ML20249C209

Forwards RAI to Complete Review of Licensee Request Re CENPSD-1041 & CENPSD-1045

ML20249C209
Person / Time
Issue date:	06/23/1998
From:	Stewart Magruder NRC (Affiliation Not Assigned)
To:	Pilmer D C-E OPERATING PLANTS OWNERS GROUP
References
PROJECT-692 NUDOCS 9806260191
Download: ML20249C209 (11)

Text

,

,N n

a

~ June 23, 1998 w

Mr. David F. Pilmer, Chairman

- CE Owners Group -

San Onofre Nuclear Generating Station

' 14300 Mesa Road '

San Clemente, CA 92672

SUBJECT:

. REQUESTS FOR ADDITIONAL INFORMATION REGARDING CENPSD-1041 -

AND CENPSD-1045

Dear Mr. Pilmer:

In a letter dated March 11,' 1998, the Combustion Engineering Owners Group (CEOG).

submitted CENPSD-1041, " Joint Applications Report for High Pressure Safety injection System -

Technical Specification Modifications," for staff review, in a letter dated April 6,1998, the CEOG also submitted CENPSD-1045, " Joint Applications Report, Modifications to the

. Containment Spray _and Low Pressure Safety injection System Technical Specifications," for staff review. Enclosed are requests for additionalinformation on these submittals. These i requests identify items of concern that must be resolved for the staff to complete its review.'

If you have any questions regarding this matter, you can contact me at 301-415-3139.

Sincerely, Original Signed By:

Stewart L. Magruder, Project Manager Generic lasues and Environmental Projects Branch Division of Reactor Program Management Office of Nuclear Reactor Regulation Project No. 692

Enclosures:

As stated l

Distribution:

CentruiFiles PUBLIC PGEB r/f JRoe TEssig -

FAkstulewicz

/

p

. SMagruder.

MWohl -

@fy $f L

MReinhart.

Document Name:g:\\simi\\ceogpsa.ral

[I

~ ~ ~, '

(2.

SC:PNDR [

OFFICE:

PM:PGES:DRPM SPSB:DSSA :

NAME-SMagruderkw$ MW M FAksih" M 'J( b 7 (p Q C

DATE 6/aF98 6Atf98 kJ/98 J.,si 0[8@ R$@@j$gggp' b.

m"#U e=, '"il 1

h c

s

[

I h$

[

d

p2ctw, pe d

UNITED STA'ES g

j NUCLEAR REGULATORY COMMISSION

'2 WASHINGTON, D.C. 20666 4 001

%,=,***4 June 23, 1998 0

Mr. David F. Pilmer, Chairman CE Owners Group San Onofre Nuclear Generating Station 14000 Mesa Road San Clemente, CA 92672

SUBJECT:

REQUESTS FOR ADDITIONAL INFORMATION REGARDING CENPSD-1041 AND CENPSD-1045

Dear Mr. Pilmer:

In a letter dated March 11,1998, the Combustion Engineering Owners Group (CEOG) submitted CENPSD-1041, " Joint Applications Report for High Pressure Safety injection System Technical Specification Modifications," for staff review. In a letter dated April 6,1998, the CEOG also submitted CENPSD-1045, " Joint Applications Report, Modifications to the Containment Spray and Low Pressure Safety injection System Technical Specifications," for staff review. Enclosed are requests for additionalinformation on these submittals. These requests identify items of concem that must be resolved for the staff to complete its review.

If you have any questions regarding this matter, you can contact me at 301-415-3139.

Sincerely, M %.

d Stewart L. Magruder, roject Manager Generic lasues and Environmental Projects Branch Division of Reactor Program Management Office of Nuclear Reactor Regulation Project No. 092

Enclosures:

As stated

RAls FOR CENPSD-1041 i

l l

l

1. On page 2,you state:

l "In implementing the ~~ dad AOT, the CEOG member utilities will, prior to entermg these l

modified Action statements associated with the extended AOT, use a proceduralized PRA-informed evaluation to manage the risk=====ud with the applicable plant configuration."

Are you implying that you perform PRA evaluation of the e=%nration only upon entry to the extended portion of the AOT?

i If so, please provide your reasomng for decomposing the new proposed AOT into two windows: Old AOT and Extended AOT.

