ML20127B062
Text
.
'6
[ene.
f UNITED STATES NUCLEAR REGULATORY COMMISSION
~'D a
,q i
wasmwoTow, o. c.nosse
- /
NOV 171983 j
HFMORANDUM FDR:
Darrell G. Eisenhut, Director Division of Licensing, NRR
- .11 Edward Christenbury, Director & Chief Counsel
,j Hearing Division, OELD
- j Roger J. Mattson, Director Division of Systems Integration, NRR FROM:
Themis P. Speis, Director Division of Safety Technology, NRR
SUBJECT:
DRAFT DES SUPPLEMENT - LIMERICK This memo is in response to a DL request (Tom Novak, 11/10/83) for DST to concur in the publication of the DES supplement for Limerick.
Frank Rowsome discussed our concerns with Tom Novak, Joe Rutberg, Jerry Hulman, et al, on Tuesday, 11/15/83.
In that meeting it was agreed that the problems wTth the draft DES supplement were of two kinds:
- 1) minor editorial changes inethe draft that can and will be worked out between Hulman, Rowsome, and Thadani and need not concern us further, and 2) policy problems associated with the packaging of the hearing testimony on risk, which should be elevated to the division director level for guidance. This memo addresses these broader Concerns.
4 In essence, the packaging of the risk testimony, as currently planned, may -
set some poor precedents and may make it unnecessarily difficult for the board to limit frivolous PRA-based contentions. However, ary plan to reorganize our risk testimony to avoid these disadvantages in the current plan will entail slipping the promised delivery date for the PRA-related material to the board by weeks or months. The board was planning to take up the DES shortly after it convenes on December 7,1983. The filing date, for the DES has already slipped; it was November 1,1983.
The question before us is this:
Is it worth the further disturbance in the hearing schedule to reorganize our documentation of the inquiry into risk at l
l Limerick in order to avoid the problems with the current packaging of this material'l The current plan for documentation and filing hearing testimony on risk at Limerick, and the disadvantages associated with it, are described in l'
Attachment I to this memo. These need review and comment by the involved NRR division directors and ELD to judge the value that should be accorded to the options for reorganization in Attachment 2.
We also need an estimate from ELD and DL of the impact of the hearing schedule perturbations I.
1 t' g m I 2%
(
16
4-ij - l
.t l
j. NOV 171983 associated with the options.
I suggest that we meet within a week to l
compare notes and settle upon a course of action.
We think Option 3 is preferable but if the schedule impact is severe, we
- )
should elect Option 2.
Frank Rowsome and Ashok Thadani are available to help implement our decision.
I h~
4 - --
= - - - -.
Themis P. Speis, Director Division of Safety Technology, NRR
- - -. Attachments:
As stated cc w/ attachments:
J. Scinto J. Rutberg R. Purple T. Novak R. Martin D. Muller J. Hulman M. Spangler F. Rowsome FT H. Thompson
- ...r.
1, _,,,
-s
. - _ -. -- _~
S 9 4DO.
.Me=
me-L.
e.
G
,.,. 7
?
- j i7
,u ll ATTACHMENT 1
.1 j
Plan for Documentation of the Inquiry into Risk at Limerick R
BACKGROUND N.
The staff has committed to the Limerick board
- to utilize the inquiry into risk at ' Limerick in four ways:
1)
The plant specific PRA results will be used in place of the i
generic release categories and accident likelihoods in the severe accident environmental statement.
2)
The PRA will be utilized to look for instances of noncompliance with the regs.
3)
In light of the high population density surrounding Limerick, the PRA's will be used to identify whether plant-specific vulnerabilities to severe reactor accidents pose undue risk warranting plant-specific requirements beyond.the norm.
l 4)
The Limerick PRAs may be used as an evaluation tool in generic standards development, e.g., the Severe Accident Rule.
Current Documentation Plan At present, the staff plans to publish a DES supplement which goes beyond the commitment to use plant-specific release categories and frequencies in the otherwise usual severe accident considerations under NEPA. The draft contains extensive' documentation of the staff's severe accident consequence analysis (similar to a portion Indian Point hearing testimony on Commission Question 1) plus an extensive inter plant comparison of bottom line risk estimates (going well-beyond that done in Indian Point hearing testimony).
It does not document the basis for the staff's estimates of accident likelihoods or containment performance.
It does not document the staff's consideration of whether the risk warrants requirements above or beyond compliance with regulations.
It does not document the staff's inquiry into whether the PRA suggests instances of noncompliance.
The original plan called for this DES board' submittal to be followed by a short memo from DST to Denton, which would be submitted to the board, which would address 1) whether the PRA revealed instances of noncompliance with the regs., and 2) whether plant-specific vulnerabilities warrant requirements beyond those in the regulations. This memo is expected in
-mid-January, 1984.
""NRC Staff Response to the Licensing Boards Order of March 21, 1983",
dated 4/13/83 and supplemented 5/24/83.
--r-y y,,,
-_g_y,_,
,_y_
-w
- j.;.
_a 1
- b.
j.=-
c
- l -
1
-z-1 Problems with th Documentation Plan l
- 1.. Nsproportionate documentation of the inquiry into risk 3
Were we to follow the current plan, we would not be giving as much.
c.i coverage to accident likelihoods or containment analysis as we are to ff consequences. The conception of a brief memorandum on PRA-based
.d decision making leaves the record thin on how the staff arrived at its
- 3 PRA-based recommendations in this precedent-setting case.
2.
