ML19353A445
| ML19353A445 | |
| Person / Time | |
|---|---|
| Issue date: | 11/25/1980 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | Advisory Committee on Reactor Safeguards |
| References | |
| ACRS-1774, NUDOCS 8101090524 | |
| Download: ML19353A445 (18) | |
Text
/2C/3 - /7/ 7 7
,f..
x,[ *
/
ACRS SUBCOMMITTEE SAFETY PHILOSOPHY, TECHNOLOGY, AND CRITERIA s
V WASHINGTON, D.C.
/
4 5
SEPTEMBER 3,1980
[
.,Y h, e, The AC,RS Subcommittee on Safety Philosphy, Technology, and Criteria held a on'eday meeting at 1717 H Street, N.W., Washington on September 3,1980. The purpose of this meeting was to discuss matters relating to NRC management philosophy in developing licensing requirements and to discuss cascading failures in nuclear plants.
Notice of this meeting was published in the Federal Register on Tuesday, August 19, 1980.
A copy of this notice is included as Attachment A.
A schedule for this meeting is included as Attach-ment C.
The, entire proceedings were held in open session.
Portions of the materials provided to the Subcommittee at this meeting are included as Attach-ment 1.
A complete set of the materials provided to the Subcommittee is in the ACRS files.
No oral statements.were made by members of the public nor were there any written statements submitted by members of the public. Dr. D.
Okrent, Subcommittee Chairman, and the Subcommittee members Mr. M. Bender, Mr.
J. Ebersole, Mr. H. Etherington, Dr. W. Kerr, Mr. W. Fathis, Mr. J. Ray, and Dr. M. Plesset were present. The Designated Federal Employee for this meeting was Dr. R. Savio, ACRS Staff.
l DJ0 3 es,o S M o
The purpose of the meeting was to discuss with the NRC management the ap which will be used in addressing the ACRS concerns on cascading failures in nuclear plants and the approach which is being used to achieve safety / safety The specific areas discussed are management improvements in a number of areas.
described on page 1 of Attachment 1.
The principal participants in the meeting were as follows:
ATTENDEE S ACRS Member J. Ray ACRS Member W. Mathis ACRS Member W. Kerr ACRS Member H. Etherington ACRS Member J. Ebersole ACRS Member M. Bender ACRS Member D. Okrent, Chairman ACRS Member M. Plesset ACRS Member C. P. Siess NRC/NRR E. Case NRC/NRR H. Denton NRC/ DOL D. Eisenhut NRC/RES D. Basdekas NRC/IE R. Vollmer NRC/NRR V. Stello NRC/DSI D. Ross NRC/DHF S. Hanauer NRC/ DST F. Schroeder ACRS Staff R. Savio*
The meeting was attended by approximately 30 persons five of whom were me of the public.
Dr. Okrent summarized the purpose of the meeting and asked the other memb of the Subcommittee for comments which they might have on the matters w The issue of the. advantages /
were scheduled to be discussed at this meeting.
disadvangtages of plant standardization was briefly discussed.
There was general agreement that standardization would bring improvements to t review process in that fewer resources would be required for an even de safety review but disagreement as to the degree that this could lead to improvements in the safety,of plant design and plant operation.
- Designated Federal Employee l
l
It was noted that serious incidents often have their root in hardware /proce-dural deficiencies which have rather obvious remedies.
It was additionally noted that although these deficiencies at first sight appeared to have rather -
trivial remedies and they often point to more serious, more far reaching weaknesses in the management of the design / operations processes.
The development of the criteria for the NTCP plants was discussed.
A number of options had been discussed ranging from licensing the plants under the pre-TMI CP requirements to suspending tiie CP reviews until the completion of the post-TMI rulemaking. The Staff has recommended an approach to the Commis-sion which would require that the NTCP plants would be required to confonn to the pre-TMI CP requirements as augmented by the Action Plan for NTOL plants and, in addition, would be required to maintain some degree of flexibility for accommodating TMI rulemaking decisions.
The policy statement discussing these options and the Staff's recommendations has been approved by the Commis-sioners for issuance for public comment (SECY 80-348).
The Subcommittee criticized the policy statement as not being specific enough and has stated that a more specific NRC proposal would be required to obtain the maximum benefit from the comment process.
