Tests of low head, high volume farm pumps

Material Information

Tests of low head, high volume farm pumps
Stephens, J. C.
Craig, A. L.
Speir, W. H.
Place of Publication:
Gainesville, Fla.
University of Florida Agricultural Experiment Station
Publication Date:
Copyright Date:


Subjects / Keywords:
The Everglades ( local )
Pumps ( jstor )
Speed ( jstor )
Power efficiency ( jstor )


General Note:
Technical bulletin - Florida Agricultural Experiment Station ; 565

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
003246891 ( Electronic_Aleph )
60385575 ( Electronic_OCLC )


This item has the following downloads:

Full Text

AUG 19 1 55

Bulletin 565

June 1955

(A Contribution from the Everglades Experiment Station)
Cooperating with
Soil and Water Conservation Research Branch

Tests of Low Head, High Volume

Farm Pumps



Single copies free to Florida residents upon request to


INTRODUCTION ................- ....... .
DESCRIPTION OF PUMPS ...........-..--
Test Site ................... ....... .. ... .. .
Driving Unit ......................... ......
Speed Variation and Measurement ...
Lift Variation and Measurement ......
Discharge Measurement .....-.........-
Power Measurement .........................
Test Procedure ........ ........ ................
DISCUSSION ................. ---- ... .....
EXPERIMENTAL RESULTS ---...........---.--
Centrifugal Type .........-..-.... ---
Helical Type .........................
Propeller Type ............... .- ... -.
Identical Pumps ........... ..........---
Similar Pumps of Different Size ......

................ -.-.. -. -..- -.. -. -...- .-- 3
---.-.-.....-..-...-..... -.. ... -- -- ----- 5
-..-- ---- ---- - ----- ---------- 8

-...... ...---.-.- ...-... ............-- .. 12
....... .. .. ... .. ....... ---.-----. 12
....... ...........--- ------- --....-- 13
........................ -----...... ......... 13
S------ - --- 14
..-..-..-- ...- ...-- . ------..--. --..- 15
-- -- ------ ----- 16

........................ .-... .. ... 1520
.--... .. .. ------- ------ 20
...... ................... -- -- 20
-... . -.. ... --... --- -... -. -..- -- 20
..........- -------------.... 27
... ............... ....------- .... ... 27
..... .... .. -. .- .. -----.----- --. 31


Operation tests of three locally manufactured low lift, high
volume farm pumps typical for the Everglades area were per-
formed over wide ranges of speed for lifts of from 3 to 9 feet.
The pumps tested were a 20-inch double-suction centrifugal
type, a 20-inch helical type and a 24-inch propeller type.
The testing procedure, including speed, head, discharge and
power measurements, is described.
A discussion outlines some of the factors influencing the effi-
ciency of a low head pump.
Test results are presented in tabular and graphical form. For
normal operating conditions, the pool-to-pool efficiency of the
centrifugal type pump varied from 30 to 40 percent and the
propeller type pump from 45 to 60 percent. For a 5-foot lift,
the efficiency of the helical type pump varied from 28 to 37
percent. Sample problem solutions utilizing the graphs are
included, and relations are given for approximate solutions of
problems involving similar type pumps of other sizes.

Tests of Low Head, High Volume

Farm Pumps


The capacity, power requirements and efficiency of pumps
vary widely with the speed, lift and design of the pumps. It
is customary for the major pump manufacturers to study the
relationship of these variables, and to supply performance
curves on which the head, horsepower and efficiency are plotted
as functions of the discharge. This enables the purchaser to
select the most suitable unit and to operate it most economi-
cally under given pumping requirements.
The majority of farm pumps in the Everglades, however, are
made by local fabricators who do not have facilities to conduct
such performance tests. Inadequate knowledge of pump char-
acteristics has hindered users in selecting the unit best suited
for their needs and in providing the proper power unit to drive
the pump at the most economical speeds for various lifts and
In response to requests from various agencies, performance
tests were made on three types of pumps in common use for
water control on farms and ranches in this area. These pumps
were manufactured locally and were tested under identical
conditions, insofar as possible, with sufficient ranges in speed
and lift to give adequate information for normal pumping op-
erations in the Everglades area.
The Soil and Water Conservation Research Branch2 of the
Agricultural Research Service, U. S. Department of Agriculture,
the Central and Southern Florida Flood Control District and
the Florida Agricultural Experiment Station cooperated in per-
forming the tests. These were made at the Everglades Experi-
ment Station pump house near Belle Glade in 1952.
Research Project Supervisor, Agricultural Research Service and
Drainage Engineer, Everglades Experiment Station; Hydraulic Engineer,
Agricultural Research Service; and Engineering Aide, Agricultural Re-
search Service, respectively.
SFormerly Division of Drainage and Water Control, Research Section,
Soil Conservation Service.

Propeller Type Pump

Fig. 1.-Partial-sectional side views of pumps tested.

