ML20024B438

Minutes of 791108 & 30 Staff Meetings to Review Findings & Recommendations of Technical Review Committee Established After TMI-2 Accident

ML20024B438
Person / Time
Site:	Crane
Issue date:	12/07/1979
From:	Ball R BABCOCK & WILCOX CO.
To:	Macmillan J BABCOCK & WILCOX CO.
References
TASK-02, TASK-03, TASK-05, TASK-06, TASK-07, TASK-08, TASK-09, TASK-2, TASK-3, TASK-5, TASK-6, TASK-7, TASK-8, TASK-9, TASK-GB GPU-0232, GPU-232, NUDOCS 8307080729
Download: ML20024B438 (29)

Text

{{#Wiki_filter:jt . -... = b kt.(r. O i q bM O THE BAECOCK & Wit.COX COMPANY 4 POWERGEHj.RATIONGROUP ]TJtt J. H. Itac!!illan, Vice President - HPGD DEC 11.1979 F,;fryi To l g 5.:- n . V 30. Fron R. H. Ball, Manager - Prcduc} Development sos sea.a

f. Q d d {3 { %

File No.

-,g

. Cast. or Ref. [ v 7 Subj. Date l MIt.'tJTES OF !!EETING O!! !!PGD STAFF REVIEN December 7,1979 I i g ~ l ~ On flovember 8 and 30, mea'bers of your staff met with you to review the Findings and Recer:1endaticas of the Technical P.eview Cormittee established L af ter the TTt!-2 accident. These minutes document the planned response i by the Division. [ 1-1 Control Room / Operator Interface stematic transient 0 1) Staff concurs with the recorrendation on sy' carried out by the i analysis. The recor= ended action is being Anticipated Transient Operating Guide (ATOG) program for both L the Operating Plants and the bactisg. The mini-WASH - 1400 JM i study (at Crystal River) will.provice a source of events for g . t the program. ACTION: Dr. Roy is rescensible for the engineering analysis associated with the prog'am and R. E. Xos'iba is responsible

• for the trainir.g and gperating plant interface.

~ l 2) The Staff ccacurred with the recocccndation to improve rocci - ' design. Some werk nust be cone to define the HPGD control in i control rcom design. ACTION: Dr. Roy has engineering and design responsibility. thM I ' An initial aperoach will be presented to the Staff by Q P.D. 'F jy '[ 12/19/79 by Ken Suhrke.,* 1 ACTION: N. S. E.-brey will resort on the institutional in-fluences affecting centrol rcen design. He will define roles j(~ for the utility the architect engineer and the Bailey Con-E t trols Ccepany relative to F.PG':. 3) The Staff concurs with the reccerer.dations and notes that the tie specific trstrr.cnts rentiered in the Report have r been realized as procucts ar.d are being sold.' The item is [ cceplete. I l 3 [ l NO k ForiD Ex. 5 O/? g?/ J.R. Danyo [- t 8307050729 791207 PDR ADOCK 05000289 -w- = e - P HOL .

• A la e

- r ~ l g, 3 hh i December 7,1973 R. M. Ball to J. W Macf411an 2

1:mt o

W ' N 4 ,, 4). The Staff concurs with the recommendation and notes that the 1 0 1 ATOG program is the starting point of the implementation. Staff recommends further that a policy decision be considered h a on NPGD review of the Customer's control room procedures for s t i concurrence with the design guidelines. - O 1 ACTION: J. H. MacMillan to consider a policy directive on j ~- this issue. D. W. Berger will provide information by January 15. 1980 on a stap-by-step apprcach to accceplish iy [ the policy directive. .i j Steps mentioned: Organizational Clarity. ATOG implementation. Training review and personnel requirements. The Staff recom- ['

I mends that as a present minimum that EW establish agreement s'

with the Customer on the training procedures and that the user control room procedures be used in simulator training. ACTION: Dr. Roy will review the Engineering Department responsibility in the preparation of procedures and review: What disciplines are used? What is, or should be the organi-zation? How do we close the loop with the user? What documents does Engineering review? 5) The Staff concurs with the reconnendation to enhance fundamental operator knowledge through trainin,. g ACTION: R. E. Kosiba will doctrient and provide evidence by F January 15.19E0. Some of the actions already noted are re- .N% visions of the training prograri, greater involvement with 46 i 9 the customer training program, interaction with the Institute for Nuclear Plant Operations (I!;PO), and hiring of Professor 4 Robeson to review training pregrams. 6) The Staff concurs with tha reca:::endation and notes that there [. is no specific program required but that the work will be part f [ of f.he individual evaluations in the ATOG program and the i mini-WASH 1400. 1 ACTION: Dr. Roy will assure that auto ation alternatives are considered in the evaluation of the ATCG results. y [ 1 1-2 Off-Normal Systems Analysis (* 1) The Staff concurs with the reconcendation"and notes that it will be implemented in the ATCG and = int-EASH 1400 studies. ( ACTION: Dr. Roy will arrange for a EW man to be part of * {~ the liRC team.. I 2) The Staff concurs with the recomendation and notes that a i Failure P.cdes and Effect Analyst: (FTEA) has been completed on the Integral Control System and that a reliability study exists for the Auxilliary Feedwater (AFW) system. The need for a reliability study en the " sin Feedwater System and on the Pilot Cperated Relief Valve (FORV) will be evalua,ted by Engineering. s = v-- _n_. n._, ..A W .W k j ( J

-_...a-a__ z __.x i ~ l t. 31). p . y, n R. M. Ball to J. H. MacMillan 3 December 7,1979 [9y .c

p TNT

.;p=== i 3) The Staff believes that no specific action is required by h this recocuendat*on relative to TMI-2. Administrative con-7 trols and sorveillance procedures are being addressed by O the utilities and will be part of the INPO overview. P 'Q { i Reactimeter .2-1 s The Staff concurs with the recommendation and notes that the . Q-L RECALL system is being made available.. i l ACTION: R..E. Kosiba will report by January 15, 1980 on utility acceptance and B&W action. This is a necessary item in the, plants. c. 22 Pilot Operated Relief Yalve (PORV) .1) The Staff does not concur with the reconcendation. The r change recocuended has been implemented but it should be s 0 considered a short tenn and temporary measure which will he subject to additional analysis and review by NPGD and [ c. the NRC. r

