ML17305B636

Suppls 880921 Response to Action Item 3 of NRC Bulletin 88-004,consisting of Methodology for Determining Mechanical Seal Life.Analysis of Condensate Transfer Pump Test Data Does Not Indicate Trend Toward Pump Degradation

ML17305B636
Person / Time
Site:	Palo Verde
Issue date:	07/01/1991
From:	Conway W ARIZONA PUBLIC SERVICE CO. (FORMERLY ARIZONA NUCLEAR
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
161-04032-WFC-M, 161-4032-WFC-M, IEB-88-004, IEB-88-4, NUDOCS 9107120005
Download: ML17305B636 (18)

Text

REGULATORY INFORMATION DISTRIBUTION SYSTEM (RIDS)

ACCESSION NBR:9107120005 DOC.DATE: 91/07/01 NOTARIZED: NO 'OCKET FACIL:STN-50-528 Palo Verde Nuclear Station, Unit 1, Arizona Publi 05000528 STN-50-529 Palo Verde Nuclear Station, Unit 2, Arizona Publi 05000529 STN-50-530 Palo Verde Nuclear Station, Unit. 3, Arizona Publi 05000530 AUTH. NAME AUTHOR'FFXLIATION CONWAY,W.F. Arizona Public Service Co. (formerly Arizona Nuclear Power RECIP.NAME RECIPIENT AFFILIATION Document Control Branch (Document Control Desk)

SUBJECT:

Suppls 880921 response to Action Item 3 of NRC Bulletin D 88-004,consisting of methodology for determining mechanical seal life. Analysis of condensate transfer pump test data does not indicate trend-toward pump degradation. S DISTRIBUTION CODE: IE37D TITLE: Bulletin 88-004 re COPIES RECEIVED:LTR Potential Safety-Related J ENCL Q Loss Pump SIZE:

A NOTES:STANDARDIZED PLANT 05000528 Standardized plant. 05000529 D Standardized plant. 05000530 D

RECIPIENT "COPIES RECIPIENT COPIES ID CODE/NAME LTTR ENCL XD CODE/NAME LTTR ENCL PD5 LA 1 0 PD5 PD 1 1 TRAMMELL,C 1 1 THOMPSON,M 1 1 INTERNAL: AEOD/DOA 1 1 AEOD/DS P/TPAB 1 1 ALEXXON,T 13E21 1 1 MCCOY, M 8E23 1 1 NRR/DET/EMEB 7E 1 1 NRR/DOEA/OEABll 1 1 NRR/DOEA/OGCB11 1 1 NRR/DREP/PEPB9D 1 1 NRR/DS 8E2 " 1 1 NRR/PMAS/ILRB12 1 1 Rgg~F1Ti 2 1 1 RES/DSIR/EIB 1 1 1 1 EXTERNAL NRC PDR 1 1 NSIC 1 1 NOTES: 1 1 D

D D

NOTE TO ALL "RIDS" RECIPIENTS:

PLEASE HELP US TO REDUCE iVASTE! CONTACT THE DOCUMENT CONTROL DESK, ROOihl Pl-37 (EXT. 20079) TO ELIiVl!NATEYOUR NAME FROM DISTRIBUTION LISTS FOR DOCUMENTS YOU DON'T NEED!

TOTAL NUMBER OF COPIES REQUIRED: LTTR 20 ENCL 19

~ I Arizona Public Service Company P.O. BOX 53999 ~ PHOENIX, ARIZONA85072-3999 WILLIAMF. CONWAY 161-04032-WFC/MEP/JMQ EXECUTIVEVICE PRESIDENT NUCLEAR July 1, 1991 Docket Nos. STN 50-528/529/530 U. S. Nuclear Regulatory Commission Attn: Document Control Desk Hail Station Pl-37 Washington, D. C. 20555

References:

A) Letter from E. E. Van Brunt, Jr., APS, to NRC, 161-01159, dated July 8, 1988.

Subject:

Response to NRC Bulletin No.

88-04.

B) Letter from D. B. Karner, APS, to NRC, 161-01270, dated August 29, 1988.

Subject:

Additional Response to NRC Bulletin No.

88-04.

C) Letter from W. F. Conway, APS, to NRC, 161-02345, dated September 21, 1989.

Subject:

Supplemental Response to NRC Bulletin No. 88-04.

D) NRC Bulletin 88-04 from C. E. Rossi, NRC to All holders of operating licenses or construction permits for nuclear power reactors, dated Hay 05, 1988.

Subject:

Potential Safety-Related Pump Loss.

Dear Sirs:

Subj ect: Follow up to the Letter from W. F. Conway, APS, to NRC, 161-02345, dated September 21, 1988 - Reference C and Response to Action P3 of NRC Bulletin 88-04.

File: 91-056-026 References A, B, and C submitted partial responses to the subject bulletin.

Attachment 1 explains the change in methodology for determining mechanical seal life and is a follow up response to Reference C. Attachment, 2 includes a detailed response to Action P3 of NRC Bulletin 88-04. Action 43, of the Bulletin requested an evaluation of the adequacy of the minimum flow bypass lines for safety-related centrifugal pumps with respect to damage resulting from operation and testing in the'inimum flow mode.

This completes our response to the subject Bulletin.

9."L07l20LIO~ 5'l0701 F'DR AGGCf; 0 I.IOI.I 28 I.J PDR

161-04032-'k'PC/NEP/JllQ uly 1 , 1991 U. S . Nuclear Regulatory Commission Attn: Document Control Desk NRC Bulletin 8 8 - 04 Page Two If you should have any ques tions , please contact Michael E. Powell of my staff at ( 602 ) 340 - 49 8 1 .

S incerely ,

WFC/MEP/JMQ Attachment CC: J. B. Martin D. H. Coe D. Corporandy A. C. Gehr A. H. Gutterman

161-04032-WFC/HEP/JMQ uly 1, 1991 ATTACHMENT 1 Follow up to the Letter from W. F. Conway, APS, to NRC, 161-02345, dated September 21, 1988 Reference C Ingersoll-Rand verified the specified mini-flow rates for the PVNGS Safety Injection pumps,'n their evaluation of the potential damages resulting from extended run time on minimum-flow. The vendor's evaluation of the Safety Injection pumps, indicated that the specified miniflow rates are below the minimum continuous stable flow point; however, the vendor never intended and recommends against running the pumps on miniflow for exten'ded periods of time.

At PVNGS, administrative controls and operating procedures are such that the Safety Injection pumps are not operated for extended periods of time at miniflow in accordance with the vendor technical manual recommendations. The vendor further stated that the potentially adverse effects of running a pump on minimum flow for extended periods of time would primarily affect the mechanical seals.

Their resulting equation for calculating seal life was conservatively based on empirical data of vibration at low flow with the effects of starts and stops factored in by assuming a maximum of 500 starts and stops for useful life. The implication of their evaluation is that ensuring adequate mechanical seal life will minimize the effects of extended run time on miniflow.

As stated in Reference C, Ingersoll-Rand provided an equation by which the effect of pump operation on miniflow can be accounted for in determining when seals should be replaced. This equation was intended for the Safety Injection pumps and not the Condensate Transfer pumps, because the Condensate Transfer pumps do not have mechanical seals. Also, PVNGS intended to use the Emergency Response Facility Data Acquisition and Display System (ERFDADS) to log run time hours for input into the Ingersoll-Rand equation to compute seal life.'pon careful evaluation, the Ingersoll-Rand equation appeared to be overly conservative and not specifically applicable to PVNGS. APS also consulted'he mechanical seal vendor, Durametallic to develop an applicable seal life equation. Both the Durametallic and Ingersoll-Rand equations were used to develop a PVNGS specific mechanical seal preventative maintenance task frequency. The periodic seal replacement schedule is based on a) the vendor's equation (Ingersoll-Rand and Durametallic) as guidance, b) the review of operating experience c). the surveillance test data, d) maintenance histories, and e) design basis considerations.

Therefore, APS did not utilize a component evaluation system that would take the necessary data from the plant computer and ERFDADS data acquisition system to compute seal life.

161-04032-I}PC/I IEP/JN(}

uly 1, 1991 ATTACHMENT 1 CONT. ,

Preventative maintenance tasks are being developed which replace Low Pressure Safety Injection mechanical seals every other scheduled refueling outage, and replace High Pressure Safety Injection and Containment Spray mechanical seals every third scheduled refueling outage. This schedule will ensure that adequate seal liEe is available for both normal and post accident operating conditions.

The preventative maintenance frequencies are considered conservative with regard to seal life therefore, surveillance of actual run hours is not necessary, and the ERFDADS system would not be required.

Reference C also stated that until a trending system 'could be developed, a periodic routine taintenance task-would be instituted at the required frequency (based on -Ingersoll-Rand data) to ensure seal lifetimes are not exceeded. Based on the new methodology for determining seal liEe, a trending system was determined not to be necessary. The preventative maintenance tasks program will replace seals in a proactive manner. In addition, APS stated in Reference C, that the Engineering Evaluation Department would review each oE the affected pumps'aintenance history to determine when the seals were last replaced, and would issue work requests, as appropriate, to replace seals prior to exceeding their useEul liEetime. Surveillance test vibration data would also be reviewed for each pump to ensure that no increase in vibration has occurred. The initial review Eor the purpose of meeting the above commitment was completed on September 18, 1989. Subsequent reviews of maintenance histories, surveillance tests, and control room logs have been performed to ensure adequate mechanical seal life has been maintained. Work reouests have been written for those mechanical seals requiring replacement. Work requests will be written to replace the remaining seals as required,'ntil the Preventative Maintenance Program is in place.

I I 1 i

161-04032-WEC/MEP/JMQ uly. 1 , 199 1 ATTACHMENT 2 Response To Requested Action ¹3 Of NRC Bulletin 88-04 Action ¹3 of NRC 'Bulletin 88-'04 requested an evaluation of the adequacy of the minimum flow bypass lines for safety-related centrifugal pumps (i. the Safety Injection pumps and ii. the Condensate Transfer pumps) with respect to damage resulting from operation and testing in the minimum flow mode. The following items were included in the evaluation: Consideration of the effects of cumulative operating hours in the minimum flow mode a) over the life of the plant b) during postulated accident scenario involving the largest time spent in that mode.

i. a) At PVNGS, during a typical 18 month cycle of. full power operation, the High Pressure Safety Injection Pump, Low Pressure Safety Injection Pump, and Containment Spray are run on miniflow for surveillance testing, once per quarter.

Typically the duration is for less than one hour of continuous operation as recommended in the vendor technical manual.

i. b) In post Loss of Coolant Accident (LOCA) procedures the Safety Injection Actuation Signal (SIAS) actuated equipment is reset when the reset criteria is met, shutting off the non-injecting Safety Injection pumps, thus limiting operating time under minimum flow conditions.

ii.a) At PVNGS during a typical 18 month cycle of full power operation, the Condensate Transfer pump is run on miniflow for surveillance testing on a quarterly basis. Typically the duration is one to two hours, and the 100 gpm flowrate during testing is, sufficient to preclude any damage from low flow operations. The Condensate Transfer pumps are also required to provide make-up water to class cooling water systems and the spent fuel pool, when normal make-up water sources are not available. However, these instances are so rare as to make this mode of operation negligible. Note that during make-up operation, the passive miniflow recirculation line is in service. ii.b) The Condensate Transfer pumps are normally in the standby mode with both pumps starting, automatically with the receipt of a SIAS,. Loss of Offsite Power (LOP), Control Room Essential Filtration Actuation Signal (CREFAS), and a Control Room Ventilation Isolation Actuation Signal (CRVIAS).

