CENTRIFUGAL PUMPS INSPECTION AND TESTING inod P. Patel ‘Senior Principal Machinery Engineer M. W. Kellogg Houston, Texas Jim Bro ‘Test Engineer BW/IP International Ine. ‘Tulsa, Oklahoma Vinod Patel has been Senior Principal Machinery Engineer for the MLW. Kellogg Company for over twenty years. In this as signment, he is responsible inthe prepara: tionandaudltingofspecificaions equipment ‘evaluation, engineering coordination, test- ing and insallationstartup of rotating and special equipment. He hus worked in the \arious applications of rotating machinery inthe petrochemical and refinery processes including LNG, olefins, eat cracking and Iydrotreating for domestic and international projets. Mr. Patel received BS. and MS. degrees (Mechanical and Metallurgical Engincering) from M.S. University and Youngstown State Univers ty. respectively He isa registered Professional Engineer inthe State of Texas James R. Bro isa Tes Engineer with BW/ IP International, Pump Division, Tulsa, (Oklahoma. He isresponsibl for supervising the testing of vertical pumps, horizontal pumps, submersible pumps, and submers: tle motors. Mr. Bro has been associated with BWP International, Panp Division fr 15 yearsin the areas of design application, factory and {field testing, and field problems involving vibration, cavtationand pump performance. He coordinates the installation and testing of Research and Development projects. He received a BS. degree (Engineering Technology) from Oklahoma State University (1979), He isa mem- Der of ASME and ISA INTRODUCTION ‘The general requirements for inspection and testing of centri ‘weal pumps are presented, The successful completion of inspec: tion and testing verifies pump design and specification compliance prior to release of pump for shipment and helps avoid field problems. ‘Scope: A general overview of inspection and testing require ‘ments and guidelines i established forthe following ~ Tnspeetion and testing requirements based on ASMEJANSI Standards ANSI B73.1M and B73.2M—1991 Horizontal and Vertical Cenviugal Pumps, and APIG1O, 7th edition. 7 Specifying inspection und testing requirem + Shop test acceptance iteria ration and use of Inspection and Testing checklist. + Review of shop perf + Shop test reporting. ince and NPSH test procedures. Prior to issue of the pump bid inquiry, the requirements of inspection and testing of pumps shall be clearly established inthe pump specification and datasheets. These requirements are subdi- Vided into the following categories: Inspection Requirements, Shop Test Requirements, Inspection and Testing Checklist for Centrifugal Pumps; and Shop Test Procedures Taspection Requirements + Materials (physical properties and che composition) + Castings Casting Repairs Welding Nondestructive Testing mpacr + Visual Inspection Shop Tests Requirements + Hydrostatic Tests ‘Shop Performance and NPSH Tests + Acceptance Inspection and Testing Checklist for Centrifugal Pumps + Single Stage ANSI and General Service Pumps + Heavy Duty Centrifugal Pumps (API) Shop Test Procedures + Performance Test Procedure + NPSH Test Procedure + Test Report4s PROCEEDINGS OF THE TWELFTH INTERNATIONAL PUMP USERS SYMPOSIUM INSPECTION REQUIREMENTS Materials General “The materials of the major pump components shall Be identified inthe pump data sheets with ASTM or ANSI Standards based on API Standard 610, 7th edition, Appendix H or equivalent. When such designation is not available, a tabulation of the physical properties and the chemical composition shall be provided by vendor inthe proposal “Tests for physical properiiesand chemical composition Fomine heat from which material is supplied, i. for pressute containing parts, shafts and impellers are normally fecommended for: «Internals of pump component exposed to hydrogen sulfide (8.8) including traces of HS. + Pumps in highly corrosive serve. + Pumps in eryogenie service (below -S0°C) The material of construction can be further classified as pee following + Cast ion pressure casing material fs generally limited to cooling water and similar services to avoid brittle failure caused from thermal shock, + Sour water service shall be constructed from cast sel mate rial as a minimum, + Step by step procedure of repair itcluding heat treatment where applicable and method of inspection after repair. Norepairsby peening, plugging or imp-egnation are allowed to any ferrous pressure cont ASTM A278, $36, oF 395, respectively. Weldable grades of ste! autings may be repaired based on Scetien VII, Division 1, and