Mr. Denton discussed his general approach to establishing safety margins.
He indicated that he viewed safety in terms of a goal which the agency strives to achieve rather than a plateau which is achieved and maintained.
H.e expressed a willingness to seek improvements on plants as could be reasonably achieved l
in the licensing process.
Mr. Denton also indicated that he was willing to accommodate local concerns in the licensing of plants.
When this resulted in l
increased margins of plant safety.
l l
l I
The development of criteria for new (beyond NTCP) plants was discussed.
Mr.
Denton indicated that no significant resources would be allocated to this work and stated that the reason was that nn new applications were expected within a time scale that would require the allocations of such resources in the near future.
The DOE projections of a slow growth in the number of new electrical generating stations were considered in arriving at this decision.
Require-ments are expected to be developed within the context of the post-TMI-2 rulemaking. The Subcommittee disagreed with this approach.
It was noted that the post-TMI-rulemaking could reasonably be expected to take a long period of time.
It was additionally noted that projections of need for electrical generating capability have usually been some what uncertain and that in any case some new electrical generating capability appears to be required.
Deci-sions would have to be made by utilities as to the type of electrical genera-ting plants that would be built.
It can be reasonably presumed that the choice would be between coal and nuclear plants.
Opinions were expressed by Subcommittee members that the lack of established interim criteria for new plants would create such uncertainty as to assure that there were no applica-tions for new nuclear plants regardless of the public need.
This would in l
effect be a de facto moratorium on new applications by the NRC.
The validity of the single failure criteria was discussed.
The Staff's plans are presently to use the single failure criteria as augmented by probabilistic l
l analysis to develop a sufficiently reliable plant design.
Opinions were expressed by Subcommittee member-that while this approach was adequate for the existing plants some improved approach should be developed for new plant designs.
The NRC Staff indicated that they expected to gain somlinsight in their experience with the reviews of existing plants.
i l
t O
O
The generic implications of the Browns Ferry-3 event were discussed.
The question was raised to what the discovery of serious hardware deficiency in BWR systems which had extensively reviewed in the context of the ATWS work-
~
impl ied.
It was noted that the hardware deficiency resulted from a previously identified failure mode which was estimated to have a very low probability of occurri ng.
The deficiency was one which was easily recognized and could have been quite easily remedied in the design process.
Mr. Denton indicated that the details of the reactor design were reviewed only on an audit basis under the current methods used by the NRC Staff and that it would not be possible to move beyond this in a regulatory review.
He stated that he believed that' significant improvements could be achieved by renuiring more design / construction review of AE and vendor equipment at the utility level.
He indicated that trial programs of this type were in place and that the results were promising.
It wa s Mr. Denton's expectation that the scope of these programs would be expanded.
The development and use of quantitative criteria in determining the schedule for the resolution of safety issues and the implementation of the required modifications was discussed.
The NP,C Staff approach to resolving the nonseis-mic auxiliary feedwater systems issue and the probabilistic methods proposed by PAS (See Attachment 1,
ref: ltr dtd 7/11/80, Rowsome to Bernero) were pointed out as being examples of this type of process.
Specific problems have been addressed using this process but no policy exists for app' lying these techniques or for utilizing unifonn criteria in the various applications.
The process by which a decision is made to order a shutdown of the plant for safety modifications was discussed.
The degree to which the NRC understands it, the event which drew attention to the safety problems, the availability of O
systems to mitigate the consequences of the event, and the perceived probabi-lity of recurrences were identified factors in the decision process.
Individ-ual judgnent rather than quantitative criteria is used.
It appeared that uniformity of judgment existed to the extent that the same individuals would be involved in the different decision processes.
The reliability with which estimates could be made of the probability of events was discussed.
It was noted that the reliability of these types of estimates was affected by the ability of the individuals making the estimates to assess the range of circumstances which might lead to an event and by their knowledge of the details of the systems which were involved in these events.
Dr. Hanauer described NRC's program for reviewing existing control rooms and developing control room requirements.
The approach used is to conduct an onsite review of the control room using an NRC review team.
