Professional Documents
Culture Documents
T h is p a p e r g iv es a n e x p o s itio n o f t h e v a r io u s i t e m s w h ic h p r e f e r a b ly a s a p e r c e n t o f t o t a l e n e r g y t h e o r e t i c a l l y a v a i l
go t o m a k e u p t h e le a v in g v e lo c ity a n d e x h a u s t lo s s o f a a b le f o r c o n v e r s io n t o s w it c h b o a r d p o w e r.
s te a m tu r b in e . T h e i m p o r t a n c e o f t h i s lo s s a n d t h e r a W it h a p a r t i c u l a r e x h a u s t o p e r a t i n g a t fixed s t e a m c o n
p id ity w i t h w h ic h i t in c r e a s e s a t h i g h lo a d s c a u s e i t t o b e a d i t i o n s , t h e le a v in g v e lo c ity a n d e x h a u s t lo s s in c r e a s e s
d e te r m i n i n g i n f lu e n c e i n fix in g t h e e c o n o m ic r a t i n g o f a r o u g h l y a s t h e s q u a r e o f t h e lo a d ( p a ra b o lic r u l e ) .
m a c h i n e . T h e se v e ra l e le m e n t s n e c e s s a r y f o r a n a n a l y s i s W ith a p a r t i c u l a r e x h a u s t p a s s i n g a fixed flow , i n c r e a s in g
a r e e a c h e v a lu a te d i n a f a ir l y d i r e c t , a l t h o u g h s o m e ti m e s t h e t o t a l a v a ila b le e n e r g y i n t h e h i g h e r s t a g e s o f t h e t u r
a p p r o x im a te , m a n n e r . M o re d e ta i le d a n d p r e c is e e s t i b in e b y im p ro v e d s te a m c o n d itio n s c o rre s p o n d in g ly r e
m a te s m ig h t b e m a d e b u t a re b e y o n d th e i n t e n t o f th is d u c e s t h e p e r c e n ta g e lo s s i n t h e e x h a u s t ( h y p e rb o lic r u l e ) .
p a p er. W it h a fix ed p e r c e n ta g e lo s s i n a p a r t i c u l a r e x h a u s t t h e
T h e lo ss i n q u e s t i o n o c c u r s i n t h e e x h a u s t h o o d b e tw e e n p o w e r m a y b e i n c r e a s e d b y i m p r o v e d s t e a m c o n d it i o n s a s
t h e l a s t w h e e l e x it a n d t h e e x h a u s t fla n g e t o t h e c o n d e n s e r . t h e !/i-power o f t h e t o t a l a v a ila b le e n e r g y b y i n c r e a s in g t h e
I t is m a d e u p b o t h o f k i n e t i c e n e r g y lo s s a n d o f p r e s s u r e flow t o t h e c o n d e n s e r .
lo ss t h r o u g h t h e h o o d a n d e a c h e ffe c t v a rie s w i t h l o a d a n d B a s e - lo a d o p e r a t i o n j u s t if i e s m o r e l i b e r a l e x h a u s t a r e a s
w ith l o c a t io n a r o u n d t h e w h e e l a n n u l u s . M o i s t u r e is t h a n p e a k - l o a d s e rv ic e b u t i n t h e u l t i m a t e t h e h e a t - r a t e -
a llo w e d f o r ; s u p e r s a t u r a t i o n n e g le c te d . lo a d c u r v e is t h e c h a r a c t e r i s t i c w h ic h is o f m o s t i m p o r
T h e t o t a l lo ss m a y b e e x p re s s e d i n B t u p e r p o u n d flow ta n c e to th e o p e r a to r a n d se c o n d o n ly to th e re lia b ility o f
to c o n d e n s e r o r a s a p e r c e n t o f a d ia b a tic h e a t d ro p o r p e rfo rm a n c e .
made. The idea is that each comparison should be made on its In evaluating these items it is necessary to consider the follow
own merits. ing:
R feu M ^ o f Item s (5) Moisture content of the steam
The several items of the leaving velocity and exhaust loss may (6) Possibility of supersaturated expansion.
be listed as follows; this list being intended to cover all losses We shall rule out from this discussion:
(7) Consideration of radial velocity in an axial-flow annulus
(8) Eddy loss associated with edge thickness of buckets.
This last is properly chargeable to the nozzle and bucket
efficiency and the stream is supposed to have healed into a cylin
drical jet on emerging from the wheel annulus. This latter
ruling is, of course, arbitrary. The N.E.L.A. Prime Movers
Committee Report on Turbines, No. 234, July, 1932, recommended
correcting for bucket-edge thickness, thus in effect charging the
bucket-edge loss to the exhaust loss rather than to the bucket.
Suffice it to say that in any case it should not be charged twice
and the manner used should be clear in any particular case.
T y p ic a l C urves
F ig . 3 S p e c if ic V o l u m e o r S t e a m A p p r o a c h in g 1 In. Hg
A b s P r e s s u r e a n d 10 P e r C e n t M o is t u r e weight flow to con dense/? lb / sec.
(3 5 ,0 0 0 -k w tu rb in e .)
F ig . 5 R e l a t io n B e t w e e n A n n u l u s P r e s s u r e a n d W e ig h t
If1 F l o w t o r 1 I n . H g A b s P r e s s u r e a t t h e F l a n g e f o r a 3 5 ,0 0 0 - K w
Ss T u r b in e
8 (T he d o tte d lines A, B,_C, and D for c o n sta n t annulus volum e flows corre
spond to th e velocity diagram s in Fig. 4 and show th e relation betw een an
* nulus pressure and w eight flow for varying exhaust flange pressures.)