• 2)

The Staff concurs with the recoccendation and notes that it has been implenented and has met the NRC requirement date .of January 1, 1980. The system was tested to the require- . ments of the "Short Term Lessons Learned" HUREG 0578 NRC [ l doctment.- L h . 3) The Staff does not concur with the recoceendation and notes b that it will be further evaluated in conjunction with the E study to restore the PORV function (The study referfed to [ in,1 above). ~ ~ 4). The Staff concurs with the receraendation and notes that the work has been comple1;ed.. 1 5) The Staff "re[ects the recorrendation on the basis that the PORY has a valid purpose in the operation of the MSS. [ q h 6) The Staff concurs with the recoceendation and notes that the action will be taken as part of the R&D program associated with the recovery of TMI-2. g ACTION: G. E. )(ulynych will be contacted by R. M. Ball to i assure that this is part of the program. St.W will also be i. part of the valve testing program now being initiated by EPRI. ( r 7) The Staff concurs with the recensendation and notes that the I action has been completed in response to the MRC request on all B&W plants. l 8) For emphasis, the Staff reccenends as an added item that a program be surted and r.onitored to restore the PORY function h as origitally intended with the added emphasis on the system' i ~ reliability. l-L. _-__-.-__y_., ~ b 5

l ~ ~~ - u-- ~ ~ w 4 ? .T i c, di R._]f. Ball to J. H. MacMillan 4 December 7, 1979 I (c@ d PW7 1 ACTION: Dr. Roy will report to the staff on the existence k- { M a program. l p ~ 2-3 Main feedwater Reliability i[ i g The Staff concurs with the recossendation. E ACTION: Dr. Roy will assure that the s'tudy is completed and L, also develop a plan for fr. groved feedwater *;ontrol. E" 2-4 Aux 1111ary Fluid Systens Evaluation i t The Staff concurs with.the reconcendation that the system be r reviewed although it considers this a low priority) action. [ The specific areas recocuended for review are: (3)(baron addition 1 Letdown. degas and seal injection. (2) sample lines. to the reactor coolant system, and (4) possible additional F shielding. ACTION: Dr. Roy will establish a review plan. 2-5 Valve Status I 1.2.3) Thestaffrejectstherecessendai:ionsonthebasisthatthe i proper way to respond to the finding should be through tighter control and discipline in the administration of procedures tiy j the utility. System monitoring and status display should be r incorporated in the control room study identified in 1-1. [ 4) The Staff considers that this recocuendation in redundant i with 1-2 which studies the reliability of the AN. The } AN study should censider the autanatic actuation of the i block valve on loss of feedwater (LON). ACTION: Dr. Roy will assure that this actuatio'n is considered. 2-C Vapor and Gases in the Primary Systen j la) The Staff concurs' with.the receraendation and notes that two phase flow studies will be part of the evaluation with the = .. revised SMC code. ACTICN: Dr. Roy will determine when this capability will be i L avaT1able. He will also assure that the ncn-condensible gas effects are included'in the study and determine if any difff-E culties would be solved by the presence of J-leg vents and ( evaluate what detection methods would be required to determine i the presence of non-condensible gases. 1b) The Staff concurs with the recernendation and notes that some L studies have been completed. I. L ACTION: Dr. Roy will assure that the studies have defined tne decay heat removal methods, operational requirements and capabilities and ifmitation with respect to time (when) and - i l 5 P w.a, ,w ,_m. _y-m m= 4 ma J,'

1. e p------d-----

aey,-t- -mu-,-m -r--- e w w eee------P=--'w-&----w-*---e- -w-m-- ---ese-------m-- - --=*-

= a i: t; g jg 4 R. M. Rait en.1 t MacMt11=n 5 December 7,'1979 (: 1 ltY h. I fC 1 p duration, of' use. 2a} The Staff concurs with only that portion of the recomendation dJ - P dealing with the J-leg vent and notes that this is part of the l F required NUREG 0578 action. The recomendation to vent the o 1-i dome of the reactor vessel is rejected. I h 2b) The Staff does not concur with the recomendation. It believes 5 g that the function will be accomplished by the AFW functiori with s* appropriate reltahility and control. [ 2c) The Staff concurs with the recoamendation and notes that the implementation of the T-Sat (saturation tenperature and pressure) ' r system and the use of the incore therinoccuples satisfies the ' E. l' recomendation, i 2d) The Staff rejects the recommendation as too vague. 2e) The Staff concurs with the recommendation and notes that it is sat-1sfied using the T-Sat meter and the results of 2-6. la). [ 2?) The Staff concurs with the recomendat'ian and notes that tha' i AT06 analysis will define the methods available to assure adequate core cooling. 2g) The staff concurs with the recomendation but notes that no issediate response is required. The impleentation is incor-porated in the response to*the Centon 10/24/79 order. Review of the AFW control also responds to the thtent of this recom-I mandation. E. L 2h) The Staff concurs with the recor.mendation and it has been L impimented on all plants. 2-7 Decay He.at Renoval g 1) The staff rejects the recocmendation and notes that the Technical ' Product Evaluation Program (TPEP) will evaluate system alternatives for decay heat renoval over the full pressure range. 2-5) The Staff concurs with the reconzendation and notes that eval-uation proposals have been issued to both 205 and 177 owners. The 205 proposals have been accepted. 2-8 High Pressure Injection Lock-In [ ' t-1.3and4) The staff rejects the recornendations and notes that the proper i response is the enforcement of energency procedures which have t been defined as a result of the ATCG program. ~ ~ 2) The Staff concurs with the reccmendation and notes that the ATOG results provide the information. g i [ _m g,.. M m_ 1 i 1 L

~ - u - ,...w, i ; 4 1 J l3 C31 4 j R. M. Ball to J. H. MacMillan 6 Dec eber 7, 1979 t i 1 At!! { ACTION: R. E. Kostba will report on the implainentaMon of h*9 s a7IeTd ' change package which will provide for a reset of the [ e E5FAS system after there has been an operator override, a 1, i 1 l 2-9 Pressurizer Loop Seal The Staff concurs with the recomendation and notes that it is f being included in the TPEP work. '., e 2-10 Radioactive Liquid Wasta Handling t i active wasta transfer at THI-2 has invalidated the finding. The l [.: The Staff rejects the recomendation. Later information on radio- ! 7 Staff also notes that the systems are not part of the B&W business. 2-11 Enviromental Stress on Components f i s