The review team typically consists of six peopl e, is multi-disciplined, and includes human factors experts as consultants. Typically the team spent about a week in each control room. Deficiencies ranged from subtle aspects of the control and read out arrangements to lighting and noise levtl deficiencies.
It was r:oted that one control room was found in which the noise level from the ventilation system was so high as to make normal voice communication impossible. Experi-l l
ence with the control rooms in French reactors indicates they utilize levels i
l of technology f ar advanced from what is used in the control rooms of U.S.
reactors.
Some of the control rooms reviewed had deficiencies which could be remedied by the normal backfitting process. There are some control rooms that will require er, tensive modification.
For these plants, the installation of a secondary control room encompassing only essential control and instrumentation is being considered as an alternative to the modification of the original control room.
O
. The proposed rule on degraded core cooling was discussed.
The rule addresses hydrogen control for accidents of the TMI-type but does not address core melt accidents involving badly degraded cores. The proposed rule requires inerting for Mark I and Mark 11 containments, hydrogen recombiners, and the applicants The evaluation of the containment capability to withstand hydrogen releases.
Subcommittee disagreed with this approach in that it did not address the whole issue of Class 9 accidents and badly degraded cores. The Committee has in the past recommended that such issues be addressed on an interim basis in the near The Staff's preser,t approach is to leave these issues for resolution in term.
the anticipated post-TMI rulemaking.
The Staff's efforts Hydrogen control in ice condenser plants was discussed.
TVA has in this regard are currently concentrated on the Sequoyah plant.
proposed a controlled ignition system to enhance the plant's capability for dealing with hydrogen releases.
In addition, TVA and the NRC Staff are conducting research programs to evaluate controlled hydrogen burning and other methods for dealing with hydrogen releases.
It is expected that this will result in additional requirements for ice condenser plants.
The public risk associated with the Sequoyah plant (in the context of the ice condenser plant on a remote site) was assessed in a PAS-sponsored study and was judged, on a preliminary basis, to be comparable to the risk associated with the WASH-1400 pl ants.
PASNY has The Staff's approach to systems interaction studies was discussed.
performed a systems interactior study for Indian Point 3 which was based on a review of selected LERs.
This approach has been judged to be unsatisfactory.
The seismic systems interaction study performed at Diablo Canyon was based largely on plant walk-through inspections by experienced personnel.
This O
.i.
The Staff is cur-type of approach appears to havr yielded useful results.
rently engaging contractors to review the state-of-the-art for systems inter-sction methodology and is developing specific criteria for the perfomance of studies of these type.
The Staff expects to have a draft addressing the methodology to be used in these types of studies by early 1981.
Page 2 of 1 outlines the Staff's schedule for developing criteria and for Attachment applying it to selected plans.
It is expected that six plants (beyond Diablo Canyon and Indian Point-3) will be involved in this first effort.
The responsibility for the work on accident cescades has been assign'sd to the Systems Interaction Branch.
Seven positions and $482K in program support has been requested for FY 81.
NRC Staff resources totaling 14 man / years, $632K in program support has been requested for FY 81.
Approaches which might be Mr.
used to approach the study of nuclear accident cascades were discussed.
Schultz and Mr. Ebersole agreed to develop plans as to how this work would be approached for discussion at a later date.
Additional details are available in the meeting transcript on file in the NRC l
Public Document Room, 1717 H St., N.W., Washington, D.C. or f rom Alderson i
A complete set of Reporting Inc., 400 Virginia Ave, S.W., Washington, D.C.
all slides used are on file in the ACRS Office with the record copy of the minutes.
l l
l t
L*
3 o(statQ b6WilriVoRt0EIIDELMSTdTo%e 7)to of Meiting P i Open-Open 9/4-9 00 (20d:
t a m 1212 00 p m-Clorid 9/4-1M p m to This dett*minati:n was made by th:r discussed, whether the meetmg has
.J 4 00 p m -Closed 9/F9 o0 a m. to 12 00 Committee min:pement 05rer pursuant t3 been c ncelled or rescheduled the p m Fort Open-Optn 9/F100 p m to 4 00 pcosioons of Secten 10(d) of Pl. 9;-4G3.The Chiirman's ruhng on requests for the f
p m-Comm.ttee httnigement O!!.ccr was (pportunity to present oral statements Cortac' person Mrs Darre) Hgns. Room dele;sted t% authorit) to n. ke such and the timt allotted therefor con be 1:50. Na tro,a! Sdence Foundanon.
determmations by the D.rceter. NSF. or. July obtained by a pre paid telephone call to
.l Washington DC2030 Telephone 200/
e,1g g' 35*-95r2 Persons planning to attend the cognizant Desigr.ated Federal
- H should notif) htrs it ;;:pns by August:8.
