ML20056D985

Comments on Regulatory Review Group Rept.Suggests That Removal of Requirements for PASS Be Utilized as Test Case

ML20056D985
Person / Time
Site:	Yankee Rowe
Issue date:	07/23/1993
From:	Denise Edwards YANKEE ATOMIC ELECTRIC CO.
To:	Gillespie F NRC - REGULATORY REVIEW GROUP, NRC OFFICE OF THE EXECUTIVE DIRECTOR FOR OPERATIONS (EDO)
Shared Package
ML20056D986	List: ML20056D985 ML20056D993
References
FRN-58FR29012, FRN-58FR33285 FYC-93-018, FYC-93-18, NUDOCS 9308190139
Download: ML20056D985 (43)

Text

{{#Wiki_filter:f YANKEE ATOMICELECTRIC COMPANY "'y"'l'?lgT,'$"g'"

• Q-WO 580 Main Street, Bo!!on, Massachusetts 01740-1398 Yauxts July 23, 1993 FYC 93-018 SPS 93-069 Dr. Frank T. Gillespie, Chairman Regulatory Review Group Office of_the Executive Director for Operations United States Nuclear Regulatory Commission Washington, DC 20555

Subject:

Request for Comments on the Regulatory Review Group's Report to the Executive Director for Operations (58FR29012) and Comment Period Extension (58FR33285)

Dear Dr. Gillespie:

Yankee Atomic Electric Company (Yankee) appreciates the opportunity to comment on the subject report. Yankee is the owner of the nuclear power plant in Rowe, Massachusetts. Our. Nuclear Services Division also provides engineering and licensing services to sponsor companies in New England and other nuclear power plants in the United States. Yankee has been specifically charged by our sponsors to aggressively pursue revisions to regulatory processes as well as relaxation of specific requirements. Many of the specific areas identified by our sponsors as deserving attention have been. discussed in the subject report. The Regulatory Review Group (RRG) is to be congratulated for an outstanding effort. The scope of review is comprehensive, and, for the time available, the depth of review of many issues is noteworthy. The innovations presented as alternatives to some of the current obsclete. regulatory paradigms are both original and refreshing. Finally, the extension of the comment period to 60 days from the original 30 was_a relief for those reviewing this-extensive document. Discussion The Review Group noted, at the beginning of its report, that, during'the review, special emphasis would be placed on the feasibility of substituting performance-based requirements whenever reasonable. To this end, one of the cnun b af 9308190139 930723 PDR. ADDCK 05000029 ak P PDR v

( Dr. Frank T. Gillespie July 23, 1993 Page 2 most profound statements contained in the Review Group's report appears in Section 2.3.9 under III

• Regulatory Coherence" (p. 67).
• We believe that a commitment is to the basic articulation of the requirement and not to the method by which the licensee initially stated a commitment would be met."

This statement highlights the essence of a performance-based regulatory system and at the same time the major barrier to its achievement which exists currently. Although the obligation of licensees to meet all " requirements" (i.e., rules) is unquestionably clear, the differentiation between satisfaction of the objective of a requirement and the method of carrying out this obligation or commitment has historically been very difficult. A further complication, which makes characterization of commitments much more difficult, is the practice-(which has grown in the recent past) of issuing requirements in forms other than as rules (e.g., BWR Mark I containment "hard" vents). Clarification of the distinction between compliance with a requirement and the individual licensee's prescription as to how compliance is to be achieved is a fundamental attribute, we feel, of a performance-based regulatory system. The Review Group Report provides a practical example of an approach to that problem. (Section 2.3.29 - P. 133). "The security plan is required. However, the plan is a demonstration of how the applicant will comply with 10 CFR 73 and is not, therefore, a requirement in and of itself. Changes to-the plan that do not reduce the licensee's ability to meet 10 CFR 73.55 effectively but actually reduce the resources [ required to do so] should be acceptable without prior approval. The burden is then on the staff to establish not that the plan commits less resources, but that 10 CFR 73 is not being effectively met." The variety of " commitments" that exist today makes the achievement of uniformity in this area, consistent with the above example, a difficult matter. Our contribution to this discussion is included in Attachment 2 in the response to the recommended definition for " commitment". We' commend the Review Group for their insight and for their effort to deal with this issue. Structure of the Response There are 60 recommendations presented in the report. Most appear to be very good suggestions and should, we believe, be implemented as soon as reasonably practicable. The ones which we strongly support in the form presented are listed in Attachment 1. In this and the other attachments, the CM\\341 -r

