American Technical Society.

Cyclopedia of engineering : a general reference work on steam boilers, pumps, engines, and turbines, gas and oil engines, automobiles, marine and locomotive work, heating and ventilating, compressed air, refrigeration, dynamos motors, electric wiring, electric lighting, elevators, etc. (Volume 2) online

. (page 25 of 30)
Online LibraryAmerican Technical SocietyCyclopedia of engineering : a general reference work on steam boilers, pumps, engines, and turbines, gas and oil engines, automobiles, marine and locomotive work, heating and ventilating, compressed air, refrigeration, dynamos motors, electric wiring, electric lighting, elevators, etc. (Volume 2) → online text (page 25 of 30)
Font size
QR-code for this ebook

would consume less than 2% of the power.

The most important losses are due, first, to the friction of the steam
jet against the vanes and guides, which will be approximately propor-
tional to the cube of the velocity of the steam relative to the vanes
or guides, and second, to the considerable amount of friction of the
disks as they revolve in the chamber filled with steam. This friction
generates heat which raises the temperature of the steam and metal
parts and thus causes the re-evaporation of some of the condensed
moisture. Since this adds some heat to the expanding steam, the
expansion is not absolutely adiabatic. The smoother the revolving
wheels are made, the less will be this friction, a fact well illustrated by
a reported improvement of about 1% in steam consumption which
was effected in a well-known make of turbine by making the riveting
on the revolving disks perfectly flush. To these losses may properly
be added the' generator losses which, of course, are a factor of the
speed of revolution.

With either reciprocating engines or turbines, the steam economy
is much better in large than in small units, and especially is this
true of the turbines of the reaction type. In small turbines of this
type, the steam friction is high and the leakage large, and this makes
it undesirable to build this type of turbine in sizes much below 500
kw. For the impulse type of turbine, these losses are not as impor-
tant in the smaller powers, and DeLaval, Curtis, and Rateau turbines
of comparatively small power can be built to give nearly as good


steam economy as larger turbines of the same type, and can easily
excel small reciprocating engines.

The steam consumption of the turbine depends naturally
enough upon the vacuum, steam pressure, degree of superheat,
variation in load, and variation in speed. It has already been
explained that the turbine can utilize the lower ranges of vacuum
far better than can the reciprocating engine, but it could not, in all prob-
ability, use the higher pressure ranges with as good economy as the
best reciprocating engines. If the turbine runs at a vacuum of 27 in.,
its steam consumption will be practically on a par with that of the
reciprocating engine, and it will show a gain of about one-half pound
of steam per kw.-hr. for each extra inch of vacuum, below 25 in. But
from the saving effected by this one-half pound of steam must be
deducted the extra cost of maintaining the high vacuum, if the real
economy is desired. Not only can the turbine theoretically utilize
the greater vacuum to better advantage, but it has an advantage also
in a practical way, because with the reciprocating engine, a very high
vacuum cools the cylinder walls and thus causes a relatively large
initial- condensation, which difficulty is not met with in the turbine,
the high vacuum having no detrimental effect. It thus has both a
theoretical and practical advantage.

Superheated steam, whether used in the reciprocating engine
or in the turbine, will reduce the steam consumption; but in the re-
ciprocating engine, superheating cannot be carried very high, as
the cylinder lubricant is likely to be burned, and there will be little
condensation in the cylinder to help out the lubrication. The tur-
bine is not handicapped in this way, but nevertheless high degrees
of superheat are likely to cause trouble due to unequal expansion in
the casing, the temperature at the high-pressure end being so much
greater than that at the low-pressure end. This expansion is trouble-
some, but should be provided for in the design.

Superheat affects the economy of the steam engine in two ways ;
it carries additional heat units into the cylinder, and lessens con-
densation. It also helps in the turbine in two ways; it carries
additional heat into the turbine, and, being less dense than saturated
or moist steam, causes less friction within the turbine, and thus
effects a mechanical as well as a theoretical gain. It is generally
reported that the gain is 10% for each 100 of superheat, but tests


which appear to be thoroughly reliable do not seem to bear out this
claim. 1\% to 8% is a better figure.

