ML18096A736

LER 91-036-01:on 911213,determined That AFW Flow Values May Have Had an Adverse Impact on Licensing Basis Slb.Caused by Inadequate Design Review.Cycle 10 Operation,W/New AFW Flows Has Been reviewed.W/920522 Ltr

ML18096A736
Person / Time
Site:	Salem
Issue date:	05/22/1992
From:	Pollack M, Vondra C Public Service Enterprise Group
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
LER-91-036, LER-91-36, NUDOCS 9206010031
Download: ML18096A736 (9)

Text

OPS~~

Salem Generating Station May 22, 1992 U. s. Nuclear Regulatory Commission Document Control Desk Washington, DC 20555

Dear Sir:

SALEM GENERATING STATION LICENSE NO. DPR-70 DOCKET NO. 50-272 UNIT NO. 1 LICENSEE EVENT REPORT 91-036-01; SUPPLEMENT This Supplemental Licensee Event Report is being submitted pursuant to 10CFR50.7J. This supplement modifies the root cause and corrective action sections based upon completed Engineering investigations *. Also, the 10CF5.7J reportability identification code on the NRC Form 366 has been corrected.

tZAlJ-

c. A. Vondra General Manager -

Salem Operations MJP:pc Distribution

'1 /

,**/ -1.., f,r

/

I ,

/ \ 'v

-9~G&Gl0031 PDR ADOCK s

!<AC 1.1\/l,,j lM US ljlJCI I AA 111 llUI ATOR* COMMl'.\St<l"

""'"'. ' "lllt \ .. h' ....

'*"' 111.'*' It fl f41*llt*t **** , fl UI :...,,..._~, 'i' I I , ..... L. "' *** flt&',

1*.*' ,.,,.,,. 11 I'\ ~ I*,, I I' ' I .... Ill' Jt *l ... ' .. ,. ,l ,,.,.. I' *II\' .\11(*

LICENSEE EVENT REPORT (LERI ,.,),,,,, ~47', 14! ' 1 hUl!1,,: t4ull'.I( *~I C., 1 1\*I\ If ' ' ' ! I l l II!' r.'Jlil~ *

.\'\i' ut **~*II'\ "-'°""°'\.fl ~I t"ll,\~1 11 :1* *, ~ ** . '. .,., ;1'1 f o\11

~11 **, *. , ,\ 1 1111°' I '\'1~*:111** ...,,\ 1 0 111~'*'1'"" j1,' ,'ll'*'*'* A"~!',,'

~"I fo\l'!U~\'*11111. Hl(hllJ*l'fl'tl(IJI:? 111'*1\1!1'-I IHllff 1

.11 \t,\:-*\ql \'I ".,1 ,\"',;11 IH*Lu.",f f w\to\~t11 ..,*}*:"'*  ;

1

• • .'"'**'1

'"'CILITY ...... , ,,, lO'nC*ll "IJM81A " ' I PA(*I il*

TITll I**

Salem Generating St.it ion ~---~'.'.2.~-~--------*----*--------- *.-*-*--.. ---*** ~l_J~ __ L.'~J_O_J_~J_::l.~l :*..Ld~J!:_L~

Di sco\*cry o ( nonconsC"' rv.i t i Vl' i1ux i l i a ry F1~(**lw<1 t t~ r Sys t L'r.* f i o;,* r,1 t cs in St car: t. i Ill' 11 t*'.1 k.

~~.,l. Pollack - J.EP Coon"!in,1tor 1--------------------------------------*--*-- *---,**-- ___ i . l .

---

------------~1r: ::.*:;=.: :~~-=~~l':'~~~:~ f ..:.-L_L.. L-L ... ____ ._..._

CO~PL(T[ 0~( ll~f S:QA l4C'1 C01i4PO'il""tt f'AILUAf DfSCRlllO '"" TUIS AfP()AT '1]*

~*:~-u_*~~t~~s~*~ *~!-_.-!.+~*~- --'.'-...*_-"_- .)_*-._,~*-~-~-+lj~ -,_.-__.-:-_.,-_. _;-_~-*--~-.jf_'_1 ~-'*_: ,_:_J.a,),_=:_~ 1------.....;.,+-'_**J:=J_~~:~~~:~. --~~l- - -~- ~-.'.-~,-~ - :*___--_~:_-4;-~- ~* ~-: -=~;~ ~ I~~~~~~~~~~~=

.. t_. _.'