a i Dr. Frank T. Gillespie July 23,1993 l Page 3 i i page reference is to Volume I and the section references are to the other volumes. ] Several of these recommendations, however, warranted specific comments, in our opinion, because they dealt with very significant issues for which we would like to suggest alternatives or additions to the proposed action. The discussion of " commitments" is an example of this kind of issue. These comments are provided in' Attachment 2. There were a few comments contained in the report which we feel deserve qualification or rebuttal. These are discussed in Attachment 3. Finally, one aspect of the report deserving attention are the issues not i substantively addressed or fully developed. The three prominent ones in'this category, in our opinion, are: Inspection Program Improvements, Seismic Design Requirements, and Generic Treatment of Relief Requests. Each of these are mentioned but none, other than relief requests, briefly discussed in conjunction with a revision to 2.802 (p.8 and Section 2.3.17), is developed to the degree that we believe they should. The following sections of this letter summarize discussions and recommendations for these issues which are included as separate enclosures. issues Deserving of More Attention in the Report inspection Program improvements (Enclosure 1) l The inspection process is "where the rubber meets the road" so to speak for implementation of the regulatory process and, to a large degree, for manifestation of regulatory philosophy. Any modifications to achieve a performance-based system and to reduce regulatory burden must, ultimately, be expressed as changes in the NRC's inspection program. It would be interesting and informative to obtain the Review Group's comments on the inspection program. To be effective, many of the Review Group Recommendations must have an inspection program component. Additionally, based on discussion at the Regulatory Information Conference, it is not clear that the NRC Task Force report on inspection program reform will be'made available for public comment. Hence, the availability of the Review Group's i comments may also_be questionable. Because of the significance of the inspection program to reform, it is imperative that both of these reports-be made public as soon as they are available. Without any. indication of what might be contained in either report, we have developed a series of proposals for improvement in what is acknowledged i usuu

Dr. Frank T. Gillespie July 23, 1993 Page 4 i as an overly burdensome inspection process. Fully developed in Enclosure 1, i these include the following suggestions-l 1. More realistic criteria and controls for major team inspections should be established. 2. The NRC should perform regular reviews of the effectiveness of special issue major team inspections. 3. The NRC should improve management oversight of inspection activities to provide assurance that individual inspectors do not impose inappropriate regulatory actions or positions. 4 There should be a strong correlation between licensee performance and the inspection burden. 5. Proposed rulemaking packages and changes in regulatory guidance -l should also include the related changes to the NRC inspection manual. 6. The expenditure of resources for each inspection should be clearly documented. Seismic Desian Recuirements (Enclosure 2) Over the last 15 years, the Lawrence Livermore National Laboratory (LLNL), at the direction of the NRC, has developed and refined probabilistic models to assess the likelihood of exceeding the seismic desiga basis at existing nuclear facilities in the eastern United States (EUS). During the eighties, with encouragement from the NRC staff, the Electric Power.Research Institute (EPRI) also developed a state-of-the-art seismic hazard methodology. Comparisons of the 1985 and 1989 EPRI and LLNL results at typical EUS sites showed dramatically different perceptions of seismic hazard, with the EPRI results being much more favorable. As a result of continued discussion between scientists, as well as evolutions in the state-of-the-art, today's LLNL results (1992) compare favorably to the EPRI results. at most EUS sites. Significantly, while the plant-specific hazard associated with seismic events computed-by LLNL and, thus, the perception' of seismic hazard that it connotes, has been steadily evolving in a direction of reduced hazard, the staff has continued to press for detailed seismic reviews of all plants (the j seismic IPE). In fact, with the older vintage plants are being required to perform not one, but two seismic review programs (IPE/EE and a review to satisfy Unresolved Safety Issue A-46). l 1 CN\\342. r

Dr. Frank T. Gillespie July 23,1993 Page 5 This _ topic not only should be included in the report, but deserves special commentary. As is explained in Enclosure 2, there has been a transition in the Staff's belief that the seismically induced design basis accident was a low probability event (pre-1380). to the belief that it might be highly likely (early 80's), to a consensus (LLNL and EPRI results) that this design basis accident is, in fact, a severe accident of very low probability. This issue constitutes a unique opportunity to relieve an enormous regulatory burden with no impact on the current level of plant i safety. Generic Relief Process The Review Group Report highlighted the value of a generic safety evaluation as the means to improve the response time for licensee relief requests (p.5). A generic safety evaluation which could " envelope" many rMants is a very worthwhile tool which deserves immediate development to its full potential. A few changes to the generic Technical Specification or license improvement process currently employed could result in rapid action in most cases with a significant reduction in duplication of staff efforts. This process is described in Enclosure 3. l Briefly, the process would involve: 1. Issuing a draft generic letter containing a generic safety evaluation and offering a specific improvement opportunity for public comment. 2. Asking in the draft for an expression of interest by licensees and for an estimate from them of the time required to prepsre plant-specific submittals. 3. Issuing the final generic letter containing a filing deadline by which interested licensees could be involved in a " batch" for generic treatment. 4 Reviewing and processing of the individual requests ty a single Issue Manager who would prepare the staff's "no significant hazard" analysis utilizing the elements of the submittals received. 5. Presenting a single notice for public comment in the Federal Register listing all the plants for which the notice has been prepared. (76\\343

