Document text
US Army Corps
of Engineers
Waterways Experiment
Station
Technical Report EL-94-15
October 1 994
Reliability, Availability, and Maintainability
Testing of High Pressure Pumping Units
for Mobile Army 600 Reverse Osmosis
Water Purification Unit
by Wayne W. Sharp, Howard K. Bell Consulting Engineers, Inc.
Charles R. O’Quinn, Michael G. Channel I, WES
Approved For Public Release; Distribution Is Unlimited
1 9941 1 28 1 07
DTIC QX
IE
afT0‘
5
Prepared for Belvoir Research, Development, and Engineering Center
The contents of this report are not to be used for advertising,
publication, or promotional purposes. Citation of tradenames
does not constitute an official endorsement or approval of the use
of such commercial products.
PRINTED ON RECYCLED PAPER
Technical Report EL-94-15
October 1 994
Reliability, Availability, and Maintainability
Testing of High Pressure Pumping Units
for Mobile Army 600 Reverse Osmosis
Water Purification Unit
by Wayne W. Sharp
Howard K. Bell Consulting Engineers, Inc.
P.O. Box 546
Lexington, KY 40585
Charles R. O’Quinn, Michael G. Channell
U.S. Army Corps of Engineers
Waterways Experiment Station
3909 Halls Ferry Road
Vicksburg, MS 39180-6199
Final report
Approved for public release; distribution is unlimited
Accesion For
NTIS CRA&I
OTIC TAB
Unannounced □
Justification
By .
Distribution /
Availability Codes
Avail and/or
Dist | Special
I M
Prepared for Belvoir Research, Development, and Engineering Center
Fort Belvoir, VA 22060-5606
US Army Corps
of Engineers
Waterways Experiment
Station
HEADQUARTERS
BUILDMG
FOR N FORMATION CONTACT
PUBLIC AFFAIRS OFFICE
U. S. ARMY ENGINEER
WATERWAYS EXPERIMENT STATION
3909 HALLS FERRY ROAD
VICKSBURG. MISSISSIPPI 39180-6199
PHONE: (601)634-2502
AREA OF RESERVATION - 2 7 sq km
Waterways Experiment Station Cataloging-in-Publication Data
Reliability, availability, and maintainability testing of high pressure pump¬
ing units for Mobile Army 600 Reverse Osmosis Water Purification Unit /
by Wayne W. Sharp ... [et al.] ; prepared for Belvoir Research, Develop¬
ment, and Engineering Center.
157 p. : ill. ; 28 cm. — (Technical report ; EL-94-15)
1 . Military hydrology — Equipment and supplies. 2. Water — Purifica¬
tion — Reverse osmosis process. 3. Pumping machinery — Performance.
4. Saline water conversion — Reverse osmosis process. I. Sharp,
Wayne W. II. United States. Army. Corps of Engineers. III. U.S. Army
Engineer Waterways Experiment Station. IV. Environmental Laboratory
(U.S. Army Engineer Waterways Experiment Station) V. Belvoir Research,
Development, and Engineering Center (U.S.) VI. Series: Technical re¬
port (U.S. Army Engineer Waterways Experiment Station) ; EL-94-15.
TA7W34 no. EL-94- 15
Contents
Preface . .
Conversion Factors, Non-SI to SI Units of Measurement . vii
1— Introduction . 1
Background . 1
Pressure . 1
2 — Pumping Unit Characteristics . 3
3 — Facility Design . 5
4 — Testing Procedures . 8
Initial Inspection . 8
Start-up Testing . 8
Endurance Testing . 9
5 — Test Results . 12
Initial Inspection . 12
Start-up Testing . 17
Endurance Testing . 22
Reliability . 26
Maintainability . 26
Availability . 28
RAM Results . 28
6 — Summary and Conclusions . 34
Appendix A: Wheatley Pump . A1
Appendix B: Union Pump . B1
Appendix C: Sundstrand Pump . Cl
Appendix D: FMC Pump . D1
Appendix E: Copper-Nickel Test Results . El
SF 298
List of Figures
Figure 1. Mobile Army 600 ROWPU . 2
Figure 2. Test facility plan view . 5
Figure 3. Wheatley quintuplex pump . 12
Figure 4. Union pump fluid end . 13
Figure 5. Sundstrand pump . 14
Figure 6. Ingersoll Rand pump . 15
Figure 7. White Rock damper . 16
Figure 8. Young Engineering damper . 16
Figure 9. Accelerometer locations . 19
Figure 10. Power consumption during endurance testing . 24
Figure 11. Endurance test results and RAM characteristics . 25
List of Tables
Table 1. Pump and Motor Description . 3
Table 2. Pulsation Damper Description . 4
Table 3. Average Start-up and Operating Amperage . 17
Table 4. Suction and Discharge Pressures . 18
Table 5. Acceleration/Vibration Test . 20
Table 6. Noise Level . 21
Table 7. Temperature and Salinity at Beginning of Test . 21
Table 8. RAM Results . 33
Preface
The report herein describes testing procedures and results applicable to
reverse osmosis water purification high pressure pumping units. This work
was sponsored by the Department of Army, Belvoir Research, Development,
and Engineering Center, Fort Belvoir, VA. The technical monitor for this
work was Mr. Bob Shalowitz.
The work was conducted by the U.S. Army Engineer Waterways Experi¬
ment Station (WES), Vicksburg, MS. The test facility for reliability, availabil¬
ity, and maintainability testing was located at the U.S. Army Engineer District,
Vicksburg, Harbor Project. This report was written by Mr. Wayne W. Sharp,
Howard K. Bell Consulting Engineers, Inc., and Mr. Charles R. O’Quinn,
Engineering and Construction Services Division, WES, and Mr. Michael G.
Channell, Environmental Restoration Branch (ERB), Environmental Engineer¬
ing Division (EED), Environmental Laboratory (EL), WES.
Mr. O’ Quinn was responsible for the physical operations and construction
of the test facility and daily data collection during all test phases at the Harbor
Project. At the time of testing, Mr. Douglas Lee was plant supervisor at the
Harbor Project.
Electronic data collection during start-up testing was provided by
Messrs. Joe Savage, Leo Koestler III, and Richard Floyd, Instrumentation Ser¬
vices Division, WES.
Scanning electron microscope testing and results were reported by
Messrs. Jerry P. Burkes and Sam Wong, Structures Laboratory, WES.
The study was conducted under the supervision of Mr. Norman R.
Francingues, Chief, ERB; Dr. Raymond L. Montgomery, Chief, EED; and
Dr. John W. Keeley, Director, EL.
At the time of publication of this report, Director of WES was
Dr. Robert W. Whalin. Commander was COL Bruce K. Howard, EN.
This report should be cited as follows:
Sharp, W. W„ O’Quinn, C. R„ and Channell, M. G. (1994).
“Reliability, availability, and maintainability testing of high
pressure pumping units for the Mobile Army 600 Reverse
Osmosis Water Purification Unit,” Technical Report EL-94-15,
U.S. Army Engineer Waterways Experiment Station, Vicksburg,
MS.
The contents of this report are not to be used for advertising, publication,
or promotional purposes. Citation of trade names does not constitute an
official endorsement or approval of the use of such commercial products
Conversion Factors, Non-SI to
SI Units of Measurement
Non-SI units of measurement used in this report can be converted to SI units
as follows:
Multiply
By
To Obtain
cubic feet
0.02831685
cubic meters
Fahrenheit degrees
5/9
Celsius degrees or kelvins'
feet
0.3048
meters
gallons (U.S. liquid)
3.785412
liters
horsepower (550 foot-pounds
(force) per second)
745.6999
watts
inches
2.54
centimeters
miles (U.S. statute)
1.609347
kilometers
pounds (mass)
0.4535924
kilograms
pounds (force) per square inch
6.894757
kilopascals
quarts (U.S. liquid)
0.9463529
liters
' To obtain Celsius (C) temperature readings from Fahrenheit (F) readings, use the following
formula: C = (5/9) (F - 32). To obtain kelvin (K) readings, use the following: K = (5/9) (F - 32)
+ 273.15.