Please provide your reasoning for not requiring risk evaluation upon entry into the LCO. Please provide your response in relation to the requirements of the three-tiered approach Aar===ad in draft SRP Chapter 16.1.

2. In addition to the proposed changes to AOT for at powennode of operation, thejoint application also proposes modifications to the ECCS action statements for lower mode operation. The joint application argues that by allowmg extended operation in these modes with inoperable HPSI trams, the risk associated with unscheduled mode-to-mode transitions is minieni-d (page 2.) In Section 6.3.3 of the submittal, thejoint applicaten claims that the risk of transitioning the plant from "at power" to lower modes of operation is relatively high (i.e., comparable to the risk of remaining at power.) 'Ihe basis of this claim comes from the transition risk models developed by CEOG, which assume unavailability of only one HPSI train (Tables 6.3.3-1 and 6.3.3-2.)

However, one of the proposed modifications to the TS for HPSI in lower modes is applicable to LCO configurations in which two HPSI trams are unavailable. How can you justify relaxing the TS requirements for HPSI in lower mode operations (e.g., mode 3) in light of the risk results presented in Section 6.3.3 of thejoint application, which claims relatively high r;-k with only one HPSI train available?

3. According to Table 5.1-1, Palisades has two HPSI pumps. The "Y" entry in the column " Spare HPSI Pump Powered by Either EDG" implies there is (was) a third HPSI pump Under " Comment" heading, it is stated that the third HPSI pump is converted to a spare AFW pump. Please clarify this apparent discrepancy.

4. 'Ihe modeling of hot-leg injection in CE plant's IPEs is not consistent. Some IPEs did not analyze hot-leg injection in the LOCA analysis and some discounted the failure of bot-leg injection as non-threatening. Please address the impact of this modeling uncertamty on the AOT risk resuhs.

1

5. In section 6.2.1 of thejoint application for the LPSI (Page 12), you referenced the following resuhs l

based on deteHM analyses :

l The function of LPSI is not required for mitigation oflarge LOCAs; and e

e' The operation of one HPSI and a Sfr can mamtam the Appendix K cntena during a design basis large LOCA scenano In Section 6.2.1 of the HPSIjoint application, you represent LPSI as the backup to HPSI to provide i

inventory control and long term heat removal it appears that the HPSIjoint application gives more June 8,1998 1

h weight to the role of LPSI in LOCA mitigation than the LPSI application. Please comment on this observation.

6. 'Ihe IPEs for CE plants generally did not take credit for using LPSI in mitigation of SGTR events involving the loss of HPSI. Please specify whether this is a convenient modeling assumption or ifit is based on an engineerms analysis. Have you performed any analyses to establish the effectiveness of the strategy to rapidly depressurize the RCS via the secondary and use LPSI instead of HPSI to control reactor coolant inventory for mitigation of SGTR events?

7. In Section 6.3.2.4 you state:

~

"In generat, for purposes of the PSA evaluation, it was conservatively assumed that the spare pump was unavailable. Thus, only one sub train with a single HPSI pump available was assumed when calculating the conditional CDF."

In your comparative study you further state that Waterford 3 and Fort Calhoun took credit for the spare pumps for the AOT risk calculation. Please respond to the followmg questions:

Which CE plant took credit for the spare HPSI pump in its IPE?

Is the treatment of the spare HPSI pump the same in the IPE and the AOT risk analysis?

8. Based on the results of the AOT risk and the design of HPSI system, Table 6.3.2-4 of the joint application groups the CE plants into three categories. For categories in which the AOT risk exceeds 1.0E-6, the application provides the following comments (Page 36:)

A long PM that disables a HPSI sub-train is not yd; and Entry into LCO action statement for HPSI pump repair is expected to be an infrequent occurrence.

Are you implying that additional procedural controls (licensee's commitments) will be in place to limit the duration of the outage and/or the frequency of entenng into LCO for plants whose single AOT risk exceeds 1.0E-57 Please elaborate further.