Section IX ofthe ASME code, Welding ‘Welding of pressure containing parts, wetted parts, piping, and any weld repairs to these components shall be performed and inspected in accordance withthe requirements of Section VII, Division 1, and Section IX of the ASME Code, All pressure containing welds, including suction and discharge les of pumps with fabricated casingsare generally full-fusion, il penetration welded, and the quality of welding is verified by iondestructive examination. Fabricated casing welds require post weld heat treatment in accordance with Section VII, Division I, of the ASME Code. Piping-to-casing welds are often stress relieved locally. All welds shall be post weld heat treated in aecordance with Section VII, Division 1, UW40 of the ASME Code, Welding of the pressure containing components shal be done by qualified weld- fs in accordance with Section IX of the ASME Code. All welding not covered by Section VIM, Division 1, of the \SME Code such as all noneritical welding, ie, structural weld +The material of specific components Tn hydrogen service shall conform to NACE Standard MR-O17S, ide jing, baseplates, etc, shall be performed inaccordance with Amer jean Weiding Society (AWS) DLL Stan Castings Castings shall be free from any defect such as porosity, racks, blow holes, shrink holes, sale, and similar injurious defects, For critical pump services, pumps application in eryogenic services, of for a specialty casting, inspection is recommended for ny major casting defect repair as follows + Check mill est reports for physical properties and chemical composition property compliance, + Check post repair heat treatment charts, + ChecKwitness nondestructive examination as specified + Checkireview welding procedures equired and Classification of Casting Repairs Major repairs can be defined per ASTM Standards a8 follows + The casting failure to meet hydrostatic test requirements, + Repairs for which the depth of any cavity prepared for repair by welding exceeds 20 percent ofthe wall thickness oF 1.0 in (25, min), whichever isles. + The cavity prepared for weldingis greater than approximately 10 in? (65er + Any repair to cast iron compon ‘major repai, Procedure for Major Repairs In general, all major repairs as previously defined require written acceptance of repair procedure by purchase prior to pro ‘ceeding withthe repair. The procedute generally should cover: + Type of defect and how it was discovered, + Sketch or drawing showing location and depth of the defect. + Method of repair (welding, Critical pump services, and any major repaits as defined earlier, ‘generally requte review and written acceptance of welding proce: dure and procedure qualification, Nondestructive Testing Theextentof nondestructiveexaminationisnormally determined by the pump size, type, materials of construction, or criticality of the pump application, Nondestructive testing is normally agplied to nozzle welds, castings repairs and critical pump materials for components such as casings, impellers, and shafts to determine defects in fabric and any repairs Nondestructive testing shall be perforred in accordance with [ASME Code, prior tothe hydrostatic test as per the guidelines Table | Table 1. Nondestructive Testing Guidelines = — a oe oe‘TUTORIAL ON CENTRIFUGAL PUMPS INSPECTION AND TESTING us Impact Test Carbon steel materials in low temperature service below the 100 hp). + The purchaser’s written approval is required repairs performed bectuse ofa failed hydrostatic test. to any Shop Running Tests Shop running tests are performed to prove the mechanical teri ofthe unit, alignment, vibration level, and to verify pump performance, ‘The requirements for pump performance, vibration, and NPSH testing should be clesrly defined in applicable specfication/ standards and in data sheets prior to bid inquiry issue. For de- tailed summary of testing requirements as specified in API 610, ANSI BT73.1M/B73.2M, and Hydraulic Institute Standard refer to Table 2 "These requirements vary depending onthe pump standards used and requirements specified by EPC contractors and pump users as per following, + ANSIASME B73.1M and B73.2M (1991) specification for horizontal end suetion and vertical nine centrifugal pumps donot ‘mandate performance, mechanical, or NPSH testing. However, ‘when esting isspecified, the acceptance criteria isthat of Hydrau- Tic Institute Standards, unless more stringent limits are imposed, Often, pumps built per ANSI specifications are manufactured as Table 2. Comparison of