1 Introduction
Background
Mobile reverse osmosis water purification units (ROWPUs) are used by the
Army for production of potable water, particularly from saline, brackish, and
fresh sources. Figure 1 shows the Mobile Army 600 gal/hr1 ROWPU. The
reverse osmosis is a membrane process requiring a high pressure (up to
1,000 psi) feed stream. Therefore, a high pressure pump is a key ROWPU
component. The work described in this report involves testing high pressure
pumps from various vendors to evaluate their use as ROWPU components.
Pressure
This report provides data obtained by a unified testing program applicable
to all high pressure pumping units that are candidates for use as ROWPU
components. The testing program consists of three test phases (initial inspec¬
tion, start-up testing, and endurance testing) to assess the following pumping
unit characteristics (“pumping unit” refers to a pump-motor-pulsation damper
combination):
a. Physical dimensions and weight.
b. Noise generation data.
c. Acceleration (vibration) data.
d. Operational characteristics (flow rate and pressure delivered).
e. Reliability.
/. Availability.
1 A table of factors for converting Non-SI units of measurements to SI units is presented on
page vii.
Chapter 1 Introduction
Figure 1 . Mobile Army 600 ROWPU
g. Maintainability.
h. Compatibility with ROWPU.
Test procedures are described in Chapter 4. Results from all test phases are
presented in Chapter 5.
Chapter 1 Introduction
2 Pumping Unit
Characteristics
A general description of each pump and motor evaluated is presented in
Table 1. All pumps were selected to deliver 60 gal/min at 1,000 psig. All
motors were supplied three-phase, 60-Hz, 460-V electrical service. A motor
control center common to all pumping units was utilized for electrical control.
The manufacturer’s off-the-shelf manuals (not included in this report) were
utilized for installation, operation, and repairs.
Table 1
Pump and Motor Description
Pump (Model)
Motor
HP
Drive
Inlet
NPT
Outlet
NPT
Length
in.
Width
in.
Height
in.
Weight
lb
(Pump and Motor)
Wheatley
Quintuplex
General
Electric
40
Belt
3 in.
1.5 in.
40.0
32.25
46.0
1,663
Union TD 30
Triplex
Duty
Master
40
Belt
2 in.
1.5 in
59.75
35.75
35.50
1,572
Sundstrand
Sunflo P-2000
Marathon
75
Shaft
2 in.
1.5 in.
48.0
18.0
19.5
640
FMC L1 1
U.S.
Electric
40
Belt
2.5 in.
1.25 in.
32.31
20.25
14.31
855
Ingersoll Rand
HDP-31
(Hammellman)
Lincoln
40
Belt
2 in.
1.5 in.
59.58
30.63
41.50
900
Note: NPT = National Pipe Threads.
A pulsation damper located immediately downstream of the pump absorbs
pressure surges and protects reverse osmosis (RO) elements. Dampers used
during testing are identified in Table 2. Pumping unit selection was made by
the U.S. Army Belvoir Research, Development, and Engineering Center
(BRDEC).
Chapter 2 Pumping Unit Characteristics
3
Table 2
Pulsation Damper Description
Model
Overall
Length
in.
Outside
Diameter
In.
Mechanism
Young Engineering
(Zemark)
36
12
Reservoired stainless steel
acoustical
White Rock
24
12
Reservoired stainless steel
acoustical
Greer
16
10
1-qt Bladder - nitrogen charged
Greer
10
7
4-qt Bladder - nitrogen charged
4
Chapter 2 Pumping Unit Characteristics
3 Facility Design
The ROWPU test facility was constructed in an outdoor shelter (roof and
back wall only) located at the U.S. Army Engineer District, Vicksburg, Harbor
Project, Vicksburg, MS. Figure 2 shows the test facility design. The facility
was constructed on a poured concrete pad and fenced on all open sides with
access gates.
Figure 2. Test facility plan view
Chapter 3 Facility Design
The test fluid was a saltwater solution composed of bulk salt dissolved in
potable water to create a saline solution between 1.5 and 3.0 parts per thou¬
sand. Five 250-gal, coated steel tanks served as the test fluid reservoir. One
tank was converted into a fan-driven cooling tower to help dissipate the heat
transferred into the test fluid from the high pressure pumping loops. A
freshwater intake was also designed to replace lost fluids because of system
leaks and evaporation. A liquid level switch was installed in the reservoir to
completely shut off power in the event of fluid loss to protect the pumping
units.
A booster pump designed to deliver 60 gpm at 42 psig and 180 gpm at
32 psig was installed downstream of the reservoir. This booster pump sup¬
plied the tested pumping units with the required net positive suction head. An
in-line strainer downstream of the booster pump provided protection to the
pumping units and associated appurtenances.
Four test loops were constructed to accommodate start-up and endurance
testing requirements (Chapter 4). Individual test loops include a strainer for
debris removal, high and low pressure switches to cut power to individual
motors for personal and mechanical safety, pressure gauges to manually moni¬
tor suction and discharge conditions, a pressure-regulating valve to control dis¬
charge pressure from the pumping units, ball valves, unions, and necessary
fittings to meet data acquisition requirements. Discharge piping was
configured to approximate the shape and length of the Mobile Army 600
ROWPU. All discharge piping was constructed of 1.5-in. (inside diameter)
NPT, 316 stainless steel rated at 1,200 psig with one exception. Copper-nickel
piping (90-10) as constructed for the Mobile Army 600 ROWPU was installed
in one endurance test loop for materials-testing purposes.
The start-up test loop was constructed to allow the potential for 25 pumping
unit combinations to be tested. Flexible, high pressure hoses were utilized to
make the necessary attachments between suction and discharge piping,
pulsation damper, and pump for each pumping unit combination. A data
acquisition structure with heating and air conditioning was installed to protect
and house the electronic equipment (computers, oscilloscopes, etc.) necessary
to acquire start-up test data. Pressure transducers and accelerometers were
installed to measure pressure differentials and mechanical vibrations,
respectively.
The endurance test loops (three total) were constructed for the individual
pumping unit to be tested and are less data intensive. Design priorities were
established for access ease and maintenance considerations. Data acquisition
for endurance testing was manual (not electronic). Chapter 4 details data
requirements for both start-up and endurance testing.
The test facility was powered by Mississippi Power and Light Company.
Initiating from a 13,800-V service, two transformers stepped down the supply
voltage to the 230/460-V service required for testing purposes. All motors
Chapter 3 Facility Design
were supplied a 460-V, three-phase, 60-Hz service. All lighting, safety
switches, data acquisition equipment, etc., utilized an existing 115-V service.
A motor control center (MCC) was installed to electrically control the test
facility. All pumping units and related controls (pressure switches and liquid
level switch) were governed through the MCC. All electrical data (amperage
and kilowatt hour) were obtained through equipment installed in the MCC.
The timer installed for endurance testing was housed in the MCC where the
individual starters to each pumping unit could be controlled.
Chapter 3 Facility Design
7
4 Testing Procedures
Initial Inspection
Initial inspection involved disassembling and visually inspecting pumping
unit components. Components that were inspected included valves, packings,
bearings, crankshaft, belts, driver, exterior of unit, pulsation damper, and skid.
All information presented in Tables 1 and 2 was verified or determined during
initial inspection. Initial inspection also allowed testing personnel to familiar¬
ize themselves with the pumping units. Disassembly of the pumps and
pulsation dampers was done in strict accordance with the manufacturer’s off-
the-shelf manuals.
Start-up Testing
Start-up testing included an extensive collection of electrical, fluid, vibra¬
tion, and noise data for each possible combination of five pumps and four
dampers as well as no damper (25 possible combinations).
Each pumping unit combination was installed in the start-up test loop
(Figure 2) and operated under similar conditions for a 6-hr period. The fol¬
lowing data were obtained for each pumping unit tested.
a. Temperature and salinity of test fluid.
b. Start-up amperage (instantaneous peak).
c. Operating amperage at 1,000 psi, 60 gpm.
d. Suction pressure (Appendixes A-D).
e. Discharge pressure (Appendixes A-D).