9. In the joint application you define Change Factor (CF) to be the ratio of the conditional C9F when the selected equipment is removed from service to the baseline CDF. The calculated CFs associated with removing a HPSI train for preventive maintenance ranges from 2 to 7. The average CF for a HPSI train is about 4.5. This value is higher almost by a factor of 4 relative to CF as@*ad with removing a LPSI train for PM (the CF for LPSI is very close to unity.) The high CF for HPSI indicates the importance of HPSI in accident mitigation From the risk point of view, the AOT for HPSI should be lower than that of LPSI. Please comment on why you are proposing the same AOT for both systems

10. The current staff position as outlined in the DraA Regulatory Guide (DG-1061) is that while implementing risk-informed decision making, the proposed change should meet several key principles includag.

Defense-in-depth is maimminad Sufficient safety margins are maimmined

• - Over-reliance on programmatic activities should be avoided.

June 8,1998 2

Please explain how these principles are met, especially for plants that do not have a spare HPSI pump.

11. In Section 63.5,you present an " Assessment of Early Large Release." While some of the qualitative arWs appear reasonable, namely that there will be little or no increase in the LERF due to taking a HPSI train out for PM or CM, these justifications need to be judged agamst LERF values, consistent with the (draft) Regulatory Guide 1.174 (DG-1061). Please provide LERF values or arguments that they are not needed, consistent with the staff guidelines for LERF (See Section 2.4.2.2 of DG-1061).

State your definition of Large Early Release and the implications of any differences in your definition compared to the NRC dermition.

i l

Junes,199:

3 1

____m

RAls FOR CENPSD-1045

1.'

'Ihe submittal chooses the CDF as the risk measure to analyze the AOT risk for the containment spray (CS) system We recognize that the CS system interacts with the ECCS and its unavailability affects

'he CDF, but the primary function of the CS is to mitigate consequences of accidents including core damage accidents. The AOT risk analysis of the CS system should have addressed frequency of releases quantitatively. Section 4.3 of the draA Regulatory Guide DG-1065 proposes evaluation of incremental conditional core damage probability (ICCDP) and incremental conditional large early release probability (ICLERP). The latter measure is more appropriate for systems such as the containment spray system Please explain why the submittal did not choose the "6tquency of releases" as a risk measure to 1

analyze the AOT risk of the CS system

2. Related to Question No.1, in Section 6.3.5, as in Section 6.3.5 of your CEOG HPSI report, you present an " Assessment of Large Early Radiological Releases." While, as with the HPSI report, some of the qualitative argn=~da appear reamanahle, thesejustifications need to bejudged against LERF

' values, consistent with the (draA) Regulatory Guide 1.174 (DG-1061). Please provide LERF and l

deha LERF values for each of the CEOG NPPs or arpmade that they are not needed, consistent with I

4 the staff guidelines for LERF and delta LERF (See Section 2.4.2.2 of DG-1061).

Please state your definition oflarge early release and the implications of any differences in your

' de6nition compared to the NRC dermition.

l I

3. While recognizing that the main focus of contamment<spability issues is on the conditional and absolute LERF values, it is also true that the probability oflate contamment failure is usually much larger than the probability of early containment failure. It is also true that the containment heat removal capability plays an important role in keeping the contamment from failing. Please discuss the impact of taking a CS train out for PM or CM on the change in late contamment failure probability and large late release frequency.

4. The AOT risk results presented in the submittal are based on the IPE models that generally analyze the progression of accidents for a mission time of 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />. This assumption can impact the AOT risk results for systems whose performance is needed beyond the IPE mission time (e.g., CS.) Please address how the CE plant IPEs handled accident scenarios whose end states werejudged to be "OK" at the end of the mission time but had a pr*~*61 o challenge the reactor vessel and contamment t

integrity.

5.

In section 5.2.1 (page 14, last sentence of the 1" paragraph), you referenced the adnumstrative guidance, which requires the performance of risk assessment for PM mamtenance. Please elaborate on the scope and nature of the risk assessment.

- 6. In~section 5.2, you referenced LER # 94 005 01 (your reference 13, parj 67) which forced Palo Verde Unit 2 to shutdown due to inability to repair an MOV valve whin the present AOT. The same LER is referenced in the HPSI subimtsal Are you referencing the same valve? Please clarify.