F ig . 4 V e lo c ity D ia g r a m s A , B , C , a n d D f o r S te a m E m e r g in g
F r o m L a s t B u c k e t o f a 3 5 ,0 0 0 - K w T u r b i n e
(Figures are velocities in f t p e r sec except for th e h e a t eq u iv alen ts in B tu
per lb of th e kinetic energy of ab so lu te ex h au st velocity. E a c h d iag ram re p
resents a p a rtic u la r an n u lu s volum e flow, as in d icated . T he volum e m ay be
m ade u p of a larger w eight of denser stea m or a sm aller w eight of m ore rarefied
steam . See Fig. 5.)
existing in the exhaust hood between the exit from the wheel
annulus and the exhaust flange: AnnuLus v o m r iE flo w , c u ft. p e r s e c .
F ig . 6 K i n e ti c - E n e r g y L o s s in T h a t F r a c t i o n o f t h e S te a m
(1) Normal velocity loss W h ic h C r o s s e s t h e A n n u l u s o f a 3 5 ,0 0 0 -K w T u r b i n e a t S p e e d
(2) Loss due to tangential component, or whirl loss (T he m oisture moves a t very low speed. F o r 1 in. H g abs pressure a t the
(3) Eddy losses, associated with non-uniformity of flow exh au st flange and the exhaust-hood drop show n in Fig. 5, the points A,
B, C, and D correspond to the a ctu al kinetic energy leaving loss and also
(4) Pressure drop through the hood itself. to th e several diagram s in Fig. 4.)
FUELS AND STEAM POWER FSP-56-10 517
A V A IL A B L E E H E / ? $ Y L O S S BTU /LB T O T A L F L O V J
TO COM D E H S E F ?
F i g . 10 A v a i l a b l e E n e rg y L oss D u e to P re ssu re D ro p
T h ro u g h t h e E x h a u s t H o o d F ro m t h e W h e e l A n n u lu s o f a
35,000-Kw T u r b i n e t o t h e E x h a u s t F l a n g e
(If the wheel annulus could exhaust directly at 1 in. Hg abs pressure,
the adiabatic energy available ahead of the last wheel exit would be increased
by the amount shown.)
F ig . 7 V a r i a t i o n o f A n n u l u s P r e s s u r e A r o u n d t h e P i t c h
C i r c l e o f t h e A n n u l u s o f a 35,000-Kw T u r b i n e f o r T w o
S p e c i a l W e i g h t F l o w s , 85 L b p e r S e c a n d 45 L b p e r S e c
(The irregularity is due to the quarter turn of the hood and its internal
bracing. Lack of symmetry is due to the tangential component in the ex-
h a u s t v e lo c ity .)
F i g . 11 V a r i a t i o n A r o u n d t h e A n n u l u s o f A v a i l a b l e E n e r g y
L oss D u e to P re s s u re D ro p T h ro u g h th e E x h a u s t H ood o f a
3 5 ,0 0 0 - K w T u r b i n e f o r T w o W e i g h t F l o w s , 8 5 L b p e r S e c a n d 4 5
Lb p e r Sec
(T his v a ria tio n is due to th e v a ria tio n of p ressure show n in Fig. 7.)
F ig . 8 V a r i a t i o n o f K i n e t i c - E n e r g y L o s s p e r P o u n d o f
S te a m a t S p e e d A r o u n d t h e P i t c h C i r c l e o f t h e A n n u l u s
Fig. 7. We arrive finally at Fig. 9 which shows the kinetic-
o f a 3 5 ,0 0 0 - K w T u r b i n e f o r t h e S p e c i a l W e i g h t F l o w s , 8 5 L b energy loss plotted against the various loads in kw.
p e r S ec a n d 45 L b p e r S ec Returning to the r&um6 of items, the method of arriving at
(The w e ig h t flow is p ra c tic a lly u n ifo rm a ro u n d the a n n u lu s . The v o lu m e Fig. 9 has taken care of (1) normal velocity, (2) tangential
flow v a rie s in in v e rse re la tio n to th e p re s s u re a n d c o n s e q u e n tly th e v e lo c ity
a n d k in e tic e n e rg y V ary a s sh o w n .) component, and (3) non-uniformity of flow. It remains to
518 TRANSACTIONS OF TH E AMERICAN SOCIETY OF MECHANICAL ENGINEERS
evaluate the loss of available energy due to pressure drop through efficiency of 75 per cent has been assumed, whereas more accurate
the exhaust hood itself. estimates use true efficiency curves, or true integrated available
Fig. 10 shows the relation to be nearly linear so that Fig. 11, energy.
which illustrates the distribution around the annulus, is not The total loss might be divided up in another way by defining
absolutely essential for a satisfactory preparation of Fig. 12 the net exhaust-hood loss as the extra loss occasioned by having
which shows the available energy loss plotted against the various the specified exhaust or condenser pressure occur at the hood
loads in kw. flange instead of at the wheel annulus. This viewpoint has a
certain merit in setting up a standard for comparison of no
A b s o l u t e V a l u e o f L o s s e s I n s id e E x h a u s t H ood
pressure drop through the hood. The net hood loss viewed this
When it comes to expressing the total effect it is necessary to way is less than the actual loss due to the pressure because the
take account of the quantities involved. The kinetic-energy loss, pressure drop reduces the kinetic-energy loss at the annulus.