/. Vertical acceleration on pump frame and floor.
g. Horizontal acceleration on pump frame and floor.
Chapter 4 Testing Procedures
h. Noise levels (db).
i. Flow rate.
Temperature and salinity of the test fluid were recorded before initiating
each test. Start-up and operating amperage readings were obtained through the
MCC. Each pumping unit was started at maximum test pressure (1,000 psig) a
minimum of four times to obtain the average start-up (instantaneous peak)
amperage required. Operating amperage was obtained multiple times through¬
out each 6-hr test.
Suction and discharge pressures were obtained manually and electronically.
Manual readings from 4-in. dial, fluid-dampened pressure gauges mounted on
both the suction and discharge side of the pump gave visual indication to the
tester and allowed proper pressure regulating valve (PRV) adjustment to ensure
proper operation of the pumping unit. Transducers, installed flush-mounted
with inside walls of suction and discharge piping, electronically recorded
(820 readings per second) suction and discharge pressures (after the pulsation
damper if applicable). These readings were recorded for a 15-min duration of
each pumping unit test. These pressure readings help quantify the effects of
the pulsation damper.
Vibration data were obtained by the use of accelerometers mounted on the
pump frame and on the concrete pad to which the pumping unit was anchored.
Four 7/8-in. by 3-in. concrete anchors bolted through the mounting brackets on
the skid were used. A polypropylene pad 3/8-in. thick separated the pumping
unit from the concrete pad. These accelerometer readings quantify the
mechanical vibrations caused by the pumping unit. This data were recorded
for a 15-min duration of each pumping unit test.
Noise levels for each pumping unit were obtained by a hand-held decibel
meter. Three locations, each 2.0 ft (horizontally) from the pump frame and
3.0 ft (vertically) from the pad, were used to obtain an average noise level (db)
for each pumping unit.
Flow rate produced by each pumping unit was verified by an in-line disk
flowmeter. All pumping units tested delivered 60 gpm at 1,000 psig.
Data obtained from start-up testing are presented in Chapter 5.
Endurance Testing
At the conclusion of the start-up testing, three pumping units were selected
for endurance testing by U.S. Army Engineer Waterways Experiment Station
(WES) and BRDEC personnel, based on start-up test results. The selected
pumping units were installed in parallel in the test loop and operated by a
mechanical timer, to provide approximately 60 gpm at 1,000 psi for 20 hr per
day, 7 days a week, until a total of 2,000 hr of operation were reached, or a
Chapter 4 Testing Procedures
failure event necessitated cessation (Figure 2). Pumping units were operated
and maintained in strict accordance with manufacturer’s off-the-shelf manuals.
Efforts were made to confine routine and preventive maintenance to the 4 hr of
daily scheduled downtime. A daily log was kept for each pump unit being
tested, and the following parameters were recorded daily:
a. Suction pressure.
b. Discharge pressure.
c. Flow rate.
d. Fluid temperature and salinity.
Suction and discharge pressures were obtained manually from 4-in. dial,
fluid-dampened pressure gauges. Chart recorders were utilized to obtain a con¬
tinuous record of discharge pressure for each pumping unit.
These records helped to indicate failure scenarios during periods when no
testing personnel were on site.
All pumping units were operated at 1,000 psig (±50 psig) discharge pres¬
sure controlled by the downstream pressure regulating valve. The suction
pressure supplied to each pumping unit was 32 psig (±2.5 psig) from the
booster pump.
The flow rate produced by each pumping unit was measured by a dedicated
disk flowmeter (accumulator). All pumping units delivered 60 gpm at
1,000 psig. Fluid temperature and salinity were recorded daily.
Power consumption for each pumping unit was recorded daily from a dedi¬
cated kilowatt hour meter located at the motor control center. Increasing
power consumption during testing could indicate decreasing efficiencies.
Failures during endurance testing were defined as any malfunction that
caused or may cause inability to commence operation, cessation of operation,
degradation of performance below designated levels, or serious personnel
safety hazards.
Any malfunction that the operator could remedy was not considered a fail¬
ure provided that the repair was authorized or prescribed as an operator func¬
tion and could be accomplished in 30 min or less using only controls and
small hand tools. Whenever a pumping unit failure occurred, the time, nature,
and cause of the failure was documented. Pumping unit reliability, avail¬
ability, and maintainability (RAM) were examined and quantified.
In conjunction with the RAM endurance tests, copper-nickel (Cu-Ni) piping
was utilized on the discharge side of the Wheatley pump test loop. This pip¬
ing was procured from MECO, Inc., manufacturer of the Mobile Army
10
Chapter 4 Testing Procedures
600 ROWPU, and is nearly identical to the discharge piping used in the
600 ROWPU. Additional 1/4-in. NPT parts were manufactured into the pipe
assembly to meet endurance test equipment requirements (pressure gauges,
etc.). Two complete assemblies were procured, such that used and unused pipe
could be evaluated for ROWPU compatibility upon completion of RAM endur¬
ance testing. A scanning electron microscope was used to quantify corrosion
and corrosion products in the Cu-Ni pipe. Appendix E documents the findings
from this investigation.
Chapter 4 Testing Procedures
5 Test Results
Initial Inspection
The following observations were recorded during the initial inspection for
each pump and pulsation damper. Tables 1 and 2 list overall characteristics
verified by initial inspection. All units were disassembled and inspected
within 90 days of arrival.
a. Wheatley Pump (Figure 3).
(1) Pump was disassembled easily in accordance with the manufac¬
turer’s off-the-shelf manual.
(2) Fluid end was free of standing water.
Figure 3. Wheatley quintuplex pump
Chapter 5 Test Results
(3) Internal materials appeared in good condition, and all parts
including the crankshaft were easily accessible.
(4) Valves, packings, crankshaft, belt, coupling, driver, and pump
exterior were in good shape.
(5) Bearings were burred and scratched. Metal shavings were
observed in the oil reservoir and removed with a magnet.
b. Union Pump (Figure 4).
(1) Pump was disassembled easily in accordance with the manufac¬
turer’s off-the-shelf manual.
Figure 4. Union pump fluid end
(2) Rust was forming on cast iron surfaces around plunger on the
liquid end of the pump.
(3) Water was retained in the liquid end of the pump even with plug
extracted.
(4) Rust was forming on valves, couplings, driver, and on the exterior
of the pump.
(5) Packings, crankshaft, and belts appeared to be in good condition.
Chapter 5 Test Results
13
c. Sundstrand Pump (Figure 5).
(1) Pump was disassembled very easily in accordance with the manu¬
facturer’s off-the-shelf manual. Compared with the positive
displacement pumps, there are less moving parts, and disassembly/
reassembly was easier.
Figure 5. Sundstrand pump (Greer 4-qt damper)
(2) No rust was apparent internally or externally.
(3) Entire pump unit (as a whole) was easily removed.
(4) Impellers, gears, and exterior of the pump were in excellent shape.
d. Ingersoll-Rand (Hammellman) Pump (Figure 6).
(1) Pump was easily disassembled in accordance with the manufac¬
turer’s off-the-shelf manual.
(2) No rust was apparent on internal or external surfaces.
(3) A slight tear was apparent in the center cylinder rubber packing.
(4) Crankshaft was difficult to access — motor was removed to access.
(5) Packings, valves, pistons, belts, coupling, driver, and exterior of
the pump were in good condition.
Chapter 5 Test Results
Figure 6. Ingersoll Rand pump (Greer 1-qt damper)
e. FMC Pump (photo not available). The FMC pump was not initially
inspected because of its late arrival.
/. Pulsation Dampers.
(1) The White Rock damper (Figure 7) is a cylindrical, stainless steel,
acoustical damper. It is the heaviest damper, but not the largest.
Infeasible to disassemble.
(2) The Young Engineering damper (Figure 8) is a cylindrical, stain¬
less steel, acoustical damper. It is the largest damper. Infeasible
to disassemble.
