ML17329A154
| ML17329A154 | |
| Person / Time | |
|---|---|
| Site: | Cook  |
| Issue date: | 08/30/1991 |
| From: | Fitzpatrick E AMERICAN ELECTRIC POWER CO., INC., INDIANA MICHIGAN POWER CO. (FORMERLY INDIANA & MICHIG |
| To: | Davis A NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| AEP:NRC:0966Q, AEP:NRC:966Q, GL-89-10, NUDOCS 9109100094 | |
| Download: ML17329A154 (47) | |
Text
ACCELERATED DISTRIBUTION DEMONSTIWTION SYSTEM rA'EGULATORY INFORMATION DISTRIBUTION SYSTEM (RIDS)
ESSXON NBR:9109100094 DOC.DATE: 91/08/30 NOTARIZED: NO DOCKET CIL:50-315 Donald C. Cook Nuclear Power Plant, Unit 1, Indiana 6 05000315 50-316 Donald C. Cook Nuclear Power Plant, Unit. 2, Indiana 6 05000316 AUTH. NAME AUTHOR AFFILIATION FITZPATRICK,E. Indiana Michigan Power Co. (formerly Indiana 6 Michigan Ele FITZPATRICK,E. American Electric Power Co., Inc. R RECIP.NAME RECIPIENT AFFILIATION DAVXS,A.B. Document Control Branch (Document Control Desk)
SUBJECT:
Forwards response to unresolved item from insp repts 50-316/91-09 re generic ltr l
50-315/91-09 !!'4 89-10, "Motor-Operated Valve Test Accuracy."
DISTRIBUTION CODE'064D COPIES RECEIVED'LTR TITLE: Response to Generic Ltr 89-10, "Safety-Related ENCL L MOV SIZE Testing & urveill /
NOTES' RECIPIENT COPIES RECIPIENT COPIES D ID CODE/NAME LTTR ENCL ID CODE/NAME LTTR ENCL PD3-1 LA 1 0 PD3-1 PD 1 1 D COLBURN,T. 1 1 INTERNAL: BAER, B RES 1 1 NRR GODY,A., JR 1 1 NB/~ MEB 7E 1 1 NRR/LPEB/BC 1 1 EG FILE g 1 1 RES/DSIR/EIB/B 1 1 RNAL: NRC PDR 1 1 NSIC 1 1 D
S D
D NOTE TO ALL "RIDS" RECIPIENTS:
PLEASE HELP US TO REDUCE WASTE! CONTACT THE DOCUMENT CONTROL DESK, ROOM P 1-37 (EXT. 20079) TO ELIMINATEYOUR NAIVE FROM DISTRIBUTION LISIS FOR DOCUMENTS YOU DON'T NEED!
TAL NUMBER OF COPIES REQUIRED: LTTR 11 ENCL 10
. Power Company P.O. Box 16631 r Coiumbus, OH 43216, rg AEP:NRC:0966Q Donald C. Cook Nuclear Plant Units 1 and 2 Docket Nos. DPR-58 and DPR-74 License Nos. 50-315 and 50-316 RESPONSE TO UNRESOLVED ITEM FROM INSPECTION REPORT 50-315/91009 (DRS) and 50-316/91009 (DRS):
MOTOR-OPERATED VALVE TEST ACCURACY U. S. Nuclear Regulatory Commission Document Control Desk Washington, D.C. 20555 Attn: A. B. Davis August 30, 1991
Dear Mr. Davis:
Inspection Report 50-315/91-009 (DRS) and 50-316/91009 (DRS) discussed the results of an inspection of our Generic Letter 89-10 (Motor-Operated Valve) program. The inspection report contained one unresolved item (50-315/91009-01 and 50-316/91009-01), that concerned testing of the OATIS motor-operated valve data acquisition system by Idaho National Engineering Laboratory. The information requested is contained in the attachment to this letter.
This document has been prepared following Corporate procedures that incorporate a reasonable set of controls to ensure its accuracy and completeness prior to signature by the undersigned.
Sincerely, EF~p Vice President EEF/eh Attachment 0600~9
- "waar SOO'Oi4 'iioS~0 F'DR ADGCK 050003i 5 9 PDR
Mr. A. B. Davis AEP:NRC:0966Q cc: D. H. Williams, Jr.