Centrifugal Pump Testing Requirements standard produets and sold as off the shelf commodities without any shop tests. + API 610, 7th Edition mandates all punps to be tested for ‘mechanical and hydraulic performance veriFcation, + The process industry practices (PIP) for ANSI pumps recom- ‘mends that the performance test be carried out if any of the ‘ollowing conditions are met: ‘Pumps operating in parallel Suction specific speed above 11,000 [Normal flow isless than 10 percent above minimum contin- uous flow. ‘An NPSH testis generally required when available NPSH is Within two t0 six feet of NPSH,. Refer to test specification summary Table 2. Complete unit tests are not normally specified except for ex- ‘tremely critical pumping systems, and when driver size and volt- age level are compatible with that available in the vendor's shop. ‘Shop Test Acceptance Criteria ‘Typical Shop Test Acceptance Criteria arelisted in the Hydrau- lic stitute Standard and AP1610. For summary of testacceptance criteria, refer to Tables 3 and 4 INSPECTION AND TESTING CHECKLIST FOR CENTRIFUGAL PUMPS ‘Typical inspection and testing checklists ae shown in Tables 5, 6,7, and 8 for: Table 3. Comparison Centrifugal Pump Performance Test Accep- tance Criteria,146 PROCEEDINGS OF THE TWELFTH INTERNATIONAL PUMP USERS SYMPOSIUM Table 4. Centrifugal Pumps Vibration Acceptance Criteria Summary. + Single stage ANSI and general service pumps (Table Sand 6), + Heavy duty centrifugal pumps. (APD (Tables 7 and 8) ‘The checklists provide @ convenient way to specify test and inspection requirements and extent of purchaser's participation, ‘This document can be included in the bid inquiry to communicate requirements tothe vendor. Table 5. Cenerifugal Pumps Inspection and Testing Checklist. (Single tage ANSI and general service.) SHOP TEST PROCEDURES Introduction Both EPC contractors and users of centrifugal pumps require shop tests that not only must comply with industry standards for Table 6. Centrifugal Pups Inspection and Testing Checklist. (Single stage ANST and general service) ‘centrifugal pumps, but must also comply with ther specifications Shop tests are more involved now than they were afew years ago. ‘Thisisduetoenhancement of pump reliailty programsand major advancements test equipment, personal computers, and software Performance and NPSH shop testing fas always been a part of| ‘the centrifugal pump industry. Testing methods may vary lightly for each manufacturing facility. If shop test results are wit acceptance criteria guidelines and the pup is properly installed, then the pump should function correctly durin field operation, Related Specifications and Standards ‘The majority ofall specifications and standards refer to either APL Standard 610 or Hydraulic Institue Standards when shop testing requirements are needed, Other testing standards that may be used ate ANSIAWWA E101-88 and ANSIASME B73.1M- 1991. Prior to shop testing of pumps, ensure that purchaser and responsible vendor test personnel are in agreement with the test procedure and acceptance criteria. Test Objectives Centrifugal pumps are shop tested to determine if they fall ‘within the acceptance criteria specified by the purchase order, Acceptance criteria for dynamic testing may cover varioustypesot tests + Performance test + NPSH test + Complete unittest + Auniliary equipment test‘TUTORIAL ON CENTRIFUGAL PUMPS INSPECTION AND TESTING 17 Table 7. Inspection and Testing Checklist—Heavy Duty Centra Table 8 Inspection and Testing Checklist Heavy Duty Centrif gal Pumps (API). ‘gal Pumps (API). sults donot fll within acceptance eritria, then the user o-aecep the test or retest after necessary corrections are made by the manufacturers Horizontal and vertical pumpsare performance tested in various configurations, Different configurations are usually duc to the result of Himitations of the shop test stan. ‘A typical horizontal pump test stand configuration (Figure 1) may utilize either a shop driver ora contract driver. Some shops. ‘may use open loop configurations withand without booster pumps. Vertical pumps may be tested in a closed loop test stand configuration (Figure 2) or an open loop test stand configuration (Figure 3). Shop test stand imitation may allow testing ofthe bowl assembly only. pump length is not excessive, then the complete tuit could be tested with or without the contract driver. = |> d — Performance Test Procedure | ass | | ‘Typical shop test stand limitations ae follows: + Loop pressure + Loop flowrate Figure 1. Horizontal