I 161-04032-WPC/HEP/JW(} ul y 1 , 1 99 1 ATTACHMENT 2 CONT ~ The Condensate Transfer pumps provide makeup water to the Diesel Generator (DG) cooling water surge tank, Essential Chilled (EC) water expansion tank and the Essential Cooling Water (EVi') surge tank during postulated accident scenarios. The makeup water is provided to the DG, EC, and EW systems when the normal makeup to these systems is not available. The Condensate Transfer pumps can operate simultaneously with the Auxiliary Feedwater pumps during accident scenarios. The junction of the common minimum flow recirculation header for the Auxiliary Feedwater and Condensate Transfer pumps is close to the condensate storage tank. Consequently, the backpressure imposed by the Auxiliary Feedwater pump on the Condensate Transfer pump is negligible (less than 1 psi.). The evaluation should be based on best current estimates of potential pump damage of pumps involved - from test data and field experience. A review of Safety Injection pump surveillance test data did not indicate any trends toward degrading Safety Injection pump performance. The surveillance test data showed that vibrations on miniflow are well below the acceptable maximum vibration limit set by the vendor. The PVNGS Surveillance tests performed on the Safety Injection pumps monitor the performance of the pumps such that any trends toward a degraded performance are immediately identified. The mechanical seal vendor's evaluation showed that operating the pumps at high temperatures and high pressures accelerates seal face wear. A review of control room logs determined that during full power operation, the Safety Injection pumps are operated at low temperatures and low pressures under normal operating conditions. The Low Pressure Safety Injection pumps are the only Safety Injection pumps that experience high temperatures and pressures under normal operating conditions, when shutdown cooling is placed in service in mode 4 and prior to mode 3 entry during heatup. However, the amount of time they are exposed to those conditions is administratively limited. Analysis of Condensate Transfer pump surveillance test data does not indicate any trend toward degrading pump performance. The Surveillance test data shows vibration levels are well below acceptable vibration limits. The surveillance tests performed on the Condensate Transfer pumps monitor the performance of the pumps so that any trends toward degraded performance at design operating conditions are immediately identified.

t 161-04032-VFC/MEP/ JMg uly 1, 1991 ATTACHMENT 2 CONT. The Condensate Transfer "A" pump in Unit 1 has been disassembled. There are no signs of impeller low flow damage and the wear ring clearances are within tolerances. The Condensate Transfer "B" pump will be disassembled prior to the next scheduled Unit 1 refueling outage. The frequency of this inspection will be established after the "B" train disassembly. It should also include verification from the pump suppliers that current miniflow rates are sufficient to ensure no pump damage from low flow operation for short periods of time.

i. Ingersoll-Rand verified the Safety Injection mini flow rates are sufficient to ensure no pump damage from low flow operation for short periods of time.

Ingersoll-Rand provided an equation to determine mechanical seal life for the Condensate Transfer pump. Because the Condensate Transfer pump does not have mechanical seals, APS has confirmed with the vendor that the Ingersoll-Rand equation does not apply. Ingersoll-Rand has concurred that the specified minimum flow of 30 gpm is not an absolute value and that operation for sh'ort periods of time at lower flow rates would not lead to immediate pump damage. APS has considered the following information in'the evaluation of the minimum flowrates for the Condensate Transfer pumps: vendor supplied information calculated estimates for minimum flows low energy size of the pumps seal features of the pumps (packing as opposed to mechanical seals) low maintenance history low run time of the pumps surveillance testing of the pumps pump inspection All of these factors considered together have led APS to conclude that the theoretical analyses alone are not sufficient to determine if low flow damage has occurred to the pumps. Ingersoll-Rand has concurred that the only effective detection of low flow damage to the internals of the pump is from pump disassembly and inspection. The inspection was performed on the Unit 1 Condensate Transfer "A" pump, as stated in 2.ii., with no signs of impeller 'low flow damage and the wear ring clearances are within tolerances. The frequency of this inspection will be

I I

161-0!i032-'i<PC/HEP/JNQ uly 1, 1991 ATTACHMENT 2 CONT. established after the "B" train pump disassembly. The "B" train pump wi11 be disassembled prior to the next scheduled Unit 1 refueling outage.

l}}