No'e:His not:ce is be$ng amended Employee. Afr. Pal Boehnert (telephone 1960 primanly to ch nge the times of the open and 200/634-32671 between 8.15 a.mfand i
closed sessions.The agenda is basically the 5.00 p.m EDT.
Summary M4nutes May be obtained from the same as that pubbshed m the Federal
'3 Contact person, at the above stated I
address.
- Register on Fnday. August 15.1980.
Dated. August 13.1980.
Dated. ALFust 14.1960-john C. [to}le, Purpose of Committee: To provide advice.
i recommendations and osersight M. Rebetta WinLler, Advisory Committee Management Ol pen.
conteming wpport for activit,es related to Comm;ttee Management Coor.fmator.
p1t ner. so nene r.!,4 s-ts.a> a os em; the Foanda%n's progra iin mformation a goog wg science and technolegy.
g I
Ageeda Open-Septemer 4.1980
[
Advisory Committee on Reactor
- N o m-12W P m.
Safeguards, Subcommittee on Saf ety NUCLE AR REGULATORY Philosophy and Criteria; Meeting M
Welcome and Introducto*y Remarks.
Reuew of the Daision of information COMMISSION k' i Science and Technology Activities for FY he ACRS Subcommittee on Safety Advisory Committee on Reactor
. Philosophy and Critena will hold a 1
1980 h
Working Group on Informatiert Safeguards. Subcommittee on Reactor meeting on September 3.1980 in room E
Technology. Discuss.on of Meetmg and Fuel; Meeting ING,1717 H St, NW. Washington. DC to
(([c'e!! anes' The ACRS Subcommitice on Reactor review the NRC Staff's general approach Fuel will hold a meetir.g on September 3, to developing licensing requirements in a number of areas. includmg Near-Term n;
Closed-September 4.1960 1980 in Room INS 1717 H St., NW.
~
Construction Permit (NTCP)
H Wash!n;:fon. DC., to complete its redew of NUREG-0630. "Claddmg Swelimg and requirements. the effect of nonsafety wpmaoopm.
i Review and comparison of declined proposals (and supportmg documentation; Rupture Afodels for LOCA Analysis."
systems on plant sefety functions, and with the successful aw ards under the Notice of this meeting was pubhshed the requirements for future LWRs.The Subcomnuttee also plans to begin a Division of Information Science and July 25.1980.
review of the nticlear accident cascades Technology. including resiew of peer review In accordance with the procedures issuo j
materials and other g nileged material outlined in the Federal Register on October 1.1979 (44 FR 5M08), oral or In accordance with the procedures
)
Closed-September 5.1980
)
written statements may be presented by outlined in the Federal Register on
- c.m.-J2 00 p.m.
members of the public, recordmgs will October 1.1979. (44 FR 56406). oral or f
Continuation of review and comparison of be permitted only during those portions written statements may be presented by dechned proposals (and suppertmg of the meeting when a transcript is being members of the public, recordings will
[
documentation) with the successful awards kept, and questions may be asked onJv be permitted only during those portions by mb of & ShMM N
- of the meeting when a transenpt is being onder the Daision of Information Science consultants and Staff. Persons desiring kept, and questions may be asked only to make oral statements should notify by members of the Subcommittee,its te iew ri en o er ge ma terial.
Report to the Director.
the Designated Federal Employee as far consultants and Staff. Persons desiriag to make oral statements should notify l
in advance as practicable so that Open-Sep;enmber 5.198o appropriate arrangements can be made the Designated Federal Employee as far 2 cop.mA 00p.m.
to allow the necessary time during the in advance as practicable so that i
Organizational 14cus ofIST.
meeting for such statements.
appropriate arrangements can be made j
).