I I

! I I

I I .I I

I i I I

I I

I !

I

! I I I i

j_i___~c_-1 _Lj__!~-

__l_ __..:..

I i .J 1--..

- -...---~---4 s_u_""~l_l_.. _1"_'_A...;l_A_£_PO'::__! .. 'lCT(O 1 1

---~ ***t*  !-=::.:.::+_:_:_::___ ;

n

.. Es,, **** :c1 .. *p**~* t *"CCTED

. .. ~.*.-ll'-'*S~* l-:..., *.,~ *c

---

c- U -------*-----*----

i

! *1:

--~:*.*"::**.

______ I_::_l:~---- __ -*--- -*- . _~--- __ L..-.. _____________ _j_ ___ i ___ i_ ... J .... l.: __ J__

i I i On 12/13/91 it was determined that the AFW flow values may have had an

~dverse impact on.the licen~ing basis Steam Line Break (SLB) analyses, placing both Salem Units in an "unanalyzed cohdition". This wris discovered during the assessment of a recently developed Salem-specific model used to generate maximum AFW flows as a function of steam generator pressure. A reanalysis of SLB cases, indicated that the Salem Unit 1 core ~nd containment response would remain within the licensing basis limits thereby justifying continued operation; however, until the AFW flow assumptions used in the SLB analyses were corrected, and the effects of the new flow values on unit operation assessed, Salem Units 1 & 2 were considered to have been in an unanalyzed condition. The new flow values are significantly and nonconservatively different than those-descrihed ih the Salem UFSAR, therefore Salem is considered to have been in a .

condition 6utside of its design hasis. The root cause of the inaccurate AFW flow values (dating back. to 1978) is inadequate design review. The calculations were apparently based on incorrect ~odeling assumptions for AFW system hydraulic conditions. current £tandards require detailed verification for all safety related calculations which should preclude this event from occurring today. A project team was assembled to develop and implement corrective actions to s.ibport future cycles of operation for* both Salem tJni ts. l OCFR :J.1 has been shown to not be applicable.

IWRC Form )66 ih l't~I

. :*~-

LICENSEE EVENT REPORT (LER) TEXT CONTINUATION Salem Generating Station DOCKET NUMBER LER m.JMBER PAGE Unit 1 5000272 91-036-01 2 of 8 PLANT AND SYSTEM IDENTIFICATION:

Westinghouse - Pressurized Water Reactor Energy Industry Identification system (EIIS) codes are identified in the text as {xx}

IDENTIFICATION OF OCCURRENCE:

Discovery of nonconservative Auxiliary Feedwater System flow rates used in Steam Line Break analyses Discovery Date: 12/13/91 Report Date: 5/22/92 I.

This report was initiated by Incident Report No.91-909.

CONDITIONS PRIOR TO OCCURRENCE:

Unit 1: Mode 1 ReaEtor Power 100% - Unit Load 1150 MWe Unit 2: Defueled - 6 h Refueling in progress DESCRIPTION OF OCCURRENCE:

To increase the margin between Salem's containment ~afety analyses results and the licensing basis limits for containment temperature and pressure; PSE&G implemented a Containment Margin Program with .

Westinghouse in July 1991. The program involves reevaluation of the containment response during postulated events, with the expectation of identifying more realistic (but conservative) assumptions.

On~ area investigated was the Auxiliary Feedwater system. {BA} flow.

UFSAR section 15.4.8.2 indicates that the AFW flow assumptions used in the Steam Line Break (SLB) analyses (as given in UFSAR Table 10.4-2) are bounding values associated with a completely depressurized faulted steam generator. The SLB cases which produce the most limiting containment response are typically smaller po~tulated breaks (less than l .*4 square feet) , which would have relatively low depressurization rates' for the faulted steam generator. Therefore, it was anticipated that a more realistic model of AFW flow vs. steam generator pressure could justify a reduction in total integrated AFW flow reaching the faulted steam generator.