a a 4 Dr. Frank T. Gillespie July 23, 1993 i Page 6 6. Resolving of comments, if any, and issuing the revised pages to all the plants included in the notice. t It is our belief that nothing in the 'Sholly" amendment to Section 189a of the Atomic Energy Act precludes the treatment of more than one plant at a time in the steps required to satisfy the legal requirement of the Act. Additionally, such a procedure would enhance rather than inhibit public involvement. In effect, the proposed change would be publicly available in draft form and final form as a Generic Letter and once as a public notice for all of the plants involved. E i We suggest that the staff consider applying the proposed process on a trial basis. Specifically, we suggest that removal of the requirements for the Post-Accident Sampling System be utilized as the test case. The NRC has i already developed the necessary technical justification in NUREG/CR-4330. We believe that relief on this issue will afford many licensees significant cost savings with respect to on-going training and maintenance requirements. Conclusion We are most hopeful that the Regulatory Review Group's Report heralds the onset of a fundamentally new approach to regulation: that is one in which the regulator and the licensees can work to establish a common set of performance objectives. This would set the stage for a process in which real advantage can be taken of the insights from risk analysis. The first step of this approach is acknowledgement that the current body of regulations constitute a set of requirements which oo beyond the assurance of adecuate protection of public health and safety. Two existing regulatory programs (Marginal to Safety, and this report) focusing at regulatory excess, constitute this acknowledgement. The next step is appropriate adjustment of the regulatory burden on the regulated community. Progress on this next step must begin immediately. Very truly yours, D. W. Edwards Director. Industry Affairs DWE/dhm Attachments Enclosures r IM O4) 4 [

4 ~ ATTACHMENT #1 SIGNIFICANT RECOMMENDATIONS WE STRONGLY SUPPORT Do not revise or modify 10 CFR 50 Appendix B (p. 7) (Sections 2.3.13, 2.3.14. and 4.4). The implementation of Appendix B in a performance-based and graded manner can readily be accomplished by revising the guidance and implementing documents using insights from risk assessments. Revise 50.54(a)[3), 50.54(p)(2) and 50.54(g) to make the program charge process for 0A, security and emergency plans consistent (p. 9) (Section 2.3.9). These are also manifested in plant-specific recommendations (p. 15 Section A.3.2.1 and p. 16 Section B.3.2.2). A standardized treatment for changes to licensee programs which results j in allowing changes that do not decrease program effectiveness without NRC prior approval is a worthwhile improvement. However, this endorsement should not be construed as supporting the addition of other programs to this set that are not currently included in regulation. Revise the frequenc; ef fitness for duty performance information submitted from semia qual to annual (p. 10) and program audits from annual to triennial licenses and 18 months for contractors (p. 10) (Sections 2.3.4 and 2.3.6) Both the reporting frequency and audit frequency are nnnecessarily burdensome. Reconsider the need for numerous reporting requirements which are generic and plant-specific (p.11) (Section 2.3.16). The three recommendations listed in this regard should be pursued to end unnecessary and burdensome reporting. Replace the guidance in Generic Letter 88-16 with guidance directing that the reference of specifically-approved topical report revision be in the core operating limits report (p. 12) (Section 2.3.8). This would eliminate an unnecessary Technical Specification change submittal. Upgrade the Regulatory Agenda to provide clear scheduled and accurate status information for all rulemaking activities (p. 13) (Section 2.3.17). D6i38C l