Pubhc Participation.
De entire meeting will be open to to allow the necessary time during the Public attendance.
meeting for such statements.
i l
[
ne agenda for subject meeting shall The entire meeting will be open to Renon for Closing he meeting wi!! deal with a review of be as follows:
Public attendance.
l g
li'ednesday. September 3.1983. J:00 The agenda for subject meeting shall,
tee w)f re ie7va af ontaining i
P.m. until the conclusion of business, be as follows:
I" the names of applicant institutions and prindpal investigators and privileged During the initial portion of the it*ednesday, September 3.1930. P:00 information contained in dechned proposals.
meeting, the Subecmmittee, along with o.m. until the conclusion of business.
((
hi6 meetmg will also include a review of any ofits consultants who may be During the initial pcrtion of the
[
peer resiew documentation pertainir g to present, will exchange preliminary meeting, the Subcommittee, along with
[
spplicants Any non-esempt material that viewr regarding matters to be any ofits consultan's who may be i
enay be d.scussed at this eneeting (proposals considered during the balance of the present, will exchange preli ninary tot ha}e been awarded) will be inextncably meeting.
Views regarding matters to be
't pt The Subcommittee will then hear considered during the balance of the f[
presentations by and hold discussions meeting.
et n le an of r se rat on a practical E
nese matters are within esemptions (4) with representatives of the NRC Staff.
He Subcommittee will then hear
[
rnd (6) of 5 U.S C. 55:b[c) the Govemment in their corsultants, and other interested presentations by and hold discussions the Sunshine Act persons regarding this review, with representatives of the NRC Staff.
W
-~
_y a
g
. e
~,'T
'v-
r
-w
--,m
.,,.___p,,
av
,q y,
.y
(
7 y
?
a e. & s o-r
- *. t i
s
s PROPOSED SCHEDULE FOR SEPTEMBER 3, 1980 MEETING OF THE ACRS SUBCOVMITTEE ON
, SAFETY PHILISOPHY, TECHNOLOGY AND CRITERIA 9:00 - 9:15 am 1.
EXECUTIVE SESSION (OPEN) 9:15 - 1:00 pm Discussions with NRC Management on the Management 2.
philosophy and approach associated with the following topics LEAD RESPONSIBILITY 00L - D. Eisenhut a) General approach to establishing NTCP requirements b) General approach to the reevaluation of the Indian 00L - D. Eisenhut Point, Zion, and Limerick plants DST - F. Schroeder General approach to determining the need for modifi-c) cation and modifying the single failure criteria and the General Design Criteria DST - F. Schroeder d) General approach to defining requirements for future (beyond NTCP) LWRs and Standard Plants DST - F. Schroeder Basis for development and use of quantitative e) criteria for allocating resources to and resolving regulatory issues DST - F. Schroeder f) Basis for developing schedules for completion of IREP-like studies ILE - V. Stello g) Generic implementations of the Browns Ferry Event DSI - D. Ross h) Approach to devcloping requirements for ice condenser containment plants DSI - D. Ross Approach to evaluating the effect of the failure of 1) non-safety systems on plant safety functions DSI - D. Ross Preliminary approach by the NRC Staff for developing j.
recommendations for action on degraded core rulemaking DHF - S. Hanauer i
Approach to developing control room' requirements k.
1:00 - 2:00 pm l
LUNCH 2:00 - 4:00 pm Discussions on cascading failures in nuclear plants 3.
REF: ACRS from Plesset to Ahearne, August 12, 1980 D. Ross/ J. Stolz - DSI l
SUBCOMMITTEE MEMBERS _
D. Okrent k'. Kerr J. Ebersole W. Mathis M. Bender J. Ray H. Etherington e
6 b
i 0
e Y
5>
e N
N
=*
ng E
ca Eh 7C 3
% yn - G '."=*
en =W t.fl' 49 C
sut N em e
g
=
3c d, 69 N9 5 te *=
m
.=
es c.
e=. e esp
==
e
]
W*
be. 1==
ere g M i
ge
.D l
00 r
W C.
E *==
G e.J n,
C e
r9
.. J.C.