Reduced flow would result in less inventory contributing to the mass and energy release out of the break, which would result in lower calculated containment te~perature and pressure.

A Salem-specific model was developed and used to generate maximum AFW flows as a function of steam generator pressure. The preliminary results were provided to Westinghou~e (by letter dated November 22, 1991), to determine the sen~itivity of the SLB results.to the AFW flows. Their results indicated that containment response could be

LICENSEE EVENT REPORT (LER) TEXT CONTINUATION Salem Generating Station DOCKET NUMBER LER NUMBER PAGE Unit 1 5 0 0 0 2 7.,!0.2_ _ _ _---"9'--"1'--_,0::...::3~6:<...--0~1_ _ __.,,3:..-.;:0._.f,__,8"--_

DESCRIPTION OF OCCURRENCE: (cont'd) significantly and adversely affected by the new AFW flow predictions.

Based on .the results of the preliminary sensitivity study, PSE&G took the following immediate actions:

1) Met with Westinghouse to review the AFW flow input assumptions for the licensing basis SLB. The.results of the review are summarized under "Analysis of Occurrence."

2) The Salem-specific AFW model was completed including independent review. This review revealed a modeling err6r, which when corrected, resulted in approximately 5% lower predicted flows. However, these flows remained higher than the licensing basis analysis values.

3) The impact of these higher AFW flows, on safe operation of Salem Unit 1, was evaluated (because Unit 2 was shutdown,*

it was not.an immediate concern). The excess available shutdown margin specifi9 to Salem 1, Cycle 10 was judged sufficient t,o offset the effects of the increased AFW flow. At the request of PSE&G, Westinghouse began reanalyses of the SLB cases to confirm this judgement.

On December 13, 1991 it was determined that the AFW flow values may

• have an adverse impact on th~ licensing basis SLB analyses results.

A one hour notificatiori was made (for both Salem Units), to the Nuclear Regulatory Commission, pursuant to Code of Federal Regulations 10CFR50.721b) (1). The reanalysis of SLB cases using the Salem 1, Cycle 10 specific shutdown margin, indicated that the reactor core and containment response would remain within the licensing basis limits thereby justifying its continued operation at full power. The evaluation of the impact of increased AFW f lciw upon the Salem 1, Cycle 10 design basis was transmitted by Westinghouse to PSE&G on December 20, 1991 and confirmed the initial judgement outlined in item 3 above.

APPARENT CAUSE OF OCCURRENCE The r6ot cause of th~ use of an iriaccurate AFW flow value (dating back t6 1978) is inadequate design review. The calculations were apparently based on incorrect modeling ass*1:npt ions. .f.or AFW system hydraulic conditions. Curr~nt standards require detailed -

verification for all safety related calculations which should preclude this event from occurring today.

Three (3) basic AFW maximum faulted loop flow

, ~ssumptions have been utilized in Salem's licensing basis:

1. 2040 gpm - This value was initially used in the Salem licensing basis SLB analyses. Westinghouse performed the r

LICENSEE EVENT REPORT (LER) TEXT CONTINUATION Salem Generating Station DOCKET NUMBER LER NUMBER PAGE Unit 1 5000272 91-036-01 4 of 8 APPARENT CAUSE OF OCCURRENCE (cont'd)

SLB analyses using multiple single failure assumptions thereby reducing the number of cases by 75%. This value (associated with accident scenarios involving an assumed single failure ~f the AFW runout protection system) was calcul~ted by PSE&G and transmitted to Westinghouse in 1978. In the initial SLB analyses, this AFW maximum flow value was assumed in all SLB cases.

2. 1840 ~pm - This value was calculated by PSE&G and transmitted to Westinghouse circa 1978. This value was conservatively associated with accident scenarios not involving an assumed single failure of the AFW runout protection system (as specified in the Final Safety Analysis Report (FSAR) section 15.4.8.2 through FSAR Revision 6)~ However as identified above, Westinghouse ran SLB cases with multiple single failure assumptions and used the 2040 assu~ption.

