - VACUUM
RELIEF VALVE
REGULATING VALVE
.. CHLORINE
FEED RATE INDICATOR INJECTOR DISCHARGE
MANUAL
FEED
ADJUSTER
COMBINATION
INJECTORa DIAPHRAGM
CHECK VALVE
INJECTOR WATER
SUPPLY INLET
Courtesy Wallace & Tiernan Inc.
Figure 8-4 • V-notch Chlorinator
interference from various substances cause it reduction potential. Additional instruments

to be used with certain caution. A modifica- and methods of control are continuously being
tion of this test known as the OT A test has developed for this very important application.
been widely used although certain inherent
weaknesses continue to exist. The micro- REFERENCES
titration of free chlorine with methyl orange "Chlorine Manual", 3rd Ed., The Chlorine Insti-
or xylene cyanole has been successfully used, tute, Inc., New York, N.Y. ( 1959)
especially in cooling water systems. A test for A. E. Griffin and N. S. Chamberlin, "Some Chemi-
cal Aspects of Breakpoint Chlorination", Journal,
titrating chlorine with a reducing agent has New England Water Works Assoc., Vol. LV, pp.
been developed and is generally known as the 371-383 (1941)
amperometric method and is gaining in ap- E. J. Laubusch, "Chlorination of Water", Water &
plication and popularity. Fortunately, resid- Sewage Works, Vol. 105, pp. 411-417 (1958)
E. W. Moore, "Fundamentals of Chlorination of
ual chlorine recording can now be made Sewage and Waste", Water & Sewage Works, Vol.
with the instrumentation presently available. 98, pp. 130-136 (1951)
Some industrial waste chlorination treat- C. N. Sawyer, "Chemist~y for Sanitary Engineers",
pp. 246-256, McGraw-Hill Book Company, Inc.,
ment can be controlled by the oxidation- New York, N.Y. (1960)
9
Hydrogen Sulfide Ren1oval
Difficulties Encountered. . . . . . . . . . . . . . . . . . . . . 56
Methods for Removal . ..................... 56
Aeration . ............................. 56
Aeration with pH Reduction . . . . . . . . . . . . . . . 57
Chlorination. . . . . . . . . . . . . . . . . . . . . . . . . . . 58
Anion Exchange . . . . . . . . . . . . . . . . . . . . . . . . 58
Analysis for Sulfide. . . . . . . . . . . . . . . . . . . . . . . . 59
Application and Limitations. . . . . . . . . . . . . . . . . 59
56 BETZ HANDBOOK
YDROGEN sulfide removal involves the Figure 9-1

H use of mechanical or chemical processes
to secure the elimination of this undesirable
Percentage of Total Sulfide Present as H2S
gas from solution. pH

Hydrogen sulfide may be present in a water 5.0 .................................... 98
supply due to purely chemical or purely bi- 6.0 ..................•................. 83
ological causes. The presence of hydrogen 6.5 ..••.•.•••..••...•.••.•...••.......• 61
7.0 .................................... 33
sulfide is more common to well supplies than 7.5 ..................................... 14
to surface supplies. In well waters the hydro- 8.0 .................................... 4.8
gen sulfide results generally from the location 9.2 .................................... 0.32
such as Boulder clay areas or "sour gas" oil
fields. In other supplies, it is generally the re- a corrosive character to the water.
sult of contamination due to local atmos- Hydrogen sulfide accounts for a well oc-
pheric conditions, the presence of some types casionally yielding a water that is black due
of bacteria in the absence of oxygen, or the to ferrous sulfide. Ferrous sulfide is usually ex-
contamination of mine waters by sulfur. tremely finely divided and only a small amount
Bacteria often encountered in industrial is required to blacken the water. Excessive
waters may be responsible for the presence hydrogen sulfide concentrations exhibit an ag-
of hydrogen sulfide. The most common of the gressive tendency toward iron, steel, and cop-
sulfur bacteria are those that split hydrogen per alloys, even in the absence of oxygen.
sulfide from organic matter containing sulfur. Greensand zeolites are damaged by the pas-
Most protein matter contains sulfur, therefore sage of water containing hydrogen sulfide,
hydrogen sulfide will be produced when these Cases have been reported where the mineral
compounds are broken down by bacteria in has been so badly attacked in a few months
the absence of oxygen. One group of sulfur by water containing only 2 ppm of hydrogen
bacteria produces hydrogen sulfide by there- sulfide that the bed could not be recondi-
duction of inorganic sulfates, sulfites and sul- tioned. Special minerals for the catalytic re-
fur itself. moval of iron and manganese from water are
Hydrogen sulfide possesses a definite solu- also affected by the presence of hydrogen sul-
bility in water which is approximately three fide. With the use of these materials, manu-
times that of carbon dioxide and imparts a facturer's guarantees are based on the absence
characteristic taste and odor (rotten eggs) to of this gas.
the water. This odor becomes objectionable Another effect of hydrogen sulfide is to re-
when present in a concentration of 1 ppm act with chemicals used for the protection of
or more and can be detected when present in systems against corrosion thus increasing the
amounts less than 1 ppm. dosage required and, in some instances, in-
If the water has a high pH, the odor may creasing cost to an extent that treatment of the
be slight, since most of the sulfide may be system with positive corrosion inhibitors is not
present as alkaline sulfides rather than as economically feasible.
hydrogen sulfide. Figure 9-1 shows the per-
centage of total sulfide present as hydrogen sul- METHODS FOR REMOVAL
fide with the difference in the total sulfide con-
tent being that present as the alkaline sulfides. AERATION. The removal of hydrogen sulfide
by aeration follows Henry's law. Hydrogen sul-
DIFFICULTIES ENCOUNTERED fide is relatively soluble in water while the par-
tial pressure of this gas in the normal atmos-
Removal of hydrogen sulfide is desirable not phere is practically "zero." The establishment
only from the standpoint of odor, but because of a state of equilibrium between the water
even small concentrations of this gas impart and the air will result in the saturation of the
water with nitrogen and oxygen. The basic sulfides and the sulfide as discussed previously.
fundamental in securing removal of gases from By reducing the pH of the water, the ioniza-
water by aeration is the establishment of a tion constant is shifted and more of the total
state of equilibrium between the gases in the sulfide is converted to hydrogen sulfide in
water and the gases in the surrounding at- which form it is available for removal by
mosphere. aeration.
Reduction in hydrogen sulfide content is In recent yea~s flue gases which are high in
obtained with many different means of aera- carbon dioxide content have been employed
tion. The absorption of oxygen by the water in reducing the pH value sufficiently to liber-
during aeration also assists hydrogen sulfide ate hydrogen sulfide. The quantity of gas re-
removal by accomplishing oxidation of a part quired for pH reduction will depend upon
of the hydrogen sulfide and alkaline sulfides the alkalinity of the supply being treated.
to free sulfur. Forced draft aeration has been One type of equipment for this process con-
found desirable where there is an appreciable sists of a double stack with a water seal sepa-
hydrogen sulfide concentration to be removed. rating the upper carbonating chamber from
Experience has indicated the desirability of the lower aeration chamber. The flue gas is
considerable cross-sectional area and a tall introduced at the bottom of the carbonating
aeration tower. With such equipment hydro- chamber and mixes with the influent water
gen sulfide can be reduced to 1-2 ppm. The flowing counter-current. Most of the hydro-
amount of free sulfur precipitated may prove gen sulfide is removed at the reduced pH in
objectionable and require filtration for its the upper section. In the lower section, which
removal. is similar in design to the carbonating cham-
Two general types of aerators are employed ber, a counter-current of air provides addi-
for hydrogen sulfide removal; one a pressure tional scrubbing of the water with air alone
type aerator and the other a gravity type aer- for the removal of carbon dioxide and further
ator. The pressure type has the advantage removal of any residual hydrogen sulfide. Also,
that line pressure is not broken. The gravity it is possible that any remaining alkaline sul-
aerator requires that line pressure be broken; fides will be oxidized to sulfur. This type of
however, more efficient aeration is obtained. hydrogen sulfide degassifying apparatus is il-
At present the tendency is toward the installa- lustrated by Figure 9-2. The introduction of
tion of gravity aerators. carbon dioxide at the bottom of the carbonat-
In the use of aeration alone for the removal ing chamber permits a counter-current flow
of hydrogen sulfide, complete removal of this of gas and water through the entire length of
gas is not obtained. At higher pH values the the chamber, thus permitting maximum con-
sulfide content will be present chiefly as al- tact time between the gas and the water. The
kaline sulfides which will result in incomplete taller the tower used the greater is the con-
removal by aeration alone. This condition re- tact time provided.
sults because carbon dioxide is less soluble than Reduction in pH of the water may also be
the hydrogen sulfide and is removed by aera- obtained by the feed of a mineral acid. The
tion more rapidly than hydrogen sulfide. As feed of acid is controlled to obtain the desired
the carbon dioxide is removed, the pH rises pH reduction. This method provides a posi-
and the equilibrium between alkaline sulfides tive control over pH reduction.
and hydrogen sulfide is altered so that the re- By combining pH reduction with aeration,
action proceeds in the wrong direction for the the size of the forced draft aerator can be
most complete removal of hydrogen sulfide. considerably reduced. More effective hydro-
AERATION WITH pH REDUCTION. Removal of gen sulfide removal is also secured. However,
hydrogen sulfide by aeration is most effective with this method some precipitation of flowers
at reduced pH in the range of 4.0 to 5.0 be- of sulfur can be expected. Coagulation and
cause of the equilibrium between the alkaline filtration may be necessary if the initial total
58 BETZ HANDBOOK
aeration or aeration with pH adjustment may

not provide sufficient treatment. Complete re-
moval of hydrogen sulfide frolJl water can be
secured by means of chlorination or a com-
bination of aeration and chlorination. One
reaction with chlorine is as follows:
HzS + Clz =
hydrogen sulfide + chlorine =
2HCI + S
hydrochloric acid + sulfur
In this reaction, the formation of flowers

of sulfur is shown. This fine colloidal precipi-
tate should be removed by coagulation and
filtration to prevent reversion to hydrogen sul-
fide by the action of certain microorganisms.
Theoretically, 2.1 ppm of chlorine are re-
quired for the removal of each ppm of hydro-
gen sulfide.
When used in greater quantities chlorine
will further oxidize hydrogen sulfide to sul-
furic acid without the precipitation of sulfur.
This reaction is shown below:
H2 S + 4C)z + 4H 20 =
hydrogen sulfide + chlorine + water =
HzS04 + 8HCI
sulfuric acid + hydrochloric acid
Theoretically, 8.4 ppm of chlorine are re-
quired for the removal of each ppm of hydro-
gen sulfide. This high chlorine consumption
considerably increases the cost of this method
Figure 9-2 • Hydrogen Sulfide Degasi.fier of hydrogen sulfide removal. Consequently, it
is usually the more economical procedure to
sulfide content of the water is high. remove as much hydrogen sulfide as possible
by aeration with reduced pH and to use com-
CHLORINATION. While aeration alone and aer-
plete chlorine oxidation, if necessary, for re-
ation combined with pH reduction will reduce
moval of the residual hydrogen sulfide after
the hydrogen sulfide content of a water sup-
aeration.
ply to a relatively low value, depending upon
the amount of gas originally present, the hy- ANION ExcHANGE. Anion adsorbers have re-
drogen sulfide content of a water can not be recently been developed which are applicable
duced to "zero" by either of these methods. for the removal of hydrogen sulfide from water
Depending on the use of the water, complete supplies. Highly basic resins have been pre-
removal of hydrogen sulfide may be required pared which, when regenerated with salt and/
as even small quantities produce an offensive or sodium bicarbonate depending upon the
odor and result in increasing the corrosive pH of the water, exchange their ions for sul-
characteristics of the water. Consequently, fides, sulfates, etc. This reaction is illustrated
by the following equation in which "R" repre- acetate solution ( 220 g Zn ( C 2 Ha0 2 ) 2 .2H2 0
sents the resin matrix: per liter of distilled water) and 1N sodium
R4N.OH + H2S hydroxide solution (40 g NaOH per liter of
basic exchange resin + hydrogen sulfide = distilled water) . To fix a sample, place 1 ml
R4N.HS + H20 of zinc acetate in a 16 oz bottle. Collect the
sulfide resin + water sample (approximately 16 oz) in this bottle
with a minimum of agitation. Add 1 ml so-
In accomplishing this removal a low cost
dium hydroxide, stopper and mix. The fixed
regenerant is employed, line pressure is not
sample should be analyzed within 24 hours
broken, excess oxygen is not added to the
for maximum accuracy. The method of analy-
water, and a minimum of equipment is re-
sis usually employed is colorimetric, based
quired.
upon the reaction between paraaminodi-
The equipment employed is a pressure type methylaniline, ferric chloride and the sulfide
softener shell designed for counter-current ion to produce methylene blue.
flow utilizing highly basic anion exchange
resins such that dealkalization will also be ob- APPLICATION AND LIMITATIONS
tained and the chloride content of the water
will be increased. By counter-current flow is In general, the method or combination of
meant the up flow of water during the service methods employed for hydrogen sulfide re-
run and the down flow of water and regen- moval involves individual study. The method
erant for rinsing and regeneration. of removal is determined by such factors as
At present the use of the highly basic anion the total sulfide content of the water, alkalin-
resins for hydrogen sulfide removal has been ity of the water, desired effluent sulfide con-
limited to domestic uses though it is concentration, use or purpose for which the water
sidered that they are satisfactory for some in- will be employed, existing plant equipment,
dustrial uses. They have the limitations of low etc.
capacity, low flow rates, and iron must be ab- The methods discussed in detail above are
sent from the water. It is preferable for the the ones in general use. However, there are
influent water to be low in hardness. However, other methods employed in specific cases
the process can be employed with the use of which are too limited for general application
a dual bed to both soften the water and re- and which in general are indirect methods of
move hydrogen sulfide. It is also possible hydrogen sulfide removal. Examples of these
in some cases for existing zeolite softeners to methods are the removal of sulfide by ozona-
be converted to perform both hydrogen sulfide tion, usually accomplished as an indirect re-
removal and softening of the water. sult of the sterilization of water, bacterial ac-
tion on sulfides over long periods of time fol-
ANALYSIS FOR SULFIDE lowed by sterilization and filtration, and slow
aeration in an open reservoir followed by the
When dissolved gases are present in a water use of a coagulant, sterilization, and filtration.
supply, it usually is necessary to analyze the
water at the point of origin to obtain accurate REFERENCES
results. For example, this procedure is manda- M. E. Flentje, "Aeration", Journal, Am. Water
Works Assoc., Vol. 29, pp. 872-880 (1937)
tory if the gases are carbon dioxide or dis-
J. E. Foxworthy and H. K. Gray, "Removal of Hy-
solved oxygen. The determination of hydro- drogen Sulfide in High Concentrations from Water",
gen sulfide and total sulfide is a rather in- Journal, Am. Water Works Assoc., Vol. 50, pp. 872-
volved analysis which is not readily performed 878 (1958)
S. T. Powell and L. G. von Lossberg, "Removal of
in the field. Fortunately, samples for sulfide Hydrogen Sulfide from Well Water", Journal, Am.
can be "fixed" at the source and then trans- Water Works Assoc., Vol. 40, pp. 1277-1289 (1948)
ported to a laboratory for analysis. S. W. Walls, "Hydrogen Sulfide Problems of Small
Water Systems", Journal, Am. Water Works Assoc.,
The fixing reagents employed are 2N zinc Vol. 46, pp. 160-170 (1954)
60 BETZ HANDBOOK
10
Iron Ren1oval and
Iron Retention
Methods of Iron Removal . .................. 61
Aeration, Sedimentation and Filtration. . . . . . . . 61
Coagulation, Sedimentation and Filtration . . . . . . 62
Lime and Lime-Soda Softening . . . . . . . . . . . . . 63
Cation Exchange . . . . . . . . . . . . . . . . . . . . . . . . 63
Contact Filtration. . . . . . . . . . . . . . . . . . . . . . . 64
Iron Retention . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
Iron Bacteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Manganese Removal . ...................... 67
Applications and Limitations. . . . . . . . . . . . . . . . . 67
RON removal is the process by which objec- change surfaces become fouled with iron oxide
I tionable amounts of iron are removed from
water either in one step or a series of steps.
deposits. Valves, lines, etc. will be plugged by
the precipitated iron. Ferric iron present in
Iron retention is the treatment of iron- water to be zeolite softened results in fouling
bearing water with surface active agents, of the zeolite beds and loss of efficiency.
capable of forming a complex with iron, Where iron is present in boiler feedwater it
thereby retaining iron in solution and pre- constitutes another source of potential scale
venting precipitation which would otherwise forming tendencies.
occur. In many industries, the presence of iron in
process water may be highly undesirable. In
the paper, tanning, textile, beverage and ice
manufacturing industries, for example, the
presence of iron is highly detrimental to the
product.
METHODS OF IRON REMOVAL
Iron can be removed from water by several

different processes and combinations of proc-
esses. In fact, there are so many different
methods that can be employed for iron re-
moval that careful study is required to select
a method most suitable to any individual case.
Selection of an iron removal method will be
influenced not only by the amount of iron in
the raw water and the form in which the iron
is present, but also by the ultimate use of the
water and the degree of softening that may
Figure 10-1 • Typical Iron Deposition in Pipe Line also be desired.
AERATION, SEDIMENTATION AND FILTRATION.
Iron in water will deposit in water works Aeration is a process which consists of inti-
distribution systems increasing the frictional mately admixing water and air in some man-
resistance through pipes; water meters will ner. Where aeration is employed for iron re-
become heavily coated with iron oxide and moval, two actions occur simultaneously. First,
must be cleaned regularly in order to register carbon dioxide is removed by aeration, thus
correctly. If water containing large amounts increasing the pH of the water. Secondly, the
of iron is used in the home, it will injure the oxygen added to the water by aeration enters
color and taste of many edibles and will des- into the chemical oxidation of the iron. As a
troy the sanitary appearance of sinks and result of these two actions, precipitation of
cooking utensils. Clothes washed with such a the iron occurs as ferric hydroxide, illustrated
water will develop yellow stains. Crenothrix by the following equation.
or "iron bacteria" frequently accompany 4Fe(HC03h + 0 + 2H 0 =
2 2
waters of high iron content. These organisms ferrous bicarbonate + oxygen + water =
established in a city distribution system cause 4Fe(0Hh + 8C0 2
clogging of service mains and are responsible ferric hydroxide + carbon dioxide
for disagreeable odors. Iron is frequently present in well waters in
Where well waters containing iron are used the ferrous or unoxidized state. Increase in
in industrial cooling water operations, deposi- pH in the absence of oxygen would precipi-
tion of iron from solution will occur. Heat ex- tate ferrous hydroxide, but since ferrous hy-
62 BETZ HANDBOOK
TO SERVICE
FILTER
AERATOR
WELL PUMP
SEDIMENTATION
BASIN
Figure 10-2 • Iron Removal by Aeration, Settling and Filtration
droxide is more soluble than ferric hydroxide, ing relatively small amounts of iron and car-
complete iron removal requires oxidation of bon dioxide.
ferrous iron to the ferric state, with precipita- Under some circumstances, chlorine and
tion of the highly insoluble ferric hydroxide. other oxidants have been employed instead
Complete precipitation of iron by aeration of aeration in order to oxidize and precipitate
is not instantaneous. For this reason, it is ad- iron. Usually, such methods cannot compete
visable to employ a retention and sedimen- economically with aeration.
tation tank following the aerator to provide While the insoluble ferric hydroxide pre-
time for the precipitation and settling of ferric cipitated from an iron-bearing water may be
hydroxide, prior to filtration. Sludge settled removed by settling alone if sufficient time is
in this tank is removed periodically. provided, it is better practice to follow with
Even with an adequate supply of oxygen, filtration in order to insure complete iron re-
the precipitation of ferric hydroxide is hin- moval in a reasonable period of time. Stand-
dered by low pH values. For effective iron re- ard pressure or gravity filters are employed
moval by aeration the pH should be increased for this purpose, using either sand or anthra-
to approximately 7.5. If aeration alone is un- cite coal as the filter medium.
able to raise the pH to this point, the supple- CoAGULATION, SEDIMENTATION AND FILTRA-
mental feed of an alkali is required, following TioN. Some waters, usually surface supplies
aeration, to complete the precipitation of fer- and particularly those from swampy areas,
ric hydroxide. may contain appreciable organic matter which
Coke-tray aerators are frequently used in acts to retard or prevent the precipitation of
the removal of iron. Other types of aerators iron. Even following aeration and pH adjust-
such as cascades, cones, nozzles and towers ment, little removal of iron may be secured.
are also employed. Provision must be made to Chlorination may be of value because of its
periodically clean the unit of the precipitated destructive action on the organic present, but
iron. Figure 10-2 shows a flow sheet for iron may be costly and not fully effective. Where
removal by aeration, sedimentation and fil- such waters are encountered, coagulation is
tration. usually the most effective measure, followed by
To avoid re-pumping, aeration may be con- sedimentation and filtration. The coagulant
ducted under pressure. Carbon dioxide is not most widely employed is alum although it is
removed and an alkali feed may be required necessary to conduct tests to determine the
to precipitate the iron. This method of iron most effective and economical coagulant for
removal is applicable only to waters contain- each particular water.
LIME AND LIME-SODA SoFTENING. Where Removal of iron by cation exchange is simi-
either lime softening or lime-soda soften- lar to and simultaneous with the removal of
ing is employed, the pH of the treated calcium and magnesium. The cation exchange
water is elevated sufficiently high to provide material may be green sand, carbonaceous or
the softening reactions and iron is precipitated resinous. When operated on the sodium cycle,
as ferric hydroxide. Aeration is usually pro- iron removal takes place as follows:
vided prior to softening to supply any oxygen Fe(HC03h + NazZ
needed for the conversion of ferrous to ferric ferrous bicarbonate +
sodium zeolite =
iron. Aeration is frequently desirable prior to FeZ + 2NaHC03
cold process softening, regardless of any iron ferrous zeolite +
sodium bicarbonate
problem, in order to remove carbon dioxide Regeneration reactions with salt are as follows:
and thereby decrease lime requirements. FeZ + 2NaCI
The aerator frequently is located over the ferrous zeolite + sodium chloride =
lime-soda softener. As the pH of the water is Na2Z + FeCI2
increased by the feeding of lime, oxidation of sodium zeolite + ferrous chloride
iron is rapid. Precipitation and sedimentation Similar softening and regeneration reac-
can take place in the same unit, followed by tions take place when the cation exchanger is
filtration as illustrated in Figure 10-3. operated on the hydrogen cycle in which case
Iron removal, of course, is readily accom- regeneration is secured with the use of sulfuric
plished under the temperature and pH condi- acid.
tions existing in hot process lime or lime-soda While iron removal by cation exchange
softening. Iron, precipitated in such units, will possesses a field of usefulness where complete
serve to aid coagulation. hardness removal is also necessary, it is not
Iron removal in conjunction with cold proc- suited to those applications where hardness
ess lime or lime-soda softening may be fol- removal is not required. It is possible, how-
lowed, on occasion, by acidification and zeolite ever, to overrun the cation exchanger from the
softening where the use of the treated water standpoint of hardness removal while con-
requires low hardness. tinuing to secure iron removal.
CATION ExcHANGE. With certain waters, cat- In order to avoid fouling the exchanger bed
ion exchangers may be employed for the re- with ferric hydroxide, it is necessary that the
moval of soluble iron from water. The cation iron content of the influent water be present
exchange process is limited essentially to well only as soluble ferrous iron. Air must be ex-
waters where the iron content is in the ferrous cluded from contact with the well water prior
state. to the exchange unit.
RAW GRAVITY FILTER

WATER PUMP
,PUMP TO SERVICE
CLEAR WELL
Figure 10-3 • Iron Removal by Lime Treatment

64 BETZ HANDBOOK
Cation exchange is applied only to waters of in the process of iron removal. Backwashing
limited iron content and is not recommended of the precipitated iron from the bed is also
for the removal of high iron concentrations. necessary. Filter rate is 3 gallons per square
CoNTACT FILTRATION. Contact filtration ac- foot per minute and backwash rate is a mini-
complishes oxidation and filtration of the iron mum of 8 gallons per square foot per minute.
in one step. Two types of contact filtration are Potassium permanganate is an expensive re-
employed. In one type, the material of the fil- generant chemical and manganese zeolite in-
ter bed simply acts as a catalyst for the reac- stallations are usually limited to waters of low
tion between the iron and oxygen in the water. iron content or employed as a polishing de-
In the other type, the filter bed supplies the vice, following other processes which have re-
oxygen needed for oxidation of the iron, and moved the major portion of the initial iron
this oxygen must be periodically replenished content of the water.
by use of a regenerant chemical.
IRON RETENTION
Catalytic beds usually involve a filter me-
dium containing manganese dioxide. Stand- For many industrial water uses, it is not neces-
ard pressure or gravity filter shells are em- sary that the iron be removed, provided the
ployed with the 30-36 inches catalytic bed water can be stabilized to prevent or retard
supported on gravel. Flow rates are limited precipitation. Well waters, high in iron con-
to 4 gallons per square foot per minute or less tent, are frequently employed for cooling pur-
and backwash for removal of precipitated iron poses on a once-through basis. By stabilizing
is at the rate of 12-15 gallons per square foot the iron content, deposits can be prevented
per minute. Units of this type present a simple throughout the system. The cost of stabilizing
method of iron removal since no chemical the iron is frequently less than the cost of iron
treatment is required. Proper functioning of removal and, in addition, the investment in
such units, however, requires that the water equipment is avoided.
be low in organic matter, free of hydrogen Stabilization of iron can be secured with the
sulfide, possess a minimum of pH of 6.5 and use of surface active agents, such as polyphos-
contain sufficient oxygen to react with the phates and organic sequestrants. A complex
iron present. If there is inadequate oxygen· is formed between the iron present in the
content of the raw water, oxygen must be sup- water and the surface active agent added. By
plied by an aerating device. This aeration can this complex formation, precipitation of ferric
be accomplished under pressure, without need hydroxide can be delayed or, with sufficient
for repumping. treatment concentration, prevented indefi-
Manganese zeolite is an example of contact nitely. Usually, economic considerations limit
filtration where the filter bed itself supplies such stabilization to once-through systems. In
the needed oxygen for iron removal. Manga- recirculating cooling water systems, the con-
nese zeolite is green sand which has been tinuous aeration and long retention period so
treated with manganous chloride to convert it increases the quantity of stabilizing agent re-
to manganous zeolite. Then, treatment with quired as to make iron removal from the
potassium permanganate results in the pre- makeup water the more desirable practice.
cipitation of the higher oxides of manganese The quantity of surface active agent re-
on the zeolite grains. quired to retain iron in solution will vary ac-
No removal of hardness is involved with the cording to such factors as the iron content, al-
manganese zeolite process. The oxidation re- kalinity, pH and solids content of the water.
actions occur between the natural iron con- Another important factor is the temperature
tent of the water and the manganese oxides to which the water will be heated. For higher
on the zeolite grains. Regeneration of the unit temperatures, increased treatment is required
with potassium permanganate is required, as to overcome the increased tendency for pre-
the oxidizing ability of the bed is consumed cipitation. In general, the organic iron reten-
tion agents are more effective than the poly- most economical, it· is necessary that the sta-
phosphates. Another advantage of the organic bilizing agent be fed to the water prior to a
agents is that they can be used at higher point where precipitation has already occurred
concentrations, on waters of high iron and and that the amount be limited to that neces-
high hardness content, without danger of pre- sary to retain iron in solution under the con-
cipitating calcium phosphate. Under certain ditions of that individual system.
circumstances, a combination of organics and
polyphosphates represents the most advan-
IRON BACTERIA
tageous treatment.
;fhe retention time of the water in the sys- The iron bacteria are an unusual group of
tem is another important variable affecting microorganisms found in industrial waters,
the treatment concentrations required. If, for streams, lakes, wells and in potable water
example, a simple system of well water supplies. These organisms may occur alone or
pumped through a single heat exchanger and in different combinations with other species
then to waste is involved, then there is little of bacteria, fungi and algae.
retention time in the system. Only sufficient Although the iron bacteria have not been
treatment may be needed to retain iron in studied as completely as some of the other
solution for 10-15 minutes. On the other hand, organisms this lack of study does not indicate
in a large plant with many varied cooling that they have not caused serious difficulty in
operations, and numerous smaller lines at low industrial and fresh water systems. There are
flow rates, it may be necessary to feed suffi- many case histories regarding the effect of
ci~nt stabilizing ~gent to retain iron in solu- these iron bacteria in city and industrial water
tion for 24 hours. The cost of iron retention systems. The published reports on the effect
treatment increases as the retention time in of iron bacteria in city water systems are many
the system increases and all such data are re- and are justifiably referred to as "water ca-
quired for accurate treatment estimations. lamities." The universal distribution of the
The proper surface active agents are ef- iron bacteria is borne out by the development
fective in preventing iron precipitation re- of these water calamities in Europe and in the
gardless as to whether the iron is in the ferrous United States. Recently these microorganisms
or ferric state. It is advisable, however, to have gained recognition in industry by their
introduce the stabilizing material to the water effect upon process equipment and on the
prior to its exposure to air and prior to chlori- finished product. Usually the affected water
nation in order to permit complex formation will become slightly turbid or acquire a light
with the iron as quickly as possible. reddish tint and/or objectionable odor. As
Equipment for feeding iron retention agents these microorganisms increase in number the
can be relatively simple. Where the raw water water may become more turbid, and the color
is at a constant flow rate, a chemical solution of the water will become brick-red. Hence the
tank for one day's treatment supply can be common reference to "red water." In addition
used together with a small constant rate to discoloring the water this group of micro-
chemical pump. If the raw water flow is vari- organisms produces undesirable accumula-
able, then a proportional method of chemical tions in pipes, nozzles, spray ponds, etc. These
feed should be employed so as to permit the deposits in time will slough and plug lines,
most economical use of treatment. foul pumps, valves and/or affect the quality
Iron retention agents are available which of the finished product.
possess such powerful ability to form iron com- The growth and development of iron bac-
plexes that iron can be retained in solution teria may manifest itself in several different
practically indefinitely. Such agents also pos- ways. For example, it has been reported that
sess the power to dissolve previously precipi- these microorganisms have reduced the ef-
tated iron. However, for iron retention to be fective area of a 6 inch pipe to that of a 2 inch
66 BETZ HANDBOOK
pipe in a matter of weeks. The accumulation Crenothrix, Clonothrix, Gallionella, Sidero-

in pipes, ponds, nozzles and so forth may be capsa, Siderosphaera, Sideronema, Ferribac-
hard and crusty or it may be relatively light terium, Sideromonas, N aumqnniella, Ochro-
and spongy in appearance. Often other bio- bium, Siderococus, Siderobacter and Ferro-
logical slimes may be associated with iron bacillus.
bacteria and thus further reduce or restrict In general, the iron bacteria are extremely
the flow of water through pipes or entrap other difficult to culture in the laboratory using
debris. The association of other species of the usual culture media. Therefore it is not
microorganisms with the iron bacteria will in- possible to evaluate or to determine the num-
crease turbidity, increase color, give rise to ber of these microorganisms in different
objectionable odors and tastes in domestic water supplies with the usual bacteriological
water supplies and also cause sloughing of de- procedures. A satisfactory procedure for de-
posits in industrial systems which may affect termining the population index of these micro-
the process or the product being produced. organisms in water supplies is a direct micro-
The iron bacteria are usually considered scopic analysis of the water. This analysis,
as typical fresh water organisms. Recently, however, will only determine the number of
however, some of the iron bacteria have been the more obvious species of this group of
isolated from high brines, which indicates the microorganisms.
versatility of this class of organisms. These In general, the iron bacteria show prefer-
microorganisms are usually considered as ence for the lower temperatures; these organ-
aerobic, but they have also been found to grow isms have been observed to grow at tempera-
in waters with low oxygen content. Although tures which range from 0 C to 40 C. How-
the iron bacteria are widely distributed in na- ever, their optimum temperature will range
ture and are usually found in wells, cool from 6 C to 25 C.
springs and brooks they have also been isolated Hydrogen ion concentration has a great
from marine waters, and connate waters (high effect upon the growth and development of
brines) . Although they are usually found in these bacteria. Although published reports
waters which are reasonably high in iron they show that the different species have different
may be occasionally found in waters which requirements, the pH range for growth will
contain a very low amount of iron. vary from 5.5 to 8.2. The optimum pH ap-
The prinicipal distinguishing characteris- pears to be approximately 6.5. Much has been
tics between the iron bacteria and other types written regarding the sensitivity of the iron
of microorganisms is that they have the capac- bacteria to light but their general reaction to
ity to absorb and to accumulate iron and/or light appears to be doubtful. These organisms
manganese when grown in environments have been found to grow in exposed areas, in
which contain these elements. These organ- shade and in complete darkness such as in
isms deposit iron and manganese salts around pipes, wells, etc.
'their cells which results in the characteristic Obviously the growth and development of
reddish brown to black color. iron bacteria and the deposition of iron in
These autotrophic (self-sufficient) organ- their sheaths is influenced by the amount of
isms oxidize iron compounds as a source of dissolved iron in their immediate environ-
energy. However the amount of energy re- ment. The iron bacteria have been found in
leased by the oxidation of iron is relatively waters with as much as 30 ppm of iron and
small. they have also been found in waters which
On the basis of their morphological charac- contain as low as 0.1 ppm of iron. In the low
teristics alone the iron bacteria can be placed concentrations the growth and development
into several different categories. The most of the iron bacteria is dependent upon the
common iron bacteria fall in the following flow of water or where the iron supply is
genera: S phaerotilus, Leptothrix, Toxothrix, continually renewed.
The control of iron bacteria in fresh and for its precipitation. However, coagulation,
salt water systems is possible with the use of lime and lime-soda softening, cation exchange
biocides. The selection of the most appropriate and contact filtration are effective methods
biocide will, of course, depend upon the en- for removal of manganese, as well as Iron.
vironment in which the organism is growing,
the general conditions of use of the water, and APPLICATIONS "'NO LIMITATIONS
whether the water is for industrial or domestic
use. Obviously the use of control agents in There are many combinations of treatment
domestic systems must be limited to those methods that can be employed for iron re-
which have the approval of the Public Health moval. A detailed study of each individual
Service. Under the usual operating conditions plant problem is necessary in order to select
for city water supplies the most suitable agent the particular system best suited to the in-
is chlorine. However, in industrial systems it dividual case. An iron removal method can-
is possible that biocides other than chlorine not be selected solely on its efficiency with
will give satisfactory microorganism control. respect to iron, but must be correlated with
In some industrial systems, where the use of other quality requirements of the water, such
chlorine is not indicated because of precipita- as the need also for low hardness in the
tion of iron and manganese, it may be neces- treated water.
sary to use a special proprietary formulation. Where there are both potable and indus-
These proprietary formulations, although usu- trial uses involved, it may be desirable to em-
ally more expensive than chlorine, will pro- ploy two different methods of iron removal.
duce the desired effect. It is, of course, Also, in some cases, iron removal may be
essential that these industrial biocides be com- employed for potable and process water,
patible with iron retention formulations and while the larger volume of cooling water is
with corrosion inhibitors. In addition, the treated for iron retention rather than iron
biocides must not affect the quality or manu- removal. A careful study, relating all treat-
facture of the product. ment methods to the individual plant needs,
is highly advisable because of the many dif-
ferent methods that can be employed both for
MANGANESE REMOVAL
iron removal and iron retention.
Manganese removal is closely associated with
iron removal. Fortunately, manganese is not REFERENCES
encountered as frequently as iron in water
S. B. Applebaum, "Iron and Manganese Removal",
sources and usually is present in smaller Proceedings, Engineers Soc. of Western Penna., pp.
quantities. However, the methods used for 49-57 ( 1955)
manganese removal are essentially the same R. B. Conlan, "Correcting the Iron Problem in
as for iron removal. In most waters of ap- Water Systems", Heating and Ventilating, Vol. 48,
pp. 77-79 (Dec. 1951)
preciable manganese content, iron is also "Water Quality and Treatment", 2nd Ed., pp. 359-
present so that the treatment must be such 368, Am. Water Works Assoc., New York, N.Y.
as to remove both undesirable constituents. ( 1950)
R. S. Wolfe, "Cultivation, Morphology and Classi-
Manganese is less readily oxidized than fication of the Iron Bacteria", Journal, Am. Water
iron by aeration and a higher pH is required Works Assoc., Vol. 50, pp. 1241-1249 (1958)
68 BETZ HANDBOOK
11 Chemislly Involved. . . . . . . . . . . . . . . . . . . . . . . . 69
Use of Lime and Soda Ash . . . . . . . . . . . . . . . . 69
Lillle-Soda Softening
Use of Lime . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
Use of Caustic Soda . ..................... 71
Use of Gypsum . ........................ 72
Propagating the Chemical Precipitating Reactions. . 72
Intermittent or Batch Cold Softeners. . . . . . . . . 73
Continuous Sedimentation Type. . . . . . . . . . . . . 73
Continuous Sludge Contact Type. . . . . . . . . . . . 73
Continuous Hot Process Lime-Soda Softening. . . 74
Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Limitations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
TIME-SODA softening is the process by which very large quantltres of water are softened,
L the calcium and magnesium salts con- such as a city plant; it is usually more eco-
stituting the hardness content of a water are nomical to employ unslaked lime-GaO-
chemically precipitated and removed through which is first slaked to form hydrated lime at
the use of lime (calcium hydroxide) and the plant in accordance with the following
soda ash (sodium carbonate) . This process reaction:
may be carried out at normal raw water {1)
temperatures, in which case it is referred to CaO + H20 =
as "cold process" or at temperatures near or unslaked lime (calcium oxide) + water =
above the boiling point, referred to as "hot Ca(OHh
process." hydrated lime (calcium hydroxide)
The lime-soda ash process is based upon Soda ash-Na 2 C03 -of 98% purity is em-
four essential factors: ployed in lime-soda softening at the usual
Proper selection of specific reagents to ef- specification of 58% Na2 0 content.
ficiently precipitate part of the impurities UsE OF LIME AND SoDA AsH. Hydrated lime
to be removed in the form of an insoluble reacts to chemically precipitate the carbonate
sludge and to convert the remainder to a hardness present in water, thereby producing
soluble form. insoluble precipitates of calcium carbonate-
Correct proportioning of the specific chem- CaCOa - and magnesium hydroxide -
ical reagents to the raw water. Mg(OH) 2-without the production of any
·Propagating the chemical precipitating re- soluble by-products. These suspended precipi-
actions by a suitable combination of coagu- tates are removed by sedimentation and fil-
lation, sedimentation, heat or sludge con- tration and thus a marked reduction in solids
tact. content is obtained.
Filtration of the chemically softened or The carbonate hardness of a water is
finished water to remove traces of turbidity composed of the carbonates and bicarbonates
or suspended matter resulting from the of calcium and magnesium. Their precipita-
process. tion by means of hydrated lime proceeds in
Lime-soda softening is employed for the accordance with the following reactions:
removal of hardness in order to minimize (2)
scale and sludge in boilers, reduce calcium Ca(HC03h + Ca(OHh =
carbonate deposition in heat exchange sys- calcium bicarbonate + calcium hydroxide =
tems and cooling water systems, and in gen- 2CaC03 + 2H 20
eral to remove hardness wherever hardness calcium carbonate + water
in water is a source of difficulty either in in- (3)
dustrial process water, boiler feedwater or Mg(HC03h + 2Ca(OHh
city water supply. magnesium bicarbonate + calcium hydroxide
Incidental to the removal of hardness in Mg(OHh + 2CaC03
the proper operation of lime-soda softening, = magnesium hydroxide + calcium carbonate
iron, free carbon dioxide and turbidity are + 2H20
removed. +water
Hydrated lime also reacts with the mag-
CHEMISTRY INVOLVED nesium salts constituting a portion of the
non-carbonate hardness of water. In these
Hydrated lime-Ca(OH) 2-of 90% purity reactions, insoluble magnesium hydroxide-
is usually employed in hot process lime-soda Mg(OH)2-is produced. This precipitate is
softening, or in cold process softening where removed in softening process by sedimenta-
the quantity of water softened is not exception and filtration. However, a soluble by-
tionally large. In cold process systems where product of calcium sulfate, calcium chloride,
70 BETZ HANDBOOK
etc., is formed as illustrated below: tation, a reduction in solids content will re-
(4) sult. If, on the other hand, the hardness
MgS04 + Ca(OHh content of the water is chi~fly in the non-
magnesium .sulfate + calcium hydroxide = carbonate form a marked reduction in the
Mg(OHh + CaS04 solids content of this supply through lime-
magnesium hydroxide + calcium sulfate soda softening will not occur.
(51 It will be noted that two different forms
MgC12 + Ca(OHh = of precipitates are produced in the lime-soda
magnesium chloride+ calcium hydroxide = process. These are calcium carbonate-CaC0 3
Mg(OHh + CaCh -and magnesium hydroxide-Mg(OH) 2 • In
magnesium hydroxide + calcium chloride contrast with the crystalline nature of cal-
Such soluble by-products of calcium sul- cium carbonate, magnesium hydroxide is
fate, calcium chloride, etc., must be removed gelatinous in character and may serve as a
in the softening process, since these salts, to- coagulating agent aiding the settling of these
.. gether with the calcium sulfate and the cal- precipitates in the sedimentation tank. Mag-
cium chloride naturally present in a water, nesium oxide is frequently employed as a
constitute the calcium non-carbonate hard- coagulant in both hot and cold process lime-
ness. soda softening. In addition, the magnesium
hydroxide precipitate, formed through the
Calcium sulfate, calcium chloride, etc.,
use of this coagulant or formed from the
whether produced as by-products in the
precipitation of magnesium naturally present
above reactions or naturally present in a
in the raw supply, possesses the property of
water, are chemically precipitated as calcium
adsorbing soluble silica from solution, thereby
carbonate-CaC0 3-by the use of soda ash
effecting its removal.
(sodium carbonate) as follows:
(6) UsE oF LIME. For certain applications, com-
CaS04 + Na 2C03 plete treatment with lime and soda ash is
calcium sulfate + sodium carbonate = not required since it may be advisable to re-
CaC03 + Na2S04 duce only the calcium or the alkalinity of
calcium carbonate + sodium sulfate the raw water. Under such circumstances,
(7) soda ash may not be needed and the use of
CaCh + Na 2C03 = lime alone may suffice. In the hot lime-hot
+ sodium carbonate =
calcium chloride ion exchange process lirne only is used to re-
CaC03 + 2NaCI duce alkalinity and calcium content prior to
calcium carbonate + sodium chloride the ion exchangers.
In the above reactions while the calcium Other applications of lime softening are in
content is precipitated as insoluble calcium city water plants where the objective is the
carbonate, soluble by-products in the form of reduction of hardness at the most economical
sodium sulfate and sodium chloride are pro- cost, in the beverage industry where high
duced. While the use of soda ash in lime-soda alkalinity is a prime objection and in the
softening effects removal of the hardness, a treatment of cooling tower makeup water
decrease in solids content is not obtained be- where calcium bicarbonate hardness may be
cause of the production of soluble sodium the chief scale forming factor. In these ap-
salts. The previously given reactions with plications, cold process lime softening is em-
lime, however, do result in a reduction of the ployed and the use of a coagulant with the
solids content. In the application of lime- lime is usually recommended.
soda softening to a given water supply, there- The degree to which hardness reduction
fore, it is evident that if the hardness content can be secured with lime alone depends on
of that water is chiefly in the carbonate the bicarbonate and calcium content of the
form, which requires lime for its precipi- water. Non-carbonate calcium hardness re-
quires soda ash for its removal and cannot even under these conditions, but the objec-
be removed though the use of lime. Non- tives of the treatment have been secured.
carbonate magnesium hardness, as shown by Calcium has been reduced to a relatively low
equations (4) and (5) can be reduced by value with a minimum of chemical treatment
lime, but only at the expense of adding ad- cost.
ditional non-carbonate calcium hardness. The third column illustrates still further
Therefore, in softening with lime alone, use is reduction in calcium, but to secure this re-
made only of the reactions shown in equa- duction it has been necessary to add soda ash
tions ( 2) and ( 3 ) . to further force calcium precipitation. The
Lime softening and selective calcium sof- additional calcium removal has been achieved
tening are not necessarily synonymous. Selec- at the expense of slightly higher softener ef-
tive calcium softening, as the name implies, fluent alkalinity.
is used where it is desired only to reduce the Complete lime-soda softening, including
calcium content without using additional reduction in magnesium, is shown by the
chemical for removal of magnesium. Lime fourth column. To secure this reduction in
softening will reduce the carbonate calcium magnesium it has been necessary to again in-
hardness, but soda ash is required for re- crease the softener effluent alkalinity and to
moval of non-carbonate calcium hardness. employ a considerable increase in lime and
Therefore, if it is desired to precipitate cal- soda ash.
cium as completely as possible, it is sometimes Each of the chemical balances illus.trated
necessary to also employ soda ash in selective is the most desirable for certain specific uses
calcium softening. of the treated water. Other variations are
Fig. 11-1 also possible. For example it may be desired
Cold Process Softening Balances to secure maximum alkalinity reduction, or
to secure maximum magnesium removal even
Selective
Lime Calcium at the expense of additional alkalinity in the
Softening Softening Lime-
for With Lime Soda Ash effluent.
Raw Calcium and So.ften-
Water Removal Soda Ash ing UsE oF CAUSTIC SoDA (somuM HYDROXIDE).
Under some circumstances, caustic soda (so-
Total Hardness as CaC03 . 160 80 69 52 dium hydroxide) can be substituted in place
Calcium as CaC03 ...... 110 35 24 24
Magnesium as CaC03 ... 50 45 45 28 of lime and soda ash if the carbonate and
"P" Alkalinity as CaC03 . 0 17 22 60 non-carbonate hardness of the raw water are
"M" Alkalinity as CaC03 . 130 50 60 82 in the correct proportions. The reaction of
Sulfate as S04 ......... 48 48 48 48 caustic soda with calcium bicarbonate hard-
Chloride as Cl .......... 16 16 16 16
ness is illustrated below:
Chemical Treatment Requirements (8)
Ca(HC0 3)z + 2NaOH ~
lime . . . . . . . . . . . . . . . . . . 98 98 156 calcium bicarbonate+ sodium hydroxide =
22 64
Soda Ash ............. .
CaC03 + Na2C03 +
calcium carbonate + sodium carbonate +
All values in parts per million.
2H20
Figure 11-1 illustrates the chemical bal- water
ances obtained in softening the same water Soda ash (sodium carbonate) is produced
with different objectives. The second column as a by-product in this reaction which be-
shows anticipated softener effluent balances comes available for the precipitation of cal-
where only lime is used in the softening proc- cium non-carbonate hardness as shown by
ess and where it is desired to remove only the equations (6) and (7). The use of caustic
calcium and not the magnesium. A small soda for softening is rather limited because
amount of magnesium will be precipitated of the necessity for a definite relationship
72 BETZ HANDBOOK
between the carbonate and non-carbonate PROPAGATING THE CHEMICAL

hardness in the raw water and the lack of PRECIPITATING REACTIONS
flexibility in caustic soda treatment if the
raw water is subject to material fluctuations. The chemical precipitating reactions are con-
sidered to be instantaneous. This fact is not
UsE OF GYPSUM (CALCIUM SULFATE). Highly
evident in the practical operation of softeners
alkaline waters containing appreciable quan-
because all of the insoluble reaction products
tities of sodium bicarbonate-NaHC0 3-are
(sludge) are not immediately precipitated in
characteristic of certain sections of the coun-
a form permitting rapid and efficient sedi-
try. This type of water possesses a total alka-
mentation or separation from the water. A
linity considerably greater than the total
portion of the sludge formed is comprised of
hardness. To condition this type of water
relatively large floes or crystals, readily sepa-
(particularly for boiler makeup) lime and
rated by sedimentation. The remainder of
gypsum are employed. These waters do not
the sludge will be of smaller particle size and
require the use of soda ash. Lime reacts to
will not readily settle. Some of these particles
precipitate the carbonate hardness as in
are present as colloids and are difficult to
equations ( 2) and ( 3) . In the presence of
separate from the water. It is necessary to
sodium bicarbonate, lime reacts to precipi-
coagulate these smaller particles into agglom-
tate calcium carbonate and to form sodium
erates capable of producing a sludge of the
carbonate as shown:
required characteristics.
(9)
2NaHC0 3 + Ca(OHh If a sufficient amount of magnesium is
sodium bicarbonate + calcium hydroxide = naturally present in the raw water, in most
CaC03 + Na2C03 + cases satisfactory coagulation results because
calcium carbonate + sodium carbonate + of the coagulating properties of the magne-
2H20 sium hydroxide precipitated in the softening
water process. In the absence of sufficient natural
For the removal of the sodium carbonate coagulation, recirculation of softener sludge
thus formed, gypsum (calcium sulfate) is em- or the addition of coagulants such as mag-
ployed to precipitate calcium carbonate as nesium oxide, sodium aluminate, or ferric sul-
follows: fate should be employed.
(10) Temperature is an important factor di-
NazC03 + CaS04 rectly affecting the rate of sedimentation
sodium carbonate + calcium sulfate =
which is in turn governed by the density of
CaC03 + Na S042
the water and the time required to effect
calcium carbonate + sodium sulfate
optimum particle size. At higher tempera-
It will be noted that the use of gypsum for tures, the density of the water is less and
the removal of sodium bicarbonate does not more rapid settling of the sludge is permitted
decrease the solids content of the water but in the absence of convection currents. Agglom-
instead replaces the sodium bicarbonate with eration of smaller particles into larger size
an equivalent amount of sodium sulfate. In particles capable of rapid settling is acceler-
most cases this sodium sulfate is less objec- ated by heat. For example, conventional cold
tionable than the sodium bicarbonate. process lime-soda softening usually requires
In a few rare occasions barium salts have a sedimentation time of 4 hours. At the tem-
been employed in order to precipitate and perature of hot process operation ( approx.
remove sodium sulfate in the form of in- 212 F), a one hour sedimentation period is
soluble barium sulfate. Barium salts are rela- sufficient. For consistent results, however>a
tively expensive and poisonous. Because of constant temperature must be maintained.
these limitations, this process is rarely em- Recirculation of sludge provides a contact
ployed. medium for proper precipitation. The use of
lime and soda ash in the cold results in the vantages are the long sedimentation period
formation of a super-saturated solution, or, required, the considerable manual operations
in other words, the water contains more cal- necessary and the fact that the effluent is
cium carbonate in solution than would nor- super-saturated with calcium carbonate and
mally be dissolved in that water at a given is unstable, particularly with increase in tem-
temperature. If, however, particles of cal- perature.
cium carbonate sludge are brought into con- CoNTINUous SEDIMENTATION TYPE LIME
c
tact with water super-saturated with calcium AND SoDA SoFTENERS. Continuous sedimen-
carbonate, crystallization of calcium carbon- tation type units may be of basin design simi-
ate from solution will occur on these par- lar to coagulation and sedimentation basins.
ticles. Units of this type are frequently employed
EQUIPMENT EMPLOYED for city water softening. The raw water is
thoroughly and continuously mixed with the
Lime-soda softeners may be generally classi- proportioned lime, soda ash and coagulant in
fied as follows: a flash mixer consisting of essentially a pro-
Cold Process Softening. peller type agitator confined in a relatively
a. Intermittent or Batch. small chamber. Effluent from the mixer flows
b. Continuous-Sedimentation Type. through a flocculator basin, employed to in-
c. Continuous-Sludge Contact Type. sure formation of proper par~icle size by the
Continuous Hot Process Softening. action of slowly rotating paddles. The water
INTERMITTENT OR BATCH CoLD PRocEss then flows to a basin designed to permit ef-
SoFTENERS. Intermittent softeners are oper- fective sedimentation of the sludge for ap-
ated by the fill and draw method. The sys- proximately a four hour period. Various types
tem consists of two or more vertical tanks. of sludge collection systems are employed.
One tank is employed to supply softened Advantages of this system are its ready
water while the other tank is being cleaned, adaptation to any cold process softening
filled, treated and settled. In some cases only problem, the minimum of maintenance and
one tank is provided if sufficient storage ca- supervision required and the fact that the
pacity is available to permit repeated soften- system may easily be converted to cope with
ing cycles without interruption of a soft water taste and odor control by pre-chlorination or
supply. super-chlorination. Disadvantages are the
The system is operated on a definite cycle. large space requirements, relatively high con-
The calculated quantities of lime, soda ash struction cost and the fact that the effluent
and coagulant are made into a slurry in a must be recarbonated to achieve stability.
chemical mixing tank and added to the tank Chemical requirements are comparable to
while it is being filled with raw water. The other cold process methods discussed.
full tank is agitated for a period varying from CoNTINuous SLUDGE CoNTACT TYPE. Im-
15 minutes to one hour, agitation stopped proved results in cold process lime-soda sof-
and the water permitted to settle for as long tening have been made possible by the
a time as available. The clarified water is development of sludge contact units. Basi-
drawn off by means of a swingpipe floating at cally, these units operate on the principle of
the surface and discharged to filters usually crystallization from a super-saturated solution
of the gravity type. by providing a contact medium for proper
The intermittent or batch softener is nor- precipitation of the sludge. The raw water
mally employed for small plants. It comprises is intimately mixed with previously precipi-
the original softener design and is rarely tated sludge and with the lime, soda ash and
recommended for new construction. Advan- coagulant. Initial precipitation of calcium
tages of this system are its low initial cost and and magnesium occurs and joins the slurry
comparatively simple operation. Its disad- pool of previously formed precipitates of cal-
74 BETZ HANDBOOK
cium and magnesium by contact. In this specified effects appreciable savings in equip-
manner, precipitation and equilibrium are ment cost, construction and floor space when
quickly established. The newly formed solids compared with other cold process methods.
do not separate individually as small parti- Chemical requirements are low and com-
cles, instead the fresh precipitate deposits on parable to the requirements of hot process
the slurry already present by accretion. The softeners. The hardness of the finished water
nature of this sludge permits ready separation is lower than can be obtained with other cold
of the treated water from the slurry. A defi- process units employing a given amount of
nite line of separation exists between the chemical.
slurry bed and the clarified water with the Sedimentation chambers are eliminated in-
treated effluent virtually filtered through a asmuch as all sludge particles are kept in
bed of its own sludge. The entire volume of suspension due to velocity of upward flow.
the tank is slowly but constantly circulated A gradual decrease in velocity permits a
by an agitator. definite point of separation of sludge and
One design of sludge contact unit is shown clarified water. The sludge bed is maintained
by Figure 11-2. Softeners of the sludge con- at definite density by a constant sludge bleed
tact type are rated in capacity much the same and therefore expensive sludge collection
as filters. Units are available to produce from systems are eliminated. Another design of this
1.0 to 3.0 gallons per square foot of surface type is shown in Figure 11-3.
area per minute, dependent on the design. CoNTINuous HoT PRoCEss LIME-SooA SoF-
Turbidities of the finished water of 5.0-10.0 TENING. The hot process softener is normally
units (without filtration) are not uncommon. employed as a specific external treatment for
This is much lower than obtained by most boiler feedwater. The units are designed to
other methods of cold process softening. The operate at temperatures of 212 F or above
maximum retention time of 1 hour usually and employ live or exhaust steam as the
INFLUENT EFFLUENT
I COLLECTING
TROUGH
. SLUDGE FILTER
zONE
SWINGING BAF'f'L(S
Courtesy -Pfaudler Permutit Inc.
Figure 11-2 • Sludge Contact Softener

SLOW MIXING FLOC FORMATION
TREATED WA
EFFLUENT
CLARIFIED
WATER
CLEAR WATER
SEPARATION
RAW WATER
INFLUENT
SLUDGE RECIRCULATION SEDIMENTATION
SLUDGE REMOVAL
Courtesy Graver Water Conditioning Co.
Figure 11-3 • Sludge Contact Softener
source of heat. The design of one type of hot blown down to remove the precipitates and
process unit is shown in Figure 11-4. Other the clarified and softened water is drawn off
types of different design are procurable, but through the internal take-off and subjected
the softening principles are similar in all. to pressure filtration. Since silica is quite
While varying slightly, dependent upon the soluble in hot alkaline solutions, anthracite
particular. design of different manufacturers, or some other non-siliceous material is nor-
in a hot process softener the cold raw water mally specified as a filter medium. Filters are
enters through a float controlled regulating backwashed with hot softened water from
valve to maintain a given water level in the the sedimentation chamber and backwash
sedimentation tank. The flow of raw water water is returned to this chamber. No ap-
to the softener actuates a proportioning de- preciable thermal loss is entailed in back-
vice to continuously feed lime, soda ash and washing, therefore, and both chemicals and
coagulant. The raw water is heated to or · treated water are conserved.
above 21'2 F in a heater which may be of a Hot process softening possesses the advan-
tray, jet or deaerating design. Chemical re- tage of treating large quantities of water in
actions to form calcium carbonate and mag- a relatively small compact unit. A retention
nesium hydroxide are practically instanta- time of 1 hour is normally specified. Maxi-
neous and due to the effect of the higher mum efficiency of chemical treatment is ob-
temperature sedimentation takes place rap- tained by virtue of the high operating tem-
idly. The sludge collecting cone is periodically perature. Economies in chemical treatment
76 BETZ HANDBOOK
DIRECT CONTACT
GAS CONCENTRATOR
VACUUM BREAKER
RECORDING
THERMOMETER ELEMENT
RAW WATER
INLET
REGUlATING
VAlVE
FlOAT BOX
.. WASH
RETURN
AUTOMATIC
WASH WATER DESlUDGING
OUTlET VAlVE
Courtesy Cochrane Division-Crane Company
Figure 11-4 • Downflow Type Hot Process Softener Figure 11-5 • Up.fiow Sludge Blanket
Type Hot Process Softener
can frequently be effected by recirculation of secured in the sludge blanket aids the effi-
softener sludge. ciency of silica removal. Softening and tur-
In recent years, the sludge contact princi- bidity removal are also favored by this design.
ple, originally developed in cold process work, The height of the sludge blanket can be de-
has been applied to hot process softening. termined by use of sampling cocks at differ-
Sludge blanket hot process softeners are parent positions on the sedimentation tank. A
ticularly applicable where silica reduction is disadvantage of the sludge blanket design is
required. Sludge blanket units differ from the difficulty in maintaining a uniform sludge
conventional designs in that, as shown by blanket, without carryover, under varying
Figure 11-5, the raw water, mixed with the load conditions.
added chemicals, is delivered to the bottom The hot process lime-soda softener can be
of the tank by means of a central downtake. used to combine in one complete unit the
Flow is reversed at this point and the water functions of:
rises slowly up through a blanket of previ- Softening by removal of calcium and mag-
ously formed sludge. The intimate contact nesium salts.
Silica removal through the use of magne- LIMITATIONS

sium salts.
Complete de aeration of the boiler feed- The lime-soda softening process does not re-
water. duce the hardness of a softened water to as
Filtration for the removal of suspended low a value as does the ion exchange soften-
matter. ing ~rocess. Hardness can be reduced to ap-
proximately 10 to 30 pprn as calcium car-
The effluent from the hot process softener
bonate dependent upon the temperature of
is low in residual hardness and is relatively
operation and the alkalinity maintained on
stable. Marked increases in temperatures such
t?e softener efflmmt. Consequently, \vhere
as in economizers, closed heaters, etc., will
hme-soda softened makeup water is em-
cause some after-precipitation. Under these
ployed for boiler feedwater, internal chemi-
conditions, stabilization of the softener effiu-
~al. treatment requirements are higher than
~nt by acidification or surface active agents
1f 10n exchange softened makeup \Vater is
lS necessary to prevent objectionable deposits.
employed. Because hardness can be reduced
to no lmvcr than 10 to 30 ppm this process
APPLICATIONS of softening offers little value where raw
water hardness is less than 50 ppm.
Lime-soda softening is the type of treatment
most suitable to the softening of high hard- Particularly in cold process softening, the
ness supplies, particularly those of predomi- use of special coagulants may be required to
nately carbonate hardness. Marked reduction obtain the proper softening and hardness re-
in solids content results where the raw water moval and to avoid formation of complex
contains high carbonate hardness. soluble forms of hardness.
Iron or suspended matter in the raw water On waters of high non-carbonate hardness
can be removed by lime-soda softening and there will be no decrease in solids content ef-
pre-filtration of such supplies is not required. fected by the lime and soda process and on
\Vhere the raw water contains appreciable raw waters containing only a small amount
color and dissolved organic matter, through of natural alkalinity, an increase in solids
the .use of proper coagulant it is frequently may result.
possible to effect a marked reduction in the . C?ntrol of lime-soda softening, in general,
organic content. IS slightly more difficult than the control of
Lime softening alone, without the use of ion exchange softening and more frequent
soda ash, can be employed advantageously testing of the softener effluent may be re-
quired with more frequent adjustment of
~here it is desired to reduce only the calcium
bicarbonate content. In the cold, typical ap- control.
plications are in the treatment of makeup REFERENCES
water for cooling systems, preparation of
V. J. Calise, "Factors in Hot-Process Softener De-
?rocess wate~ for the beverage industry and sign and Operation", Combustion, VoL 22, pp. 49-
m the softenmg of city supplies. In the hot, 55 (Aug. 1950)
lime softening provides an efficient method A. A. Kalinzke and J. M. Kahn "Applying Steam
Lift Circulation to Hot-Process S~ftcncrs" Proceed-
for reducing alkalinity and solids prior to ings, Midwest Power Con£., Vol. XII pp. 71-78
hot ion exchange units. (1950) '
Hot process lime-soda softening provides E. Nordell, "Water Treatment for Industrial and
Other Uses", 2nd Ed., pp. 489-54 7, Reinhold Pub-
the ideal conditions necessary for the most lishing Co., New York, N.Y. ( 1961)
economical and efficient removal of silica S. T. Powell, "Water ~onditioning for Industry",
from boiler· feedwater. In cases where raw pp. 84-145, McGraw-H1ll Book Co. Inc. New York
N.Y. (1954) ' ' '
water hardness is relatively low, magnesium
J. D. Yoder, "The Sludge Blanket Hot Process
oxide and lime may be used primarily for the Softener", Engineers Soc. of Western
value in silica removal rather than softening. Penna., pp. 65-80 (
78 BETZ HANDBOOK
12
Hot Process
Phosphate Softening
Chemical Reactions. . . . . . . . . . . . . . . . . . . . . 79
Equipment Requirements . ................... 80
Applications and Limitations . . . . . . . . . . . . . . . . 80
OT phosphate softening is a process acid itself may be used depending upon the
H whereby the calcium and magnesium
salts constituting the hardness of water are
characteristics of the water to be treated. For
control of chemical treatment, an excess of
chemically precipitated and removed through 5-10 ppm phosphate is maintained in the
the use of phosphate, normally disodium softener effluent. A pH of approximately 9. 7
phosphate, in conjunction with caustic soda. is required for precipitation of calcium, but
The calcium hardness is precipitated as tri- removal of magnesium and silica requires a
calcium phosphate while the magnesium slightly higher pH approximating 10.1. The
hardness is precipitated as magnesium hy- softening reactions may be represented as
droxide. The softening operation is carried follows:
out in the hot at temperatures of 212 F or (1)
above. Since both the calcium and magne- 3Ca(HC03h + 6NaOH 3CaC03 +
sium precipitates formed are very insoluble, a calcium sodium calcium
water of practically "zero" hardness is pro- bicarbonate+ hydroxide carbonate +
duced by this softening process. JNa2C03 + 6H20
Hot process phosphate softening is em- sodium + t
carbonate wa er
ployed as a specific form of external treat-
ment for boiler feedwater-particularly for (2)
high pressure boilers. 3CaC03 + 2Na3P04 = Ca3(P04h +
calcium + trisodium tricalcium +
This process is applicable to very soft carbonate phosphate = phosphate
waters of the order of hardness not exceed-
JNa2C03
ing 60 ppm. It is part'icularly well adapted sodium
to turbid waters low in hardness where coag- carbonate
ulation and softening can be brought about It can be noted from equation ( 1) that
in a single apparatus. calcium bicarbonate is first converted to cal-
Where waters are high in hardness this cium carbonate which is fairly insoluble as
system may also find application by employ- discussed under lime-soda softening. In the
ing two stage softening. Thus, a greater por- second equation the calcium carbonate is
tion of the hardness may be removed econom- converted to the more insoluble tricalcium
ically by lime-soda softening as the first phosphate. The magnesium hardness is pre-
stage and the remaining hardness may be re- cipitated as the hydroxide in accordance with
moved by subsequent phosphate softening in the following equation:
the second stage. (3)
Soft waters high in silica, likewise are Mg(HC03h + 4NaOH = Mg(OHh +
suited to this process. Most efficient silica re- magnesium + sodium = magnesium +
duction is accomplished in the hot phosphate bicarbonate hydroxide hydroxide
softener by magnesium sulfate. It is therefore 2Na2C03 + lH20
possible to decrease silica to a low value and sodium + water
at the same time accomplish softening to a carbonate
"zero" hardness. The tricalcium phosphate and magnesium
Where a reduction in alkalinity is desired, hydroxide precipitates which form are re-
particularly to insure a low carbon dioxide moved in the softening process by means of
content in the steam, phosphoric acid and/or sedimentation and filtration.
sulfuric acid can be used prior to softening. From the above reactions it can be noted
that a soluble by-product of sodium carbon-
CHEMICAL REACTIONS ate alkalinity results, increasing the natural
alkalinity from one equivalent on the un-
Any of the various phosphates, from the al- treated water to two equivalents on the
kaline trisodium phosphate to phosphoric treated water. Where the natural alkalinity
80 BETZ HANDBOOK
is already high this condition is undesirable although a retention period of only 45 min-
and may be offset by the use of an acid utes or less has been used with phosphate
phosphate or by use of sulfuric acid. Alka- softening compared with the 1 hour required
linity reduction by means of phosphoric acid for hot process lime-soda softening.
is illustrated as follows:
(4)
Ca(HC03h + H3P04 CaHP04
calcium phosphoric_ calcium
bicarbonate+ acid -acid phosphate
+ 2C02 + 2H20
carbon
+ d'IOXI'd.e + water
The carbon dioxide thus formed may be
removed by deaeration, thereby minimizing
the carbon dioxide content of the steam.
Following deaeration, the soluble calcium
acid phosphate is precipitated by the addi-
tion of sodium hydroxide. This reaction is Figure 12-1 • Single Tank Two Stage Softener-Hot
illustrated as follows: Lime-Soda Followed by Hot Phosphate
(5)
While a calcite filter medium can be used
3CaHP04 + 3NaOH = Ca3(P04h in conjunction with lime-soda softening, re-
calcium sodium tricalcium
acid phosphate+ hydroxide= phosphate action of calcite with a phosphate softener
effluent will result in cementing the entire
+ Na3P04 + JH20 filter bed. Anthracite coal is the most desir-
+ trisodium + water
phosphate able filter medium with phosphate softening.
All of the preceding reactions illustrate The two stage water softeners may take
the removal of carbonate hardness. The re- one of several forms as, for example, the in-
moval of non-carbonate hardness proceeds in corporation of two stages in a single sedimen-
accordance with the following reactions: tation tank as illustrated by Figure 12-1.
(6)
APPLICATIONS AND LIMITATIONS
3CaS04 + 2Na3P04 = Ca 3(P0 4h +
calcium+ trisodium _ tricalcium + Hot process phosphate softening is primarily
sulfate phosphate- phosphate
useful in the treatment, for boiler feedwater
JNa2S04 purposes, of turbid waters that are low in
sodium hardness and alkalinity. It is possible with
sulfate
hot phosphate softening to combine turbidity
(7)
removal, softening and deaeration in a single
MgCh + 2NaOH = Mg(OHh unit. Where silica removal is also necessary,
magnesium sodium _magnesium+
chloride + hydroxide - hydroxide hot phosphate softening provides a single
2NaCI unit to accomplish all of these objectives.
sodium Through the use of phosphoric or sulfuric
chloride acid, reduction can be secured in potential
carbon dioxide content of the steam.
EQUIPMENT REQUIREMENTS The effluent from a hot process phosphate
softener, while quite low in hardness, will
In general the equipment employed for hot cause after-precipitation when subjected to
process phosphate softening is very similar to marked increase in temperature, such as en-
that used in hot process lime-soda softening, countered in economizers, closed heaters, etc.
Stabilization of the softener effluent by acid Until the development of the styrene base
or the application of organic surface active ion exchange resins, the combination of hot
agents is necessary to prevent objectionable lime-soda softening, followed by hot phos-
deposits. phate softening, was considered · the best
The fine particles of calcium phosphate method for producing a low hardness boiler
formed in the sedimentation tank do not feedwater from a hot process treatment
settle as rapidly as the precipitates formed in plant. However, as discussed in the chapter
lime-soda softening. Greater difficulty is en- on hot lime-hot ion exchange softening, this
countered in maintaining a low turbidity new process offers several advantages over
softener effluent with hot phosphate soften- the hot lime-soda hot phosphate process.
ing. The presence in the raw water of or- With the hot ion exchange method, more
ganic matter, which would tend to retard the complete hardness removal can be secured
precipitation reactions, may result in rela- than with hot lime-soda hot phosphate sof-
tively poor hardness removal and plugged tening, together with a lower carbon dioxide
filters. content of the steam.
The following is a list of the common commercial forms of phosphate that may be used
for hot phosphate softening. In selecting the type of phosphate to be employed there are two
important factors to consider: (a) cost per pound P 2 0 5 and (b) alkalinity of the water to
be treated. For example, although anhydrous monosodium phosphate has a greater P 2 0s
content than anhydrous disodium phosphate, the cost is higher and it is generally more eco-
nomical to purchase the latter phosphate. However, if the alkalinity of the water to be treated
is very high, it may be desirable to use anhydrous monosodium phosphate, which is acidic, in
spite of its greater cost so as to reduce the alkalinity of the treated water. The general state-
ment can be made that for the majority of cases anhydrous disodium phosphate is the most
satisfactory phosphate to be employed. However, each specific application should be studied
to determine this point.
Fig. 12-2
Supplementary Data-Hot Phosphate Softener
Pounds
Lbs CaC03 NaOH Required
Removed to Convert 100
Per 100 tbs lbs of the Phos-
Name of Phosphate Chemical Formula %P20s of the Phosphate phate to Na3 P04*
Monosodium Phosphate, Anhydrous ........•.. NaH 2 P0 4 58% 122 lbs. 671bs.

Monosodium Phosphate, Monohydrate ......... NaH 2P0 4 .H 20 51% 108 lbs. 58 lbs.
Disodium Phosphate, Anhydrous .............. Na 2HP0 4 48•/. 101 lbs. 28 lbs.
Disodium Phosphate, Crystalline .............. Na 2 HP0 4.12H 2 0 19% 40 lbs. 11 lbs.
Trisodium Phosphate, Monohydrate ........... Na 3 P04.H 20 39% 82 lbs. 0 lbs.
Trisodium Phosphate, Crystalline .............. Na 3P04 .12H 20 19% 40 lbs. 0 lbs.
Phosphoric Acid, 75% Solution ..........•••. H3 P04 54% 114 lbs. 921bs.
* This value is the theoretical NaOH requirement for each phosphate used. However, the natural alkalinity or acid-
ity of the water to be treated will vary the total NaOH requirement.
REFERENCES
E. W. Ellis and H. E. Carlson, "Lime-Soda- gineers Soc. of Western Penna., pp. 79-91 (1941)
Phosphate Feedwater Treatment", The Petroleum F. N. Kemmer, "Conditioning Makeup by High-
Engineer, Vol. 26, pp. C29-32 (June 1954) Temp., Sodium Ion-Exchange Excess Calcium, Hot
C. E. Joos, "Ramifications of the Hot Process Lime-Zeolite Process", Combustion, Vol. 21, pp.
Method of Water Conditioning", Proceedings, En- 59-62 (April 1950)
82 BETZ HANDBOOK
13
Silica Ren1oval By
Magnesiun1 Salts
Factors Controlling Silica Removal . ........... 83
Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Retention Time . . . . . . . . . . . . . . . . . . . . . . . . . 83
pH .................................. 83
Sludge Recirculation . . . . . . . . . . . . . . . . . . . . . 83
Various Magnesium Compounds . ........... 84
Mechanism of Reaction. . . . . . . . . . . . . . . . . . . 85
Equipment Required. . . . . . . . . . . . . . . . . . . . . . . 86
Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
Limitations.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
HE removal of soluble silica from water The removal of silica from boiler feed
T in the hot is accomplished with the use
of magnesium compounds under relatively
water is accomplished in order to prevent or
to minimize silicate boiler scales such as cal-
simple controlled conditions of temperature, cium and magnesium silicate and also to pre-
retention time, pH and sludge recirculation. vent the formation of the complex sodium-
Hot process silica removal is carried out alumino-silicate scales such as analcite,
employing a conventional hot process lime or Na 2 0.Al 2 0 3 .4Si0 2 .2H 2 0. For the prevention
lime-soda softener. It is advantageous to thus of siliceous turbine blade deposits, it may be
combine in one unit the functions of soften- advisable to limit the silica content of the
ing and silica removal. boiler water thus involving the removal of
The magnesium compound employed may silica from the boiler feedwater.
be Epsom Salt (magnesium sulfate), dolo-
mitic lime, calcined magnesite, magnesium FACTORS CONTROLLING SILICA REMOVAL
carbonate or magnesium oxide. Since mag-
Effect of Temperature-The higher the tem-
nesium oxide possesses the greatest silica re-
perature of operation, the more efficient is
moval efficiency and does not increase solids
the removal of silica from the solution. Fig-
content, it is usually employed.
ure 13-1 illustrates the increase in silica re-
moval efficiency using magnesium oxide with
100 increase in temperature up to 95 C, the ap-
proximate temperature of operation of hot
process softeners.
90
Effect of Retention Time-At 95 C almost
complete silica removal is accomplished by
80 magnesium oxide in 15 minutes retention
time. At lower temperatures there is a grad-
0
UJ
> ual increase in the quantity of silica removed
~ 10 with increased retention. Figure 13-2 illus-
UJ
a: trates this effect of retention. Since hot proc-
<(
~ 60
ess lime-soda softeners are designed for one
iii hour retention, it is evident that complete
..J
<( silica removal can readily be accomplished
~ 50 in this period of time.
a:0 Effect of pH-The influence of pH on sil-
lL
0
40
ica removal by magnesium oxide is illustrated
1-
by Figure 13-3. It can be noted that the op-
~
u timum pH, in each of the three cases shown,
a:
UJ
D..
30 is approximately 10.1. This pH value is
readily obtained with either hot or cold proc-
ess lime-soda softening. The greater the per-
20
centage of silica removed the less important
is the control of pH.
15 MINUTES RETENTION
10
ON ALL TESTS Effect of Sludge Recirculation-During the
silica removal process, if the partially spent
0
magnesium oxide sludge is brought back in
oL-~--2~0~~4~0--~-6~0--~-8~0~~1~00~--
contact with fresh incoming water high in
TEMPERATURE , C
silica, a considerable increase in silica re-
moval efficiency can be achieved. A portion
Figure 13-1 • Effect of Temperature of the sludge collecting in the sludge cone of
84 BETZ HANDBOOK
100
0
w
> 90
0
~
w 80
c:r:
<(
(.) 70
.J
CJ) 60
.J
<(
z
(!)
c:r:
0
1.1,;·
0 8 TEMP= 95C
1-
z e TEMP= 23C
w
(.)
0::
w
a..
180 240
TIME OF RETENTION, MINUTES

Figure 13-2 • Effect of Retention Time
the softener is recirculated by means of a ash needed to precipitate the magnesium or

small pump to the upper section of the sedi- calcium compounds thus introduced. How-
mentation tank. Here it comes in contact with ever, the major disadvantage of their use is
the raw makeup water and, through addi- the increase in solids content of the boiler
tional silica adsorption, considerably reduces feed water.
the quantity of fresh magnesium oxide that Dolomitic lime can be used to provide
must be employed. It is possible by sludge economical silica removal where the lime
recirculation to reduce magnesium require- requirements for softening will introduce suf-
ments by as much as 60%. The increased ficient magnesium oxide to accomplish the
silica removal efficiency is shown by Figure desired degree of silica removal. Where mag-
13-4. There is no change in the chemical nesium oxide requirements for silica removal
balances maintained on the softener efHuent are relatively high and lime requirements
and recirculation affects the characteristics low, if sufficient dolomitic lime were fed to
of the softener effiuent only with respect to a secure the desired silica removal, it would be
reduced silica content. necessary to feed additional soda ash to re-
Effect of Various Magnesium Compounds act with the excess lime thus introduced. In
-A number of different magnesium com- such cases, dolomitic lime can be used in an
pounds have been investigated for use in the amount sufficient to supply lime requirements
removal of silica from water. A number of and additional magnesium oxide fed to secure
these compounds such as Epsom Salt (mag- the necessary silica removal.
nesium sulfate) and dolomitic lime may con- Magnesium oxide has the advantage of
siderably increase the solids content of the causing no increase in the solids content of
treated water. They may also require large the treated water. In addition, no lime or
increases in the quantity of lime and soda soda ash is required to effect its precipitation.
100 combining weights. The removal of silica

from the solution by magnesium compounds,
however, is not a stoichiometric reaction but
proceeds by adsorption from the solution.
Figure 13-6 illustrates silica removal plotted
BO
against the qv.antity of magnesium oxide
used on a water supply high in silica. These
0 70
data were obtained from actual operating
w results on a hot process lime-soda softening
6
::;;
w system rated at 270,000 pounds per hour and
.0:
":::;
60
conditioning water for 450 psi boilers. It can
iii
be noted that increased quantities of magne-
..
..J
z
50 sium oxide do not effect proportionate in-
creases in the quantity of silica removed. In
'"
;;:
0
u. 40
other words, as larger amounts of silica are
0 removed, larger and larger quantities of mag-
~ nesium oxide are required to effect removal
of the last traces of silica. This action is
~ 30
characteristic of adsorption reactions.

20
The empirical equation of Freundlich
closely fits adsorption data. This equation
may be expressed as:
10
X
KCII"
m
9.0 9.4 9.8 10.2 10.6 ll.O
pH
0.5
Figure 13-3 • Effect of pH
/
While different forms of magnesium oxide
vary in their silica removal efficiencies, 0
specially prepared adsorptive forms of mag- ~ 0.4
nesium oxide, have been found to possess the 1-
a::
greatest silica removal efficiency and to be ~
the most economically feasible reagent. a::
11..1
0..
Figure 13-5 illustrates the silica removal 0 0.3
11..1
capacity of a number of different magnesium >
0
compounds under the same conditions. Be- ::;;
11..1
cause silica removal by magnesium com- a::
pounds is an adsorption reaction, those ma- ~ 0.2
terials which effected over 80% silica :::i
removal in this test possess very high silica en
removal capacities.
0.1
Ell WITHOUT RECIRCULATION
Mechanism of Silica Removal Reactions-
41 WITH RECIRCULATION
In the chemical reactions usually encountered
in water conditioning there is a stoichiometric
relation or constant reacting value. For ex- 6 B 10 12 14 16 IB 20 22 24 26
ample, a definite quantity of soda ash will SILICA REMOVED, PPM
always react with a definite amount of cal-
cium sulfate in accordance with the law of Figure 13-4 • Effect of Sludge Recirculation
86 BETZ HANDBOOK
MgS04·7H20
ORIGINAL SILICA = 29.8 PPM
U.S.P. MAGNESIUM HYDROXIDE 150 PPM EACH REAGENT USED
DOLOMITIC LIME
CALCINED MAGNESITE
MAGNESIUM CARBONATE
TECHNICAL MAGNESIUM OXIDE
U.S.P. LIGHT MAGNESIUM OXIDE
.U.S.P. HEAVY MAGNESIUM OXIDE
SPECIAL ADSORPTIVE MAGNESIUM OXIDE
--'- --'- I
0 10 20 30 40 50 60 70 80 90 100
PER CENT OF ORIGINAL SILICA REMOVED
Figure 13-5 • Comparison of Various Forms of Magnesium Oxide
where K and 1/n are constants. feeding equipment is required. Magnesium

x = quantity of silica adsorbed oxide can simply be added to the same mix-
m = quantity of adsorbent ing tank where the lime and soda ash are fed.
C = residual concentration of adsorbed
material at equilibrium.
54
A straight line results from plotting the
logarithm of ~ against the logarithm of C

rn '
when the reaction proceeds by adsorption.
Silica removal data -thus shows the reaction
to be one of adsorption similar to the re-
N
moval of color by activated carbon. 0
<i)
EQUIPMENT REQUIRED
24
For the most efficient removal of silica by

magnesium compounds, a combination of the 18
following factors is required:

Temperature close to 212 F. 12
Retention time, 15 minutes to one hour.

pH value approximately 10.1.
Recirculation of sludge.
0
These conditions are most readily secured o~--------~2S~--------5~0--------~75~
MAGNESIUM OXIDE , PPM
through the use of a conventional hot process
softener which provides the ideal unit for the Figure 13-6 • Field Tests on Silica Removal
silica removal process. No special chemical with Sludge Recirculation
Recirculation of sludge is simply provided by When magnesium oxide is employed in

tapping the sludge cone and circulating a conjunction with hot process lime-soda sof-
portion of the sludge to the top of the sedi- tening, the additional advantages secured in
mentation tank by means of a small centrif- addition to the above are:
ugal pump. No additional capital expenditure is re-
While silica removal can be secured in con- quired since, both softening and silica re-
junction with hot lime or hot lime-soda sof- moval are accomplished in the one opera-
tening, the use of lime and soda ash is not tion, requiring no additional equipment.
necessary for the successful operation of the Quantities of lime and soda ash are not
silica removal process. On a water low in increased by the use of magnesium oxide,
hardness, a conventional hot process softener unlike the use of Epsom Salt, dolomitic
can be employed for silica removal feeding lime or iron salts which require lime and
only caustic soda and magnesium oxide. soda ash for their precipitation.
No increase in the hardness or alkalinity
of the softened water is occasioned by the
APPLICATIONS
use of magnesium oxide.
Hot process silica removal by magnesium
compounds can be successfully applied to any LIMITATIONS
type of water. The use of magnesium com-
pounds is advantageous with the hot lime-hot In order to efficiently accomplish the removal
zeolite process, providing removal of silica of silica by magnesium compounds and to
to as low as 1 ppm in conjunction with hard- comply with the conditions of temperature,
ness removal to the range of 0.5-2.0 ppm. retention time and sludge recirculation, a hot
Magnesium oxide is preferred as the -silica re- process softener complete with chemical pro-
moval agent in hot lime or hot lime-soda portioner, heater, sedimentation tank, filters
softening, while more efficient results are and backwash facilities is required.
secured in hot phosphate softening by the use Where such equipment is already installed,
of magnesium sulfate. silica removal can be very simply applied.
Silica removal is particularly desirable in Where a hot process softener is not already
conditioning water for use in high pressure installed in a plant, its installation is necessary
boilers both from the standpoint of reducing before silica removal by magnesium com-
scale deposits in the boilers and also to re- pounds can be accomplished.
duce siliceous deposits on turbine blading.
When the magnesium compound employed REFERENCES
is magnesium oxide, the additional advan- L. D. Betz, C. A. Noll and J. J. Maguire, "Re-
tages secured are : moval of Silica from Water by Hot Process", Ind.
No increase in the solids content of the and Eng. Chem., Vol. 32, pp. 1323-1329 (1940)
L. D. Betz, C. A. Noll and J. J. Maguire, "Adsorp-
treated water. tion of Soluble Silica from Water", Ind. and Eng.
More silica is removed per part of reagent Chem., Vol .33, pp. 814-820 (1941)
employed than in any other method. C. A. Noll and J. J. Maguire (to Betz Laborato-
Magnesium oxide is itself an excellent coag- ries, Inc.) U.S. Patent 2,307,466 (Jan. 5, 1943)
J. D. Yoder, "Removing Silica from Boiler Feed
ulant and eliminates any need for other Water", Proceedings, Midwest Power Conf., Vol.
coagulants. XI, pp. 164-170 (1949)
88 BETZ HANDBOOK
14
Sodiulll Zeolite Softening
Zeolite Exchange Materials. . . . . . . . . . . . . . . . . 89

Applications. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
term zeolite is applied to insoluble, damental reaction in sodium zeolite softening:

T
HE
solid materials which have the property (1) Ca++ + Na2Z CaZ + 2Na+
of exchanging various ions with which they calcium + sodium = calcium + sodium
come in contact. The sodium zeolite soften- ion zeolite zeolite ion
ing process, also referred to as cation ex- When this reaction has proceeded to the
change on the sodium cycle, exchanges so- point where the zeolite will release no more
dium ions for all cations of two or more sodium in exchange for calcium, it becomes
positive charges, when operating on the sof- necessary to regenerate the zeolite bed. This
tening cycle. When water containing the regeneration is accomplished by washing with
hardness salts of calcium and magnesium is a strong salt solution (sodium chloride). The
passed through the cation exchange bed, the high concentration of the sodium ion in the
sodium ions of the bed replace the hardness salt solution reverses the above reaction re-
ions with a resultant effiuent water of close sulting in regeneration of the zeolite bed.
to "zero" hardness. Sodium zeolite softening This regeneration reaction may be written as
is, by far, the most common of the exchang~ follows:
reactions involved in water conditioning and (2) CaZ + 2NaCI = Na2Z + CaCI2
is generally the one referred to vvhen the calcium sodium sodium calcium
term "zeolite softening" is used without zeolite + chloride = zeolite + chloride
further qualification.
The calcium placed in solution by regen-
The primary purpose of sodium zeolite sof-
eration is run to waste as calcium chloride.
tening is the removal of the scale forming
The fundamental chemical principle of mass
ions of calcium and magnesium, replacing
action determines if the reaction proceeds as
these ions with an equivalent amount of
in equation ( 1) or equation (2). In the sof-
sodium ions. By removal of the calcium and
tening cycle, where the calcium ions pre-
magnesium ions, which normally constitute
dominate, the reaction proceeds as in ( 1) .
the hardness content of a water, this process
In the regeneration reaction, where the so-
is useful in the preparation of water for boiler
dium ions predominate, the reaction proceeds
feedwater, laundry use, various industrial
as in ( 2) . The net result is that sodium chlo-
processes and in the reduction of the hard-
ride (common salt) has been consumed and
ness content of city supplies.
hardness has been removed from solution.
ZEOLITE EXCHANGE MATERIALS As it is necessary to have an excess of salt
present for regeneration, an equivalent
Natural zeolite minerals possess a complex amount of hardness is not removed for a
chemical structure and usually consist of so- given amount of salt consumed.
dium-aluminum-silicate. Other cation ex- The capacity of zeolites, that is the amount
change materials are made synthetically and of hardness that can be removed before re-
have as a base such chemicals as sulfonated generation, will vary greatly. The synthetic
coal (carbonaceous) , phenolic resins and zeolites generally have a higher exchange
polystyrene resins. Regardless of how the capacity than the natural zeolites, but are
zeolites are produced, all operating on the more expensive per unit volume. The capac-
sodium cycle contain a complex molecule ities for synthetic zeolites range between 6000
with sodium attached. Accordingly, it is pos- grains per cubic foot and 32,000 grains per
sible to represent sodium zeolite in a chemical cubic foot. Natural zeolite capacities vary
equation by using the symbol Na 2 Z where Na between 2500 and 5000 grains per cubic foot,
stands for the sodium and Z for the complex depending on the processing. Greensand zeo-
zeolite radical. For simplification in the fol- lite is more resistant to waters of low pH
lowing reactions, the calcium ( Ca ++) in solu- than the synthetic aluminum-silicate zeolites.
tion will be considered as representing all the Greensand, carbonaceous and resinous zeo-
hardness. The following is therefore the fun- lites are less likely to increase the silica con-
90 BETZ HANDBOOK
tent of the softened water and, consequently,

are better applied to the softening of waters
of low silica content. Synthetic zeolites of the
sodium-aluminum-silicate type may impart
considerable silica to waters that are low in
initial silica content.
Generally, synthetic zeolites are employed
in waters of high hardness where the ex-
change value of this material is of importance
in permitting the use of a smaller unit. For
low hardness waters, greensand~the natural
zeolite~is still occasionally employed.
The effluent from a zeolite softener will

rise quite rapidly in hardness as it approaches
the limit of its capacity. In some cases, where
minimum hardness is essential, the softener
will be regenerated before it reaches this
point. In other cases, the hardness is allowed
to rise only to a predetermined level before
regeneration. As the concentration of anions
such as chloride and sulfate increase in the
Courtesy Rohm & Haas Company '"'·ater to be softened, there is the tendency
Figure 14-1 • Microscopic View of Resin for hardness leakage to take place. With the
Beads (20-50 mesh) of a Sulfonated older types of zeolite minerals and synthetics,
Styrene-Divinyl Benzene Cation Exchanger the amount of hardness leakage could not be
controlled since this leakage was a function
of the influent ion concentration. However,
with the advent of the polystyrene type res-
ins, the amount of hardness leakage can be
accurately controlled by adjusting the salt
regeneration level. For example, in a normal
water supply where the total hardness and
sodim11 salts were in the range of 700-800
ppm, using a carbonaceous zeolite the mini-
mum effluent hardness was found to be in the
range of 15 ppm with a salt regeneration rate
of 3.5 pounds of salt per pound of hardness
removed. Usmg a polystyrene resin at the
same salt rate produced an average effluent
hardness of 8 ppm. By doubling the salt rate,
the effluent hardness was reduced 2 ppm.
Figure 14-3 also illustrates the effect on
capacity at different salt regeneration levels.
From this table it can be readily understood
that sodium zeolite softeners, using a poly-
styrene base resin, can be designed for low
Courtesy Rohm & Haas Company capital investment and high operating cost or
Figure 14-2 • Microscopic View of Beads (20-50 mesh) higher capital investment with lower operat-
of a Carboxylic Exchanger Resin ing cost for the same number of gallons of
water with constant characteristics. The ap- EQUIPMENT EMPLOYED

plication of this type of resin is also advan-
tageous when used with waters having wide Natural zeolite minerals have a sandy nature
seasonal variations in hardness since all that while the synthetic zeolites are in the form of
is necessary is the adjustment of the salt rate small crystals or beads. A zeolite softener is,
to obtain softening cycles of predetermined therefore, very similar in construction and
length. appearance to ,a sand filter excepting that it
Fig. 14-3 is filled with a bed of zeolite instead of sand.
Figure 14-4 illustrates one type of zeolite
Effect of Salt Regeneration on Exchange
Capacity softener.
Although there are different modifications,
Salt Requirement Polystyrene Resin in general the softener consists of a steel
lbs sall/kilograin Exchange Capacity
hardness removed grains/cu ft shell, the bottom of which contains graded
sizes of gravel or anthracite. Above the
0.285 ............................. 18,000
0.323 ............................. 21,000 graded base is located the zeolite material,
0.385 ............................. 24,500 usually 24 inches or more in depth. In the
0.440 ............................. 27,000
0.490 ............................. 29,000
majority of softeners, the water enters the
top of the tank and passes down through the
PRESSURE WATER
TO EDUCTOR
SALT DISSOLVING
AND
STORAGE TANK
WATER
Courtesy Cochrane Division-Crane Company

Figure 14-4 • Zeolite Softener with Single Control Valve and Header Lateral Underdrain
92 BETZ HANDBOOK
zeolite. The size of the zeolite bed will de- turning a valve or valves. Semi-automatic
pend upon the exchange capacity of the zeo- operation entails pushing a button for each
lite, the hardness of the water being treated, regeneration step to actuate motorized valves.
the amount of water to be softened between The motive force can be electric, pneumatic
regenerations, and the flow rate. The bed or hydraulic. Completely automatic operation
must also be of sufficient size to allow proper requires no attendance. The usual method
contact time between the water and the zeo- consists of having a water meter preset so
lite and flow rate is usually regulated at 5-8 that when a certain number of gallons of
gallons per square foot per minute. In addi- water have been softened, an electrical con-
tion, it is necessary to provide a tank to hold tact is made in the meter. This electrical con-
the salt solution for regeneration, proper contact actuates a system of counters and timers
trol valves and interconnecting piping. which in turn actuates the valves controlling
In regenerating the unit, it is first back- the various regeneration steps.
washed to clean the bed. In backwashing,
water is first passed up through the bed and APPLICATIONS
sent to waste until it runs clear. The rate of
backwash should be sufficient to lift and ex- Zeolite softeners have a definite application
pand the bed to thoroughly clean it, but not in any process where soft water is an advan-
high enough to wash the zeolite out of the tage. Laundries find a decided use for these
unit. The required rate is a function of the softeners to reduce their soap consumption.
type of zeolite, water temperature and free- Softening offers an added advantage in im-
board in the tank above the zeolite bed and proving the appearance of the wash as dis-
the usual range is between 4 and 10 gallons colorations, due to soap curds-those insoluble
per square foot of bed area per minute. A particles of soap formed by the reaction of
clear water, free from suspended matter, soap with the hardness in the water-are
should always be available for backwashing. eliminated. Dye houses and other textile
A strong salt solution is next passed into the plants where finishing is accomplished also
softener at the top. Rinse water is next added find these softeners beneficial for the same
to the top of the softener to force the salt reasons, eliminating many causes for uneven
water or brine through the zeolite and out dyeing.
the bottom to waste. At first, this water will In certain districts where the water sup-
contain large quantities of calcium and mag- plies are quite hard, many small zeolite sof-
nesium chlorides together with unreacted teners are used in the home. In some cities
sodium chloride. As the bed becomes regen- softeners of the zeolite type have been in-
erated, the calcium and magnesium chlorides stalled to supply the public with softened
in the effluent are eliminated, and the excess water.
salt is gradually washed out by continuing to In boiler feedwater conditioning, the use
add further· quantities of fresh water, which of zeolite softening is frequently attractive
acts as a rinse. Rinsing is continued until all because of the simplicity of operation and
of the ex<;ess salt is washed from the bed and control. An outstanding advantage of zeolite
the unit is then returned to service. Many softening is the low hardness obtainable in
units are provided with a master valve which the treated water. Marked reduction in the
controls the various operations by merely scaling properties of a water is secured by
turning a hand wheel in steps as indicated at proper zeolite softening.
points on the rim of the valve. With the use of resinous and carbonaceous
Softeners can be manufactured for man- exchangers, it is possible to operate on the
ual, semi-automatic or completely automatic sodium cycle and to secure the advantages of
operation. In manual operation the regen- a low hardness effluent without danger of
eration steps are accomplished by manually silica pickup from the zeolite bed. Even
though such exchange materials are some- should be removed prior to softening. In such
what more expensive than greensand zeolite, cases, the additional expense of an iron re-
this increased installation cost is frequently moval system is required.
justified by salt savings and by avoiding any In boiler water conditioning, zeolite has a
silica increase while softening. marked advantage over lime-soda softening
insofar as it more completely removes hard-
LIMITATIONS ness. Whereas lime-soda softening may reduce
the hardness to as low as 10-30 ppm, depend-
Sodium zeolite softening should not be con- ing on the water, zeolite softening will show
sidered as a cure-all simply because it re- "zero" hardness or very close to "zero." Never-
moves scale forming solids from the water. theless, scale formation can take place in a
Sodium zeolite softeners require very little boiler using water softened by the zeolite proc-
maintenance and are quite simple to operate. ess. Invariably, some internal treatment is
Plant operators can be trained to operate necessary in a boiler following sodium zeolite
them properly in a very brief_ period because softening in order to prevent scale formation.
of this simplicity; however, many zeolite sof- Naturally, the quantity of internal treatment
teners have been installed to operate on required after zeolite is less than that required
. waters entirely unsuited for their use. There after lime-soda softening. As the amount of
are many types of waters and conditions of precipitated sludge formed in the boiler by the
operation for which sodium zeolite softening internal treatment depends on the original
is the proper solution; however, there have feedwater hardness, it follows that the amount
been numerous installations whereby the ap- of suspended matter in a boiler following the
plication of zeolite softening alone has in- use of zeolite softening will be small. Accord-
creased cost of operation and added several ingly, there is less chance of the sludge baking
problems which could be considered more on to the surface to form a scale than with
serious than those initially present. lime-soda softening or when using internal
Turbid waters should not be passed through treatment only.
a zeolite softener as the accumulated deposits One of the principal disadvantages of so-
will coat the zeolite grains and reduce effi- dium zeolite softening is that the hardness is
ciency. Water for a zeolite softener should removed without a reduction in alkalinity or
have a turbidity of less than 10 units. City total solids. This defect can be seen from equa-
supplies are generally filtered and accordingly tion ( 1). Calcium is removed, but is replaced
are suitable for zeolite softening. Most well by two ions of sodium, weighing slightly more
water-s are relatively free from turbidity as than the original calcium. With zeolite soft-
passage through the earth usually filters out ened water, since the hardness has been re-
suspended matter. Surface waters generally moved and the calcium exchanged for sodium,
tend to be turbid and in most cases must be the entire alkalinity is free to concentrate in
filtered prior to zeolite softening. Filtration, the boiler. Unless the natural alkalinity is
to be properly effective, should be preceded relatively low in comparison with total solids,
by coagulation. With these requirements, the zeolite softening results in high boiler water
zeolite system becomes more complex due to alkalinity and high carbon dioxide concentra-
the necessity of a coagulation and filtration tion in the steam. This condition often re-
system, increasing capital cost, space require- quires high rates of blowdown to maintain
ments and operating cost. In certain cases, fil- boiler water alkalinities within reasonable
tration will be omitted, particularly in small bounds. High alkalinities are objectionable as
installations, with the understanding that re- they tend to promote both carryover and
duction in efficiency and life of the zeolite are caustic metal embrittlement ( intercrystalline
to be expected. cracking), unless special precautions are taken
Ferric iron is detrimental to zeolite and with respect to chemical treatment of the
94 BETZ HANDBOOK
boiler water. investigation of these factors. Zeolite soften-

Where silica is a problem, there is the defi- ing may be coupled with other methods to
nite possibility of silica being picked up from provide a satisfactory solution to a problem.
the zeolite by the water, thereby adding to Zeolites fill a very important place in water
the problem. This undesirable condition can conditioning, but should be used only with a
be overcome by using a nonsiliceous zeolite full knowledge of their limitations. Entirely
material. too many installations have been made in the
Sodiuni. zeolite softened water is normally mistaken belief that sodium zeolite softening
corrosive, with the severity of the corrosive offered a final solution to a problem. Espe-
characteristics dependent on the amount of cially in boiler feedwater conditioning, the
dissolved gases, such as carbon dioxide and proper solution to a problem is rarely achieved
oxygen. In large cold water distribution sys- by the use of sodium zeolite softening alone.
tems, and ~n domestic hot water systems, seri- However, in combination with other external
ous corroswn may result following softening and internal treatment methods, zeolites per-
unless these corrosive gases are reduced to low form a very useful water conditioning func-
values by adjusting pH, deaeration or apply- tion.
ing positive corrosion inhibitors. REFERENCES
Many of the disadvantages outlined here
for zeolite can be overcome by methods in- L. F. Collins, "A Study of Contemporary Zeolites",
Journal, Am. Water Works Assoc., Vol. 29, pp.
volving other types of zeolite operating on 1472-1514 (1937)
other than the sodium cycles. R. Kunin, "Ion Exchange Resins", 2nd Ed., John
Whether zeolite should or should not be Wiley & Sons, New York, N.Y. (1958)
S. T. Powell, "Water Conditioning for Industry"
used depends on many factors and any indi- pp. 146-165, McGraw-Hill Book Co., Inc., Ne._,;
vidual case can be judged only following an York, N.Y. (1954)
15
Cation Exchange by
Hydrogen Zeolite
Principles of Cation Exchange . . . . . . . . . . . . . . . . . . 96
Hydrogen and Sodium Zeolite Units in Parallel. . . . 96
Hydrogen Zeolite with Alkali Neutralization . . . . . 97
Hydrogen Zeolite with Raw Water Neutralization. 98
Equipment Arrangements. . . . . . . . . . . . . . . . . . . . . . 98
Hydrogen and Sodium Zeolite Units in Parallel. . . . 98
Hydrogen :(eolite with Alkali Neutralization . . . . . 99
Hydrogen Zeolite with Raw Water Neutralization. 99
Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101
Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701
96 BETZ HANDBOOK
YDROGEN zeolite is the name given to a ready for use. Sulfuric acid is generally used
H group of non-siliceous organic materials,
either natural or synthetic, which have the
for the regeneration process due to its low
cost, although hydrochloric acid also is fre-
property when operated in the acid cycle, to quently used, particularly when high calcium
exchange hydrogen for cations such as cal- water results in the precipitation of calcium
cium, magnesium, sodium, etc. Upon exhaus- sulfate in the zeolite bed.
tion of the hydrogen zeolite bed, it is regen- The following equations illustrate the re-
erated with acid. Non-siliceous or carbona- actions which take place when sulfuric acid
ceous zeolite may also be operated in the is used for regeneration:
sodium cycle by regenerating with sodium
chloride (salt) and- when used for sodium (5)
zeolite softening possesses some advantages
over the siliceous zeolites.
ea IZ + H2S04 = H2Z + leaMg
Mg
An example of natural hydrogen zeolites lNa2 Na2
-are the various lignites found in the Dakotas;
From reaction ( 1) it is evident that all of
while sulphonated coal, coke, and synthetic
the bicarbonates are converted into carbonic
resms are examples of the man-made prod-
uct. acid (H2 C0 3 ). This acid is very unstable in
water solution and can easily be removed by
aeration or degasification.
PRINCIPLES OF CATION EXCHANGE
Thus, in addition to completely removing
When water containing calcium, magne- the hardness the bicarbonate alkalinity is also
sium and sodium ions is passed through a removed without substituting other salts in
hydrogen zeolite these ions are exchanged for its place and accordingly it is possible by this
hydrogen, and the bicarbonate, sulfate, ni- process to reduce the total solids. The re-
trate and chloride radicals are converted to moval of bicarbonate alkalinity without sub-
their respective acids, viz., carbonic acid stitution of any other ion is the outstanding
(H2C0 3 ), sulfuric acid (H 2S04 ) , nitric acid advantage in the use of hydrogen zeolite.
(HNOa) and hydrochloric acid (HCl) as Is Since these materials are non-siliceous in
shown by the following equations: nature no pick up of silica is experienced.
( 1) Furthermore, with this type of softener it is
~~
Na2
I
(He03h
possible to reduce to any predetermined
value not only the alkalinity due to calcium
and magnesium bicarbonates, but also the
alkalinity due to sodium bicarbonate.
(2)
While it is possible to rid the system of
I ~~ 1so4
Na2
(3)
carbonic acid by degasification, sulfuric, ni-
tric and hydrochloric acids cannot be re-
moved by this method. These acids must be
I ~~ I I
neutrali:;r:ed, otherwise severe acidic corrosion
Na2
el2 + H2Z = ~~~
Na2
Z+ 2Hel
of the metal in the system would result.
Three general methods for the neutraliza-
(4) tion of these acids are used.
HYDROGEN AND SoDIUM ZEoLITE UNITS IN
PARALLEL. A pre-determined percentage of
raw water is passed through a sodium zeolite
When the hydrogen zeolite becomes ex- which is operating in parallel with the hydro-
hausted it is backwashed and regenerated gen zeolite so as to obtain the desired alka-
with acid. After rinsing, the bed is again linity in the mixed effluent. This is accom-
plished with simple automatic rate-of-flow 4 ppm of S04 as S0 4 X 1.04 =

controls. 4.16 ppm of H2 S04 as CaC0 3
The correct amount of water to pass 2 ppm of N0 3 as N0 3 X 0.81 =
through the hydrogen zeolite unit depends on 1 .62 ppm of HN0 3 as CaC03
the factors: 6 ppm of Cl as Cl X 1.41 =
8.46 ppm of HCI as CaC03
The alkalinity of the influent ( Ar) .
The free mineral acidity (FMA) of the Free mineral acidity =
14.24 ppm as CaC03
hydrogen zeolite effluent (which depends
upon the sulfate + nitrate + chloride con- Substituting these values in the formula,
tent of the influent) . we have
The desired alkalinity in the mixed efflu- 34- 15
ent (Am). % H2Z treated = 100 X 39.5
34 + 14.24
The proper proportion can readily be cal-
% Na 2Z treated = 100 - 39.5 = 60.5
culated according to the formula:
Ar- Am HYDROGEN ZEOLITE WITH ALKALI NEUTRALI-
Percent H2 Z treated= 100 X - - - - - ZATION. Neutralization of the acidity of the
Ar + FMA hydrogen zeolite effluent is accomplished with
The use of this formula will be clarified by any suitable alkali, such as caustic soda
an example. Consider a raw water of the (NaOH), soda ash (Na 2 C0 3 ), etc. This
following characteristics: method is economical where the FMA in the
ppm
effluent is low or where the total volume of
Total Hardness as CaC03 = 36
water is small so that the cost of the chem-
Sulfate as S04 4
Nitrate as N03 = 2 ical used is negligible.
Chloride as Cl = 6 With the water used in the example above,
Methyl Orange Alkalinity as CaC03 = 34 the quantity of caustic soda required to es-
Calcium as CaCOl = 24 tablish a residual alkalinity of 15 ppm, as
Magnesium as CaC03 = 12 CaC0 3 is determined as follows:
Carbon Dioxide as C0 2 = 15
Add an amount of caustic soda equal to
It is desired to maintain an alkalinity of the FMA present plus enough to give the
15 ppm as CaC03 in the mixed effluent. desired alkalinity in the effluent.
We are given Ar = 34 ppm as CaCOl The amount of caustic soda required can
We want Am = 15 ppm as CaCOl
readily be calculated according to the for-
All that is required before applying the Inula:
formula is the value for FMA. Calculation
of the FMA value is based upon the fact that ppm NaOH = (0.8 X FMA) + (0.8 X Am)
the mineral acidity value is equal to the sum (The factor 0.8 in the equation is em-
of the acids of the sulfates, nitrates, and ployed to convert the FMA as CaC0 3 and
chlorides present in the water. In addition alkalinity as CaC0 3 to NaOH).
to converting the sulfates, nitrates and chlo- We have calculated FMA = 14.24 ppm, as
rides to their respective acids, they must also CaC03
be converted to a common denominator ·we are given Am = 15 ppm as CaC03
so that they can be added together; the com- ppm NaOH = (0.8 X 14.24) + (0.8 X 15)
mon denominator CaC0 3 is chosen because = 23.4
Ar and Am are already in terms of CaCOa. The advantage of this alkali neutralization
The constants necessary for these conversions method is that only a hydrogen zeolite unit
are as follows: is required. The additional investment cost
S04 - 1.04; N0 3 - 0.81; Cl - 1.41 of a sodium zeolite unit is avoided. Chemical
Therefore, the FMA value can be calcu- treatment cost is higher than with the hy-
lated: drogen-sodium zeolite units in parallel. Ac-
98 BETZ HANDBOOK
cordingly, the double unit process is usually generation level for the exchange capacity
employed for all systems except those where desired.
the total sulfate, nitrate and chloride content When the carbonaceous z~olite is operated
is very low or where the installation is only in the acid cycle, acid requirements for re-
of small capacity. generation will average 0.25 pound of 66°
HYDROGEN ZEOLITE-RAW WATER NEUTRA- Baume sulfuric acid per kilograin of cation
LIZATION. Neutralization of acidity of the hy- exchange, or 1. 75 pounds of acid per million
drogen zeolite effluent is accomplished by di- pounds of water treated for each part per
lution with the raw unsoftened water itself. million of cation exchanged.
The elimination of a sodium zeolite unit is
an advantage secured also by this system.
However, the hardness of the mixed effluent EQUIPMENT ARRANGEMENTS
is increased in direct proportion with the
quantity of raw water used for neutralization. HYDROGEN AND SomuM ZEOLITE UNITS IN
. This method is usually feasible only on a PARALLEL (Fig. 15-1) . Major equipment in-
water of low total hardness and high sodium cludes:
bicarbonate alkalinity. Rubber-lined steel pressure tank to con-
A non-siliceous resinous or carbonaceous tain hydrogen zeolite, with single valve
zeolite, operating in the sodium cycle, is re- control.
generated by passing a strong sodium brine Steel pressure tank to contain sodium zeo-
solution through the bed in the same manner lite, with single valve control.
as in an ordinary zeolite. The sodium salt Acid regeneration tank for hydrogen zeo-
may be sodium chloride, sodium sulfate or lite.
sodium nitrate, but sodium chloride is gen- Salt regeneration tank for sodium zeolite.
erally used because of its lower cost. This Control for automatically maintaining de-
type of zeolite softener requires the same sired ratio of flow to the hydrogen and so-
amount of sodium chloride for regeneration dium units.
as the conventional zeolite. Polystyrene type Closed wood degasifier with trays which
resins may be employed on the sodium or break up treated water into series of small
hydrogen cycle. The salt or acid requirements streams. Air from blower then comes into
for the sodium or hydrogen cycle may be intimate counter-current contact with the
quite variable and will depend on the re- water and scrubs out the carbon dioxide.
VENT TO
RAW WATER ATMOSPHERE
DEGASIFIER
HYDROGEN SODIUM
ZEOLITE FOR C02
ZEOLITE
UNIT REMOVAL
UNIT
SATURATED NaCI
DILUTE H2S04 FOR BRINE FOR Na2 Z
H2Z REGENERATION REGENERATION
TO USE
RUBBER LINED PIPE FOR
H2 Z AND MIXED EFFLUENTS
Figure 15-1 • Hydrogen and Sodium Zeolite Units in Parallel

RAW WATER VENT TO

r---'"--_., ATMOSPHERE
ALKALINE
DEGASIAER
FOR C02 NEUTRALIZER
REMOVAL
HYDROGEN
ZEOLITE
UNIT
RUBBER LINED PIPE FOR

H2Z EFFLUENT STORAGE FOR AERATED
NEUTRALIZED H2Z EFFLUENT
Figure 15-2 • Hydrogen Zeolite with Alkali Neutrali;;ation
HYDROGEN ZEOLITE WITH ALKALI NEUTRALI- counter-current contact with the water
ZATION (Fig. 15-2). Major Equipment In- and scrubs out the carbon dioxide.
cludes: A neutralizing feeder.
Rubber-lined steel pressure tank to con-
tain hydrogen zeolite, with single valve HYDROGEN ZEOLITE WITH RAW WATER NEU-
control. TRALIZATION (Fig. 15-3). Major Equipment
Acid regeneration tank for hydrogen zeo- Includes:
lite. Rubber-lined steel pressure tank to con-
Closed wood degasifier with trays which tain hydrogen zeolite, with single control
break up treated water into series of small valve.
streams. Acid regeneration tank for hydrogen zeo-
Air from blower then comes into intimate lite.
RAW WATER
VENT TO
ATMOSPHERE
DEGASIFIER
FOR <Xl2
REMOVAL
HYDROGEN
ZEOLITE
UNIT
DILUTE ACID FOR BLOWER

H2Z REGENERATION
RUBBER LINED PIPE FOR TO USE

H2Z EFFLUENT
STORAGE FOR AERATED
NEUTRALIZED H2Z EFFLUENT
Figure 15-3 • Hydrogen Zeolite with Raw Water Neutrali;;ation