- P =, 2 8
O
.= z g 0"
N El m
s l
l Er i
2 m
- v. e l
es*
A E
C l
VL. =z=
.=
>= 6 6 >
e E
8 t
b E
= = =
Ee A
sd' Z ae.G. e
-l sa a-c ts:==
w v
- w =.
e 6=
=
E
- b. e E. ew ar*
v'
~
9 E
C J - &.T 3,1 es > J 4>.=:
z : : e a
c c1 =
L. C-wq-b
-l e
.o v= ~
C
=>-
ae z E.
== T.
39 1
, e. r m,.
s.
I er L
% e
.ea F
C e.J s.
g-
==
-r n-
==. 2 < f.
g
- v. L 6e
= g
"" 49 E
E g z O
es es 3 em aC.7 =L 2 =
I Q c..=.
ec
= =e s
.c.=.
>==
- =t Ca4 3J r"*
e to e C
6 c:
C i
. =,
go w aC e.
wC g
E M
==
at==
l r;
s s.
e'C 6-
- o.
l 2
- g e.
C g
a
=
C.
h= v
.=
e-C ec l
l C
== **
L'
= e e
E.
(* I O
@h C" (
- g
- e= w== r :=
*N" eC== et=
C. g
$ I.
.f%=
6 "...
m E'
,o.
- L
.6 4
== a
!C
)
L L
W
- 2..
- ma a C:
e'-
C
==
C' M
d b
! E M b' W e 5
~)
l e
c 7
C ne. -
==
C C
==
- 7. f."
L.*
= O
."T Cj g,
e c.=.
g,;====..r:=
C s* aC e=-
y
-L l
- e 2.=.0.*
J==
- 6. s N
i
=
w e.
6==
.C=
w 4.
C $ =. =.:'
==r
. =
s.;.*: -
g i
a as
==
C*
$3 C
C O
e.=..
=
C m.=
p w c Nh c> J'" "I m
w L
e e
c
=.
a C
a a 69 g3 C sso 4
8.
=.J C'
On C. E O*
XX L* l 8=
f
>=
at, es.
C (S
- E ZF 8
C
%== e E
eur==. aC
-h.e N
L h
.% w %
za f*=
i, =;l e
=>
O V
4
"': 1
.en 3
b l
l d
.C.
== '
==
w e
O W
- P
>==
ed w
te *==
5 E
3 8"*
l 2C G en E.J Y*
u
=a
== 3 i>
- M M=
= * "
- eo Z 8
e...*=.
ann.
set =*
.s P"t w e M.*
.e g
t'e si e
C.s b
C.
a bd WOM w
w e p O.
si m
t c
C= #
sy E=
EC eJ *
==s C, e.'
)
9 T E
a e-2
=W a
=.4 G
g 37 w =D 6
art ts f.2 s
ONE g-8 e E 2
.3
=
.J===
g i
- g*=.
== 1
- 4. #
= 3 se w g o.s E
- t. *:
-=
.:l:.-
se
==
h.
en #.
m c
- =
~
- u g/
A@w em ens
(,
d m.
'M h'
...../.nL..d iV.&.A i1%....-
1 t
'o ocas g UNITED STATES s
af g
PJUC' EAR REGULATORY COMMISSION E
3.. gf ;
W ASHINGT ON, O. C. 20555
/
.E
%,o'..'[+/
quL 31 ELO o
MEMORANDUM FOR:
Robert M. Bernero, Director Probabilistic Analysis Staff Office of Nuclear Regulatory Research Frank H. Rowsome, Deputy Director FROM:
Probabilistic Analysis Staff Office of Nuclear Regulatory Research BACKFIT DEADLINES BASED ON HYPOTHETICAL INTERIM
SUBJECT:
ACCEPTABLE RISK NUMBERS We might accept up to a 0.1% chance / unit of a significant Hypothesis 1:
accident, i.e., one of the seriousness of THI or worse, in the interval between the discovery of vulnerability to a short tem fix, another 0.1% chance while we decide upon a midrange policy on backfits, and a third 0.1%
chance for the rest of the service life of a unit.
A worst case plant would have a lifetime probability of
.003 or 0.3% chance of a significant accident (after Corollary 1.1:
discovery).
A population of 100 worst-case plants would have a 75%
chance of living out their days without an accident.
