ML20235Q712

Safety Evaluation Re Review of WCAP-10271,Suppl 2 & WCAP-10271,Suppl 2,Rev 1 on Evaluation of Surveillance Frequencies & out-of-svc Times for ESFAS

ML20235Q712
Person / Time
Issue date:	02/22/1989
From:	Office of Nuclear Reactor Regulation
To:
Shared Package
ML20235Q711	List: ML20235Q709 ML20235Q712
References
NUDOCS 8903030005
Download: ML20235Q712 (6)

Text

ENCLOSURE 1 SAFETY EVALUATION BY THE OFFICE OF UUCLEAR REACTOR REGULATION REVIlFDTRI~STINGHOUSE REPDRTlClP-10271 SUPPLERENT 2 AND WCATID'271 SUPPLEMEhT2 ~Plil310N 1 OTITIA[DAtlDhlF SUhVETIEANCE FREQUEffCT13_] ADD ~DDT OF SER ENGINEERED SAF Q FEATURES ACTUA, TION SVSTg i

1.0

SUMMARY

The staff has reviewed the Westinghouse Topical Reports WCAP-10271, Supplement 2 and WCAP-10271 Supplement 2, Revision 1 " Evaluation of Surveillance Frequencies l and Out of Service Times for the Engineered Safety Features Actuation System",

supported by the Westinghouse Owners Group (WOG) for purposes of proposing extensions of surveillance test intervals (STIs) and test and maintenance allowed outage times (A0Ts) for the Engineered Safety Features Actuation System (ESFAS). {

Specifically, bases were provided for increasing the STI for the analog channels l from 1 nionth to 3 months; no STI changes were requested for the combinational logic, or the master or slave relays.

It was also proposed that 1) the A0Ts for test for the analog channels be increased from 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> to 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> for both solid state and relay systems, 2) the A0Ts for test for all components be increased to 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> in solid state systems, and

3) in relay systems, the A0Ts for test for the logic trains and master relays  ;

be increased to 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> and for the slave relays to 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br />. j Additionally, it was reouested that the A0T for maintenance for all components be extended to 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> for both relay and solid state systems. All components except the analog channels would be in bypass during the maintenance A0T, with an analog channel tripped after spending 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> in bypass.

Finally, it was reccested that a staggered test requirement not be implemented for analog channels in the ESFAS and that this requirement be removed for analog channe h in the Reactor Protection System (RPS) (Ref. 1), many of which are common with ESfAS channels.

The staff has concluded that the analyses presented in WCAP-10271 Supplement 2 and WCAP-10271 Supplement 2, Revision 1, augmented by a Brookhaven National Laboratory (BNL) technical evaluation report (TER) are acceptable for resolving the STI and A0T extension issues, subject to any limitations and conditions presented  ;

herein. '

Additionally the staff concludes that a staggered test strategy is no longer reouired for RPS analog channel testing, as originally stipulated in Ref. 1.

@ OpOFRPocos a90222 EMVWEST C: . ,

PDC '

l i

2.0 BACKGROUND

Item 4.5.3 of Generic Letter 83-28 (Ref. 2) requested that all licensees and applicants review the existing RPS on-line functional test intervals required by Technical Specifications (TS). They are to ensure that current and proposed intervals (Ref. 1) for such testing are consistent with a goal of achieving high RPS availability. Extensions to RPS STIs have been granted for Westinghouse PWR plants.

The ESFAS shares some common instrumentation with the RPS. On the average, the number of ESFAS analog channels sensing either process or nuclear parameters is

58. with 20 channels dedicated to the ESFAS and 38 channels common between ESFAS and RPS. It is therefore worthwhile from an operational viewpoint to consicer extensions of STIs for all ESFAS analog channels. Additionally, plant operational effectiveness is entanced by considering STI extensions for the ESFAS logic l cabinets and master and slave relays. At the same time, consideration of extension of test and maintenance A0Ts will allow more effective test ano maintenance operations. This will reduce human error rates in these activities and the number of inadvertent actuations of engineered safety features.