A. A. Blind - Bridgman J. R. Padgett G. Charnoff NFEM Section Chief NRC Resident Inspector - Bridgman
ATTACHMENT TO AEP:NRC:0966Q INFORMATION REGARDING IDAHO NATIONAL ENGINEERING LABORATORY TESTING, OF OATIS MOTOR-OPERATED VALVE DIAGNOSTIC EQUIPMENT
I r IU I
7 BACKGROUND Page 1 At the January, 1990 meeting of the Motor Operated Valve Users Group (MUG) a plan was developed for validation of vendor accuracy claims for Motor Operated Valve (MOV) diagnostic test equipment. The NRC provided funding for the validation effort, and Idaho National Engineering Lab (INEL) was selected to conduct the testing. The MOV diagnostic equipment utilized at Cook, Nuclear Plant is the third generation Operations Analysis 6 Test Interpretive System (OATIS) developed by ABB Impell. The NRC audit of our Generic Letter 89-10 program (March 18-28, 1991) recognized that the OATIS ZII equipment accuracy is a significant portion of the overall program, and considered the review and submittal of the INEL test results an unresolved item (50-315/91009-01; 50-316/91009-01). The OATZS IZZ equipment was tested at ZNEL the week of April 22, 1991, and an interim report with limited data was issued at the MUG summer meeting the week of July 29, 1991. Several data points from the INEL testing were outside the OATZS published instrument accuracy, and on that basis a Problem Report was originated within AEPSC on August 8, 1991. On August 16,
'BB Zmpell originated a Nonconformance Report which will initiate their internal corrective action and reportability reviews. The MUG anticipates having a final report ready for issuance at the winter meeting during the first quarter of 1992, that will contain full stroke data and graphical overlays to compare INEL and vendor data.
SUMMARY
OP RESULTS Attachment 1 is a summary of the vendor's stated measurement accuracy for the OATIS ZZI equipment. Attachment 2 provides a brief description of the OATZS ZII equipment, and compares close stroke thrust at torque switch trip and final load between OATIS and the INEL Motor-Operated Valve Load Simulator (MOVLS). Attachment 3 contains unedited comments from ABB Impell and INEL regarding the overall validation effort and test methodology. Attachments 1, 2, and 3 are taken directly from the MOV User's Group "Progress Report of the Validation Committee" dated July, 1991.
DISCUSSION OP RESULTS I
The results in Attachment 2 are a time-based comparison of discrete poi.nts in the closing stroke of the MOVLS between OATZS III and the INEL standard. These results show deviations in thrust measurement which are large in magnitude, and apparently independent of test parameter manipulation. In evaluating this data, several limitations of the ZNEL testing must be considered:
- 1) The INEL MOVLS was designed to simulate operational and dynamic conditions such as line pressure, differential pressure and rate of loading.
- No comparison to static diagnostic test results was made.
- 2) The SMB-0-25 actuator selected for testing is not typical of those installed at Cook Nuclear Plant. Only 10% of the actuators in the Generic Letter 89-10 program are SMB-0 type, and only 3% are SMB-0 with a 25 ft-lb motor. Zn contrast, 64% of the actuators in our program are SMB-00 type.
- 3) A relatively small amount of data was taken, and limited to a single actuator type, stem configuration, spring pack, and gearing. Three different transducer combinations were stroked nine times at three different loading conditions, for a total of twenty-seven strokes.
- 4) Commentary from different vendors in the MUG Report challenged whether the upper bearing assembly of the MOVLS was defective, that the compressive loads on the stem did not allow for relubrication as may occur during normal valve operation, and that the stem lubricant used has demonstrated poor performance characteristics in the EPRZ sponsored lubrication study. These factors will tend to degrade the torque to thrust conversion and introduce variability in the stem factor and friction coefficient.
Page 2 Despite these limitations which may restrict the widespread applicability of the data, the INEL testing does provide valuable information for comparing equipment accuracy and variables associated with actuator repeatability. ABB Impell has purchased the DaDisp 2.01 software utilized by INEL for data analysi's, and this has already proven to be a powerful tool for evaluation.
Review of the INEL data by AEPSC, Cook Nuclear Plant, and ABB Impell has resulted in the following observations:
- 1) There are timing differences between OATIS and the INEL standard. This is a significant factor in the variability shown in Attachment 2, as was acknowledged by a representative of INEL during the recent MUG meeting.
- 2) OATIS consistently predicted thrust higher than that from the INEL MOVLS.
- 3) Measurement of spring pack displacement at the end of running load, torque switch trip, and final load is within OATIS published accuracy for all data points.