Corollary 1.2:
One can define an allowable delay for a short term fix Corollary 1.3:
from the 0.1% criterion:
As t (.001
%.001, t,t.001
+ 1 3 t
A 3
3
+ t 1 10 years
- s = 10 /yr.
s
)
A, = 10-3/yr.
+t1 1 year s
A, = 10-2/yr
+t1 1 month s
~
JUI. I l 530
' Robert'M. Bernero A ten year period to settle upon and implement a long cy impi%s the midrange acceptable Corollary 1.4:
termretroff,tpoy/ year.
rate is A,
- 10 A new plant with a 40 year planned service life must meet, during the 30 year period after the retrofit Corollary 1.5:
policy goes into effect, a long term rate A
=.00 3.3X10' / year t
30
~
-2 We can be fairly confident that A actual 1
/ year Observation 1.1:
outliers among the operating plants.
To meet the goals overall, one should calculate the allowable short term exposure with the time-to-discovery Observation 1.2:
E10'geethereprobablyaresome, then Si added to the time-to-fix.
IREP is moving too slowly to meet 0.1% chan:e/ plant outliers with A actual However, we are likely to be within a decade of the go The population of reactors had a failure rate for sign accidents, A actual, distributed as follows before TMI:
Hy: othesis 2.:
AA_
f Units _
f Unit Years 28.57 10-2 5
Avg. 5.71 year / unit 57.14 3X10 10 for 400 reactor years
-3 57.14
-3 10 among 70 units 10 85.71 3X10~4 15 85.71 10 15 85.71 3X10-5 400 Unit Years 15 Total:
70 Units n
The occurrence rate for serious accidents in the 70-pla A =.096 per year.
populationisthenAindustry=lng Corollary 2.1:
g This implies that the Mean Time Between failures.(M
[
1 HTBF = 10.4 calendar years.
m
JUL 111980
.- Roter.t M. Bernero The probability of no accidents in 400 reactor years is
{
~*ind
=.57
-}
e
+ P (one or more accidents) =.43 in 400 r yr.
This appears to be consistent with the TMI experience Observation 2.1:
and the occurrence of several close calls.
Suppose the THI experience and ratchets lowers the top
$ pothesis 3.:
I two x's by X / each.
3 What then is the industry A and MTBF7 Corollary 3.1:
A ind =.043/yr -* MTBF = 23.2 years What, then, would be the probability of going 3 calendar
-A d
- 3 yr " "U Corollary 3.2:
P (3 yrs) = e 1=
years after TMI accident free?
o 1
P (one or more accidents in 3 years) =.32 conversely We can test the plausibility of hypothesis 3 by considerin In the 1.3 years Observation 3.1:
the experience with precursors since TMI.
since TMI we have had one close call (Browns Ferry AT which may have come within a decade of TMI-level severity (one of ten such events might produce extensive co Other precursors, such as the February incident at CR-3 ar 1 think - more like two decades or more from ex damage.
Suppose 1 (incl. precursors)= 10 X A (ind core damage
}
Hypothesis 4.:
Together with %pothesis 3 we would expect A incl. pre =.43/yr. MTBF pre = 2.33 yr.
l l
1
JUL 11580 Rcbert.M. Eernero The likelihood of getting one or more precursor events of the severity suggested in Observation 3.1 and Hypothesis 4 Corollary 4.1:
- j
~
in the 1.3 years since TMI is
-Apre X 1.3 yr,,43 P(14) = 1 -e This seems reasonably congruent with the experience.
Observation 4.1:
At exactly 4 calendar years after TMI. all plant Hypothesis 5.:
cases and 80'; of the 10'3 cases
-3 catches 90: of the 3X10 and 70% of the 3X10~4 cases and - where identified reduces these to 10'4 cases, i.e.
Before - IREP +++++ After A actual = 3X10'3 P =.1 A actual = 10'3 A act. = 3X10'3 P =.2 A actual = 3X10'4 P =.3 A actual = 10'*
P *.4 Before w 1 REP - After P =.2 Aa = 10
-3 A act. - 10 la = 3X10'4 P =.3 la = 10 P =.5 A a = 3X10~4 P =.3 4
~4 A act. = 3X10 P =.7 A a = 10 1
Aa,_ 10' before are unchanged.-
~
The model..as thus far developed, stands as follows:
w th5r.d f;. Bernero After TitI, Before IREP After IREP_
(4 years) 9A Before TNI Effective f Unit i of Units # A fof! Units 9A y.