100 BETZ HANDBOOK
Figure 75-4 • Laboratory Testing of Ion Exchange Resins

Control for automatically maintaining de- The cost of chemicals must be considered.
sired ratio of raw water and hydrogen In one locality, acid and salt may be expen-
zeolite flow. sive while in another they may be inexpen-
Closed wood degasifier with trays which sive. Installation, maintenance, depreciation,
break up treated water into series of small etc., all must be taken into consideration.
streams. Air from blower then comes into An engineer, specializing in water treatment,
intimate counter-current contact with the should carefully evaluate the many variables
water and scrubs out the carbon dioxide. and make the final recommendations for the
equipment which will provide optimum re-
APPLICATIONS sults with a minimum of cost.
The properly blended and degasified hy-
Cation exchange by hydrogen zeolite is drogen zeolite-sodium zeolite effluent is low
utilized in boiler feed water conditioning in hardness and alkalinity, but is saturated
principally on a water low in hardness and with oxygen and is consequently of a cor-
high in bicarbonate alkalinity. The removal rosive nature. Corrosion resistant pipe or
of sodium and bicarbonate ions by the hydro- pipe specially painted or lined may be re-
gen zeolite process constitutes the outstand- quired to handle the water and deliver it to
ing advantage of this process inasmuch as the feed water heater.
this removal is secured without the substitu-
In general the hydrogen zeolite method for
tion of other ions. It is thereby possible to
removing calcium and magnesium bicarbo-
effect marked reduction in the solids content
nate is more expensive than the lime-soda
of a high bicarbonate water.
process, but provides an effluent of lower
The hydrogen zeolite process finds exten- hardness. The operating personnel must be
sive use in the manufacture of raw-water ice carefully trained inasmuch as the handling
and in the carbonated beverage industry. To of sulfuric acid is somewhat hazardous and
produce clear raw-water ice rapidly and to the control tests necessary are very important
prevent a brittle product, the water must be for proper operation.
low in alkalinity. In the carbonated beverage
This type of softening for treatment of
industry, low alkalinity is necessary because
boiler makeup water is not applicable on
many soft drinks are prepared with acid ex-
water of high hardness and low alkalinity or
tracts that require certain pH values to ob-
on a water high in silica. In cases where the
tain the desired flavor.
raw water silica concentration is high, the
concentration of silica becomes the con-
LIMITATIONS trolling factor with respect to boiler concen-
Before definite recommendation for this tration control. In such cases, hot process sof-
type equipment can be given it is necessary tening methods in which silica is easily
to make a thorough plant study inasmuch as removed are more suited for use in the
operating cost is composed of several factors. preparation of boiler feedwater.
If the water is muddy, sedimentation, co-
agulation and/or filtration will be necessary REFERENCES
because a clear influent must be delivered to
R. F. Goudey, "Removal of Salts from Water",
the hydrogen zeolite unit otherwise the sus- Journal, Am. Water Works Assoc., Vol. 32, pp.
pended matter will coat the grain surfaces 435-455 ( 1940)
and materially reduce efficiency. Further- R. Kunin, "Ion Exchange Resins", 2nd Ed., John
more, the influent should be free from ferric Wiley & Sons, New York, N.Y. (1958)
E. Nordell, "Water Treatment for Industrial and
iron because iron will also be precipitated on Other Uses", 2nd Ed., pp. 445-458, Reinhold Pub-
the grain surfaces. lishing Corp., New York, N.Y. (1961)
102 BETZ HANDBOOK
16
De111ineralization
Ion Exchange Resins. . . . . . . . . . . . . . . . . . . . . . 103

Demineralization Equipment. . . . . . . . . . . . . . . . 105
Applications and Limitations . ......... 0 0 0 0 0 • 108
EMINERALIZATION is the process of re- did not gain wide acceptance because of the
D moving the mineral salts from water
by ion exchange. Deionization is generally
problems involved in feeding the fluoride and
in disposing of the waste water from the re-
considered synonymous with demineralization. generation operation. The waste water has
Only those substances which ionize in water a high fluoride content and is toxic in high
can be removed by this process. With most concentration. With the development of the
industrial water sources, it is possible to use highly basic anion exchange resins, the use
demineralization to remove mineral salts to of fluoride is no longer necessary. The highly
the same extent or to a greater extent than basic anion resins are capable of removing
would be possible with distillation. Deminer- silica from demineralized water and the proc-
alization involves two ion exchange reactions. ess now competes directly with distillation as
The cations such as calcium, magnesium and a source of makeup water for high pressure
sodium are removed in a hydrogen cation ex- boilers. In many cases, demineralization is
changer. The process and equipment are the more economical than distillation and can be
same as those used for hydrogen zeolite sof- operated in such a manner that the makeup
tening. The anions, such as sulfate and chlo- water will be of superior quality to distilled
ride, are removed in an anion exchanger. water.
Various arrangements of equipment are em-
ION EXCHANGE RESINS
ployed dependent upon the requirements of
the installation. Demineralization is carried out with the use
Demineralization and distillation are the of two types of materials. The cation ex-
two most commonly used methods for pro- change material replaces the metallic cations
ducing large quantities of water with a low with hydrogen ions. This process is identical
solids content. In comparison with the cost to hydrogen zeolite softening. Sulfonated coal
of producing distilled water, demineraliza- is sometimes used as the cation exchange
tion by ion exchange will frequently be more material for economy in the use of acid, but
economical for many industries. For this rea- it is more common to use one of the organic
son, the process of demineralization has found cation exchange resins in which the active
wide acceptance among industries requiring group is a sulfonic. acid group. The remainder
process water of low solids content. of the resin is of such a nature that it is in-
Prior to the development of the highly soluble, chemically and physically stable, and
basic anion exchange resins, demineralization permeable to diffusion of ionsfrom the water.
of a water by ion exchange did not lower the Ion exchange is an equilibrium reaction
silica content. Without silica removal, de- which is reversible. In the hydrogen cation
mineralization possessed limited application exchange resins, the active group shows a
for the preparation of boiler feedwater. greater affinity for other cations in preference
Water of quality comparable to that of dis- to hydrogen ion. This process is reversed
tilled water is required for high pressure when the resin is regenerated but the reverse
boiler operation. The presence of silica in process takes place only when the hydrogen
such water is most undesirable, since the ion is present in high concentration. The
silica may permit the formation of boiler tube position of equilibrium depends upon the
deposits or vaporize with the steam and cause relative concentrations of cations present in
turbine blade deposits. It is therefore neces- the water.
sary for demineralized water to have a very During the exchange operation, the princi-
low silica content in order to be useful for pal reaction is indicated by the following
boiler feedwater purposes. equations. In these equations, the symbol
With the weakly basic anion exchange res- R · S0 3 is used to represent the portion of the
ins, it is possible to use fluorides to obtain silica cation exchange material which does not go
removal with demineralization. This process into solution.
104 BETZ HANDBOOK
(1) Strongly basic anion exchanger:

R·S0 3H + NaCI = R·S03Na + (4)
hydrogen + sodium sodium +
exchange resin chloride resin
R4N ·OH + H2Si03 = R4N ·HSi03
basic + silicic silicate
HCI exchange resin acid resin
hydrochloric + H20
acid +water
(2) In the operation of a strongly basic anion
2R·S03H + CaS04 = (R·S03)zCa exchanger with downward flow, it is believed
hydrogen + calcium = calcium
that the anions of the strong acids will be
exchange resin sulfate resin
removed at the top of the bed with silicate
+ H2S04 representing one of the last anions to be re-
+ sulfuric
acid moved. The top of the bed is exhausted first
and the portion of the bed which is effective
!he ani~n. exchange resins are organic
in removing silica is gradually displaced
·· re~1.ns contammg active amine or quaternary
downward. Eventually the bottom layer in
am1?e gr~ups. The anion exchange resins are
the bed will no longer effectively remove
ava1lable m two types known as weakly basic
silica and the effluent will show increasing
and strongly basic anion exchangers. The
concentrations of silica. When this condition
weakly basic type will remove sulfate chlo-
is reached the bed will still remove other an-
ride and nitrate, but will not remove silica or
ions, with the result that silica break-through
carbon dioxide. The weakly basic resins can
will take place before there is a significant rise
~e. use.d to advantage where the presence of
in conductivity of the effluent.
s1~1ca m the demineralized water is not sig-
m?cant. The .s~rongly basic anion exchangers The purity of demineralized water and
wlll remove s1hca, sulfide and carbon dioxide distilled water is frequently expressed in terms
as well as the other common anions. of specific resistance or specific conductance.
One micromho of specific conductance will
The weakly basic anion exchangers can
normally indicate a dissolved solids content
remove silica only with the feed of sufficient
of 0.5-0.6 ppm. Completely pure water would
fluoride to convert the silicate to a fluosili-
cate. Silicate acts as the anion of a weak acid have a ~pecific conductance of approximately
0.04 m1cromhos due to a small quantity of
~nd is not removed by the weakly basic an-
lOll exchanger. The fluosilicate acts as an
hydrogen ion and hydroxide ion. These ions
anion of a strong acid and can be removed result from the dissociation of a small portion
by a weakly basic anion exchanger. This of the water molecules. Water of such purity
pr?cess is not applied to the preparation of Fig. 16-1
bmler makeup water at the present time. Quality of Low Solids Water from
The strongly basic anion exchangers will Various Sources
(Laboratory Scale Tests)*
remove the anions of both strong acids and
weak acids. For this reason, direct silica re- Specific
Conductance,
moval is possible. The following equations micromhos
can be used to illustrate the manner in which
the anion exchange resins operate. Calculated Maximum Quality ............... 0.038
Single Distilled Water (in glass) . . . . . . . . . . . . . 2.0
Weakly basic anion exchanger: Triple Distilled Water (in glass) . . . . . . . . . . . . . 1.0
Monobed Demineralization ................. 0.055
(3) 28 Distillations in Quartz .................. 0.044
RNH 3·0H + HCI = RNH 3·CI * '' Amberlite Mono bed Deionizotion,'' Rohm and Haas Company
Philadelphia, Po., 1950. '
basic + hydrochloric = chloride
exchange resin acid resin
+ H20 can not be produced by methods which are
+water commercially available at the present time.
The purity of various types of demineralized ployed only when the bicarbonate alkalinity
and distilled water are outlined in Figure of the raw water is low. Otherwise, the cost
16-1. of removing carbon dioxide in the anion ex-
changer becomes a sizeable factor.
DEMINERALIZATION EQUIPMENT An alternate arrangement is to place an
aerator between the hydrogen cation ex-
The equipment employed for demineraliza- changer and the highly basic anion ex-
e
tion is similar to that employed for hydrogen changer, as illustrated in Figure 16-3. This
zeolite softening. The piping and exchanger arrangement is suitable for water with high
shells are generally lined with synthetic rub- bicarbonate alkalinity, since the carbon di-
ber or plastic materials. Small fittings which oxide is liberated to the atmosphere in the
can not be conveniently protected by coatings aerator. In place of an aerator, it is possible
are constructed of Monel, plastic, or stainless to install a vacuum deaerator which will re-
steel. move nearly all dissolved gases such as oxy-
The aeration towers which are used for gen, in addition to the carbon dioxide.
removal of carbon dioxide are frequently of Another arrangement is to pass the raw
wooden construction with any exposed metal water into the hydrogen cation exchanger,
parts of Monel, plastic, or stainless steel. Ce- then into a weakly basic anion exchanger fol-
ramic materials are also being employed. lowed by an aerator and then a highly basic
Final effluent water lines and storage tanks anion exchanger. This system is illustrated in
should be protected with a suitable paint or Figure 16-4. It is necessary to regenerate
lining to prevent corrosion. Such corrosion the highly basic anion exchanger with caustic
will arise from the dissolved oxygen content soda, but the weakly basic anion exchanger
of the demineralized water and would con- may be regenerated with soda ash or with
taminate the water with iron oxide. the waste regenerant from the highly basic
Demineralization with silica removal may anion unit. As a consequence, the operating
be carried out in several steps. The simplest cost for chemical regeneration is lower with
arrangement is illustrated in Figure 16-2 this type of arrangement when the raw
in which the raw water passes through a water has relatively high sulfate or chloride
cation exchanger directly to a highly basic content.
anion exchanger. This arrangement is em- Maximum purity of demineralized water
RAW WATER
l 1
ACID CATION ANION CAUSTIC SODA

REGENERANT EXCHANGER EXCHANGER REGENER ANT
~ 1 T
~
TO SERVICE
Figure 16-2 • Two-Bed System for Demineralization and Silica Removal

106 BETZ HANDBOOK
I
RAW WATER
1 l
AERATOR
CATION ANION
EXCHANGER EXCHANGER
ACID CAUSTI C SODA

REGENER ANT REG ENE RANT
~ T T
UJ
TO SERVICE
CLEARWELL_~
Figure 16-3 • Two-Bed System with Aerator for Demineralization and Silica Remoual
can be obtained by placing a number of ex- types of resins are intermixed. The resins are
changer units in series. A similar result can of different density and during backwashing
be obtained by mixed bed demineralization. will separate into two layers. Each layer is
This arrangement uses a mixture of cation regenerated and rinsed. After rinsing the
and highly basic anion resins in the same resins are remixed by air agitation and the
shell. Figure 16-5 illustrates the internal ar- exchange operation is resumed.
rangement of the resin beds during regenera- The theoretical purity of the effluent water
tion. from a mixed bed demineralizer is very high.
During the exchange operation the two The equipment requires a complicated ar-
RAW WATER
j_ _L
AERATOR
ACID WEAKLY STRONGLY
CATION ALKALr BASIC BASIC CAUSTIC SODA
REGENERANT ANION ANION
EXCHANGER REGENERANT
Q~="'-
~
EXCHANGER
T 1
~
TO SERVICE
Figure 16-4 • Three-Bed System with Aerator for Demineralization and Silica Remoual
RAW WATER INFLUENT to growth of bacteriological slime. Depending

RINSE INFLUENT upon characteristics of the water source, the
BACKWASH EFFLUENT influent water to the demineralizing system
CAUSTIC SODA REGENERANT may require a preliminary treatment of co-
AIR agulation, filtration and possibly chlorination.
I EFFLU~ NT However, excess chlorination will cause de-
_...... -......... terioration of the cation resin so that some-
times dechlorination is required. The usual
methods of dechlorination involve the feed-
ing of sodium sulfite or passing the water
through a bed of activated carbon ahead of
1- the cation exchanger. In some cases, cold
z process softening is applicable prior to the
<(
ct:: demineralizer, particularly if there are other
IJ.J
1-z plant water requirements for which demin-
ANION EXCHANGE zW eralized water is not necessary.
RESIN
wfB
3ct:: The operation of equipment is similar to
... 1..1..0
ZG
IJ.JU
(f)-
zct::
_:::>
<(
the operation of zeolite softening equipment.
Operating water flow will usually be in the
range of 5-8 gallons per minute per square
foot of bed surface. A minimum bed depth
CATION EXCHANGE ct::I.L.
....J of 30 inches is usually recommended and
RESIN :::> the bed depth will frequently be much
(/)
greater in order to obtain the desired ex-
change capacity for the unit. Operating flow
__.... rates are· sometimes expressed in terms of
AIR
INFLUENT
f J
DEMINERALIZED EFFLUENT
gallons per minute per cubic foot of bed. In a
moist condition, the exchange resins will
weigh approximately 30-50 pounds per cubic
RINSE EFFLUENT
SPENT REGENERANT foot. The cation exchange resins are heavier
BACKWASH INFLUENT than the anion exchange resins.
Figure 16-5 • Arrangement for Mixed Bed Backwash rates of 4-8 gallons per minute
Demineralization per square foot are frequently employed, de-
rangement of valves in order to permit using pending upon density of the resin and tem-
caustic soda in one regeneration step and acid perature of the backwash water. Higher back-
during the other step. There is also likely to wash rates are required for the heavier resins
be a variation in capacity of the unit, unless and for higher water temperature, due to
remixing of the resins after regeneration is reduced viscosity of the water. Backwash rate
thorough and uniform. is normally adjusted to provide between 50
Regardless of the relative arrangements of and 100% expansion of the bed. In some
the exchangers. it is essential that the influent plants, where the influent water is not com-
water be free of suspended solids. Silt and pletely free of suspended solids, it has been
organic matter are particularly objectionable, found advantageous to install filter bed agita-
since these materials can deposit in the ex- tors to obtain maximum cleaning of the resin
changer beds and reduce the capacity of the during backwash.
units by coating the beads of exchanger resin Regenerant flow rates are usually low and
with films which either prevent or retard the close to approximately one gallon per minute
movement of anions and cations through the per square foot. With strongly basic anion
resins. Deposits of organic matter may lead exchangers, maximum removal of silica in a
108 BETZ HANDBOOK
given period of time will be obtained if the gen cation exchanger will have an effluent pH
caustic soda solution is heated to 90-100 F, or in the acid range and a certain value for free
slightly higher. mineral acidity. Operation of ,the cation unit
With the cation exchangers, the concen- can be followed to some degree by means of
tration of the regenerant solution may be in a continuous pH recorder. Sodium will break
the range of 1-10% for hydrochloric acid and through the hydrogen cation unit a short
1-5% for sulfuric acid. When using sulfuric time before there will be a significant increase
acid to regenerate a resin cation exchanger in pH. For maximum purity of the effluent
it is sometimes necessary to start at a low water from the system, it is necessary to check
concentration of acid, which is increased in the hydrogen cation exchanger effluent for
steps if the influent water to the unit con- sodium content when the end of the expected
tains an appreciable calcium content in com- run approaches. If maximum purity is not
parison to the other cations present. This essential, determination of pH and free min-
procedure is necessary to avoid precipitation eral acidity are used as the control tests.
of calcium sulfate upon the resin. Hydro- The quality of the effluent from the
chloric acid does not require step-wise re- strongly basic anion exchanger can be checked
generation procedure. It is more difficult to by means of a conductivity recorder. Silica
handle than sulfuric acid and for this reason and carbon dioxide will break through the
its use is avoided. unit somewhat ahead of other anions. For
In the regeneration of strongly basic anion minimum silica in the effluent water, silica
exchangers, the strength of the caustic soda determinations must be conducted as the end
solution is in the range of 2-4%. With weakly of the expected run approaches.
basic anion exchangers, the strength of the
sodium carbonate regenerant will normally
be in the range of 2-4%. Ammonia can also
be used as a regenerant if silica removal is Demineralization with silica removal has
not required. proved to be an economical method for pro-
Following introduction of the regenerant viding makeup water for high pressure boiler
solution, the unit is rinsed to remove excess operation. In many cases, it is a more eco-
regenerant. Rinsing will generally require 50 nomical method of makeup water prepara-
to 150 gallons of water per cubic foot of tion than distillation using single effect evap-
resin. About 5 to 10% of the influent water orators. Demineralization either with or
will be wasted from the system as a result of without silica removal has proved to be a
backwash and rinse operations. It should be useful and economical process for many in-
noted that the hydrogen cation exchanger dustries. The use of demineralized water has
can be rinsed with raw water, but the anion been applied to the manufacture of chemical
exchanger must be rinsed with effluent from and pharmaceutical products. Demineralized
the hydrogen cation exchanger. Rinse water water is used in some textile processes, in the
from the anion unit can be recycled to the silvering of glass, and in some electro-plating
raw water supply when the conductivity of processes. It has also been used widely in
the rinse water decreases to the conductivity various manufacturing processes where qual-
of the raw water. supply. A small portion of ity of the finished product may be affected
the initial rinse water from the hydrogen by the presence of mineral salts in water
cation exchanger can be reclaimed and used coming in contact with the product.
at the beginning of the next regeneration At the present time, hydrogen cation ex-
cycle, since this water will have an appreci- changers are available which can tolerate
able acid content. operating temperature as high as 250 F.
Control of demineralization plants requires Strongly basic anion exchange materials must
determination of several factors. The hydro- generally be operated at a water temperature
not exceeding 105 F. As a consequence, it is more precise methods of operation control.

necessary to carry out demineralizing proc- These changes will permit production of de-
esses at relatively low temperatures at the mineralized water with greater purity and at
present time. When the raw water source is lower cost.
very high in dissolved solids, evaporation will
frequently prove to be the most economical
method of obtaining water which will be REFERENCES
nearly free of mineral salts. If the raw water
S. B. Applebaum, "Experiences With Silica Re-
is contaminated with substances of an or- moval Demineralizer Plants", Proceedings, Mid-
ganic nature, distilfation may be required for west Power Conf., Vol. XIII, pp. 275-289 (1951)
purification. Coagulation, chlorination and H. W. Frazer, "Special Operating and Regenerat-
filtration will frequently reduce organic con- ing Techniques for Ion Exchange", Proceedings,
Engineers Soc. of Western Penna., pp. 83-96
tent of water to very low values. However, (1958)
not all organic materials can be completely N. W. Frisch and R. Kunin, "Long-Term Operat-
removed by such methods. Many organic ing Characteristics of Anion Exchange Resins",
Ind. and Eng. Chem., Vol. 49, pp. 1365-1372
substances can not be removed by deminerali- (1957)
zation, since they do not ionize in water. M. E. Gilwood, C. Calmon and A. H. Greer, "Sil-
Several years ago the fluosilicic acid process ica Removal Characteristics of Highly Basic Anion
Exchangers", Proceedings, Engineers Soc. of West-
for silica removal was the only one which ern Penna., pp. 119-130 (1951)
was available for practical application to J. Harlow, V. J. Calise and M. Lane, "Boiler
the preparation of boiler makeup water Feedwater Treatment For A High-Pressure, High-
Makeup, Power And Steam Producing Plant",
with demineralization. This process is now Proceedings, Midwest Power Conf., Vol. XIII, pp.
obsolete, having been superseded by a more 290-303 ( 1951)
efficient and economical process. Since the R. Kunin, "Ion Exchange Resins", 2nd Ed., John
Wiley and Sons, Inc., New York, N. Y. ( 1958)
present methods of demineralization are com-
F. C. Nachod and J. Schubert, "Ion Exchange
paratively new and being widely applied, it Technology", Academic Press, New York, N. Y.
can be expected that considerable progress ( 1956)
will be made within the next few years. Such R. I. Smith and H. D. Reppin, "Four Years Of
Operating Experience With A 3200-GPM Demin-
progress will most likely result in more effi- eralizer", Proceedings, Am. Power Conf., Vol.
cient use of regenerant chemicals and in XXIII, pp. 635-649 ( 1961)
110 BETZ HANDBOOK
Dealkalization by
Chloride-Anion Exchange
Use for Dealkali::;ation. . . . . . . . . . . . . . . . . . . . 111

Chemistry Involved. . . . . . . . . . . . . . . . . . . . . . . 111
Equipment Required. . . . . . . . . . . . . . . . . . . . . . 113
Abplications. . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Limitations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
EALKALIZATION by chloride-anion ex- USE FOR DEALKALIZATION

D change is a process whereby a sodium
zeolite softened water is passed through a The basic usefulness of this new process lies
second exchange unit which removes anions in its ability to reduce the alkalinity of the
such as sulfate, nitrate, carbonate and bi- treated water without the use of acid.
carbonate. These anions are replaced by The alkalini~y of a water is unchanged in
chloride. Salt is used to regenerate the chlo- passage through a sodium zeolite softener.
ride-anion exchanger. Since carbonate and Concentration of feedwater alkalinity in the
bicarbonate ions are exchanged for chloride boiler may develop high boiler water alka-
in this process, alkalinity reduction is achieved linities, with possible danger of foaming and
without the use of acid. carryover of solids with the steam. In addi-
It is theoretically possible to employ un- tion, decomposition of alkalinity under boiler
softened water as influent to the chloride- water temperatures will produce carbon di-
anion exchanger. However, at this stage in the oxide in the steam. Carbon dioxide, thus
development of this process, it is the con- formed, is a major source of steam and co.n-
sensus among equipment manufacturers that densate line corrosion.
it is desirable to precede the dealkalizer with a These disadvantages to sodium zeolite sof-
sodium zeolite softener. With unsoftened tening can be minimized by acid treatment
water there is the danger of precipitating of the zeolite effluent or by employing hy-
calcium carbonate and magnesium hydroxide drogen zeolite softening. Both of these treat-
in the dealkalizer bed. In addition, the chlo- ment methods involve the use of acid, which
ride-anion exchange bed is even more critical is regarded with disfavor in small plants,
with respect to suspended matter than the without adequate personnel.
conventional sodium zeolite cation exchanger. While not as efficient or economical as
Because the anion exchange resin is lighter acid treatment or hydrogen zeolite softening,
than the cation exchanger, it can be back- dealkalization by chloride-anion exchange
washed only at low rates, insufficient to re- possesses the advantage of securing similar
move suspended matter. The use of a sodium alkalinity reduction without the necessity of
zeolite, preceding the anion exchange, serves handling acid.
as an additional safeguard for the removal
of suspended matter. Even for the sodium CHEMISTRY INVOLVED
zeolite unit, turbidity of the influent water
should not exceed 10 ppm. Dealkalization by chloride-anion exchange
RAW WATER
SALT SODIUM CHLORIDE SALT e.

CAUSTIC
REGENERANT ANION REGENERANT
ZEOLITE
1-::-'::. I:1
~ -~1
RESIN
UNIT
UNIT
L..==.:J
TO USE
Figure 17-1 • Sodium Zeolite-Chloride Anion Exchange

112 BETZ HANDBOOK
makes use of highly basic anion exchange of chloride. Unlike demineralization, solids
resins employed in demineralization. In de- are not removed by· dealkalization, but con-
mineralization, a raw water is first passed verted to the sodium chloride form. Reac-
through a cation exchanger, where cations tions in the chloride-anion exchangers are as
are exchanged for hydrogen ions. Calcium, follows:
magnesium and sodium are removed and the (HC0 3) - + R·CI R·HC0 3
cation exchanger effluent contaiils sulfuric bicarbonate + chloride = bicarbonate
acid, hydrochloric acid, carbonic acid, etc., ion anion anion
which are then adsorbed in passing through resin resin
the second unit, the anion exchanger. The + Cl-
cation unit is regenerated with acid and the +
chloride
anion unit with alkali. This demineralization ion
process results in the complete removal of (504)= + 2R·CI = R2·S04 + 2CI-
both cations and anions, producing an efflu- sulfate chloride sulfate chloride
ent comparable to distilled water. ion + anion = anion + ion
resin resin
Chloride-anion exchange differs from de-
mineralization in that cations and anions are Of course, there is no advantage secured
not removed from solution, but are ex- in exchanging sulfate for chloride. Nitrate is
changed for an equivalent amount of other also exchanged for chloride in this process.
ions of less undesirable characteristics. In de- Where the sulfate and nitrate content of a
alkalization, two separate units are employed. water is high, the cost of the dealkalization
The first unit is a conventional sodium zeo- process is increased by the exchange of these
lite softener, regenerated with sodium chlo- ions, from which exchange no advantage is
ride, which exchanges calcium or magnesium secured.
ions for sodium ions. The reactions are as However, as indicated in the preceding re-
follows: actions, bicarbonate is removed by passage
Ca++ + Na 2Z = CaZ + 2Na+ of a sodium zeolite softened water through
calcium + sodium = calcium + sodium the chloride-anion exchanger and this re-
ion zeolite zeolite ion moval of alkalinity is the chief objective of
Mg ++ + Na2Z = MgZ the process~namely removal of alkalinity
magnesium + sodium = magnesium without the use of acid.
ion zeolite zeolite The chemical reactions of regenerations
+ 2Na+ with this process are as follows:
+sodium
ion 1st Unit~Sodium Zeolite
Of course, as previously mentioned, anions CaZ + 2NaCI = Na2Z +
CaCI 2
are not affected by passage through a sodium calcium + sodium = sodium + calcium
zeolite unit. Alkalinity, sulfate, chloride, etc., zeolite chloride zeolite chloride
pass through unchanged. 2nd Unit~Chloride-Anion Exchanger
In the second unit, the exchange material R·HC0 3 +
NaCI = R·CI
is a highly basic resin anion exchange ma- bicarbonate + sodium = chloride
terial, also regenerated with sodium chloride. anion chloride anion
In passage through this unit the sodium salts resin resin
present in the influent, such as sodium bicar- +NaHC03
bonate and sodium sulfate, are converted to + sodium
anequivalent amount of sodium chloride. In bicarbonate
other words, the dealkalization unit removes The exchange capacity of the dealkaliza-
the anions present in the influent by exchang- tion unit is affected by several variables. Data
ing these anions with an equivalent amount have been presented to the effect that ca-
pacity for the total exchangeable anions (bi- should not exceed five gallons per square foot
carbonate, carbonate, sulfate and nitrate) per minute.
will decrease with an increasing chloride con- Salt regeneration rates range between 3.0-
tent in the raw water. The total anion 4.0 pounds of salt per cubic foot of resin. It
exchange capacity (including the raw water is desirable to employ softened water in pre-
chloride) is not affected by variation of the paring regenerant brine. Rinse and regener-
percentage of the total anions which are ant rates of 0.5 gallons per minute per cubic
present in the chloride form. However, since foot of resin are recommended.
the purpose of the process is the removal of
bicarbonate alkalinity, the capacity of the APPLICATIONS
resin bed for bicarbonate removal will de-
crease with increasing chloride concentrations The major usefulness of dealkalization by
of the raw water. ion exchange is in relatively small plants
The process is most efficient when the where the hazard of acid handling is a con-
anions are low with the exception of bicarbon- trolling factor in deciding the type of ex-
ate. There is no advantage in exchanging sul- ternal treatment process to be installed.
fate and nitrate with chloride, and yet regen- Where neutralizing amines are employed, the
erant must be employed for this purpose. use of a chloride-anion exchanger may per-
Consequently, for low cost operation, it is mit economy, since any reduction in the C0 2
desirable that the sulfate, chloride and nitrate content of the steam will be reflected in a
content of the raw water be low in compar- dec.reased requirement for the neutralizing
ison with the bicarbonate. amme.
Exchange capacities of 5-10 kilograins ex- Only one regenerant chemical (salt) is re-
changeable anions (expressed as CaC0 3 ) per quired, although as mentioned previously,
cubic foot of resin have been reported. Im- improvement in the exchange capacity may
provement in exchange capacity has been be achieved through the use of caustic soda
noted when sodium hydroxide is used with along with the salt. There is no hazard in-
the sodium chloride in the regenerating solu- volved in the handling of salt and it is a
tion, in the ratio of approximately one pound cheap, readily available material. Where suf-
of caustic soda for each ten pounds of so- ficient competent personnel are not available
dium chloride. for controlling the other methods of alkalinity
At one of the first installations of the de- reduction, the chloride-anion exchange proc-
alkalization process, it was reported that the ess is definitely advantageous.
addition of sodium hydroxide to the sodium
chloride regenerant resulted in an increase LIMITATIONS
in the exchange capacity of approximately
With the installation of proper equipment
45-60%. At this plant, raw water alkalinity
and adequate personnel, there is little hazard
as calcium carbonate is 145-150 ppm and the
involved in handling acid. Direct acid treat-
dealkalization unit is taken off line and re-
ment of a sodium zeolite effluent or the use
generated when the effluent alkalinity IS m
of hydrogen-sodium zeolite softening will ac-
the range of 27-31 ppm as CaC0 3 •
complish alkalinity removal more efficiently
and economically than the chloride-anion ex-
EQUIPMENT REQUIRED
change process. In addition, in plants of suf-
The chloride-anion exchanger is similar in ficient size to justify the installation, the hot
construction to a conventional sodium zeolite lime-hot ion exchange process will prove
unit and equipped with the same accessories more economical than chloride-anion ex-
for regeneration. A resin bed depth of 30-36 change.
inches is recommended. Operating flow rate Like other treatment methods, dealkaliza-
114 BETZ HANDBOOK
tion by chloride-anion exchange, reduces but quently, in a plant with high carbon dioxide
does not eliminate the co1 content of the content of the steam and satisfactory corro-
steam. Dependent on the type of equipment sion control with the filming amines, the in-
in which steam is condensed and the degree stallation of a chloride-anion exchanger to
of stratification of carbon dioxide in condens- lower the carbon dioxide in the steam would
ing equipment, severe corrosion problems can not necessarily permit a reduction in the
result even from relatively low C0 2 concen- amine rate of feed.
trations in the steam. Where filming amines
are used for control of carbon dioxide cor- REFERENCES
rosion, the amine feed rate is substantially in-
dependent of C0 1 concentration. The quan- S. B. Applebaum, "Experiences with Chloride An-
ion Exchangers for Reducing Alkalinity Without
tity of amine required is only sufficient to Acid", Proceedings, Engineers Soc. of Western
provide a non-wettable film on the metal Penna., pp. 1-8 (1952)
surfaces. Except under unusual circumstances, A. E. Kittredge, "Dealkalization of Boiler Feed
it is ,not necessary to increase the feed rate Water in Small Power Plants", Proceedings, Am.
Power Conf., Vol. XV, pp. 567-572 (1953)
of the filming amines because of higher car- E. Maher, "Dealkalization System", Industry and
bon dioxide content of the steam. Conse- Power, Vol. 63, pp. 102-103 (Sept. 1952)
18
Hot Lin1e-Hot Ion
Exchange Softening
Use of Process 0 0 0 0 0 0 0 0 0 o 0 o o 0 0 o 0 0 0 0 0 o o o o o 0 116

Chemical Reactions 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 o 0 o o 0 o 116
Equipment Employed 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 o 0 0 o o o 0 117
Applications 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 o 117
Limitations 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 o , 0 o 0 o 0 o 118
116 BETZ HANDBOOK
ITH the development of styrene base quires close attention and frequent control
W resins that can withstand waters of
high pH and high temperature, a two-stage
tests to assure proper operation at all times.
Hot lime softening alone will reduce the
softening process has been developed, em- carbonate and magnesium hardness as well
ploying hot lime softening as the first stage as the total solids of a raw supply. Only lime
followed by hot ion exchange softening as is required when the calcium exceeds the
the second stage. In many cases, this method total alkalinity. On high alkalinity waters, it
of treatment offers definite advantages over may be found necessary to apply calcium
the usual two-stage hot process softeners using chloride or calcium sulfate to reduce the al-
lime and soda ash in the first stage and phos- kalinity to the desired range. The effluent
phate in the second stage. from the lime treater, whether filtered or un-
Although the ion exchange unit will oper- filtered, is passed through a sodium ex-
ate satisfactorily on the effluent from hot changer containing the styrene base resin
process lime-soda softening, it is usual to use which removes the residual calcium and
only lime treatment in the first stage soften- magnesium salts, thereby producing a water
ing, so as to minimize softener effluent alka- with a total hardness in the range of 0.5 to
linity and thereby produce less carbon dioxide 2 ppm.
in the steam.
CHEMICAL REACTIONS
USE OF PROCESS
Hydrated lime is employed to react with the
For waters containing high concentrations of bicarbonate alkalinity of the raw water, pre-
hardness, alkalinity, total solids and silica, it cipitating calcium carbonate as shown by the
was often found necessary to resort to hot following reaction.
process lime-soda softening to reduce these
( 1)
factors to a reasonable range for boiler feed-
Ca(OH)z + Ca(HC03lz = 2CaC03
water purposes, even for relatively low boiler hydrated + calcium = calcium
pressures. At higher boiler pressures, hot lime bicarbonate carbonate
lime-soda softening alone is not completely
satisfactory due to the relatively high softener
+ 2H20
+water
effluent hardness, resulting in boiler sludge To reduce silica, the natural magnesium
accumulations even with proper internal of the raw supply can be precipitated as mag-
treatment. To overcome this difficulty, hot nesium hydroxide which adsorbs silica from
phosphate softening following hot lime-soda solution. Precipitation of magnesium requires
softening has often been employed to reduce a slight excess of lime and this reaction is
the makeup water hardness to approximately shown by the following equation.
2-3 ppm.
(2)
Due to variations in composition common Ca(OHlz + MgCh = Mg(OHlz
to many water supplies, it has often been hydrated + magnesium = magnesium
found difficult to control chemical balances lime chloride hydroxide
in the effluent from a hot lime-soda softener. + CaCh
When such variations occur, boiler water + calcium
balances are upset, resulting in poor internal chloride
conditions. With two-stage hot lime soda-hot If the quantity of natural magnesium in
phosphate softening, any variations in the the raw supply is not sufficient to reduce the
primary softener are immediately reflected in silica to the desired level, it will then be
a poor effluent from the secondary softener. necessary to employ an activated magnesium
From the operating standpoint, either single oxide or dolomitic lime depending on the
stage hot lime-soda softening or two-stage type water to be treated. No soda ash is
hot lime soda-hot phosphate softening re- employed to reduce the noncarbonate hard-
ness since this is automatically removed by posits. With the use of hot phosphate soften-
the ion exchange softener. The reactions oc- ing, difficulties with adherent boiler sludge,
curring in the hot exchange softener follow- principally in the form of magnesium phos-
ing the hot lime softener are identical to those phate are encountered. The use of the hot
outlined in the chapter on sodium zeolite ion exchange process reduces this factor to
softening. a minimum.
The normal hot lime-soda process requires
EQUIPMENT EMPLOYED maintaining an excess of carbonate alkalinity
in the softener effluent to reduce calcium
The equipment employed in this process is hardness to a minimum. The high carbonate
similar to that described in the chapters on alkalinity of the softener effluent increases the
hot lime-soda softening and sodium zeolite carbon dioxide content of the steam, render-
softening with only minor modifications. Fig- ing the resulting condensate more corrosive.
ure 18-1 illustrates a typical hot lime-hot ion In addition, a high blowdown rate often is
exchange softening process. necessary to restrict boiler water alkalinity
within safe limits. With hot lime softening,
APPLICATIONS the carbonate alkalinity of the softener efflu-
ent can be maintained at a minimum and, as
This system of treatment is desirable for high a result, the carbon dioxide content of the
pressure boilers ( 450 psi-1000 psi) since it steam will be lower than results with lime-
produces a water of low total hardness, low soda softening. Blowdown for alkalinity con-
silica and low alkalinity. trol can be maintained at a minimum.
Low total hardness is essential for high Following hot lime-soda softening or hot
pressure operation because of the suspended lime-soda, hot phosphate softening, calcium
solids limitation in preventing adherent de- carbonate or calcium phosphate deposits will
TO BOILER
ION EXCHANGER
SALT
DISSOLVER
SLUDGE SLUDGE
RECIRCIJLAT ION BLOW DOWN
Figure 18-1 • Flow Sheet of Typical Hot Lime-Hot Ion Exchange Process
118 BETZ HANDBOOK
Figure 18-2-Typical Softener Effluent Analyses

Hot Lime-Soda, Hot Phosphate Hot Lime, Hot lon Exchange
lsi Stage 2nd Stage lsi Stage , 2nd Stage
Raw Water (Lime-Soda) (Phosphate) (Lime) (Exchanga Resin)
Total Hardness ........ as CaC03, ppm ... . 150 18 0-2 70 0-2

Calcium ............. as CaC03, ppm ... . 120 15 0-1 67 0-1
Magnesium ........... as CaC03, ppm ... . 30 3 0-1 3 0-1
"P" Alkalinity ........ as CaC0 3 , ppm ... . 0 40 30 10 10
"M" Alkalinity ........ as CaC03, ppm ... . 100 60 50 20 20
Silica ................ as Si02, ppm ..... . 15
Hot Lime-Soda, Hot Lime-

Treatment Requirements Hot Phosphate ion Exchange
lime, lbs/1000 gal .................................................. . 1.0 0.9

Soda Ash, lbs/1000 gal, ........... , ........................•......... 0.8
Phosphate (D.S.P.), lbs/1000 gal ....................................... . 0.18
Salt, lbs/1000 gal. ......................................... , ........ . 2.0
*Can be reduced to approximately 1 ppm with use of magnesium oxide or dolomitic lime.
occur in stage heaters or economizers unless be installed to supplement an existing hot

the pH is reduced by the addition of an process softener, the question of turbidity of
acidic material or dilution with appreciable the influent to the ion exchange softener is
returned condensate. Following hot lime-hot of primary importance. Excessive turbidities
ion exchange softening, there is no danger of in the influent to the exchange softener will
such deposits because the calcium and mag- tend to reduce capacity and form deposits
nesium salts are practically "zero." In fact, within the unit. When such plugging occurs,
with the use of certain types of surface active it becomes necessary to acid wash the ion
agents, the required amount of internal boiler exchanger to remove the deposits.
water treatment may be fed directly to the Even though a properly designed lime
feedline without danger of deposition. The treater will normally produce a low turbidity
internal chemical feed system is thus simpli- water, there are occasions when high tur-
fied since individual high pressure chemical bidities can occur, such as: quick load
feed pumps for each boiler are unnecessary. changes, overload conditions, mechanical
Silica deposits or silica carryover with the failure of softener controls, improper chemi-
steam are important factors to be considered cal balances, and rapid changes in the raw
with high pressure boiler operation. The silica supply. To prevent excessive turbidities from
concentration in the boiler feedwater can be interfering with the hot exchange softening
accurately controlled in the first stage sof- process, filters are normally recommended.
tener and there is no pick-up of silica as the If the exchange units are used as filters,
water passes through the hot ion exchange the flow rate in gallons per square foot per
softener. Figure 18-2 illustrates typical efflu- minute must be low to obtain proper filtra-
ent analyses for a hot lime-soda, hot phos- tion. Lack of filters requires a larger volume
phate softening process and also a hot lime- of the costly resin than for the softening
hot ion exchange softening process. alone since proper softening can be obtained
at flow rates considerably higher than the
LIMITATIONS acceptable filtering rate. It will also be neces-
sary to backwash the hot exchange units at
Whether an entirely new system of hot lime- more frequent intervals than they require re-
hot ion exchange softening is being installed, generation in order to avoid break-through of
or whether a hot ion exchange system is to hardness, due to channeled beds. In the ab-
sence of filters, it can be generally stated that suitable for process use and consequently the
more operating attention is required. While installation of a hot lime softener is generally
some installations operating without filters not economical. In such cases, cold ion ex-
have reported little difficulty, it is the opinion change softening is generally used to com-
of the manufacturers that it is safer to have plete the treatment for boiler makeup water.
filters installed. From the chemical cost viewpoint, it is
When hot exchange softening is being necessary to c~lculate the exact lime and salt
considered to supplement an existing hot requirements and the lime-soda requirements
process softener, there are many items which in each case in order to determine if any
require detailed study. One consideration advantage exists. Also, the cost of internal
should be the turbidity in the effluent from treatment following the various types of ex-
the first stage softener and its effect on the ternal softening systems should be correlated
exchange resin in the event that the installed in order to determine any economic advan-
filters were converted to hot exchange sof- tage. ·
tening units. Like any other system of water treatment,
The hot exchange resins must be back- a hot lime-hot ion exchange system has its
washed during the softening phase to loosen limitations which should be thoroughly
the bed, due to the long softening runs re- studied and evaluated in order to assure that
sulting from the low influent hardness and the system is practical and economical to
high exchange capacity. Higher backwash meet the requirements of the individual
rates are normally required to properly ex- boiler system.
pand the bed, because of the lower density
of the hot backwash water. The quantity of REFERENCES
backwash water and rinse water per unit of
time should be accurately calculated since S. B. Applebaum, "Hot Lime Treatment Followed
by Sodium Zeolite", Proceedings, Engineers Soc. of
this water must come from the hot lime sof- Western Penna., pp. 127-145 (1950)
tener tank and therefore may impose a seri- W. S. Butler, "An Inspection Report After Eighteen
ous overload condition. Months with Hot Lime Zeolite", Proceedings, En-
gineers Society of Western Penna., pp. 1-13 (1951)
In plants operating at lower boiler pres- V. J. Calise, "Hot Zeolite Softening", Power En-
sures with a reasonable quantity of return gineering, Vol. 54, pp. 54-59 (Sept. 1950)
condensate, the installation of a hot lime-hot J. E. Harden and G. R. Hull, "Operating Experi-
ion exchange system represents a consider- ences with a Large Hot Lime-Zeolite System for
1500 PSI Boilers", Proceedings, Am. Power Con£.,
able capital investment that may not be Vol. XIX, pp. 672-684 (1957)
economically justified. B. E. Varon and S. B. Applebaum, "An Application
of Hot Lime Zeolite to Moderate High-Pressure
In many industries, the entire water sup- Boiler Operations", Proceedings, Am. Power Con£.,
ply must be lime treated to produce a water Vol. XIX, pp. 660-671 (1957)
120 BETZ HANDBOOK
19
Oil Ren1oval
Difficulties Due to Oil. . . . . . . . . . . . . . . . . . . . . 121

Oil Removal Methods. . . . . . . . . . . . . . . . . . . . . 122
Oil Removal from Exhaust Steam. . . . . . . . . . 122
Oil Removal from Liquid Condensate. . . . . . . 122
Analysis for Oil. . . . . . . . . . . . . . . . . . . . . . . . . 126
IL removal is a step in the preparation Where exhaust steam is used for process
0 of water used for process or boiler feed
by which this contamination is removed by
work involving heat exchange, it is desirable
that the steam be free of oil in order that
either mechanical or chemical means. oily films may not develop on the heat ex-
Oil which is employed for steam cylinder change surfaces of radiators, calendering
lubrication is usually mineral oil containing rolls, heater coils, etc. The presence of oil
a small percentage of a compounding ma- decreases efficiency by impeding the flow of
terial or fatty oil. The use of a compounded heat, requiring higher back pressures on the
oil provides better lubrication by emulsifying engine exhaust in order to maintain sufficient
with the moisture in the steam and prevent- heat flow.
ing the oil film from being washed off the Compounded oils may cause foaming of
lubricated surfaces. While lubrication is aided the boiler water. Animal fatty oils employed
by the compounding oil, the emulsion so for compounding purposes are saponifiable in
formed makes the removal of oil from the the presence of normal boiler water alkalin-
condensed steam more difficult. ities, thus forming soap and developing a
foaming condition of the boiler water. The
impure steam resulting from such a boiler
DIFFICULTIES DUE TO OIL water may foul superheater tubes and turbine
blading, destroy cylinder lubrication, clog
The presence of oil in boiler feed water usu-
lines and traps, etc.
ally originates from three sources:
Due to the difficulties that may result from
Exhaust or condensed steam from recipro-
the presence of oil in boiler water, the pre-
cating engines, pumps, etc.
vention of oil contamination of the boiler
Steam or water used in certain process
feedwater is recognized as highly important.
heating equipment.
The American Boiler Manufacturers Associa-
Industrial wastes.
tion in its standard guarantee on steam pur-
The removal of oil from boiler feedwater ity formerly specified that the total quantity
is desirable and necessary because of the un- of oil or grease, or substances which are ex-
predictable and objectionable effects it may tractable either by sulfuric ether or by chlo-
have on boiler operation. roform, shall not exceed 7 ppm in the boiler
The presence of even thin films of oil on water when the sample being tested is acidi-
the boiler heating surfaces will seriously re- fied to 1% hydrochloric acid, or 7 ppm in the
tard heat transfer. Oil acts as a heat insula- feed water when the sample being tested is
tor and preveB.ts the rapid transmission of first concentrated at low temperature and
heat from the metal to the water. The re- pressure to the same ppm total solids as the
sultant increase in metal temperature may boiler water.
be sufficient to cause blistering and loss of Because of the importance of avoiding the
boiler tubes and plates, particularly at high introduction of oil into the boiler feedwater
boiler ratings. system, the decision must be made as to
The combination of oil with boiler water whether oily condensate or exhaust steam
sludge may result in the formation of a should be discharged to waste or oil removal
pasty mass. Pellets of greasy sludge may form, equipment installed. Since considerable heat
which may be oval or spherical in appear- will be contained in the oil contaminated
ance, or the entire mass may be deposited on steam or condensate, and since this steam or
the boiler surface. Circulation and heat trans- condensate will usually be of a high degree
fer may be seriously impaired, resulting in of purity with respect to mineral solids and
starvation of boiler tubes or overheating due otherwise desirable for boiler feed purposes,
to the heat insulating effect of the oily the installation of oil removal equipment is
sludge. usually justified.
122 BETZ HANDBOOK
OIL REMOVAL METHODS rator reservoir. Frequent inspection of trap

operation should be made. If the trap is not
Oil removal equipment may be divided into functioning properly the reservoir will fill
two types dependent upon the field of ap- with the oil-water mixture and overflow, in
plication. Baffle or centrifugal type separators which case the oil separator ceases to func-
can be used for the removal of oil from tion. Separator efficiency is dependent on the
steam, while various adsorbents, filters and proper removal of the oil-water mixture from
chemical coagulation can be employed for the reservoir.
the removal of oil from liquid condensate. Separators operating on the centrifugal
OrL REMOVAL FRoM ExHAUST STEAM. Baffle principle are also employed. Receiver and
type separators are used to separate oil from purifier type separators operate on the same
steam. Steam containing oil impinges on a principle as baffle type separators but may
baffle and the condensed moisture and oil are have more than one baffle, may be filled with
thrown out on the baffle and drain down the steel wool, and may be drained from more
vertical ribs into the reservoir of the separa- than one compartment.
tor. From this point the oil-water mixture is Mechanical separators of this type are, of
drained or trapped to waste. The direction course, of no value for the removal of oil
of steam flow is altered horizontally and from liquid condensate, but are of definite
passes around the baffle. Figure 19-1 illus- usefulness in the removal of oil from ex-
trates a unit of this type. haust steam. It is usual for such separators to
For the proper operation of such separa- be installed in the exhaust lines from engines,
tors, it is important that the oil-water mix- particularly if the exhaust steam is to be used
ture be continuously removed from the sepa- for direct process work or in heat exchange
operations. Open type feedwater heaters are
customarily provided with baffle type separa-
tors on the exhaust steam inlet.
Baffle type oil separators cannot be em-
ployed with superheated steam, but must be
used with saturated steam since the action
of the separator depends on condensation
taking place on the baffles. If oil is present in
superheated steam, a de-superheater may be
used to provide saturated steam or feedwater
may be sprayed into the superheated steam
prior to the separator.
OrL REMOVAL FRoM LIQUID CoNDENSATE.
Free Oil. When oil is present as a mechanical
mixture in water, direct filtration may be
employed. The oil bearing water is passed
through various devices, using for a filter
medium some coarse material as loofa
sponges, burlap, terry cloth, coke or excelsior.
Equipment using the above materials as fil-
tering media operates on the principle that
the semi-porous mass, when thoroughly wet
with water, will permit water to pass through
bm will retain the oil. Hot wells may be
Courtesy Cochrane Division-Crane Company filled with sponges or coke designed to re-
Figure 19-1 • Baffie Type Oil Separator tain on its surface the oil present in the
water. Sponges are usually changed fre- ployed do not differ from those used in con-
quently and reinstalled after boiling with a junction with hot or cold process lime-soda
caustic soda or soda ash solution for removal or hot phosphate softening. The optimum
of the adsorbed oil. Coke beds are usually re- filtration rate for an oil filter is two gallons
placed with fresh coke when inspection re- per square foot of filter area per minute, this
veals the bed to be fouled with oil. rate applying to the standard pressure type
Cloth strainers must also be removed fre- filter. As the mat of coagulated oil and ad-
quently for cleaning in an alkaline solution sorbing floc accumulates on the filter bed,
and may be provided in the double unit type the resistance to flow increases until a point
with provision for one unit out of operation is reached where the filter must be taken out
at a time. of service and backwashed. The frequency
As with skimming devices, units of this type of this operation will depend upon the load
are effective only in the removal of free oil imposed upon the filter. The usual practice
from the water. Emulsified oil is too finely is to backwash when the loss in head through
dispersed to be removed by such straining ac- the filter reaches 4 to 5 psi. For backwashing
tion. In some cases, the labor and handling purposes, hot water should be specified since
required is undesirable. Since oil contamina- the use of cold water tends to. "set" the co-
tion from reciprocating engines is usually in agulated oil retained on the filter bed.
the emulsified form, units of this type have a A caustic scouring tank is sometimes pro-
restricted field of application. vided for intermittent hot scouring of the
Emulsified Oil. Several methods are avail- filter for the removal of oil which has not
able which will satisfactorily break the oil- been removed by the normal backwashing
water emulsion, causing the oil particles to operations. The hot caustic solution remains
coalesce and be readily removed by the ad- in contact with the bed for approximately
sorbing medium. The usual systems involve one hour, dissolving and saponifying the oil.
the use of aluminum and iron salts for the The frequency of caustic scouring is depen-
purpose of adsorbing the oil. While ferric dent upon the quantity and the type of oil
sulfate has been used successfully, aluminum contamination of the condensate.
sulfate is the more commonly used material In certain cases where the volume of con-
for this purpose. Caustic soda or soda ash is densate is large and a high degree of oil
fed in addition to the coagulant in order to contamination exists, it is found advanta-
produce the floc of aluminum hydroxide or geous to employ flocculating basins in the oil
ferric hydroxide which coagulates and en- removal process. The chemistry and me-
meshes the oil so that it can be removed by chanics of this operation are fundamentally
filtration. The pH of the treated water is the same as previously described under co-
regulated at a point that insures the most agulation with the flocculation and coagula-
efficient coagulation. tion of the oil being accomplished in basins
In one type of unit, the coagulant and al- equipped for chemical feed and agitation. A
kali are fed to the filter influent in propor- retention time of one hour is usually pro-
tion to the flow of condensate. Conventional vided. If free oil is present in the condensate
pressure type filters are usually employed together with emulsified oil, mechanical
with the filter medium of anthracite coal. skimmers may be used to remove as much oil
Sand or gravel as a filter is undesirable due as possible prior to chemical treatment in the
to the pickup of silica from the filter bed by flocculating basin.
the hot condensate. Calcite filter medium is A variation of the above system involves
also undesirable because some of the filter the application of a certain amount of pre-
bed would dissolve in the condensate, intro- formed floc to the filter bed to form an initial
ducing calcium hardness. adsorbing mat, followed by the continuous
In design and operation the filters em- feed of a small amount of the preformed floc
124 BETZ HANDBOOK
CAUSTIC
TANK
L
,--~------------¥,
------±-- EFFLUENT
METER PACED CONTINUOUS \ __ , - ... WASH AND RINSE

FEED PUMP COLD WATER WASH WASTE
{SURFACE WASH
NOT SHOWN)
PREFORMED FLOC FEED FILTERS

Courtesy Infilco, Incorporated
Figure 19-2 • Preformed Floc Type Oil Removal Plant
Courtesy R. P. Adams Co., Inc.

Figure 19-3 • Diatomaceous Earth Filtration System
to the condensate passing to the filters. In specially designed pumps are available to
this instance, the coagulant and necessary handle such material. At the end of the filter
quantity of alkali are added in the predeter- run, newer designs of the filters permit the
mined quantity to a chemical mixing tank sloughing off of the filter cake by backwash
where the floc is formed and aged. Figure or air pressure with the filter mat dropping
19-2 illustrates a system of this type. into a hopper bottom and discharging to
This method of preparation and injection waste. After thorough rinsing, the unit is
of the floc eliminates the need for proportion- again ready for the precoat application.
ing the chemicals and results in appreciable In the event improper operation causes
savings in initial equipment costs. The filters the septa to become fouled, they are readily
used with this type system are usually the removed for chemical or mechanical clean-
standard pressure type although some manu- ing. The diatomite filters operate with con-
facturers supply units equipped with rotary siderably higher pressure drops than do
surface washers or rake agitators. These standard pressure filters. They may operate
special features are usually employed where with head losses as high as 50 psi before re-
the oil content of the condensate is high and quiring backwashing. This operation is pos-
difficulty in removing the oil and floc mat sible since the rigid support of the septa pre-
from the filter medium is expected. vents channeling and the filtrate will remain
During World War II, the need for im- clear even though the throughput decreases
proved oil removal methods was a primary as the pressure drop increases.
one in the marine field. This need spurred In oil removal practice, it is usual to em-
the application of diatomaceous earth as a ploy a precoat of filter-aid approximating
filter-aid and several types of filters were de- one to two ounces per square foot of filter
signed specifically for the purpose of oil re- area. The body coat will vary depending
moval on shipboard. Development has con- upon the quantity of oil to be removed. It
tinued with units of this type being designed is necessary to employ a body coat whenever
for stationary plant application. Most of the oil is present to any extent in order to main-
latest designs have eliminated former obtain the porosity of the filter cake and pre-
jections and provide easy removal and dis- vent separated oil from clogging the inter-
posal of the filter cake in a backwashing stices. In general, the length of run obtained
operation that does not require removal of on a diatomite filter depends upon the filtra-
the head of the filter or the individual septa tion rate, amount of filter-aid used and the
under ordinary circumstances. amount of foreign matter, such as oil, present
in the water.
The diatomite filters employ a permeable
base or septum to support the filter cake. A suitable grade of filter-aid specially
These septa include various types of filter processed for adsorption of oil is supplied by
cloths and papers, porous stone tubes, wire several manufacturers and adsorbs approxi-
cloth and wire wound tubes. The usual pro- mately one pound of oil for each pound of
cedure involves application of a precoat on material used. An application of this type
the septa with the filter-aid recirculated filter has produced a filtered condensate of
through the unit for several minutes to form practically "zero" oil content with an in-
a porous mat. The unit is then operated to fluent oil content as high as 50 ppm.
waste for several minutes to make certain Operation of these filters requires reason-
the effluent is clear before cutting it into the able care and adherence to operating instruc-
system. The usual practice in oil removal tions. However, their operation is no more
systems is to follow the precoat with a complicated than that of standard filters.
"body" coat or continuous feed of the filter- Chemical coagulation and filtration or use
aid into the condensate passing to the filter. of diatomite adsorption and filtration are the
Even though diatomaceous earth is abrasive, only certain means for the removal of ernul-
126 BETZ HANDBOOK
sified oil from liquid condensate. Naturally, The determination of oil or grease is based
systems of this type are a more costly initial upon refluxing through a trap for volatile
installation and require a greater degree of materials followed by solvent ,,extraction of
supervision than do baffle type separators or the remainder of the sample. Commonly
towel filters. Consequently, adsorption and used solvents are hexane, petroleum ether,
filtration systems are usually installed so as benzene, chloroform or carbon tetrachloride.
to handle only the contaminated condensate. These solvents exert selective extraction of
Oil free condensate such as turbine conden- specific greases and oily constituents. How-
sate is returned directly to the feedwater ever, nonoily materials such as phenolic com-
heater without passing through the oil re- pounds or colloidal sulfur may be selectively
moval system. By limiting the size of the extracted to varying degrees. The method of
equipment to the flow of contaminated con- sampling is of utmost importance. Solvent-
densate, the initial cost may be minimized. washed glass containers should be employed
The adsorption systems described have using a closure with an impervious liner such
certain limitations. Those systems using a as aluminum.
coagulant and alkali require rather rigid pH The definition of grease and oily matter
control and do not function well when the must be based upon the procedures used, the
pH of the condensate is variable, requiring source of contamination, the type of solvent
constant adjustment of the chemical feed. and other factors. The interpretation of the
This defect is true also for the preformed analytical results will vary accordingly. In
floc system since introduction of the pre- view of the multiplicity of the variables en-
formed floc to a condensate which does not countered, the Joint Committee on Uniform-
have a satisfactory pH value will result in ity of Methods of Water Examination has
the solution of the floc and resulting failure concluded that uniformity of methods for
of the system to coagulate and adsorb the oil. the determination of grease and oily matter
Diatomite filters eliminate the necessity for is not practicable on the basis of present
pH control, since no coagulant is employed. technical knowledge.
Since diatomite is a siliceous material, some
silica will be dissolved in the hot condensate. REFERENCES
Usually, this silica pickup is less than 1 ppm.
G. R. Bell, "Removal of Oil from Condensate by
Diatomite Filtration", Proceedings, Engineers Soc.
of Western Penna., pp. 123-137 (1952)
ANALYSIS FOR OIL V. J. Calise, "Removing Oil from Water by Floc-
culation and Filtration", Power Engineering, Vol.
The method to be used for the removal of 58, pp. 86, 87, 110, 138 (Apr. 1954)
oil and grease from water usually is predi- J. J. Maguire, "Oil Removal from Heating Sur-
faces", Southern Power & Industry, Vol. 74, pp.
cated upon a chemical analysis to determine 52, 54, 64, 67 (Aug. 1956)
the amount and type of contamination. Oil "Manual on Industrial Water and Industrial Waste
or grease is of animal, mineral or vegetable Water", 2nd Ed., pp. 400-402, Am. Soc. for Test-
origin and may be present in water as an ing Materials, Phila., Pa. ( 1959)
emulsion, free-floating or in some cases a
J. H. Richards, "Maybe It's Oil that Needs to be
Removed from Boiler Feedwater", Industry and
light petroleum fraction may be in solution. Power, Vol. 47, pp. 62-66 (Dec. 1944)
20
Deaeration
Theory Involved. . . . . . . . . . . . . . . . . . . . . . . . . 128

Removal of Carbon Dioxide and Ammonia. . . . . 130
Equipment. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
Applications and Limitations. . . . . . . . . . . . . . . . 131
Condenser Deaeration. . . . . . . . . . . . . . . . . . . 133
Vacuum Deaeration . .................... 134
128 BETZ HANDBOOK
rium from being established. Therefore, the

M ECHANICAL deaeration, as applied to
boiler feedwater, is the removal of
dissolved gases such as oxygen, carbon di-
reaction will proceed to completion until
either the metal or the gas in ,the system has
oxide and ammonia by the process of raising been entirely exhausted.
the water to the saturation temperature for Ferrous hydroxide is an alkaline compound
the pressure under which the process is con- and its rate of solution depends upon the pH
ducted; employing an equipment design that of the water with which it is in contact. The
secures intimate mixture of the water and the lower the pH of the water, the more rapidly
scrubbing steam; and providing means for ferrous hydroxide goes into solution. Con-
the liberal venting from the system of the densate may contain carbon dioxide ( C0 2 )
gases released. in solution, which, by forming carbonic acid
The necessity for removing these corrosive (H 2 C03 ) , may lower the pH to a range such
gases from boiler feedwater in order to pre- as 5.6-6.5. This low pH is more conducive
vent corrosion is now fully recognized. Dis- to the solution of ferrous hydroxide. Hence,
solved oxygen will cause pitting and corrosion it is highly desirable to rid the system of car-
of the boiler feed lines, closed heaters, econ- bon dioxide so as to increase the pH of the
omizers, boilers, steam and return lines. If water.
carbon dioxide is also present, the two gases Ammonia may be extremely corrosive to
acting simultaneously may be considered copper and copper bearing alloys in the
from 10% to 40% more corrosive than would presence of moisture. Corrosion of valves,
be expected for the same quantity of the fittings, turbine blading, condenser tubes,
two gases acting individually. etc., may take place with sufficient ammonia
It has been determined that oxygen is contamination. In contact with copper, am-
relatively 5 to 10 times more corrosive than monia forms a complex copper-ammonium
equal quantities of carbon dioxide, and that ion [Cu (NH 3 ) 4 ++ ] and solution of copper
temperature is also an important factor. For and copper bearing alloys may be quite
example, a condensate in which is dissolved rapid. A source of ammonia in the system
equal molar concentrations of each gas at may be from breakdown of nitrogenous mat-
194 F is 2.0 to 2.5 times as corrosive as that ter under boiler temperatures and the evo-
same condensate at 140 F. Where corrosion lution of ammonia with the steam.
is caused largely by carbon dioxide, the
structure and composition of the steel are THEORY INVOLVED
factors in determining its life.
Dalton's Law states that the total pressure
When iron or steel is brought into contact
of a mixture of several gases is equal to the
with water, iron goes into solution. A soluble
sum of the pressures which each gas would
compound, ferrous hydroxide is formed as
follows: exert were it alone present in the volume oc-
cupied by the mixture.
Fe++ + 2H 20 = Fe(OH)z + 2H+
Henry's Law states that the concentration
ferrous ion + wa t er _ ferrous +hydrogen
- hydroxide ion of the dissolved gas in the solution is directly
proportional to the partial pressure of that
If dissolved oxygen is present in the sys-
gas in the free space above the liquid, with
tem, it will combine with the ferrous hy- the exception of those gases which chemically
droxide to form the insoluble compound, unite with the solvent. For example, carbon
ferric hydroxide, as shown:
dioxide and ammonia unite chemically with
4Fe(OH)z + 0 2 + 2Hz0 = 4Fe(0Hh water.
ferrous f- + wa t er = ferric Henry's Law may be written:
hyd rOXI'd e - oxygen h yd rOXI'd e
c
The oxygen, by removing the ferrous hy- -=K
droxide from the solution, prevents equilib- p
where K = solubility coefficient (constant for falling streams.

any one temperature) Oxygen similarly can be removed from
c = concentration of dissolved gas in water by passing another gas counter-flow to
solution falling streams of water. Air could not be
p = pressure of gas used for this purpose because of the high
For example, if 0.002 mol of a gas is dis- percentage of o;-;:ygen (21%) in air. Nitrogen
solved in a certain volume of water under could be employed but cost would make the
one atmosphere pressure, 0.004 mol will be process impracticable.
dissolved if the pressure is doubled. The most desirable and practicable method
Deviations from Henry's Law may be ex- of removing dissolved oxygen from boiler
pected when the dissolved gas undergoes a feedwater involves the use of the vapor of
secondary reaction with the solvent, as is the the solvent or steam as a scrubbing gas. This
case when ammonia gas is dissolved in water. choice of a scrubbing gas is logical since
Henry's Law applies only to the dissolved steam is the most economical gas that could
ammonia gas (NH 3 ) but the concentration be employed, and in the process of gas re-
in the solution is determined by analytical moval a thermodynamic advantage is secured
methods which include both the ammonia by heating the water with the steam.
gas and the ammonium hydroxide (NH40H) The mechanism of gas removal from water
formed. through the use of the scrubbing action of the
In accordance with the basic principles vapor (steam) of the solvent (water) in-
contained in Dalton's and Henry's Laws, it is volves two separate mechanisms as follows:
evident that removal of a dissolved gas from 1. Mechanical separation of gas bubbles re-
water can be effected by reducing the partial sults when saturation conditions are achieved
pressure of that gas in the surrounding at- and when the vapor pressure of the liquid
mosphere, regardless of the total pressure on solvent and the solution pressure of dissolved
the system. The simplest manner in which gases appreciably exceed the pressure im-
this can be accomplished is to bubble another posed on the system. Under these conditions,
gas through the water or to spray the water all gases becom,e insoluble. However, satu-
into a countercurrent flow of another gas, ration conditions alone do not assure com-
supplying in each case free venting of the gas plete gas removal. If only saturation condi-
from the system. In this manner, the con- tions were necessary for gas removal, there
centration of the dissolved gas in the sur- would be no need for the complicated tray
rounding atmosphere is reduced by dilution or jet design of feedwater heaters, since an
with the scrubbing gas. The partial pressure open tank, filled with water at 212 F at
of the dissolved gas in the surrounding atmos- atmospheric pressure, would accomplish com-
phere decreases and, in accordance with plete deaeration.
Henry's Law, to establish equilibrium the 2. Diffusion of gas particles to the sur-
concentration of dissolved gas in the water rounding atmosphere (without bubble forma-
decreases in direct proportion with its de- tion) results when the total internal pressure
crease in partial pressure. of the liquid solvent and dissolved gases is
Practical application of this theoretical less than or slightly more than the pressure
principle is obtained in the removal of free imposed on the system. Ammonia removal,
carbon dioxide from water by blowing air because of the extreme solubility of this gas,
countercurrent to falling streams of water as must be accomplished by diffusion rather
in a degasifier. Such a unit, as described than mechanical separation.
under cation exchange by hydrogen zeolite, In the removal of dissolved oxygen from
consists of slatted wooden trays to break up water, 90 to 95% of the initial oxygen con-
the falling water into thin streams and a low tent will leave the water immediately and
pressure blower to force air up through the without difficulty by mechanical separation
130 BETZ HANDBOOK
when the water is heated to the saturation nism of removal of these gases is primarily
temperature for the pressure imposed on the molecular diffusion rather than mechanical
system. The remaining 5 to 10% of the separation.
initial oxygen content must be removed When carbon dioxide is dissolved in water
through molecular diffusion. In ridding the it forms a relatively unstable compound, car-
system of this last 5 to 10% of the initial bonic acid (H 2 C03 ) • This acid ionizes into
oxygen content, it is obvious that the rapidity the bicarbonate radical (HC0 3-) and the
of oxygen removal will be a function of the hydrogen ion (H+). Further ionization of the
heater design and also of the rapidity with bicarbonate radical takes place to form the
which the surrounding atmosphere is changed carbonate radical (CO a= ) and the hydrogen
and gases vented from the system. ion ( H + ) as illustrated.
The practical considerations involved in COz + HzO P HzC03
the removal of dissolved oxygen from boiler carbon dioxide + water P carbonic acid
feed water may be briefly summarized as: HzC03 P H+ +
Heating the water to the boiling tempera- carbonic acid p hydrogen ion +
ture for the pressure under which the Heal-
process is conducted (saturation condi- bicarbonate ion
tions). Hco3- P col= +
Providing a heater design that secures in- bicarbonate ion p carbonate ion +
timate mixing of steam and water. H+
Continuously venting from the system the hydrogen ion
mixture of gases and steam. The hydrogen ion concentration or pH
controls the distribution of carbon dioxide
REMOVAL OF CARBON DIOXIDE AND AM-
among these three forms. It is only in the
MONIA. Unlike oxygen, carbon dioxide and
form of carbonic acid (H2C0 3 ) that a gas
ammonia do not obey Henry's Law because
solution pressure is exerted and that carbon
they unite chemically with water. At elevated
dioxide can be removed by deaeration. Fig-
temperatures their deviation from Henry's
ure 20-1 illustrates the availability of carbon
Law decreases due to decreased chemical
dioxide for removal by deaeration and in-
combination with the solvent. The mecha-
dicates it is at the more acid pH values that
the most efficient removal of carbon dioxide
can be obtained.
When ammonia gas is dissolved in water,
pH !.s. "1.-Haco, OF roiAL co~ it forms an unstable compound, ammonium
100
r-.;;: l". hydroxide (NH 4 0H). Ammonium hydroxide
90
ionizes to form the ammonium ion (NH 4 +)
0 '\ '\. and the hydroxyl ion (OH-) as illustrated.
870 1\ r\ NH3 + HzO P NH40H
.J
g. \ \ ammonia + water p ammonium hydroxide
..
,_ 0
\ NH40H P NH4 + +
0 •0 21S•F
r--' - 77•F
ammonium hydroxide + ammonium ion +
~40
.. \ OH-
_i \ hydroxyl ion
20
\ The hydroxyl ion concentration controls
10 '\ l\.. the distribution of ammonia between these
......... ~ two forms. It is only in the form of ammo-
0
3.0· 4.0 6.0
pH
7.0 B.O 9.0 nium hydroxide (NH4 0H) that a gas solu-
tion pressure is exerted and that ammonia.
Figure 20-1 • Per Cent C02 Available for Removal can be removed by deaeration. Figure 20-2
pH VS. '% NH40H OF TOTAL NH3 The conventional type open heater is not as
100
/~
v efficient in the removal of oxygen and carbon
90
V; dioxide as is a deaerating unit. With deaerat-
1/ v
£ eo ing units, extraJactors of safety are incorpo-
z
;;t 70 rated in the design and a vent condenser is
b
t I employed so that a liberal amount of steam
0
60
77" F I /100°F may be vented from the heater without heat
~ 50
I II waste. Ordinary open heaters are not usually
..
z 40
30
I
designed to provide as intimate mixture of
steam and water and as efficient venting as is
20
I I required for more complete deaeration.
10
/ / The Standards of the Heat Exchange
0
~/ Institute define a deaerating heater as an
6.0 7.0 8.0 9.0 10.0 11.0 12.0
equipment for heating water by means of
pH
steam, designed so as to insure the removal
Figure 20-2 • Per Cent NHa Available for Removal
from the water of the dissolved oxygen in
illustrates the availability of ammonia for re- excess of 0.03 cc per liter.
moval by deaeration and indicates it is at A deaerator is defined as an equipment for
the more alkaline pH values that the most heating water by means of steam and de-
efficient removal of ammonia can be ob- signed so as to insure the removal from the
tained. water of the dissolved oxygen in excess of
EQUIPMENT
In accordance with the Standards of the

Heat Exchange Institute a standard deaerat-
ing unit for operation at atmospheric pressure
and above includes:
Deaerating unit.
Condenser.
Integral float or floats.
Piping between deaerating heater and/or
water inlet regulating valve and the vent
condenser.
Water storage capacity not less than two
minutes.
Weir type metering device with recording
and/ or integrating instrument.
Material of construction is usually cast iron
for working pressures up to 20 psi with rolled
steel plate for higher operating pressures.
It is standard practice to provide the
storage compartment of a deaerating unit
with a manhole, gauge glasses and fittings,
supporting feet and openings for water con-
nections conforming to the deaerating unit Courtesy Cochrane Division-Crane Company
capacity. Figure 20-3 • Tray Type Deaerating Heater
132 BETZ HANDBOOK
Figure 20-4 • Jet Type Deaerating Heater Courtesy Cochrane Division-Crane Company
0.005 cc per liter. "zero" oxygen feed water must be qualified

There are two principal types of deaerat- as to the specific test to be employed in
ing heaters-the tray type and the jet atomiz- measuring the oxygen content.
ing type. The tray type deaerating heater, as CoNDENSER DEAERATION. In central station
illustrated by Figure 20-3 spreads out the operation, deaeration is sometimes secured
water in thin films by means of baffles or by the use of surface condensers with de-
trays so as to insure intimate contact of the aerating hot ~ells. In such installations, a
steam with the water and to provide thorough separate deaerating feedwater heater is not
agitation. In addition, a relatively large employed. The condensate leaving the sur-
amount of steam is vented from the unit, face condenser is guaranteed not to exceed
passing to the vent condenser where the 0.03 cc per liter oxygen content. The con-
steam is condensed and returned to the denser design is such that the condensate
heater while the noncondensable gases are spills into the hot well through a steam at-
discharged to the atmosphere. The flow of mosphere, with liberated gases removed by
steam may be counterflow to the waterflow the air ejector. Because of the danger of air
or parallel flow may be employed. Satis-
factory deaeration is obtained with each de-
sign.
In the jet atomizing deaerating heater, as
illustrated in Figure 20-4, the water is sprayed
into a steam atmosphere and thereby heated
to within 2 to 3 F of the steam temperature.
This operation removes 90 to 95% of the
oxygen. The water is then delivered into a
high velocity steam jet and the water is
broken up, or atomized, by the action of the
steam jet securing complete deaeration in
this operation. Steam flow is countercurrent
to waterflow. As in a tray type heater, the
gases are passed through a vent condenser
and the condensate returned to the heater.
The atomizing deaerator, while subject to
a 0.5 psi pressure drop across the nozzle, is
especially suited to handling turbid waters
and those of high carbonate hardness which
would tend to plug the trays in a tray type
unit. The atomizing deaerator is also suited
for marine use inasmuch as the deaeration
will not be adversely affected by the roll of a
ship which would cause uneven water distri-
bution over the trays of a tray type unit.
In general it can be definitely stated that
the production of a deaerated water of a
"zero" oxygen content is not physically pos-
sible although the dissolved oxygen content
may be reduced to a value below that which
can be detected by chemical tests. Since an
absolute "zero" oxygen in a deaerator efflu- Courtesy Cochrane Division-Crane Company
ent is not possible, any specification for a Figure 20-5 • Vacuum Deaerator
134 BETZ HANDBOOK
leakage into the system, both on start-up and either by steam jet eductors or by vacuum
during normal operation, this system of de- pumps.
aeration is not considered as good a practice These units are usually not designed for
as the use of a deaerating feedwater heater. complete deaeration inasmuch as the size
VAcuuM DEAERATION. It is occasionally de- and cost of the apparatus increase sharply
sirable to deaerate water at temperatures be- for the removal of the last traces of oxygen.
low the atmospheric boiling point of water. Oxygen content has been reduced to 0.25 cc
The object is to render non-corrosive the per liter with the maintenance of a 28.5 inch
water in long pipe lines, for industrial proc- vacuum. Removal of small residual oxygen
ess uses or in certain power plant cycles. The concentrations can be secured economically
deaeration of cold water requires that an with the use of catalyzed sodium sulfite.
equipment design be employed which will
divide the water into small particles to fa- REFERENCES
cilitate gas removal, provide venting of the
· g<;tses from the unit and maintain a vacuum H. A. Grabowski, H. D. Ongman, W. B. Willsey
in the deaerator corresponding to the boiling and W. Nelson, "Problems in Deaeration of Boiler
Feedwater", Combustion, Vol. 26, pp. 43-48 (Mar.
pressure for the water temperature involved. 1955)
The incoming water is introduced usually A. L. Jones, "Deaerating Feedwater Heaters -
Spray Type vs Tray Type", Proceedings, Engineers
through a spray pipe and then falls over Soc. of Western Penna., pp. 41-55 (1949)
staggered wooden trays. The object of this A. E. Kittredge, "The Removal of Dissolved Gases
arrangement is to reduce the water into thin from Boiler Feedwater", Proceedings, Engineers
films and particles, aiding gas release. Water Soc. of Western Penna., pp. 105-114 (1941)
G. T. Skaperdas and H. H. Uhlig, "Corrosion of
vapor and gases released from the water are Steel by Dissolved Carbon Dioxide and Oxygen",
removed through the vacuum maintained, Ind. and Eng. Chern., Vol. 34, pp. 748-754 (1942)
Wet or Solution Type Feeders. . . . . . . . . . . . . . . 136

Shot Feed Systems .. ' .................... 136
Pot Type Feeders .. ................... 136
Water Jet Eductor... . . . . . . . . . . . . . . . . . 137
Automatic Shot Feed. . . . . . . . . . . . . . . . . . 140
Constant-Rate Feeders. . . . . . . . . . . . . . . . . . . 140
Controlled Volume Pump . .............. 140
Drip Feeder. . . . . . . . . . . . . . . . . . . . . . . . . 142
Proportional Feed Systems . ............... 142
Pot Type Proportional Feeds. . . . . . . . . . . . 142
Positive Displacement Pump .... ......... 142
Electric-Motor-Driven Pumps . .......... 142
Meter and Timer Controlled Pumps .. ..... 143
21 Thymotrol . ......................... 143
Chelllical Feed Systellls

Pneumatic Proportioning Control. . . . . . . . . 144
Hydraulic Feeders. . . . . . . . . . . . . . . . . . . . 144
Decanting Feeders . ................... 144
Double-Orifice Feeders. . . . . . . . . . . . . . . . . 146
Dry Type Feeders. . . . . . . . . . . . . . . . . . . . . . . . 148
Volumetric Feeders .. .................... 148
Gravimetric Feeders. . . . . . . . . . . . . . . . . . . . . 148
Automated Feeding Equipment . .............. 149
Continuous Monitoring ................... 149
Boiler Systems. . . . . . . . . . . . . . . . . . . . . . . . 149
Recirculating Cooling Systems. . . . . . . . . . . . 149
Control Mode...................... . . . . 149
Sampling . ............................. 150
136 BETZ HANDBOOK
T HE proper introduction of chemicals is a

very important phase of the treatment of
water and sewage. A predetermined concen-
water passes through stage heaters or econo-
mizers, etc. For example, the feed of sodium
sulfite for oxygen elimination should be con-
tration of the chemicals must be established tinuous to the storage section of the heater or
and maintained to secure the desired reac- the suction side of the boiler feed pump for
tions. Thus, accurate and satisfactory treat- most effective results. A suitable chemical
ment application has required the develop- pump or similar device is required for the
ment of many types of chemical feeders. continuous feed of the sulfite solution. On the
Chemical feeding equipment can be di- other hand, the phosphates normally used for
vided into two general classifications: dry boiler scale prevention may be shot fed
feeders and wet or solution feeders. Either type through the feedlines to the boilers, provided
may be further classified on the basis of method stage heaters or economizers are not used and
of feed, that is, constant rate or proportional. the "number of boilers is restricted to two or
Dry feeders are manufactured in two designs, three.
volumetric and gravimetric. Solution feeders SHoT FEED SYsTEMS. These systems are sub-
may be either of the gravity or pressure type divided into pot type feeders, eductor sys-
and under some circumstances are used as a tems and automatic shot feed arrangements,
"shot feed" rather than constant rate or pro- in which the object of the feeding device is to
portional. rapidly introduce a unit charge of chemical
Dry chemical feeders are usually applied in treatment. Usually these systems are em-
external treatment where large quantities of ployed to avoid feedline deposits that might
chemicals must be handled. The application result from continuous chemical feed.
of this equipment is common to sewage and Pot Type Feeders. Probably the oldest meth-
trade waste treatment plants as well as water od used for the introduction of chemicals to
works and large industrial water conditioning boilers is the pot type shot feeder which
plants. In these installations the demand for consists of a pressure tank and fittings as
chemicals such as alum, lime and soda ash is shown in Figure 21-1. The chemical charge is
frequently quite large. prepared by dissolving the chemicals in a
The solution type of feeder provides greater bucket and then filling the pressure tank with
convenience and accuracy when smaller the solution. It is necessary, of course, to drain
quantities of chemicals are to be employed. water from the tank and open the air vent
Also, some types of chemicals can not be prior to pouring the chemical solution into
handled satisfactorily with a dry feeder. Solu- the tank. The illustration shows the feeder
tion feeders are used exclusively for the addi- connected across the boiler feed pump. Feed-
tion of secondary or internal treatment to water from the pressure side of the pump
boiler feedwater systems for the prevention of forces the chemical solution into the suction
scale, corrosion and carryover. The primary side of the pump. Within a few minutes, the
or external treatment of boiler makeup solution will be washed out of the pressure
water, such as softening, also usually employs tank and will be injected into the boiler
the solution type of feeder since the quantity shortly thereafter.
of chemicals required is relatively small. A system of this type is usually satisfactorily
applied when the number of "shots" per day
WET OR SOLUTION TYPE FEEDERS does not exceed three or four. If more fre-
quent "shots" are required, it is possible to
The type of boiler chemical feed system that install a large mixing tank above the pressure
should be employed in any individual instance tank, similar to the tank shown in Figure
depends primarily upon the amount and type 21-2, thus eliminating the inconvenience of
of chemicals and for what purpose they are dissolving chemicals each time an addition of
being used; type of boilers; whether the feed- treatment is necessary.
being taken from the pressure side of the

boiler feed pump, the outlet discharging with-
in the boiler drum through a separate chemi-
cal feedline. With the difference in pressure
between feed pump discharge and boiler
drum, this type of installation is quite work-
able. By this means, the necessary treatment
for each boiler can be injected at will. The
system, while practical, is rather cumber-
some in that an individual charge must be
prepared for each boiler.
As is indicated by the foregoing discussion,
shot feeders are not suitable for the continu-
ous feed of chemicals such as sodium sulfite
(for oxygen removal), caustic soda or soda
ash (for pH adjustment) or any of the surface
active organic and inorganic chemicals (for
CIRCULATING
prevention of deposits in distribution lines) .
The shot feed of precipitating chemicals
through equipment such as economizers or
PUMP DISCHARG£---._ stage heaters is not recommended, inasmuch
as the water flow through units of this nature
Figure 21-1 • Pot Type Shot Feeder is sufficiently slow to permit premature re-
action of the treatment with the incrusting
salts, forming deposits in heat transfer equip-
Pot type feeders are frequently installed, in ment. This condition is also accelerated by
the case of small boiler plants, on the suction the increase in temperature associated with
side of return condensate pumps or in a by- the operation of units of this type.
pass arrangement on the raw water line, Water Jet Eductor. A second method of shot
where raw water pressure forces the chemical feed utilizes a water jet eductor, installed as
charge into the boiler upon by-passing the shown in Figure 21-3. The eductor works on
water through the feeder. the same principle as an injector, using a
Shot feeders are convenient and economical stream of water under pressure instead of
for injecting precipitating chemicals such as steam. Water from the pressure side of the
phosphates, caustic soda and soda ash to boil- boiler feed pump passes through a pressure
ers, thus reducing the hazard of feedline de- nozzle within the eductor body, producing a
posits caused by premature reaction of these high velocity jet. The high velocity jet creates
chemicals with the incrusting salts present in a vacuum, drawing in the treating solution,
the feedwater. If the chemical treatment is and injecting it into the suction side of the
added frequently, the treatment will usually boiler feed pump. The eductor of the size
average itself among the boilers on line. How- ordinarily used has a capacity of approxi-
ever, the number of boilers treated in this mately 6.5 gpm.
manner generally should not exceed three, Although the eductor is of advantage in
otherwise it will be extremely difficult to that any amount of treatment may be in-
maintain consistent chemical balances in all jected, as compared to the restricted quantity
boilers. of the "pot" type feeder, the disadvantages of
Where a large number of boilers are in use, shot feed are inherent in this system as well.
it is relatively common practice to install a In addition, a pressure differential of at least
feeder for each boiler, the inlet to the feeder 4 to 1 between the pressure and suction sides
138 BETZ HANDBOOK
FILLING LINE
(preferably OPERATION:-PREPARE 24 HR SUPPLY OF CHEMICAL
condensate) SOLUTION IN LARGER TANK. WITH ,VALVES NO. 2 AND
NO. 3 CLOSED AND WITH VENT OPEN, FILL SMALLER
TANK THROUGH VALVE NO. 1.
REMOVABLE COVER
CLOSE VENT AND VALVE NO. 1 OPEN VALVE NO. 2,
REMOVABLE THEN NO. 3. AFTER SEVERAL MINUTES, CLOSE VALVE
Va" SCREEN NO. 3, THEN VALVE NO. 2. OPEN VENT AND DRAIN
SMALLER TANK.
GRADUATED GAUGE
GLASS--.t
APPROXIMATELY 3"
ABOVE BOTTOM
VENT
\DRAIN
SUCTION~ DISCHARGE
Figure 21-2 • Pot Type Shot Feed System Employing Treatment Storage Tank
of the pump must be available for satisfactory do not prove suitable for continued use, since
operation of the eductor. Also, it is not pos- efficiency greatly depends upon proper align-
sible to introduce the eductor discharge direct ment of internal parts and freedom from de-
to the boiler drum, in the average plant, be- posits. Chemical treatment may attack the
cause of insufficient pressure differential. metal used in construction of the internal parts
Steam injectors are sometimes used in the and contact of high-temperature steam with
same manner as eductors. Usually, injectors treating solution frequently causes deposits.
- FILLING LINE (preferably condensate)
REMOVABLE IRON Vs" SCREEN
GRADUATED
GAUGE GLASS
DISCHARGE LINE TO BE - - GLOBE VALVE (#2)

APPROX 3" ABOVE BOTTOM.
DRUM SHOULD BE APPROX ONE FOOT ABOVE

EDUCTOR TO PROVIDE POSITIVE HEAD. PRESSURE
DIFFERENTIAL ACROSS EDUCTOR SHOULD BE AT
LEAST 4 : 1
THIS DISCHARGE LINE OF
-
1" X 1%"
THE EDUCTOR SHOULD BE COUPLING
AS SHORT AS POSSIBLE %"PIPE
CHECK VALVE EDUCTOR
llfz" PIPE
SUCTION - DISCHARGE
TO START FLOW OF CHEMICAL SOLUTION FROM DRUM TO STOP FLOW OF CHEMICAL SOLUTION FROM DRUM
TO BOILER FEEDWATER SUPPLY TO BOILER FEEDWATER SUPPLY
FIRST: OPEN VALVE NO. 1 FIRST: CLOSE VALVE NO. 2
SECOND: OPEN VALVE NO. 2 SECOND: CLOSE VALVE NO. 3
LAST: OPEN VALVE NO. 3 LAST: CLOSE VALVE NO. 1
Figure 21-3 • Shot Feed System Employing Eductor

140 BETZ HANDBOOK
Automatic Shot Feed. A small gear pump,

timer controlled, is frequently employed for
automatically shot feeding treatment to the
suction side of the boiler feed pump. The
timer is regulated to operate the pump for a
variable length of time up to a maximum of
ten minutes, once each hour. Water charac-
teristics would determine whether more fre-
quent operation would be possible and
whether a "shot" of as long as ten minutes
duration would be permissible.
This automatic shot feed method permits
charging of the chemical tank once per day.
Due to the frequent additions to the boilers, a
··larger number of boilers can be treated more
efficiently than in the case of the pot and
eductor feeders.
Gear pumps of this nature usually operate
most efficiently with discharge pressures of
100 psi or less.
CoNSTANT-RATE FEEDERS. A constant-rate
feeder is one which will continuously add a
treating solution at a fixed rate for any given
control setting. The rate of feed is adjustable
by manual control but is not in proportion to
changing water flow. Controlled volume
pumps and drip feeders are the most common
units falling under this general classification.
Controlled Volume Pump. The most common
example of this type of feed employs a motor-
driven chemical feed pump to inject a specific
quantity of chemicals into a system at ·a given
time.
A chemical feed system of this nature is
used most frequently in the continuous Figure 21-4 • Package Chemical Feeding System
feed of sodium sulfite to feedwater heaters or
feedlines and for discharge of precipitating feeding system. Figure 21-5 shows the recom-
chemicals such as phosphates within the boiler mended piping for any applicable positive dis-
drum against boiler pressure. When the rate placement pump with separate chemical tank.
of flow is relatively constant, a system of this It is necessary to supply one pump or one
nature can be used to introduce treatment to pumping cylinder for each point of applica-
once-through cooling and water distribution tion of treatment. The capacity of the pumps
systems. This system is also used quite fre- used is generally in the range of 3.0-10.0
quently to insure introduction of the proper gph per cylinder. It is impossible to throttle
amount of chemicals to a circulating cooling the discharge from one cylinder satisfactorily
water system. so that several points will receive an equal
Figure 21-4 illustrates a pump of suitable amount of treatment. The discharge rate is
type, shown as a part of a complete packaged too low to permit satisfactory division of flow,
CONDENSATE 1/4" RELIEF VALVE

LINE SET 50 LBS ABOVE
DISCHARGE PRESSURE
CHEMICAL PUMPING
TANK CHAMBER
1/2" LINE TO
BOILER DRUM
BLOW BACK
VALVE
PUMP SUCTION TAP - 3/8"

PUMP DISCHARGE TAP -1/4"
RELIEF VALVE
1/2" TO 1/4"
REDUCER
1/2" TO 3/8" 1/2" LINE TO

REDUCER BOILER DRUM
CHEMICAL
TANK
Figure 21-5 • Piping Diagram for Chemical Pump Installation
even when using various types of spt;cially trol tests are conducted with sufficient fre-
designed flow control valves. quency, so as to maintain chemical balances
Employment of this system of feed readily in the system within prescribed limits. While
approaches proportional feed where the there will be periodic variations, this type of
chemicals are injected into a large body of system will maintain chemical balances within
water, such as contained in a boiler or cir- limits, on the average. Usually, a suitable
culating cooling system. It is common practice factor of safety is incorporated when pre-
to maintain certain chemical residuals in the scribed limits are established by the water con-
water being treated. In the average plant con- sultant. During the course of operation, there-
142 BETZ HANDBOOK
fore, sufficient treatment is present at all times The degree of proportionality is question-
and adjustments are readily made if the con- able at all times, for there is little control over
trol tests so indicate. These adjustments in- the solution rate of the briquettes or of the
volve changing the length of stroke on the chemicals incorporated in them. This system,
pump or changing the strength of chemical while it is classified as proportional, is con-
solution being pumped. sidered only roughly so and cannot be applied
Drip Feeder. A rather crude type of constant- where appreciable accuracy of feed is re-
rate feeder, described as a drip feeder, con- quired.
sists of a suitable tank with bottom take-off A further adaptation of the pot type pro-
and control valve for governing solution dis- portional feeder is the type commonly used to
charge. By controlling the flow of treating proportion alum and alkali to water systems
solution by means of a needle valve, an almost for coagulation, or for oil removal in feed-
constant volume of solution is added over a water systems. This system also consists of a
given period of time. As the tank empties, the pot in which potash alum or ammonium alum,
solution head pressure is reduced, thus re- in lump form, or sal soda (crystalline
ducing the rate of feed. Dependent on total sodium carbonate) are placed. An orifice is
solution head pressure above the point of ap- installed in the raw water line, and a line from
plication, this reduction in feeding rate may the upstream side of the orifice leads to the
or may not be of significance. This feeder is a under-side of the tank. From the topside of
gravity type and finds application frequently the tank, a discharge line returns to the water
in the addition of acid and other chemicals to line just after the orifice. The resulting pres-
circulating cooling systems, air washers in air sure differential causes a small stream of water
conditioning systems, and addition of sulfite to flow through the pot feeder, gradually dis-
and like chemicals to some point in a feed- solving the crystalline chemicals contained
water system permitting gravity flow. therein and introducing the treating chem-
While a feeder of this type is readily fabri- icals to the water line.
cated, in actual use it requires constant at- This type of feeder is roughly proportional
tention to make certain the discharge line is and can be used in any installation where ac-
not plugged with some foreign material. curate proportioning is not necessary and
PROPORTIONAL FEED SYSTEMS. There are a where the chemicals to be fed are obtainable
considerable number of systems for achieving in large crystalline or lump form. Dry or
feed of chemicals proportional to the flow of powdered materials cannot be used in a feeder
water, involving a great number of mechan- of this type.
ical and electrical principles in their opera- Positive Displacement Type Pump. Pumps
tion. This discussion will be confined to the are available that can be driven by a ratchet
more commonly used devices and the prin- arm connected to a reciprocating pump or
ciples upon which they are based. engine. These pumps can also be driven by a
Pot Type Proportional Feeds. Probably the belt and pulley for connection to motor driven
simplest type of proportional feeder is the pot pumps which operate intermittently or under
type feeder with. a special opening in one constant head conditions. By so connecting
end to permit charging with chemicals in either type of pump, feed of treatment is ac-
briquette or lump form. A portion of the complished in proportion to the flow of water
water to be treated is allowed to pass through through the main pump. Such a system is
the tank, gradually dissolving the chemicals. quite suitable for the continuous and propor-
Feeders of this type are used for continuous tional feed of chemicals to boiler feedlines,
and proportional feed of treating chemicals once-through cooling systems, and the like.
to boiler feedlines for protection of feedlines, Electric-Motor-Driven Pumps. An electric-
to hot and cold water distribution systems and motor-driven positive-displacement pump can
various types of cooling systems. be used to provide proportional feed of chem-
icals to a well supply under certain conditions. ing upon the time cycle, the three-way valve
The chemical pump can be controlled by the permits chemical solution or condensate to
same electrical circuit that operates the well flow to the suction of the chemical pump.
pump, so that the chemical pump operates It is intended that this system prevent
only when the well pump is operating. Where plugging of the chemical feedlines by flushing
several well pumps are used, the system may periodically with condensate. However, con-
not be applicable. A system of this nature, siderable maintenance is usually involved with
when adaptable, provides accurate propor- operation of three-way valves. If the chemical
tioning of chemicals for treatment of the well solution is made up originally with condensate
water for prevention of corrosion, scale for- and the solubility of the chemicals used is not
mation and to retain iron in solution. exceeded, no difficulties will be experienced
Meter and Timer Controlled Pumps. This with plugging of chemical feedlines. In addi-
system utilizes a water or steam meter tion, the chemical feedline must be properly
equipped with an electrical contacting device located within the boiler drum, in order to
that makes contact after passage of a specific further prevent the possibility of plugging of
amount of steam or water in the lines. Upon the line.
each contact, a stop-cycle timer is started, Thymotrol. An electronic system employing
which operates a complete cycle. The timer the Thymotrol, as developed by the General
is adjusted to permit the operation of the Electric Co., has been adapted to the regula-
motor-driven pump a pre-set number of min~ tion of positive-displacement type pumps. Al-
utes for each contact. The time cycle is readily ternating current available at the plant is con-
adjustable to provide flexibility. For the appli- verted to direct current, full wave rectifica-
cation of chemical treatment to boiler water, tion being employed. A potentiometer is
the control system is applied to steam, feed- mounted in the water meter and governs the
water or makeup water meters. The system speed of the direct-current motor operating
is usually established so that the pump will the pump. The setting of the potentiometer is
operate approximately 75 per cent of the total immediately changed with any change in rate
time, allowing ample capacity for emergencies of water flow. Since the speed of a pump is
and unusual demands on the system. constant for any potentiometer setting, re-
This type of system is not suitable for pro- gardless of load, and since the quantity of
portional feed of treatment to once-through liquid pumped by any positive-displacement
systems unless the treatment is injected ahead pump is proportional to the speed of the
of a large storage tank where adequate mixing pump, accurate proportioning of chemical
can be accomplished. Also, for proportioning feed to the water flow is achieved with imme-
phosphates and other treatment chemicals di- diate compensation for changes in rate of flow.
rect to the boiler drums, it is necessary to This system has been applied in a number
supply a meter, timer, and pump for each dis- of instances requiring considerable accuracy,
charge point; this makes such an installation such as proportioning of lime and soda ash in
rather expensive. If makeup water character- the lime and soda ash softening process. The
istics are changeable or if there is a wide vari- system can be applied quite readily in the
ation in amounts of return condensate, boiler treatment of large quantities of water, such as
water balances may prove to be quite variable, once-through cooling supplies that vary in
even though treatment is added proportion- flow, and where a suitable mixing tank or
ally. reservoir is not available. Typical examples of
Another variation of this type of system application of Thymotrol proportioning con-
permits the motor-driven pump to operate sist of the addition of acids or alkalies in the
continuously. A meter controls the timer as adjustment of the pH of a treated water,
before, but in this instance the timer controls and the addition of stabilizing agents for
the functioning of a three-way valve. Depend- prevention of scale in heat exchange units
144 BETZ HANDBOOK
and similar equipment. and motor-driven agitator for handling slur-

Pneumatic Proportioning Control. Standard nes.
meters are available which transmit control- The chemical solution is drawn off either
ling air signals to a stroke adjusting mecha- at a constant rate, or proportionately, by
nism installed as part of a constant speed means of a swing pipe lowered by a cable
positive displacement pump. The adjustable and drum arrangement, which in turn is
stroke can be linked to a flow meter or other controlled by a small fractional-horsepower
unit, such as a pH controller-recorder. When motor geared to very low speed. Figure 21-8
flows in the line are quite variable and it is shows a feed system of this type. This par-
desired to control pH, a flow meter as well as ticular type is especially applicable for the
a pH controller are used to provide a signal to feed of corrosive solutions such as acids, alum
the pump stroke mechanism. Both variables and ferric salts. By the use of rubber tubing,
are transmitted to a ratio controller which the contact of the draw-off system with the
then sends an air signal adjusting the stroke corrosive liquid is limited to the overflow
· in relation to both flow rate and pH. Figure box, which is lowered by the control device.
21-6 shows such a controlled pump, respond- When operating as a constant-rate feeder,
ing to two variables. the decanting arm is lowered at a constant
Hydraulic Type Feeders. Figure 21-7 illus- rate by the control mechanism. This rate is
trates an economical and effective feed system manually adjustable. If proportional feed is
frequently utilized for chlorination and am- desired, a suitable water meter, equipped
moniation of water supplies, pH adjustment, with an electrical contactor, which makes
sterilization of sewage treatment plant effluent contact upon passage of a specific quantity of
and similar applications. The system may be water, is used to control the operation of the
used for the proportional feed of chemicals lowering motor. A predetermined amount of
to water lines for any purpose if the ma- solution is thereby drawn off for each contact
terial used in constructing the reagent head made by the water meter.
of the pump is selected accordingly. The discharge from the decanting arm can
Most of the standard meters may be be allowed to gravitate to the point of treat-
adapted as the control medium. The meter ment application. Frequently, however, the
does not supply the power to operate the discharge is introduced to a pump suction
pump but simply controls the inflow of pres- box. Several pumping arrangements exist,
sure water to the hydraulic cylinder. The depending upon the service requirements. In
design of the pump is such as to permit in- the case of lime and soda ash proportioning,
jection of treating solution into the pressure for example, a centrifugal pump is used
line. The pump is normally actuated by the which operates continuously, pumping the
pressure of the fluid present in the main line decanted solution to the point of discharge
but can be separately actuated by air, steam, with addition of dilution water, as necessary,
oil or water. controlled by a float in the suction box.
The hydraulic system is usually applicable A proportional-feed system of this type is
to lines operating at pressures of 100 psi or sometimes used for adding phosphate and
less. The addition of the pump control mech- other treating chemicals to boilers, boiler
anism does not affect the meter accuracy to feedlines or other water lines, using an inter-
any appreciable degree. mittently-operated piston type pump, con-
Decanting Type Feeders. This type of chemi- trolled by a float switch in the suction box.
cal feed system involves the use of a constant- In some cases, a high-capacity chemical
diameter tank fabricated of material suitable pump is employed, discharging to a common
for the chemicals to be handled. The tank header equipped with flow control valves to
will be flat-bottomed for handling solutions permit division of flow to a number of dis-
or will be equipped with a rounded bottom charge points while using only one pump.
FOXBORO PNEUMATIC FOXBORO M/40 STABILOC FOXBORO DYNALOC

SQUARE ROOT CONVERTER RECEIVER RATIO pH RECORDER
TRANSMITTER CONTROLLER CONTROLLER
FOXBORO ORIFICE PLATE
FLOW TYPE
ELECTRODE
I ASSEMBLY
I
I
I
I
I
I
I
I
I
'--:::::;;;;===~'-=-:I'~'-------===--- ----- _J
SOLUTION TANK
A ... AIR SUPPLY
Courtesy The Foxboro Company
Figure 21-6 • Automatic pH Control with Wide Flow Variations
Courtesy B-I-F Industries, a Division of The New York Air Brake Company
Figure 21-7 • Meter-Controlled Hydraulic Type Proportional Feeder

146 BETZ HANDBOOK
the pressure lines shunts a portion of the

~ Electrical
Contact Head flowing water into the top of a pressure-tight
chemical feed tank, in which the chemical
solution occupies the complete volume. As
flowing water enters the top of the pressure
solution tank, an equal quantity of chemical
solution is released from the tank at a bottom
connection, entering the low pressure tap on
the downstream side of the secondary orifice.
The equipment is so designed that the
amount of chemical solution injected is rea-
sonably proportional to flow.
When recharging the feeder, chemicals
are dissolved in an upper tank and then
drained into the pressure solution tank. The
density of the chemical solution is sufficient
to support a special float installed in a gage
glass. The float will normally sink in water.
Since a definite line of demarcation exists
between the water and chemical solution at
all times, the float will show the amount of
chemical solution present.
Figure 21-8 · Decanting Type Proportional Feeder
Final control over the chemical solution
The decanting type feeder is used most is obtained by adjusting a needle valve
frequently as a proportioning feeder for line, governing the water flow into the pressure
soda ash, coagulants, acids and other chemi- solution tank. The feeder usually must be
cals applied in the external conditioning of recharged at least once daily.
water. A definite advantage is the fact that Double-orifice type feeders are frequently
only one meter and contactor is needed to used for the application of various chemicals
control the operation of several decanting to once-through cooling water and domestic
feed mechanisms. It is not generally applied water systems to condition these waters for
for addition of phosphates and other chemi- the prevention of scale and corrosion. This
cals to the boiler systems due primarily to type of feeder is also employed for propor-
availability of other equipment which is more tioning of alum and soda ash ahead of coagu-
convenient and less expensive. Also, dis- lation and oil removal systems.
charge of phosphates to feedlines is undesir- Use of these feeders is generally restricted
able and proportioning feed of treatment to in boiler feed systems by pressure limitations.
more than one boiler with one pump and one The feeder manufactured by the average
control meter again introduces the objections supplier is usually built to operate at pres-
outlined in the discussion of systems operating sures not greater than 100 psi. Also, the pres-
in an essentially similar manner. sure drop over the primary orifice is sufficient
Double-Orifice Type Feeders. The double- to cause steam binding of the boiler feed
orifice type feeder is a solution feeder which pumps if a sufficient static head does not
normally proportions with reasonable per exist between the feedwater heater and the
cent accuracy. The principle of operation suction of the feed pumps. Application of
involves passage of water through a primary these feeders for continuous feed of phos-
orifice, creating a pressure differential over phates and other precipitating chemicals to
the orifice and a flow in the lines connected boiler feedlines is not desirable because of
to the pressure taps. A secondary orifice in the danger of feedline deposits.
Fig. 21-9-Methods of Feeding for Common Treatment Chemicals
CHEMICAL PROCESS USE TYPE FEEDERS POINTS OF APPLICATION
Alum Coagulation Dry Coagulation and sedimentation systems,

Color removal Pot Type proportioning prior to pressure filters for removal of
Double Orifice suspended matter and oil.
Decanting
Pumps
Sodium Aluminate Coagulation Dry Usually added with soda ash Ia softeners.
Decanting Used to some extent for internal boiler
Pumps water treatment.
Shot Feeders
Coagulant Aids Coagulation Pumps As primary coagulants, to chemical mix-

ing secfion.
As coagulant aid, to flume after primary
chemical mixing or to flocculation zone of
clarifier.
Ferric Salts Coagulation Dry Prior to coagulation and filtration systems.

Color removal Decanting
Oil removal Double Orifice
Pumps
Lime pH adjustment Dry Prior to coagulation systems; to softeners;

(Hydrated) Softening Decanting to treated water lines for adjustment of
Special pumps (slurry) pH.
Soda Ash pH adjustment Pot Type Prior to pressure filters.

(Crystalline) proportioning
Soda Ash pH and alkalinity All types of wet To domestic systems, feedlines, softeners,
(Anhydrous) adjustment and dry feeders coagulation and filtration systems; boilers.
Softening
Caustic Soda pH adjustment All types of solution To softeners; oil removal systems; domes-
Alkalinity adjustment feeders tic water systems; boilers.
Softening
Acid Feed pH adjustment Decanting Treated water lines, prior to degassifiers

(H2S04) (Sulfuric Acid) Reduction of Pumps or deaerating heaters; H3P04 to phos-
(H3P04) (Phosphoric Acid) alkalinity phate softeners (for both softening and
(NaHS04) alkalinity reduction); tower sump or basin.
Surface Active Prevent calcium All types of Treated water lines.