- 4) The data indicates improved accuracy from strain gage based systems compared to LVDT based systems measuring spring pack displacement, such as OATIS.
To address the timing difference, the ZNEL data was evaluated using an event based comparison rather than time based. The event based evaluation consists of comparing thrust values at defined signature events (end of running load, torque switch trip, and final load) that are extracted from the full stroke traces, rather than comparing thrusts at a given point in time. The event based evaluation is consistent with how OATIS is applied for conducting diagnostic testing in the field, and removes. the inaccuracy induced by inconsistent time computations between INEL and OATIS. This evaluation resulted in substantial improvement, with 81% (22 of 27) of the INEL data points at torque switch trip falling within OATIS published instrument accuracies.
It was also observed that OATZS,consistently predicted higher thrust than that from the INEL standard, based on limited statistical evaluation. This was true even after the timing differences were normalized using the signature event comparison. The root cause of this phenomenon has been investigated, but no conclusions have been reached yet. In order to apply the ZNEL data to our thrust calculations, it has been decided to shift the OATIS accuracies upward by 4% of the given thrust range values. When this correction is applied, all 27 INEL data points fall within the revised OATIS accuracies at torque switch trip. ABB Impell is independently reviewing the INEL data, and may restate their published instrument accuracies as part of their Nonconformance Report evaluation.
HOV CAPABILITY ASSESSMENT Cook Nuclear Plant has already completed a substantial amount of static diagnostic testing, and a limited amount of differential pressure testing of MOVs under the Generic Letter 89-10 program. This testing was based on thrust calculations which assigned target windows using the instrument accuracies published by ABB Impell for the OATZS ZIZ equipment. Additionally, NRC 1'nspection Report 50-315/91009(DRS);50-316/91009(DRS) caused us to reevaluate and revise our methodology for degraded voltage conditions and mispositioning.
Therefore, it became necessary to assess the impact these changes had on the valves we have tested under the Generic Letter 89-10 program.
Using revised degraded voltage conditions, differential pressures, and diagnostic equipment accuracies, new target thrust windows were produced for
1
)
Page 3 comparison to as-left values from diagnostic testing. In some cases, stem factor was also revised to reflect our review of the INEL test results and the refurbished condition of the actuators. This review showed that 93% (89 of 96) of the valves which have been tested under the Generic Letter 89-10 program have as-left torque switch trip settings which are greater than the minimum required for the new target thrust windows, and therefore are capable of performing their intended design function. Seven valves have as-left torque switch settings slightly below the new minimum target thrust. One of these seven valves was successfully tested at, differential pressure during the 1990 Unit 2 outage, and passed with substantial margin. This result demonstrates the inherent conservatism of our thrust calculation methodology. By comparing the differential pressure results to the other six valves with low torque switch setpoints, we have determined that they will be capable of performing their design function and there is no need for immediate corrective action to reset the torque switches. Four of the six valves which have not been differential pressure tested are 4" diameter and less gate or globe valves with low design basis differential pressures (180 psid or less). The two remaining valves (2-FMO-201 and 2-FMO-203) are 14" diameter gates with design basis differential pressures of 1310 psid. These valves have torque switch settings which are only 1% below the'ew minimum target thrust. Based on the high inertial load which was measured during static diagnostic testing, it is judged these valves will close and perform their intended design function.
CONCLUSIONS
- 1) Review of the MOVs at Cook Nuclear Plant which have been tested under the Generic Letter 89-10 program has determined they are capable of performing their design function aft'er consideration of revised degraded voltage conditions, differential pressures, and diagnostic equipment accuracies.
- 2) Future use of the OATIS III equipment, at Cook Nuclear on instrument accuracies which account for the INEL test Plant will be based results. Pending completion of AAB Impell's review, we plan on'applying the 4% of thrust range correction described herein.
- 3) We are evaluating the use of strain gage based diagnostic test equipment for the improved accuracy which was demonstrated during the INEL testing.
- 4) The torque switch settings for the six valves which are below the new minimum target thrust will be reset during the 1992 refueling outages.
- 5) AEPSC, Cook Nuclear Plant, and ABB Impell plan on continuing our review of the timing deviation and high thrust measurement bias in the OATIS III equipment, performing further review of the INEL data using the DaDisp software, and monitoring industry developments for further insights regarding diagnostic equipment accuracy and actuator repeatability.