Fig. 9, affects only the steam at speed, and since in this case 10 In order to divide the total loss in this way it is necessary to
per cent of the condenser flow is in the form of moisture moving compute the leaving velocity loss at the annulus with the specified
at low velocity, this is to be applied to 90 per cent of the eon- exhaust pressure occurring at that location. The balance be
tween this and the total may be thought of as the net amount due
to the presence of the exhaust hood. This is not the same as the
approximate computation suggested below, because true per
formance (as nearly as can be estimated) under the supposed
conditions is computed. The result especially depends on both
the relation between the size of annulus and the hood and the
load, and the method of computation employed for estimating
what would happen under the assumed conditions. Thus a
heavily loaded annulus discharging into a liberal hood will suffer
very little additional loss because of pressure drop through the
hood as compared with free discharge without any hood. On
F i g . 14 T o t a l L e a v i n g - V e l o c i t y a n d E x h a u s t L oss o f a
3 5 ,0 0 0 - K w T u b b i n e , E x p r e s s e d a s a P e r c e n t a g e o f T o t a l T h e o
r e t i c a l l y A v a i l a b l e E n e r g y i n A l l S t e a m , D i v i d e d So a s t o
S how th e N e t L o ss C h a rg e a b le to th e P re ssu re D ro p
T hbough th e H ood
(In this case th e velocity loss is th a t which would occur w ith exhaust p res
sure a t the wheel itself.)
the annulus goes part way to compensate for the neglect of Thus it is correct to speak of a loss of 30 Btu per lb flow to the
pressure drop through the hood with its corresponding loss of condenser. But such a statement is not very significant. If
available energy. the adiabatic heat drop is 500 Btu, the loss may be said to be 6
per cent of the adiabatic heat drop. But there is still an un
S u p e r s a t u r a t io n
certainty as to the effect on the power generated since the 6
In dealing with the results so far set down no account has been per cent loss applies only to the steam going all the way through
taken of any effects which may be caused by supersaturated the turbine. If 6/ 6 of the power generated comes from steam
expansion, that is, by expansion of the steam without condensa which goes all the way through the turbine, then the true loss
tion to a momentarily cooler, denser condition. In considering
Fig. 13 as representing absolute values, this reservation has to be
kept in mind in addition to the minor inaccuracies purposely
assumed for simplicity.
It has been noted in comparing Fig. 14 with Fig. 13 that the
true loss through the exhaust hood is accompanied by a reduc
tion of velocity loss in the denser medium at the annulus. The
two effects are, to a certain extent, compensating.
Similarly supersaturation, if present, results both in a reduc
tion of energy made available for conversion and in a reduction of
the leaving-velocity loss. With the amount of moisture present
there is not likely to be any high degree of supersaturation in
this particular case. However, in a different case with, say only
2 or 3 per cent of moisture theoretically present, supersaturated
expansion should be allowed for.
C o m p a r is o n o f T u r b in e s
For instance, take the same loss of 30 Btu per pound flow to the
condenser (see Fig. 15). Such a loss amounts to 10 per cent
of the adiabatic heat drop of an ancient low pressure turbine
with 300 Btu available energy while it is only 5 per cent of the
adiabatic heat drop of a modern high pressure resuperheating
turbine with 600 Btu available. Similarly a full use of stage
extraction for feed heating so increases the power generated from
a particular exhaust that the importance of a fixed leaving-
velocity and exhaust loss may be decreased as much as 20 per
cent in this manner; an effect which is not shown at all by ex
pressing the loss in terms of adiabatic heat drop, which does not
change with extraction.
The use of the mercury turbine in conjunction with the steam
turbine still further reduces the percentage importance of a fixed
size loss.
In other words, modern cycles warrant the use of much
F ig . 15 P e r C e n t L e a v i n g - V e l o c i t y a n d E x h a u s t L o s s W i t h
higher absolute losses per pound of steam exhausted to conden
V a r io u s S te a m C y c le s , i n E a c h C a s e W i th 3 0 - B tu L o s s p e r
L b T o t a l F lo w t o C o n d e n s e r ser, because much less steam is being exhausted per kw hour
(If the sam e tu rb in e ex h au st is used for th e sam e kw c a p a c ity while sub generated.
s titu tin g m odern steam conditions for low er pressures an d older cycles,
the percentage leaving loss will be g reatly reduced an d th e ex haunt will I n c r e a s e d C a p a c it y
appear w astefully generous in size for th e b e tte r stea m conditions.)
Fig. 16 is the counterpart of Fig. 15 showing how the capacity
Correct values of the loss may be expressed in terms of heat obtainable from the steam entering a given exhaust may be
equivalent in Btu per lb flow to the condenser, or as a per cent increased by the use of modern cycles. The relative capacities
of the adiabatic heat drop, or as a per cent of the total theoretical for constant percentage loss are based on the approximate rela
energy available within the flowing steam between throttle inlet tion that, for a given cycle, the absolute value of the leaving-
and the several extraction and exhaust flanges, for conversion to velocity and exhaust loss increases as the square of the flow.
switchboard power. This leads directly to the conclusion that relative capacity for a
520 TRANSACTIONS OF TH E AMERICAN SOCIETY OF MECHANICAL ENGINEERS
constant percentage loss increases as the 3/ V p o w e r 0f the total ever, this method is somewhat difficult to apply and if a simpler
energy theoretically available. method could be devised for comparing turbines it would be
helpful in evaluation procedure. It would be interesting if the
C o n d it io n s o f O p e r a t io n
author could indicate the probable error which might be intro
The type of service and conditions of operation also are very duced in comparing the exhaust end of a variety of turbines
important in evaluating the amount of leaving velocity and ex on the basis referred to in the N.E.L.A. Prime Movers Committee
haust loss that is acceptable. This is because of the rapid change Report No. 234 on Turbines.
at high loads. Thus a base load machine which is to run most The author speaks of load curves and conditions for which a
of the time at its maximum rated capacity requires a more turbine is chosen but these conditions are so subject to change
that it is difficult to predict conditions far in advance.
