ML20071L879
Text
/l DCT AP-January 16, 1991 r
L.
$~-
~~
NOTE T0:
Joseph Gray FROM:
Martin G. Malsch O
SUBJECT:
SECY-90-377 You asked for a brief note describing the procedural complications arising from adoption of Staff's recommendation in SECY-90-377 that a so-called " tier 3" j
level of design detail be available for audit in the certification review i
pursuant to 10 CFR 52.47(a)(2).
The two complications which I see are both sketched out in our November 15, 1990 memo and are discussed in more detail i
below.
1.
Staff's proposal would make all of the tier 3 information legally material to the certification rulemaking. This is because if tier 3 is needed for the staff safety finding then tier 3 must be needed for the Commission's certification finding as well.
This means that all parties to the certification rulemaking must have access to all of tier 3 and all of tier 3 is open to review, interpretation, and dispute among the parties.
This could add additional issues to the certification process and thereby require more resources and time to complete certification.
2.
The safety finding to be made on the basis of tier 3 will need to be carefully considered from two perspectives. First, the finding will control the form of contentions about tier 3 in the certification rulemaking since contentions on tier 3 will need to be raised as alleged inabilities to make the safety finding.
Second, too broad a finding could intrude into the later staff ITAAC conformance review. For example, if the finding on tier 3 to be made in the certification rulemaking is whether tiers 1 and 2 have been properly translated into design products, this finding could not later be revisited in the review for conformance with ITAAC (or, put another way, there could be no ITAAC testing whether tiers 1 and 2 had been properly translated into design products).
This result could change if the D.C.
Circuit's decision on Part 52 stands since the decision would allow reopening at the pre-operational stage based on new issues.
I disagree with the NUMARC view (their December 21 letter) that tier 3 would remove issue preclusion by allowing for reopening of issues previously resolved in the certification.
This view is based on a misreading of SECY-90-377 to suggest that revisions in tier 3 would automatically call into question the validity of tiers 1 and 2.
SECY-90-377 did not intend this result: tiers 1 and 2 by definition will control tier 3 and any changes in tier 3 which are 940B030309 940629 l
PDR COf91S NRCC i
CORRESPONDENCE PDR
t 2
inconsistent with tier 1 or 2 are invalid oer se.
In short, if there is an inconsistency tier 3, and not tiers 1 or 2, would be called into question.
t Incidentally, if the certification finding on tier 3 is that tiers 1 and 2 have i
in fact been properly translated into the design products in tier 3, then so long as tier 3 is not changed, the matter of consistency among the tiers is resolved and cannot be reopened.
In this sense review and finding on tier 3 will give rise to more issue preclusion rather than less.
i 4
Mar in G.
sch i
Deputy General Counsel for Licensing & Regulation cc:
Varen Cyr hjron Karman Steve Crockett EDO NRR l
l I
.o f
i 4
MEMORANDUM FOR:
Commissioner Curtiss FROM:
K.
Connaughton
SUBJECT:
SECY-90-377, REQUIREMENTS FOR DESIGN CERTIFICATION UNDER 10 CFR PART 52 j
b Backaround In the subject SECY paper, the staff responds to several j
questions posed by the Commission concerning the level of design detail necessary for certification, the two-tiered approach for J
formulating a design certification rule, the need for design flexibility, finality of issues, and a description of how staff would ensure standardization in its design certification reviews.
1 4
}
Discussion - General
}
The staff has done a remarkably good job in defining an accroach
]
for implementing the requirement for an essentially complete design as a prerequisite to design certification.
Throughout the paper, the staff discusses what design detail "can" or "could" be developed prior to certification (i.e. what is technically feasible).
Additionally, staff discussions are littered with statements concerning the promotion of standardization and its j
j attendent safety and economic benefits.
The staff does not go to any length to discuss just what those benefits are -- though they a
have been enumerated many times elsewhere, such as in the j
Commission's standardization policy statement.
j i
In my mind, there is nothing.rong with these portions of the staff's discussion; however, they have served to turn readers' i
attention away from the more important issues of just what level l
of detail is required by 10 CFR Part 52 as a procedural matter and, perhaps more importantly, what does staff need to make the i
safety findings required by Part 52, irrespective of whether or 4
not the information is explicitly referenced in Part 52.
In i
SECY-90-377, the staff addresses these important matters, l
although the relevant portions of the staff's discussion are
]
obscured by the sh r length of the paper and its Attachments.
)
These issues are discussed in the following excerpts from SECY-l 90-377:
1 l
pg. 2 "The staff examined the design process in detail to i
determine the degree of design completion necessary for
}
the staff to make its safety judgement (which will include the benefits of standardization) while recognizing what is feasible and practical under 10 CFR i
Part 52."
4 pg. 5 "The staff believes that an applicant can achieve a j
high degree of design finality (approximately 85 1
f percent) by expending approximately 50% of the engineering hours at design certification.
Such an effort would complete nearly all the engineering necessary to prepare procurement specifications."
pg. 6 "The level of detail to be developed will not exceed that normally contained in procurement specifications and. construction and installation specifications."
pg. 7 "In developing the Tier 1 requirements, the staff i
sought to standardize design details to the maximum extent practical, considering the procurement and 4
i design reconciliation process.