Phosphates carbonate deposits proportional
Eliminate "red water" solution feeders
Orthophosphates Prevent scale in Pumps Added continuously to boiler drums; shot

(Monosodium Phosphate) boilers Shot feeders fed to drums or boiler feedline.
(Disodium Phosphate)
(Trisodium Phosphate)
Chelating Agents Prevent scale in Pumps Added continuously to feedline, on pres-

boilers and sure side of feed pumps.
economizers
Sodium Sulfite Prevent corrosion All types continuous Storage section of deaerating healer;
due to oxygen in solution feeders suction or pressure side of boiler feed
boilers, feedlines, pumps.
economizers
Sodium Nitrate Inhibition of All types of solution Any point in boiler feedlines or direct to
embr'ittlement feeders boilers.
148 BETZ HANDBOOK
CHEMICAL PROCESS USE TYPE FEEDERS POINTS OF APPLICATION
Sodium and Potassium Corrosion inhibitor All types of To brine systems and various circulating
Chromotes solution feeders cooling and hot water systems.
Reactive Colloids Coagulation All types of To boiler feedlines; circulating cooling

(Sodium Mannuronate) Particle absorption solution feeders systems.
Protective Colloids and adsorption
(Starches)
(Tannins)
(Lignins)
Amines and Related Prevention of Shot Feeders Application depends upon material used.
Organic Compounds return line Pumps Some materials may be added Ia boilers
corrosion and volatilize with steam; others are
added to steam line direct, requiring
pumps.
DRY.TYPE FEEDERS tency of the chemical being fed. If the

Dry-type feeders are generally classified as density or weight of the chemicals per cubic
volumetric type and gravimetric type; and foot is subject to any great variation, then
they are usually warranted in the larger-size this same variation will appear in the rate of
plants where the quantities of chemicals that chemical feed. Certain chemicals tend to
must be added to a system exceed approxi- "arch" or "hang" in the feed hopper, re-
mately 10 lb. per hr. quiring agitators or vibrators to be installed
in the hoppers. Care must then be exercised
VoLUMETRIC FEEDERS. In one type of volu- to prevent packing of the chemical, for this
metric feeder, the chemical is discharged would directly affect the accuracy of propor-
from a hopper onto a circular table, which tioning. The normal accuracy is within 5 per
revolves at a slow and constant rate of speed. cent.
One or more multi-section feed knives extend
GRAVIMETRIC FEEDERS. Gravimetric feeders
on the table, deflecting an amount of chemi-
measure the chemical to be fed by weight,
cal dependent upon the position of the knives.
resulting in accuracy of proportioning within
The feeder can be calibrated for use with
any chemical of relatively uniform composi- 1 to 2 per cent. A typical gravimetric feeder
tion and density. Once calibrated, the num- is shown in Figure 21-10. The chemical is
ber of pounds of chemical that will be de- fed to a traveling belt, the amount of chemi-
livered per unit of time, for the various cal admitted being controlled by a gate,
settings of the feed knives on the rotating which in turn is regulated by the total weight
table, will be constant. of the belt assembly and chemical. A constant
Chemicals rernoved from the rotating table weight of chemicals is maintained on the
are dissolved or suspended in slurry form by belt and variation in rate of feed to the
introduction to a dissolving pot agitated by treater is obtained by varying the speed of
a vortex of dissolving water or by a small the belt.
motor-driven agitator. There are several variations of this prin-
Still another type of volumetric feeder em- ciple, but essentially gravimetric feeders are
ploys an oscillating plate below the chemical scales, balanced in such a manner as to as-
hopper. The length of stroke of the plate and sure the delivery of the desired weight of
depth of chemical on the plate are both ad- chemicals to the system. The chemicals dis-
justable in order to obtain varying rates of charged by a gravimetric feeder are placed in
discharge. solution or suspension form, in the same
Accuracy of any volumetric type feeder is manner as for a volumetric feeder.
dependent to a large extent on the consis- Gravimetric feeders are considerably more
expensive than volumetric types, hence are addition which the analysis represents.
applied only to large systems where the ad- Conductivity, which may be related to the
ditional accuracy is required or when chemi- dissolved solids content of a solution, is also
cals are used whose variable density prevents a widely used measurement. This determina-
use of volumetric feeders. tion may be applied for continuous monitor-
ing of the total dissolved solids concentration
AUTOMATED FEEDING EQUIPMENT
of boiler water or for automatic control of
Many installations are now equipped with continuous blowdown. Specific conductance
some type of device which automatically can serve also as the basis for monitoring
controls the operation of feeding equipment. condensate, and to actuate an alarm or to
CoNTINuous MoNITORING. The first step in divert the condensate stream when impuri-
providing complete and automatic control of ties are present.
the treatment program is the installation of Recirculating Cooling Tower Systems. The
continuous indicating and/or recording ana- class of analyzers noted in the case of boiler
lyzers. These analyzers are being applied not systems can be applied for the monitoring
only in monitoring plant influents and ef- and control of treatment programs for cool-
fluents, but also in the monitoring of external ing water systems. The differences are the
and internal water treating systems. Invalu- range of measurement and the parameters
able assistance can be gained by the judicious most critical to cooling systems (e.g. chro-
selection and installation of a continuous mate level, pH and conductivity).
analyzer to monitor critical variables in a Measuring instruments for pH and conduc-
water system. tivity utilize electrodes which are specific for
Analyzers commonly use a sampling tech- the parameters desired. In the area of chro-
nique whereby a discrete sample of the water mate, an electrode has been perfected which
is obtained, reagents are added in proportion does not require the addition of any reagent
to the sample and sufficient time is allowed to provide the measurement. This analyzer
for reactions. The sample-reagent solution is determines the chromate level on the basis of
then discharged into a measuring cell where the natural color imparted to the water by
a photocell determines the concentration of a chromate treatment.
specific parameter. The measurement is in- It appears that more investigation has
dicated and/or recorded, and serves as the taken place in the area of providing a com-
basis for manual adjustments or for remote- pletely automatic system for the control of
automatic regulation. cooling tower treating programs than in any
Continuous monitoring provides an extra other treatment area. Progress is evidenced
measure of safety by providing around-the- by the development of a completely inte-
clock surveillance of the system. Closer regu- grated system of critical parameter measure-
lation of treatment parameters is gained, ment and control, as shown by Figure 21-11.
thereby ensuring proper system protection CoNTROL MoDE. Most treatment control sys-
with minimum expenditure for treatment tems function properly using the simple on-
chemicals. off control mode. The method is suitable due
Boiler Systems. Continuous monitoring and to the large capacity of the water system
control devices are available for the measure- when compared to the relatively small quan-
ment of constituents, such as silica (control tity of chemical addition or blowdown re-
of demineralizer regeneration) , hardness quired by the treatment program. In fact,
(softener regeneration), phosphate (internal the sampling system used to provide a repre-
treatment control) and sulfite (oxygen scav- sentative sample of the water being controlled
enger control) . These parameters are con- is often the most distorting part of the whole
tinuously monitored, and serve as the basis analysis and control system.
for control of the specific unit or treatment Not all treatment systems can be controlled
150 BETZ HANDBOOK
with the on-off mode. There are occasions,

such as feed of sulfite to a deaerator and
treatment to a once-through cooling system,
where proportional control is a requisite.
However, on-off control can be applied quite
successfully on most systems.
SAMPLING. The sample to the analyzer must
be representative of the water in the system
under control. Care and precaution must be
taken in the selection of the sampling point
and in "sample preparation prior to introduc-
tion into the analyzer or to the electrode.
This sample preparation may involve filtering
of the sample, pressure and temperature re-
duction and, in some cases, flow control.
The manufacturer of the analyzer often
provides the sampling system. Regardless of
the supplier, every recommendation by the
manufacturer of the control system regarding
sampling should be followed to ensure ac-
curate and reliable measurement and control.
REFERENCES
R. P. Lowe, "Chemical Feed Systems," Proceed-
ings, Engineers Soc. of Western Penna., pp. 155-
170 (1949)
J. H. Richards, "Chemical Feed Systems in
Modern Water Treatment," Power Generation,
Vol. 52, pp. 62, 63, 122-125 (Aug. 1948) and pp. Figure 21-11 • Modular Automation System
61-63, 118, 120, 124, 126 (Sept. 1948)
R. T. Sheen, "Liquid Chemical Feed Systems,"
Presented at 8th Ann. Water Conf., Engineers
Soc. of Western Penna., (1947)
VARIABLE SPEED DRY FEED SENS-A-GRAM GATE
TRANSMISSION MATERIAL INLET CONTROllER
RATE
INDICATOR
Courtesy B-1-F Industries, a Division of The New York Air Brake Company
FEED
DISCHARGE OPENING
Figure 21-10 • Belt Gravimetric Feeder
Mechanism of Scale Formation .... ........... 152

Effect of Boiler Scale. . . . . . . . . . . . . . . . . . . . . . 153
Mud, Silt, etc. . . . . . . . . . . . . . . . . . . . . . . . . . . . 154
22
Boiler Scale Control
Calcium and Magnesium Salts .. ............. 154
Silica. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
Iron and Copper Oxides. . . . . . . . . . . . . . . . . . . . 156
Oil . .................................. 156
Identification of Scale Deposits . .............. 157
Internal Chemical Treatment. . . . . . . . . . . . . . . . 158
Phosphate Control. . . . . . . . . . . . . . . . . . . . . . 158
Carbonate Control. . . . . . . . . . . . . . . . . . . . . . 160
Organic Agents. . . . . . . . . . . . . . . . . . . . . . . . 160
Chelating Agents . ...................... 163
Chemical Testing. . . . . . . . . . . . . . . . . . . . . . . . . 164
Feeding of Internal Treatment Chemicals. . . . . . . 164
Feeding Methods for Clzelants. . . . . . . . . . . . . . . 165
152 BETZ HANDBOOK
HE formation of scale and sludge de- 22-1 illustrates the more common external
T posits on boiler heating surfaces is the
most serious water problem encountered in
treatment methods that are employed and
the range of residual hardness that usually
steam generation. The object of the majority results.
of the external treatment processes is to re-
move from the boiler feedwater those ob- MECHANISM OF SCALE FORMATION
jectionable substances which will contribute
to scale or deposit formations in the boiler. The primary cause of scale formation is the
Subsidence, coagulation and filtration may fact that the solubilities of the scale-forming
be employed to remove suspended solids from salts decrease with increase in temperature.
the boiler feedwater and so to prevent the Consequently, the higher the temperature
incorporation of mud, silt, etc., in boiler de- (and pressure) of boiler operation, the more
posits. Lime-soda softening, zeolite softening insoluble the incrusting salts become. No
and hot phosphate softening are used to re- method of external chemical treatment oper-
move from the boiler feedwater the incrust- ates at a temperature as high as that of the
ing salts of calcium and magnesium. Iron re- boiler water. Consequently, when the feed-
moval and silica removal processes have as water is elevated to boiler water temperature
their object the prevention of boiler scale and concentrations, the solubility of the
difficulties through the removal of iron and scale-forming salts is exceeded and they crys-
silica, both of which may .cause heat insulat- tallize from solution as scale on the boiler
ing deposits on the boiler surfaces. heating surfaces.
However, no system of external softening, The actual mechanism of scale deposition
regardless of a high degree of efficiency, is in can be viewed as taking place in two distinct
itself adequate protection against boiler scale operations. First, the precipitation from solu-
without the use of supplementary internal tion of a salt such as calcium carbonate
chemical treatment of the boiler water. In (CaC03) and the sedimentation of this al-
all cases, there will remain some small ready formed precipitate on the boiler heat-
amount of incrusting solids present in the ing surfaces where it bakes in place as scale.
boiler feedwater. In addition, it is also pos- Scale formation due to this mechanism un-
sible at time of softener regeneration, for doubtedly does take place particularly in fire-
appreciable hardness to enter the boiler feed- tube boilers and where the boiler feedwater
water system. Incomplete rinsing of a zeolite is high in hardness content. However, it can
bed also will introduce to the boiler feed- be stated that in general only a minor por-
water considerable quantities of scale-forming tion of scale formation develops in this man-
calcium and magnesium chlorides. Figure ner.
The chief mechanism of scale formation is
due to crystallization of scale-forming salts
from a solution locally supersaturated. The
thin viscous film of boiler water immediately
adjacent to the heating surface tends to be-
come more concentrated than the main body
of boiler water. Consequently, the solubility
of the scale-forming salts is first exceeded in
this thin film on the heating surface rather
than in the main body of the boiler water.
Crystallization of scale results directly on the
heating surface even when the solubility of
Figure 22-1 • Common External Processes the scale-forming salts has not been exceeded
qf Boiler Feedwater Preparation in the boiler water proper.
EFFECT OF BOILER SCALE graphically illustrated by Figure 22-2. For

simplification purposes no temperature drops
Boiler scale creates a problem in boiler oper- through gas or water films have been shown.
ation because the different types of scale Section A shows a cross-section of the tube
formed all possess a low degree of heat con- metal with a completely scale-free heating
ductivity. The presence of scale is therefore surface. There is a temperature drop across
equivalent to spreading a thin film of insula- the tube metal 'from the outside metal tem-
tion across the path of heat travel from the perature T 2 to the temperature of the boiler
furnace gases to the boiler water. The pres- water. Section B illustrates this same tube
ence of this heat insulating material will re- after the development of a heat insulating
tard heat transfer and cause a loss in boiler scale layer. In addition to the temperature
efficiency. Stack gas temperatures may in- drop from T 2 to T 1 , there would be an addi-
crease as the boiler absorbs less heat from the tional temperature drop through the scale
furnace gases. layer from T 1 to T 0 • This condition would,
There have been figures published show- of course, result in a lower boiler water tem-
ing large losses in boiler efficiency resulting perature T 0 • However, boiler water temper-
from even thin films of scale on the boiler ature is fixed by the operating pressure and
surfaces. Some of the early estimates of effi- operating conditions require that the same
ciency loss were erroneously high. The me- boiler water temperature T 1 be maintained
chanical structure of the scale, its porosity, and as before the development of the layer of
thickness as well as the design and method scale.
of operation of the boiler are all factors in-
TUIJ~ WALL
fluencing the amount of heat lost. However,
it has been shown that the rate of heat trans-
mission may be reduced as much as 10-12%
METAL T.
by the presence of scale. While an equivalent TEMP~Itltrtllll
loss in heat utilization does not result, it is

possible for a scale of approximately Ys" T. BOIL~R
WATI!R
T. IJOII.Eil
WATirR.
Tl!MP. TI!MA
thickness to cause an over-all loss in boiler
efficiency in the neighborhood of 2.0-3.0%.
During the course of a year an increase of
2.0-3.0% in fuel consumption will amount to Sl!CTION A SEC'YION B ~I!CTIONC
an appreciable monetary loss. Figure 22-2 • Effect of Scale on

Even more important than the effect of Boiler Heating Surfaces
scale in causing heat loss, however, is the
effect of scale in causing overheating of the Section C illustrates the actual condition
boiler metal and consequent tube failures. that develops. Starting at the base boiler
Costly repairs and boiler outage are the re- water temperature of T 1 , the temperature
sult of such a condition. In modern boilers increase through the scale layer is repre-
with high rates of heat transfer, the presence sented by the line from T 1 to T 3 • The further
of even extremely thin deposits of scale will temperature increase through the tube wall
cause a serious elevation in the temperature is represented by the line from T 3 to T 4, the
of tube metal. The scale coating retards the outside metal temperature. It can be noted
flow of heat from the furnace gases into the that the outside metal temperature T 4 is now
boiler water. This heat resistance results in a considerably higher than the temperature T2
rapid rise in metal temperature to the point which was the outside metal temperature
at which failure results. prior to the formation of scale on the tube
The action which takes place in the blister- surfaces.
ing of a tube because of scale formation is If continued scale deposition takes place,
154 BETZ HANDBOOK
increasing the thickness of the heat insula- duction in calcium and magnesium content
ting deposits, further increases will take of the boiler feedwater. Internal chemical
place in the tube metal temperature until the treatment is used to preve~t deposit and
safe maximum temperature of the tube met- scale formation from the residual hardness
al is exceeded. Usually this maximum tem- concentrations remaining in the feedwater,
perature is taken as 900 F. and also to maintain clean boiler heating
With higher boiler pressures and attendant surfaces in those cases where no external
higher boiler water temperatures, there is softening is employed and where all chem-
less temperature difference between the boiler ical treatment for scale prevention must be
water temperature and the failing temper- applied directly to the boiler itself.
ature of the metal. Consequently, tube fail- Perhaps one of the most common sources
ures can result at higher boiler water pres- of scale is the breakdown of calcium bi-
sures with scale of considerably less thickness carbonate to form calcium carbonate under
than would cause failures at lower pressures. the influence of heat. This action is illus-
This condition explains the need for the ex- trated by the following equation:
tremely close control of boiler water condi- Ca(HCOlh + b.. = CaCOl +
tions in high pressure boiler operation. . calcium + heat = calcium +
bicarbonate carbonate
MUD, SILT, ETC. H20 + C02
water + carbon
The use of turbid river waters as makeup to dioxide
boiler systems is undesirable from several The precipitation from solution of calcium
standpoints, among which is the formation of carbonate to form boiler scale readily takes
boiler scale and sludge deposits. Settling of place where the boiler feedwater contains
mud at points of low flow may restrict boiler any appreciable quantity of calcium bicar-
water recirculation and cause water starva- bonate. The action illustrated above may
tion of water wall tubes. Baking of mud also be the cause for the formation of feed-
in place on tubes and sheets will result in line and economizer deposits and deposits on
their overheating and failure. Mud and silt open heater trays.
may be incorporated in scale deposits, add- Calcium sulfate is a more soluble salt than
ing to their volume and heat insulating calcium carbonate and is generally found
effect. only in boilers that have not received any in-
The removal of such suspended solids is ternal chemical treatment or have received
necessary prior to the use of the water for only an inadequate quantity of internal
boiler feedwater purposes. This removal is treatment. The presence of calcium sulfate
accomplished by the processes of subsidence, in a boiler scale can usually be taken as defi-
coagulation and filtration. No internal chem- nite evidence that internal chemical treat-
ical treatment of the boiler water is designed ment has been used only to an inadequate
to overcome difficulties due to suspended extent, possibly due to a flaw in the system
solids in the makeup water. The removal of chemical control. All methods of internal
of these impurities by external treatment chemical treatment rely on precipitating the
processes is required. calcium salts in some form other than cal-
cium sulfate.
CALCIUM AND MAGNESIUM SALTS Figure 22-3 illustrates the solubility of cal-
cium sulfate together with similar data for
The salts of calcium and magnesium are the calcium carbonate. The curves illustrated
most common source of difficulty with boiler show the solubility of the pure materials in
scale. The use of external treatment proc- distilled water solution. The presence of
esses such as lime-soda, zeolite and hot phos- other salts in the solution will affect the
phate softening have as their object the re- solubilities shown and consequently it is
40 salts commonly employed for this purpose.

Magnesium salts present in the boiler feed-
32 1\ water are usually more easily prevented from
:z:
C>
\ forming scale than are the calcium salts.
:::;
;i ~ Normally, the magnesium salts found in-
.. -
:5
24
16 ........... CALCIUM SULPHATE

corporated in boiler scale are magnesium hy-
droxide, magnesium silicate or magnesium
"'.... N phosphate. The latter salt, magnesium phos-
..."'< 8
CALCIUM
1-- ~CARBONATE
.! r- 1--- phate, tends to form a very sticky deposit,
not very hard in nature, but requiring tur-
0 bining for its removal.
360 400 440 480 520 560 600
TEMPERATURE IN DEG. F.
SILICA
Figure 22-3 • Calczum Solubilit)', for Calcium
Su(fate and Calcium Carbonate As mentioned previously, calcium silicate de-
posits may prove quite troublesome in boiler
necessary to employ certain factors of safety operation. However, another source of con-
in interpreting these data. For example, the cern is the tendency for silica to form scales
usual boiler water will contain a variety of of a complex nature, which may or may not
salts in solution, tending to alter the theo- include calcium and magnesium in their
retical solubilities illustrated. Specifically, in structure. Particularly in combination with
the case of calcium sulfate, an increase in the aluminum, silica tends to form complex
quantity of sodium sulfate present will lower scales which have been identified by X-ray
the calcium sulfate solubility. diffraction methods. Formation of such com-
Because of the high solubility of calcium plex deposits usually occurs in boilers oper-
sulfate as shown in Figure 22-3 it is rela- ating at the higher pressures and proper
tively simple to introduce chemical treat- treatment of the boiler water requires the
ment to the boiler water and, with proper maintenance of both silica and aluminum
control, to avoid deposition of calcium sul- concentrations at a minimum. Under such
fate scale. operating conditions, the prevention of com-
The use of sodium carbonate (soda ash) plex silica scales is primarily a problem for
was one of the first materials employed to external treatment. The maintenance of high
prevent calcium sulfate scale formation. As phosphate and alkalinity concentrations in
can be seen from Figure 22-3 calcium car- the boiler water is an aid in control of the
bonate is considerably more insoluble than problem, but only limited silica concentra-
calcium sulfate and accordingly will be pref- tions can be controlled. Blowdown can be
erentially precipitated. However, calcium used to reduce silica concentrations to toler-
carbonate in itself can form a hard adherent able levels.
deposit and is frequently found as a constit- The concern for silica is not always be-
uent of boiler scale. cause silica concentrations are too high. In
Calcium phosphate and calcium silicate some cases, silica can be too low. This situ-
are frequently deposited as boiler scale. Cal- ation occurs primarily with waters that con-
cium silicate may be formed either from the tain high magnesium concentrations and low
combination of the calcium and silicate ions silica values. In these special cases, magne-
naturally present in the boiler feedwater or sium will tend to precipitate as magnesium
as a result of the misguided use of sodium phosphate which produces a sticky sludge.
silicate as internal chemical treatment. Cal- Precipitation in this form is aggravated when
cium phosphate, on the other hand, is in- the feedwater enters the boiler in a region of
variably the result of treatment of the boiler low alkalinity and poor circulation. Under
water by the use of the various phosphate these conditions, increasing the silica content
156 BETZ HANDBOOK
of the feedwater to avoid magnesium phos- line corrosion has eliminated this problem in
phate precipitation will prove helpful. How- several instances.
ever, this practice must be applied within Copper and copper bearing. metals consti-
the normal silica tolerances established for tute a large proportion of the metallic sur-
the boiler operating pressure. Other factors faces exposed in a modern power station.
that must be considered are increased boiler Solution of copper from surface condensers,
water alkalinity, the use of special organic pump impellers, stage heaters and evapo-
sludge conditioning agents and better mixing rators takes place as a result of corrosive
of the feedwater within the boiler. action. Dissolved oxygen, ammonia and car-
bon dioxide exert the major influences in the
IRON AND COPPER OXIDES corrosion of copper. Ammonia forms a sol-
Deposits of iron oxide, metallic copper and uble complex cupric ammonium ion which
copper oxide are frequently found in boilers is broken down on heating and can result in
operating with very pure feedwater. In such deposition of copper oxide in the boiler.
plants, the makeup water requirements may Where conditions are favorable to the depo-
be only 1-2% with evaporators or demineral- sition of metallic copper on the boiler metal,
izers supplying the makeup water needs. De- it is possible for localized pitting to take place
posits in drums and tubes may consist essen- beneath the deposit.
tially of iron oxide, copper oxide and metallic While internal corrective measures such as
copper. The usual sludge forming calcium the use of organic sludge conditioning agents
and magnesium salts may amount to less than may be applied to minimize the problem of
5% of the boiler deposit. iron oxide and copper oxide deposits, the
basic solution to the problem lies in cor-
The usual source of such deposits lies in
recting the corrosive condition.
corrosion external to the boiler. Corrosion of
iron and steel can result in the solution of In some cases, the source of iron oxide
iron by the condensate or feedwater, with its boiler deposits is not the result of corrosion
subsequent precipitation under the higher external to the boiler. Corrosive attack on the
temperature and alkalinity conditions of the boiler metal by high caustic concentrations
boiler water. The usual causes of such cor- or dissolved oxygen may be the cause of this
rosive action are dissolved oxygen and carbon condition.
dioxide. The prevention of the iron oxide
OIL
deposit in the boilers requires elimination of
the corrosive effect of these gases. The source The presence of oil in boiler feedwater is
of dissolved oxygen should be determined particularly undesirable. One of the most
and removed through the use of mechanical obvious difficulties resulting from oil con-
andjor chemical deaeration. If evaporators tamination is carryover of solids with the
are employed, carbon dioxide should be re- steam. However, oil may also contribute
duced to a minimum by suitable pretreatment greatly to the formation of heat insulating
or by changing the point to which evaporator scale deposits and is frequently the cause of
distillate is discharged. The use of ammonia, blistered tubes. Oil may be carbonized to
neutralizing amines and filming amines aid form a hard asphaltic scale or may be ab-
in the control of this problem. sorbed by porous scale already present. Ad-
Iron dissolved in the condensate and thus herence of such deposits to the heat transfer
introduced into the boiler can cause addi- surface may cause local overheating and
tional difficulties besides the deposition of failure of the metal.
iron oxide. Combination of iron with silica Oil should always be externally removed
present in the boiler water can result in the from the boiler feedwater. There is no
formation of iron silicate boiler scales. The method of internal chemical treatment spe-
use of filming amines to control condensate cifically adaptable to overcoming the hazards
presented by oil contamination of the boiler Fig. 22-4

water. While certain methods of internal Crystalline Scale Constituents, Identified
treatment may slightly aggravate or slightly by X-ray Diffraction
minimize the problem caused by oil, there is
NAME FORMULA
no method of internal treatment which can
be relied upon to cope with any appreciable Acmite NazO.Fez03.4SiOz
oil contamination. Since the extent of the Analcite NazO.AizOa.4Si0z.2HzO
Anhydrite CaS04
difficulties caused by oil are more or less un-
Aragonite CaC03
predictable, external removal of oil from the Brucite Mg(OH) 2
boiler feedwater is the only safe policy. Calcite CaC03
Concrinite 4NazO.Ca0.4Aiz03.2COz.9SiOz.3H20
Hematite Fez03
IDENTIFICATION OF SCALE DEPOSITS
Hydroxyapatite Calo(OH)z(P04)6
Magnetite Fe304
In addition to the microscope and gravi- Nose lite 4Naz0.3Aiz03.6Si0z.S04
metric chemical analyses, a useful tool in Pectolite Naz0.4Ca0.6Si0z.H20
identifying boiler scale deposits is the X-ray Quartz Si0 2
diffraction method. Any material possessing Serpentine 3Mg0.2Si0z.2Hz0
a crystalline structure will yield a distinctive Thenardite NazS04
Wollastonite CaSi03
X-ray pattern which will always be the same Xonotlite 5Ca0.5Si0z.Hz0
and which can be compared with the pattern
of standard materials on file. It is thus pos- ture whereas only the hypothetical structure
sible to identify complex chemical com- could be obtained from the usual chemical
pounds and their individual crystalline struc- analysis.
Figure 22-5 • Analysis of Boiler Deposits by X-ray Fluorescence

158 BETZ HANDBOOK
Figure 22-4 illustrates some of the com- whether or not the precipitate will tend to
mon and uncommon constituents of boiler tightly adhere to the boiler heating surfaces
scale which have been identified by X-ray -the chemical characteristics of the precipi-
methods. Some of the salts could readily be tate are relatively unimportant. Thus, from
determined by the conventional chemical a chemical standpoint sodium silicate will
analysis. Others, such as the complex precipitate the calcium and magnesium salts
alumino-silicates, could hardly have been just as well as a phosphate or carbonate.
identified except with the use of X-ray However, it is the physical characteristics of
diffraction methods. The use of X-ray fluo- the precipitate that are of importance, for a
resence also is developing. It is possible that precipitate of calcium silicate will tightly ad-
X-ray fluoresence eventually will prove more here to the heating surfaces.
useful in the identification of boiler deposits
than X-ray diffraction. Today it is no longer necessary to depend
on industrial application for experimentation.
This work is now performed in laboratory
11\JTERNAL CHEMICAL TREATMENT
experimental boilers and carefully studied in
For the internal chemical treatment of boiler each stage of development before application
water many strange and varied materials to the industrial field. One type of laboratory
have been employed ranging from the earlier experimental boiler used in this work em-
use of potato peelings, kerosene, molasses, etc., ploys ring circulation with electrical resistance
as "cure-ails" to the present use of specific heating, using specially wound coils. Elec-
chemicals applied to specific problems. It tronic water level control maintains water
can be accepted as axiomatic that no one level at +0.25 inches. The entire operation
particular material is universally applicable is automatic with automatic feedwater and
and that a "cure-all" is a physical impos- treatment pump control and automatic con-
sibility. Present day internal chemical treat- tinuous blowdown based upon boiler water
ment uses commercial chemicals with known conductivity. Electrical input is measured for
chemical properties to accomplish certain correlation with steam production. Upon
specific objectives. Many of the chemicals conclusion of a test run, the scaling surface
commonly employed are available to all on is removed from the boiler, the scale dis-
the open market; others are proprietary in solved in inhibited acid and analyses then
nature or their use may be subject to patent conducted on the acid solution for the vari-
licensing. Continued research in the field of ous scale constituents. Opportunity is pro-
boiler water conditioning is constantly de- vided for study of crystal structure and ob-
veloping new materials of value and finding servation of scale adherence prior to re-
new uses for well known products. moving scale from the heating surface. Boiler
The reason for the earlier use of "boiler water samples secured during operation are
compounds" with their usually disastrous re- analyzed for correlation with the scale analy-
sults is that the present day highly technical sis data.
science of internal boiler water treatment As previously stated, scale is caused pri-
developed slowly and to a large extent by marily by calcium and magnesium salts,
"trial and error" methods. Too often, the silica and oil. Each is a separate problem
"trial and error" method was applied to and must be dealt with individually.
industrial boiler plants without preliminary PHosPHATE CoNTROL. For the prevention of
laboratory and pilot plant experimentation. calcium and magnesium salts from baking on
It is well to realize that the prevention of boiler evaporative surfaces, the internal treat-
boiler scale cannot be predicted by any basic ment consists of precipitating the calcium
chemical principle. Where a precipitating and magnesium salts as sludge, and main-
type of treatment is employed, it is the taining this sludge in a fluid form to be re-
physical characteristics of the precipitate moved by the boiler blowdown. Calcium is
formed in the boiler water that determines considered to present a more difficult prob-
Figure 22-6 • Research Experimental Boilers
lem than magnesium. This is because the phate to be used depends upon economics
magnesium is readily precipitated by the alka- and upon the alkalinity of the boiler water.
linity of the boiler water to form a sludge of In some instances it is desirable to blend
magnesium hydroxide. two phosphates to obtain the desired alka-
The most common chemicals employed for linity. In others, it may be necessary to add
the precipitation of the soluble calcium salts alkaline materials to the boiler water such
are the sodium phosphates-trisodium phos- as caustic soda or soda ash.
phate, disodium phosphate, sodium meta- Once the calcium and magnesium salts
phosphate and monosodium phosphate. Once have been precipitated as a sludge, it is
any of these phosphates enter the boiler, necessary to maintain this sludge in a fluid
their action is exactly the same. They are all form. Generally, this is accomplished by the
converted into trisodium phosphate and as use of one or more forms of organic material.
such react with the calcium to form tri- After a fluid sludge of calcium and magne-
calcium phosphate, a rather flocculent pre- sium precipitates has been formed, it is
cipitate. In order for this reaction to take necessary to remove this sludge from the
place, sufficient alkalinity must be present in boilers by blowdown. The amount of blow-
the boiler water. At a pH value of approxi- down required depends upon the amount of
mately 9.5 or less the calcium is not pre- sludge formed and the nature of the sludge.
cipitated properly. The choice of the phos- Fluid sludges require less blowdown.
160 BETZ HANDBOOK
CARBONATE CoNTROL. Another type of in- adherent of the inorganic precipitates. How-
ternal treatment for the prevention of scale ever, even these precipitates will form scale
is commonly known as carbonate-organic and sludge on boiler heating surfaces, de-
treatment. This type treatment has definite pendent on many factors such as suspended
uses and limitations. Its principal application solids concentration, boiler pressure, heat
is on the so-called self-purging waters- transfer rate and chemical balance of the
waters in which the alkalinity is approxi- boiler water.
mately equal to or greater than the hard- Organic agents are required to prevent
ness. This type of treatment employs rela- scale and adherent sludge deposits resulting
tively high concentrations of active organic from these inorganic precipitates. It is proper
materials in the boiler water together with internal conditioning to use inorganic agents
the use of sodium carbonate (soda ash) to to precipitate the incrusting solids in the
maintain proper alkalinity concentrations. In most desirable inorganic form. Then the in-
cases of low feedwater hardness, this type organic treatment may be supplemented with
of treatment is generally not as applicable as organic agents to complete the formation of
~ phosphate type of treatment. However, in
a fluid sludge and to prevent the inorganic
a feedwater of high hardness, that is over 60 precipitates from adhering to the boiler heat-
to 70 ppm, and with high alkalinity con- ing surfaces. The organic agents commonly
centrations, carbonate-organic treatment can used in internal treatment of boiler water are
produce better results than phosphate treat- tannins, lignins, starches, and seaweed deriva-
ment, primarily through the elimination of tives.
the heavy sludge condition which usually ac-
companies phosphate treatment of high- Tannins, which are extracted from wood
hardness feedwaters. In some instances a and bark, are organic substances of high
combination of both types of treatment has molecular weight and complex structure. It is
produced better results than either one singly. characteristic of these compounds that they
are benzene derivatives and also are optically
ORGANIC AGENTS. The use of organic agents active. Tannins possess a glucoside structure
for the prevention of scale and sludge de- in that sugar is produced on fermentation or
posits is a necessary part of the internal hydrolysis. While synthetic compounds of
boiler water conditioning for most industrial tannin structure have been prepared, the
plants. Where external feedwater condi- tannins usually employed in boiler feedwater
tioning systems result in practically com- conditioning are of natural origin. Tannins
plete removal of incrusting solids and where may be conveniently grouped into catechol
very close chemical control is exercised, there and pyrogallol types and mixtures of the two.
may be no need for organic treatment. Central Among the catechol tannins are cutch, que-
station operation is an example of where or- bracho, hemlock, larch, and gambier. Pyro-
ganics are required only for special problems. gallo tannins include gallnuts, sumac, myro-
However, in a great majority of industrial balans, chestnut, valonia and dividivi. Mixed
plants, particularly where feedwater hard- tannins include wattle bark, English oak and
ness is high, organics are a definite necessity chestnut oak bark. These different tannins
if optimum conditions are to be maintained. are some of the most common, but many
As mentioned previously, a precipitating others also exist. Commercial products in
type of internal boiler water treatment is de- some cases contain mixtures of different tan-
signed to precipitate the calcium salts as tri- nins.
calcium phosphate and the magnesium salts Lignin is also a complex organic substance
as magnesium hydroxide. Under some cir- and is the chief non-cellulose component of
cumstances it has been found desirable to wood. The small portion of the total lignin
precipitate a portion of the calcium as cal- present which is in a free state and which
cium carbonate and some of the magnesium can be isolated with the use of solvents is
as magnesium silicate. These are the least referred to as "native lignin." Native lignin
contains four methoxyl groups, four hydroxyl some extent a coating action on precipitated
groups and one carbonyl group for each unit inorganic solids, decreasing their tendency
of molecular weight of 840. Commercial both to cohere and adhere to boiler surfaces.
methods of lignin extraction cause some Of prime importance, however, is the effect
chemical modification in structure. One of of organic agents in distorting crystal forma-
the most useful forms in which lignin is com- tion and thus inhibiting scale deposits directly
mercially available is as the lignosulfonate, at the point of formation. Regardless of the
produced by the sulfite digestion process. theory chosen to explain the action of or-
Starch is a high polymer carbohydrate. ganics, the end result with proper organic
The structure is presumed to consist of glu- concentrations in the boiler water is the
cose, joined by glucoside linkages. There is formation of a fluid boiler water sludge,
considerable range in the size of the mole- readily removed by blowdown. With proper
cules even in one variety of starch and the organic treatment of a boiler water there is
chemistry of starch is quite complex. In the also considerably less sludge of any sort re-
natural state, starch is insoluble in water, maining in the boiler upon opening.
but on boiling the individual cell walls burst Figure 22-7 shows the low and high heat
and a portion is dissolved, forming a milky transfer scaling surfaces (left-28,000 Btujsq
solution of starch paste. Commercially, starch· ftjhr and right-180,000 Btufsq ftjhr) from
is prepared from corn, wheat, potatoes and an experimental boiler run without the use
rice. The U. S. Navy standard boiler treat- of an organic sludge conditioning agent. The
ment formulation specifies the use of corn- deposit on the low heat transfer scaler was
starch for sludge conditioning and anti-foam 1.20 gramsjsq ft and on the high heat trans-
action. fer scaler 5.80 grams/sq ft.
Of the seaweed derivatives, the most prom- In contrast, Figure 22-8 shows the effect
inent are dispersions of sodium mannuronate, of an organic sludge conditioner under other-
Irish Moss, sodium alginate and agar-agar. wise identical test conditions. The low heat
'There are over 1100 kinds of seaweed from transfer scaler developed a deposit of 0.03
which a variety of water soluble gums can grams/sq ft and the high heat transfer
be extracted. Laminaria digitata, laminaria scaler a deposit of 0.12 gramsjsq ft.
longicruris and laminaria saccharina are Figure 22-9 illustrates data typical of the
varieties of seaweed widely used for extrac- results secured in experimental boiler studies
tion purposes. The molecular weight of sea- by superimposing various types of organic
weed derivatives employed in boiler feed- agents on phosphate control: While different
water conditioning varies with the degree of feedwaters were employed in the three series
polymerization and has been measured up to of tests covered by this figure, in each case
250,000. Sodium mannuronate dispersions the control run employed conventional phos-
will precipitate with polyvalent positive ions phate internal boiler water treatment. By re-
such as calcium and magnesium, forming a peating the test with all conditions held
flocculent precipitate which will adsorb sus- constant with the exception of the additional
pended inorganic precipitates. organic treatment of the boiler water, data
There are probably several different mech- were secured on the scale reducing properties
anisms by which organic substances function of the organic agents under the test condi-
to hinder boiler scale form~tion. With the tions employed.
sodium mannuronate polymers, a coagulant Figure 22-10 illustrates similar studies on
effect is secured with adsorption of inorganic the reduction of calcium carbonate type scale
precipitates in the organic floc. With tannins where no phosphate treatment was employed
and lignins it is believed that a major part of and the carbonate and hydroxide ions were
their action is in dispersion of precipitates. the only inorganic precipitating agents used.
All of the organic agents probably exert to Imposing organic agents under such condi-
162 BETZ HANDBOOK
Figure 22-7 • Scalers without Organic Treatment Figure 22-8 • Scalers with Organic Treatment
Fig. 22-9
Phosphate Scale Control Studies
Organic Cone. Type Suspended Solids Heat Transfer Operating Scale Formed %Scale
in Boiler Water Organic in Boiler Water Rate Pressure grams/ft 2 Protection
ppm ppm Btu/ft 2/hr psi
0 400 57,000 260 1.89

190 lignin 420 57,000 260 1.06 44.0%
180 Lignin 380 57,000 260 1.66 12.2%
180 Lignin 400 57,000 260 0.20 89.5%
0 250 24,000 320 0.684
150 Tannin 300 24,000 320 0.526 23.1%
150 Tannin 250 24,000 320 0.453 33.7%
150 Tannin 300 24,000 320 0.092 86.6%
0 900 57,000 650 3.50
180 Seaweed Derivative 800 57,000 650 1.98 43.5%
Fig. 22-10
Carbonate Scale Control Studies
Organic Cone. Type Suspended Solids Scale Formed
in Boiler Water in Boiler Water %Scale
ppm Organic ppm grams/ft2 Protection
0 400 4.10
200 Tannin 380 2.95 28.0%
200 Tannin 400 2.13 48.0%
185 Tannin 380 1.39 66.1%
Operating Pressure: 300 psi, Heal Transfer Rate: 57,000 Btu/f12/hr.

tions, which are similar to many types of Proper application of qualified chelating
railroad operation, can result in a marked agents involves optimum levels of suitable
decrease in boiler scale and sludge. While it supplementary dispersing agents and the
can be demonstrated by such experimental maintenance of favorable alkalinity as well
work that a major reduction of scale and as pH conditions. Where employed in a com-
sludge deposits is obtained by superimposing bination program, phosphate must be satis-
the proper type of organic agent on inorganic factorily regulated. With these factors in
control, the information secured on the balance, existing precipitated matter is re-
grams of scale per square foot of heating r110ved in a controlled manner and with a
surface is not the sole criterion for judging controlled degree of rapidity while prevent-
the applicability of a specific organic for in- ing new accumulations. Even certain complex
ternal boiler water treatment. Some organic scales, such as analcite and acmite, can be
agents have shown a fair degree of scale re- prevented or gradually removed with the ap-
duction, but have resulted in the develop- plication of a properly balanced chelant
ment of a thin, tightly adherent scale, high in program.
organic, and very difficult to remove from The applicability of the chelant type of
the heating surface. It is necessary to corre- treatments is generally in the area of low
late with the scale formation, information on feedwater hardnesses. That is, the treatment
the crystal structure of the deposit, its tend- is economical for most systems where the
encies for adherence and a complete chemi- feedwater hardness is 1.0 ppm or less. A full
cal analysis of the scale itself for proper chelant program is seldom economical when
interpretation of the tests. hardness in the feedwater exceeds 2.0 ppm.
CHELATING AGENTS Recent developments indicate that, under
certain conditions of application, chelating
Chelating or complexing agents have been agents provide dispersive powers of a nature
applied for a number of years to remove de- which are additive to the properties of nor-
posits of calcium and magnesium salts from mal organic dispersants. These characteristics
waterside surfaces of boilers. Developments can advantageously supplement standard
over the past few years have shown that these phosphate type treatments, particularly in the
agents can also prevent certain types of scale cases of systems where the feedwater hard-
and sludge accumulations in industrial boilers. ness is at a higher level. This property ex-
The completely new approach to the rather tends the useful range or the applicability of
common problem of preventing objectionable properly selected chelants.
deposits employs the tetrasodium salt of Chelating agents are useful for not only
ethylenediamine tetraacetic acid and the tri- problems of complex scale and sludge on
sodium salt of nitrilo triacetic acid as the waterside surfaces of industrial boilers, but
basic and most common chelating agents. A also for the control of iron and copper de-
number of complexing and sequestering posits in high purity feedwater systems. How-
agents are used in conjunction with these ever, chelating agents are not a cure-all for
basic agents in order to provide a more com- all types of boiler water problems. As an
plete and satisfactorily controlled scale and example, carryover problems are not cor-
sludge prevention program. Application is rected to any measurable degree in the aver-
in a continuous fashion on a long-term basis. age plant when a conversion to chelant
By their chemical nature, chelating agents treatment is made. In practically all instances,
form complex soluble salts with calcium, any carryover condition existing under a
magnesium, iron and copper as well as other phosphate treatment continues unabated with
metallic ions in exchange for sodium. There- a chelant approach. When mechanical or
fore, in the practical application, precipitates chemical characteristics quite apart from the
are avoided by the formation of soluble salts. small amount of suspended solids present
164 BETZ HANDBOOK
cause carryover, the use of chelants does not stances where contamination has been severe,
solve the problem. Field studies, based upon superheater problems due to carbonized oil
the sodium tracer technique, have shown also deposits have continued. It is well to note
that the application of chelating agents does that oil contamination is an undesirable situa-
not contribute to a carryover condition. tion with respect to boiler feedwater charac-
Use of a proper chelant program on a teristics. The best possible solution to an oil
continuous basis provides certain other ad- problem is to eliminate the source of con-
vantages which usually justify a higher cost. tamination or to remove the oil completely
Because of the reduction in suspended solids from the feedwater before introduction to the
and the elimination of precipitates, higher boiler.
total solids in the boiler water and hence CHEMICAL TESTING
lower blowdown rates are possible in many
instances. A further advantage of chelating The use of internal chemical treatment, re-
treatments is the fact that temporary upsets, gardless of the excellence of the materials
which may cause a scaling condition, can be employed, is valueless unless it is closely con-
mitigated with respect to aftereffects. That is, trolled by means of daily chemical tests of
the maintenance of proper treatment balances the boiler water by the plant personnel. De-
following the upset period causes the removal pendent upon the nature of the treatment
of scale. employed and the individual plant condi-
Longer and safer operation between turn- tions, the control of chemical treatment and
arounds is possible with the advance of che- blowdown may be based on tests for hard-
lant treatments. Some states allow longer ness, alkalinity, chloride, sulfite, phosphate,
operating periods between internal inspec- chelant, silica, nitrate, specific conductance
tions when waterside conditions are satis- and total and dissolved solids. Not all of these
factorily controlled, and such extension may tests are necessary in any one plant, but de-
be an incentive to apply a suitable chelant pendent upon the specific problems encoun-
program. Also, since less work is required for tered, several of these tests may be required
cleaning waterside surfaces with an effective for proper control. In addition, special tests
chelant program, less time is needed for a are occasionally necessary.
turnaround. Labor requirements are reduced. FEEDING OF INTERNAL TREATMENT
Of course, close control must be maintained CHEMICALS
in order to fully gain the advantages of a
properly designed chelant program. In order A satisfactory method of feeding chemical
to aid control, the development of accurate correctives is second only in importance to
and adequate testing methods is a continuing the consistent control of chemical balances.
effort in this new field of boiler water treat- However, some feedwater problems dictate
ment. the necessity of highly efficient feeding meth-
As indicated previously, chelants are not ods before adequate chemical control can be
cure-all treatments. The application of a established and maintained.
chelating program does not correct corrosion Obviously for proper results and to insure
problems due to oxygen and/or carbon di- maximum efficiency from any chemical treat-
oxide. Failure due to steam blanketing is not ment, continuous methods should generally
eliminated, and the potential for embrittle- be employed. Many chemicals, such as so-
ment must still be considered. As discussed in dium sulfite and the majority of the organics,
other chapters of this publication, treatments may be fed continuously to the suction of the
of a specific nature are required. boiler feed pump without fear of inviting
In some instances, the reduction of sus- feed line deposits, regulator troubles, etc.
pended solids has greatly reduced the effect On the other hand, phosphate treatment
of oil contamination. However, in isolated in- and in certain instances sodium carbonate
and sodium hydroxide, when applied in a

manner similar to the above, cause a prema-
ture reaction to take place in the boiler feed
line. Precipitates such as tricalcium phosphate
and calcium carbonate which result, readily
interlock on water distribution and regula-
tor surfaces to form adherent scale. These
premature reactions may frequently be ac-
celerated by high feedwater temperatures or
by chemical characteristics such as high feed- Figure 22-11 • Section of Boiler Feedline Containing
water hardness, excessive feedwater alkalin- Phosphate Deposit, Resulting from Continuous Feeding of
ity and high pH values. Phosphate to Cold Process Lime-Soda Sqftened Feedwater
Serious difficulty can be anticipated from
the consistent formation of feed line deposits. ment purposes is by pumping the phosphate
Usually, erratic operation of boiler feedwater solution direct to a boiler drum, utilizing for
regulators is the first symptom, followed by this purpose a small chemical pump. This
thermal loss in economizers or the necessity method of chemical feed is also suitable for
for increased feedwater pressure. Eventually the introduction of most other treatment
a point may be reached where insufficient materials. Sodium sulfite is usually pumped
feedwater can reach the boilers regardless continuously to the suction side of the boiler
of feed pump pressure. Sudden forced shut feed pump or to the storage section of the
down and boiler outage results, with conse- feedwater heater. Organic sludge condition-
quent possible equipment damage or certain ing agents may be fed continuously to the
loss in production schedules. boiler feedwater or direct to a boiler drum.
Chemicals added for alkalinity adjustment,
In the case of phosphates, the tendency
such as caustic soda and soda ash, may in
toward premature precipitation is not con-
some cases be fed continuously to the boiler
fined to the orthophosphates such as triso-
feedwater and in other cases, dependent on
dium, disodium and monosodium phosphate.
feedwater characteristics, must be fed directly
The problem of feed line deposition is also
to the boiler drum to avoid feedline deposits.
prevalent with the use of molecularly dehy-
drated phosphates.
FEEDING METHODS FOR CHELANTS
In their original or unreverted states, the
molecularly dehydrated phosphates will not The basic chelating agents commonly used
precipitate as readily with calcium as the in boiler water treatment have an affinity for
ortho materials. This property is frequently boiler system metals under certain conditions.
utilized, in conjunction with certain surface Copper bearing materials, in particular, are
active organics, to prevent calcium carbonate subject to attack. Therefore, contact with
deposition. However, it should be remem- copper bearing alloys in feedwater stage
bered that such formulations (normally used heaters, control valves and check valves must
in relatively small quantities) are designed to be avoided. Copper cooling coils for sampling
stabilize the natural deposition tendencies of should be changed to a suitable stainless
a water. They are not expected to cope with steel, such as 304 or 316. Monel and similar
the problem of feed line deposition resulting nickel alloys are not satisfactory for this
from the continuous feed of relatively high service.
concentrations of chemical precipitants, such Suitably diluted chelant solution is usually
as sodium carbonate, sodium hydroxide and injected into the boiler feedwater through a
all phosphates. properly designed stainless steel quill. This
The ideal method of adding phosphate to quill is best located on the pressure side of
a boiler water for internal chemical treat- the feedwater pump, and preferably at a
166 BETZ HANDBOOK
point where the temperature is no greater REFERENCES

than the normal feedwater temperature. R. C. Corey, "What About Copper in Boilers,"
Because of their great activity, chelating Combustion, Vol. 16, pp. 43-46. (June 1945)
agents should be fed to a boiler system only J. A. Holmes, "The Development of Organics for
Water Treatment," Proceedings, Midwest Power
in a continuous fashion. This method pro- Conf., Vol. XIII, pp. 238-245 (1951)
vides continuous dilution. In addition, direct J. J. Maguire, "Organic Agents in Boiler Scale and
feed to boiler drums should be avoided, even Sludge Control," Proceedings, Midwest Power
though a suitably designed stainless steel Conf., Vol. XIII, pp. 246-256 (1951)
distribution line may be installed within the J. Van Brunt, "Tube Failures in Water-Tube Boil-
ers," Combustion, Vol. 17, pp. 32-37 (July 1945)
drum. The breaking of the line by chance
could lead to serious and undetected cor- J. R. Metcalf, "Chelant Treatment Programs for
Boiler Water," The Betz Indicator, Betz Laborato-
rosion problems. ries, Inc., Trevose, Pa. (June, 1967)
23
Boiler Corrosion Control
Control of pH . .......................... 168

Control of Dissolved Oxygen . .......... , . . . . 168
Protection of Idle Boilers. . . . . . . . . . . . . . . . . . . 171
Dry Storage. . . . . . . . . . . . . . . . . . . . . . . . . . . 171
Wet Method. . . . . . . . . . . . . . . . . . . . . . . . . . 172
Superheater Storage . ...... ; . . . . . . . . . . . . . 172
168 BETZ HANDBOOK
major consideration in the design of a

A feedwater conditioning system is pro-
tection of boilers and accessory equipment
CONTROL OF DISSOLVED OXYGEN
Dissolved oxygen can be int~oduced into the

against corrosion, both during operation and system not only in the makeup water but also
when the equipment is out of service. Except due to air infiltration of the condensate sys-
in special cases, corrosion in boilers, feed tem. When dissolved oxygen is present in the
lines and economizers is due either to the feedwater, an attack of the feed line, closed
influence of low pH or dissolved oxygen. heaters and economizer can be expected with
the severity of the problem dependent on the
concentration of dissolved oxygen and the
CONTROL OF pH temperature involved. One of the most seri-
ous aspects of oxygen corrosion is that it
The solubility of iron in water increases as generally occurs as pitting so that the attack
the pH decreases below the neutral point. is con-centrated in a small area of the total
The attack on steel by water with an acid metal surface. With this type of corrosion,
pH is of a general nature over the entire failures can occur even though only a rela-
surface with little tendency toward localized tively small portion of the metal has been
corrosion. The obvious method of preventing lost.
such corrosion is neutralization of the acid The influence of temperature on the cor-
with an alkali. Soda ash and caustic soda are rosivity of dissolved oxygen is particularly
commonly employed for this purpose in the important in such equipment as closed heat-
treatment of boiler feedwater. Except in ers and economizers where the water temper-
special cases of corrosion in high pressure ature is increased very rapidly. Under such
boilers, experience has shown that it is de- conditions an additional driving force for
sirable to maintain a minimum boiler water ' reaction is present and for th1s.
the oxidation
pH of approximately 10.5. This pH is suf- reason, even very small quantities of dis--
ficiently high to stifle acidic attack. In the solved oxygen in feedwater can cause severe
proper treatment of boiler water for control corrosion in such equipment.
of scale and sludge deposits, the desired
When oxygen is present in the feedwater
boiler water alkalinity and pH for prevention
entering the boiler, a portion will be flashed
of acidic corrosion is automatically estab-
and will leave the boiler with the steam. The
lished. Therefore, except under unusual op-
erating conditions, pH is not a factor in remainder of the dissolved oxygen can attack
boiler water control. the boiler metal. While the point of attack
will vary with the boiler design and feed-
Trade waste contamination of the makeup water distribution, oxygen pitting is usually
water, sugar contamination of the condensate concentrated adjacent to the water level in
or acidic contamination from acid pickling the feedwater drum.
or plating baths are some of the most com-
The first and most important step in elimi-
mon sources of acid contamination of boiler
feedwater. Acid contamination can also enter nating the corrosive influence of dissol~ed
the system due to improper operation or oxygen is mechanical deaeration of the bmler
control of hydrogen zeolite softeners. feedwater. Efficient deaeration will reduce
the dissolved oxygen content of the boiler
If the makeup water or condensate is sub- feedwater to a very low value. It is advisable
ject to periodic acidic contamination, record- to follow mechanical deaeration by chemical
ing or indicating instruments are desirable deaeration in order to remove the last traces
to maintain a check on the purity of the of dissolved oxygen. Where mechanical de-
makeup or condensate. Such devices can be aeration is not employed, chemical deaera-
used either to operate automatic dump valves tion must be used for the removal of the en-
or to warn the operators to discharge the tire oxygen content of the feedwater.
contaminated water to waste.
Sodium sulfite is the chemical agent most
of the boiler feed pump. While intermittent

application is generally not recommended, it
has been found in some low pressure systems
that adequate protection is provided as long
as the additions of sodium sulfite are made
with sufficient frequency to continuously
maintain the proper residual concentration in
the boiler water.
Testing of the boiler water for sulfite resi-
dual and recording the quantity of sulfite re-
quired serves also as a quick check on heater
deaeration efficiency in those plants where
the oxygen content of the feedwater is not
Figure 23-1 • Solubility of Dissolved Oxygen determined regularly. Any decrease in boiler
water sulfite residual and consequent need
for increased feed of sodium sulfite, is an in-
commonly employed for chemical deaeration dication that heater operation should be
due to its low cost, ease of handling and its checked to ascertain and correct the reason
lack of scale forming properties. The oxygen for increased oxygen content of the boiler
scavenging characteristic of sodium sulfite is feedwater.
illustrated by the following reaction:
The speed of the sulfite-oxygen reaction is
2NazSOl + Oz. = 2NazS04 affected by a number of factors, the most
sodium sulfite + oxygen = sodium svlfate important being temperature. The reaction
The removal of 1.0 ppm dissolved oxygen time decreases with increased temperature.
theoretically requires 7.88 ppm of chemically In general, the reaction speed doubles for
pure sodium sulfite. However, use of a every 10 C increase in temperature. At tem-
technical grade of sodium sulfite combined peratures of 212 F and above the reaction is
with handling and blowdown losses as en- quite rapid. It has also been found that the
countered in actual plant operation usually presence of an excess or overfeed of sodium
requires the feed of approximately 10 lbs of sulfite will increase the reaction rate. Several
sodium sulfite for each pound of oxygen. investigators have shown that the reaction
Requirements will also depend on the con- proceeds most rapidly at pH values in the
centration of excess sulfite maintained in the vicinity of 9.0-10.0.
boiler water.
To assure complete oxygen removal, it is
necessary to maintain a residual concentra-
e:.
tion of sulfite in the boiler water. The resid-
ual required depends on a number of fac- ~
f3 10 0
tors such as the method of feed and the point ::; I/

~ 90
SEA WATER
of application, the dissolved oxygen con- ...i50 •0
f.-f.-
70
centration and the variation in the dissolved [....-
~
DISTILLED WAlE~
60
oxygen concentration of the feedwater. 1/ :;:::;
ill "'
Ill"'
[....-
Continuous feed of sodium sulfite is gener- 40
v
~ I
o
ally required for complete oxygen removal. 3
1/ v
In the majority of plants, the most suitable ~ 2
0
.....
10
INHIBITORY SURFACE WATER
point of application is the storage compart- 4 • • 10
TIME, MINUTES
ment of the deaerating or open heater. In
other plants, sufficient reaction time will be Figure 23-2 • Effect of Di.fferent Waters on Time of
allowed with application to the suction side Oxygen-Sulfite Reaction
170 BETZ HANDBOOK
It has also been shown that some natural Catalyzed sodium sulfite is used in low
waters possess an inhibitory effect on the temperature systems for oxygen removal and
oxygen-sulfite reaction. Figure 23-2 shows the also finds application in boiler systems where
reaction rates at room temperature between the feedwater temperature is low, where me-
oxygen and sodium sulfite for three different chanical deaeration is not complete or where
types of waters. The reaction is fairly rapid it is essential to obtain rapid reaction for
with sea water, complete oxygen removal prevention of pitting in feed lines, closed
being secured in two minutes. The reaction heaters and economizers.
is much slower in distilled water, only 75%
of the oxygen being removed after 10 minutes
contact time. With a surface supply possessing I TI
inhibitory properties, after 10 minutes con- 10.0 I I I I
i 9. or-
t- t-
tact only approximately 3% of the oxygen SODIUM SULFITE -
f 8.0
t-
~
has been removed. o1 t- f---+..-
~ 7.
Research directed toward increasing the ~ 6.
OH I
I -~
~ !1.0
-speed of the oxygen-sulfite reaction has de-
termined that certain materials act as cata-
lysts in speeding this reaction to completion.
I :: ofr"CATALYZED SODIUM SULFITE
0 I
2.0 i
The most suitable catalysts are the heavy I. 0
0
metal cations of two or more valences. Iron, ~ 20
2 3 4 !I 6 7 8 II 10
SECONDS
copper, cobalt, nickel and manganese are TIYE IN MINUTES
among the more effective catalytic aids to Figure 23-3 • Comparison of Sodium Sulfite vs Catalyzed
the oxygen-sulfite reaction. Combinations of Sodium Sulfite on Time of Oxygen Removal
several of these heavy metal cations have
proved most effective in providing a con- As boiler operating pressures have in-
tinuously active influence on the speed of creased, two disadvantages in the use of so-
reaction. dium sulfite as a chemical deaerant have
As a result of research on catalytic aids for become evident. The product of the sodium
oxygen removal, a catalyzed sodium sulfite sulfite-oxygen reaction is sodium sulfate
has been developed. Through the incopora- (Na2S04 ) which increases the dissolved
tion of suitable catalysts and sodium sulfite in solids concentration of the boiler water.
one formulation, a material is available While the increase in dissolved solids in
which will consistently provide practically low or medium pressure boilers is generally
instantaneous oxygen removal, even when the of little consequence, it can be significant in
water possesses natural inhibitory properties. high pressure boilers. Also, at high pressure,
the sulfite in the boiler tends to decompose
Figure 23-3 illustrates test results using
to form acidic gases, sulfur dioxide ( so2)
ordinary commercial sodium sulfite for oxy-
and hydrogen sulfide (H2S) which can con-
gen removal at room temperature with re-
tribute to corrosion in the return system.
sults obtained with catalyzed sodium sulfite.
As can be seen, the original oxygen content Hydrazine is a reducing agent which does
of 9.8 ppm has been reduced by sodium sul- not possess these disadvantages for high pres-
fite after 10 minutes contact to only 6.6 ppm. sure operation. Hydrazine removes dissolved
Little reduction has been made in the cor- oxygen in accordance with the following
rosive nature of the water. With treatment reaction:
by catalyzed sodium sulfite, however, com- N2H4 + 02 = 2H20 + N2
plete oxygen removal has been accomplished hydrazine + oxygen = water + nitrogen
in only 20 seconds contact time. Complete Since the products of this reaction are water
elimination of the corrosivity of the water and nitrogen, no solids are added to the
due to dissolved oxygen has been obtained boiler water. The decomposition products of
in this short contact time. hydrazine are ammonia and nitrogen. The
ammonia is alkaline and therefore will not in the boiler water to remove the normal
attack steel. However, if present in sufficient concentrations of oxygen without danger of
quantity, it can attack copper bearing alloys excessive ammonia liberation.
when oxygen is present. With proper appli- When feedwater is employed in desuper-
cation, the concentration of ammonia in the heating, it is desirable to feed sodium sul-
steam can be controlled so that the danger fite after the point from which the feedwater
of attack of copper bearing alloys will be for desuperheating is taken so as to avoid
minimized. At the same time, the ammonia adding solids to the water. When hydrazine
will neutralize carbon dioxide so that return is used as the chemical deaerant, this precau-
tion is not necessary since the hydrazine will
not add solids to the desuperheated steam.
PROTECTION OF IDLE BOILERS
Unless proper storage procedures are fol-

lowed, severe corrosion may occur in idle
boilers. The method of protecting idle boilers
depends primarily on the length of down-
time. Dry storage is preferred when the
boiler will be out of service for a month or
more while wet storage may be suitable for a
Figure 23-4 • Oxygen Pitting in a shorter period of downtime.
Two Inch Generating Tube
DRY STORAGE. The boiler should be drained,
line corrosion due to carbon dioxide will be thoroughly cleaned and dried completely by
reduced. means of hot air or a small wood fire. Close
Hydrazine is a toxic liquid and therefore attention should be given to complete elimi-
must be handled with the usual care observed nation of moisture from nondrainable super-
in the handling of all toxic materials. Pure heater tubes. A suitable moisture absorbing
hydrazine has a low flash point so a 35% material in a water tight container should be
solution, which has no flash point, is usually placed in the boiler drums or on top of the
employed. Theoretically, 1.0 ppm hydrazine flues in a fire tube boiler. The most widely
is required to react with 1.0 ppm of dissolved used moisture absorbants are quick lime and
oxygen. However, in practical application, silica gel. Silica gel is more efficient in ab-
hydrazine is usually required at approxi- sorbing moisture and can be regenerated by
mately 1.5-2.0 parts per part of oxygen. To heating so that it can be used over and over
avoid excessive ammonia concentrations in again. It is not a caustic substance like lime
the steam, the residual hydrazine concentra- so it can be handled more easily and safely
tion in the boiler water is usually maintained and is therefore generally preferred.
at 0.1 ppm or less. Requirements for quick lime approximate
In some instances, the feed of both sodium 20-30 lbs per 100 boiler horsepower or 7-10
sulfite and hydrazine is advisable in high lbs per 1000 lbs per hour boiler rating. Silica
pressure installations. In these cases, a very gel is recommended at the rate of 4 lbs per
low sulfite excess is carried in the boiler water 100 cubic feet of internal volume.
so that the potential for corrosion from After placing the quick lime or silica gel
acidic gases is minimized but a reserve of de- in the boiler, all openings should be tightly
aerant is available to react with an oxygen closed. The unit should be checked every
surge. In conjunction with the feed of so- two or three months, or as experience dic-
dium sulfite, the hydrazine is fed in a quan- tates, for renewal of the lime or regeneration
tity sufficient to maintain a very low residual of the silica gel.
172 BETZ HANDBOOK
WET METHOD. The boiler should be cleaned Nitrogen or other inert gases may also be
and inspected and then filled to the normal used for storage purposes. A slight positive
water level. If deaerated water is not avail- pressure of the gas is maintained after the
able, dissolved gases should be expelled by boiler has been filled to the operating level
boiling the water for a short time with boiler with deaerated feedwater. This method has
vented to atmosphere. The boiler water been employed in some high pressure utility
alkalinity should be adjusted with caustic stations or similar installations where short
soda to a minimum of 400 ppm. Sufficient periods of standby service are frequently re-
sodium sulfite should also be added to pro- quired.
duce a minimum sulfite residual of 100 ppm. SuPERHEATER STORAGE. In most modern
The quantities of caustic soda and sodium boilers it is not possible to separate the super-
sulfite required will approximate 3.0 and 1.5 heater section from the rest of the boiler.
lbs respectively per 1000 gals of water in the Accordingly, it is necessary to follow the
boiler. After the boiler is cooled and before same storage procedure for the superheater
a vacuum is created, the unit should be filled section as for the other portions of the boiler.
completely and all connections closed. Wet storage of drainable superheaters is rela-
Tests should be conducted on a weekly tively simple while wet storage of nondrain-
basis and additions of the treatment chemi- able superheaters is more complicated. In
cals should be made as necessary to maintain dry storage, care must be taken to remove
the recommended minimum concentrations. all of the moisture from the nondrainable
When treatment additions are required, the superheaters by reheating the superheaters
boiler water should be circulated by means sufficiently to evaporate all the water. This
of an external pump or by lowering the water may be accomplished by means of a small
to the operating level and steaming the boiler fire in the boiler furnace. In some cases it
for a short time. The boiler should then be may be possible to dry the nondrainable
completely flooded as outlined previously. superheaters with hot air diverted from the
The temperature of the boiler should be air heaters of one of the operating boilers.
maintained as low as possible since the cor- Depending on the actual design, there may
rosion rate increases at higher temperatures. be a choice as to whether the dry air is
When the boiler is returned to service, a directed over the external surfaces or used
high rate of blowdown should be maintained internally.
initially so that alkalinity and sulfite will be Since a residue will be left in nondrainable
reduced to the normal operating levels superheater tubes after boiling out, if the
rapidily. superheater has been flooded with water
In some small installations or where weekly containing boiler water salts, it is desirable
testing is not practical, chromate salts can be to employ a method of wet storage which
employed to protect idle boilers against cor- does not involve the use of solid chemicals.
rosion. The concentration maintained should Volatile chemicals or inert gas can be used
be 2000-2500 ppm as sodium chromate. The in the superheater section. The volatile chem-
boiler should be filled completely and closed icals recommended are hydrazine and am-
tightly. To assure good mixing, circulation monia or neutralizing amine. If high purity
of the water with a pump is recommended. water is not available to fill the entire boiler,
Boilers stored in this manner should be blown the superheater tubes can be filled with
down heavily to dissipate the chromate color, condensate or demineralized water from the
before being returned to service. The use of outlet end. The recommended treatment
chromate is not recommended if steam is to concentrations are approximately 100 ppm
be used in cooking, sterilization or in the of hydrazine and sufficient ammonia or
processing of food products since chromate neutralizing amine to elevate the pH to ap-
salts are toxic. proximately 9.0-10.0.
Figure 23-5 • Bottom of U-Bend in Superheater Tube Shol!Jing Pitting and Perforation
REFERENCES
H. D. Ongman, "A Study of Reducing Solutions "Steam: Its Generation and Use", pp. 21-18, 19,
at Steam Boiler Temperatures", Combustion, Vol. The Babcock & Wilcox Co., New York, N.Y.
24, pp. 40-44 (Feb. 1953) (1955)
R. L. Reed, "Corrosion: Its Effect in Boiler Sys- "Suggested Rules for Care of Power Boilers",
tems", Combustion, Vol. 19, pp. 28-33 (May 1948) ASME Boiler and Pressure Vessel Code Section
and pp. 43-49 (June 1948) VII, Am. Soc. of Mech. Engineers, New York,
C. J. Schafer, "Protection of Idle Turbines and N.Y. (1954)
Superheaters," The Betz Indicator, Betz Labora-
tories, Inc., Phila., Pa. (March 1962)
174 BETZ HANDBOOK
24
Carryover
Effects of Carryover . ...................... 175

Causes of Carryover. . . . . . . . . . . . . . . . . . . . . . . 175
Mechanical . .......................... 175
Chemical. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 175
Prevention of Carryover. . . . . . . . . . . . . . . . . . . . 177
Mechanical. . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Chemical. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
Anti-Foam Agents . ..................... 180
Carryover Studies. . . . . . . . . . . . . . . . . . . . . . . 181
HE terms carryover, foaming and priming

T are closely associated and each term is
employed in describing conditions which
CAUSES OF CARRYOVER
Carryover may result from mechanical and
cause the entrainment of boiler water solids chemical causes or from a combination of
with the steam. Because priming, foaming the two. Among the mechanical causes of
and carryover are sometimes loosely used carryover are boiler design, high water level,
interchangeably, it is well to define their method of firing and load characteristics.
exact meaning. Among the chemical causes are high total
Foaming is the condition resulting from solids of the boiler water, high suspended
the formation of bubbles on the surface of solids, high alkalinities, oil and improper
the boiler water. The foam produced may type of external or internal chemical treat-
entirely fill the steam space of the boiler or ment.
it may be of relatively minor depth. In either MECHANICAL. Boiler design will influence the
case, this foaming condition causes appre- solids content of the steam. Certain types of
ciable entrainment of boiler water with the boilers are known for their ability to con-
steam. sistently produce clean steam while certain
other types are traditionally recognized as
Priming is a more violent and spasmodic
subject to carryover.
action, resulting in the throwing of "slugs" of
boiler water over with the steam. This action Turbulence of the boiler water is a definite
is similar to the "bumping" experienced contributing factor to the mechanical carry-
when water is boiled in an open beaker. over of boiler water solids. In older boiler
designs the steam carrying or riser tubes dis-
Carryover is the term applied to the con- charged below the water level resulting in
tinuous entrainment of a relatively small severe turbulence. To minimize this condi-
quantity of boiler water solids with the steam. tion modern designs have the steam gener-
The term carryover is also often used to cover ating tubes discharge above water level or
the general conditions more exactly divided through steam separating devices. These de-
into foaming, priming and carryover. vices have become quite complex and are
generally very efficient. However, some pres-
EFFECTS OF CARRYOVER sure drop is experienced across the steam
purifier and thus any leak in the equipment
When boiler water solids are carried over in will result in impure steam.
the steam, deposits will occur in non-return High water level is another frequently en-
and stop valves, superheaters, steam piping countered cause of mechanical carryover
and engines or turbines. Such deposits can while boiler circulation is yet another factor.
promote failure of the superheater tubes and For example, at low load one burner may
loss in engine or turbine efficiency. Depo- cut out and the resultant disturbance of the
sition on governor valves can permit over- boiler circulation may influence the tendency
speeding and wrecking of the machines. to carryover.
Foaming or priming of the boiler water Obviously, operation of a boiler at loads in
can result in a false water level reading while excess of its designed rating will provide
the thermal shock resulting from slugs of more cause for carryover than will operation
water entering the steam distribution system within the designed rating. Sudden load
as a result of priming can seriously endanger changes will cause more difficulty than steady
all steam driven equipment. It should also be loads.
noted that when moisture is carried over CHEMICAL. The American Boiler Manufac-
with the steam a loss in thermal efficiency is turer's Association stipulates in its standard
experienced, as such carryover corresponds to guarantees on steam purity the following
additional blowdown through the steam line. boiler water concentrations.
176 BETZ HANDBOOK
Oil present in boiler water is highly un-

Total Suspended
Operating Total Solids, Alkalinity, Solids, desirable from the standpoint of carryover.
Pressure, psig ppm ppm ppm
Saponification of compounding agents by the
0 - 300 3500 700 300 alkalinity of the boiler water will produce a
301- 450 3000 600 250 soap, thereby causing a foaming effect.
451- 600 2500 500 150
601- 750 2000 400 100 Internal boiler water treatment with phos-
751- 900 1500 300 60 phate control is standard for most types of
901-1000 1250 250 40 feedwater. However, there is a tendency for
1001-1500 1000 200 20
1501-2000 750 150 10 the precipitated, finely divided calcium phos-
2001 and higher 500 100 5 phate to stabilize boiler water foam, thus
contributing to carryover of boiler water
The effect of dissolved solids, suspended solids with the steam. This carryover tend-
solids and alkalinity on carryover has been a ency must be controlled by increased blow-
subject of considerable research. In some down to lower suspended solids concentra-
plants alkalinity is the determining factor tions or by the use of organic agents to con-
while in other plants it is the total dissolved trol this increased foam load.
solids that limit the boiler water concentra- The increasing amounts of various syn-
tions that can be maintained without carry- thetic detergents and wetting agents, both
over. In general, it is agreed that part for for industrial and domestic use, may well
part suspended solids have a much more lead to increased foaming tendencies for
pronounced effect in causing carryover than many surface supplies to which wastes are
dissolved solids. discharged. Unlike soaps, which react with
The constituent having the most pro- the hardness present in water to form in-
nounced effect upon carryover-suspended soluble precipitates, these synthetic agents do
solids-is far less frequently shown on a boiler not react with hardness and their foaming
water analysis than alkalinity, dissolved solids, qualities are not thus eliminated, except by
sulfate, etc. Alkalinity is determined by titra- dilution. Contamination of surface waters by
tion with a standard acid; dissolved solids these agents in various industrial and do-
can be determined by conductivity methods mestic waste waters has caused difficulty
or by gravimetric analysis, but suspended with boiler foaming.
solids can only be determined by a tedious In many areas where the water supply is
gravimetric procedure. If sufficient data are critical, sewage plant effluents have been
available, however, it is possible to calculate used by industry for both boiler and cooling
the suspended solids content of the boiler water supply. Usually, the high concentra-
with reasonable accuracy. tion of organic wastes present in the sewage
plant effluent will lead to a foaming boiler
In certain sections of the country where
water condition unless controlled by anti-
the raw makeup water contains a high dis-
solved solids content it would be impossible foam agents.
to operate the boilers within the limits speci- The conventional mineral analysis of a
fied in the standard steam guarantee of the water does not indicate whether foaming
American Boiler Manufacturer's Association. tendencies are to be expected from organic
Boiler manufacturers are cognizant of this contamination. Even a determination of the
condition and will, in cooperation with the organic content of the water does not pro-
purchaser, supply boilers with extra large vide this information since many surface sup-
steam drums and with special baffling and plies from heavily wooded areas contain a
steam purifying devices in order that the relatively high organic concentration of
unusually high solids concentration of the harmless or beneficial organics of the tannin
boiler water can be maintained without type.
detrimental carryover. While all of the factors mentioned above
possess a definite bearing on the foaming drum internal purifying devices has become
characteristics of a boiler water, because of increasingly more complicated.
the inter-relation of these various influences, Primary separation of steam and boiler
it is not possible to generalize to the extent water is secured by means of abrupt changes
of specific maximum limits with respect to in direction of flow, utilizing the difference
alkalinity, suspended solids, dissolved solids, in density of water and steam as the means
etc. Minute traces of organic surface active of separation. Impingement of the steam and
contamination may materially alter the water mixture against a baffle for gravity
boiling characteristics of a concentrated boiler separation may be employed or centrifugal
water and yet the concentration of the sur- force may be used in a separating device
face active contaminants present may be so which causes the water to be thrown from a
small as to defy detection by chemical whirling mixture. The majority of the boiler
analysis. water impurities is removed in this primary
Operating conditions also vary to such an separation which also serves the purpose of
extent and are affected by so many factors removing steam from the recirculated boiler
that, except in a general fashion, it is not water returning to the tubes. Any appreciable
possible to predict the maximum water solids amount of steam in the downcomer circuits
that can be tolerated without carryover will reduce the head available for circulation
under certain operating conditions. Specific and will reduce the boiler circulation rate as
limits for any one set of operating conditions a result.
can usually be set only after a detailed indi- Primary separation can be secured by rela-
vidual study of operation at that particular tively simple devices, such as "end around"
plant. baffles over the discharge of the riser tubes
or curtain baffles protecting a simple dry
pipe. More complicated arrangements are
PREVENTION OF CARRYOVER
used to improve the efficiency of primary
The means that can be taken for the pre- separation. In many industrial boilers, pri-
vention of carryover can be divided into mary separation only is incorporated in the
mechanical and chemical methods. Of course, steam drum internals. For high pressure tur-
mechanical correction is required for the bine operation, however, it is advisable to
minimization of carryover resulting from follow primary separation with a secondary
mechanical causes, such as change in boiler separation step so as to reduce the solids con-
design, installation of baffles, etc. In addi- tent of the steam to the lowest value possible.
tion various mechanical devices are em- Secondary separation is also termed steam
ployed for the separation from the steam of scrubbing and steam drying. In this process,
the droplets of boiler water carried over. small amounts of moisture must be separated
Chemical methods for the treatment of the from large amounts of steam. This separation
boiler water are ineffective where the cause is accomplished by providing a large amount
of the carryover is mechanical. However, of contact surface and frequent reversals of
where carryover has a chemical origin, such steam flow so that the mist of boiler water
as high calcium phosphate concentrations in may collect and be drained from the sepa-
the boiler water, proper chemical treatment rating unit. Steam velocity is kept low to
methods can be highly effective in elimi- avoid reentrainment of the separated boiler
nating this condition. water and to insure maximum contact for
MECHANICAL. As pressures approach critical deposition of the boiler water.
the difference in steam and water density be- Steam washing is also sometimes employed
come progressively smaller and thus with to lower the solids content of the steam.
the trend to higher steam pressures and Steam washing employs the incoming feed-
higher steam generation, the design of steam water to mix with and dilute boiler water
178 BETZ HANDBOOK
solids with feedwater of lower solids content, Figure 24-2 illustrates another design of
thereby securing a lowering not of the mois- steam drum internals in which double rows
ture content of the steam, but of its solids of primary separators separate the entering
content. Various steam washing arrange- steam-water mixture. The second stage of
ments include bubble cap washers and feed- purification consists of so-called steam con-
water sprays through which the steam must densers which are U shaped finned tubes
pass. A condensing type steam washer is also through which feed water flows. Because the
used in which feedwater does not directly tubes and fins are of lower temperature than
contact the steam, but in which steam is con- the saturated steam a small portion of the
densed on finned tubes cooled by incoming steam condenses and maintains wet surfaces.
feedwater. In this design, the steam is washed Steam is thus effectively scrubbed and
in its own condensate. washed in pure condensate. The final stage
Figure 24-1 shows an installation of cen- of steam purification is secured in the dryer
trifugal primary separators in a boiler drum. cartons.
Steam and water from the risers enter the In still another design of steam drum inter-
separator tangentially. The water moves nals for large natural circulation boilers, the
downward in a helical path on the inside steam entering the drum (along with 20-40%
wall of the cylinder. Centrifugal force on the water by volume) is collected in a compart-
mixture boiling around the cylinder separates ment formed by internal baffles. The first
the steam from the water. The steam then stage of steam purification is in the centri-
flows through the hollow space in the center fugal separator which spins the steam-water
and passes out of the separator into cor- mixture, throwing water to the outside and
rugated plates at the top of the cylinder, steam to the inside. A skim-off lip above the
where moisture removed from the steam is skimmer collects the layer of water from the
arrested and returned to the cylinder. Sec- outer wall and returns it to the drum through
ondary separation is secured in the steam the annulus surrounding the inner chamber.
scrubber section. The steam then proceeds vertically upward
to the next stage.
STEAM SCRUBBER {
Secondary separation is secured in two
Scrubber Element TO REDUCE SOliDS CARRY-OVER Scrubber Element
Corrugated Plates } TO EXTREMELY LOW VALUES Plates opposed banks of closely spaced thin cor-
rugated sheets which direct the steam in a
tortuous path at low velocity. Separated
moisture drains from the corrugated plates.
Additional separation takes place in the drum
space below the dryers and in the drying
screen.
Figure 24-3 is a line drawing of another
steam purifying device. In this separator the
centrifugal principle is again employed for
primary steam-water separation. The water
is removed through the bottom of the sepa-
rator through a slot discharging below mini-
mum water level while secondary steam puri-
fication is accomplished in a unit chevron
dryer as shown in Figure 24-4.
CROSS SECTIONAL ELEVATION CHEMICAL. As discussed earlier the principal
chemical causes of boiler water carryover are
Courtesy Babcock and Wilcox Co. high total solids, high alkalinity levels and oil
Figure 24-1 • Cyclone Separators contamination. By definition the total solids
STEAM TO
SUPERHEATER
STEAM SAMPLERS
DRYER
SCRUBBERS
(/)
V)
UJ UJ
co co
::::J ::::J
I- I-
UJ UJ
(.!) (.!)
~ ~
<{ <{
::c ::c
u u
(/)
(/)
0 0
~ ~
UJ UJ
I- I-
<{ <{
:;;: :;;:
~ :i
<{
<{
UJ UJ
I- I-
(/) (/)
FEED DISCHARGE NOZZLES THERMAL SHIELD
DRYER CARTON
DRAIN EJECTORS VORTEX BREAKER
Courtesy Riley Stoker Corp.
Figure 24-2 • Steam Purifier for Large Capacity Boilers
of a boiler water is the sum of the dissolved mally be accomplished either by adjustment
and suspended solids. Of these, the sus- of boiler water controls, such as lowering con-
pended solids are considered to be the more centrations through additional blowdown, or
objectionable as their effect on foam stabiliza- by change or alteration of the external feed-
tion is more pronounced than that of the dis- water treatment facilities. For example filtra-
solved solids. If solids are determined to be tion following a hot process lime and soda
the cause of carryover, correction will nor- ash softener may be inadequate permitting a
180 BETZ HANDBOOK
facilities may be required such as the instal-

lation of a hydrogen zeolite unit to operate
in parallel with a sodium unit thereby re-
ducing the total alkalinity of the treated
water.
The prevention of carryover difficulties
due to the presence of oil in the boiler water
is a problem that must be handled entirely
external to the boiler itself. No method of
internal treatment can be relied upon to over-
come a carryover problem caused by oil
contamination and the only safe policy that
can be followed is the removal of the oil
from the boiler feedwater.
ANTI-FOAM AGENTS. Frequently the cause of
a carryover problem cannot be economically
corrected through adjustment in boiler water
balances or installation of additional external
Courtesy Foster Wheeler Corp. treatment facilities. In many such instances
Figure 24-3 • Horizontal Steam Separator the use of effective anti-foam agents has
provided the desired result.
During recent years highly effective or-
ganic anti-foam agents have been developed.
In general the polymerized esters, alcohols
and amides are the most effective agents yet
developed. Various theories of foam inhibi-
tion by these compounds have been proposed
and one of the most logical is to the effect
that adsorption of the anti-foam agent on
the metal surfaces results in hydrophobic
conditions at the boiling surface. Fewer but
larger steam bubbles are generated under the
same heat transfer conditions and these
bubbles readily coalesce in place on the
heating surface, even prior to leaving that
surface and rising through the boiler water.
For the same amount of total steam produc-
Courtesy Foster Wheeler Corp. tion, tiny bubble formation is avoided and
Figure 24-4 • Unit Chevron Dryer non-foaming boiler water is the result of this
action.
higher amount of suspended solids in the With the use of these anti-foams it has
boiler feedwater than can be tolerated. been possible to develop boiler water con-
Alkalinity as a contributing factor to carry- centrations in laboratory experimental boilers
over can also be corrected by adjustment or in excess of 150,000 ppm before carryover
changes in internal and/or external treat- was experienced. While no practical use can
ment facilities. If the alkalinity is a result of be made of such anti-foam power it clearly
the internal treatment program the use of demonstrates the effectiveness of the modern
less alkaline materials will correct the condi- anti-foam. Figure 24-5 graphically illustrates
tion. Should the cause be external, additional the results of one such laboratory experiment.
COMPARATIVE TEST RESULTS

STEAM PURl TY
but also for techniques to measure such im-
purities in fractional parts per million. This
has been. achieved through the development
0 of highly sensitive conductivity equipment
and the sodium tracer technique which are
5
0
v
PHOSPHATE CONTROL
~NTJ~ FOAM AGENT
PHOSPHATE CONTROL
WITH ANTl-F.OAM AGENT
.......
discussed in another chapter.
Carryover can be a serious problem in the
modern steam plant and frequently the
source of the carryover can only be estab-
0
'../ ./- lished by exhaustive studies employing sensi-
2000 4000 6000 8000 10000 12000 14000 16000
tive sampling and testing techniques. Science
TOTAL SOLIDS IN BOILER WATER, PPM.
has provided the water treatment engineer
with these tools and through their proper use
Figure 24-5 • Test Results Indicating Effect of
Anti-Foam Agent
he has been enabled to assist the plant
operator in the generation of maximum
While the prime purpose of anti-foam purity steam consistent with minimum blow-
application is the generation of high purity down requirements and clean boiler internals.
steam additional economic benefit is fre-
quently experienced through reduced blow- REFERENCES
down requirements. This has been proved on
numerous occasions by actual plant studies 0. de Lorenzi, "Combustion Engineering", pp. 20-
where anti-foam was applied initially to 1, 2, 3, Combustion Engineering-Superheaters,
Inc., New York, N.Y. (1949)
overcome a chemical cause of carryover and, M. H. Kuhner, "How Steam Is Released in
subsequent to the anti-foam feed, boiler water Water-Tube Boiler Drums", Power Plant Engi-
concentrations could be carried safely at neering; Vol. 48, pp. 92-94 (Sept. 1944)
much higher values without detrimentally R. A. Mumford, "A Theory of Carryover", Com-
bustion, Vol. 18, pp. 39-41 (Feb. 1947)
affecting steam purity. P. B. Place, "Investigation of Carryover Problems
CARRYOVER STuDIEs. The design of modern and Identification of Types", Proceedings, Engi-
neers Soc. of Western Penna., pp. 43-62 (1947)
steam driven equipment is such that toler- J, H. Richards, "Solving an Unusual Case of
ance for impurities in the steam is practically Carryover", lvfodern Power and Engineering, Vol.
non-existent. Thus there is an ever increasing 51, pp. 101, 102, 190-193 (Mar. 1957)
demand, not only for more effective steam C. D. Shields, "Boilers: Types, Characteristics,
and Functions", pp. 227-244, F. W. Dodge Corp.,
purifying equipment and anti-foam agents, New York, N. Y. (1961)
182 BETZ HANDBOOK
25
Silica Turbine Blade Deposits
Occurrence of Silica Deposits. . . . . . . . . . . . . . . . 183

Cause cif Silica Carryover. . . . . . . . . . . . . . . . . . . 184
Removal of Deposits.... . . . . . . . . . . . . . . . . . . . 184
Prevention ~~Silica Deposits. . . . . . . . . . . . . . . . 186
URBINE blade deposits develop from a culties may have numerous causes such as
T number of causes, but the general effect
is the same in all cases. That is, an adherent
condensate contamination or incorrect feed-
water treatment, they are not considered fur-
deposit develops in the steam passages and ther in this discussion. In recent years, a third
distorts the original shape of turbine nozzles and more troublesome type of carryover has
and blades. The deposits are frequently some- been observed, and this is the specialized
what rough or uneven at the surface, causing case of silica carryover.
increased resistance to the flow of steam. Dis- Non-selective carryover of boiler water
tortion of the steam passages will alter solids occurs most frequently in units operat-
steam velocities and pressure drops, and the ing at lower pressures, and various methods
overall effect will be one of causing less effi- of correction are available depending on the
cient recovery of the energy which would be cause. The deposits from such carryover have
available from the steam. Under sufficiently a large proportion of soluble salts, and wash-
severe conditions, the deposits may develop ing of turbines is usually sufficient to restore
unevenly and unbalance the turbine rotor. full efficiency. On the other hand, when
Corrosion or carryover are the principal selective carryover of silica takes place, the
causes of turbine blade deposits. Corrosion turbine blade deposits frequently contain up
may be caused by incomplete feedwater treat- to 95% silica, with the result that simple
ment, or by incorrect lay-up procedure when water washing is no longer adequate and
the turbine is not in service. The principal specialized cleaning must be used.
cause of corrosion is oxygen in the presence
of moisture, and less frequently, high carbon
dioxide content of the steam together with OCCURRENCE OF SILICA DEPOSITS
moisture. Deposits may also originate with Turbine blade deposits in which silica is a
normal carryover, gross carryover, or selective predominant constituent and which generally
carryover of boiler water solids with the tend to be insoluble are very rarely encoun-
steam. tered below 400 psi. As a result, such de-
With the deposition of solids on turbine posits were not recognized as an individual
blading, rapid loss in efficiency will occur. problem until comparatively recent years.
Records for stage pressures in the turbines Since that time, considerable investigation
will indicate progressively higher pressures. has been given to the problem and a number
Stage pressures will increase beyond the of groups are now cooperating to find a com-
manufacturer's recommendations, and fur- plete and effective solution to the problem.
ther operation will not be considered advis- The problem becomes increasingly severe
able. Most important will be the increase in with increase in operating pressure. At pres-
pounds of steam per kilowatt hour of power sures of 900 psi and above, the turbine blade
generated. deposits which contain a high percentage of
The term "normal carryover" refers to the insoluble silica are the rule instead of the
carryover which occurs in any boiler operat- exception.
ing under the best conditions. It is simply a With the development of fast steaming
matter of boiler design and solids content of high pressure boilers, advancement was also
the feedwater and the boiler water. With made with improved ability of the units to
modern, efficient steam generators, such generate steam of high purity. Limitation of
carryover is less than 1 ppm of total solids. total solids in the steam to a maximum value
Industrial plants, at times, encounter a of 1.0 ppm is now a standard specification
condition which may be described as gross and most large high pressure units consis-
carryover, but which is caused by priming or tently operate under such conditions. When
foaming of the boiler water, or by leakage of this value is compared to the solids content
steam purifying equipment. Since such diffi- of the boiler water, it is seen that the process
184 BETZ HANDBOOK
of purification is carried out to a high degree. dioxide in the steam on deposits of sodium
On the other hand, even a very small solids silicate resulting from normal carryover.
content of the steam will amount to carry- Straub has experimentally investigated the
over of several thousand pounds of boiler possibility of solubility of boiler water salts
water solids per year from a high capacity in steam, as well as vaporization of boiler
unit. Of this quantity, only a very small per- water· salts. The results of his investigations
centage need remain in the turbine steam indicate that the common boiler water salts
passages in order to form objectionable de- are soluble to an insignificant degree at pres-
posits. The problem is further complicated by sures below 2400 psi. Sodium chloride ap-
the fact that there is considerable evidence pears to be the most soluble common ma-
to prove that silica is carried selectively from terial, and appreciable amounts may be
boiler water at higher pressures in such a dissolved in steam when the steam is super-
manner that normal steam purification equip- heated and can come in contact with solid
ment cannot prevent contamination of the salt. This condition is not common in normal
steam. boiler water system operation. Since solubility
is based on temperature and pressure of the
CAUSE OF SILICA CARRYOVER steam, deposition of the salt would be ex-
pected by reduction in temperature and pres-
A number of theories have been proposed to sure, such as occurs in a turbine.
explain silica carryover and silica -deposits in
Evidence was also obtained to support the
turbine steam passages. Some of the better
theory that silica will actually leave the
known explanations are reviewed.
boiler as a vapor. The quantity in the steam
Normal carryover is due to mechanical will vary with boiler pressure and with con-
entrainment of minute droplets of boiler centration of silica in the boiler water. It was
water in the steam. Such droplets of water will also noted that pH value of the boiler water
evaporate to dryness in the superheater sec- has considerable influence on the quantity
tion of the boiler, and the solids will con- of silica in the steam. A decrease in silica
tinue to be carried along with steam in a carryover was noted with an increase in pH
manner similar to that by which dust par- value.
ticles are carried in air. If such carryover
Although silica is found on turbine blades
resulted in turbine blade deposits, it is to be
as crystalline and amorphous silica ( Si 0 2) it
expected that the deposits would have a
has been proposed that vaporization from the
composition similar to that of the boiler
boiler takes place as silicic acid (H2Si0s).
water. It is apparent that other factors enter
This would account for the variation in silica
the picture, since deposits containing up to
carryover with pH of the boiler water. To
95% silica are common.
some degree, this theory is supported by other
Experimental evidence has been developed experimental results. Other investigators have
to support the theory that silica is selectively reported a definite relationship between tur-
carried from the boiler as a vapor, or in solu- bine stage temperatures and the form in
tion in the steam. Other theories propose that which silica deposits. These results throw
the difficulty stems from mechanical entrain- some doubt on the theory that silica carry-
ment, but assume various methods by which over is in the form of silicic acid.
the deposits that result are principally silica.
The proposed theories are based on melting REMOVAL OF DEPOSITS
points of various salts as they pass through
the superheater tubes, adhesive properties of Two procedures are in widespread use for
sodium hydroxide particles resulting from in- removal of insoluble silica turbine blade de-
complete drying while passing through the posits. These are caustic washing of the tur-
superheater tubes, and the action of carbon bine and mechanical cleaning. Where the
facilities are available, and where operating Turbine manufacturers and central power
conditions permit, it is desirable to attempt station operators have developed procedures
caustic washing, since mechanical cleaning for turbine washing which experience has
generally takes a longer time and proves to shown can be safely performed without
be more costly. damage to the turbine. The procedures are
It is_ desirable to precede caustic washing available from a number of sources, and the
by plain washing, in order to determine the operation must- be carried out as recom-
extent to which the deposit consists of soluble mended in order to avoid the possibility of
salts. Shutting down and starting up a tur- damage to the turbine from mechanical and
bine after permitting the machine to cool thermal stresses.
accomplishes a certain amount of washing Caustic washing follows the same general
due to the condensation which takes place. procedure as water washing. Frequently, the
Where temperature of the steam varies with washing is started by using plain water, and
load, a certain amount of washing can also a solution of sodium hydroxide is gradually
be accomplished by operation at very light substituted. A 10 to 20% solution of caustic
load. Where washing is desired under con- soda is generally used, although under some
trolled conditions, it is preferable to use water circumstances, a stronger or weaker solution
injection in order to desuperheat the steam. may be desirable.
Figure 25-1 • Blading Deposit Analyzed to be 90-95% Silica

186 BETZ HANDBOOK
eo
With simple washing, the progress of clean- 40 1"-
ing can be followed by measurement of
specific conductance of the condensate. How-
30
'
"
E
ever, when caustic washing is being used, it "'
"'20
""
is preferable to follow the progress of the "'
0
iii
cleaning by silica determinations on the con- <l
() 10
densate. Progress of the cleaning must be fol- :::; 9
u; a
lowed closely, in order to determine that all 0:: 7 ......
UJ 6
deposits have been removed to the greatest ~ 5
possible extent.
Some plants are not in a position to resort
0::
~
0
4
"f'..
"
<D
to caustic washing, since large quantities of
2
condensate are wasted during the operation. .........
In addition, if the deposits are sufficiently I'
heavy, it may be desirable to use mechanical I
500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700
cleaning in order to insure thorough removal BOILER DRUM PRESSURE, psig ,,
of all deposits.
Mechanical cleaning can be conducted by Figure 25-2 • Approximate Boiler Water Silica to
Limit Silica in Steam to 0.02 ppm
hand operations such as filing and scraping,
but is generally conducted by sand blasting.
Both fine sand and fly ash are widely used Steam purification systems which wash the
as the abrasive agents. During such an opera- steam with incoming boiler feedwater which
tion, care must be taken to prevent damage is low in silica, will slightly reduce the silica
to the metal surfaces. content of the steam. However, in most cases,
It is apparent that insoluble turbine blade the time of contact is too short for significant
deposits may be removed after they have de- reduction.
veloped, but it is also apparent that the The problem is essentially one of providing
procedure is costly, and in many cases highly feedwater with the lowest possible silica con-
inconvenient. Forced outage for cleaning may tent, and of preventing silica contamination
require overloading of other equipment, or from condenser leakage. After silica has en-
considerable expense for purchase of power tered the boiler water, the usual corrective
from an outside source. In the case of in- action will be blowdown within the ability
dustrial installations, plant shutdown may of the system to replace the blowdown water
even be necessary. The basic problem is with suitable makeup water.
therefore not one of removing deposits, but Internal treatment for removal of silica is
of preventing them in the first place. normally not practiced, but an interesting
application has been reported. Magnesium
PREVENTION OF SILICA DEPOSITS
oxide is widely used in hot process softeners
The experimental work of Straub comes for silica reduction with good results. Several
closest to fitting the facts of actual operating years ago, magnesium oxide was used intern-
experience, and on this basis, may be applied ally in a 1200 psi boiler. With feed of magne-
to actual operating conditions by making sium oxide, it was necessary to discontinue
suitable modifications, depending upon the feed of sodium phosphate, but particular care
circumstances of the individual problem. In was taken to practically eliminate calcium
units operating at 900 psi and above, it is contamination in the makeup water.
probable that the most significant factor of Condenser leakage is most readily detected
corrective action is maintenance of silica by providing instruments to continuously re-
content of the boiler water at least as low as cord specific conductance of the condensate.
10 ppm, but preferably less than 5 ppm. Such equipment must be maintained in good
working order and periodically checked. The circumstances, evaporated makeup IS too
chemical characteristics of the condenser costly and one or more of the other systems
cooling water must be kept in mind in inter- of treatment will provide satisfactory results
preting conductivity results on the conden- with favorable economy.
sate. For exarnple, the silica content of sea Evaporators may be obtained with per-
water is extremely low in comparison to the formance guaranteed so that the distillate
total solids content. At the other extreme, does not contain more than 0.5 ppm total
the silica content of some ground waters will solids with fresh water being used as evapora-
be quite high in comparison to the total tor feedwater. If sea water is used as feed-
solids. In addition to depending upon conduc- water the guarantee may specify that the
tivity measurements, it is also advisable to distillate will not contain more than ;,-4 grain
make periodic analyses for the silica content per gallon of total salts. Carryover from an
of the condensate. Determination of silica in evaporator corresponds to carryover from any
very low concentrations can be subject to low pressure boiler, and consists of mechani-
considerable error, unless the determination cal entrainment of the shell water in the
is conducted by an experienced analyst using evaporator vapor. The carryover from low
the proper laboratory facilities. pressure boilers is subject to many mechani-
The other principal source of silica con- cal factors, but evaporator design and opera-
tamination will originate with the makeup tion usually tend to eliminate or reduce the
water. The degree to which such contamina- factors which promote carryover. The most
tion can develop will depend upon the ex- important factor affecting vapor purity is con-
ternal system of treating the makeup water. centration of the shell water. Limiting concen-
The makeup water will usually be obtained tration may be specified by the evaporator
from a source such as condensate from a sys- manufacturer, and if it is not, then competent
tem operating at lower pressure, from evapo- advice should be secured.
rators, or from chemically treated water. Feedwater to the evaporator should be
Since it is necessary to maintain control deaerated and depending upon characteristics
over other factors aside .from silica content, of the feedwater, chemical treatment may
makeup water not from evaporators or con- also be advisable to eliminate scale formation
densate is usually treated to have essentially and sludge accumulation within the evapora-
"zero" hardness. This requirement calls for tor shell.
final treatment such as zeolite softening, de- The use of anti-foams to reduce carryover
mineralization, or hot phosphate softening. in high pressure boilers is well known. Similar
However, these processes may be preceded anti-foams may be used in evaporators to
by coagulation and filtration, hot or cold improve steam purity. The need for anti-
process softening using lime and other agents, foams is increasing due to expanding use of
and acid treatment. Choice of the exact treat- synthetic detergents which are present in in-
ment system is based upon chemical character- creasing quantities m many surface water
istics of available sources of makeup water supplies.
considered in conjunction with economy of The technology of demineralization has
appropriate systems. advanced to the extent that demineralization
Where a high pressure system operates is generally the most common method of
under conditions such that very little makeup preparing makeup water for high pressure
water is required, evaporation or deminerali- utility boilers. Present day demineralization
zation of the makeup water is generally the systems will produce water with very low
the most satisfactory system. Many industrial solids content including silica. In plants
installations use high pressure steam for proc- where makeup requirements are more than
ess requirements, and appreciable quantities about 10%, demineralization may not be the
of makeup water are required. Under these most economical method. Organic contamin-
188 BETZ HANDBOOK
Fig. 25-3
Occurrence of Silica in Natural Waters
Buffalo Elko Laurel Haverhill Savannah Warren Sea
N.Y. Nevada Miss. Mass. Ga. Ohio Phil. Is. Water
Hardness as CaC03, ppm .................. 118 180 10 30 107 160 84 6250

Calcium as CaC03, ppm .................. 85 135 7 25 79 105 41 1000
Magnesium as CaC03, ppm ................ 33 45 3 5 37 55 43 5250
"M" Alkalinity as CaC0 3, ppm....... ······ 90 436 66 17 109 24 130 115
Sulfate as S04, ppm ...................... 20 119 13 11 9 137 16 2650
Chloride as Cl, ppm 0 •••••••••••••• 0 •••••• 19 45 7 6 6 9 5 19000
...................
Silica as Si02, ppm -··· 1.6 80 50 0.8 53 0.2 90 0.02-4.0
Source ....... ············· .............. Lake Well Well Lake Well River Well Ocean
ation of the available water source may make Figure 25-3 illustrates the wide variation
th.e use of ion exchange resins questionable in silica content of natural waters which may
unless thorough pretreatment is provided. be encountered. Some of the waters are of
The other systems of zeolite softening, hot course rarely encountered, but serve to illus-
and cold process softening are well known trate the fact that where water treatment is
and widely used. Each system has certain concerned, the problem calls for individual-
inherent limitations, and these are the guid- ized attention.
ing factors in selecting a suitable system. As The means by which silica is transported
an example, the silica removal which can from the boiler water to the turbine steam
be obtained with each system can be sum- passages, and the means by which deposition
marized in general as follows: takes place are problems which are still be-
a. Silica reduction to approximately 1 ppm ing widely investigated, and on which con-
with hot process softening. tinuing progress can be expected. A generally
b. Forty to seventy percent silica reduction applicable, clear-cut solution is not available
with cold process softening. for the problem and each plant presents an
individual problem. In addition to the prob-
c. No silica reduction with zeolite soften- lem of preventing insoluble silica turbine
ing, and the possibility of silica increase with blade deposits, a plant may be faced with
a siliceous type exchange medium. the problems of preventing corrosion, scale
With any of the systems of treatment out- or embrittlement. These complicating factors
lined, design and operation of the equipment also influence selection of the proper system
as well as choice of treatment chemicals de- of treatment.
termines whether optimum results are ob-
tained.
REFERENCES
Some plants are fortunate in having more
than one source of makeup water available. E. E. Coulter, E. A. Pirsh and E. J. Wagner, Jr.,
"Selective Silica Carry-Over in Steam", Trans-
When this situation arises, a careful study actions, Am. Soc. of Mech. Engineers, Vol. 78,
should be given to each source of makeup pp. 869-873 (1956)
water, in order to avoid future difficulties F. G. Straub, "Steam Turbine Blade Deposits",
and provide best economy. Eng. Exp. Station Bul. 364, University of Illinois,
Urbana, Illinois ( 1946)
26
Measurelllent of Stealll Purity
History. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
Total Solids Determination. . . . . . . . . . . . . . . . . 190
Routine Mineral Analyses. . . . . . . . . . . . . . . . . . 190
Specific Conductance. . . . . . . . . . . . . . . . . . . . . . . 190
Tracer Techniques. . . . . . . . . . . . . . . . . . . . . . . 191
Steam Sampling. . . . . . . . . . . . . . . . . . . . . . . . . 193
190 BETZ HANDBOOK
ACCURATE measurement of steam purity is TOTAL SOLIDS DETERMINATION

1""\..of considerable significance in the oper-
ation of a steam plant. It is particularly so The collection of a sample of condensed
for the modern higher pressure plants where steam and the evaporation of 'the sample to
tolerances for deposition can be very critical. dryness and weighing of the solids residue is
With advancement in technique it is now another method employed in the determina-
possible to determine steam purity in parts tion of steam purity. This method is cumber-
per billion and it seems safe to state that some and time consuming, not lending itself
present knowledge permits reliable measure- to routine control. Satisfactory accuracy can
ment of steam contamination to a degree be obtained only with careful handling and
which should satisfy the demand of critical technique. The presence of 1 ppm of solids
high pressure operation. Thus it is possible in the steam will yield a weighable residue
to accurately check on boiler guarantee or of only 0.001 gram from a 1 liter sample.
effectiveness of steam purifying equipment. Naturally, therefore, a larger sample must
be taken in order to avoid the error in weigh-
ing and the sample must be carefully pro-
HISTORY OF METHODS OF DETERMINATION
tected from dust, etc. The large size sample
OF STEAM PURITY
must be evaporated down to dryness, under
The importance of steam purity measurement carefully controlled conditions. ASTM Ten-
in boiler operation was investigated as early tative Methods of Test for Suspended and
as 1865 and for some fifty years steam quality Dissolved Solids in Industrial Water (ASTM
determinations were accomplished mainly by Designation: D-1888-6lT) is intended for
steam calorimetry. Of the many types of application to water containing 5 ppm or less
calorimeters developed for this purpose the of total solids, and for which samples of
most familiar was the throttling calorimeter more than 4 liters must be evaporated. The
in which a sample of saturated steam is ex- American Boiler Manufacturer's Association
panded through an orifice to atmospheric stipulates that the total solids in the steam
pressure. The total heat in the steam sample for referee analyses shall be determined by
is greater than the total heat of the saturated testing no less than 10 samples of condensed
steam at atmospheric pressure, and the excess steam and taking the average of all results.
heat in the steam sample will superheat the
expanded steam or evaporate the moisture ROUTINE MINERAL ANALYSES
originally present in the steam sample.
The use of the conventional mineral analyses
From measurements of the pressure and covering such determinations as hardness,
temperature of the inlet and outlet steam, it alkalinity, chloride, pH, sulfate, etc., is of
is possible to calculate the moisture present
little significance in the analysis of condensed
in the original steam sample. The accuracy steam samples except of course where the
of the throttling calorimeter is usually taken
contamination of steam is excessive. The
as -+- 0.2% moisture which can mean an ap-
limit of accuracy of various test methods
preciable error in the solids content of the
such as hardness, chloride, etc. is such that
steam. For example, at a boiler water con-
they cannot be applied where the condensed
centration of 3500 ppm, an error of 0.2%
steam sample is reasonably pure.
moisture would correspond to an error in the
calculated total solids of the steam of 7.0
SPECIFIC CONDUCTANCE
ppm.
The wide use of superheated steam along The specific conductance of a condensed
with the need for more precise measurement steam sample, which is the measure of the
of steam purity and our broadening knowl- flow of an electric current through the
edge of steam properties have practically sample, is probably the most universally used
eliminated calorimetric measurement. method of determining steam purity. Pure
water has a very low conductance, but as and apply correction to the conductance
dissolved solids in a steam sample increase reading and the other is to degas the sample
so does conductance. This method of de- before measuring conductance. Accurate cor-
termining purity is rapid and reasonably rection curves are available to establish the
accurate. effect on conductance of carbon dioxide or
One of the disadvantages to the use of ammonia in a sample. While constant im-
specific conductance and which materially provements have been made in effectiveness
affects accuracy is the fact that some dis- of degassing equipment, traces of these gases
solved gases have a marked effect on the con- can remain. A further advancement in the
ductance of water. Such gases as carbon di- removal of dissolved gases has been the use
oxide, ammonia and hydrogen sulfide ionize of hydrogen ion exchange which can effec-
in water solution and thus interfere with the tively reduce ammonia to a negligible value.
use of conductance as a measure of dis- However, should the sample being tested
solved solids present. This interference can contain solids which are primarily in the
be appreciable when determining purity on form of sodium sulfate and sodium chloride,
a high quality steam sample. For example, in hydrogen ion exchange will produce acids
a sample containing less than 1 ppm dissolved which will result in 1an increase in conduc-
solids, specific conductance may be in the tance. There is no doubt that proper use of
range of 1.0 to 2.0 micromhos. The presence conductance and proper sample preparation
of any ammonia or carbon dioxide in this have greatly increased the reliability of this
sample will materially increase the conduc- method of steam purity measurement. Never-
theless the results of steam purity studies
using specific conductance and recently de-
veloped tracer techniques do indicate that
when measurement for solids in the order of
1. 0 ppm or less is desired, the accuracy of
conductance becomes questionable.
TRACER TECHNIQUES
While the steam purity guarantee for the

modern high pressure steam generators states
that the steam will contain not more than
1.0 ppm of total solids for given boiler water
conditions, it is now generally accepted that
steam purity of this order is not satisfactory
for high pressure operation if deposition in
superheaters or turbines is to be avoided.
Tracer techniques involve testing for a com-
Figure 26-1 • Specific Conductance Flow Cell and
ponent of the solids present in the boiler
Recorder with One of the Earliest Dega.r.rer.r
water, which component can reasonably be
tance reading. Ammonia will impart con- expected to be present in the steam at the
ductance of 8.0 to 9.0 micromhos per ppm same ratio to the total solids as is the case for
of this gas, while carbon dioxide will increase the boiler water. For tracer technique studies
conductance on an average of 5.0 micromhos published so far radioactive phosphorus
per ppm of carbon dioxide in the lower (p 32 ), which was added to the boiler water
range of concentration. and the sodium ion which is normally present
Two procedures may be employed to off- as the ion in greatest concentration in the
set the effect of dissolved gases. One is to boiler water, have been employed. While the
measure the content of the gas in the sample use of radio-active phosphorus resulted in
192 BETZ HANDBOOK
Figure 26-2 • Sodium Recording Equipment Set Up for Plant Study
extreme sensitivity, it presents a problem of densed steam with a precision of 0.0004 ppm.
handling and disposal of the boiler water and An average ratio of 2.5 to 3.0 ppm of solids
would be too costly to be practical for actual per ppm of sodium has been established for
plant steam purity studies. For this reason boiler waters so the sensitivity of this tech-
the work on the use ·of tracer techniques as nique is in the range of 0.001 ppm of solids.
a means of determining steam purity has Initially, the use of the sodium tracer tech-
concentrated on the use of the sodium ion. nique for steam purity evaluation possessed
Employing a flame spectrophotometer it is the disadvantage of the steam sample having
possible to detect sodium content of con- to be tested in the laboratory. Extreme
Figure 26-4 • Strip Chart Showing Effectiveness

Figure 26-3 • Close-up of Burner Housing of Air Filter
care had to be taken in the sampling pro- and it was generally assumed that with the
cedure to avoid contamination by sodium use of either an ASTM or an ASME steam
which is always present in the atmosphere to sampling nozzle, correctly installed, good
some extent. This procedure was time con- steam sampling would be experienced. As
suming and did necessitate frequent resam- methods of measuring steam purity become
pling to verify results. Steady progress, how- increasingly sensitive, it has become obvious
ever, has been made in the technique of that our analytical techniques are far ahead
measuring sodium by flame spectrophotom- of our sampling techniques. A great deal of
eter. It is now possible to operate the experimental work on steam sampling
equipment in the boiler plant with contin- methods has been conducted recently in an
uous measurement and recording of the so- attempt to improve knowledge of this sub-
dium content of a constant flowing condensed ject. This work has been conducted under
steam sample. The problem of contamination the supervision of the Steam Contamination
from sodium in the air was eliminated Subcommittee of the Joint Research Com-
through the use of a highly effective air filter mittee on Boiler Feedwater Studies of the
which provides a sodium free air to the ASME.
burner of the spectrophotometer at a slight
positive pressure. Experience gained with this REFERENCES
equipment in many actual steam studies has
shown that steam purity evaluations to the R. V. Cobb and E. E. Coulter, "The Prevention
level of 0.001 ppm of solids can readily be of Errors in Steam Purity Measurement Caused
by Deposition of Impurities in Sampling Lines",
obtained. Proceedings, Am. Soc. for Testing Materials, pp.
It is of interest to note that of the many 1386-1395 ( 1961)
W. A. Crandall and W. Nacovsky, "The Develop-
plants which operate equipment for the con- ment and Operation of An Ultrasensitive Record-
tinuous monitoring of steam purity by con- ing Flame Photometer", Proceedings, Am. Power
ductivity, some steam generating plants op- Conf., Vol. XX, pp. 726-738 (1958)
perating at extremely high pressures have J. J. Maguire, "Sodium Test Measures Steam
Purity Accurately", Power Engineering, Vol. 62,
found the use of conductivity to be of in- pp. 81-82 (Sept. 1958)
sufficient accuracy for their needs. As a re- S. 0. Meyer, "Measurement of High Purity Steam
sult there are now several installations where by Continuous Sodium Recording", The Betz In-
dicator, Betz Laboratories, Inc., Phila., Pa. (Oct.
continuous monitoring of steam purity is ac- 1959)
complished by constant measurement and re- T. A. Miskimen, "Results of Steam Sampling Noz-
cording of the sodium content of the steam. zle Tests on Evaporator Vapor", Paper Number
59-A-301, Am. Soc. of Mech. Engineers, New
York, N.Y. (1959)
STEAM SAMPLING H. Phillips, "Closing Our Gaps in Our Knowledge
of Steam Sampling", Paper Number 59-A-287,
For our present knowledge of steam purity Am. Soc. of Mech. Engineers, New York, N.Y.
determination to be of practical value it is ( 1959)
essential that the sample being tested is truly E. A. Pirsh and F. G. Raynor, "Instrumentation
for the Determination of Steam Purity", Pro-
representative of the steam being generated. ceedings, Engineers Soc. of Western Penna., pp.
Detailed instructions for the installation and 79-88 ( 1956)
construction of steam sampling nozzles are J. H. Potter, "Steam Calorimetry", Combustion,
Vol. 29, pp. 51-55 (July 1957)
available in the ASME Power Test Code
J. K. Rice, "Steam Quality Measurements by
and the ASTM Standards. For many years Flame Photometer", Proceedings, Engineers Soc.
these standards have been widely accepted of Western Penna., pp. 89-101 (1956)
194 BETZ HANDBOOK
27
Condensate Return
Line Corrosion
Causes of Return Line Corrosion. . . . . . . . . . . . . . 195
Sources of Dissolved Oxygen. . . . . . . . . . . . . . . . . 196
Sources of Carbon Dioxide. . . . . . . . . . . . . . . . . . 197
Corrosive Effects of Carbon Dioxide. . . . . . . . . . . 197
Reduction of Carbon Dioxide Content of Steam. . . . 199
Internal Methods of Correction for Carbon Dioxide. 199
Ammonia . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
Neutralizing Amines. . . . . . . . . . . . . . . . . . . . 200
Filming Amines . . . . . . . . . . . . . . . . . . . . . . . 202
Heat Transfer. . . . . . . . . . . . . . . . . . . . . . . . . 203
problem of corrosion in return con-

T HE
densate systems can be brought under
control in most industrial systems by one of
several proven methods. This problem, once
responsible for expensive maintenance costs,
need no longer be of concern. In new plants,
precautionary measures should be taken dur-
ing design stages, rather than waiting until Figure 27-2 • Typical Grooving Attack by
expensive shutdowns show the need for Carbon Dioxide
correction.
CAUSES OF RETURN LINE CORROSION
Failures most frequently occur at threaded

joints because the metal thicknesses have been
reduced at these points. Figure 27-1 is an ex-
ample of such attack. Horizontal lines are
usually more severely attacked than vertical
lines due to incomplete draining of the con-
densate in systems that operate intermittently.
The corrosive action manifests itself in the
form of grooving or channeling for the most
part. Figure 27-2 is a cutaway section of pipe
illustrating the typical grooving attack caused
by carbon dioxide. Figure 27-3 shows another
pipe sample typical of carbon dioxide attack.
Figure 27-4 shows a condensate line in
which the corrosive attack is primarily in the
form of pitting. Such pitting attack is typical Figure 27-3 • Example of Pipe Wall Grooving ~y
of oxygen. Carbon Dioxide
The principal causes for steam and con-
densate corrosion are the dissolved gases car-
bon dioxide and oxygen. These gases must
be in solution before they become aggressive.
There are other influencing factors such as
rate of condensation, time of contact, tem-
Figure 27-1 • Carbon Dioxide A/lack at