In case of doubt as to the future it may be wise to lean
in the direction of capital savings at the expense of
economy. Until quite recently growth of load through
out the country was rapid and new units were in
stalled at fairly frequent intervals. Each succeeding
installation carried design improvements rendered pos
sible by the advance in the art. With new units of
superior economy added to the system, the capacity fac
tor of the earlier machines dropped and they are op
erated only at peak loads. For example, on one system
new machines purchased some years ago were operated
at 58 per cent capacity factor for about 6 years and
after this period the use of these machines diminished
F i g . 17 T y p i c a l A n n u a l - L o a d C u b v e s
almost annually until it finally reached a capacity
(Since th e leaving-velocity a n d ex h au st loss increases roughly in a parabolic m anner
w ith th e load, a base load sta tio n such as B will experience a g reater in te g ra te d loss factor in the neighborhood of 5 per cent in about 20
th a n a sta tio n like A which shares light loads. T u rbines for base load service w a rra n t years. Experiences of this kind cannot be safely taken
m ore liberal ex h au st areas.)
as a basis for purchase of new machines since it
liberal exhaust with smaller absolute loss than a machine de presupposes a continuation of load growth such as we have had in
signed for a broad range of service and the expectation of running the past.
at maximum capacity only a short part of the year. In the absence of continued load growth, the capacity factor
Roughly speaking, the higher the annual capacity factor the and use factor of the more recent machines will increase materially
lower should be the fixed loss in the exhaust but this is just an above that which might be anticipated from earlier experience.
other way of saying the more you run a machine the more ef Under such conditions machines evaluated for a relatively low
ficient it should be. For careful comparison actual load re capacity and use factor may ultimately become base-load units of
quirements should be analyzed. For instance, Fig. 17 illustrates the system. This is a situation which cannot always be foreseen
two types of service, the total energy generated being the same and should not be lost sight of in the purchase of new units, es
in each case, but Station A is taking the swings and does the bulk pecially as it influences exhaust areas.
of its operation at half and three-quarters load while Station B is Another important consideration in the choice of exhaust areas
on base-load service and operates mostly around three-quarters in the turbine is the cost of steam-generating capacity to meet
to full load. If these two stations were each equipped with a full-load requirements. With a lower exit loss less steam-generat
single turbine of the type represented by Fig. 13, the integrated ing capacity is required and this should be carefully considered
loss due to hood effects in Station B would be 50 per cent more in evaluating turbine performance. Whether the value of the
than in Station A, the difference amounting to approximately additional boiler-plant capacity required by a less efficient turbine
1 per cent of fuel requirements. From this it may be inferred is taken on a pro-rata base or increment cost base is largely a
that with medium coal prices a purchaser could afford to pay matter of judgment, but it is an important consideration if the
some 10 per cent more in the case of Station B for a turbine with highest economy in the use of capital is to be achieved.
more liberal exhaust and a different load curve from that which Summarizing, the influences to be considered in the design of
would be suitable for Station A. the turbine-exhaust end are the capacity factor, cost of boiler and
This brings attention directly to the load curve as being con condenser capacity, cost of steam, and quality, quantity and
sidered a most important feature by operators in the selection temperature-duration curves of circulating water available.
of a turbine. It is second only to reliability of operation and The design is influenced also by the relation of the system de
with turbines as dependable as they have been of late years, mand period to the circulating water temperature, and if the
attention (perhaps too often) is likely to be concentrated en period of maximum demand of the system coincides with the
tirely on the load curve. Together with throttle loss at light period of minimum circulating water temperatures, the condi
loads, which dwindles out at full load, the leaving-velocity and tions are favorable for a design with relatively low exit loss.
exhaust loss, which increases from very little at light loads, is The paper clearly brings out the influence of higher initial
one of the most effective tools the turbine designer has at his steam conditions and other improvements in reducing the per
disposal in producing the type of machine suited to the operating centage exit loss, nevertheless there is often potential capacity
conditions. available which might have been purchased at an attractive
price. In latitudes favored with an abundance of cold condens
Discussion ing water during the period of maximum demand, it is probable
that there are cases where available capacity might have been
W. E. C a l d w e l l .* The paper presents a comprehensive purchased in the last wheel and condenser well below the unit
method of calculating the exit loss in a condensing turbine. How- cost of the plant.
• Efficiency Engineer, The New York Edison Co., New York, As the art advances progress in improving efficiencies will
N. Y. Mem. A.S.M.E. diminish as the more attractive possibilities have been exhausted
FUELS AND STEAM POWER FSP-56-10 521
and we may find it profitable in the future to resort to more The question arises as to whether it would be desirable and
liberally designed exhaust areas in steam turbines, as well as economic to incorporate diffuser designs in the exhaust hoods of
more liberal condensers and auxiliaries. With each successive all turbines. Some savings would result from such an ideal
installation closer cooperation is developed between the engineers design.
of the manufacturers and those of the power producers and it is Regarding supersaturation, Mr. Robinson indicates that this
through the mutual understanding of the common problems in only needs consideration when moisture contents of 2 to 3 per cent
volved that the greatest progress may be made in achieving a occur at exhaust. Generally the moisture content ranges from 8
well-balanced design in the ultimate plant. to 11 per cent. Can any supersaturation exist at the last row of
blades under these conditions? The late Professor Callendar and
A. G. C h r i s t i e .4 As the author states in his opening para H. M. Martin, both of England, have advanced the opinion
graph, the combined leaving-velocity and exhaust loss constitute that supersaturation will persist even to exhaust. R. Colburn
the most important single loss in condensing steam turbines. and the writer concluded two years ago that supersaturation
These losses have attracted the serious attention of the plant- might exist at high moisture contents for there appear to be
designing and operating engineers only in the last few years. indications that this was a contributing factor in blade erosion
Leaving losses were considered a few years ago by the Prime from moisture. But the question is by no means settled.