A clear ~ and current rulemaking record (at least in abstract) would enhance the value of the Regulatory Agenda for all users. -i Review existing security requirements to determine if they can be expressed in a more performance-based manner (p. 14) (Section.2.3.18). 4 Security offers a good opportunity to employ performance-based requirements. Ensure there exists a disciplined control process to preclude introducing nonrequirements into licenses as a matter of practice (p. 14) (Section A.3.2.1, Vol III). This past practice has circumvented all safeguards such as the Backfit Rule,10 CFR 50.109, and should not be permitted to continue. j Permit line-item improvements in accordance with Improved Standard Technical Specifications for all plants (p. 16) (Section B.3.2 4., Vol. III). The Technical Specification Improvement Program provides flexibility to licensees on a line-item-basis without ccmpromising safety. Expand the use of performance-based requirements for both license conditions and Technical Specifications (p. 16) (Section 3.2.4, Vol. III). The use of performance-based requirements allows.the incorporation of 4 prescriptive provisions into licensee-controlled documents where administration can remain entirely the purview of the licensee. + i 6 'I P k. 06\\340 i 1 _..J

r S e ATTACHMENT #2 RESPONSE TO SIGNIFICANT RECOMMENDATIONS WARRANTING COMMENT Define

• Current Licensing Basis" and "'>esign Basis" (p. 6) (Details Section 2.3.10)

The proposed definition of " Commitment" may be useful in refining the definition of current licensing basis (CLB) which appears in Part 54. The definition of CLB contained in 10 CFR 54 (Section 54.3) conveys a scope that is overly broad. (See discussion of the definition of " commitment" on Pages 6-8 of this attachment). Recommendation - The proposed definition of " commitment

• should be adopted and the current definition of CLB should be modified to be consistent with it.

Clarification of the term

• design basis
• may not be necessary.

However. the results of the industry-wide design basis recovery initiative sponsored by NUMARC will both answer the question as to the need for revision as well es provided the appropriate wording changes. Recommendation - Any clarification of the term " design basis" should await completion of at least the majority of design basis recovery programs currently underway. Revision of 10CFR Part 21 to allow gradation of the rigor applied to dedication of parts based upon safety significance (p. 6) (Section 2.3.1). This is an important change because it will increase greatly licensees

• ability to procure and utilize commercially available parts. The dedication process which qualifies them for use in safety-related applications can be tailored to the relevant safety function and the degree of safety significance for each part instead of a blanket uniform review. The result is a much more cost-effective program.

NUMARC has proposed a specific set of wording changes which accomplish the review group's recommended changes in a very effective manner (Petition for rulemaking filed June 21. 1993). Recommendation - Adopt the changes proposed by NUMARC_into 10 CFR 21. Develop PRA methods for use in 10 CFR 50.59 evaluations (p. 7) (Section 2.3.19). This recommendation. though made more clear in Section 2.3.19, could be interpreted as an ADD-0N to 50.59 evaluations. We believe that 50.59 evaluations are currently being performed adequately and need no more - Of\\337 i

i required analysis. The development of ouidelines (not

• methods" as i

stated on page 7) might be helpful for the use of PRA as an ootional technique for portions of the evaluations. We have found that often the { best use of PRA insight is as an overview mechanism to check for reasonableness of the conclusion after a 50.59 evaluation has been completed. This has not been needed, however, for all evaluations. Recommendation - Clarify the above recommendation in the report that PRA should not become an add-on requirement. Redirect the Marginal to Safety Program to be only a response to petitions program (p. 8) (Section 2.3.17). The report recommends that the Marginal to Safety Program focus on and be responsive to " specific and detailed petitions for rulemaking that are performance-based, propose to eliminate regulatory burden, and are l safety neutral." The report further recommends that 'the industry should take advantage of the petition for rulemaking process and submit complete. technically sound petitions in accordance with the NRC Regulatory Analysis Guidelines that the staff could publish expeditiously as proposed rules and request public comments." We concur i that it is reasonable to look to the expected beneficiaries of improvements in the regulations to play a major role in the development of proposals and supporting justification for said proposals. i However, regardless of the quality and degree of detail provided in a rulemaking petition, commitment of NRC resources is still required to evaluate and process such petitions. At the recent Marginal to Safety Workshop and at the Regulatory Information Conference we were unsuccessful in obtaining any clear indication of the resources that the NRC can make available to support this activity. If licensees are to expend their resources in developing rulemaking proposals, there must be a reasonable expectation that these proposale ill be processed in a timely manner. Our concerns have been heightened by a recent newsletter account of a meeting between the NRC Staff and NUMARC representatives to discuss the Marginal to Safety Program. The newsletter quoted the Staff as indicating that they only have resources to work on proposed changes to 10 CFR 50, Appendix J and one other issue. With respect to the question of developing rulemaking proposals which I have sufficient detail and bases to expedite the rulemaking process, the NRC Staff has been working on revisions of NUREG/CR-0058, " Regulatory Analysis Guidelines for the U.S. Nuclear Regulatory Commission" and i NUREG/CR-3568, *A Handbook for Value-Impact Assessment" for an extended period of time. j Recommendation - Timely completion and publication of pending revisions to NUREG/CRs-0058 and --3568 would significantly assist those who intend to develop rulemaking petitions. In i C76\\337 j