3. 1320 gpm - Circa 1984, Westinghouse performed SLB analyses in suppoFt of Boron Injection Tank (BIT) removal. These analyses required the separation of single*failure assumptions to meet the containment safety limits.

Therefore, for the AFW runout protection failure cases, Westinghouse continued to use the 2040 gpm assumption. For all other single failure assumption cases, Westinghouse took credit for a lower AFW flow assumption (1320 gpm), as per Westinghouse letter PSE-92-052, dated March 5, 1992

("Maximum AFW Flow Assumptions in SLB Licensing Basis Analyses"). The basis of this 1320 gpm value was the sum of the rated AFW flows with a conservatively calculated flow split to the faulted loop. In addition, the use of this 1321 gpm value was not initially conveyed to PSE&G.

Use of the value was revealed during a Containment Margins Project resulting in the issue of LER 272/91-036-00.

ANALYSIS OF OCCURRENCE:

The AFW flow rates assumed in the Salem SLB analyses, particularly in those cases for which AFW runout protection remains functional, were

• found to .be significantly lower than predicted using a Salem-specific AFW model. Using low flow rates in the SLB analyses i*

nonconservative in that it underestimates the degree of primary system cooling and return to ~riticality, and also underestimates the mass/energy released inside c.ontalnment.

Westinghouse currently uses a maximum flow of 1320 gpm for the cases without runout protection system failure. Assuming a constant flow of 1320 gpm to the faulted steam generator results in underestimation of total int~grated flow. The Apparent Cause of Occurrence Section

LICENSEE EVENT REPORT (LER) TEXT CONTINUATION Salem Generating Station DOCKET NUMBER LER NUMBER PAGE Unit 1 5000272 91-036-01 5 of 8 ANALYS!S OF OCCURRENCE: (cont'd) summarizes the results of inve?tigations as to the basis for AFW flow values.

Salem 1. Cycle 10 The effect of the increase in maximum AFW flow has been evaluated for Salem Unit 1 Cycle 10 operation. This increase in maximum AFW flow is non-conservative for the SLB analyses which address core response (DNBR), containment pressure, and containment temperature safety limits. This evaluation took credit for the following increases in minimum shutdown margin which are available for Salem Unit 1, Cycle 10:

1) Excess shutdown margin. Salem's }icensing basis analyses assume a shutdown margin of 1600 pcm. Because a minimum shutdown margin of 2570 pcm is available for Salem lr Cycle 10, the additional 970 pcm was* credited in the reanalyses.

2) Reduced uncertainty for Rod Worth Calculations. Salem's licensing basis analyses assume 10% uncertainty on rod

• worth calculations for determining shutdown margin.

W~stinghouse methodology supporting a 7% uncertainty (applicable to all Salem l and 2 cycles of operation), was used, resulting in an additional 200 pcm of shutdown margin for Salem Unit 1, Cycle 10.

With respect to SLB core response, the higher AFW flow increases the amount of RCS {AB} cooldown, which would result in a positive reactivity insertion. The positive reactivity insertion would result in an increased return to pciwer level following the SLB, which has a negative impact on Departure from Nucleate Boiling (DNB). The SLB reanalyses showed only a slight negative impact on DNB, vs. a large margin between the DNB safety limit and the predicted worst case DNB for the SLB. The impact of the higher AFW flow is easily accommodated by the excess DNB margin for the SLB event.

With respect to containment integrity, Westinghouse reanalyzed the limiting SLB cases for containment temperature and pressure.. The limiting pressure case, a o.9oa sq. ft. split bre~k from 70% power, with an emergency diesel failure, result~d in a peak pressure of 45.8 psig, vs. containment design pressure of 47.0 psig. It was also concluded that the* containment temperature response would not exceed the present Salem EQ temperature envelope.

• No other non~LOCA event. and no LOCA events ar~ adversely impacted by an increase in the maximum AFW flow.