Type Pump

Helical Type Pump

Tests of Low Head, High Volume Farm Pumps

Three pumps, each representative of a different type, were
tested. Partial sectional views of these pumps are shown in
Figure 1. All were installed in typical open-pit settings.
The first pump tested was a panel-mounted, modified-cen-
trifugal type having a vertical shaft and double intake open-
ing with peripheral flow produced by an impeller within a
volute casing. This impeller consisted of three straight verti-
cal vanes of 8-inch height enclosed between two annular rings
of 20-inch inner and 26-inch outer diameter. (Figure 2.)
Intake openings in the horizontal casing above and below
the impeller both were 20 inches in diameter. Flow produced
by rotation of the impeller was discharged into a volute cas-
ing of rectangular cross-section which terminated at a flap
valve 47 inches wide and 16 inches high. Figure 3 is a side
view of this pump, raised in its mounting grooves, showing the

Fig. 2.-View of discharge side of centrifugal pump, showing termination
of volute casing, impeller and lower intake opening.

- .I -~~

Florida Agricultural Experiment Stations

upper intake opening, the volute casing and the raised flap
valve. Direction of flow can be reversed by lifting the pump
and rotating it 180 degrees.
The second pump tested was a panel-mounted, vertical, heli-
cal type. The impeller was a continuous helix of 20-inch diam-
eter, 20-inch pitch and 8-foot length. A 3-inch-diameter steel
shaft through the axis of the helix supported the impeller, which
rotated within a cylindrical steel tube 9 feet long.
Flow entered the pump through a rectangular cut 15 inches
high extending around half the circumference, in the submerged
end of the tube. Flow from this tube entered a curved dis-
charge box at the top which changed the direction of flow from
vertical to horizontal.
Since the discharge opening was above the maximum pool
stage, the variation in head for the tests on this pump was
limited to varying stages in the intake canal. This pump was
built specifically for these studies, and, after observing the
Fig. 3.-Side view of panel-mounted centrifugal type pump, showing volute
casing, upper intake opening and discharge flap valve.

mur: (EEU

Tests of Low Head, High Volume Farm Pumps

results of a series of tests, the manufacturer improved the in-
take and discharge transitions. Only the results of the tests
after these modifications were made are presented herein. The
discharge side of this pump is shown in Figure 4.
The third pump tested was an axial flow, vertical, propeller-
type pump mounted in a double-decked chamber with intake

Fig. 4.-Replacing centrifugal type pump on left with helical type
pump for testing. View of discharge side of helical type pump shows
rectangular discharge box at top.

Florida Agricultural Experiment Stations

from the lower chamber and discharge to the upper chamber.
The 3-bladed impeller of varying pitch, with an outside diameter
of 24 inches, 10-inch hub diameter and 8-inch depth, was of
cast and welded construction. Short flow-straightening vanes
were mounted in front of and behind the impeller. Details of
this pump appear in Figures 5, 6 and 7. Direction of flow is
controlled by shifting stop logs within the grooves of the upper
and lower chambers.

Test Site.-The pump station which served as the site of
these tests is the discharge portion of the water control sys-
tem for the Everglades Experiment Station tract of about 800
acres of peat soil. The drainage system is comprised of mole
drains leading to lateral ditches which, in turn, drain into the
main canal which terminates at the pumphouse. (Figure 8.)

Fig. 5.-View of disassembled propeller type pump. Intake bell with
lower bearing and lower flow-straightening vanes in foreground.

Tests of Low Head, High Volume Farm Pumps

Since this ditch system also supplied all irrigation water
to the experimental tract, nearly all of the water pumped dur-
ing the tests was withdrawn from channel or soil storage within

Fig. 6.-View of lower end of disassembled propeller type pump, showing
three-bladed impeller and upper flow-straightening vanes.

Florida Agricultural Experiment Stations

the tract. During the tests it often was necessary to allow
water to flow back through the pumphouse into the ditch sys-
tem during the night to replenish storage for the next day's
tests. This condition confined the tests to the wetter months
of the year so that enough water would be available for pump-
ing without seriously upsetting the water table in the experi-
mental lands.
The pumphouse has a concrete floor with three pump wells
built to accommodate farm pumps as typically installed in open
pit settings in the Everglades. Two of the pump wells contain
vertical grooves for panel-mounted pumps, and the third has
two decks with floor fittings for the axial flow, chamber type
pump. Each pump well has vertical grooves for stop logs at
the intake and discharge bays.
The pumps were set to discharge into a stilling basin or
sump. This sump is an oval-shaped basin 85 feet long by 65
feet wide and 9 feet deep, surrounded by muck embankments.

Fig. 7.-View of disassembled propeller type pump, showing its shaft
in right foreground and discharge bell, float valve and upper mounting
assembled on shaft tube.

Tests of Low Head, High Volume Farm Pumps


SR 80 II



Fig. 8.-Plan of test site.