addition, NRC resource commitment and accountability for results (i.e., relief actions completed in a timely fashion) must be a prominent part of any program modifications. Utilize an " integral approach" to licensing actions (p. 8) (Section 2.3.17) On Page 132, the report suggests that amendments could contain a number of modifications that "... overall have no effect on the current level of safety at the plant." The idea appears to be a very good one. It would allow the tradeoff of various requirements to maintain a level of safety represented by the plant as currently licensed, but would permit increased flexibility in operational decisions. An example might be that the allowed outage times for a particular piece of safety equipment could be extended to permit maintenance with insignificant impact on at-power calculated risk. This, in turn, would increase availability of that equipment during shutdown. The net effect would be to reduce shutdown risk and, thereby, reduce overall plant risk. However, this type of example presumes that the safety envelope of importance is the total risk presented by the plant and that inter-mode tradeoffs are possible. However, review of Volume Four " Risk Technology Application" indicates that the type of tradeoff envisioned by the Staff may only be intra-mode. (pages 4-36 and 4-37.) Since PRA methodology represents an effective way to quantitatively estimate the

• level of safety
• for a plant (and changes thereto), and, since the specific treatment of tradeoffs in the PRA section of the report appears limited to single mode (e.g., during operation), the concept of an
• integrated approach" does not appear to have nearly the usefulness expressed in the summary explanations in Vol. I.

Recommendation - Clarify the concept of " integrated risk approach" in a manner that is consistent throughout the report. Revise 10 CFR 2.802 to clearly distinguish between Petitions for Relief versus those for Health and Safety reasons (p. 8) (Section 2.3.17) The suggestion, which is detailed on P. 131 would *... draw a distinction between a petition for rulemaking proposed for public health and safety reasons and one that is made to eliminate burden." The recommendation goes on to describe that the accompanying guidance for the regulation should outline an acceptable threshold of information and the detail needed for each type of petition. Although not stated, the inference could be drawn that there would necessarily be higher information threshold to introduce a petition for change to the regulations which would provide relief to licensees in some way as compared to the threshold to petition for a change which might impose 3 C76t337

further regulatory burden. This implied dual standard, presumably tilted toward increased regulatory requirements, seems entirely inappropriate. The justification necessary to support the change in the regulation should be measured to a single standard for level of information and detail, independently of the possible effect of the change. The level of detail and the nature of information required for backfitting of a requirement is clearly established in regulation. It would seem that, with regard to level of detail and the kind of information necessary, a petition for removal of a requirement should conform to a similar standard. We have never experienced difficulty in developing a petition in conformance with 2.802 as it is currently written. This is not to say that a well-founded and adequately prepared petition for relief should ever be " hostage" to a process which, by its nature, delays appropriate actions. We would endorse a change to 2.802 which had, as its result, assurance of expedited relief from an unnecessarily burdensome requirement which has little or no safety benefit. 1 Recommendation - This suggestion should be modified as suggested in the paragraph above. Add a definition for " Commitment' to 10 CFR 50.54 (p.9) (Section 2.3.2) The term " commitment

• does appear in the regulations. Though not defined there, a " commitment" is mentioned, incidentally, in Appendix C to 10 CFR 2 (VI.D. Related Administrative Actions). The inference in this section is that a commitment is different from an " obligation" and may not be "a legally binding requirement". Though, perhaps useful for some purposes, this type of connotation is not as useful as it could be for the conceptual development of a performanced-based regulatory system.

For the purposes of conceptual development, a commitment could be viewed as the licensees