7. In section 6.1, page 19," Statement of Need"you state:

" Preventive maintenance of the CS components during "at power" operation will reduce the maintenance burden during the shutdown and allow increased availability of shutdown significant components (SDC heat exchanger, CS pump as a backup for SDC) during lower mode plant operation."

l June 15,1998 lL

O This statement implies that the risk of mode 4 and 5 is high enough that it warrants four levels of redundancy for SDC pumping function (two LPSI purtp and two CS pumps.) If the proposed AOT extension is approved, are you implementing programmatic procedures to ensure availability of these four pumps during modes 4 and 57

8. Your submittal highlights the unique design features of the CS system for Palo Verde units. As evident in the risk results reported in Tables 6.3.2-1 and 6.3.2-2, these design features are responsible for higher vulnerability of Palo Verde units to increase in unavailability of CS as a result of the new proposed AOT. 'Ihe Single AOT risk for Palo Verde units (both PM and CM) exceeds the numerical guidelines of DG-1065. Please provide your rationale for including Palo Verde in the joint application.

9.

In Section 6.3.1, page 27, you state:

"For this evaluation, continuation of at power operation with the LCO ACTION statement is compared with the risk of proceeding with a plant shutdown. An estimate of this risk was evaluated by modifying the reactor trip core melt scenario for a representative CE NSSS PWR...

This statement implies that a ratio technique was employed to predict the transition risk for other CE plants. This observation contradicts the data presented in Table 6.3.3-1, page 47 of the submittal. In this table, the estimates of transition risk appear to be plant specific. Please address this apparent discrepancy.

10. Please describe the process of predicting the new average CDF based on the assumption of the new average unavailability of 0.022 per a single CS train (Section 6.3.2.3). In your response, please provide the following for each IPE:

What is the baseline unavailability estimate assumed for a CS train?

Are there any cutsets in the master list of core damage cutsets that relate to the unavailability of CS due to maintenance?

How did you account for maintenance unavailability of the second CS train in the re-e quantification process?

11. In Section 6.3.2.4, page 36, the submittaljustifies the high AOT risk of ANO-2 on the rigid success criterion assumed in the ANO-2 IPE. With respect to the CS system, the success criterion is the same for all CE plants. "Ihat is, one of the two trams needs to be functional. The AOT risk calculation analyzes the risk impact associated with the unavailability ofone train of CS train. The AOT risk does not affect the assumptions of the IPE for the fan coolers. 'Iherefore, the change in CDF should only reflect the change in CS unavailability. One plausiblejustification for this behavior is that the cutsets related to containment systems failures were not truncated in the ANO-2 IPE. This is because ANO-2 IPE assumed a more rigid success criterion for containment cooling than other IPEs. Please address the affect of tmncation on AOT risk results.

12. In Section 6.3.2.4, page 37, in discussing the interaction of CS system with the ECCS for Palo Verde, the submittal references the ability for ECCS recovery that was not credited in the IPE. Please elaborate further.

13. In Table 6.3.4-2, the submittal attempts to rationalize that a system with four !cvels of redundancy is more reliable than a system with three levels of redundancy. This concept can easily be conveyed June,15.1998 2

rm

i qualitatively without a need to resort to building a simple model of shutdown operations that are complex in nature.

14. On page 11, you state that,"These studies [ assuming that you mean the three IPE studies - your references 5,6, & 7] demonstrate that containment protection can be provided with either one CS subtrain or one fan cooling unit." Do all the CEOG IPE studies demonstrate this (with, of course, the exception of Palo Verde)? In this context, what do you mean by " containment protection"? Is containment protection assured for Palo Verde with one CS subtrain?

~

5.

June 15,1998 3

1 1

i l

CE OWNERS GROUP Project No. 692 cc:

Mr. Gordon C. Bischoff CEOG Project Manager ABB Combustion Engineering l

M.S. 9615-1932 2000 Day Hill Road Windsor, CT 06095 Mr. David Pilmer, Chairman CE Owners Group San Onofrc Nuclear Generating Station 14300 Mess Road San Clemente, CA 92672 Mr. Ian C. Rickard, Director Nuclear Licensing ABB-Combustion Engineering, Inc.

Post Office Box 500 2000 Day Hill Road Windsor, CT 06095 Mr. Charles B. Brinkman, Manager Washington Operations ABB-Combustion Engineering, Inc.

12300 Twinbrook Parkway, Suite 330 Rockville, MD 20852

-