12 Florida Agricultural Experiment Stations

A rectangular concrete flume, 8 feet wide, 10 feet deep and
10 feet long, provided with straight wingwalls, conducts drain-
age water from the sump to a 77 x 57-inch oval culvert dis-
charging into the Hillsboro Canal. At the sump end, the sides
of this flume contain vertical grooves for installation of stop
logs and/or a measuring weir. (Figure 9.) Without controls
in the flume the water level in the sump is approximately equal
to the Hillsboro Canal level.
Driving Unit.-To permit accurate determination of power
input to the pumps, an electric motor was used as a driving
unit. This was a delta connected, 50 horsepower, 2,300 volt,
10 pole, 720 rpm., 3-phase 60-cycle squirrel cage induction mo-
tor with stator windings paralleled to permit operation on 230
Electric power was supplied from local utility lines through
a watt-hour meter, an oil-immersed disconnect switch and a
manual auto-transformer type starter.
Mechanical power was transmitted from the horizontal mo-
tor shaft to the vertical pump shaft by five or six, as required,
C section V-belts operating through a quarter-turn.
The wide range in positions needed for changing pump speeds
required an easily adjusted base for the motor. Two 6 x 2-inch

Fig. 9.-Measuring weir in place in concrete discharge flume. View
from pumphouse.

-- .- S _

Tests of Low Head, High Volume Farm Pumps

by 6-foot rolled channels were anchored, web back up and ap-
proximately 5 feet apart, to the pumphouse floor across a com-
mon radial line from the pump under test. A 5/g-inch steel base
plate, 8 feet long by 4 feet wide, was laid across these chan-
nels. The motor was either bolted or clamped with heavy C-
clamps to this plate. Two more channel sections, across the
base plate and bolted to the channels on the floor, locked the
base plate in position after the proper motor placement had been
attained by jacking the base plate and motor.
Speed Variation and Measurement.-Pump speeds were varied
by changing V-belt sheaves to alter the ratio of the diameter
of the motor sheave to the pump sheave. The speeds employed
ranged from low speeds that did not produce appreciable dis-
charge at the pump's design lift, to high speeds that resulted
in a motor overload of about 50 percent during the test interval.
Pump shaft and motor shaft speeds were determined by use of
a direct reading, multi-range mechanical tachometer.
Lift Variation and Measurement.-The lift was varied by
inserting or removing stop logs beneath the measuring weir
in the sump outlet flume and by pumping down the supply
ditch. The variation in lift was limited by three factors. First,
the proximity of the pumphouse to crops growing in adjacent
fields limited the height at which water could be maintained in
the supply canal for low lifts. Second, the stage of the Hills-
boro Canal controlled the minimum discharge levels, since it
was desirable to have the weir crest high enough above canal
stage to provide free over-fall for discharge measurements.
High canal stages encountered on a few occasions at the end
of the testing period were cared for by using the weir in a
submerged position. Third, maximum lift was limited to the
head obtained by raising the discharge stage to the level of the
pumphouse floor and drawing down the intake canal to a level
so low that, as a result of reduced canal cross section and stor-
age, it was unable to supply sufficient water to maintain reason-
ably stable intake water elevations during a test run. This
condition occurred at a lift of about 9 feet.
The elevation of the intake water at the pump was measured
to the nearest 0.01-foot by level-rod readings from a bench
mark on the pumphouse floor. The elevation of the discharge
water, also the head on the weir, was determined by use of a
vernier hook gage mounted on the headwall of the concrete
flume far enough from the weir crest to insure accurate read-

Florida Agricultural Experiment Stations

ings of total head on the weir. Readings of this gauge were
made to the nearest 0.001-foot, and were converted to M.S.L.
elevations by adding them to the hook point elevation at 0.000
gauge setting. A stilling well around the gauge hook was used
when required.
Discharge Measurement.-Discharge from the sump was
measured by the use of a sharp-crested rectangular weir of
6-foot crest length with end contractions. It was built in accord
with plans and specifications developed for weirs constructed
and tested for both free and submerged flow in the Hydraulic
Laboratory at the Carnegie Institute of Technology. For free
fall, weir discharge was obtained from the relation
Q = 3.33 (L-0.2H)
where Q = discharge in cfs. (cubic feet per second)
L = weir crest length in feet
and H = total head on weir in feet
For the few cases of submerged flow, the weir discharge was
corrected for submergence by use of the relation

Q/Q = 1.00 -0.40
L 2 (10 10s)
where Q = corrected discharge in cfs.
Q = free discharge of the weir at the given
f headwater elevation
and S = submergence factor -
where a = area of segment of weir notch below the
level of the tailwater pool
h = tailwater depth above crest
A = area of segment of weir notch below the
level of the headwater pool
H = headwater depth above crest

The weir itself was built of 18-gauge sheet steel mounted on
timber backing for stability. The sump stabilized flow so that
turbulence and approach velocity at the weir were negligible.
Since the weir discharge did not include the water pumped
but lost by seepage through the boundaries of the stilling basin,
this loss was determined in the following manner: A cross-sec-
tional survey of the discharge sump was made and from this a
curve of increments of storage versus one-tenth of a foot in-

Mavis, F. T., Submerged Thin-Plate Weirs, Engineering News Record,
Vol. 143, No. 1, July 7, 1949.