Pump Test Stand Configuration. + Well depth and volume + Crane heisht i ‘The test will require vrioustypes of equipment and instrumen: + Crane loa capacity tation, Test loop will be constructed ofthe following types of + Available NPSH equipment: + Driver power and speed + Storage tank or pit for test uid + Supply power (voltage and current) + Teststandto upportthe pump, driverandauxiliary equipmentus PROCEEDINGS OF THE TWELFTH INTERNATIONAL PUMP USERS SYMPOSIUM Figure 2. Vertical Pump—Closed Loop Configuration - Figure 3. Vertical Pump—Open Loop Configuration. + Piping system equipped with a flow control valve an flowmeter + A negative and postive pressure source mounted on storage tank for NPSH testing + Power source for pump driver ‘The following types of instn + Monitor pressure with hourdon gauge, liquid manometer, of electronic transducer station could be use + Monitor flowrate with venturi meter, pitottube, orifice meter, weight tank, propeller meter, or magnetic meter + Monitor electrical power input with wattchour meter or watt meter + Monitor horsepower requited at pump shaft witha torqueme ter or calibrated motor. + Monitor speed with a stroboscope, hand held tachometer, of key phasor with readout meter + Monitor voltage and current with votage and current trans formers connected to a readout deviee. + Monitor vibration with noncontacting proximity probes or with seismictype velocity sensors mounted onthe bearing housing or held by hand, Allinstrumentation must be calibrated intervals of six months or less, oras determined by the shop testing facility. A certificate of calibration mustbe kept on file, A calibration sticker should also be attached to each piece of instrumentation. Test Stand Setup and Performance Testing ‘The pump first connected to the test sand mounting base and. piping foop. Driver is then mounted to the pump. Alignment ‘etween pump and drivers adjusted toa value within acceptance criteria. Mechanical seal piping for shop testing is connected, Vibration probes are placed at agreed loeations to monitor pump and moor bearing housing vibration levelsand shaft displacement levels. Flow, pressure, speed, and power ronitoring instrumenta- tion isselected and activated. The pump isthen started andtest data can be recorded once the pump and motor temperatures stabilize Recorded test data will usually consist of the following information + Flowrates| + Discharge pressure + Suction pressure + Elevation corrections + Test uid temperature + Test uid specific gravity + Power reading + Voltage at the driver + Current tothe driver + Frequency of supply voltage + Vibration levels + Speed Five to seven test points are taken to cover the flow range from zet0 flow to 120 percent of design flow, depending on which test standard is specified. The number of vibretion reading taken also ‘depends on which test standard is used. See Table 2 for details on typically used test standards. When the test is completed, all recorded data must be converted into a fermat (Figures 4 and 5) that can be easily compared to acceptance criteria, AI pressure readings must be converted to head values. Pump efficiency must be calculated from power readings at each flowpoint. Shop test results may have tbe speed corrected tothe predicted operational field speed. An as tested performance curve is then prepared and included with test report. Fora summary of acceptance criteria inthe Hydraulic Institute ‘Test Standards and the API Standard 610, refer to Tables 3 and 4 ‘These tolerances usually cover flowrate, head, efficiency, vibra tion Tevels, brake horsepower, speed, and oil temperatures in ‘bearing housings. Seal piping arrangementfor shop test may differ from field arrangement. During shop test, seals are inspected for leakage. Leakage rate isa function of test uid, pressure output of ‘pump, test speed, and temperature. Seals are normally tested for zero visible leakage; however, if leakaze rate appears to be excessive, the seal must be repaired or replaced. I'a pur fails the shop test the manufacturer must take corrective action with &‘TUTORIAL ON CENTRIFUGAL PUMPS INSPECTION AND TESTING 149 Aone wl oped wa ia toed Bea ed NPS ied asthe ordinate and flowrate as abse'sa ‘“STypieal recorded test data will usually const of he following information: + Flowrate + Discharge pressure + Suction pressure + Elevation corrections + Test uid temperature + Test fluid specific gravity + Test uid vapor pressure + Speed + Barometric pressure NPSH Testing ‘The most common method used