Previous Unit 1 and Unit 2 Operation The results of the Salem 1, Cycle 10 safety analyses may be applied

LICENSEE EVENT REPORT (LER) TEXT CONTINUATION Salem Generating Station DOCKET NUM.BER LER NUMBER PAGE Unit 1 5000272 91-036-01 6 of 8 ANALYSIS OF OCCURRENCE (cont'd) to all other Salem cycles whose excess shutdown margin can be demonstra.ted to be greater than or equal to 1170 pcm.

C~cles with Boron Injection ~ank CBtTl Available The Boron Injection Tanks were functionally removed during Salem 1, 7th refueling and Salem 2~ 4th Refueling. The reacti~ity worth of the BIT volume at 20,000 ppm boron is a minimum of 1200 pcm for the conditions associated with the SLB event (End of Cycle, o ppm RCS boron concentration). The BIT benefit alone is greater than the 1170 pci:n used in the Salem 1, Cyc1e*10 evaluation. In addition, the cycles with the BIT available would benefit from an additional 100 pcm (minimum) excess shutdown margin, as we11* as a 100 pcm (minimum) from the reduction in rod worth uncertainty.

It.is therefore concluded that the conclusions from the Salem 1, Cycle 10 evaluation are applicable to Salem 1, Cycles 1 through 7, and Salem 2, cycles 1 through 4.

Cycles Without BIT Available I

Salem l ,* cycles 7, 8, and 9 Because of the Salem 1 control rod pattern, this unit has historically had significant quantities of excess available shutdown margin. Unit 1, Cycle 8 had a minimum shutdown margin of 2550 pcm (20 pcm.less than cycle 10), and a reduced rod worth uncertainty of 198 pcm (2 pcm less than Cycle 10). This results in cycle 8 having a total of 22 pcm less shutdown margin than Cycle 10. This small difference (0.8% of total Cycle 10 shutdown margin) results in a negligible increase in the return to power during a SLB event.* Given the excess margin to the containment limits shown by the cycle 10 analyses (i.e., 45.8 psig vs. 47.psig), as well as the ample margin to the DNBR limit available, it is concluded that the results of the cycl* 10 analyses demonstrate acceptable ~esults for a postulated Salem 1, Cycle 8 SLB. .

Cycle 9 had 2950 pcm *shutdown margin, which exceeds the value for

• Cycle 10 (2570 pcm). The reduced rod wbrth uncertainty available Cycle 9 exceeds the 200 pcm assumed for Cycle 10. Therefore, it is .

. for concluded that the results of the Cycle 10 analyses are applicable to Unit 1, Cycle 9.

Salem 2, Cycles 5 and 6 The contr~l rod pattern at Salem 2 provides less excess shutdown margin than that of Unit 1. Salem 2, Cycle 5 had the lowest excess shutdown margin (80 pcm) and reduced rod worth uncertainty benefit (166 pcm) of all Salem cycles subsequent to BIT removal. The tot~l

LICENSEE EVENT REPORT (LER) TEXT CONTINUATION Salem Generating Station DOCKET NUMBER LER NUMBER ?AGE Unit 1 5000272 91-036-01 7 of 8 ANALYSIS OF OCCURRENCE (cont'd) excess shutdown margin for Cycle 5 was 246 pcm, which is considerably less than the 1170 pcm used in the Salem 1, cycle 10 evaluation.

However, the licensing basis shutdown margin calculations for all cycles is reduced by the reactivity worth of most reactive rod (i.e.,

it is assumed to be stuck). For Salem 2, Cycle 5, the postulated

• maximum worth stuck rod had a reactivity of 940 pcm. Based on rod drop time testing per Technical Specifications, successful rod motion testing every 31 days (Technical Specification Surveillances 4.1.3.1.2

( and 4.1.3.2.2) and control rod performance during actual d~mands (reactor trips), there have been no indications of stuck rods at Salem during Cycles 5, and 6. Therefore, taking credit for this results in a total excess shutdown margin of 1186 pcm. This was the minimum excess shutdown margin associated with any of the Salem Unit 2 cycles without the BIT.