(3) The Greer 1-qt and Greer 4-qt bladder dampers were disassembled
according to the manufacturer’s off-the-shelf manual. The 4-qt
(1-gal) model was found to be in good condition internally and
externally. Both models require a compressed gas (nitrogen)
precharge before use.
(4) The Greer 1-qt bladder was damaged upon inspection. The manu¬
facturer could not supply additional bladder in the necessary time
frame (2 weeks). Thus, the Greer 1-qt pulsation damper was elim¬
inated from further testing.
Chapter 5 Test Results
Figure 7. White Rock damper (Union pump)
Figure 8. Young Engineering damper (Wheatley pump)
16
Chapter 5 Test Results
Start-up Testing
Start-up tests were conducted as described in Chapter 4. The Ingersoll-
Rand pump experienced mechanical problems with valves sticking in the
pump, rendering the pump inoperable. The pump was disassembled, the intake
valves were shaved according to the manufacturer’s off-the-shelf manual and
Ingersoll-Rand personnel, and the pump remained inoperable.
Start-up testing proceeded with the other pumping units. A total of
16 combinations remained to be tested (four pumps, four pulsation dampers).
The following data summarize start-up test results from these 16 pumping unit
combinations. All pumping units tested delivered 60 gpm at 1 ,000 psig
successfully.
Start-up and operating amperage
Start-up and operating amperage readings were obtained from the MCC
(460- V) for each pumping unit combination starting and operating at
1,000 psig and 60 gpm. A minimum of four recordings were obtained for
each combination, and the average of these readings is shown in Table 3.
Table 3
Average Start-up and Operating Amperage
Amperage (amps)
Pump Model
Damper
Start-up
Average Run
Wheatley
White Rock
190
38
Young
200
33
Greer
210
37
None
190
38
Union
White Rock
190
40
Young
200
41
Greer
205
40
None
190
39
Sundstrand
White Rock
416
80
Young
416
77
Greer
416
80
None
416
80
FMC
White Rock
260
39
Young
260
40
Greer
250
40
None
260
40
Starting the pumping units at 1,000 psig created a current (amperage)
demand 500 to 625 percent greater than normal operating amperage. Given all
pulsation damper combinations, the Union pump averaged approximately
Chapter 5 Test Results
500-percent increase. The FMC pump was highest, averaging an approximate
625-percent increase.
The Sundstrand pump requires twice as much amperage to operate because
of its 75-hp motor, which is nearly twice as large as the other pumps and half
as efficient.
Suction and discharge pressure
Suction and discharge pressures were recorded 820 times per second with
pressure transducers for each pumping unit combination. These results help
quantify the effects of the different pulsation dampers with each pump.
Appendixes A-D, respectively, show a 0.5-sec recording of suction and
discharge pressures for each pumping unit combination (Appendix A is the
Wheatley pump combinations; Appendix B is the Union pump combinations;
Appendix C is the Sundstrand pump combinations; Appendix D is the FMC
pump combinations). Table 4 summarizes information from Appendixes A-D.
The maximum, minimum, average, and change in pressure delivered by each
pumping unit is shown.
Table 4
Suction and Discharge Pressures
Pump
Model
Damper
Pressure, psi
Discharge
Suction
Maximum
Minimum
Average
AP
Maximum
Minimum
Average
AP
Wheatley
White Rock
1,064
997
1,036
67
78
31
47
47
Young
1,102
1,015
1,060
88
91
40
57
50
Greer
1,123
1,001
1,058
123
77
37
43
41
None
1,148
898
1,063
250
68
23
38
45
Union
White Rock
1,085
1,020
1,056
66
94
29
27
65
Young
1,089
1,008
1,047
81
107
43
25
64
Greer
1,116
925
1,041
192
126
-13
27
138
None
1,167
906
1,045
262
143
-8
32
151
Sund¬
strand
White Rock
1,057
1,025
1,041
32
36
17
26
19
Young
1,079
1,044
1,060
35
77
54
65
23
Greer
1,047
1,014
1,029
34
68
47
56
22
None
1,061
1,020
1,040
41
30
15
23
15
FMC
White Rock
1,115
1,036
1,080
79
89
-16
24
106
Young
1,067
1,001
1,036
67
114
7
45
107
Greer
1,097
1,034
1,063
64
111
6
44
105
None
1,231
906
1,093
325
115
4
41
111
18
Chapter 5 Test Results
Evaluation of the change in pressure for each pumping unit with a pulsation
damper and without a pulsation damper shows significant decreases in dis¬
charge pressure surges can be obtained with the use of a pulsation damper.
However, Sundstrand pump performance was not significantly affected by any
pulsation damper and was generally 50 percent less than all other pumping
units with pulsation dampers (35 psig compared with 70 psig). Suction pres¬
sure surges did not seem to be significantly affected by pulsation dampers.
Vertical and horizontal acceleration (vibration)
Vibration data were obtained electronically in the vertical and horizontal
directions by accelerometers located in two positions: (a) on the pump skid
and (b) on the floor immediately below the pump skid. Figure 9 shows these
locations on the Wheatley pump. Appendixes A-D also show a 0.5-sec record¬
ing of horizontal and vertical accelerations on the pump and floor for each
pumping unit tested except the Sundstrand. Only acceleration data from the
floor position were obtained for the Sundstrand pump because of the absence
of a skid for the pump. Table 5 summarizes information from Appendixes A-
D. Acceleration is given in gravity force equivalents (G’s).
Figure 9. Accelerometer locations (start-up tests)
Based on the Table 5 data, generally, accelerations appear independent of
the presence or type of pulsation damper. The Union pump tested significantly
lower (several hundred percent) for acceleration data obtained from die skid
(and thus the floor). The Sundstrand pump accelerations were also signifi-
candy lower based on the floor-mounted accelerometer data.
Chapter 5 Test Results
Table 5
Acceleratlon/VIbratlon Test
, ■
Acceleration, G's
Vertical on Skid
Horizontal on Floor
Vertical on Floor
Horizontal on Floor
Model
Damper
Maximum
Minimum
Maximum
Minimum
Maximum
Minimum
Maximum
Minimum
Wheatley
White
Rock
17.5
-18.7
8.1
-9.0
0.8
-1.0
m
-0.5
Young
16.3
-16.6
8.3
-7.4
1.0
-1.2
0.5
-0.6
Greer
17.5
-15.8
8.9
-8.4
1.1
-0.9
0.5
-0.4
None
13.7
-15.2
H
-8.7
0.9
-0.7
EH
-0.5
Union
White
Rock
4.0
m
3.6
-4.0
■1
0.1
-0.1
Young
3.3
—
-3.5
3.4
-2.9
0.2
-0.2
0.2
-0.1
Greer
1.8
-1.7
2.3
-2.7
0.3
-0.3
0.1
-0.1
None
5.5
-5.2
6.1
-6.3
0.4
-0.3
0.2
-0.2
Sund-
strand
White
Rock
N/A
N/A
N/A
N/A
0.9
-0.9
0.4
-0.3
Young
N/A
N/A
N/A
N/A
0.6
-0.6
0.2
-0.3
Greer
N/A
N/A
N/A
N/A
H
-0.6
N/A
N/A
None
N/A
N/A
N/A
N/A
0.7
-0.6
0.3
-0.4
FMC
White
Rock
14.8
-17.9
19.0
-18.0
N/A
N/A
1.0
-1.0
Young
12.9
-14.0
15.7
-15.3
1.1
-1.1
N/A
N/A
Greer
12.3
-10.3
14.5
-16.2
1.5
-1.4
1.2
-1.2
None
8.8
-9.0
9.7
-11.2
1.3
-1.2
0.8
-0.8
Noise level
Decibel (db) readings were taken at three locations, 2 ft horizontally and
3 ft vertically from the bottom of the pumping unit. Reference can be made to
Figure 2 for the locations of noise level readings. Table 6 summarizes average
noise level readings for all pumping units tested.
