ML17335A387

LER 98-031-01:on 980610,potential Common Mode Failure of RHR Pumps Were Noted.Caused by Inaccurate Values.Accurate Miniflow Numbers Have Been Determined by Flow Testing

ML17335A387
Person / Time
Site:	Cook
Issue date:	12/11/1998
From:	Boston D INDIANA MICHIGAN POWER CO.
To:
Shared Package
ML17335A386	List: ML17335A385 ML17335A387
References
LER-98-031, LER-98-31, NUDOCS 9812150146
Download: ML17335A387 (5)

Text

NRC Form 366 U.S. NUCLEAR REGUlATORY COMMISSION APPROVED BY OMB No. 3150%104 EXPIRES 06l30/2001 1998)

EsrsIATED IKEIOEN pER REspoNsE To cotsK.Y wiTH Ttes MAIIATORY 00'ORMATICN COLLECTION REOVEST: 50.0 tOtS. REPORTED LESSONS LEARNED ARE NCORPORATED elTO Tt& UENNSSIO PROCESS AND FED SACK TO SIVSTRY.

LlCENSEE EVENT REPORT (LER)

(See reverse for required number of RECORDS MANAGEMENT SRANCH IM COMMISSCN. WASISNCTOH. DC 005550001. AICI TO THE P~

FORWARD COMMENTS REOARDSIO SVRDEN ESTSIATE TO THE IIFORMATICN ANO AS). U.S. NUCLEAR RECIAATORY PROIECT o1500100, OFFCE OF MANACEMEN'r AIE5 SVDCET, WASteNOTOIL DC 20505 EDVCllON digits/characters for each block)

FACIUlYNAME (I) DOCKET NUMBER (2) PAGE (5)

Cook Nuclear Plant Unit 1 05000-315 1 of 3 llTLE(5)

Potential Common Mode Failure of Residual Heat Removal Pumps Due to Use of Inaccurate Values EVENT DATE (5) LER NUMBER (6) REPORT DATE (7) OTHER FACILITIES INVOLVED(8)

ILI V SEQUENTIAL REVISION MONTH DAY YEAR YEAR NUMBER NUMBER DAY Cook Unit 2 05000-316 YEAR 06 10 1998 1998 031 01 12 1998 OPERATING THIS REPORT IS SUBMITTED PURSUANT TO THE REQUIREMENTS OF 10 CFR g: (Check one or more) (11)

MODE (9) 20.2201 (b) 20.2203(a)(2)(v) 50.73(a)(2)(i) 50.73(a)(2)(viii)

POWER 20.2203(a)(1) 20.2203(a)(3)(i) 50.73(a)(2)(ii) 50.73(a)(2)(x)

LEVEL (10) 00 20.2203(a)(2)(I) 20.2203(a)(3)(ii) 50.73(a)(2)(iii) 73.71 20.2203(a)(2)(ii) 20.2203(a)(4) 50.73(a)(2)(iv) OTHER 20.2203(a)(2)(iii) 50.36(c)(1) 50.73(a)(2)(v)

Specify in Abetrect bekee 20.2203(a)(2)(iv) 50.36(c)(2) 50.73(a)(2)(vii) or DNRCFcrlnSSEA LICENSEE CONTACT FOR THIS LER (12)

NAME TELEPHONE NUMBER (Inc4dehree Code)

Mr. Dan Boston, Safety Related Mechanical Engineering Superintendent 616/465-5901, X1 863 COMPLETE ONE LINE FOR EACH COMPONENT FAILURE DESCRIBED IN THIS REPORT (13)

REPORTABLE REPORTABLE TO CAUSE SYSTEM COMPONENT MANUFACTURER TO EPIX CAUSE SYSTEM COMPONENT 'ANUFACTURER EPIX SUPPLEMENTAL REPORT EXPECTED (14 EXPECTED MONTH DAY YES SUBMISSION (If Yes, complete EXPECTED SUBMISSION DATE) NO DATE (15)

Abstract (Limit to 1400 spaces, I.e., approximately 15 single-spaced typewritten lines) (16)