Tests of Low Head, High Volume Farm Pumps

crements of stage was plotted. Before pump tests were begun
each day stop logs in a consistently maintained sequence, with
the weir on top, were inserted in the flume grooves. Pumps
were operated until the discharge sump reached the maximum
stage limited by the pump house floor elevation. This stage
was held by pumping for approximately two hours until the
muck embankment around the sump reached a high degree of
saturation. Pumping then was stopped and backflow through
the pump prevented.
When the water level fell below the weir crest, observations
were begun, utilizing the hook gauge and a stop watch, of the
time required for the water surface to drop one-tenth of a
foot from maximum level at regular intervals of about 7 inches
down to the Hillsboro Canal stage. A similar observation also
was performed at the end of each day's tests. By dividing
values from the storage versus stage curve by the time of fall
in seconds for 0.10-foot increments of stage, morning and eve-
ning curves were plotted as seepage loss in cfs. versus sump
The seepage for each test during that day was interpolated
between these two seepage curves as they were affected by
differential head and degree of saturation of the confining em-
bankment. The leakage, thus determined, was added to the flow
measured through the weir to obtain the total pump discharge.
Power Measurement.-A hook-on type wattmeter was used
to obtain the power input to the electric motor. The "two-
wattmeter" method of obtaining 3-phase power values was used
on all three possible 2-wire combinations of the power leads.
The three power readings were then averaged to give the power
input to the motor. These values were checked by observing
the rotation speed of the disk in the commercial watt-hour
meter in the power supply line and applying the meter's power
constant to the observations. Excellent correlation between
the power meter and the energy meter was obtained for all sets
of readings.
After the pump tests were completed the electric motor was
subjected to a dynamometer test over a load range of from no
load to 150 percent of full load, to determine the motor's load-
efficiency characteristics. Power input to the motor during this
test was measured in a similar manner with the same watt-
meter used during the pump tests which was monitored by two
other wattmeters taking duplicate readings. All power readings

Florida Agricultural Experiment Stations

checked within 1 percent. Torque was measured on a spring
scale accurate to 1 percent at full load.
Dynamometer and motor speeds were taken with the tachom-
eter used in the pump tests, which checked within 1/ percent
at full speed. Since the dynamometer was V-belt driven from
the motor, power transmission losses comparable to the belt
losses during the pump tests were obtained. A curve of shaft
power at the dynamometer versus electric power input to the
motor was plotted from these data and values of shaft horse-
power input to the pumps for efficiency computations were
taken directly from this curve.
Test Procedure.-Each pump was inspected and approved by
its manufacturer prior to the tests. When the maker desired,
it was removed from the pit, sent back to the factory and re-
built to new pump specifications before testing. Manufacturers
and their representatives were invited to be present during the
A standard procedure was followed for all tests. Morning
seepage measurements were first made as previously described.
The weir was then placed as low as possible without interfering
with aerated free fall from the crest. The weir crest elevation
was established from the bench mark on the culvert headwall
by use of an engineer's level, and this datum was transferred
to the hook gauge.
The pump being tested was started and allowed to run until
steady flow over the weir was established. A four-man record-
ing team took readings as simultaneously as possible of dis-
charge stage, intake stage, motor speed, pump shaft speed and
electric power input. These readings were repeated at 5-minute
intervals until four series of readings had been completed.
The pump was then shut off, the weir was removed, about
one foot was added to the height by inserting more stop logs
in a consistent sequence, and the weir was replaced. The new
elevation of the weir crest was then determined and the above
testing procedure repeated. Tests at increasing head were con-
tinued until the maximum lift was reached. The pump was then
shut off and evening seepage measurements were made. Sheaves
were changed to give the desired speed for the next day's tests
and the motor and pump were serviced. On the following day
another series of tests at a different speed was performed on the
pump. This procedure was followed until all tests were com-
pleted on a specific pump.

Tests of Low Head, High Volume Farm Pumps


The efficiency of a machine, usually expressed in percentage,
is defined as the ratio of useful work delivered by the machine
to the total work done on the machine and may be expressed
by the relation:
useful work output power
Efficiency -- or (1)
total work input power
When water is pumped from one level to a higher level, the
pump must perform work. The useful work done is equal to
the weight of the water pumped times the lift, or:
E = WH (2)
in which E = energy in foot-pounds
W = weight of water in pounds
and H = lift in feet
Work and kinetic energy are equivalent, and power is the rate
of doing work. Since one horsepower is equal to 550 foot-pounds
per second, the power required to lift the water may be com-
puted by the formula:
Q H (3)
p -
in which p = water horsepower, and
H is as above
and Q = weight rate of flow in pounds per second

Where the water is measured by volume instead of by weight,
the weight is replaced by the product of the volume Q in cfs. and
the weight of a cubic foot of water, which is 62.4 pounds, then
62.4 Q H Q H
S= or p (4)
550 8.81
In addition to the useful work done by the pump in lifting the
water, it also must do work and expend power in overcoming
the inertia of the water as it enters, flows through, and dis-
charges from the pump, as well as accounting for the drag
effect between interior surfaces of the pump and the moving
water. The power required to overcome these kinetic energy
losses is dependent upon the velocity of the flowing water, and
the relation can be expressed as
h = (K) (5)