All pumping units tested approximately 90 db. The Wheatley pump was
the loudest, averaging nearly 97 db. Noise levels appear to be independent of
the presence or type of pulsation damper used.
20
Chapter 5 Test Results
Table 6
Noise Level
Pump Model
Damper
Average Decibel Reading, db
Wheatley
White Rock
98
Young
97
Greer
96
None
96
Union
White Rock
91
Young
81
Greer
92
None
Sundstrand
White Rock
88
Young
93
Greer
89
None
88
FMC
White Rock
91
Young
89
Greer
90
None
90
Temperature and salinity
The temperature and salinity of the test fluid at the start of each pumping
unit test is summarized in Table 7. Test fluid temperatures rose to approxi¬
mately 35 to 40 °C for all pumping units after approximately 6 hr of contin¬
uous operation.
Table 7
Temperature and Salinity at Beginning of Test
Pump Model
Wheatley
Damper
Sundstrand
White Rock
Young
Greer
None
White Rock
Young
Greer
None
White Rock
Young
Greer
None
White Rock
Young
Greer
None
Salinity, %
Water Temperature, °C
Chapter 5 Test Results
Start-up test summary and endurance test pumping unit selection
Based on results from the initial inspection and start-up tests, three pumping
units were selected for endurance testing.
Given acceptable test results from the four pumps tested during start-up
tests, the Ingersoll-Rand pump, experiencing valve problems, was eliminated as
a candidate for endurance testing.
The FMC pump experienced significant leaking around the leather packing
in one of the cylinders from the onset of start-up testing. The packing was
removed, reassembled, allowed to soak (leather packings swell and seal), and
continued to leak for the duration of start-up testing. All other pumps tested
performed acceptably. Given this situation and the qualitative assessment that
the FMC pump neither tested superior or inferior to the other pumps according
to the data obtained during start-up testing, the FMC was eliminated as a can¬
didate for endurance testing.
The three pumps to be used during endurance testing were then defined:
Union, Wheatley, and Sundstrand.
Based on data presented in Table 4 concerning pressure, the damper provid¬
ing the least amount of pressure surge protection (greatest delta P) for the
Wheatley and Union pumps is the Greer 4 qt. No damper is required for
the Sundstrand pump. The White Rock damper showed smaller pressure
surges in both the Union and Wheatley pumps than in the Young Engineering
damper and thus was selected to accompany the pump exhibiting the highest
pressure surge with no damper, which was the Union pump. Thus, endurance
testing pumping units are defined: Union pump with White Rock damper,
Wheatley pump with Young engineering damper, and Sundstrand pump with
no damper.
Endurance Testing
The following pumping units were selected for endurance testing and
installed in the test facility (Figure 2).
a. Wheatley pump with Zemark (Young Engineering) damper.
b. Union pump with White Rock damper.
c. Sundstrand pump with no damper.
Each pump was anchored through the manufacturer’s mounting brackets
with four (3-in. by 7/8-in.) concrete anchors. A 3/8-in. polypropylene pad
separated each pump from the concrete pad.
Chapter 5 Test Results
Each unit was targeted to operate 20 hr per day, 7 days a week for 2,000 hr
or until failures necessitated the end of testing. Daily records were kept for
each pumping unit concerning all operations, including routine maintenance
and failure events. Parameters such as suction pressure, discharge pressure,
flow rate, test fluid temperature and salinity, and supply voltage were the same
for all pumping units and are summarized below.
a. Pump suction pressure: 32 psig (manual gauge reading).
b. Pump discharge pressure: 1,000 psig (manual gauge reading).
c. Flow rate: 60 gal per min.
d. Test fluid temperature and salinity: Temperature varied with seasonal
changes. However, regardless of the starting temperature, the test fluid
reached a steady state temperature after approximately 3 hr of daily
operation of 115 °F. The temperature remained constant for the dura¬
tion of the daily test because of the cooling tower and freshwater
supply. Salinity varied each day from a morning high of approximately
2.2 percent (salt added) to a low of 1.5 percent as the freshwater source
coupled with system leaks lowered the salinity.
e. Supply voltage: Three phase, 60 Hz, 460 V.
Parameters such as power consumption, failure scenarios, and routine main¬
tenance are described for each individual pumping unit.
Figure 10 displays power consumption for each pumping unit during endur¬
ance testing. The linear nature of all pumping units power consumption indi¬
cates pumping efficiencies remain constant. Wire to water efficiency for each
pumping unit can be calculated from Figure 11 in the following manner:1
Hp = 1.34 Kw (1^
where
Hp = motor horsepower
Kw = kilowatt (power) usage (dy/dx Figure 11)
Hp = Qtfy/550 Ew/w
(2)
1 Streeter, V. L., and Wylie, B. E. (1985). Fluid mechanics. 8th ed., McGraw-Hill, New
York.
Chapter 5 Test Results
23
KW HOURS
EXDURANCE TESTING
POWER CONSUMPTION (KW HR)
15000 +
10000 4-
POWER REOUIREUENT (KW)
5000 +
sunostrand
AWtEATUK
UNION
<Jy/d*
WIRE TO WATER
EFflOENCr
SUNSTRAND
35 KW
47 77.
WHEATLET
26 KW
67.
UNION
y kw
84 5X
KMEA7EEY
<1.216 Hr* )
500
1000 2000
TIME. (HR.)
2000
2500
Figure 10. Power consumption during endurance testing
where
Q = delivered flow rate, cfs (60 gpm = 0.134 cfs)
H = delivered head, ft (100 psig = 2,308 ft)
Y= fluid density, lb/ft (approximately 62.4 lb/ft)
Ewlw = wire to water efficiency
Substituting the first equation into the second, wire to water efficiency can be
expressed
Ewfw = 26. 1 85/ATw
This results in the following efficiencies for each pump:
Wheatley 93.6 percent
24
Chapter 5 Test Results
Union
84.5 percent
Sundstrand 47.7 percent
Figure 1 displays endurance testing results for the three pumping units
tested. Failure scenarios are indicated by an upper case “F” with the hour of
operational failure indicated below. Routine maintenance (oil changes) is
indicated. Failure scenarios are described for each pumping unit and indicate
the failure, cause of failure, remediating actions, time to repair (man-hours),
availability of spare parts, and any comments about the failure.
Other parameters such as the mean time between failures (MTBF), mainte¬
nance ratios, and inherent availability are evaluated from Figure 10 and daily
records. These parameters quantitatively describe the reliability, availability,
and maintainability of the pumping units.
Reliability
Reliability can be defined as the probability that a product will perform a
specific function under specific conditions for a stated period of time.1 Quan¬
titatively, reliability can be expressed as the MTBF for a particular product
under specific conditions for a stated period of time. The specific conditions
for the operation of the high pressure pumping units have been previously
detailed; thus, MTBF can be expressed as
MTBF = T/n
where
T = total test time, hr
n = total number of failures
Maintainability
Maintainability can be defined as the probability that a failed system is
restored to operable condition in a specified downtime when maintenance is
performed under specified conditions. Downtime may have three components:1
1 Babcock, D. L. (1991). Managing engineering and technology. Prentice Hall, Englewood
Cliffs, NJ.
Chapter 5 Test Results
a. Administrative and preparation time.
b. Logistics time.
c. Active maintenance time.
Administrative and preparation time was not considered in this study
because of the variances between testing conditions and actual field use condi¬
tions. Logistics and active maintenance times were utilized to determine the
three maintenance ratios (MR) that are used to define maintainability. They
are as follows:
MR1 = TMT/T (3
where
MR1 = total maintenance time ratio (includes routine preventive and
corrective maintenance)
TMT = total maintenance time, hr
T = total test time of the pumping unit, hr
MR2 = CMT/T
where
MR2 = corrective maintenance time ratio
CMT = total corrective maintenance time, hr
T = total test time of the pumping unit, hr
MR3 = PMT/T
(5)
where
MR3 - preventive maintenance time ratio
PMT = total preventive maintenance time, hr
T = total test time of the pumping unit, hr
Chapter 5 Test Results
Availability
Availability can be defined as the probability that a system will operate
satisfactorily when required. The inherent availability (Ai) of a system con¬
siders only corrective maintenance under ideal support conditions, not consid¬
ering administrative or logistical delays. Inherent availability can be expressed
as:1
Ai =
MTBF
MTBF + MTTR
(6)
where
MTTR = mean time to repair (active corrective maintenance only)
RAM Results
Wheatley pump with Zemark (Young Engineering) damper
Figure 10 shows the timetable for the Wheatley pumping unit, which ran a
total of 1,216 hr with seven failure events. This establishes an MTBF of
173.7 hr. The seven failure events are described below:
FI (215 hr): Oil leak in middle cylinder.