-2
~3 5
10
-3 0.5 3X10 5
3X10
-3
-3 10 3X10 5.0 10 m '4 f
20 10
-3 10 10
~4 m
-4 15 3X10 10'4 15 3X10 37.5 15 10
-5
-4 15 10
-5 15.0 _
3X10
-5 IS 3X10
_15_.
3X10 79,n 10 70 Aind. =.0143/ year.
Aind. =.043/ year Aind. =.095/ year MTBF = 70 cal. yr.
MTBF = 23.2 cal. yr.
= 4900 Unit yr.
HTBF c 10.4 cal.yr.
= 1624 Unit yr.
= 729 Unit yr.
P (10 calendar P'(4 calendar years years w/o P (400 reactor years w/o accident) accident) w/o accidents)
=.84
=.87
=.57 P.(14 ace) =.13 P (1+ ace) =.16 P (1+ acc) =.43 Uhat, then, is the probability of remaining accident-free long term from TMI through 4 years to IREP an,d_ 10 years thereafter to th Observation 5.1 P (14 accident free y' ears) =.84 X.87 =.73 "f.x"?
P ( 1 or more accidents in 14 years) =.27 according to this uncertain model.
-s l
f R
,/ s,
Frank H. Rowsome Deputy Director Probabilistic Analysis Staff Office of Nuclear Regulatory Research t
M. Cullingford, PAS cc:
I
?
July 25,1980
_t
~
a i
"l NOTE T0: Robert M. Bernero, Director Probabilistic Analysis Staff, RES Frank H. Rowsome, Deputy Director FROM:
Probabilistic Analysis Staff. RES ADDENDA TO "BACKFIT DEADLINES BASED UPON HYPOTHETICAL
SUBJECT:
INTERIM ACCEPTABLE RISK NUMBERS" I have added a few more corrolaries to my calculation.
Corollary 5.1 The model developed in Hypothesis 5 pennits us to answer the question:
What fraction of the statistical exposure to one or more significant accidents in the 14 year period after TMI is due to:
Answer % 50%
IREP moves too slowly?
Answer % 33%
a.
b.
IREP misses too much?
the 10-4/yr interim criterion is too Answer % 17%
c.
lenient?
Observation 5.2 The balance of risk between IREP speed and thoroughness is likely to be However, I an artifact of the assumptions, i.e., not very robust.
suspect that an uncertainty analysis would show that the 10-4 interim criterion is unlikely to be a weak link in the strategy, i.e., the smallness of risk associated with its leniency (as distinct from IREP completeness) is a robust conclusion.
Observation 5.3 One could replace the assumption of an instantaneous short tenn fix associated with IREP in Hypothesis 5 with the phased fixes described in Hypothesis 1 and then assess where the weaknesses of t If we do this over with a proper treatment of uncertainties, we tuning.
might then do sensitivity studies on hypothetical strategies like 1
Hypothesis 1 and 5.
m-,-
_~,, - - - -
, July 25,1980 Robert M. Bernero Observation 5.4 s
- to alter the model to address only killer accidents, one
-)
If we were-would expect the following changes in results:
l failure rates and probabilities would be of the order of 1/10 those o
cited for significant core damage IREP screening effectiveness is probably a little better than for j
core damage scenarios, i.e., a smaller completeness error for o
killer accidents, at least at comparable l's If we leave the criteria of Hypothesis 1 unaltered but consider the balanceofriskdistributedamongIgEPspeed, thoroughness,andthe
}
o interim criterion for killer L.
residual risk associated with a 10 accidents only, we would find a more nearly unifortn balance, i.e..
This IREP is probably well-optimized for public health and safety.
refinement, too, should be considered if we elect to do this job again with uncertainty analysis.
'/ J
. Frank H. Rowsome, Deputy Director Probabilistic Analysis Staff Office of Nuclear Regulatory Research
... =
, -..