3.0 APPP0ACH l

The Westinghouse Owners Group (WOG) approached resolution of this issue generically. The unavailabilities of the ESFAS signals were calculated by i Westinghouse /WOG (Ref.'s 3 and 4) for both relay and solid state systems. The i analyses show that the unavailabilities of the relay and solid state ESFAS signals are of similar magnitude.

The WOG originally evaluated the impact of the proposeo STI and A0T changes on core damage f requency (CDF) and public health risk on the Hillstone Unit 3 piant.

This plant has a solid state ESFAS with 2-out-of-4 (2/4) logic. The staff and its contractor. Brookhaven National Laboratory (BNL), had a concern that Millstone Unit 3 might not fully bound the change in CDF due to the proposed STI and A0T changes for all Westinghouse plants. Some plants have either a 2-out-of-3 (2/S) logic or a combination which may have higher unavailability than that associated I with a 2/4 logic such as at Millstone Unit 3. In response to this concern, l Westinghouse performed an analysis, documented as WCAP-10271 Supplement 2.

Revision 1. Aodendum 2 to determine the effect on the change in the Millstone

3 CDF of an assumed change of the ESFAS logic from 2/4 to 2/3. This resulted in l

a CDF increase for the 2/3 logic over the 2/4 logic of less than 1 percent of the base case CDF for the solid state system. The staff concludes that the relay plants would exhibit similar relative CDF changes with respect to the impact of 2/3 vs. 2/4 logic.

4.0 NRC ACTION The staff engaged the services of Brookhaven National Laboratory (BNL) to review the approach useo and the analyses performed in the Westinghouse reports. This

l review was performed to determine the adequacy of the methods used to establish the technical buses for the proposed modifications of STis and A0Ts for the Westinghouse PWR ESFAS instrumentation and actuation relays.

The BNL review calculations yielded, for the proposed ESFAS STl/A0T changes, a CDF increase of 2.8% for solid state plants, which is in good agreement with the 2.4% increase calculated by the WOG.

BNL performed a variety of parametric CDF increase calculations. Among the results was a relay plant CDF increase of 4% assuming concurrent slave relay testing. Another BNL sensitivity study yielded a CDF increase of 5.7%,

assuming a very conservative sequential testing scheme which is not used in practice.

BNL also determined that use of Millstor6 Unit 3 as a reference plant may not fully bound the change in CDF aue tu the proposed STl/A0T changes because of its 2/4 ESFAS logic, which yielded the 2.4% CDF increase. The 2/3 ESFAS logic WOG analysis, oiscussed earlier, yielded a 3.3% CDF increase.

lhe staff concludes, therefore that an overall upper bound for the CDF increase due to the proposed STl/A01 c'unges is less than 6% for Westinghouse PWR plants. )

The staf f also concludes that actual CDF increases for incividual plants are expected to be substantially less than 6%. The staff considers this CDF increase to be small compared to the range of uncertainty in the CDF analyses and therefore acceptable.

Based on the Westinghouse /WOG analyses ano the BNL audit and sensitivity '

i analyses, the staf f concludes that the proposed S11 and A0T changes for the ESFAS would have only a small and therefore acceptable impact on plant risk. BNL issued a technical evaluation report (Enclosure to this Safety Evaluation) presenting the details and results of its reviews.

~

Additionally the staf f concludes that a staggered test strategy need not be implemented for ESFAS analog channel testing and is no longer required for RPS analog channel testing, as originally stipulated in Ref. 1. This is based on the small relative contribution of the analog channels to RPS/ESFAS unavail- 4 ability, process parameter signal diversity, and normal operational test spacing.

5.0 CONCLUSION

S Based on a review of the BNL technical evaluation report (TER), the staff concludes that a 6% CDF increase due to the proposed STI/A0T extensions represents an upper bound. For realistic testing strategies, the CCF increase will be substantially less than this. The staff therefore concludes that the analyses presented in WCAP-10271 Supplement 2 and WCAP-10271 Supplement 2, Revision 1, augmented by the TER, form an acceptable basis for increasing the STI for ESFAS analog channels from 1 month to 3 months.