Cause: Oil seal failure.
Fix: Seal removed, inspected, and reinserted. Seal was not replaced
because leak was not severe at this time.
Corrective time: 1 hr.
Availability of part: Not applicable.
F2 (309 hr): Oil leaks in middle three cylinders - Operation of pump
terminated.
Cause: Oil seal failures.
Fix: All five seals removed and inspected. New manufacturer’s seals
ordered for all cylinders.
Corrective time: 2 hr.
Availability of parts: Parts were available and delivered to testing site
within 5 working days of order.
F3 (309 hr): Plunger coating found to be cracking on two cylinders.
Cause: Unknown - Discovered only because of response to F2.
Fix: Plungers removed, inspected, and reinserted.
1 Babcock, D.L. (1991). Managing engineering and technology. Prentice Hall, Englewood
Cliffs, NJ.
Chapter 5 Test Results
Corrective time: 2 hr.
Comment: At this time, there were no apparent side effects of this
failure. It was ruled a failure because of the potential to cause
cessation of operation. Verbal communications with the manufac¬
turer indicated this was a known problem for this particular
plunger coating. Verbal communications with BRDEC also indi¬
cated that this type of failure has been observed before on the
operating Wheatley pumps.
F4 (330 hr): New oil seals exhibiting severe leaking.
Cause: Oil seal failure.
Fix: Manufacturer seals removed and nonmanufacturer seals ordered
(see comment).
Corrective time: 3 hr.
Availability of parts: Seals were available for order from local retailer
and were on site within 4 working days.
F5 (680 hr): Electric motor mount broken.
Cause: Structural failure at weld between motor and mount because of
mechanical vibration.
Corrective time: 16 hr.
Availability of parts: Not applicable.
Comment: This was a very odd failure. Since the motor did not
receive an external load at anytime during the testing, the struc¬
tural failure must have been caused by fatigue because of mechani¬
cal vibration. Since the motor is mounted vertically above the
pump (and thus foundation), it does seem logical that this configu¬
ration is more likely to receive higher mechanical vibration than
the traditional pump/motor combination on the same horizontal
plane.
F6 (885 hr): Oil seal failures in middle three cylinders (severe).
Cause: Oil seal failures.
Fix: Middle three seals removed and reinserted trying to establish new
seat for existing seals.
Corrective time: 3 hr.
Availability of parts: Not applicable.
Comment: Oil leak requiring additional 1 to 2 qt of oil per day to be
added for continued operations.
F7 (1,195 hr): Water leaks developing at three packing glands.
Cause: Deteriorated plunger coating on three cylinders.
Fix: New plungers and new packing assemblies needed for these three
cylinders immediately and most likely on all five in immediate
future. Tightening of packings was attempted and did not alleviate
the problem.
Corrective time: 8 hr.
Availability of parts: Plungers and packings were available from man¬
ufacturer but not ordered.
Chapter 5 Test Results
29
Comment: The corrective action needed for this failure coupled with
the recurring oil leaks necessitated the need to stop pump testing
because of excessive repair costs and time. The corrective time of
8 hr is an estimate based on previous repair times for this pump.
Summing all the corrective times for the Wheatley pumping unit gives a
total of 35 man-hours devoted to corrective maintenance. This yields a correc¬
tive maintenance time ration of the following:
MR2 = 35 hr/ 1,216 hr
MR2 = .029
Preventive maintenance time was found to be 19 hr (1 hr for every 63 hr of
pump operation) making routine checks on all pumping unit components. This
establishes a preventive maintenance time ratio of the following:
MR3 = 19 hr! 1,216 hr
MR3 = 0.016
Total maintenance time is the sum of preventive and corrective mainte¬
nance, which is 54 hr. This yields a total maintenance time ration of the
following:
MR1 = 54 /jr/1,216 hr
MR1 = 0.044
The MTTR is the total corrective maintenance hours (35) divided by the
number of failures (7). Thus, MTTR = 5.0 hr. The Ai is established as
follows:
(173.7 + 5)
Ai = 0.972
Chapter 5 Test Results
Union pump with White Rock damper
Figure 1 1 shows the timetable for the Union pumping unit, which ran for a
total of 405 hr with two failure events. This establishes an MTBF of 202.5 hr.
The two failure events are described below.
F8 (292 hr): Knocking sound on cylinder farthest from suction end
accompanied by moderate water leak from packing.
Cause: Largely unknown, but suspected to be first signs of improper
alignment between power and fluid ends.
Fix: Alignment was checked according to manufacturer’s operations
manual; adjustments were not deemed necessary, as all compo¬
nents appeared to be acceptable. Packing was tightened slightly,
but leak persisted.
Corrective time: 4 hr.
Availability of parts: Not applicable.
F9 (405 hr): Piston failure in same cylinder as F8. Broken where
piston meets rod (threaded connection).
Cause: Possible misalignment and/or mechanical vibration.
Fix: New piston and rod assembly necessary as well as packing gland
assembly. All necessary parts were ordered and were not deliv¬
ered to sight for 54 working days after a purchase agreement was
established (see comment). Pumping unit never fixed because of
manufacturer delays.
Corrective time: 6 hr (estimate based on previous maintenance).
Availability of parts: Manufacturer had no parts readily available and
had to machine the parts. This resulted in unexpected delays of
testing and ultimately the cessation of the Union pumping unit for
this test.
Comment: Verbal communications with the manufacturer revealed the
pump to be singularly made for BRDEC. Spare parts were not
readily available because of this and had to be manufactured to
meet replacement requirements. Availability of spare parts, in this
instance, is poor.
Summing all the corrective times for the Union pumping unit gives a total
of 10 man-hours devoted to corrective maintenance. This yields a corrective
maintenance time ratio of
MR2 = 8 hr/405 hr
MR2 = 0.020
Preventive maintenance time was found to be 6.5 hr (1 hr for every 63 hr
of pump operation) making routine checks on all pumping unit components.
This establishes a preventive maintenance time ratio of
Chapter 5 Test Results
31
MR3 = 6.5 hr/405 hr
MR3 = 0.016
Total maintenance time is the sum of preventive and corrective mainte¬
nance, which is 16.5 hr. This yields a total maintenance time ratio of
MR1 = 16.5 hr/ 405 hr
MR1 = 0.041
The MTTR is established as (10/2) 5.0 hr. Thus, the inherent availability is
defined.
Ai =
202.5
(202.5 + 5.0)
Ai = 0.976
Sundstrand pump with no damper
Figure 1 1 shows the timetable for the Sundstrand pumping unit, which ran
for a total of 2,353 hr with no failure events. This establishes a MTBF of
2,353 hr/0 (theoretically undefined; an infinitely large approximation will be
made for the MTBF).
A total of zero man-hours were devoted to corrective maintenance since
there were no corrective measures taken. This yields a corrective maintenance
time ratio of
MR2 = 0 hr/ 2,353 hr
MR2 = 0.0
Preventive maintenance time was found to be 12.5 hr (1 hr for every 189 hr
of pump operation) making routine checks on all pumping unit components.
This establishes a preventive maintenance time ratio of
Chapter 5 Test Results
MRS = 12.5 hr/ 2,353 hr
MR3 = 0.005
Total maintenance time is the sum of preventive and corrective mainte¬
nance, which is 12.5 hr. This yields a total maintenance time ratio of
MR1 = 12.5 hr! 2,335 hr
MR1 = 0.005
The MTTR is established as zero. Thus, the inherent availability is defined
as MTBF/MTBF:
Ai = 1.0 (by definition )
Table 8 summarizes the reliability, availability, and maintainability test
results.