for NPSH testing isto run the ‘pump at constant flowrate and speed while lowering. suction pressure. The suction pressure (static head) is decreased until specified reduction in head is reached at each flowrate. The ‘numberof readings may range from five to twelve. Depending on the test standard used, the numberof floweatescould vary from one 10 Four Fora typical NPSH test fog and curves, refer to Figures 6 and 7 which indicate head reduction for each flowrate point. Figure 5. Pump Test Performance Curve, possible retest, The most common causes ofthese failures are Incortect impeller diameter, rotating elements incorrectly bal anced, poor purnp-to-motor alignment; uncalibrated instrumenta tion, poor quality surface finish on impeller and cases, testing errors and misinterpretation of test requirement, oF acceptance criteria. ‘The test log will consist of recorded test data, caleutated result and an as tested performance curve. The performance curve will consist of a graph of the test data with te total differential hea pump efficiency, NPSH required, and brake horsepower plotted as ‘ordinate with flowrate as abscissa, NPSH Testing ‘Test stand configurations for horizontal pumps and vertical ppumpsare usually inclosed loop systems asshown in Figures 1 and. 2. Some testing facilities may use open loop configurations if pump NPSH requirements are not too excessive. In open loop configurations, the suction pressure to pump is usually adjusted ‘witha suction thro valve, The equipment and instrumentation specified previously for performance testing are also used during NPSH testing, ‘The test log will consist of recorded test data, calculated results and an as tested NPSH curve, Graphical presentation for each Figure 6, NPSH Test Data, “The Hyaraulle institute Standards derine NPSH, (start of cavl- tution) as any change in performance, i. ether a drop in head, iency ata given capacity of a shange in the sound level or vibration. However, because of the difficulty in determin- ‘ng just when the change start, a drop in heal of three per cent is usally accepted as evidence of performance impairment, For ‘multistage pumps, total head is divided by number of stages to Approximate Fist stage head. Dismantling to correct NPSH, re- quires retest Pump Disassembly After Test Pump disassembly for bearings and seal inspection is generally notrequired orrecommended except in specia circumstances such as inconsistent fest results, prototype pump cesign, etc.150 PROCEEDINGS OF THE TWELFTH INTERNATIONAL PUMP USERS SYMPOSIUM Figure 7. NPSH Test Curve. Test Report ‘This report should contain all graphic representations and test, Jog sheets for both the performanceand NPSH test. All test results shouldbe certified by the testing facility. Test results must comply with acceptance criteria specified inthe purchase order. Adjust- ‘ments may be made to actual test results to simulate the "as built™ pump performance characteristics, i, friction loss or viscosity correction of speed adjustment. CONCLUSION Specifying correct inspection and testing requirements, plus verification of pump performance during shop tests, are equally important as the purnp design specification and selection of the right pump forthe specified service. ‘Several alternatives available for centrifugal inspection and testing were explored The review of the typical pump performance and NPSH test ‘procedutes and actual shop practices further emphasize the impor: tance of inspection and testing prior t release of pumps for shipment. BIBLIOGRAPHY API 610, 7th edi Refinery Service, ANSI/ASME Standard B73.1M (1991), Specification for Hori zontal End Suction Centrifugal Pump: for Chemical Process. ANSI/ASME Standard B873.2M (1991), Specification for Vertical Centrifugal Pumps for Chemical Process. AWS.- D11 - Structural Welding Code Stee ic Institute Standards for Centrifugal, Rotary and Reeipro- Pumps - Lith Edition (1983), Hydraulic Institute Test - Standards for Centrifugal Pumps - 1.6. (1988). Process Industry Practices RESP73HV - Specification for Vertical Centrifugal Pumps for Chemical Process - ASME B73.2M 99). MSS SP-55 - Quality Standard for Stesl Castings for Valves, Flanges and Fitings and Other Pipiag Components (Visual Method), NACE MR.O1-75 - Sulfide Stress Corrosion Cracking Resistant Material for Oil Field Equipment Process Industey Practices (PIP) RESP7Si - Exceptions to ASME B.73.1M, Specification for Horizontal 3nd Suction Centrifugal Pumps for Chemical Process - (1991), (1989) - Centrifugal Pumps for General