I-1)

The Staff concurs with the recomendation and notes that a i i plan must define which components are to be studied and under ~~~ I t uhat conditions. The emphasis should be on critical components. This is considered to be a low priority item. ACTION: Dr. Roy will have the plan developed within engineering. ~ l, 'L 2) The Staff concurs with the recamendation. i ACTION: Dr. Roy will assure that the engineering recomendation My) is comunicated to customer servica, j"

i 2-12 Plant Computer - Alarm Recorder

~ The Staff concurs with the recomendation. Higher speed printers i ; ) i were. offered to customers. It is also reconmended that the customer be offered a special identificattor, capability for nuclear safety ( t i The broader aspects of alam/cperator interface will be alaras. incorporated in the control room design studies. I 2-13 Readout Ranges The Staff concurs with the recossendations. ACTION: Dr. Roy is to initiate a program in Plant Design. the y results of which will be comunicated to customer service for r implenentation by the customer. 2-14 Contairment Isolation Systan t The Staff concurs with the recomendation and notes that we 1) have provided recornendations to all customers on the isolation [ criteria. l F Q I L h 's j = - - - - - ~ 7 ..y-= q. ?? " -, t ..k r 4 1 .? v--- + _ _ - -.,, - - - - ~ -, -..m ,,-.,y_ .--7-,yy_ , _ _ _ _,.ww ,-p7,- -,

^+ . =.,.; ' ~ ~ - ..... ~ _ i k.6 l i . b.., t 3 ) R. M. Ball to J. H. Mae illan 7 Decenber 7,1979 f. w p g I I.h t v igy i ACTION: Dr. Roy will determine that we have defined.which [.-P

1 lines should be isolated and under what conditions and what lines should not be isolated for each backlog and operating 1

t plant. 3-1 Comunication Between NPGD and the SITE 'C x

2 ne l

The Staff concurs with the recomen'dation. ![ iL ' - - - - I ACTIOM: R. E. Kosiba is charged with responsibility of putting !L -l in pTee an implementation program.

T-

'l 3-2 HPGD Organization for' Site Emergencies t-The Staff concurs with the recormendation and notes that the Babcock Emergency Site Tea:n (BEST) proposal (per D. W. Berser) j! will help define the prcgram.

k l

ACTION: R. E. Kostba will prepare the NPGD plan. 4-l' Factors which Influence.the Product e The Staff recoccends the rejection of the.reconnendations as too general and institutional. The points raised were discussed by i the Staff relative to fGGD Operations.., 6-e t e s ~ L 9 .O R. M. Ball. Manager Product Developnent e ( m cc: D. H. Roy 'k E.'DeCarli l A. l.. Mercado l l R. E. Kosiba r l D. W, Montgocery N. S. Embrey [ p D. W. Berger [ J. C. Deddens i E Attacteent ?; s NOTI: The Final Report of the TMI-2 Occur *ence Technical Review Comittee is E. a " Company Confidential - 1.fmited Distribution" docunent. Section 2 of j that report (attached) should be distributed and handled on a strict a "need-to-know" basis, and not disseminated outside of B&W. [ s 1. A F ~. .--;,-y-.---- _t--i r ) l l.

~ 1 d _- - " M. -- jpydg, - =" i ! C;1 ' c. 3 ~ [.q 1 E.c UT \\ 1 !SM i E. "l s n. .e 6 : f-l i P i I SECTION 2 -.h y q . i - j [g I, t ] This section is a compilation of the "Tindings and Recommendations" p i of the Technical Revicw Comittee (TRC) grouped by categories.. l } r = 4 t 3 ( " Findings" are statements of fact or conclusions drawn from a p o collection of facts. d d 4 s. .1 o "Reconsendations" are statenants of action which the TRC believes i W ] are the appropriate response to the finding. [ I 1, ? a In some cases the recomendation includes the establishment of a program to i lenent the recoe:endatics. These prograns are j j numbered and are included in later seccions vhich have assembled ~ e J l the prograns into related' areas. L 9 J b 1 ( F 1. 1 t a I, t e ( j 9 1 1 .. -- 2 -4 W I e t -- ~ " ~^'

.-,..s.- .:. ~ n. t.; ~ - +. - - - 3 j

f. G4

g.

FII DD CS A :D REco S!ENDAT70 3 1-1 (r-& i CD:TIROL R00:i/0Pff.AT01 IL'IERTACT. 3 a. Q \\ 4 gy'. l 1 s 3 Operators were apparently unable to understand and interpret h;- j Pinding - l plant condition from svallabic control room information. They 4 had no " mental insce" of the 555 system charseteristics when ' E Q boiling occurred in the primary loop. Their analysis of l 5 system conditions led to action decisions which caused the incident to result in core ds= age.

w-L.

l [ e.

• Reconsendation'-

1) Use systematic esgineering tools to extend transient I

4 j analysis in scope duration and events considered. l l P

2) Improve control room design through human engineering r

. j' .to better couple the operstor with the infor=stion 1 available and important control situations. Provida integrated data to operstor en system status for [ selected syste=s. Use dispisy techniques to focus t h the operator's attention rapidly on the most important conditions. I F 3) 3 ' Provide additional infor=stion to reduce operator need for implied analytical decisions. Examples include: j a) 1 1-r1 r vs. r,o,. m b) Positive flow indicatio.ns hs opposed to valve [ position indications. t i lQ *

4) Provide greater input to operating procedures developed J

by the utility to correctly direct the operstor during F both noresi and emer.gency situdti9ns. Prgvida,means to q assure thst' all operating procedures *sre consistent with L the capabilitics of pisnt equipment and assumptions made [ in safety analyses. Insure procedures are designed to p 'l l saxinize probsbility of correct implementstion and ) execution. (

5) Provide enhanced basic funds =cntal knowledge and

[ l understanding to operator through training. J [ l

6) Evaluate vorth of additions 1 improved automation of t

l nornsi p3snt operation ersding operator oversight assinst [ boredos and insttention. Increased cephssis on the trade-offs betvuon operator response to upsets and conclugencies E versus autonsted respense with littic or no msnual over- ' t rides required. ~ t 3, t 6 b h' 2-1 t r ^ = ~.. 7... ._a 9 j i

=.-.