Florida Agricultural Experiment Stations

in which h = velocity head losses in feet
V = velocity of the water in feet per second
g = acceleration of gravity = 32.2 feet per
second, per second
K = a coefficient depending upon design
There is also slippage of water through the impeller. At a
given lift, there is a minimum speed required to balance the
slippage and only above this speed will the pump produce an
effective flow. The amount of this loss depends upon the geo-
metrical design of the impeller and varies with the speed of the
pump and with the square root of the lift. In the Everglades
area where the lift is low, this is usually relatively small com-
pared to the velocity head losses.
The velocity head and slippage losses together make up the
hydraulic losses and the power required to overcome them is
equal to Q h'
p' (6)
in which p' = horsepower required to offset hydraulic
h' = sum of hydraulic head losses through
pump in feet
Q = volume of water in cfs.
In addition to hydraulic losses, there is some loss in friction
between the pump shaft and bearings, although this is minor
and should not amount to more than 5 percent of the shaft
horsepower in a well-constructed and properly lubricated pump.
From (4) useful work done by the pump is equal to
and from (6) power required due to hydraulic losses is equal to
Q h' output power
Then, from (1), efficiency equals and in-
8.81 input power
eluding mechanical friction losses the pump efficiency is then
determined by the relation
e =- (7)
QH Qh'
+ + (power lost by mechanical friction)
8.81 8.81
Since the denominator is the total power input at the pump
shaft, the pump efficiency in these studies was calculated from
the expression

Tests of Low Head, High Volume Farm Pumps

( QH 1 )
e' = -- X 100 (8)
8.81 P
in which e' = pump efficiency in percent, pool to pool
Q = pump discharge in cfs.
H = static lift in feet
P = shaft power input in horsepower
Energy to offset the hydraulic losses in pump operations due
to momentum changes in the water is consumed by turbulence
and eddying. Therefore, from a practical standpoint in pump
design, turbulence and eddying must be held to a minimum,
which means that flow must be kept streamlined and the vel-
ocity as low as practical to achieve high efficiencies. Keeping
the hydraulic losses down is especially important in the case
of low lift pumps, since there is a limiting pool-to-pool efficiency
as a result of velocity head losses.
To illustrate, assume a pump operates at a 4-foot lift with a
discharge velocity of 8 feet per second. The loss of head for a
pump discharging directly into a reservoir is very nearly equal to
V 8 X 8
or -.994, or approximately 1 foot. Neglecting
2g 32.2 X 2
mechanical friction and slippage losses, from formula (7) it
will be seen that the relation between efficiency, lift, and vel-

ocity head losses can be expressed by e cc1 Then,

disregarding entrance losses due to bends, skin friction, etc.,
the maximum efficiency possible would be or 80 per-
cent. If the lift were increased to 10 feet and other factors re-
mained the same, the maximum possible efficiency would then be
or 91 percent.
10 +1
To show the effect of increasing the speed of the water pumped,
assume that the same lifts and other factors are equal, but that
the discharge velocity is doubled to 16 feet per second; the head
16 X 16
loss would then be or 4 feet. The maximum possi-
ble efficiency for the 4-foot lift = 50 percent, and for
the 10-foot lift or 71 percent.
10 +4

Florida Agricultural Experiment Stations

The axial flow or propeller type pump which imparts low
velocity energy per unit of flow to the fluid being pumped is
theoretically most efficient at low lifts and rapidly loses effi-
ciency at high heads and ordinarily is used at heads not ex-
ceeding 25 feet.
The radial flow type pump which imparts moderate to high
velocity energy, depending upon the geometry of the impeller,
is theoretically most efficient at high heads and is employed at
heads of 30 feet up.
Mixed-flow design pumps, which combine axial and radial
flow to varying degrees so that continuous trend exists, may
be used efficiently in the transition range.

Data obtained from the test runs including speed, lift, dis-
charge, shaft power and plant and pump efficiencies for the
three pumps are presented in Tables 1, 2 and 3. From these
tables, performance graphs, Figures 10, 11 and 12, have been
plotted for each pump. Curves of test information appear in
solid lines. Curves for 25 rpm. increments of speed have been
interpolated and appear as broken lines.
Centrifugal Type Farm Pump.-The pool-to-pool efficiency of
this pump (Fig. 10) varies from 30 to 40 percent over a wide
range of operating conditions. The efficiency is only slightly
affected by changes in lift or speed over the central portion of
the graph. The power requirement is determined primarily
by the speed at which the pump is operated and varies only
slightly as the lift is changed. At constant speed discharge
decreases as the lift increases.
Helical Type Farm Pump.-The impossibility of varying the
static head of the helical pump (Fig. 11) limited the amount of
information available from the test data. The efficiency varied
from 28 to 37 percent. At a relatively fixed head of 5 feet,
discharge increases proportionately to speed, efficiency decreases
with increasing speed, and the power requirement increases
rapidly with increasing speed.
Propeller Type Farm Pump.-The pool-to-pool efficiency of
this pump (Fig. 12) varies from 45 to 60 percent over the typi-
cal operating range. Extreme efficiency values of 39 to 68
percent represent abnormal pumping conditions. The efficiency
responds to changes in speed and head. The power requirement