Movers Committee of N.E.L.A. Data on various turbines were One of the Power Test Code Committees should consider a
collected and referred to the writer for analysis. In many cases standard definition for the combined leaving and hood loss. Mr.
arbitrary assumptions had to be made regarding the amount of Robinson indicates the different ways in which the loss can be
steam to exhaust and other items. Certain of these such as the expressed but has not given any method preference over the
allowance for blade-outlet thickness on the last set of blades are, others. It is highly desirable that the expression of this loss be
as pointed out by Mr. Robinson, open to question. Manu standardized so that all engineers can refer to it in the same
facturers at that time were somewhat reluctant to discuss leaving terms.
losses. There appeared to be no standard or other method of A further comment may be made. These losses are fixed by
expressing leaving losses. The writer therefore proposed as a the original design of the turbine and are inherent in its con
tentative measure the expression of leaving loss as the equivalent struction. They cannot be increased or decreased for given
of the absolute velocity from the last row of blades assuming operating conditions by any efforts of the station operators.
axial flow and found by assuming the exhaust pressure and It is therefore incumbent upon the plant designer to give the
volume at the blade annulus instead of at the exhaust nozzle. fullest consideration to the economics of these losses in the
In other words the losses in the exhaust hood were not con initial selection of the turbine equipment. Mr. Robinson points
sidered. The Prime Movers Committee then considered the out certain factors such as the character of load, etc., which
question of a standard method of expressing leaving loss. How influence the economic effect of leaving losses.
ever, differences of opinion developed and a suggestion was made There are differences of opinion in regard to the allowance for
that a leading authority on turbine design be asked to discuss leaving losses in determining the true end-point of the condition
this whole subject in a paper before A.S.M.E. Mr. Robinson’s curve. In some cases, the whole of the leaving loss has been
concise and enlightening paper is the result of this suggestion. A deducted from the total heat to exhaust to find the end-point of
careful analysis of its contents will aid plant designers and op the condition curve. Others estimate the probable stage ef
erators to give intelligent consideration of these losses and to ficiency of the last stage and only deduct from the heat to exhaust,
their influence on plant economy. the product of this stage efficiency and the leaving losses. Ob
Mr. Robinson shows very clearly that the leaving-velocity loss viously the latter method gives the higher end-point. Can Mr.
and the pressure loss in the hood are interdependent and both Robinson indicate which method more nearly approaches the
should be considered. The method of indicating leaving loss true end-point?
used in the N.E.L.A. report does not give all the facts. This paper will prove valuable to designing and operating engi
In Fig. 13 Mr. Robinson shows that the total leaving-velocity neers as it discusses a hitherto little understood subject in a clear
and exhaust-hood loss at 28,000 kw, the most efficient load on the and comprehensive manner.
35,000-kw turbine under consideration, exceeds the loss cal
culated by the N.E.L.A. method by 0.6 per cent but he also S a b i n C r o c k e r .6 Mr. Robinson’s paper presents valuable
points out that this may not be true for turbines with diffuser information on turbine-exhaust loss as seen by a turbine designer.
exhausts. Few of the turbines in the N.E.L.A. report have As such, the presentation of these data is an excellent and timely
diffuser exhaust hoods so that the computed losses are probably work. The writer would like to present the turbine user’s
not as large as the sum of the true leaving loss and hood loss. view of these data as they may be applied to turbines in power
However, it is apparent that the computed figures published plants.
by the writer in the N.E.L.A. report can only be used with reser As power-plant heat rates have been improved through the
vations. development of more efficient equipment and the adoption of
Leaving loss depends upon the length of the last row of tur more favorable heat-utilization cycles, it has become necessary
bine blades and the quantity of steam flowing to the condenser. for engineers charged with the design and operation of power
Hence the amount of this loss when the maximum length of plants to go progressively to greater refinements in order to
blades is used, depends upon the output rating of a given casing continue an improvement of thermal efficiency within the eco
as fully discussed in the writer’s paper before the World Power nomic limitations of the case. Consequently in order to obtain
Conference a few years ago. further improvements at the present time it is necessary to con
Losses through the hood will also be dependent upon the out sider comparatively small differences and quantities, such, for
put of a given casing but as Mr. Robinson states these will also instance, as are involved in the return of low-grade heat through
depend upon exhaust outlet design. Certain exhaust hoods the turbine feedheating circuit, in changes in condensing equip
exert a diffuser effect so that the leaving velocity of the steam is ment to obtain better vacuum, or other changes which may affect
partly converted into pressure head to overcome hood losses. the exhaust loss in different ways.
4 Professor of Mechanical Engineering, Johns Hopkins University, 5 Engineer, Engrg. D iv., Detroit Edison Co., Detroit, Mioh.
Baltimore, Md. Mem. A.S.M.E. Mem. A.S.M.E.
522 TRANSACTIONS OF TH E AMERICAN SOCIETY OF MECHANICAL ENGINEERS
Operating companies make many such studies in which the capacity including reheat, is 160 Btu/lb at 29 in. vacuum of
exhaust loss becomes an important item. Under present prac which 105 Btu/lb can be converted into useful work. Then
tise it is necessary to obtain specific data from the manufacturer by virtue of limited exhaust dimensions the pressure at the exit
for each case considered. This results in great expense to both of the last row of blades is reduced 0.5 in. vacuum to 28.5 in.
parties, loss of time in correspondence, and a restriction in the Then the reduction in available energy will be (from Fig. 10)
number of comparisons that can be made in a given study. 21.6 Btu/lb and the resulting Ai, will be 139.4 Btu/lb at 28.5-
In consequence, while the manufacturer usually shows a willing in. vacuum of which 100 Btu/lb can be converted into useful work.