...:.~,.1 : --:_ _

~ s i } J O g' t I"M c (ttm. t h.8T*\\ l rIMDICS A*:D RECT *NT12t3 1-2 I l ,l t orr-rom:at. sTsms m: wists = .g

j

.j .) All studies and investigations to date have concluded that 'P Finding - the TMI-2 incident resulted from a conbination of factors as opposed to a single mode failure. F w. ,I i Recommendation -

1) Extend studies of*off-normal sequences to identify h

impro tesents in components (valves, etc.) instruman- }- 4 tation, control nodes, and system configuration which will reduce challenges, improve system reliability. t F and provide greater protection against the progression 2 of event sequences to uncontrolled conditions. The [ recommended approach is a mechanistic event tree meth-odology which provides a realistic evaluation of system k response to event seguances. W i 6. f.

2) Improve the reliability of normally operating systems F

and components to reduce the ferequacy of initiating events. Particular attention should be given to feed-i water syste=s and electric power supplies. t

3) Emanine testing and maintenance requirements and procedures to assure positive impact on system and component avail..bility. Particular attention to task F

analysis and human engineering for maintenance is [ suggested. r e b L r j ? L. L i I ' I 3 { l t i ~' F . i l 2-2 ~ q 6 8 I .,, c. w. -, ,e, b,,,.-m i h 6 s....

r .n... g_, .L ~... T.. 1.-t Q }.Q f -]- -i . v ^ M147 I, CW r TI?:DI?:CS A "D Rtcor'*"*: DATIO?tS 2-1 .4 p ,-} REACTIMETI2 J V.C '= 1 g .e h Much of the understanding of the TMI-2' accident is based C finding - ], on data supplied by a special recording system (The [ "Reactimetar") which was in operation for experimental ,L purposes. g h ? I Recommendation - 3&W should design and install at each plant a recording c. system - to record key variables which can be queried remotely. Typical systc=s have recently been acquired a'j for the Custo=ce Service Depart =ent which can sample data. - I . digitize it and put it on pemanent records - such as tape or disc. This system can aisc transmit the infor=ation over i } phoos lines to re=ote locations. Transiant data recorded i should, in the future, include all injection flows to the 5 primary system and stean generators. J s y ~ ~ ~- l ] t V L 6 -i I a i [ t -I I ? J -1 rf C J., r L ],.. I t Y, r-ly I 5 s ,h ~ '6 r 23 ig ,F em a r v =- w g - Mer -w WN, -,- -w-- 3=m-p,. -. ew -'EW-% _-{',"r, m-- N r .----w no u- "_ m 1 I >] S m t l, i, bh .=8* e

_,..__.._a.--.___ --c?r. m :-. 3 l__ r sk WI i E' l t ~ p aq r

. J.O-A'

~ 1 L -c .r j ( e TINDINCS A' D RECC'0 C:DATIONS 2-2 P11.OT OPERATED RELIEP VALVE, ( O' 'm I V Finding - The Pilot Operated Relicf Valve (PORV) opened as designed at the actpoint of 2355 psig. The PORV was designed to t reseat at 2205 psig but did not. The valve renMned open 3 E.- until isolated by closure of the PORV block valve 142 F W U minutes into the transient. I[ i j ]' d-Recoc=endation - Determine alterations,to the B&W systen that would alleviate or eliminate unwanted PORY action. Items to be considered l in future or present designs include:

1) Change in reactor /PORV trip and set points to reduca j f 3

probability of PORY being required. 7 g. j

2) Provide positive valve position indication for PORY.

L.' [ { E

3) Provide auto =atic isolation of.PORY by closing PORY f

block valve on low reactor coolant systes pressure. t, ( ? l

4) Provide positive flow indication and alarm for r

j ) auxiliary feedvatar and PORY. [s 4 6 l s \\ h

5) Consider elininstion of PORY on new dest,gns.

t

6) The causes of the failure to resent should be established i

and the cause corrected. Any redesign valve should be [ . qualified for cyclic secan, two phase and liquid flow g-under condition: 'siculating the actual conditions. p t

7) Add anticipatory trip for loss of main feedvacer.

t r e b' .4 y E E y 2-4 I l ~ Er .:n _ ;1 l r

~. - _ _ . ) se r e s {- % = Y W i t e 3 0.3fri.. g; - q k' L-A e FINDIMCS AND REC 0!NE*: DATIO."S 2 'l [f = ) 'J M 1H FEEDM EER REI.MBH.I'If 2 g i 1 VY I Finding - The tripping of a condensate pu:sp resulted in the loss of [ ] all main feedvater. Retrospective review shows main Di upsets lead to 3-4 reactor tripa per reactor year. t. i Recoc:mendation - Improve reliability of the main feedwater system to reduce y 3j the frequency of plant transients. l i i Items which should be considered includes ] f.

1) Desig;: such that single fault will not cause the loss of all main feedvater.

"[ p Provide greater fault tolerance on new designs. J 2)

3) Review guidelin'es for the conduct of maintenance

+ '1 J during system operation. r i t Reference - Program 2-3-1 L 7 3 / ' J t -] r b l I ] r e i L I 2-5 ,3 e 3 u a a-. D ( l I

l - - - - - _.. -... 1 ~ 1 En 'l ..41

i

'i } - i j - Q .., tm

• 5 5

".r*hw 0 G. Ci FINDU:CS A?:D RECO O!CMTIO*:S 2-4 I 1 W Auxt!.;ARY FLUID SYSTCf3 EVALUATIQtt' n l Q = c r r I FDIDIEC - During the TMI-2 incident the letdown flow path became ?" i partially clogged, although flow was sufficient to balance the seal injectica flow to the RC pumps. ( E, i' i F Rece-endation - Review the design of auxiliary fluid systems to iden*1fy y components subject to clogging and provide improvements l where they are needed. i u I { (- I w i.fe,ence - rro.,- 2 + t i I. 4 - e% T p: L". t. 3 l-e s