Table 8
RAM Results
MTBG
MR1
MR2
MR3
Ai
Sundstrand
a
0.005
0.0
0.005
1.0
Wheatley
173.7
0.016
0.029
0044
0.972
Union
202.5
0.041
0.020
0.016
0.976
Chapter 5 Test Results
33
6 Summary and Conclusions
The Sundstrand pump clearly tested superior through all phases of testing.
Centrifugal technology is markedly different from positive displacement and
has advantages and disadvantages. Most notably, the Sundstrand was half as
efficient as the Union and Wheatley pumps, thus requiring twice the power to
operate under similar conditions. However, because of its light weight, lack of
pulsation damper, superior RAM characteristics, and low vibrations, other
technologies such as energy recovery systems and smart motor control devices
may lessen the power requirement (generator size) to start and operate centri¬
fugal pumps in conjunction with ROWPU technologies.
RAM results from the Union and Wheatley pumps required scrutiny
because of differing test lengths. RAM ratios shown in Table 8 (and
Figure 11) may tend to change with additional test hours. Maintenance ratios
would tend to increase with additional operation and MTBF’s would tend to
decrease.
The effects of increased operation hours on inherent availability are
unknown.
Chapter 6 Summary and Conclusions
Appendix A
Wheatley Pump
No Pulsation Dampener
Appendix A Wheatley Pump
RTLEY PUMP
A 2
Appendix A Wheatley Pump
WHEATLEY PUMP, NO DAMPENER, INPUT PRESSURE
HISTORY TEST 12 11/30/88
©
in
A3
Appendix A Wheatley Pump
TIME-SEC
WHEATLEY PUMP, NO DAMP., V. ACCEL. PUMP
WHEATLEY PUMP, NO DAMP., V. ACCEL. FL
'ORY TEST 12 11/30/88
Young Pulsation Dampener
Appendix A Wheatley Pump
>~
—
LU
cd cn
_J
>- CM ©
1—
CC SI CD
az
o — s
UJ
(7)
rc
•—
rs
X
LU
X X CD
•—CO
i — ^ ar r
•091
09
Appendix A Wheatley Pump
Appendix A Wheatiey Pump
All
A12
Appendix A Wheatley Pump
White Rock Pulsation Dampener
Appendix A Wheatley Pump
A15
Appendix A Wheatley Pump
.05
Greer 1-Gal (4-Qt) Pulsation Dampener
A22
Appendix A Wheatley Pump
Appendix A Wheatley Pump
A27
IME-SEC
Appendix B
Union Pump
No Pulsation Dampener
Appendix B Union Pump
UNION PUMP. NO DAMPENER. H. ACCEL. FL.
I ME HISTORY TEST 30 01/27/89
AX 0. 19 G'S MIN -0. 16 G'S
VG 0.00 G'S RMS 3.04 G'S
UNION PUMP, NO DAMPENER, V. ACCEL. FL.
ME HISTORY TEST 30 01/27/89
X 0.36 G'S MIN -0.31 G'S
B7
Appendix B Union Pump
TIME-SEC
UNION PUMP, YOUNG DRMP . . OUTPUT PRESSURE
TIME HISTORY TEST 27 12/19/88
TIME-SEC
©
in
BIO
Appendix B Union Pump
UNION PUMP, YOUNG DRMP . , V. RCCEL
White Rock Pulsation Dampener
Appendix B Union Pump
B15
TIME HISTORY
MAX 1085.
RVG 1055.
B16
Appendix B Union Pump
TIME HISTORY TEST
Appendix B Union Pump
B19
UNION PUMP, WHITE ROCK DRMP. , H. RCCEL. FL.
HE HISTORY TEST 24 12/14/88
X 0. 14 G'S MIN -0. 13 G'S
G 0.00 G'S RMS 0.03 G'S
UNION PUMP, WHITE ROCK DAMP., V. ACCEL. FL.
IME HISTORY TEST 24 12/14/88
Appendix B Union Pump
B21
TIME-SEC
Greer 1-Gal (4-Qt) Pulsation Dampener
B22
Appendix B Union Pump
UNION PUMP. GREER DRMP . . OUTPUT PRESSURE
HE HISTORY
UNION PUMP, GREER DAMP., H. RCCEL . PUMP
IME HISTORY TEST 29 01/26/89
Appendix B Union Pump
TIME-SEC
UNION PUMP, GREER DAMP., V. ACCEL . PUMP
IME HISTORY TEST 29 01/26/89
RX 1.76 G'S MIN - 1.66 G'S
VG 0.03 G'S RMS 0.43 G'S
B26
Appendix B Union Pump
TIME-SEC
UNION PUMP, GREER DRMP. , V. RCCEL. FL
TIME HISTORY TEST 29 01/26/89
MAX 0.29 G'S MIN -0.27 G'S
cn AVG 0.00 G'S RMS 0.07 G'S
B28
Appendix B Union Pump
TIME-SEC
Appendix C
Sundstrand Pump
No Pulsation Dampener
Appendix C Sundstrand Pump
SUNSTRRND PUMP, NO DRMPENER, OUTPUT PRESSURE
TIME HISTORY TEST 36 2/24/89
MAX 1061.41 PSI MIN 1020.27 PSI
RVG 1040.45 PSI RMS 0.00 PSI
C2
Appendix C Sundstrand Pump
.00
Young Engineering Pulsation Dampener
Appendix C Sundstrand Pump
SUNSTRAND PUMP, YOUNG DAMP . * OUTPUT PRESSURE
HISTORY TEST 34 2/22/89
Appendix C Sundstrand Pump
TIME-SEC
SUNSTRRND PUMP, YOUNG DRMP . , V. RCCEL. FL.
HE HISTORY TEST 34 02/22/89
X 0.58 G ' S HIN -0.59 G ' S
G 0.01 G' S RMS 0.18 G ' S
White Rock Pulsation Dampener
Appendix C Sundstrand Pump
SUNSTRAND PUMP, WHITE ROCK DAMP., H. RCCEL. FL
E HISTORY TEST 35 02/23/89
SUNSTRRND PUMP, WHITE ROCK DRMP . , V. RCCEL . FL
Appendix C Sundstrand Pump
Greer 1-Gal (4-Qt) Pulsation Dampener
C16
Appendix C Sundstrand Pump
SUNSTRRND PUMP, GREER DRMP. , OUTPUT PRESSURE
Appendix C Sundstrand Pump
C17
IME-SEC
Appendix D
FMC Pump
No Pulsation Dampener
Appendix D FMC Pump
FMC PUMP, NO DAMPENER, OUTPUT PRESSURE
©
ID
Appendix D FMC Pump
TIME-SEC
FMC PUMP, NO DAMPENER, H. ACCEL. PUMP
TIME HISTORY TEST 49 03/07/89
MAX 9.69 G'S MIN -11.21 G'S
RVG -0.59 G'S RMS 1.39 G'S
Appendix D FMC Pump
TIME-SEC
FMC PUMP, NO DAMPENER, V. ACCEL. PUMP
D5
Appendix D FMC Pump
IME-SEC
Young Engineering Pulsation Dampener
D8
Appendix D FMC Pump
FMC PUMP, YOUNG DAMP., V. ACCEL. PUMP
IME HISTORY TEST 45 03/06/89
AX 12.92 G'S MIN -13.95 G'S
VG -0.22 G'S RMS 1.46 G’S
D12
Appendix D FMC Pump
TIME-SEC
Appendix D FMC Pump
White Rock Pulsation Dampener
D15
Appendix D FMC Pump
Greer 1-Gal (4-Qt) Pulsation Dampener
Appendix D FMC Pump
D21
s
in
isd-3ynss3ud
D22
Appendix D FMC Pump
FMC PUMP, GREER DRMP . , V. RCCEL . FL.