On June 10, 1998, with both units in Mode 5, it was determined that the Residual Heat Removal (RHR) pump minimum flow (miniflow) controls for both units had a potential design deficiency. Westinghouse Nuclear Safety Advisory Letter 98-002 stated that during a LOCA of a size to allow the RHR/Low Head Safety Injection pumps to inject into the reactor coolant system (RCS) at less than required miniflow, the miniflow valves might cycle repeatedly from open to close until the valves or the valve motors failed. Available miniflow is a combination of accident mitigation flow and bypass flow through the miniflow valves. If the failed valves prevented adequate miniflow, the associated RHR pumps could fail. In accordance with 10CFR50.72(b)(2)(i), "Any event, found while the reactor is shut down, that, had it been found while the reactor was in operation, would have resulted in the nuclear power plant being in an unanalyzed condition that significantly compromises plant safety," and 1.0CFR50.72(b)(2)(iii), "Any condition that alone could have prevented the fulfillment of the safety function of a system needed to [m]itigate the consequences of an accident," an ENS notification was made at 1140 hours0.0132 days <br />0.317 hours <br />0.00188 weeks <br />4.3377e-4 months <br /> EDT. An interim LER was submitted in accordance with 10CFR50.73(a)(2)(ii) and 10CFR50.73(a)(2)(v).

The primary cause of this event was use of inaccurate miniflow numbers in calculating the valve control set points. It is not known how long or why inaccurate flow was used for the set point calculations. The values had not been verified by sting. Accurate minifiow values have been determined by flow testing. These numbers will be used in calculating set oints for new instruments that will be installed. Other systems were evaluated for similar concerns. Several programs ave been initiated or improved to identify or prevent similar concerns.

Overall evaluation of this low probability condition determined that the health and safety of the public were not endangered.

98i2i5014h 9812ii PDR ADQCK 050003i5 S PDR

RCFORM366A Us Nu L E 6-1998)

LICENSEE EVENT REPORT (LER)

TEXT CONTINUATION FACILITYNAME (1) DOCKET NUMBER(2) LER NUMBER (6) PAGE (3)

YEAR SEQUENTIAL REVISION Cook Nuclear Plant Unit 1 05000-315 NUMBER NUMBER 2 of 3 1998 031 01 TEXT (Ifmore spece is required, use edditionel copies of NRC Form (366A) (17)

Conditions Prior to Event Unit 1 was in Mode 5, Cold Shutdown Unit 2 was in Mode 5, Cold Shutdown Descri tlon of Event On June 10, 1998, during a review of Westinghouse Nuclear Safety Advisory Letter (NSAL)98-002, engineers determined that the Residual Heat Removal (RHR) pump minimum flow (miniflow) controls for both units had a potential design deficiency. The NSAL stated that during a LOCA of a size to allow the RHR/Low Head Safety Injection pumps to inject into the reactor coolant system (RCS) at less than required miniflow, the miniflow valves might cycle repeatedly from open to close until the valves or the valve motors failed. Available miniflow is a combination of accident mitigation flow and bypass flow through the miniflow valves. If the failed valves prevented adequate miniflow, the associated RHR pumps could fail.

An earlier condition report (CR) investigation had determined that the flow measurement instrumentation that controls the RHR pump miniflow valves would need to be replaced by instruments with different design characteristics. This work had begun in October 1996 when a CR was written because the Unit 2 East RHR Pump failed a post maintenance test. The CR investigation determined that the flow measurement instrumentation that controls the RHR pump miniflow valves would eed to be replaced by instruments with different design characteristics. New instruments were being procured when SAL-98-002 was received. Review of NSAL-98-002 prompted instrumentation and controls (l&C) engineers working on he instrument replacement to focus on potential valve cycling problems. To prevent valve cycling, it was necessary to have an accurate value for the flow through the miniflow line to property set the open and close setpoints. Ultrasonic flow measurement equipment was used to determine that actual miniflow was approximately 508 gallons per minute (gpm) for Unit 1 and 535 gpm for Unit 2. Once miniflow was known, it was possible to review historical miniflow instrument calibration data and determine if cycling could have occurred in the past. The review showed that the open and close set points, with flow instrumentation calibrated to the historical standards, did not have enough separation to prevent cycling, given the accident scenario presented in NSAL-98-002. The typical open setpoint was about 455 gpm and the typical close setpoint was about 939 gpm. Calibration records showed that with instrument drift and uncertainty, there were periods when the set points did have enough separation to prevent cycling.

Cause of Event The primary cause of this event was use of inaccurate miniflow numbers in calculating the valve control set points.

Determination of the proper control set points depends on accurate knowledge of full flow in the miniflow lines. The actual flows are approximately 508 gpm for Unit 1 and approximately 535 gpm for Unit 2, as determined by recent ultrasonic flow meter testing. The value used historically, which had not been verified by testing, was approximately 463 gpm. It is not known how Iong or why inaccurate flow was used for the set point calculations.