• g P

1 ]' l' .i 'l f i j 1 i l i L r i L L L i L 2-s t L. r t -'~~*'17 y -. _, e 9 m'A S* r S 4 4 9 - - ~ ~

.j ~ a 3 l e c

us 3

m e 1 l;m a:

• .:e I

FINDINCS A'm RF.CO*C'UDATIONS 2-5 I. - VALVE STATUS E I .I. ' $ g: l p 2 l Finding - Contrary to the plant technical specification, the aur411mry [" feedvater block valves were closed preventing delivery of (- -) anuf1tary feedvater to the steam generator until the operator discovered the valves were closed and opened them. k",.. h 1 ( Recommendation -

1) Provide status conitoring instrunentation to indicata

-I the operability status.of all safety systems and alarm when the status is incorrect, including all valves 5 necessary for system operation. f il

2) Provide automatic positioning of all renotely operated valves to the operating position when a system i'nitiat-

} ing signal is present. k

3) Provide self-checking instru:sentation and alarms to

= ,9 verify that the system is delivering flow to'the staan I .i generator following system initiation. I . 4) Review functional reliability of all Aly systems. f, 1 4 l S q i Reference - Program 2-5-1. i i )q E 7 i i I h <f J J d C w-1 Ef E v i E 2 1-1 e I i l .t 3' -wg= -ms- .. _ ~. 1 i i-i

. -.. __. a -.. _.3- - w

• e Qi

.{ 0 l

.l.$..

(-. tm. r-i :- $NT ' j '-P (e ,1 E 4 l I f l FII:DI!!CS AND RECO?cT.'DATT0!!S 2-6 l l l o t VAPOR / CASES IN THE PRDIARY SYSTIM "- [ 1, t,e

• O l

r i lt .f Finding - Void formation in the core, venting of the pressurizar steam bubble, the collection of steam and noncondensible gases in the reactor vessel dome and upper parts of the ""i [ A J-legs were.found to be phenonena which can result in E degradation of heat removal and cora cooling. E F.' s 'my f t r Recommendation -

1) Develop an improved understanding of the modes of

,i T L heat renoval when forced circulation in the reactor .. _... coolant system is not availabis. l [ g. a). Investigate the therr.a1 hydraulics of natural circulation in single phase water and two l phase mixtures and in the presence of noncon-i, '~ densible s'asas. b) Define the espebility of present decay heat 9 renoval systens which are operable at high g pressure and the conditiona under which they are effective. m. E ~ ,g e ' Steam generator cooling with main, and a~d"=7 feedwater. lv v Righ pressurizer injection including the recir-h~ F 6 culation node. g

2) Deter =ine C.terstions to the 35U system that would E

improve the heat ro= oval capability under.off-nornal N [ conditions. B a) Means of venting reactor done and top of J__less. i b) Hanns of increasing thermal capacity of secondary [ system hest sink inventory to increase time for [ l operator to react. g [ I j L l h L 2-8 I a e8 i J i l 1 \\ f L

_s .._. - ~- .my s%4 i et 1 FINDI!!CS M D RICC:2 FEN 3ATIONS 2-4 Page 2

- -(Go

[. FT. e l M g l' Recom:acadation - c) Frovida instrumentation to detect and provide ] (Cont'd) direct indication of inadequate core cooling. .O 1 d) Investigate means to improve the inherent systes response to anticipated transients. e) Y.eens of nonitoring natural circulation capability [ ] of the reactor coolant system and indicate when the capability does not exist. i '3 f) Instrumentation to confirm that adequate natural [ 4 circulation exists. l g) Define i=provements to post trip control and system response following plant trip. g L h) Provide an anticipating reactor trip on loss of 3 } main feedwater. g 1 i Reference - Program 1-1-1 l a' L: .] o i r F r !] 3 i 3 1 l l } i j 1'j l q* e 2-9 ~ i i' es ? J.m,"k l c ll [

-- '~" ,WI ..m.- ?.. f i .-:. g. \\ ~ Cm i FINDINGS AND RICGCTI':DATImIS 2-7 DECAY HEAT W.07AT. V l, H e cu=r. nc:o c b.. l es 3F I ,Pinding - I The decay heat re=ovat system is currently designed to operate with 12 failed fuel. Based on the 21I-2 incident, g

b. J it is desirable to evaluate long tern decay heat removal

[j.,, at higher activity Isvels. v. .Q. ~.K. !,~ \\ g Recommendation'- 1) Develop a decay heat removal system for new plants i r h which limits release of radioactivity under post-accident conditicus. ':his system should be capable C 1 of performing its function over the full range of ( q reactor coolant system pressures, and should, preferably 4 g bg independent of the normal heat transport systema. ) {.' 2) Provida connecticus on decay heat cooling system to f L allow enerzency connection of reliable alternate cool-ing system. 3) m =+a-te sources of' decay' heat removal system leakage k;o l or provide for collection and conn 4 - e of leakage. g.

4) Provide shielding to pernit auxiliary building access E

j during decay beat removal system operation with con-taminated c6clant. c I )8 l

5) Provide shielding or re=ote operation provisions to 5

permit long-ters decay heat renoval system operation E e.g. remote motor oil level surveillance and remoto y ( m ann to add oil if required. t L' L t; Reference - Programs '2-7-1, 2-7-2, 2-7-3, 2-7-4 }

i..