/k
" Tier 2 will include information demonstrating how Tier pg.
l 1 criteria are implemented in the design and will be of sufficient detail for the staff to make its safety determination as to the adequacy of the design as described in Tier 1" pg. [Qg "As stated in 10 CFR 52.47 (a) (2), that information l
/
normally contained'in certain procurement specifications and construction and installation specifications will be developed and-available for audit if such information is necessary for the NRC to make its safety determination.
To ensure that Tier 1 and 2 criteria have been properly translated into l
design products, the staff proposes allowing' applicants l
for design certification and COL to develop and l
finalize the design in a graded approach and have this material available for audit.
If, during the audit, the staff finds a part of this material necessary to make its safety determination (about the adequacy of design information in Tier 1) this part will be docketed in the application."
pg. 14 "In the Part 50 process, the staff relies on an inspection of a physically constructed facility to provide assurance that licensing commitments and regulations have been met.
In lieu of a constructed facility, the proposed regulatory guide will request the applicant to develop and make available for NRC audit information such as that normally provided in certain procurement specifications and construction and installation specifications if it is necessary for the staff to make its safety determination."
l l
App.A, pg.5 "Certain aspects of the plant are required to be I
in an advanced state of design completion for the l
purposes of staff safety review.
These relate principally to the primary system and associated protection systems grouped in Attachment C under
" nuclear island."
For these systems, the level of detail recommended for the supporting design
i products may not be significantly greater than that level which is required for the safety review in the 10 CFR Part 50 review process.
App.A, pg.6 Next in the hierarchy of safety significant systems, structures and components, are those relied upon to mitigate the consequences of a postulated accident, their support systems, and related structures.
These systems are defined in Attachment C as formingthe " balance of the nuclear island".
engineering has progressed to the point where the applicant could prepare almost all of the procurement specifications.
...The staff i
also recommends that the level of detail neceussry for the turbine island at the time of design certification equal the level required for the balance of the nuclear island.
This uniform level of detail has the added safety benefit of standardizing the systems whose malfunctions could challenge the safety systems or make recovery from off-normal conditions more difficult."...
"The level of detail recommended for design certification pursuant to Part 52 is slightly higher for systems grouped under the nuclear island than previously required of equivalent systems for a nuclear power plant licensed under 10 CFR Part 50.
In contrast, the level of detail for systems grouped in the balance of the nuclear island or the turbine island is substantially higher for a 10 CFR Part 52 design certification than was requirea for the equivalent systems of a power plant licensed under 10 CFR Part 50.
The largest difference between the two licensing processes involves the variancein the level of detail for the systems grouped'in the turbine island."
Because the NRC has no actual constructed facility on which to base a licensina decision, both test and acceptance criteria (ITAACs) and a hiah level of desian completion for most aspects would be needed, even without standardization considerations, to provida the staff with necessary confidence that these systems, structures, and components will fully implement the criteria specified in the application and will perform their intended functions. [ emphasis added]
App.A, pg.11 The staff expects no direct review effort related solely to standardization.
Att.A, pg.A-4
...the 10 CFR Part 52 design process will enable the design approach finality by the time of design
certification, because approximately 50% of the engineering hours will have been expended to produce the level of detail sufficient to allow the staff to complete the review and audits necessary to make its safety determination.
From the above excerpts, it is clear that the staff sees a safety nexus and a procedural nexus to all that it is recommending.
The safety nexus and procedural nexus both establish and limit the level of design detail to that which is necessary to develop procurement specifications and construction and installation specifications, where such information is necessary for the staff to make its safety findings.
The staff explains that, unlike licensing decisions under Part 50, design certification decisions under Part 52 will not rely on staff access to a constructed facility.
Hence the staff will rely on the development and examination, by way of NRC audits, of more design detail for design certification decisions than it did under Part 50.
rt. ct;t;;...t; ~~~ ~a i ng d 4 --"- 4 7 +Me r :.. 41
__e.
1 SUGGESTED QUESTIONS FOR GE RE: LEVELOfDESIGNDETAIL 1)
What design information recommended by the staff *sy in GE's view, goes beyond that which would normally be contained in procurement, construction, or installation specifications?
2)
What design information recommended by the staff has not traditionally been available for audit under the Part 50 process at the time the staff renders its safety j
determination that all matters associated with the design (at OL issuance) are resolved?
l 1
3)
GE has apparently adopted the industry's relatively well-defined position as to the level of design detail necessary
)
for certification.
How is it possible for GE to define its position on this matter without a proposed set of ITAAC on the table?
The NUMARC ITAAC document is embrionic, at best.
4)
Is GE planning to prepare or participate with industry in the preparation of any more detailed comments or counterproposals concerning the staff's recommendations?
1 5)
How would GE propose to minimize the likelihood of l
adjudication of design issues at the CP/OL or operating stages invul' ring the implementation of the certified design?
6)
Does GE think that the staff's proposal as it least adequate to address the flexibility issues (commercial competition, obsolesence, construction fittup, etc.)?
7)
What purpose would be served if the Commission were to use a lesser standard for FDA issuance, as opposed to design certification?
1 1
I l
l l
l V
l yj. LI A GV 1
1 i
i
)
l l
1 l
- - - - -...