Anal sls of Event This condition was determined to be reportable in accordance with 10CFR50.72(b)(2)(i), as a condition which was found while the reactor is shut down, which if it had been found while the reactor was operating, would have resulted in the nuclear power plant being in an unanalyzed condition and with 10CFR50.72(b)(2)(iii), as a condition that alone could have prevented the fulfillment of the safety function of a system needed to mtigate the consequences of an accident. An ENS notification was made on June 10, 1998, at 1140 hours0.0132 days <br />0.317 hours <br />0.00188 weeks <br />4.3377e-4 months <br /> EDT on June 10, 1998. An interim LER was submitted on July 10, 1998, in accordance with 10CFR50.73(a)(2)(ii) and 10CFR50.73(a)(2)(v). This LER is submitted as an update.

The safety function of the RHR system is to provide low head emergency core cooling flow during a LOCA. RHR injection ay be precluded during a small break LOCA. In such a situation, the miniflow controls play an important role in pump rotection by regulating flow through the miniflow lines.

NRC FORM 366A (6-1 998)

NRC FORM 366A U.S. NUCLEAR REGULATORY COMMISSION (6-1998)

LICENSEE EVENT REPORT (LER)

TEXT CONTINUATION V

'ACILITYNAME (1) DOCKET NUMBER(2) LER NUMBER (6) PAGE (3)

YEAR SEQUENTIAL REVISION Cook Nuclear Plant Unit 1 05000-315 NUMBER NUMBER 3of3 1998 031 01 TEXT (Ifmme speceis mquinrd, use edditionel copies of NRC Form (366A) (17)

Analysis of Event (continued)

The event had a low probability of occurrence because multiple conditions would have had to occur in specific sequences to have caused a common-mode failure of the RHR pumps. Cycling could have occurred only if the flow through an RHR train was within a narrow range of values. The approximate flow through a train would have had to have been between 390 and 470 gpm. This flow would have only occurred while the RHR pumps were discharging to the suctions of the safety injection (Sl) and centrifugal charging pumps while the RCS was at a relatively high pressure. Even if the flow would have been within the range, the systems were not always susceptible to cycling. Flow instrument drift caused the actual differential between the open and close set points to vary. If cycling had occurred, the valve would have had to have failed closed to deprive the pumps of minNow. Even with the valve fully closed, flow would have been at least 390 gpm.

Westinghouse had informed AEP that the miniflow requirement for similar pumps at another nuclear power plant was approximately 330 gpm. Although this cannot be directly applied to D. C. Cook, it is reasonable to believe that an RHR pump can su'Ivive at flows less than the 500 gpm given in the vendor manual.

Finally, there is no reason to believe that cycling would have caused both valves to fail at the same time. The failure of one valve and pump would have allowed RCS pressure to decrease as input flow was reduced. This would have caused the other pump's flow to increase beyond the range where cycling would have occurred. The flow Into the RCS would have ad to decrease back to the cycling range before the other valve and pump could have failed. The time between the two vents would have given the operators time to take corrective actions. The combined effect of the above conditions was to reduce the probability of a common-mode failure.

Overall evaluation of the condition determined that the health and safety of the public were not endangered.

Corrective Actions Accurate miniflow numbers have been determined by flow testing. These numbers will be used in calculating set points for new flow control instruments that will be installed. The calculation, ECP-12-I3-01, has not yet been completed, however the methodology is complete and is not expected to change. The calculation will serve as the record for how and why the set points were established.

During a review of other systems, engineers determined that the centrifugal charging pumps and Sl pumps might be subject to similar conditions. Evaluation of the Sl pumps determined that they were not susceptible to the same failure mechanism because there is no automatic control scheme. Evaluation of the charging pumps determined that the associated miniflow system was not as tightly coupled as the RHR miniflow control system. The potential for cycling had been considered during preparation of Calculation ENSM 971023CV, which had established the charging pump miniflow control set points. The calculation basis will be maintained through the new calculation procedure, 800000-LTG-5400.02 "Calculations".

Control and documentation of changes to plant instrument set points have been improved and are controlled by procedure PMP,6065.ISP.001, "Plant Instrument Set Point Control Program."

The operating experience review program, system readiness reviews, restart walkdowns, the calculation verification program, and the set point control and instrumentation uncertainty review, will provide additional assurance that issues similar to the miniflow valve cycling issue are corrected or prevented.

revlous Similar Events one NRC FORM 366A (6-1998)

vs~

0