E. s p g 2-10 [ ( r e.-- w m w as- 'W"" A'* I 4 5 1

s. p-1 = - - - -.- l h* ~ ^' 1 .,u q i 1 -g,d 'l .. IR O ) p~;FT ' e 3 $..4 if

• c;t

.e 3 i i FINDINGS AND RECCCENDATIONS 2-8 9 c l \\1 _ n i 6 IICE FRESSURE INJECTION LOCK II J .g s ,\\ m.. J , 8 h Finding - Operator manually shut down EPI pumps following start on I.c

I low pressure.

m. 4 O j 9 i w Re-: - - = tion -

1) Consider supply of interlocks or permissive switches r

once L-which would prevent operator from stopping EPI'4 they were started by primary coolant low pressu t ^'

2) Define and limit the conditions under which an operator may terminate an emergency safety action once inititated

~ ... ~

3) Consider self-checkinginstrumentatica to verify proper system performance when an ESF actuation signal is l

present. t. Consider automaticpositioning of all reente operated l 4)_ valvse to operating position when sa EST signal is, 6 present. .*.s= ,-e. ..s .~ ---~ l c. [ 3aference - Program.2-8-1. L r I r u I i s i L i 1 ' r. 7 C E ~l [ t p r g.. i 2-11 [ ~ l E m __**Y ' L%.1 J 4 G 'l 1.

y. s.. j -O-3 ..g 1 9. v. v f. e ,~~';ttT.. \\ k ' t im I s ] g? FINDINCS AND RIGT. wATIONS 2 ' r, 'I FIESSURIZE1 LOOP SEAL L u. W r. r ' y ? e k Tinding - Pressurizar level indicated full during the greater portion [ f of the transient. The indication.was real. The pressurizar j L piping loop seal prevented the pressurizar from dr=4M,g 6. f T when loop pressure was higher than Psag of the pressurizar and the relief valve was closed. When the pressurizer relief [ valve was open, the insurge from the surge line provided a 1

• sourca of watar to keep the pressurizar fmLL.

[ p- [ L, Recommendation - 1elocate pressurizar and surge line penetration on new t designs.*to =14 4*=te piping loop seal. l j I i. ? e Refairen.ce - Program 2-9-1. 3 ~~ e u t i e ~ e t e =

• f*

= (

• ;.7-i l

t ~. l j E f .l L f 48 g y .SW 1 ie e t .u f t p

a f

• i k

2-12 L j y 1 (, s, ,= - _q v- -,.. -,, -.. -- ,-v .r e ~' o . 4 - =- m t ,._%c ,-.m. r,.7,- p., _,7s _Y_-.F, _.-o e

-. w.,:..~ _ .x l O..gl 4 i ta - (-

d. ',&,

'. i}Tj$ t L i yINDINCS AND REC 0!C D DATIONS 2-13 L, e*** l k RADIDACTIVE LIQUID WASTE BANDLING h -t g t ... j E .I Finding - The weste management system transferred contaminated reactor coolant from the reactor building suny to the vaste canagemene tW 1 {

• 7' system in the auxiliary building. The volume of coolant trans-I.-

~ ferred exceeded the capacity of the vasta ha m M system. l

• ~^
• i This ultimately resulted in a release of activity to the r

environment. L r'. i j ...u. tv..... s F ~ Recorneridation -

1) Provide confinement systems on new designs to enclose l

systems which handle radioactive. materials. L r . s )

2) Designers of vaste management systems should initiate

? prograns to understand the performance of the vasta I.'., management systems at TMI-2 and define needed i y....- i mests.

3) B&W balance of plant criteria which specify wasta

......".. volumes to be considered in radvaste system design . '., - should consider vaste volumes generated in off-i s. notual operatioc. t

4) Develop radiation monitoring systems and isolation E

criteria to prevent transfer of radioactive matarial I to a potential release point. o.' [ r e

... ~.

~. Inference - Program 2-10-1. t .j l ? J 1 g j ~ e 2 13 c t w, 4 e w t o r

~ [.. 'g t 1 W 1 ?[-l i. 4 ,j p.- O i Te4 ..'M 1 j ~ . 3 e E, a yDfDINGS A::D REC 0 NATIONS 2-11 s .C C-c, r s I IhTIRONMDITAL STRESS ON COMPONENTS .ytadiny - During and after the accident at 2MI-2, control and recovery l actione vers aided by the avsilability of information con- [ : carning principal plant variables. There was concern that l ;g. these indications would becone inoperative due to the g environment existing in the cont =imaat consisting of high r i; water level, high humidity level, temperature, and radiation. k ~ [ Reco m ad= tion -

1). Review the complete scope of the equipment wheae operation is i=portant to control and safe recovery.

i-to assure that the equipment is reasonably able to withstand the kind of environment which night be , created by a snell ste'an or ECS leak within the ', ( ca.. 4 t.

2) Specifically review the radiation resistance of 7,

starting capacitors on the reactor coolant pump notors . to deter.ine acceptability of these devices for op, era- - 1 tion in a post-accident

• radiation environment.

r = l r a a g + r .T. t -j y-I 1 c' 5 I ) u [ 2-14 c,

t.

7 s l r =. w _a ~ t 1 ~ c -- -.-..,-

g. :. '-

s'

• e

)t

k

ti) f g d

i y ? or i J. ' !, M + .I y ' 2 W r, a : O FDfDINGS A.:D RICO'0'J!3".DATIONS 2-11 P 4 s m \\ s C c g-n P r ENTIILON! ENTAL STRISS ON COMPONENTS i L . Finding - During and after the accident at THI-2, control and recovery I actions were aided by the availability of information con-1 carning principal plant variables. There was concern that l [ 3 these indications would become inoperative due to the t ) environment existing in the conthinnent consisting of high F. water level, high humidity level, temperature, and radiation. [ .i i t Racommendation -

1) Review the complete scope of the equipment whoes i

operation is important to control and safe reccreery. l t to assure that the equip =ent is reasonably abla to withstand tha kind of environment which might be created by a small staan or '.45 laak within the caneatn-ame. l-

2) Specifically review the radiation resistance of starting c'apacitors on the reactor coolant pump motors

[ to determine acceptability of these devices for opera-j 3 time in a post-secident radiation environment. l g g l g e g g b l E. g g r [ ..- -.... - i. l r l-V j [- i .y e I i t i 2-14 [ S t s p t m A

• n, 1

1 e- ,.-.,.,-n

a-m _-.xx .%.: -:ww. ' !{< .e u. U]61 ] ( ' OT FMtr ll. = d P % g., FTNDTNCS AND ltICCT. auATIONS 2-12 IC i I; / l-FUJET COMPUTER - AIAI21 RECCIDER 9 v' Finding - The alam typer of the plant cociputer system became saturated with information during the course of the accident and was unable to gather all the information received and to keep it - currently printed out for the infomation of the operator. _r p r- ~ o 1 Re - =d= tion - Modify the functional capabilities of alars typer systems to assure that they have reserve capacity for the storage F of data and the capability to keep th*e operator currently 1. .n.. informad at laast of those alarns which are most important. } ..~.:. ~ p L n (* '.g a e. l r F [ n l O 9 I g E. I p f j e I t I-E I 2-15 1 e 7