Threaded Joints Figure 27-4. • Pz'tting Attack Typical of Oxygen
196 BETZ HANDBOOK
perature, metal heterogenity, galvanic action (5)

and others but only because they supplement 2Fe(HC03h + lf2 Dz = FezOl +
the basic factors. ferrous bicarbonate + oxygen = ferric oxide
The corrosion reaction for dissolved oxygen + 4COz + 2Hz0
carbon dioxide + water
may be illustrated as follows:
(1) A combination of reactions ( 4) and ( 5)
4Fe + 6H 0 +
2 30 2 = 4Fe(0Hh can be written to show ·the production of
iron + water + oxygen = ferric hydroxide Fe 3 0 4 , the magnetic oxide of iron (magnetite)
as illustrated by equation 6.
The ferric hydroxide formed may later re-
vert to iron oxide rust as follows : (6)
(2)
3Fe(HC03h + lf2 Oz = Fe304 +
ferrous bicarbonate + oxygen = magnetite
2Fe (OHh = FezOl + 3Hz0 + 6COz + 3Hz0
ferric hydroxide = ferric oxide + water carbon dioxide + water
When carbon dioxide dissolves in water it
forms carbonic acid which is a weak acid and In some cases, the decomposition of ferrous
produces corrosion in keeping with the fol- bicarbonate has resulted in the production of
lowing equation: deposits of ferrous carbonate:
(3) (7)
Fe + 2HzC03 = Fe(HCOllz Fe(HCOlh FeC03 +
iron + carbonic acid = ferrous bicarbonate ferrous bicarbonate = ferrous carbonate +
+ Hz COz + HzO
+ hydrogen carbon dioxide + water
The above reaction proceeds rapidly at pH From the above reactions, it can be seen
values below 5.9. The reaction product, fer- that deposits of FeO, Fe20a, Fea04, FeCOa
rous bicarbonate, tends to elevate the pH and and mixtures of these can be found present
as the pH increases above 5.9, the reaction in steam and condensate systems and are the
rate becomes reduced. The hydrogen formed type deposits usually responsible for plugging.
in the reaction will also have a retarding
effect particularly when the pH exceeds 5.9. SOURCES OF DISSOLVED OXYGEN
The ferrous bicarbonate formed in this
reaction is soluble and can be carried along Oxygen in steam and condensate systems may
in the condensate. However, at points of pres- originate directly from the boiler feedwater
sure drop or reduced carbon dioxide content or may enter at various points in the con-
in the vapor phase it will precipitate from densate system. Oxygen is present in most
solution in keeping with the following re- makeup waters and unless the makeup water
action: is brought to saturation temperature in a
properly vented open or deaerating heater,
(4)
appreciable amounts will remain in the feed
Fe(HCOlh FeO + water entering the boilers. When the feed
ferrous bicarbonate = ferrous oxide +
2COz + HzO water is subjected to boiler temperature the
carbon dioxide +water oxygen will be released and a portion of it
can be entrained with the steam and find its
When dissolved oxygen is also present, it way into the condensate system. Oxygen that
will influence the reaction rate by removing · enters in this fashion can be coped with
the hydrogen to form water. The ferrous bi- readily by use of mechanical or chemical
carbonate will precipitate as follows: deaeration of the feedwater.
In vacuum return systems or in gravity (8)

type return systems there is ample opportu- 2(HC0l)- + heat = (COli= +
nity for oxygen infiltration. While condensate bicarbonate + heat = carbonate +
that is returned under pressure will usually C02 + H20
be free of oxygen, infiltration can take place carbon dioxide +water
where such systems operate intermittently. (9)
(COli= + ~hO + heat = 2(0H)- +
It is not uncommon for cooling water to be carbonate + water + heat = hydroxide +
injected into the vacuum pumps to lower the C02
temperature of the condensate. This practice carbon dioxide
is also followed in certain gravity type return Reaction (8) proceeds to 100% completion
systems where the condensate is collected in whereas reaction (9) is only partially com-
a centrally located receiving tank for pump- plete with decomposition taking place to ap-
ing back to the boiler plant. The cooling proximately 80%. Based on these reactions
water is usually laden with oxygen and its it can be calculated that for each 1 ppm of
introduction into the condensate generally bicarbonate alkalinity (expressed as calcium
creates a corrosion problem that manifests carbonate) in the boiler feed water, 0. 79 ppm
itself in the form of pitting. Where cooling of carbon dioxide will be evolved. For each
water must be added, it should be controlled
1 ppm of carbonate alkalinity (expressed as
thermostatically so that the amount intro- calcium carbonate) 0.35 ppm of carbon di-
duced is held to an absolute minimum. To
oxide will be evolved.
avoid corrosion from the introduction of cool-
ing water, catalyzed sodium sulfite, which The other possible and usually minor
permits rapid reaction of the sulfite with sources of carbon dioxide are the free, gase-
oxygen in the cold, can be proportioned to ous carbon dioxide that is dissolved in most
the cooling water to react with the oxygen natural waters and the carbon dioxide which
and eliminate it from the water. forms from decomposition of soda ash that
may be used for regulation of boiler water
In general, the elimination of oxygen from alkalinity. Free carbon dioxide is almost com-
condensate systems can be accomplished pletely eliminated by efficient feedwater de-
through relatively simple means and for this aeration and consequently is rarely a factor
reason corrosion and pitting due to dissolved in steam and condensate corrosion. The use
oxygen is readily controlled. of soda ash is inadvisable where return line
corrosion is being experienced and should be
SOURCES OF CARBON DIOXIDE replaced by caustic soda.
Carbon dioxide is the usual cause of steam CORROSIVE EFFECTS OF CARBON DIOXIDE
and return line corrosion and the corrosion
is characterized by a general thinning of the The curves illustrated by Figure 27-5 show
pipe wall or grooving along the bottom of the that other conditions being fixed, corrosion is
pipe. proportional to the carbon dioxide concentra-
The chief source of carbon dioxide is the tion of condensates.
bicarbonate and carbonate alkalinity of the The rapid increase in corrosion rate with
makeup water to the boilers. The bicarbon- increase in carbon dioxide concentration and
ate and carbonate alkalinity when subjected with increase in flow rate is illustrated
to boiler temperature undergo thermal de- graphically.
composition and liberate carbon dioxide Even at low carbon dioxide concentrations,
which becomes entrained with the steam. The however, corrosion can take place at isolated
decomposition is illustrated by the following points in the condensate system. For example,
equations: the curve of Figure 27-5 shows that for any
198 BETZ HANDBOOK
z
0
i=
<(
0::
1- 40
LLJ
zo
LLJO
0..0
CJ)
l.LJX 30
:I:
0 a::
z
I ~
LLJ
. r: >- 20
<1:·· 0::
O::LLJ
z a..
0
CJ) 10
0
0::
0::
0
0 0+-----~------~-----r------~-----r------r------i
0 100 200 300 400 500 600 700
CONDENSATE FLOWING OVER SPECIMENS- LB/HR. (I INCH PIPE)
Figure 27-5 • Effect of Condensate Flow and Carbon Dioxide Content on Rate of Corrosion
concentration of carbon dioxide, the rate of negligible amount of carbon dioxide dissolved
corrosion increases with the quantity of con- in it, due to the fact that the partial pressure
densate flowing, or in other words, corrosion of the carbon dioxide in the vapor phase is
is a function of the total pounds of carbon low. Assuming that the incoming steam con-
dioxide. Therefore, in a given system corro- tains 20 ppm of carbon dioxide, the initial
sion may only be a serious problem where the condensate discharged by the unit will have
quantity of condensate handled is high, such less than a fraction of 1 ppm. The remaining
as in the main return line. carbon dioxide will accumulate in the vapor
In order for carbon dioxide to be corrosive phase above the condensate in the unit. As
it must be in solution. The laws of Henry condensation continues the accumulated car-
and Dalton govern the solution of gases in bon dioxide will develop a partial pressure
liquids and state in simple form that the con- sufficiently high to permit re-solution of more
centration of a gas dissolved in a liquid is and more carbon dioxide, until equilibrium is
proportional to its partial pressure (or con- attained, under which condition the carbon
centration) in the contacting gaseous phase. dioxide content of discharged condensate will
The application of these laws serves to ex- equal that of the incoming steam.
plain severe corrosion that occurs in units It can be calculated that the quantity of
condensing steam that operate intermittently. carbon dioxide that must accumulate in the
For example, when a drier or unit heater is gaseous phase will be in the range of several
first supplied with steam and condensation hundred to more than a thousand ppm before
occurs, the resulting condensate will have a equilibrium conditions are attained. If such
units operate intermittently, on cooling and as in the anion exchange resin. Normally, de-
pressure on the unit decreases considerable mineralization is limited to those plants where
quantities of carbon dioxide will go into solu- high percentage of returned condensate is
tion. The pH of the condensate will be expected, due to the cost of this process.
lowered and accelerated corrosion will take In addition to hydrogen zeolite neutralized
place. It is apparent, therefore, that corrosion with caustic, raw water or blending with
in isolated units can take place even though sodium zeolite, other ion exchange processes
the concentration of carbon dioxide in the can be used. Softeners operating on the
steam is relatively low. Because of the ac- sodium cycle do not reduce the bicarbonate
cumulation of carbon dioxide that is possible, alkalinity of the raw w;ater. However, acid
venting of the units continuously during oper- treatment followed by aeration can be used
ation or shortly before shut-down would following the softening process. Dealkaliza-
prove of material benefit. It is also essential tion by chloride-anion exchange is another
that such units be provided with proper method, in which the makeup water is first
drainage. passed through a sodium unit and then
through a second anion exchange unit re-
generated with salt. In this latter process,
REDUCTION OF CARBON DIOXIDE CONTENT alkalinity is reduced without the use of acid,
OF STEAM but chemical efficiency is poor compared to
other systems. The use of chloride-anion ex-
change is generally limited to small plants
The various softening processes in which lime where objection to handling acid is the in-
is employed, such as lime softening, lime-soda, fluencing factor.
and lime-gypsum softening, secure a reduc-
tion in alkalinity by precipitation of calcium By selection of the proper external treat-
and magnesium bicarbonate by the lime. Hot ment processes, it is possible to produce a
lime-hot ion exchange is another example of minimum carbon dioxide content of the
this type of process. It should be emphasized steam in the range of approximately 5 ppm
that while these processes reduce the feed- when using 100% makeup water. In most
water alkalinity, they do not completely systems, except at points where stratification
eliminate it. With normal balances, utilizing may occur, corrosion should be kept to a
hot process lime and soda softening and satisfactorily low level. Also, at this low level
100% makeup, the carbon dioxide content of carbon dioxide, neutralizing amines can be
economically considered.
of the steam will be in the range of 17 - 21
ppm.
The various processes using acid, such as INTERNAL METHODS OF CORRECTION
direct acid neutralization or hydrogen zeolite FOR CARBON DIOXIDE
softening, will reduce the alkalinity of the
treated water to a relatively low value. With Of historical interest only was the attempt to
hydrogen zeolite softening, chemical efficiency use inorganic chemicals to control carbon
is lower, but softening is accomplished simul- dioxide corrosion in condensate lines, as these
taneously. Where very low carbon dioxide methods proved to be ineffective. These
values are required, necessary alkalinity for methods involved the use of an alkali, such
good boiler operation can be obtained by as sodium hydroxide to neutralize carbon
caustic soda addition. dioxide or polyphosphates to form an iron-
Almost complete carbon dioxide removal phosphate film. The injection of these in-
is secured by demineralization, in which both organic materials to the steam was objec-
cations and anions are removed either by tionable for the same reasons that carryover
aeration, vacuum deaeration, or adsorption of boiler water solids is undesirable.
200 BETZ HANDBOOK
AMMONIA industrial plants, and consequent higher am-

monia requirements, the problem of copper
Ammonia has been employed for neutraliza- attack limits the application qf ammonia.
tion of carbon dioxide and elevation of pH.
Ammonia can be fed to the system as am- NEUTRALIZING AMINES
monium hydroxide (a solution of ammonia
gas in water) or an inorganic salt such as Neutralization of carbonic acid can safely be
ammonium sulfate can be used. At boiler accomplished by certain volatile amines such
temperatures, the ammonium sulfatq will be as cyclohexylamine, CsHuNH 2 and morpho-
decomposed with the liberation of ammonia line C 4 H 9 NO. These amines when fed to a
gas with the steam. boiler volatilize with the steam and combine
Ammonia has been successfully employed with the carbon dioxide in the condensate
for control of C0 2 corrosion and iron pick-up to neutralize its acidity. Unlike ammonia the
in central stations with low percentage make- amines in low concentrations are not corro-
up and low carbon dioxide concentration in sive to copper and zinc bearing metals. Con-
the steam. While a reduction in corrosion of trol of treatment is usually based on feeding
ferrous metals can be secured in this manner, sufficient amine to raise the pH of the con-
copper and zinc bearing metals can be seri- densate to 7.0. Satisfactory reduction of car•
ously corroded, particularly when oxygen is bon dioxide corrosion is obtained with the
also present and there is opportunity for use of the neutralizing amines in the absence
ammonia to concentrate. Because of the of oxygen.
higher C02 concentrations usually found in Figure 27-6 shows cyclohexylamine require-
9.---------------------------------------------------------~
25 50 75 100 125 150 175 200 225

CYCLOHEXYLAMINE ( 40%), ppm
Figure 27-6 • pH AdJustment with Cyclohexylamine
9r--------------------------------------------------------------.
::c
a.
225
MORPHOLINE (40%), ppm
Figure 27-7 • pH AdJustment with Morpholine
ments for different concentrations of carbon in this section of the turbine to minimize cor-
dioxide in the condensate. Increased carbon rosion and iron pick-up. Morpholine is par-
dioxide concentrations require increased ticularly suited to this use, because it is less
quantities of cyclohexylamine. Cyclohexyl- volatile than ammonia or cyclohexylamine.
amine requirements are shown for the 40% The distribution ratio of ammonia and
solution since this is a common commercial the amines can be used as one means of
strength employed to minimize fire hazard selecting the proper material. Distribution
with the material. Approximately 3.0 ppm ratio is the ratio of the concentration in the
cyclohexylamine (40%) are necessary for steam compared to the concentration in the
each 1.0 ppm carbon dioxide to elevate the condensate. Ammonia has a distribution ratio
pH of the condensate to 7.0. of 10, cyclohexylamine about 3 and morpho-
Morpholine requirements for different line only 0.4. This favorable distribution ratio
levels of carbon dioxide are shown in Figure for morpholine makes it the more applicable
27-7. Morpholine is usually employed as a amine for central stations to protect the wet
40% solution to ;minimize fire hazard. Ap- steam sections in high pressure turbines, since
proximately 3.6 ppm morpholine ( 40%) are high pH is imparted to initial condensed
required for each 1.0 ppm carbon dioxide moisture at a minimum feed rate. Cyclohexyl-
to elevate pH to 7.0. An interesting applica- amine can be expected to remain with the
tion of the neutralizing amines has been in steam. At points of initial condensation,
the prevention of corrosion in the wet end higher rates are necessary to maintain desired
section of high pressure central station tur- pH levels. This property of cyclohexylamine,
bines. In these sections of the turbine, con- however, is of benefit in extended systems
densed moisture is traveling at high velocity utilizing reduced pressure wet steam, espe-
and tends to remove any protective oxide cially where excess condensate is frequently
films. It is desirable to develop a pH of 9.0 trapped off.
202 BETZ HANDBOOK
This difference in the action of these two

most commonly used neutralizing amines can
be utilized to select the proper material for
any given plant. Blended neutralizing amines
are available, possessing the desirable char-
acteristics of morpholine and cyclohexyl-
amine. Such a blend may be the most ap-
plicable material in many industrial plants
where protection is desired both at points of
initial condensation, as in turbines, and in
low pressure steam systems used for heating.
Neutralizing amines do not protect return
lines against oxygen attack. However, with
proper deaeration of feedwater and other
steps to prevent oxygen pick-up, this dis-
advantage can be considered minor. Since
the neutralizing amine requirement is a func-
tion of the carbon dioxide content of steam,
judicious selection of external treatment
methods may be necessary to achieve greatest
economy.
FILMING AMINES
The filming amines function on a completely

different principle from the neutralizing Figure 27-8 • Water Repellancy of Octadecylamine
amines. The filming amines do not neutralize Film at Left. Untreated Specimen at Right.
carbon dioxide. Instead, they function by
forming on the metal surfaces contacted an
impervious non-wettable film that acts as a a filming amine. It can be seen how the
barrier between the metal and the conden- amine film causes the water to assume drop-
sate, protecting against both oxygen and car- let form, on both the large and the small
bon dioxide attack. When the filming amines drops. The untreated specimen at the right
are adsorbed on a metal surface, water will does not repel water and the water spreads
not wet that surface. The film formed and wets the metal surface.
through the use of these amines is of sub- The filming amines of value in the preven-
stantially monomolecular thickness and does tion of corrosion are the high molecular
not increase in thickness with continued treat- weight amines and amine salts having straight
ment. The amine will penetrate corrosion carbon chains containing 10-18 carbon atoms.
products present on a metal surface and Octadecylamine (C1sHa1NH2), hexadecyl-
deposit a protective film on the metal. While amine (C 16 HaaNH2) and dioctadecylamine
continuous treatment is advised, the adsorbed (C 36 H 74 NH) are examples of useful materi-
film on the metal surface is quite durable als of this type. In plant practice the most
and is not removed during short periods of successful results have been obtained with
discontinued treatment. the dispersed octadecylamine.
Figure 27-8 is a photograph of two test Octadecylamine and octadecylamine ace-
specimens on which a few drops of water tate are available in either dry, flaky form or
have been placed. The specimen on the left in the form of a dispersion which simplifies
has been exposed to condensate treated with the problem of solution and chemical feeding.
Condensate should be used in preparing the from the standpoint of improved heat trans-
chemical soultion which is then fed with a fer is the ability of the filming amines to
small chemical pump to the main steam line loosen and remove old corrosion films. Dur-
at a point where it will be well mixed with ing initial start-up, the amines are applied at
the steam. The material is dispersed readily a low rate to permit gradual removal of
with the steam and is carried through the corroswn products in the system.
·steam and condensate lines, forming a non-
wettable film on the metal surfaces contacted. REFERENCES
Unlike the neutralizing amines which must
be fed in direct proportion to the carbon di- A. A. Berk, "Treating Steam Chemically to Reduce
Return Line Corrosion", Industry and Power, Vol.
oxide concentration of the steam, filming 53, pp. 79-81 (1947)
amines are required at a rate sufficient to L. F. Collins and E. L. Henderson, "Corrosion in
establish and maintain the desired corrosion Steam Heating Systems", Heating, Piping and Air
Conditioning, Vols. 11 and 12, Sept. 1939-May
resistant film on the metal surfaces. 1940 incl.
E. Elliott and P. J. Gaughan, "Plant Stops Return
Line Corrosion-Saves $14,200 a Year", Power
HEAT TRANSFER Engineering, Vol. 55, pp. 104-109 (Aug. 1951)
H. L. Kahler and J. K. Brown, "New Polar Film
As previously stated, filming amines function Treatment for the Control of Return Line Corro-
as a barrier treatment forming a non-wettable sion", Proceedings, Engineers Soc. of Western
Penna., pp. 115-126 (1949)
film of substantially monomolecular thickness.
G. A. Mierendorf, "Protecting Heating Units and
This film does not impede heat transfer due Condensate Lines Against Corrosion by the Use of
to its very thin nature. Also, due to the non- Filming Amines", Proceedings, Am. Power Conf.,
wettable feature, dropwise condensation is Vol. XIV, pp. 453-458 (1952)
W. A. Tanzola and J. G. Weidman, "Film Forming
promoted actually producing an increase in Corrosion Inhibitors Also Aid Heat Transfer", The
heat transfer. Of even greater importance Paper Industry, Vol. 36, pp. 48-50 (April 1954)
204 BETZ HANDBOOK.
28
Elllbrittlelllent of Boiler Metal
Causes of Embrittlement. . . . . . . . . . . . . . . . . . . 205

Embrittling Characteristics. . . . . . . . . . . . . . . . . 206
Treatment Methods . ...................... 207
Status of Problem . . . . . . . . . . . . . . . . . . . . . . . . 208
cracking or caustic
I
NTERCRYSTALLINE
metal embrittlement generally occurs· in
boilers along riveted seams and at tube ends.
Cracking of the metal may cause dangerous
weakening of its structure and subsequent
boiler failure.
CAUSES OF EMBRIITLEMENT
There are three factors necessary to produce

intercrystalline cracking of boiler metal.
These are:
1-Leakage of the boiler water must take
place so as to permit escape of steam and Figure 28-1 • Embrittlement Cracks in a 1 Inch Plate.
concentration of the boiler water at the point The Cracking in this Drum Was So Extensive that the
of leakage. Drum Had to be Replaced.
2-The boiler metal must be subjected to
high stress. This stress may be internal, re-
sulting from cold working of the metal, or the
stress may be external, due to expansion and
contraction.
3-The concentrated boiler water must
possess embrittling characteristics and chemi-
cally attack the boiler metal.
Leakage of boiler water from riveted seams
or rolled-in tube ends can produce the re-
quired concentration of the boiler water. In-
vestigations have shown that intercrystalline
cracking is not produced at "normal" boiler
water concentrations, but that continued
evaporation may build up local concentra-
tions of sodium hydroxide to a range such as Figure 28-2 • Cracks in the Rolled End qf a Tube.
75,000 to 500,000 ppm which will cause in- These Cracks Were Thought at First to be the Result qf
tercrystalline cracking under stress and in the Excessive Rolling, but Further Investigation (and many
absence of an inhibitor. tube failures) Showed the Cause to be Embrittlement.
Even in boilers with welded drums, inter-
crystalline cracking may manifest itself at
rolled-in tube ends. In general, therefore, the
power plant engineer has little control over
the factors of leakage and stress.
Inasmuch as all three factors of leakage,
stress and embrittling character of the boiler
water are required for the production of in-
tercrystalline cracking, if it can be definitely
shown that one of these three factors is ab-
sent, it can reasonably be assumed that the
danger of embrittlement is non-existent in Courtesy Mutual .Boiler and Machinery Insurance Co.
that instance. The factor that can best be Figure 28-3 • Mud Drum of Boiler Which Exploded as
shown to be present or absent is the embrit- a Result of Caustic Metal Attack Weakening the Seam.
206 BETZ HANDBOOK
Courtesy Mutual Boiler and Machinery Insurance Co.
Figure 28-4 • Wreckage of the Power House Caused by the Explosion of the Boiler Shown in Figure 28-3.
tling character of the boiler water. If a boiler must be taken to render it non-embrittling.
water is naturally non-embrittling or has been
made non-embrittling by the addition of
EMBRITTLING CHARACTERISTICS
embrittlement inhibitors, the power plant en-
gineer can neglect the other two factors of Naturally, it is desirable that chemical con-
leakage and stress as affecting the embrittle- trols of some nature be established so that
ment problem. However, should it be defi- chemical analysis of the boiler water will
nitely shown that the boiler water does pos- indicate the ability of that boiler water to
sess embrittling characteristics, the definite crack steel. However, the methods of study-
possibility exists for intercrystalline cracking ing embrittling characteristics today are more
should the additional factors of leakage and direct. They consist of placing a steel speci-
stress also be present. men in contact with a boiler water under ex-
Good engineering practice, therefore, dic- treme concentrations and simultaneously
tates that the boiler water be investigated for placing stress on the steel specimen in order
embrittling characteristics. If the boiler water to reproduce under test conditions the same
is found to be embrittling, naturally steps circumstances which will cause cracking in
shown that cracking of the specimen will re-

sult if the boiler water is embrittling.
BOILER
WATER IN The Embrittlement Detector is most con-
veniently installed in the continuous blow-
down line of a boiler. If continuous blowdown
is not installed, boiler water may be taken off
at any convenient point and either returned
to the boiler or run to waste. The require-
ment is that water at substantially boiler
temperature circulate rapidly through the
block.
The use of the Embrittlement Detector,
while requiring attention once daily, possesses
the advantage of operation over a thirty,
sixty or ninety day period which tends to
take into account fluctuations in water char-
acteristics during that time. In addition, the
unit is operated on the actual boiler water at
the plant. Occasionally, the specimen will
show definite cracking on its removal from
BOILER
the block although in many instances the
WATER OUT cracking is revealed only on further bending
in the course of a subsequent examination of
the specimen bar in the laboratory of the
Figure 28-5 • Embrittlement Detector supervising organization.
TREATMENT METHODS
the boiler itself.
This testing is conducted with the use of The use of sodium nitrate is practically stand-
the Embrittlement Detector as developed by ard for the inhibition of embrittlement. While
the United States Bureau of Mines. This nitrate is naturally present in many water
testing unit, covered by U.S. Patents 2,283,954 supplies, the inhibition of embrittlement re-
and 2,283,955, is illustrated in Figure 28-5. quires a definite ratio of nitrate to the caustic
It consists of a rectangular block 5 inches alkalinity present in the boiler water. The
long, 214, inches thick and 3;;2 inches wide formula for calculating this sodium nitrate/
with a % inch hole bored through it, as illus- sodium hydroxide ratio in the boiler water is:
trated, through which boiler water circulates. NaN0 3/NaOH ratio=
The test specimen is a steel bar 5 inches Nitrate as N0 3, ppm X 2.14
long, % inch wide and ;;2 inch thick which M.O. alkalinity as CaC03, ppm-
is bolted into a slot in the block by a clamp- Phosphate as P04, ppm
ing plate and four studs. The clamping of
The ratios recommended by the U. S.
the specimen into place creates a high stress
Bureau of Mines depend on the boiler oper-
in the metal surface in contact with the
ating pressure and are as follows:
block. Then, by adjustment of the clamping NaN0 3 /NaOH ratio
plate and the adjusting screw, water is up to 250 psi 0.20
allowed to leak out very slowly under the up to 400 psi 0.25
specimen with escape of steam leaving a con- up to 700 psi 0.40
centrated solution in contact with the stressed Figure 28-6 illustrates the 'before and after'
surface. Laboratory and plant tests have appearance of two test specimens exposed in
208 BETZ HANDBOOK
control of this method of preventing em-

brittlement, it is advisable to check on the
hydroxyl ion content of the boiler water.
STATUS OF PROBLEM
A large portion of our present knowledge

of intercrystalline cracking and of the means
that can be taken to combat this condition is
due to the study of this subject at the U. S.
Bureau of Mines under the sponsorship of
the Joint Research Committee on Boiler Feed
Water Studies. Development of the Embrittle-
ment Detector by W. C. Schroeder and A. A.
Berk of the U. S. Bureau of Mines was a
great step forward in this work. Industry as a
whole is indebted to these research investi-
gators for the continuous logical develop-
ment of the basic principles involved and for
the continuous follow-up on methods of pre-
Figure 28-6 • Detector Specimen Bars venting embrittlement.
from a 300 psi Boiler
From the viewpoint of the power plant
engineer, therefore, the status of the em-
brittlement problem may be briefly sum-
a plant operating at 300 psi. The first bar marized. Embrittlement may strike unex-
cracked badly. Sodium nitrate was then em- pectedly, when we least anticipate it. It may
ployed for correction of the embrittling char- occasion costly repairs as well as serious inter-
acteristics of the boiler water. The second ruption in production. Each engineer should
bar, which exhibits no cracking, was obtained know whether his boiler water can cause
after the use of sodium nitrate. intercrystalline cracking of the boiler metal if
Coordinated pH-phosphate control pre- the required leakage and stress conditions are
vents the boiler water from developing em- also present. The embrittling characteristics
brittling characteristics by eliminating 'free'
sodium hydroxide from the boiler water.
Since it is desirable to avoid low pH in the IU
boiler water, the proper pH is provided
through the use of an agent such as trisodium
phosphate. Figure 28-7 illustrates the ap-
~ 10.5
11.0
IL
~ -
proximate pH of solutions of trisodium phos- /
phate of various P0 4 content. To insure the ;::
absence of 'free' sodium hydroxide and the a. 10.0
_j_
absence of any alkalinity other than that of I
trisodium phosphate, it is necessary to main- 9.5
J
tain the pH of a boiler water below the curve.
This coordinated pH-phosphate method of 9.0 J
20 40 60 80 100 120 140 160 180 200
preventing embrittlement is primarily appli- PHOSPHATE AS PO
cable to plants employing evaporated or de-
ionized makeup water. Although the curve in Figure 28-7 • Approximate pH Values of Trisodium
Figure 28-7 can be used as a guide in the Phosphate Solutions
of the boiler water can readily be checked in REFERENCES

the plant itself through the use of the Em- A. A. Berk and W. C. Schroeder, "A Practical
brittlement Detector under the direction of a Way to Prevent Embrittlement Cracking", Trans-
supervising laboratory. If embrittling char- actions, Am. Soc. of Mech. Engineers, Vol. 65, pp.
701-711 (1943)
acteristics of the boiler water are found, there
A. A. Berk, "The Prevention of Embrittlement
are definite established methods available for Cracking", Transactions, Am. Soc. of Mech. En- .
overcoming such undesirable conditions. Ap- gineers, Vol. 73, pp. 859-864 (1951)
plication of specific chemicals or methods of J. D. Betz, "Preventing Boiler Metal Embrittle-
ment", Petroleum Refiner, Vol. 24, pp. 98-102
treatment can be tested by further embrittle- (July, 1945)
ment test runs until specimens no longer 0. H. Preis, "Embrittlement A Modern Menace",
crack in the Detector. It is then possible to Power Engineering, Vol. 64, pp. 68-69 (Apr. 1960)
establish controls over the embrittlement in- S. F. Whirl and T. E. Purcell, "Protection Against
Caustic Embrittlement by Coordinated Phosphate
hibitor and to so control the boiler water as pH Control", Proceedings, Engineers Soc. of West-
to maintain it in a nonembrittling condition. ern Penna., pp. 45-53 ( 1942)
210 BETZ HANDBOOK
Purpose for Blowdown. . . . . . . . . . . . . . . . . . . . 211
29 Manual Blowdown . .................... 212

Continuous Blowdown . .................. 213
Boiler Blowdown Control

Equipment Employed . ..................... 213
Manual Blowdown . .................... 213
Deconcentrators . ....................... 215
Surface Blowdown. . . . . . . . . . . . . . . . . . . . . . 215
Continuous Blowdown . .................. 215
Blowdown Control. . . . . . . . . . . . . . . . . . . . . . . . 216
Total Solids. . . . . . . . . . . . . . . . . . . . . . . . . . 216
Dissolved Solids. . . . . . . . . . . . . . . . . . . . . . . 217
Sulfate, Silica and Alkalinity . ............ 220
Chloride . ............................ 220
Specific Gravity. . . . . . . . . . . . . . . . . . . . . . . . 220
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 220
term blowdown means to eject or

T HE
drain a portion of the water from a
system. Boiler water blowdown is the removal
k
100 y
X
Since the total chlorides entering boiler must
from a boiler of some of the concentrated
equal total chlorides leaving boiler
water to be replaced with feedwater, thus
effecting a general lowering of concentra- xa = yb
tions in the boiler. 100
Multiplying both sides of equation by---
Boiler blowdown is usually expressed m equals xb
percentage and is defined as follows:
100 a 100 y
Quantity blowdown water
- - - - - - - - - ( 1 0 0 ) =%Slowdown b X
Quantity feedwater
100 y
Thus a 5% blowdown means that 5% of Since by definition--- = k, then
X
the water fed to the boiler is removed by
blowdown. The percent blowdown normally 100 a
k ----or
is determined by means of the chloride test. b
The formula and its algebraic derivation is Cl in feedwater
shown in Figure 29-1. X 100 = % blowdown
Cl in boiler water
PURPOSE FOR SLOWDOWN The American Boiler Manufacturer's Asso-
ciation in its standard guarantee of steam
Under normal operating conditions the only purity stipulates the following limits for
reasons for boiler blowdown are to lower the solids content of a boiler water:
suspended solids and dissolved solids content
of the boiler water. These solids are brought Operating Total Solids,
Suspended
Solids,
in by the feedwater and since the steam Pressure, psig ppm ppm
leaving the boiler is relatively pure, concen-
0- 300 3500 300
tration of solids will develop in the boiler 301- 450 3000 250
water. Excessive amounts of suspended solids 451- 600 2500 150
will cause deposition of sludge and may alter 601- 750 2000 100
751- 900 1500 60
the surface tension of the boiler water This 901-1000 1250 40
effect, in turn, may be severe enough to 1001-1500 1000 20
1501-2000 750 10
cause carryover in the absence of suitable 2001 and higher 500 5
anti-foam agents. Excessive amounts of dis-
solved solids have the same deleterious effects These recommended limits for solids are
as suspended solids and, in addition, when not entirely applicable to each individual
the solubility of the salts present in the boiler
case. Rather, they are limits of a general
water is exceeded, scale w:ll be deposited.
nature, providing a suitable basis from which
Fig. 29-1 to vary solids in accordance with the specific
Algebraic Proof of Slowdown Formula needs of a particular plant.
In the past few years, highly effective anti-
Let x = quantity feedwater foam agents have been developed and in
y = quantity blowdown water many instances, where they have been prop-
a = chloride concentration in feed- erly applied, excellent steam purity has been
water produced even though the total solids content
b = chloride concentration in boiler of the boiler water was maintained con-
water siderably in excess of the range stipulated by
k = percent blowdown the A.B.M.A. In fact, in many cases the
Therefore: by definition of percent blowdown effect of boiler water solids on steam purity
212 BETZ HANDBOOK
need no longer be considered in controlling lowering the dissolved solids concentration of

boiler water balances and therefore the sole the boiler water, it will also remove a portion
controlling factor in boiler blowdown is the of the sludge which is generally more con-
need to control sludge concentrations in the centrated in the lower part of the boiler.
boiler for the prevention of deposits on the When continuous blowdown is also in-
boiler heating surfaces. stalled, the primary purpose of manual blow-
The total solids content of the boiler water, down is to remove the suspended solids or
however, is only one factor controlling boiler sludge which may have accumulated in the
blowdown. In many cases, especially boilers area. In most cases, it is necessary to employ
operating above 600 psig, blowdown is based two or more short sludge blowdowns per day
upon limiting the silica content of the boiler in order to procure proper sludge removal.
water below certain values so as to limit the However, in cases where the boiler feed-
amount of silica vaporized with the steam. water is exceptionally pure, such as resulting
Vaporization of silica with the steam is not from an evaporator or with a high per-
usually a problem where boiler pressures are centage of returned condensate, blowdown
maintained below 400 psig. Unless silica can be employed less frequently as very little
values are properly restricted, a siliceous type sludge is formed in the boiler.
of scale may be deposited on the boiler When continuous blowdown is not in-
heating surfaces. Therefore, it is quite possible stalled, it is necessary to use the manual blow-
to find a plant operating with an apparently down to control concentrations in the boiler
high blowdown rate even though A.B.M.A. water. Whenever the dissolved solids or sus-
limits might permit higher solids and the use pended solids approach or exceed pre-
of an anti-foam might permit a very high determined limits it is necessary to use manual
total solids content before carryover occurred. blowdown to lower these concentrations. In
In addition to silica, other chemical factors practice, the valve of the bottom blowdown
which frequently limit or control the final is opened periodically in accordance with an
percentage of blowdown are alkalinity and operating schedule and/or chemical control
suspended solids. The two objectives gov- tests. From the standpoint of control, econ-
erning the boiler water limits which are omy and results, frequent short blows are
established are the production of pure steam preferred to infrequent lengthy blows. This
and the obtaining of clean boiler heat trans- is particularly true when the suspended solids
fer surfaces. content of the water is appreciable. Figure
In addition to the chemical balances main-
tained in the boiler water, there are many
mechanical factors which influence the total
solids concentrations which can be carried
safely. These include boiler design, rating,
water level and load characteristics and
whether water level and load variations are
relatively constant or subject to extremely
rapid variation.
MANUAL BLOWDOWN. There are two princi-
pal types of boiler water blowdown-inter-
mittent manual blowdown and continuous
blowdown. Manual blowdown, or sludge
- - - INFREQUENT LENGTHY BLOWS
blowdown, is necessary for the operation of a -----FREQUENT SHORT BLOWS
•••••••••• CONTINUOUS SLOWDOWN
boiler regardless as to whether or not con-
TIME
tinuous blowdown is also installed. The blow-
down take-off is usually located at the lowest Figure 29-2 • Effect of Different Types of Blowdown on
part of the boiler so that in addition to Boiler Water Concentrations
29-2 illustrates a hypothetical case. With heat exchanger can be used, providing a
the use of frequent short blows, a more uni- definite savings in initial cost of equipment
form concentration of the boiler water is since a considerable portion of the heat re-
maintained and in general a smaller quantity coverable from the boiler blowdown is con-
of water is blown down because of the better tained in the amount of flash steam obtained.
control. Figure 29-3 illustrates typical calculations
Bottom blowdown, properly controlled, ac- for determining the savings involved in a
complishes the lowering of the dissolved heat recovery system utilizing a low pressure
solids and suspended solids, providing for flash tank and heat exchanger. Installation
satisfactory boiler operation. It is frequently of heat recovery equipment is of value only
undesirable, however, from an operating when the low pressure steam or heat re-
standpoint, to blow down a boiler manually covered from the blowdown water can be
with sufficient frequency to properly control utilized for increasing the feedwater tem-
concentrations. This condition is especially perature. If an excess supply of exhaust or
true if the boiler feedwater is high in solids low pressure steam is already available, natu-
content. rally, there would be no savings and instal-
lation of heat recovery equipment would not
CoNTINUous BLOWDOWN. Continuous blow-
be justified.
down as the name implies is the continuous
removal of concentrated water from the Other applications of continuous blow-
boiler. It offers many advantages not pro- down water and methods of recovering the
vided by the use of bottom blowdown alone. heat from this water are practiced. Under
Continuous blowdown may be viewed as an certain conditions, an advantage is obtained
extension of the practice of frequent short in the recirculation of boiler water blowdown
to hot process softeners. In other instances, a
blows. In actual plant practice, periodic ad-
small portion of the blowdown water is re-
justments are made in the setting of the con-
trol valve in order to increase or decrease the turned to the feedwater heater for the pur-
amount of water blown out of the boiler in pose of elevating the feedwater pH value
accordance with controlled test results, thus and thus decreasing the possibility of cor-
making it possible to maintain close control rosion in the feed lines.
of boiler water concentrations at all times.
Another important advantage of continu- EQUIPMENT EMPLOYED
ous blowdown is that heat recovery is possible.
With the use of an efficient heat exchange MANUAL BLOWDOWN. The equipment em-
unit the only heat loss is the terminal differ- ployed for manual blowdown of the boiler is
ence between the incoming cooling water considered a part of the boiler and is installed
and the blowdown water to the sewer. This along with the unit. This equipment generally
terminal difference usually amounts to ap- consists of a take-off line, a quick opening
proximately 20 F. valve and a shut-off valve. The take-off line
The use of the flash steam principle prior is always located in the lowest part of the
to a heat exchange unit further increases the boiler proper where it is anticipated the
possible savings which can be realized by greatest concentration of sludge will form
continuous blowdown. By flashing a portion and also to permit proper draining of the
of the boiler water into steam at a lower unit.
pressure, the resulting low pressure steam Water tube boilers of various types gener-
can be used quite readily for process heating ally have more than one blow-off connection,
or, as is the usual case, for feedwater heating one or more being installed so as to blow
by direct introduction to the feedwater down from the mud drum and additional
heater. With the use of a flash tank in con- connections may be made to various headers.
junction with a heat exchanger, a smaller Blowdown connections are installed on the
214 BETZ HANDBOOK
Figure 29-3-Typical Calculations on Savings Possible through

Use of Heat Recovery on Continuous Slowdown. Basis-1 Day
Evaporation = 5,000,000 lbs steam
Boiler pressure = 200 psig
Feedwater temperature = 200 F (live steam used)
Makeup water temperature = 60 F
Rate of blowdown = 5.0%
Boiler efficiency = 75%
Fuel·= Oil at 150,000 Btu/gal, 7.0¢/gal delivered and fired
Feedwater = Steam plus blowdown
5,000,000 lbs
Feedwater 5,263,000 lbs
0.95
Blowdown = 263,000 lbs
Employing a flash tank at 5 psig, the quantity of 5 psig steam available may be calculated
from the formula:
Hb- Hr
% flashed steam = X 100
Vr
heat of liquid at boiler pressure, Btu/lb

heat of liquid at flash pressure, Btu/lb
latent head of vaporization at flash pressure, Btu/lb
362- 196
% flashed steam = X 100 = 17.3
960
Flashed steam available at 5 psig = 263,000 lbs X 0.173 = 45,500 lbs
Total heat of flashed steam at 5 psig = 1,156 Btu/lb
Heat savings in flashed steam = 1,156 Btuflb - 28 Btu/lb X 45,500 lbs = 51,324,000 Btu
The drain water from the flash tank is passed through the heat exchanger and thence to
the sewer. It is assumed that the temperature of the water leaving the exchanger is 20 F
above the incoming makeup water or 80 F.
Heat of liquid at 80 F = 48 Btuflb
Heat of liquid at 5 psig =.196 Btu/lb
Heat recovery = 148 Btu/lb
Heat savings from heat exchanger = 263,000 lbs X 0.827 X 148 Btujlb 32,190,000 Btu
Total heat savings = 51,324,000 Btu plus 32,190,000 Btu = 83,514,000 Btu
At a boiler efficiency of 75% there is actually utilized 150,000 Btu/gal X 0.75 120,000
Btu/gal oil.
83,514,000 Btu
Fuel savings 696 gal
120,000 Btu/gal oil
Daily savings = 696 gal X 7¢/ gal oil = $48.72
headers for the chief purpose of blowing similar device in an attempt to make the
down these sections of the boiler in order to installation more effective. A further refine-
remove suspended solids which may accu- ment of this idea involves the application of
mulate at such points and cause a restriction a floating trough which is designed to re-
of circulation. Usually, the boiler manu- move water at a point close to the boiling
facturer will stipulate certain restrictions in surface regardless of any variations in water
the blowdown of these water wall headers level in the boiler. Skimming devices, in gen-
and, in all instances, the recommendations of eral, have been more commonly applied in
the manufacturer should be followed. marine practice and have not found wide
DECONCENTRATORS. A deconcentrator is a application in stationary plants.
cylindrical tank so connected to the boiler as CoNTINuous BLOWDOWN. Continuous blow-
to receive concentrated boiler water. By down equipment is usually considered aux-
means of special baffiing, a settling of the iliary boiler equipment although many manu-
suspended solids takes place and a clarified facturers now install the internal collection
water is returned to the boiler. Periodic blow- pipe or at least will make provision for such
down of the deconcentrator results in the an installation during the course of manu"
blowdown of water containing a higher per- facture. The exact location for the continuous
centage of suspended solids as compared to blowdown take-off line within the boiler de-
the normal boiler water. A deconcentrator in pends primarily upon the water circulation
no way affects the dissolved solids or alka- circuits which are involved. The take-off line
linity content of the boiler water. Theoreti- is installed in such a manner as to assure the
cally, deconcentrators should be of value in removal of the most concentrated water from
cases where excessive suspended solids or the boiler. It is essential also that possible
sludge develops. In actual plant practice, short-circuiting of boiler feedwater or chemi-
however, they can be recommended only on cal feed solution into the take-off line be
rare occasions since installation of softening avoided. The size of the lines and control
equipment can be more readily justified in valve depends upon the feedwater character-
most such instances. istics, quantity of blowdown required and the
SURFACE BLOWDOWN. Connections are in- boiler rating. Figures 29-4 and 29-5 illustrate
stalled occasionally with the take-off line the method of installation in typical boilers.
located slightly below the working water level In most boilers, the take-off line is located
for the purpose of skimming sediment and approximately two inches below low water
oil from the surface of the water and the level although in certain boiler designs, the
name, surface blowdown, is generally applied take-off line is located close to the bottom of
to such an installation. The end of the take- the drum at a point where the most con-
off line may be equipped with a funnel or centrated water is found.
216 BETZ HANDBOOK
2nd pau
CONTINUOUS /
BLOWDOWN
CONTRO.!:__l
VALVE Q
CHEM. FEED
SPECIFICATIONS SECTION AA SECTION BB
CONTINUOUS BLOWDOWN:-Take-off pipe should be _ _ , extra heavy iron pipe lo-

cated under cyclone separators in 1st gas pass and at least 2" below low water level. End
of pipe should be capped and __ , hole drilled therein. On top side of pipe, _ _ equally
spaced __ , holes should be drilled. External to drum, install globe valve and reduce line
to _ _ " pipe. Install forged steel _ _ " _ _ - V port flow control valve similar to the Han-
cock Flocontrol at convenient location.
CHEMICAL FEED: -Distribution line should be _ _ , extra heavy iron pipe installed be-
neath cyclone separators in 3rd gas pass and at least 2" below low water level. Line
should remain open ended, discharging at point between 2nd and 3rd gas passes, as indi-
cated.
NOTE:-Use iron or steel valves and piping on all continuous blowdown and chemical feed
piping.
Figure 29-4 • Continuous Blowdown and Chemical Feed forB & W Type F and FH Boilers
BLOWDOWN CONTROL blowdown rate, it is essential that suitable

boiler water tests be conducted in order to
It is important that blowdown be such that obtain a constant check on concentrations
boiler water concentrations do not exceed developing in the boiler water. Regardless of
certain limits. Likewise, it is equally impor- the factor limiting the boiler water concen-
tant that excessive blowdown be avoided be- trations, the control of boiler blowdown is
cause of the heat units that are wasted as usually based on the determination of total
well as the additional chemical treatment re- solids content of the boiler water, determined
quired to react with the additional incoming gravimetrically, the dissolved solids as meas-
feedwater. Therefore, it is essential that a ured by specific conductance, the chloride,
system of control be established so as to main- sulfate, silica or alkalinity content of the
tain blowdown within determined limits. In boiler water or any combination of these
order to illustrate the savings which may be tests.
obtained by proper blowdown control, Figure ToTAL SoLIDS. Determination of the total
29-6 shows the amount of heat saved when solids content of the boiler water provides an
assuming a reduction in blowdown from 6o/o excellent control over blowdown. While usu-
to 2%. In order to maintain an economical ally applicable from a technical standpoint,
CONTINUOUS BLOWDOWN
CHEMICAL FEEDLINE "A"
I
(
\
')
_}
CHEMICAL FEEDLINE "B"
CONTINUOUS BLOWDOWN
A'
I
Globe Valve
Reducer _,.
SPECIFICATIONS Flow Control Valve
CONTINUOUS BLOWDOWN: Take-off line should be __ inch pipe located at least 2

inches below low water level in top, rear drum as indicated. Cap end of take-off and drill
with __ inch hole. In top side of the take-off, __ , _ _ inch holes spaced equally should
be drilled within length A- A'. Immediately external to the drum, install globe valve and re-
duce line to __ inch. Insert __ inch, forged steel, _ _ - V port flow control valve at an
easily accessible location.
CHEMICAL FEED: Distribution line should be __ inch pipe discharging open ended
approximately 2 inches below low water level and external to trough if provided. With tee
arrangement as in illustration "A", the open ends should discharge at third points of drum.
NOTE: Use extra heavy iron or steel piping and fittings for all continuous blowdown and
chemical feed lines.
Figure 29-5 • Continuous Blowdown and Chemical Feed for Three-drum, Bent-tube Boiler
it is rarely employed as the analysis requires carbonates liberating carbon dioxide gas, etc.
much time and is too difficult for a routine DisSOLVED SoLIDS. Determination of the
control. There are objections to the use of specific conductance of a boiler water will
the total solids test for calculation of the provide an accurate measure of the dissolved
rate of blowdown. A comparison of the total solids content and usually can be applied to
solids content of the boiler water with the blowdown control with considerable advan-
total solids content of the boiler feedwater tage. However, calculation of the rate of
will not necessarily provide an accurate blowdown on the basis of the relative specific
measure of the concentration that has taken conductance of the feedwater and boiler
place within the boiler due to the boiler water is not technically correct. Not only will
water samples not showing a representative specific conductance be affected by loss of
suspended solids content, effect of internal carbon dioxide with the steam and the solids
treatment, breakdown of bicarbonates and introduced by internal chemical treatment,
218 BETZ HANDBOOK
Figure 29-6-Typical Calculations on Sav- but the specific conductance of a dilute solu-
ings Possible Through Reduction in Blow- tion such as feedwater and a concentrated
down. Basis-1 Day solution such as boiler water cannot be com-
pared directly. The specific conductance of a
Evaporation . . . . . = 2,000,000 lbs. steam sample is caused by ionization of the various
Boiler Pressure . . . = 200 psig gage salts present in the sample. In dilute solu-
Feedwater Temper- tions, ionization increases and greater specific
ature . . . . . . . . . . =215 F conductance results. In more concentrated
Makeup Water solutions, ionization is repressed and specific
Temperature . . . = 60 F conductance decreases per part of salt pres-
Present Blowdown ent. This effect is illustrated by the fact that
Rate .......... =6% the solids content of very dilute solutions
Reduced Blowdown such as condensate may be calculated by
Rate .......... =2% using a factor of 0.5 to 0.6 ppm of dissolved
solids per micromho of specific conductance.
Boiier Efficiency . . = 80% In more concentrated solutions, such as boiler
Fuel = Oil at 150,- water, the factor will vary between the limits
000 Btu/gal., of 0.55 and 0.9 ppm of dissolved solids per
7.0¢/ gal. delivered micromho of specific conductance. If gallic
and fired
Feedwater . . . . . . . =Steam plus blowdown
Feedwater at 6.0%
This chart is used to calculate the percent
Blowdown . . . . . = 2,128,000 lbs.
of boiler water discharged by a continuous
Feedwater at 2.0% blowdown system that can be flashed into
Blowdown =2,041,000 lbs. steam at a reduced pressure and recoverable
Saving m Feed- as low pressure steam for heating or process.
water . . . . . . . . . = 87,000 lbs. ExAMPLE: A boiler operates at a pressure of
Heat of Liquid at 450 psig. Continuous blowdown amounts to
Boiler Pressure, 10,000 pounds per hour. What percentage of
Btu/lb. .. .. .. .. =362 blowdown water can be recovered as flashed
Heat of Liquid at steam at 5 psig?
60 F, Btu/lb. . . . =28 SoLUTION: Locate 450 psig on lefthand axis.
Heat Saving, Follow horizontally toward the right to the
Btu/lb. . . . . . . . . = 334 intersection with 5 psig "flash" curve (point
Total Heat Saving A) . Drop vertically downward to the bottom
= 87,000 lbs. x axis and read 25.5%. (25.5% of 10,000
334 Btu/lb. . . . . = 29,058,000 Btu pounds per hour blowdown = 2550 pounds
per hour of flash steam at 5 psig pressure.)
At Boiler Efficiency
of 80% There is These curves have been prepared from the
Actually Utilized formula:
150,000 Btujgal. He- HF
Percent flashed steam - - - X 100 where
x 0.80 . . . . . . . . . = 120,000 Btu/gal. VF
HB = heat of liquid at boiler pressure m
29,058,000 Btu Btu/lb
Fuel Savings 242 gal.
120,000 Btu/gal. HF = heat of liquid at flash pressure m
Btujlb
Daily Savings = 242 gal. oil x $0.07/ gal. V F = latent heat of vaporization at flash
=$16.94 pressure in Btujlb
%FLASHED
1\ 'I\ I I \ II if 'I I I
I\ \ \ II I\ II Ill/ 1/ 1/ II
\ 1 1\ 1 \ ~ I I II I I I
700 ~~~~~~++~~+1\wi\~I~X
1
~11~~1\~II~hHJ~II~II~I~I1~+4•~o
I~ \ 1\ I 1\ J JII II
\ \ \ I\ I II! II \ I II I I I
' 1 I~- - ft ~fll-· -11- ·f'r-
FLASH {'RES. l/1/E CVRVES \ ~ IV IJ
POUNDS PER SQUARE IIKH 6.1/U6E 1\ I~ 1\ l '\ Ill I I 1300 g
g 6oo \ 1J 1 \ I/ I 1\ 1 11 ~
~ 1/ 11 ' 1\ '1]}1111 J
i
:0
1
1~ '
i5
~
~
\/1\ 1\ 1\ \
V\1\ 111
l/
t\ Jlii!IV1 l' lii
I/
~
~
soo' H-l--1--+-+-1-+--+-+-+-+-+-+IH-1
/\.1--\t'+--llr-\-l\-
\+J[J4i-\l--
1/ ~
II I'
\lf.l/.f-ltr\Vh4-'-vlf-HI--t-+-t-t-H 1200 ':'
,1\1\ ~
'11/IAI/
IY 1' A 1/1\,
g
g
:g 1-· - -I- --- +· -1--- - - rt _, f-1 -, -t-. ~~if If I'V 1\ ~
~ J7 ~ '1\ 1/ I IJ\ :0
~ J.J..--+\I\r+l\¥1r\-\+\-l\-+,\A!,'Jlf-JA/IY-III/+1JI
40oH-l-t-++-l-+-+-+-i+J+-+-+I+- f-1 H-+-+1\-\t-t-t-t--H--t 1100 ~
~ 11 I\ lA ' \ OC . 1 \ ~· ~
~ II J ) lllJ\ ~ v II I I I\ z
8: I II 1\ 1\ V\ IV l~ I K I I\ £
~ V II \lXII '\r\ I/ \d 11 ~
g 300~4-~+-+~V~~~~~r-t-~'+-++A~~VAV~~~I~v~~'-r+\~-M'\~-t-~looo ~
I/ I II Vl"il~l\ ~~ \
J 1/ v ~ ~\ ~ 1\
1\
\ 1\
100
'
~o:::I/..LJ.._L.J-4-v.L.--L...~:::V:...r:V:..AV~Vx..I/..J...oQ.11~L.....L,_...1..-1..-+-JI..l...l....l\l\.1-l\;u..l\~l\u...~-~iL....J....:\~.~....~-....~.-~1...u...
\.1-j.U 800
0 5 10 15 20 30 3S
%FLASHED
Figure 29-7
Flashed Steam Recoverable from Continuous Blowdown Systems
NOTE: For boiler pressures between 100 ond 800 psig, use "flash" pressure curves slanting from lower
lefthand to upper righthand corner and the bottom axis. For boiler pressures above 800 psig, use
"flash" pressure curves slanting from lower righthand to upper lefthand corner and top axis.
220 BETZ HANDBOOK
acid is employed to neutralize the phenol- content will cause an appreciable error in the
phthalein alkalinity, the specific conductance calculated rate of blowdown.
multiplied by the factor 0.9 will give a very SPECIFIC GRAVITY. The specific gravity of a
close approximation of the actual dissolved boiler water is proportional to the dissolved
solids content of the boiler water. solids. However, the accuracy in determining
SuLFATE, SILICA AND ALKALINITY. Under dissolved solids by means of a hydrometer
certain circumstances, the sulfate, silica and measuring the specific gravity of a boiler
alkalinity content of the boiler water can be water is so poor that it cannot be recom-
used for blowdown control where they are mended for proper blowdown control.
factors which, in that particular case, limit
the boiler water concentrations that may be CONCLUSION
safely tolerated. Usually, these determina-
tions are not suitable for calculation of the It is to the benefit of any plant to maintain
rate of blowdown since both sulfate and boiler water concentrations consistently at
silica, may be removed from solution as scale. the optimum value. This control is necessary
The use of sodium sulfite will increase the to assure the production of pure steam and
sulfate content of the boiler water and alka- maintain heat transfer surfaces free of scale
linity is usually affected by either variable and other objectionable deposits. The eco-
softener balances or by the internal treat- nomics of blowdown requires careful study in
ment employed. All these conditions prevent order to make certain that blowdown is es-
accurate determination of the relative feed- tablished upon a proper basis, taking into
water and boiler water concentrations which account all chemical and mechanical factors
is the basis for calculation of the rate of affecting the problem. Material savings in
blowdown. fuel and water consumption may be made
CHLORIDE. The chloride determination can either by the application of proper chemical
be used not only for blowdown control, but treatment methods or by the installation of
it is also suitable for the accurate calculation heat recovery equipment. The entire subject
of the rate of blowdown provided the chlo- of blowdown is of sufficient importance to
ride concentration of the feedwater is suffi- justify in every case a suitable study of all
cient to permit accurate measurement. Since factors pertaining to it.
chloride is not subject to precipitation in the
REFERENCES
boiler water, the determination of the rela-
tive chloride concentrations of the feedwater 0. de Lorenzi, "Combustion Engineering", pp.
and boiler water provides the most accurate 21-17 & 18, Combustion Engineering-Superheat-
basis for calculation of the rate of blowdown. ers, Inc., New York, N.Y. (1949)
W. L. Nieland, "Control and Economics of Boiler
Therefore, the percent blowdown can be Blowdown", Combustion, Vol. 16, pp. 34-38 (Oct.
calculated substituting the determined chlo- 1944)
ride values in the formula as shown in Fig- W. L. Nieland and J. H. Richards, "Boiler Blow-
ure 29-1. down Control", Industry and Power, Vol. 57, pp.
64-68 (Mar. 1954)
The chloride test is unsuitable for this cal- R. T. Sheen, "Continuous Blowdown on Boilers",
culation only in those cases where the feed- Power Plant Engineering, Vol. 49, pp. 90-93 (Jan.
1945)
water chloride is quite low, as in the range
C. D. Shields, "Boilers: Types, Characteristics,
of 0.5-1.0 ppm. In this range a slight analyti- and Functions", pp. 227-244,. F. W. Dodge Corp.,
cal error in determining feedwater chloride New York, N.Y. (1961)
30
Chemical Cleaning of Boilers
Alkaline Boil-out. . . . . . . . . . . . . . . . . . . . . . . . 222

Inhibited Acid Cleaning. . . . . . . . . . . . . . . . . . . . 224
Hydrochloric Acid . ..................... 224
Phosphoric Acid. . . . . . . . . . . . . . . . . . . . . . . 226
Removal of Copper Deposits . . . . . . . . . . . . . . . . 226
0 ther Acids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 228
222 BETZ HANDBOOK
0 NE of the most serious problems encoun-

tered in steam generation is the forma-
tion of scale and sludge deposits on boiler
sate from steam engines or pumps. It may
also originate from steam or water used in
certain processes. Even thin films of oil on
heating surfaces. The presence of oil on boiler boiler heating surfaces may seriously interfere
heating surfaces may also give rise to serious with normal heat transfer and the resultant
difficulties. In high pressure generating sta- increase in metal temperature may be suffi-
tions boiler deposits containing metallic cop- cient to cause blistering and loss of boiler
per and copper oxide have been experienced. tubes and plates. ·
This condition may result in boiler tube fail- For the removal of oil and grease from
ures. In addition to mill scale, new boilers boiler heating surfaces an alkaline boil-out is
also contain deposits of oil, grease and pro- used. With new boilers the alkaline boil-out
tective coatings. It is important that the removes the protective coatings applied dur-
proper steps be taken to remove deposits from ing erection together with oil, grease and any
the boiler heating surfaces. Removal of de- dirt that may have accumulated during erec-
posits may be effected by one or a combination. The boil-out should be so formulated to
tion 'of chemical methods, depending on the provide the greatest detergent, emulsifying
particular nature of the deposit. and dispersive action on the greases and oils
normally found in new boilers. With boilers
ALKALINE BOIL·OUT that have been in operation, the boil-out
should exert the same effect on oil present as
The presence of oil in the boiler feedwater a result of feedwater contamination.
originates from the exhaust steam or conden- While it is not possible to obtain maximum
Figure 30-1 • Thin Scale Particles Cemented into a Dangerous Mass

emulsifying and dispersive action by the use rate in accordance with recommendations of
of inorganic chemicals alone, readily avail- the manufacturer. When steam is flowing
able inorganic chemicals may be used when freely from the vents, all vents should be
a specially prepared boil-out mixture cannot closed except those specified by the manu-
be obtained. One of the most satisfactory facturer on boilers containing superheaters,
combinations of inorganic chemicals includes where superheater protection is required.
3 pounds of caustic soda, 3 pounds of diso- The pressure 'is then raised to a maximum
dium phosphate and 1 pound of sodium ni- of 25 psi and maintained for approximately
trate for each 1000 pounds of water required 48 hours. For high pressure boilers it is usual
to fill the boiler to operating level. Because to boil-out at one-half the operating pressure
of the high alkalinity developed during the to obtain circulation. In all cases the recom-
boil-out, the sodium nitrate is included in mendations of the manufacturer should be
the combination as a precaution against caus- followed regarding boil-out pressure. The
tic metal embrittlement. boiler should be blown down one-half gauge
Prior to boiling-out a boiler, it is advisable glass approximately every eight hours through
to replace the normal high pressure gauge the bottom blow-off valve. Where there is
glasses with temporary glasses to prevent at- more than one blow-off connection, alternate
tack on the glass. After addition of the re- valves should be used. After each blowdown
quired quantity of boil-out material, all man- the boiler should be refilled to the top of the
holes should be closed and the boiler filled to glass.
the top of the gauge glass with the vents After completion of the boil-out the unit
open. The boiler should be fired at a low is cooled slowly, drained and flushed thoroughly
Figure 30-2 • This Superheater Tube Showing Carryover Deposits is an Application for Chemical Cleaning
224 BETZ HANDBOOK
with a high pressure hose. If inspection sists of black magnetic iron oxide. The pres-
should indicate that any scum or oil remains, ence in the boiler water of the black iron
the boil-out procedure should be repeated. oxide suspension may cause. a new boiler to
Some loose mill scale may be removed from run "black water" for several weeks under
a new boiler during the boil-out process. some conditions. These iron oxide deposits
However, removal of mill scale is not the also may accumulate to a point where they
primary objective of the boil-out since acid impede circulation and interfere with heat
cleaning is required for effective removal of transfer. They may also cause obstructions
mill scale. in steam separators and dryers, possibly af-
fecting the flow of steam in the drums. A
INHIBITED ACID CLEANING carryover condition may be aggravated,
causing iron oxide to deposit in superheaters,
The principal cause of scale formation on strainers and turbine blading. In the presence
bbiler heating surfaces is the fact that the of the alkaline boiler water, concentrated cor-
solubility of scale forming salts decreases with rosion can occur at breaks in the mill scale
an increase in temperature. When the feed layer and tube failures may result due to
water temperature is elevated to boiler op- these pits. Before acid cleaning, a new boiler
erating temperature, the solubility of scale should be given an alkaline boil-out so as to
forming salts is exceeded and crystallization · remove oil, grease, protective coatings and
takes place on the boiler heating surfaces. any dirt that may have accumulated on these
The use of inhibited acid for the removal of surfaces during erection.
scale from boiler and metal surfaces has been Prior to acid cleaning the usual protective
practiced for many years. Within the past measures should be taken such as replace-
few years, however, there has been a con- ment of any brass or bronze parts temporarily
siderable development in the technique and with steel or steel alloy equipment. Steps
a number of new inhibitors are now avail- should be taken to provide for proper vent-
able. Acid cleaning requires no great dis- ing of acid vapors and any valves connecting
mantling operation and labor cost is held to the boiler with the main steam header should
a minimum. Scale is removed from all sur- be closed.
faces reached by the acid solution and the HYDROCHLORIC Acm. Inhibited hydrochloric
cleaning, in most cases, is more complete acid is most widely used in acid cleaning be-
than could be obtained by mechanical means. cause of its relatively low cost and the num-
By acid cleaning the internal metal sur- ber of satisfactory inhibitors presently avail-
faces of a boiler are left metal clean and as able. Two general methods used in acid
such the true condition of the metal is re- cleaning of boilers are the "circulation"
vealed. Opportunity is provided for more method and the "fill and soak" method. In
complete inspection of questionable areas the first method, the cleaning solution is
within the boiler tubes, headers and drums. circulated continuously through the unit un-
In many cases pits and pitted areas may go til the cleaning action is completed, as shown
unnoticed until inspected after acid clean- by tests on the effluent. This method is suit-
ing. Cleaning down to the metal surface will able for units with positive liquid flow paths
help reveal any stress areas and cracks which and offers the advantage of regular checking
may be due to fatigue, embrittlement or during the cleaning operation. In the second
corrosion. method the cleaning solution is permitted to
Acid cleaning of new boilers is considered soak in the unit for a prescribed length of
desirable by some for the removal of mill time. This method is used where positive
scale. Mill scale is the term generally applied circulation of the cleaning solution cannot
to the oxide film produced when iron and be readily obtained.
steel surfaces are heated and cooled during In the "circulation" method the unit is
manufacture and fabrication. Mill scale con- filled to overflow with clean warm water at
the maximum temperature permitted for the metal surfaces from corrosion. This protec-
particular inhibitor used. This water is circu- tion can be accomplished by employing either
lated until a constant temperature is obtained the wet or dry lay-up procedures.
on the effluent which is returned from the In the "fill and soak" method inhibited
highest point of the unit to a tank equipped acid is added in the proper ratio with warm
with a clean water line, steam line and drain. water or condensate (at the maximum tem-
For circulation, a corrosion resistant pump perature permitted for the inhibitor) to give
should be employed. Inhibited acid is added a cleaning solution of the desired acid strength
in sufficient quantity to provide an acid solu- and the unit is filled to overflow. The vent is
tion of proper strength entering the unit and opened and soaking is carried out for a pre-
this feed is continued until tests on the efflu- determined length of time, depending upon
ent solvent indicate a maximum acid content the type and estimated quantity of deposit
has been reached. The concentration of acid and on the rate of reaction between the de-
used will generally vary from 3 to 6%, deposit and solvent, as indicated by laboratory
pending upon the particular conditions. tests.
Should continued testing of the effluent sol- At frequent intervals samples can be drawn
vent reveal a 1% drop in acid content, suffi- from available points and tested for acid
cient inhibited acid should be added to main- strength. However, these samples will not be
tain the desired acid content of the solvent indicative of conditions throughout the unit
influent. Circulation of the solvent through although they will give an indication as to
the unit continues until the acid content of the amount of deposit removed. Vent gases
the effluent is constant, thus indicating the may contain hydrogen and therefore open
end of the reaction between the solvent and flame is not permitted in the venting area.
the boiler deposit. With new boilers the soaking period, using
Following acid cleaning, the boiler is this method, may be approximately six hours
drained and flushed with clean warm water although treatment may be discontinued after
until tests on the returned flushing water a shorter period, depending upon control
show it to free of acid and soluble iron salts. tests.
Flushing is followed by a neutralizing or At the end of the acid cleaning period the
metal conditioning boil-out. Soda ash is gen- unit is flushed with clean warm water as
erally used for this purpose although triso- described previously. The neutralizing solu-
dium phosphate, sodium tripolyphosphate, tion is added and the unit is permitted to
sodium chromate, caustic soda or other chem- stand for a short period of time. The level
icals may be used. The alkaline boil-out can then be lowered to normal and the unit
serves to neutralize any residual acid, re- fired as described under the "circulation"
leases from the metal surface the hydrogen method. Following this neutralizing boil-out
generated during the cleaning procedure, and the unit should be drained and flushed, as
provides a slightly protective film on the previously described. Also, if the boiler is to
metal surface. To aid in neutralization the stand idle for more than a few days, proper
unit can be fired at 50 psi or somewhat steps should be taken to protect the metal
higher after circulation of the neutralizing surfaces.
solution has been stopped and the water level In addition to the properly inhibited acid
dropped to normal. During this period the it may be desirable under some conditions to
unit should be vented on occasion to permit include certain inorganic chemicals which
escape of the liberated gases. will aid in the removal of one particular
At the conclusion of this boil-out the unit constituent of the scale. The use of the proper
is drained and flushed with clean warm wetting agent along with the inhibited acid
water. If the boiler is to stand idle for more serves to reduce the rate of corrosive attack
than a few days before being returned to on the metal during cleaning. The lower the
service, steps should be taken to protect the degree of wetting of the metal by the inhib-
226 BETZ HANDBOOK
ited acid, the greater is the effect of the wet- REMOVAL OF COPPER DEPOSITS
t~ng agent in reducing the amount of corro-
siOn. The increased use of copper .and copper alloy
Sulfuric acid and nitric acid may also be equipment in the feedwater cycle has given
used in chemical cleaning. Sulfuric acid is rise to a new problem of boiler deposits in
economical and may be easily inhibited. high pressure generating stations that have
However, it does form insoluble salts, such low makeup water requirements. Deposits
as calcium sulfate, during the cleaning proc- have been composed principally of iron
ess. Although nitric acid is relatively inex- oxide, metallic copper and copper oxide.
pensive, there are only a few suitable inhibi- While acid cleaning is the answer to the
tors available. Both sulfuric acid and nitric problem of removing iron oxides, difficulties
acid are dangerous to handle and therefore may arise due to acid cleaning when copper
extreme care must be taken when these acids or copper oxides are present in the sludge.
are used. Although metallic copper is relatively in-
soluble in hydrochloric acid, it tends to go
PHOSPHORIC Acm. Inhibited phosphoric acid into solution during acid cleaning due to the
has been used successfully for removal of presence of ferric ion in the acid solution.
mill scale from new boilers. This agent pos- Copper can then replate on the boiler sur-
sesses certain advantages over inhibited hy- faces and it may adhere to these surfaces.
drochloric acid when used for chemical
cleaning. Inhibited phosphoric acid can be When the boiler is returned to service, the
boiled in the unit by direct firing of the boiler copper may be loosened from the boiler sur-
with neglible attack on the boiler metal and faces and deposit in areas where circulation
it does not create a problem of noxious or is less rapid. Under these conditions normal
corrosive fumes. Natural circulation resulting circulation may be affected and over-heating
from this direct firing promotes good distribu- may occur, resulting in tube failures. While
tion of the cleaning solution. Boiler surfaces copper deposits may be removed mechani-
cleaned with phosphoric acid are resistant to cally, this is a long, expensive process and as
corrosion whereas metal surfaces freshly such would nullify any benefits gained in
cleaned with inhibited hydrochloric acid re- acid cleaning. As in any mechanical process
act with the atmosphere to form rust if not there are locations that cannot be cleaned
properly protected during draining, and es- mechanically and thus deposits not removed
pecially if left in a moist condition, such as would be potential sources of difficulty dur-
before neutralization. ing later operation.
Prior to cleaning with inhibited phosphoric Various chemicals have been used for re-
acid, a new boiler should be given an alka- moval of copper deposits from boilers. In
line bbil-out, as previously described. A 5% some cleaning operations, the solvents were
inhibited phosphoric acid solution boiling at ammonia base oxidizing agents and the ef-
atmospheric pressure has been successfully fectiveness of these solvents is based on the
used in the removal of mill scale from new reaction of the oxidizing agent with metallic
boilers. Following the cleaning, some authori- copper to form cupric oxide. This oxide is
ties recommend that an acid rinse be em- dissolved in ammoniacal solution to form the
ployed for removal from the boiler surfaces familar cupric-ammonium complex ion. This
any iron that may have deposited upon series of reactions is reversible and it is neces-
draining the acid cleaning solution. A 0.1% sary that certain fundamental conditions be
solution of phosphoric acid may be used as maintained. During the treatment period ex-
a rinse. The final step in the cleaning process oxidizing agent must be available in the
cedure consists of an alkaline treatment for solvent and the ammonia present must be
purpose of neutralizing the remaining acid maintained at a uniformly high level of con-
and for conditioning of the metal surfaces. centration. The cupric-ammonium ion con-
centration must be uniform throughout the the principal oxidizing agent. The solvent is
solvent. circulated continuously through the boiler at
The first solvents used for removal of cop- a temperature of 140 F to 160 F. While this
per from boilers were those containing chlo- type of solvent dissolves copper at a less
rate as the major oxidizing component. One rapid rate than solvents containing persul-
of this type of solvent, made up of sodium fate, it is faster. acting than the chlorate type
chlorate, ammonia and ammonium sulfate, solvents. During the cleaning action analyti-
must be used at temperatures in excess of cal control can be set up to measure rate of
175 F in order to rapidly oxidize and dis- oxidant consumption and quantity of dis-
solve copper. Another type solvent, contain- solved copper in the solvent.
ing sodium chlorate, slight amounts of potas- More recently, ammoniated citric acid has
sium bromate, ammonia and ammonium ni- been employed for removal of water side de-
trate, must be used at temperatures between posits containing high percentages of copper
140 F and 190 F. With this solvent the bro- oxide. Also, complexing agents specific for
mate reacts first and limited amounts of copper have been used in an acid medium
chlorate react depending on the temperature for single stage cleaning of metal surfaces
maintained. containing significant quantities of copper
With the use of these solvents the boiler oxide in a multi-component deposit.
is generally fired intermittently to obtain re-
quired temperature and to induce circulation. OTHER ACIDS
Certain disadvantages are connected with the
use of these solvents. Where over-heating In the past few years, sulfamic acid
occurs at localized areas of the boiler, in- (NH 2 S03H) has come into use for chemical
creased attack on the boiler metal by the cleaning of equipment. This material, blended
oxidizing agent will occur. In over-head with an inhibitor, is available in powder form
areas, the ammonia content may be depleted and becomes active as an acid cleaner when
causing copper already taken into solution to placed in solution. In the cleaning operation
replate on the boiler metal. In the cleaning the solution of inhibited sulfamic acid is
process, any ammonia driven off into the air handled in a manner similar to that described
space in the boiler drums decreases the con- previously. Inhibited sulfamic acid has an
centration of the solvent, rendering it less advantage over hydrochloric acid inasmuch
effective. as it is easier to handle, does not produce
Another type of solvent contains ammo- noxious fumes on dissolving and is less corro-
sive than inhibited hydrochloric acid, particu-
nium persulfate, ammonia and sodium hy-
droxide. This solvent is highly reactive with larly at higher concentrations and tempera-
tures. Sulfamic acid, while more expensive
metallic copper even at room temperatures.
than either hydrochloric or sulfuric, is ef-
The solvent is recirculated in the boiler for
fective in removing calcium carbonate scale
a short period, then the unit is drained and
as well as iron oxide and organic deposits.
flushed with clear water. While the solvent
However, its rate of reaction is slow against
may be used at room temperature there are
calcium sulfate. In chemical cleaning, sul-
some disadvantages inasmuch as it reacts
famic acid has been used primarily for small
rapidly with iron as well as copper. As such,
boilers and other plant equipment such as
accurate chemical control of both oxidant
condensers and compressors.
consumption and quantity of copper dissolved
is difficult. Citric acid (C6Hs0 7 ) may also be used in
Another type of solvent that has been used chemical cleaning of power plant equipment.
s~ccessfully contains potassium bromate, Recirculation of a 3% solution of inhibited
shght amounts of sodium chlorate, ammonia citric acid at elevated temperature has been
and ammonium carbonate. The bromate is successful in removal of deposits from boilers
228 BETZ HANDBOOK
and other power plant equipment. The ease conditioning practices.

of removal depends on the composition and Although the technique of acid cleaning
quantity of deposit. The addition of a wetting has progressed to a point where there is little
agent to the cleaning solution may hasten danger from corrosion during the cleaning
removal of deposits. Where a boiler deposit process, it is advisable that the procedure
is found to be particularly resistant, the addi- be performed under the supervision of ex-
tion of ammonia to the citric acid solution in perienced personnel. During acid cleaning
sufficient quantity to elevate the pH, tends to every safety precaution must be exercised.
speed removal. In recent years, citric acid has The ever present danger of hydrogen ex-
been used successfully in cleaning of stain- plosion exists because hydrogen is evolved in
less steel superheater tubes of utility boilers. the reaction of acid with iron. In pockets
Citric acid may be used for removal of where hydrogen mixes with air, any spark or
mill scale from new boilers. Prior to use of open flame will ignite the mixture. Other
the acid the boiler should be given an alka- gases may also be generated during the clean-
·Iine boil-out to remove oil, grease, protective ing process, such as hydrogen sulfide. There-
coatings and any dirt accumulated during fore, adequate venting is essential.
erection. Following draining and flushing of Personnel handling any acid as well as
the boiler a 3% inhibited citric acid solution most inhibitors, should be fitted with proper
at elevated temperature is added and circu- safety clothing, including protective goggles
lated throughout, until all mill scale is re- and rubber gloves. Protective masks are also
moved. Periodically the iron content of the necessary for some vapors produce a suffo-
cleaning solution is checked to determine ef- cating effect and irritate the mucous mem-,
ficiency of the cleaning action. branes.
REFERENCES
CONCLUSION
S. Alfano, "Cleaning The Iron System Using Am-
Deposits of oil and grease as well as boiler moniated Citric Acid", Combustion, Vol. 33, pp.
scale, iron oxide and copper deposits may be 17-24 (11ay 1962)
removed from the boiler metal surfaces by P. H. Cardwell and L. H. Eilers, "Use of Wetting
Agents in Conjunction with Acid Inhibitors", Ind.
chemical cleaning. When employing any of and Eng. Chern., Vol. 40, pp. 1951-1956 (1948)
the chemical cleaning methods the proper "Chemical Cleaning of Internal Heating Sur-
technique must be followed and the neces- faces", Bulletin G-39A, The Babcock & Wilcox
Company, New York, N.Y. (1946)
sary precautions taken. With regard to boiler C. 11. Lqucks, "Chemistry Tackles Plant 11ainte-
scale, it should be recognized that acid clean- nance", Chemical Engineering, Vol. 69, pp. 103-
ing is not a substitute for proper feedwater 120 (11ar. 5, 1962)
31
Cooling Water Treatlllent:
Once-Through Scale Control
Scale Formation. . . . . . . . . . . . . . . . . . . . . . . . . 230

Prevention of Calcium Carbonate Scale . . . . . . . . 232
Iron Deposits . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
230 BETZ HANDBOOK
HE term "cooling water" is applied "scale" applies to deposits which result from
T wherever water is circulated through
equipment to absorb and carry away heat.
crystallization or precipitation of salts from
solution. Wasting away of a, metal is the re-
This definition includes air conditioning sys- sult of corrosion. While a deposit results in
tems, engine jacket systems, refrigeration both cases, the mechanism of formation is
systems as well as the multitude of industrial different and different corrective methods are
heat exchange operations, such as found in required to prevent the deposit.
oil refineries, chemical plants, steel mills, etc. Calcium carbonate is usually the chief in-
The once-through system, as the name gredient of scale formed in once-through
implies, is one in which the water is passed cooling water systems. Calcium carbonate is
through the heat exchange equipment and more insoluble than calcium silicate and
the cooling water is then discharged to waste. calcium sulfate and this fact accounts for its
Usually, a once-through system is employed more frequent appearance in the form of
only where water at suitably low temperature scale. Figure 31-1 illustrates analyses of sev-
is. readily available in large volume and at eral typical scales formed in once-through
low cost. The usual source of once-through cooling water systems. It can be noted that
cooling water is from wells, rivers and lakes the combination of calcium and carbonate
where the cost involved is that of pumping accounts for the major portion of each de-
only. In a once-through system, no evapora- posit.
tion takes place and consequently the water The factors which affect scale formation
does not concentrate. Circulating water char- are temperature, rate of heat transfer, the
acteristics are the same as the makeup water. calcium, sulfate, magnesium, silica, alkalinity,
dissolved solids and pH of the water. In a
once-through system, the magnesium, silica
SCALE FORMATION
and sulfate concentrations are normally suf-
Deposits in lines, heat exchange equipment, ficiently low that these factors do not enter
etc., may originate from several causes. For into the problem.
example, the precipitation of calcium carbon- In 1936, the work of Prof. Langelier was
ate will form scale, but products of corrosion published dealing with the conditions at
also result in a deposit of iron oxide. In speak- which a given water is in equilibrium with
ing of deposits which form in cooling water calcium carbonate. The use of the equation
systems, it is Important to bear in mind the developed by Langelier made it possible to
mechanism causing the deposit, otherwise predict the tendency of natural or condi-
confusion may result. In general, the term tioned water to either deposit calcium car-
Figure 31-1-Analysis of Typical Scales from Once-Through Cooling Water System
% % % %
Calcium as CaO 47.2 34.4 43.5 52.3
Magnesium as MgO 2.1 1.7 0.6 4.1
Iron as RzOa 3.1 22.7 10.72 3.00
Aluminum as Al 2 0a 1.5 4.2 0.88 0.20
Carbonate as C02 34.2 22.9 34.0 35.2
Sulfate as SOa 0.9 0.1 1.4 0.3
Silica as Si02 4.9 9.8 2.1 1.1
Loss on Ignition 5.9 4.1 6.4 2.1
'C' SCALE
......
2.7 2.6 2.5 2.4 2.3 2.2 2.1 2.0 1.9 1.8 1.7 1.6 1.5 1.4 1.3 1.2 1.1 ~
cUl
pCa >-1
:::0
......
5ooo I I I I Ill I I I I I\ I I I I I I ~ I I I I I 1\ I I I I I ft I I I I l I I I Ill I I I Ill I I I ~ I I I l I I I I I I I I I I l'ol I 1\ I I II I I ~ I I I I A I I A 5000 >
t"'
:::: IIIII NII II Nlllllllllll M1111\1111\111 NIll ill IN IIIlii NII NII NIf.J Itllltll II:::: ~
>
>-1
t;::l
:::0
z 1000 1000 "tl 0
0 > 0
::::; ::u zt;
.... -1
......
i 500 500 (II
>-1
~ 300
...
"tl ......
0
Q.
300 ::u z......
en 200
.....
200
....!,... z0
a::
<
Q.
0
100 100 z
50 50
,_-
30 30
20 20
10 10
4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0
pALK AND pCa SCALE
I
pALK
To DETERMINE:
pea: LOCATE PPM VALUE FOR Ca AS CaC0 ON THE PPM SCALE. PROCEED pCa = 2.70
3
HQRIZONTALLY TO THE LEFT DIAGONAl liNE DOWN TO TH~ pCa SCALE. pALK= 2.SO
ExAMPLE:
pALK: LOCATE PPM VALUE FOR 'M'ALK AS CaC03 ON THE PPM SCALE. PROCEED C AT 140F = I.S8
TEMP.= 140F. pH= 7.80 SUM= pHs = 6. 76
HORIZONTAllY TO THE RIGHT DIAGONAl LINE DOWN TO THE pALK SCALE.
Ca HARDNESS= 200 PPM ACTUAL pH = 7.80
TOTAL SOLIDS: LOCATE PPM VALUE FOR TOTAL SOLIDS ON THE PPM SCALE. PROCEED M AlKALINITY: 160 PPM DIFFERENCE = + 1.04
HORIZONTALLY TO THE PROPER TEMPERATURE LINE AND UP TO THE 'c' SCALE. TOTAL SOLI OS= 400 PPM = SATURATION INDEX
N
Figure 31-2 • Langelier Saturation Index Chart <.>:>
232 BETZ HANDBOOK
bonate from solution or to dissolve calcium While a high hardness water, with a positive
carbonate with which the water is brought Saturation Index will definitely lead to cal-
in contact. The equation takes into considera- cium carbonate scale formation, a low hard-
tion readily obtained analytical values such ness water, with the same positive Saturation
as pH, calcium, total alkalinity, dissolved · Index, may not form any appreciable calcium
solids and temperature. carbonate scale.
In simplified form, and for waters in the In an attempt to secure a quantitative
pH range of 6.5-9.5, Langelier's equation for index, Ryznar has proposed the Stability
the pH at which the water is in equilibrium Index. This index is empirical, based on a
with calcium carbonate (pHs) may be study of actual operating results with waters
written as follows: of various saturation indexes.
pHs = (pK2 t - pKs 1 ) + pCa + pAik Stability Index = 2(pHs) - pH
The two latter terms of the above equa- With waters having a Stability Index of
tion are negative logarithms of the molal 6.0 or less, scaling increases and the tendency
··· a~d equivalent concentrations of calcium and to corrosion decreases. When the Stability
titratable base respectively. The terms pK 2 1 Index is above 7.0, a protective coating of
and pKs1 are respectively the negative loga- calcium carbonate scale may not be de-
rithms of the second dissociation constant for veloped. Corrosion would become an increas-
carbonic acid and the activity product of ingly greater problem as the Stability Index
calcium carbonate. The difference (pK 21 - increases above 7.5 or 8.0. In some cases, the
pKs 1 ) varies with ionic strength and tem- use of the Stability Index along with the
perature. The term pHs represents the pH Saturation Index will help in predicting more
at equilibrium assuming no change has taken accurately the scaling or corrosive tendencies
place in the water bringing it to equilibrium. of a water.
The Saturation Index is defined as the
algebraic difference between the actual meas- PREVENTION OF CALCIUM CARBONATE SCALE
ured pH of the water and the calculated
pHs at saturation with calcium carbonate. Because of the large volumes of water used
Saturation Index = pH - pHs in once-through systems, external softening
This index shows qualitatively the tend- of the raw water is not usually economically
ency for deposition or solution of calcium feasible as a means of scale prevention. How-
carbonate. A positive Saturation Index will ever, a softening plant may be installed to
indicate a tendency to deposit calcium car- supply softened water for a variety of plant
bonate. A negative Saturation Index will purposes which may also include some once-
indicate an undersaturation condition with through cooling uses. In such cases, the
respect to calcium carbonate and hence a treated water balances are controlled so as
tendency to dissolve any existing calcium car- to provide a Saturation Index that will pre-
bonate. A "zero" Saturation Index denotes vent calcium carbonate deposition at the tem-
that the water is exactly at equilibrium with peratures involved.
respect to calcium carbonate. However, once-through cooling water is
Langelier's Equation has been slightly generally not softened by external treatment
modified by other investigators. Various nom- processes even though such softening facil-
ographs have been prepared to simplify the ities may be installed for other plant uses.
calculation of the pHs. A nomograph suit- The chemical treatment cost in softening the
able for this purpose is shown in Figure 31-2. large volumes of once-through cooling water
It should be emphasized that the Saturation is usually prohibitive. The investment cost in
Index is only a measure of the directional the treatment plant for handling such large
tendency or driving force of a water. It is in volumes of water is also another factor dis-
no way intended to be a measure of capacity. couraging such softening.
The most common method employed to systems of this type, the feed water of definite
inhibit calcium carbonate scale formation in composition is pumped at a constant rate
once-through cooling water systems is the through a scaling chamber, and scale forma-
use of various anti-nucleating agents. Except tion develops on a cartridge heater. Temper-
under unusual conditions, these materials ature, rate of flow and electrical input are
provide very satisfactory inhibition of scale automatically controlled.
formation, at a low chemical treatment cost. Consistent results are obtained with the use
An anti-nucleating agent employed for this of the same type of heating surface. Figure
putpose is one which possesses the property 31-6 illustrates a system of this type.
of preventing crystal growth and thereby scale
Figure 31-3 is illustrative of the data se-
formation. In effect, these agents broaden
cured under such controlled conditions in the
or increase the solubility of the scale forming
prevention of calcium carbonate scale forma-
salts and permit an oversaturated condition
tion. While most of the starches tested ex-
to exist without precipitation from solution.
hibit relatively low power in preventing cal-
The materials most commonly used include
ciun'l carbonate scale under test conditions,
the polyphosphates, tannins, lignins, starches,
appreciable power is shown by some of the
polyacrylates and combinations of these ma-
tannins and lignins. The most effective re-
terials. By blending several of the various
sults have been secured with blends of sev-
agents it is possible to secure the additive
eral different agents. For example, some
benefits of each for application to that par-
agents exhibit very high scale prevention
ticular problem.
powers per ppm in the low treatment range
In addition, chelants can be used in certain of 1-5 ppm. Increasing amounts of treatment
types of once-through cooling water systems above this range does not result in any ap-
for the purpose of chelating metal ions, such preciable increase in scale prevention powers.
as calcium and magnesium. Chelants react Other materials show practically a straight
stoichiometrically, and the reaction products
are usually very stable. Typical agents of this
type are ethylenediamine tetraacetic acid, Fig. 3.1-3
citric acid and gluconic acid. Effect of Surface Active Treatments in
Crystal distortion has been advanced as Reducing Calcium Carbonate Scale
one of the main forces in reducing carbonate
scale by the use of anti-nucleating agents. Heat
Treatment Type Transfer Rate CaC03 Scale % Scale
Figure 31-4 illustrates crystals of pure cal- Cone., ppm Organic Btu/1! 2 /hr Grams/1!2 Reduction
cium carbonate at 450 magnifications. Figure 0 16,000 26.8
31-5 is a photomicrograph of crystals of cal- 50 Starch A 16,000 23.3 13
cium carbonate formed in the experimental 50 Starch B 16,000 26.8 0
50 Starch C 16,000 18.2 32
laboratory system shown in Figure 31-6. The
crystal distortion shown by Figures 31-7, 31-8 0 24,000 18.7
and 31-9 illustra,tes the effect of various treat- 50 Tannin A 24,000 16.4 12
50 Tannin B 24,000 6.6 65
ments. Laboratory correlation of crystal dis- 50 Tannin C 24,000 12.3 34
tortion with the amount of scale actually
formed on the heat transfer surface indicates 0 17,000 34.6
that those which do not produce crystal dis- 50 Lignin A 17,000 25.6 26
50 Lignin B 17,000 38.1 -lQ
tortion also do not reduce the amount of 50 Lignin C 17,000 10.7 69
scale formation.
0 17,000 37.7
Study of the inhibitory effects of various 2 Blend A 17,000 18.1 52
agents for the prevention of calcium carbo- 25 Blend B 17,000 5.6 85
nate deposition can be made with relatively 25 Blend C 17,000 2.7 93
50 Blend D 17,000 0.4 99
simple once-through laboratory systems. In
234 BETZ HANDBOOK
Figure31-4 • Pure CaC0 3 (Merck), 450 Magnifications Figure 31-7 • Distorted CaCOa Crystals with Tannin
Treatment
Figure 31-5 • CaCOa Crystals (17,000 Btu/Ft2/hr), Figure 31-8 • Distorted CaC0 3 Crystals with
450 Mag. No Treatment Polyphosphate Treatment
Figure 31-6 • Once- Through Cooling Water Figure 31-9 • Distorted CaC0 3 Crystals with
Experimental Unit Polyphosphate- Tannin Treatment Blend
line relationship with scale properties in- ing iron bearing well waters. This precipita-
creasing in direct proportion to the concen- tion of iron can be prevented by the use of
tration of treatment employed. By combin- sequestrants or dispersants, which are very
ing two or more such agents into a blended specific in their activity on ferrous or ferric
treatment, it is possible to secure their addi- ions. Dispersants are adsorbed on the iron
tive benefits over a wide range of treatment particles while the iron is still in the ionic
concentrations. The results in Figure 31-3 for form. Residual charges on the adsorbed parti-
such a combination treatment indicate the cles are similar, and hence cause a stable,
high degree of scale preventive power that dispersed condition. Polyphosphates, lignin
can be incorporated into a single blended derivatives and organic phosphates are ex-
treatment. amples of treatments employed for this pur-
With the use of anti-nucleating agents, it pose.
is possible to secure complete prevention of Iron deposits also result from corrosion
calcium carbonate precipitation unless the products. In this case, the basic problem is
water is highly supersaturated with respect prevention of the corrosion, although poly-
to calcium carbonate at the temperature of phosphates may be employed to minimize the
use. In such cases, it is necessary to decrease "red water" problem, as discussed more fully
the positive Saturation Index, usually through in the next chapter.
the use of acid treatment. Complete elimina-
tion of the positive Saturation Index is not REFERENCES
required under such circumstances, but only H. L. Kahler, "Once-Through and Recirculating
enough adjustment in this index to bring the Cooling Water Studies," Proceedings, Engineers
water thus treated within the effective range Soc. of Western Penna., pp. 39-62 (1944)
W. F. Langelier, "The Analytical Control of Anti-
of surface active agents. Corrosion Water Treatment," Journal, Am. Water
Works Assoc., Vol. 28, pp. 1500-1521 (1936)
IRON DEPOSITS J. W. Ryznar, "A New Index for Determining
Amount of Calcium Carbonate Scale Formed by a
A special case of scale formation from fer- Water," Journal, Am. Water Works Assoc., Vol.
36, pp. 472-486 (1944)
rous bicarbonate may be encountered on J. M. Donohue, "Organic Topping Agents," South-
once-through cooling water systems employ- ern Engineering, Vol. 85, pp. 52-54 (Jan. 1967)
236 BETZ HANDBOOK
Causes of Corrosion . . .......... /' . . . . . . . . . . 237

Water Characteristics. . . . . . . . . . . . . . . . . . . 237
Effect of Temperature ................ ··: : 238
32 Effect of Flow Rate . ..... : . . . . . . . . . . . . . . 238

Dis~imilar Metals . .................... ; 238
Cooling Water Treatment:

Once-Through
Corrosion Control
Prevention of Corrosion . ................... 240
CaC0 3 Protective Scale. . .. . . . . . . . . . . . . . . . 240
Mechanical Deaeration. . . . . . . . . . . . . . . . . . 240
Chemical Deaeration. . . . . . . . . . . . . . . . . . . . 240
Corrosion Inhibitors . .................... 241
Polyphosphates. . . . . . . . . . . . . . . . . . . . . . 241
Silicates. . . . . . . . . . . . . . . . . . . . . . . . . . . 242
Dian odic-:(inc Dianodic . . . . . . . . . . . . . . 243
C
o~RosroN of ferrous metals in. distribu- 0>0
tiOn systems, once-through coolmg water

systems, plant piping, etc. is a serious prob- "" v
I
lem from several standpoints. Not only is the 0.20 /
line replacement costly in itself, but the labor !"
cost of removing and replacing buried lines,
or those imbedded in walls, is very high. Cor-
"" ,v <f)Y
/
/
v
0.10
rosion products also reduce the carrying ca- / v

pacity of lines, increase frictional resistance
and pumping costs. The iron oxide formed as
a result of corrosive action occupies a volume
000
~ r:.--
I
/
---
[._.....----
I--
' 4
-
PPM OXYGEN
~
'
- 48 F
----
10
many times greater than the original metal

and eventually may result in complete block- Figure 32-1 • Effect of Oxygen Concentration on Corro-
age. "Red water," due to corrosion products, sion at Different Temperatures
gives an unsightly appearance to the water
and may interfere with processes sensitive to geri at the cathodic surface, forming water
iron. and depolarizing the surface, thus permitting
additional iron to dissolve. Figure 32-1 shows
CAUSES OF CORROSION the effect of oxygen concentration on the cor-
Corrosion is defined as the destruction of a rosion of low carbon steel specimens at dif-
metal by chemical or electrochemical reaction ferent temperatures. Philadelphia (Delaware
with its environment. In the corrosion of fer- ·supply) tap water was used in these tests,
rous metals in contact with moisture, rever- with a pH of 6. 7. Linear rate of flow past
sion of the metal to the oxide form takes the test specimens was 2.5 feet per second.
place. The driving force or the tendency to The chief s~urce of dissolved oxygen in
corrode is electrochemical. However, the rate water is contact with the atmosphere. Deep
at which corrosion proceeds is dependent on well waters are usually devoid of oxygen, while
the resistance to continued attack caused by most surface supplies are saturated with oxy-
the products of the corrosion reactions. gen. The solubility of .oxygen in water is de-
In water systems, the principal factors in- pendent on the temperature and atmospheric
fluencing the corrosion of ferrous metals are pressure. Increased pressure increases oxygen
the water characteristics, temperature, rate of solubility, while increase in temperature low-
water flow and contact with dissimilar meters the sol~bility of oxygen.
als. The chief variables controlling the cor- Generally, oxygen produces an easily iden-
rosive clta:racteristics of a water are its dis- tified corrosion in the form of small pits or
solved oxygen concentration, carbon dioxide depressions. As corrosive attack continues, the
content, pH and dissolved solids. Other fac- pit usually increases, not only in area but also
tors are involved) of course, such as the pos- in depth, and the nodule of corrosion prod-
sible presence of free mineral acid, hydrogen ucts becomes enlarged. The detrimental ef-
sulfide, sulfur dioxide, etc. In general, how- fects of corrosive attack are more severe when
ever, dissolved oxygen and carbon dioxide are this corrosion is localized in the form of pits
the chief problems, with pH directly affected than would be the case if general metal wast-
by the carbon dioxide concentration. In- age occurred at the same rate over a large
creased dissolved solids, particularly chloride area. The centralization of corrosion at the
and sulfate, enhance the corrosive effect of points of pitting permits deeper penetration
oxygen and carbon dioxide. of the metal and more rapid failure at these
WATER CHARACTERISTICS. Dissolved Oxygen. points.
Oxygen attacks ferrous metals, causing cor- pH. At pH values below approximately 4.3
rosion and pitting. Oxygen reacts with hydro- (the region of free mineral acidity), the chief
238 BETZ HANDBOOK
controlling factor in promoting corrosion is anodic to the colder areas.

pH. Acid waters, such as those contaminated EFFECT OF FLow RATE. Increase in the rate
with acid mine drainage, are quite corrosive. of flow tends to increase the rate of corro-
Neutralization of such acidic characteristics sion. Figure 32-2 illustrates data secured
is a necessity for practically every water use. through the exposure of Y4" tubular low car-
However, in the range of most natural wa- bon steel specimens to Philadelphia (Dela-
ters, between pH 6.0-8.0, the pH of the water ware supply) tap water at a temperature of
is not the controlling corrosive factor, in the 120 F. The corrosion rate increases rapidly
presence of oxygen. with increases in the rate of flow although at
Carbon dioxide concentration, pH and al- higher flow rates this influence is less marked.
kalinity of a water are directly related. For a Similar effects have been noted on tests with
given alkalinity, the higher the carbon diox- Pittsburgh, Pa. tap water and with sea water.
ide, the lower the pH. In the absence of oxy- The influence of flow rate on corrosion is
gen, as encountered with the use of deep well probably primarily the depolarization effect
wa.ter, pH is a controlling influence on the secured by bringing larger quantities of oxy-
corrosion rate. At lower pH values, corrosion gen into contact with the cathodic surface,
increases. Carbon dioxide is a factor in such together with the effect of greater amounts
corrosion since it influences pH. In the of ions (such as chloride and sulfate) on the
presence of oxygen, however, carbon dioxide anodic surface.
like pH is not the major influence on the rate
of corrosion. 0.35
Dissolved Solids. In general, increase in the

a:
dissolved solids content of a water increases ~ 0.30
)o
the corrosivity. Higher solids content increases a:
lU
the conductivity of the water, thus favoring ... 0.25
increased corrosion. The chloride ion, in C/)
lU
:r
particular, accelerates corrosive action al- ~ 0.20
though at very high concentrations (brine) a z"
0
decrease in corrosivity is observed. Various ~ 0.15
explanations have been advanced for the cor- ....lUa:
rosive effect of higher dissolved solids and z
~ 0.10
one logical theory involves the effect of ions
lU
in destroying the coating of corrosion prod-
ucts on the metal. High chloride concentra- "":5 0.05
tions, for example, also hinder the formation ~
of protective coatings by inhibitors, such as o.o 0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
the chromates. FLOW RATE, FEET PER SECOND
EFFECT OF TEMPERATURE. The chemical re-
actions of corrosion increase in rate with in- Figure 32-2 • Influence of Flow Rate on Corrosion
crease in temperature. Reference to Figure
32-1 shows the accelerated corrosion at 90 F DisSIMILAR METALS. The corrosion that oc-
and 120 F, in comparison with 48 F for vari- curs from the contact of dissimilar metals is
ous oxygen concentrations. Increased tem- known as galvanic action. The tendency of a
perature decreases water viscosity and thus in- metal to dissolve and enter solution as an ion
creases the diffusion rate of oxygen. Elec- is measured by the electrode potential between
trical conductivity is increased at higher tem- the metal and its ions. The listing of the elec-
peratures, thus producing an effect similar to trode potentials of the various metals consti-
increase in dissolved solids content. Pitting tutes the familiar Electromotive Series. The
may be enhanced where adjacent metal areas electrode potential of hydrogen is defined as
vary in temperature, with the hotter sections zero. The metals higher in the table are con-
sidered able to displace those metals lower in Fig. 32-3

the table.
Galvanic Series of Metals and Alloys*
The Electromotive Series cannot be inter- CoRRODED END
preted as a precise indication of galvanic ac-
(anodic, or least noble)
tion. The potentials of the various metals can
be altered by ion concentration, temperature, Magnesium
and the corrosive environment. Consequently, Magnesium alloys
in practical service conditions, it is possible
Zinc
to 'reverse the order in which some adjacent
metals are listed in the Electromotive Series. Aluminum 2S
Figure 32-3 lists a galvanic series of metals Cadmium

and alloys, based bn corrosion testing experi- Aluminum 17ST
ence. While closely allied to the Electromotive Steel or Iron
Series, it is based on practical experience Cast Iron
rather than theoretical consideration alone.
It will be noted that some of the metals are Chromium-iron (active)
grouped together. Metals so grouped possess Ni-Resist
little tendency to produce galvanic corrosion 18-8-Cr-Ni-Fe (active)
of each other and from a practical standpoint 18-8-3-Cr-Ni-Mo-Fe (active)
are relatively safe to use in contact with each
other. However, coupling of two metals from Hastelloy C
different groups and distant from each other Lead-tin solders
in the list will result in galvanic, or accele- Lead
rated, corrosion of the one higher in the list. Tin
The tendency for galvanic corrosion is in- Nickel (active)
creased when the metals coupled together are
far apart in the list. While such a galvanic lnconel (active)
series is a useful guide, it should not be in- Hastelloy A
terpreted as an absolute rule under all cor- Hastelloy B
rosive service conditions. Brasses
Corrosion of ferrous metals in the presence Copper
of moisture is a truly vast subject and it is Bronzes
possible to only touch upon the various influ- Copper-nickel alloys
encing factors in this discussion. Many other Titanium
variables will affect the corrosion rate, such Monel
as the homogeneity of the metal, condition of Silver solder
the metal surfaces, ratio of surface areas ex-
posed by two dissimilar metals, static and cy- Nickel (passive)
cling stresses, oxygen concentration cells, etc. lnconel (passive)
Water characteristics also are influenced by Chromium-iron (passive)
factors such as hydrogen sulfide, chlorine, and 18-8 Cr-Ni-Fe (passive)
biological influences which can markedly in- 18-8-3 Cr-Ni-Mo-Fe (passive)
fluence the corrosion rate. The technical lit- Silver
erature on corrosion is quite voluminous. In Graphite
addition, several excellent books on this spe-
cific subject are included among the chapter PROTECTED END
references and should be consulted by those (cathodic, or most noble)

wishing to delve more deeply into this com- • "Corrosion Testing Bulletin," International Nickel Com-
plex subject. pany, 1944.
240 BETZ HANDBOOK
PREVENTION OF CORROSION duces an oversaturated condition at the point

of higher temperature. If the calcium carbo-
The prevention of corrosion in once-through nate balance is so set th(l.t only a small
systems is accomplished either by forming a amount of deposit will form at the higher
calcium carbonate scale on the metal surface temperature, none will form at the lower
to act as a protective coating against corro- temperature and no protection against cor-
sion, by elimination of the corrosive oxygen rosion will be provided at that point. In ad-
itself from the water, or by the use of rela- dition, within the limits of plant control, it
tively small quantities of organic and inor- is exceedingly difficult to control scale forma-
ganic corrosion inhibitors. tion to such an exact point as to provide pro-
CALCIUM CARBONATE PRoTECTIVE ScALE. The tection against corrosion and yet not interfere
Langelier Saturation Index is useful in pre- with heat transfer.
dicting the tendency of a water to deposit or The controlled formation of calcium car-
dissolve calcium carbonate. The Saturation bonate scale as a means of corrosion preven-
... Index can be used as a guide in the controlled tion is successfully employed in some water
deposition of calcium carbonate scale. If such works distribution systems where increase in
a scale can be deposited uniformly over the temperature is not encountered. For indus-
metal surfaces throughout the system, it will trial cooling water systems, however, where
act as a physical barrier, preventing contact heat transfer is involved, this method of cor-
of the water with the metal surface. The corrosion prevention is impractical, except under
rosive oxygen content of the water cannot ideal control conditions.
contact the metal and corrosion is prevented. MECHANICAL DEAERATION. The corrosive
In effect, a protective coating similar to qualities of a water can usually be practically
cement lining of the pipe has been forn!ed. eliminated by oxygen removal by deaeration.
In order to develop the positive Saturation Vacuum deaeration for once-through cooling
Index required for deposition of calcium car- water systems has been applied successfully
bonate, it is necessary to adjust the pH, to a number of installations. Where oxygen
alkalinity or calcium content of the water so removal is not complete, the remaining traces
as to raise the pH slightly above the calcu- of oxygen can be removed chemically with
lated pH of saturation. The materials em- catalyzed sodium sulfite.
ployed for this purpose may be soda ash, Of course, there is considerable investment
caustic soda and lime (calcium hydroxide) . cost, as well as continued operating cost, in-
Lime is usually the most economical alkali volved with vacuum deaeration. These fac-
to employ since it raises the calcium content tors have hindered the more widespread use
as well as the alkalinity and pH. A lesser of this method. Under certain circumstances,
amount is thus required to develop the pH however, vacuum deaeration is unquestion-
of saturation. ably the most effective means of corrosion
Theoretically, it is possible to control the control and gradual extension of its use is to
deposition of calcium carbonate scale to such be expected in the future.
a point that protection against corrosion is CHEMICAL DEAERATION. Sodium sulfite is
provided, and the scale film is sufficiently thin commonly used in the treatment of boiler
to be non-objectionable. However, such a feedwater as a chemical deaerant and reacts
treatment method is more theoretical than with dissolved oxygen as follows:
practical inasmuch as a definite positive Sat-
uration Index of a certain value pertains only Na 2 S03 + 0 Naz$04
for the temperature for which it has been
sodium sulfite + oxygen = sodium sulfate
calculated. With different temperatures This reaction proceeds rapidly at tempera-
throughout a cooling water system, the main- tures of 212 F and above. However, at cold
tenance of a protective film on the metal sur- water temperatures, the reaction between so-
face at the point of lowest temperature pro- dium sulfite and dissolved oxygen is quite
slow. The slowness of this reaction has pre- sults in a once-through system as it does in a
vented the use of sodium sulfite in the chem- recirculating system. However, the cost of the
ical deaeration of cold water. In a once- chemical treatment would be so high as to
through system, for example, with both oxy- be impractical. The chief problem in corro-
gen and sulfite existing unreacted, the treated sion prevention in once-through systems is
water may pass completely through the sys- not lack of satisfactory corrosion inhibitors,
tem before any appreciable oxygen removal but rather the difficulty of securing corrosion
has taken place. Of course, under these cir- protection at the low treatment concentra-
cumstances, no reduction will take place in tion economically permissible in treating such
the corrosive qualities of the water. large water flows. For this reason, the corro-
The development of catalyzed sodium sul- sion inhibitors used are fed usually only in
fite has overcome the slowness of the oxygen- sufficient quantity to minimize or control
sulfite reaction at cold water temperatures severe corrosion. Only rarely, is sufficient
and makes technically practical the removal treatment applied to stifle corrosive attack.
of oxygen in once-through cooling water sys- The reasons for this practice are not techni-
tems. In general, oxygen removal is complete cal, but economic.
within 20 seconds contact time, as discussed Polyphosphates. Polyphosphates are a special
under the subject of boiler corrosion control. class of phosphates, corresponding to molec-
Naturally it is necessary to feed the catalyzed ularly dehydrated orthophosphate. Unlike
sodium sulfite in direct proportion to water orthophosphates, the polyphosphates possess
flow in order to continuously achieve com- marked surface active properties, useful in
plete oxygen removal. the control of scale formation and in mini-
Because of the relatively large amount of mizing tuberculation. Examples of polyphos-
catalyzed sodium sulfite required for oxygen phates are sodium tripolyphosphate, sodium
removal ( 10 parts catalyzed sodium sulfite hexametaphosphate and sodium decaphos-
for each 1 part of dissolved oxygen) , the use phate.
of chemical deaeration is restricted by eco- Polyphosphates are widely used in the
nomic factors. In once-through systems using treatment of once-through systems, particu-
large volumes of water, vacuum deaeration larly city distribution systems. Remarkable
may prove more economical. In smaller sys- results can be secured from the use of as
tems, the investment cost in vacuum deaera- little as 2 ppm polyphosphate, a concentra-
tion may not be warranted and the use of tion usually employed in these systems. The
catalyzed sodium sulfite may be justified. Usu- polyphosphates are quite effective in reducing
ally, catalyzed sodium sulfite is well justified tuberculation in distribution lines and in
following vacuum deaeration to remove the minimizing "red water" troubles. However,
last traces of oxygen at a nominal cost. these results are secured without appreciable
CoRROSION INHIBITORS. The corrosion inhib- reduction in corrosion rate.
itors most widely employed in industrial In once-through systems, and with the use
once-through cooling water systems are the of such low treatment concentrations, flow
polyphosphates, silicates, the Dianodic com- rate is quite important since this is the factor
bination of phosphates and chromates, and that determines the supply of treatment to
the Zinc Dianodic combination of zinc, phos- the metal surface. Dependent upon the flow
phate and chromate. Other agents are used rate selected for test purposes, wide variations
on occasion. In fact, from a technical stand- are secured in the corrosion inhibition prop-
point only, any of the methods of corrosion erties of the polyphosphates.
inhibition employed in recirculating systems Figure 32-4 shows results secured in an
can be used satisfactorily in once-through sys- experimental laboratory once-through system
tems. For example, the use of 300-500 ppm employing Philadelphia (Delaware supply)
chromate would provide as satisfactory re- tap water at 160F and 3.8 feet per second
242 BETZ HANDBOOK
20 the carrying capacity of the lines and making

necessary periodic mechanical or chemical
cleaning. The loss in head caused by tuber-
culation requires increased p~mp pressures
15 and causes higher pumping costs.
z Figure 32-5 illustrates results obtained over
0
1- several months operation with untreated
~ 10 water and water treated with 2 ppm poly-
1- phosphate. Results were obtained at a flow
0
a:: rate of 3.2 feet per second. The loss in head
a.
with the untreated water is an indication of
the effect of tuberculation in increasing the
resistance to flow. The use of 2 ppm poly-
phosphate resulted in only minor head loss,
equivalent after eight months operation to
2 4 6 8 only 10% of the head loss encountered with-
DAYS UNDER TEST out polyphosphate treatment. In these tests,
tuberculation was effectively minimized by
Figure 32-4 • Inhibition of Corrosion with 2 ppm the use of this minute amount of polyphos-
Polyphosphate phate.
Silicates. Sodium silicate has been employed
flow rate. Polyphosphate was employed at for many years to reduce corrosion in once-
2 ppm. Comparison of weight loss data with through systems. Considering the long period
the untreated water shows the treated run of time this material has been employed,
levelling off at a corrosion rate approximately there is surprisingly little quantitative data
20% below that of the untreated water. Re- 100
duction in metal loss by this low concentra-
tion of polyphosphate is not outstanding, but
the benefits secured in reducing tuberculation
and "red water" are greatly in excess of what 80
would be expected from this reduction in t
corrosion rate. 0
Q
The ability of the polyphosphates to form
complexes with iron naturally present in the
ffiQ. 60
water, or that resulting from corrosion, re- t
sults in retention of the iron in solution and 0
the prevention of "red water" difficulties. ~ 40
::c
Corrosion of ferrous metal piping results
:!':
in the formation of iron oxide corrosion prod- en
en
ucts many times more voluminous than the g
metal lost from the pipe wall. At the usual 20
2ppm POLYPHOSPHATE
pH values in distribution systems, tubercules
of corrosion products are considerably greater
than at low pH values, where there is a
greater solution effect from the more acidic
condition. Increase in pH value of the water, MONTHS IN SERVICE
short of pH 10.0-11.0, is not effective in re-
ducing tuberculation. Figure 32-5 • Reduction in Tuberculation by
Tuberculation is responsible for reducing Polyphosphate
in the technical literature concerning its cor- minor corrosion inhibitory properties for so-
rosion inhibition properties in water systems. dium silicate, as measured by metal loss.
Sodium silicate has been recommended as Without question, however, "red water" dif-
a corrosion inhibitor primarily for the treat- ficulties are reduced with the use of sodium
ment of relatively soft waters and based on silicate, although the manner in which the
increasing the silica content of the water by silicate functions is not entirely clear.
8 ppm as Si0 2 • Sodium silicate is available Dianodic-Zinc · Dianodic. The Dianodic sys-
in different grades with varying ratios of tem of treatment, utilizing the two anodic in-
Na2 0 to Si02 and consequently varying alka- hibitors, phosphate and chromate, was used
linity. The alkalinity supplied by sodium sili- to improve results obtained with polyphos-
cate is undoubtedly useful in neutralizing car- phates and silicates. Laboratory and plant re-
bon dioxide and increasing pH. sults showed 50%-60% reduction in overall
Figure 32-6 shows results in a laboratory corrosion and tuberculation. However, treat-
once-through system under the same condi- ment residuals of 12-16 ppm were just outside
tions used in securing the data for Figure of an economical treatment level for once-
32-4. Flow rate was 3.8 feet per second, through systems. It was therefore necessary
temperature was 160 F and Philadelphia tap to search even further for an acceptable
water was used. Liquid sodium silicate (28% treatment.
Si02 ) was fed at the rate of 100 ppm and it Addition of zinc to the two anodic inhibi-
will be noted that weight loss was decreased tors resulted in a cathodic-anodic system of
approximately 17%, as compared with an un- protection. The new three component treat-
treated condition. Sodium silicate at 100 ppm ment was used in the range of 8-12 ppm and
thus provided lesser corrosion protection than the data indicated 75%-90% reduction in
polyphosphate at 2 ppm, under these test corrosive attack. By applying the principles
conditions. of pretreatment for rapid film formation even
Similar data has been secured under dif- greater protection can be secured. Many
ferent test conditions, all indicating only plants have adopted the use of phosphate-
chromate-zinc for maximum protection of
once-through cooling water systems at this
20
acceptable cost level.
Figure 32-7 shows three groups of un-
cleaned and cleaned steel test specimens. The
15 two specimens on the left were exposed to a
z test water with no inhibitor present. In each
0 of the three groups of specimens the one on
;::
the left is uncleaned while the one on the
~ 10
1- right has been cleaned. With the untreated
0 water the uncleaned surface was coated with
Ct:
ll. a bright red oxide.
~ 5 The middle group of specimens was ex-
posed to the test water with 6 ppm poly-
phosphate present as an inhibitor. Less cor-
rosion was shown although tuberculation
2 4 6 8 developed as shown by the uncleaned speci-
DAYS UNDER TEST men on the left.
The third group of specimens, on the
Figure 32-6 • Inhibition of Corrosion with 100 ppm right, were exposed to the test water treated
Sodium Silicate (28% Si02) with 8 ppm of the zinc-phosphate-chromate
combination. The uncleaned specimen had
244 BETZ HANDBOOK
Figure 32-7 • Three Groups of Uncleaned and Cleaned Test Specimens Exposed to the Same Test Water with-No
Inhibitor Treatment- Polyphosphate-,Zinc-Phosphate- Chromate
Figure 32-8 • Enlargement of Cleaned Specimen with No Inhibitor Treatment

Figure 32-9 • Enlargement of Cleaned Specimen with Polyphosphate Treatment
Figure 32-10 • Enlargement of Cleaned Specimen with Zinc-Phosphate-Chromate Treatment

246 BETZ HANDBOOK
a characteristic greenish cast, sometimes a bare mmrmum of chromate salts in this

identified as interference colors. balanced three component system, thus re-
Figure 32-8 shows an enlargement of the ducing the objection in overall plant waste
cleaned specimen with no inhibitor treat- water. Ultimate dilution to negligible quan-
ment. Nearly complete attack of the speci- tities is the usual end result in nearly all
men is shown with only a small area in the cases. This treatment is not acceptable for
center and lower left exhibiting the original potable waters. Work still continues in the
streaked appearance of a new specimen. organic line to find an acceptable, as well as
Figure 32-9 is an enlargement of a section economical, treatment for corrosion control
of the specimen exposed to the polyphosphate of water used for human consumption.
treated test water. The pits beneath the tu-
bercles are clearly apparent. REFERENCES
Figure 32-10, inhibited by the zinc-phos-
Rhate-chromate combination, shows nearly U. R. Evans, "The Corrosion and Oxidation of
cbniplete reduction of the pitting function Metals: Scientific Principles and Practical Appli-
and little evidence of general corrosive at- cations", St. Martin's Press, Inc., New York, N.Y.
(1960)
tack.
R. J. McKay and R. Worthington, "Corrosion Re-
These laboratory results have been trans- sistance of Metals and Alloys", Reinhold Publish-
ferred directly to once-through cooling sys- ing Corp., New York, N.Y. (1936)
P. E. Pallo, "A Study of Corrosion Control With
tems that have experienced severe corrosion Sodium Hexametaphosphate", Journal, Am. Water
in the past. The use of Zinc Dianodic has Works Assoc., Vol. 38, pp. 499-510 (1946)
made it possible to bridge the gap between F. N. Speller, "Corrosion Causes and Prevention",
'fair' protection and the 'uneconomical' McGraw-Hill Book Co., Inc., New York, N.Y.
(1951)
treatments for once-through cooling water H. H. Uhlig, "The Corrosion Handbook," John
corrosion problems. It is also possible to use Wiley & Sons, Inc., New York, N.Y. ( 1961)
33
Cooling Water Treatn1ent:
Open Recirculating Scale
Control
Scale Formation. . . . . . . . . . . . . . . . . . . . . . . . . 249

Scale Prevention. . . . . . . . . . . . . . . . . . . . . . . . . 251
Control of Cycles . ............... : . . . . . . . . 252
Acid Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . 255
248 BETZ HANDBOOK
use of a recirculating system, in as 1% per 10 F temperature drop through

T HE
which a cooling tower, spray pond,
evaporative condenser and the like serve to
the tower.
The circulating water becomes more con-
dissipate heat, permits great economy in centrated than the makeup water due to
makeup water requirements. With dwindling this evaporation loss. Cycles of concentration
supplies of fresh cold water available for is the term employed to indicate the degree
industry's cooling requirements, increased of concentration of the circulating water as
use must be made of recirculating systems in compared with the makeup. For example,
which the cooling water is used over and 2.0 cycles of concentration indicates the cir-
over again. A typical flow sheet of an open culating water is twice the concentration of
recirculating system is shown in Figure 33-1. the makeup water.
After passage of the circulating water While the evaporation loss tends to cause
through the heat exchange equipment, the the water to concentrate, the windage or
water is cooled in passing over the cooling drift loss, which is the loss of fine droplets
tower. This cooling effect is produced by of water entrained by the circulating air,
evaporation of a portion of the circulating tends to limit the degree of concentration.
water in passing over the tower. By virtue of The amount of windage loss varies with the
the evaporation which takes place in cooling, type cooling tower used, but the following
the dissolved solids in the water become con- may be taken as typical windage losses based
centrated. The evaporation must be replaced on the rate of circulation.
by makeup water. The quantity of water lost Spray Ponds = 1.0%-5.0%
by evaporation can be calculated from th~
Atmospheric Towers = 0.3%-1.0%
conditions of the inlet and outlet air. The
evaporation loss will amount to approxi- Mechanical Draft Towers = 0.1 %~0.3%
mately 0.85%-1.25% of the rate of circula- The windage also represents a loss of water
tion of water over the tower per 10 F drop from the system and this likewise must be
of water temperature through the tower. For replaced by makeup. Therefore, the evapora-
practical purposes, this figure may be taken tion loss plus the windage loss is equal to the
GENERAL RHRIGER·
COMPRESSOR
POWER PLANT ATIDN
JACKETS
SERVICE CONDENSERS
VAPOR VAPOR LIQUID

COOLERS CONDENSERS CONDENSERS
COOLING TOWER
PUMP
MAKEUP
Figure 33-1 • Typical Flow Diagram of Open Recirculating Cooling Water System
e.0
t-.v~ ~~~'I ~~ ,JJ !o/ ..Y SCALE FORMATION
t-+-
E v1/ lJ In open recirculating systems, where the
ot-t-
IGJ, I 1/ v I/ 1/ ~~7 water is concentrated by evaporation, the
Ifjll vv v 1/ <tc;
~<;~!-I- problem of, scale formation is increasingly
rt,l/ 1/ v
troublesome. A water that possesses scale
/ vv forming tendencies under once-through con-
W' 1/vvv v v ditions becomes much more scale forming
VIJI' v 1/
0
v when concentrated. Even a water which is
V/,VJ.vv v v not scale forming on a once-through basis
1.0
j~ Yv usually will become scale forming when con-
centrated two, four or six times.
o.o 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0 In addition, while calcium carbonate scale
PERCENT WINDAGE LOSS
formation is the primary problem in recircu-
Figure 33-2 • Relationship between Evaporation Loss, lating systems as well as in once-through sys-
Windage Loss and Cycles of Concentration tems, it is also necessary to take steps to pre-
vent the deposition of calcium and magnesium
makeup required. silicate and calcium sulfate scales in recircu-
Both evaporation loss and windage loss are lating systems. Scales of this nature usually
expressed as percentage of the rate of circu- are not a problem under once-through con-
lation. As an example, if the temperature ditions.
drop across the tower were 20 F, an evapora- The curves in Figure 33-3 will serve to give
tion loss of 2.0% would result. If the tower some idea of the relative solubility of calcium
were the atmospheric type, with a windage carbonate compared with calcium sulfate.
loss of 1.0%, the total makeup would be While the various forms of calcium sulfate
3.0% of the rate of circulation. A circulation are illustrated, the one normally encountered
rate of 10,000 gallons per minute would re- in cooling. water systems is gypsum. Due to
quire a makeup of 300 gallons per minute, or the fact that calcium carbonate is consider-
approximately 430,000 gallons per day. ably less soluble than calcium sulfate, the type
The relationship between evaporation loss, scale formed in open recirculating systems is
windage loss and the cycles of concentration principally calcium carbonate in the great ma-
the water will undergo is illustrated by Fig- jority of cases. Typical analyses of scale sam-
ure 33-2. To illustrate the use of the curves, ples are shown in Figure 33-4.
assume a 20 F temperature drop through the The main factor controlling the increased
tower. This figure corresponds to an evapora- tendency for scale formation in open recircu-
tion loss of 2.0%. From the typical windage lating systems is the increased concentration,
losses illustrated for a spray pond, assume a due to cycles of concentration. Langelier's
windage loss of 2.0%. From the curves it will Saturation Index can be used as a guide to
be noted that 2.0 cycles of concentration will predict the tendency toward scale formation.
develop. If the tower were of the atmospheric Figure 33-5 shows makeup and circu-
type and a 1.0% windage loss were assumed, lating water analyses from a plant with a
with the same temperature drop through the maximum circulating water temperature of
tower, 3.0 cycles of concentration would re- 140 F. Calculation of the pHs at 140 F on the
sult. In the case of a mechanical draft tower makeup water yields a value of 7.8 resulting
with a windage loss of 0.3% and the same in a negative or non scale-forming Saturation
temperature drop, approximately 7.5 cycles Index of -0.8. The Stability Index of 8.6
of concentration would develop. It therefore also indicates a non scaling condition. These
can be seen that the type tower limits the conclusions would pertain if there were no
maximum cycles that can develop in a cool- concentration of solids, as in a once-through
ing water system. system.
250 BETZ HANDBOOK
3.000 Figure 33-4-Typical Analysis of Scales

2,800 ~ \ From Open Recirculating Systems
2,600 \ Source Condenser,.% Pipe Cooler, %
2!'100 \\ GYPSUM )\
1\ Calcium as CoO ............ 49.79 ......... 39.85
Magnesium as MgO ......... 2.42 ......... 2.20
ceaso.·2H,O)
Iron as Fe203 .............. 0.61 ......... 6.98
2,200 Aluminum as Al203 ......... 0.21 ......... 0.52
,.,- ~ l Carbonate as C02 .......... 39.00 ......... 30.20
v
2,000 Sulfate as S03 . . . . . . . . . . . . . 1.29. . . . . . . . . 0.50
"" ,\
Silica as Si02 .............. 0.15 ......... 3.85
1,800 \ '\
I'IEMIHYDRATE
c easo, • 'i H,o) loss on Ignition ............ 6.05 ......... 15.00
1,600
\
I~
z
0
31.400 1\ At 2.0 cycles of concentration, the pHs
:i
·~ 1,200 \ \ is 7.3, equivalent to the circulating water
pH. Stability Index of 7.3 also indicates a
~
Q. \
non scaling condition. At two cycles of con-
~ 1,000
ANHYDRITE centration, calcium carbonate scale would not
a:
it 800
ceaso.)
\ \
\ be expected.
\.
\ At 3.0 cycles of concentration, a positive
600
Saturation Index of 1.5 is obtained and a
~
400
"' ......__
!'-..
Stability Index of 5.3. These values indicate
an appreciable tendency toward calcium car-
""'""'
200 bonate scale formation. It would not be safe
to exceed 2.0 cycles of concentration with
10 0
this water, without employing some form of
50 ~ treatment. With organic and inorganic sur-
,5 CALCIUM CARBONATE CCaCO,)
0
32 68
~
104
140 176 212 248 284 320 356
face active agents, it would be possible to
maintain 3.0 cycles of concentration without
TEMPERATURE- f. calcium carbonate scale formation, but this
Figure 33-3 • Solubility of Calcium Carbonate Com- concentration could not be tolerated without
pared to Calcium Sulfate such treatment.
Fig. 33-5-Circulating Water Analyses

Circulating Water Circulating Water
Makeup at 2.0 Cycles at 3.0 Cycles
Total Hardness as CaC03 ............................... . 60 120 180

Calcium as CaC03 ......................... · ... · · .. · · · · 50 100 150
Magnesium as CaC03 .................................. . 10 20 30
P Alkalinity as CaC03 ................................. . 0 0 10
M Alkalinity as CaC03 ................................. . 50 100 150
Sulfate as S04 ................ · .. · · · . · · · · · · · · · · · · · · · · · · 40 80 120
Chloride as Cl ........................................ . 10 20 30
Silica as Si02 5 10 15
Above values expressed in ppm
pH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.0 7.3 8.3
pHs (140F) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.8 7.3 6.8
Saturation Index ... ,., . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . -0.8 0.0 +1.5
Stability Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.6 7.3 5.3
Interpretation ....................................... Non scale forming Nan scale forming Scale forming
SCALE PREVENTION hold true under recirculating conditions.