(20 in. diameter double suction intake; impeller of 8 in. depth, 20 in. inside, 26 in. outside diameter, volute casing.)
1 2 3 4 5 6* 7 8 9
Work Electric Pump 10 C
Test Pump Static Discharge Sump Pump Discharge Done Power Shaft Efficiencies
Date Speed Lift (Weir) Loss (Water Used Power Plant Pump
(Rpm) (Ft) (Cfs) (Cfs) (Cfs) (Gpm) Hp) (Hp) (Hp) % %

7-26-52 291 2.88 19.09 1.15 20.24 9,080 6.62 31.4 17.4 21.2 38.0
7-26-52 291 4.02 12.73 1.80 14.53 6,520 6.62 31.1 17.2 21.3 38.5
7-26-52 291 5.10 5.50 2.70 8.20 3,680 4.75 31.4 17.4 15.1 27.3
7-26-52 291 6.12 ... 5.20 5.20 2,330 3.61 31.1 17.2 11.6 21.0

7-24-52 325 3.21 23.00 1.80 24.80 11,130 9.03 37.4 23.6 24.1 38.3
7-24-52 325 4.17 16.50 2.30 18.80 8,440 8.87 37.4 23.6 23.7 37.6
7-24-52 325 5.45 10.36 3.20 13.56 6,090 8.38 37.3 23.5 22.5 35.6
7-24-52 325 6.62 6.30 3.40 9.70 4,350 7.28 37.4 23.6 19.5 30.8

7-25-52 352 3.26 27.20 0.90 28.10 12,610 10.40 45.7 32.3 22.8 32.2
7-25-52 352 4.55 21.45 1.50 22.95 10,300 11.86 45.6 32.2 26.0 36.8
7-25-52 352 5.80 15.66 2.50 18.16 8,150 11.95 45.4 32.0 26.3 37.3
7-25-52 352 7.43 8.21 3.90 12.11 5,440 10.21 45.6 32.2 22.4 31.7
7-22-52 390 3.50 30.50 1.50 32.00 14,360 12.74 57.0 43.8 22.4 29.1
7-22-52 390 4.73 24.61 2.70 27.31 12,260 14.66 55.6 42.0 26.4 34.9
7-22-52 390 6.17 18.75 4.00 22.75 10,210 15.92 55.6 42.0 28.6 37.9
7-22-52 390 8.26 10.83 5.20 16.03 7,190 15.02 55.5 41.9 27.1 35.8
7-29-52 418 3.58 32.78 1.10 33.88 15,210 13.75 71.8 55.1 19.1 25.0
7-29-52 418 4.78 28.65 2.50 31.15 13,980 16.91 70.6 54.2 23.9 31.2
7-29-52 418 6.26 23.70 4.80 28.50 12,790 20.23 70.8 54.4 28.6 37.2 c
7-29-52 418 8.88 13.46 6.00 19.46 8,730 19.62 69.3 53.3 28.3 36.8

Column 4 plus Column 5.

(20 in. diameter, 20 in. pitch and 8 ft. length auger.)
1 21 3 4 -5 6* j 7 8 9
I Work Electric Pump 10
Test Pump Static Discharge Sump Pump Discharge Done Power Shaft Efficiencies
Date Speed Lift (Weir) Loss (Water Used Power Plant Pump
(Rpm) (Ft) (Cfs) (Cfs) (Cfs) (Gpm) Hp) ) (H (Hp) % %
8-29-52 355 4.62 10.01 1.10 11.11 4,990 5.82 29.6 15.7 19.6 37.1
8-28-52 390 4.73 12.06 1.10 13.16 5,910 7.07 33.8 19.9 20.9 35.5
8-29-52 438 4.93 14.89 1.25 16.14 7,240 9.04 38.6 25.0 23.4 36.2
8-28-52 525 5.06 19.40 1.30 20.70 9,290 11.88 51.7 38.3 23.0 31.0
8-21-52 558 5.21 20.52 1.92 22.44 10,070 13.26 60.3 46.2 22.0 28.7
8-29-52 570 5.23 22.83 1.50 24.33 10,920 14.43 63.4 48.7 22.8 29.6
8-28-52 610 5.66 24.05 1.40 25.45 11,420 16.34 73.9 56.5 22.1 28.9

*Column 4 plus Column 5.

(24 in. diameter, 3 blade, impeller of varying pitch.)