ATTACHMENT 1 Suassary of vendor Claims tor Measure>>ant Accuracy Page 1 ot 2 (Noter The values listed on this sheet have not been validatod by the HUG)
Proparmzr Edd Sa Da ~ of azf gfnatfrxrr Test. Bquf)nmant Vsodoz / Systsar ABB I 11 cor ration ohTIs III Doeczfptfoa of test mmtbodr Stem thrust is measured dirsetl fn the o n direction into load cell. S rfn ck deflection ia r>>aaured in the closed dfreetion and used to correlate thrust. Thrust values measured uate to
~ sat thrust values reater than runnin load . Clan -on AC DC current robes measure motor current. Alii a attach to limit and tor e switches to monitor switch actuation.
ator'li Present Lccnzscy S pecif fcatfousr Parameter Accuracy Repoatability Drift Plaid Variables (PV)
Thrust Sonar h lt2g3p4 Thrust Sonar B Torque Stem Position Motor Current SEE ATTACHED S,C Motor Voltage Motor Power Tine 'Base z 0.0 csee/mfn Data Sample 1000 sls/c ih Sprfng Pack r0.01 inch Deflection Notssr 1: Accuracies are measured "end-to-end" (full loop accuracies).
2r "Total" accuracy speciffcation includes the "units" (fo. 4FS, %RDG, otc.).
3: Data sample zats units aro "sanpfeslsee/~ channels" Def iuitione r - A measure of the degree by which ths actual output of a device approxfnatsa the output of Accuracy an ideal device nominally performing tho sane function.
Repeatability - Tho closeness of agreer>>nt among repeated measurements of tho output for tho sane valuo of input made under tho samo operating conditions over a porfod of time.
Drift - Gradual occur'acy doviatfon in a gfvon timo period unrelated to input or envfronnont.
Plaid Variables - Conditions vhich may aff'ect oquipnont accuzacy and ai ~ not accounted for fn ths accuracy statonsnr ~ above (is. spring pack loading rats, non-standazd atom threads, actuator mounting position, improper bolt torque, erc).
PV Botoor 1:S rin ck conditioner fati us a loadin rats rsaas reload.
2rSton atom nut conditioner lubrication non-standard threads a ackin load stem ~
3rhetuator conditioner nountfn ~ ition sar bsarin voar lubrfeatfon runnin load bolt tor ~ tor e thrust conversion.
4rosnsral conditfonsr environment line ressure durin calibration o n calibration close thrust correlation.
SrMotor conditioner volta ~ fr sn Other Soteer A: Re tabilit and dritt varfablss tall vithin the stated total aecura value.
Brm li ibis error.
values rsasnt the value selected on the si nal condftionin unit. Tho ran s on ths current C:Current ro robe remains aer. at 200 an tor an motor current value from 0-200 am . Accura values aro based u n manutactursa claims for a ran e ssttfn ot 200 an s. Based on calibration results ths actual accur values or ~ reduced to a roxinatsl ths occurs value at the 200 an ran ~ throu hout ths entire ran e. i.o. Sin son s 4. 13% from 0-200 an Fluke ~ 2.24% from 0-200 am Rev 2/91
~J Page 2 of 2 Additional Data Sheet Space preparer: Edd Sa ed Date of origination: 2 19/91 Test gqui~t Vendor / Systaa:ABB In ll Co oration ctATIS III paraneter Accuracy 'epeatability Drift Field rr Variables (FV)
Thrust Range 4000 lbs z16 ~ 8 1 RHC 8000 lbs 49 F 4 1 RNC 14000 lbs z10.1 1 RNC 24000 lbs 48 ~ 2 1 RNC 45000 lbs 46 4 1RNC 70000 lbs ZSrl 1 RNC 140000 lbs S6 F 4 \ RNC 200000 lbs 46 F 2 '1 RNC Sinpson Current ran e 10 Z80 '1 RNC 20 440 1 RNC 50 z16 4 \ RNC 100 i8,06 1 RHC 200 z4, 13 1 RNC Fluke Curzent range r10 440 1 RNC 20 z20 \ RNC 50 zer06 1 RNC 100 44 12 \ RNC 200 S2 ~ 24 '1 RNG W Nateer 6c 7s 10'ther Woteez Ds F
Rsv 2/91
4 4<
ATTACHMENT 2 i
fi
3~2 ABB IMPELL 3.2.1 E UIPMENT DESCR PT ON The OATIS MOV test system is'designed around a portable
.Personal Computer (PC). Data is input by the user or collected by transducers and transferred to the PC through a signal conditioning unit known as the Back Pack. The PC is preconfigured with. the OATIS MOV software package and a special Analog to Digital (A/D) acquisition card. The A/D card is located in a expansion chassis mounted at the rear of the PC. It gathers data with a twelve-bit resolution and stores the data in the PC's internal memory.