• recognition of the obligation to satisfy the L

regulations and should stop with that. In that construct, a licensee would commit to meet each applicable regulation and offer an explanation as to how compliance would be accomplished. The staff would review the proposal and agree or disagree in each case that the requirement had been satisfied. This interaction would, no doubt, be iterative, but eventually the licensee would have staff agreement that the regulations were satisfied - thus, all commitments would be fulfilled. The license would be obligated to maintain the means for achieving compliance into the future. If an alternative means presented itself, and the licensee could support the substitution with the assurance that the alternative was no less effective in meeting the commitment, then the means could be changed, presumably without prior staff review and approval. At some point the licensee would report the change, probably in conjunction with the FSAR update. C76037

o Inspection and enforcement would remain the independent mechanism for the agency to assure that commitments were maintained. However, the inspection process would not (contrary to the present situation) provide a pathway for introduction of alternatives or additional means for satisfying commitments. A legitimate area for review, however, would be changes in implementation as discussed above. Unfortunately, this simple model is not entirely practicable because not all applicable, and indeed, necessary, safety requirements are contained within the regulations. (e.g., BWR Mark I hardened vents, or the entire array of TMI backfits). Again, unfortunately, the impetus for this quandary has come as much, or more, from the industry than it has from the regulator. So, commitments come in different " flavors", and there has not existed, until the review groups' attempt, a clear expression of what a commitment should be and what it should not. Consequently, for the foreseeable future, the concept of commitment must include requirements and " requests" as explained in the report. Additionally, the definition must include, as has been proposed, some element of staff acknowledgement that the commitment is an integral part of a safety decision. This provides the potential for "line drawing

• with respect to the nature of a safety decision.

At present, every intended action described by a licensee is potentially viewed as an unalterable commitment. The distinction between the commitment to meet a requirement and the means to implement that obligation need to be distinguished from one another. The first discriminator should be whether the associated decision which the staff must make based on the information provided by the licensee specifically relates to the satisfaction of a requirement (or the present set of " requests" that should have been requirements). The second discriminator should be whether the decision being weighed by the Staff relates specifically and solely to the acceptability of the means to satisfy the requirement. If the decision involves staff preferences or comparisons between means employed by different licensees, or a matter other than a specific requirement, the licensee proposal or intended action should not be considered a commitment over which the staff exerts approval authority. The two tier commitment proposed in the report appears to be a reasonable approach. Alternatively, a separate term could be coined for this category. The Review Group Report attempts this differentiation on Page 9 ...where [if] a commitment is not significant enough to be elevated to a regulatory status..." licensee controls should govern the change process. The Group's proposal in this area has merit and deserves careful evaluation, discussion, and debate to facilitate incorporation of a final definition into the regulations. Recommendation - Offer the draft definition for commitment as a proposed addition to 50.54 and obtain additional usan i

6 s ~ public comment on the issue. Integrate the resulting definition and concept into the " current licensing basis

• definition in Part 54 of 10CFR.

Consider revision of 10 CFR 50.82 and Regulatory Guide 1.B6 to aodress inconsistencies and define the possession only license (p. 10) (Details Section 2.3.3). These proposed revision activities are both long overdue and very important. The information in Section 2.3.2 deals with the type of license which NRC must issue for decommissioning. Since there is a potential for hearings in any of the three scenarios discussed (i.e., amendment, new license, and renewed license) the type may not be significant to a licensee. The amended, new, or renewed license would - be issued in concert with approval of the Decommissioning Plan and issuance of the Decommissioning Order. More importantly the type of hearing process that is used deserves careful consideration. Given that the risks to the public health and safety during decommissioning are substantially less than those associated with power generation, and given that it is in the best interest of NRC and the public for licensees to maintain a possession only license until the plant is decommissioned, any hearings on a Decommissioning Plan should be far less formal than those of the full adjudicatory process. The legislative type hearings conducted by EPA. who typically (in one outing) listens to comments from all parties and then renders a decision, appears much more preferable to a protracted adversarial adjudication. The Office of General Counsel, in SECY 92-382. has also suggested other types of " hearings" to be used for the approval of Decommissioning Plans. Recommendation - Revise and update the regulations and the Reg Guide to incorporate the following provisions for a Possession Only License (POL): 1. Specific acknowledgement that the following regulations do not apply to a POL holder, 50.44. Combustible gas control system 50.46. ECCS 50.48. Fire protection t 50.49. EO 50.55a(g). Inservice inspection 50.60. Fracture prevention 50.61. Fracture toughness 50.62. ATWS 50.63. Loss of alternating current power l 50.65. Maintensoce 50.71. Maintenance of records, making of 6-tn\\nr i