Those anti-nucleating agents which cause
The problem of scale prevention in open re- crystal distortion are the most eff~ctive in
circulating systems is similar to the problem decreasing scaling tendencies. Figure 33-7
in once-through systems. However, the prob- illustrates at 450 magnifications typical crys-
lem is more severe in recirculating systems tals of calcium carbonate scale formed under
and, while essentially the same types of recirculating conditions. Figure 33-8 shows
treatment are used, higher treatment concen- the crystal distortion secured under the same
trations in the circulating water are neces- conditions with a blended phosphate-tannin
sary, because of the increased scaling po- treatment.
tential. In general, blended treatments of organic
Figure 33-6 shows a laboritory experi- and inorganic agents have shown the greatest
mental cooling tower system used in the effectiveness in scaling inhibition. The con-
study of scale formation in open recirculat- centrations required range between approxi-
ing systems. In a system of this type it is mately 30-100 ppm in the circulating water.
possible to automatically control temperature, Of course, these concentrations must be
heat input to the scaling chamber, cycles of divided by cycles of concentration in order to
concentration, blowdown and makeup. The express treatment concentration on a make-
effect of different treatment concentrations up water basis. For example, at a circulating
can be determined at different scaling loads. water concentration of 60 ppm treatment, if
the system were operated at 5.0 cycles of con-
As is to be expected the principles estab- centration, the treatment required would be
lished in the control of calcium carbonate 12 ppm or 0.1 pounds per 1,000 gallons
scale formation in once-through systems also makeup water.
Figure 33-6 • Laboratory Experimental Recirculating Cooling Water System Used to Evaluate Scaling Tendencies
252 BETZ HANDBOOK
The solubility of calcium sulfate has been once-through scale prevention. Except in an
extended in cooling systems through the use extraordinary case, calcium carbonate scale
of dispersants, sequestrants or chelants. Dis- formation can be prevented in once-through
persants prevent buildup of particle size systems by the simple device of using in-
through adsorption, which involves electro- creased treatment concentrations for increased
static forces. Sequestrants and chelants func- scaling loads. However, in open recirculating
tion through electron transfer, forming a systems it is possible to reach such a degree
water soluble complex with calcium. of oversaturation of calcium carbonate that
As an example of the power of dispersants, even increased treatment concentrations can
in one tower system a study was conducted not control the scale problem. No increase in
using a combination organic-phosphate dis- treatment, within economic limits, will con-
persant. This study indicated that the cooling trol the oversaturated condition that can de-
system could be opera ted at 145% over the velop under some circumstances at high
saturation level of calcium sulfate without cycles of concentration.
deposition occurring in the system. To avoid such an oversaturation with re-
spect to calcium carbonate, and also to con-
trol formation of calcium sulfate, calcium
silicate and magnesium silicate, it is necessary
to limit cycles of concentration. By thus
limiting the concentration of the circulating
water, the degree of oversaturation can be
restricted to within the effective range of the
treatment employed. Where the natural
windage loss is insufficient to limit cycles of
concentration, it is necessary to supplement
the windage loss by means of blowdown.
Blowdown consists in removing a portion
Figure 33-7 • CaC0 3 Scale Formation without of the concentrated circulating water and
Treatment in Recirculating System at replacing it with fresh makeup water, thus
27,000 Btu/Ft 2/Hr Heat Transfer. 450 Mag. securing a lowering of the concentration in
the system. Blowdown can be either intermit-
tent or continuous. Continuous blowdown is
preferred, controlled by a flow control valve
on a separate line to waste. The blowdown
valve control setting is varied to maintain
cycles of concentration within safe limits to
prevent scale formation. '
The amount of blowdown required to
limit cycles to any predetermined value can
be calculated through the use of the basic
formula:
% E +% B
Figure 33-8 • CaC0 3 Scale Formation with Poly- Cycles of Concentration = ~~---
phosphate- Tannin Treatment in Recirculating System %8
at 27,000 Btu/Ft 2/Hr Heat Transfer. 450 Mag. where% E = Evaporation loss, expressed as
percent of circulation rate.
CONTROL OF CYCLES
% B = Combined windage and blow-
The concentration of scale forming solids down loss, expressed as per-
that takes place in a cooling tower system cent of circulation rate.
introduces a problem not encountered in
Derivation of this formula is as .outlined
c
8
TERMS: Therefore:
E = Lbs of water evaporated. (3) Mu X CIMu = (E X Clel + (B X Cia)
B = Lbs of water lost by windage and
blowdown. However, chlorides are not lost by evapora-
Mu = Lbs of makeup water. tion, and therefore OlE = 0. Equation (3)
then becomes:
C = Lbs of water circulated.
(4) Mu X CIMu = B X Cia
ClMu = Chloride of makeup water.
OlE = Chloride of water evaporated. At equilibrium conditions the chloride con-
Cia = Chloride of circulating water. centration of the blowdown water is the
CIB = Chloride of water lost by windage same as the circulating water. Substituting
and blowdown. in equation (4) :
PROOF: (5) Mu X CIMu = B X Clc
By definition, cycles of concentration is
the term employed to indicate the degree Rearranging the terms :
of concentration of the circulating water Clc Mu
as compared with the makeup water. (6)--=--
Since chlorides remain soluble on concen- CIMu B
tration, the term cycles of concentration
is best expressed as the ratio of the chloride Substituting from equation ( 1)
content of the circulating and makeup Mu
waters. (7) Cycles of concentration = - -
B
Clc
(1) Cycles of concentration = --
CIMu Substituting from equation (2)
Since makeup water is added to the system E+ B
in an amount equivalent to the water lost (8) Cycles of concentration = - - -
from the system: B
(2) Mu = E + B
Multiplying through by percentage (%)
At equilibrium conditions, the chloride enter-
ing the system must equal the chloride leaving . %E+%B
(9) Cycles of concentration =- ----
the system. %8
Fig. 33-9-Derivation of Formula for Calculation of Cycles of Concentration

254 BETZ HANDBOOK
in Figure 33-9. While this formula has been If we take as windage loss, the fairly high
expressed in percentages, other units can be figure of 0.3%, we can calcvlate the cycles
used provided all quantities are expressed in of concentration that will develop.
the same units. The formula may also be % E + % B 2.0 + 0.3
rearranged to solve directly for blowdown as Cycles = - - - - - - - - - = 7 . 7
follows: "'oB 0.3
Obviously, additional deconcentration by
blowdown is necessary in order to restrict
%E cycles to 2.0. This blowdown can be calculated
% B=-----
(cycles-1) as follows:
As an example, if we vvere to attempt to "foE 2.0
use the makeup water in Figure 33-5 in a % B =----- ---=2.0
(Cycles-1) 2.0-1.0
cooling tower system without treatment, it
Combined windage and blowdown losses
is obviously necessary to limit cycles of con-
are 2.0% and if the windage loss is 0.3%, the
centration to 2.0.
blowdown loss must be 1.7% of the rate of
With the following basic data, calculations circulation, equivalent to 170 gpm. The flow
can be made: control valve on the continuous blowdown
Rate of circulation = 10,000 gpm can be sized accordingly. Since evaporation
Temperature drop through tower 20 F loss is also 2.0%, makeup is 4.0% or 400
Type tower = Induced draft gpm.
12.0
11.0
~
c 10.0
:2
:;"' 9.0
~
·o
c 8.0
0
"(1)
Vl
7.0
e"'
Ct:
w 6.0 50' drop
f-
<
;;:::
5.0
"-
::;:,
w 4.0
""
<
::;< 25' drop
3.0
~
2.0
10' drop
1.0
0 3 4 5 6 7 8 9 10
CYCLES OF CONCENTRATION
Figure 33-10 • Makeup Water Requirements vs. Cycles of Concentration

ACID TREATMENT water is employed, sulfuric acid is generally

used because of its low cost. The objective of
While blowdown is an effective method for acid treatment is to reduce the scaling tend-
limiting cycles of concentration and hence ency, but not to develop an acid condition in
the scaling potential of the circulating water, the circulating water. In fact, acidity is to be
it is not always possible to use excessive rates avoided. Sufficient acid is employed to re-
of blowdown to accomplish this objective. In duce, but not eliminate the alkalinity of the
many localities, the supply of fresh water is circulating water.
either limited or costly. It may be desirable By the use of acid treatment, the calcium
to resort to treatment measures which will bicarbonate is converted to the more stable
permit higher cycles of concentration in the and more soluble calcium sulfate, as illus-
circulating water, with consequent lower trated by the following reaction:
rates of blowdown.
Ca(HC0 3 h + H2S04 = CaS04 + 2C02
Figure 33-10 shows the relationship be- calcium + sulfuric = calcium + carbon
tween makeup water requirements and cy- bicarbonate acid sulfate dioxide
cles of concentration at different tempera- + 2H20
ture drops across a tower. Curves are based +water
on an average value of 1% evaporation loss It can be seen from this equation that acid
per 10 F drop across a tower. It can be noted treatment decreases the alkalinity and, by
how sharply makeup requirements decrease virtue of this reaction, minimizes the degree
as cycles are increased up to approximately 5 of oversaturation with respect to calcium
cycles. There is not usually much advantage carbonate. The calcium sulfate formed in
to be secured in reducing makeup water re- this reaction, as shown by the curves in Figure
quirements by increasing cycles of concentra- 33-3, is considerably more soluble than cal-
tion above 5.0. In some cases, however, it cium carbonate and therefore a greater con-
may be desirable to carry very high cycles of centration of this salt can safely be tolerated.
concentration, where permitted by water
An example of the advantages of acid
characteristics, in order to conserve on chem-
treatment is shown in Figure 33-11. In this
ical treatment.
plant, heavy scale formation occurred even
In those cases where lowered blowdown with low cycles of concentration, when no
rates are desirable, either for technical or chemical treatment was used. After a survey
economic reasons, it is necessary to resort of plant conditions, the use of scale inhibiting
either to softening of the makeup water to treatment, together with alkalinity adjust-
lower the scale forming solids or to use acid ment with acid, was recommended. Increased
treatment of the circulating water. cycles of concentration, and a lower blow-
External softening of the makeup water down rate, were easily tolerable under these
can be accomplished through the use of cold conditions, with complete freedom from scale
process lime, lime-soda or lime-gypsum proc- formation.
esses and by zeolite softening. The operation The first column of Figure 33-11 shows
of the softening processes is as described in the characteristics of the plant makeup
the chapters on these subjects. In general, water. The second column illustrates the
these external softening processes can not be circulating water analysis, without treatment.
justified economically in comparison with It can be seen that the system was under 3.2
acid treatment. While exceptions exist, acid cycles of concentration. With 50 ppm of cal-
treatment is usually more economical in cium hardness in the makeup water, at 3.2
continuous chemical cost. In addition, the cycles of concentration, the circulating water
large investment cost in softening equipment should contain 160 ppm calcium hardness if
is avoided. none precipitated in the form of scale or
Where acid treatment of the circulating sludge. The presence of only 70 ppm calcium
256 BETZ HANDBOOK
in the circulating water represents only 44% this treatment procedure, blowdown was re-
of the expected amount. The analysis of the duced from 152 gpm to 40 gpm, decreasing
circulating water confirmed the heavy scale makeup water requirements from 524 gpm
formation experienced in the plant operation. to 412 gpm.
In the third column is shown the charac- It is advisable where acid treatment is
teristics of the circulating water with acid employed to make the acid feeding equip-
treatment and a scale inhibiting treatment. ment as automatic as possible so that fre-
It can be seen that by virtue of the acid quent adjustment is not required by operating
treatment, the methyl orange alkalinity has personnel. In many cases, the investment in
been limited to 125 ppm and the sulfate automatic pH control for governing acid
content of the water has sharply increased. feed is advantageous.
The addition of the acid, however, has per- In cases where the sulfate content of the
mitted the maintenance of 8 cycles of con- makeup water is already quite high, the
centration, compared with 3.2 cycles of con- supplementary feed of sulfuric acid may not
. centration on the untreated water. At 8 be permissible, if calculations show that the
cycles of concentration, with 50 ppm of cal- solubility of calcium sulfate will be exceeded
cium in the makeup water, 400 ppm of in the circulating water. Under these cir-
calcium would be expected in the circulating cumstances, hydrochloric acid may be used
water. It can be seen that 385 ppm of calcium instead of sulfuric acid for alkalinity adjust-
are present, which represents 96% maintained ment. It is also in such cases that external
in the soluble form. softening of the makeup water for reduction
It would have been possible to decrease in calcium hardness becomes advantageous.
alkalinity below 125 ppm and to maintain Occasionally, a proposal is made to control
still higher cycles of concentration. In this the pH of the circulating water by means of
particular case, there was no appreciable recarbonation. The use of carbon dioxide
gain to be secured by such a procedure. will reduce pH, but it will not reduce alka-
Operating data shown indicate that by linity. On each pass of the circulating water
Fig. 33-11-Effectiveness of Acid Treatment in Eliminating Scale and

Reducing Slowdown
Untreated Treated
Makeup Circulating Circulating
Water Water Water
Total Hardness as CaC03 . . . . . . . . . . . . . . . . . . ........ ...... 85 160 655

Calcium as CaC0 3 . . . . . . . . . . . . . . . . . . . . . . . ........ ...... 50 70 385
Magnesium as CaC03 . . . . . . . . . . . . . . . . . . . . . ........ ...... 35 90 270
P Alkalinity as CaC03 . . . . . . . . . . . . . . . . . . . . ........ ...... 0 25 0
M Alkalinity as CaC03 . . . . . . . . . . . . . . . . . . . . ........ ...... 180 470 125
Sulfate as S04 . . . . . . . . . . . . . . . . . . . . . . . . . . . ........ ...... 25 80 1520
Chloride as Cl . • . . . . . . . . . . . . . . . . . . . . . . . . . ........ ...... 40 128 320
Silica as Si02 • • • • • • • • • • • • . • • • • • • • • • • • • • • •••••••••••••• 8 20 60
Above values expressed in ppm
pH ................................................. . 7.1 8.9 7.9
Cycles of Concentration . . . . . . . . . . . . . . . . . . . • . . . . . . . . . • . . . . . . . . . . . . . . . . . . . . . 3.2 8.0
Operating Data
Circulation, gpm . . . . . . . . . . . • . . . . . • . . . . . . . . . . . . • . . . . . . . . . . . . . . . . . . . . . . . . . 12,000 12,000
Evaporation, gpm (3.0%) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360 360
Windage, gpm (0.1 %) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 12
Slowdown, gpm . . . • . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 152 40
Makeup, gpm . . • . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . • . . . . . . . . . . . . . . . . . . . 524 412
over the tower, the carbon dioxide added for only rare that such systems are treated for
pH control will be aerated from the circu- scale control alone. Corrosion is an ever-
lating water, down to its equilibrium value present problem. It is not sufficient to ade-
for those particular conditions. Consequently, quately control scale formation and to permit
to secure any pH reduction from the use of corrosion to progress. Corrosion control meas-
carbon dioxide, it would be necessary to add ures must be cmnbined with scale prevention
an amount equal to that required if all the treatment, as discussed in the next chapter,
water circulated were on a once-through basis. in order to provide a satisfactory balanced
The carbon dioxide will not recycle. Con- treatment program.
sequently, for all but the most unusual cir- REFERENCES
cumstances, recarbonation is not practical as
L. D. Betz and J. J. Maguire, "Problems in the
a means of pH control in open recirculating Treatment of Cooling Water in Industrial Plants",
systems. Proceedings, Midwest Power Conf., Vol. XI, pp.
318-328 ( 1949)
Sulfur dioxide is occasionally used as a
R. P. Gulley, "Modern Cooling Tower Water Treat-
means of alkalinity adjustment. Sulfur burners ments", Petroleum Refiner, Vol. 39, pp. 165-168
generate sulfur dioxide which combines with (Apr. 1960)
water and oxygen, simply as an alternate W. A. Tanzola, "Cooling-Water Systems and Their
Chemical Treatment," Petroleum Refiner, Vol. 23,
method of supplying sulfuric acid. pp. 375-382 (Oct. 1944)
While the discussion in this chapter has J. M. Donohue and G. A. Woods, "On Stream De-
been confined to the problems and control sludging Restores Heat Transfer," 23rd Annual
NACE Conference (1967)
of scale in open recirculating systems, it is
258 BETZ HANDBOOK
Causes of Corrosion. . . . . . . . . . . . . . . . . . . . . . 259

Prevention of Corrosion. . . . . . . . . . . . . . . . . . . . 259
Calcium Carbonate Protective Scale. . . . . . . . . 259
Mechanical Deaeration. . . . . . . . . . . . . . . . . . 259
Corrosion Inhibitors. . . . . . . . . . . . . . . . . . . . . 259
Chromates. . . . . . . . . . . . . . . . . . . . . . . . . . 260
Polyphosphates. . . . . . . . . . . . . . . . . . . . . . 261
34 Dianodic Method. . . . . . . . . . . . . . . . . . . . . . 262
Phosphate Dianodic. . . . . . . . . . . . . . . . . . . 263

Open Recirculating
Corrosion Control
,Zinc Dianodic . . . . . . . . . . . . . . . . . . . . . . . 264
Fluoride Dianodic . . . . . . . . . . . . . . . . . . . . 265
Chromium Dianodic. . . . . . . . . . . . . . . . . . 265
Organic Base Treatments. . . . . . . . . . . . . . . . 265
Miscellaneous Inhibitors. . . . . . . . . . . . . . . . . 266
Chromate Removal. . . . . . . . . . . . . . . . . . . . . . . . 266
Factors Influencing Corrosion Load. . . . . . . . . . . 266
Corrosion of Non-Ferrous Metals . ........... 268
Summmy ............................... 269
HE prevention of corrosion in open re- tubes is common practice. Such galvanic in-
T circulating cooling water systems is a
problem on which much research effort has
fluences increase the problem of corrosion
control.
been expended in recent years. As a result, Open recirculating systems, to a greater
effective anti-corrosion measures have been extent than once-through systems, are sub-
developed which can be applied in most ject to outside contaminating influences. Sand
instances at reasonable cost. In earlier years, and dust may be blown into the circulating
many recirculating systems were not treated water. Fibres may break loose from the tower
for corrosion prevention because the cor- wood. Sludge and slime accumulations col-
rosion inhibitors then known were required lect in the tower sump. Where such deposits
at high concentrations with consequent high lodge on metaiiic surfaces, there is the op-
treatment cost. Today, it has been demon- portunity for development of an oxygen con-
strated repeatedly that the cost of corrosion centration ceil, and severe localized pitting.
prevention treatment is minor compared Because of the effect of ali of these cor-
with the savings to be obtained in equip- rosive factors, corrosion control in open re-
ment replacement, down time and produc- circulating systems represents an aggravated
tion loss. Corrosion control in open recircu- problem as compared with once-through
lating systems is, in reality, more important systems.
than scale control. Scale can be removed by
acid cleaning or mechanical cleaning. The
damages wrought by corrosion may require PREVENTION OF CORROSION
the complete replacement of heat exchangers,
The methods employed for corrosion control
pumps, lines, etc.
of open recirculating systems primarily in-
volve the use of corrosion inhibitors. How-
CAUSES OF CORROSION ever, controiied calcium carbonate scale
deposition, as weii as deaeration, have been
The causes of corrosion discussed in once- employed.
through systems also apply to open re- CALCIUM CARBONATE PRoTECTIVE ScALE. The
circulating systems. However, the problem in use of this treatment method was discussed
recirculating systems tends to be intensified under corrosion control in once-through sys-
as compared with once-through systems. tems. The same comments apply to its use in
The intimate contact of the cooling water recirculating systems. Basically, since proper
with air in passing over the tower renders the treatment of recirculating systems requires
circulating water continuously saturated with that heat exchange surfaces be scale free,
dissolved oxygen. This corrosive gas is, there- this method of control is impractical except
fore, continuaiiy present in maximum amount under unusual circumstances.
and is the major source of corrosion diffi- MECHANICAL DEAERATION. While vacuum
culties. Where towers are located in industrial deaeration has been applied to small cooling
areas, the air also may be contaminated with tower systems, this method of corrosion con-
gases such as sulfur dioxide, ammonia and trol is not practical for large installations.
hydrogen sulfide. The deaerating unit must be sized to remove
The concentration of dissolved solids, par- oxygen from the full flow of oxygen saturated
ticularly chloride and sulfate, on cycling in circulating water, since the oxygen content is
an open recirculating system also adds to the replenished in passing over the tower. Instal-
corrosion load and makes more difficult the lation and operating costs make this method
prevention of corrosion by inhibitors. impractical for all but the smallest installa-
Most recirculating systems contain dissim- tions although the method is highly effective
ilar metals. For example, the use of steel from a technical standpoint.
tube sheets with admiralty heat exchanger CoRROSION INHIBITORS. In cooling water
260 BETZ HANDBOOK
systems, the corrosion inhibitors most fre- pitting and the small amount of metal loss
quently employed are the chromates and that does take place is concentrated in a
polyphosphates. Other corrosion inhibitors small area.
such as nitrites, silicates, amines and various This pitting effect, and the resultant tu-
other organic agents also are employed to a berculation from the corrosion products, rep-
lesser extent. resents the major disadvantage in the use of
An anodic inhibitor is one that restrains chromate salts for corrosion inhibition. It is
the anodic corrosion reaction : possible to minimize this tendency for pitting
Fe = Fe+++ 2e by using higher chromate concentrations, par-
Examples of anodic inhibitors are sodium ticularly under severely corrosive conditions.
and potassium chromate, nitrite, phosphate However, economic factors usually limit the
and hydroxide. treatment concentration that may be em-
ployed so that the isolated pitting encountered
Cathodic inhibitors are those that restrain
with lower chromate concentrations has been
the cathodic corrosion reaction:
accepted as inevitable and the primary dis-
Oz + 2 HzO + 4e = 4 (OH)- advantage of an otherwise fine corrosion in-
Salts of metals which form sparingly soluble hibitor.
hydroxides, oxides and carbonates, such as
zinc and nickel, act as cathodic inhibitors. 1000
Calcium carbonate is also a cathodic inhibitor.
In general, any treatment which forms an
adherent coating on the metal surface func- BOO I r\
I \
:I:
(.)
tions as a cathodic inhibitor. An ideal in- z
hibitor would be one which would develop a
0 600
dense protective coating of minimum thick-
\
(/)
ness and which would deposit evenly at both

hot and cold sections of the system.
0::
w
/
a.. 400
It is characteristic of anodic inhibitors that
(/)
they are efficient corrosion inhibitors if used f-
in sufficient quantity to completely stifle cor- a.. 200
rosive action. However, if used in insufficient
amount to completely stop attack, corrosion
is severely localized in the form of pitting. 0
0 5 6 7 8 9
Failure of the metal by perforation may occur
more quickly than if no inhibitor had been PH
added. Figure 34-1 • Variation in Number of Pits with pH-
Chromates. From the standpoint of saving Low Chromate Concentrations
metal, by providing a high percentage cor-
rosion protection, the various chromate salts Figure 34-1 illustrates the influence of pH
are the most effective corrosion inhibitors in on the frequency of pitting at low chromate
use. With adequate chromate concentrations, concentrations. The number of pits per
corrosion rates can be reduced 95% or more square inch of test specimen surface, at a
in comparison with an untreated condition. chromate concentration of 60 ppm, rises to
Such corrosion prevention is substantially a maximum at pH 7.0 and then drops off
higher than can be secured by any other sharply at pH 8.0. At all pH values shown,
chemical inhibitor. A disadvantage of low however, pitting is excessive.
chromate concentrations, however, is that iso- Usually, satisfactory corrosion control can
lated pitting of the metal is not completely be secured with the maintenance of 300-500
stifled. The small amount of corrosive attack ppm chromate in the circulating water. How-
which does occur is localized in the form of ever, in many cases, concentrations of 500-
1000 ppm, and even higher, have been found urement of pit depth.
necessary to control the tendency toward pit- Figure 34-2 shows the metal loss from
ting. Naturally, the use of such high chro- steel test specimens exposed in a continuous
mate concentrations has made the use of flow system at a flow rate of 0.36 feet per
this material too costly in many instances. second at 120 F. Polyphosphate was fed at
Polyphosphates. Phosphates, particularly the the rate of 9,0 ppm, as Na12P1oOa1· The
polyphosphates, are also widely used in initial attack decreased on longer exposure as
cooling water treatment. The ability of phos- shown by the slope of the curve, but steel
phate to prevent metal loss is far inferior to loss continued at an excessive rate.
chromate and, in addition, pitting with phos-
phate treatment is encountered to an even
35 0
greater extent than with chromate. Unlike
chromate, higher phosphate concentrations
usually are not practical because of the pre- 30 0
_/
/
cipitation of calcium phosphate. Phosphate,
however, does possess the advantage of re-
ducing tuberculation, thereby minimizing
250 /
head losses due to corrosion products. Al-
though pitting will occur, the dispersing or
Ol
E
• 200 /
~
(f)
cleaning action of the phosphate prevents (f)
0
the buildup of iron oxide tuberculation prod- ...J
150
ucts, maintaining the surface in a relatively ...J
clean condition and reducing friction losses.

Because of this solubilizing action, exposed
<l:
f-
UJ
::;: 100 L
metal surfaces with phosphate treatment usu-
ally present a clean appearance, relatively 50 I
free of corrosion products. Instead of forming
iron oxide tuberculation products over a pit,
these corrosion products are continuously re- 0
0
I 2 4 6 8 10 12
moved from the surface by the phosphate DAYS EXPOSURE
concentration present m the circulating
water. Figure 34-2 • Metal Loss with High
Phosphate Concentrations
On the other hand, surfaces exposed to
cooling water containing low chromate con-
centrations may show tuberculation in the A factor limiting the use of phosphates in
form of corrosion products from scattered pit- the treatment of recirculating cooling water
ting. The steel savings of the chromate treat- systems is the reversion of polyphosphate to
ment may be 50-100 times as great as the orthophosphate. The orthophosphate form
phosphate treatment and the number and provides less corrosion inhibition than the
depth of pits only a small fraction of that polyphosphate form. In addition, orthophos-
experienced with phosphate treatment. How- phate reacts with the calcium content of the
ever, a casual observation of the metal sur- circulating water to precipitate calcium phos-
faces may erroneously i11dicate the phosphate phate, which can form a troublesome deposit
treatment to be superior to the chromate. on heat exchange surfaces. Reversion of poly-
Evaluation of the merits of the respective phosphate to orthophosphate is increased by
treatments, unless extended over a long period long retention time of the system and by
of time, can best be made from weight loss higher water temperatures. Blowdown from
data on test specimens exposed in the system the system must be adjusted so as to avoid
and from microscopic examination and meas- exceeding the solubility of calcium phosphate.
262 BETZ HANDBOOK
DrANomc METHOD. As previously stated, if

anodic inhibitors are present in insufficient
quantity, corrosion will be localized in the
form of pitting. Both chromate and polyphos-
phate are anodic inhibitors, but the different
til
action of these two materials as corrosion in- "
UJ
0:
hibitors led to an investigation of the possi- ">
UJ
bility of combining the desirable effects of

both. The Dianodic method resulting from ~
0
this research, was introduced approximately a :i

til
>-
decade ago, and employs a dual treatment of a:
"-
molecularly dehydrated phosphates and chro- 0
z
mates over a selective pH range. This method 0
>=
produces effective control over pitting and "'
:::>
0
UJ
tuberculation even with the relatively low 0:
.treatment levels of 40-60 ppm. The superior "'

corrosion protection secured and the low cost
of treatment has encouraged wide use of the
Dianodic method. Corrosion rates of 1 mil per
-315
year or less for steel are common with the pH -315
pitting function completely prevented or defi-

nitely stifled. Corrosion rates with copper and Figure 34-4 • Reduction in Pitting by Dianodic Method
its alloys are even lower-often less than 0.3 vs. Chromate
mils per year. Figure 34-3 illustrates the
results secured from an actual operating
system.
Figure 34-4 illustrates the reduction in pit-
ting secured with Dianodic at 60 ppm, com-
pared with the results for sodium chromate
at 60 ppm over the pH range of 5.0-8.0.
Figure 34-5 shows this same comparison with
60 ppm of polyphosphate. til
Continued research has permitted the de- "~

UJ
UJ
>
1.4
~
0
z
~
"~
1.2
a:
\
STEEL--- COMPARISON OF SO ppm DIANODIC
COPPER--- "-
0 WITH SO ppm POLYPHOSPHATE
1.0 z A- DIAN ODIC (50 PHOSPHATE- 10 CHROMATE)
9
z
0
0.8
g B- DIAN ODIC ( 40 PHOSPHATE-20CHROMATE)
C- OlANO DIG { 20 PHOSPHATE- 40 CHROMATE)
~ fil
~
0:
0
Q6 "'
g
>
< 0.4
Q2
20 60 80 tOO 140 160 180
DAYS EXPOSED
pH
Figure 34-3 • Corrosion Control of Steel and Copper Figure 34-5 • Reduction in Pitting by Dianodic Method
Specimens with Phosphate-Chromate Control vs. Phosphate
velopment of many variations of the original

Dianodic method to meet specific require-
ments for the individual cooling water system
and to cope with the many variables that
enter into the corrosion problem. The Dia-
nodic methods include: Phosphate Dianodic;
Zinc Dianodi'c; Fluoride Dianodic; Chro-
mium Dianodic; as well as combinations of
these methods.
Phosphate Dianodic. The Phosphate Dia-
nodic method, which makes use of polyphos-
Figure 34-6 • Chromate Pits at 18 Magnifications phate in combination with chromate over a
controlled pH range, eliminates the pitting
function that is produced by the individual
components when used separately. Figures
34-6 through 34-9 illustrate the control of
pitting by the Dianodic method compared to
phosphate and chromate when used sepa-
rately. The shallow anodic areas that may de-
velop with the use of this combination treat-
ment have no appreciable depth and are rap-
idly rendered inactive. In many cases, even
these shallow anodic areas do not develop.
Figure 34-7 • Phosphate Pits at 18 Magnifications The use of the Dianodic method to reduce
pitting and tuberculation does not involve
any sacrifice in the control of general cor-
rosion. Figure 34-10 shows the average and
maximum penetration values obtained from
weight loss data for 3 treatment methods.
Phosphate concentration was 4-8 ppm, chro-
mate approximately 200 ppm and the Dia-
nodic concentration was 60 ppm. These re-
sults show that the Dianodic method is equal
Figure 34-1 0-Plant Comparison of Three

Treatment Methods
Figure 34-8 • Dianodic Method Specimen at
18 Magnifications Treatment Average Maximum
Penetration, mpy• Penetration, mpy•
low Phosphote . . . . . . . . . . 21.0 . . . . . . . . . . 660

Chromate . . . . . . . . . . . . . . . 3.3 . . . . . . . . . . 150
Phosphate Dionodic . . . . . . 2.0 . . . . . . . . . . 20
*mpy =mils per yeor = 0.001 inch per year
to chromate in prevention of general cor-

rosion and far superior to phosphate from
this standpoint. While the major advantage
of the Dianodic method is reduction of pit-
ting, general metal loss is controlled as effec-
Figure 34-9 • Unexposed Specimen at 18 Magnifications tively as with high chromate.
264 BETZ HANDBOOK
Severe tuberculation can result from chro- Zinc Dianodic. In spite of the wide coverage
mate treatment due to the formation of iron permitted by the Phosphate Dianodic method
oxide over the isolated pits. With phosphate some limitations are encountered where cal-
treatment, there is very little tuberculation cium of the makeup water. is high or the re-
because of the dispersing action of the phos- version of polyphosphate to the orthophos-
phate. While pitting with phosphate treat- phate form is abnormal. Under these condi-
ment takes place to an even greater extent tions it may not be possible to establish or
than with chromate, the corrosion products maintain the most effective ratio of phosphate
are dispersed by the phosphate and the metal to chromate without encountering deposition
surface maintained relatively free from tu- of calcium phosphate. In these cases, Zinc
berculation. Dianodic can be used effectively. The super-
Results in the prevention of tuberculation imposing of zinc on the phosphate-chromate
with the Dianodic method are superior even combinations permits a greater latitude on
to phosphate treatment, as shown by Figure phosphate levels or, at the same phosphate
34-11. The Dianodic combination acts to dis- levels, improves corrosion protection. For ex-
ample, when normal phosphate levels cannot
be maintained due to some limiting factor,
such as high calcium or a high reversion rate
of the phosphate, phosphate level may be de-
creased and zinc superimposed. In this case,
the zinc takes up the slack and acts as a ca-
thodic inhibitor to stifle the corrosion reac-
tion. With the introduction of Zinc Dianodic,
even broader coverage is possible and this
combination of zinc, phosphate and chromate
minimizes, if not eliminates, most limitations
to use of the Dianodic method.
The Zinc Dianodic method is quite versa-
Figure 34-11 • Tuberculation qf Nipple Threads at tile and can be operated in the 'soluble
0.5 ft per sec flow rate. Left .Specimen-Phosphate. range' or the 'coating range'. In the 'soluble
Center-Chromate. Right-Dianodic ·
range' or film forming range, pH and zinc
values are set below the point where zinc
phosphate will precipitate. Under these con-
perse any corrosion products and since pit- ditions no visible coating is formed except at
ting is reduced to a minor quantity, there is. points where the pH is elevated by local cell
little corrosion product to be removed. action. For example, if local cell action takes
Plant results with Phosphate Dianodic place, the area undergoing attack is the
have been excellent. Treatment concentra- anode and the area surrounding the point
tions will vary with the individual operating of attack is the cathode. The cathodic re-
conditions and may range between 30 and action is:
100 ppm depending on water characteristics,
temperature, cycles of concentration, reten-
02 + 2H2 + 4e = AIOH)-
tion time and other factors. Well established The hydroxide thus formed is sufficient to
rules have been developed through years of elevate the pH and cause deposition of zinc
experience over a wide range of conditions phosphate which stifles the cathode reaction.
to permit the selection of the proper treat- In the coating range, the zinc and pH are
ment levels, proper treatment ratios and pH set above the pH of saturation of zinc phos-
range to assure maximum protection and the phate which results in a coating formation
most economical cost for the individual over the entire system. After the desired
system. coating is formed, the zinc and pH levels are
adjusted downward so as to just maintain extensive damage was done to the system.
the coating intact and to repair the coating With the application of Dianodic, the chro-
at any point where cell action may start. In mate component of the treatment is reduced
both cases, it can be seen that the function by these contaminants to trivalent chromium
of zinc is one of a cathodic inhibitor or bar- and leakage is readily detected. Where pos-
rier type treatment. Generally, however, the sible the leakage should be corrected immedi-
use of the 'coating range' is restricted to the ately to minimize the heavy corrosion load
treatment of once-through systems. imparted by hydrogen sulfide. In many cases,
In the control of corrosion, it is important immediate correction of the leakage is not
that rapid film formation be accomplished possible and a suitable method of treatment
to stifle the corrosion reactions. The use of in the presence of process leakage becomes
Zinc Dianodic permits rapid film formation necessary. This requirement led to the devel-
and is an excellent pretreatment for steel opment of Chromium Dianodic which makes
heat exchangers. use of chromium phosphate as a barrier type
treatment during periods of leakage.
Fluoride Dianodic. Fluoride Dianodic was de-
veloped to handle waters contaminated with ORGANIC BAsE TREATMENTs. Recent develop-
aluminum. Surface supplies that are turbid ments in research have made available organic
must be clarified to remove suspended solids base inhibitors that can effectively replace
that would otherwise settle on heat transfer chromate treatments in certain systems. Al-
surfaces. It is common practice to coagulate though not as dependable as chromate base
and settle surface supplies without the benefit treatments in systems with high corrosion
of filtration when the water is used as makeup loads, the organics can often produce satis-
to open recirculating systems. The low tur- factory results while eliminating the need for
bidity from the modern type clarifier usually chromate removal.
permits this practice. However, where alumi-
The most effective treatments of this type
num is used as a coagulant and floc carry-
include zinc and polyphosphate with organic.
over takes place, fouling of the heat exchang-
Some of the newer processed lignin deriva-
ers with aluminum deposits can result.
tives have demonstrated their abilities as
The fouling materials are most commonly cooling water inhibitors, particularly when
aluminum hydroxide and aluminum phos- supplemented by zinc. Another successful in-
phate. The presence of aluminum interferes hibitor system is a combination of a polar
with the Phosphate Dianodic method by organic sulfur, zinc and polyphosphate. This
forming a soluble aluminum polyphosphate treatment provides a barrier of a thin, but
complex. The Fluoride Dianodic method adherent, nature that protects metal surfaces
makes use of the fluoride ion to preferentially from corrosive attack. Results comparable to
complex the aluminum ion. The soluble the Dianodic method are possible under many
complex thus formed acts dianodically with conditions, but treating costs will normally
chromate and aids in avoiding fouling and run somewhat higher.
stifling of corrosion.
Polar organic sulfur compounds, zinc and
Chromium Dianodic. The original applica- polyphosphate are not compatible in a single
tions of Phosphate Dianodic to many refinery feeding system. Two feeding tanks are norm-
cooling water systems uncovered many in- ally required.
stances where leakage from the process side
was taking place. Where sour crudes are be- Lignin derivatives and organic sulfur in-
ing processed, the leakage contaminates the hibitors are not compatible with chlorination,
cooling water with hydrogen sulfide, mercap- so non-oxidizing biocides must be used. How-
tans and organic matter. Formerly, such leak- ever, non-oxidizing biocides provide effective
age went undetected for appreciable periods slime and algae control at costs approaching
of time and usually was discovered only after those of chlorination.
266 BETZ HANDBOOK
MISCELLANEOUS INHIBITORS. Sodium nitrite rosion inhibitory properties.

can be used as a corrosion inhibitor in open
circulating cooling water systems. While ni- CHROMATE REMOVAL
trites are sometimes regarded as reducing Regulations regarding the quality of plant
agents, they do possess oxidizing properties effluent may restrict the concentration andjor
and probably for this reason tend to render form of chromate in water discharged to
metal surfaces passive to corrosive attack in sewers, streams or other bodies of water.
weakly alkaline solutions. In acid solutions, However, in many cases it is better to apply
however, nitrites do not render iron passive. the Dianodic method of treatment and re-
In order to establish a protective iron move chromate from the discharge water
oxide film it is usually necessary to feed so- than it is to use an inferior non-chromate
dium nitrite at a much greater concentration treatment.
initially than normally required. The rate of Chromate removal can be accomplished
feed can be reduced to normal once the pro- with various commercially available chemi-
tective film has been established. For most cals. The procedure used is reduction of hexa-
effective results, sodium nitrite should be fed valent chromate to the trivalent chromium
continuously to the system. Decomposition form, followed by elevation of pH for pre-
difficulties frequently make it difficult to cipitation of the reduced chromium. Reduc-
maintain effective nitrite concentrations in ing agents employed most commonly for this
the circulating water without excessive feed purpose include sulfur dioxide, ferrous sul-
rates. fate, sodium bisulfate, sodium metabisulfate,
Both oxidizing and reducing agents tend sodium sulfite and hydrogen sulfide. The re-
to destroy nitrite and certain bacteria also action generally proceeds more rapidly at
cause its decomposition. Where nitrite con- lower pH levels, such as 3.0. Depending on
centrations can be continuously maintained the pH of the reaction, the reducing agents
at an effective level, satisfactory protection of can also react with oxygen in the cooling
ferrous metals is secured. Nitrite is not very water blowdown.
effective in the prevention of attack on cop-
A typical removal system is designed for
per or copper alloys.
the addition of sulfuric acid to reduce the
The polyphosphates are versatile materials pH of the discharge water to 3.0, followed
and many other agents have been employed by the injection of sulfur dioxide at ap-
in conjunction with the polyphosphates in proximately one part per part of chromate.
an attempt to improve their corrosion inhibi- A mixing facility is provided, and then lime
tory properties. Complex cyanide synergists is fed to increase the pH to 8.5. A retention
alter to some degree the typical pitting at- and sedimentation basin or lagoon allows
tack. Various organic materials of the sulfur settling of chromium hydroxide.
and nitrogen bearing type have also been
used. Mercaptobenzothiazole is an example The design of a removal system and chemi-
of this group which has been found to mini- cal requirements for a given installation must
mize the typical phosphate pitting function be based on individual plant conditions.
of ferrous metals as well as improve the pro- FACTORS INFLUENCING CORROSION LOAD
tection afforded copper and its alloys. Mer-
captobenzothiazole possesses the disadvantage It has been pointed out previously that, in
of not being compatible with chlorination. plant operation, numerous factors contribute
Various chromate base inhibitors also have to the corrosion load.
been used in which organic materials are The effect of increased temperature is
either combined with the chromate or shown by Figure 34-12. Test specimens ex-
blended into the formulation. With both the posed on the pump discharge, at a tempera-
polyphosphate-organics and the chromate-or- ture of 80-90 F showed considerably less
o:anics, zinc salts have been used to add cor- weight loss than specimens exposed at the
Fig. 34-12-Simultaneous Plant Tests at products settle on uncorroded surfaces such

Different Temperatures in the Same System as the test specimens, cell action takes' place
and serious corrosion results. The corrosion
PUMP DISCHARGE (80-90 F) load during the sloughing off period is quite
Days Exposed Loss, mgs Penetration, mpy high and temporarily imparts an added cor-
28 66 2.2 rosion load that must be dealt with during
42 64 1.4
56 78 1.3 this period.
RETURN LINE (110-120 F) Figure 34-13 shows data collected at one
Days Exposed Loss, mgs Penetration, mpy plant at which Dianodic treatment had been
28 92 3.1 applied to a system that had undergone very
42 143 3.2
56 163 2.8 serious corrosion. It will be noted that the
(Recirculating water analysis: Calcium as CaC0 3 256 corrosion rates at both the hot and cold test
ppm, Magnesium as CaC03 144 ppm, Methyl Orange rack were quite high and the losses in the
Alkalinity as CaC03 16 ppm, Sulfate as S04 1400 ppm,
Chloride as Cl 1700 ppm, pH 5.9 to 6.3.) hot were higher, as would be expected. It
should further be noted that the losses were
higher ( 110-120 F) temperature. A Dianodic not consistent. For example, both 90-day
concentration of 60 ppm was maintained in specimens show lower losses than the 60-day
the circulating water and it can be noted specimens. This condition was due to the
that the attack decreased with time, showing incidence of migratory corrosion products ac-
stifling of the corrosion reaction. cumulating on the specimens, setting up cell
action and accelerating attack. The corrosion
In the application of treatment to a system
load imparted by this condition was more
that has already suffered serious corrosion,
than could be handled by the 70 ppm treat-
invariably old corrosion products will slough
ment level and it was necessary to temporarily
off during the initial treatment period. These
increase treatment to 200 ppm to bring the
corrosion products will migrate to other sec-
attack under control.
tions of the system and some will settle at
points of slow circulation. The new areas ex- Suspended matter aside from corrosion
posed at points of sloughing off will become products produce a similar increased corro-
anodic to the adjacent corroded areas. At the sion load. Dust blown into a tower basin in
same time, when the migratory corrosion quantities and fiber thrown from tower wood
also are detrimental. Figure 34-14 will also
Fig. 34-13-Effect of Deposits on Corrosion serve to illustrate the marked corrosion load
Load and Treatment Level imparted by corrosion products and foreign
matter. These data were secured by installing
TREATMENT LEVEL= 70 ppm (DIANODIC)
Days Exposed
Pump Dicharge Fig. 34-14 Effect of Suspended Solids on
Loss, mgs Penetration, mpy
30 576 18.2 Corrosion Load
60 1451 23.0
90 1310 13.8 Pump Discharge (Unfiltered)
112 1459 12.4 Days Loss, mgs Penetration, mpy
Return Line 14 51 3.5
Days Exposed Loss, mgs Penetration, mpy 28 221 7.5
30 738 23.2 42 205 4.6
60 1777 28.0 56 509 8.6
90 1488 15.7 Return Line (Filtered)
112 2151 18.2 Days Loss, mgs Penetration, mpy
TREATMENT LEVEL = 200 ppm (DIANODIC) 14 78 5.3
28 137 4.6
Return Line
Days Exposed Loss, mgs Penetration, mpy 42 58 1.3
14 56 76 1.3
163 11.0
28 73 2.5
42 116 2.6 (Recirculating water analysis: Calcium as CaC0 3 12
56 131 2.2 ppm, Magnesium as CaC03 6 ppm Methyl Orange
Alkalinity as CaC03 60 ppm, Sulfate ds S04 1800 ppm
Chlorides as Cl 1600 ppm, pH range 6.5 to 6.71 '
268 BETZ HANDBOOK
a small homemade filter ahead of the test rosion of the less noble metal presents one of
rack in the hot return line. It will be noted the most difficult problems in plant opera-
that the losses for the cold, unfiltered water tion. Figure 34-15 shows the weight loss data
are substantially higher than losses for hot for coupled and uncoupled steel and ad-
filtered water. Normally, the hot specimens miralty specimens, properly insulated and
would show a much higher loss than the cold exposed simultaneously in the same cooling
specimens. Specimens in the hot were only box.
subjected to the corrosion load imparted by The losses for steel and admiralty, un-
the water characteristics. Specimens in the coupled, were negligible. The coupled speci-
cold line were subjected to the corrosion load mens showed a marked increase in attack on
imparted by the water characteristics and the the steel specimens. While the average pene-
suspended matter. In spite of the fact that tration was not high, most of the attack was
the water was saturated with oxygen, and concentrated. This bimetallic attack can be
the chlorides as Cl and the sulfates as S04 reduced by increasing the Dianodic treatment
were each on the order of 1500 to 1800 ppm, concentration over the 60 ppm level em-
it will be noted that this corrosion load was ployed in this system. However, only 15% of
minor compared to the corrosion load im- the heat exchange equipment in this system
parted by the suspended matter.. was vulnerable to bimetallic attack. Increase
Fig. 34-15-Weight Loss Data on Steel in chemical treatment to control a problem
in a minor part of the system may not be
Specimens Coupled to Admiralty Specimens
justified.
Metal loss, mgs Penetration, mpy Other means of controlling bimetallic cor-
Uncoupled rosion frequently are justified. Cathodic pro-
Steel . . . . . . . . . . . . . . . . . . . . 22 0.50 tection by magnesium anodes, the use of in-
Admiralty . . . . . . . . . . . . . . . . . 13 0.27 sulating gaskets and use of protective paints
Coupled are all measures that properly should be
Steel . . . . . . . . . . . . . . . . . . . . . 148 3.34
Admiralty . . . . . . . . . . . . . . . . . 7 0.15
applied to this problem. The cost of the high
level of chemical treatment required for con-
(All strips were 1 by 3 in. size exposed for 21 days trol of bimetallic corrosion usually is not
in the cracked oil tower overhead cooling box. Average justified in comparison to these other mea-
water temperature, approximately 95 F)
sures.
In many systems the foreign matter may CORROSION OF NON-FERROUS METALS
be due to sand or silt carried in with the
makeup water. Where such contamination Copper and more particularly its alloys repre-
has an adverse influence on the corrosion rate sent the non-ferrous metals most frequently
and requires high treatment concentrations encountered in cooling water heat exchang-
to stifle attack, filtering of the makeup water ers. While these alloys are considered rela-
will prove advisable. In some cases, the tively resistant to corrosion, they are subject
foreign matter originates from other sources to attack by many factors both chemical and
such as dust, lint or fibers blown into the mechanical.
tower, and filtering of the makeup water
Among the more common types of attack
will not accomplish a solution to the problem.
encountered are dezincification, impinge-
In these cases, the installation of a side ment, erosion corrosion, stress cracking and
stream filter should be given consideration. fatigue cracking. Because of the many cop-
The cost of side stream filtration may be per alloys available it is possible to select an
justified economically where excessive quan- alloy that best meets the requirements for
tities of treatment are required to stifle attack the particular service condition. For example,
due to suspended or foreign matter. in refinery heat exchangers the inhibited ad-
Bimetallic contact and the resultant cor- miralty alloys are most widely used because
of their good resistance to corrosion, particu- ing water with oxygen, the concentration of
larly dezincification. dissolved solids due to cycling and contami-
The chemical factors that cause corrosion nation of the circulating water by suspended
of the copper alloys are low pH values, am- matter.
monia, cyanides, hydrogen sulfide and sulfur Maximum heat transfer efficiency demands
compounds, as well as excessive chlorine resid- freedom from scale formation and tubercula-
uals. When attack takes place, the copper tion. Corrosion control is a necessity. While
content of the circulating water will increase. high chromate concentrations of 300-500
The increased copper content produces an ppm will control most corrosion problems, in
added corrosion load that adversely affects the many cases this treatment cost cannot be
steel parts of the system. The copper can justified.
plate out or deposit on the steel parts and The Dianodic' approach to corrosion con-
create cell action that will cause serious pit- trol has proven to be highly versatile, with
ting of the steel. excellent corrosion and scaling protection
provided even under unusual or extreme
Fig. 34-16-Typicallnhibition of Admiralty
conditions. The presence of reducing agents
by Various Treatments such as hydrogen sulfide or mercaptans need
Average no longer be considered an excuse for in-
Treatment Penetration, mpy ferior corrosion control. Periodic process
0.1
leaks, except for economic reasons for prod-
Chromate
Dianodic 0.1 uct saving, do not necessarily require im-
Palyphosphate 2.0 mediate emergency shutdown. From the
Polyphasphate-Mercaptobenzothiazole 0.3
Polyphosphate-Ferrocyanide 1.0 standpoint of corrosion control, these leaks
Untreated 2.0 usually can be tolerated with some sacrifice
in corrosion protection, by use of properly
(Philadelphia (Delaware) supply, 500 ppm added chloride,
oxygen saturated, 120 F, velocity 2.5 feel per second.
designed Dianodic treatment or revision in
~Treatment level 50 ppm.) the treating program during these emergen-
cies. Other unusual corroding or scaling con-
Chromates and chromate base materials ditions, such as aluminum and copper con-
such as the Dianodic and Zinc Dianodic taminil.tion, have also been well contained.
treatments offer maximum protection to cop- REFERENCES
per and copper alloys. Excellent protection is H. L. Kahler and C. George, "A New Method for
afforded even at pH values as low as 5.7 the Protection of Metals Against Pitting, Tubercu-
and the tolerance for ammonia, cyanides, lation and General Corros~on", Corrosion, Vol 6,
pp. 331-340 (1950)
chlorine residuals and elevated temperatures H. L. Kahler and P. J. Gaughan, "Protection of
is improved. Figure 34-16 compares the re- Metals Against Pitting, Tuberculation, and General
sults of the various treatments in protecting Corrosion", Ind. and Eng. Chem., Vol. 44, pp.
1770-1774 ( 1952)
admiralty. It will be noted that the chromate H. L. Kahler and C. George, "Decreasing Cooling
and chromate base materials offer maximum Water Corrosion", Petroleum Refiner, Vol. 34, pp.
protection. These tests were run at velocities 144-148 (July 1955)
of 2.5 feet per second. At lower velocities and H. L. Kahler (to Betz Laboratories, Inc.), U. S.
Patent 2,711,391 (June 21, 1955)
elevated temperatures the superiority of the H. L. Kahler, C. A. Bishof and W. A. Tanzola (to
chromate base materials is even more pro- Betz Laboratories, Inc.), U. S. Patent 2,848,299
nounced. (Aug. 19, 1958)
H. L. Kahler and C. B. George (to Betz Labora-
tories, Inc.), U. S. Patent 2,872,281 (Feb. 3,
SUMMARY 1959)
H. L. Kahler and W. A. Tanzola (to Betz Labora-
The problem of corrosion in open recirculat- tories, Inc.), U. S. Patent 2,900,222 (Aug. 18,
ing systems is intensified as compared with 1959)
W. A. Tanzola, "Corrosion Control in Cooling
once-through systems. Aggravating influences Water Systems", The Petroleum Engineer, Vol.
are the continuous saturation of the circulat- 25, pp. C57-60 (April 1953)
270 BETZ HANDBOOK
35
Closed Recirculating Systems
Advantages of Closed Systems . .............. 271

Corrosion Control. . . . . . . . . . . . . . . . . . . . 272
HROUGHOUT the years the problems con- controlled temperatures, can be circulated
T nected with the removal of heat from
engines and compressors have received con-
through the equipment. From the standpoint
of water caused problems, the small makeup
siderable attention from engine manufactur- water requirements of the closed recirculating
ers, plant operating personnel, and water system greatly simplify the situation. The only
treatment consultants. makeup required is to replace losses from
Engines and compressors first were de- leakage at the pump packing or from drain-
signed for use with low temperature water at age of the system for repairs. Little, if any,
low velocity. These designs resulted in a large evaporation loss takes place.
temperature rise through the system which Because the Closed recirculating system
set up unequal stresses in the metal, causing greatly simplified the water caused problems,
serious failures. To overcome this problem it was frequently assumed that no water
and to increase the efficiency of the equip- treatment was necessary. This assumption
ment, manufacturers' designs have been proved erroneous. Corrosion in closed cooling
changed so that the modern engines now water systems serving the modern engine can
operate with a rapid flow of water at a rela- lead to serious problems. As a result of metal
tively high temperature, and with a tempera- loss, leaks often develop which cause loss of
ture rise of only 10 to 15 F throughout the cooling water and seriously increase makeup
system. requirements and once again intensify the
At one time, little if any attention was water problem. The corrosion products that
given to the cooling water. It was not uncom- develop often cause plugging of the narrow
mon to use water on a once-through basis, cooling passages, restrict flow, decrease heat
taken from a stream that was subject to transfer and result in rapid deterioration of
marked variations in dissolved solids and the engine.
more particularly in suspended solids. Well An increase in water temperature will in-
waters high in hardness often were used so crease corrosion by dissolved oxygen. In a
that heavy scale formation was common. vented system this tendency is opposed by
Soon it was learned that scale formation and the decreased solubility of oxygen at higher
deposit accumulation from suspended solids temperatures. Thi3 fact, of course, is the
could not be tolerated .. basis of mechanical deaeration.
In other cases, cooling of the engines was Figure 35-1 illustrates data on the increase
accomplished with an open recirculating sys- in corrosion at higher water temperatures for
tem using a cooling tower or a spray pond two different conditions. Curve A represents
for dissipation of the heat. This method of results obtained in a completely closed sys-
cooling was a step in the right direction and tem without venting of the oxygen to the
permitted better control of the problems of atmosphere. The increase in the corrosion
scale formation, corrosion and slime. How- rates with temperature is practically a
ever, water-side problems could not always straight line. Curve B shows data secured in
be solved economically, even with this type a vented system. Note the curves are practi-
system. cally parallel up to approximately 170 F.
Curve B then drops off since the lower solu-
ADVANTAGES OF CLOSED SYSTEMS
bility of oxygen at increasing temperatures in
The closed recirculating system then became a freely vented system is decreasing the cor-
popular, and this type of system has been rosion rate faster than it is increased by the
widely accepted as one of the most desirable temperature rise. However, in many closed
ways of dissipating heat from the modern systems it is not possible to freely vent the
engine. This type of system is ideally suited dissolved oxygen which is added in the
to modern engine and compressor operation makeup water. Release of the oxygen at
because a large quantity of water, at carefully points of high heat transfer (and high tem-
272 BETZ HANDBOOK
perature) can result in severe corrosion. with is that of copper and aluminum. Buf-
The metals of construction in the modern fered chromate concentrations of 5000 ppm
engines, compressors, and cooling systems are or more usually are not adequate. It becomes
many and include cast iron, steel, copper, necessary to use other additives such as ni-
various copper alloys and aluminum as well trates or silicates to supplement the chromate
as solders. Nonmetallic components such as base materials.
natural or synthetic rubber, asbestos and In systems that utilize mechanical seals on
carbon also are involved. Wide temperature pumps, the chromate content of the circulat-
variations are encountered up to 180 F or ing water is best limited to a maximum con-
higher and down to ambient (when the en- centration of 250 ppm. This limitation usually
gine is off line) with a corresponding varia- eliminates leakage of chromate bearing water
tion in oxygen content. The waters used in through the seal. Such leakage could result
the cooling system vary in composition with in damage to the mechanical seal by the at-
geographical location. Because of these many trition effect of chromate crystals.
variables, there is ample opportunity for 25.0
oxygen pitting, galvanic action and crevice
attack. Untreated systems will suffer serious
deterioration from these forms of corrosion.
The use of a closed recirculating system 20.0
does minimize makeup water requirements
· and therefore the use of softened water,
condensate, or even demineralized makeup g. 15.0
water becomes economically feasible thus E
eliminating the problem of scale formation. (/)
(/)
The low makeup requirements also make it 0
...1
possible to economically treat the system with
adequate concentrations of effective corrosion
...
:X:
10.0
(!)
inhibitors. ILl
3:
5.0
CORROSION CONTROL
The three most common types of corrosion

inhibitors used for closed cooling water sys-
tems are the chromate base materials, the 100 140 180 220
nitrite base materials and the soluble oil in- TEMPERATURE, F.
hibitors. Of these three basic types, the chro- Figure 35-1 Effect qf Temperature on Corrosion Rate
mate base materials are most commonly used. in Open vs. Closed Systems
Buffered chromates in the range of 500-1000
ppm are satisfactory for waters low in chlo- In closed systems that must be protected
rides and where bi-metallic influences are not against freezing during winter operation al-
encountered. Where chlorides of 100 ppm or cohols or glycols are often used. While buf-
more are present, increased concentrations of fered chromates may be used with methyl
buffered chromate types of treatment may alcohol anti-freeze solutions, it is essential
be necessary. When bi-metallic couples such that the pH be maintained consistently at a
as steel and copper are also present even high level; otherwise reduction of the chro-
greater treatment levels are required on the mate will take place. Glycol anti-freezes and
order of 2000 ppm or more. When aluminum chromate base treatments are not compatible.
or its alloys are also present the problem of The non-chromate type treatments are better
preventing corrosion is even more difficult. suited for handling closed systems containing
The bi-metallic couple most difficult to cope anti-freezes.
The problem of toxicity of high chromate toxicity. Their primary use is to handle these
concentrations on occasion may restrict the situations.
use of these materials, particularly when fre- Soluble oils are good corrosion inhibitors
quent draining of the system is necessary. and afford excellent protection to ferrous
However, usually provisions can be made to metals, copper and steel couples, and are ef-
handle situations of this type. fective even wh~n aluminum is present. Cavi-
Another problem sometimes is encountered tation erosion attack also is controlled ef-
where the use of chromate base materials fectively by these inhibitors. For the most
may not be applicable. Compressor stations part soluble oils owe their effectiveness to
that handle gas high in hydrogen sulfide con- their polar properties which permit adsorp-
tent sometimes are plagued with leakage of tion of a film that acts as a barrier to corro-
the gas from the power cylinder into the sion. These agents must be used at relatively
water circuit. In these cases reduction of high treatment concentrations, on the order
chromate will take place leading to ineffec- of 1 to 2 percent or higher in some cases.
tive corrosion control and deposit formation. It is essential that these inhibitors be ef-
The nitrite base materials find frequent ficiently buffered and emulsified, otherwise
use in the treatment of closed recirculating heavy fouling of the cooling system can take
systems, particularly in those situations where place. Serious interference with heat transfer
the chromate base materials are not appli- and overheating have resulted under these
cable. Nitrite is a suitable corrosion inhibitor conditions. It is essential also that natural
for iron and steel when the pH value is rubber hoses and gaskets be replaced with
maintained in the alkaline range. Nitrite is synthetic rubber to avoid difficulties with
not an effective inhibitor for copper and its swelling and softening of these components.
alloys and can not cope adequately with
couples of these metals with steel. Accord- REFERENCES
ingly, other inhibitors that are specific cor- F. E. Clarke, "Cooling Water Treatment For Na-
rosion inhibitors for copper must be used val Diesel Engines", Proceedings, Engineers Soc. of
Western Penna., pp. 1-15 (1956)
with the nitrite. When aluminum is also
M. Darrin, "Chromate Corrosion Control For
present in the system, the corrosion problem Engine-Jacket Water", Corrosion and Material Pro-
is intensified and nitrates and silicates must tection, Vol. 4, pp. 6, 8-11 (May-June 1947)
be used to supplement the nitrite. The nitrite F. L. LaQue, "The Water Side Deterioration of
Engine Cylinders", Power Engineering, Vol. 58, pp.
base treatments are poly-component, and are 76-77 (Jan. 1954)
required usually at considerably higher treat- F. N. Speller and F. L. LaQue, "Water Side De-
ment levels than the chromate base materials. terioration of Diesel Engine Cylinder Liners",
These treatments have the advantage of be- Corrosion, Vol. 6, pp. 209-215 (1950)
B. C. Thiel, "Controlled Water Jacket Cooling Re-
ing compatible with anti-freeze solutions and duces Maintenance Costs", Petroleum Refiner, Vol.
are less objectionable from the standpoint of 23, pp. 85-91 (June 1944)
274 BETZ HANDBOOK
36
Pretreatment and
On-Stream Cleaning
Pretreatment 0 0 0 0 0 0 0 o o 0 0 0 o 0 o o o 0 0 o 0 0 0 0 0 0 0 o 275
Care of Steel Exchangers o o o o 0 o 0 o o 0 0 0 0 0 0 0 0 0 o 276
Initial Treatment 0 0 o o o 0 0 0 o o 0 0 0 0 0 0 0 0 0 0 o o 276
Other Factors 0 0 0 0 0 o o 0 0 0 o 0 0 o o 0 o 0 0 0 0 0 0 0 o 276
General Application 0 0 0 0 0 0 o 0 0 0 o 0 0 0 0 0 0 0 0 0 0 0 0 277
On-Stream Cleaning 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 277
HE advances made in the control of cor- A marked difference in the care of han-
T rosion in open recirculating cooling
water systems over the past decade have been
dling steel tubed heat exchangers compared
to handling admiralty tubed heat exchangers
phenomenal. The Dianodic methods for cor- is evident from this and similar experiences.
rosion control are capable of decreasing the The corrosion rate of steel under these ad-
corrosion rate of steel to 1 mil per year or verse conditions is many times greater than
less with the pitting function totally pre- the corrosion of admiralty and the quantity
vented or definitely stifled. Excellent proof corrosion products produced is proportion-
tection is secured with treatment concentra- ally greater. When steel tubed heat ex-
tions generally in the range of 30 to 100 ppm, changers are employed, it is essential that
depending on operating conditions and water care be exercised in the initial handling of
characteristics. Because steel can be protected the bundles to avoid corrosion during con-
economically many of the newly designed struction and initial start-up. It is mandatory
plants include all-steel heat exchangers in to clean steel bundles and to pretreat them
place of the more costly and more commonly to render the metal surfaces passive.
used admiralty tubed heat exchangers. The
cost of all-steel heat exchangers is approxi- PRETREATMENT
mately 60% of the cost of heat exchangers
constructed with admiralty tubes. Under Most methods of corrosion control are based
these conditions the capital investment for a on forming a film that acts as a barrier to
new plant is considerably less when steel is stifle corrosion. The rate at which the film
used. or barrier forms will largely determine the
However, the design engineer should be effectiveness of the treatment. Materials that
cognizant of all factors that have a bearing do not form films rapidly will permit cor-
on the service life of steel tubed heat ex- rosion to take place and the result will be in-
changers. The experiences of one plant de- complete film formation and continued cor-
signed with all-steel heat exchangers will rosion. The rate at which the film forms is
illustrate several of the precautions that are related to the inhibitor concentration.
necessary if the expected service life is to be The function of pretreatment is primarily
secured. Shortly after this new plant went to permit rapid film formation to stifle the
into operation, it was found that pressure corrosion reaction immediately by formation
drops across many of the heat exchangers of a uniform impervious film. Under these
were greater than anticipated. Examinations conditions the low treatment levels will main-
of the heat exchangers revealed that heavy tain the film intact and avoid the accumula-
tuberculation was present in the steel tubes tion of corrosion products.
and in the water boxes. Some of the tubes Reference to Figure 36-1, Curve A, shows
were blocked with tubercules that had broken the high initial corrosion rate of steel when
free from other locations. It was evident that exposed to Philadelphia tap water, fortified
the handling of the bundles during construc- with 500 ppm of chloride, and at a tem-
tion of the plant permitted corrosion to start. perature of 120 F. The application of 30 ppm
The pressure of other work attendant with of Zinc Dianodic markedly decreases the rate
start-up did not allow sufficient time for of corrosion (to less than 1.5 mils per year at
initial cleaning of· the heat exchangers for the end of 3 days) . However, if the steel is
removal of debris, grease, and corrosion prod- first pretreated for rapid film formation with
ucts that had accumulated during construc- a treatment concentration of 500 ppm of
tion. No pretreatment of the heat exchangers Zinc Dianodic for only 4 hours and there-
was employed and the intermittent operation after treated with 30 ppm Zinc Dianodic
of the system permitted frequent periods of an even greater reduction in the initial cor-
inadequate inhibitor concentrations so that rosion rate as well as the over-all corrosion
corrosion progressed at a rapid rate. rate results, as shown by Curve C.
276 BETZ HANDBOOK
60 reduction in weight loss and a negligible ac-

cumulation of corrosion products.
CURVE A- NO TREATMENT The low treatment levels normally used for
50
corrosion control in open recirculating sys-
tems should be viewed as the quantities
required to maintain the film intact and to
heal the slight breaks that may occur from
minor variations in environment. Whenever
any serious changes in environment occur
that cause destruction of the film, corrosion
products can accumulate before the film is
CURVE B -30PPM TREATMENT
re-established by the low treatment levels.
Under these conditions, in order to secure
maximum corrosion protection and to mini-
CURVEC-500PPMTREATMENT mize accumulation of corrosion products,
FOR 4 HRS ;30 PPM THEREAFTER.
treatment levels should be increased to re-
establish the protective film as rapidly as
possible.
2 3
DAYS Pretreatment of steel heat exchangers is
Figure 36-1 • Effectiveness of Pretreatment in Decreasing required ...
Initial Corrosion Rates ... for all new bundles
... whenever the exchangers are acid
The data contained in Figure 36-2 demon- cleaned
strate the superior qualities of the film formed ... whenever low pH is encountered
by 500 ppm of Zinc Dianodic compared to ... whenever serious process leakage occurs
5000 ppm of chromate. Pretreatment with ... immediately following start-up after
5000 ppm of chromate for 2 hours permitted inspections.
a 50% reduction in weight loss and a corre-
sponding reduction in the amount of cor- CARE OF STEEL EXCHANGERS
rosion products that accumulate. However,
pretreatment with 500 ppm of Zinc Dianodic INITIAL TREATMENT. The care of new steel
for 2 hours permitted more than an 80% heat exchangers should start with the writing
of the specifications. The design engineer
should include as part of the specifications, a
requirement that the units be delivered free
of debris, grease and other foreign matter
that may initiate corrosion, and that the units
be protected from the weather during transit
and prior to installation. As soon as possible
after installation, and prior to placing the
system in full operation, pretreatment should
be applied to the circulating water to permit
rapid film formation. A procedure of this
type will avoid excessive pressure drops and
assure maximum protection of the units as
well as materially aid in achieving the antici-
pated service life of an all-steel system.
Figure 36-2 • Lift-No Pretreatment. Center -5000 ppm OTHER FACTORS. Once the film is properly
Chromate. Right-500 ppm Zinc Dianodic formed, the normal treatment levels will
serve to maintain the film and prevent cor- heat exchanger may partially or totally re-
rosion under normal operating conditions. move the protective film. Debris and migra-
However, there are a number of factors that tory corrosion products may settle at various
can destroy the film to such a degree that points in the system. These and many other
higher than normal treatment levels wil! be factors will create conditions conducive to
required to re-establish the film rapidly. For corrosion. Pretreatment by the use of higher
example, when an individual heat exchanger than normal treatment levels during start-up
is acid cleaned, the film will be completely will assure rapid repair of the protective film
destroyed. Acid cleaning will leave the steel and minimize the accumulation of corrosion
surfaces in a very active state and the initial products.
corrosion rate will be very high. The normal The precautions outlined above will be
treatment levels usually will not permit suffi- instrumental in achieving maximum service
ciently rapid re-establishment of the film. life from all-steel heat exchanger systems.
Therefore, following acid cleaning and flush-
ing, the unit should be pretreated.
GENERAL APPLICATION
The modern day corrosion inhibitors func-
tion best in a pH range of 6.0-7.0. This same While much attention has been focused on
pH range permits maximum protection to the use of pretreatment for all-steel systems
tower wood and is usually the pH range re- because of the higher corrosion rates that
quired to avoid calcium deposits. Therefore, can occur, the principles of pretreatment have
acid treatment of the circulating water to application to and will prove beneficial for
control pH is a common practice. In the cooling water systems in general. It is true
great majority of these systems excellent con- that the initial corrosion rates for admiralty
trol of pH is maintained. On occasion, how- tubes or tubes of copper alloys in general are
ever, low pH values may develop. If low pH materially lower than for steel, and therefore
is permitted for any extended period of time, the effects of these less voluminous corrosion
partial or complete destruction of the pro- products are not as pronounced. However,
tective film may occur. When this occurs, where admiralty tubed heat exchangers are
treatment levels should be increased to re- employed the systems are bimetallic since the
establish the film as soon as possible. shells and water boxes are constructed of
In other systems, process leakage may intro- steel. While the admiralty heat exchanger
duce hydrogen sulfide and mercaptans. This tubes are not adversely affected, the steel
components of these units do undergo an
type contamination will have an adverse
initially high corrosion rate and produce vo-
effect on most films. Also, leakage of this
luminous corrosion products that interfere
type usually produces a drop in pH as well.
with proper and complete film formation.
Until such time that the leakage can be cor-
The application of the principles of pretreat-
rected the system should be blown down
ment for rapid film formation, followed by
heavily. If it is established that a particular
the use of normal treatment levels for film
heat exchanger is a chronic offender, it should
maintenance will prove of benefit in mini-
be equipped with blowdown facilities so that
mizing corrosion of the steel parts of the
as much of the blowdown as possible is taken
system. In general, improved heat transfer,
from the offending heat exchanger. Following longer service life and less plant maintenance
correction of the leak, treatment levels should
are the benefits that will be secured.
be increased to rapidly re-establish the film.
The periodic inspections required for most ON·STREAM CLEANING
systems will produce conditions during start-
up that may increase the initial corrosion A type of fouling that is not truly scale or
rate. For example, brushing or manual clean- biological slime is often encountered in cool-
ing of the water boxes or other parts of the ing tower systems. This fouling is that caused
278 BETZ HANDBOOK
by a combination of mud, silt and metal is passivated by a high concentration of the

oxides held together by organic matter. Gen- corrosion inhibitor.
erally, the inorganic portion is an accumula- In tower systems suffering, continuous or
tion of airborne particles or suspended matter heavy fouling, quite often it is more effective
from the makeup supply. The organic matter to apply a continual maintenance dose of an
can be caused by process contamination of antifoulant than to resort to periodic clean-
the cooling system, or by organic naturally ing. Deposition can be minimized or pre-
present in the makeup water. vented by application of the proper anti-
Unless removed or controlled, fouling seri- foulant for the conditions involved at low
ously interferes with operation of a cooling treatment levels.
water system. Outage is frequently forced by The new chemical antifoulants function to
decreasing flow rates or by insufficient heat prevent agglomeration of particles, and are
exchange through accumulated masses. most effective against foulants when still in
In the past, generally this type of fouling the ionic or colloidial forms. Chelants, such
has been removed from heat exchange equip- as ethylenediamine tetraacetic acid, seques-
ment by mechanical means or by acidizing trants, such as polyphosphates, dispersants,
with the system out-of-service. During the such as sodium lignosulfonate, and coagu-
last few years, many successful on-stream lants, such as polyacrylamide, are employed
cleaning programs have been performed with for this cleaning function. Treatment levels
the aid of newly developed dispersants or vary from 1.0 ppm to 300 ppm, depending
coagulants. on the nature of the fouling matter and the
An on-stream cleaning program is usually type of antifoulant. Products have been de-
modified to suit the individual chemical and veloped specifically for iron problems, for
mechanical requirements of a particular sys- bentonitic silt and for aluminum as well as
tem. A typical desludging program is con- for general fouling conditions.
ducted for 72 consecutive hours with the
dispersant concentration at the recommended
level. The pH is decreased 1.0-2.0 units dur- REFERENCES
ing this time, and blowdown is increased to
L. F. Probst, "Pretreatment for Effective Corrosion
lower cycles of concentration. After com- Control," The Betz Indicator, Betz Laboratories,
pletion of the cleaning program, the system Inc., Phila., Pa. (Nov. 1960)
Handling and Storage of Concentrated Acid. . . . . 280

Constant Rate Acid Feed. . . . . . . . . . . . . . . . . . . 281
37 Small Quantities. . . . . . . . . . . . . . . . . . . . . . . 281

Larger Quantities . . . . . . . . . . . . . . . . . . . . . . 281
Acid Feeding to
Cooling Water Systen1s
Proportional Acid Feed. . . . . . . . . . . . . . . . . . . . 282

Handling Dilute Acid. . . . . . . . . . . . . . . . . . . . . 283
Sulfur Burners . . . . . . . . . . . . . . . . . . . . . . . . . . 284
Proper Points for Acid Discharge . . . . . . . . . . . . 284
Acid Filters . ............................ 285
Materials of Construction. . . . . . . . . . . . . . . . . . 286
Some Mqjor Sqfety Rules Wizen Handling Acid. . . 286
280 BETZ HANDBOOK
of different systems are em- air pressure systems and pumps are typical of
A
NUMBER
ployed for the addition of acid, involving the pressure type. While most of these units
an even greater number of variations in feed are operated in a constant rate manner with
equipment. Typical systems and typical equip- manual adjustment, many methods can be
ment are discussed here but these are by no made proportional to makeup water flow or
means the only acceptable systems or equip- pH by the addition of suitable metering and
ment. control equipment.
The application of acid, usually sulfuric The primary purpose of this discussion is
acid, in the treatment of cooling tower cir- to offer suggestions and information con-
culating waters is a generally accepted prac- cerning practical acid handling and feeding
tice. Acid is a necessary adjunct in the treat- systems and to outline some of the precau-
ment of many waters used in cooling tower tions which are necessary in any well designed
systems and is used for the purpose of re- system. Because of the many factors influ-
ducing the total alkalinity concentrations encing the selection of suitable acid feed
wh_ich would otherwise develop. equipment, such as the type and quantity of
By acid treatment the formation of calcium acid to be fed, labor requirements, and pur-
carbonate scale is prevented, deposition of chase and installation costs, each system re-
calcium phosphate is controlled and delignifi- quires designing on an individual basis.
cation of tower wood is also controlled. The
advent of treating systems which function HANDLING AND STORAGE OF
most effectively in the pH range of 6.0-6.5 CONCENTRATED ACID
has increased interest in acid feed and Concentrated sulfuric acid is usually shipped
brought the need for better equipment and in glass carboys or in tank trucks or cars.
closer control. Figure 37-1 shows a typical acid storage
The types of systems or equipment for system and, when sized properly, will be able
acid addition can be divided into several to receive full tank car or tank truck ship-
categories such as proportional feed and ments while maintaining an adequate reserve.
constant rate, gravity or pressure systems, Where acid is delivered in carboys, a number
and concentrated acid feed or dilute acid of foot and hand operated air pumps are
feed. The drip feeder and decanting feeden available, from different manufacturers, for
are typical of the gravity units while eductors, the purpose of emptying these carboys. A
Figure 37-1 • Acid Delivery, Storage and Transfer System
STORAGE TANK 10,000 GALS CAPACITY
VINT z• IX, HVY, STEEL PIPE STEEL PIPE
BALL JOINT
P'IIID TANK
CAST STEEL CAR TANK ~AX. PRESSURE

PLUG AND SEAT NOT TO fXCUD 30 LIS PER SQ IN
END NEAR GROUND
WITHIN SIGHT OF OPERATOR
ACID EGG f SHELL
small electrically driven air pump is also Larger quantities are better handled by
available and suitable for emptying of car- equipment described in the next section.
boys. The pump uses a self-sealing rubber
connection in the throat of the carboy and, CONSTANT RATE ACID FEED (large Quantities)
when operated, develops approximately 2.5
psi, forcing acid through the discharge tube. Controlled air pressure, with flow measured
The corrosive liquid contacts only the tubing by a rotameter on the air line, can be used
through which it flows. to provide close' control over the feed of acid
from a clay tank to the point of discharge.
When storage tanks are employed for the
This arrangement is illustrated in Figure
handling of large scale deliveries, the tank
37-2.
is usually located so as to provide gravity
flow of acid to the day tank or point of use. Suitable decanting feeders similar to that
In emptying shipping containers, it is essential illustrated in Figure 37-3 are available. The
that excessive air pressure is not used. When larger and more expensive types are usually
transfer pumps are needed for introducing lead lined and offer the least difficulty in
acid to the storage tank, it is well to main- feeding acid when the acid used has con-
tain a pressure of approximately 5 psi on the siderable sediment. These units can be sup-
tank car or truck in order to supply a steady plied either as adjustable constant rate
flow of acid to the suction side of the feeders or the feeder may be controlled to
transfer pump. provide acid feed in proportion to makeup
Complete precautions to be followed when water addition or pH control. The smaller
handling sulfuric acid with particular refer- rotary-dip types have shown satisfactory serv-
ence to unloading of tank cars and trucks ice when properly installed and maintained.
are covered in Manual Sheet TC-1 issued by These units are usually made of polyvinyl
the Manufacturing Chemists Association of chloride (PVC) plastic for those parts in
the U. S. This bulletin may be purchased at contact with the acid, with acid resistant
nominal cost from the association at 1825 metals used wherever possible in other parts
Connecticut Avenue, N. W., Washington 9, of the equipment.
D. C. Those responsible for handling tank A type of pump commonly employed in
car and tank truck equipment and fittings the metering of acid to a cooling system is
should be completely familiar with the en- illustrated in Figure 37-4. The unit is a
tire contents of this bulletin. diaphragm type pump and acid does not
come in contact with working parts of the
CONS'[ANT RATE ACID FEED (Small Quantities) pump other than the check valves. All ma-
Simplified drip and syphon feed systems in- terials in direct contact with the acid are
volve the least expenditure of time and equip- made of acid resistant or non-corroding
ment in setting up an acid feed system. How- metals and plastics.
ever, syphon feed is generally unsatisfactory. Acid pumps, when used in metering service,
Gravity feed control is improved by em- require the removal of all suspended matter
ploying a sludge pot or strainer. A large area from the acid. Sediment present in com-
filter using glass wool provides the best mercial acid will cause the check valves to
removal of suspended matter from the acid clog and the pump will cease to function in
before the acid reaches the needle control such instances. Therefore, a sludge pot is
valve. These systems are suitable only for the desirable and a glass wool filter is mandatory
addition of small quantities of acid and where on the pump suction line if attention-free
continuous attention is possible. service is to be expected from the pump.
The above described methods of feed are Several suitable diaphragm type pumps are
usually adequate for feeding concentrated available on the market covering all sizes
acid quantities up to 2 or 3 gallons daily. and capacities which may be required.
282 BETZ HANDBOOK
MERCURY LOOP SEAL

OR MICRO:-BEAM VALVE
DISCHARGE LINE TO DILUTION TROUGH
(AVOID SIPHON),
BREAK
~FUNNEL
PRESSURE
--------TO ORA IN
AIR
/
OR GAS
ACID FROM STORAGE T~K I

EMERGENCY STOP VALVE DAY TANK DILUTION TROUGH
(NORMALLY OPEN)
Figurr 37-2 • Air Pressurized Feed System
PROPORTIONAL ACID FEED
A pH controller, equipped for starting or

stopping electrical equipment, can be used to
control the flow of acid into a system by
regulating a solenoid operated shut-off valve
or motor driven pump. A variation of this
system involves the use of a diaphragm valve
governed by the pH controller, usually
through an air operated motor. This method
of feed of acid is continuous and proportional,
the diaphragm valve opening wider to permit
more acid to flow into the system whenever
Courtesy Clarkson Co. Courtesy Lapp Insulator Co., Inc.
Figure 37-3 • Rotaty Dip Decanting Feeder of PVC Figure 37-4 • Diaphragm Pump Equipped with Auto-
Construction matic Stroke Adjustment
Figure 37-5 • Automatic pH Control Using Positioning Control Valve
RECORDING pH CONTROLLER AND AMPLIFIER
'"'""'"~"'"<)
TANK
MOTORIZED
_ - P H ELECTRODES
I
EMERGENCY STOP VALVE
(NORMALLY OPEN)
SLUDGE POT
required. Likewise, as acid requirements de- HANDLING DILUTE ACID

crease clue to decreased flow of water or
lower pH, the diaphragm valve will close When the quantity of concentrated acid to
down and decrease the feed of acid. The be fed is small and does not permit easy
control valves are constructed of acid resist- addition to a system, a dilute acid feed sys-
ant alloys and plastics also. A typical system tem may be set up. Certain precautions must
of this type is shown in Figure 37-5. be followed in order to obtain proper service
A rather deluxe system of acid control can life from the equipment. Figure 37-6 illus-
be employed when desired, by utilizing a trates a suitable method for handling the
variable speed motor driven pump together dilution problem.
with a variable stroke mechanism. Usually, In preparing dilute acid, not less than 18
in such instances, the driving motor speed is gallons of water should be used for each gal-
controlled by a flow meter with the stroke lon of concentrated acid ( 66° Baume sulfuric
variations controlled by a pH controller. An acid) . The dilute acid tank should be filled
electronic control system is used to supply the with the proper amount of water and the
power to operate the variable speed motors. mixer started. The concentrated acid must
be added to the water, as indicated in the
diagram. The dilution should proceed slowly
ACID U~ON STOIIAGl TANK
DILUTION WATIII SUPPlY
in order to avoid excessive heating, which
.,_ ~ STIILTANK
may buckle the tank lining. When draining
the concentrated acid into the water, a
INIJIOUICY ITO~ALVf.
(NOJI.YALLV OPIH)
draining time of at least two minutes per
gallon should be allowed. Dilution should
fiUIIIII Oil LlAD LJNID
ITIIL TANK
proceed more slowly, if necessary. The con-
(USl A JIOUHO TANK)
centrated acid tank should be used for meas-
uring only, not storage.
In some instances, porcelain crocks can be
used in place of lined tanks. In any case, the
crock must be placed inside a steel tank for
Figure 37-6 Acid Dilution .'l)stem safety reasons. Crocks have not proven to be
284 BETZ HANDBOOK
'BAUME sulfur, will cause some reduction of chromate

5' 100 15° 20" 25' 300 35' 40° 45' 50' 55' 60' 66'
in those systems which are treated with chro-
90 mate base corrosion inhibitors, causing an
80
70
increase in treating costs.
60
50
PROPER POINTS FOR ACID DISCHARGE
40
~~
...
..;
0:
....:::>
~ -10
30 ~
20
10
0
~2: f---1--
~r-
~~ I-
t A typical mixing trough as recommended for
the introduction of acid to a typical cooling
tower system is shown in Figure 37-8. Make-
"'II.,. -20 ~ up flows up to 500 gpm are satisfactorily acid
"'.... -30 ) +-\ treated and passed through such a trough
-40
-!50
although, in many instances, a small portion
-60
I\ \ of the makeup water, between 20 and 50
-70 gpm, is by-passed through a suitably sized
... -B?
-90 trough and provides sufficient dilution water
-100
I\JI for the average system. The discharge from
-110
a properly constructed trough of this type
0 5 10 15 20 25 30 35 40 45 !50 55 60 65 70 75 00 85 90 95 100 may be introduced to the flume leading to the
PERCENT H1 S04
cold well or into the sump close to the pump
Figure 37-7 • Freezing Points of H2S04 Solutions suction in large systems. The discharge point
should be in the range of 25 feet away from
particularly long-lived in such applications the pump suction, as a minimum.
and complete rupture of the crock with a Installations should be avoided which pro-
full load of acid could be disastrous. Also, vide for the introduction of acid, concentrated
under no circumstances, should an ordinary or diluted, into the makeup water line or
unlined steel tank be used for the dilution into the circulating water line where reaction
tank. will take place and carbon dioxide will be
The freezing point for various strengths of released. Even though the acid may be inter-
sulfuric acid is given in Figure 37-7. It will mixed with the water in special alloy spools
be noted that 93% acid (66° Baume), the or other mixing devices, severe corrosion will
strength most generally used, has a freezing develop in those portions of the lines com-
point of - 29 F. As such, insulating pre- posed of ordinary steel and very short service
cautions are needed only in the coldest areas. life can be expected. Likewise, if such lines
should discharge close to concrete surfaces,
USING SULFUR BURNERS attack of the concrete may take place as well.
Interest develops periodically in the applica- Dependence upon check valves should be
tion of sulfur burners for acid treatment of avoided completely where they are expected
various water supplies. In certain instances, to prevent the backing up of water from a
sulfur burners are ideally suited as in the makeup line, for example, into the acid line.
case of some applications in the paper in- When check valves fail in such service, very
dustry. On the other hand, for cooling water rapid corrosion of the acid line develops,
treatment, the sulfur burner has not proved particularly since the temperature is elevated
sufficiently reliable and trouble-free. The dis- by the contact of water and concentrated
advantages, in most cases, far outweigh the acid-the resulting situation can become
saving in cost of acid. In most instances, quite dangerous.
combustion difficulties create excessive main-
tenance demands. In addition, the use of It is just as important to design proper
sulfur dioxide gas, resulting from the burning methods of introducing acid to a system as
CONSTRUCT TROUGH AND BAFFLES WIDTH AND HEIGHT OF TROUGH TO BE !lASED UPON
OF REDWOOD OR CYPRESS AMOUNT OF WATER USED. FIGURE HEIGHT OF 51 DE
AS TWICE EXPECTED WATER DEPTH.
END BAFFLE
FROM CONC. ACID FEEDER
WATER
GLASS TIP END 3"

ABOVE MAX. WATER
GLASS PLATE BELOW ACID DISCHARGE
Figure 37-8 • Acid Dispersal and Dilution Trough
it is to select dependable feeding equipment.

In general, the acid should be so introduced
that no other material may enter the acid
line under any circumstances. The final dis-
charge of acid treated water should be into a
section, where low pH water cannot attack
pumps or piping materials or concrete. Any
released carbon dioxide may be readily aer-
ated from the treated water.
ACID FILTERS
Of all the pieces of equipment available,

probably the most important to the con-
tinuous trouble-free functioning of feeding
systems employing pumps and control valves
is the glass wool filter. All commercial acid
has varying quantities of sediment and it is
C]Uite possible for these small particles to
gradually clog any feeding system since the
actual flow of acid in most plants is a small
amount. Glass wool filters and sludge pots
eliminate practically all feeding interruptions
due to sediment. Figure 37-9 shows a typical Courtesy Milton Roy Co.
glass wool filter for application in concen- Figure 37-9 • Combination Sludge Trap and Glass
trated acid feeding systems. Wool Filter for Pump Suction. Line
286 BETZ HANDBOOK
MATERIALS OF CONSTRUCTION pumps or low air pressure not exceeding 3

psi. If this is not sufficient, use small self
While sulfuric acid is considered a strong priming transfer pumps.
acid, the properties of the concentrated acid
Avoid excessive pressure in acid lines.
allow certain materials to be used in contact
with it and give satisfactory service life. For No section of an acid line should be valved
example, steel is normally used for storage shut without suitable vents-action of the
acid upon piping may cause sufficient pres-
tanks and provides satisfactory service when
sure to develop to burst the line.
moisture is removed from air drawn into the
tank by means of suitable water absorption Proper drainage and reservoir facilities
materials. Steel may be used for pipe lines should be provided for any tank which may
carrying concentrated acid where the lines overflow or leak. Reservoir pits may be filled
are kept full of concentrated acid at all times. with limestone.
In some instances, it is desirable to consider A plug type valve should be used in stor-
the installation of PVC or some chrome- age tanks for safety shut-off in case of line
nickel alloys for piping in order to provide breakage. Use only one valve between the
further dependability. storage tank and day tank. Lubricate valves
Steel day tanks which are refilled daily or with acid-resistant grease and test operate
oftener should be lead lined in order to twice weekly. Ordinary grease is not acid-
provide satisfactory life because of the cor- proof.
rosion which develops with a thin film of Air vents should be piped to a point close
acid on the tank surfaces remaining in con- to the ground in order to avoid dangerous
tact with the surrounding air as the acid splashing in case the tank overflows.
level is drawn down. Discharge piping, as All measuring or day tanks should be of
previously noted, should not be connected lead or lead lined steel or other acid resistant
into any other line where back pressure may materials. All lined tanks should be circular.
force water into the acid piping or system. In Proper clothing should be worn by all
addition, where acid is dripped into contact persons working with acid. Rubber gloves,
with the water, precautions such as the rubber aprons and close fitting goggles should
glass tip on the end of the line and the glass be worn. Water in ample quantity must be
drip plate in the mixing trough add con- immediately available.
siderably to the freedom from maintenance Proper care in handling of fittings is neces-
problems. sary. Proper installation of fittings and lines
Pump bodies and valves may be made of will prevent splashing, leakage or spraying.
cast iron, but usually better service life is Fittings should never be struck with tools or
obtained when suitable alloys or plastics are other hard objects for any purpose and
employed and the increased life compensates especially where fittings are under pressure
for the increased cost several times over. or acid tanks are full.
Open flames of any kind and smoking
SOME MAJOR SAFETY RULES WHEN should be prohibited in acid areas due to the
HANDLING ACID danger of hydrogen explosions. For illumina-
In unloading tank cars, use clean, dry air at tion, use approved safety flashlights or suit-
no more than 30 psi. Contents of tanks should able electric lighting, designed for hazardous
be thawed only in strict accord with recog- areas.
nized practice and never by using torches or Proper mixing of acid with water to be
building a fire under the car or around the treated is essential. Mixing troughs should
tank. provide an intimate mixture of makeup water
When unloading carboys, use approved and acid, discharging at a point in the tower
as far removed from the intake to the circu- ment seriously interferes with the operation
lating pumps as practicable. This procedure of pumping or gravity feed operations.
will prevent stratification of the acid which Solenoid valves and check valves should
may cause considerable damage to pumps not be relied upon where large quantities of
and concrete. Makeup quantities up to 500 acid are involved.
gpm may be treated in this manner. Water !lhould never be introduced into
The ends of discharge lines from acid sulfuric acid. Feeding systems should be
feeding equipment should be equipped with designed to prevent the possible introduction
glass tips in order to prevent corrosion of the of water to concentrated acid in the event
tip in the corrosive atmosphere. A glass plate of mechanical failure or human error.
should be placed on the bottom of the mixing
trough at the point of acid introduction in REFERENCES
order to prevent attack at this point. W. J. Gossom and J. 0. Johnson, "Cooling-Tower
Suction for acid feeders should be drawn Water Treatment", The Oil and Gas Journal, Vol.
55, pp. 91-95 (Dec. 9, 1957)
at least 2 inches above the bottom of the day
J. H. Richards, "Chemical Feed Systems in
tank and the day tank periodically should be Modern Water Treatment", Power Generation,
drained completely in order to remove ac- Vol. 52, pp. 62, 63, 122-125 (Aug. 1948) and pp.
cumulating sediment. The acid purchased 61-63, 118, 120, 122, 124, 126 (Sept. 1948)
J. T. Russell, "Chemical Costs Drop Sharply in
should be sampled frequently to make cer- This Installation for Automatic Treating of Cool-
tain that a high quality acid is being obtained ing Water", The Oil and Gas Journal, Vol. 53,
with a Ininimum of sediment. Excessive sedi- pp. 109-111 (July 12, 1954)
288 BETZ HANDBOOK
Difficulties Due to Biological Fouling . ........ 289

Control of Biological Fouling. . . . . . . . . . . . . . . 289
Chlorine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
38 Chlorine Residual. . . . . . . . . . . . . . . . . . . . . . 292

Chlorination Programs. . . . . . . . . . . . . . . . . . . 292 ·
Slillle and Algae Control
F;(fect of Chlorine on Tower Wood . ........ 293

Bromine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
Acrolein . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293
Chlorinated Phenols. . . . . . . . . . . . . . . . . . . . . 294
Continuous vs. Shock Treatment. . . . . . . . . . . 295
Copper Salts. . . . . . . . . . . . . . . . . . . . . . . . . . 296
Quaternary Ammonium Compounds . . . . . . . . . 297
Miscellaneous Toxicants. . . . . . . . . . . . . . . . . 298
General Considerations in Selecting Slime
Control Agents. . . . . . . . . . . . . . . . . . . . . . 298
NDUSTRIAL users of cooling water systems

I have long recognized the need for con-
trol of biological fouling in order to avoid
loss in heat transfer and to minimize the
added corrosion load imposed by slime and
algae growths. While difficulties due to bio-
logical fouling are encountered in both once-
through and recirculating cooling water sys-
tems, the problem is more severe in open
recirculating systems. Exposure of the cir-
culating water to sunlight in cooling towers
and spray ponds favors the growth of algae.
Increases in temperature, concentration of
bacteria food and the greater incidence of
air-borne contamination increase the slime-
forming potential in open recirculating sys-
tems.
Because slime and algae problems are en-
countered to a greater extent in open re- Figure 38-1 • Serious Fouling with Slime Growths
circulating systems, this discussion has been
directed to the control of biological fouling
in such systems. However, the difficulties these deposits are various bacteria, filament-
described and the principles of their control ous fungi, yeast and occasionally protozoa.
are applicable in most cases to once-through Slime growths may also contain dead algae
systems as well. which have migrated from other areas and
have become entrapped in the mass. Slime
DIFFICULTIES DUE TO BIOLOGICAL FOULING growths may occur in illuminated areas or
in dark areas. In general, dark areas favor
Biological fouling in open recirculating cool- the growth of slime since the microorganisms
ing systems is the result of excessive growth involved do not utilize light in the manu-
and development of different members of facture of their food but are totally depend-
the lower forms of plants, namely: algae, ent upon the food available in the cooling
fungi and bacteria. In general, the principal water.
difference between algae and fungi is that Slime deposits on process equipment will
algae are capable of manufacturing their so retard heat transfer as to create a serious
own food whereas fungi cannot. Except for loss in efficiency. Biological fouling on metal
autotropic types, such as iron consuming and surfaces will create differential oxygen con-
sulfate reducing types, bacteria do not manu- centration cells and result in serious pitting of
facture their own food. the metal surfaces. Figure 38-i illustrates a
Because sunlight is necessary for growth heat exchanger that is seriously fouled with
and development of algae, abundant algae slime growths.
growths occur in spray ponds and in exposed
parts of cooling towers. The principal types
of biological growths that occur in the non- CONTROL OF BIOLOGICAL FOULING
exposed portions of a cooling system are
referred to as slime. Slime is an accumulation Many different types of chemical agents are
of microorganisms and their excretions to- employed for microorganism control in re-
gether with whatever inorganic and/or or- circulating cooling water systems. The pri-
ganic debris may become embedded in the mary purpose of the chemical agent is to kill
mass. The microorganisms usually found in and/or inhibit the growth and accretion of
290 BETZ HANDBOOK
Figure 38-2 • Common Microorganisms Found in Cooling Water Systems
_L
1-Ulothrix X250 2-Chlorella XSOO
3-Asterionella X250 4-Cosmarium X250

organisms. If a given toxicant is present in a Chlorine is highly toxic and acts quickly
sufficient quantity or is highly toxic, the bac- to kill bacteria. A chlorine residual of 0.3-
teria will be killed. However, at levels below 1.0 ppm will usually destroy most microor-
its killing concentration, the chemical used ganisms. However, since chlorine acts on all
will only inhibit the growth of the organism. oxidizable material, the chlorine demand of
At extremely dilute concentrations it is not a water is greatly increased by the presence
uncommon for the chemical agent to ac- of organic matter, hydrogen sulfide and fer-
tually stimulate the growth of microorgan- rous iron. Thus, sufficient chlorine must be
isms. Many substances will kill microorgan- added to the water to develop the residual
isms only at high concentrations, but they necessary for toxicity and to satisfy the de-
inhibit growth and development at signifi- mand of other materials present in the water
cantly lower concentrations. Such substances which are readily oxidized or possess the
are known as bacteriostatic agents whereas ability to absorb chlorine.
those chemical agents that kill bacteria are Chlorine is stable in neutral, acid and alka-
known as bactericides. Dilute solutions of a line waters although its germicidal efficiency
bactericide may act as a bacteriostatic agent is reduced somewhat under alkaline condi-
when its concentration is too low to kill bac- tions. When large amounts of chlorine are
teria. required, it is generally obtained in liquid
CHLORINE. Chlorine is probably the agent form for economy and convenience, and fed
most widely employed for control of micro- to the system by means of a chlorinator. Be-
biological deposits in recirculating systems. In cause of its corrosivity, the use of chlorine
the absence of substances which cause a high involves some hazards in handling. However,
satisfactory chlorination equipment is avail-
Fig. 38-3-General Comparison of able for this purpose.
Chlorination Programs The amount of chlorine required for the
Program Remarks control of biological fouling in any individual
Continuous Chlorination Most effective system is governed by numerous factors.
-Free Residual Most costly There may be wide variations in chlorine re-
Not always technically or eco-
nomically feasible due to high
quirements between different tower systems
chlorine de!lland. in the same plant. Chlorine requirements are
Continuous Chlorination Less effective affected by the quality of the makeup water
-Combined Residual Less costly to the system, water temperatures in the
Inadequate for severe prob-
lems. system, air to water ratio of the system, the
Intermittent Chlorination Usually effective amoQnt of contamination by reducing agents
-Free Residual Less costly than continuous and the type and amount of bacterial and
chlorination
fungal contamination.
Intermittent Chlorination Least effective
-Combined Residual Least costly The manner of chlorine feed, whether con-
tinuous or intermittent, and the residual
chlorine for control of the problem are also
chlorine demand, chlorine is usually the most
individual to a specific system and will vary
economical method of treatment. A survey of
between towers in the same plant. For con-
16 Gulf Coast refineries and chemical plants
trol of the slime and algae problem in any
shows that continuous chlorination is em-
system, it is necessary:
ployed in 24 towers and intermittent chlo-
rination in 10 towers. Commercial biocides 1. That sufficient chlorine be fed to secure
are used in 26 towers either alone or in the necessary residual chlorine content of
conjunction with chlorine. One tower is the treated water.
treated with sodium hypochlorite and in only 2. That this necessary chlorine residual be
one tower are no slime control measures maintained in the system for the time in-
employed. terval for microorganism control.
292 BETZ HANDBOOK
CHLORINE REsiDUAL. While a chlorine resid- While continuous chlorination is obviously

ual of 0.3-1.0 ppm normally is adequate to advantageous from the standpoint of ef-
kill most microorganisms contaminating cool- fectiveness, it also is the • most expensive
ing water systems, it is necessary in many cases method. In addition, continuous chlorination
that a free chlorine residual be present rather of a recirculating cooling water system is not
than a combined residual. The free available always practical and may not be economical.
chlorine residual is defined as that portion of Because of its volatility, chlorine is aerated
the total residual chlorine which will react from the water in passing over a tower or
chemically and biologically as hypochlorous spray pond. The presence of any organic
acid or hypochlorite ion. It is in this form that agents in the cooling water treatment will in-
chlorine exerts the most potent bactericidal ef- crease the chlorine demand as will the
fect. The combined available chlorine residual presence of reducing agents due to process
is defined as that portion of the total residual leaks. If contaminants such as hydrocarbons,
chlorine which will react chemically and mercaptans and sulfides are present in ex-
- b~ologically as chloramine or organic chlo- cessive quantities, it may be impossible to
ramines. In this form, chlorine is a relatively feed sufficient chlorine for control of slime.
mild bactericide and oxidizing agent. Investi- Complete sterility of the circulating water
gations have shown that for a 100% bac- is not possible because of continuous con-
terial kill with the same contact time, ap- tamination by air-borne organisms. However,
proximately 25 times as much combined sterility is not necessary for control of or-
residual chlorine is required as is necessary ganic growths. It is necessary only to reduce
with a free residual. To obtain a 100% kill, the bacterial population to the point where
using the same amount of free and combined troublesome accumulations are avoided.
available residual chlorine requires approxi- Intermittent chlorination is a common
mately 100 times as long a contact period for practice because of the economy that can be
a combined residual as is necessary for a free secured. It is usually possible to adequately
residual chlorine. control biological fouling by means of a pro-
CHLORINATION PROGRAMS. Chlorination program of intermittent chlorination. Of course,
grams used in recirculating cooling systems fall the schedule of chlorine feed and the residual
within the categories listed in Figure 38-3. maintained must be tailored to the particular
Fig. 38-4-Relative Toxicity of Different Phenolic Compounds
Test Organism
Compound Aerobacter Bacillus Aspergillus Penicillium

aerogenes mycoides niger expansum
Inhibiting Concentration in ppm

Chloro-o-phenylphenol 40 25 35 35
2-tert-butyl-4-chloro-5-melhylphenol 35 6 95 75
4, 6-dichlorophenol 30 0.7 2000 40
o-benzyl-p-chlorophenol 55 5 80 80
Sodium soli of:
o-phenylphenol 200 200 150 150
2, 4, 5-trichlorophenol 20 15 15 7
chloro-2-phenylphenol 60 30 55 30
2-chloro-4-phenylphenol 45 20 65 50
2-bromo-4-phenylphenol 60 15 150 80
2, 3, 4, 6-telrachlorophenol 400 7 20 30
pentachlorophenol 200 4 25 30
problem. However, a program that is gener- some investigations would be conducted on

ally successful is to chlorinate daily until a the usefulness of bromine in microorganism
free residual of 1 ppm chlorine has been control. Some applications of bromine have
maintained for a four hour period on the been made to recirculating cooling water
cooling water returned to the tower. The systems, particularly on the West Coast. The
period of chlorination may exceed four hours conclusions reported from several installations
because of the time required to raise the are conflicting. -.Some plants apparently have
residual to 1 ppm at which level the residual obtained satisfactory slime control through
should be maintained for four hours' time. the use of bromine alone. In other applica-
Adjustment in this basic schedule can be tions, it has been concluded that bromine is
made in accordance with the severity of the most effective when used as an intermittent
problem and the plant conditions observed. "shock" treatment in conjunction with con-
There are many different chlorination pro- tinuous chlorination.
gram variations between continuous chlorin- The use of bromine is not widespread.
ation to a free residual as the most effective Probably the most important single factor re-
and most expensive program, and intermit- tarding interest in bromine is the hazard in-
tent chlorination to a combined residual as volved in its handling. Bromine is a very ac-
the least effective and least expensive. tive oxidizing agent, both as a liquid and a
EFFECT oF CHLORINE oN TowER WooD. vapor. Serious hazard exists by contact with
the liquid or inhalation of the vapor. The
While chlorination usually will prove to be maximum allowable concentration in air is
an over-all satisfactory method for control of
less than one half that of chlorine. The
biological fouling in recirculating cooling
liquid rapidly attacks the skin and other tis-
water systems it must be used judiciously
sues to produce irritation and burns which
since excessive concentrations of chlorine
heal very slowly. The vapor concentration
have an adverse effect on cooling tower
considered safe for eight-hour exposure is less
wood. The subject of cooling tower wood than 1 ppm while a concentration of 10 ppm
deterioration is covered in a separate chap-
can hardly be tolerated. Because of its haz-
ter. In recirculating cooling water systems
ardous nature and because of its relatively
that employ cooling towers, there has been high cost, it seems unlikely that bromine will
an increased trend toward the use of non-
find wide use for microorganism control in
oxidizing biocides to replace or to supplement
cooling water systems.
the use of chlorine. The expanded use of
non-oxidizing biocides has resulted from the AcROLEIN. Acrolein, a highly reactive liquid
realization that chlorine alone does not han- agent, provides exceptionally good control
dle the complete microbiological problem of against microbiological growths in once-
a cooling tower system. Chlorine does not through and recirculating water systems. The
minimize biological attack of the tower wood material is particularly useful where chlorine
and moreover can have an adverse effect on and/or non-oxidizing biocides are ineffective
tower life. An effective slime control program for problems of high organic demand water,
should handle the complete microbiological process contamination and marine infestation.
problem. It should have as its objectives, The specially stabilized biocide exhibits a
prevention of loss in heat transfer, minimiz- broad spectrum of biological activity at ex-
ing corrosion and should protect the tower tremely low concentrations. Normally, a re-
wood. Accordingly, where chlorine is used sidual in the order of 0.2-1.0 ppm is adequate
for slime control it is advisable to periodically to kill most microorganisms that contaminate
feed a non-oxidizing biocide so that biological cooling water systems. This level is particu-
surface attack of cooling tower wood can be larly toxic to sedentary marine growths.
minimized. Another important advantage of acrolein
BROMINE. Bromine is quite similar to chlorine is its ability to function as a herbicide in
in its chemical properties and it is natural that controlling submerged and floating weeds as
294 BETZ HANDBOOK
well as algae in irrigation canals, ditches, nolic and chlorinated phenolic compounds to
ponds and other bodies of water. This biocide standard test organisms. The bacteria used in
exhibits a remarkable and unique specificity these tests were Aerobacter aerogenes, a capsu-
for aquatic plants as compared to terrestrial lated non-sporeforming organism and Bacillus
plant life. Although considerable data in dif- mycoides, a sporeforming bacterium. The
ferent environments have been accumulated, fungi employed were Aspergillus niger and
the aquatic toxicity of acrolein has not been Penicillium expansum. It can be noted that a
definitely established for all species of fish. material may be inhibitory to the growth of
The material is extremely toxic in some cases, bacteria at relatively low concentrations as
whereas the tolerance is reasonably high in the case of 4, 6-dichlorophenol, but re-
under other conditions and for certain species quires high concentrations of 2000 ppm to
of fish. exhibit inhibitory power against the fungus
Precautions must be taken when handling Aspergillus niger. The sodium salts of tetra-
acrolein. However, a specially designed pres- chlorophenol and pentachlorophenol are in-
sure feed system available for this biocide hibitory at low concentrations against the
-embodies several safety features so that the fungi and Bacillus mycoides, but concentra-
agent can be applied without hazard. tions of 400 ppm and 200 ppm are necessary
Acrolein is completely noncorrosive to all to inhibit the growth of Aerobacter aerog-
metals, which is another advantage over enes.
It should be realized that the results ob-
chlorine. Also, the biocide may be deactivated
tained by a given laboratory procedure do not
whenever desired by injecting sodium sulfite
necessarily indicate the dosage required to
into the water stream. Approximately 3.0
give satisfactory microorganism control in
ppm of sulfite per 1.0 ppm of acrolein is re-
the field nor do the data indicate the relative
quired for this purpose. The reaction is rapid,
effectiveness of each material under the dif-
and occurs over a broad pH range (5.0-7.0
ferent field conditions encountered. Rather,
optimum).
such laboratory data indicate that there are
Acrolein residuals can be determined by wide variations in the effectiveness of a toxi-
means of a specific analytical test. The cant against different organisms and that ef-
method is accurate to the range of 50 parts fective control of a microbiological complex
per billion. sometimes can be achieved only through a
From a cost standpoint, acrolein will not combination of several treatments.
replace chlorine for all industrial cooling sys- The chlorinated phenols which have been
tems in the near future. However, this liquid most widely used in recirculating cooling
material has an important economical posi- water systems are the sodium salts of tri-
tion as a biocide where high organic demand chlorophenol and pentachlorophenol. Sodium
waters are used, process contamination oc- pentachlorophenate is probably in most wide-
curs, marine and sedentary organisms plug spread use either alone or combined in com-
equipment and corrosiw attack is a problem. mercial formulations with sodium trichloro-
phenate. Sodium pentachlorophenate is a
CHLORINATED PHENOLS. Phenolic materials soluble and stable material. It does not react
have been used extensively for germicidal pur- with most inorganic or organic chemicals
poses in medicine. Many phenolic materials which may contaminate a cooling water sys-
and particularly the chlorinated phenols have tem. A broader spectrum of biological activ-
found use in industrial microbiological con- ity can be achieved by combining the use of
trol. The toxicity of these materials varies more than one chlorinated phenolic com-
widely between different species of bacteria pound such as, for example, the simultaneous
and fungi. Figure 38-4 shows data taken from use of a trichlorophenate with a pentachloro-
the work of Shema and Conkey and illustrates phenate. Additional benefits may be gained
the inhibiting concentration of various phe- by the combined use of metallic ions such as
copper and zinc. Where a concentration of the individual system conditions.