The saving in steam will be from 1.5 to 1.75 pounds per
kw.-hr. for each 100 of superheat, but the real economy resulting
from this superheat will be the difference beween the value of
this saving in steam and the cost of superheating. The superheating
plant costs more, not only for the additional expense of the super-
heater, but for piping, valves, etc. Cast-steel fittings, and valves
with nickel-steel valve stems, are usually required for high degrees
of superheat.

The usual steam pressure in turbine work is about 150 pounds
gauge. If lower than this, some gain in steam consumption may be
had by an increase in boiler pressure, but an increase over 150 pounds
does not appear to be productive of great economy. A reference to
Fig. 25 will readily show that increasing the pressure above 150
pounds will add very little to the area of available work. Fig. 30
shows the curves of economy of a 30-H. P. turbine at different steam
pressures. The gain is less and less the higher the pressure becomes,
and is small from 75 to 100 pounds. From 35 to 100 pounds the
gain is about 33 \/ c , but this gain is not due entirely to the rise in
steam pressure.

The study of steam nozzles has shown that to use steam efficiently,
the nozzle must be properly designed with reference to both the
initial and final pressures. Now, if the nozzle on this turbine were
designed for 100 pounds pressure, it could neither utilize steam
economically at 35 pounds, nor at 150 pounds pressure. To show
the real gain due to an increase in steam pressure, it would be
necessary to have nozzles in each case that were designed for the
specific pressures used. Then, and only then, would the curves
show the true gain due to increase in pressure. But a study of
Fig. 25 shows that if the theoretical gain is small the practical gain
cannot be large. It must, moreover, be borne in mind that a high-
pressure plant costs more than a low-pressure plant, and for
stationary work very high pressures will not pay. On shipboard,
where space and weight are at a premium, it may be good engineering
policy to install very high pressures, even though the first cost is

Fig. 31 shows the curves for a 600-kw. Curtis turbine with vary-



ing pressures. In this type of turbine, the same conditipns exist as
in the previous one, the nozzles being designed for only one pressure.
The economy of a turbine varies with the load, as does the econ-
omy of the reciprocating engine, but not perhaps to as marked an
extent, and the economy depends of course upon the type of tur-
bine. Turbines like the DeLaval and Curtis admit steam through
a number of nozzles which are opened and closed either automatically
by the governor or by hand. At normal load, about two-thirds of these
nozzles would be open and a 50% overload could then be carried
with all nozzles open. In the Parsons turbine, steam is admitted
all around the circumference of the drum but the admission is in-
termittent. For heavy loads the valve remains open for longer in-




/=? S


i ru


e Mo








35tbs. steam press u












5 /0 A5 c>0 ^5 30 J5

Fig. 30. Curves Showing Economy of 30-H. P. Turbine.

tervals, and when the load is such that the valve remains open all the
time, further overloads can be provided for only by resorting to a by-
pass which admits high-pressure steam to the second stage of the
turbine. In such cases, of course the economy falls off, for the
steam does not get the benefit of full expansion. At low loads, there
is not a great deal of choice between the different types of turbine,
but those that can carry a large overload without opening a by-pass
are bound to be the most economical under these conditions.

Overload is taken care of in a reciprocating engine by increasing
the cut-off, but, as this reduces the number of expansions, this method
is uneconomical. For small ranges of load, the relative economy of
turbine and reciprocator are not very different, but the effective
range of the turbine is much greater than for the reciprocating engine.



A good turbine will carry 100% overload for a short time and will
carry 50% to 60% overload on approximately 10% more steam.
Fig. 32 shows characteristic curves of steam consumption at varying

A variation in speed of the turbine within moderate limits does
not materially affect the economy. The best speed of the vanes
(see Page 17) is half the velocity of whirl (iVcosa). Moderate
departures from this speed do not materially affect the economy,
provided the entrance angles of the vanes are such that the steam
jet strikes without shock. The angle of the vane must depend upon
the speed, and once fixed,
any variation in speed Zl \
must of course cause the
steam jet to spatter and
form eddies, a source of 30
material loss. This is
entirely apart from the
question of whether or not
the designed speed of ro- ' 3
tation is the most econom-
ical. To avoid spatter-
ing and eddy losses, the fg
vane angle must change
with the speed, which is
evidently impossible

A rapid change in load will cause cylinder condensation in a
reciprocating engine, so that, on test under steady load, the engine
is likely to show up better than it would under service conditions.
With the turbine, this is not so. Here there is no such condensa-
tion, and the performance under test is far more likely to agree with
performance under service conditions. Both types of motor will
fall off under service conditions, but if an engine and turbine do
equally well under test, under such widely varying conditions as
exist in a central station, for instance, the turbine ought to show
up better in actual service.