1 2

Test Pump
Date Speed

10-3-52 400
10-3-52 400
10-3-52 400
10-3-52 400

10-2-52 468
10-2-52 468
10-2-52 465
10-2-52 465

10-1-52 510
10-1-52 510
10-1-52 510
10-1-52 505

10-6-52 600
10-6-52 595
10-6-52 590
10-6-52 590

10-7-52 683
10-7-52 675
10-7-52 675
10-7-52 655

Static Discharge
Lift (Weir)
(Ft) (Cfs)

3.73 15.67
4.78 10.66
5.74 7.94
7.02 3.27

3.94 20.17
4.48 17.04
6.97 8.68
7.79 8.05

3.93 23.13
4.60 20.58
5.79 16.19
7.33 12.88

3.96 28.60
4.90 28.40
6.08 23.39
8.20 19.66

4.16 34.23
5.66 29.16
6.75 25.03
8.89 21.62

5 -- 6*c

Sump Pump Discharge
(Cfs) (Cfs) (Gpm)

1.10 16.77 7,530
2.00 12.66 5,680
1.85 9.79 4,390
4.30 7.57 3,400

1.30 21.47 9,640
1.50 18.54 8,320
4.50** 13.28 5,960
3.50 11.55 5,180

0.80 23.93 10,740
1.40 21.98 9,860
2.70 18.88 8,470
3.50 16.38 7,350

2.40 31.00 13,910
2.20 30.60 13,730
3.00 26.39 11,840
4.80 24.46 10,980

0.90 35.13 15,770
2.10 31.26 14,030
3.20 28.23 12,670
4.30 25.92 11,630



















Plant Pump
% %

30.0 68.2
27.4 58.7
24.0 49.4
20.5 38.9

33.0 63.2
32.8 62.9
28.1 44.7
27.8 44.7

33.3 59.3
33.3 56.0
33.1 52.1
31.8 46.2

31.3 44.8
36.6 51.2
36.8 50.3
40.0 52.8

29.7 39.3
34.0 44.5
34.3 44.6
36.6 47.7

* Column 4 plus Column 5.
** Seepage includes change of storage factor due to falling stage.

3pa -- -*
?(I w ---

-n-- ---

^sn Hm11 ------------------ '

n ----- - ~ ~------------- _

20" Double Suction
Panel MoUnted, Centrifugal
Form Pump

Lift, Efficiency, & Power

2 4 6 8 10
Discharge, thousand g. p.m.

12 14

Fig. 10.-Graph of performance information and problem
modified centrifugal type pump.

solution for

Tests of Low Head, High Volume Farm Pumps

r A


20" Helical Type
Form Pump

Discharge, Efficiency, 8 Power
at fixed lift

300 400 500 600 700
Speed, R.P.M.
Fig. 11.-Graph of performance information for helical type pump.


70 0i

70~~~ 0-------------------_-- -- -----..----,-

24Prpo Te \ \


20 944

\ \ \\\

Form Pump \ ,,\ \P mp

Not@; 0 Free \ ow weir
. .a S\bereged weir

28 2 4 6 8 to 12 14 16

propeller type pump.
propeller type pump.

Tests of Low Head, High Volume Farm Pumps

is fixed mainly by pump speed but is responsive to changes in
lift. At constant speed, discharge decreases as the lift increases.

Identical Pumps.-The graphs of the pump test data with
interpolated speed curves in Figures 10, 11 and 12 can be used
in solving problems relating to speed, head, discharge, power
and efficiency for the centrifugal and propeller pumps tested.
For example, in Figures 10 and 12, solutions of a common
type of problem are shown for each pump. The problem, the
speed and power necessary to pump 9,000 gpm. against a 5.5-foot
static head and the efficiency under those conditions is solved
graphically, as shown by the short-dashed lines. The results
are: Centrifugal Propeller
Speed 361 rpm. 514 rpm.
Power Input 34 hp. 24 hp.
Efficiency 37.8% 53.2%
Similarly, to pump 14,000 gpm. against a 5-foot lift, the
results are:
Centrifugal Propeller
Speed 428 rpm. 645 rpm.
Power Input 59 hp. 40 hp.
Efficiency 34.5% 45.7%
An example of a typical irrigation problem is at what speed
maximum pump efficiency would be obtained when supplying
1/4-inch of water per day to 1,280 acres if the static lift were
4.75 feet.
Since 18.9 gpm. is equal to 1 inch on one acre in 24 hours, then
6,048 gpm. would be equivalent to the above requirement. The
graphical solution to this problem, for each pump, appears in
Figures 13 and 14. It is apparent that the pump must be op-
erated on the heavy, dashed line at 4.75 feet lift and above
6,048 gpm. discharge. Efficiencies for different speeds at 4.75
feet static lift appear as triangles in the efficiency versus dis-
charge graph. There is a wide range of speeds and of discharge
values that could be used to solve this specific problem without
causing a serious drop in pump efficiency.
Conditions for peak efficiency for each pump are:

Centrifugal Propeller
Speed 340 rpm. 450 rpm.
Efficiency 38.5% 60.3%
Discharge 8.800 gpm. 7,500 gpm.
Operating Time 16.5 hrs./day 18.5 hrs./day

31 400
604-- -- -- --

A 375


v ardged for
300 to425 RPM


20" Double Suction
Panel Mounted, Centrifugal
Form Pump

Lift, Efficiency, 8 Power

2; 2 4 6 8 10 12 14
Discharge, thousand gp m.

Fig. 13.-Problem solution for modified centrifugal type pump.

UU n m

Tests of Low Head, High Volume Farm Pumps

Fig. 14.-Problem solution for propeller type pump.