The OATIS MOV Back Pack consists of five separate signal conditioning cards, an output card, an output display device, and a triple output power supply. The signal conditioning cards are designed to isolate the incoming signal from its corresponding transducer and minimize the level of background noise generated by the surrounding environment. Each card contains a self test circuitry to verify operation.
The five cards are:
Card Transducer parameter LVDT Linear displacement Spring Pack Disp.
Strain Gauge Compression Load Cell Stem Thrust Current Current Probe Motor Current Limit Switch Test Leads Control Switch Trip Torque Switch Test Leads Control Switch Trip The Output card produces high level analog signals that are compatible with the A/D card in the PC and allows manual or automatic selection of stroke direction.
The Output Display indicates LVDT zero position adjustment, Load Cell calibration check, Current probe adjustment and power supply output.
Stem thrust is measured directly in the open direction into a load cell. Spring pack. deflection is measured in the closed direction and used 'to correlate stem thrust.
Thrust values measured equate to seat thrust (values greater than running load). Clamp-on AC/DC current probes measure motor current. Alligator clips attach to the limit and torque switches to monitor switch actuation
d
3.2.2 DIFFICULTIES ABNORMALITIES
- a. Testing for the first series of nine tests was performed with the Limit Switch Auto Control Switch being manually manipulated for the Open and Closed direction.
The Limit Switch card sensitivity'was adjusted and the next eighteen tests were conducted with the switch in "auto.
- b. Zero time reference for the OATIS testing was taken at the opening of limit switch contact number four. Due to the sensitivity problems encountered contact switch four was not able to be monitored during th'e first nine tests.
- c. Current readings for the first nine tests were changed to zero for both the final open and close direction after the data was recorded.
- d. After the first nine tests the INEL spring pack transducer was bumped from a zero setting. The data for the next eighteen tests must be zeroed to obtain the actual reading.
r
't~
0
OATIS (IMPELL) TEST DATA S/N k5 OPEN TO CLOSE STROKE VENDOR NEL ORQUE INEL BAND EVIATION CCUM CCUM STROKE OKE TCH ESIRE0 NDOR NEL ROM ERCENT L RESS OAD ENSOR NUMBER UMBER NG EASURE MENT AD EAD NEL EVIATION INCH PSIG LASS PE A1 A1 1.25 T. S. TRIP 9890 13827 13682 13972 -3792 27.7 9.5 25 L LVDTI FINAL 17570 16122 15977 16267 1303 8.0 LOAD CELL 1.25 .S. TRIP 9820 11284 11139 11429 -1319 11.8 9.5 VDTI INAL 17750 15722 15577 15867 1883 11.9 OAD CELL 1.25 . S. TRIP 10130 15375 15230 15520 -5100 33.5 9.5 90 VDTI INAL 175?0 15331 15186 15476 13.5 OAD CELL A4 1.75 .S. TRIP 15150 16847 16702 16992 -1552 9.3 VDTI INAL 24130 - 20310 20165 20455 3675 18.0 OAD CELL 1.75 .S. TRIP 15760 13502 13357 13647 2113 155 VDTI INAL 24650 19744 19599 19889 4761 23.9 OAD CELL A6 1.75 . S. TRIP 15490 11914 11769 12059 3431 28.5 170 VDTI INAL 26390 11961 11816 12106 14284 118.0 OAD CELL A7 2.25 . S. TRIP 27510 20311 20166 20456 7054 34.5 0.75 50 VDTI INAL 39750 21154 21009 21299 18451 86.6 OAD CELL A8 2.25 . S. TRIP 28080 20317 20172 20462 7618 37.2 0.75 100 VDTI INAL 37350 21240 21095 21385 15965 74.7 OAD CELL 2.25 .S. TRIP 27950 22674 22529 22819 5131 22.5 0.625 190 VDTI INAL 28640 23087 22942 23232 5408 23.3 OAD CELL