200 ppm of a single phenolic compound may In any program for the continuous main-
be required for satisfactory control of organic tenance of a given toxicant concentration in
growths, combinations of this type have been the circulating water, it is necessary that the
successful in the range of 60-100 ppm. concentration maintained be at an adequate
Because of the toxicity of the chlorinated level. One program originally suggested for
phenols to fish and animals, some precau- sodium pentachlorophenate involved contin-
tions may be necessary to control the concen- uous feeding of the material so as to maintain
trations in the blowdown water from the a concentration of approximately 20 ppm in
cooling system. Care also must be exercised the system. This method has been unsuccess-
in handling these materials since the dust is ful because 20 ppm is inadequate for inhibi-
irritating to the mucous membranes of the tion of the microbiological complexes existing
eyes, nose and throat. If the concentrated in most recirculating systems. On the other
material is permitted to remain in contact hand, programs of continuous feeding are
with the skin, a rash may develop. With or- necessary in particularly difficult cases which
dinary precautions in handling, however, no cannot be controlled by a program of "shock"
difficulties are to be expected. treatment. In one plant using a makeup
CoNTINuous vs. SHOCK TREATMENT. With water with a high degree of sewage contami-
the chlorinated phenols, as in chlorina- nation, all shock treatment programs were
tion, there are many different feeding pro- ineffective and satisfactory control of the
grams that can be employed, ranging on the problem was secured finally by the continuous
one hand from the continuous maintenance maintenance in the system of a 200 ppm
of a high concentration in the circulating concentration of the toxicant. Intermittent
water to the other extreme of intermittent ad- shock treatment at this 200 ppm level was
clition of the material in low concentration ineffective. Whether the feeding program is
at infrequent intervals. As with chlorination, continuous or intermittent, it is important
the slime control program must be fitted to that the concentration of slime control agent
Fig. 38-5-Relative Toxicity of Quaternary Ammonium Compounds
Inhibiting Concentration, ppm

Aerobacter Bacillus Aspergillus Penicillium
aerogenes mycoides niger expansum
Dilauryldimethylammonium chloride 1500 35 >3000 >3000

Dilauryldimethylammonium oleate >3000 40 >3000 >3000
Dodecyltrimethylammonium chloride 80 65 >3000 >3000
Trimethylammonium chloride 1500 35 >3000 >3000
Octadecyltrimethylammonium chloride >3000 55 >3000 >3000
N-alkylbenzyi-N, N, N-trimethylammonium chloride 500 4 >3000 2500
Mixture of alkyl-9-methyl-benzyl ammonium chlorides 150 15 >3000 >3000
Lactoxymercuriphenyla mmen iu m.lactate 40 15 >3000 800
Alkyldimethylbenzylammonium chlorides 65 6 >3000 >3000
3, 4-dichlorobenzylammonium chlorides 700 3 >3000 >3000
Phenylmercuric trihydroxyethylammonium lactate 15 1 35 45
Phenylmercuric triethanolammonium lactate 20 4 65 70
Mixture of alkyldimethylbenzylammonium chlorides 200 7 >3000 >3000
*Benzyltriethylammonium chloride >3000 >3000 >3000 >3000
*Benzyltrimethylammonium chloride >3000 >3000 >3000 >3000
*Cetyldimethylammonium chloride 1600 5 >3000 >3000
*Lauryldimethyl-benzyldiethylammonium chloride (75"/o) 350 5 >3000 >3000
*Values obtained in Laboratories of Betz Laboratories, Inc.
296 BETZ HANDBOOK
developed in the system exceed the minimum the slug of treatment should be repeated.
inhibitory concentration. Otherwise, either The formula is indepenedent of the initial
continuous maintenance of a non-inhibitory concentra:tion developed by the slug of
concentration, or shock treatment to such a treatment and holds so long as the final con-
level will not bring control of the problem. centration for repetition of the slug is taken
Since it is necessary to develop an effective as 25% of the original value. ·
concentration of the slime control agent in This formula is useful in setting a feeding
the system, whether continuously or intermit- schedule and is valid for slime control agents
tently, it is obviously desirable from an which do not volatilize in passage over the
economic standpoint to employ intermittent tower and which are not consumed by reac-
feeding if the problem can be controlled by tion with other substances in the system. It
these measures. Unless the system has an un- is desirable to check the schedule in any
usually low retention time, there will be no water for the slime control agent. However,
marked difference in the inhibitory concen- once the schedule has been checked by such
tration required with continuous versus in- analyses a regular program of shot feeding
fermittent feed. can be established, feeding the slime control
In setting a schedule for intermittent feed- agent every three, five, etc. days as necessary.
ing of the slime control agent, it is necessary CoPPER SALTS. Algae are sensitive to very
to determine the concentration to be de- small amounts of copper ion and copper sul-
veloped in the system and the frequency with fate was one of the earliest materials used as an
which the shot will be repeated. The theoreti- algicide. The copper ion in distilled water
cal depletion of the slime control agent from solutions is also toxic to bacteria, but this effect
the system can be determined by the formula: is nullified by the presence of other ions and
organic matter.
BD X T
log Ct = log C; - Although algae in general are sensitive to
2.303 v quite low concentrations of copper, there is
where: considerable variation in the killing dose for
Cr = final concentration, ppm different species.
C; = initial concentration, ppm While copper salts are toxic to fish, various
BD = blowdown and windage loss, gpm
fish exhibit different degrees of sensitivity
V = system capacity, gallons
toward the copper ion. Investigations have
T = time, minutes
shown that fish mortality is not due entirely
A program that has been found practical to the toxic effect of the copper ion in the
in a shock treatment program is to repeat treated water. The mortality has been shown
the slug of treatment when the concentration to be due partially to clogging of the gills
has been depleted to 25% of the original with dead organisms that have previously
concentration. On this basis, the formula existed in the water.
above can be simplified as follpws:
The effectiveness of copper salts in the
v control of algae is nullified at the higher pH
T = 1.385
values since the copper ion is precipitated as
where: insoluble copper hydroxide. However, by
T = retention time, days employing copper salts in conjunction with
V = system capacity, gallons stabilizing agents, toxicity is greatly increased
BD = blowdown and windage loss, gallons and the precipitation of the copper ion is
per day prevented. Further improvements include
the use of wetting agents which permit the
Solving this equation for T (retention copper ion to penetrate the various slime and
time) will indicate the frequency with which algae growths.
Figure 38-6 • Culturing of Algae under Laboratory Conditions
Copper sulfate alone is rarely used for considerable difference in the effectiveness of
control of biological fouling in recirculating various quaternaries against the test organ-
cooling systems. The copper ion is corrosive isms. All quaternaries listed were relatively
to iron, steel, aluminum and other metals. ineffective against the fungi, with the excep-
Moreover, recirculating systems rarely are tion of those containing mercury in the mole-
afflicted only with the problem of algae and cule.
other agents are required for bacterial con- It is characteristic of the quaternaries that
trol. Even specially prepared copper base al- their bacterial properties are reduced in the
gicides, that have been stabilized to avoid presence of soap, protein and high ionic con-
precipitation of the copper ion, and blended centrations. It has been found that certain
with other toxicants to control bacteria, must of the quaternaries react with or are absorbed
be used judiciously. by organic matter found in industrial re-
QuATERNARY AMMONIUM CoMPOUNDS. circulating cooling water systems and a loss
There are a great number of quaternary am- of effectiveness results. Some volatilization
monium compounds available to industry and also occurs over a cooling tower, thereby in-
these compounds are effective germicides in creasing losses from the system.
many applications. Figure 38-5, taken par- Despite these disadvantages from the stand-
tially from the work of Shema and Conkey, point of practical application to recirculating
shows data on the relative toxicity of several cooling water systems, some success has been
quaternaries. It can be noted that there is achieved with the use of quaternaries, par-
298 BETZ HANDBOOK
ticularly when combined with metallic ions, Only a small fraction of this number have
such as copper. The quaternaries are effective been applied to the control of biological foul-
wetting agents and aid penetration of organic ing in recirculating cooling water systems.
growths because of this property. In general, Many of these chemical agents are ruled
like the chlorinated phenols, their most ef- from consideration immediately because of
ficient method of use has been in shot fash- cost. Many of these materials are effective
ion. The periodic addition of high concentra- only in high concentrations such as 5%, 10%
tions to the circulating water, repeated at or more. Others are ineffective under the
regular intervals, has been found more effec- conditions of use in a cooling water system,
tive in controlling organic growths than at- due to volatility, reaction with organic mat-
tempts to continuously maintain a lower con- ter, reaction with inorganic ions naturally
centration in the circulating water. present or the chemical treatments in use for
MISCELLANEous ToxiCANTS. The inorganic control of scale or corrosion. Of the great
mercurials such as mercuric chloride are con- number of bactericidal and bacteriostatic
sidered too toxic to humans for use in recircu- agents known, few can qualify for use in re-
lating cooling water systems. Organic mer- circulating cooling systems, principally be-
curials have been developed in an effort to re- cause of the factors of cost, effective con-
tain the bactericidal properties of mercury, centrations and compatibility in this environ-
while reducing the toxicity of these materials. ment.
Some of the organic mercurial compounds GENERAL CONSIDERATIONS IN SELECTING
may be aerated readily from a cooling tower SLIME CONTROL AGENTS
system thus allowing only a short contact
time with the slime-forming organisms. Such As mentioned previously, chlorine is probably
materials should not be used in any cooling the most widely used agent for microorgan-
tower system where the spray might contact ism control in cooling water systems. The de-
humans or where, because of volatility, the cision whether to use chlorine or another
compound will be present in the air breathed. slime control agent is usually made by bal-
In general, both the inorganic and organic ancing continuous chemical operating cost
mercurials are considered too toxic for use in against equipment installation and mainten-
cooling water systems. In addition, the mer- ance costs.
cury ion, even in low concentrations, will add Where liquid chlorine can be used, chemi-
appreciably to the corrosive qualities of a cal cost will usually be lower than with other
water. toxicants if chlorine is purchased in the
Potassium permanganate, similar to chlo- larger one ton containers or tank cars, at
rine, is a powerful oxidizing agent and effec- their lower price. However, the installation
tive in the control of slime-forming organ- cost of chlorinators and facilities for handling
isms. However, potassium permanganate acts bulk chlorine shipments is quite appreciable.
on all oxidizable material present and man- Maintenance of this equipment is an added
ganese dioxide precipitates as a result of the cost.
reduction of the permanganate. The result- On the other hand, other slime control
ing sludge is the chief factor limiting the use agents are usually furnished in readily han-
of permanganate. dled containers in liquid, powder or briquette
Iodine and silver have been used to a form. No feeding facilities are required where
limited extent in the sterilization of drinking the toxicant is shot fed and the material is
water and in swimming pool control, but added only once every several days. No con-
have not been found adapted to use of cool- tinuous attention to feeding equipment is
ing water systems. necessary and there is no installation cost to
Biological and medical literature refer to be written off.
hundreds of chemical agents which are use- Where the system is sufficiently large and
ful in controlling microbiological activity. the problem sufficiently acute, the installation
and maintenance expense of chlorination and other materials may present a problem
facilities will be justified. Where the cooling of fish toxicity in the blowdown discharge.
water system is relatively small or where Copper salts are more specific to algae than
there is only infrequent necessity for slime to bacteria and are rendered ineffective by
control agents, it will be more economical to precipitation at higher pH values. Quater-
employ agents other than chlorine. The naries are affected by ions and organic
simplicity of feeding and handling liquid or matter.
briquetted materials is particularly attractive All of these chemical agents are not equally
in the small installations and where there is economical and efficient. Selection of toxi-
only periodic need for slime control measures. cants must be based on the microbiological
Even where the control of biological growths associations encountered. Frequently, more
is handled normally by the use of chlorine, than one toxicant is desirable to control, for
there will frequently exist the need of addi- example, both a bacteriological slime and/or
tional slime control agents. For example, algae problem. Methods and frequency of
during periods of contamination of the cool- chemical feeding must be varied to suit the
ing water by reducing agents, chlorination individual problem. Slime control agents
facilities may be inadequate for the quantity vary in the rapidity of their action and must
of chlorine required under these circum- be selected with a knowledge of the retention
stances. Addition of agents such as the chlo- period in the system. Compatibility of the
rinated phenols or the various commercial toxicant must be studied in relation to other
combinations, may be necessary to control treatments used for scale and corrosion con-
organic growths during such periods. Even trol. Local conditions governing discharge of
where the increased chlorine dosage can be blowdown water from the cooling system
handled by the feeding equipment installed, must be considered. Selection of the proper
it may be more economical to use another method of control of microorganisms in re-
slime control agent at such times rather than circulating cooling water systems, therefore,
to employ the greatly increased amounts of requires evaluation of numerous factors.
chlorine required during periods of con-
tamination. REFERENCES
At the present time, there is no single toxic
F. E. Hale, "The Use of Copper Sulphate in Con-
agent which is completely effective for the trol of Microscopic Organisms," Phelps Dodge
control of biological fouling in all types of Refining Corp., New York, N.Y. (1954)
industrial cooling water systems. There are B. F. Shema and J. H. Conkey, "Relative Toxicity
of Disinfectants Available for Use in the Pulp and
faults inherent in all of the toxicants in com- Paper Industry," Tappi, Tech. Assoc. of the Pulp
mon use. Chlorine is highly effective and and Paper Ind., Vol. 36, pp. 20A, 22A, 24A, 26A,
rapid in bactericidal action, but chlorine 28A, 30A, (Nov. 1953)
consumption is greatly increased by the pres- B. F. Shema and J. H .. Conkey, "Relative Toxicity
of Disinfectants Available For Use in the Pulp
ence of reducing agents. Chlorine does not and Paper Industry-1954 Supplement," Biological
afford protection to biological attack of cool- Control Committee, Am. Paper and Pulp Assoc.,
New York, N.Y. (1954)
ing tower wood. Moreover, chlorine must be
f. M. Donohue, A. J, Piluso and J. R. Schieber,
used judiciously to avoid chemical attack of '"New Type Biocide Produces Cost Savings in Slime
cooling tower wood. The chlorinated phenols Control," 22nd Annual NACE Conference ( 1966)
300 BETZ HANDBOOK
39
Cooling Tower Wood
Deterioration
Types of Wood Deterioration . . . . . . . . . . . . . . . 301
Chemical Attack. . . . . . . . . . . . . . . . . . . . . . . 303
Biological Attack. . . . . . . . . . . . . . . . . . . . . . 303
Physical and Other Factors. . . . . . . . . . . . . . . 304
Control of Wood Deterioration . . . . . . . . . . . . . . 304
Flooded Section. . . . . . . . . . . . . . . . . . . . . . . . 304
Non-Flooded Section. . . . . . . . . . . . . . . . . . . . 307
Wood Examination. . . . . . . . . . . . . . . . . . . . . . . 308
Methods of Spraying. . . . . . . . . . . . . . . . . . . . . . 309
Other Methods. . . . . . . . . . . . . . . . . . . . . . . . . . 309
VER the past decade considerable at- still subjected to the same conditions. The
0 tention has been focused on the prob-
lem of cooling tower wood deterioration-
large, open natural draft tower was the first
type used to cool water. The flow of air
and for good reason. Complete collapse of through this tower is dependent upon wind
some cooling towers has occurred. In other velocity and upon the draft produced by the
cases, supports have deteriorated and fans chimney design of the tower. In general, the
have fallen through the tower. Accidents conditions in these towers from the biologi-
have occurred due to collapse of ladders, cal point of view were not nearly so critical.
decking and ·other parts of the tower. Wood The few natural draft towers which have
deterioration has shortened the life of cooling been in service for the past 20 years or more,
towers from the anticipated 20-25 years to 10 are still in a remarkable state of preservation.
years or less in many cases. Repair and re- Because natural draft towers do not exhibit
placement costs have been excessive, and rapid loss of wood structure or strength as a
cooling tower operation has been inefficient. result of attack by decay organisms, does not
Redwood is the most commonly used ma- mean that decay is non-existent but rather it
terial for cooling tower construction. It has indicates that biological deterioration is not
been used for this purpose almost since the the limiting factor affecting the life of these
advent of cooling towers themselves. Origi- towers.
nally, this material was selected because of The development of the mechanical draft
high strength to weight ratio, ready avail- tower has resulted in greater cooling effi-
ability, ease of fabrication, cost, and natural ciency. A smaller tower is required to supply
resistance to decay. Other species are used the process with cooled water, and in addi-
for this purpose, particularly Douglas Fir and tion these towers also give more reliable and
less frequently cypress and pine. uniform performance because they are not
Wood is composed of three main com- dependent upon wind velocity for their op-
ponents: cellulose, lignin and natural ex- eration. Even though mechanical draft towers
tractives. The cellulose exists as long fibers, are more efficient than atmospheric towers,
almost identical to cotton fibers, and gives they impose a completely different set of con-
wood its strength. Lignin acts as a cementing ditions upon the wood. The general design
agent for the cellulose. The extractives con- of a mechanical draft tower is more con-
tain most of the natural compounds which ducive to biological and chemical attack than
contribute to the resistance to decay. In the natural draft tower. The drift elimina-
general, the more highly colored woods are tors, cell partitions, doorways, and the com-
the more durable. A typical analysis of red- pact design of the mechanical draft towers
wood, based on dry weight, will show ap- create a greater number of areas where en-
proximately 50% cellulose and hemi-cellulose, vironmental conditions are ideal for the
30% lignin and 20% extractives. The ex- growth and development of objectionable
tractives in redwood make this material one microorganisms. In addition, the amount of
of the most resistant to decay. Unfortunately, water and air handled per unit area is higher
the extractives present in all woods are largely and inoculation of the system with air borne
water soluble and will be leached from the organisms is at a greater rate.
wood by the circulating water. The strength
of the wood does not appear to be affected
TYPES OF WOOD DETERIORATION
by leaching. However, the wood does become
more susceptible to decay. Cooling tower wood is subject to three main
Exceptionally good service was obtained types of deterioration: chemical, biological,
from redwood during its initial use and it is and physical. It is rare when one of these
reasonable to expect that this good service types of deterioration is present without the
would be obtained today if redwood were other. In most cases, these different types of
BETZ HANDBOOK
302
Figure 39-1 • Delignified Wood-Caused by Chemical Figure 39-2 • Checked, Brash Wood-Caused by
Attack Biological Surface Attack
Figure 39-3 • Fibrillated Wood-Caused by Chemical Figure 39-4 • Internal Decay of Redwood Structural
Attack Member Showing White Fungus Mycelium
deterioration occur simultaneously. When de- The organisms that are responsible for at-
terioration occurs it is sometimes difficult to tack of cooling tower wood are those that can
determine which type of attack is predomi- utilize cellulose as their source of carbon in
nately responsible. However, it is evident that their growth and development. Degradation
physical and chemical deterioration will of the cellulose is accomplished by the secre-
render the wood more susceptible to biologi- tion of enzymes which convert the cellulose
cal attack. into compounds that can be absorbed by the
organisms. The attack tends to deplete the
CHEMICAL ATTACK. Chemical deterioration
cellulose content of the wood and leaves a
of cooling tower wood commonly is mani-
residue rich in lignin. Characteristically the
fested in the form of delignification. Since the
wood becomes dark in color and loses much
lignin component of the wood is affected and
of its strength. The wood may also become
removed by this type attack, the resultant
brash, soft, punky, cross-checked or fibril-
residue is rich in cellulose. The chemicals
lated. The principal cellulolytic organisms
most commonly responsible are oxidizing
isolated from cooling tower wood are prima-
agents such as chlorine, and alkaline ma-
rily fungi, which include the classical wood
terials such as sodium bicarbonate and so-
destroyers (Basidromycetes), and members
dium carbonate. The attack is particularly
of Fungi Imperfecti. However, bacterial or-
severe when the combination of high chlorine
ganisms that exhibit cellulolytic properties
residuals and high alkalinity concentrations
also have been isolated but their exact role in
are maintained simultaneously.
cooling tower wood deterioration is yet to be
Typically, the wood takes on a white or determined. The wood destroying organisms
bleached appearance and the surface be- are common air and water borne contami-
comes fibrillated. This attack is restricted to nants, and are widely distributed in nature.
the surface of the wood and the strength of
The classical internal decay is restricted
the unaffected area is not impaired. How-
generally to the plenum areas of the tower,
ever, severe thinning of the wood will occur
such as cell partitions, doors, drift elimina-
wherever cascading water has an opportunity tors, decks, fan housing and supports. It is
to wash away the surface fibers. In serious
cases, the loosened fibers have caused plug-
ging of screens and tubes, and have served as
focal points for corrosion at areas where the
fibers accumulate in heat exchange equip-
ment.
Chemical attack most frequently occurs in
the fill section and flooded portions of the
tower where water contact is continuous.
However, it will occur also in those areas
where alternately wet and dry conditions
develop-such as on the air intake louvers
and other exterior surfaces. Chemical attack
occurs also in the warm, moist areas of the
plenum chamber of the tower as a result of
chlorine vapors and the entrainment of drop-
lets of tower water.
BIOLOGICAL ATTACK. Biological attack of
cooling tower wood can be divided into two
basic types-soft or surface rot and internal Figure 39-5 • Structural Member Showing Sound Outer
decay. Surface and Typical Internal Decay
304 BETZ HANDBOOK
the more insidious of the two types of biologi- the soft rot organisms.
cal attack. It is characterized externally by Preferential erosion of the summer wood
an apparently sound piece of wood, which is relatively common in towe.r fill. In severe
upon breaking shows severe internal decay. cases of erosion significant losses of wood
Because the decay is internal, it is difficult to can result in very short periods of time. Ex-
detect in its early stages. Rarely is internal tremely high concentrations of dissolved
decay found in the flooded portions of the solids are to be avoided where there is op-
tower such as the fill section. In these sec- portunity presented for alternately wetting
tions, the wood is saturated completely with and drying of certain wood areas. While
water which excludes oxygen from the in- natural salts have been shown to possess
terior of the wood. The lack of oxygen little tendency to attack wood, crystallization
limits the growth and development of these of these salts in dry areas may rupture the
organisms. wood cells.
Soft or surface rot is found predominantly
in the flooded sections of the tower but also CONTROL OF WOOD DETERIORATION
occurs in the plenum areas. The water flow-
ing over the wood surfaces in the flooded The only effective method of protecting
portions contains enough oxygen to support operating cooling towers is to adopt a pre-
growth. Surface rot is detected more readily, ventive maintenance program. The preven-
and its effect is less severe than internal de- tive measures for the flooded sections of the
cay. tower where attack of the chemical and bio-
logical types is limited to the surfaces of the
In addition to oxygen, moisture and tem-
wood are relatively easy to accomplish. The
perature have a marked bearing. Locations
preventive measures for the non-flooded
where the moisture content of the wood
portions of the tower where internal decay is
range between 20 and 27% and tempera- the primary concern are more difficult and
tures range between 88 and 105 F, usually
the success of the program is largely depend-
permit optimum growth and development of
ent on adopting the necessary measures be-
the organisms.
fore infection reaches serious proportions.
PHYSICAL AND OTHER FACTORS. One of the
FLoODED SECTION. The control of chemical
major physical factors is the effect of tem-
and biological surface attack of cooling tower
perature on wood. Wood technologists have
wood in the flooded portions of the tower is
long recognized that high temperatures have
a water treatment problem that requires:
an adverse effect on wood. It is known that
continuous exposure to high temperatures 1. Control of pH of the circulating water
will produce gross changes in anatomical below 8.0 and preferably in the range of
structure and will accelerate loss in wood 6.0 to 7.0.
substance. These resultant effects will weaken 2. The use of non-oxidizing biocides alone
the wood and predispose it to biological at- for control of slime and prevention of bio-
tack, particularly in the plenum areas of the logical surface attack.
tower. 3. Where chlorine must be used-chlorine
There are other factors which also have a should be restricted to 1.0 ppm or less and
bearing on the deterioration of tower wood. preferably in the range of 0.3 to 0.5 ppm.
For example, areas adjacent to iron nails and In addition, the supplemental use of a non-
other iron hardware usually deteriorate at an oxidizing biocide should be employed to
accelerated rate. These areas invariably lose control biological surface attack.
much of their strength and the wood will In order to minimize delignification, pH
crumble easily in the fingers. Slime and al- of the circulating water should definitely be
gae growths and deposition of dust and oil maintained below 8.0 and preferably lower.
can all aid the growth and development of Fortunately, most modern treatments for con-
trol of corrosion function best when pH is cooling tower wood in systems using chlorin-
maintained around 6.0-7.0. This pH range ation has led to the conclusion that chemical
corresponds to the range best suited for attack predisposes the wood to biological at-
minimizing attack of wood. tack. In cases where carefully controlled
Surveys have shown that in tower systems chlorination has been practiced and chemical
where non-oxidizing biocides alone are used attack of the wood surfaces is low, studies
for control of slime, surface attack is at a have been conducted to check the incidence
minimum. Accordingly, when it is economi- of fungi on the surfaces of the wood and in
cally feasible non-oxidizing biocides should the subsurface sections. Microtome sections
be used in preference to chlorine. of the wood, such as that shown in Figure
Where chlorine must be used for control of 39-8, reveal a high incidence of fungi in the
slime it is necessary that chlorine be used wood from the fill and plenum areas of the
judiciously. Undoubtedly, many of the sys- tower. These studies indicate that chlorina-
tems that have shown rapid or excessive at- tion alone is not effective in controlling the
tack from the use of chlorine have lacked organisms responsible for attack of wood.
adequate control of chlorination. In many of The judicious use of chlorine requires low
these cases, the chlorine residual has been residuals be maintained. As the water cas-
improperly determined, determined at the cades through the tower the residuals are de-
wrong point for proper control, not checked pleted. The air-borne organisms can and do
frequently enough or neglected entirely. In accumulate readily on the surfaces of the
the majority of the systems using chlorine for wood and grow and develop in the sub-
slime control, the chlorine residual is checked surface sections. Similar studies conducted
(as it should be) in the water returning to for systems that use non-oxidizing biocides
the tower, not after the water cascades alone for slime control show that the inci-
through the tower. The water in passing dence of fungi on the wood surfaces is
through the tower will have its chlorine residu- negligible. Since these biocides do not react
al decreased to very low values or depleted with the wood, they are able to penetrate the
entirely.· The reduction in chlorine residual wood and effectively control the causal or-
occurs partially due to aeration but mainly ganisms. Accordingly where chlorination
through reaction with the wood. Where at- must be practiced the supplemental use of a
tempts have been made to control chlorine non-oxidizing biocide is recommended to as-
residuals after the water has passed through sure maximum tower life.
the tower, excessive chlorine residuals have One standard procedure currently utilized
been permitted to contact the wood and ac- to aid in minimizing biological attack in the .
celerated chemical attack has been the result. flooded portions of the cooling tower re-
quires adding a non-oxidizing biocide to the
In instances where intermittent chlorina- system approximately once every three
tion has been practiced, initially the duration months. The concentration of biocide used is
of the chlorination period and the rate of in the range of 60 to 120 ppm. The purpose
chlorination may have been established prop- of this treatment is to kill the organisms that
erly by testing for the chlorine residual. How- have accumulated on the surface and sub-
ever, all too often, after initially establishing surface portion of the wood in the fill or
the rate and duration, infrequent or no fur- flooded portions of the tower. Supplemental
ther testing for chlorine residual is made. benefits are also secured in that the non-
Since the chlorine demand of most circulat- oxidizing biocide will carry down into the
ing waters will vary, this practice has led to lower depths of the cooling tower sump
low residuals and poor slime control or to where it is difficult to maintain a chlorine
high residuals and excessive attack of the residual. Muck and debris accumulate in
wood. these areas and provide ample food for
The prevalence of biological attack of growth and development of microorganisms.
306 BETZ HANDBOOK
Figure 39-6 • Sectioning of Wood by Use of a Sliding Microtome
Figure 39-7 • Microscopic Examination of Figure 39-8 • Microtome Section of Wood Showing
Cooling Tower Wood Fungus Hyphae Growing through Wood Cells
These areas serve as a point of inoculation, and certainly more frequently if a preventive
particularly for sulfate reducing bacteria. The maintenance program is not in use.
supplemental use of non-oxidizing biocides Although soft rot or surface attack occurs
serves to eliminate this point of infection. in the non-flooded portions of a tower, the
In many cases, it is possible to use a effect on loss on wood structure is not as
variety of treatment programs which com- severe as in the flooded areas. This is due
bine the use of chlorine and a non-oxidizing primarily to the fact that the wood is not
biocide. Where a combination program is subject to erosion by cascading water. The
possible, chemical attack can be held to a principal and most serious problem in the
minimum and biological attack can be con- non-flooded area is internal decay. When in-
trolled effectively. ternal decay is restricted to white pocket rot,
It should be evident that if the makeup the affected areas can be very small and
water is taken from a stream, river or other easily missed. As a general rule internal de-
surface supply that is not pretreated or cay is detected usually only after extensive
clarified, high bacteria counts can enter with damage has occurred. It is very important to
the makeup water. In addition, these sources look for signs of internal decay in structural
of makeup water usually contain organic members. This is sometimes revealed by ab-
matter and other contaminants that serve as normal sagging or settling of the wood. When
food for bacteria as well as increase the chlo- obvious decay is not evident, it is advisable to
rine demand of the water. Where water of secure samples of the wood for microscopic
this type is used, the slime potential is usually examination to determine whether the in-
heavy and continuous chlorination to a free ternal areas of the wood are infected with
residual is generally necessary. To minimize fungi.
the amount of chlorine that contacts the wood Replacement of infected wood is necessary
many plants have adopted the practice of to retard spread of infection to adjacent
chlorinating the makeup to secure a total sound members. A weakened section will
bacteria kill and to minimize bacteria food cause additional weight load to be shifted to
entering with the makeup. Under these con- sound sections of the structure. These in turn
ditions, the slime potential will be substan- may crack under the increased load and be-
tially reduced and an intermittent chlorina- come more susceptible to spread of internal
tion program can be adopted for the circula- decay. The infected wood should be replaced
ting water alternating weekly or biweekly with pressure treated wood. Several different
with the use of a non-oxidizing biocide. wood preservatives are available as well as
The adoption of these measures will effec- a choice of wood species. Willa discusses a
tively minimize attack of wood in the flooded seven year study of eight different pressure
portions of the tower and will also aid in treatments. On the basis of incidence of in-
minimizing surface attack in the plenum ternal decay the treatments can be listed in
areas of the tower. the following order of effectiveness:
1. Creosote
NoN-FLoODED SECTION. The preventive
maintenance program for the non-flooded or 2. Ammonical Copper Arsenite
plenum areas of the tower requires: 3. Acid Copper Chromate and Copper
1. Thorough periodic inspections Naphthenate
2. Replacement of damaged wood with 4. Chromated Copper Arsenate
pressure treated wood 5. Pentachlorophenol
3. Periodic spraying of the plenum areas 6. Fluoride Chromate Arsenate Phenol
with fungicides. 7. Chlorinated Paraffin
A thorough inspection of the cooling tower Periodic spraying with an effective fungi-
should be conducted at least once per year cide is the third essential step in an effective
308 BETZ HANDBOOK
preventive maintenance program. The pur- determines whether any fungi present are
pose of spraying the plenum areas with a cellulolytic (wood destroying) or simply fun-
fungicide is to render the wood resistant to gus organisms.
spread and growth of wood destroying fungi. The most beneficial test when evaluating
Diffusion of fungicide into the wood at best cooling tower wood is commonly known as
permits penetration of the wood to a depth the zone of inhibition test. This test deter-
of 1/16" to 1/4". The degree of penetration mines the relative penetration of any fungi~
is dependent on the preparation given the cidal agent employed and the degree to which
wood and to some degree on the concentra- this material is present. Also determined is
tion of fungicide sprayed. The objective is to the susceptibility of the wood to support
spray a sufficient concentration of fungicide fungal growth and to become decayed if
so that the wood remains fungistatic until inoculated with wood destroying organisms.
the next semi-annual or annual inspection
Briefly described, the zone of inhibition
and spraying.
test consists of placing two ~ inch squares
WO<?D EXAMINATION of the wood specim'en on nutrient agar pre-
viously seeded with a wood rotting organism,
Since several types of wood deterioration are such as Aspergillus niger or Chaetomium
possible in cooling towers, physical inspec- globosum. The plates are then incubated for
tions along with periodic laboratory examina- seven days at 28 C. After the elapsed incuba-
tions of wood samples should be scheduled tion period, the plates are evaluated for their
regularly. In this connection, service labora- degree of protection or susceptibility.
tories are equipped with specialized equip- A complete zone of inhibition exists when
ment to determine all of the important data
no fungal spores of the test organism are
needed to comprehensively evaluate the con-
present in a clear zone around the test block.
dition of tower wood:
This preventative barrier against fungal in-
Microscopic examination of the wood re- fection develops when the effects of a fungi-
veals the degree of grooving, erosion, surface cidal application are still present in the wood
structure and depth of surface attack. Speci- and inhibit the growth of wood rotting
mens are broken to determine whether the organisms.
wood is brash, and the degree of brashness as
A partial zone of inhibition is one which
well as apparent loss in structural strength
shows some growth of the test organism
are recorded. Macroscopic study reveals the
physical aspects of the wood, whether or not around the block, but the growth of fungal
chemical and/or biological factors are present spores of the test organism is retarded. This
and, if so, to what extent. Figures 39-1 to partial zone indicates that some residual
39-3 inclusive demonstrate some of these fungicide or natural inhibitory properties are
physical characteristics that are helpful to still present in the wood.
fully evaluate the condition of cooling tower No zone of inhibition is exhibited by a
wood. specimen of low resistance, so the growth of
Microtome studies of wood cells, as shown the test organism or other organisms inherent
in Figures 39-6 to 39-8, are of assistance in in the wood occurs around the block. This
determining the extent of microbiological growth indicates that the wood is susceptible
deterioration. These microtome sections are to fungal attack. Corrective measures must
usually 25 microns in thickness, and permit be taken in order to control the spread of
viewing of the internal structure of the wood. fungus in the sound members of the tower.
This examination indicates to the analyst the Figure 39-9 shows two of the conditions
extent of infection, and whether the infection that aid the analyst in determining whether
is due to bacterial or fungal activity. From or not inhibitory power from a preservative
this study of the wood cells, the microscopist treatment or from the natural property of the
wood is at an effective level. This zone of OTHER METHODS

inhibition test may be used to evaluate the
residual effect of fungicidal treatment or the In order to decrease labor costs and to fur-
degree of resistance restored by an applica- ther minimize the hazards of direct spraying
other methods have been considered. Steam
tion of fungicide.
spraying through a permanent piping ar-
METHODS OF SPRAYING
rangement has been used. In this procedure
the fungicide is forced into the steam and is
The rnost effective method of spraying is the transported into the cell by the steam. It is
direct or manual method. This method is essential that the distribution piping be de-
very similar to painting in that the concen- signed properly in order to secure complete
trated fungicide is applied directly to the coverage. The use of steam causes dilution of
wood by the use of spraying equipment the fungicide and compared to direct spray-
handled by an operator or team of operators. ing a relatively dilute solution contacts th~
This method of application proves most wood. As a result the quantity of toxicant
effective since small areas can be covered that penetrates the wood is smaller and more
thoroughly by the fungicide, and special at- frequent spraying is necessary to maintain
tention can be given to spraying the joints, the wood fungistatic.
holes and other access points into the wood. Recently a revised method of spraying the
Direct spraying can be hazardous unless plenun1c areas of cooling towers has been
proper precautions are taken to protect the introduced. This new method is designed to
operator. Figure 39-10 shows an operator rninimize the hazards and high cost of direct
properly equipped. He is clothed in air-fed spraying while maintaining its major advan-
coveralls with the legs secured with tape at tage, namely: spraying of a concentrated solu-
the ankles. The outer clothing consists of a tion of fungicide so that the frequency of
rain suit. The rubber gloves are vulcanized spraying will be required only once or twice
to the sleeves of the suit and the legs are per year. Pneumatic spraying with a portable
secured to the arctics with tape at the ankles. rig equipped with an atomizing nozzle similar
The sandblaster's hood is in place and its air to that shown in Figure 39-11 is used. The
feed can be seen from the rear. The air-fed rig is inserted through holes in the deck of
coveralls and sandblaster's hood permit the the tower so that spraying can be accom-
operator to stay cool and comfortable while plished from outside the tower.
spraying.
Figure 39-9 • Test Blocks of Wood on the Surface of Inoculated Nutrient Agar. The Block on the Left Shows No
Resistance to the Test Organism. The Block ·on the Right, which has had Fungicide Spray Treatment, Shows a Clear
Zone of Inhibition
310 BETZ HANDBOOK
Diffusion and spray methods are limited

for the most part to penetration of the outer
surfaces of redwood. Therefore it is essential
Figure 39-ro • Direct Fungicide Spraying by

Properly Equipped Operator
that preventive maintenance be started be-

fore infection takes place and preferably be-
fore the interior parts of the wood lose much
of their natural resistance. Under these con-
ditions new infection can be prevented and
the life of the tower extended. Where the
tower has undergone a serious degree of in-
fection and it is likely that infection can-
not be contained by following the normal
preventive maintenance program, then con-
sideration can be given to sterilization of the
wood by elevating the temperature of the
internal portions of the wood to 150 F and
maintaining this temperature for two hours.
REFERENCES
W. A. Dost, "Certification of Redwood and Effect Figure 39-71 • Portable Rig for Pneumatic Spraying
of Oxidizing Agents on Redwood in Cooling Tower
Service", Proceedings, Engineers Soc. of Western
Penna., pp. 77-83 (1959)
B. F. Shema, "Clarification of Cooling Tower
Wood Deterioration", Proceedings, Engineers Soc.
of Western Penna., pp. 69-74 (1959)
Unpublished data, Betz Laboratories, Inc., Phila.,
Pa.
J. L. Willa, "Report On Field Wood Preservation
Studies", Proceedings, Engineers Soc. of Western
Penna., pp. 83-96 ( 1959)
40
Air Conditioning Systen1s
Open Recirculating Systems . ................ 313

Air Washers. . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
Closed Water Systems . .................... 315
Control of Water Balances . ................ 315
Chemical Feed Methods. . . . . . . . . . . . . . . . . . . . 316
General Considerations. . . . . . . . . . . . . . . . . . . . . 317
312 BETZ HANDBOOK
AIR conditioning is considered a modern DISCHARGE o;=:::=::'"-=+=9'/ AIR DISCHARGE

LINE FROM
1""\. luxury but it has been used for many COMPRESSOR REFRIGERANT
TO DIRECT-
years in industry for drying, humidity con- EXPANSION
COIL
trol and for dust and smoke abatement. Air
conditioning is used for health purposes and ELIMINATOR
has proven beneficial in increasing the effi- SUCTION
LINE TO
REFRIGERANT
ciency of working personnel. COMPRESSOR
COIL
Air conditioning has been defined as the

process of treating air so as to control simul-
taneously its temperature, humidity, cleanli-
ness and distribution to meet the require-
ments of the conditioned space. The tem-
perature of air is controlled by passing the
air over chilled or heated coils or by exposing
the, air to a spray of water of controlled Figure 40-1 • Evaporative Condenser
temperature. Removal of moisture from the
is defined as the capacity to remove heat at
air is accomplished by cooling the air by
the rate of 12,000 Btu per hour. Freon sys-
chilled coils or by direct water spray. De-
tems require for heat rejection the evapora-
humidification is also accomplished by sor-
tion of approximately 1.5 gallons of water
bent materials such as silica gel. Washing of
per hour per ton of refrigeration. An absorp-
the air with water sprays may also serve to
tion system, because of the additional steam
moisten the air and commonly has an im-
heat input, requires the evaporation of ap-
portant function in dirt and odor removal.
proximately 3 gallons of water per hour per
The basic components of an air condi- ton. These water evaporation requirements
tioning system are the refrigeration machine, are based on air conditioning operation at
a system of heat rejection and a cooling dis- rated capacity. The average operating load
tribution system. The two types of refrigera- on an air conditioning system during the op-
tion commonly used with air conditioning erating season will be less than rated capac-
are the refrigerant (Freon) compression sys- ity. Usually the average operating tonnage
tem and the absorption system. Absorption will be in the range of 30-60% of rated ton-
systems used in air conditioning have a lith- nage.
ium salt solution as the refrigerant and
Cooling distribution throughout the build-
steam as the energy source. ing is accomplished by a variety of methods.
In industrial or commercial size air con- Air can be passed over coils chilled by direct
ditioning systems, heat rejection almost in- expansion of refrigerant and then circulated.
variably is accomplished by water. Once- In larger systems, water is chilled and this
through cooling is sometimes used. However, water is circulated in a closed system to fan
municipal restrictions on water consumption, units where the air is blown over coils
and the water costs, usually result in evapora- chilled by the water. Sometimes these closed
tive cooling for all but very small installa- water systems are used for the dual purpose
tions. Evaporative condensers or economizers, of winter heat. The chilled water may also be
as well as cooling towers, are used for evapo- recirculated through one or more air washers,
rative cooling. Occasionally a spray pond either in closed coils, or as the spray water.
may be used. By recirculating water in an Where the chilled water is in a closed system
evaporative cooling system, the water makeup the air washers will have individual spray
requirement is reduced to less than 3% of water systems.
the water used for once-through cooling. In considering water treatment, the water
Air conditioning capacity is measured in systems associated with air conditioning can
tons of refrigeration. A ton of refrigeration be classified into the three general categories
of open recirculating cooling water systems, galvanizing in the system. Polyphosphates are
closed water systems, and air washers or of some value for scale control, but again
open chilled water systems. care must be taken in their application. Ex-
cessive feed can result in calcium phosphate
OPEN RECIRCULATING COOLING WATER deposits.
SYSTEMS The water in open recirculating cooling
systems will be- saturated with oxygen and
The water in open evaporative cooling sys-
consequently quite corrosive. In addition,
tems is subject to the problems of scale, cor-
those air conditioning systems located in ur-
rosion and slime and algae.
ban areas are often subject to the absorption
As water is evaporated in the cooling of acidic gases from the air. Acidic pickup
tower or evaporative condenser, pure vapor can be beneficial in reducing scaling tend-
is lost and dissolved salts will concentrate in encies, but excessive absorption of acid gases
the system. When cycles of concentration de- can result in severely corrosive conditions.
velop sufficiently the solubility limit of a salt,
The corrosion inhibitors used in open sys-
usually calcium carbonate, will be exceeded
tems are generally the same as those de-
and precipitation will occur. The result is
scribed in Chapter 34. Chromate base treat-
scale deposits on hotter surfaces such as in
ments are commonly used, provided there is
the tubes of the condenser. Deposits result in
no objection to the color of the water. The
high head pressure and reduced efficiency.
drift loss from a cooling tower with high
Control of scale in open cooling systems is chromate treatment will cause staining. Chro-
discussed in detail in Chapter 33. In air con- mates may also be limited in application
ditioning systems provision for continuous from the standpoint of disposal of the blow-
blowdown, or bleed, is usually adequate for down water.
scale control. The importance of continuous
bleed, as opposed to periodic draining, can- Nitrite base treatments have been used
not be overemphasized for the water content where chromates cannot be applied. Nitrite is
of most air conditioning cooling systems is colorless, but it has several disadvantages. A
quite small compared to the amount of water treatment level approximately twice that of
evaporated. Excessive solids can develop in chromate must be used for equivalent pro-
less than a day of operation. Continuous tection. In addition, nitrites are subject to
blowdown is usually provided by running a biological degradation requiring the greater
valved line from the discharge of the recircu- use of biocides to maintain concentration of
lating water pump to the sewer. the corrosion inhibitor and to control devel-
When the water is extremely hard and opment of biological slime and algae. Pro-
high in alkalinity, feed of sulfuric acid or an prietary formulations have been developed
acid salt, in addition to blowdown, may be combining the biocide with nitrite.
necessary for scale prevention. Feed of an The use of certain organic base treatments
acidic material requires care in handling and in place of nitrite is a more desirable ap-
good control and is not used unless the blow- proach to the problem of providing an ade-
down rate otherwise necessary would be ex- quate non-chromate inhibitor. Treatment
cessive. Zeolite softening of the makeup water levels are essentially the same as with nitrites,
is another method of scale control. However, but the problem of biological degradation is
zeolite softening does not decrease the alka- no longer encountered. Some organic base
linity of the makeup water. When alkalinity treatments impart a light brown color to the
as well as hardness is high in the makeup treated water, while others still produce a
water, a higher blowdown rate still will be colorless solution.
necessary to prevent high alkalinity and pH Since the heat load on air conditioning
of the recirculating water and consequent at- systems is generally quite variable, depending
tack of cooling tower wood, copper alloys and upon weather conditions, water balances tend
314 BETZ HANDBOOK
to be irregular. Consequently, in order to and therefore less corrosive. This condition

provide a protective margin, inhibitor treat- is generally true except for the important fact
ment levels are usually maintained higher that the air being washed quite frequently
than in industrial cooling tower systems. has sufficient acidic contamination to lower
Algae and biological slime in open cooling the pH of the spray water to a point where
water systems are usually controlled by chlo- severe acidic corrosion can occur. This con-
rinated phenols, fed in a shot fashion. Appli- dition requires the feed of an alkaline ma-
cation of chlorine gas can be effective, but terial, such as soda ash. It is also good prac-
this requires chlorination equipment and tice to use a corrosion inhibitor in air washers
specific controls, which are not practical for both winter and summer.
most air conditioning systems. Chlorine also One of the functions of air washers is to
can be applied as a hypochlorite. However, remove dust, smoke and odor from the air.
hypochlorites must be used frequently and Usually the air being washed will contain
with care so that corrosion will not be in- numerous microorganisms together with ma-
ereiised and tower wood affected by excessive terials which act as food for the bacteria.
chlorine. Consequently, the development of biological
slime is a usual problem in air washers. Un-
AIR WASHERS fortunately, the more effective slimicides such
Air washers can function to moisten the air

in the winter and dehumidify the air in the AIR RETURNING
OUTSIDE TO
summer, in addition to controlling air tem- AIR BUILDING
l
AIR TO
perature. In winter operation, spray water STEAM BUILDING
evaporates so that the washer then operates

like a cooling tower or evaporative condenser.
Cycles of concentration develop and scale
COIL
l
can occur. Scale is not common, however,
since operating water temperatures are usually
ambient or lower. Dehumidification is
achieved by maintaining the temperature of
the spray water below the dew point of the
entering air. In such summer operation the WATER SUPPLY
condensation of water from the air results in CONTROLLED BY CHILLED WATER
FLOAT VALVE FROM COOLER
dilution of mineral salts in the recirculating STEAM
TO COOLER -
water and overflow of water from the sump RETURN SUPPLY
PUMP
or pan of the washer.
Chromate base treatments provide opti- Figure 40-2 • Air Washer with Open Chilled Water
mum corrosion protection. However, because System
of the possibility of airborne contamination
from chromate dust collected in air ducts,
under certain circumstances use of materials
approved by the United States Food and
Drug Administration is desirable. These ma-
terials include the polyphosphates and sili-
cates.
When washers are operating as dehumidi-
fiers the corrosion potential will be less than
when humidifying. Since the water condensed Figure 40-3 • Air Conditioning System: Cooling
from the air contains no dissolved salts, the Tower and Freon Refrigeration with Closed Chilled
water will be low in electrical conductivity Water System and Air Washer
as chlorine, trichlorophenates and organic sul- by attacking system metals and corrosion no
fur compounds impart an odor to the air so longer occur. However, the systems in fact
that they cannot be used in air washers. commonly have sufficient water loss and air
Sodium pentachlotophenate can be used ju- leakage to justify the cost of protective treat-
diciously although it is not particularly effec- ment.
tive unless applied in combination with other The inhibitors used are chromate or nitrite
toxicants. The quaternary ammonium com- based. The amount of inhibitor used will
pounds have little odor, but again they are depend upon the temperature of operation of
more effective when used in conjunction the system with higher temperatures requiring
with other biocides. Periodically, air washers higher levels of treatment. Since these sys-
should be sterilized by stopping air flow and tems use no additional chemical treatment
recirculating a solution of hypochlorite or a after the initial charge, it has been common
non-oxidizing biocide. The washer should to use relatively high treatment concentra-
then be thoroughly hosed, washing off de- tions for a margin of safety.
posits of slime and miscellaneous dirt, all of There have been indications that high
which should be cleaned from the pan of chromate concentrations shorten the life of
the washer. mechanical seals on pumps. Where pumps
are equipped with mechanical seals, lower
chromate concentrations in the range of 200-
300 ppm are sometimes used. This chromate
level reportedly is low enough so that the
seal life is not effected. However, there is a
possibility that this concentration of chro-
mate will not be sufficient to stifle corrosion
in the system, particularly if there are areas
of stagnant conditions or bimetallic couples.
OVERFLOW'
The development of corrosion products re-
sulting from continuing corrosion may pos-
sibly shorten the life of mechanical seals. The
Figure 4G-4 • Air Conditioning System: Evaporative data available at this time regarding this
Condenser with Freon Refrigeration by Direct Expansion matter are highly contradictory and incon-
clusive.
CLOSED WATER SYSTEMS CONTROL OF WATER BALANCES
Closed systems are not subject to scale ex- Weather changes cause water concentrations
cept in the case of abnormal water makeup in open cooling water systems to be erratic,
to a hot water system when the water has a particularly in air washers. Air conditioning
scaling tendency. Under these circumstances systems are designed for considerations other
zeolite softening of the makeup water is than water treatment. To reduce weight cool-
necessary. ing towers located on roofs may have no water
The corrosion potential in closed water sump. The lower water capacity of the sys-
systems is not great but the systems are often tem then causes more rapid change of water
very sensitive to corrosion because of fouling concentrations because of shorter retention
of automatic valves and vents by corrosion time. Also, evaporative condensers and air
products. Theoretically, closed recirculating washers have pans of small water capacity to
water systems should not require corrosion minimize space and weight requirements.
inhibitors. The oxygen introduced with the Cooling towers may be located near smoke-
initial fill of water should be soon depleted stacks where pickup of dirt and acidic gases
316 BETZ HANDBOOK
is excessive. Air conditioning units are some- 3000 - SHOT FEED

times installed and operated in a manner E
a.
a.
such that considerable overflow of water
z 2aoo
from the system occurs on shutdown and 0
there are instances where additional makeup ~
(l:
water has been used during hot weather to lZ 2000 l'llll AREA OF EXCESS
TREATMENT
reduce water temperature. These are some UJ
Q
z
of the factors which can complicate a water 0
Q 1300
treatment program. 1-
z
In open systems, continuous blowdown, UJ
::;:
con.tinuous feed of the corrosion inhibitor and !:( 1000
UJ
daily testing of the water are definitely neces- 0::
1-
sary for best protection. Some systems are ~00
treated and controlled on the basis of weekly
tests, but this requires good proportioning of
chet:nical treatment at higher cost for feeding
12 24
equipment and will require more chemical TIME AFTER SHOT FEED (HOURS)
treatment to assure adequate protection at Figure 40-5 • Waste of Chemical Treatment
all times. with Shot Feeding
CHEMICAL FEED METHODS

Chemical treatment can be fed continu-
Unless an open cooling water system has a ously by direct exposure of the dry chemical
large water content, shot feed of the cor- treatment to water in the basin of the air
rosion inhibitor is not satisfactory. The effec- conditioning unit. Rate of feed is adjusted by
tiveness of treatment will be in direct rela- controlling the amount of chemical exposed
tion to the amount of inhibitor in the water to water. Sometimes the treatment can be
so that it is necessary to maintain a certain formulated to dissolve slowly with the rate of
minimum concentration of chemical treat- addition being controlled by using more or
ment. Continuous blowdown is necessary to less of the formulation submerged in the
prevent scaling and, correspondingly, con- basin. Regulation requires trial and error
tinuous chemical feed is necessary to main- since several factors, such as water turbulence,
tain adequate treatment levels. If treatment affect the rate of solution. Serious overfeed
is shot fed once per day, it must be fed far in can occur if the air conditioning unit is shut
in excess of its required treatment level if the down without removing the chemical treat-
minimum treatment concentration is to be ment.
maintained until the next shot feed. One of the simplest continuous chemical
Figure 40-5 illustrates the rate of decrease feed devices is the drip feeder, a chemical
in treatment concentration under average solution tank with a valved discharge line at
conditions with continuous blowdown. It will the bottom. The treatment is dripped into
be seen that a shot feed of 3,000 ppm must the sump or pan of the air conditioning unit
be made in order that there will still be a by adjustment of the valve. This feeding de-
minimum concentration of 200 ppm at the vice is seldom entirely satisfactory; plugging
end of 24 hours. If treatment were to be of the valve occurs, there is difficulty in
maintained within the normal limits of 200- properly adjusting the valve, and the rate of
400 ppm at all times, shot feed every six flow decreases as the tank empties. When
hours would be necessary. More than three such feeders are used, it is advisable to make
times as much treatment is required when the installation so that the regulating valve
shot feeding once a day as when treating con- is always flooded and the tank raised as high
tinuously in order to maintain the same pro- as possible with respect to the valve (See
tection. Figure 40-6) .
FILL ployed for automatic blowdown control, the

LINE
electric motor of the diaphragm pump can
SHUT -OFF VALVE
be connected with the controller to effect
~ 50- 100 MESH STRAINER
automatic chemical feed.
DRAIN
GENERAL CONSIDERATIONS
The importance of properly beginning chem-

MAXIMUM ical treatment cannot be overemphasized.
DISCHARGE HE"AD
New installations will have mill scale on
metal surfaces and will contain oil, pipe dope
.............. DISCHARGE POINT and other miscellaneous debris. Systems which
have been operated without protective treat-
BASIN OF
AIR CONDITIONING UNIT ment contain corrosion products which
slough-off when protective balances are estab-
lished. These materials can impede water
NOTE'
1. THE INSTALLATION SHOULD BE MADE SO THAT THE MINIMUM DISCHARGE
flow, cause fouling and will increase the cor-
HEAD IS AS LARGE AS POSSIBLE COMPARED TO MAXIMUM DISCHARGE HEAD.
PLACE THE CHEMICAL FEED TANK AS HIGH AS POSSIBLE ABOVE THE
rosion load by cell action. Automatic valves
DISCHARGE POINT AND USE A LOW, WIDE TANK WHEN POSSIBLE.
and controls can malfunction and suspended
2. THE DISCHARGE CONNECTION OF CHEMICAL FEED TANK IS TO BE A
MINIMUM OF 2 ~ABOVE TANK BOTTOM. solids can seriously shorten the life of mech-
3. THE REGULATING VALVE IS TO BE AT A LEVEL BELOW THE DISCHARGE
POINT. USE EITHER A NEEDLE VALVE OR A FLO~ CONTROL VALVE.
anical seals on pumps.
These systems should first be chemically
cleaned, drained and then flushed as thor-
Figure 40-6 • Drip Feeder Installation oughly as possible. Cleaning agents are com-
monly polyphosphates, synthetic detergents or
Chemical feed pumps are quite depend- combinations of these materials. It is im-
able although some types present problems portant to establish the permanent protec-
in regulation and maintenance. Piston type tive treatment immediately after cleaning
pumps are prone to some leakage which can since the metal surfaces in the system are
be particularly objectionable when feeding particularly vulnerable to corrosion.
colored materials such as chromate.
Air conditioning systems quite often do not
A water-motorized proportioning feeder is
operate in winter so that measures should be
particularly applicable for the feed of treat-
taken for protection during the idle period.
ment to air conditioning systems. This unit
Open cooling water systems should be com-
requires no electrical connections whatsoever.
pletely drained if possible. It is good pro-
The device is merely installed in the makeup
cedure to open condensers at the end of an
water line and adjusted to feed a prescribed
air conditioning season, not only to inspect
rate of chemical in direct proportion to the
conditions, but also to provide for drying of
quantity of makeup. the equipment. When the water cannot be
Up to six pumping heads can be handled removed from an idle system it is advisable
by one water-motor base. This feature is ad- to use a higher concentration of the corrosion
vantageous when at least two incompatible inhibitor in order to provide the additional
materials, such as a corrosion inhibitor and a protection necessary when water is not flow-
slimicide, must be fed in a proportional ing. Basins of cooling towers or air condi-
fashion from two tanks. tioning units should be cleaned of any sedi-
Compact diaphragm chemical pumps also ment. The application of protective paints to
provide precise and reliable service for the metal surfaces is generally not satisfactory
addition of treatment to air conditioning sys- unless more than one coat is used and a fully
tems. Where a conductivity controller is em- continuous film is achieved. The development
318 BETZ HANDBOOK
of holidays can result in serious penetration cleaning procedures may also be required
of the metal through localized corrosion. several times during the air conditioning sea-
Whenever ethylene glycol is added to a water son when conditions are particularly severe.
system for freeze protection, any chromate Since the treated water does not reach these
treated water should first be drained from areas during operation in reasonable amounts,
the system since these materials are not the treatment chemicals added to the water
compatible. Chemical feed systems should be cannot be expected to provide protection.
drained and cleaned and the power supply
to chemical pumps disconnected to prevent REFERENCES
the possibility of inadvertent operation. The
packing on piston type pumps should be re- S. M. Elonka and Q. W. Minich, "Standard Re-
placed before start-up. frigeration and Air Conditioning Questions and
Answers", McGraw-Hill Book Co., Inc., New York,
The eliminator sections of cooling towers N. Y. (1961).
are subject to build-up of salts during opera- "Heating Ventilating Air Conditioning Guide 1960"
tiqn as the result of partial or intermittent Vol. 38, Am. Soc. of Heating, Refrigerating and
Air Conditioning Engineers, Inc., New York, N.Y.
wetting. These salts and other dirt and debris (1960)
also tend to accumulate in the eliminator J. R. Schieber, "Chemical Feed Systems for Air
sections of air washers and evaporative con- Conditioning", The Betz Indicator, Betz Labora-
tories, Inc., Phila., Pa. (Nov. 1961)
densers. These areas should be hosed down J. R. Schieber, "Water Treatment Control", Plant
and where appreciable dirt is present me- Engineering, Vol. 13, pp. 146, 148, 150 (June
chanical cleaning may be necessary. These 1959)
41
Industrial Waste Treatn1ent
Industrial Water Use . .................... 320

~ffects Caused by Waste Discharge . .......... 321
Methods qf Waste Treatment . .............. 322
Mechanical Methods . ................... 323
Chemical Methods . ..................... 324
Biological Methods . .................... 324
Advanced Waste Treatment . . . . . . . . . . . . . . . . . 326
Sludge Disposal . . . . . . . . . . . . . . . . . . . . . . . . . 326
Plant Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
320 BETZ HANDBOOK
ATER differs from other natural re- Fig. 41-1

W sources, such as fossil fuels and food, in
that it is not consumed or destroyed by use.
Abridged Classification of
Industrial Wastes by Industries
It is a reusable resource.
Water has extraordinary chemical stability, INORGANIC ORGANIC
but it also possesses strong solvent powers. Metallurgical Vegetable Canning
Therefore, when used in home and factory_, Brass and Copper Corn Products
Iron and Steel
water absorbs substances that can alter its Plating Alcohol Production
overall quality. Water becomes wastewater Sulfur Slaughterhouse
and contains unwanted by-products of human Sand Washing Dairy and Milk
activities. Acid Manufacturing Beet Sugar
The basic-if oversimplified-concept of Salls Citrus Fruit Canning
wastewater treatment is to remove from Alkalies Brewery
Coal Mines Meat Packing
wastewater those foreign substances which it
co,ntains, thereby returning it along the road MIXED
toward pure water to be used for the same or Textile Petroleum

other purposes. The extent to which foreign Dyeing Plastics
substances are removed depends on practical Coke and Gas Wastes Rubber Reclaiming
considerations of the intended uses for the Paper Board and Allied Natural and Synthetic Rubber
reconditioned water. Man-made wastewater Laundry Explosives
tanning Vegetable Dehydration
treatment facilities are designed to augment
Cutting Oils Wool Scouring
and assist the inherent ability of natural Dye Manufacturing
flowing waters to cleanse themselves.
In handling wastewater discharges, in-
dustry is faced with a multiplicity and di- pated. It is reasonable to expect that future
versity of problems. These problems are demands upon water resources will necessi-
largely the outgrowth of technological tate more widespread application of present
changes, industrial expansion and urbaniza- technology, plus the development of new
tion. Industrial waste disposal problems are methods to handle new situations.
not only technical in nature, but also have
legal and public relations aspects that must INDUSTRIAL WATER USE
be considered.
Since the passage of the Water Quality The general subject of industrial waste is of
Act of 1965, there has been increased public broad scope. This discussion is primarily di-
concern and activity at all levels of govern- rected to liquid or waterborne industrial
ment in matters relating to pollution control wastes which are eventually discharged into
and abatement. This Act provides that criteria a watercourse.
be established on all interstate waters so as to Substantially all industries contribute some
enhance quality and to protect public health liquid or waterborne industrial waste to a
and welfare. waterway, either directly or indirectly. Treat-
Water quality criteria promulgated for the ment is usually required to prevent the im-
enhancement of water resources may require position of an excessive pollution load on the
increased performance from existing waste- receiving stream, lake, river or waste treat-
water treatment facilities as well as the re- ment plant.
moval of constituents which normally pass Certain industries are unaware of enormous
through these processes. With increasing losses of valuable by-products in their wastes.
density of population and industrial expan- Similarly, industries are often not cognizant
sion, the demands for cleaner water and more of the volume and effect of their waste upon
stringent water requirements can be antici- the receiving streams or sewers. For these
reasons, industrial waste problems may exist EFFECTS CAUSED BY WASTE DISCHARGE
and may not be recognized by an industry
having such problems. In order to measure the effects caused by
waste discharges and to ascertain that water
Most industries use water in their pro- quality criteria are being met, the legally
cessing operations in such manner that some designated regulatory agency generally estab-
of the water becomes contaminated. Figure lishes some form of qualitative and/or quan-
41-2 indicates the tremendous quantities of titative specifications for each waste dis-
water employed by industry. These data are charge. Where standards are set, they may be
not meant to indicate consumptive use of 'effluent standards' or 'receiving water stand-
water, but are illustrative of the large quanti- ards.' In some instances, both types of stand-
ties of total water required by various types ards are simultaneously employed, with one
of industry. Portions of the water, such as that covering general circumstances and the other
used for cooling purposes, may be uncon- covering specific aspects.
taminated. But, substantial portions will re-
Effluent standards state precisely what is
quire treatment to avoid imposing a pollution
allowed in a waste discharge as the discharge
load upon the receiving waters.
is introduced into a receiving water. These
standards may specify the amount of any
Fig. 41-2 particular quality constituent that an effluent
Abridged List of may contain, or they may specify the per-
Industrial Water Requirements
centage of removal required from the waste
treatment facility. Through effluent standards,
GALLONS OF WATER a portion of the assimilative capacity of the
INDUSTRY PER UNIT OF PRODUCTION
receiving water is, in effect, allocated to each
CHEMICAL INDUSTRIES waste discharge so as to insure that the
Corn Refinery . . . . . . . . . . . . . 333 per ton of starch quality criteria set for the receiving water will
Alcohol . . . . . . . . . . . . . . . . . . 20,000 per ton of grain be met during normal low flow conditions.
Ammonium Sulfate ........ 200,000 per ton
Gun Powder . . . . . . . . . . . . . . 200,000 per ton Receiving water standards establish quality
Lactose . . . . . . . . . . . . . . . . . . 220,000 per ton conditions which must be met after adequate
Hydrogen . . . . . . . . . . . . . . . . 660,000 per ton
admixture of a waste discharge with the re-
FOOD INDUSTRIES ceiving water in order to comply with criteria
Beet Sugar . 20,000-25,000 per ton of sugar promulgated for the receiving water. Re-
Canning . . . 300-25,000 per 100 cases of No.2 cans
Meat Pack- ceiving water standards normally contain no
ing ..... 16,000-55,000 per 100 live units specific limitations on the composition of the
Milk
Products ..
waste discharges, provided that the quality
200-250 per 100 lb
Vegetable of the receiving water is maintained within
Dehy- specified limits. This approach takes ad-
dration ... 500-2,000 per 100 lb
vantage of the varying natural flows and as-
TEXTILE INDUSTRIES similative capacities of receiving waters. The
Cotton ................... . 1,000- 3,800 per 100 lb receiving water must be examined to ascer-
Linen .................... . 10,000 per 100 lb
Rayon ................... . 8,000-10,000 per 100 lb tain that these standards are being met.
Wool Scour ............... . 2,000-12,000 per 100 lb The effect of individual wastes upon the
MISCELLANEOUS receiving waters will vary, dependent upon
Air Conditioning 6,000-15,000 per person, per season the volume, concentration and character of
Automotive .. . 15,000 per car the waste as well as the flow, location and
Oil Field ..... . 18,000 per 100 bbl crude
Oil Refinery .. . 77,000 per 100 bbl crude downstream use of the receiving waters. Ob-
Steam Power .. 80 per kwh viously, the discharge of large volumes of
Steel ........ . 20,000-35,000 per ton relatively highly concentrated waste into a
Sulfur ...... . 3,000 per ton
very large receiving body of water, where ap-
322 BETZ HANDBOOK
preciable dilution will result, may have a Sewage treatment plants are operated by
lesser effect than the discharge of a smaller many cities, some treating a combination of
volume into receiving waters where a mini- domestic sewage and industrial trade waste.
mum of dilution can be obtained. If industrial wastes change in volume or
The direct effect on the receiving waters of character, serious interference with operation
the discharge of industrial waste is dependent of the sewage treatment plant may result.
upon the character and specific types of ma- This operating interference may take many
terials present in the waste. Solids may pre- forms, from overloading of solids removal
cipitate with the result that all flora growth facilities to drastic interference with biologi-
is· killed. Solids which settle out may shroud cal stabilization. Excessive solids, presence of
fish breeding grounds, and minute particles certain toxic metals or oil, or excessive con-
may lodge in fish gills and cause suffocation. centrations of organic constituents will cause
Oil waste discharge will float to the surface one or more of the problems indicated.
and may prevent reoxygenation of the water. METHODS OF WASTE TREATMENT
Toxic metals and salts, acids and alkalies may
cause flora and fauna casualties. The wastes The technology of industrial wastewater treat-
may produce taste and odor contamination ment practice has not advanced to a point
which can render the water unfit for livestock comparable to industrial process and domestic
or for human consumption, even following potable water treatment. For the solution of
normal treatment. certain waste treatment problems, vast
The problems of industry due to contami- amounts of specialized research are required
nation of its water supplies by industrial to devise a practical treatment method.
waste discharge are numerous. In many cases, Minor deviations in manufacturing practice
treatment of an industrial water supply for among plants in the same type of industry
process, boiler or cooling water uses is more and variations in local conditions may neces-
difficult and costly when a polluted surface sitate an entirely different approach to the
supply is used. Frequently, specific corrosion waste treatment problem. Therefore, generali-
problems in industrial plant water circulating zations regarding treatment methods for vari-
systems or product damage can be traced to ous types of industries cannot be drawn with
contamination of the water supply by up- accuracy. Each problem must receive specific
stream discharges of industrial waste. attention, consideration and evaluation for
its solution.
Fig. 41-3
Average Analyses of Domestic Sewage
in Milligrams per Liter
Courtesy Link-Belt Company
ANALYSIS WEAK MEDIUM STRONG
Tolal Solids .................. . 430 720 1230 Figure 41-4 • Straight-line grit chamber for removal
Total Volatile Solids .... , ...... . 240 420 810 of inorganic solids.
Suspended Solids ............. . 98 200 372
Suspended Volatile Solids ...... . 72 133 220 The engineer working in the field of in-
Settleable Solids (ml/1) ... , ..... . 2.1 3.8 6.4 dustrial waste treatment has available certain
Biochemical Oxygen Demand ... . 96 212 413 tools with which to attack the industrial
Oxygen Consumed ...... , , .... . 74 162 267
Total Nitrogen ............... . 13 28 40 waste problem. These tools consist of certain
Ammonia Nitrogen , ........... . 4.2 12 22 unit processes or combination of processes.
Chlorides .............. , , .... . 18 38 79 These processes may be divided generally into
Sulfates ..................... . 42 22 34
Soaps and Fats .......... , .... . 6 13 23 mechanical treatment methods, chemical
Total Alkalinity .......... , , . , .. 31 40 18 treatment methods and biological treatment
pH ..•...•...•.•.•••.•..•.•... 6.9 7.4 7.1
methods. Only on very rare occasions is dilu-
tion alone an acceptable method for the dis- cal treatment. Air may be satisfactorily ap-
posal of industrial waste. In the treatment of plied by spraying the waste through suitable
some industrial wastes, the need to segregate nozzles, adding air through diffuser tubes or
certain portions is paramount, while with cascading the wastes over suitable aerators.
others it is highly desirable to mix the wastes There are also other available mechanical
and thereby utilize the beneficial effects of aerators which are applicable for certain
interaction. plant installations.
Occasionally, there are wastes which can Flotation is a specialized form of mechani-
be satisfactorily discharged to large holding cal treatment in which the solid material en-
lagoons from which evaporation, ground traps air and rises to the surface for removal.
percolation and drying represent all the treat- Flotation may be carried out either under
ment that is required. In other cases, wastes atmospheric pressure or under vacuum.
may be disposed of by field irrigation, re- Chemical aids are sometimes employed to
sulting in benefits to the soil, and thus avoid- assist in this operation.
ing the use of more sophisticated treatment
plant facilities.
MECHANICAL METHODS. The simplest form
of mechanical treatment for waste involves
screening for the separation of the larger solid
particles. Screening has its limitations in that
the use of fine screens presents mechanical dif-
ficulties in operation. Certain wastes contain
varied amounts of particulate inorganic solids
which can be effectively removed by the use
of grit chambers. The debris from screens and
grit chambers may differ considerably. The
collection from grit chambers is usually in-
organic and nonputrescible, and can be taken
to dry disposal. Screen residues may be both Courtesy Komline-Sanderson Engineering Co.
inorganic and organic in nature and require
Figure 41-5 • Vacuum .filter for sludge dewatering.
burning or burial. Suspended solids remain-
ing after screens or grit chambers may be Filtration of the wastes may be applied
further reduced by sedimentation, either by either for the clarification of the wastes or
natural subsidence or by chemical coagulation. for dewatering the separated solids. Centri-
In addition to solids, many wastes contain fuges, either batch or continuous, may be ap-
varied amounts of oil, either in the free state plied for similar purposes, treating either the
or emulsified form. Preskimming chambers previously separated sludge or the entire
for readily separable oil are quite effective, volume of waste as the particular case may
the skimmed oil being either recovered for require.
use or disposed of by burning or other suit- Conventional sedimentation is one of the
able means. Oil in the emulsified state is basic waste treatment methods. When natural
much more difficult to remove and may re- or mechanical sedimentation does not pro-
quire acid cracking or precoalescence. Oc- duce the desired results, or when conditions
casionally, greater separation of both emulsi- of the waste are such that the pollution is in
fied and free oil can be accomplished by a colloidal or soluble form, it is necessary to
preaeration. add chemicals to produce suitable sedimenta-
Aeration, which is highly effective for cer- tion. Equipment is commercially available
tain waters, may also be included in mechani- which accomplishes both settling and flota-
324 BETZ HANDBOOK
tion within the same unit. In such a unit, ing or reducing components require neutrali-
two layers of solids, one from the surface and zation of these factors prior to discharge or
one from the bottom, are separately removed, further treatment. Chemical conversion of
eliminating the necessity of forcing a readily some constituents can be accomplished by
settleable material to float or to settle a dif- strong oxidizing or reducing agents, render-
ficult settleable component. ing them unobjectionable. Certain wastes re-
Evaporation or drying is applied to certain quire that adsorptive materials be added for
wastes. Normally, these methods cannot be the removal of undesirable characteristics.
economically applied unless by-product re- Some wastes, following preliminary treat-
covery is realized from the process. However, ment, may be treated through ion exchangers
there are occasions when low flows of particu- prior to final discharge. Sterilization of the
larly difficult wastes are encountered, and waste may also be necessary and chlorine is
drying or evaporation is the only practical usually employed for this purpose.
... known method.
Fig. 41-6
CHEMICAL METHODS. Chemical flocculation Reduction Efficiency of Units and
and precipitation find application in the treat- Plants on Domestic Sewage
ment of certain industrial wastes, either alone
or followed by other treatment. These proc- PERCENT
SUSPENDED PERCENT PERCENT
esses are carried out in a manner similar to UNITS AND PLANTS SOLIDS BOD COLIFORM
the treatment of water and in most cases in Sedimentation (2 hours) 45-60 30-45 40-60
similar types of equipment. Chemical treat- Imhoff Tanks (2-3 hours) .... 45-60 20-45 40-60
Trickling Filters (Standard) .. 10-30 55-75 60-70
ment is usually effected by means of a suitable Activated Sludge , .••..•••. 85-95 80-95 90-96
mixing apparatus and straight line or upflow Chemical Treatment ........ 65-90 45-75 60-90
type clarifiers. Chemical treatment may be Sedimentafion and
handled in either batch or continuous fashion,
Sand Filters ... ·········
Sedimentation and
90-98 85-92 85-95
or a combination thereof. Chemical precipi- Trickling Filters ......... 75-85 70-90 80-90
tation is popular, due largely to its ability to
operate unaffected by greatly varying loads BIOLOGICAL METHODS. Biological treatment
of toxic material and to its usually lower capi- may be generally divided into two classifica-
tal expenditure in comparison to more in- tions. Those methods requiring the presence
volved biological processes. of oxygen are known as aerobic processes. The
Flocculation and coagulation are used two aerobic treatment processes most com-
primarily as aids in sedimentation. The pur- monly employed are the trickling filter process
pose of coagulation is to obtain the removal and the activated sludge process. Both proc-
of suspended and colloidal solids and also to esses are used for the treatment of industrial
effect precipitation of materials which may wastewaters. Proponents of the trickling filter
be present in solution in the waste. The method hold that it is less susceptible to shock
coagulants used must be selected for the resulting from change in rate of flow or char-
particular waste undergoing treatment, and acteristics of the waste being treated. It has
are usually aluminum sulfate, iron salts, been shown, however, that well-designed and
sodium aluminate, calcium chloride or lime. operated activated sludge plants have effec-
Certain wastes may be highly acid or tively handled high proportions of various
highly alkaline, and will require neutraliza- industrial wastes. In general, both processes
tion as part of the treatment. There are times have applicability for industrial waste treat-
when neutralization may be accomplished by ment. The choice is determined by circum-
combining the various wastes from a plant. stances particular to each situation.
Industrial wastes which contain highly oxidiz- The trickling filter and activated sludge
processes are similar in principle in that both biota are supplied with oxygen from the air.
depend on biochemical oxidation of com- The activated sludge is similar in character,
plex organic matter in wastes. Soon after a composition and action to the biological film
filter is placed in operation, the surface of the in trickling filters. When wastes are mixed
filtering media becomes coated with zooglea, with a sufficient amount of biologically active
a viscous substance containing bacteria and sludge, rapid clarification takes place. The
other biota. Under favorable environmental finely divided suspended, colloidal and dis-
conditions, the zooglea absorb and utilize sus- solved solids in the wastes are transferred to
pended, colloidal and dissolved organic mat- the surface of the sludge floc, in which large
ter from the wastes which pass in a relatively numbers of living organisms have their habi-
thin film over its surface. Eventually, a popu-
tat. These organisms, supplied with oxygen
lation equilibrium is reached. As biota die,
from the air, oxidize the organic material.
they, together with the more or less partly
The floc grows and settles out, leaving a
decomposed organic matter, are discharged
from the filter. This sloughing of material clear, relatively stable liquor. To produce
may occur periodically or continuously. Gen- rapid results, the floc must be kept in sus-
erally, secondary settling is provided to re- pension by agitation and contact with oxygen.
tain the settleable solids sloughed from the A continuous supply of active sludge must
filter. be mixed with incoming wastes, hence a por-
The activated sludge process also makes tion of the settled sludge is returned to the
use of biological organism populations. These head of the process. Excess sludge not needed
to maintain the process is wasted.
There are possible occasions when natural
oxidation ponds may have application for
the treatment of some industrial wastes.
Usually, industrial wastes are too strong for
this approach, and require such large land
areas as to make another aerobic treatment
method more attractive economically.
The second classification of biological
treatment takes place in the absence of air
and is designated as anaerobic. This method
finds limited use in liquid waste treatment,
being normally confined to the treatment of
sludge resulting from some of the other
processes.
Combination of trickling filters and acti-
vated sludge treatment is often required on
high organic wastes to effect complete treat-
ment. In biological treatment, as in chemical
precipitation, it may be necessary to add
chlorine to the final waste prior to discharge.
In those locations where geologic sub-
Courtesy Link-Belt Company structures are satisfactory, deep wells have
Figure 41-7 • Straightline bar screen for removal of been employed to dispose of difficult to treat
large solids. industrial wastes.
326 BETZ HANDBOOK
Courtesy Infilco-Division of Fuller Co,
Figure 41-8 • Trickling filter for biological stabilization.
ADVANCED WASTE TREATMENT trodialysis, evaporation, freezing, foaming,

chemical oxidation, solvent extraction, ion
Considerable research is being conducted on exchange and reverse osmosis.
advanced waste treatment, or tertiary, meth- Approaches under study for the disposal of
ods. One of the objectives of this research is concentrated impurities include the recovery
to renovate wastewater for direct and de- of usable products from the wastes, dumping
liberate reuse. Advanced waste treatment is of highly concentrated wastes in waste sinks,
considered to be a two-step process: ( 1) subsurface disposal through deep-well injec-
separating concentrated contaminants from tion, and converting the refractory contami-
the purified water, and ( 2) disposing of these nants into innocuous material and using them
contaminants in a way that will render them for such purposes as land fill.
totally innocuous. The need for advanced
waste treatment technology stems largely SLUDGE DISPOSAL
from the fact that some of the inorganic and
organic contaminants now entering receiving Practically any waste treatment system will
waters resist every phase of present day treat- ultimately present the problem of disposal of
ment-waste treatment, natural purification sludge. This phase of treatment often presents
in streams and water treatment. There is ap- greater problems than are encountered in
prehension over the long range effect that treating the liquid wastes.
these substances may exert. Some of the Sludge dewatering schemes have the ob-
processes and principles being studied in- jective of rendering the sludge suitable for
clude: adsorption on activated carbon, elec- further dewatering or for disposal. Ultimately,
the residual sludge product is disposed to the the sludge has the proper charactertistics for
land, to oceans or to the atmosphere. All digestive action. Toxic metals and salts, oils
sludge handling and dewatering procedures and greases and excessive inert materials in-
are conceived to prepare a product that will terfere with successful digestion. Provided
be suitable for practical disposal at the point conditions are optimum, digestion produces
of collection. The form of disposal that is a readily dewaterable, non-odorous sludge
practical at a particular site is dependent that can be handled without difficulty.
upon geographical, political and other con- Sludge cake from vacuum filtration or
siderations. underdrained drying beds is normally suitable
Various methods have been employed to for disposal to land by dumping or by burial.
accomplish sludge disposal. Where adequate In some cases the product is spread on agri-
ground is available, solar drying of sludge cultural land. Heat dried sludge product may
may be accomplished in lagoons or sludge at times be marketed as a soil conditioner or
drying beds. Rotary or spray dryers can be plant food.
applied where space limitations prevent the Liquid sludges near coastal areas may be
use of solar drying. Some sludges dewater disposed of in the ocean by means of barges
readily, and this may be accomplished by the or by pumping through long outfall lines.
above mentioned methods or by vacuum fil- Liquid sludges occasionally have been spread
ters or centrifuges. With the recent develop- on agricultural land from tank trucks.
ment of improved dewatering devices, it is
now possible to incinerate raw sludge. PLANT DESIGN
Incineration or wet combustion processes
dispose of gaseous combustion products to the The design of a successfully operating waste
atmosphere. The ash residue is usually treatment plant depends upon the selection
dumped on land or used for fill. Such proc- of the proper unit processes to accomplish
esses entail high capital outlay, but feature a the required degree of treatment and purifi-
cation of the waste. The degree of purifica-
tion required is a function of State regula-
tions, which take into account local conditions
and requirements. The degree of treatment
required is such that the waste discharge will
not accomplish degradation of the receiving
water nor interfere with its downstream uses.
Each waste treatment problem, therefore, re-
Fig. 41-10
Waste Produced per Employee
of Each Industry
!Domestic Sewage-1.0)
Courtesy of Rex Chain Belt Inc.
Figure 41-9 • Sludge flotation clarifier. OXYGEN SUSPENDED

INDUSTRY DEMAND SOLIDS
minimum of residue to be disposed of to the Tannery ................... · · · 25.3 12.0

Chemical Manufacturing ....... . 22.5 7.4
land. Thickened sludges or partially de- Organic Waste ............... . 2.9 2.4
watered sludge cake often have sufficient Mill Pickling ................. . 3.6 3.8
heat value to support combustion without Dye Wastes .................. . 44.2 5.4
Laundry ................... · · · 22.8 22.2
the addition of auxiliary fuel. Distillery .................... . 470.0 181.0
Digestion of sludge has long been recog- Dairy ..................... · · · 44.6 5.7
Miscellaneous ................ . 13.2 6.8
nized as a suitable method of stabilization, if
328 BETZ HANDBOOK
quires individual attention and study to de- most wastes do not contain a sufficient
termine and apply appropriate treatment. amount of recoverable material to offset the
Probable changes in plant processes which cost of treatment. On the other hand. some
would affect the composition of industrial industries do waste recoverable products. The
wastes must be taken into account to assure economics of waste treatment is not neces-
the adequacy of the treatment facilities as sarily associated solely with the recovery of
installed. Changes in, or additions to, the reusable or saleable substances. In certain
products manufactured may also affect the plants, where water is in extremely short sup-
industrial wastes. Such manufacturing modi- ply, treatment of plant waste is economical
fications should be anticipated, if possible, for salvaging reusable water. In other cases,
when the treatment plant is designed. the cost of the plant water supply is such as
The matter of plant housekeeping is of to render the recovery of water by treatment
utmost importance. Vast economies in treat- of plant wastes attractive. Frequently, a de-
ment costs, both for capital and operation, tailed study of plant water uses will reveal
can be realized by the proper practice of that water used in certain operations can be
adequate industrial housekeeping rules. reused in other operations with little or no
In certain cases, a city sewer system can be treatment in the interest of operating econ-
used as the receiver for industrial waste omy. Total waste flows requiring treatment
water. Dependent upon the effect of the in- may be thereby reduced, resulting in lower
dustrial waste upon operation of the treat- treatment cost.
ment plant, varying degrees of pretreatment
of the waste may be required. In most cases, REFERENCES
where wastes are discharged to the city
sewers, a charge is assessed for receiving and C. F. Gurnham, "Principles of Industrial Waste
Treatment", John Wiley & Sons, Inc. (1955)
treating wastes to render them acceptable for
R. H. Marks, "Industrial Waste Treatment", Power,
ultimate discharge. Many considerations Vol. 109, pp 185-192 (May, 1965)
need to be properly evaluated before deter- R. Neboline and E. J, Donovan, "Treating Meth-
mining that this will serve as a satisfactory ods Available for Industrial Waste Water", Plant
method for waste disposal. The type of waste Engineering, Vol. 20, pp 117-121 (May, 1966)
W. Rudolfs, "Principles of Sewage Treatment", Na-
must be such as to cause no interference, and tional Lime Association, Fourth Edition ( 1955)
the city treatment plant must be of proper G. D. Symons, "Industrial Waste Disposal", Sewage
type and adequate capacity for the added Works Journal, Vol. XVII (1945)
load. "Advanced Waste Treatment Research Program",
U. S. Public Health Service ( 1964)
Not every waste contains valuable re- "Water in Industry", National Association of Manu-
coverable by-products; as a matter of fact, facturers and U. S. Chamber of Commerce ( 1965)
42
Analytical Methods
and Equipment
N the conditioning of industrial water, it RECOMMENDED ANALYTICAL METHODS
I is necessary that analyses be made to gov-

ern the treatment processes. Water treatment
In this book, all recommended test methods
are those which are the most applicable for
without control analyses would be useless and
control analyses by the plant personnel. The
sometimes harmful. For control purposes, a
methods have been selected for their simplic-
complete analysis of the water usually is not
ity, rapidity and convenience with minimum
required. For example, in sodium zeolite sof-
sacrifice of accuracy. While suitable for use
tening the important determinations are hard-
by an experienced chemist, the tests also can
ness and chloride. The hardness test is em-
be employed by personnel with no formal
ployed to determine the end of the softening
chemical training.
cycle and the chloride test to determine the
end of the rinse cycle. A complete water Methods are given for those tests com-
analysis to control a zeolite softener would monly conducted for the control of various
be pointless. Similarly, in lime-soda softening softening processes, internal boiler water
the important tests are hardness, phenol- treatment and the conditioning of industrial
phthalein alkalinity and methyl orange alka- cooling water. Tests made infrequently, or
linity. These three tests provide proper con- under unusual circumstances, such as sodium,
trol of the softener and additional tests are potassium, fluoride, boron, etc. have been
unnecessary under normal circumstances. omitted. No tests have been included for
When the proper tests have been selected, bacteriological or sanitary purity or for trade
it is necessary that the analyses be conducted wastes.
by some responsible person at the plant. The Referee methods of analysis are contained
results of the control tests determine the ad- in such authoritative works as the "Manual
justments required in chemical treatment. on Industrial Water", published by the Amer-
Treatment methods and practices should be ican Society for Testing Materials and
checked by complete analyses at periodic in- "Standard Methods for the Examination of
tervals by the supervising laboratory. How- Water and Wastewater", published jointly
ever, the actual daily control of any treat- by the American Public Health Association,
ment process should be in the hands of the American Water Works Association, and the
plant personnel. The tests must be conducted Water Pollution Control Federation. The
promptly after sample collection so that any referee methods published in these references
treatment adjustment necessary can be placed are intended to be the most precise and ac-
into effect without delay. While many proc- curate known. It is recommended that these
esses require more frequent tests and treat- publications be consulted where the utmost
ment adjustment, the minimum frequency precision and accuracy are required, such as
is usually once daily. for research work or where litigation may be
330 BETZ HANDBOOK
involved. example, in sampling from a boiler, the point

of sampling should be blown thoroughly be-
RECOMMENDED ANALYTICAL EQUIPMENT fore the sample is taken. It is desirable to
sample the boiler from the continuous blow-
In titration methods, only common labora- down rather than from the gage column. It
tory glassware usually is required. However, is also important to secure the sample with
in colorimetric work a specific color compara- as little flashing of steam as possible, employ-
tor has been recommended in each case. ing a cooling coil for this purpose. In the
While color comparisons may be made with case of a sample from a tap or valve, the
the use of Nessler tubes or a photometer, it water should be allowed to run for a few
is more rapid and convenient to use one of minutes before collecting. Sampling points
the several types of color comparators com- should always be selected so as to secure a
mercially available. These units are relatively representative sample rather than from a
inexpensive and are suitable for most in- "dead spot". The sample should be collected
dustrial control water analyses. in a clean container, which has been rinsed
'Therefore, recommendations have been several times with the water to be sampled.
made under the individual analysis for spe- The sample should be cooled to room tem-
cific commercial analytical equipment when perature ( 60-80 F) and any suspended mat-
desirable. For example, Taylor color com- ter permitted to settle. Analysis is made on
parators have been recommended for certain the clear supernatant liquid. Other necessary
tests. These are rugged, practical units that precautions have been noted under the in-
have been found suitable for control work. dividual tests.
Other satisfactory units are available from
different manufacturers.
CARE OF EQUIPMENT
COMPOSITION OF REAGENTS Too much emphasis cannot be placed upon

the necessity for keeping the apparatus clean
The chemical reagents specified, such as at all times in order to obtain accurate re-
"Sulfuric Acid, Concentrated" under the in- sults. Bottles and containers should be cleaned
dividual tests should all be reagent grade carefully after each using. If dirt or grease
laboratory chemicals. Where standardized re- accumulate on the glassware, it can be re-
agents have been specified, the composition moved by cleaning with a strong solution of
of these reagents is shown under the section chromic acid, made by slowly adding one
of this book entitled "Composition of Pre- liter of concentrated sulfuric acid to a 35 ml
pared Reagents". However, standardized re- saturated solution of sodium dichromate.
agents should be prepared only in a well-
All measuring apparatus such as pipettes
equipped laboratory, under the supervision
of an experienced chemist. It is advisable to and graduated cylinders should be carefully
rinsed before using, preferably with a portion
purchase standardized reagents from a labo-
of the water to be tested. Flasks, casseroles
ratory supply house if complete laboratory
and similar vessels in which titrations are
facilities are not available for their prepara-
made should be rinsed with distilled water,
tion.
but a portion of the water to be tested may
be used if distilled water is not available.
PREPARATION OF SAMPLE
If the results obtained from an analysis are to TITRATION METHODS

be of any value, it obviously is necessary to
secure a sample that is truly representative The most common method of plant control
of the condition of the water at the point analysis is by titration. This method of testing
from which the sample was obtained. For is based on the use of a burette from which an
amount of standard solution is added to the the bulb to pump air into the bottle until
sample until an "endpoint" is reached. The the solution has filled the burette (see Figure
endpoint usually is a color change. 42-1). Release pressure by removing thumb
For plant use, automatic burettes should and the excess solution in the burette will
be employed. An automatic burette is con- return to the reservoir bottle, the solution
structed so as to start each titration at "zero" level in the burette remaining at the "zero"
reading, thus eliminating any subtraction to mark. Some types of automatic burettes are
determine the amount of standard solution equipped with a flexible plastic reservoir bot-
employed. tle instead of an air tube and bulb. This type
In using burettes, always employ the same of burette is filled by squeezing the bottle
Figure 42-1 • Typical Automatic Burette Design Figure 42-2 • Meniscus
one for any given standard solution. In the until the reagent rises above the "zero" mark
tests decribed, 25 ml automatic burettes have and then releasing the pressure. In titrating
been specified for convenience in handling. from any burette, be certain that the outlet
Smaller automatic burettes of 15 ml and 10 tip is filled with solution before starting a
ml capacity may be used, although usually titration. The reading of the liquid level in
with less convenience. Always arrange buret- either a burette or graduated cylinder should
tes in some standard order to avoid confusion. always be made at the bottom of the curved
If automatic burettes are employed, fill the surface, called the "meniscus", as noted in
reservoir with standard solution to be used, Figure 42-2. This illustration is typical of the
first rinsing the reservoir bottle and the bu- appearance of a meniscus in an ordinary bu-
rette with a small amount of the solution. rette or graduated cylinder. In most auto-
Place the thumb on the open glass tube to matic burettes, the center tube slightly affects
shut off the air from the reservoir bottle. Use the curvature of the meniscus.
332 BETZ HANDBOOK
In using burettes in a plant, they should some cases for routine plant control.
be protected against breakage and maintained One of the most useful tools available to
in a position ready for use. Most plants have the analyst in this search for more sensitive
found the use of a test cabinet containing all methods is the filter photometer (sometimes
of the equipment to be the most satisfactory also referred to as photoelectric colorimeter,
procedure. However, in protected laborator- photoelectric comparator or photoelectric
ies, offices, or similar areas, a simple titration photometer) . In general terms, a photometer
stand for holding the burette is satisfactory. can distinguish differences in color intensity
not apparent to the human eye and can con-
COLORIMETRIC METHODS sistently provide a reliable reproducible read-
ing. In addition, a spectrophotometer can
Colorimetric tests are particularly useful for measure light absorption in the ultraviolet
plant control because of the simplicity of such and infra-red ranges, not visible to the hu-
methods. These methods are based on the man eye. A flame accessory for the spectro-
.. development of a color in the sample which photometer normally is used in the determin-
is' proportional to the substance to be de- ation of minute quantities of such elements
termined. The concentration present in the as sodium, lithium and potassium.
sample is determined by comparison with
PRINCIPLE OF OPERATION. White light, pass-
color standards.
ing through a solution, is partially absorbed.
The basic method of color comparison is The unabsorbed portion will produce a sensa-
with the use of Nessler tubes. These tubes tion of color to the human eye. The strength
are available in matched sets made from or depth of this color will vary as the inten-
uniformly drawn tubing. Nessler tubes are sity of the unabsorbed light. A photocell can
used in color comparison work by comparing be used to measure the intensity of the un-
the color produced by the unknown sample absorbed light, thus determining the light ab-
with the color of standard solutions of dif- sorptive capacity of the solution. Consequently,
ferent concentrations, after treatment with with photometers it is the light absorptive
the color-developing reagents. Since standard properties of the solution which are measured.
solutions of known concentration must be The instrument does not measure color, but
prepared for each set of tests, Nessler tubes instead the intensity of light transmitted
are not adapted to rapid control analyses. through the solution.
For color comparison work in control water Photometers operate by responding to the
analyses, it is recommended that one of the unabsorbed light which is passed through the
several types of commercial color compara- solution in the absorption cell and which
tors be employed. Color comparators are strikes a photocell. The photocell transforms
available in which the color standards are
sealed in glass ampoules. Other comparators e
use colored glass discs as standards. Usually,
these instruments carry a nonfading guaran-
tee for the color standards. Units of this type
are to be preferred to Nessler tubes for con-
trol purposes.
PHOTOMETRIC METHODS M-Micro~ammeter
P-Photo-cell
As water treatment problems and the meth- C-Absorption Cell R-Spherical Mirror
F-Filter Disc ¥-Voltage Stabilizer
ods for their solution have grown more com-
plex, there has arisen the need for more ac-
curate analytical determinations-not only in Figure 42-3 • Typical Photometer-
laboratory and research analyses - but in Principle of Operation
Figure 42-4 spaced standards. For example, if the phos-

FILTER DATA phate content of the boiler water is con-
trolled between 30-60 ppm as P04 , the use
Peak Transmission, Effective Width, of a conventional color comparator is satis-
millimicrons millimicrons
factory and there is no need for the addi-
415 ................................ 77 tional accuracy secured by a photometric
445 54 method. However, where phosphate may be
460 54
490 54 controlled between 3-6 ppm in some once-
520 57 through cooling water systems or within nar-
535 57
550 45 row limits in some high pressure boiler
580 44 waters, there is considerable advantage m
610 38 the use of a photometric method.
640 40
Some titration methods for chromate in
cooling water systems are subject to interfer-
ence by various contaminants. The diphenyl-
the light into an electric current which is
carbohydrazide test is to be preferred because
proportional to the intensity of the light.
of its freedom from such interference. Com-
This current is measured by a microammeter,
parator methods using fixed standards are
as shown in Figure 42-3 to produce a reading
unsatisfactory because of the rapid deteriora-
on the dial. For those solutions following
tion of the reagent. The photometric method
Beer's law, the dial reading thus obtained is
possesses such a high degree of accuracy that,
proportional to the intensity of the color in
even when high dilutions of the sample are
the solution, which in turn is proportional to
made, no significant error is introduced.
the concentration of the ion being measured.
In analytical measurement, it is desirable Similarly for those situations where small
to measure the light absorptive capacity of concentrations of nitrate, copper, iron and
the solution at a wave length where there is sulfate are significant, the use of photometric
maximum light absorption by the solution. methods is advisable.
Instead of passing white light through the With photometric methods it is important
solution in the absorption cell, filters are to be aware of the effect of temperature on
used which absorb all light except that of color development. With practically any test
the desired wave length. By using such a there is a definite effect of temperature on
properly selected filter, maximum sensitivity color intensity and consequently on the pho-
can be obtained. Theoretically, of course, for tometer dial reading. To secure reliable re-
the most precise work monochromatic light sults, all samples should be tested at the same
(light of one wave length only) should be temperature used in preparing the calibra-
employed. Spectrophotometers can produce tion curve, namely room temperature. The
nearly monochromatic light, but normally are normal variation in room temperature is not
too expensive and fragile for plant control. significant in most tests. However, where the
Filter photometers employ glass filters which utmost accuracy is required, as for example
pass light in a narrow range of wave length in the silica tests, all standardization and test-
only. Figure 42-4 shows the peak transmission ing should be conducted at exactly the same
value and effective width at 50% transmis- temperature.
sion of the ten filters supplied with a typical
photometer. OTHER EQUIPMENT AND METHODS
APPLICATION IN WATER ANALYSES. As pre-
viously mentioned, the use of photometers There are other methods of analyses which
is most pertinent where it is desired to secure are of importance to the water treatment
greater accuracy than is possible with visual engineer but which are too tedious and com-
color comparators which employ fairly widely plicated to lend themselves to plant control.
334 BETZ HANDBOOK
Such methods as flame photometry, X-ray meter. This instrument measures electrical
diffraction, microscopic and gravimetric an- conductance which is proportional to dis-
alyses can be important in the laboratory. solved solids. The method of test is particu-
However, there are other special instruments larly important in the control of boiler blow-
which may be required for plant control for down and as a check on steam and
certain tests. The turbidimeter is such an condensate purity.
instrument. Where turbidity is an important Finally, the battery operated or line op-
control, such an instrument is desirable. erated pH meter is sometimes employed
Another special instrument, which is im- where greater accuracy in pH measurements
portant in plant control, is the conductivity and alkalinity titrations is required.
Figure 42-5 • A Modern Laboratory for Water Anarysis

43
Expression of
Analytical Results
N an analysis of a sample of water, it is 1.0 ppb.
I necessary to determine the presence of
various substances which usually are found
All methods of water analyses in this book
contain the calculations required to obtain
in extremely small amounts. It is for this results in parts per million (ppm), this being
reason that the results of a water analysis the method accepted by the American Chemi-
usually are expressed in parts per million cal Society, the American Society of Mechani-
(ppm) instead of percentage. One part per cal Engineers, and the American Society for
million equals one ten thousandth of one Testing Materials. Test procedures and cal-
percent (or 0.0001%). Thus, the expression culations of results are based on the milliliter
of results in percentage would require the use (ml) rather than the more common cubic
of cumbersome figures. centimeter ( cc) . The defined distinction be-
One part per million means one part in a tween the two terms is so slight that it has
million parts; for example, one ounce in a no practical significance, but technically the
million ounces of water, or one pound in a expression milliliter is to be preferred in this
million pounds of water. It makes no differ- work. By definition, a milliliter is the volume
ence what units are used as long as the occupied by one gram of water at four de-
relationship between the substance reported grees centigrade in vacuo, whereas a cubic
and the water is the same. As a further centimeter is the volume enclosed within a
example, one part per million does not equal cube one centimeter on each side. ( 1 ml =
one pound in a million gallons of water since 1.000028 cc)
the units are not similar.
Just as the use of percentage is avoided in EQUIVALENTS PER MILLION
favor of parts per million when reporting the
results of a water analysis, there are circum- Another unit sometimes employed in ionically
stances which may make the use of parts per reporting water analyses is equivalents per
million somewhat cumbersome. When ele- million ( epm) . This method is closely allied
ments are present in minute or trace quanti- to the use of parts per million and consists in
ties, the use of parts per million results in reducing all constituents to a common de-
decimal values. It is therefore more con- nominator.
venient to use parts per billion (ppb) in these The use of equivalents per million is not
cases. One part per billion is equal to one recommended for normal plant control. Parts
thousandth of one part per million (0.001 per million is a more simple form of expres-
ppm). For example, in conducting studies of sion of results and is accepted as the com-
steam purity using a flame spectrophotometer mon standard basis of reporting a water
to measure the sodium content, values as low analysis. However, whenever extensive calcu-
as 0.001 ppm are not uncommon-a result lations must be performed the use of equiva-
of this type is more conveniently reported as lents per million greatly simplifies the mathe-
336 BETZ HANDBOOK
matics since all constituents are on a either in ppm or epm. For example,
"chemical equivalent weight" basis. The re- total hardness, acidity, and alkalinity.
mainder of this section provides a detailed
(c) The concentration of undissolved or
discussion of parts per million and equiva-
lents per million for those who desire a suspended solids should be reported
working knowledge of these means of ex- in ppm only.
pression for purposes of calculations. (d) The concentration of organic matter
A part per million (ppm) is a measure of should be reported in ppm only.
proportion by weight and is equal to a unit
weight per million unit weights of solution. (e) The concentration of dissolved solids
Since many different constituents are deter- (by evaporation) should be expressed
mined · in a water analysis, some of these as ppm only.
constituents are reported on a common unit (f) Total dissolved solids by calculation
weight basis. For example, calcium is usually may be expressed both in ppm and
expressed in terms of calcium carbonate epm.
( CaC0 3 ) ; in other words, as the unit weight
of another substance-calcium carbonate in (g) Concentration of individual gases dis-
the example given. When constituents are on solved in water should be reported in
the same unit weight basis, they can be ppm. The total concentration of each
directly added or subtracted. For example, gas when combined in water may be
ppm total hardness as CaCOa minus ppm calculated to its respective ionic con-
calcium as CaC03 equals ppm magnesium centration in either ppm or epm.
as CaC0 3 • In every case, it is necessary to
define the unit weight basis of the results
CALCULATION OF DISSOLVED SOLIDS BY EPM
such as "ppm alkalinity as CaC0 3 ", "ppm
sulfate as S04 " or "ppm silica as Si02 ".
Starting with a reasonably complete water
Where the unit weight basis is different,
analysis, dissolved solids may be calculated
calculations must be based on the use of
by means of epm. In a complete watet an-
chemical equations or equivalents per million.
alysis, the negative epm will equal the posi-
An equivalent per million ( epm) is a unit tive epm. Where there is an excess of nega-
chemical equivalent weight per million unit tive epm, the remaining positive epm usually
weights of solution. Concentration in epm is is assumed to be sodium or potassium or
calculated by dividing concentration in ppm both, and for the sake of convenience gener-
by the chemical equivalent weight of the ally is considered to be sodium. Where there
substance or ion. This unit also has been is an excess of positive epm, the remaining
called milliequivalents per liter and milligram negative epm usually is assumed to be nitrate
equivalents per kilogram. provided the analysis is rather complete.
The following general rules illustrate where To calculate dissolved solids, convert the
epm can be used and where ppm must be various constituents from ppm to epm and
used. total the various cations (plus) and anions
(minus) . The cations should equal the an-
(a) The concentration of all dissolved salts ions. If not, add either sodium (plus) or
may be expressed either in ppm or nitrate (minus) to balance both columns.
epm of the individually determined Convert to ppm as the individual ionic
ions. weights and then total to obtain ppm dis-
solved solids.
(b) Two or more ions of similar properties In the example shown in Figure 43-1, to
whose joint effect is measured by a convert 150 ppm calcium as CaC03 to epm,
single determination may be reported divide by 50 (the equivalent weight of cal·
Figure 43-1
cium carbonate) and obtain 3.0 epm. To
convert 96 ppm sulfate as so4 to epm, divide epm
<+> (-)
by 48 (the equivalent weight of sulfate) and ppm cations anions Ionic ppm
obtain 2.0 epm. After balancing the cations Calcium as CaC03 150 = 3.0 60 as Ca
and anions by the addition of sodium, con- Magnesium as CaC03 50= 1.0 12 as Mg
Sulfate as S04 96 = 2.0 = 96 as so4
vert to ionic ppm by multiplying the epm Chloride as Cl ' 18= 0.5 = 18 as Cl
by the particular ionic equivalent weight. Bicarbonate as CaC03 120 = 2.4 = 146 as HC03
For example, to convert 3.0 epm calcium to Sodium as Na
ppm calcium as Ca, multiply by 20 (the (difference) 0.9 = 21 as Na
equivalent weight of calcium) and obtain 60 Total Dissolved Solids 353
ppm calcium as Ca. To obtain the ppm dis-
4.9 4.9
solved solids, total the individual ions.
Figure 43-2- Conversion Table
Equiv- Equiv-
alent alent
Formula Weight Formula Weight
POSITIVE IONS Calcium Chloride ............ CaC1 2 55.5