Florida Agricultural Experiment Stations

To obtain peak plant efficiencies, the drive from the power
unit must be set so that the power unit itself is running at
peak efficiency. The proper speed for internal combustion en-
gines may be obtained from the performance curve supplied by
the manufacturer, which shows the relation between engine
speed, horsepower and fuel consumption. The sheave or gear
ratios should be selected to match the pump speed which gives
maximum efficiency at the desired discharge for the average lift
with the engine speed which gives the required power with
minimum fuel consumption.
Where the average and maximum discharge and lift require-
ments are known or can be estimated, the most efficient pump
speeds can be obtained from the performance graph and the
most economical power unit can then be selected, using the
performance curves for the engines.
For most industrial engines the specific fuel consumption,
which is the unit of fuel used per brake-horsepower-hour, does
not change significantly over a fairly wide range of engine
speeds. This allows considerable tolerance in adjusting engine
speeds to suit pumping demands before the plant efficiency
is appreciably affected by engine characteristics. However,
when pumpage requirements vary a great deal in lift and dis-
charge over fairly long periods, it may prove advantageous to
have several sheaves or gears available so that the ratio may
be changed to give the most efficient pump speed. This is es-
pecially true for the propeller type pump where the efficiency
responds to changes in speed and lift.
Electric motors are coming into use in the Everglades area
to supply power for irrigation pumps. The three-phase, squir-
rel-cage induction motor is used primarily because simplicity
of construction makes initial and maintenance costs low. Since
these are inherently constant speed machines, the horsepower
requirements should be known in advance and not vary too
widely, since the power factor and efficiency of the motor drops
off rapidly at low loads. Changing sheaves on the shafts is
the most economical method of changing speeds theoretically,
even though belt drives reduce the efficiency of the unit by 3
to 5 percent as compared with direct connection. Practically,
however, operators are reluctant to change sheaves and by
running the pumps at high speeds at low or medium lifts waste
10 to 25 percent of the electric power.

Tests of Low Head, High Volume Farm Pumps

Similar Pumps of Different Size.-The performance of other
sizes of pumps of the same series as those tested can be pre-
dicted, from the results, by an approximate method of analy-
sis. This technique is strictly correct only for the comparison
of pumps of exact geometrical similarity, but minor shape
variations are permissible if less precise results will be ac-
Two pumps of different size but of the same series will op-
erate at the same efficiency when pumping under similar con-
ditions 3 if
n 1 D n2 D,2
Q1 Q.
nl D1i n2 D,.
P1 P.
and (10)
n13 D1 n?3 D.5
where H = pool-to-pool head
n = speed
D = impeller diameter
P = shaft power input
subscript refers to one pump
subscript. refers to the other pump
Theoretically, the head, H, should be the change in total head,
but the pool-to-pool head, or lift, can be used for a slightly less
precise, but practical solution.
One common type of problem, the speed, discharge and power
of a 30-inch centrifugal pump operated at an 8-foot lift, is
solved by the use of relations (8), (9) and (10). The test results
(Fig. 10) for the 20-inch centrifugal pump show that, when
n =- 390 rpm. and H1 = 6.8 feet, then Q = 9,000 gpm. and P1 = 42 hp.

D H 20 8
From (8), n n = n - = 390 X- = 282 rpm.
D2 HJ 30 6.8 1
3 3
n2 D2 282 30
From (9), Q2 = 9,000 X 22,000 gpm.
ni D1 390 20
3 5 3 6
( n\ D, I 282 \) 30
From (10),. P = 42 X 120.5 hp.
nI D I 390 2\ 20 I
3 Rouse, Hunter. Elementary Mechanics of Fluids. Wiley, 1946, p. 298.
Stepanoff, A. J. Centrifugal and Axial Flow Pumps. Wiley, 1948, pp.

Florida Agricultural Experiment Stations

Therefore, the 30-inch pump would deliver 22,000 gpm. at 282
rpm. at an 8-foot lift and would require 120.5 horsepower input.
The efficiency for either the 20-inch or the 30-inch pump is 35
percent for the stated operation conditions.
Another common type of problem, to determine the required
size, speed and power of a propeller pump to deliver 6,000 gpm.
at a 6-foot lift, is solved through the use of the same relations.
The test results (Fig. 12) for the 24-inch propeller pump show
that at

n, = 550 rpm. and H1 = 4.2 feet, Q, = 12,000 gpm. and P = 25 hp. From (9),
3 2 %
Q / D, \ (H Qs
n = n, -- -- substituted in (8) gives D. = D1 --
Q D, H, Q,

4 4.2 / 6,000\ 1
24- I-
L 6 12,000/
=15.7 inches, say 16 inches
6,000 24 1
then n = 550 X ---- 928rpm.
12,000 16
3 5 3 5
( n D 928 \ 16
From (10) P2 = P1 -- = 25 = 15.8 hp.
SnX D, 550 2\ 24

Therefore, a pump to deliver 6,000 gpm. at a 6-foot head should
be of 16-inch diameter, should be operated at 928 rpm., and
would require 16 horsepower input. The efficiency of either
the 24-inch or the 16-inch pump is 54 percent for the stated
operation conditions.
Speed, head, discharge and power data used in solving the
above problems were extracted from portions of the graphs
showing high efficiencies at or near operating conditions for the