A reciprocating engine is usually designed for a low average
load and, therefore, it will permit a relatively large increase in load,
but it is generally working on a slight underload, and hence at less

31. Economy Curves of 600-K. W. Curtis Tur-
bine with Different Steam Pressures.



than the maximum efficiency. The turbine, on the other hand, is
usually designed for its normal and most economical load, taking
care of overload by opening more nozzles at theoretically the same
efficiency, or by opening a by-pass at somewhat less efficiency. This
should give the turbine a still further advantage at the end of the
day's work.

Tests. Tests of reciprocating engines usually give steam in
pounds per indicated horse-power per hour, but there being no









s\ i





? TL











I lo









* I


of curves

















Fig. 32. Economy Curves of Turbines and Compound Corliss Engine with
Varying Loads.

indicated horse-power for a turbine, the comparison must be made
on some other basis. Brake, or shaft horse-power may readily be
obtained for a turbine, and in engine tests, where the brake horse-
power has been determined, there is of course opportunity for a
direct comparison. However, since engineers are in general more
familiar with steam rates per I. H. P., it seems well to consider
how a comparative I. H. P. may be had for the turbine. Various
tests to determine the relation between brake and indicated power


on reciprocating engines seem to show that 02% is a fair figure
for a good engine. 92% then of the steam rate per brake horse-
power would give the rate per comparative indicated horse-power.

The largest field for the steam turbine being central station
work, it follows that by far the larger number of turbine tests are
quoted in terms of electrical units. It is costly to fit a brake for
a large turbine and entirely useless when the power delivered at the
switchboard can be read off at once. For electrical work, of course
reciprocating engine tests are often quoted in the same electrical
units, in which case, there are abundant opportunities for direct

Suppose, however, that it is desired to compare steam per
I. H. P. with a corresponding rate per kilowatt-hour at the switch-
board. 1 kw. = 1.34 electric horse-power measured on the switch-
board, which is evidently shaft or brake output less losses in the
generator. Since the efficiency of a good generator is not far from
95%, the brake horse-power will be equal to the electric horse-power

divided by -TOO' ^ e ma ^ sa ^' therefore, in ordinary cases, that

B. II. P. 92
since we assume, that T TJ~~ FT = Tnn'

1 34 X kw
we have I. H. P. = 95 x ~092" = L53 kw ' a PP roximateIv -

Steam per kw.-hr. then, divided by 1.53 would give the steam per
comparative indicated horse-power per hour, or

steam per I. H. P.-hr. X 1-53 = steam per kw.-hr.
A turbine using 20 Ibs. of steam per kw.-hr. would be about on a par

with a reciprocating engine using ^-^ =13 Ibs. per I. H. P.

In comparing the performance of one engine with the perform-
ance of another, or one turbine with another, or an engine with a
turbine, pounds of steam per horse-power per hour is generally the
rough basis of comparison, but this is very crude and often mislead-
ing. For instance, one test may be made with superheated steam
and another with saturated or even moist steam, or one may have a
higher steam pressure, or the vacuums may be different.



To get an approximately intelligent comparison, all tests should
be reduced to a standard degree of superheat, pressure, and vacuum,
or better still, if the comparison is between two, correct both to the
average conditions of the two. The corrections applied are more
or less arbitrary, and it is manifestly unfair to apply them all to
either test. If each is corrected for half the difference, a much more
reliable comparison is likely to result.

It is generally accepted that the steam consumption will de-
crease about 8% for each 100 of superheat, about 5% for each inch
of vacuum below 28 in., and about 5% for 50 Ibs. rise in steam pres-
sure between 100 and 150 Ibs., and 3% for similar rise between
150 and 200 Ibs. The manufacturer usually gives guarantees of
steam rates for various pressures, vacuums, and degrees of super-
heat. When such figures are available, it is probable that their
use would lead to more satisfactory results than if the rough approxi-
mations mentioned above were used, but such figures would be cor-
rect only for the one individual turbine, and in the large majority
of cases the engineer is compelled to use the approximations. They
are in most cases fair and satisfactory in the absence of definite data.

