ML19219B001
| ML19219B001 | |
| Person / Time | |
|---|---|
| Issue date: | 01/28/1981 |
| From: | Murley T Office of Nuclear Reactor Regulation |
| To: | Harold Denton Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML19219B002 | List: |
| References | |
| FOIA-81-417, TASK-A-09, TASK-A-9, TASK-OR NUDOCS 8102270733 | |
| Download: ML19219B001 (3) | |
Text
a 4r 4-o-
S D
4 Distri b ticn,/
Central ;ilev :.Saer SHanaLer AI~ad:ni JAN 2 S 331 SPEE File iciel O e" EEi S *t "0055 Tvarley
.Narncn vErnst JPanciera
?TLTFAfPtF FIF:
k artle I. Perten, Cirecter, Cffice of t ericer Terc*cr Peerlatice TFCl*:
Itoras E. t urley, Directer, Civision cf Fafety Tectnclecy, ' TF SUCJECT:
CST PEVIEk CF FFC STITF S2FETY Ev/LPATICr FEPCET CF EbF SCPtr DI! CHARGE SYSTEF Pe f erence s :
1.
"em reneur fcr C. Laines, T. Foval, P. Teresec, UE?TC frcr F. Check, OfEPC, FSP Screr Discharre Systar Safety Evaluatier, recert er 1, irrc.
v rcrencun fer Farold F'. Ernter, USEFC f rer ". Ernst, 2.
eUStTC, r!T Evelnetier rf tFe / eteretic Air Fraerr rtrr cr foiling Unter Teacters, Decerter F, 1PPC.
2.
/.nticirated Trensirrts UitFcut Screa for tirbt Peter Peacters: Eesciutice of l'rresolved Safe +y Isser TIP A o, FLTEC-CzF0,l'ElIC Steff.
Tie Civisicn of Safety Techr.clery has reviewed the fIC Stef # Ezfatv Evaluatim
?crcrt (SEF) entitled 'EFF Scram Cischarte Syster Eafety Evelvetich," of
!!ece-ber 1,19FC (Fefererc.< 1). Ve tave previcusly (Feference 2) trevided you with ccr coments en tre instellation ef an interir auterstic air "reder cun syster recorrerded in that safety evaluaticn. We fcurd trat tFe additien of this syster cculd provice en irrertert increase in eve safety.
Cur corrents en the re-ainder of tre SER ere civen telcw.
The SER rreposed a functierel criterien and severel safety criterie, creratire criteria, design criteria and surveillance criter19 fer the EUR screr discFerre syster. The inter.t cf these criteria was to previde cuiderce to FLP licersees to assure that their screr eischerre systers Trevide the decree of safety tbcught to be previded when these reacters were licensed. Father than rrc-cosinc specific solutiens tc the prcblers which were identified durint the review of the Browns Ferry L' nit 3 partial scran event, tFese criteria ere to te used as guidance te allow tre EUR licensees to resclve the problers in e way best suited to their specific desiens.
These criteria were first prcpesed by a surcomittee of the Cvrers Grcup fer ce to werk with the f:PC cn this problev'. The f.PC staff rede cre sirrificent advi-tien tc the criteria by recuiring tret the scrar level irstrtrentation be designed fer corren cause f ailt;re as kall as rarde'- f ailure (by the aeditien cf diversity into the desien). Le agree witP tHs additicnal criterien since cer.cn cause f ailure reces are, in reneral, sienificart centributers te the
,,I everall less cf functicn and aise since irsterCEs Cf COPen Ceuse failure of h
this instrerertaticn wera actually cbserved (e.g., crusted floats on level 9)'
f,; i~
instrurenta tion).
, m k.{ j t
5 pe p& U b u W.j d D W;rA;i d
"'20 t.
q.
.o v.m q l
. s *E k j
.e OFFICIAL RECORD COPY m & n,. c
.s ; " "
s-m mr v7 2"?
ry 2
4 tr
<?
e s
d 16 ta Harold R. Denton
-Z-y y 2 g 1991 ns noted above, the SER criteria for the Scram Discharge System were divided into safety criteria, cesign criteria, operating criteria and surveillance c ri teria.
In at least one case it is not clear that the civision of tne criteria into citferent categories is useful. Tnis case is Cesign Criterion I wnich statrs (in part):
The scram discharge beacers snall be... hydraulically couplec to the instrumented volume (s) in a manner to pernit operability of the scram level instrumentation prior to loss of system function.
This appears to be a safety criterion as much as a design criterion. Care should be taken to assure that all criteria are met regardless of whether they are called safety criteria or given another cesignation.
ke have reviewed these criteria and find that they adequately acdress all the issues raised curing the investigation of the Browns Ferry event.
The SEri lists appropriate ways of meeting these criteria (called " Acceptable Compliance") which will minimize subsequent staff review. he have reviewed these methods of complying with the criteria and find these to be appropriate with the exception of the two alternatives listed for acceptable compliance with the level instrumentation diversity requirement (pages 39-40 of the SER).
The first alternative requires level sensors ecploying a cifferent operating principle for automatic scram from those currently useo. The second alterna-tive requires that the cause of damaged floats be identified and corrected (or compensated for) and goes en to require either:
1.
Acditional (or substitute) level switches for automatic scram wnich are made by a different manufacturer; or 2.
Demonstration that the diverse indication of SDV water accumulation pro-vided to the operator by an alamed continuous monitoring system can pro-vide adequate backup protection if the level sensors (providing automatic scram) should fail.
(The operator's response to a fast SDY fill event must be considered).
,ie recommend that the BWR licensees not be of fered the alternative which involves operator action. A precedent for our reco::rendation is the ATWS requirement that no crecit for operator action be given for at least 10 minutes (Reference 3).
In addition, we consider it doubtful that the licensee could show that the opera-tor could take timely action based on a level alarm. Therefore, we question whether this second alternative is really a choice at all. Furthermore, what choice would the operator actually have.if. the automatic system had not scrammed His the reactor but the operator had received an alarm f rom the backup system?
Since only reasonable action would be to manually scram the reector imediately.
the reactor operator's first action upon receiving the alarm must be to manually trip the reactor, and he therefore will not have tir.e to diagnose the cause of the alarm (e.g., to determine if it is spurious) wnen the reactor nas not tripped automatically, the alarm does not seem to provide any benefit to the licensee.
- nui l
\\
)_
l l
.I.
m m >i l
f I
l
.l..
p 3
u r r m n m w:s. c:"
OFFICIAL RECORD COPY -
~ ~
w Harola R. Denton AN 2 81551 The first alternative of two automatic scram systems based en sensors using different operating principles seems to us to be a better choice than level switches nace by two cifferent manufacturers because of ne even greater diver-sity inat it of fer3.
Because of the importance of this systea, we recomend that the licensees not be of fered the choice of level switches made by a dit-terent manufacturer as an cption to the first alternative of two sensors based on dif ferent operating principles.
Experience with a sensor cperating en a dif ferent principle than the ;urrently used level switches has been limited so f ar to ultrasonic techniques.
There have been nany cperating problees itn systems using ultrasonic transduce 's but pro-gress is being made in solving then.
In addition, at least one other c1 verse technique has been proposed.
Therefcre, it is our opinicn that the first alter-native of two civerse level sensing systems basec on dif ferent operating principles in the scram discharge system is preferacle.
.le cid not atten.pt a cost benefit analysis of methcds of acceptable ccepliance with these criteria; however, it is our judgement tnat such an analysis would that the safety benefit of implementing these criteria would Justify the shcw cost of implementation, o wat sign.s by
-r Thomas E. t;urley, Director Division of Safety Technology Office of Nuclear Recctor Regulation a(
\\
- r>
DST:SPEB DST:S B D
AD i Djh1 l
! TMarley
,K)nst Th81 5~M RLobel :ao RLBagr-1N S1 l$'31 lQEl
'~
CFFICIAL RECORD COPY m r:
e c. m: v..;