Aluminum .................. At+++ 9.0 Calcium Hydroxide .......... Ca(OH)2 37.1
Ammonium ................. NH4+ 18.0 Calcium Oxide ............. CoO 28.0
Calcium .................... Co++ 20.0 Calcium Sulfate (anhydrous) .. CaS04 68.1
Copper .................... Cu++ 31.8 Calcium Sulfate (gypsum) ..... CaS04.2H20 86.1
Hydrogen .................. H+ 1.0 Calcium Phosphate .......... Ca3(P04l2 51.7
Ferrous Iron ................ Fe++ 27.9 Carbon Dioxide ............. C02 22.0
Ferric Iron ................. Fe+++ 18.6 Chlorine ................... Cl2 35.5
Magnesium ................. Mg++ 12.2 ferrous Sulfate (anhydrous) ... FeS04 76.0
Manganese ................. Mn++ 27.5 Ferric Sulfate .............. Fe2(S04l3 66.7
Potassium .................. K+ 39.1 Magnesium Oxide ........... MgO 20.2
Sodium .................... Na+ 23.0 Magnesium Bicarbonate ...... Mg(HC03)2 73.2
Magnesium Carbonate ....... MgC03 42.2
NEGATIVE IONS Magnesium Chloride ........ MgCI 2 47.6
Bicarbonate ................ HC03- 61.0 Magnesium Hydroxide ....... Mg(OHl2 29.2
Carbonate ................. C03-- 30.0 Magnesium Phosphate ........ Mg3(P04l2 43.8
Chloride ................... CI- 35.5 Magnesium Sulfate (anhydrous) . MgS04 60.2
Fiuoride ................... F- 19.0 Magnesium Sulfate
lodide ..................... !- 126.9 (Epsom Salts) ............. MgS04.7H20 123.3
Hydroxide .................. OH- 17.0 Manganese Hydroxide ....... Mn(OHl2 44.4
Nitrate .................... N03- 62.0 Silica ..................... Si02 30.0
Phosphate (tribasic) .......... P04--- 31.7 Sodium Bicarbonate ......... NaHC03 84.0
Phosphate (dibasic) .......... HP04-- 48.0 Sodium Carbonate .......... Na2C03 53.0
Phosphate (monobasic) ....... H2P04- 97.0 Sodium Chloride ............. NaCI 58.5
Sulfate .................... S04-- 48.0 Sodium Hydroxide ........... NaOH 40.0
Bisulfate ................... HS04- 97.1 Sodium Nitrate ............. NaN03 85.0
Sulfite ..................... S03-- 40.0 Trisodium Phosphate ......... Na3P04.12H20 126.7
Bisulfite .................... HS03- 81.1 Trisodium Phosphate
Sulfide ..................... s-- 16.0 (anhydrous) .............. Na3P04 54.7
Disodium Phosphate ......... Na2HP04.12H20 119.4
COMPOUNDS Disodium Phosphate
Alum ...................... AI2(S04)3.18H20 111.0 (anhydrous) .............. Na2HP04 47.3
Aluminum Sulfate (anhydrous) .AI2(S04l3 57.0 Monosodium Phosphate ...... NaH2P04.H20 46.0
Aluminum Hydroxide ........ AI(OH)3 26.0 Monosodium Phosphate
Aluminum Oxide ............ AI203 17.0 (anhydrous) .............. NoH2P04 40.0
Ammonia .................. NH3 17.0 Sulfuric Acid ............... H2S04 49.0
Sodium Aluminate ........... Na2AI204 27.3 Sodium Metaphosphate ...... NaP03 34.0
Calcium Bicarbonate ......... Ca(HC03l2 81.1 Sodium Sulfate ............. Na2S04 71.0
Calcium Carbonate .......... CaC03 50.0 Sodium Sulfite .............• Na2S03 63.0
338 BETZ HANDBOOK
44
Acid (Free Mineral)
T is unusual to encounter a raw water de- color change of methyl orange indicator ( ap-
I void of alkalinity and containing free
mineral acid. On some streams subject to
proximately pH 4.3) is taken as the endpoint,
representing a definite point to which the
acid mine drainage or pickling acid contam- acidity of the sample has been reduced by
i!lation, free mineral acid will be found. Free the addition of the standard sodium carbon-
mineral acidity is due to the presence of acids ate solution.
such as sulfuric, nitric and hydrochloric and APPARATUS REQUIRED.
does not include carbonic acid formed by the 1-Burette, automatic, 25 ml
combination of carbon dioxide and water. 1-Casserole, porcelain, 210 ml
Naturally, free mineral acid renders a water 1-Cylinder, graduated, 50 ml
quite corrosive. Elimination of the corrosive 1-Stirring rod, glass
effect of this acidity is secured by neutraliza-
tion with an alkaline agent, usually lime, CHEMICALS REQUIRED.
caustic soda or soda ash. Sodium Carbonate, Nf50
Methyl Orange Indicator
Leaks in heating coils in pickling tanks
may result in the contamination of boiler PROCEDURE FOR TEsT. Measure a clear 50
feedwater systems with acidic condensate. ml sample of water in the graduated cylinder
Where the possibility of such contamination and transfer to the casserole. Add 4 drops of
exists, alarm systems should be installed on methyl orange indicator. Titrate from the
the returned condensate. burette with standard N/50 sodium carbonate
until the color changes from pink to yellow.
The effluent from a hydrogen zeolite ex- Record the ml of sodium carbonate required.
change unit will contain free mineral acid in
CALCULATION OF RESULTS.
proportion to the anions present in the raw
Formula:
water. This acidity is neutralized by blending
either with a sodium zeolite effluent or with ppm free mineral acid as CaC01 =
raw water. Alkalies, such as caustic soda, 1000
ml N/50 sodium carbonate X - - - -
may also be used for neutralization. Deter- ml sample
mination of the free mineral acid content of
the hydrogen zeolite effluent serves to control Using a 50 ml sample, the free mineral
the operating run of these units as well as to acidity in parts per million as CaC03 is equal
indicate when regeneration is required. to the ml of N /50 sodium carbonate required
multiplied by 20.
LIMITATIONS OF TEsT. The procedure out-
FREE MINERAL ACID
lined is satisfactory for control purposes.
Heavy metals can interfere. For most ac-
THEORY OF TEsT. This test is based on the curate results it is advisable to eliminate the
determination of the acidity of a sample by color indicator (methyl orange) and to titrate
titration with a standard sodium carbonate to a definite pH value using electrometric
solution. In this measurement the point of pH measurement.
45
Alkalinity,
HE alkalinity determined by titration with For drinking purposes the U. S. Public Health
T a standard acid may be present due to a
large number of different substances. The
Service Drinking Water Standards limit the
alkalinity for chemically treated waters only,
phenolphthalein and methyl orange alka- and in accordance with a scale based on
linity determinations do not provide a direct both the total alkalinity and the pH value
measure of any one specific ion present in the of the treated water.
water such as is the case with the sulfate and Alkalinity may be an undesirable constitu-
chloride tests. Usually several ions present ent of a water supply used in industrial pro-
contribute to the alkalinity. For the sake of cess work. High methyl orange alkalinitymay
simplicity, it is usual to consider alkalinity influence the flavor of carbonated beverages
as due to the presence of bicarbonate, car- and is to be avoided in this industry. Also,
bonate and hydrate ions although other ions in ice manufacture, waters containing a high
such as phosphate and silicate may partially methyl orange alkalinity from the presence
contribute to the alkalinity. of calcium and magnesium bicarbonate are
The points of change in color of phenol- undesirable since cloudiness is imparted to
phthalein and methyl orange indicators ( ap- the ice and deposits and scum may also be
proximately pH 8.3 and pH 4.3) simply pro- formed.
vide standard reference points which can be In boiler feedwater conditioning, the pres-
employed universally to conveniently express ence of a high methyl orange alkalinity in
the acid neutralizing power of a water. the boiler feedwater is to be avoided for
In natural surface waters, phenolphthalein several reasons, one of which is the resultant
alkalinity usually is absent and the methyl carbon dioxide content of the steam. Under
orange alkalinity is relatively low. Well waters the influence of heat in the boiler, bicar-
usually contain a higher methyl orange alka- bonate will break down producing carbonate
linity than surface waters, but again phenol- and liberating free carbon dioxide with the
phthalein alkalinity ordinarily is absent. The steam. The carbonate formed undergoes fur-
alkalinity of a natural water normally con- ther decomposition, producing hydroxide and
sists of calcium and magnesium bicarbonate liberating additional carbon dioxide with the
although, on occasion, some sodium bicar- steam. Carbon dioxide usually is responsible
bonate may be present. The origin of the for the corrosion of the steam and return
calcium and magnesium bicarbonates lies in lines. Since the quantity of carbon dioxide
the solvent action of carbon dioxide present evolved with the steam is directly propor-
in rain or surface water reacting with the tional to the feedwater alkalinity, it is desir-
minerals present in the earth such as calcite able from this standpoint to obtain feed-
and dolomite to form calcium and magne- water of as low an alkalinity as possible.
sium bicarbonates. The alkalinity of a boiler water should be
Carbonate and particularly hydrate ions sufficiently high to protect the boiler metal
are rarely encountered in untreated waters against acidic corrosion, but should also be
although they may be introduced where water sufficiently low so as not to produce a carry-
is softened by lime or by lime and soda ash. over condition. While subject to variation
340 BETZ HANDBOOK
dependent on individual plant requirements, boiler water and the cost of feedwater con-
for relatively low pressure boiler operation a ditioning are increased.
minimum alkalinity of 300 ppm ordinarily In circulating cooling water systems, alka-
is specified. An upper limit for boiler water linity is of major importance since the total
alkalinity, in order to avoid entrainment of alkalinity of a water is one of the factors
boiler water solids with the steam, is not so that must be considered in predicting the
easily specified since many factors besides tendency for a water to precipitate calcium
alkalinity are involved. In order to establish carbonate scale. The calculation of the Sat-
the maximum safe alkalinity limits, it is neces- uration Index by Langelier's Equation re-
sary that condensed steam samples be ana- quires the use of total alkalinity as one of
lyzed frequently for purity as boiler water the factors. With all others factors equal, the
alkalinity concentrations are gradually in- higher the alkalinity the greater will be the
creased. When contamination of the steam tendency for the precipitation of calcium
becomes evident from analysis, the maximum carbonate scale. Where acid treatment of the
permissible boiler water alkalinity has been circulating water is employed, alkalinity is
established. With the development of modern an important control test for governing the
anti-foam agents, it now is possible to toler- rate of acid feed.
ate quite high boiler water alkalinities with- The lime-soda softening process can effect
out carryover. reduction in the methyl orange alkalinity of
For control purposes a range of boiler a water although in the softening process an
water alkalinity is specified which will pre- increase in the phenolphthalein alkalinity will
vent both acidic corrosion and carryover of result. This reduction in total alkalinity by
solids with the steam and which also will the lime-soda process is a definite advantage
provide the proper environment in the boiler not possessed by sodium zeolite softening
water for the precipitation of the scale which neither increases nor decreases the
forming salts. For example, with phosphate alkalinity of the water being softened. The
treatment certain minimum alkalinity con- use of hydrogen zeolite softening completely
centrations must be present in the boiler to removes all alkalinity from the water, this
insure precipitation of calcium as tricalcium property being one of the outstanding fea-
phosphate. If insufficient alkalinity is pres- tures of this process. By the use of a hydrogen
ent, a soluble form of calcium phosphate is zeolite in parallel with the sodium zeolite,
produced which will permit the simultaneous the disadvantage of the sodium zeolite can
presence of both soluble hardness and soluble be overcome and the alkalinity of the mixed
phosphate. Under such conditions, the proper effluent reduced to any desired value. De-
protection against scale will not be achieved. alkalization by chloride-anion exchange is an-
other method of reducing alkalinity.
High boiler water alkalinities also are un-
desirable from the standpoint of intercrystal-
ALKALINITY, DIRECT TITRATION METHOD
line cracking (caustic metal embrittlement).
It has been established that the presence of THEORY OF TEST. This test is based on the
high hydroxyl ion concentrations is the pri- determination of the alkaline content of a
mary cause of intercrystalline cracking. Even sample by titration with a standard acid
where inhibitors are employed to overcome solution. In this measurement, the endpoints
the ernbrittling characteristics of a boiler are taken as points of change in the color of
water, it is desirable that excessive alkalinities organic indicators; phenolphthalein ( approxi-
be avoided since such inhibitors are fed in mately pH 8.3) and methyl orange ( approxi-
proportion to the alkalinity content of the mately pH 4.3) represent definite points to
boiler water. If boiler water alkalinity is which the alkalinity of the sample has been
high, a greater quantity of the inhibitor must reduced by the addition of the standard acid
be fed and both the solids content of the sqlution.
APPARATUS REQUIRED, If the water sample is colored, such as

1-Burette, automatic, 25 ml one containing chromate, methyl purple in-
1-Casserole, porcelain, 210 ml dicator may be substituted for methyl orange
1-Cylinder, graduated, 50 ml indicator to provide a more definite end-
1-Stirring rod, glass point. The color change with methyl purple
CHEMICALS REQUIRED. is from green to gray to purple. The end-
Sulfuric Acid, N I 50 point is taken as the first change to a definite
Phenolphthalein Indicator purple. This sample cannot be used for the
Methyl Orange Indicator chloride test. A second sample must be used
Methyl Purple Indicator for chloride and neutralized to the P reading
PROCEDURE FOR TEST. Measure a clear 50 as described under Chloride, Mohr Method
ml sample of water in the graduate and CALCULATION OF RESULTS.
transfer to the casserole. Add 4 to 5 drops Formula:
of phenolphthalein indicator. If the sample ppm alkalinity as CaC0 3 =
is an alkaline water, such as usually is the 1000
case with boiler water, it will turn red. If the ml Nl50 sulfuric acid X - - - - - -
ml sample
sample is a raw or natural water, it usually
Using a 50 ml sample, the phenolphthalein
will remain colorless. Add the standard N I 50
alkalinity in parts per million as CaC03 is
sulfuric acid from the burette drop by drop
equal to the ml of N I 50 sulfuric acid re-
to the sample in the casserole, stirring con-
quired for the P reading multiplied by 20.
stantly until the point is reached where one
The methyl orange alkalinity is equal to the
drop removes the last trace of red color and
ml of N I 50 sulfuric acid required for the M
the sample becomes colorless. Stop and re-
reading multiplied by 20.
cord the total number of ml to this point as
the P reading. LIMITATIONS OF TEST. It is preferable to ex-
press the results of the alkalinity determina-
Add 4 drops of methyl orange indicator (if tion in terms of P and M alkalinity as above.
no red color develops on the addition of the However, results are sometimes calculated in
phenolphthalein indicator to the original terms of bicarbonate, carbonate and hydrate
sample, the titration may be started with the on the assumption that titration to the P
methyl orange indicator at this point) . Con- endpoint is equivalent to all the hydrate and
tinue adding the acid drop by drop until one one-half the carbonate alkalinity, and that
drop changes the color from a yellow to a the titration to M is equivalent to the total
salmon-pink. Record the final burette read- alkalinity. Many factors such as the presence
ing as the M reading. This is a more difficult of phosphate, silica, organic and other buffers
endpoint and some practice may be required. affect this titration and the calculation of
The general tendency is to add too much the form of alkalinity present may be in
acid. If too much acid is added, the sample error. Under normal circumstances in plant
will change from a salmon-pink to a definite control, expression of results as P and M
red. Record the titration to the P point and alkalinity is entirely satisfactory and is to
the total titration to the M point as the P be preferred from the standpoint of sim-
and M readings respectively. (Note that the plicity.
M reading will always be greater than the P If an electrometric pH machine is avail-
reading inasmuch as the P reading is in- able, the titration may be made without in-
cluded in the M reading.) dicators to pH 8.3 and pH 4.3. Particularly
If the water sample does not settle clear, where the color of the water sample obscures
the M reading can be determined on a fil- the indicator endpoint; the electrometric titra-
tered sample. The P reading, however, al- tion method-provides greater accuracy. While
ways should be determined on an unfiltered most authorities on water analyses agree that
sample. phenolphthalein changes color at almost ex-
342 BETZ HANDBOOK
actly pH 8.3, a range is indicated for methyl color change of methyl orange to be at pH
orange of pH 3.9 to 4.6. The procedure for 4.3 so that a consistent reference point is
alkalinity as outlined above presumes the provided for control purposes.
Figure 45-1 • Test Set for Hardness, Alkalinity, Chloride, Sulfite, and Phosphate
46
Alkalinity (Hydrate)
YDRATE alkalinity is that portion of the HYDRATE ALKALINITY, BARIUM CHLORIDE
H total alkalinity due to the presence of
hydroxyl ions. Hydrate alkalinity is not found
METHOD
in natural waters unless highly contaminated THEORY OF TEsT. This test is based on the
by alkaline trade wastes. The hydrate ion determination of the hydrate alkalinity of a
may be introduced to treated waters by the sample by titration with a standard acid
lime-soda softening process or by the intro- solution. Barium chloride is added to remove
duction of alkaline treatment chemicals such the carbonate ion from solution thereby per-
as caustic soda or lime. Even where lime or mitting direct titration of the hydrate alka-
caustic soda may be added to a water to in- linity. In this measurement, the endpoint is
crease its pH value and overcome its possible taken as the point of change of color of
corrosive character, it is rare that sufficient phenolphthalein indicator (approximately pH
treatment is added to produce a definite hy- 8.3) representing a definite point to which
drate alkalinity. the alkalinity of the sample has been re-
Under ordinary circumstances the control duced by the addition of a standard acid
of alkalinity on the basis of phenolphthalein solution.
and methyl orange tests is sufficient for con- APPARATUS REQUIRED.
trol purposes. Direct determination of the 1-Burette, automatic, 25 ml
hydrate alkalinity is of importance in rela- 1-Casserole, porcelain, 210 ml
tively few instances. However, where inter- 1-Cylinder, graduated, 50 ml
fering substances are present which affect the 1-Stirring rod, glass
methyl orange test, a direct determination of 1-Pipette, delivery, 5 ml
the hydrate ion may be of considerable value
in controlling lime-soda softening. A definite CHEMICALS REQUIRED.
hydrate alkalinity usually is desired on a Hydrochloric Acid, N /50
lime-soda softener effluent for most efficient Barium Chloride, 10%
hardness removal. While this usually can be Phenolphthalein Indicator
accomplished by control of phenolphthalein PROCEDURE FOR TEsT. Measure 50 ml of the
and methyl orange alkalinity, where organic sample in the graduated cylinder and trans-
contamination is present direct measurement fer to the casserole. By means of the pipette
of the hydrate alkalinity is to be preferred add 5 ml barium chloride solution and stir
for best control. vigorously. Add 4 to 5 drops phenolphthalein
It occasionally is desirable to determine indicator and stir. If the sample does not
the hydrate alkalinity of a boiler water by turn red, hydrate alkalinity is not present,
direct measurement. This is particularly true the test is discontinued at this point and the
in high pressure operation where hydrate alka- hydrate alkalinity is recorded as "zero". If
linity concentrations may contribute to cor- the sample turns red, add N /50 hydrochloric
rosion and hydrogen evolution or where co- acid continuously from the burette, drop by
ordinated pH-phosphate control for embrittle- drop with constant stirring, until one final
ment is maintained. drop changes the color from red to colorless.
344 BETZ HANDBOOK
Disregard any reappearance of the red color. ml of N I 50 hydrochloric acid required multi-
Record the ml of N I 50 hydrochloric acid re- plied by 20.
quired. LIMITATIONS OF TEsT. This method for hy-
drate alkalinity is considered accurate only to
CALCULATION OF RESULTS. within 10%. High silica, organic matter and
Formula: aluminates can interfere with the determina-
ppm hydrate alkalinity as CaC03 = tion. Should the sulfate content of the sample
1000 be high, the turbidity produced will tend to
ml Nl50 hydrochloric acid X - - - - mask the endpoint. This method, however,
ml sample
affords a rapid means for checking hydrate
Using a 50 ml sample the hydrate alkalinity alkalinity for routine control and can be of
in parts per million as CaC0 3 is equal to the value in the control of lime-soda softening.
Figure 46-1 • Test Set for Alkalinity, Chloride, Sulfite and Phosphate
47
Alllrnonia
ITROGEN may be reported in a water AMMONIA NITROGEN,
N analysis in four different forms. These
are albumenoid ammonia, free ammonia,
DIRECT NESSLERIZATION METHOD
nitrite and nitrate. THEORY OF TEST. This test is based on the

development of a color varying from yellow
Free ammonia is determined more fre-
through amber to red on the addition of
quently than albumenoid ammonia. This lat-
Nessler Reagent to water containing the am-
ter determination is of significance primarily
monium ion. The color developed is com-
in the sanitary analysis of water. Direct Nes-
pared with a color standard slide and the
slerization of steam condensate will determine
ammonia content expressed as N in parts per
the free ammonia content.
million is read directly from the slide.
The presence of free ammonia in the steam
condensate is caused by the breakdown of or- APPARATUS REQUIRED.
ganic nitrogenous compounds under boiler 1-Taylor Water Analyzer complete (ni-
conditions. These compounds enter the boiler trogen slide range 0.0-1.0 ppm)
system as organic impurities contaminating 1-Cylinder, graduated, 50 ml
the makeup water. 1-Pipette, Mohr, 1 ml
1-Flask, Erlenmeyer, 250 ml
Ammonia nitrogen is of importance in con-
nection with the determination of the specific CHEMICALS REQUIRED.
conductance of steam condensate. Ammonia Nessler Reagent
possesses a high specific conductance and PROCEDURE FOR TEsT. The sample should be
where this gas is present in steam condensate, freshly obtained with as little contact with
conductance measurements must be corrected air as possible. Measure carefully 50 ml of
for the conductance due to ammonia. Other- the sample with the graduate and add to the
wise, erroneously high solids content of the flask. Add 1 ml of Nessler Reagent. Allow
steam would be indicated. Where effective the sample to stand for 10 minutes. Fill one
degassing equipment is employed for the re- of the comparator tubes to the mark with
moval of ammonia and carbon dioxide prior prepared sample and place it in the middle
to the measurement of steam purity, ammonia compartment of the Analyzer. Fill the other
may be reduced to such a low value in the two tubes to the mark with the original un-
test sample as to require no correction for its treated sample and place them in the outer
conductance effect. compartments of the Analyzer.
The determination of ammonia may also Place the instrument so that the com-
be of significance where corrosion of copper parator tubes point toward a source of day-
or copper bearing alloys is taking place. Am- light. Read the instrument by viewing the
monia will attack copper, placing it in solu- reflection in the mirror and not by looking
tion as the copper ammonium ion. Removal down through the tubes. Move the slide in
of ammonia from a boiler feedwater system front of the test samples until a color match
may be accomplished by the use of break- is obtained.
point chlorination or cation exchange by hy- If the ammonia content of the sample ex-
drogen zeolite. ceeds the range of the slide, discard test,
346 BETZ HANDBOOK
take a smaller sample such as 25 ml or 10 iron, turbidity and sulfides. Special treatment
ml, dilute to 50 ml with ammonia free dis- can be employed for the removal of these
tilled water, repeat the test and multiply the interfering substances.
reading by the appropriate factor. Where the ammonia content is low and
CALCULATION OF RESULTS. Using a 50 ml greater accuracy is desired than can be ob-
sample, ammonia as N is read directly from tained from the permanent slide standards,
the color standard slide in parts per million. it is necessary to prepare standards for the
desired range using standard ammonium
LIMITATIONS OF TEST. The direct Nessleriza- chloride in distilled water, employing a Nes-
tion of the ammonium ion is applicable to sler rack and tubes. For most accurate results
waters of low mineral content such as con- in the determination of the ammonia the dis-
densed steam and condensate samples. Re- tillation method, which is a laboratory pro-
sults are affected by calcium, magnesium, cedure, should be employed.
Figure 47-1 • Taylor Water Analyzer

48
Calcium
HE determination of calcium in a water Saturation Index which predicts the tendency
T analysis is closely allied to the determina-
tion of hardness, since hardness is caused pri-
of a water to form calcium carbonate scale-
the scale most easily formed in cooling water
marily by the presence of calcium and mag- systems. In employing the Langelier Index,
nesium. the determination of calcium is required. On
Under certain circumstances determination this basis, a plant control can be established.
of calcium is made so as to subdivide the Most softening processes are primarily con-
total hardness into calcium hardness and cerned with the removal of total hardness
magnesium hardness. The total hardness (calcium and magnesium hardness) . How-
minus the calcium hardness equals the magne- ever, under some circumstances the calcium
sium hardness. hardness only is removed. The magnesium
The determination of calcium usually is hardness is allowed to remain in solution. Cal-
not made as a control test. However, this cium softening is most applicable where the
determination is necessary for making chem- water is to be employed in heat exchange
ical calculations for treatment chemicals. units. Since calcium hardness is the prime
When employing internal boiler water treat- offender in such cases, it may not be neces-
ment of the phosphate type, the calcium sary to remove the magnesium hardness from
hardness is precipitated by the phosphate the water and a savings in treatment chem-
whereas the magnesium hardness is precipi- icals is obtained by this partial softening for
tated primarily as magnesium hydroxide the removal of calcium hardness only.
either by the natural alkalinity of the boiler
water or by the use of alkaline materials such
as caustic soda or soda ash. This same prin- CALCIUM, TITRATION METHOD
ciple applies to the hot phosphate softener.
Therefore, in calculating the quantity of ma- THEORY oF TEST. This test is based on the
terials required, it is necessary that the hard- determination of the calcium content of a
ness be subdivided into calcium hardness and sample by titration with a sequestering agent
magnesium hardness. in the presence of an organic dye sensitive to
In lime-soda softening, it is necessary to calcium ions and insensitive to magnesium
know the calcium hardness in order to calcu- ions under the conditions of test. The end-
late the initial quantities of lime and soda point occurs when all the calcium ions are
ash required. However, once the softener is sequestered and the endpoint is a color
in operation changes in chemical treatment change from salmon-pink to orchid-purple.
are usually controlled by analyzing the efflu- APPARATUS REQUIRED.
ent for total hardness, phenolphthalein and 1-Burette, automatic, 25 ml
methyl orange alkalinity. 1-Casserole, porcelain, 210 ml
In the treatment of cooling water the de- 1-Cylinder, graduated, 50 ml
termination of calcium as a control test plays 1-Stirring rod, glass
a more important part. An important basis 1-Pipette, safety bulb, 2 ml
for all cooling water treatment is Langelier's 1-Measuring cup, brass
348 BETZ HANDBOOK
CHEMICALS REQUIRED. further color change occurs.

Calcium Indicator CALCULATION OF RESULTS.
Hardness Titrating Solution, 1 ml = 1 mg Formula:
CaC0 3 ppm calcium as CaC0 3 =
Sodium Hydroxide, 1.0 N 1000
PRoCEDURE FOR TEsT. Measure 50 ml of ml titrating solution X - - - -
ml sample
sample and transfer to casserole. Add 2 ml
of 1.0 N sodium hydroxide to sample and Using a 50 ml sample, the calcium in parts
stir. With measuring cup provided, add one per million as CaC03 is equal to the ml of
level measure (approx. 0.2 g) of calcium indi- titrating solution employed multiplied by 20.
cator and stir. If calcium is present, the sample The magnesium hardness can be deter-
will turn salmon-pink. Add the hardness mined by subtracting the calcium hardness
titrating solution (same titrating solution from total hardness.
as used in hardness test) from the burette LIMITATIONS OF TEsT. This method is sub-
with ,continued stirring. When approaching stantially free from ionic interference in the
the endpoint, the sample begins to show a concentrations normally encountered in in-
purple tinge. The endpoint is a final change dustrial water conditioning. Orthophosphate
to orchid-purple. Once the endpoint is above 5 ppm will precipitate calcium at the
reached, further addition of hardness titra- pH of the test. By dilution of the sample,
ting. solution will not produce any further ionic interference can be minimized. Because
color change. Always check endpoint by of the accuracy possible with both the hard-
adding one additional drop of hardness ness and calcium methods, a reliable magne-
titrating solution and observe whether any sium value can be secured by difference.
49
Carbon Dioxide
REE carbon dioxide is the term used to tion of a feedwater heater unless the free
F designate carbon dioxide gas dissolved
in water. The designation "free" carbon di-
carbon dioxide content of the heater influent
is abnormally high. In these cases, such as
oxide is used to differentiate a solution of result from combined sodium-hydrogen zeo-
carbon dioxide gas from combined carbon di- lite softening, the use of a degasifier is re-
oxide present in the form of bicarbonate and quired to reduce the carbon dioxide content
carbonate ions. of the heater influent to a range suitable to
Free carbon dioxide in water is caused by the conventional type feedwater heater.
the decay of organic matter, solution of car- Corrosion from carbon dioxide in boiler
bon dioxide from underground sources and systems is most frequently encountered in
solution from the atmosphere. Since the steam and return lines. Even though a boiler
carbon dioxide content of the normal atmos- feedwater contains no free carbon dioxide,
phere is quite low (less than 0.04%), this is appreciable concentrations of this corrosive
not a source of appreciable carbon dioxide gas may be present in the steam from the
except where flue gases containing a high boiler. The source of this carbon dioxide is
percentage of carbon dioxide are brought in the bicarbonate and carbonate content of the
relatively close contact with the water. boiler feedwater. Under the influence of heat
Well waters usually contain an appreciable in the boiler, the bicarbonates and the car-
quantity of free carbon dioxide while sur- bonates are decomposed, liberating free car-
face waters normally are relatively low in bon dioxide with the steam. Sodium hydrox-
free carbon dioxide. ide (caustic soda) is a by-product of this
reaction. Normally, the carbon dioxide evolved
Corrosion is the principal difficulty caused from the decomposition of the bicarbonate
by carbon dioxide. This gas will ionize on and carbonate of the boiler feedwater is the
solution in water producing carbonic acid. major source of the carbon dioxide present in
Because of the low pH value and corrosive the steam, since the free carbon dioxide of
character contributed to a water by carbon the boiler feedwater is usually reduced to a
dioxide, this gas is undesirable. Where oxygen minor value by feedwater deaeration.
also is present, the low pH value due to car-
Severe corrosion of steam and return lines
bon dioxide enhances the corrosive effect of
caused by the carbon dioxide of the steam is
the oxygen.
to be expected where sodium zeolite softening
Severe corrosion of heat exchangers, pip- is used with a raw water of high methyl
ing, valves, etc., can result where appreciable orange alkalinity. For approximation, the
quantities of free carbon dioxide are present carbon dioxide content of the steam with
and have given the water a corrosive char- zeolite softening can be calculated by multi-
acter. plying the feedwater methyl orange alkalinity
In boiler systems, it is desirable to reduce by a factor of 0. 79 which takes into account
the free carbon dioxide content of the boiler 80% sodium carbonate decomposition. For
feedwater as low as possible. This usually is example, if a zeolite softened boiler feed-
accomplished by the normal deaerating ac- water possesses a methyl orange alkalinity of
350 BETZ HANDBOOK
100 ppm as calcium carbonate, a carbon di- tubing discharging at the bottom of the 100
oxide content of the steam of 79 ppm can be ml cylinder. Allow the sample to overflow
expected. This high carbon dioxide content for a few minutes and withdraw tubing while
of the steam would create an extremely cor- sample is flowing. Flick cylinder to throw
rosive condition particularly in units where off excess sample above the 100 ml mark.
accumulation of carbon dioxide can take Add 5-10 drops phenolphthalein indicator.
place, such as unit heaters, jacketed kettles, If the sample turns red, no free carbon di-
etc. oxide is present. If the sample remains color-
Aeration is the most common method em- less, titrate rapidly into the cylinder with
ployed for the removal of large quantities of N /22 sodium carbonate, stirring gently with
free carbon dioxide. Neutralization of the the stirring rod until a definite pink color
acidic nature of this gas can be accomplished develops throughout the sample, permanent
by the feeding of alkaline chemicals such as for 30 seconds. This color change is the end-
lime, soda ash and caustic soda. Lime-soda point.
softening effects removal of free carbon di- Where the free carbon dioxide of the
oxide: Free carbon dioxide may be removed water sample is high, there may be some loss
from a boiler feedwater by deaeration since of carbon dioxide to the atmosphere even
the solubility of this gas in water decreases with this titration technique. To check this
with increased temperature. The filming point it is advisable to secure a second sam-
amines and the neutralizing amines are em- ple as before and, after reducing sample to
ployed in controlling the problem of corro- the 100 ml mark, to immediately run in the
sion in steam and condensate lines. full amount of N/22 sodium carbonate used
in the first titration. Add 5-10 drops of
FREE CARBON DIOXIDE, phenolphthalein indicator and if the sample
DIRECT TITRATION METHOD remains colorless, add additional N/22 so-
dium carbonate to the endpoint, accepting
THEORY OF TEsT. This test is based on the this second test as the more accurate titra-
titration of a sample with standard sodium tion.
carbonate solution in the presence of phenol-
phthalein indicator. The carbon dioxide re- CALCULATION OF RESULTS.
acts with the sodium carbonate to form so- Formula:
dium bicarbonate, which is colorless to ppm free carbon dioxide as C02 =
phenolphthalein. As soon as an excess of 1000
ml N/22 sodium carbonate X - - - - -
sodium carbonate is present, a light red ml sample
color develops and this is taken as the end-
point. Using a 100 ml sample the free carbon
APPARATUS REQUIRED. dioxide in parts per million as C0 2 is equal
1-Burette, automatic, 25 ml to the ml of N /22 sodium carbonate re-
1-Cylinder, graduated, 100 ml quired multiplied by 10.
1-Tubing, sulfur free rubber, 10 feet LIMITATIONS OF TEsT. Free mineral acid, if
1-Stirring rod, glass present, will be measured in this test. Heavy
CHEMICALS REQUIRED. metals such as iron, chromium and aluminum
Sodium Carbonate, N /22 salts interfere. During titration some loss of
Phenolphthalein Indicator free carbon dioxide may take place.
PRoCEDURE FOR TEST. Reliable results in the Diagrams and nomographs have been pro-
determination of free carbon dioxide can be posed for the calculation of the carbon di-
secured only on a freshly obtained sample, oxide, bicarbonate, normal carbonate and hy-
secured under careful sampling conditions. droxide content of natural and treated waters.
Obtain the sample through gum rubber Knowing the pH, total alkalinity, tempera-
1000
~
900
800
1\.
1"\. ' '1\ ' ' 1\. ' 1\1'\' 1\ 1"\. '~ ~
100
600
500
' ' ' ·'
1\ 1\' 1\ 1\1\ r\
"
\
1\
1\' 1\ 1\ ~ ,,, '

'I\
1\
400 '
[\1\ 1\ 1\ r\ [\ r\ 1\1\ r\ !\ r\ ~
300 ~
'
r\ r\ 1\ 1\ ~ 1\ r\ [\ 1\1\ i\ i\ r\ [\
[\ ~ I\ I\ ~ \ I\ i\ \ \ \ \ ~ r\ 1\ ~
:LOO
1\ i\ ~ i\ 1\ i\ \ ~ \ \ \ i\ [\r\ \ \ i\ ~ .)\
1\ i\ l\ i\ \ \ \ '[\ i\ r\ l\ r\ K \ \ i\ ~ 0~
\ [\ \ 1\ 1\[\ ~
r\ ' \ \ 1\ r\ r\ 1\' r\ 1\"~ i\' \ [\ \ 1\1\1\1\ ~
\ \ 1\ 1\
1\.
1'\
'"' ~
100
90 1\. 1\ 1\ 1\
so
'' '
1\ "
1\
1\ 1\' '
1\ 1\ ' 1\' 1\
1"\.
' '
1\
1"\.
1\ ' 1\' 1\ 1\. ' 1\1\. ~
~
a.
70
60
,,
'1\ ' 1\ I\ ' 1\I\ i\' I\ I\ i\ I' I' 1\1\l\1\ ,\
' 'l' ~
" ,'
a.
I
so 1\
4o 1\ 1\ I\ 1\ 1\ [\ 1\ i\ ""[\ ~ 1\i\1\1\1\ 1\ ~~0~ '
0
('I
1\ 1\ r\ [\ [\ [\ 1\ 1\1\ f\i\1\[\1\1\;. ~ ~[\ ~I
30
I\ I\ I\ 1\ 1\1\ l\l\1'~~0fo~ i\1' \l\ ~
u
UJ 20
I\ I\
'
1\ 1\ 1\ 1\ 1\ 1\ 1\ 1\~~~~~~1\1\ 1\1\1\[\ ~~~ ~
w
' ""~
a:LL 1\ 1\ 1\ 1\ 1\ ~~ 1\i\i\[\1\[\ f\1\1\[\ ~ 1-
I\ 1\ I\ ~~\ 1\1' 1\1\1\[\1\l\ 1\l\ \[\ ..;
10
[\ [\ ~i\1\f\ 1\ [\1\ r\1\r\:\r\[\ i\1\r\r\
~ ~~ 1\.1\. 1\.IX
' X' ~ ' ' f\' ' ' ' ' 1\.1\
9
a
' "' ' ' ' \' \ ' \1\.
'
7
6
I\ \
1\
" I\' 1\ I\ I\ 1\ ' \' '\ \' \ \ \ \ \'
\ I\ I\ I\ 1\. \.
5
4
1\
\ \ ' '\ i\ I\ \ r\ \ \ \ \ ~ \ 1\ ~ \ '
3
1\ 1\ 1\ 1\ 1\ i\ r\ 1\ [\ i\ 1\ i\ [\ 1\ [\ r\
I\ I' I\ I\ I\ I\ I\ I\ I\ I\ r\ I\
2
I'
1\
I'
1\ I\ 1\ I\ 1\ 1\ 1\ 1\ 1\ [\ 1\ "
r\ I\ \ 1\ 1\ (\ i\ 1\ 1\ i\ i\ 1\ [\1\
r" i\ i\ i\ I\ l\ I\ \ l\ I\ I\ I\ l\
I
i\ r\ 1\ r\ 1\ 1\ 1\ f\ 1\ 1\ 1\ r\
4.5 5.0 5.5 6.0 6.5 7.0 7.5 a:o
pH VALUE
352 BETZ HANDBOOK
ture and total mineral content of a water any cedures for these calculations are shown m
or all of the alkalinity forms and carbon di- "Standard Methods for the Examination of
oxide can be determined. The proper pro- Water and Wastewater."
Figure 49-1 • Relation of Alkalinity and pH to Free Carbon Dioxide in Raw Waters
This graph indicates approximate relation- sample and read the free carbon dioxide con-
ships between pH, alkalinity and free carbon tent of the water at the side of the chart as
dioxide and when any two of these three determined by the horizontal intersection.
factors are known, the third can be deter- CAUTION: The chart should not be depended
mined. To determine free carbon dioxide, upon for a high degree of accuracy, but is
locate the intersection of the vertical line for close enough_ for average use. For truly ac-
the pH value, with the slanting line for the curate results the 00 2 should be actually
total alkalinity (methyl orange) of the water determined by chemical test.
50
Chloride
HE chl?rides of calcium, ~agnesium, so- The following facts should be kept in mind
T dium, Iron and other catwns normally
found present in water are extremely soluble.
when considering blowdown figures based on
the above formula:
The principal importance of the chloride test
in industrial water conditioning is based on the The feedwater chloride determination
high solubility of the chlorides. Since no pre- must be an average. A spot sample may
cipitation of the chlorides takes place, the be misleading, particularly if feedwater
chlorides present in boiler water and cooling chlorides vary to any extent.
water are proportional to the concentration
Where sodium chloride is used to re-
of the water.
generate an ion exchanger, complete
In natural waters high chloride values rinsing following regeneration must be
sometimes indicate the presence of animal employed, otherwise the boiler feedwater
pollution. However, the chloride test is chloride will be erroneously high.
merely an indication and animal pollution
should be confirmed by bacteriological and The chloride test is unsuitable for blow-
sanitary analyses, as a high chloride value down calculation only in those cases where
can be due to other reasons such as contamin- the feedwater chloride is quite low, such as
ation by oil field brines and industrial wastes. 0.5 ppm or 1.0 ppm. In this range, a slight
In addition, high chlorides can result simply analytical error in determining feedwater
because the water has passed through a chloride will cause an appreciable error in
natural salt formation in the earth. The U.S. the calculated rate of blowdown.
Public Health Service Drinking Water Stand-
ards for potable waters recommend a maxi- Another use for the chloride determina-
mum chloride content of 250 ppm as CI. tion is to estimate the percent makeup
present in a boiler feedwater. To make this
In industrial water conditioning the prin-
estimate it is necessary to determine the
cipal application of the chloride test is in the
chloride concentration of the makeup water,
control of blowdown in boiler and cooling
the condensate, and the boiler feedwater. The
systems. In addition, the chloride test is also
chloride content of pure condensate will be
suitable for the accurate calculation of the
"zero". Assuming a pure condensate, 4 ppm
rate of boiler blowdown provided the chlo-
chloride in the boiler feedwater and 10 ppm
ride concentration of the feedwater is of suf-
chloride in the makeup water, then the chlo-
ficient magnitude for accurate determination.
ride of the feedwater is 40% of the chloride
Since chlorides are not subject to precipita-
of the makeup water. Therefore, the feed-
tion, the determination of the chloride con-
water contains 40% makeup and 60% con-
centration of the feedwater and the boiler
densate.
water provides an accurate basis for calculat-
ing the rate of blowdown using the following In cases where there is a measurable chlo-
formula: ride content of the condensate due to carry-
over or condenser leakage, the percent makeup
Cl in feedwater
- - - - - - - - X 100 = "'o Slowdown can still be calculated, employing the Tea
Cl in boiler water Mixers rule or algebraic computations. How-
354 BETZ HANDBOOK
ever, in estimating the percent makeup where CHLORIDE, MOHR METHOD

the chloride content of the makeup water and
feedwater are very low the error introduced THEORY OF TEsT. This test is based on the
by the accuracy of the chloride test itself can titration of a water containing the chloride
be appreciable. ion with a standard silver nitrate solution,
In open recirculating cooling water sys- using potassium chromate as an indicator.
tems, control of blowdown usually is based on The chloride ion is precipitated by the silver
maintaining cycles of concentration within a nitrate as silver chloride. As soon as the chlo-
definite range. The chloride test is employed ride has been completely precipitated, further
in determining cycles, using the formula: additions of silver nitrate will result in a red
color due to the reaction of the silver ion
Cl in Circulating Water with the chromate ion and this change is
taken as the endpoint. In the sulfite modifi-
Cl in Makup Water cation of this test, the hydrogen peroxide
oxidizes the sulfite to sulfate which, unlike
= Cycles of Concentration
sulfite, does not interfere with the chloride
test.
The degree of concentration of the circu-
lating water, as compared with the makeup
APPARATUS REQUIRED.
water, can be determined by the relative
1-Burette, automatic, 25 ml
chloride content whether blowdown is em-
1-Casserole, porcelain, 210 ml
ployed or if natural windage losses are suf-
1-Cylinder, graduated, 50 ml
ficient to restrict cycles.
1-Bottle, with dropper, 8 oz
The chloride test can be of advantage in 1-Stirring rod, glass
the control of ion exchange softeners. After 1-Pipette, Mohr, 5 ml
the regeneration of a sodium unit, the excess
salt must be washed out of the exchanger bed
before returning the unit to softening opera- CHEMICALS REQUIRED.
tion. If the salt is not washed out entirely, a Sulfuric Acid, 1 N
sample of softener effluent will possess a high Silver Nitrate, 1 ml = 1 mg Cl
chloride content. In this case, the washing Phenolphthalein Indicator
should be continued until the chloride of the Potassium Chromate Indicator
softener effluent is reasonably close to the Hydrogen Peroxide, 3%
chloride of the raw water entering the sof-
tener. PROCEDURE FOR TEsT. If the P or M alka-
Finally, the chloride test is useful in deter- linity has already been determined, the same
mining condenser leakage, especially aboard sample may be used directly for the chloride
ships or other places where sea water is used test without further neutralization. If the P
for condensing purposes. If leakage is occur- or M alkalinity has not been determined, the
ring, the condensate will show a material in- sample must be properly neutralized with sul-
crease in the chloride content. furic acid. Take a 50 ml sample of the water
Generally speaking no attempt is made to to be tested, add 4 to 5 drops of phenolphtha-
remove the chloride ion from water in in- lein indicator. By means of a dropper add
dustrial water conditioning since chlorides just sufficient sulfuric acid to neutralize to the
are very soluble. However, for certain special colorless side of phenolphthalein as described
industrial uses such as the water employed under alkalinity. In this case, it is not neces-
in the silver plating of mirrors, the chlorides sary to accurately measure the amount of
can precipitate silver in the form of silver acid required for neutralization.
chloride. In cases such as this, the chloride Add to the neutralized sample 5 drops of
can be removed by demineralization. potassium chromate indicator which will
turn the solution a bright yellow. Add silver silver nitrate required minus a blank of 0.2
nitrate from the burette slowly to the sample, ml, multiplied by 20. When using a 50 ml
stirring constantly, until 1 drop produces a sample and silver nitrate of the strength 1
permanent reddish color. This reddish color ml = 5 mg Cl, the chloride in parts per mil-
is to be taken as the endpoint and not a brick lion as Cl is equal to the ml of silver nitrate
red color that will develop if additional silver required multiplied by 100.
nitrate is added.
A blank of 0.2 ml for this strength silver LIMITATIONS OF TEsT. This method of chlo-
nitrate solution ( 1 ml = 1 mg Cl) should be ride determination is unaffected by sulfate,
subtracted as this is the quantity of silver total alkalinity, phosphate, silica, iron and
nitrate required to produce the endpoint with color in the concentrations normally encoun-
50 ml distilled water. tered in industrial waters. Sulfite interference
Where brackish waters are encountered, a can be eliminated as outlined. In the de-
stronger solution of silver nitrate should be termination of chloride in the low concentra-
employed ( 1 ml = 5 mg Cl) . No blank is tions existing in many natural and boiler
necessary with this stronger solution. feedwaters, however, an error in titration
with silver nitrate of 0.1 or 0.2 ml may be
MoDIFICATION WHEN SuLFITE IS PRESENT.
an appreciable percentage error. This error
When the sulfite content of the water tested
is significant if the percentage boiler blow-
exceeds 10 ppm, the sulfite should be oxidized
down is calculated on the relative chloride
to sulfate by hydrogen peroxide ( 3%) to
content of feed and boiler waters. The chlo-
avoid interference with the chloride test.
ride test is the one used most frequently in
After neutralization to the colorless side of
calculating and controlling cycles of concen-
phenolphthalein, add 2.0 ml of hydrogen
tration in recirculating cooling systems. If the
peroxide (3%) and stir well. Add potassium
chloride content of the makeup water is low
chromate indicator and titrate as outlined
and the chloride endpoint is overrun, the
above.
cycles of concentration calculated will be er-
This procedure should not be- used on a roneously low. The actual cycles may be con-
sample already neutralized to methyl orange. siderably greater than calculated and thus a
If M alkalinity has been determined, discard scale-forming condition can develop even
the sample and neutralize a fresh sample to though cycles apparently are properly re-
the phenolphthalein endpoint only, before stricted. The chloride concentration of the
adding the hydrogen peroxide. makeup water must be determined accurately
CALCULATION OF RESULTS. to avoid this error.
Formula: For greater accuracy in low chloride con-
ppm chloride as Cl = (ml silver nitrate - ml centrations, a microburette for silver nitrate
1000 is recommended. Accurate determination of
blank) X strength AgN03 mg/ml X - - - - chloride above 1000 ppm is not feasible with
ml sample this method except by dilution of the sample
Using a 50 ml sample and silver nitrate of with distilled water so as to lower the chlo-
the strength 1 ml = 1 mg Cl, the chloride in ride concentration below 1000 ppm in the
parts per million as Cl is equal to the ml of sample to be tested.
356 BETZ HANDBOOK
51
Chlorine
HE test for chlorine is a test for the hy- mild bactericide and oxidizing agent.
T drolysis products of chlorine gas in
water. The chlorine test should not be con-
Because of the significance of the form in
which chlorine residual is present, it is be-
fused with the test for chloride which is a coming increasingly important to determine
negative ion. Like other gases, such as carbon both the free and combined available chlo-
dioxide and ammonia, chlorine gas possesses rine residuals. For example, at a pH of 6.0-
a definite solubility in water forming hypo- 8.0, a free available chlorine residual of 0.2
chlorous acid and the hypochlorite ion. It is ppm after ten minutes contact will provide
in the form of hypochlorous acid and the excellent bacterial reduction while for the
hypochlorite ion that chlorine exerts its disin- same results a combined available chlorine
fecting and oxidizing properties. residual of 1.0 ppm after sixty minutes con-
Chlorine is not present in natural waters tact may be required.
and is found only as a result of chlorination
of a water supply. Chlorine gas or hypochlo- CHLORINE, TAYLOR·ENSLOW METHOD
rites are added to potable water supplies for
the destruction of harmful bacteria and to THEORY OF TEsT. The determination of resi-
render the supply safe for human consump- dual chlorine in water samples by colori-
tion. In industrial water conditioning, chlo- metric means is based on the fact that such
rination is employed primarily for the con- a water, when treated with an acid solution
trol of slime and algae although chlorine also of orthotolidine, will impart a definite colora-
may be used to assist in coagulation, taste, tion to the resulting mixture. Small amounts
odor, color and iron removal problems. of chlorine give a yellow color and larger
Formerly, no differentiation was made be- amounts give an orange color.
tween the forms of chlorine residual present By comparing the color developed in the
in a treated water. In recent years it has been sample with that of prepared standards rep-
recognized that there is a marked difference resenting certain known concentrations of
in the germicidal properties of a chlorine chlorine, the residual chlorine content of
residual, dependent on the form in which the unknown can be determined. Nitrites,
chlorine exists in the treated water. ferric iron and manganic compounds will
The free available chlorine residual is de- develop a color with orthotolidine and may
fined as that portion of the total residual yield a false residual chlorine value.
chlorine which will react chemically and To correct for interference and to deter-
biologically as hypochlorous acid or hypo- mine the free available and combined avail-
chlorite ion. It is in this form that chlorine able portions of the residual chlorine, sodium
exerts the most potent bactericidal effect. arsenite is employed, (OTA test). Sodium
The combined available chlorine residual arsenite reduces free or combined chlorine
is defined as that portion of the total residual but does not affect false residuals. A flash
chlorine which will react chemically and reading within ten seconds indicates the free
biologically as chloramines, or organic chlora- available chlorine residual and a five minute
mines. In this form, chlorine is a relatively reading yields the total residual of both free
and combined available chlorine. free distilled water. Multiply results by ap-
propriate factor.
RESIDUAL CHLORINE
APPARATUS REQUIRED. FREE AND COMBINED AVAILABLE

1-Taylor-Enslow Slide Chlorimeter com- RESIDUAL CHLORINE
plete (chlorine slide range 0.0-1.0 ppm
or 0.1-4.0 ppm) ADDITIONAL CHEMICALS REQUIRED.
Sodium Arsenite, 0.5%
CHEMICALS REQUIRED.
Orthotolidine Reagent PRoCEDURE FOR TEsT. As outlined under the
Residual Chlorine Test, place the three (3)
PROCEDURE FOR TEsT. Rinse the three (3) cells containing the sample in the base of the
cells thoroughly with the sample of water to chlorimeter. To each of the outside cells add
be tested and fill all cells to the mark with 0.5 ml of sodium arsenite solution. Mix and
the sample. add 0.5 ml of orthotolidine reagent. To the
Place the cells in the three ( 3) openings in middle cell add 0.5 ml orthotolidine reagent
the base of the chlorimeter making certain only and mix quickly. Read the chlorine
that the etched side of each cell is next to value within ten seconds. This flash reading
the rectangular opening in the back of the yields the free available chlorine residual cor-
base. rected for interference. After five minutes,
Add 0.5 ml of the orthotolidine to the mid- again read the chlorine value which will
dle cell by means of the dropper and mix be the total residual chlorine corrected for
thoroughly. A yellow color will develop if interference. The difference between the two
chlorine is present. The color standard slide readings will be the combined available chlo-
should then be placed on the base and moved rine residual. In the absence of interfering
in front of the test samples until either an substances, the sodium arsenite may be elim-
exact match is made with one of the stand- inated and the flash test used to differentiate
ards or the color is determined to be at a between the free and combined available
point between two adjacent standards. In chlorine residuals.
obtaining a color falling between two stand-
ard colors, the resulting concentration should CALCULATION OF RESULTS. If a comparison
be recorded as the average standard concen- of colors can be accomplished when testing
trations. As noted, interfering substances may the original sample, the amount of residual
cause the residual chlorine value to register chlorine in parts per million can be read
erroneously high. However, if the chlorinated directly from the color standard slide.
water contains no more than 0.3 ppm iron, If the original sample must be diluted,
0.01 ppm manganic manganese andjor 0.1 then the final value found for the diluted
ppm nitrite nitrogen, the characteristic yellow sample must be multiplied by the dilution
color with orthotolidine may be accepted as factor to give the amount contained in the
being due to chlorine. original sample.
Sunlight causes fading of the orthotolidine- LIMITATIONS OF TEsT. In order to obtain ac-
chlorine color and tests should never be made curate results, it is necessary to maintain a
in direct sunlight. Exposure of samples to pH value of 2.0 or less in the sample treated
direct sunlight before or during testing is not with orthotolidine. If the pH is above 2.0,
advisable. A standard artificial light source is reliable colors cannot be obtained. There is
best for consistent color comparisons. If not sufficient acid in the orthotolidine reagent
available, North daylight can be used. to give a pH of 1.3 when 0.5 ml is added to
If the chlorine content of the sample ex- 11.5 ml of drinking water or swimming pool
ceeds the range of the slide, discard test. Take water. In the case of sewage or process waters,
a smaller sample and dilute with chlorine- where alkalinity may be high, it may be
358 BETZ HANDBOOK
necessary to use 1.0 ml of orthotolidine. 1-Bottle, with dropper

Orthotolidine reagent should not be re- 1-Stirring rod, glass
tained for periods longer than three months CHEMICALS REQUIRED.
as an arbitrary precaution against the effect Xylene Cyanole, 0.02%
of possible occasional exposure to high tem- Hydrochloric Acid, 5 N
peratures or direct sunlight. PROCEDURE FOR TEsT. Measure a 100 ml
To avoid interferences due to iron, man- sample in the graduate and transfer to the
ganese, chromates and nitrates, the ampero- casserole. Add 2 or 3 drops of 5 N hydro-
metric titration method is recommended. This chloric acid. Add xylene cyanole from the
procedure permits an accuracy of + 0.01 burette drop by drop with constant stirring.
ppm. The required equipment includes a In colorless water, the endpoint is taken as
chlorine detecting cell and microammeter. the first appearance of a permanent light
Phenylarsene oxide is used as the titrating blue color; in water containing chromate,
solution. Amperometric residual chlorine re- the endpoint will be green (the shade of
cor:ders are also available. green is a function of the amount of chro-
mate present) .
FREE CHLORINE, MICRO-TITRATION CALCULATION OF RESULTS.
METHOD Formula:
THEORY OF TEsT. This test is based upon the ppm free chlorine as Ch = ml xylene cyanole X
reaction of xylene cyanole with free chlorine 30
to form a stable soluble compound. After the ml sample
free chlorine residual of the sample has been
depleted, further additions of xylene cyanole Using a 100 ml sample, the free chlorine in
produce a blue color which is taken as the parts per million as Cl2 is equal to the ml of
endpoint. xylene cyanole multiplied by 0.30.
APPARATUS REQUIRED. LIMITATIONS OF TEsT. This method measures
1-Burette, automatic, 25 ml only free chlorine and does not measure com-
1-Casserole, 210 ml bined chlorine. Ferric iron and nitrite up to
1-Cylinder, graduated, 100 ml 25 ppm do not cause interference.
Figure 51-1 • T aylor-Enslow Chlorimeter

52
Chron1ate
HROMATES are rarely present in untreated Chromate salts are not suitable for use in
C water, except as a result of industrial
contamination. However, chromates are
potable waters and the U. S. Public Health
Drinking Water Standards stipulate a limit
among the most widely used chemicals for for chromium of 0.05 ppm as Cr (hexa-
control of corrosion in industrial cooling water valent.
systems. The determination of chromate con- Where it is necessary to remove chromate
centration usually is made in order to regulate from plant waste discharge, this removal us-
treatment concentration to the level neces- ually is accomplished by reducing the hexa-
sary for satisfactory corrosion inhibition. valent chromate to trivalent chromium and
The most commonly used chromate salts then precipitating chromic hydroxide. Fer-
are sodium and potassium chromate and di- rous sulfate and lime commonly are used
chromate. The dichromates are acidic while for this purpose.
the chromates are alkaline. It is only in the
chromate or hexavalent form that chromium CHROMATE, TITRATION METHOD
salts function as corrosion inhibitors. Tri-
valent chromic salts do not inhibit corrosion. THEORY OF TEST. This test is based on the
Chromates are anodic inhibitors, and as oxidizing property of hexavalent chromium
such, can intensify pitting if used in insuffi- to liberate free iodine from potassium iodide
cient concentration to completely stifle cor- in acid solution. The liberated iodine is
rosive attack. Normally, concentrations of titrated with sodium thiosulfate solution in
300-500 ppm chromate are employed for the presence of a starch-type indicator. The
corrosion control in open circulating systems, disappearance of the blue color is taken as
while it is usual to employ 1000-2000 ppm the endpoint.
chromate in closed systems. However, the APPARATUS REQUIRED.
effect of other ions present may alter the 1-Burette, automatic, 25 ml
concentration required. High chloride hin- 1-Flask, Erlenmeyer, 250 ml
ders the development of a protective film by 1-Cylinder, graduated, 100 ml
chromate and, under some severe conditions, 1-Cylinder, graduated, 10 ml
chromate concentrations as high as 10,000 1-Measuring dipper (brass)
ppm have been required to stifle corrosive 1-Measuring dipper (plastic)
attack. CHEMICALS REQUIRED.
The combination of phosphate and chro- Sodium Thiosulfate, N /10
mate in the proper ratio and under con- Iodide Crystals
trolled pH conditions in the Dianodic method Starfamic Indicator
permits the use of relatively low chromate Sulfamic Acid
concentrations. The field of usefulness of
chromate corrosion inhibitors has been ex- PROCEDURE FOR TEST.
panded greatly by the development of the (Range 10-100 ppm as Cr0 4 )
Dianodic method. Inhibition of pitting as The N/100 sodium thiosulfate required to
well as general corrosion, is secured at rela- determine chromate in this range is not
tively low treatment concentrations. stable due to absorption of carbon dioxide
360 BETZ HANDBOOK
from the air. Therefore, this reagent must (Range 100-1500 ppm as Cr0 4 )
be freshly prepared (or restandardized) at Measure 50 ml of the water sample in the
least every two weeks. To 90 ml of distilled graduated cylinder and transfer to a 250 ml
water that has been boiled and cooled to Erlenmeyer flask. Use the plastic dipper to
room temperature, add 10 ml of N/10 sodium add two (2) dipperfuls (approx. 2.0 g) of dry
thiosulfate. Mix thoroughly and store in a sulfamic acid. Swirl the flask until the acid
tightly stoppered bottle. Avoid unnecessary is completely dissolved. Allow the sample to
exposure of the solution to air. stand for two (2) minutes to eliminate any
Measure 50 ml of the water sample in the nitrite interference. Then use the brass dip-
graduated cylinder and transfer to a 250 ml per to add two (2) dipperfuls (approx. 0.50 g)
Erlenmeyer flask. Use the plastic dipper to of iodide crystals. Swirl until the iodide is
add two (2) dipperfuls (approx. 2.0 g) of dry completely dissolved. Allow the sample to
sulfamic acid. Swirl the flask until the acid stand for two (2) minutes. Titrate the
is completely dissolved. Allow the sample to sample with N/10 sodium thiosulfate until
stand two (2) minutes to eliminate any ni- the yellow-brown color of iodine has almost
trite interference. Then use the brass dipper disappeared. Add one ( 1) brass dipperful
to add one (1) dipperful (approx. 0.25 g) of (approx. 0.15 g) of Starfamic Indicator and
iodide crystals. Swirl until the iodide is com- swirl to dissolve (all of the indicator may not
pletely dissolved. Allow the sample to stand dissolve, creating a slight haze in the sample) .
for two (2) minutes. Titrate the sample with Continue to titrate with N /10 sodium thio-
N /100 sodium thiosulfate until the yellow- sulfate until the blue color which developed
brown color of iodine has almost disappeared. upon the addition of the indicator first dis-
Add one (1) brass dipperful (approx. 0.15 g) appears. Disregard any reappearance of the
of dry Starfamic Indicator and swirl to dis- blue color. Record the ml of the sodium
solve (all of the indicator may not dissolve, thiosulfate used.
creating a slight haze in the sample) . Con-
tinue to titrate with NjlOO sodium thiosul- CALCULATION OF REsULTS. Using a 50 ml
fate until the blue color which developed sample, the parts per million of chromate as
upon the addition of the indicator first dis- Cr0 4 equals ml N/10 sodium thiosulfate re-
appears. Disregard any reappearance of the quired multiplied by 77.4.
blue color. Record the ml of the sodium thio- (For Brines)
sulfate used. Measure 10 ml of the brine sample in the
CALCULATION OF RESULTS. Using a 50 ml graduated cylinder. Make up to a total vol-
sample, the parts per million of chromate as ume of 50 ml with distilled water or tap
Cr0 4 equals ml N/100 thiosulfate required water and transfer to a 250 ml Erlenmeyer
multiplied by 7.74. flask.
In some cases it may be desirable to use a Use the plastic dipper to add two ( 2) dip-
sample larger than 50 ml. When this modi- perfuls (approx. 2.0 g) of dry sulfamic acid.
fication is required, the quantities of dry re- Swirl the flask until the acid is completely
agents and the multiplication factor should dissolved. Allow the sample to stand for two
be as follows: (2) minutes. Then use the brass dipper to
add two (2) dipperfuls (approx. 0.50 g) of
Sample Size Sulfamic Acid Iodide Crystals
iodide crystals. Swirl until the iodide is com-
100 ml 4 dippers 2 dippers pletely dissolved. Allow the sample to stand
200 ml 8 dippers 4 dippers for two ( 2) minutes. Titrate the sample with
Starfamic Indicator Factor N/10 sodium thiosulfate until the yellow-
brown color of iodine has almost disappeared.
1 dipper 3.87 Add one ( 1) brass dipperful ( approx. 0.15 g)
1 dipper 1.94
of dry Starfamic Indicator and swirl to dissolve
(all of the indicator may not dissolve, crea- chromate in the samples to be tested. The
ting a slight haze in the sample) . Continue dilutions of the standard should be treated
to titrate with N/10 sodium thiosulfate until in exactly the same manner as that shown
the blue color which developed upon the below for analysis of the water samples.
addition of the indicator first disappears. Dis- Each time a determination is made the
regard any reappearance of the blue color. calibration curve should be checked to estab-
Record the ml of sodium thiosulfate used. lish a correcti'on factor. This procedure is
CALCULATION OF REsULTS. Using a 10 ml necessary to insure that the results are ac-
sample, the parts per million of chromate as curate since reagent age and stability as well
Cr0.1 equals ml N /10 sodium thiosulfate re- as temperature can affect the results. The
quired multiplied by 387. curve should be checked with chromate-free
LIMITATIONS OF TEsT. These tests are not water and also at a dilution of the chrom-
specific for chromate, but rather are a meas- mate standard that approximates the middle
ure of the oxidizing substances present in the of the chromate range covered by the curve.
water. Ferric iron and organic matter inter- All reagents as well as the "check" samples
fere. These procedures are employed when and the actual sample to be analyzed must
chromate is present over approximately 10 be at the same temperature.
ppm and where sources of appreciable inter- Diphenylcarbohydrazide reagent employed
ference are absent. in this test must be prepared fresh daily. Add
Where chromate is to be determined in one level measuring cup of diphenylcarbo-
low concentration and also when interfering hydrazide (approx. 0.1 g) to 50 ml of ace-
substances may be present it is preferable to tone reagent. Stir until dissolved. A pink
employ the diphenylcarbohydrazide proce- color may develop in this reagent, but is of
dure, a Nessler tube or photometric method. no significance.
Prepare a "zero" reference blank by adding
CHROMATE, PHOTOMETRIC METHOD to a beaker 25 ml sample, 1 ml distilled water
and 1 ml sulfuric acid reagent. Use this refer-
THEORY OF TEsT. This test is based on the
ence blank to set meter at "zero" immediately
development of a reddish color produced by
prior to test.
the reaction of hexavalent chromium with
diphenylcarbohydrazide. While the color To a second beaker add 25 ml sample, 1 ml
comparison can be made with Nessler tubes, sulfuric acid reagent and 1 ml of diphenyl-
more accurate and satisfactory results can be carbohydrazide reagent. Allow to stand ex-
obtained with the use of a filter photometer actly 1 minute and immediately obtain dial
or a spectrophotometer. reading.
APPARATUS REQUIRED. CALCULATION oF RESULTS. Compare the
Filter photometer complete with assorted photometer dial reading with the calibration
laboratory glassware curve and read the results directly in parts
CHEMICALS REQUIRED. per million of chromate.
Acetone Reagent LIMITATIONS OF TEST. This procedure is not
Diphenylcarbohydrazide, C. P. affected by the ions normally found in water.
Chromate Standard, 10 ppm Cr04 The test determines only hexavalent (yellow)
Sulfuric Acid Reagent and not trivalent (green) chromium. Up to
PROCEDURE FOR TEsT. This procedure em- 2.5 ppm chromate as Cr0 4 can be deter-
ploys a wave length of 535 mu and a mini- mined without dilution of the sample. Con-
mum light path of 10 mm. Prepare a calibra- centrations above 2.5 ppm can be determined
tion curve for the photometer using succes- by diluting the original sample with distilled
sive dilutions of the chromate standard to water, following the procedure outlined, and
adequately cover the anticipated range of multiplying result by the appropriate factor.
362 BETZ HANDBOOK
The range covered by this chromate meth- ever, the accuracy of this chromate method
od is only 0-2.5 ppm and the normal pro- is so high that no appreciable error is intro-
cedure is to dilute the sample in order to duced even by high dilution factors such as
determine higher chromate contents. How- 1 ml of sample diluted to 50 mi.
Figure 52-1 • Filter Photometer-Chromate Anarysis

53
Color
HE measurement of the color of a water The usual methods for color removal in-
T does not provide any information con-
cerning the nature of the substance pro-
volve the use of iron or aluminum coagulants
at a low pH value followed by filtration. Acti-
ducing that color. Instead, similar to the vated carbon may also be employed.
measurement of turbidity, an arbitrary stand-
ard scale is used as a means for comparison COLOR, PLATINUM-COBALT METHOD
of color intensity with the water sample. A
color of 5 units means that the intensity of THEORY OF TEST. This test is based on the
the color of the water is equal to the in- comparison of the color of a water sample
tensity of the color of a sample of distilled with an arbitrary standard color scale. One
water containing 5 milligrams of platinum unit of color is defined as that produced by
(as potassium chlorplatinate) per liter. 1 mg of platinum per liter. The "true" color
Waters usually vary from colorless to a of a water sample is that due only to the
deep brown. Color in natural water gener- presence of soluble substances. The "ap-
ally is found to be due to the presence of parent" color is that due to the presence of
organics such as tannin from decaying vege- both soluble and suspended substances.
table growth or from various industrial wastes. APPARATUS REQUIRED.
The U. S. Public Health Service Drinking 1-Taylor Water Analyzer complete (color
Water Standards recommend that the color slide range 0-70 units)
of potable water should be less than 15 units.
In industrial water, color is most harmful PROCEDURE FOR TEST. Fill one of the com-
where the water is to be used in process work. parator tubes to the mark with sample and
For example, in the manufacture of paper, place in the middle compartment of the
highly colored waters can discolor the paper base. Fill the two comparator tubes on either
being manufactured. In boiler feedwater con- side to the mark with distilled water. Place
ditioning, the organic can concentrate and the standard slide on the base. Place the
reach a point where it makes the analysis of instrument so that the comparator tubes point
the water most difficult by masking end- toward a source of daylight. Read instrument
points and color comparisons. The organic by viewing the reflection in the mirror and
materials in solution may aggravate a carry- not by looking down through the tubes.
over condition in boiler water, particularly Move the slide in front of the test sample
if the organic material is from industrial until a color match is obtained.
wastes. CALCULATION OF REsULTs. The color value
In industrial cooling water, a high color of is read directly from the color standard slide
the makeup water may indicate the presence in terms of units of color.
of organic agents which may reduce cor- LIMITATIONS OF TEST. The true color of a
rosion inhibitors such as chromate and nitrite. water sample is due to soluble substances
Organic pollution may also provide an en- and its accurate determination where sus-
vironment favorable to the growth of slime pended matter is present is practically im-
and algae. possible. Filtration should not be used be-
364 BETZ HANDBOOK
cause of the decolorizing action. The best Photometers are not recommended for the
results can be obtained by centrifuging prior determination of color, because of interfer-
to color observation. ence from turbidity.
Figure 53-1 • Organic Identification by Infrared Spectroscopy

54
Conductance (Specific)
HE specific conductance of a water is a recommended since such a calibration intro-
T measure of the ability of the water to
conduct an electrical current. This property
duces an error into the instrument reading
itself. The conversion factor from micromhos
is of no consequence in itself with respect to of specific conductance to parts per million
water treatment. However, from a control will vary slightly with different waters. To
standpoint, the conductivity test is important include a constant conversion factor in the
as a direct measure of the total ionizable instrument calibration is to introduce an un-
solids in the water. The conductivity test necessary source of error.
provides an accurate measurement of steam The conductivity test provides an accurate
purity as well as a simple control for boiler and simple method of blowdown control.
water solids. Conductivity also may be used However, certain limitations must be con-
for blowdown control in recirculating cooling sidered. While the conductivity test measures
water systems. the total ionic concentration, the hydroxide
Specific conductance is inversely propor- ion has a much higher conductance than the
tional to electrical resistance. Pure water is other ions present. Thus, for accurate results
highly resistant to the passage of an electric the sample must be neutralized before the
current and therefore has a low specific con- conductivity test is made.
ductance. However, if the water contains Conductivity is an exceedingly sensitive
ions, the water becomes a better conductor test and is accurate down to the level of ap-
of electricity and the specific conductance is proximately 0.5-1.0 ppm ionizable solids. Un-
increased. Inorganic compounds such as so- til development of the flame spectrophotom-
dium chloride and sodium sulfate dissociate eter method for determining low sodium con-
into positive and negative ions, which will centrations, conductivity was the most ac-
conduct electricity in proportion to the curate method of determining steam purity.
amount of ions present. The conductivity The newer sodium method is preferred where
test, therefore, is not specific for any one ion, solids in the steam are 1 ppm or less.
but rather a measure of the total ionic con-
centration. CONDUCTANCE OF BOILER WATER,
The basic unit of electrical resistance is the CONDUCTO BRIDGE METHOD
ohm. Since electrical conductivity is the re-
ciprocal of resistance, the unique term "mho" THEORY OF TEsT. The ionizable solids in
("ohm" spelled backwards) was chosen as boiler water have the ability to conduct an
the basic unit of conductivity. In the con- electric current through a solution. This
ductivity test, small amounts of electrical property of electrical conductance of solids
conductance are measured and the instru- makes it possible to accurately measure the
ment is usually calibrated in micromhos (a quantity of solids in solution by suitable con-
micromho is a millionth of a mho). To cali- ductance equipment.
brate a conductivity instrument to read di- APPARATUS REQUIRED.
rectly in parts per million of dissolved solids 1-Conducto Bridge (choice of various
(or some specific ion or compound) is not ranges)
366 BETZ HANDBOOK
1-Dip cell (constant 2.0)

1-Cylinder, rimmed glass, not graduated
!-Thermometer, dial type (0-220 F)
1-Measuring cup, brass
CHEMICAL REQUIRED
Gallic Acid, c. p.
PROCEDURE FOR TEST. Pour approximately
50 ml of distilled water or steam condensate
into the rimmed glass cylinder and insert the
conductivity cell. Move cell up and down
several times to wash off any solids present
on the cell. Discard the water in the cylinder.
Pour approximately 50 ml of boiler water
in!o the cylinder (use a settled or filtered
sample) . Add two dippers of gallic acid
( approx. 0. 2 g) to the sample. (Note-if a
small amount of the gallic acid remains un-
dissolved, the conductivity test will not be ad-
versely affected.) Measure temperature of
the sample and adjust the temperature cor-
rection dial on the Conducto Bridge to the
proper temperature. Insert conductivity cell
and move up and down several times to in-
sure equilibrium. Measure the specific con-
ductance on the Conducto Bridge by turning
the conductivity dial until the electric "eye"
is at its widest black angle.
NOTE-Two dippers of gallic acid will
neutralize approximately 1300 ppm P alka-
linity. On some highly alkaline boiler waters,
additional gallic acid may be required. A de-
sirable precaution is to add approximately
four drops phenolphthalein indicator to the
sample and delay taking the conductivity
reading until the pink color has been com-
pletely discharged by the addition of gallic
acid.
CALCULATION OF RESULTS. The specific con-
ductance in micromhos is read directly from
the calibrated scale as indicated by the pointer
on the conductivity knob, when the "eye" is
at its widest black angle.
The relationship between specific conduct-
ance and the dissolved solids content of a
boiler water depends on the characteristics
of each individual boiler water and there-
fore may be slightly different for each plant. Figure 54-1 • Determination of Boiler Water
Using the gallic acid neutralization method, Conductivity
an average value determined over a wide

range of operating conditions is that one
micromho is equivalent to 0.9 ppm dissolved
solids. This value is sufficiently accurate for
the average industrial plant.
The exact relationship between micro-
mhos and solids can be individually estab-
lished for each plant by determining both
the conductance and solids content of a
series of approximately ten samples taken
over a two week period.
LIMITATIONS OF TEsT. The conductance
method affords a rapid means of checking
the dissolved solids content of a sample. The
effect of hydroxide in causing high conduc-
tivity is minimized by the gallic acid neutral-
ization, thereby securing a consistent rela-
tionship between solids and conductance. The
conductance method does not measure non-
electrolytic solids such as organic matter, and
in order to express results in terms of parts
per million of boiler water solids it is neces-
sary to use a conversion factor. This con- Figure 54-2 • Conductivity Test Set
version factor is described above.
perature correction dial on the Conducto
CONDUCTANCE OF STEAM CONDENSATE, Bridge to the proper temperature. Insert con-
CONDUCTO BRIDGE METHOD ductivity cell and move up and down several
times to insure equilibrium. Measure the
THEORY OF TEST. The ionizable solids and specific conductance on the Conducto Bridge
gases in condensed steam have the ability to by turning the conductivity dial until the
conduct an electric current through a solu- electronic "eye" is at its widest angle.
tion. This property of electrical conductance CALCULATION OF RESULTS. With the "eye"
of solids and gases make it possible to detect at its widest black angle, the scale on the
small quantities of these materials in solution conductivity knob indicates directly the spe-
by means of suitable conductance equipment. cific conductance in micromhos.
APPARATUS REQUIRED. An average value for converting micro-
1-Conducto Bridge ( 0-60 micromhos) mhos of conductance to dissolved solids in
1-Dip cell (constant 0.1) condensed steam samples is approximately
1-Cylinder, rimmed glass, not graduated 0.5-0.6 ppm dissolved solids per micromho
!-Thermometer, dial type (0-220 F) of conductance. This factor is subject to some
PROCEDURE FOR TEsT. Pour approximately slight variation dependent upon the indi-
50 ml of the steam condensate into the vidual steam condensate measured, but for
cylinder. Insert the conductivity cell, moving normal work this factor can be employed.
it up and down several times to wash off any In evaluating the conductance of a con-
solids present on the cell. Discard the water densed steam sample and expressing conduct-
in the cylinder. ance in terms of dissolved solids through the
Pour approximately 50 ml of the steam use of a conversion factor, it is necessary to
condensate into the cylinder. Measure tem- check the steam condensate for dissolved
perature of the sample and adjust the tem- gases such as ammonia and carbon dioxide
368 BETZ HANDBOOK
E
which impart conductance. Conductance cor- 0.
0.
rection for the presence of these gases may z

(/)
be determined by subtracting from the ob- ""

~
served conductance, the conductance due to z
0
ammonia and free carbon dioxide as indi- ..; "'"'

cated from the curves showing the conduct- 0
z ""
~ 4t---__;;:;.--r:
ance of these two gases when present to- (.)
=>
0
gether in steam condensate (see Figures 54-3, z
0
(.)
54-4 and 54-5) . Subtracting the value for (.)
u:
conductance due to dissolved gases from the ~ oL---~----~~--~----~--~~--~
~
observed conductance will yield the conduct- 0 5 10 15 20 25 30
FREE CARBON DIOXIDE AS C0 2, ppm
ance due to solids content of steam conden-
Figure 54-3 • Conductance Correction Curves for
sate. Ammonia and Carbon Dioxide
LIMITATIONS oF TEsT. The conductance
method is rapid and accurate for determin- u:,....
ing solids of 1 ppm or more in steam con- t:
10.0
densate. However, where any appreciable 0

(/)
::t: /
::E 8.0
quantity of ammonia or free carbon dioxide
~ E::::
0
a::
is present in the steam, it is necessary to de- 9 .........:
termine the concentration of these gases and "' 6.0
~
~~
w
0
make suitable correction, as otherwise the z
dissolved solids content calculated from the ""=>
1- 4.0
!?"
I~ b?
(.)
0
conductance reading would be erroneously z
8 2.0
high. In many cases for routine control, only
~
0
u: 91
negligible amounts of ammonia or free car- ~
0. 0
bon dioxide may be present and a check for Ul 0 ~ M U M M M ~ M U ~
AMMONIA AS N, ppm
these gases need be made only at infrequent
Figure 54-4 • Conductance Correction Curves Based on
intervals. Where a high percentage of con-
Ammonia and pH (Low Range)
taminated makeup is employed in a plant,
it usually is necessary to make frequent
checks for the ammonia and free carbon di- u: eo
....
t:
oxide content of the condensed steam. VI
0
::t:
::E 64
CONDUCTANCE CORRECTIONS FOR 0
a::
0
AMMONIA AND CARBON DIOXIDE i
w· ••
0
To properly interpret the conductivity of a z
condensed steam or condensate sample, it is g"" ~~--4---~---4~~~~~~~---4.-~

0
z
necessary that corrections be made for the 80 16~--~~~~~;_~r=--l----t---1--~
conductivity due to any gases present. The u:
purpose of the conductivity determination is iii
~ 4
to obtain a measure of the solids present in AMMONIA AS N, ppm
the condensed steam sample as an indication Figure 54-5 • Conductance Correction Curves Based on
of the degree of carryover of boiler water Ammonia and pH (High Range)
solids. In many cases, the presence of gases
such as ammonia and carbon dioxide will im- bon dioxide present and to make correction
part a conductivity far greater than that due for the conductivity due to these gases.
to the solids present. To interpret the con- Figures 54-3, 54-4, and 54-5 are based on
ductivity of the condensed steam sample, it is the data of R. 0. Parker and R. J. Ziobro
necessary to determine the ammonia and car- ("Comments on Corrections to Steam Con-
ductivity Measurements"-Presented at the Free carbon dioxide as C0 2 = 10 ppm

1956 N a tiona! Meeting, American Society for Correction due to gases = 6.0 micromhos
Testing Materials). Corrected specific conductance: ( 8.0-6.0)
Figure 54-3 shows the conductivity correc- = 2.0 micromhos
tion for ammonia in the range of 0-1.0 ppm Solids content of sample: 2.0 micromhos X
and carbon dioxide in the range of 0-30 ppm. 0.5 (average £actor) = 1.0 ppm solids
For example, if a condensed steam sample Figure 54-4 is useful in determining the
contained 10 ppm carbon dioxide and 0.6 conductivity correction where no free car-
ppm ammonia as N, the conductivity cor- bon dioxide is found present. The conduc-
rection would be 6.0 micromhos. This cor- tivity correction for ammonia is a function
rection must be subtracted from the ob- both of the ammonia concentration and the
served conductivity of the sample to obtain pH of the sample. For example, in a sample
the conductivity due to the solids present. containing 0. 7 ppm ammonia as N, with a
pH of 8.0, the correction value would be 6.0
micromhos specific conductance.
EXAMPLE
Figure 54-5 covers a higher range of am-
Specific conductance = 8.0 micromhos monia concentration and the same procedure
Ammonia as N = 0.6 ppm is used to obtain the conductivity correction.
370 BETZ HANDBOOK
55
Hardness
HE presence of calcium and magnesium by one accustomed to using a water with
T salts causes water to be hard, with the
degree of hardness being directly proportional
300 or 400 ppm hardness, whereas, one ac- ·
customed to using a water with less than 50
to the quantity of these heavy metals that are ppm hardness may call waters with a hard-
present. Hardness of natural water will vary ness of 100 ppm rather hard. A water with
considerably, depending upon the source a hardness of 100 ppm is not soft in terms
from which it is obtained. Sections that have of soap consumption in cleaning, washing and
limestone formations generally have a high laundering operation. The accompanying
hardness content in the water. Since surface table gives a general classification of waters.
waters are diluted by rainfall, well water in Hardness Classification
the same area will normally have a much
higher hardness than surface water since the Less than 15 ppm Very Soft Water
15 Ia 50 ppm Soft Water
flow is underground over rock layers and 50 to 100 ppm Medium Hard Water
through sand strata. 100 to 200 ppm Hard Water
Greater than 200 ppm Very Hard Water
The terms "temporary hardness" and
"permanent hardness" have been superseded In industry, high hardness is undesirable
by the expressions "carbonate hardness" and for laundries, metal finishing, dyeing and
"noncarbonate hardness". When the total textile plants, food processing, pulp and
hardness is greater than the carbonate and paper, bottle washing, photography, leather
bicarbonate alkalinity, the amount of hard- goods and many others. Hardness is also the
ness equivalent to the alkalinity is termed source of scale formation in boiler feedwater
carbonate hardness; the remainder of the heaters, feed lines, and economizers. Boilers
hardness is noncarbonate hardness. When the also will be heavily scaled due to precipita-
sum of carbonate and bicarbonate alkalinity tion of calcium and magnesium salts unless
equals or exceeds the total hardness, all hard- properly treated. In cooling water systems,
ness is carbonate hardness and there is no scale will develop in heat exchange equip-
noncarbonate hardness. ment, engine jackets, piping and in general
Hardness will cause numerous detrimental wherever water circulates and is exposed to
effects domestically, such as excessive soap a temperature change.
consumption in the home and laundries, as Hardness can be removed by lime-soda
well as scums and curds formed on equip- softening, ion exchange softening, hot lime-
ment; yellowing of fabrics; toughening of hot ion exchange and various combinations
vegetables; film formation in tea; and scale of processes. It may be removed in internal
formation in hot water heaters, pipes and boiler water conditioning by employing inor-
utensils. ganic salts such as phosphate and carbonate
The classification of water supplies as soft, in conjunction with either protective or re-
moderately hard, hard and very hard is active organic materials, which insure the
rather unsatisfactory even where the domestic precipitation of hardness from solution as a
use of water is concerned. Water with a hard- fluid nonadherent sludge.
ness of 100 ppm may be called a soft water Hardness may be controlled in internal
Fig. 55-1-Range of Recommended treatment chemicals employed. However, the

Values for Hardness in Industrial hardness usually is within the range of 10 to
Water Supplies 30 ppm for a hot lime-soda softener and 25
to 50 ppm for a cold process .softener.
Limiting or
Recommended
Industry and Process Values, ppm HARDNESS, TITRATION METHOD
Boiler Feedwater
At 0-150 psi . . . .. .. .. .. .. .. .. . . .. .. . .. .. 80 THEORY oF TEsT. This test is based on the
At 150-250 psi . . . . . . . . . . . . . . . . . . . . . . . . . . 40
At 250-400 psi . . . . . . . • . . . . . . . . • . • . . . . . . . 10 determination of the total calcium and mag-
Over 400 psi • . . . . . . . • • . . . . . . . . . . . . . . . . . . 2 nesium content of a sample by titration with
Brewing ......................•........... 200-300 a sequestering agent in the presence of an
Carbonated Beverages ..................... 200-250
Cooling . . . . . . • . . . . . . . . • • . . • . . . . . . . . . • . . . . 50 organic dye sensitive to calcium and mag-
Food Canning and Freezing nesium ions. The endpoint occurs when all
General . . . • . . . . . . . . • . • . . . . . . • . . . . . . . . . . 50-85
Legumes . . . • • . . . . . . . . . . . . . . . . . . . . . • . . . . . 25-75
the calcium and magnesium ions are seques-
Fruits and Vegetables .................... 100-200 tered and the endpoint is a color change
Peas .................••........•.•..... 200-400 from red to blue.
Food Equipment Washing . . . . . . . . . . . . . . . . . . • 10
Food Processing, General •................. 10-250 APPARATUS REQUIRED.
Laundering • . . . . . . . . . . . . . . . . . . . . . . . • . . . • . . 0-50 1-Burette, automatic, 25 ml
Pulp and Paper Making
Groundwood Pulp ....•.....•..•.......... 200 1-Casserole, porcelain, 210 ml
Soda Pulp •••..........•..•...•.•....... 100 1-Cylinder, graduated, 50 ml
Kraft Pulp, Bleached ....•••..............• 100 1-Stirring rod, glass
Kraft Pulp, Unbleached .....•.•..•...•...• 200
Fine Paper Pulp ......................•.. 100 2-Measming dippers, brass
Rayon CHEMICALS REQUIRED.
Pulp Production .......•..•....••... _. .•... 8
Cloth Manufacture .•......•........•..••. 55 Hardness Indicator
Steel Manufacturing ....•.....•...•.......•. 50 Hardness Buffer Reagent
Synthetic Rubber •.•.......•....••.•..•...• 50
Tanning
Hardness Titrating Solution, 1 ml = 1 mg
Beam House •......••.•.••...........•.. 513 CaCOs
Tan House •...................••...•...• 50-135
Textile Manufacturers ..........•...•........ 0-50 PROCEDURE FOR TEST. Measure 50 ml of
sample and transfer to the casserole. With
Source-"Water Quality Criteria" the measuring cup provided, add 1 level
State Water Pollution Control Board
Sacramento, California (1952) measure (approx. 0.2 g) of hardness buffer
reagent to sample and stir. Add 1 level mea-
cooling water treatment by employing organic sure ( approx. 0.2 g) of hardness indicator
and inorganic surface active agents which and stir. If hardness is present, the sample will
frequently are used in conjunction with acidic turn red. Add the hardness titrating solution
materials such as sulfuric acid. Here the pur- slowly from the burette with constant stir-
pose of the treatment is to retain the hard- ring. When approaching the endpoint, the
ness in solution and prevent its precipitation sample begins to show some blue coloration,
as scale. but a definite reddish tinge can still be ob-
Hardness in an ion exchange softener ef- served. The endpoint is the final discharge
fluent should not exceed approximately 5 of this reddish tinge to produce a definite
ppm. When this type of softening is used for blue color. Further addition of the hardness
makeup to high pressure boiler systems, the titrating solution will not produce any color
effluent hardness should be controlled to change.
much lower values to minimize sludge forma- With the procedure given above the hard-
tion in the boilers. The hardness of the efflu- ness titrating solution must be added slowly
ent from a lime-soda softener will vary de- since the endpoint is sharp and rapid. For
pending upon the temperature at which the routine hardness determination it is suggested
process is carried out and the quantity of that 50 ml of sample be measured, but only
372 BETZ HANDBOOK
approximately 40-45 ml be added to the cas- titrating solution employed to the second
serole at the start of the test. Add the hard- decimal place. Multiply result by 10 to ob-
ness buffer reagent and hardness indicator as tain hardness in parts per million as CaC0 3 •
described above and rapidly titrate to the
endpoint. Then add the remaining portion LIMITATIONS OF TEsT. This method of hard-
of the sample. The hardness present in the ness determination is more accurate than the
remainder of the sample will turn the con- "soap method" and is considerably more
rapid. Hardness values up to 1200 ppm can
tents of the casserole red. Continue titrating
slowly until the final endpoint as indicated be titrated with an accuracy of 2% without
above is reached. Record the total number of dilution of the sample. In low hardness con-
milliliters of hardness titrating solution em- centrations (0-5 ppm) an accuracy of 0.1
ppm can be obtained.
ployed.
Interference results with the presence of
CALCULATION OF RESULTS. sodium hydroxide over 300 ppm, copper over
Formula: 10 ppm, iron over 20 ppm, manganese over
· p'pm hardness as CaC03 = 2 ppm, aluminum over 20 ppm, orthophos-
1000 phate over 100 ppm and polyphosphate over
ml titrating solution X - - - - - 25 ppm. Strontium titrates as hardness. Other
ml sample
ions in the concentration normally encoun-
Using a 50 ml sample, the hardness in tered in industrial water conditioning produce
parts per million as CaCOa is equal to the no interference.
ml of titrating solution employed multiplied The effect of interfering ions can be re-
by 20. duced by dilution of the sample. If interfer-
ence occurs, measure a 25 ml sample, dilute
Low HARDNEss PROCEDURE. In low hard- with 25 ml of distilled water and proceed in
ness waters (0-5 ppm), an accuracy of 0.1 the normal fashion. Multiply results by 40
ppm hardness can be obtained by employing instead of 20 as described under "Calcula-
the following special procedure. tion of Results". This procedure will reduce
The same strength chemicals and appara- any interfering ions, usually enabling a sharp
tus given above are used with the exception endpoint to be obtained.
that a microburette is employed instead of a When waters contain free mineral acid,
regular burette. such as the effluent of a hydrogen zeolite, the
In the determination of low hardness, 100 sample should be neutralized with 1 N so-
ml of sample is taken. Add 2 level measures dium hydroxide. Litmus paper can be used
( approx. 0.4 g) of hardness buffer reagent as the indicator, adding sufficient sodium
and stir. Add 1 level measure ( approx. 0.2 g) hydroxide until a blue color is obtained.
of hardness indicator and stir. The titrating In the presence of certain ions, the color
solution is added very slowly from the micro- change may not be from red to blue. For
burette. example, 500 ppm chromate causes no test
The endpoint is the final disappearance of interference. However, with chromate present
any reddish tinge in the sample and the de- the endpoint is from red to green, due to the
velopment of a pure blue color which will color of the chromate. The true endpoint in
not change upon further additions of hard- each case in taken as the dissappearance of
ness titrating solution. the reddish tinge of the sample, regardless of
Record the number of ml of hardness the final endpoint color.
Figure 55-2 • Hardness Test Set

374 BETZ HANDBOOK
56
Hydrazine
As boiler operating pressures have in-
creased, two disadvantages have become
evident in the use of sodium sulfite as a chemi-
based upon samples of boiler feedwater or
boiler water, depending on plant operating
conditions.
cal deaerant. Sodium sulfite reacts with oxy-
gen to produce sodium sulfate which increases HYDRAZINE, TAYLOR METHOD
the dissolved solids concentration in the boiler
water. In addition, there is some tendency for THEORY oF TEsT. This test is based on the
sulfite to decompose at high pressure to pro- reaction of paradimethylaminobenzaldehyde
duce sulfur dioxide and hydrogen sulfide, both with hydrazine to produce a yellow color.
of which are acidic gases and can contribute The intensity of the yellow color is propor-
to corrosion in condensate return systems. tional to the amount of hydrazine present
in the sample and follows Beer's law.
Hydrazine (N2Hi) reacts with dissolved
oxygen to produce nitrogen and water. There- APPARATUS REQUIRED.
fore, the products of the hydrazine reaction 1-Taylor Water Analyzer complete (hy-
do not add solids to the boiler water. In the drazine slide 0.00-0.30 N 2 H 4 )
event there should be some decomposition of 1-Bottle, amber glass, 4 oz.
hydrazine, the products of the decomposition 1-Bottle, clear glass, 2 oz, calibrated 50
reaction are ammonia and nitrogen. Since ml, glass stoppered
nitrogen is an inert gas and ammonia is al- 1-Cylinder, graduated, 50 ml
kaline, there is no attack on ferrous metals. 1-Measuring dipper, plastic
However, if ammonia is present in sufficient 1-Pipette, graduated, 2.5 ml
quantity, it will attack copper and copper CHEMICALS REQUIRED.
bearing alloys. Taylor Hydrazine Reagent A
Hydrazine is a toxic liquid and must be Taylor Hydrazine Reagent B
handled with care. Pure hydrazine has a low PRoCEDURE FOR TEsT. The single reagent
flash point so that a 35% aqueous solution used in this test is prepared from Hydrazine
usually is employed in water conditioning to Reagent A (a powder) and Hydrazine Re-
avoid fire hazard. Theoretically, 1.0 ppm hy- agent B (a solvent) . Measure 44 ml of Re-
drazine is required to react with 1.0 ppm dis- agent B and transfer to the amber bottle. Add
solved oxygen. Therefore, the theoretical two ( 2) plastic dipperfuls ( approx. 1.6 g)
quantity of the 35% solution of hydrazine is of Reagent A and mix thoroughly. Use the
slightly less than 3.0 ppm to 1.0 ppm dis- amber bottle for storage; keep tightly closed
solved oxygen. In practical applications, hy- and avoid unnecessary exposure to air. Due
drazine usually is required in quantities some- to limited stability prepare fresh reagent
what greater than theoretical. To avoid ex- every two weeks.
cessive ammonia concentrations in the steam To test for hydrazine fill the 2 oz glass
the residual hydrazine in the boiler water usu- stoppered bottle to the mark (50 ml) with
ally is maintained at 0.1 ppm or less. In some the water sample. Use the pipette to add 2.5
cases, a residual of hydrazine as low as 0.01 ml of the prepared Hydrazine Reagent to the
ppm is employed. The control ranges are bottle. Stopper the bottle and shake to mix
thoroughly. Allow the treated sample to stand should be prepared for use in this hydrazine
in the stoppered bottle for 30 minutes. test. These reagents and the chemicals used
At the expiration of 30 minutes fill one of to prepare them should be handled with care.
the tubes with the treated sample and place When it is necessary to use a pipette, always
it in the middle compartment of the Analyzer. use an aspirator bulb.
Fill the other two tubes with the original un-
Reagent A--Dissolve 4 g of paradimethyl-
treated sample and place them in the outer
aminobenzaldehyde powder in
compartments of the Analyzer.
200 ml of methyl alcohol.
Place the instrument so that the compara- When dissolved, add 15 ml of
tor tubes point toward a source of daylight. concentrated hydrochloric
Read the instrument by viewing the reflec- acid.
tion in the mirror and not by looking down
Reagent B'--Mix 200 ml methyl alcohol
through the tubes. Move the slide in front of
and 15 ml concentrated hy-
the test samples until a color match is ob- drochloric acid.
tained.
Reagent C-Dilute 74 ml concentrated hy-
CALCULATION OF REsULTS. Using a 50 ml
drochloric acid to one liter
sample, hydrazine as N 2 H4 is read directly with distilled water.
from the color slide in parts per million.
Reagent D-Pipette 10 ml hydrazine stand-
LIMITATIONs oF TEsT. The ions normally
ard solution into a one liter
present in industrial water do not interfere
volumetric flask. Make up to
with this test. If oxidizing agents are col-
one liter with Reagent C. One
lected with the sample or absorbed prior to
ml of Reagent D is equal to
testing, the hydrazine content may be dimin-
0.001 mg hydrazine as N2H,.
ished. This method of analysis is not suitable
for water samples that are colored. If color This method for the determination of hy-
is present, the photometric method should be drazine employs a wave length of 460 mu
employed. and a light path of 20 mm. Prepare a cali-
bration curve in the following manner. Mix
HYDRAZINE, PHOTOMETRIC METHOD 20 ml of Reagent B and 25 ml of Reagent
C; use this mixture to set the photometer at
THEORY OF TEsT. This test is based on the "zero" on the dial. Mix 20 ml of Reagent A
reaction of paradimethylaminobenzaldehyde and 25 ml of Reagent C; use this mixture to
with hydrazine to produce a yellow color. obtain a reference point for hydrazine-free
The intensity of the yellow color is propor- water ("zero" hydrazine) .
tional to the amount of hydrazine in the
Reference points for the remainder of the
sample and follows Beer's law. J;he method
calibration curve are obtained by a series of
can be modified to eliminate interference
dilutions of Reagent D to a total volume of
from color.
200 ml, using a volumetric flask and Reagent
APPARATUS REQUIRED. C as the diluent. For example, 1.0 ml Re-
Filter photometer complete with assorted agent D diluted to 100 ml in a volumetric
laboratory glassware. flask with Reagent C is equivalent to 0.01
CHEMICALS REQUIRED. ppm hydrazine. Reference points are obtained
Paradimethylaminobenzaldehyde Powder by treated 25 ml portions of the dilutions
Hexyl Alcohol with 20 ml Reagent A as shown below.
Hydrazine Standard Solution, 1 ml=O.l Each time a determination is made the
mg N2H1 calibration curve should be checked to es-
Hydrochloric Acid, Concentrated tablish a correction factor. This procedure is
Methyl Alcohol (absolute) necessary to insure that the results are accu-
PROCEDURE FOR TEsT. The following reagents rate since reagent age and stability as well as
376 BETZ HANDBOOK
temperature can affect the results. The curve at the end of one minute.
should be checked with Reagent C and also To determine hydrazine in a colored sam-
at a dilution of Reagent D that approximates ple-measure 50 ml of sample (collected as
the middle of the hydrazine range covered by per above) in a graduate and transfer to a
the curve. All reagents as well as the "check" scparatory funnel. Add 10 ml hexyl alcohol
samples and the actual sample to be analyzed by means of a pipette and shake to mix. Al-
must be at the same temperature. low the layers to separate and drain bottom
To determine hydrazine in a colorless sam- layer into a beaker for use as the sample.
ple-add 15 ml concentrated hydrochloric Proceed as under "colorless sample".
acid to the sample bottle and collect sample CALCULATION OF RESULTS. The hydrazine
with minimum exposure to air until the 200 content in parts per million as NzH4 is ob- ·
ml mark is reached. Measure 25 ml of the tained by reference to the calibration curve.
sample and add 20 ml of Reagent B. Use this LIMITATIONS oF TEST. The ions normally
mixture as a blank to set photometer at present in industrial water do not interfere
"zer9". Measure a second 25 ml portion of with this test. If oxidizing agents are collected
the sample and add 20 ml of Reagent A. with the sample or absorbed prior to testing,
Place in the photometer cell and read the dial the hydrazine content may be diminished.
57
Iron
RON exists in two states-ferrous iron water must be considered in interpreting the
I and ferric iron. In the ferric state, the
iron is completely oxidized, whereas in the
results of the test. For example, if a boiler
water sample contains appreciable quantities
ferrous state it is only partially oxidized. In of total iron, it should not be immediately
the total dissolved iron test, both the ferrous concluded that serious corrosion of the boiler
and ferric iron are measured. metal is occurring. An examination of the
Since iron is one of the most common ele- natural iron content of the feedwater should
ments in the earth's crust, it frequently is be made since iron will concentrate in a
found in natural waters. In surface waters, boiler similar to other ions.
iron is in the ferric state since it is completely Iron removal can be accomplished by aera-
oxidized by the oxygen in the atmosphere. tion, coagulation and filtration, lime and
However, in well waters the iron usually is in lime-soda softening, cation exchange and
the ferrous form. Upon removal from the contact filtration. Frequently it is more ad-
well, exposure to air and release of carbon visable to retain iron in solution rather than
dioxide, the iron in the water is converted to to remove it. Iron retention is secured with
the ferric state. the use of various surface active agents.
Iron is classified as an incrusting solid simi-
lar to calcium and magnesium since iron will TOTAL IRON, PHENANTHROLINE METHOD
precipitate out of solution and can form a
scale. Iron in water also is objectionable from THEORY OF TEsT. This test is based on the
the standpoint that it will form stains destroy- development of an orange-red complex from
ing the sanitary appearance of sinks, lava- the reaction of ferrous iron with phenanthro-
tories, etc. The U. S. Public Health Drinking line. The color developed is matched with
Water Standards recommend a maximum of that of freshly prepared standards using Ness-
0.3 ppm iron as Fe in potable waters. ler tubes to make the comparison. A filter
In addition to the "natural" iron content photometer or a spectrophotometer also may
be used at a wave length of 510 mu and a
of a water, iron is placed into solution when
minimum light path of 10 mm. When either
the corrosion of iron and steel surfaces oc-
of these instruments are employed, the values
curs. Thus, the iron test occupies a unique
position inasmuch as the interpretation of of iron in parts per million are read directly
this test must consider the factors of both from calibration curves.
scale and corrosion. For example, an iron de- APPARATUS REQUIRED.
posit can occur in a cooling water system as a 1-Nessler tube rack ( 12 hole)
scale problem (deposition of the natural iron 12-Nessler tubes (2 matched sets of 6
in the water) or as a corrosion problem each) , 100 ml
(wasting away of the metal surfaces). 1-Flask, volumetric, 100 ml
In industrial water conditioning, the deter- 1-Flask, Erlenmeyer, 250 ml
mination of iron can reveal information on 1--Pipette, delivery, 100 ml
corrosive action taking place within a system. 1-Pipette, safety, 2 ml
However, the natural iron content of the 1-Pipette, delivery, 10 ml
378 BETZ HANDBOOK
1-Pipette, Mohr, 1 ml other analyses. If the hydroxylamine reagent