reports 50.120. Training and qualification of NPP personnel Part 50, Appendix G. Fracture toughness Part 50, Appendix H Materials surveillance program Part 50, Appendix J. Containment leak testing Part 50 Appendix K, ECCS Part 50,' Appendix R. Fire protection 73.21, Protection of safeguards information 171.15 Annual fees 2. Make specific revisions to the following regulations: 50.47 Emergency Plans (identification of the individual sections that continue to apply - Change conclusion in Vol. II, App. R, P.134) i 50.54 Conditions of License (identification of specific sections that continue to apply - Add to list in Vol. II, App. A, p. 149) 50.59 Changes, Tests and Experiments (acknowledgement that this_ regulation is applicable to the POL situation - i Identified in'Vol. II, App. A, p. 167) Part 140 Financial protection requirements and Indemnity Agreements (explicit reduction of primary liability coverage and exemption from sect-sy liability coverage) In addition, the NRC should adopt a legislatiot hearing format for the POL process in 10CFR2. Address design basis testing and test frequency based on risk techniques (p. 10) (Section 2.3.7). It is not entirely clear what is intended by the recommendation. Certainly, the efforts by the staff to facilitate uses by licensees of updated code editions and addenda has been very worth while. As the report points out. safety has benefitted by the work. Certainly, any: refinement of testing frequency based on risk techniques should have a net benefit. We believe that.this will also provide a benefit to -safety, and we encourage and support efforts in this direction. , tih337

Implementation of In-Service Testing (IST) at plants has been very costly and has resulted, in some cases, in unanticipated system transients due to the unusual configurations required. As revisions to codes continue and more systems are added to the scope of testing, plants will be required to develop more testing procedures for. systems not originally designed to support testing of this nature. Plants usually perform these tests with valving configurations (in lieu of modifying the system). In some cases these configurations can introduce more risks to equipment than benefits from the test results. Continued extension of testing in this regard could lead to an incident of 1 equipment damage at a facility. This issue seems to center on the suggestion that more testing should reflect accident conditions rather than normal conditions (usually such testing is conducted with the plant shutdown and system conditions are not bounding ones). Implementation of simulated accident condition testing at operating plants will be tenuous without, in many cases, major system modifications, for which the cost / benefit appears uncertain.'at best. The industry has accumulated much actual experience, at a great deal of expense, during the implementation of the MOV testing program. As the industry continues to try to test MOVs under design basis conditions, either through in-situ testing or by means of diagnostic testing, it is becoming apparent that in-situ design basis testing should not, in general, be performed on an operating plant. The original design verification test program conducted during construction and startup may offer the best opportunity for_ some boundary condition testing using plant systems. In the ideal, verification of component capability to perform under design basis conditions should have been performed by the equipment manufacture as part of the procurement process. Once these conditions have been demonstrated, during this initial test, testing of the component at the operating f acility should be restricted to normally expected operating ranges. An example of this ideal condition would be as follows: A new pump and motor purchased from the manufacture would be required by the procurement specification to be tested under all aspects of design basis conditions. This testing would verify that the pump and motor will perform to specification. The data obtained from the testing (i.e., pump performance curve) could then be used at the plant to verify that it is still performing within original design conditions but the data need not be taken under boundary conditions. There may be critical components where this process was not followed or not followed rigorously enough. In those cases, great effort should be l expended to obtain test data in test facilities. Expanding ISI testing to address this shortcoming is not a preferred approach. C76\\327

Recommendation - Approach the imposition of design basis testing carefully such that test conditions requiring bounding design are not imposed into power station In-Service Test (IST) programs. Add a definition for

• alteration" to 50.2 to clarify 10 CFR 50.23, 50.45, 50.56, and 50.92(a)(p.12) (Section 2.3.10).

The suggestion that the term

• alteration" be made synonymous with the

" design basis" is entirely inappropriate. The term " alteration" generally refers to

• material alteration" which has its origin in the Atomic Energy Act and means a conceptual change to an already licensed facility. There have only been two material alterations since the beginning of licensee history:

(1) Redesign of a zero power research reactor to a 50 MW design and (2) redesign of the West Valley fuel reprocessing facility to triple its output (the Construction Permit proceeding begun to facilitate this alteration culminated in the demise of the facility). Since a material alternation, by statute, requires the issuance of a new Construction Permit (and the attendant adjudicatory hearing process), there should be no connection whatsoever between the term

• alteration" and the term
• design basis".

Design basis is, and will remain, a detailed description of the engineering implementation of the licensing c requirements. It will continue to change over the life of a facility for numerous reasons such as vendor product line changes. None of these types of changes even remotely approach the realm of a material alteration in the context of the AEA. Recommendation - This proposal should be dropped. Policy statements should be deleted (11 cited) or revised into rulemaking proposals (6 cited) (p.13) (Section 2.3.11) This

• housecleaning
• should be pursued on a time available basis, with the exception of the statement on " Availability and Adequacy of Design Basis Information at Nuclear Power Plants." This policy statement contains all the trappings of a full-blown requirement complete with inspection and enforcement provisions.