OATIS (IMPELL) TEST DATA S/N ¹2 OPEN TO CLOSE STROKE VENDOR NEL ORQUE INEL BAND DEVIATION CCUM CCUM STROKE OKE TCH ESIRED NDOR NEL FROM ERCENT L RESS OAD NSOR NUMBER UMBER NG EASUREMENT EAD EAD NEL EVIATION INCH PSIG LASS E B1 2.25 . S. TRIP 25652 21935 21790 22080 3572 16.2 0.625 VDT/
INAL 30030 28352 28497 1533 5.4 OAD CELL B2 2.25 . S. TRIP 22921 3001 13.1 0.625 100 VDT/
INAL 30940 28996 5.6 OAD CELL 2.25 .S. TRIP 24241 24096 24386 1707 7.0 0.625 190 VDT/
INAL 26510 23958 23813 24103 2407 10.0 OAD CELL B4 4 1.75 .S. TRIP 15250 12299 12154 12444 2806 VDT/
INAL "
22940 19435 19290 19580 3360 17.2 OAD CELL B5 1.75 .S. TRIP 15959 11919 11774 12064 3895 32.3 5 100 VDT/
INAL 23010 19228 19083 19373 3637 18.8 OAD CELL B6 1.75 . S.TRIP 15347 12375 12230 12520 2S27 5 170 VDT/
INAL 23180 19453 19308 19598 3582 18.3 OAD CELL B7 1.25 .S. TRIP 9815 6151 6006 3519 55.9 9.5 25 VDT/
INAL 17720 14426 14281 14571 3149 21.6 OAD CELL BS 1.25 .S. TRIP 9815 5790 5645 5935 3880 65.4 9.5 YDT/
INAL 17720 13871 13726 14016 OAD CELL B9 1.25 .S. TRIP 9987 6801 6656 6946 3041 43.8 9.5 100 VDT/
INAL 17720 14238 14093 143S3 3337 23.2 OAD CELL
0 A
'I
OATIS (IMPELL) TEST DATA S/N 43 OPEN TO CLOSE STROKE VENDOR ORQUE INELBAND EVIATION GGUM CGUM STROKE ROKE TCH ESIRED ENDOR NEL ROM ERGEMI L RESS OAD NSOR NUMBER'EL UMBER NG EASUREMENT EAD EAD NEL EVIATION INCH SIG ITS E C1 1.25 .S.TRIP 5257 5112 5402 3532 65.4 9.5 25 VDTI INAL 14860 15229 15084 15374 -224 1.5 OAD CELL G2 1.25 . S. TRIP 4625 4480 4770 4430 92.9 9.5 VDTI 16110 14096 13951 14241 1869 OAD CELL 1.25 .S. TRIP= 7109 7399 1801 9.5 ~
100 VDT/
INAL 16330 7278 7133 7423 8907 120.0 OAD CELL 1.75 .S.TRIP 14030 10302 10157 10447 3583 50 VDTI INAI 20590 18765 18620 18910 1680 8.9 OAD CELL 1.75 . S. TRIP 14030 10051 9906 10196 3834 37.6 5 100 VDT/
INAL 20270 18690 18545 18835 OAD CELL 1.75 . S. TRIP 13410 11797 11652 11942 1468 12.3 5 170 VDTI INAL 20680 19580 19435 19725 955 4.8 OAD CELL Cj 2 .S.TRIP 17530 14582 14437 14727 2803 19.0 50 VDTI INAL 24110 22478 22333 22623 1487 6.6 OAD CELL 2 .S.TRIP 17920 14629 14484 14774 3146 21.3 1 100 VDTI INAL 24530 22861 22716 23006 . 1524 6.6 OAD CELL C9 2 .S.TRIP 18160 17693 17548 17838 322 1.8 170 VDT/
INAL 18480 . 17216 17071 17361 1119 6.4 OAD CELL
4 4 k 0
OATIS (IMPELL) TEST DATA AVERAGES DEVIATION{PERCEN)
LVDT/ LOAD CELL DATAUSED STROKES ATAPOINT MIN ALLSTROKES 27 STROKES . S. TRIP 1.8 92.9 27 STROKES INAL 1.5 120.0 S/N ¹'1 9 STROKES . S. TRIP 9.3 37.2 9 STROKES INAL 8.0 118.0 S/N ¹2 9 STROKES . S. TRIP 7.0 65.4 9 STROKES INAL 5.4 26.4 S/N ¹3 9 STROKES . S. TRIP 1.8 92.9 9 STROKES INAL 1.5 120.0 LOW LOADING 9 STROKES . S. TRIP 11.8 92.9 9 STROKES INAL 1.5 120.0 MED LOADING 9 STROKES . S. TRIP 9.3 37.6 9 STROKES INAL 4.8 118.0 HIGH LOADING 9 STROKES .S.TRIP . 1.8 37.2 9 STROKES INAL 5.4 86.6
ATTACHMENT 3
~ l" 1h 1 I 1'
~ ~
ABB IMPEL~
ABB Impe ll Corporation has reviewed the MOV Users Group (MUG)
Subcommx b
t's
'ttee ee on Test Equipment and Method "Progress Report on Equipment Valxdatzon
.Corpora son Valxdatxon re p ort ABB Impell currently performing a detailed evaluation of the data available from the testing of OATIS at comparing thee Preliminary Report with the detailed evaluation, two items have been identified /
which require consid eration.