PSE&G is not proposing to remove the stuck rod assumption from the lem licensing basis. However, for the limited purpose of assessing U

fety significance of increased AFW flow on Salem 2, Cycles 5 and 6 d taking credit for all rods operable, this results in sufficient cess shutdown margin to conclude that the results of a postulated P would have been bounded by the Salem 1, Cycle 10 eValuation.

Until the AFW flow assumptions used in the SLB analyses were corrected, and the effects of the new flow values on unit operation assessed, Salem Units 1 & 2 were considered to have been in an -

unanalyzed condition~ The new flow ~alues are significantly higher (nonconservative) than those described in the Salem UFSAR, therefore Salem is considered to have been in a condition outside of its design basis. Therefore, this event is reportable in accordance with 10 CFR

50. 73 (a) (2) (ii).

CORRECTIVE ACTION:

The Westinghouse evaluation of Salem 1, Cycle 10 operation, with the new AFW flows, has been reviewed by PSE&G. Based on the Westinghouse input, PSE&G completed a Safety EvaluationJ which included a 10CFR 50.59 evaluation, demonstrating-that Salem Unit 1 cycle 10 operation is acceptable.. Salem Unit 1 procedures h~ve been revised to ensure excess shutdown margin, required by current evaluation, is maintained.

PSF.&G has formed a project team to develop and implement corrective actions to support future cycles of operation for both Salem Units.

As part of normal station practices, the Salt~ Unit 2 fuel load and

~ubsequent mode change~ (in iupport of the 6

• r~fueling outage) was dependent on the issuance* of a Reload Safety Evaluation (RSE) . The RSF. accounted for the pr~ger AFW flow value. Corrective actions implemented during tbe 6 . refueling outage and cre~ited in the RSE for Salem 2, Cycle 7 operation consisted of 1) reduction of the Containment High-High pressure setpoint to 15 psig (from 23.5 psig)

~ *

--~-*-----

~:- -~H **

~ ... - :. ~~

• _ -** ** -* *~ ****'-** __** :.

f

• *** ~ ,4

. -. . . .* . . _* * *.:....... - ... -~**:.: _1

• **' I 1

LICENSEE EVENT REPORT (LER) TEXT CONTINUATION

-*~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Salem Generating Station DOCKET NUMBER 1 LER NUMBER PAGE Unit l 5000272 91-036-01 8 of 8 CORRECTIVE ACTION: (cont'd) and 2) administrative controls to ensure that suhtdown margin is maintained 250 pcm above the current Technical Specification value.

The project team is i~glementing the same corrective actions during the present* Unit 1 10 refueling outage for Salem ~git 1, cycle 11 operation as was done for Salem Unit 2 during its 6 r~fueling outage.

The project team is currently eva1uating ~lternatives for permanent implementation ~gring the Salem Unit 2 7t refueling outage and the Salem Unit 1 11 refueling outage.

Westinghouse (at PSE&Gs request), investigated to determine if any generic implications (i.e., 10CF*R21 issues) associated with their use of non-conservative AFW flow assumptions in the Salem licensing basis analyses. Westinghouse found that no generic impl'ications exist because the assumptions used were based either on information obtained from PSE&G or on Salem specifc -AFW system information.

In 1991, PSE&G proactively initiated the development of a Salem specific Accident Analysis Profile (AAP) with Westinghouse. This project, scheduled for completion in 1992, will provide a detailed description of the key accident analyses input assumptions used in our licensing basis. The primary purpose of AAP is to provide on-site engineering staff with additional information with regard to accident analysis assumptions and aid in the evaluation of safety significant issues. In addition, the development of AAP, coupled with the on-going Configuration Baseline Documentation development

• process, will aid in determining if similar discrepancies exist between analysis inputs and actual plant* performance.

/? . /

,.!'~~:'/,/ ,..)~ //' //

. ~* /'; *.,,-/-~~/* /:

~ ~.d Ir/:/

y_ (,

I '{f/~rt!L-General Manager -

Salem Oper~tions MJP:pc SORC Mtg.92-062 I