== - i _.- __ ~ .~M z__ e .-? - -~-- - -w- .y. ~ umb h=

• L, r

4 E 1.4 Y I

e.. .1..- : ' r ~maramg4h:.1 Q~~~}.;, ';.: ' - 1,. s

• ' ' ~

-%, w.- ~ r l [ p. p".7,'j te FDfDDICS AND RECOMMINDATICNS 3-2 t,!.-- l l i NFGD ORGANIZATION FOR SITE EFN"55" j t 'O t ,,e Finding - 'tHI-2 support efforts by B&W Lynchburg during the post-accidene racovery period could have been note effectiva r had an organizational structure and responsibilities been .k defined prior to the accident to deal with emergency -4 ~ e4 -. r Ese--dation As organizational structure should be established and responsibilities at least broadly defined for 4-=145 l [ with a wide-ranging set of ciressastances similar to t i 'tMI-2. Considerations should be given to the needs q for up-to-data plant configuration information and responsibility for main mining sad retrieving that e

• information should be defined. Consideration should also be given to the need for facilities at the sita
• to promptly acconnodate temporary personnal from NPGD.

s 7 .,.:,.c .' s e. 3eferences - Program 3-1-1. t b g g ... =........,. 1 : *.; = P .... ; * -:. - r.:u. - i (. 1 l i F I 2-19 l l 8 ) _3,. -N ---em g- - - ,5 ~ ' --- " " - ' - 4 C p-p N mh a __M w 4 .J 4 f'.

--s r -,- -, _p_- 7 _t s. g y. j t g + W f W.- 3' FINDINGS AND RECOPF.DiDATIONS te-1 I7 cm 1 .Ie 25 @ YACTORS WHICE INILLUCE THE PRODUCT

i 3M

. ({*; $E"" T' l L r; yinding - There are environmental and organizational factors inherent l,.. P .J in the nuclear industry which influence the end product. They are long-standing factors which have evolved over many k-q O years and vill be difficult to change. l [.,", c. (o. 13 f. Reconnendations -

1) Promote industry recognition of the influence these L,f [

b factors can have upon the end product ar.d support [n." industry efforts to develop and imple.ent improvements. {

e..

C

2) Promote understanding of these factors within B&W and I

'a f, 4s. take steps to develop and inplenant chacces.

h I

8 b t 4 ( 3ackground luforsation-The Comnittee has identified the following factors which are .I,. h' ~; *.F believed to deserve attention.

1) Inordinate emehasis has been olieed ueen desienint for, very lov erobability events. The regulations which

. Q[ strongly influence the industry call for equipment designed to prevent and nitigata against large 1DCA's, , 4* large steam line breaks and large earthquakes. The ' k.l d higher probability events which c'n and have frequently ({3 a T led to upsets in the primacy or secondary systems havn not received the same level of attention in the design and evaluation process. This would seem to be due to Ti the fact that licensability is a major require-g. ment and naasure of project progress, and the licensing 7 process has traditionally emphasized designing for what 1, was considered the worst or birkest and not the most s. likely. p B&W should support the developnent of a better understanding L. {, r( of conbinations of off-normal events and should take stepe to "close the loop" betv. a the system designers, system behavior analysts, procedure writers, trainers and plant k' - t'.. operators. This effort is further discussed in yinding 1-2. {' y s r

2) Strict classifiestion of all structures, syste:s. and P

.ic equiecent sceerdin? to satety function has tended to de-- f! 3 emphas1=c tne te?ortance or itcas not having a safety I p classi'ication according to the erescribed regulator? k R' definitions. The regulatory enpnasis on items which '( h protect cladding integricy, reactor coolant s}ates boundary g p i F } i 5 i 3-2-20 L c ~ ~ .,7 ---.e-s -"'T = m .M -._s 4

• e'

... t _~~_ t : -*...:. u.w u :..: ;..:. . _;..= _ _

• v ~A j

p-, c g. j s y.' y.- gp l 4-1 yactors Ehich Influence the Product (Cont'd)

2) integrity or mitigate against accidents has tended to establish a "nind set" which classifies various parts

.j u s. of the plant as important or unimportant in terms of O the safety classification given to that iten. Industry operating experience prior to IMI-2 and IMI-2 itself underscore the safety significance pad important role j played by equipment and systems cur. side the traditional [E"*~ safety classificactions. ] Bei should sup;nrt a reassessment of the safety classi-a L fications and safety functions perfarned by equipment s b outside the current safety classification definitions. ~~~

• Equipment installed for safety purposes has received excessive attention in the regulatory process while equipment and systens which can initiate a demand for safety equipment performance has received very little

[ ,,,, attantion.' " -e. M ~'~ ~~ .... r i Bei should support and propose to its utility customers, s programs aimed at minimizing the challenges to safety systems; i.e., preventing upsets which can lead to safety [ systems operation. This may require involvement in areaa i ~ of the plant which are outside the NSS scope of supply. i '~

3) The manner in which the QA proeram is inclemented empha=

i'

• • 'C sized fots rather than substance. Regulations require I

that'a formai progran be defined which is auditab1.a in j

• detail. Considerabl,e effort is required to proevce the requirM paper work and make it traceable. The ERC in-r fluence has caused undue emphasis on the process rather than the content. The auditors are generally not quali.

fled to evaluate content and hence the audit findings must j frequently deal viti; form rather than substance tending to f de-esphasize real quality. B&W should continue to empha-L i eice the positive aspects of QA programs and the promotion of. technical excellence. Efforts should be made to c I simplify the QA procedures. for clant d'sian is subdivided acons E

4) The responsibtitty e

the A/E, the utility, the 'sSS vendor and the NRC. This leads to a tendency to treat systa=s interaction problems with less e=phasis. For example, interactions between BOF and N55 or between safety and non-safety systems or between [ the man and the e.achine are, in many cases, also scope-of-supply interfaces. Coc=ercial pressures have tended to L inhibit involve =ent by the ::55 vendor in areas outside his p 1 1 e 2-21 r. i. l ( _m- .. -~. _a ma t I e