ness to cooperate in such matters, the results are not always Consequently from this consideration the actual reduction in
satisfactory. heat to work will be 5.0 Btu/lb, not corrected for moisture losses
The present paper is a notable beginning toward clearing up which, for such small changes, could be neglected.
these difficulties in that it enumerates all the items involved in the Carrying this example to a conclusion by adding the effect
exhaust loss of a steam turbine. A detailed statement of the of moisture, we see on the i-s diagram that the average moisture
turbine-user’s need for data of this type should likewise help the for the expansion to 29-in. vacuum is 9.25 per cent while the
situation. Briefly, he should be able to determine the exhaust average moisture for the expansion to 28.5-in. vacuum is 9.0 per
loss on his turbines for any possible operating condition. In this cent. Assuming no moisture removal and applying a correction
regard Figs. 4 and 7 from which can be determined the magnitude
of the loss for a given turbine are of particular interest. It would
seem that such information should be made available to a turbine
owner in either of two different ways: (a) on request furnish
him curves similar to Figs. 4 and 7 applying to his particular tur
bine and (b) make available to him the methods for computing
such curves from the basic data for his type and size of turbine.
Such material is indispensible to the turbine operator for an in
telligent solution to his problems in power-plant design. Ap
parently Fig. 4 can readily be used for any desired operating
condition but it appears that Fig. 7 would have to be given for a
series of different condenser pressures and exhaust flows before it
would be of much use to a plant designer. This would entail a
large amount of detailed work on the part of the manufacturer,
which it would seem he could eliminate by making more general
ized computation methods available to the turbine user.
The writer is quite aware that many variables are involved in
computing exhaust losses, and that turbine manufacturers and
their designers prefer to pass out information piecemeal rather
than to give an operating company’s engineers sufficient informa
tion about a given turbine for them to compute the necessary
correction factors for that turbine themselves. Nevertheless,
it would seem that the author could well afford to give the user’s
requirements further consideration in closing an otherwise com
mendable paper.
available energy converted into work at approximately the same The use of these factors would considerably alter the author’s
efficiency as in the case of the 29-in. expansion. Fig. 13 and also call attention to the necessity of using weighted
In regard to the question of designating leaving losses it averages over the entire year in connection with the author’s
appears that the most rational method of evaluating them Fig. 17, showing typical annual load curves, in order to compute
would be to consider the loss in kw. By the use of kw an ab the total annual losses from this source.
solute loss is immediately determined and its magnitude is left
without question. P. H. K nowlton.8 This paper is an excellent presentation of
Another point which should be considered is the method of the calculation of the total exhaust loss from a turbine. It
carrying peak loads. Some installations carry peak loads with appears worth while to supplement this work by a statement as to
the bleeder heaters cut out of service and this condition lends the background of the method and the reasons for believing that
itself to separate consideration by virtue of the added flow passed such calculations are adequate.
through an already crowded exhaust. In the first place, for any particular turbine the characteristics
Fig. 18 shows a typical “leaving-loss” curve for a reaction ma of the last stage wheel and buckets are, of course, known to the
chine of 35,000 kw. designer. The calculations necessary for Fig. 4 are fairly simple,
involving the aforesaid characteristics together with the laws of
H. G. H iebelek.7 We are pleased to see such a thorough flow for wet steam as they are understood.
definition and explanation of a subject which has caused much Fig. 7, however, requires something more than well-known
comment by power engineers within the past few years. rules. The flow of steam through the ordinary downward
We feel that it is a function of the turbine designers to define exhaust type of hood is rather complicated and resource must be
what should be understood by such losses and to show their rela had to tests to determine the characteristics of various types.
tive magnitude. Two ways of testing are open, namely, by means of models or by
From a strictly operating viewpoint, once the selection of a testing hoods on actual turbines in operation. In either case,
machine has been made, these losses are beyond control except to the test consists of measurements of the annulus static pressure
a limited degree by the plant men. In southern stations, such together with the steam flow through the last stage wheel.
as thq Deepwater (Houston) plant, the high circulating-water Of the two means, the model tests are easier and more in
temperatures which prevail throughout most of the year limit structive. We are able to make models from Ve to V 12 size or
the vacuum obtainable. In the north, however, with colder smaller, depending upon the size of the actual hood in question.
water conditions, particularly in the winter time, absolute pres We test these models using air as the flowing fluid, and are able
sures between 0.50 to 0.75 in. Hg abs may prevail for several to observe very closely the flow characteristics. This can, of
weeks in extreme cases. In some instances, due to the pressure course, be done in advance of the construction of the full-sized hood
from operating departments together with the natural conser in the factory.
vatism of designers, condensers are purchased for summer condi Any model tests should be checked if possible on the actual
tions. With such installations undoubtedly the magnitude of full-sized apparatus and we have been able to check in this case
these leaving velocities and exhaust losses is very great. From by making annulus pressure measurements in actual turbines
the operating viewpoint, attention should be called therefore to operating in some of the power stations in the country. The
the fact that many units are unable to use such high vacua effec agreement between model and full-sized hood is very good and
tively or that there is a limit beyond which it is not economical to justifies the model tests.