Additionally, the staff finds that 1) A0Ts for test for the analog channels may be increased from 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> to 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> for both solid state and relay systems, 2) the A0Ts for test for all components may be increased to 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />

l in solid state systems, 3) The A0Ts for test for the logic trains and master relays may be increased to 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> and the A0T for the slave relays to 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> in relay systems, and 4) the A0T for maintenance for all components may be extended to 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> for both relay and solid state systems. Additionally, ,

all components except the analog channels are to be in bypass during the main- I tenance A0T, with an analog channel tripped after spending 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> in bypass.

Further, the staff will not require a staggered test strategy for ESFAS analog '

channel testing, and will no longer require a staggered test strategy for RPS analog channel testing, as stipulated in the staff SER of February 21, 1985 (Ref. 1). The removal of the staggereo test requirement is based on the small relative contribution of the analog channels to RPS/ESFAS unavailability, process parameter signal diversity, and normal operational test spacing, which is neither staggered nor sequential, but yields some of the benefits of staggereo testing.

Table 1 lists plant-specific conditions that each licensee or applicant must meet to make any proposed STI or A0T changes fully acceptable. Table 2 sumarizes the approved changes.

6.0 REFERENCES

1. Safety Evaluation by the Office of Nuclear Reactor Regulation WCAP-10271,

" Evaluation of Surveillance Frequencies and Out of Service Times for the .

Reactor Protection Instrumentation System," February 21, 1985.

2. Eisenhut, D. G., NRC Letter to All Licensees of Operating Reactors, Applicants for Operating License, and Holders of Construction Permits,

" Requested Actions Based on Generic Implications of Salem ATWS Events,"

July 8, 1983.

3. Andre, G. R., Howard, R. C., Jansen, R. L., and Leonelli, K., " Evaluation of Surveillance Frequencies and Out of Service Times for the Engineered Safety Features Actuation System," WCAP-10271, Supplement 2, February 1986.

}

4 .. Andre, G. R. , Howard, R. C. , dansen, R. L. , and Leonelli, K. , " Evaluation of Surveillance Frequencies and Out of Service Times for the Engineered Safety Features Actuation System," WCAP-10271, Supplement 2, Revision 1, March 1987, i

l l

l l

TABLE 1 .

CCNDITIONS TO CLOSE OUT PLANTS 4 For plant-specific application of the TS changes for the Engineered Safety Features Actuatior System (ESFAS) that are proposed, the licensee must:

(1) Confirm the applicability of the generic analyses to the plant.

,( 2) Confirm that any increase in instrument drift due to the extended STis is properly accounted for in the setpoint' calculation methodology. (For additional information on this issue, see letter from C. E. Rossi to R. F. Janecek. dated April 27. 1988.)

l l

TABLE 2 APPROVED SURVEILLANCE TEST INTERVAL (STI) AND ALLLOWED OUTAGE TIME U,01) CHANGES RELAY SYSTEM SOLIO STATE SYSTEM LOGIC CABINETS STI 1 mo - 1 mo 2 mo---2 mo TEST A0T 3 hr-8 hr 1.5 hr--4 hr MAINTENANCE A0T 2 hr -12 hr 2 hr -12 hr MASTER RELAYS STI 1 mo--I mo 2 mo---2 mo TEST A0T 3 br-8 hr 1. 5 h r--4 h r ,

MklNTENANCE A0T 6 hr--12 hr 2 hr-12 hr SLAVE RELAYS STI 3 mo - 3 mo 3 mo- 3 mo TEST A0T 6 hr-12 hr 4 hr-4 hr MAINTENANCE A0T 6 hr -12 hr 2 hr -12 hr ANALOG CHANNELS STI 1 mo-- 3 mo 1 mo - 3 mo TEST A0T 2 h r --~ 4 h r 2 hr-~4 hr MAINTENANCE A0T 1 hr-12 hr 1 hr -12 hr

-_ - - _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _