ML20126A758
| ML20126A758 | |
| Person / Time | |
|---|---|
| Issue date: | 09/13/1979 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | Advisory Committee on Reactor Safeguards |
| Shared Package | |
| ML20126A760 | List: |
| References | |
| ACRS-1641, NUDOCS 8002200042 | |
| Download: ML20126A758 (43) | |
Text
o 2 -
\) /qC/O- M W gh \k\g .f. Q DATE ISSUED; SEP f M\
b'.m5 $.\ 'h- QK';)
\
MINUTES OF THE ACRS 1 3 7973
'd
\ .3 k e fis LICENSEE EVENT REPDRT SUBCOMMITTEE
'\"6 y pG ,c D WASHINGTDN, D.C.
MAY 24-25, 1979 The ACRS LER Subconmittee held a meeting on May 24-25,1979, in room 1046, 1717 H St., N. W., Washington, D.C. The purpose of this meeting was to continue the ACRS review of Licensee Event Reports (LERs) submitted to the NRC by operators of commercially licensed power plants, the NRC's role in the analysis of the LERs, and tne means for using this information to institute improvements in nuclear power clants. The -
notice of the meetir.g appeared in the Federal Register, Volume 44, on March 23, April 20, and May 11, 1979. Copies of these notices are included as $ttachment A. A List of attendees for this meeting is included as Attachment C and the schedule for this meeting is included as Attachment C. Selected portiens of the meeting handouts are also included as Attachment D. A complete set of handouts has been included in the ACRS file. No written statements or requests for tiae to give oral statements were received from members of the Dublic. The meeting was attended by Dr. D. Moeller, Subcommittee Chairman, '4r. H. Ethering-t on, Subcommittee member, Dr. J. C Mark , Subcommittee member, Mr. W.
Mathis, Subcommittee memoer, r. J. Ray, Subcommittee member. Dr. R.
Savio and Dr. A. Bates, ACRS Staf f, and the ACRS consultants, Mr. J.
Arnold, Dr. R. Burns, Dr. I . Catton, Mr. S. Cromer, Mr. S. Ditto, Mr.
E. Epler. Dr. M. First, Mr. A. Grendon, Dr. R. Seale, Dr. Z. Zudans, and Dr. W. Lipinski were also present. The meeting was opened at 8:30 a.m.
on May 24, and continued tnrough 5:30 pm and was reconvened at 8:30 a.m.
on May 25, and was adjourned at 3:00 p.m. on that date.
EXECUTIVE SESS'ON The objectives for this two day meeting were reviewed oy Dr. Moeller, Review of the Licensee Event Report Subcommittee's draft report and the review of the individual consultant's reports were named as being speci_fic goals. (See Attachment D). The draft report of the Subcommit-tee was to be revised on such a schedule as to per. nit discussion at the June meeting of the LER Subcommittee. The discussion of and the follow-through of Dr. Okrent's comr.:ents (See Attachment D) was also named as a specific goal for this meeting. Comments from the members and ccnsul- p tants were solicited at this point. These were as follows: Q '7 -
8002200
_.-_________-____m._-_-_m-.
t 1 LER May 24-25, 1979
- 1. The imoact of LER experience on t're Tecnnical Specifica*.1cns was discussed. :: a;; ears that the Technical Specifications are at least cnanged ;c accommodate whatever nardwa re modifications mign: ce made .ts a result of action on LERs. This matter is ::
De discussed at a future time.
- 2. The advisaoility of imposing penalties (fines, use of tne da:a in a rating syste , etc.) on the utility for the events wnicn are re;orted in LEDs was discussed. It was noted :na; fcr the
- - re:Or ing to De *ne most effective, it neede: 10 be cb;e: 1.'e.
1 Tc mai.e it self-incriminatinc wo.i d cecrease its eff e::'venest.
The Succortittee, opinion wa s that the issuance of penalties sho;ld rest with !&E and should not be ass 0ciated wi n LEE < ,' ,
i re:or ia; 3:: 1.ities. ,
1
- 3. It was ncted that the LERs snould address past occurrences anc the rotential for recurrence of the reported event, and tha; l i
s e atte :: as ic :nc nei;ntin; ;f tne sericusness of tne leu !
l event would be usefsl (e.g., a 1-10 rating systec). It was noted :na: the respcnsio11ity for mak ing this evaluatica should no: be assi;ne to too many cifferent individuals.
- a. :: was noted ina :ne cos;-cenefit as;ects sh0uld be consicered when specifying :ne remedy for an LER event.
DISCUSSION OF THE DRAFT REPOPT, ENTITLED " COMMENTS ON THE LER REPORT '.3 ,
S Y ST E M",,,,
0 ATE D 'dM 5,. 19 7 9, The subject report was discussed by the Subcommittee (See par: 1 of Attachment D). The report contained nine comments which were discussed On an item-by-item basis by the Subcommittee. A summary of the discussion is as follows:
Comment 1: Use of the LER system. The Subcommittee recommended that the purpose of the LER system needs to be clearly identified oy %k the NRC and recommended that one group of people within NRC, 4
f
'. . 0 .'- 'a. . .v74 .? 5 1 :.7 C.
consisting crimarily of representatives of :ne cffices using the LERs, should te assigned tne overall responsicility for tne LER systen, ne needs of :ne syster users. anc the types of uses to whicn the system snould be applied. In tne discus-sion of tr.1s comment, 1t was noted that this LER group snould be se;arate from the Inspection and Enforcement organization.
Tne im:or:ance of ge;;ing :ne results of ne LER analysis to One e:ul- en* des':eer an: n;1e ented int: nis ces'g was stressec. ;;in" ors ser exgressed :na; s;e:ifi: :r1:2ri a raj be ne:essary if :ne fixes suggested by the LER analysis are ::
be im;1erente: in a '..ure des 1gr. 'ne infor ati:r generatyd
- , ice ? ref:-;4 3; a"d analysis shesid alsc te available' to
- 2. - ,; ..3 -4 5 u, . . ,. .c'.,, -.
- w. 1..
.. , ~., -A.a.-*
... r n,-=.r.:..rs. .? . n.*-> .rw.
'
- a. ."
tna; it would be difficul ., if not inpassible tu generate criteria whicn w0sid 5:ecify the ce: ailed design changes. The even; re;orting syster r;n by E :'. wa s :15cssse:. 'nis syst cusers all ;1 ant com:onents and its 00jective is to improve pl a nt reliability bj anelyzing plan: c:erating experience and mating tne infernation available to ne utilities.
^
onmen: 2: Revisions of :ne ER reperting form.
Specific revisions recor. men:ed were a n;m:ering scheme whicn would f acilitate tracing the utility actions, including information giving tne age of the plant component involved, ne time of occur-ren:e, an: the significance of :ne event relative to the particular plant systems. Discussions of similar failures.
l tne potential for recurrence, and the expected reliability of the components involved were also thought to be useful.
Comment 3: Problems in the reporting of the LER data. The lack of a clear definition of the LER purpose was though* to be a particular problem. The Subcommittee believes that more .%
I
W
- 1, 0, 0 i .d.
- .s.m y. . . d .. n. , ,
tnorougn investigations into the C0urse of the LER events were required. Unambiguous descriptions of tne systems and the event should be included in :ne LERs along with descriptions Of the appropriate follow-up actions. It was noted that the lessons learnec from tne LER analysis shoulc be input as appro;riate into training crograms. !* was notec that the led.s currently su mi*ted provide informati:n 0".ly as to the f recaen-
- j of *.ne #3ilures. anc :: n:t rovide infor .ati:n as to the nJnLer u# :0..;;nents us : a nd t he f reque ncy a t vd. . h *.'.e, ere Called u Gr.. PrOviding these dat3 Would enaDie tne reviewer to tet*.er assess ne consecuences of the cor:1onent f311gres. .f 2e: : rt i a: tne ex ected reliatility cf the af'e:*.et co :nen*.
re '. c *.1. e t. One :tserve: fail re 'reauen:j a:s als: *n:.gn. ::
be useful.
Ca..?e . .
Iclectic and Jesignati:n Of LEPs. Tne Sa;;;~-' t .ee f el t some means was required for separatin; .important LER events frc the , mass of data that was ty? i cally involved in the LER syste . Tne estaclishment of special recorting en categories (e.g. . wa*.cr harmer events) and the assignment of a s1gr.1f1-cance to LERs eitner by a central organizatior or oy wcl1-coor-dinated field organi:ations were suggested. The need for updating the utility support f0r tne esta lishment of a consistent. co!grehensive, and cojective LER reporting sys*.e das stressed.
Comment 5: Coding of LERs. The Subcommittee has observed tnat tne coding ,
of LER cata should be improved. It has also been the experience of the Subcommittee that the data recording systen used by the Nuclear Safety Information Center is su erior to that used by NRC Headquarters organization.
s i
i i
i 11 O.
(h m e- C O t re C
- GJ 9 e *
( 31 ru @ .
- C
- +* C 3 E 13 e- 11 en C - L QJ in t3 ni 10 8T3 +* O O O L. to -
a x E C on F U C O s m L ,- n r-g e E, e m as y c--
r n en r-v y \,s. e (3 O c g . , -
r O
ou s3 r
-g . , -
D L T1 -r- en
- 3 U sh GJ O 3 e +s e- f.
q) m nu s r 3 L m C C c- r fNJ 13 ni to ea to - LA c- ni U I r U n U @ QJ % 3 Ub TJ X U O 3 T1 f .n r t' CL m ,a L r at tJ E O *+- 3 GJ GJ e- LYa U W C C L <e- +* O a) -e- +* e -*- I h - ( t. C e.n & U QJ 41 +2 4* tas
% 3 r L QA r r p QJ GJ O tn
'O L @ c- ---
G3 r 48 e- - +> <tn tn r- _s h G an t-3 O n m 3 c-- en an L 3
- 3. na L e .* LD GJ TJ 6- an tn ei nj 1 en as c GJ Os 11 C p CL Q _3 r n3 nn su c- ea C r C 3 c-an GJ C en C QJ - . - du e- 3 e-C n O C *ti E sn p U O O e- r- G7 4 +' L +8 = n - o +e $e
- 4) -r- Q) e O C e- 3 CJ
+n (- C O C n q) C rp e2 -c- .* O *U I3 U C f3 4) Q1 4-
- O U +* I G3 O 80 +* O U C GJ GJ T3 O O W C G1 i L C L 3 e* O v c-- U O P U e- O W 3 O +* 3 fL _) n ( s' ,J L O +* C C G3 O in O G X O C CL tal O 4- L ++ r- T1 -r- +s +.8 G3 3 QJ e-
+8 in e- QJ G Q) h O O 3 62
- L L +J T) +8 3 *e-4* -'1 D L *e- L- G) * +2 O al 8 L **- O 11 L L +> O QJ do O L Ga
- n
> ra t. ) as e % u 13 C >
e 'e- E-L J3 U C o h al O E m C TJ 3 to sn in GJ G3 3 3 sn Gs W L T1 G7 es L C C e du i; t. )
O su r r e s Gi e O O qs n) qr - [l QJ L I: e- an - O' E U -
o +> m +> C L n i. m O n L as O sn GJ E o m r O O w o O L r en G> 63 GJ O U C1 Q) +> U R3 48 c- r- L T) C ') 45 43 L C c- U f: O h U ** O. .'
O rn os GJ C v O Ti O vo e 3 -
n r- . sn . -
3 C L On en G7 +8 L to C (n L I QJ U 3 m GJ Gt +* 3 C L +3 f -c-O ea Gi
%- Gs e* M C O o C U - n3 m O <t s as 02 GJ r- 3 C L GJ
@ W 4* -r- U > 3 L O h - - e L E E +> r O 3 > C C U C 4- L QJ tt e- r- 4 5 11 C U ti e-- 't)
T1 r-
+3
% O (t: o 61 1 m O f n n1 C U +2 vu c-- TJ - QJ O
crt E r H +* *
- r- (1 O c1 O 3 % ^ GJ Gt Q) sn O nn n1 U C .- ra (1 O O ^ an C (r T) p 3 3
-c O in 11 O U en U h r--
Ii e-- c- +* GJ fr G 61 43 D en U @ O La s O r
- D e- e- e- L C tra 11 GJ to in GJ -- (n - tv c--
> +* v e _J +8 C O t5 L 1 143 48 en -e-49 3 ' + - % QJ O _J C % L to W
4
- C v- 'O GA O U +2 3 CJ m O +* O O @ C ,J n1 .3 O
+n en O O T)
G 3 i; C C GJ a3 T3 C ru O D h 9 I ru ( 3 Q3 p E Q) +a % ei tt C O Qt G [_ O F- -e- T3 n C GJ L +* e <0 r
+' _C GJ x rn i p L p 40 U sn 0: E C
3 O +a g L U u .c m ey c .. O en % C E 40 v- Q3 r 8 e- Z +i +*
O 8- en L L v du r- U (1 CD C cu . 4- G =
C GJ %
e
-e- e- O G7 O 1as L C O (1 L 41 t. (T C () O +2 en O- h C GP LD T3 T3 en L
_I W r L q) QS = U U WO C L G3 w-O O.
s2 L QJ L e C
C +2 O
r GJ * * -
L O G8 ci C
e-- ea as
- 1 Gr h at L .' C sn N
0 3 T) r-e-
+n
+3 e--
3 T) C
-e-C O
O +*
C1 -r- O,%
O m C. 4s 4s I cs na GJ en n3 ^ --
- L (1 X U S.a gn
'11 1t tn 6- ru ,, L e- 6* a2 L -e-L e C O. as -r- 3 L ru n3 e- en (13 41 43 e 3 en QJ TJ q)
GJ O 1 <n QJ @ a1 - G1 a.:
QJ 4) r- v % CD U QJ au en g U 1 U GP O CA U GJ G7 -e-L GJ so in O 6* --- e f-C t
4J () ia e-- +D
- L @ +3 = G1 3 1 L +* 3 6J Of C GJ O = C- c L sn T3 , .a @ an L gn m p C U 3 TJ 3 e o <t f O L Os L' O 09 +J M3 QJ @ C v GJ C 43 40 >g an O T3 @ +8 +.8 n' (! S (7 GJ y .r.
T3 3 {i e-- f-in .'* _ m t1 e- 4a O L ,s 3 > T3 q3 L w ,_.
C O G3 L c-- GJ 3 7- % I r, O C % U C 3 E en 3 C au E st2 e-- L O O e- O GA Gs C C e* 42 <t 3 v - r-- 3 GJ e - U p C
& & cu m U r r- Q3 e +s D in t O% as W M3 CD n
sn ** ([ L 44 - U L L -r* in w-O O LD C GD L Q) m +8 .C 4I ni nl en G3 O -e- T3 C O 3 R3 QJ r C
> L C L +' C 93 en 3 m 1 % C -r- G1 0 GJ Gs 23 ** t's an G r- % O at -r- m e % v 3 Cf -e-e e- r GJ ,~ es sn - C as n e - C w- p . LaJ +2 I u U - += L += Gs 11 .o % C i- i ni ti in (1 -"
G7 *F 80 GJ -f- t/b L J M3 f GJ de in T3 +8 19 41 () L (3 L
() GJ '3 O GJ e- e C C m tn v .n in <a m % m r- aJ O
- C U s+ - m m <o +e 1 .> L C - W ^ 3 e- X L GJ QJ U Qt W L' +8 tn O 09 +2 QJ GA @ U Of aa t. _) e8 1a L L % OJ C U% <0 0 -- L C CL O C C L e C C .* O U GJ G8 0 o O O r L en T3 La s -- f'
- en C e- (1 D -C O U 6- v O O 3 %
La. 3 (1 3 fu J 1 in e- O 3 e- at Ge 90
+J 48 L-C G)
QJ L:
I: E
- n. <- O est O L)
_t L3 a
LER -E- May 24-25,1979 bad design. The neea for more work on man /macnine relationships was stressed. Simulator design and simulator training were also identified as areas for possible ir:rovement. The -
relationshio between degree of reactor operator experience and human error was suggested c3 an area for study. Questions were raised as to whether or not practices existed which would result in tne assignment of the more senior coerators to the vore desiratle work snifts, tnus weignting the less desiracle wort assign ents witn tne less ex;erienced Operators. The nee for consideraDie teaching ability in the choice of training personnel was emphasized. .-
Co. nen. : :nsestiga.1:n of special situationc. t nu-ber of s;ecific areas were identifiec wnich the Subcommittee thought were worthy of special study. Iney were: (a) Browns Ferrj an 7":
accidents, (c'; nuclear plant Organizational struct;re. (c) the possibility that reactor components may Oe over-tested. (d) tne possible over-;rotection Of reactor systems (e.g.. tne use of low-rating f ases and circuit breakers). Witn regard to the Erowns Ferry and TM accidents, tne Subcommittee believed tnat much could be learned through a systematic evaluation of the LERs whicn ; receded these accidents. Opinions Were expressec
- nat such studies mignt have prevented at least the Three Mile Island accident. Tne Suecommi+. tee emphasited tne need for more interaction between emergency and operating personnel (civil defense units, state units, fire fignting personnel).
Opinions were expressed that nuclear power plants should not be operated by smaller utilities which may not have the re-sources that the larger utilities would have. It was suggested l
l l
t - _ _ _
LER May 20-25, 1979 i
- nat sucn :lants coulc tetter be operatec Dy larger orgam:ations whicn woulc speciali:e in the opera icn of nuclear power olants.
Commen 9; Computer analysis. The 5;ocommittee indicate :na: it was planning to conduct several computer-type analyses usir: the current data banx on an exploratory easis. Specific exam? l es were: (a) cnecking of LERs for clustering of events, (b) checking he fre;;en:j of occurrence for LERs of a given ty;c of ;1 art versus ancther. (c) studying :ne rela:icnshio be:neen 1 1
LERs and 013n 0; era ing times, an ', d) stucyi ng :ne f rcuency of occurrence of LERs oy plant size. !
l
' l q f. 5m .); *7
' s q q 'is .' ', .. .,
'e..'. . . ?. q ' a t_
- Y<qi-.
- e. ~ ,;(.
- * * *
- 8 m7Df9*q r
Dr. Bes ' re:v: was cis: asset firs (see :ar: 2of ::ac"en: 2). J This recort cealt wi .n a statistical analysis of :ne freauency of Lens ;
involving rea: tor coolant systems in 29 ;WRs between 1976 and 197E. The stucy is not ye: completet, however. :he following conclusions were drawn (1) :ne distriou; ion of nese LERs was sta:15:1cally homogeneous in time with the average lek rate of 3.5/ year / plant. (2) Arkansas Nuclear One, Uni: I reported tne most LE?,s (24 in 3 years) witn a average rate cf d per year.
(It was noted :nat if all reactors were equal anc were using tne same reporting criteria. there would De a .3% probability that any reactor from j
- nis sample would repor: 24 or more events /3 years.) (3) Turkey Poin 3 reported the fewes: LERs with two LERs in the three-year perico. The cnance of at least one reactor reporting two LERs or less during the tnree years is si if the reactors and reporting criteria are equal. (4) The I
reporting of events for subsystems is statistically homogeneous with time.
(5) More LERs are associated with some subsystems than witn 0:ners and certain of the subsystems dominate for individual plants (feecwater systems and controls in D. C. Cook, Unit 1; main steamline isolation systems and controls in Zion 1, reactor coolant cleanuo systems and controls in Arkansas Nuclear One Unit 1, and RHR systems and controls in Rancho Seco and Indian Point 2). Dr. Burns' analysis incicated the ex'istence of
LER -B- May 2'-25.197 9 two statistically inde enden: groups within :ne 29 reac :rs stuc1ed. He incicated that he would additionally study: (1) :ne derendence of LE; frecuency on reactor su ; lier, (2) the dependence of LER freauency on reactor size, anc (3) wnetner the distribution of LERs is consisten; witn the assumption tna all reactors are using the same reporting criteria. Table 1 of Dr. Burns' report (See Part 2 of At:acnmen D) lists
- ne num er of LERs relative to reactor size and nuclear steam supply i
system supplier. Dr. Burns indicated that preliminary studies show that j the ,ariation among reactor vendors and reactor sizes can be explained by randon effects. He notec however, tnat the ca:a sa ple si:e was sucn
- na; roor res:1;; ion wouic be exrected. His preli-i nary result s ire -
cetet no correlation witn plan; age. ,
1 Dr. Mceller's recor: dealt witn a study of several of classes of incivil- .
ual events (see par *.3Of A :acn en: D). Pour classes of events were analy:ec. They were: (ll f all re of monitoring systems m nin c0ntaire" . l (2) containment purging and airborne releases, (3) bypassing of moritors !
- nat actuate containment isolation, and (4) isolat10n of hign pressure coolan 1r;ection anc reec:cr core isolation ec ling syste s in .Ms. ;
}
Tnese classes cf events wi:n hign recurrence frequencies hac been encoun- i terec at a number of f acilities. There appears to be a clear need to institute administrative and hardware solutions for these classes cf events.
I Dr. Zucans' report dealt with events involving leakage througn valves i 1
and pumps (See part 4 of Attacnmen: D). Dr. Zudans briefly reviewed inree additional classes of LER events. These were: (1) leakage across the boundaries of interconnected fluid systems, (2) f ailure of components because of use beyond normal expected end Of life, and (3) design error resulting because of incomplete design criteria. Dr. Zudans also notec the existence of errors in the LER computer file and suggested that the files should be checked and, to the extent possible, such errors snould be corrected. Dr. Zudans also indicated that the design of fluid systems should be such as to accommodate the leaks that would occur in the valves in the fluid systems. He also questioned the soundness of some of the
> current designs which involved interconnections of high- and low-pressure
. systems. I
~
LER
- -y- May 20-25, 1979 Dr. First noted tna; i was his belief :na: :he cesign and testing of al-dampers was very weak and ina. work was needec in :nis area.
Dr. Catten's re;;r*. ceal; wi n the interrelationsni between LER events anc c: era:ce error (See ; art 5 cf Attachment D). He nc ed that 0:erator error was recognizec as a maj r contributor to LER events. Two proolems were identified witn :ne review of LERs resulting from human error.
ine first was
- hat psychological f actors may prevent the recognition of human err:r. The seconc was tha. e;ui; men- engineering is Cf ten such as to ccrtric;;e :r ;erna;s eve m3ke inev ta:le tre numan errors and nat i
- n. #, 2- v a. r .: . . r c. e. r ^. r. m.
a 4..e ..4.#.,~ "w.c. H. e^.. a-*.2
.. ec. wn'. r r e- ^ r . . m. :- - - ..
led to i'.. C om;1 e e. ca r:1 systems, ;;orly placed controls and read-outs, anc inade;uate inf or 3;ica dis: lay 5 were iden: 1 fled as major causes of
/
err:r. $ncetages f ::e-a:ing cerscor.el leaciag to inaceauate cc- ;r.fcatic-Detoeer ; era :ng sv f:s acre ais; icentifiec as sources cf ER e<ents. i full- ,me pr fessional training staff, indecencent of tne plant opera-tional staff, was iden.* fled as a use#.1 organi:ational f eature. It was sugges.ec :na: it wosic be alsc useful : stucy tne relationship ce: ween LER e e-.s and ne aual; .s cf :ne utilit.y t ra i ni nc. crc.a ni:ati on. ,
1
, i l
The meeting was ac;0urned at 5:20 p.m. on inis day (May 24,1979) and was recensened at i:20 a.m. On :ne f:llowing day (May 25, 1979).
Mr. Arnold's re ce. cealt w :n tne LE: reporting sys*er (See par: 6 of ,
I Attacnment D). Mr. Arnold noted tne present imbalance between the numoer '
of LERs that are submitted and ne resources whicn are available for l
- analy:ing and utili:ing tne information wnicn tney generate. He recor-mended ina: this be remediec by a combination Of improving the efficiency of the LER reporting system and increasing the available resources. He also recommended that it would be useful to identify special areas in which resources might be concentratea. The organization of these resources was discussed.
i
^* - . _..__.. _ _ " " ' ' ~w ~~= * *- me m w v m,, .A
LER May 24-25, 1979 4
- Was agre90 that the LER review gr0uD should te comp;sec Of Octn reg.la-
- orf and industry personnel an :na: arrangements wi:n eitner the NR: or incustry taking ne leac were possible. It was suggested ina: the effi-ciency of tne LER system could be improved by better criteria and guidance being given to the utilities, anc by the improved definitier cf plant safety margins. The role of equipment testing in LER events was discussed.
- It was noted that the use of by-passes and the altering of safety configu-rations for testing rurposes resulted in many LER events and that it was hignly cesireble to design plant systems to minimize the need for such
- r
- :ecures. It was add ti:nally noted that analysis cf some LERS had resealed ccnd1: bns wnerein so'e detailed featur'e of tne design na re-sulted in the v1olation cf what would appear in a detailed analysis to be fully recundant systems. 0:inions were expressed :nat the fixes im:lemented
/
as a resalt cf manj LER events only tended :: fi x ine part i:al a r e sent>;u t ci: no: ceal ai*.n tne princ1;al cause of the ;rcele . <r. Ar"D': sta.ed nat :ecnn ; es for assessing the degree to whi:h :ne safety margin was compromisec by particular events needed to De deseloped before the full po.ential of tne LE; systec can be realitec. Vr. Arn:!: alsc indicate:
- na*., for an LER analysis sys*.em to be fully effective. tnat there must be a hign degree cf cooperation and understanding between NRC and industry in
- ne:r develo: rent of improved c;Dlic safety and plant reliability.
Mr. Ditto's report :eal with nis review of LER events affecting con:r:1 and protecti:n systems (See part 7 of Attachment D). Mr. Ditto noted that, relative to this class of events, a broad spectrum of LERs mus: be examined very carefully, since what might appear to De relatively minor events could indicate significant deficiencies. He specifically identifie the use Of undersi:ed protection devices (fuses and circuit breakers) as a
- problem reauiring some attention. He noted that devices were sometimes l
i sized so as to protect equipment with conservative margins rather than for assuring that the equipment could perform its intended safety functions for the maximum range of conditions. He also recommended that all LERs for a given system be analyzed for common mode failures and adverse system 1
LER May 24-25,1979 interaction effec:s and noted tha such effects could oe easily -issec during intensive studies of specific failures. He add"tlonally recommended tna: there be an cnooing effor: outside the licensing and regulatory agency to study and learn from the LER events. He also ex;ressed an opinion tra:
tne structure of the LED system should not be al:ered to imo-ove its applicability to sta:istical analysis at the expense of losing good infor-mation that could im?r0ve safety.
Mr. Cromer's repor cealt witn LER events involving boron systems. (See p a r: E of At tacn en: D.i ersonnel and proced.res ;roolems were a major cause of LERs for all tcron systems. Instr; ment's and controls and valse failures were major causes of LER events in ECC systems and valve failures were a major cause of LER events in RHR systems. He noted that aporoxi-mately 1/A of the LERs suom ::ed fcr reactor coolant systems involved'~ 1 observec ocrcn concertrati;rs wnich #all outside of the tecnnical s;ecifi-cations. Vr. Cromer a:ditionally indicated that a significant number of LERs involved inadvertent safety injection events and recommended that the causes and conse:uences of sucn events be studieo.
l Dr. Lipinski's report discussed :ne class of LER events having to do ,
with set-;oint drif t in instrunentation (see cart 9 of Attachment D). Dr. I Lipinski indicated that SWRs seem to have substantially more problems with instrumen: set-point drif t than PWRs. The causes whicn were identi-fied were set-point drif t caused by rando.m component failures and instru-ments whicn were not compatible with the technical specification require-ments. The utilities solve the first class of failure by repair and recalibration and the second class by choosing and justifying a more appropriate set-ocint. The possibility of using automatic recali-bration systems was discussed. Opinions were expressed that the use of such systems would have a negative effect on safety and reliability.
)
1 I
\
May 2t-25. 1979 LER Dr. Seale's repor: deal: with LER events resulting in a icss of containmen integrity or violations of acceptable containment leak rates (See part 10 of A::achment D). Valve failures and cersennel errors were the leading causes c' LER events in these categories. Delicerate by-cass of air-lock intericc.s was discussed. M was suggested ina: improved crocedures and better accommodations for material transfer would reduce the number of )
1 events of this type. Dr. First reiterated nis recommendation for the development of improved air dampers and the need for improvements in test i procedures. such as pronibiting the cycling of the dampers before testing.
Boror dilution esents and One use of defec*.1ve materials in construction were suggested as anotner class of LER events worthy of 5;ecial s*.cdy.
l Mr. Grenden suggested that the use of unlicensed operating personnel ,
without adec. ate supervision and the replace en of failec cocoonents' I
without correcting :ne causes of :ne failures are classes of LER events !
worthy of special study (See part 11 of Attachmen: D). The training and l
)
qualification recuirements for auxiliary operator and supscrt personnel were discussed. h was noted inat a large numcer of LERs ap; ear to involvo l non-licensec ;erscnnel. Training requirements for non-licensed personnel will be discussed at some future meeting. Better c;erating crocedures, more em;hasis on wri* ten instructions, improved communications, and improved l 1
1 equipment design were icentifie; as means of reducing LER events. j 1
l Mr. Epler iden:ified two additional classes of LERs worthy of study. (See pa rt 12 of Attacnmen D) They are: (1) the separation of a control rod from its drive in a BWR and failure of the defenses directed agains: the trans tent caused by dropping of the separated rod, and (2) f ailures of :he vital instrument bus causing scram. Mr. Epler in his report noted that the emphasis currently placed on neutronics in the training of plant operators was related to the early history of nuclear power. The early reactors were small and the heat removal problem correspondingly small. Thermal hydraulics thus received less attention in the training of plant operators. As reactors 1
x
\
l l
l 4
1 LER May 24-25,1979 l 1
1 l
increased in size, it became apparent inat temperat;re coefficients l would terminate or limit the most severe reactivity transient, r.e nc e . j
- ne removal of residual heat emerged as the dominant problem. Mr. Epler l indicated that estimates of the system reliability rested almost entire-i ly on confidence in the single f ailure criterion. He noted inat tne .
Browns Ferry fire demonstrated the inadequacy of tne single failure l l
j criterion. in that a fire generated by a human error disabled many I
systems. We indicate nat Three Mile Island again demonstrated the cossitility o' m ltiple independent failures disabling a number cf syste.s. r. E;1er strongly er.:Orse: ;se of a dedicated heat removal
. l l
systeT f:r L.Os. He also recommended that the vital instrutent cus I should be so designed that it could not both ca;se a scram and ce unable to cope aitn the effects. g, Dr. First re c o-me nt e : that acre attenticr ce pai: to ne isolation of control-roo.- breathing-air suoplies and the interconnection of the breath-
-air syste witn other oossibly contaminated air supplies. He noted that many of the LERs associated witn air-cleanin; systers involved air-cam er failures associate: with poor design, control syste- failures, and degra-dation of gasket sealing surf aces. He noted tnat the performance of air !
dampers has generally been poor and tnat there is a need for research and development in this area. The equipment had been developed f rom conven-tional heating and ventilating equipment, wnere leakage is permissiDle.
The ; resent designs co not lend themselves to tne low leakage re;uirements of reactor applications. He noted that charcoal filters were nighly suscepticle to poisoning and that the filters installed during construction were of ten damaged by the ;ainting and cleaning activities associated with construction. Improvements in filter designs, such as the use sacrificial filters to protect the safety-grade filters, were suggested.
Dr. Okrent's letter of March 28, 1979, to the Subcommittee Chairman was discussed (see part 13 of Attachment D). Dr. Okrent's letter dealt with a
l l
l LER -la- May 24-25,1979 1
1 review of selected Quac Cities 1 and Peacn Sottom 2 LERs and the recent LER l cealing with the degradation of the ANO-2 safety systems. The following actions were taken on Dr. Okrent's comments :
(a) Cracking in pipes and in feedwater and control-rod-crive no: les in BWRS. The Subcommittee decided to rely on the work done by the pioe-crack study grouo and to look further into this are3
. only if there was information in the LERs that had not been l considered by this grouo.
(b) Main steam relief valse failures. Consideration cf this i'.e-aas assigned to '4r. Micnelson, Mr. Cromer. and Dr. Cat cn.
(c) RCIC systen out of service undetected. repaired in c 1/2 days no information on how long undetected. The Subcommittee decided to ,
discuss the generic implications of this at a later date.
(d) vse o# flar able materials in relays. Consideration of nis 'te-was assigned to Dr. Lipinski, Dr. Catton, anc Mr. Epler.
(e) Equipment not seismically qualified. Discussion of :nis item was postooned to a later meeting.
(f) Improper labeling of circuit board resulting in an off-gas flow-loop being removed from service. Discussion of this item was ocstponed to a later meeting.
(g) Beckflow of radioactive liquid from radwaste system to service ai r. Consideration of this was assigned to Dr. Zudans.
(h) Failure of nydraalic snubbers. Consideration of this item was assigned to Dr. Zudans.
(i) Equipment f ailures caused by high steamline tunnel temperatures.
Consideration of this item was assigne> to Dr. Moeller.
(j ) Pipe cracks in recirculation pump discharge by-pass line. !
Consideration of this item was assigned to Dr. Zudans.
(k ) Main steamline relief valve failures. Consideration of this item was assigned to Dr. Zudans. _
(1) Cooling-water suction header air-lock due to air in-leakage from pressurized pump seals. Consideration of this item was l l
assigned'to Dr. Zudans. l l
1 1
1
LER May 24-25.1979 (m) Short-period scrams on nign flux curing start-up testing. Consi-deratior. Of this item was assigned to Mr. Epler.
(n) Obstructed sensing lines on instrumen; rack. Consiceration of this item was ass:3hed to Dr. Moeller. )
(o) RCIC pum; problems. Discussion postponed to a later date.
(p) Witndrawal of two control rods simultaneously in violatien of established procedures. Assignment of tnis item was made to Mr. Epl er.
(;) Interlock failures on interlocked doors. Consideration of tnis ite was assignec to Mr. Ray.
(r; Degra:a: 1:n of engir.eered safety features a: Arkansas 'Uclear 1 Jni 4 Consideration of this item was assigned to Mr. Ray.
/
'e This con:1.c:: : e 5. c:--ittee discussions. Eusiness was acjourned wi'.h :ne ex;e::at::m that a f:110w-:r. meetir.g woulc be held in late Juno 1979.
NOTE: A copy of a transcri;t of this meeting is available in :ne NRC Putli: Docu ent Room, at 1717 H. St. . N. W. , Washington, D. C. , or may oe ott ained f ro: Ace-Fe:eral Reporters. 444 North Capital St. ,
N . ri . . Wasnington, L. C.
a i
1 4
~. .
, . - f Federal Register / Vol. 44. No. 93 / Friday, May 11,1979 / Notic.es 4h2M5 "T w data, such as sa! aries: and personal pertinent to this rev;ew. The Subcommittee ye and tirne: May 3L and June 1.19 9: em may then caucus to determme whether the a m. each day. information concerning individuals gue: National Science Foundation.1 Boo C associated with the pr9posals. These matters idenuSed in the init:al sess6on have 5;teet. N.W., Washingtert. D C. Room 30s. matters are within exemptico (4) and (8) of bee:1 adequately cosered and whether the me of rneeting Closed. 5 U.S C. 55:b(c). Cov emment in the project is ready for review by the fu!!
Cract person: Dr. Elijah B. Romanoff. Senshine Act. Committee.
P ; yarn Director for Metabolic Biology, Aasenty to c!ose =reting This deter =mation was made by the Committee In addition. it may be necessary for Jtsom 331. Nationa! Science Foundanon. .
Management Ofncer pursuant to provisions the Subcommittee to hold one or more Washmston. D C. :c550. (20218.32-4312. closed sessions for the pu pese of
-- p,epose of subcomtrjttee: To provide advtes of Section told) of Pub. L 9:-483. The and recomrrendauons cencerning support Committee Management Off:cer was exploring matters involving proprietary 9 !ar research in Meiabohe 34clegy. de!egated de authonty to make such information. I has e determmed,in 1 Agenda: To review and evaluate research - de*erminatier.s by the Acting Direct:r. accordance with Subsection 10(d) of preposals and ;rosects as part of the NSF. on February 18.19*7. Pub. L 92-463. that, should such selecuen procen for awards. Dated. May 8.19 9. "
sessions be required. it is necessary to Jtessen for cicsint The propesals besng u e,s wm, c!csa these sessions to protect reuew ed inch.'de information of a caeann vmper cme""" proprietary information (5 US C.
, prepnetary of cor.fidential nature. ;nt cor. m ras rued s.to.ra a o amt includeg tech 6 cal information. financal suo coos issus-et 532McN.
data. such as salaries: and personal Further information regard,mg topics information concerning individuals to be discussed, whether the meeting sisos:iated with the proposals. These has been canceUed or rescheduled, the matters are within exemptions (4) and (8) NUCt.E AR REGULATORY Chairman's ruling on requests for the of 5 '.? S C. 55:b(cl. Coverr. ment in t.be CCMMISSION opportunity to present eral statements and the time allotted therefor can be Aut rt o c se meeting This Advisory Committee on Reactor obtamed by a prepaid telephone call to Permination was made by the Committee Saf eguards; Subcommittee on the Designated Federal Employee for Management Off;cer pursaani to provtsions Safeguards and Security; Meeting this meeting. Mr. Richard K. Majer, of Section 10(d) of Pub. L 92-453. The
[ Committee Management Off;cer deiegated The ACRS Subcommittee on (telephone 202-634-1414) between 8:15 the authority to rnake such determinations Safeguards and Secunty will hold a a.m. acd 5:00 pm, EDT.
by the Acung Drector, NSF. on February meeting on May 23.19 9 in Room 1046. ' Backpound information concerning 18.19'? 1717 H St., NW, Washmgton. DC 20555 items to be considered at this meeting Date1 May 8.1sts. to discuss recent safeguards events, can be found in documents on fue and use wm advice from its consultants, and the 19~9 as ailable for public inspection at the ca nu Atmu=ar c.=4a= Review and Evaluation of the NRC NRC Public Document Room.1 17 H tra ow. w e m rad 3. n au aal Safety Research Program. Notice of this Street, NW, Wa shington. DC 0555.
. 8*'G coot 753HW meeting was published March 23 and Dated. May L 17 s.
4 April 20,1979,(44 FR 17&37 and 44 FR f 23609 tespeetively). 3*" CN
- Subcommittee on Social and #""*"""*"'""" '
In accordance with the proceddFes Cevelopmental Psychology; M eeting I"' " "' I outlined in the Federal Register on " * ** ' #"'~' "
In accordance with the Federal October 4,1978 (43 FR 459:5), oral or Advisory Cornmittee Act, as amended, written statements may be presented by
) Pub. L 9NS3 the National Science members of the public, recordsgs will Advisory Committee on Reactor
- Foundation announces the fotowing be permitted only during those portions Safeguards; Subcommittee on
' meeting: of the meeting when a transc .pt is being Evaluation of Ucensee Event Reports; I kept, and questions may be asked only Meeting ,
Name: Subcommittu on Social and Developmental Psychology of the Advuory by mernbers of We Subcornmittee,its Cornmittee for Behavioral and Neuraj consultants, and Staff. Persons desiring The ACRS Subcommittee on Sciences. to make oral statements sbou!d notify Evaluation of Ucensee Event Reports Date and time: May 31.137-June 1,in 9:00 the Designated Federal Employee as far wi!! hold an open meeting on May 24 a.m. to 5m p m. each day. In advance as practicable so 6at and 25,1979 in Room 1046.1717 H St.
P! ace: Room 8:3. National Science appropriate arrangements can be made NW, Washington, DC 20555. Notice of Foundanon. taco G Street, N.W. to allow the necessary time during the this meeting was pub!!shed March 23 Wa shing'on. D C. :0550- s meeting for such staternents, and AprJ 20,1979 (44 FR 17837 and 44 The agenda for the subject meeting FR 23609, respect:vely).
oNact e: nIDr y C. Shaver. P oram In accordance with the procedures Director. Social and Developmental shall be as follows:
Psy chojegy. Roorn 317. National Sc:ecce outlined in the Tederal Register en Foundation. Washing'on. D C. :oSM, Wedas Octobpr 4.1978 (43 FR 459:6), cral or Cocclus, ionday, May :3,19 J. 8:30 am I'c:f1 the of Business wTitten statements may be presented by telephone (Oc;) 630-m4. , '
Purpose of abcommatee:To provide advice The Sabcommittee may meet in T.aecuthe rnembers of the public, recordings will and recommendations concernir's support Session, with any of its consultants who may be permitted only during Sose portjons for research in Social and Developmental be present. to eniore and exchange their of the meeting when a transcript is being Psychology. prehminary op:n: ens regarding matters which
. Agenda.To review and evaluate research shoWd be considered during the meeting and k* '""d "I b' A'#
proposals as partJf the selection process to formulate a report and recommendauon to by members of the Subcommittee,i,ts, for awards. .
the full Committee, consultants, and Staff. Persons desmng Reason for closir.g The proposals being At the conclusion of the Executive Session. to make oral statements should notify reviewed ine!ude !nformation of a the Sabecmmittee will hear presentauons by the Designated Federal Employee as far propne'ary or confidential nature, and hold discussiorts with representauses of in advance as practicable so that inc!uding technical information. financial the NRC Staff. and their consultants, appropriate arrangements can be =ade a
en M
1 i
l ACRS LER SUBCOMMITTEE MEETING MAY 24-25, 1979 WASHINGTON, D. C.
.~. _
I ATTENDEES LIST ACRS NRC l D. Moeller, Chairman J. I. McMillen
. C. Mark, Member H. H. Scott H. Etherington, Member W. Mathis, Member '
J. Ray, Member '
J. Arnold, Consultant R. Surns, Consultant l I. Catton, Consultant S. Cromer, Consultant S. Ditto, Consultant E. Epler, Consultant j M. First, Consultant ,
A. Grendon, Consultant R. Seale, Consultant Z. Zudans, Consultant W. Lipinski, Consultant ;
R. Savio, Staff I A. Bates, Staff
- i D. Johnson, ACRS Fellow McGRAW-HILL l l
Sob Aderman KANSAI POWER Kunihiro Ota REPRESENTING SELF R. H. Leyse .
..
- l I
i Attachment B
. \
l
27776 n-Federal Registee / Vol. 44. No. 93 / Friday. (fay IL 1979 / Notices .
a to a!!ow the necessary time d[ing the meeting fer such statements. Commission's rules and regulatices in to 11 ,
CFR Chapter L which are set forth in the The agenda for rubject meetics sc, all license amendments. Prict pubhc cotice be as follows: ,
In the course of!ts evaluation to date of these amendments was notrequired
~
- of the accident at the Three MI!e Is?and Thursday and Friday.May O' and 25. u S. since the amendinents do not involve a Unit No. : faci'ity, which utifi:es a B&W sigmficant hazards consideration. desiEed PWR the Nuc.' ear Regulatory g
s 30 a rn. (lntil the Conclusion of Businesa The Cct: mission has determined that Commission staff has ascertained !.'.at the issuance of these amendments w!!1 B5W designed reacters appear to be ,
tis o a wo s anH th not resu!! in any sigmficact unusuaMy senJitive to certain off. noct:a! - '
representatnes of the .@y b p transient conditions criginating in the ,
C Sts5 and their environmental!mpact and that purscant consu!Iants. to concinod its rev:ew of secondary system. The features of the ,
tacenree Event Reportisub=ir'ed dtu.ng the to 10 CFR 5t.3(dl(4) an environmental B&W des;gn that centrbate to this .
per od 1975-19taL impact statement or negatfve declaration and environmentalimpact sensitivity are:(1) The design of steam The Subccrnmittee wiD also hold one ce generators to cperate w;6 relatively appraisal need not be prepared in censee Ev nt or j a le 8
e nnection with issuance of these smallliquid volumes :.n the secondary i Subcomrnittee report to the ful! Corn uttee, amendments, side: (2) The lack of direct initiation of reactor trip upon the occurrence of cff- {
Further informauen regarding topie* For further details with :espect to this ,
action see (1) the applications for norma) conditions in the feedwater to be discussed. wheder the meetinE system:(2) Reliance on an integrated has been cancelled or rescheduled, the amendments dated Apn! 30,19 9. (2) control system (ICS) to automa ticafly 4 Chairman's ruling on requ'ests for the Amendment Nos. 43 and M to Ucense regulate feedwater l'ow. (4) Actuatica opportunity to present oral statements Nos. DPR-39 and DPR-48, and (3) the before reactor trip cf a pi!ct operated and the time allotted therefot can be Commission's related Safety Evaluation. relief valve on the primary system obtained by a prepaid telephone call to r All of these items are availab!e foe pressurizer (which.if the salve sticks the Designated Federal Employee for public inspec:icn al the Comraissioo's open. can asyavate the event)! and (5) [
this meeting. Dr. Andrew L. Bates I Public Document Room.1717 H Street.A low steam generator e!evation 3 ;
(telephone 202/634-3:67) between 8.15 N.W., Washington, D.C. 20555 and at the (relatlee to the reac*cr vessel) which a.m. an 5.00 p.m.. EDT. Zion Benton Public Ubrary District. : 500 provides a smaller driving head for d t Backgrcund information concerning F.rnmars Avenue. Zion. Clincis 60099.eatural A circulatioa r items to be considered at Llus meeting copy cfiterns )
obtained upon(21 and (3) may request addressed to the be Because of dese features. B&W ir v able fo ubi e r7pection t th designed reactors pface :cre reliance ; s:
U S. Nuclear Regu!atcry Cctemisalon, on the reliability and performance Washington. O C. 20335. Attentice; re-NRC Public Document Room.1 17 H St.*
NW Washin8 ton, DC 00555. characteristics of the auxiliary -l Dated May 8.19 9.
D recter. Divisica of Operating Reacters. . feedwater systern, t e ICS, and the g
bas C Heyls Dated at Bedesda. Mary!and 6ts 3rd day di.
e,wr c.umu wow ,,,,,,,c % ,
performance to recover from fre(quentem ii Id For the Nuclear Regulaton Cocunisswn. ant lcjpg{gd tygng[ents, such as ICss of ' or irn o m.m:n.co. m so : offsite power and loss of nor=al Sn.Lmo coes r38MH8 , 4. sa.",= ] sur Ch,el ce wrmy bcws erm:h Na r. Dre:.arare orcW feedwater, than do other PWR designs.
" is mo This,in turn.p! aces a Iarge burden on OEI Commonwealth Edison Co,; Issuance of Amendments to FacIty Cperatin9 Ucenses Y[r[ .
a m .o coc4 n m.n.as ew the plant operstors in de event of off.
n rmal system behador during such j
8
e As a result of a preliminary review of $ foilc The U.S. Nuclear Regula tory s the Three Mile Island Unit No. : 'gP Commission (the Commission) has Duke Power Co. accident chrenology, the NRC staff j reha issued Amendmect Nos. 49 and 46 to it:itially identiEed several human errors "
- 8
Facility Operating Ucense Nos. DPR- 3 in the =atter of Dule power and DPR-48 issued to Commocwea16Company. Ococae Nuclaar Station, that oc=rred during de accident and i III Edisen Company (the licensee) which Units Nos.1. 2 aod 3. contnbuted sigmficantly to its severity. 1 han -
revised Techmcal SpeciRcations for Order All holders cf operati:g !!canses were d **^'
' subsequently instructed to take a i the pi operation of the Z;on Station. Unit Nos. .:
I number ofirnmediate actions to avoid .eedw I and 2. located Zion. Ilhnots. ne repetition of dese errors. in accordance "
- amendments are effective as of the : fate The Duke Power Company (the with bu!!etins issued by the of issuance. t I2I /.
!icensee). Is de holder of Fae:!ity Commission's OfSce of I.~spection and ' . ' establ.
These amendments revise de Enforcement Technical Specifications to require . Operating Ucenses Ncs. CPR-38. DPR- staff began an(TE). In addition the NRC of bod actuation of safety injection based on : - 47 and DPR-33 which audorize da immediate reevaluation of the design feafures ofiMWreactors
' "' 1 out of 3 channels ofIcwpressiu.zer operation of the nuclear power reactors -
he wa pressure known as Oconee Nuc! ear Station. Caits to determ:re whether addificnal safety .acket Nos.1. 2 and 3 (the facilities, or Oconee correct'.ons or improsements were O' C a The application for the amendments necessary wuh respect to these reactors
- 1. : and 3), at steady state power leveis 8 an au-compfies with the standards and requirements of the Atomic E'nergy Act not in excess of:568 megawatts therma) This evaluation involved numerous (rated power) for each unit.The meetings mth S&W and cerfain of the N"'
"*E IO of 1954.-as s'nended (the Act), and the affected Ecensees.
Commission s rules and regula tions. ne facilities are Babcock & Wiicex(B&W) 8 In ac designed pressuri:ed water reactors The evaluation identined design Commission has made appropriate st findings as equired by the Act and the (PWR'sf located at the licensee's site in features as discussed above which to Oconee County. South Carolina. . indicated that C&W designed reactors '
are unusually sensitive to cer*.ain off- I tw
- feedwa te g .
e
- t. [ N' P =g --g4 Oh"* r T.-
l SCHEDULE FOR THE LER SUBCOMMITTEE MEETING l
WASHINGTON, DC i 1
MAY 24-25,1979 !
l
.' \
l THURSDAY - MAY 24, 1979 I
8:30 a.m. - 5:00 p.m. Subcommittee Review and Discussions of LERs and Start of Draft LER Report.
FRIDAY - MAY 25, 1979 8:30 a.m. - 5:00 p.m. Subcommittee Review and Start of Draft LER Report e
.O N
I i
i 1
L ATTACHMENT 0 PAF.TS 1-13 9
t e
4 m
e w
- e Mw P
6
- w.- "'6m*pseq,
.g r 7/ \ twM/AS/DHJ May 8, 1979
. CCtd?INTS CN TF2 ER Fd.:CRT2G SYSTT.M l . '7The purpose of the ER System needs to be clearly identified by the NRC. Various NRC Cffices are using de system with different objectives. It appears that the data available in the ER's does not completely meet the needs of any of the NRC
, . groups attempting to' use the data. "he information included in ER's should be made to confom to the parpose of the system.
It would be beneficial if one group of people within NRC, consist-ing primarily of representatives of the offices using ER's, had overall responsibility for identifying de purpose (s) of the ER System, the needs of the System users, and de types of uses which '
the system should have. ,
- 2. The ER Fem Needs Revision
- a. ~his is exemplifief by several facts. Many utilities prefer to ,
report an event in narrative style on one or more sheets attached to the ER fom, rather dan to use the fom, itself.
i Users of de information on the form of ten find it inadequate and of insufficient detail to identify the true problem,
- b. It would be helpful if every ER could contain e single number so it could be readily traced for additional infoy.ation. Other
}
data that would be helpful includes the age of the component in which the ER occurred, the age of the ple.nt in which it occurred, ,
and the time of day at which the event oc:.urred. .
- c. The significance of a ER may be obscured if the reader is not intimate with the details of the particular plant; reports of failures of components with plural functions should list systems.
of secondary function which are also affected.
l = '
l 4
p '
~
fwaS 2
'fa g .
- . . ~
.-+'m--w e a'- ' + ' - e -: O
- 3. , Problems in the Reporting of Data
[a. Particularly troubleseme are the LERs for which the cause is
~~
listed as " unknown." Al'so needed is a better definition of "cause." For example, was the cause of an e/ent the failure of a motor or the relay controling the motor? In addition, it is not clear to what degree the cause should be attributed to " human error," particularly if it is due to a design error or to a lack of maintenance. Another example of a simple, but troublesome, problem is the fact that when doses
. are reported in " mrem," versus " millirems," this is frecuently reported on the computer printout as ".92.M," which leaves the reader to guess whether it is mega, milli, or micro rem.
Another item is the subject of the reporting of effluent releases.
Pather than simply reporting the nu-ter of curies released, it would be far more helpful if the licensee provided a rough estimate of the associated population dose (in person-rem) . Also, it might be noted thot many LERs say exactiv. the ocposite of what they mean.
More careful editing is obviously needed. ,
1 Lastly, it would be helpful if the LER reporting system could be designed so that the name of the manufacturer of the failed com-ponent and the model number were reported.
- b. LERs as currently submitted provide information only on the .
frequency of failures. They provide no information on the number of the given components or systees that are in opera-tion or on the frequency with which they are called upon to 4
respond--data that are needed if failure rates are to be com- -
I l
put ed . While it is acknowledged that the SPROS system is )
2 gathering information on failure rates, attention should be -'
directed to determining whether simple modifications or addi-tions to the LER reporting system might provide additional 1 ;
1 1 p
I I
i 1
l
_ - . -- . _ __. a
._u... -- . . . . . . ._
J, data along these lines. Failure rate data are essential to __
the assessment of the risks associated wid failures in given systems. '
- c. Cne of the best approaches for solving many of these problems would be to establish a training program for utility personnel '
who are involved in the preparation and submission of ERs.
We degree to which such a course could add standardization to the system would be most helpful.
- 4. Selection and Designation of ERs
- a. Some mechanism needs to be developed to remove from the reporting system dose events which are unimportant and data wtach are trivial.
Se abundance of unimportant events'tends to hide de events of signi-ficance and produce an a .titude on de part of the tac and *.icensee dat ER's are a paper pushing exercise dat is not important to safety.
- 1) One suggestion is that scme form of screening be done at the tac Fegional Offices to remove from the system those ERs that are not important. Compounding the problem of the excess of ERs is the fact that events of " potential plic interest" must be reported regardless of their true or pctential impact on health and safety (see Regulatory Guide 1.16) .
- 2) Another approach weeld be to classify URs into several groups.
Cnly those in de upper categories would then be analyzed in
., detail. 2 assist in ranking ERs according to such a scheme, perhaps a system for categorizing dem according to deir "po-tential" versus "actml" imper ance from the standpoint o; health and safety could be developed. B is could be fur cer subdivided into their importance on-site versus off-site. 3e
[. current system of selecting certain ERs, and upgrading , ,
~
- them to Abnormal Cccurrence Reports (ACRs) for submission to the U.S. Congress on a gaarterly basis, is a step in this direction. However, much more needs to be done and and many more ERs need to be analyzed in depth.
4 . ...
I
. 1 1
[b. It should be recognized that the number of events actually
^ occurring probably totals about ten times those reported; in fact, data show that, in many cases, similar events **hich foreshadowed an ER, were not reported because dey did not violate any of the Technical specifications. Because ERs are so closely tied into de Technical Specifications, there is no assurance dat all events of ptential importance to l health and safety are being reported. .
'c. It was found that utilities hold a variety of different ;
opinions and attitudes toward reporting events as ER's.
Some felow the pnilosophy of reporting "everything" '*hile others :eport only ahat they believe dey must. Cne should be cautious about criteri:ing a utility for re;crting a large number of ER's unless the ty;es of items subcitted are re-viewed. Cn de other hand a paucity of ER's on de part of a utility should not necessarily produce a feeling of ccafort.
It would seen that time and effort should be expended toward ,
i the goal of developing a more unifot nerly positive attitude !
toward reporting of events to the NRC. 2.e NRC is also encouraged to avoid rating utilities based upon the number of ER's; one should go beyond and look at actual perfomance at the plant as well as the capability of the management and operating personnel. .
l
- d. In certain cases, an applicant files an ER only after being )1 l
recuested to do so by the NRC Fe;ional Staff. In oder casas, a prospective ER is undoubtedly " rationalized" out of existence.
It would be useful to know how many ERs or potential ER's fall into each of these categories and how much difference dere is .
in the negotiating process from Region to Region within de SRC 1
I structure.
'J
- i -
I t 4 ~~~ 'w' -n~~ , , _ _ _ _ _ _ _ _
1 1
, . 't .* * ..a.-......-
l 1
i
- C e- .
i Meowding of ER Ca a . .
c., . - . . . -
-'a. Be Subcoccittee has observed that the s-cuding of ER data
~
is poor. Bis is particularly true with respect to the classi-fication of ERs according to system and c:cponent. A typical computer printout, for example, will often include only about 50% of the ERs known to involve a given component or system.
B e h7.C Staff, itself, estimates that at least 25% of the ERs are miscoded. Bis stems partially from de fact that plant components often serve multiple functions and can consecuently be classified under a number of different systems or subsyste .s.
Another . source of confusion is de system coding scheme used wie de standard ER form which is vendor oriented.
- b. 2e experience of the Subcoccittee, although limited, clearly indicates that de data recording system of de Nuclear Safety Information Center is superior to that of de NRC Headquarters Staff. A part of the reason for dis may be dat the NSIC Staff is Femitted to take more liberties in correcting infomation in ERs that is obviously incorrect, whereas the Headcuarters NRC Staff is not pemitted such liberties. Se Headquarters NRC Staff has also experienced restrictions in the degree to which they are able to follow through on a given ER to obtain missing infomation or to correct obvious errors. 2 e NSIC System is also able to identify " Key ords" which are valuable in doing computer searches for particular types of events. ,
- 6. Follow-up and Use of ER Ca a
- a. .vany URs contain a promise of "follew-up" act. ion but it is difficult to check whether the premised follew-up was ever accome. lished . Presently when a follow-up rec. ort is subaitted
~
3 it is used to replace de proceeding report in the NRC data' l file. m:C merely adds the follow-up report to its file,
- which is secuentially arranged according to the data of receipt, 4 *
-+..m 6 - ga.- 4r -pme -e.e-i we
i.
~
., with no alterations made in or references added to de original l_
repo rt. It is felt that the entire sequence of reperts for a given event may contain useful infomation (e.g. regarding the adecuacy of utilMy or NRC staff Actien) and dat this sequence should be stored consecutively in its entirety, the importance of this may be noted by the fact that a review of one group of ;
ERs for one pewer plant system showed that 4% contained a l
promise of more data and an additional 5% contained a premise of . follow-up action. Yet, within the current system, there is no way to detemine whether de promised action has ever been acceeplished. When !G reviews an ER and recuests additional l infomation, there should be a femal system to determine when !
the request has been met, as well as when de corrected ;
i infomat.icn has been recorded widin the data bank. In short, l l
there should be a formal system to assure that every 2R is - '
monitored until it is completed and has been properly recorded in de data bank. In this regard, there is an urgent need to grant clear authority to de data recordirg group to enable them to check back on incomplete GRs and have voids filled or
, inadequacies clarified.
, b. The NRC is currently considering an expansion in their program
, staff to review and analyze GRs. While in this process, the
' Subcommittee recommends that they consider the establishment of a supporting laboratory in which sarnples might be analyzed and tests conducted to obtain additional details relevant to specific failures. Such supporting work might well be done on a contractual basis. In a .y event, however, the establishment of a capability for 'such analyses should be considered.
- y. The procedures and NRC requirements for operator training, licens .
- i. ,, and re-qualification were discussed with* the SRC Staff. At e
the present time, ~' -~ utilities have requested that they be i 4
l 3
e l
- .-.- -~ . - - . - - . . - . - .
.n ...
.. -~ . - .: . ._
1 1
i provided with copies of all ERs pertaining to deir plant, or to--
. . -~. given systems and components in given classes of plants. In .
i addition, the NRC operator licensing staff indicates that they use .
1 ERs in reviewing the operating experience at given facilities.
However, the large number of ER's which indicate the cause as being operator error, procedural error, and maintenance error
.culd seem to indicate that a femalized program of ER review should be instituted within the operator training and re-qualifi-cation programs. It is hoped that this would reduce the number of ER's caused by human error. The Sebecomittee was also quick to mte that dere is little coverage, at all, in current operator examinations on de subject of fluid flow and hea: transfer. I: 1 is believed that upgraded training leading to increased cperator understanding of the themal/ hydraulic and system interaction
, effects of plant equipment would improve plant safety. Considera-tion should also be given to the training and qualification of maintenance ;ersonnel who are de source of a significant number of ER's d a: originate through human errors. I 1
l In summary, there does not appear to be any systematic plan whereby the benefits or lessons to be learned from ERs are fed l k
back into the training programs for plant personnel. Formali::ation j of plans for such feed-back would appear to be a worthy s:ep to l implement. -
s
. \
l
- 7. heas for hploration l In the course of its review, the ACRS Subcommittee has cbser/ed a number of areas which appear to be worthy of more detailed investi- -
,gation and evaluation. Bese include: ~
- e
{
l.
l ,_.
~. - -
__~_,.... -._-.. ..._ _ .
1 l ,. , . .%....- _. . - s . .a _ - - -
l l l
1 1 -
1 l
l 1
1 t 1 i - a. Be Impact of Human Errors !
I
- 1) It muld be helpful to know how many LERs occur because of l procedural errors, how many occur because of improper di- l l
rections, how many occur bacause of inadecuate training, and how many are perhaps associated with bored:rn?
- 2) It would also be interesting to know the frecuency of operator errors by time of day, by operators whose licensing examinations scores were low versus high, by i SRos versus PCs, and by the estimated degree of stress at the time they occurred.
- 3) The !GC and other organizations, such as 5.P:, have s
conducted studies of control roce design a.d layout in tems of optimum o,terator response. Sese reports indicate dat consideration of de man-ma dine interface in nuclear plants lags behind the state of the art in the aerospace industry. It would be interesting to detemine the degree to which the results of these studies have been fed back into control room design and whe der errors in the improved control rooms are less than those in the older designs.
- 4) Operator training is often perfomed on one of the limited number of simulators available. B ese simulators cannot duplicate the location and place-ment of dials, switches, and de controls actually ,
present in de individual reactor control rooms.
3e degree to which this affects operator performance could be usefully reviewed.
e m e
4 6 -
4 We 4
O 4
4
l I
, , , 2 -
- l 1
.___l
[ 5) Infomation provided to de Subcommittee revealed tha-O plant operators, in training status, were frequently {
assigned other duties, some of ohich carry significant i responsibilities for which they may not be fully cualified.
It would be interesting to know the degree to which this influences the number of ERs associated wid various i
, power plant systems. It would also be interesting to have tabulations of the contributions of human error to various operations within a power plant such as waste management , reactor perfornance, radiation protection, etc.
6)
Some utilities are known for running a " tight ship,"
others for operating under somewhat less strict condi-tions. It would be interesting to detemine the i:npact of such attitudes on de frecuencies of ERs. It would also be interesting to see if rewards or incentives for good performance lead to a lesser frequency of ERs.
- 7) Although the SRC is currently conducting studies on human I factors as related to reactor safety, none of the members !
of de SRC Operator Li'. nsing Group is involved in dese t studies. Bis situation should be corre:ted. (i
- b. Pipe Cracks and Leaks d Review of the ER's reveals a number of events of pipe cracks '
J and leaks. dese are produced by a variety of mechanisms in- - y
{
cluding stress corrosion cracking and fatigue. Se NFC staff l and de utilities have taken action on systems containirrg posi- .
tive displacement charging pt:nps *hich will reduce or eliminate i
~
/ ll the associated fatigue cracking events. Be degree to whfch the t1 5 NRC and industry use ER's to identify other areas where leaks have occurred on a generic bases is not clear.
e l
- I
. . .- w * " * ~ "
- . c . Instrumentation Failures ..
Many t.T.es instreent failures are attributed to heat and frecuent.'y the " cure" is to install a larger ventilation system. It sculd be interesting to know what effort has been made to work with instrument manufacturers to design and construct instruments that will ret fail under heat stress.
Cn a generic basis, this raises what appears to Subcor.nittee members to be a basic problem. Sat is, too frequently the "fix" for a problem reported in an ER is not truly a "fix."
Bat is, it represents at best only a temporary solution to the prcblem. It is not clear from ERs whether lorg tem pemanent fixes are later instituted.
- d. Long Term Relationships of Failures It would be interesting to know how many deficiencies that are noted during construction are followed by failures in the same system during operation which can be directly traced to l
the construction deficiency. Unfortunately, the current E R system does not pemit this type of relationship to be easily evalua ted .
S. Investigations of Special situations
- a. Broms Ferry and TMI Accidents '
Subcommittee members believe that much could be learned through a systematic examination of the history of ERs that preceded the accidents at the Broms Ferry and "hree Mile Islard Units.
They believe that careful study should be made not only of the -
history of ERs at both of these plants but also at all plants ,
T of a similar type (same vendor) . l l
l 1
. l
. . a,. . . . .. .. _.
- l 1
l l
. 1
.l 1
1
.. l 4
'Ihe Brew.s Ferry fire was pt eceded by a number of similar )
snall fires started by candle which were not reprted as 1 ER ' s . 'Ihe Three Mile Island accident was preceded by loss of feedwater - stuck open relief valve events on other B&W reactors that '*ere reprted as ER's and which were investigated by the staff.
It would appear that even when these prior events are noted by the NRC there has not been adequate feedback to the utilities, industry, and cperators so as to eliminate the problen.
- b. Nuclear Plant Crganizational Structure Discussions indicate that, in many cases, utility personnel
, ,, e ' organitations may prallel those of fossil fueled plants, s
~d g
y rather than being set up to handle the more com:licated and demanding requirements of a nuclear plant. In prticular the 7
degree to which the utility engineering staff interacts on a day to day basis with the plant operators following initial
. startup may affect reactor perfonnance. Although not directly a part of the ER study, several Subcommittee members suggested S
that this matter be explored in further detail to see what, if any, implications it has.
- c. Cver-Testing of Selected Components
':he Technical Specifications for a nenber of nuclear power plants require that certain systems be tested at a certain frequency. Subcommittee members have asked to what degree .
such testing might lead to failures in specific compnents.
3- Testing of diesel generators, for example, requires rapid s
1%
' . ::= . . .. .
7-i = = =_5g;.: -
- "= '
J., coming up to power with minimum lubrication; testing of cir- 1 cuit breake's, r on the other hand, may add to the efficiency of their operation; testing of valves frequently results in their being left in the wrong position.
NRC Research has developed methods for determinire the effects of the test interval and duration upon equipent
, y availability and failure rates. The use of these methods i
t or extensions of these methods to in-plant testing require-rents would seem to be appropriate.
- d. Cver-Protection of Systems subcor=ittee members have expressed the opinion that some nuclear systems or components may be overly protected and ;
that this may lead to problems. Examples are fuses of a si:e smaller than they need . to be. In some areas it has been found that the equipment monitoring instrumentation rather than the equipent itself becomes the primary mode l of failure. Bis is a matter which a4 tht-be7xplored. . .
s%
- 9. Computer Analyses ,
Although it is recognized that the current computer data bank does i not lend itself to detailed analyses, the Subcommittee is planning to conduct several computer type analyses on an exploratory basis.
l Examples of analyses being planned include:
1
- a. Checking the ERs for clusters of events, or events that occur I l
in a given time sequence. I 1
b.
Checking the frequency of occurrence of GRs of a given typ>a )
- at one plant versus another. The purpose here would be to .
T look for frequencies that appear cut of bounds (too high or too low) in comparison to the average.
- w*. #
, = - - r , w -
- c. Conducting an examination to determine if design deficiencies -
increase or decrease with plant operating time. Suppo sedly, if corrections are made, design deficiencies should decrease with time.
- d. Conducting an examination to detennine how many MPs are repetitive. Again, if the system is operating properly, identical EPs should not occur on a repetitive basis at a single plant, and hopefuly those of major consequence will ret subsequently occur at other plants, once the data on the initial event has been shared throughout the industry and factored into personnel training programs.
Repetitive EPs can also be used to investigate the adequacy of NRC staff action (at least one such repetitive sequence consistantly refers to a Facility change Request su citted by the utility) .
1 l
l l
}
. I l
l i
t [
ORAFT Revneu of the frequency of LERs Pertaining te Reacter Caslant Syscens in 29 pWRs _R. Burns This report is part of a study involving LERS from U.S.
commercial nuclear reactors during the period of time fram 1976 through 1978. The number of LERs related to specific -
systems and subsystems varies among the reactor facilities, and the number reported from individual facilities varies from year to year. There are a variety of explanations for these differences, and they are summarized below.
(1) The technical specifications vary among reactor facilities, due to differences in reactor suppliers, architect engineers, constructors, and to changes in designs over the years. This can cause variations in reporting requirements among reactor facilities.
(2) There may be a tsndency at some facilities to report occurances more readily than others in cases of marginal reportability. (These pertain te matters of small significance). This tendency can also change with time.
(3) Some LERs report recurrence of previous events. Thus, the same problem at an individual reactor facility may be reported several times.
(4) The mode of operation affects the frequency of various kinds of inspections and susceptability of different reactor systems to problems. The amount of reactor doun-time, for example, may impact LER fraquency.
(5) Actual higher incidence of problems in a system at an individual facility should result in more frequent LERs.
(6) Reportably occurances are random events. There is some probability associated with their occurance, which is subject to the types of individual plant characteristics listed above. It is possible - in fact, probable - that even among identical reacter facilities with identical probabilities of reportable occurances there will be random variatiens in the number of LERs associated with each facility in a given time perled. Further, if variations among individual plant characteristics ,
are minor, the random variations may dominate the variability in the numbers of LERs.
Uhen studying LERs, therefore, one should expect variations in frequency of reporting among facilities and in time.
However, before singling out one er a group of reactor f acilities uhich appear to deviate from the rest in their numbers of LERs,
. and before attempting to rank the safety performance of facilities according to their relative frequency of reporting, sne should be able to recognize whether the deviations among facilities (or in time) are attributable to dominating random effectsyte lfi$k2/
m.o
1 differences in plant characteristics, or to actual higher incidence of problems.
The purpose of this work is to determine where random effects l dominate the variations in LER frequency. This is the first in a series of reports pertaining to a specific LER type. A total of 305 LERa from reactor coolant systems forallPURsincommer-4},og\& p cial operation at the beginning of the year 1976, were studied. e The work involves applications of probability theory and cceputer i simuletion of random events. The approach is to assume in a l given situation that random effects dominate, determine the amount of variability to be expected in the numbers of LERs, and reject the assumption only if actual observed data is not consistent with the predictions.
(As an example, considtr a group of 20 plants in which 20 l
, reportable events occur in a given time period. The average number of LERs per plant is ... However, if these are randem events and each one that occurs has an equal chance of occuring at each of the 20 facilities, the probability that each facility l will realize exactly one event is only 20 in a billion. In fact, the most probably configuration is zero events in 7 plants, one event each in 8 plants, two events each in 3 plants, and three events each in 2 plants. further, it is entirely possible (5% chance) that one plant can have 5 or more events.)
Results from t.his study are listed below in question and answer form.
- 1. Do LERs for Reactor Coolant Systems and Connected Systems occur more (or less) frequently in time (1976-S) among 29 PURs?
No
~~ A monotonic trend between time and total number of LERs was not found. The Xandall Tau correlation coefficient value of 0.54 is not significant. The average rate is 3,5 LERs/ year / plant.
35-No. LCRs ' Note: All pWRs During 30-were in commer-Quarter 25-
. cial operation prier to 1976 20-15 1 1 1 . . < i & 1 1975 1977 1978
- 2. Do any of the 29 pVRs report more often than is consistant uith 3.5 LERs/ year / plant?
_Yes Arkansas Nuclear One-1 reports 24 - average is 10.5 in 3 years. The chance of any one of 29 reactors reporting 24 or more events is 1/3 of 1%, if all reactors are equal (i.e. random effects dominate). The next 4
o
highest is Zion 1 with 19 LERs. This is consistent with 10.5 per plant in 3 years, since there is a 1/3 chance of at least one of the remaining 2S reactors exceeding or equalling 18 reports, due solely to random effects. J
- 3. Do any of the 29 PURs report less often than is consistent -
with 3.5 LERs/ year / plant? l No Turkey point 3 reported 2 LERs in 3 years. The chance l of at least one reactor reporting 2 or less times in i 3 years is 8%. While this is unlikely, it is not inconceivable.
4 Do any subsystems (CA-CJ in LER manual) get reported more 1 often one year to the next?
Ne The widest yearly variation in. number of reports for l a subsystem if for Reactor Vessels and Appurtenances (CA) where 5 were reported in 1977 and 13 in 1978.
The 3 year total is 27, for an average of 9 por year.
If the average is 9 per year, the chance of having 5 or less in one of three years is 30%, and of 13 or more is 12%. While 12% is unlikely, it is the wcrst dev-iation from the average among 10 subsystem categories.
The chance of the worst deviation being 12% is 70%
(likely).
- 5. Are LERs associated with some subsystems more frequently than with others?
Yes Reactor Core Isolation Coolant Systems and Controls (CC) and Reactor Coolant Pressure Boundary Leakage Detection Systems (CI) have 1 and 6 LERs, respectisely, in 3 years. The average per subsystem is 30.5 in 3 years.
The chance of any one af the ten subsystems being reported 6 sr less times is 1 in a million, if all have an equal probability of occurance. The average for the remaining 8 subsystems is 298 LERs/8, or 37.3 per subsystem in 3 years. Two subsystems are reported 6 - Reactor Coolant less oftenSystems Cleanup than the andremaining Controls(CG) and Main Steam Isolation Systems and Controls, witn 20 and 21, respectively. The chance of any one of B reporting only 20 is about 2% (unlikely). The six remaining subsystems were reported from 27 to 71 times. Onl Recirculation Systems and Controls (C8)yis Coolant reported more often than the others, with 71 LERs. Hence,,________--
v a r i a t i o n a m o Dg._s.u.b s y_s t e m s_i s _ n o.t_d o m i_n a_t e d . c y_r a n d o m_ __
'eT f e _cjs .
- 6. Are some subsystems reported more often (than is consistent l with the assumption of dominating randem effects) in some '
of~the 29 PURs than in others?
Yes (a) Feedwater Systems and Controls are reported more often in D.C. Cook 1, (b) Main Steam Isolation Systems and Controls are reported more often in Zion 1,
(c) Reactor Coolant Cleanup Systems and Controls are reported more often in Arkansas Nuclear One-1, (d) Resideal Heat Removal Systems and Controls are reported more often in Rancho Seco 1.and Indian Point 2.
Further, all other subsystems uere reported uith the -
same frequency in all 29 PURs.
Additional questions remain to be studied. These will involve computer simulation. They include:
(1) Does the LER frequency depend on the reactor supplier (U, B&V, or C-E)7 (2) Does the LER frequency depend en the size of the reactor (i.e. small (<5005Us), medium (600-1000nue), and large
(>1000MU e ) ) ?
(3) Is the actual distribution of LERs among the 28 PURs (excluding Arkansas Nuclear One-1) consistent with the assumption that all report with the same frequency?
l l
l l
l j
e
TABLE 1 LERs Pertaining'to PUR' Reactor Coolant Systems, 1976-1978 No. LERs
> 1000 MUs Reactor Sucolier O. C. Cook 1 17 U Zion 1 18 U _
Zion 2 12 W 600-1000 MUs Arkansas 1 24 B&W Calvert Cliffs 1 14 C-C Indian Point 2 17 V Maine Yankee 4 C-E Millstone 2 3 C-E Oconee 1 12 B&W Oconee 2 7 B&U Oconee 3 4 B&W Rancho Seco 1 17 B&W Surry 1 14 W Surry 2 5 W TMI 1 15 B&W Turkey Point 3 2 U urkey Point 4 5 W
< 600 MUs C-E C ,.Ft H Calhoun
- 8. Robinson 1 2 17 8
U Haddem Neck 1 16 U ,
Keuaunes 6 W L -Palisades 1 7 C-E l Point Beach 1 8 U l Point Beach 2 7 U i Prairie Island 1 3 U l Prairie Island 2 8 V l Robert E. Ginna 1 13 W l l
San Onofre 1 9 W i Yankee Rous 13 U '
l 1
. 1 29 Reactors 305 LERs r,i./ *';/,.. , r, m ,
~
i e
y(: :.c .. -
\
u --
g..
~
3, . , ~ . -
( v'~T?(./
e", 7^-.^f ^* ,n , x .
Class of Event / '
,. ,. 4,A C ' ' ' ' ' ~ '
Failure of -cont:inment monitoryvu 6 V 5 '
General Description Several instances have recently been reported in which monitors within con- -
tainment (including those designed to monitor post-accident conditions) have failed due to high .-hient. temperatures. . ,
, m v, s. . . 't. ./~ e a. s u . s. .
Frecuency of Occurrence '
t ,_ . . . .- . * * <- 1 '~ U..e-: . , -<U.
Seven such events' were reported at the Davis-Besse Power Plant Unit 1 in 1978. Six of these involved g post-accident radiation monitos At least two additional failures of post-accident radiation monitors were possibly caused by high ambient temperature. A similar failure of a post-accident hydrogen analyzer occured in the Joseph M. Farley Unit 1 in June 1978.
A containment atmosphere particulate monitor failed at the Calvert Cliffs Unit 1 facility in September 1978 due to high ambient containment humidity.
J_mplications Recardino Safety The perfomance of an instrument that fails due tohigh ambient temperatures would be suspect in a post-accident environment. Without adeq: ate post-accident monitoring, the actuation of isolation systems may not occur. In addition, in the case of hydrogen analyzers, an explosive mixt .Je could develop without the knowledge of plant personnel.
1 Corrective Action
- a. The causes of the failures appear to have clearly been determined.
Basically, the failures were attributed (in addition to the heat) to design errors in the monitors, themselves, or their sampling systems. ;
- b. The fix called for better cooling and a reevaluation of the syste-s.
This appears to be a problem that should never have occurred.
Reseer&=ay-be necessary-to-determine--a positive solution to- the . . 1 problenr. ;c ~ , . ~
- 1
.- ,, ,? . L
. l
. \
LA2 1
i 4
3;,c .,
Class of Event Containment pargirs and airborne releases. -
General Description There has been a number of MRs in recent years that relate to containment _
. purging and its relationship to containment integrity and excessive airborne releases for PG installations. ~ For example, excessive airborne releases at-one plant in 1977 led to a decision to reduce the frequency of contain-ment purging. A factor entering into this decision was that the plar. had 920 m-diameter purge lines, and the NRC prefers not to permit cor.' cinous purging unless smaller 200 mm-diameter lines have been installed.- As a '
result of the reduction in frequency of purging, airborne releases from the plant were reduced. At the same time, however, this led to a reduction in the frequency with which the containment could be entered for visual inspection of safety-related equi nent, such as piping, snubbers, etc.
In a similar situation at another Ps in 1978, minimizing the frequency ,
of purging led to a buildup of radioactive materials within the primary containment to the point where both the gaseous and particulate monitors were at or near full-scale indication. As a result, the monitors became incapable of detecting further increases in airkcene activity that could have occurred as a result of a significant increase in reactor coolant leakage. The . solution was to increase the frequency of purgirg, Frequency of Occurrence Although only a few MRs are reported annually in this category, the prob-lem appears to be generic in nature and the number of event involved may be grater than those revealed by the MRs.
Implications Regardine Safety In the first case, the reduction in the ability and frequency with which inspections can be conducted within containment could led to a reduction in overall safe plant operation. In the second case, the fact that person-nel would purge containment to prevent gaseous and particulate monitoring devices from going off scale appears to reveal a deficiency in the ranges of such monitoring equipnent. Although presumably other higher range units are available within these containments for post-accident monitoring, this was not made clear. -
Corrective Action
- a. These problems do not appear to have been completely resolved.
Problems related to inadequacies in the ranges of the radiation monitors
. inside containment were presumably corrected cn a generic basis through the development of Regulatory Guide 1.97. In supporting the develop-ment of' this Guide, the ACRS has consistently urged that more attention be directed to providing instruments with ranges suf ficient to assess a ful range of conditions, including those accompanying a major accident.
i j
l
.- - . . - . .. - . -.---. . . . ... . . . ~ . . _ - - . -
- b. It is evident that this matter needs further study and that a generic approach to its . solution is required. Although there may be an '
administrative solution to the problem, it is not obvious that it '-
has been found.
Footnote The events reported above, and those relate? *.o failure of monitors within containment (due to high ambient temperaturec snd to the isolation of the monitors that actuate the closing of containme. t p.:rge valves in case of a IDCA, are all inter-related. It is quite possible that all three of these classes of events could be discussed'under one heading or caegory of LER.
a I
h F
l.
+~ . + , . . :-w..,. ._.__,m,.,,.__ , . , , _ _ , , , , _ . , . _ , . _ , , , , _ ,
Class of Event Bypassing of monitors that acutate containment isolation. ,
General Description To maintain the pressure below the Technical Specification Limit, a number _
of licensees vent the containment through purge valves. In certain instances, such venting has occurred when the containment particulate monitor isolation signal to the purge valves was bypassed. As a result of this procedure, the purge valves would not have closed in the event of a loss-of-coolant accident.
Frecuency of Occurrence At least two events, one at Salem Unit 1 and one at Millstone Unit 2, occurred in 1978 .
Imolications Rearding Safety This situation could result, in case of a LOCA, in an excessive release of airborne radioactive materials into the atmosphere.
Corrective Action
- a. The NRC Staff is fully aware of these events as is exemplified by the fact that t:oth of the events cited above were determined by the NRC to be abnormal occurrences. Licensees have modified their procedures to preclude venting of the containment through the purge valves if a containment high particulate alarm occurs. The fix, however, appears to be administrative rather than physical or mechanical.
- b. The situation appears to call for a detailed review to determine how the possible occurrence of such a sequence of events could have been overlooked. Although neither event violated any Technical Specifica-tions in effect at the time, both occurrences resulted in the degrada-tion of the pr' y_ containment boundary and appear to have resulted from violation of basic safety precautions.
I 1
. l l
m
- 4 9
4
- ~ '"'* -
. V : :a .. . ,.
Class of Event Isolation of high-pressure coolant injection (HPCI) and reactor core isolation cooling (RCIC) systems. This problem is unique to Eh7s.
General Description
~
l In the events reported, isolation of the two systems occurred as a result i of failures in the air ventilating systems. Be basis of the problem is I that the areas through which the steam lines ts the HPCI and RCIC systems l passes are equipped with temperature sensors that are designed to isolate '
the systems in case there is a leak in the lines. If there is a malfunction i in the ventilating systems for these areas, or a sudden change in the out- )
door temperature which leads to the sensor indicaltng a steam leak, the two '
systems are automatically isolated.
l Frecuency of Occurrence
)
Nineeventsreportedin1976;elevenfin1977,severalin1978.
Imo.lications Recardinc Safetv.
- - l l
With the systems isolated, coolant injection is not available for small breaks l in the primary system piping. j Corrective Action
- a. The cause of the failure was clearly determined. l i
- b. The fix was to increase the flow in the ventilation systems for i the affected areas. Pbwever, this does not appear to address the basic problem which is that the temperature sensors do not respond solely to a leak in the steam lines.
- c. A more permanent solution to the problem appears to be needed.
?!r 'ill probet1/ .squ rei 9a daveleoment of a_nedMgn-fef-,
the-leek ser. w .
1 l
s a
4 4
.