4.2.1 ITEM 1 TIME DEVIATXONS The statement on Signal Conditioning Delays provided xn the Progress Report does identify the possibility that variations f t' t and may result in differences between vendor and XNEL thrust values. It also specifies that the evalua the impact on vendor to INEL comparisons is beyond the scope of the preliminary report. During the detailed evaluation, has been determined that a discrepancy in time does exist it between INEL and OATIS. The impact of the time discrepancy is quate sxgni'ficantican and was evaluated along with additional methods of comparison for applicability to test data.
Although there may be additional means of performing comparisons of o thee data as determined by the INEL standard and OATIS diagnostic equipment, the two most logical appear ear too bee time referenced and event referenced.
INEL Time Referenced Method The use of point to point comparison at an instant zn tame 0 removes the inaccuracy associated with transducer output interpretation (signature analysis). If the comparative time computations between standard and test were negligible, then this method would provide highly accurate comparisons of thrust. Deviation in the time calculations, which may affect dis esp layed a time'etween the standard and test, will have proportional effects on the thrust comparisons.
any thrust comparisons made in a time reference based method will include error induced by deviations in time from the
'ny standard.
d urging While this time discrepancy may be of importance comparisons or evaluations of time, its impac on thrust determination may be negligible and only induced by method of comparison.
Si nature Event Referenced Method The use of event comparisons removes the inaccuracy induced by inconsistent time computations.
associated with event evaluations.
It Asdoestechnicians induce error are responsible for the evaluation, of transducer waveforms, and the waveforms can vary during similar event occurrence, an
I/
i ~
inaccuracy associated with misinterpretation or inconsistent interpretation is induced. This. inaccuracy is variable but can be minimized through consistent and defined interpretation techniques.
Method Com arison While both of the comparison methods are valid and have merit, one must be chosen for continued use during evaluation of data. The time referenced method may incorporate error which is unrelated to thrust and cannot be controlled. The event method introduces error which may be variable and based on subjective evaluation. If the time references then time referenced and method presentations are compatible, a would yield the more accurate and consistent results. An evaluation of the time reference compatibility was performed to determine applicability of the INEL method.
Time Com arison Between INEL Standard and OATIS The method of comparison utilized by the MOV Users group, Test Equipment and Method Validation Committee, Test Plan 91-1 is based on time. Comparisons are made and evaluated through system time comparisons of event (Torque Switch Trip, Final Load) activities. To review the accuracy of this thrust comparison, an evaluation of variance between INEL monitored Torque Switch Trip times and OATIS Torque Switch Trip times, as normalized to the INEL time reference, was performed.
Evaluation of the compared data leads to the following conclusions:
5 The times deviated by a mean value of 0.049 seconds with a standard deviation of .058 seconds.
In 85.24 'of the test cases the OATIS I
5 normalized torque switch trip times were larger than the INEL torque switch trip times.
While the time evaluation performed ind'icates that there is some deviation between OATIS and the INEL standard; it was inconclusive whether this . is the result of overall discrepancies or deviations of the OATIS torque switch or limit switch indication values. If the discrepancy is due to misrepresentation of switch trip time or a time shift between transducer representations, then the resulting error would greatly impact the accuracy of thrust reported at switch trip.
However, if the discrepancy is an overall variance between OATIS and the standard, and all OATIS transducer responses are represented on a consistent time scale, then the impact will
t
~ I 0
be isolated to time referenced comparisons. All representations by OATIS will not be affected and only the representation of time may be in question.