To illustrate this method, consider a turbine at 177.5 Ibs. (gauge)
steam pressure, vacuum 27.3 in., superheat 96 F., consuming 15.15
Ibs. steam per kw.-hr., and another using 179 Ibs. steam pressure,
29.5 in. vacuum, and 116F. superheat, consuming 13 Ibs. of steam
per kw.-hr. The average conditions are 178.2 Ibs. steam pressure,
28.40 in. vacuum, and 106 F. of superheat.

The work will appear clearer if arranged in tabular form as
in Table I.

Steam Consumption Tests









177.5 Ibs.

179 Ibs.

178.2 Ibs.


27. 3 in.

29.5 in.

28.4 in.








4 0.8%


15.15 Ibs.

13 Ibs.

-6.3% cr

+6.3% or



14.19 Ibs.

13.82 Ibs.

-O.G6 Ib.

+ 0.82 Ib.



The correction for steam pressure, being only for .7 Ibs., is too
small to be of consequence in this case. The vacuum correction is
1.1 inches, and at 5 % per inch (the decrease in steam consumption
for each inch of vacuum, as explained on Page 52), the correction would
be 5|%. The superheat correction is for 10, or, as the decrease
in steam consumption for 100 of superheat is 8%, this will be T V of
8% = 0.8%. The sum of these corrections gives 6.3%, making
.96 Ibs. to be subtracted from turbine #1, and .82 Ibs. to be added to
turbine #2. The final steam consumptions, then, which should be
compared are 14.19 Ibs. and 13.82 Ibs. instead of 15.15 Ibs. and 13
Ibs. Turbine #2 appears, therefore, to use about 3% less steam
than turbine #1 under similar conditions.

Another and perhaps more satisfactory method of comparison
is by means of the heat units used. This computation may be made
readily from the steam tables. Using the same tests as given above,
turbine #1 uses steam at 177.5 Ibs. gauge pressure = 192.2 absolute,
at which pressure each pound of dry saturated steam contains 1197
B. T. U. If we allow \ B. T. U. for each degree of superheat, then,
for 96 F. we should add 96 X. 5 = 48 B. T. U., and each pound
would then contain 1245 B. T. U. at admission. If this steam is
condensed at a pressure of 27.3 in. vacuum = 1.33 Ibs. absolute,
each pound of the condensation will contain 80 B. T. U. which will
be returned to the boiler in the feed water. The net amount then
consumed by the turbine and carried away by the cooling water
of the condenser is 1245 - 80 - 1165 B. T. U. per pound. 15.15
Ibs. would represent 15.15 X 1165 = 17,650 B. T. U. per hr. or 294
B. T. U. per kilowatt per minute.

Turbine #2 uses 13 Ibs. of steam at 179 Ibs. gauge pressure and
116 F. superheat, condensing at 29.5 in. vacuum. In this case, each
pound of dry steam at admission would contain 1197.3 B. T. U.
and 116 F. superheat would add about 58 B. T. U. more, making
1255.3 B. T. U. per pound. Condensing at 29.5 in vac. = .25 Ibs.
absolute, each pound of condensed water would contain 27 B. T. U.
to return to the boiler in feed water, leaving 1255.3 27 = 1228.3
B. T. U. to be used by the turbine. 13 Ibs. would represent
13 X 1228.3 = 266 B T n _ per min to compare with 294 in the

previous case.


Here, again, the direct comparison is likely to be misleading,
unless due account is taken of the difference in conditions. The
gain is apparently about 10% in favor of turbine #2 on the heat unit
basis taken under the actual working conditions of each, but the fact
must not be lost sight of, that turbine #2 is working under more favor-
able conditions of vacuum and ought to show a much better efficiency.
It appears from this discussion that both turbines work under the
conditions of design with but little difference in actual economy.

Turbine manufacturers are in the habit of reporting tests of
the turbine only, no account being made of the auxiliary apparatus.
This is manifestly misleading, for with a 29-in. vacuum, the power
consumed by auxiliaries may easily be twice what it would be for a
27-in. vacuum. This extra power and the cost of maintaining it
in a measure goes to offset the gain due to the higher vacuum.