go. This would suggest reduced speeds of the circulating pumps, The curve marked “total loss” in Figs. 13 and 14 can be checked
resulting in a saving in auxiliary power. Further, on some types as to shape by still another means which is open to us. A tur
of condensers, refrigeration losses of the condensate are increased bine can be operated under test conditions so that the weight
under high vacuum conditions. flow through the last stage buckets is at a constant rate. Then
The author’s paper considers the performance of the unit as the turbine condenser pressure can be varied by bleeding air to
suming a constant back pressure of 1 in. Hg abs at the flange. the condenser, or by other means, and the variation of turbine
We believe it would be of interest to call attention to the variation output can be measured. This variation of output is really a
in the performance of modern condensers with variations in load difference between the change in available energy due to the
and with variations in circulating-water temperatures, such as change in exhaust pressure and the change in exhaust loss oc
occur from season to season. The following will illustrate the casioned by the change in exhaust volume flow. The change in
typical performance of a condensing unit for a 35,000-kw machine available energy is readily calculated from the steam chart,
under the conditions prevailing at Houston: leaving a change in exhaust loss determined. Whenever pos
•----- Steam condensed, lb p e r b r------- sible, therefore, we obtain these variable vacuum curves at
T e m p e ra tu re of 360,000 315,000 235,00 170,000 constant flow, as valuable aids and checks on our calculation
in let circ. w ater /------B ack pressure, in. H g abs----- -
50 deg F 0 .8 6 0 .7 8 0 .6 4 0 .5 9 methods. It is not always possible, since turbines of moderate and
65 deg F 1 .3 0 1 .2 0 1 .0 2 0 .9 4 large capacities must be tested, if tested at all, in the operator’s
80 deg F 2 .0 0 1 .8 5 1 .6 0 1 .4 4
95 deg F 3 .0 4 2 .8 2 2 .4 8 2 .2 1 power station and operating conditions or other factors may
make extended tests impracticable.
In the tabulation above, the flows which correspond to maxi It is evident from the foregoing that means of testing have been
mum load (40,000 kw), full load (35,000 kw), three-fourths and developed and are used for checking and substantiating the
half load happen to correspond very closely to the flows assumed methods of exhaust loss calculation as presented by Mr. Robin-
by the author. This tabulation will show a variation of approxi
mately 25 per cent in the specific volume of the steam to the
condenser between half load and maximum load for the same H. V. R asmussen.9 The author has written a very interesting
water conditions and a change of approximately 300 per cent paper which is of vital interest to the turbine designer, as the
for the same load between summer and winter conditions. This of
course neglects the consideration of moisture. 8 Turbine Engineering Department, General Electric Co., Schenec
tady, N. Y. Jun. A.S.M.E.
’ Assistant Superintendent of Power, Houston Lighting and Power • Westinghouse Electric and Manufacturing Co., South Phila
Co., Houston, Tex. Mein. A.S.M.E. delphia Works, Philadelphia, Pa. Mem. A.S.M.E.
524 TRANSACTIONS OF THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS
dimensioning of the last spindle row and the turbine-exhaust considerably larger pressure drop through the casing than the
casing has a deciding influence on both the turbine performance bare exhaust cylinder. However it was found possible to design
and the manufacturing cost. a deflector that was shaped as part of a rotative body and
If a turbine element in the high-pressure end of a turbine is not that reduced the pressure drop as it diminished the concentra
very efficient, up to one-half the losses may be recovered in the tion of flow at the condenser side of the exhaust opening.
rest of the turbine due to reduced moisture and increased heat The pressure measurements around the periphery showed a
drop, but losses in the exhaust end of a turbine are irretrievably somewhat similar distribution to that shown on Fig. 7 of the
lost to useful work. Any improvement obtainable in the ex author’s paper.
haust end will have a direct bearing on the over-all performance of
the turbine. R o n a l d B. S m i t h . 10 Low-pressure end losses consist of fric
While the performance of a turbine is improved by a large tion due to wall resistance and curvature losses in the exhaust
last-row annulus with a correspondingly low kinetic leaving loss, hood, and eddy friction resulting from the attempt to convert
the manufacturing cost of a turbine goes up rapidly with the kinetic energy at the blade annulus to potential energy at the
increase of the dimensions of the last row. Also, blade and condenser flange. Physically these losses are an evaluation of the
spindle stresses place a definite limitation on the physical dimen well-known relation
sions involved. A proper compromise between these various
factors must, therefore, be established in practical turbine design.
The present tendency toward large single-cylinder turbines
makes it necessary to employ high peripheral blade speed in the
last row, with a resulting high steam speed and a large kinetic in which R is the total low-pressure end loss, i the enthalpy of
leaving loss. The place for improvement, consequently, is in the the steam, A the Joule conversion factor, with the subscripts
exhaust casing which should be designed to offer the smallest 1 and 2 referring to conditions at the exhaust annulus and the
possible resistance to flow from the last-row annulus to the exhaust exhaust flange, respectively. In the design of the exhaust we
opening. are faced with four characteristic solutions of the relation ex
It would be interesting to know how the author arrived at the pressed by Equation [1], First, there may be complete recon
annulus pressure drop shown on Figs. 5 and 7, and also how it was version of the kinetic energy at the exhaust annulus into poten
established that some exhaust casings are actually diffusing. tial energy. In this case we have an isentropic change in which
Measurements on actual turbines are very difficult to obtain as
they involve measurements of static pressure in the high velocity
jet. Some attempts were made to measure the pressure drop
in an exhaust casing of a large Westinghouse turbine by tapping and the low-pressure end losses AR become zero. This inter
the cylinder casing at various points in the cover and base. change of energy is contrived through a completely reversible
No conclusive results were obtained from this investigation as it process and while it represents the ideal in achievement it is,
was obvious that a velocity head created by the steam impinging in the light of our present knowledge of the physical laws, quite
against the measuring hole obscured the results and, at the very beyond the realm of possibility.
best, these measurements would only record the pressure exist The next possible solution to Equation [1] lies in the recon
ing at the periphery of the casing. They could not disclose the version of a portion of the kinetic energy of discharge by a diffusor-
pressure distribution in the middle of the casing. Measurements shaped exhaust chamber. It is represented physically by the
with pressure-measuring tubes, such as the Fechheimer tube, were boundary condition > pi and by (i2 — ii) > A f vdp. Prac
given up as impracticable, as the moisture in the steam would tically, this solution is the aim of all good designs since the total
partly fill the passages and cause incorrect readings. low-pressure end losses become less than the leaving-velocity
Another approach to this problem is to conduct tests with loss at the blade annulus. Unfortunately, mechanical restric
small-scale models. A number of exhaust model experiments for tions, imposed largely by the purchaser, have prevented many
various exhaust-cylinder designs were run by the Westinghouse exhaust-end designs of this type in the United States, although in
Company. Wooden models of the exhaust casing were made to Europe the practise is used frequently to advantage.