~. .~-......:... - _. 4 .- y . -i. c.=, g l u -'@ l E.M ,/ 3. 4-1 Factors Which Tnfluence the Product (Cont'd) 9 O +. 1 y4 I'M [9, f;s #

4) scope-of-supply. It is nou apparent all pqrties in-l volved in the design, construction an1 cperation of 1

a nuclear plant must increase their understanding of '. ' ' F ] these interfaces and the general matter of systems ( V, intaractions. . r. p i.s Thu B&W NSS has been co=hined with too many different (,. - h 7 p balance-of-plant (30P) designs for which B&W has little or ao responsibility. i.- ? w.~ i

1. p_

It is ree - dad that ^1) For future contracts. B&W should require standardization r F @Q of those parts of the 30P which are crucial to the operation of the 205 plaats. 6 [I

2) Da existing and backlog plants whereia no wholesala

${ changes to the BQF are possible B&W should recoasnand g r. or propose a design review of the BOP to assure that A satisfactory system operation can be obtained. . p, p t

3) 3&W should establish agreements with' customers allow-y ing approval rights over future changes in the 30P.

h.- i a

3) The influence of reculatory and commercial issues esa h

I inhibit the develon=ent of a full understanding of an (' issue. Frequent escalacion, and expanded interpreta== i. tion of regulations and requests for more information C by the regulatory agencies tend to restrict investiga - g b tions to the " -~~. required to meet the requirements. ' r As mentioned above, cos=arcial issues also promota ( marrou definition of a.given issue.- c 6 i r B&W should promote, through industry working groups. i efforts to reduce thm detailed content of regulatory,, (* requirenants. l r B&W should support self-sonitoring by industry. p As mentioned above. 3&V should also promote broad undar= L 6 standing of plant perfor=ance under upset conditions even E i though this =sy bring about involvenant in areas outside i the normal B&*J scope-of-supply. [

6) Internal oresniestionst structure cac influence the t

product desien bv excessive resoonsibility subdivision. f thereby crest:n? tne need for excessive interfsce control i ( ard the 1 J.-:'ive:3nt o: too a:2n ? nersons in trie resolution .f of any tiven orcbles._ la teviewing pre-Eil events, \\{i ' L 1 2-22 t (t r i v- ..-y%=m*-, x g -,.,- -.. - nr--- A L_ 4 j 0 v.,nww,- v-. -w ,-2m-- w e w------ --e

.~,-.. -.. _ _ _.... t-m ; _- -.c. i il s- - m 'i 'Ml V< 1 4-1 Tsetors Ehich Influence the Product (Cont'd) 3 ../ QTt' I questions have been raised about the ability of the NPCD 4" P } to quickly resolve technital questions which could impact N the design or operation of NPCD-designed nuclear steam [ systems. Thus a review of the organizational interfaca procedures and attitudes is in order. I. ? ? / Consideration should be given to using an outside consul-D the problems that exist with the organization better than (E. .._O, tant-to perfom a review. Although NPCD personnel know l '[ ~ l s anyone else, an outside overview nay sort the data diffe.rently { and reach different conclusions. If used.' it is ' suggested t" j the outside consultant review the following typical questions: i

1) Is NPCD organized in an optinun manner?'

E..

2) Is NPCD oriented toward the business as currently

[ defined? C.

3) Are responsibilities consistent with nrnagement

} j. ~ .5' objectives? ' 's... '.z. v

4) Is responsibility for safety adequately defined

[ and clearly assignedt t r. 1 [ TEC recommends consider 4*: e organization i j t .,- The responsibility for plant safety and the over-sight of safety issues v1 thin NPGD is not as clearly defined as it should be. The resolution 'of safety issues has, at times, taken too long and the cocpletion of resolution action, partien-larly where customer action is required, is not [c assursd. - Consideration should be given to forr.ing a safety [ l review group co= prised of canagement personnel to oversee and audit the processing of safety-related g. issues through to couplete resolution. The purposa of this group.vould be to assure cinely and appro ; priatd acti.on on safety nattars, t Procedures Civen that responsibilities assigned to mnsgers and others are clearly defined.'a review of the Procedures should be made to provide for core judgement consistent [ vith the =inicun QA requirements. This rev$ew abould include: a) A basic review for st=plicity both in structure and inple=entatica. j k 2-D I -ree v'w.- .___a__

• 'W77*7..

v o. " ~ ~ ~ y_.. , __ rl * ) 8 b I

.y.:...if Q

• _

- ~ ~Y * "" -

3 F.

% ~ --.

) :. c-m, R* -C'- W; 'h a; l T.: ,( @ l 4-1 Factors 1;hich Influence the Product (Cont'd) i

.y k '; W.-

l" $y'- Y p-Procedures (Cont'd) 'm b) A review to assure that the e.anagement philosophy I [b d reflected in the Procedures is consistent with the ,[ j Division's management objectives. This item could 4.. b be reviewed by the outside consultant, if e= ployed, Q l by means of a brief audit. . {.-

,]

c. c) Additional emphasis on providing operating guide-9-- Y' lines to first-line managers for their units' b{ 'I; activities. ,( I d) A review of all NPCD procedures should be made to determine the need for clarification and '/2 [; strengthening in regard to the handling of safety-related issues. Procedures should be revised to call for closure of a particular issue to assure ?* required actions are completed and/or implemented. ( ? f'

Q.

Definition of responsibilities for safety matters Gi [ should be reviewed and clarified as needed. -* x - i h. a UC m. c.h.

• g-E M

, &-t }' C rN; _{ f , y,3- >:u. e..v.

• h

('5, 5.% g 7.1 L'3 t.3 t 6+. c .h y l 4 r., i, i g g. ~,4 fy v:X Mt V ao. b N f k3 i i 2-n p a. c 7_' ~ p.' -,g ,,,,_ y --,, -, _ - y-- m mmg-p.w,-- ,v,y p_ v y ggun eg-,-,---( e Ag___- A _ s a .E f' a $~}}