*In this description of commercial turbines it will be convenient
to classify them as follows:

'Single-Stage Type

f Compounding by Velocity Steps only
I Compounding by Pressure Stages only


Compound Type

Compounding by both Velocity Steps

[ and Pressure Stages


Probably the simplest type of turbine is the one with a single
stage, that is, a single set of nozzles and a single rotating wheel,
but, as already pointed out, the velocity of rotation in a turbine of
this sort is so great under ordinary conditions that some device must
be employed to reduce the rotational velocity to more convenient
speeds. This may be done in t\vo ways.

As has been pointed out before, the feature of importance is
not the rotative speed, but the peripheral velocity of the wheel, this
being somewhat less than one-half the steam velocity. Maintaining
this peripheral velocity constant, turbine rotors of comparatively
small diameter may be used, the high rotative velocity being reduced
by means of gearing; or, the diameter of the turbine rotor may be
increased the rotative speed thereby being reduced in the same ratio
that the diameter of the wheel is increased.

*Many writers group by themselves all turbines using buckets of the Pelton
type, but this does not seem to be a proper classification, as it is the action of steam in
the turbfhe that makes it belong to a certain type, ami not the style of bucket that is
used. Turbines using Pelton buckets may belong to any of the impulse groups.

Copyright, l'jof>, by American School of Correspondence.



Both of these methods have been employed in turbines that have
been put upon the market, the first method being characteristic of the
DeLaval turbine, and the second being employed in the earlier forms
of the Riedler-Stumpf machine. The manufacturers of the latter
discarded this scheme in their later designs in favor of some sort of
compound turbine.

DeLaval Turbines. The turbine designed and developed by
Dr. Gustav DeLaval, of Sweden, was among the first to be commer-
cially successful. His first turbine, which was used to run the famous

Fig. 33. Principle of Operation of DeLaval Steam Turbine.

DeLaval cream separators, was of the pure reaction type, similar
in action to the old Hero engine. This turbine was not economical
in steam consumption, but, as it was used for very small powers only,
this factor was not important and, commercially, the machine was very
successful. This success led to the desire to build larger turbines,
and in their development the reaction principle was abandoned.

The essentials of the motor element of the DeLaval turbine are
illustrated by their familiar trade mark, shown in Fig. 33. They
consist of a rotating disk, having vanes on its periphery; a number of
nozzles in which the steam is expanded from boiler pressure to the



pressure in the exhaust chamber and delivered in a jet against the
vanes; a long slender shaft to which the rotating wheel is fixed, so ar-
ranged that at high speeds the rotating element can revolve about its
own center of gravity* instead of its geometrical center; and a set of
reducing gears to reduce the high rotative speeds to the desired amount.
It is an impulse turbine with a single wheel, carrying one row of
buckets, and is a single-stage turbine in all respects. The steam is
directed against the vanes from nozzles with flaring sides, so designed
as to give it the maximum velocity and to expand it within the confines
of the nozzle to the pressure in the exhaust chamber, thus transforming
all of the heat energy of the steam into kinetic energy. The nozzles
deliver the steam jets at the side of the wheel, and for a maximum
efficiency should make as small an angle as possible with the plane of
rotation. f In the DeLaval machine this angle is 20.

For small turbines, the entrance and exit angles of the vanes
are 32, increasing to 30 for the larger sizes. Under these conditions,
the best peripheral velocity will be about 1900 feet per second when
the velocity of the steam issuing from the nozzles is 4000 feet per sec-
ond. In most impulse turbines the peripheral velocity varies from
1400 in the larger sizes to 500 in the smaller sizes. These speeds are
high, even for turbine work, and necessitate the solution of very in-
teresting engineering problems. These velocities, with the diameters
used for DeLaval machines, mean revolutions of about 10,600 per
minute in the larger sizes, and 30,000 per minute in the smaller sizes,
these speeds being reduced by helical gears to approximately 900 and

Online LibraryAmerican Technical SocietyCyclopedia of engineering : a general reference work on steam boilers, pumps, engines, and turbines, gas and oil engines, automobiles, marine and locomotive work, heating and ventilating, compressed air, refrigeration, dynamos motors, electric wiring, electric lighting, elevators, etc. (Volume 2) → online text (page 25 of 30)