1-Pipette, Mohr, 5 ml should become cloudy, filter before use.
CHEMICALS REQUIRED. Preparation of Standards. Prepare a series
Hydrochloric Acid, Concentrated of standards in increments of 0.1 ppm Fe for
Hydrochloric Acid, 10% use in the Nessler tubes. A typical series
Iron Standard Solution, 10 ppm Fe would cover the range of 0.0 to 1.0 ppm Fe.
Hydroxylamine Reagent (modified to elimi- Fill the 100 ml volumetric flask approxi-
nate necessity for ammonium acetate buf- mately half full of iron-free distilled water.
fer) Add the iron standard solution in incre-
Orthophenanthroline Reagent ments of 1 ml using a graduated Mohr pip-
PRoCEDURE FOR TEsT. All glassware (pipettes, ette. For example, if only 1 ml is added, then
beakers, etc.) must be washed with concen- the standard will be equivalent to 0.1 ppm
trated hydrochloric acid and rinsed with Fe; if 2 ml are added, the standard will be
iron-free distilled water prior to use. Clean- equivalent to 0.2 ppm Fe, etc. Add 2 ml of
ing the equipment in this manner is neces- 10% hydrochloric acid to the flask and fill
sary to remove any iron oxide which may be to the 100 ml mark with distilled water.
present as a result of the use of glassware for Stopper and invert several times to mix
Figure 57-1 • Iron Determination-Phenanthroline A1ethod

thoroughly. Do not use any rinse water. Pipette 10 ml

Transfer the contents of the volumetric hydroxylamine reagent into the flask and
flask to a 250 ml Erlenmeyer flask. Pipette swirl to mix. Pipette 10 ml orthophenanthro-
10 ml hydroxylamine reagent into flask and line reagent into the flask and swirl to mix.
swirl to mix. Pipette 10 ml orthophenanthro- Fill a Nessler tube to the 100 ml mark with
line reagent into the flask and swirl to mix. the prepared sample and place in the rack.
Fill Nessler tubes to the 100 ml mark with Allow 10-15 minutes for color development.
the series of prepared standards and place in Then compare with the iron standards by
the rack leaving every other hole empty for looking down through the tubes, preferably
insertion of the Nessler tube containing the using daylight as a source of light.
prepared sample. The dilute iron standards CALCULATION OF RESULTS. 'When a color
should be prepared fresh daily since they are match is obtained, record the value as parts
not stable for any appreciable length of time. per million total iron as Fe. If it was neces-
To Analyze for Total Iron. Mix the sam- sary to dilute the sample before starting the
ple thoroughly and pipette 100 ml into a 250 analysis, multiply by an appropriate factor.
ml Erlenmeyer flask. If the sample contains LIMITATION OF TEST. Phosphates will inter-
more than 2 ppm iron, dilute an aliquot por- fere particularly in the form of polyphos-
tion with distilled water so that the value will phate. Chromium interferes; zinc in concen-
be 2 ppm or less. Pipette 2 ml of 10% hydro- trations ten times that of iron; copper and
chloric acid into the sample and boil until the cobalt above 5 ppm; nickel above 2 ppm;
volume is reduced to approximately 20 mi. bismuth, silver, cadmium, mercury and mo-
Cool and transfer the sample to the volu- lybdate are precipitated by phenanthroline.
metric flask, using distilled water to rinse If much color or organic is present, it may
from the Erlenmeyer to the volumetric flask. be necessary to evaporate the sample, gently
Dilute to the 100 ml mark with distilled ash, and redissolve in acid. Total iron can be
water, stopper and invert to mix. determined in the range of 0.02 to 4.0 ppm.
Transfer the contents of the volumetric Higher concentrations of iron require the use
flask into a clean 250 ml Erlenmeyer flask. of aliquots.
380 BETZ HANDBOOK
58
Manganese
M ANGANESE is frequently encountered
in iron bearing waters although it usu-
ally is present to a lesser extent than iron.
permanganate. The oxidation is carried out
in the presence of silver nitrate and mercuric
sulfate in an acid medium. The color devel-
The difficulties caused by the presence of each oped is matched with that of freshly prepared
ele_ment are very similar and the methods standards using Nessler tubes to make the
employed for iron removal usually also effect comparison. A filter photometer or a spectro-
the removal of manganese. photometer also may be used at a wave
Manganese is encountered most frequently length of 525 mu with a minimum light path
in well waters and usually is not present in of 10 mrn. When either of these instruments
excess of 3.0 ppm. Like iron, this element will is employed, the values of manganese are
deposit from solution, clogging piping, valves, read directly from calibration curves.
etc., with a gray or black deposit of man- APPARATUS REQUIRED.
ganese hydroxide. The presence of mangan- 1-Nessler tube rack (12 hole)
ese causes a water to produce undesirable 12-Nessler tubes (2 matched sets of 6
stains in process work. For most industrial each), 100 ml
purposes, the manganese content of a water 1-Flask, volumetric, 100 ml
supply should not exceed 0.1 ppm and for 1-Flask, Erlenmeyer, 250 ml
many purposes, such as the manufacture of 1-Pipette, delivery, 100 ml
fine papers, even this quantity of manganese 1-Pipette, Mohr, 5 ml
represents excessive contamination. 1-Measuring dipper (plastic)
The U.S. Public Health Service Drinking 1-Pipette, Mohr, 1 ml
Water Standards recommend that the man- CHEMICALS REQUIRED.
ganese content of a potable water should not Ammonium Persulfate
exceed 0.05 ppm. as Mn. Manganese Standard Solution, 10 ppm Mn
The methods used for manganese removal Manganese Special Solution
include aeration followed by filtration, con- PROCEDURE FOR TEsT. Manganese may exist
tact filtration, lime or lime-soda softening and in a soluble form in a neutral water when a
zeolite softening. As is the case with iron re- sample is first collected, but readily oxidizes
moval, for sodium zeolite softening to oper- and precipitates from solution or becomes ad-
ate successfully it is necessary that contact sorbed on the walls of the container. Man-
with air be avoided prior to softening. If oxi- ganese should therefore be determined soon
dation of the manganese occurred prior to after sample collection. Otherwise, the sample
sodium zeolite softening, the precipitation of should be acidified at the time of collection.
manganese hydroxide would take place, clog- Preparation of Standards. Prepare a series
ging the zeolite bed. of standards in increments of 0.1 ppm Mn
for use in the Nessler tubes. A typical series
TOTAL MANGANESE, PERSULFATE METHOD
would cover the range of 0.0 to 1.0 ppm Mn.
Pipette 100 rnl distilled water into a 250
THEORY OF TEST. This test is based upon the ml Erlenmeyer flask. Add the manganese
oxidation of manganous compounds to form standard solution in increments of 1 ml using
a graduated Mohr pipette. For example, if the value will be 1 ppm or less. Starting with
only 1 ml is added, then the standard will be the addition of 5 ml of manganese special
equivalent to 0.1 ppm Mn; if 2 ml are added, solution, prepare the sample in exactly the
the standard will be equivalent to 0.2 ppm same manner as that shown in the prepara-
Mn, etc. Pipette 5 ml of manganese special tion for manganese standards.
solution into the flask. Boil until the volume
Compare the ~olor of the sample with that
is reduced to approximately 90 mi. Add one
of the standards by placing the sample in
( 1) level plastic dipper ( approx. 0.8 g) of
the rack and looking down through the tubes.
ammonium persulfate and boil for one ( 1)
It is preferable to use daylight as a source of
minute. Cool and transfer to a 100 ml volu-
light.
metric flask using a very small amount of dis-
tilled water to rinse from the Erlenmeyer CALCULATION OF RESULTS. When a color
flask into the volumetric flask. Fill the volu- match is obtained, record the value as parts
metric flask to the 100 ml mark with dis- per million total manganese as Mn. If it was
tilled water. Stopper and invert to mix thor- necessary to dilute the sample before starting
oughly. the analysis, multiply by an appropriate fac-
Fill Nessler tubes to the 100 mlmark with tor.
the series of prepared standards and place in LIMITATIONS OF TEsT. The persulfate method
the rack leaving every other hole empty for is preferred for the determination of man-
insertion of the Nessler tube containing the ganese in unknown samples because it is less
prepared sample. The dilute manganese subject to chloride interterence and is more
standards should be prepared fresh daily since rapid for low manganese concentrations.
they are not stable for any appreciable length Only minute amounts of bromide or iodide
of time. may be present. Small amounts of organic
To Analyze for Total Manganese. Mix do not interfere if the period of heating and
the sample thoroughly and pipette 100 ml the amount of persu!fate is increased. In the
into a 250 ml Erlenmeyer flask. If the sample absence of interference, manganese may be
contains more than 1 ppm manganese, dilute determined in the range of 0.05-1.5 ppm
an aliquot portion with distilled water so that with an accuracy of 1%.
382 BETZ HANDBOOK
59
Nitrate
HE nitrate ion is present in natural waters is subject to seasonal vanatron, it is
T waters in relatively small quantities.
Nitrogenous compounds may be introduced
still necessary to run control tests on the
boiler water for nitrate concentrations in
with sewage with the subsequent oxidation order to insure adequate nitrate content for
of these compounds to nitrate. inhibition of embrittlement.
The ,U. S. Public Health Service Drinking
Water Standards place no specific limit on NITRATE IN BOILER WATER, BRUCINE METHOD
the nitrate content of a water. Recently it
has been found that high nitrate concentra- THEORY OF TEsT. This test is based on the
tions ( 10 ppm and above) appear to be the development of a yellow-amber coloration
cause of methemeglobinemia in infants (blue produced by the reaction in sulfuric acid
babies) . Such high nitrate concentrations solution, of the brucine reagent with the
seem to be confined to rural dug wells sub- nitrate content of the water sample. The
ject to surface influence. color developed is matched with nitrate
In industrial water conditioning, the ni- standards and the nitrate content in parts
trate ion does not possess much significance per million is read directly from the slide.
except as affecting the quality of boiler water APPARATUS REQUIRED.
from the standpoint of embrittlement. Re- 1-Betz-Taylor Nitrate Comparator, com-
search has shown that the maintenance of plete (nitrate slide range 0-100 ppm)
certain sodium nitrate-sodium hydroxide ra- 1-Beaker, 100 ml
tios in boiler water can inhibit intercrystal- 1-Cylinder, graduated, 10 ml
line cracking (caustic metal embrittlement) . 1-Dropper, calibrated, 0.5 ml
Because of the widespread use of sodium
CHEMICALS REQUIRED.
nitrate to control embrittlement, the deter-
Brucine Reagent
mination of the nitrate ion in boiler water
Sulfuric Acid, Concentrated
has assumed increasing importance.
The nitrate concentration of a boiler water PRoCEDURE FOR TEsT. Measure by means of
necessary to overcome embrittling tendencies the 10 ml graduate, 5 ml of the sample to be
will vary with the boiler water alkalinity and tested and pour into the 100 ml beaker. By
with the pressure of operation. It is customary means of the dropper add exactly ten drops
to determine the sodium nitrate-sodium hy- of brucine reagent. Carefully measure 10
droxide ratio required for each individual ml of sulfuric acid and slowly add to the
plant by running tests with an embrittlement contents of the beaker, avoiding spattering.
detector. Once the proper ratio has been Swirl the beaker gently to mix, being care-
determined, the sodium nitrate feed can be ful to avoid spilling since the solution is now
controlled on this basis. Occasionally, it will strongly acid. This solution will become
be found that the natural nitrate content of warm following the addition of the acid.
the raw water is sufficient after concentration Allow to stand five minutes after the ad-
in the boiler, to provide the necessary ratio. dition of acid (four minutes minimum; six
However, since the nitrate content of natural minutes maximum). Then add 10 ml dis-
Figure 59-1 • Nitrate Determination-Brucine Method
tilled (or nitrate-free) water. Swirl carefully million as N0,1 is read directly from the
to mix. The solution may be cooled at this values on the color standard slide.
point if desired. Pour into one of the 5ml LIMITATIONS OF TEsT. This test is unaffected
test tubes. Place this tube in the middle com- by ions in the concentrations normally en-
partment of the comparator base. Fill two countered in boiler water. Chloride, sulfite,
of the tubes with distilled water or clear tap phosphate, iron, alkalinity, ammonium, cal-
water and place on either side of the sample cium, magnesium and silica do not interfere.
(for most accurate results with highly colored High tannin concentrations in excess of 800
water, prepare the two outside tubes with 5 color units cause interference, but this effect
ml sample, 10 ml sulfuric acid and 10 ml can be eliminated by dilution of the sample.
distilled water as outlined above, omitting Although the slide covers the range of 0-
the brucine reagent) . 100 ppm nitrate, it is difficult to distinguish
Place the color slide on the base and hold- the color difference above 50 ppm. For this
ing the instrument toward a source of day- reason, it is recommended that a diluted
light, move the slide in front of the test sample be taken if the observed nitrate value
samples until a color match is obtained. exceeds 50 ppm.
If nitrate exceeds 50 ppm, discard test. For the most accurate determination of
Take a 1.0 ml sample, dilute to 5 ml with nitrate, a photometer procedure is recom-
distilled water and proceed as before, multi- mended. The same reagents and essentially
plying the observed nitrate reading by five to the same procedure are employed, but the
obtain parts per million nitrate as N03. nitrate concentration is obtained from the
CALCULATION OF REsULTS. When a color photometer dial reading by reference to a
match is obtained, the nitrate in parts per calibration curve.
384 BETZ HANDBOOK
60
Nitrite
ITRITE is present in water as an interme-
N diate compound in an oxidation or re-
duction pr'ocess and represents a stage of the
NITRITE, DIRECT TITRATION METHOD
THEORY OF TEsT. This test is based upon the

nitrogen cycle. In surface supplies, trace determination of the nitrite content of a
amo11nts of nitrite may indicate the presence sample by titration in an acid medium with
of pollution. Nitrite also may be produced in a standard oxidizing agent, potassium per-
treatment plants, water distribution systems manganate. The persistence of a definite pink
or cooling systems as a result of the action of color for one minute is taken as the endpoint.
bacteria or other organisms on ammonia ni- APPARATUS REQUIRED.
trogen. 1-Burette, automatic, 25 ml
1-Casserole, 210 ml
Sodium nitrite can be used as a corrosion
1-Cylinder, graduated, 10 ml
inhibitor in open circulating cooling water
1-Stirring rod
systems. While nitrites are sometimes regarded
1-Timing device
as reducing agents, they do possess oxidizing
properties and probably for this reason tend CHEMICALS REQUIRED.
to render metal surfaces passive to corrosive Potassium Permanganate, N I 100
attack in weakly alkaline solutions. In acid Sulfuric Acid, 5%
solutions, however, nitrites do not render Note-Where facilities are available for accu-
iron passive. rate dilution, Nl10 potassium permanganate
may be used. This stronger solution is diluted
In order to establish a protective iron oxide in the ratio of 100 ml of Nl10 potassium per-
film, it may be necessary to feed sodium ni- manganate to 900 ml of distilled water, thus
trite initially at a greater concentration than preparing 1000 ml of N I 100 potassium per-
normally required. The rate of feed can be manganate solution.
reduced to normal once the protective film PROCEDURE FOR TEsT. Measure a 10 ml sam-
has been established. For most effective re-
ple in the graduate, cool to room temperature
sults, sodium nitrite should be fed continu-
and transfer to the casserole. Measure 3 ml
ously to the system. Decomposition problems of the 5% sulfuric acid in the 10 ml graduate
frequently make it difficult to maintain effec-
and add the sample in the casserole. Use a
tive nitrite concentrations in the circulating
100 watt bulb directly over the casserole for
water without excessive feed rates.
illumination. Add the standard N I 100 potas-
Both oxidizing and reducing agents tend sium permanganate from the burette drop by
to destroy nitrite and certain bacteria also drop to the sample in the casserole, stirring
cause its decomposition. Where nitrite con- constantly until a pink color starts to persist.
centrations can be continuously maintained At this point of the determination add the
at an effective level, satisfactory protection permanganate 1 ml at a time with constant
of ferrous metals is secured. Nitrite is not stirring. Continue to add the permanganate
very effective in the prevention of attack on at the rate of 1 ml until a definite pink color
copper or copper alloys. persists for one minute. The time interval of
one minute should be measured with a sec- potassium permanganate multiplied by 23.
ond hand or a stop watch. LIMITATIONS OF TEsT. This method is rapid
CALCULATION OF RESULTS. and adaptable for control purposes where so-
Formula: dium nitrite is used as a corrosion inhibitor. It
ppm nitrite as N02 = ml N/1 00 is affected by any oxidizable substances in the
water such as orga:qic matter, sulfides, hydro-
230
potassium permanganate X - - - - gen sulfide and mercaptans. The presence of
ml sample these interfering substances may cause the ni-
Using a 10 ml sample, the nitrite in parts per trite obtained from this titration to be shown
million as N02 is equal to the ml of N /100 as a higher value than actually present.
386 BETZ HANDBOOK
61
Oxygen (Dissolved)
oxygen in water is caused by the water comes in contact. In cold and hot
D IS SOLVED
the solubility of atmospheric oxygen.
The atmosphere consists essentially of nitro-
water lines, failure of piping may occur and
the lines may become blocked with the prod-
gen and oxygen and while both these gases ucts of corrosion. "Red water" and iron
are soluble to a certain extent in water, stains in process work may result because of
nitrogen is an inert gas and of minor import- the iron brought into solution by the corrosive
ance. The term "dissolved oxygen" represents attack of dissolved oxygen. Increased tem-
the amount of oxygen gas actually dissolved peratures and low pH values accelerate oxy-
in a water and is in no way related to the gen attack.
combined oxygen present in the water mole- In boiler systems, corrosion may result in
cule, H20. feed lines, heaters, economizers, boilers, steam
Dissolved oxygen is present in all surface and return lines. In cooling water systems,
waters and rain waters due to their contact corrosion of heat exchangers, engine jackets,
with the atmosphere. If sufficient agitation pumps and piping may result.
has taken place in contact with air, the dis- In the majority of corrosion problems, the
solved oxygen content of the water will cor- dissolved oxygen content of the water is the
respond to the saturation value for the tem- principal factor influencing corrosion. While
perature involved. The solubility of oxygen corrosion may take place in the absence of
in water follows Henry's law which states that oxygen, the usual procedure in the investiga-
the concentration of dissolved gas in solution tion of any corrosion problem is to first ob-
is directly proportional to the partial pressure tain data on the dissolved oxygen content
of that gas in the free space above the liquid. throughout the system.
The mineral constituents of the water Elimination of the corrosive effect of dis-
slightly affect the solubility of oxygen. Dis- solved oxygen can be accomplished by both
tilled water will absorb more oxygen than will direct and indirect means. The direct means
waters containing higher solids content. Sea involves the actual removal of dissolved
water, because of its high solids content, will oxygen from the water by mechanical or
dissolve less oxygen than will fresh waters. chemical deaeration and such direct action
Well waters usually contain less dissolved is usually applied to boiler feedwater systems.
oxygen than surface waters and in deep well Indirect means are employed where the re-
supplies, dissolved oxygen may be absent. moval of oxygen is not feasible either from
In addition to the natural occurrence of a technical or economic standpoint such as
oxygen in water supplies, aeration is fre- in recirculating cooling water systems. In
quently employed in the removal of other such cases, the indirect means employed in-
gases such as carbon dioxide and hydrogen volves the use of corrosion inhibitors which
sulfide. Efficient aeration results in saturation exhibit a passivating influence on the metal
of the water with dissolved oxygen. surface. Alkaline chemicals may be used for
Dissolved oxygen is objectionable in water the development of a protective scale film to
used for industrial purposes because of the prevent contact of the dissolved oxygen with
corrosive effect on iron and steel with which the protected surface.
Removal of dissolved oxygen from water in application and is theoretically satisfactory

by mechanical deaeration is based on Henry's for only one temperature. Close control is re-
and Dalton's Laws and involves raising the quired to maintain the desired protection
water to the boiling temperature and con- against corrosion without development of an
tinuously venting the mixture of gases and undesirable formation of heavy scale.
steam from a properly designed heater. This
is the method of oxygen removal employed
DISSOLVED OXYGEN, WINKLER METHOD
in the preparation of water for boiler feed
purposes. Vacuum deaeration may be used THEORY OF TEsT. The determination of dis-
on occasions where an increase in water tem- solved oxygen in water is based on the ab-
perature is undesirable. sorption of oxygen by a flocculent precipitate
Chemical deaeration is usually accom- of manganous sulfate and alkaline potassium
plished by the use of sodium sulfite, which iodide. The oxygen reacts with manganous
reacts with oxygen to form sodium sulfate. At hydroxide to form manganese hydroxide.
cold water temperatures, catalyzed sodium Following acidification by sulfuric acid, free
sulfite is required in order to secure suffi- iodine is released in direct proportion to the
ciently rapid reaction between sulfite and amount of oxygen absorbed. The free iodine
oxygen. is then titrated with standard sodium thio-
The use of a chromate salt is an example sulfate in the presence of starch indicator.
of one of the indirect means taken for the The disappearance of the blue color is taken
elimination of the corrosive effects of oxy- as the endpoint.
gen. By formation of a mixed oxide film on APPARATUS REQUIRED.
the metal surface, that surface is rendered 1-Bottle, ground-glass stoppered, 10 oz.
passive to oxygen attack. A continuous sup- 2-Burette clamp
ply of passivating agent to the metal surface 1-Flask, Erlenmeyer, 250 ml
is required in order to heal any breaks in the 1-Burette stand, iron
film. Because of the high concentration of 1-Thermometer, armored (0-220 F)
passivating agent required for the prevention 2-Pipette, safety bulb, 1 ml
of corrosive attack, this method of treatment 1-Pipette, safety bulb, 2 ml
is usually only applicable to recirculating 1-Microburette, self-filling, 2 ml
cooling water systems. Combination of phos- 1-Cylinder, graduated, 100 ml
phate with chromate in the Dianodic process 1-Tubing, glass, 10 inches
permits the use of low treatment concentra- 1-Cooling coil, copper, 10 feet
tions and extends the field of application for 1-Tubing, sulfur free rubber, 10 feet
corrosion inhibitors. 1-Stirring rod, glass
A principle used for preventing the attack CHEMICALS REQUIRED.
of dissolved oxygen in municipal pipe lines, Sulfuric Acid, 50%
once through cooling water systems, etc., in- Manganous Sulfate
volves rendering the water supersaturated Alkaline Potassium Iodide
with respect to calcium carbonate by the ad- Sodium Thiosulfate, N I 10
dition of alkaline treatment chemicals. Con- Starch Indicator
trol is usually based on Langelier's Saturation PREPARATION OF Nl100 SomuM THIOSUL-
Index. The theory behind this system of treat- FATE SoLUTION. Sodium thiosulfate, Nl100,
ment relies on the formation of a thin film is not stable due to absorption of carbon di-
of calcium carbonate preventing contact of oxide from the air, and should be freshly
the dissolved oxygen with the surface to be prepared (or restandardized) at least every
protected. If not carefully controlled, heavy two weeks. To 90 ml distilled water (boiled
scale may develop. This method of treatment, and cooled to room temperature) add 10 ml
similar to the use of sodium silicate, is limited N I 10 sodium thiosulfate solution. Mix thor-
388 BETZ HANDBOOK
oughly and place in tightly stoppered bottle. overflow the outside glass container.
Avoid unnecessary exposure of the solution PROCEDURE FOR TEsT. Fill a 1 ml pipette
to air. with manganous sulfate solution. To prevent
PREPARATION OF SAMPLING EQUIPMENT. If the introduction of air to the sample be cer-
the sample is above 70 F a copper coil should tain a drop of manganous sulfate hangs from
be employed. The inlet side of the cooling the tip of the pipette. Insert the tip of the
coil should be connected to the sampling pipette through the water layer and well into
point by means of a brass nipple, brass valve the sample bottle. Permit exactly 1 ml of
and suitable length of copper tubing. All of manganous sulfate to flow into the sample.
these connections should be of a size equiv- Slowly withdraw the pipette and immediately
alent to the size of tubing comprising the coil. repeat this procedure with alkaline potassium- ·
Iron should not be used. The discharge side iodide solution, employing a second clean 1
of the cooling coil should be connected to a ml pipette.
gl~ss tube with a convenient length of sulfur Wet the glass stopper and gently drop
free rubber tubing. The entire system described straight into the neck of the sample bottle
above must be air tight. The sample must be without removing the sample bottle from
secured continuously from a point of the the outer container. Allow the stopper to
system at which the pressure is greater than "seat" by its own weight, then press down
atmosphere. · firmly. Remove the sample bottle from the
Place the cooling coil in a bucket or simi- outer container and mix the contents by
lar container into which a regulated flow of gently rotating. Do not shake. Examine for
cooling water can be discharged and the air bubbles and if present the sample is
overflow run to waste. worthless and the entire procedure must be
Secure a clean glass (or metal) container repeated. If no air bubbles are present, re-
of such proportion that the capacity is ap- place the tightly stoppered bottle in the glass
proximately 3 or 4 times the volume of the container and allow to stand three minutes
10 oz glass stoppered bottle used in this test. to permit the floc to settle.
The height of the container should be at least Carefully remove the stopper (under water)
1 inch higher than the height of the bottle and add exactly 2 ml of 50% sulfuric acid
with the stopper inserted. into the bottle following the procedure for
SECURING THE SAMPLE. Place the 10 oz bot- pipetting outlined above. Again replace the
tle in the center of the glass container. In- stopper. Remove the stoppered bottle and
troduce the glass tube into the 10 oz bottle. rotate gently to dissolve floc. The sample is
Place the thermometer in the bottle. now "fixed" and ready to titrate. The titra-
Turn on the sampling line and by means tion must be completed within five minutes
of a valve adjust rate of flow into the bottle to minimize errors from any interference.
so that the bottle overflows approximately Measure 200 ml of the fixed sample and
once per minute. Make certain no air bubbles pour into a flask. If the sample is colored
are being discharged into the sample. Con- yellow add Nl100 sodium thiosulfate from
trol the cooling water to the coil in the the microburette with constant swirling, until
bucket to reduce and maintain the tempera- the yellow color is almost discharged. Add 1
ture of the sample to 70 F or less. The cooled ml (approximately 20 drops) of starch in-
sample will overflow the sampling bottle and dicator. The sample should turn blue. Con-
in turn will continue to fill and overflow the tinue to add N I 100 sodium thiosulfate until
outer container. Remove the glass tube from one final drop turns the solution from blue to
the bottle and place it in the outside space colorless. This is taken as the endpoint. Re-
between the bottle and the outer container. cord to the second decimal the total number
Remove the thermometer in the same fash- of ml of N I 100 sodium thiosulfate required.
ion. The sample should continue to run and If the 200 ml fixed sample is not colored yel-
low and does not turn blue upon addition of determination are of utmost importance.
the starch indicator, the dissolved oxygen is This method for determining dissolved
recorded as "zero by the Winkler Method". oxygen is suitable for rapid determinations
CALCULATION OF RESULTS. where the greatest precision and accuracy
Formula: are not required. An accuracy of approxi-
cc per liter dissolved oxygen = mately 0.05 ppm may be expected. The
1000 method is not recommended for dissolved
ml N/100 sodium thiosulfate X - - - - oxygen concentrations below 0.1 ppm and is
ml sample not suitable for checking performance guar-
X .056 antees of deaerating feedwater heaters.
Using a 200 ml sample the dissolved oxy- A colorimetric procedure for the deter-
gen in cc per liter is equal to the ml of NjlOO mination of dissolved oxygen in low concen-
sodium thiosulfate multiplied by 0.28. trations, 0 to 100 ppb, is available. This
Formula: method of testing uses the dissolved oxygen
ppm dissolved oxygen = present in the sample to oxidize a reduced
1000 solution of indigo carmine. As the reduced
ml N/100 sodium thiosulfate X - - - - indigo carmine is oxidized, the color changes
ml sample
progressively from yellow through various
X .08
shades of orange, pink and red to purple.
When a standard amount of reduced indigo
Using a 200 ml sample the dissolved oxy- carmine is present in a known volume of
gen in parts per million is equal to the ml of sample, the resulting color indicates the
N/100 sodium thiosulfate multiplied by 0.4. amount of dissolved oxygen in parts per
LIMITATIONS OF TEST. Nitrites, sulfites, fer- billion.
ric iron and certain types of organic matter Where more precise methods for dissolved
interfere with this test. The proper securing oxygen determinations are required, it is
of the sample, free from contamination by recommended that ASTM Tentative Method
air, and the proper technique in making the D888-49T be consulted.
Figure 61-1 • Test Set for Dissolved Oxygen- Winkler Method

390 BETZ HANDBOOK
62
pH
defined, pH is the logarithm of liter are liberated or 10 times the amount of
E
XACTLY
the reciprocal of the hydrogen ion con- a pH of 7. In a similar manner, a pH of 5
centration. A more simple explanation is indicates 10 times the number of hydrogen
that the pH is a number between 0 and 14, ions liberated as compared to those available
d~noting various degrees of acidity or alka- at a pH of 6 and 100 times the number as
linity. Neutral water has a pH of 7. Values compared to a pH of 7. In short, each num-
below 7 and approaching 0 are increasingly ber on the pH scale is 10 times the concen-
acid while values from 7 to 14 are increas- tration of hydrogen or hydroxyl ions when
ingly alkaline. compared with its adjoining number. The
Pure water ionizes to produce hydrogen or following table will serve to further clarify
acid ions (H+) and hydroxyl or alkaline ions this point.
(OH-) as illustrated.
H20 = H+ + OH- H-ion Concentration
When pure water ionizes in this manner (gm Ionizable H+ per liter}
0.0000001 grams of hydrogen ion are liber- pH
ated per liter. This number, which can also
be written as 1 X 10-7 is inconvenient to 0.0 1.0 10 0
1.0 0.1 1Q-1
handle, but on the pH scale is written simply 2.0 0.01 1Q-2
as 7. 3.0 0.001 1Q-3
4.0 0.0001 1Q-4
5.0 0.00001 1Q-5
OH- 6.0 0.000001 1Q-6
7.0 0.0000001 1Q-7
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 8.0 0.00000001 1Q-8
9.0 0.000000001 1Q-9
On the pH scale a value of 7 is exactly 10.0 0.0000000001 1Q-10
11.0 0.00000000001 1Q-11
neutral. Proceeding to the left on the dia- 1Q-12
12.0 0.000000000001
gram and approaching "zero", indicates an 13.0 0.0000000000001 1Q-13
increasing number of hydrogen ions present 14.0 0.00000000000001 1(}-14
and increasing acidity. Proceeding to the
right on the scale and more closely approach-
ing a pH value of 14 indicates an increasing The pH of most natural waters will fall
number of hydroxyl ions and increasing alka- within the range of 6.0-8.0, although more
linity. acid conditions and lower pH will result
The further the pH is located from the where the water contains high concentrations
neutral point of 7, the greater is the concen- of free carbon dioxide or acid mine drainage.
tration of hydrogen or hydroxyl ions, but not A pH above 8.0 is seldom encountered except
in direct arithmetical proportion to the dis- where pollution by alkaline trade wastes
tance between each number. A pH of 7 exists or where the water has been chemically
means 0.0000001 grams of ionizable hydrogen treated such as by the lime-soda process.
are liberated per liter. A pH of 6 means Since pH is a measure of the relative
0.000001 grams of ionizable hydrogen per acidity or alkalinity of a water, it is a most
important factor influencing scale forming exchangers, economizers, boilers, turbines,

or corrosive tendencies. A low pH wiii give etc., pH is one of the most important control
rise to corrosion of equipment with which factors.
the water comes in contact. High pH values
may cause precipitation of calcium carbonate pH, COLORIMETRIC METHOD
from solution as a scale on the surfaces of
pipe lines, heat exchange equipment, con- THEORY OF TEsT. The numerical expression
densers, etc. In the calculation of the stabil- of pH value denotes the degree of acidity or
ity of a water in accordance with Langelier's alkalinity of a sample. A pH of 7.0 represents
equation, the determination of pH is re- the neutral point; values below 7.0 indicat-
quired. ing acidity and those above 7.0 indicating
alkalinity. The colorimetric determination of
In clarification processes, probably the
pH is based on the use of different indicators,
most important determination is pH since
each indicator covering a portion of the pH
the control of coagulation is primarily a
range. Each indicator gives different color
matter of pH control, with every coagulant
reactions from the indicator immediately
possessing an optimum pH range for most
adjacent to it and gives different shades of
efficient operation. In lime-soda softening,
color over its own range.
pH is not usually required for control pur-
poses because the proper pH is automatically APPARATUS AND CHEMICALS REQUIRED.
established with correct control of phenolph- 1-Taylor pH Comparator
thalein and methyl orange alkalinity. In sof- 2-Adjustment tubes
tening for calcium removal, however, pH Standard Slides and Indicators
control is quite important in order to insure
proper calcium removal without also pre- pH Range of Slide Indicator
cipitating magnesium hydroxide. 0.2- 1.8 Acid Cresol Red
For zeolite softening employing siliceous 1.2- 2.8 Acid Meta Cresol Purple
2.4- 4.0 Benzo Yellow
exchange materials, it is generally considered 3.0-- 4.6 Bromphenol Blue
that the pH of the influent water should be 3.8- 5.4 Bromcresol Green
in the range of 7.0-8.3 and between 4.3-8.3 4.4- 6.0 Methyl Red
5.2- 6.8 Chlorphenol Red
where carbonaceous zeolite materials are em- 6.0-- 7.6 Bromthymol Blue
ployed. Some newer ion exchange resins can 6.8- 8.4 Phenol Red
tolerate high pH and high temperature. De- 7.2- 8.8 Cresol Red
7.6- 9.2 Meta Cresol Purple
velopment of exchange materials of this type 8.0-- 9.6 Thymol Blue
made possible the use of the hot lime-hot ion 8.6-10.2 Phthalein Red
10.0--11.6 Acyl Red
exchange process. 11.0--12.6 Parazo Orange
Control of pH is an important factor in 12.0--13.6 Acyl Blue
such processes as iron removal, recarbonation
and acid treatment. The pH value of a PROCEDURE FOR TEsT. The determination of
boiler water is usually adjusted to a minimum pH preferably should be made on a freshly
of 10.5 to prevent acidic corrosion and to obtained sample, tightly stoppered to prevent
provide a sufficiently high pH for the precipi- contact with air. The sample should not be
tation of the scale forming salts. filtered. If stored, samples should be kept
In general, it can be stated that wherever in Pyrex or other resistant glass bottles.
a water analysis is to be interpreted with Plastic bottles such as polyethylene also are
respect to scale forming or corrosive tenden- suitable.
cies, such an analysis is incomplete without a Fiii three 5.0 ml test tubes to the mark
determination of pH. In the control of with sample and place in the comparator
various treatment processes and in the pre- base. To the middle test tube, add 0.5 ml of
vention of scale and corrosion in piping, heat indicator by means of the calibrated dropper
392 BETZ HANDBOOK
and mix thoroughly. (Do not place thumb indicators wherever possible.
over end of test tube.) Place the color LIMITATIONS OF TEsT. Particularly with
standard slide on the base and holding the weakly buffered solutions it is .essential that
instrument toward a source of daylight, move the proper indicator be employed as the pH
the slide in front of the test samples until a of the indicator may change the· pH of the
color match is obtained. solution under test. Indicators and slides
CALCULATION OF RESULTS. When a match should be chosen so that the pH of the
is obtained, the pH is then read directly from sample is in the middle of the range. A color
the values on the color standard slide. match with the first or last standard in any
PROCEDURE FOR STANDARDIZATION OF INDI- slide should never be takeh as an accurate
CATORS. The pH of indicator solutions may determination. Certain specific substances in
change due to direct sunlight, long exposure the water to be tested, such as free chlorine
to air and acid or alkaline fumes. It is pos- and hypochlorites, affect the color produced
sible to check this point and restandardize by the indicator and may yield erratic results.
th~s~ indicators in aqueous solution. Under proper conditions the sensitivity of the
The indicators in aqueous solutions are: colorimetric pH method is approximately 0.1
Bromphenol Blue 3.0-4.6 pH pH.
Bromcresol Green 3.8-5.4 pH
Chlorphenol Red 5.2-6.8 pH pH, GLASS ELECTRODE METHOD
Bromthymol Blue 6.0-7.6 pH
Phenol Red 6.8-8.4 pH THEORY oF TEsT. The pH or hydrogen ion
Cresol Red 7.2-8.8 pH concentration of a solution may be measured
Meta Cresol Purple 7.6-9.2 pH by determining, with a potentiometer, the
Thymol Blue 8.0-9.6 pH voltage developed by two electrodes which
About 0.5 ml of the indicator solution is .are in contact with the solution. The voltage
placed in the larger of the adjustment tubes, of one electrode known as a calomel half-cell
the smaller tube filled with distilled water is fixed, while the voltage of the other elec-
and inserted. A thin layer of indicator is trode varies with the pH of the sample.
thus formed between the two tubes and the Several types of electrodes may be used, but
shade and intensity of this layer should cor- in general a glass electrode is the most ap-
respond to the color standard in the middle plicable for industrial use. A glass surface of
of that pH range. As an example, with brom- the proper composition separating two solu-
thymol blue indicator, the proper pH is 6.8. tions gives potentials which are directly pro-
The pair of tubes is then placed in the com- portional to the pH. Due to the high resist-
parator base and compared with the slide. If ance of the circuit, it is necessary to employ
the pH read is too low, approximately 0.1 N some type of electronic device to amplify the
NaOH is added a drop at a time to 100 ml current.
of stock indicator solution, the solution being APPARATUS REQUIRED.
thoroughly mixed and tested after each addi- 1-Electrometric pH Meter
tion. If the pH read is too high, the same PRoCEDURE FOR TEsT. The exact mechanical
procedure is applied with approximately0.1 N procedure for determining pH will vary
HCI. slightly with the instruments furnished by
The other pH indicators, of the alcoholic different manufacturers. However, in general,
type, cannot be checked by means of adjust- the instrument is first standardized by using
ment tubes. Their accuracy can be checked a buffer solution of known pH. The tempera-
by comparing the indicated pH of a water ture of the sample to be tested is observed
sample versus the pH obtained electrometri- and a temperature correction dial on the
cally. Because the alcoholic indicators cannot instrument is adjusted. The electrodes are
be adjusted, it is advisable to use the aqueous then inserted in the water sample and the
Figure 62-1 • Taylor pH Comparator
pH is read directly on a dial on the instru- In addition to the determination of pH,

ment. an electrometric pH meter can be employed
in titrations in place of the various organic
LIMITATIONS OF TEsT. Glass electrode po- indicators. A titration is carried out to the
tentials are not affected by oxidizing or re- pH value equivalent to the point of color
ducing agents, gases, dissolved organic com- change of the indicator. (Example: Phenol-
pounds, colloids or suspended matter. The phthalein indicator changes at approximately
pH of viscous solutions or highly colored pH 8.3.) This method of titration is particu-
solutions may be obtained by this method. ljlrly useful where the sample is highly
Accuracy is superior to colorimetric methods. colored, masking color endpoints.
In high concentrations of sodium ions, an Electrometric pH meters are available
error is introduced which can be corrected either as battery operated or line operated
by suitable curves, or by standardizing the models. In general, the line operated unit
instrument for that range. will prove most satisfactory.
394 BETZ HANDBOOK
63
Phosphate
phosphate ion rarely occurs naturally tricalcium phosphate, an insoluble precipi-
T
HE
in a raw water. The determination of tate. However, it is possible to form two
phosphate usually is made in order to con- soluble forms of calcium phosphate-mono-
trol chemical treatment containing phosphate. calcium phosphate and dicalcium phosphate.
While the phosphate test is employed most In other words, in order to completely pre-
frequently in boiler water conditioning, the cipitate the calcium hardness, it is necessary
determination of ortho and polyphosphate is to form tricalcium phosphate. Otherwise,
used also as a control test in the treatment of part of the hardness will remain in solution
water in cooling systems, distribution lines, as monocalcium and dicalcium phosphate.
etc. where polyphosphates are employed for In order to accomplish this precipitation, it is
scale and corrosion control. necessary to maintain a pH value of 9.5 or
In cooling water systems, the surface active higher.
properties of the polyphosphates are em- The soluble phosphate which is deter-
ployed to advantage in preventing the pre- mined by the phosphate test is sometimes
cipitation of calcium carbonate. By use referred to as the "excess" phosphate con-
of these materials and organic surface active tent. The soluble phosphate is that phosphate
agents as well, it is possible to retain calcium which is present over and above that re-
carbonate in solution at a positive saturation quired to precipitate the calcium hardness.
index which would otherwise result in scale It is generally desirable to maintain a soluble
formation. In water distribution lines, munic- phosphate content of approximately 30-60
ipal systems, etc., the polyphosphates are ppm to drive the reaction of calcium and
used for control of tuberculation and some phosphate to completion, and to assure the
reduction in corrosion. The metal savings immediate removal of any calcium hardness
through the use of low concentrations of the entering the boiler in the feedwater. How-
polyphosphates is not outstanding, but these ever, operating conditions may alter the con-
materials are useful in controlling tubercula- centrations of soluble phosphate which must
tion and head losses. Phosphate, in combina- be maintained in the boiler water. For ex-
tion with chromate, is widely used in corro- ample, when it is not economically feasible
sion control in open recirculating cooling to externally remove high silica concentra-
water as an essential component of the Di- tions present in the boiler feedwater, the
anodic method. This combination of two formation of silica scale can be minimized by
anodic inhibitors, in the proper ratio and carrying soluble phosphate concentrations in
under controlled pH conditions effectively the boiler water.
inhibits pitting and tuberculation and pro- The phosphate type internal treatment,
vides superior protection compared to the use while possessing many advantages when prop-
of either of these inhibitors alone. erly employed also has the disadvantage of
In boiler feedwater conditioning, the phos- increasing the tendency of a boiler water to
phate test is employed for control of internal carryover. However, it is not the soluble or
conditioning. Under favorable conditions excess phosphate measured in the phosphate
phosphate will react with calcium to form test which increases the tendency to carry-
over. Instead, it is the small particles of cal- color slide on the base and holding the in-
cium phosphate precipitate which tend to strument toward a source of daylight, move
stabilize boiler water foam. the slide in front of the test samples until a
color match is obtained.
ORTHOPHOSPHATE, COLORIMETRIC METHOD If the sample is highly colored, as with or-
ganic present, cplor comparison will be af-
THEORY oF TEsT. This test is based on the fected. Instead of filling the two side tubes
formation of phosphomolybdic acid through with filtered sample, these blanks should be
the reaction of the molybdate reagent with prepared by mixing 5 ml of filtered sample,
the phosphate present in the water. The 10 ml of molybdate reagent and 2.5 ml of
phosphomolybdic acid is then reduced by distilled water.
stannous ion to give a blue color, the inten-
sity of which is proportional to the amount (Range 0-25 ppm as P04)
of phosphate present. Prepare the filtered sample as described pre-
APPARATUS REQUIRED. viously. Use a mixing tube graduated at 10
1-Taylor Phosphate Comparator (slide and 14 ml. Fill the mixing tube to the first
range 5-100 ppm P04) mark ( 10 ml) with the filtered sample. Add
1-Funnel, glass, 3" diameter molybdate to the second mark ( 14 ml) . Stop-
1-Filter paper, Whatman No. 5, 12.5 em per and mix well. Add 1 level brass dipperful
1-Mixing tube (5, 15, 17.5 ml, graduated) of stannous reagent. Stopper and mix well. A
CHEMICALS REQUIRED. blue color will develop, the intensity of which
Molybdate Reagent is proportional to the phosphate present.
Stannous Reagent, Dry (with dipper) Read the phosphate value in 1 minute.
PRoCEDURE FOR TEST. The temperature of CALCULATION OF RESULTS. When a match is
the sample should be in the range of 70- obtained, the phosphate in parts per million
100 F. Filter the sample through No.5 What- as P04 is then read directly from the values
man filter paper. Discard the first 10-20 ml on the color standard slide.
of filtrate since there is slight adsorption of LIMITATIONS OF TEsT. This test is unaffected
phosphate by fresh filter paper. Filter the by most of the ions in the concentrations
sample until clear, pouring the filtrate back normally encountered in boiler and cooling
through the filter paper if necessary to obtain water. Up to 150 ppm chromate as Cr04 can
a clear filtrate. It is important that no sus- be tolerated. It is essential that the sample
pended matter be present in the filtered sam- taken for test be free of suspended precipi-
ple. tates since calcium phosphate, if present,
(Range 5-100 ppm as P04) would be placed in solution by the acidic
Use a mixing tube graduated at 5, 15 and nature of the reagents. The phosphate value
17.5 ml. Fill the mixing tube to the lowest thus obtained would be erroneously high due
mark (5 ml) with the filtered sample. Add to the phosphate precipitate dissolved in this
molybdate reagent to the second mark ( 15 manner.
ml) . Add distilled or phosphate-free water to The use of a photometer with essentially
the third mark ( 17.5 ml). Stopper and mix the same procedure will provide additional
well. Add 1 level brass dipperful of stannous accuracy. Where phosphate in relatively low
reagent. Stopper and mix well. A blue color concentration is to be determined, as in cer-
will develop, the intensity of which is pro- tain cooling water applications, the photom-
portional to the phosphate present. Read the eter procedure is preferred for close control.
phosphate value in 2 minutes.
Place the mixing tube in the middle com- TOTAL PHOSPHATE, PHOTOMETRIC METHOD
partment in the base. Fill the two test tubes
on either side with filtered sample. Place the THEORY OF TEsT. This test is based on the
396 BETZ HANDBOOK
formation of phosphomolybdic acid through as temperature can affect the results. Each
the reaction of the molybdate reagent with curve should be checked with phosphate-
the phosphate present in the sample. The free water and also at a dilution of the phos-
phosphomolybdic acid is then reduced by the phate standard that approximates the middle
stannous ion to yield a blue color which is of the phosphate range covered by the curves.
proportional to the phosphate content of the It is very important that the "check" samples
sample which is measured by comparison of are analyzed at the same time, under the
the photometer dial reading with a calibra- same conditions and treated in the same man-
tion curve. ner as the actual water samples. Do not omit
APPARATUS REQUIRED. any of the steps such as the conversion pro-
Filter photometer complete with assorted cedure, etc.
laboratory glassware. Phosphate must be determined on a filtered"
CHEMICALS REQUIRED. sample, using a filter paper such as Whatman
No. 5. Discard the first 10-20 ml of filtrate
Hydrochloric Acid, Concentrated, C.P.
since there is a slight adsorption of phos-
Molybdate Reagent
phate by fresh filter paper.
Phenolphthalein Indicator
Stannous Chloride, Crystals To Determine Orthophosphate. Prepare a
Standard Phosphate Solution, 45 ppm P0 4 "zero" reference blank by adding a beaker
Sodium Hydroxide, 7N 5 ml of clear sample, 10 ml molybdate re-
Sulfuric Acid, 50% agent and 2.5 ml distilled water. Use this
solution to set the photometer at "zero" im-
PRoCEDURE FOR TEsT. A fresh concentrated mediately prior to test.
stock of solution of stannous chloride should
To a second beaker add 5 rnl clear sam-
be prepared once each month. For this pur-
ple, 10 ml molybdate reagent and 2.5 ml
pose add 12 gms stannous chloride crystals to
dilute stannous reagent. Allow to stand one
88 gms of C.P. concentrated hydrochloric
acid. Store in an amber bottle away from ( 1) minute and then immediately obtain
photometer dial reading.
light. Keep container tightly closed.
To Determine Total Phosphate. In order
The dilute stannous chloride reagent used
to determine total phosphate, it is necessary
in this test must be prepared fresh daily. The
to convert all polyphosphate in the sample to
dilute reagent consists of 1.0 ml concentrated
orthophosphate. Place two 25 ml clear sam-
stannous chloride diluted to a total volume
ples in separate 125 ml Erlenmeyer flasks.
of 40 ml with distilled water.
One sample is to be used as a blank and the
This procedure employs a wave length of other for analysis. To each flask add 2.5 ml
610 mu and a light path of 5 mm. Prepare 50% sulfuric acid. Boil both the blank and
calibration curves for the photometer using the sample vigorously for at least 30 minutes.
successive dilutions of the phosphate stand- Add distilled water periodically so that the
ard to adequately cover the range of phos- volume does not fall below 5 ml. If the
phate in the samples to be tested. Two curves volume does fall below 5 ml, the sample
are required-one for orthophosphate and must be discarded. If it is not convenient to
one for total phosphate. The dilutions of the observe the sample continuously during boil-
standard should be treated in exactly the ing, reflux condensers may be employed.
same manner as that shown below for analy- Cool the blank and the sample to room
sis of the water samples. temperature. Add 3 drops phenolphthalein
Each time a determination is made the indicator to each flask and neutralize with
calibration curves should be checked to es- 7N sodium hydroxide (approximately 5 ml
tablish a correction factor. This procedure is will be required) until a faint permanent
necessary to insure that the results are ac- pink appears. Add 50% sulfuric acid, drop
curate since reagent age and stability as well by drop, until the solutions turn colorless.
Figure 63-1 • Phosphate Comparator
Since some heat may be generated during from their respective calibration curves. The
neutralization, re-cool and then adjust the polyphosphate concentration is obtained by
volumes to exactly 25 ml with distilled water. subtracting the value for orthophosphate
A precipitate may form at this point but do from the value for total phosphate.
not filter.
LIMITATIONS OF TEST. This procedure is not
The total phosphate now can be deter- affected by most of the ions normally present
mined. Measure 5 ml of the blank and 5 ml in water. Up to 150 ppm chromate at Cr04
of the sample after shaking to be sure a can be tolerated. Up to 40 ppm phosphate
representative amount of any precipitate is can be determined without dilution of the
included. Proceed in exactly the same man- sample. Concentrations above 40 ppm can
ner as shown for orthophosphate. be determined by diluting the original sam-
CALCULATION OF RESULTS. The orthophos- ple with distilled water prior to treatment
phate and total phosphate values in parts following the procedure outlined, and mul-
per million as P04 are obtained directly tiplying result by the appropriate factor.
398 BETZ HANDBOOK
64
Silica
can be present in water in two systems.
S
ILICA
distinct and sep~rate forms. It is con- The presence of silica is particularly ob-
ventionally expressed in a water analysis as jectionable in water used for boiler feed
silicon dioxide ( Si0 2 ) • Actually, silica exists purposes as it may lead to the formation of
in water in both the crystalloidal and colloi- hard dense scales which have unusually high
dal Jorms. In the colloidal form, silica is not resistance to heat transfer. In boilers oper-
in solution, but is in suspension in a very ating at relatively low pressures, the silicate
finely divided form. This form of silica usu- scale formed is mostly calcium silicate. In
ally can be removed by proper coagulation such instances, softening of the water for re-
and filtration. Silica in the crystalloidal or moval of the calcium ions usually is sufficient
soluble form, however, is extremely difficult to control such incrustation. In higher pres-
to remove and it is in this form that it oc- sure boilers, the solution to the problem is
casions the greatest difficulty with scale for- not so easily obtained. The silica present will
mation. combine in such instances with other ele-
Silicon is the second most abundant chemi- ments forming a hard scale even though cal-
cal element found on the earth and, there- cium and magnesium are not present in any
fore, this element in the form of at least one appreciable amounts in the boiler feedwater.
of its many compounds is found in greater or Complex sodium alumino-silicate scales have
lesser amounts in all natural waters. The been identified in high pressure boilers and
solid crust of the earth is made up of ap- the formation of these scales may take place
proximately 80 to 90% silicates or other com- in high pressure boiler operation even with
pounds of silicon. Water passing through or a relatively low hardness and silica content
over the earth's surface dissolves silica as one of the boiler feedwater.
of its impurities. In addition to the difficulty of boiler scale
The silica content of natural waters will formation, a very serious problem encount-
vary to a considerable extent depending on ered in high pressure operation is in the dep-
the locality. There is a general tendency, how- osition of siliceous materials on turbine
ever, for silica to be higher in waters where blades and in superheaters. Serious loss on
the hardness is low and the alkalinity some- turbine efficiency results from insoluble silica
what high. The Pacific Northwest and many turbine blade deposits caused by vaporization
Southern states exhibit waters of this charac- of silica from the boiler water.
ter, as do certain sections of New England. Soluble silica is relatively inert and cannot
Where waters are high in alkalinity, there be precipitated by other chemicals to form
are a large number of sodium ions present; insoluble compounds, as can be done with
such ions combine with silicon forming a most of the inorganic chemicals present as
soluble sodium silicate. water impurities. It is this fact which has
Silica may cause difficulty by forming cal- made the problem of silica removal rather
cium and magnesium silicate scale where difficult. Successful processes for silica re-
water is used in once through heat exchange moval involve adsorption of silica from solu-
processes and in recirculating cooling water tion. Hot process silica removal by magne-
sium salts is a process frequently employed. Place the instrument so that the compara-
Silica also may be removed by the use of tor tubes point toward a source of daylight.
highly basic anion exchange resins in con- Read instrument by viewing the reflection in
junction with demineralization. the mirror and not by looking down through
the tubes. Move the slide in front of the test
SOLUBLE SILICA, TAYLOR COMPARATOR sample until a color match is obtained.
METHOD If the silica exceeds 25.0 ppm, discard test.
Take a smaller sample, dilute with distilled
THEORY OF TEST. This test is based on the water to 100 ml and proceed as before, mul-
development of the yellow silico-molybdate tiplying observed silica value by the proper
color produced by the reaction of ammonium factor to obtain parts per million.
molybdate in acidic solution with the silica
The color due to phosphates can be des-
present in the water sample. The color de-
troyed by oxalic acid. When phosphates are
veloped is matched with silica standards and
present, proceed as above, permitting the
the silica content in ppm is read directly
maximum color to develop in the Water
from the slide.
Analyzer tube ( 1-2 minutes). Then, add 1
APPARATUS REQUIRED. ml of 10% oxalic acid, mix and read as usual.
1-Taylor Water Analyzer complete (silica CALCULATION OF RESULTS. Using a 100 ml
slide range 0.0-25.0 ppm) sample, silica as Si0 2 is read directly in parts
2-Flasks, Erlenmeyer, 250 ml per million from the slide.
1-Cylinder, graduated, 50 ml LIMITATIONS OF TEsT. This method does not
1-Pipette, Mohr, 5 ml determine suspended silica; tannin and re-
2-Pipettes, Mohr, 1 ml ducing substances interfere. Strong mineral
CHEMICALS REQUIRED. acids will prevent the development of maxi-
Ammonium Molybdate Reagent mum color. Hydrogen sulfide interferes and
Hydrochloric Acid, 50% must be removed by boiling.
Oxalic Acid, 10%
SOLUBLE SILICA, PHOTOMETRIC METHOD
PROCEDURE FOR TEST. Add 1 ml of hydro-
chloric acid and 4 ml of ammonium molyb- HIGH RANGE (4·40 ppm)
date solution to 100 ml of clear sample. If THEORY OF TEsT. This method tests for solu-
the sample is not clear before addition of re- ble silica by formation of the yellow silico-
agents it should be filtered. Stir well and molybdate complex followed by reduction to
after allowing the sample to stand for 1 the molybdate blue color. The blue color
minute (do not allow the sample to stand produced is proportional to the silica content
longer than 10 minutes) transfer enough of of the sample and is measured by the pho-
the sample to one of the comparator tubes tometer dial reading.
to bring the level of the liquid up to the APPARATUS REQUIRED.
mark on the tube. A yellow color will de- Filter photometer complete with assorted
velop if silica is present. Place this tube in the laboratory glassware.
middle compartment of the Analyzer. The
tubes on each side should contain the same CHEMICALS REQUIRED. -l(·
water as that being tested to blank out any Hydrochloric Acid Reagent
color present in the sample. In order to Ammonium Molybdate Reagent (high
compensate for the dilution in volume range)
through addition of reagents, the blank on Sodium Sulfite Reagent
either side of the sample being tested should Silica Standard, 50 ppm Si02
be prepared by the addition of 1 ml hydro- *All reagents should be stored in silica-free con-
chloric acid and 4 ml of distilled water per tainers such as polyethylene bottles.
100 ml sample. PROCEDURE FOR TEsT. This procedure em-
400 BETZ HANDBOOK
ploys a wave length of 690 mu and a cell phosphate, possesses an advantage over com-
with a light path of approximately 10 mm. parator and photometric methods employing
Prepare a calibration curve for the photom- the yellow silica-molybdate color.
eter using successive dilutions of the silica Up to 40 ppm silica can be determined
standard to adequately cover the anticipated without dilution of the sample. Concentra-
range of silica in the samples to be tested. tions above 40 ppm can be determined by
The dilutions of the standard should be diluting the original sample with distilled
treated in exactly the same manner as that water, following the procedure outlined, and
shown below for analysis of the water sam- multiplying result by the appropriate factor.
ples.
This method is not sufficiently accurate for
Each time a determination is made the the determination of silica in condensed
calibration curve should be checked to estab- steam samples where silica vapori2lation may
lish a correction factor. This procedure is be a problem. The low range silica procedure
necessary to insure that the results are ac- is required for such samples, covering the
curate since reagent age and stability as well range of approximately 0.0-3.0 ~pm silica.
as te~perature can affect the results. The
curve should be checked with silica-free
water and also at a dilution of the silica SOLUBLE SILICA, PHOTOMETRIC METHOD
standard that approximates the middle of LOW RANGE (0.0·3.0 ppm)
the ·silica range covered by the curve. All re- THEORY OF TEST. The low range silica test
agents as well as the "check" samples and also relies on the measurement of the molyb-
the actual sample to be analyzed must be at denum blue color from the photometer dial
the same temperature. readings. However, amino-naphthol-sulfonic
Prepare a "zero" reference blank. To a acid is employed as the reducing agent in-
beaker, add 10 ml of the clear sample, 5 ml stead of sodium sulfite.
hydrochloric acid reagent, 5 ml distilled water APPARATUS REQUIRED.
and 10 ml sodium sulfite reagent. Use this Filter photometer complete with assorted
blank to set the photometer at "zero" im- laboratory glassware
mediately prior to the test of a sample.
CHEMICALS REQUIRED.*
To a second beaker add 10 ml sample, 5 Ammonium Molybdate Reagent (low
ml hydrochloric acid reagent and 5 ml of range)
ammonium molybdate reagent. Allow to Oxalic Acid, 3%
stand at least one minute, but not more than Amino-Naphthol-Sulfonic Acid
five minutes. Then add 10 ml of sodium sul- Silica Standard, 50 ppm Si02
fite reagent. Allow to stand exactly one *All reagents should be stored in silica-free con-
minute and immediately obtain dial reading. tainers such as polyethylene bottles.
CALCULATION oF REsULTS. The silica in PROCEDURE FOR TEsT. This procedure em-
parts per million as Si02 is obtained by ployes a wave length of 690 mu and a cell
reference to the silica calibration curve. with a light path of 40 mm. Prepare a cali-
LIMITATIONS OF TEsT. This procedure is not bration curve for the photometer using suc-
affected by phosphate, iron, sulfate, chloride cessive dilutions of the silica standard to
and organic matter such as tannins. Alka- adequately cover the anticipated range of
linity above 350 ppm as CaC0 3 interferes silica in the samples to be tested. The dilu-
slightly depending on the silica content. tions of the standard should be treated in
Phosphate above 150 ppm gives slightly low exactly the same manner as that shown be-
results. The effect of high alkalinity and low for analysis of the water samples.
high phosphate can be completely eliminated Each time a determination is made the
by testing a diluted sample. This method, by calibration curve should be checked to estab-
eliminating the interference of tannins and lish a correction factor. This procedure is
necessary to insure that the results are ac- tometer at "zero" immediately prior to the
curate since reagent age and stability as well test of a sample.
as temperature can affect the results. The To a second beaker, add 50 ml sample, and
curve should be checked with silica-free 5 ml ammonium molybdate reagent. Allow
water and also at a dilution of the silica to stand approximately 5 minutes. Add 5 ml
standard that approximates the middle of oxalic acid reagerit. Wait approximately one
the silica range covered by the curve. All re- minute and then add 5 ml sulfonic acid re-
agents as well as the "check" samples and agent. Allow to stand exactly two minutes
the actual sample to be analyzed must be at and immediately obtain dial reading.
the same temperature. CALCULATION OF RESULTS. The silica con-
The amino-naphthol-sulfonic acid reagent centration in parts per million as Si02 is ob-
used in this test is not stable and should be tained by reference to the prepared silica
prepared once each week. Dissolve 1.0 g of calibration curve.
1-amino, 2-naphthol, 4-sulfonic acid in 4.5 LIMITATIONS OF TEsT. This procedure is
ml of lN sodium hydroxide. Add with 60 g primarily intended for the determination of
sodium bisulfite and 2 g sodium sulfite to silica in the low concentrations that may be
900 ml distilled water. Dilute to 1.0 liter with present in condensed steam samples. The
distilled water. method is not affected by the presence of
Prepare a "zero" reference blank. To a ions such as phosphate, in the concentrations
beaker, add 50 ml of the clear sample, 5 ml which would be present in samples of such
oxalic acid, 5 ml sulfonic acid and 5 ml dis- purity. A precision of approximately 0.02
tilled water. Use this blank to set the pho- ppm silica can be secured.
402 BETZ HANDBOOK
65
Solids
(Total, Dissolved, Suspended)
solids are those solids which quality based on the turbidity test effectively
S
USPENDED
are not in true solution and which can limit the suspended solids content.
be removed by filtration. Total solids repre- The dissolved solids of a natural water are
sent the sum of the dissolved and suspended usually composed of the sulfate, bicarbonate
solids. The value for total solids is frequently and chloride of calcium, magnesium and
defined as Total Residue. sodium. Each of these individual ions may
The origin of the dissolved solids present produce an effect, specific to that ion, de-
in a natural water supply lies in the solvent pendent on whether the water is employed
action of water in contact with the minerals in process work, or as boiler feed or cooling
in the earth. Suspended solids are contributed water.
by small particles of insoluble matter, me- Limits on the total or dissolved solids con-
chanically introduced by turbulent action of tent of a water may be imposed where the
the water on the soil. Both suspended and cumulative effect of various ions may make
dissolved solids may be introduced by do: a water undesirable even where no individual
mestic and industrial wastes. constituent has exceeded a tolerable limit.
The U. S. Public Health Service Drinking For example, where water is used in beverage
Water Standards recommend that the total manufacture or ice making, it is customary
solids of a potable supply be limited to 500 to stipulate limits on the dissolved solids
ppm for a water of good quality. However, content of the water. In boiler feedwater
these standards also state if such water is not conditioning, the American Boiler Manufac-
available, a total solids content of 1000 ppm turer's Association guarantees on steam purity
may be permitted. While the standards are specify certain maximum total solids content
so worded with respect to total solids that no of the boiler water, dependent on the opera-
limits are placed directly on dissolved and ting pressure. Even though no individual ion
suspended solids, a practical limit on sus- may have reached a dangerous concentra-
pended solids content is covered by the addition, the additive effect of the various boiler
tional restriction that the turbidity of the water constituents may produce a tendency
water should not exceed 10 units. for carryover.
Suspended solids are objectionable in proc- The removal of suspended solids is accom-
ess work, boiler feed and cooling water con- plished by the processes employed for the re-
ditioning. Since the determination of sus- moval of turbidity, namely the use of sub-
pended solids is a gravimetric procedure, it sidence, coagulation and filtration.
is customary to set up standards of water Dissolved solids, depending on their char-
quality on the basis of the more rapid tur- acter, may be reduced by various treatment
bidity test. Since turbidity is caused by the processes. Sodium zeolite softening, which re-
presence of suspended solids, limits on water places calcium and magnesium ions with
sodium ions, does not lower the dissolved duce the turbidity of the filtrate to less than
solids content. Dependent on the water char- 1 unit and then 100 ml of the filtered sample
acteristics, sodium-hydrogen zeolite softening is evaporated to dryness as under the total
and lime-soda softening may affect a con- solids test.
siderable reduction in dissolved solids. For Suspended solids may be calculated as the
complete removal of dissolved solids from a difference betw~en total and dissolved solids.
water however, it is necessary to employ de- Suspended Solids.
mineralization or distillation.
Suspended solids may be determined di-
rectly by filtering the sample using a Gooch
TOTAL, DISSOLVED AND SUSPENDED SOLIDS, crucible and an asbestos mat. Sufficient
GRAVIMETRIC METHOD sample should be filtered to yield 50-100 mg
suspended residue. Dry one hour at 103 C,
THEORY OF TEsT. These tests are based on cool in a desiccator until the weight is
the evaporation of a water sample to dryness constant. This direct determination of sus-
under standard conditions and the weighing pended solids is to be preferred for accuracy
of the residue after drying. to the calculation of suspended solids.
APPARATus REQUIRED. A well equipped lab- LIMITATIONS OF TEST. Although a 100 ml
oratory is necessary for these determinations. sample was specified in the determination of
PROCEDURE FOR TEsT. total and dissolved solids, it is necessary to
alter the size of the sample taken in order to
Total Solids.
obtain a weighable residue of sufficient mag-
Evaporate to dryness in a weighed plat- nitude. The size of the sample taken for test
inum dish on a steam or water bath, 100 ml should be such that the residue after evap-
of a thoroughly shaken, unfiltered sample. oration amounts to 50-100 mg. Special pre-
Dry the residue at 103 C for one hour. The cautions are necessary in the determination
increase in weight of the platinum dish in mg of the solids content of condensed steam
multiplied by 10 equals ppm total solids. samples. Special apparatus is required and
Dissolved Solids. sufficient sample is taken to yield a residue of
The sample should be filtered so as to re- approximately 25 mg.
404 BETZ HANDBOOK
66
Sulfate
HE sulfate ion is present in all natural In cooling water conditioning, calcium sui- .
T waters due to the solvent action of
water on the minerals in the earth. The
fate scale is prevented by maintaining the
calcium sulfate in solution by the use of sur-
quantity of sulfates present in natural waters face active materials and by limitation of
will vary considerably in cliff erent sections concentrations.
oCthe country depending upon the mineral In process water where it is sometimes
content of the soil in any particular locality. necessary to remove the sulfate ion, this can
Some industrial wastes such as acid mine be accomplished through demineralization.
drainage and pickling liquors are high in sul-
fates, contributing to the natural sulfate con-
SULFATE, THQ METHOD
tent of the raw water.
The U. S. Public Health Service Drinking THEORY oF TEsT. This method is based on
Water Standards recommend that potable the precipitation of sulfate by barium
water contain less than 250 ppm of sulfate chloride in alcoholic solution in the presence
as so4. of THQ indicator. As soon as the sulfate has
The principal objection to the sulfate ion been completely precipitated, further addi-
in water is its ability to combine with cal- tion .of barium chloride will react with the
cium to form calcium sulfate scale. The most THQ indicator resulting in a color change
common constituent of boiler scale in an un- from yellow to rose red.
treated boiler is calcium sulfate. While cal- APPARATUS REQUIRED.
cium sulfate scale is not as common in cool- 1-Burette, automatic, 25 ml
ing water systems, it can occur when high 1-Cylinder, graduated, 50 ml
concentrations of calcium and sulfate exist 1-Flask, Erlenmeyer, 125 ml
simultaneously. 1-Measuring cup, brass
Sulfate may be either beneficial or detri- 1-Pipette, delivery, 25 ml
mental in waters used for manufacturing. In 2-Bottles, with droppers, 8 oz
the brewing of beer, calcium sulfate aids in 1--Pipette, Mohr, 5 ml
producing desirable flavor. Conversely, in the CHEMICALS REQUIRED.
ice industry calcium sulfate may cause the Barium Chloride, 1 ml = 1 mg S04
formation of white butts in the ice core. Barium Chloride, 1 ml = 4 mg S04
In the conditioning of boiler feedwater, it Hydrochloric Acid, N/10
usually is not necessary nor practical to re- Phenolphthalein Indicator
move sulfate from the water. If the hardness Bromcresol Green Indicator
is removed or controlled, the sulfate scale THQ (Tetrahydroxyquinone) Indicator
cannot form. In other words, the prevention Isopropyl Alcohol, 99%
of calcium sulfate scale is possible by the Silver Nitrate, 2%
removal of either the calcium or sulfate. In Sodium Hydroxide, N/10
industrial water conditioning, it is much sim- PROCEDURE FOR TEsT. Measure 25 ml of clear
pler and more economical to remove the sample into the Erlenmeyer flask. Add four
calcium than the sulfate. drops of phenolphthalein indicator and by
Figure 66-1 • Sulfate Determination- THQ Method

406 BETZ HANDBOOK
means of the dropper neutralize with N/10 indicator, acidifying with the N /10 hydro-
hydrochloric acid until the red color just dis- chloric acid until the green color is just dis-
appears. If the sample contains less than 50 charged to a straw color. Then 'proceed as
ppm total alkalinity, a more dilute hydro- outlined. By this modification, the titration
chloric acid can be employed for greater ac- is carried out at a lower pH at which
curacy in neutralization. If the sample is a up to 60 ppm phosphate as P0 4 does not
natural water containing no phenolphthalein interfere. For samples containing higher con-
alkalinity, adjust with sodium hydroxide and centrations of phosphate, a 5 or 10 ml sample
hydrochloric acid so as to just neutralize to should be diluted to 25 ml with distilled
the acid side of phenolphthalein. water, neutralized and titrated as indicated
After neutralization, add one level meas- above. The results should be multiplied by
ure (approx. 0.2 g) THQ Indicator (for sam- the appropriate factor.
ples containing less than 50 ppm sulfate, only CALCULATION OF RESULTS.
~ dipper of THQ Indicator should be em-
Formula:
ployE:d}. Swirl flask to dissolve indicator.
Then, add 25 ml of isopropyl alcohol (or ppm sulfate as 504 = (ml barium chloride
- ml blank) X strength BaCb mg/ml X
ethyl alcohol denatured by Formula 3A or
1000
30). Titrate the resulting yellow solution
with barium chloride solution, with thorough ml sample
shaking, until the color changes from a yel- Using a 25 ml sample and barium chloride
low to a rose red. of the strength 1 ml = 1 mg so4, the sulfate
To Sharpen Endpoint-The sharpness of in parts per million is equal to the ml of
the endpoint is increased by a small amount barium chloride required minus the blank of
of silver nitrate. Use 1.0 ml of 2% silver 0.2 ml, multiplied by 40. When using a 25 ml
nitrate solution when chlorides are low and sample and barium chloride of the strength
3.0 ml of silver nitrate solution when the 1 ml = 4 mg so4, sulfate in parts per million
chlorides are high. The silver nitrate is added as so4 is equal to the ml of barium chloride
during the titration. required multiplied by 160.
Caution-If silver nitrate is used in excess LIMITATIONS OF TEsT. The THQ method for
of chloride, an intense cherry color will de- sulfate determination is not affected by
velop. When this color is produced by the chlorides, silica, sulfites, tannins, soluble
introduction of silver nitrate, discard and magnesium, soluble calcium and ferric or
repeat the titration with less silver nitrate. ferrous iron. Phosphate up to 60 ppm can be
For natural waters, use barium chloride tolerated using the modified method, and
solution 1 ml = 1 mg so4 and subtract a above this concentration by dilution and the
blank of 0. 2 ml from the barium chloride modified method. It is advisable not to em-
titration. For boiler waters or other waters ploy the THQ method for sulfates lower
high in sulfate, use barium chloride solution than 20 ppm as SO •. Temperature has little
1 ml = 4 mg S0 4 • Using this strength solu- effect below 95 F.
tion no blank is necessary. For extremely Sulfate also may be measured by the use
high sulfate concentrations use a 5 or 10 ml of a photometer. The analysis is conducted
sample, diluting to 25 ml with distilled water at a 415 IIlLI wave length with a light path
and multiplying the results by the appropri- of 20 mm. The method is based on the de-
ate factor. velopment of turbidity which is proportional
Modification When Phosphate is Present. to the sulfate content, by precipitation of the
When soluble phosphate is present in the sulfate as barium sulfate. This procedure is
sample, as is the case with many boiler not affected by the ions normally present in
waters, use bromcresol green indicator in the water and sulfate up to 40 ppm can be
neutralization in place of phenolphthalein determined without dilution of the sample.
67
Sulfite
determination of sulfite usually is tions imposed by the dissolved oxygen con-
T
HE
made only on boiler waters or on waters tent of the feedwater. If appreciable quanti-
that have been treated with catalyzed sodium ties of dissolved oxygen are permitted to
sulfite for corrosion prevention. Generally enter the boiler, costs will be high when re-
speaking, sulfite is not present in natural lying entirely upon sodium sulfite as a means
waters. of oxygen removal. The safest practice is to-
In boiler feedwater conditioning, sodium remove as much oxygen as possible by me-
sulfite is fed to a boiler to remove dissolved chanical means, through the use of efficient
oxygen and thus prevent pitting. Under types of boiler feedwater heaters and then to
proper conditions sulfite will react with dis- maintain a residual sulfite in the boiler water
solved oxygen to form sulfate, thus removing to react with any oxygen which has not been
the dissolved oxygen from the system. In mechanically removed.
order for this reaction to proceed rapidly and For the prevention of corrosion and pitting
completely, it is necessary to maintain an ex- in feed lines, closed heaters and economizers,
cess sulfite concentration at an elevated tem- it is desirable that the sodium sulfite be fed
perature. To assure the rapid and complete continuously to the boiler feedwater rather
removal of any dissolved oxygen entering a than direct to the boiler. Reaction between
boiler in the feedwater, a soluble (or excess) sulfite and dissolved oxygen is not instantan-
sulfite content of at least 20 ppm is generally eous and the completion of this reaction is
maintained in the boiler water. aided by the longer contact period provided
Theoretically, it requires 7.88 pounds of by feeding sodium sulfite through the boiler
chemically pure sodium sulfite to remove one feed line. The preferred point of application
pound of oxygen. Sodium sulfite supplied for of the sodium sulfite is to the storage com-
the purpose of oxygen removal need not be partment of the feedwater heater.
chemically pure unless products such as cer- The development of catalyzed sodium sul-
tain foods are being processed. The average fite has increased considerably the field of
purity of the technical grade usually em- usefulness of this material for oxygen re-
ployed for oxygen removal is about 90% moval. Catalyzed sodium sulfite will react
sodium sulfite. The efficiency of the oxygen almost instantaneously with dissolved oxygen
removal is estimated at 75% to allow for even at cold water temperatures. Because of
oxidation in contact with air, blowdown this property, catalyzed sodium sulfite is find-
losses, etc. Therefore, it is estimated that 10 ing increased use in the treatment of cooling
pounds of commercial sodium sulfite are re- water, process water, distribution systems,
quired for each pound of oxygen removed etc., for prevention of oxygen corrosion. In
(or 10 ppm sulfite per 1 ppm dissolved the prevention of economizer corrosion, it
oxygen). frequently is more desirable to employ cata-
The use of sodium sulfite as a chemical lyzed sodium sulfite in order to obtain the
deaerant is economical within certain limita- advantage of rapid oxygen removal.
408 BETZ HANDBOOK
SULFITE, POTASSIUM IODIDE·IODATE METHOD dipperful of sulfite indicator at a time to the

sample. Stir thoroughly between each addi-
THEORY OF TEsT. This test is based on the tion of sulfite indicator (all of the indicator
titration of a water sample containing sulfite may not dissolve creating a slight haze in the
by means of a standard potassium iodide- sample). Continue to add indicator in this
iodate solution. Free iodine is produced when manner until the red color disappears. It is
the sulfite has been completely oxidized, re- not necessary to exactly neutralize the phenol-
sulting in a blue color in the presence of a phthalein alkalinity. When the sample is
starch-type indicator. This blue color is taken colorless then add one (1) additional dipper-
as the endpoint. ful of sulfite indicator and stir.
APPARATUS REQUIRED. Titrate with potassium iodide-iodate solu-
1-Burette, automatic, 25 ml tion until a faint permanent blue color de-
1-Casserole, porcelain, 210 ml velops in the sample. This color change is
1-Cylinder, graduated, 50 ml taken as the endpoint. Record the ml potas-
· 1'--Stirring rod, glass sium iodide-iodate solution used.
1-Measuring dipper (plastic) CALCULATION oF REsULTS.
CHEMICALS REQUIRED. Formula:
Phenolphthalein Indicator ppm sulfite as sol = 500
Potassium Iodide~Iodate, 1 ml = 0.5 mg ml potassium iodide-iodate X
m1 samp 1e
so3 Using a 50 ml sample, the parts per million
Sulfite Indicator
of sulfite as SOs equals ml potassium iodide-
PROCEDURE FOR TEsT. The water sample iodate required multiplied by 10.
should be freshly obtained with as little con- LIMITATIONS OF TEsT. This method is rapid
tact with air as possible. Do not filter. Cool and adaptable to field determinations. It is
to room temperature ( 70-80 F) . affected by any oxidizable substances in the
Measure 50 ml of the water sample with water such as organic matter, sulfides and
the graduated cylinder and transfer to the nitrites. The presence of these interfering
casserole. Add 3-4 drops of phenolphthalein substances may cause the sulfite obtained
indicator to the sample. The sample will turn from this titration to be shown as a higher
red. Use the plastic dipper to add one ( 1) value than actually present.
68
Turbidity
can be interpreted as a lack of struments.
T
URBIDITY
clearness or brilliance in a water, but Water taken from. a river or turbulent
should not be confused with color. A water stream usually contains appreciable quantities
may be dark in color but still clear and not of turbidity. The water from lakes and ponds
turbid. generally is less turbid because of the settling
Turbidity is due to suspended matter in a action which takes place in these bodies.
finely divided state. Clay, silt, organic matter, Springs and wells usually are low in turbidity
microscopic organisms and similar materials because of the filtering action of the ground
are contributing causes of turbidity. through which the water flows.
Although the terms "suspended matter" The U. S. Public Health Service Drinking
and "turbidity" are closely allied, they are Water Standards recommend that the tur-
not synonymous. Suspended matter is the bidity of potable water be less than 5 units.
quantity of material in a water which can The turbidity of industrial water should
be removed by filtration. Turbidity is a mea- be as low as possible. This is particularly true
surement of the optical obstruction of light of boiler feedwater. The turbidity, which is
passed through a water sample. caused by small particles in suspension, will
The standard unit of turbidity formerly concentrate in the boiler water and may settle
was defined as the optical obstruction of out in the form of a heavy sludge or "mud".
light caused by 1 ppm of insoluble silica This condition can be minimized to some ex-
(diatomaceous earth or Fuller's earth) in tent by increase in blowdown. The turbidity
distilled water. However, the lOth edition of the boiler feedwater also limits the cycles
( 1955) of "Standard Methods" abandoned of concentration carried in the boiler since
the terms "silica scale" and "parts per mil- one of the factors limiting these concentra-
lion" for turbidity. The primary standard for tions is the suspended solids of the boiler
turbidity is the Jackson candle turbidimeter water. Finely divided particles introduced
and it is recommended that turbidity read- into the boiler water by the turbidity of the
ings be expressed as follows: boiler feedwater also may be responsible for
a foaming and priming condition. Process
Turbidities shall be recorded water, particularly in the food, dyeing and
greater than: but less than: to the nearest:
units units units paper industries should have a low turbidity
or the quality of the finished products manu-
0.0 1.0 0.1
I 10 I factured may be affected.
10 100 5 Turbid makeup water to cooling water
100 400 10
400 700 50
systems may cause plugging and overheating
700 100 where solids settle out on heat exchange sur-
faces. Corrosive action may be increased
Turbidity suspensions standardized by the under such deposits which hinder penetration
Jackson candle method may be used in the of corrosive inhibitors.
calibration of other turbidity measuring in- Turbidity may be partially removed by
410 BETZ HANDBOOK
sample itself, the use of standard solutions

and their preparation is eliminated.
APPARATUS REQUIRED.
1-Hellige Turbidimeter, complete with
calibrated curves for turbidity
PROCEDURE FOR TEST. First thoroughly shake
the sample of water and then pour it into
one of the glass cells to the mark on the side.
Insert the plunger carefully, avoiding forma-
tion of bubbles on the under side. Place the
cell on the mirror in the slot provided. Close
the door of the instrument and then turn on
the light.
View turbidity through the ocular, and
revolve the dial on the side of the instrument
until the black circle matches the surround-
ing field. The dial reading is noted.
This method covers a turbidity range of 0
to 100 units without dilution of the water
Figure 68-1 • Hellige Turbidimeter sample. This range is covered by use of
various size cells and filters, a standard curve
being furnished for each set of conditions.
CALCULATION OF REsULTs. The number of
subsidence, but it usually is necessary to em-
units of turbidity are determined by compari-
ploy coagulation and filtration in order to
son of the dial reading with the calibrated
produce a treated water of low turbidity.
curve.
LIMITATIONS OF TEsT. Turbidity varies not
TURBIDITY, HELLIGE METHOD
only with the nature of the suspended solids
THEORY OF TEsT. Turbidity of water results present, but with their particle size. There-
from finely divided suspended matter. Mea- fore, there is no relation between turbidity
surement of turbidity is based on the optical and gravimetrically determined suspended
obstruction of light rays passed through a solids. For turbidities above 100 units, dilu-
sample when compared, under the same tion with "zero" turbid water is necessary.
conditions, with an arbitrary standard tur- Dial readings should be taken as soon as
bidity scale. The Hellige Turbidimeter is possible to avoid error introduced by the
based on the comparisoP of a beam of light settling of large particles.
with the Tyndall effect produced ·from the Due to the inability of a photometer to
lateral illumination of a specimen by the blank out the initial color of a sample, this
same light source. As the method involves a instrument normally is not recommended for
comparison of Tyndall effect on the test the measurement of turbidity.
69
Composition -of
Prepared Reagents
~
AGENT grade chemicals should be used BARIUM CHLORIDE, 1 ml = 4 mg so4
in the preparation of all solutions. These A solution of approximately 1.02% barium
reagents should be prepared only in a well chloride in distiiied water, standardized so
equipped laboratory, under the supervision of that 1 ml = 4 mg so4.
an experienced chemist. It is advisable to BARIUM CHLORIDE, 10%
purchase these solutions from a laboratory A 10.0% solution of barium chloride m
supply house if complete laboratory facilities distilled water.
are not available for their preparation. BROMCRESOL GREEN INDICATOR
ACETONE REAGENT A solution prepared by triturating 0.4 g
A 50% by volume solution of acetone m bromcresol green in 3-5 ml distilled water,
distilled water. then diluting to 25 ml with distiiied water.
ALKALINE PoTASSIUM IoDIDE Add 1.6 ml of 1N sodium hydroxide, dilute
to 1.0 liter and adjust pH to 4.6+0.1 with
Dissolve 700 g potassium hydroxide in
electrometric pH meter.
700 ml distilled water and cool. Dissolve
150 g potassium iodide in 200 ml dis- BRUCINE REAGENT
tilled water and cool. Mix the two solu- A 5.0% solution of brucine alkaloid
tions and dilute to 1.0 liter with distilled (C2aH2sN204•4H20) in reagent grade
water. chloroform.
AMMONIUM MoLYBDATE REAGENT (silica, CALCIUM INDICATOR
high range - Taylor and photometer A dispersion of 0.20 g ammomum pur-
methods) purate in 100 g c.p. potassium chloride,
A 10.2% solution.of ammonium molybdate ground so it wiii pass through a 30 mesh
(81% Mo0 3 ) in distilled water. Note- screen.
This solution will become cloudy from time CHROMATE STANDARD, 10 ppm as Cr0 4
to time and any turbidity should be filtered A standardized solution of potassium chro-
prior to use. mate, 10 ppm as Cr0 4 •
AMMONIUM MoLYBDATE REAGENT (silica, HARDNESS BUFFER REAGENT
low range-photometer method) The dry hardness buffer is prepared using
A 4.0% solution of ammonium molybdate special solvents, mixing, evaporation and
(81% Mo0 3 ) in 10% by volume con- grinding. Special equipment is necessary.
centrated hydrochloric acid. A liquid buffer reagent, stable for approxi-
BARIUM CHLORIDE, 1 ml = 1 mg so4 mately three months, can be prepared by
A solution of approximately 0.255% bar- dissolving 10 g sodium hydroxide, 10 g
ium chloride in distiiled water, standard- sodium sulfide, 40 g sodium tetraborate
ized so that 1 ml = 1 mg S04. (Na 2B40 7 •10H20) and 10 g sodium potas-
412 BETZ HANDBOOK
sium tartrate (NaKC 4H40 6 •4H20) in a ride in 200 ml distilled water and neutral-
liter of distilled water. In the hardness test, ize to pH 8.0 with 20% sodium hydroxide.
0.5 ml of the liquid buffer is used. Dilute to 600 ml with distilled water and
HARDNESS INDICATOR add 100 ml glacial acetic acid. I>issolve
The dry hardness indicator is prepared 200 g aluminum nitrate (Al2 (N0 3 ) 3 '
using special solvents, mixing and grinding. 9H20) in 150 ml distilled water and add.
Special equipment is necessary. Dissolve 4 g Sequestrene N a 2 in 50 ml
A liquid hardness indicator, stable for ap- distilled water and add. Dilute to 1.0 liter
proximately three months, can be prepared with distilled water.
by dissolving 10 g Eriochrome Black T in IRoN STANDARD SoLUTION, 1 ml = 10ppmFe
300 ml of 0.2% sodium carbonate solution. Prepare a concentrated stock solution by
Then dilute to 1.0 liter with isopropyl dissolving 7.022 g of crystallized ferrous
alcohol (99%), adjusting pH to 9.0 with ammonium sulfate (FeSO. (NH4) 2SO..•
0.1 N sodium carbonate. In the hardness 6H2 0) in 500 ml distilled water and 200
test, use 4-6 drops of the liquid indicator. ml concentrated sulfuric acid. Completely
HARDNEss TITRATING SoLUTION, 1 ml = oxidize with 0.1 N potassium permanga-
1 mg CaCOa nate and dilute to 1.0 liter with distilled
water. Dilute this stock solution as required
A distilled water solution of 4.0 g disodium
ethylenediaminetetracetate dihydrate and to produce an iron standard solution,
0.86 g sodium hydroxide per liter, stand- 1 ml = 10ppmFe.
ardized so that 1 ml = 1 mg CaC0 3 • MANGANESE SPECIAL SoLUTION
Dissolve 75 g mercuric sulfate in 400 ml
HYDRAZINE STANDARD SoLUTION, 1 ml =
0.1 mg N 2 H4 concentrated nitric acid and 200 ml dis-
tilled water. Add 200 ml 85% phosphoric
A solution of 0.328 g hydrazine hydro-
acid and 0.035 g silver nitrate. Cool and
chloride (N2H 4•2HC1) in 100 ml distilled
dilute to 1.0 liter with distilled water.
water and 74 ml concentrated hydro-
chloric acid. Dilute to 1.0 liter with dis- MANGANESE STANDARD SoLUTION, 1 ml =
tilled water. 10 ppm Mn
HYDROCHLORIC Acm, N/50 Prepare an aged 0.1 N potassium per-
manganate solution. Calculate the quantity
A solution of hydrochloric acid in distilled
of this solution required to prepare 1.0
water standardized to 0.02N.
liter so that 1 ml = 10 ppm Mn. Acidify
HYDROCHLORIC Acm, N/10 the calculated volume of permanganate
A solution of hydrochloric acid in distilled with 2-3 ml concentrated sulfuric acid.
water standardized to 0.1N. Add sodium bisulfite solution dropwise ( 10
HYDROCHLORIC Acm, 5 N g dissolved in 100 ml distilled water) until
A solution of hydrochloric acid in distilled the permanganate color disappears. Boil
water standardized to 5.0 N. to remove S02. Cool and dilute to 1.0
HYDROCHLORIC Acm 10% liter with distilled water.
A 10.0% by volume solution of concen- MANGANous SuLFATE SoLUTION
trated hydrochloric acid in distilled water. A 48% solution of manganous sulfate
HYDROCHLORIC Acm, 50% . (MnS0 4 •4H20) in distilled water.
A 50% by volume solution of concentrated METHYL ORANGE INDICATOR
hydrochloric acid in distilled water. A 0.05% solution of methyl orange in dis-
HYDROCHLORIC Acm REAGENT tilled water.
A 2.0% by volume solution of concentra- METHYL PuRPLE INDICATOR
ted hydrochloric acid in distilled water. A proprietary dispersion of an organic dye
HYDROXYLAMINE REAGENT sensitive to essentially the same pH range
Dissolve 100 g hydroxylamine hydrochlo- as methyl orange indicator. (Fleisher
Chemical Co.) Na2COa.

MoLYBDATE REAGENT (for phosphate) SILVER NITRATE, 1 ml = 1 mg Cl
A solution of sodium molybdate in 5.0% An approximately 0.03 N solution of silver
by volume sulfuric acid standardized to nitrate in distilled water, standardized so
0.145 N. that 1 ml = 1 mg Cl.
NESSLER REAGENT SILVER NITRATE, 1 ml = 5 mg Cl
Dissolve 100 g of mercuric iodide and 70 g An approximately 0.14 N solution of silver
of potassium iodide in a small quantity of nitrate in distilled water, standardized so
ammonia-free water, and add slowly, with that 1 ml = 5 mg Cl.
stirring, to a cool solution of 160 g sodium
hydroxide in 500 ml of ammonia-free SILVER NITRATE, 2%
water. Dilute to 1.0 liter with ammonia- A solution of 2.0% silver nitrate in distilled
free water. water.
0RTHOPHENANTHROLINE REAGENT SoDIUM ARSENITE, 0.5%
Dissolve 1.0 g orthophenanthroline mono- A 0.5% solution of sodium meta arsenite
hydrate in 1.0 liter distilled water. in distilled water.
0RTHOTOLIDINE REAGENT SoDIUM CARBONATE, N j50
A solution prepared by adding 50 ml of a A solution of sodium carbonate in distilled
0.270% solution of orthotolidine dihydro- water, standardized to 0.02 N.
chloride in distilled water to 50 ml of a SoDIUM CARBONATE, N /22
30% by ·volume solution of concentrated A solution of sodium carbonate in distilled
hydrocholoric acid. water, standardized to 0.0455 N.
OxALIC Acm, 3% SoDIUM HYDROXIDE, 1.0 N
A 3.0% solution of oxalic acid in distilled A solution of sodium hydroxide in distilled
water. water, standardized to 1.0 N.
OxALIC AciD, 10%
SoDIUM HYDROXIDE, 7 N
A 10.0% solution of oxalic acid in distilled
A solution of sodium hydroxide in distilled
water.
water, standardized to 7.0 N.
PHENOLPHTHALEIN INDICATOR
SoDIUM SuLFITE REAGENT
A 0.5% solution of phenolphthalein m
A solution of sodium sulfite m distilled
50% ethyl alcohol.
water, standardized to 2. 7 N.
PoTASSIUM CHROMATE INDICATOR
SoDIUM THIOSULFATE, N/10
A 5.0% solution of neutral potassium
A solution of sodium thiosulfate in distilled
chromate in distilled water.
water, standardized to 0.1 N.
PoTASSIUM IoDIDE-IODATE, 1 ml = 0.5 mg
STANNoUs REAGENT, Dry
SOs
A solution of 0.31 g sodium bicarbonate, A dispersion of 2.5% stannous sulfate.
4.35 g potassium iodide and 0.45 g potas- STARCH INDICATOR
sium iodate per liter, standardized to A 0.5% solution of c.p. starch in distilled
0.0125 N. water.
PoTASSIUM PERMANGANATE, N/100 STARFAMIC INDICATOR
An aged solution of potassium perman- A proprietary dispersion of cold water sol-
ganate in distilled water, standardized to uble starch in a crystalline acid medium.
NjlOO. (U.S. Patent No. 2,963,442).
SILICA STANDARD, 50 ppm as Si0 2 SULFITE INDICATOR
A standardized solution of sodium silicate, A proprietary dispersion of cold water sol-
50 ppm as Si02. The sodium silicate is uble starch in a crystalline acid medium.
prepared by fusion of pure Si0 2 and (U.S. Patent No. 2,963,442).
414 BETZ HANDBOOK
SuLFURIC Acm, N/50 trated sulfuric acid in distilled water.

A solution of sulfuric acid in distilled water THQ (TETRAHYDROXYQUINONE) INDICATOR
standardized to 0.02N. A proprietary dispersion of disodium tetra-
SuLFURIC Acm, 1.0 N hydroxyquinone dye.
A solution of sulfuric acid in distilled water TAYLOR HYDRAZINE REAGENT A
standardized to 1.0 N. Paradimethylaminobenzaldehyde powder.
SuLFURIC Acm, 5% TAYLOR HYDRAZINE REAGENT B
A 5.0% by volume solution of concentrated A solution containing 10% concentrated
sulfuric acid in distilled water. hydrochloric acid and 90% isopropyl
SuLFURICAcm, 50% alcohol.
A 50% by volume solution of concentrated XYLENE CYANOLE, 0.02%
sulfuric acid in distilled water. A 0.02% solution of xylene cyanole m
Acm REAGENT
SuLFURIC distilled water.
A 15.0% by volume solution of concen-
Conversion Factors
WATER ANALYSIS
Parts per Parts per Grains per Grains per
Million 100,000 U.S. Gallon Imp. Gallon
part per million = 1.0 0.10 0.058 0.07
part per 100,000 10.0 1.00 0.585 0.70
grain per U.S. Gal = 17.1 1.71 1.000 1.20
grain per Imp. Gal = 14.3 1.43 0.835 1.00
HARDNESS
Parts per Grains per Clark French German
Million U.S. Gallon Degrees Degrees Degrees
part per million = 1.0 0.058 0.07 0.10 0.056
grain per U.S. Gal = 17.1 1.000 1.20 1.71 0.958
Clark degree = 14.3 0.829 1.00 1.43 0.080
French degree 10.0 0.583 0.70 1.00 0.560
German degree = 17.9 1.044 1.24 1.78 1.000
MISCELLANEOUS
1 U.S. Gallon = 0.1337 cu ft
1 U.S. Gallon = 231 cu in
1 U.S. Gallon = 3.785 liters
1 U.S. Gallon (water) = 8.34 lbs
1 cubic foot = 7.48 U.S. Gallons
1 cubic foot (water) = 62.4 lbs
1 part per millon = 8.33 lbs per million gallons
1 grain per gallon = 143 lbs per million gallons
1 lb per million gallons = 0.12 parts per million
1 lb per million gallons = 0.007 grains per gallon
1 lb per thousand gallons = 120 parts per million
1 cu ft per second = 646,316 gallons per day
1 cu ft per second = 448.83 gallons per minute
1 gal per minute = 0.00144 million gallons per day
1 ft head water = 0.434 psi
1 psi = 2.31 ft head water
Of = {°C X 1.8) + 32
°C = {°F - 32) X 0.55
416 BETZ HANDBOOK INDUSTRIAL WATER CONDITIONING
417
Index
AcETONE reagent, 411 closed water systems, 315
Acid open recirculating systems, 313-314
concentrated, handling and storage of, water balances, control of, 315-316
280-281 winter protection, 317-318
construction materials used for equipment Air diffusion method of aeration, 21-22
in contact with, 286 Air washers, 314-315
dilute, handling, 283-284 Algae control, 289-299
discharge, proper points for, 284-285 Alkaline boil-out method of boiler cleaning,
feeding to cooling water systems, 280-287 222-224
constant rate, 281 Alkaline potassium iodide, 411
proportional, 282-283 Alkalinity, 339-342
filters, 285 adjustment with acid, 255
free mineral, 338 difficulties caused by, 16
difficulties caused by, 16 hydrate, 343-344
test for, 338 test for, 343-344
treatment, 16 methyl orange, 256, 339
handling of, safety rules concerning, 286- phenolphthalein, 339
287 test for, 340-342
sulfur burners, use of, 284 treatment, 16
Acid cleaning, of boilers, 224-228 Alum, 30
Acid feed, feeding method, 147 chemical formula, 30
Acid treatment, in scale elimination, 255-257 feeding method, 147
Acrolein, biological fouling control, 293-294 Aluminum sulfate
Aeration, 18-22 chemical formula, 32
applications, 22 commercial strength, 32
defined, 19 grades available, 32
hydrogen sulfide removal by, 56-58 handling materials, 32
iron removal by, 61 weight, 32
limitations, 22 Amines
methods, 19-22 feeding method, 148
theory involved, 19 filming, 202-203
uses, 19 neutralization of carbonic acid by, 200-202
Aerators, 20-21 Ammonia, 345-346
coke tray, 20, 62 chemical formula, 17
forced draft, 20 difficulties caused by, 17
water fall, 20 employed for neutralization of carbon di-
wood slat, 20 oxide, 200
Air conditioning systems, 312-318 methods of removal, 50-51, 130
air washers, 314-315 test for, 345-346
chemical feed methods, 316-317 treatment, 17
Ammonium alum Blowdown

chemical formula, 32 continuous, 213
grades available, 32 defined, 211
handling material, 32 equipment employed, 213-215
weight, 32 manual, 212
Ammonium molybdate reagent, 411 method for limiting cycles of concentra-
Analytical equipment tion, 252
care of, 330 purpose of, 211
recommended, 330 surface, 215
Analytical methods, 329-334 Boiler blowdown control, 216-220
recommended, 329 equipment employed, 213-215
Analytical results, expression of, 335-337 Boiler corrosion control, 168-172
Anion absorbers, hydrogen sulfide removal dissolved oxygen control, 168-171
by, 58-59 idle boilers, protection of, 171-172
Anion exchange resins, 104 pH control, 168
Ani~n exchangers, 104 Boiler metal, embrittlement of, 205-209
Anodic inhibitors, 243, 260, 262 causes, 205-206
Anthracite coal, as filter medium, 36, 38 characteristics of, 206-207
Antifoam agents, use in carryover preven- status of problem, 208-209
tion, 180-181 treatment methods, 207-208
Asbestos, as filter-aid, 44 Boiler scale, effect of, 153-154
Automated feeding equipment, 149-150 Boiler scale control, 152-164
Automatic shot feed, 140 calcium salts, 154-155
chelating agents, 163-164
feeding methods, i6s..:t66
BACTERIA chemical testing and, 164
iron, 65-67 copper oxide, 156
removal by subsidence, 24 identification of scale deposits, 157-158
Barium chloride, 411 internal chemical treatment, 158-163
method for determining hydrate alkalinity, carbonate control, 160
343-344 feeding of chemicals, 164-165
Bentonite organic agents, 160-163
grade available, 32 phosphate control, 158-160
handling material, 32 iron and copper oxides, 156
weight, 32 magnesium salts, 155
Biological fouling, 289-299 mud, silt, etc., 154
acrolein treatment, 293-294 oil, 156-157
bromine treatment, 293 scale formation, mechanism of, 152
chlorinated phenols, use of, 294-295 silica, 155-156
chlorination programs, 292-293 Boilers
chlorine residual treatment, 292 chemical cleaning of, 222-228
chlorine treatment, 291 alkaline boil-out, 222-224
continuous vs. shock treatment, 295-296 hydrochloric acid, 224-226
control of, 289-299 inhibited acid cleaning, 224-226
copper salts, use of, 296-297 phosphoric acid, 226
defined, 289 copper deposits, removal of, 226-227
difficulties due to, 289 idle, protection of, 171-172
quaternary ammonium compounds, use of, Breakpoint chlorination, 49, 50-51
297-298 Bromcresol green indicator, 411
toxicants, miscellaneous, 298 Bromine, biological fouling control, 293
INDEX 419
Brucine reagent, 411 feeding methods, 147, 165-166

Burettes, use of, 327-329 Chemical feed systems, 136-150
Burners, sulfur, 284 See also Feeders; Pumps
Chemical testing, boiler scale control and, 164
Chloramines, 49
CALCITE, as a filter medium, 38 Chloride, 353~355
Calcium, 34 7 chemical formula, 16
test for, 347-348 difficulties caused by, 16
Calcium carbonate protective scale, 240, 259 test for, 354-355
Calcium carbonate scale, prevention of, 232- treatment, 16
235, 252 Chloride-anion exchange, dealkalization by,
Calcium indicator, 411 111-114
Calcium salts, boiler scale control, 154-155 applications, 113
Calcium sulfate, 72 chemistry involved, 111-113
Carbon, as filter-aid, 44 defined, 111
Carbon dioxide, 349-350 equipment required, 113
chemical formula, 16 limitations, 113-114
content of steam, reduction of, 199 use for, 111
correction of, internal methods of, 199-203 Chlorinated phenols, biological fouling con-
corrosive effects of, 197-199 trol, 294-295
difficulties caused by, 16 Chlorination, 48-54
removal of, 130 ammonia removal by, 50-51
sources of, 197 biological fouling control, 292-293
test for, 350, 352 breakpoint, 49, 50-51
treatment, 16 chemistry of, 48-51
Carbonate control, boiler scale, 160 chlorine-ammonia process, 50
Carbonic acid, 130 chlorine requirements, 49-50
Carryover, 175-181 dechlorination, 52-53
causes of, 175-177 defined, 48
chemical, 17 5-1 77 hydrogen sulfide removal by, 58
mechanical, 177 hypochlorination, 52
defined, 175 laboratory control, 53-54
effects of, 17 5 marginal, 50
prevention of, 177-181 Chlorinator, V-notch, 54
antifoam agents, 180-181 Chlorine, 48, 356
chemical causes, 178-180 biological fouling control, 291-293
mechanical causes, 177-178 disinfecting action, theory of, 48-49
silica, cause of, 184 effects of, on tower wood, 293
studies, 181 effects of, 51-52
Cathodic inhibitors, 260 feeding equipment, 53
Cation exchange tests for, 356-358
by hydrogen zeolite, 96-101 Chlorine dioxide, 51
applications, 101 Chlorine gas, 53
equipment arrangements, 98-101 Chromate, 359
limitations, 101 removal of, 266
iron removal by, 63-64 tests for, 359-362
principles of, 96-98 Chromate standard, 411
Caustic soda, 71-72 Chromates, as corrosion inhibitors, 260-261,
feeding method, 147 273
Chelating agents, boiler scale control, 163-164 Chromium Dianodic method of corrosion
control, 265 Controlled volume pumps, 140-142

Clarifiers Conversion factors, 415
sludge contact, 28 Cooling tower wood deterioration, see Wood
upflow, 29 Cooling water systems
Citric acid, use in chemical cleaning of power acid feeding to, 280-287
plant equipment, 227-228 continuous monitoring of, 149
Closed recirculating systems, 271-273 See also Closed recirculating systems; Once-
advantages of, 271-272 through cooling water systems; Open
corrosion control, 272-273 recirculating cooling water systems
Coagulant aids, 33-34 Cooling water treatment
feeding method, 147 closed recirculating systems, 271-273
Coagulants, 30-32 once-through corrosion control, 23 7-246
alum, 30 once-through scale control, 230-235
copperas, 30-31 open recirculating corrosion control, 259-
-ferric, 31 269
m~gnesium oxide, 31-32 open recirculating scale control, 248-257
sodium aluminate, 30 Copper and copper' alloys, corrosion control,
Coagulation, 27-34 268
coagulant aids, 33-34 Copper deposits, removal of, 226-227
coagulants, see Coagulants Copper oxide, boiler scale control, 156
defined, 27 Copper salts, biological fouling control, 296-
factors involved in, 27 297
iron removal by, 62 Copperas, 30-31
theory of, 27-28 chemical formula, 32
Colloids, reactive, feeding method, 148 commercial strength, 32
Color, 363 grades available, 32
difficulties caused by, 16 handling material, 32
test, 363-364 weight, 32
treatment, 16 Corrosion
Colorimetric tests, 332 causes of
Condensate return line corrosion in once-through cooling water systems,
carbon dioxide 237-239
content of steam, reduction of, 199 in open recirculating cooling water sys-
corrosive effects of, 197-199 tems, 259
sources of, 197 control
causes of, 195-196 in boilers, 168-172
correction of, internal methods of, 199-203 in closed recirculating systems, 272-273
dissolved oxygen, sources of, 196-197 in once-through cooling water systems,
Condenser deaeration, 133-134 237-246
Conductance, specific, 365 pretreatment for, 275-276
tests for, 365-369 in return condensate systems, 195-203
Conductivity load, factors influencing, 266-268
difficulties caused by, 17 of non-ferrous metals, 268-269
treatment, 17 prevention of
Conductivity meter, 334 in closed recirculating systems, 272-273
Constant-rate feeders, 140-142 in once-through cooling water systems,
Continuous monitoring automated feeding 240-246
equipment, 149-150 in open recirculating cooling water sys-
Control mode, 149-150 tems, 259-266
Control water analyses, 10 Corrosion inhibitors, 241-246, 259-266, 273
INDEX 421
Cyckhexlamine, 200-201 calculation of, by epm, 336-337

difficulties caused by, 17
test for, 403
DALTON's Law, 128 treatment, 17
Deaerating heaters, 131, 132-133 Double-orifice feeders, 146
Deaeration, 128-134 Drip feeder, 1~2
applications, 131 Dry-type feeders, 136, 148-149
chemical, corrosion control by, 240-241 gravimetric, 136, 148-149
condenser, 133 volumetric, 136, 148
equipment, 131
limitations, 131
mechanical, corrosion control by, 240, 259 EnucToR, water jet, 137-139
theory involved, 128-131 Electric motor-driven pumps, 142-143
vacuum, 134 Embrittlement of boiler metal
Deaerator, 131 causes, 205-206
atomizing, 133 characteristics of, 206-207
vacuum, 133 status of problem, 208-209
Dealkalization by chloride-anion exchange, treatment methods, 207-208
111-114 Exchangers, steel heat, care of, 276-277
applications, 113 "English" filters, 38
chemistry involved, 111-113 Equipment, see Analytical equipment
defined, 111 Expression of analytical results, 335-337
equipment required, 113 External treatment, as phase of industrial
limitations, 113-114 water conditioning, 9
use for, 111
Decanting feeders, 144-146
Dechlorination, 52-53 FEED systems, chemical, 136-150
Deconcentrators, 215 See also Feeders; Pumps
Deionization, see Demineralization Feeders
Demineralization, 103-109 constant-rate, 140-142
aP,plications, 108-109 decanting, 144-146
defined, 103 double-orifice, 146
equipment, 105-108 drip, 142
ion exchange resins, 103-105 dry-type, 136, 148-149
limitations, 108-109 gravimetric, 136, 148-149
Deterioration of wood, see Wood hydraulic, 144
Dianodic method of corrosion control, 243, pot type shot, 136-137, 138
246, 261-266 proportional feed systems, 142-146
Diatomite filtration, see Precoat filtration shot feed systems, 136-140
Diatomaceous earth filtration, see Precoat fil- volumetric, 136, 148
tration wet (solution), 136-146
Dioctadecylamine, 202 Feeding equipment, automated, 149-150
Disinfecting action, theory of, 48-50 Ferric chloride, 31
Disinfection, by chlorination, 48 chemical formula, 32
Dissolved gases, removal by deaeration, 128 commercial strength, 32
Dissolved oxygen, 386-387 grades available, 32
boiler corrosion control, 168-171 handling material, 32
sources of, 196-197 weight, 32
test for, 387-389 Ferric coagulants, 31
Dissolved solids, 402 Ferric hydroxide, 128
Ferric salts, feeding method, 147 removal by lime-soda softening method,

Ferric sulfate, 31 69-77
chemical formula, 32 test for, 371-373
commercial strength, 32 treatment, 16
grades available, 32 Hardness buffer reagent, 411-412
handling material, 32 Hardness indicator, 412
weight, 32 Hardness titrating solution, 412
Ferrous hydroxide, 128 Heat exchangers, steel, care of, 277-278
Filming amines, 202-203 Heaters, deaerating, 131, 132, 133
Filters Henry's Law, 128
acid, 285 Hexadecylamine, 202
"English," 38 Hot lime-hot ion exchange softening, 116-119
gravity, 37, 38-39 applications, 117-118
precoat, 43-46 chemical reactions, 116-117
-·pressure, 39-40 equipment employed, 117
"slow sand," 38 limitations, 118-119
Filtration, 36-41 use of process, 116
applications, 40-41 Hot process phosphate softening, 79-81
contact, iron removal by, 64 applications, 80
defined, 36 chemical reactions, 79-80
filter operation, 40 equipment requirements, 80
filters, types of, see Filters limitations, 80
limitations, 41 Hydrate alkalinity, 343•344
precoat, see Precoat filtration Hydrated lime, 69
theory involved, 36-38 feeding method, 148
Flocculation, 27, 29 Hydraulic feeders, 144
Fluoride Hydrazine, 169, 374
chemical formula, 16 tests for, 374-376
difficulties caused by, 16 Hydrazine standard solution, 412
treatment, 16 Hydrochloric acid, 48, 412
Fluoride Dianodic method of corrosion con- use in boiler cleaning, 224-226
trol, 265 Hydrochloric acid reagent, 409
Foaming, defined, 175 Hydrogen sulfide
Fouling, biological, see Biological fouling chemical formula, 17
Free chlorine test, 358 difficulties caused by, 17
Free mineral acid, 338 removal, 56-59
difficulties caused by, 16 aeration method of, 56-57
test for, 338 aeration with pH reduction method,
treatment, 16 57-58
anion exchange method, 58-59
application, 59
GRAVIMETRIC feeders, 136, 148-149 chlorination metnod, 58
Gravity filter, 37, 38-39 limitations, 59
Ground water, 14 treatment, 17
Gypsum, use in removal of sodium bicarbon- Hydrogen zeolite, cation e:kchange by, 96-101
ate, 72 applications, 101
equipment arrangements, 98-101
limitations, 101
HARDNESS, 370-371 Hydroxylamine reagent, 412
difficulties caused by, 16 Hypochlorination, 52
INDEX 423
Hypochlorites, 52 Iron retention, 61, 64-65

calcium, 52 defined, 61
sodium, 52 Iron standard solution, 412
Hypochlorous acid, 48, 49
jET atomizing deaerating heater, 132, 133
IoLE boilers, protection of, 171-172
Impurities, in water, 13, 15-17 LANGELIER Saturation Index Chart, 231
Industrial waste treatment, 320-328 Lignin, use in boiler scale control, 160-162
advanced, 32o Lime, hydrated, 69
methods of, 322-327 feeding method, 147
biological, 324-325 Lime-soda softening
chemical, 324 applications, 77
mechanical, 323-324 caustic soda, use of, 71-72
plant design and, 327-328 chemistry involved, 69-72
sludge disposal, 326-327 cold process, 69
Industrial water defined, 69
defined, 9 equipment employed, 73-77
problems, 10-11 gypsum, use of, 72
use, 320-321 hot process, 69, 74-77
Industrial water conditioning iron removal by, 63, 69
defined, 9 lime, use of, 70-71
external treatment, 9 lime and soda ash, use of, 69
internal chemical treatment, 9-10 limitations, 77
Inhibitors, corrosion, 241-246, 259-266, 273 propagating the chemical precipitating re-
Internal chemical treatment, as a phase of actions, 72-73
industrial water conditioning, 9-10 Lime softening, 70
Internal treatment chemicals, feeding of, Lime treatment, iron removal by, 63
164-165
Ion exchange resins, 103-105
Iron, 377 MAGNESIUM oxide, 31-32
chemical formula, 17 chemical formula, 32
difficulties caused by, 17 commercial strength, 32
test for, 377-379 grades available, 32
treatment, 17 handling material, 32
Iron bacteria, 65-67 weight, 32
Iron deposits, encountered in once-through Magnesium salts
cooling water systems, 235 boiler scale control, 155
Iron oxide, boiler scale control, 156 silica removal by, 83-87
Iron removal, 61-64 applications, 87
aeration, sedimentation and filtration equipment required, 86-87
method, 61-62 factors involved, 83-86
applications, 67 limitations, 87
cation exchange method, 63-64 Magnetite, as filter medium, 38
coagulation, sedimentation and filtration Manganese, 380
method, 62 chemical formula, 17
contact filtration method, 64 difficulties caused by, 17
defined, 61 removal, 67
lime and lime-soda softening method, 63 test for, 380-381
limitations, 67 treatment, 17
Manganese special solution, 412 corrosion load, factors influencing, 266-268

Manganese standard solution, 412 non-ferrous metals, corrosion' of, 268-269
Manganese zeolite, 64 scale control, 248-257
Manganous sulfate solution, 412 Organic agents, use in boiler scale control,
Meniscus, 328 160-163
Mercurials, slime and algae control with, 298 Organic base treatments for corrosion control,
Metal, boiler, see Boiler metal 265
Meter and timer controlled pumps, 143 Orthophenanthroline reagent, 413
Methyl orange indicator, 412 Orthophosphates, feeding method, 147
Methyl purple indicator, 412-413 Orthotolidine reagent, 413
Modular Feeding System, 150 Oxalic acid, 413
Molybdate reagent, 413 Oxygen
Morpholine, 201 chemical formula, 17
Mud, ~oiler scale control, 154 difficulties caused by, 17
dissolved, 386-387
boiler corrosion control, 168-171
NESSLER reagent, 413 sources of, 196-197
Nitrate, 382 test for, 387-389
chemical formula, 16 treatment, 17
test for, 382-383
treatment, 16 PERLITE, as filter-aid, 44
Nitrite, 384 pH, 390-391
as a corrosion inhibitor, 273-274 boiler water control and, 168
test for, 384-385 defined, 390
tests for, 391-393
()CTADECYLAMINE, 202 treatment, 16
Oil pH meter, 334
analysis for, 126 Phenanthroline method for determining total
boiler scale control, 156-157 iron, 377-379
difficulties caused by, 17, 121 Phenolphthalein indicator, 413
removal, 121-126 Phenols, chlorinated, biological fouling con-
from boiler heating surface, 222 trol, 294-295
from exhaust steam, 122 Phosphate, 394-395
from liquid condensate, 122-126 commercial forms for hot phosphate soften-
treatment methods, 17 ing, 81
On-stream cleaning, 277-278 feeding method, 147
Once-through cooling water systems, 230-246 tests for, 395-397
calcium carbonate scale, prevention of, Phosphate control, boiler scale, 158-160
232-235 Phosphate Dianodic method of corrosion
corrosion control, 23 7-246 control, 263-264
iron deposits, 235 Phosphate softening, hot, 79-81
scale control, 230-235 Phosphoric acid, use in boiler cleaning, 226
scale formation, 230-232 Photometers, 332-333
Open recirculating cooling water systems, Photometric methods of analysis, 332-333
248-269 Pneumatic proportioning control, 144
acid treatment, 254-257 Pollution control, 320-328
control of cycles, 252-254 Polyelectrolytes, 33
corrosion control, 259-269 Polyphosph,ates, as corrosion inhibitors, 241-
INDEX 425
242, 261, 266 calcium carbonate protective, 240, 259

Positive displacement pump, 142 control
Pot type shot feeder, 136-137 in boilers, 152-166
Potassium chromate indicator, 413 in once-through cooling water systems,
Potassium iodide-iodate, 413 230-235
Potassium permanganate, 413 in open recirculating cooling water sys-
slime control, 298 tems, 248-257
Precoat filtration deposits, identification of, 157-158
applications, 45-46 formation
defined, 43 in once-through cooling water systems,
equipment employed, 43 230-232
filter operation, 44-45 in open recirculating cooling water sys-
limitations, 46 tems, 349
theory involved, 43-44 mechanism of, 152
Pressure filters, 39-£10 prevention
Pretreatment, for effective corrosion control, in once-through cooling water systems,
275-276 232-235
Priming, defined, 17 5 in open recirculating cooling water sys-
Pumps tems, 251
controlled volume, 140-142 Seaweed derivatives, use in boiler scale con-
electric motor-driven, 142-143 trol, 161
meter and timer controlled, 143 Selective calcium softening, 71
positive displacement, 142 Separators, oil, 122
Shot feed systems, 136-140
Silica, 398-399
QuARTZ, as filter medium, 38 boiler scale control, 155-156
Quaternary ammonium compounds, bio- carryover, cause of, 184
logical fouling control, 297-298 chemical formula, 17
·removal by magnesium salts, 83-87
RAIN water, 14 application, 87
Reagents equipment required, 86-87
composition of, 330 factors involved, 83-86
prepared, composition of, 411-414 limitations, 87
Recirculating cooling tower systems, 149 tests for, 399-401
Referee methods of analysis, 329 treatment, 17
Residual chlorine test, 357 turbine blade deposits, 183-188
Resins, ion exchange, 103-105 occurrence of, 183-184
Results, analytical, expression of, 335-337 prevention of, 186-188
Return line corrosion, see Condensate return removal of, 184-186
line corrosion use in precoat filtration, 43
Silica standard, 413
Silicates, as corrosion inhibitors, 242-243
SAMPLE, preparation of, 330 Silt, boiler scale control, 154
Sampling, 150 Silver nitrate, 413
Sand, as filter medium, 36 Slime control, 289-299
Scale "Slow sand filters," 38
boiler, effect of, 153-154 Sludge contact clarifier, 28
calcium carbonate, prevention of, 232-235, Sludge disposal, industrial waste treatment,
252 326-327
Soda ash, 69 Solids, see Dissolved solids; Suspended solids;

feeding method, 147 Total solids
Sodium aluminate, 30 Solution feeders, see Wet (solution type)
chemical formula, 32 feeders
commercial strength, 32 Spraying, of cooling tower wood, 309-310
feeding method, 147 Spring water, 14
grades available, 32 Stannous reagent, 413
handling materials, 32 Starch, use in boiler scale control, 161
weight, 32 Starch indicator, 413
Sodium arsenite, 413 Starfamic indicator, 413
Sodium bicarbonate, removal by use of gyp- Steam purity measurement, 190-193
sum, 72 history of, 190
Sodium carbonate, 413 routine mineral. analyses, 190
Sodium chlorite, 51 specific conductance, 190-191
Sodit,1m hydroxide, 71, 413 total solids determination, 190
Sodium nitrate tracer techniques, 191-19 3
feeding method, 147 Steam sampling, 193
use for inhibition of embrittlement, 207 Steel heat exchangers, care of, 276-277
Sodium nitrite Stoke's Law, 24
as a corrosion inhibitor, 266
Storage of idle boilers
Sodium silicate
dry, 171-172
as corrosion inhibitor, 242-243
superheater, 172
chemical formula, 32
wet, 172
commercial strength, 32
Subsidence, 24-25
grades available, 32
applications, 25
handling materials, 32
defined, 24
Sodium sulfite, feeding method, 147
equipment employed, 25
Sodium sulfite reagent, 413
limitations, 25
Sodium thiosulfate, 413
purpose of, 24
Sodium zeolite softening, 89-94
theory involved, 24
applications, 92-93
Sulfamic acid, use for chemical cleaning of
equipment employed, 91-92
equipment, 227
limitations, 93-94
purpose of, 89 Sulfate, 404
chemical formula, 16
Softeners
lime-soda, 73
continuous hot process, 74-7 6 test for, 404-406
continuous sedimentation type, 73 treatment, 16
continuous sludge contact type, 7 3-7 4, Sulfide, analysis for, 59
75 Sulfite, 407
intermittent (batch cold), 73 test for, 408
zeolite, 89-94, 96-101 Sulfite indicator, 413
Softening Sulfur burners, 284
hot lime-hot ion exchange, 116-119 Sulfuric acid, 414
hot phosphate, 79-81 Sulfuric acid reagent, 414
hydrogen zeolite, 96-101 Surface water, 14
lime, 70-71 Suspended solids, 402
lime-soda, 69-77 difficulties caused by, 17
selective calcium, 71 removal by filtration, 36
sodium zeolite, 89-94 test for, 403
INDEX 42.7
TANNINS, use in boiler scale control, 160-162 water treatment

Taylor hydrazine reagent A, 414 Water
Taylor hydrazine reagent B, 414 chemical formula for, 13
THQ (Tetrahydroxyquinone) indicator, 414 ground, 14
Thymotrol, 144 impurities in, 13, 15-17
Titration, 330-332 pure, 13
Total solids, 402 rain, 14
difficulties caused by, 17 sources of, 13-15
test for, 403 surface, 14
treatment, 17 Water jet eductor, 137-139
Tray-type deaerating heater, 131, 133 Wet (solution type) feeders, 136-146
Tuberculation, 242, 263 constant-rate, 140-142
Turbidimeter, 334 decanting, 144-146
Turbidity, 409-410 double-orifice, 146
difficulties caused by, 16 hydraulic, 144
removal pot-type proportional, 142
by filtration, 36-41 pot-type shot, 136-137, 138
by lime-soda softening method, 69 proportional feed systems, 142-146
by subsidence, 24, 25 shot feed systems, 136-140
test for, 410 Wood, cooling tower
treatment, 16 deterioration, 301-310
Turbine blade deposits (silica), 183-188 biological, 303-304
causes, 183 chemical, 303
effect of, 183 control of, 304-308
occurrence of, 183-184 physical, 304
prevention of, 186-188 types of, 301-310
removal of, 184-186 examination, 308-309
spraying methods, 309-310
zone of inhibition test, 308-309
U PFLOW clarifier, 29-30
XYLENE cyanole, 414
V-NOTCH chlorinator, 54
Vacuum deaeration, 134
Volumetric feeders, 136, 148 ZEoLITE exchange materials, 89-91
Zeolite softening, see Sodium zeolite softening
Zinc Dianodic method of corrosion control,
WAsTE discharge, difficulties caused by, 321- 243, 246, 264-265
322 Zone of inhibition test, cooling tower wood,
Wastewater treatment, see Industrial waste- 308-309

BETZ Handbook of Industrial Water Conditioning

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

BETZ Handbook of Industrial Water Conditioning

Uploaded by

Copyright:

Available Formats

._... ~ .._,__ ... .,..,.. - -....• ,._...,. •• tr., .,.,.--...~----~...,....._...... _ _ _ ~-· -·-~· ~ - ~ .... .....

TREVOSE, PENNSYLVANIA 19047

BETZ LABORATORIES, INC.

TREVOSE, PENNSYLVANIA 19047

This book was prepared by the engineering and technical

35. Cooling Water Treatment: Closed Recirculating Systems ...................... 270

CONTROL WATER ANALYSES AND THEIR INTERPRETATION

What Is Industrial Water Conditioning?. . . . . . . . 9

used directly or indirectly in an

INDUSTRIAL WATER PROBLEMS

Following is a list of the more common water

Boiler Water Problems

Steam and Condensate Problems

Carbon Dioxide Grooving

Cooling and Process Water Problems

Preliminary Treatment Problems General Problems

Sources of Water . ......................... 13

ing. Whatever the source, water always con- SOURCES

Figure 2-1 • Calcium Sulfate Crystalline Scale in 5 inch Line

Figure 2-2 • Pipe Section Pitted from Dissolved Oxygen

The water from deep wells usually contains

The impurities that may be present in a

Fig. 2-4-Common Impurities Found in Water

CONSTITUENT CHEMICAL FORMULA DIFFICULTIES CAUSED MEANS OF TREATMENT

CARSON Corrosion in water lines and par- Aeration. Deaeration. Neutraliza-

SULFATE ....... Adds to solids content of water, Demineralization. Distillation.

Fig. 2-4-Common Impurities Found in Water (cont.>

CONSTITUENT CHEMICAL FORMULA DIFFICULTIES CAUSED MEANS OF TREATMENT

MANGANESE . . . Mn + + Same as iron. Same as iron.

HYDROGEN Cause of "rotten egg" odor, Cor- Aeration. Chlorination. Highly

AERATION is a process which consists of bon dioxide and no measurable amount of

Courtesy lnfilco, Incorporated

Courtesy Pfaudler Permutit Inc.

Figure 3-2 • Forced Draft Aerator

APPLICATIONS Theoretically, at 20 C it is possible to re-

0.01 mm or more. Smaller particles have It is necessary in basin design to make

Theory of Coagulation . .................... 27

Courtesy Graver Water Conditioning Co.

Figure 5-1 • Sludge Contact Clarifier

SLURRY POOL INDICATED BY SHADED AREAS

Courtesy lnfilco, Incorporated

Figure 5-2 • Upjfow Clarifier

Fig. 5-3-Data on Coagulants and Coagulant Aids

Sodium Crystals 60 Iron Stabilized

Figure 5-4 • Jar Test Coagulation Studies

COAGULANT AIDS ionic, cationic or nonionic. Polymers whose

Courtesy Roberts Filter Mfg. Co.

0. 72 mm, with a uniformity coefficient of moval of accumulated suspended matter is

Courtesy, Infilco, Incorporated

IATOMA_CEous earth filtration is a proc:ss

Because diatomaceous earth is an abrasive

Figure 7-2 • Industrial Type of Diatomaceous

HLORINATION is the process in which chlo- tion and oxidation.

CHLORINE RESIDUAL, PPM

Figure 8-2 • Ammonia Removal by Chlorination

CHLORINE SOLO CHLORINE SOLD AS CHLORINE

I# OF CHLORINE POUNDS OF DRY GALLONS OF SODIUM HYPOCHLORITE

Courtesy Chlorination Topics, Spring 1956

be processed and consequently dechlorination Least detectable odor 3.5

(from U.S. Bureau of Mines Paper 248)

FEED RATE INDICATOR INJECTOR DISCHARGE

Courtesy Wallace & Tiernan Inc.

Figure 8-4 • V-notch Chlorinator

interference from various substances cause it reduction potential. Additional instruments