Vs of full size and air was blown through the models. A rotating A solution almost akin to the one just described lies in a partial
blade row, mounted on a disk and driven by a motor, represented reconversion into pressure which is subsequently lost through wall
the last spindle row. Pressure measurements were taken with a friction in the exhaust chamber. In this case pi = p2, and the
Fechheimer tube at a number of points around the blade annulus low-pressure end loss is equal to the kinetic energy of the fluid
and the velocity distribution over the exhaust opening was from the last-blade annulus. This assumption was widely used
recorded with a Prandtl impact tube. in turbine design up to the past few years but increasing demands
These experiments disclosed a number of interesting facts. for economy have necessitated a solution more in keeping with
First of all, it was found that steam is distributed most unevenly the actual condition. This solution usually consists not only in a
throughout the casing and the exhaust opening. The steam complete loss of the kinetic energy of discharge but in an addi
clings to the generator side of the exhaust casing and also crowds
tional pressure drop as well (p2 < pi). In this case the a J^2 vdp
this side of the exhaust opening, while the part of the exhaust
opening that is nearest to the turbine is hardly filled. If a part is negative. This is the condition to which the author devotes
of the exhaust opening is located under the bladed part of the himself, and while far from the ideal it is at present the most
cylinder, the pressure drop through the casing will increase con common. The real problem in this design lies in the determina
siderably. tion of the pressure drop. Since measurements on actual ma
The tests also showed that ribs and steam deflectors in an chines are practically impossible, models are usually prepared.
exhaust cylinder might improve the distribution of the steam over However, the requirements of similarity are not wholly satisfied
the exhaust end, but generally accomplishes this at a cost of an in the model due to the extreme velocity and pressure conditions
increased pressure drop from the last row to the exhaust opening. that exist in the usual exhaust end.
The older exhaust-cylinder design with a number of separate 10 Experim ental Engrg. D ept., W estinghouae Elec. and Mfg. Co.,
passages from the last row to the exhaust opening had thus a South Philadelphia W orks, Philadelphia, Pa. Jun. A.S.M .E.
FUELS AND STEAM POWER FSP-56-10 525
As the author has already pointed out an expression for the benefit of a generously dimensioned exhaust annulus may be lost
low-pressure losses in per cent of the adiabatic enthalpy change unless the hood is properly proportioned and dimensioned.
means little unless one is familiar with the type of cycle. A It would be of great interest to know by what means the pres
term more closely representing the efficiency of the low-pressure sure distribution shown by Fig. 7 was obtained. We have made
end design proper would be the ratio of the average kinetic attempts at similar measurements and have been forced to con
energy at the exhaust flange to the total frictional loss. clude that it is exceedingly difficult to get the pressure readings
inside the exhaust of an actual turbine. Fig. 7 indicates a de
gree of precision which I know is very difficult to reach.
A t jt h o r ’ s C l o s u r e
pressure-volume energy, temperature energy, velocity energy, velocity and exhaust loss by itself. Although the distinction
and if there is supersaturation, it may be necessary to include might be thought of as only a matter of definition, it is none the
surface-tension energy. As a short cut from the true end-point less important.
to the wheel exit, one might draw a horizontal line to the left to Mr. Hiebeler points out the important effects of varying cool-
annulus pressure and thence down the pressure line so as to de ing-water conditions on the vacuum and consequently on the
duct the velocity energy fraction of the leaving loss. That magnitude of the leaving losses. The diagrams in the paper have
would locate what might be called the “chart condition” of the- been based on 1 in. Hg abs back pressure. There is a certain
steam at the wheel exit, which would be useful to give the mois back pressure below which better vacuum fails to yield any more
ture content. This is to be distinguished from what would be power. Refrigeration, in itself, is detrimental. In deciding
the end-point of the machine if it had zero-leaving loss. The how much water to pump it is necessary for an operator to bal
important point in this matter is to base predictions upon exactly ance his pump requirements and refrigeration losses against the
the same conventions used in analyzing performance. additional power shown by his vacuum corrections.
Mr. Crocker has pointed out how necessary it is for an operator Mr. Knowlton’s discussion anticipates the questions raised
to have all the essential information about his machine. Cer by Messrs. Rasmussen and Soderberg about the pressure meaure-
tainly the manufacturer intends to furnish all such information ments in exhaust hoods as shown in Figs. 5 and 7 and it does not
and if it has not been done in the past, that must have been due seem necessary for the author to add anything further on that
to a misunderstanding of the needs in the particular case. subject.
Mr. Franck has called particular attention to a number of Mr. Soderberg also mentions the importance of proper consider
items and rightly expresses his concern about the flow area of ation of moisture losses and here again it may help in clarifying
the last row of blades. That is usually the most important fea the situation to point out that moisture effects have been consid
ture of all, and, if the author did not show proper concern about ered in the paper only in so far as they affect the leaving-velocity
it, that was because of its general recognition. Mr. Franck’s and exhaust loss itself. Changing degrees of moisture in pre
remarks about the effects on the preceding stages help to bring ceding stages affect the turbine performance and have to be con
out the distinction which is necessary between the total net effect sidered in predicting vacuum corrections but such effects are
on turbine performance of a change of vacuum, and the leaving- not a part of the leaving loss even though in part due to it.