To determine the cause of this apparent time discrepancy was necessary to evaluate additional data which can provide it isolation of time deviations. If the representation of switch indication were invalid, then it should be expected that comparisons of relative spring pack displacement at. torque switch trip would be discrepant. It was determined that the variations measured between INEL spring pack displacement at INEL torque switch trip and OATIS spring pack displacement at OATIS torque switch trip are all within reputed equipment accuracies. The lack of variation in measured displacement may not be representative of switch indication accuracy, however, either;
~ the INEL standard and OATIS are equal in time variance,
~ the OATIS measurement of spring pack displacement is out of tolerance and the time variance makes acceptable, or it appear
~ the variation lies in the computational deviations of time. To further investigate the cause of time variance, a comparison of overall stroke times was performed.
The result of stroke time comparisons indicate that stroke time deviation averaged .072 seconds and the time discrepancies are not limited to switch trip indication but are present throughout data display.
Conclusions On Method'Anal sis E
From this evaluation, the time discrepancy bounds the entire signature region. As this is the case the more appropriate method of thrust comparison would be signature event analysis.
While the use of an event base method will correctly ignore overall time discrepancies, and reflect any misrepresentations in switch trip indication, it will induce error associated with data interpretation. The impact of this error can be minimized through consistent analysis techniques and definitions.
4.2.2 ITEM 2 GENERIC APPLICABILITY While the results of this test should be considered as valuable and informative, they should not be considered as the basis for the establishment of revised tolerances. The MUG validation has been designed, for determining the
t II
conformance/nonconformance of results to a specific test and not to establish equipment tolerances for all ranges and applications. A small population sample. (one set of MOVLS specifications) and a limited number of cycles at any specific test condition impose finite limitations on evaluation of this data.
In addition to the Items noted, ABB Impell and OATIS users are continuing with detailed data evaluation to determine the most applicable analysis and applications of the INEL testing. The results of this analysis will be reconciled with MOV programs of OATIS users upon completion of data evaluation. ABB Impell does note that the limited data sample provides analysis restrictions in the areas of rate of loading and stem factor repeatability. Review of the wealth of data as collected by INEL throughout equipment validation, may prove insightful and provide the industry with a better understanding of these phenomena. The review of this data could be initiated by release of all the INEL MOVLS obtained data to public domain.
w I, ~ ~
I
INEL COMMENTS phd INEL's role in the MUG validation was one of service. We did not make policy 'or enforce it. All vendors chose to test at our facility and they accepted the risks knowingly. Policy and outcome for remarks should be reserved MUG.
Many comments concern event timing and measured values. Timing is qritical when evaluating measurements at some event during the stroke such as torque switch trip. Measurement channels with signal processing time delays or transducer response time variations may overestimate the torque and/or force available.
With timing delays, measured values of force and torque include momentum effects that may not respond the same at higher loadings.
The real question is how accurately a vendor can determine event timing with respect to his system's time response. Small errors in the timing scale itself are trivial, as long as the sequential spacing and the relationship of the event to time is maintained.
Several vendors questioned the INEL stem torque measurements. The INEL typically looks at the torque balance across the operator.
The spring pack force multiplied by the effective moment arm length gives the input torque. The difference between this and the measured stem torque is equal to the losses in the operator and the losses not accounted for in the calibrations (e.g. MOVLS lower thrust bearing). Figure 1 shows the INEL torque spring pack force measurement for a typical MOVLS closure stroke. Figure 2 shows the INEL stem torque measurement for the same stroke. By dividing the stem torque measurement. by the spring pack force measurement we can determine the apparent moment arm length. Figure 3 compares this apparent moment arm length with the theoretical length obtained from Limitorque. The two curves lie on top of one another indicating no significant losses between input and output torque.
Figure 4 shows the difference between input torque (measured spring pack force times moment arm length) and output torque (INEL stem torque measurement). This data shows that essentially there is no error due to losses in the operator or lower thrust bearing. This relationship remains constant from the lowest to the highest MOVLS loadings, convincing us that the MOVLS stem torque measurement methodology is sound.
0
-200
-400
-600
-800
-1,000 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 Time (s)
F(gure $ . Tbrquesprfng pack forca M371 kqd4791-01
V' I
0
0
-50 Xl 0)
D
-100
~O E
Q 40
-'t 50
-200 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 Time (s)
Flguro 2. Stem torque M371 kad4791-02
C r 4
~ ~
Ch Apparent Q
Theoretical E I iill (
l lit Q
E O
0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 Time (s)
Figure 3. Apparent vs theoretical moment arm length M371 kgd4791-03
4 C
V 0
II
0
-5
-10 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 Time (s)
Figure 4. input vs output torque error A@71 kRd4791-04
r, V
tl N