TIME HISTORY TEST 48 03/07/89
MAX 1.45 G'S MIN -1.42 G'S
Appendix D FMC Pump
D27
TIME-SEC
Appendix E
Copper-Nickel Test Results
EDX unit. Scanning electron microscope (SEM) micrographs of inner pipe
wall features were produced, elemental composition of deposits on the surface
were identified, depth of pitting was determined, and a high resolution X-ray
map showing distribution and concentration of elements was created.
The untested straight pipe shows preexisting voids and grain boundaries
(Figures El and E2), while the untested elbow shows similar features and
striations likely to have resulted from the bending process (Figures E3, E4, and
E5). Preexisting voids were generally less than 25 pm long and less than
5 pm wide. Some spherical voids were present ranging from 2 pm, to less
than 1 pm in diameter. The voids in the elbow appear to be stretched in the
long-axis direction of the pipe. Grains were approximately 5 pm in size.
Surface debris associated with the grain boundaries range from 0.1 to 0.8 pm
in diameter (Figure E2).
Surface debris (scalelike material) was evident on the inside surface of the
tested pipes (Figures E6 and E7). Large pits were also observed (Figures E8
and E9). These large pits were typically less than 200 pm in diameter. Pits
were commonly 0.5 pm in diameter and believed to have similar depths as
diameter. Figures E10 and Ell show the inside surface of the straight pipe
and elbow, respectively, after the scale was removed.
Elemental analysis of untested pipe (Figure E2) using EDX indicated that
copper (Cu) and Nickel (Ni) were the major elements present. Iron (Fe), man¬
ganese (Mn), oxygen (O), and carbon (C) were present as trace elements (Fig¬
ure E12). EDX data collected from the untested elbow shown in Figure E5
indicate the same chemistry as that of the untested straight pipe (Figure E13).
The spectrum shows chemistry similar to what was collected from scale in the
elbow. Aluminum (Al), magnesium (Mg), silicon (Si), sulfur (S),
chlorine (Cl), sodium (Na), zinc (Zn), oxygen (O), and carbon (C) were
present.
High resolution X-ray maps were made of tested pipes. Figures E14, E15,
and El 6 are hard copies of X-ray maps made from the inside surface of the
elbow. The area was partially coated with the scale material. Figure E14
Appendix E Copper-Nickel Test Results
shows the distribution of copper, Figure E15 shows the distribution of nickel,
and Figure El 6 the distribution of chloride.
Based on the SEM and EDX findings, the following conclusions can be
made:
The small particles shown in Figures E2 and E5 may be a copper/nickel
oxide that has formed over the surface.
Small pits have formed in the surface of the used pipe. The depths of the
pits found in the elbow are 25 to 50 pm. Pits of this size do not pose a signif¬
icant structural threat to the integrity of the pipe. Extrapolation of pit growth
to yield a useful life approximation based on the known information is not
warranted.
The X-ray maps of Cu, Ni, and G show Cu to be common throughout the
pipe. The map of Ni parallels that of the G map (Figure El 6), suggesting the
G is now a Cu/G phase.
The other elements detected in trace amounts are associated with the added
salt and potable drinking water comprising the test fluid.
Appendix E Copper-Nickel Test Results
Appendix E Copper-Nickel Test Results
Figure El . X700. Irregular voids were common to inside Figure E2. X5,000. Round voids like those seen in photo¬
surface of pipe. “Mosaic” design can be seen on micrograph were also quite common. Small par-
pipe surface tides on surface are probably Cu/Ni oxides
Appendix E Copper-Nickel Test Results
Figure E3. XI 50. Inside surface of elbow was striated Figure E4. X700. Micrograph of untested elbow shows same
“mosaic” design and irregular voids present in
straight pipe
Appendix E Copper-Nickel Test Results
Figure E5. X2,500. Higher magnification showing “mosaic" Figure E6. XI 00. “Scale” collected on inside surface of tested
design in elbow. Voids in elbow appear to be pipe. Clear looking surface in center contains pits,
stretched in direction of long axis Figure El 3 shows chemistry of scale
Appendix E Copper-Nickel Test Results
Figure E7. X250. Shows deposits inside elbow. Surface of Figure E8. X220. View is typical of larger pits. Note that
pipe is also shown. Much of original surface is smaller pit is inside larger pit
gone
E7
Appendix E Copper-Nickel Test Results
Figure E9. X600. Several large pits were found on inside sur- Figure El 0. X3,000. Shows damage in straight pipe caused
face of elbow. This one is about 150 by 100 pm in by saline solution. Pits average approximately
sjze 0.5 pm in diameter. “Valleys” that create
boundaries for mosaic design have been
deepened, and pits are in valleys
Appendix E Copper-Nickel Test Results
Figure Ell. X2.000. Shows damage done to elbow. Pits are
larger in elbow area than they were in straight
pipe
Figure El 2. Spectrum of unused straight piece of pipe
Appendix E Copper-Nickel Test Results
E9
Figure El 3. Spectrum of deposits in used straight pipe
E10
Appendix E Copper-Nickel Test Results
Figure El 4. EDX high resolution X-ray map for copper in a tested pipe. Light areas show
homogeneous distribution of copper in sample
Appendix E Copper-Nickel Test Results
Figure El 5. EDX high resolution X-ray map showing distribution of nickel. Light areas indicate
concentrations of nickel in some areas
E12
Appendix E Copper-Nickel Test Results
Figure El 6. EDX high resolution X-ray map of chlorine. Orientation of concentration of
chlorine is similar to that of nickel shown in previous figure
Appendix E Copper-Nickel Test Results
E13
REPORT DOCUMENTATION PAGE
Form Approved
OMB No. 0704-0188
1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE
October 1994
3. REPORT TYPE AND DATES COVERED
Final report
4. TITLE AND SUBTITLE
Reliability, Availability, and Maintainability Testing of High Pressure
Pumping Units for the Mobile Army 600 Reverse Osmosis Water Purification
Unit _ _
6. AUTHOR(S)
Wayne W. Sharp, Charles R. O’Quinn, Michael G. Channell
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)
Howard K. Bell Consulting Engineers, Inc.
P.O. Box 546, Lexington, KY 40585
U.S. Army Engineer Waterways Experiment Station
3909 Halls Ferry Road, Vicksburg, MS 39180-6199
9. SPONSORING /MONITORING AGENCY NAME(S) AND ADDRESS(ES)
Belvoir Research, Development, and Engineering Center
Fort Belvoir, VA 22060-5606
8. PERFORMING ORGANIZATION
REPORT NUMBER
Technical Report
EL-94- 15
10. SPONSORING /MONITORING
AGENCY REPORT NUMBER
11. SUPPLEMENTARY NOTES
Available from National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161.
12a. DISTRIBUTION /AVAILABILITY STATEMENT
12b. DISTRIBUTION CODE
Approved for public release; distribution is unlimited.
13. ABSTRACT (Maximum 200 words)
Mobile reverse osmosis water purification units (ROWPUs) are used by the Army for production of potable
water from saline, brackish, and freshwater sources. The reverse osmosis is a membrane process that requires a
high pressure feed stream; therefore, a high pressure pump is a key component of the ROWPU.
This report provides data collected from a unified testing program applicable to all high pressure pumping units
that have potential use as ROWPU components. The testing program consists of three phases (initial inspection,
start-up testing, and endurance testing). All pumping units evaluated during this testing were subjected to the three
test phases.
14. SUBJECT TERMS
Feed stream
High pressure pumping units
Reverse osmosis
17. SECURITY CLASSIFICATION 18.
OF REPORT
UNCLASSIFIED
NSN 7540-01-280-5500
SECURITY CLASSIFICATION
OF THIS PAGE
UNCLASSIFIED
19. SECURITY CLASSIFICATION
OF ABSTRACT
15. NUMBER OF PAGES
157
16. PRICE CODE
20. LIMITATION OF ABSTRACT
Standard Form 298 (Rev 2-89)
Prescribed by ANSI Std Z39-18
298-102