It sends the wrono message to industry 3
Introducing the severe accident iisk assessment.of Limerick in the context of bottom line risk rather than in the context of lessons learned for matters of design and conduct of operations will strengthen the industry's perception that the NRC will accept PRA as bottom line propaganda. Although the PRA submitted by the licensee has been rebaselined by the staff to eliminate nonconservative biases and omissions, such a presentation emphasizes absolute and comparative bottom-line applications. The Commission, on commenting upon the safety goal evaluation progress last month, urged the staff to take advantage of the insights PRA affords while holding off on bottom-line conclusions, pending the evaluation. Thus, it is a poor practice,to surface the staff's inquiry into Limeris;k risk first in the context of bottom-line results.
3.
It may set a precedent tending to require plant specific PRAs in all environmental statements Surfacing the PRA first in the DES arena will add substantially to the,
weight of precedent behind requiring plant-specific PRAs of all plants going for OLs. Hearings for Midland and Shoreham have already pointed in this direction. The staff has maintained that the Commission policy statement doc:; ::ot require plant-specific PRAs, although case after case chips away at,the credibility of this position. This could be the straw that breaks the camel's back. While a case can be made that PRAs are desirable for all plants, I would rather see the case madts.in the safety arena--as a desigr)'and conduct of operations tool.
4.
NEPA The Commission, in-its policy statement on severe accident considerations under NEPA, wished to make gesture to acknowledge the credibility of core melts.
NEPA was never meant to address very low
}
probability environmental impacts. We have come a long way since then i
in the use of PRA in safety. The framework we have established to deal L
with high population density sites is quite satisfactory and is n9t tied to NEPA. By amplifying the distortion of NEPA, the Limerick j.
sets a bad precedent not only for the nuclear industry but for al), DES contexts covered by NEPA.
I
=
_..-,-,,,-,-.--..._~,.-----.-.v--.--.
m.
+
O T
l l P 4
5.
Overemphasis on inter-plant comparative risk The Limerick draft DES supplement displays comparisons of Limerick
- j results with many ESs of plants without a plant-specific PRA as well as plants with full PRAs. The explanation of the limitations of this analysis is skimpy. Staff testimony at the IP hearing indicated that i
such comparisons were, in large part, " poise" rather than " signal,"
due to methodological and scope differences. The IP board was far more emphatic in rejecting such comparisons as meaningless. Thus, the draft DES supplement employs a largely discredited comparison that will be difficult to defend without more extensive caveats and serves little purpose.
6.
Basis for board evaluation of the admissibility of contentions The board has announced its intention to limit PRA-based contentions on safety to identified specific design problems and to exclude issues going solely to PRA methodology.
PRA-based contentions on environmental issues may go to'the bottom line.
If the staff files bottom-line NEPA material well in advance of the remainder of the risk testimony, and. embraces more than we need to under NEPA, then we make it more difficult for the board to identify the boundaries of admissibility and to screen contentions.
8
- 7. J.. __ -.
.;.7..
'_':._ _ c._;~..
' =a -:+:. --..: =
= :.:. T ;~s
. T= '"=.=.' = : -.
-' ~ -
9 9 e
e p
s
.e y9 y
- n
.y,
.m, e.
y
,w
Q..-
- ,1, n..
4 d
ATTAC MENT 2
?a!
Options to Reorganize the Documentation of the Staff's 1.1 1
Inquiry into Risk at Limerick i
"l Option 1: Stick with current plan.
I N
We might take some-care to minimize the undesirable signals or f
precedents in profiled testimony, and we could move in other arenas to signal the industry that the Commission is looking to PRA as a risk management tool rather than as bottom line propaganda. The hearing schedule would be unaffected beyond the already overdue filing of the DES supplement. However, most of the problems would remain.
Option 2:
File the DES supplement,'d[tleting the excessive inter plant comparisons of bottom line risk. Replace the internal memorandum with a NUREG report on a) basis for the staff's accident likelihood estimates,,
b) basis for the staff's containment ~ analysis, and c) fully develop basis for the staff's conclusions on requirements for Limerick based upon the high population density concern with disproportionate risk.
Some edited version of the inter plant comparative risk assessment would-be included in the NUREG.
Schedule Impact:
.The DES would not be much later than it would be with Option 1.
The second, NUREG filing would be delayed around three months from late January to April.
Effect on Concerns:
The disproportionate ' documentation would be alleviated, although it would be out-of-sequence and thus difficult to litigate, with consequence analysis preceding by four months the likelihood /
source term basis for the consequence analysis.
Such a sequence of filings would continue to convey disproportionate emphasis on PRA bottom lines, and make it hard for the board to screen contentions. The misuse of NEPA would be partially mitigated.
~
. ~
.s.r
- )
2-Option 3:
'l Withhold the DES supplement for a simultaneous filing with the NUREG
.L report described in Option 2.
/
Schedule Impact:
<L y
The DES filing would be delayed roughly four months and the rest t "'
of the PRA material.three months compared with current plans. The board would see neither DES supplement nor other rirk material
.y until April.
Effect on Concerns:
The e.oncarns could be fully addressed. The PRA inquiry could be cast iin terms of design and operational insights, rather than giving undue attention to bottom lines.-- The NEPA considerations could be given the subsidiary treatment they deserve. The logical development of accident likelihoods, followed by containment analysis, followed by consequence analysis, followed by staff recommendations in one simultaneously-filed package would not perturb the boards judgment of the admissibility of PRA based contentions.
g t
=
e.m.
~
se l.
~
S 9
9 4e o
m
--+4 m
66 l
F-T
--w--b y
72-=
e w
r-eV
--w-=-
J vm r
erw
- - - - * * - - -