We objected vigorously when this statement was proposed in draft form (5BFR15885) on the basis that the response it required imposed a significant burden on licensees and would r*sult in providing information generally available to the staff but in a different format. Additionally, we maintained that the request had not been justified to warrant the burden (Enclosure 4). We agree with the Review Group's assessment that the entire regulatory program such as was envisioned by this policy statement should have more substantitative regulatory underpinnings (i.e., receive the careful scrutiny and public debate associated with a rule). We disagree strongly, however, that this particular statement should become a rule. The industry initiative .g. G6\\337

~ 4 to undertake design basis reconstitution programs in conformance to the NUMARC guidelines for this activity obviate the need for any NRC requirement or guidance in this area. i Recommendation - Delete the Policy Statement on " Availability and Adequacy of Design Basis Information at Nuclear Power Plants" as duplicative of existing industry programs. The training and retraining in behavioral observation for aberrant [ behavior should not focus solely on substance abuse induced behavior (p. 13) (Section 2.3.5). In our comments on the staff's proposed reduction in the random testing rate we highlighted the value of supervisor behavior observation programs in terms of detection and deterrence. (Enclosure 5). This recommendation is consistent with our belief that these programs deserve acknowledgement as beneficial in assuring a highly reliable work force. (As noted in our comments, the proposed high rate of random testing (50% i for employees and 100% for contractors) does not accomplish this end.) The notion that aberrant behavior can only be substance related is (as noted in Section 2.3.5) indeed. unfortunate and needs to be corrected. However, we agree with the Review Group's assessment that no further recuirements are necessary. but merely a clarification in intent via generic correspondence. Recommendation - Issue the proposed Information Notice Provide a discussion of the Regulatory Research prioritization system in the pref ace of the Regulatory Agenda (p.13) (Section 2.3.17). The whole subject of prioritization methods deserves consideration. The methods employed by the agency to set priorities for regulatory focus should be readily understood by both employees and licensees. Describing the method in a readily accessible public document is a minimal step. The method-itself deserves evaluation from the standpoints of ease of use, ease of understanding, comprehensiveness of t weighing factors (health and safety as well as utilization of agency and licensee resources), comparative importance between issues and ease of relative ranking. The tool of risk analysis can facilitate some of these comparisons. Prioritization methods used by some licensees may provide useful input in this regard. The point is to be sure that the prioritization method is relevant and useful and then that it actually is used. Recommendation - Review the prioritization method for rulemakings for consistency with agency-wide methods. t 1 neun

i Drop any rulemaking for which staff resources are not anticipated to be available (p. 14) (Section'2.3.17). ? These intended rule changes should probably not be dropped unilaterally. The intended improvement to the regulatory agenda (p.13) should permit explicit discussion of the subject rulemakings. Public input should then be sought on the advisability of discontinuing these proceedings. Should the industry, or other members of the public, feel that c.7y of these warrant continued effort, the mechanism discussed on Page 8 of the Review Group's report, industry development of a rule change package, should be considered. Recommendation - Avoid any unilateral dropping of intended rulemaking activities. [ t -i L i i , C74\\337 -l l

ATTACHMENT #3 ISSUES RAISED WHICH WARRANT CRITICISM Section 2.3.4 of the report appears overly critical of the industry's efforts, or lack thereof, to "take full advantage of the available options for enhanced flexibility..." (P.20). In the opinion of our fire protection experts, the NRC Regional Offices often do not appear to have qualified fire protection engineers on staff who can understand and interpret the myriad of very complex NRC regulations and guidelines and the fire protection codes / standards applicable to a specific licensee. As a consequence the guidelines and the regulations are sometimes inconsistently interpreted and applied. This practice often reinforces the perception that, as a matter of common practices, licensees will be penalized via inspection for any program reductions. One additional recommendation that might stem from this comment is that qualified fire protection specialists be readily available to each of the NRC regional offices. Section 2.3.5 endorses the Staff's proposed rule change for a reduction of random testing from 100% to 50% only for utility employees. We have strongly disagreed with this reduction urging a change for all nuclear plant workers to 10% on the basis of the effectiveness of behavior monitoring programs. (Enclosure 5). Section 2.3.10 contends that the staff has interpreted the terms alteration or material alteration to mean a modification of the design basis. We are unaware of any instances when the synonymous definition I has been used and. if it has, we believe that it is erroneous. Material alteration is an extensive and fundamental modification to a facility such that it is 50 different that the licensing process must be repeated beginning with a new Construction Permit. Section 2.3.13 states " Manufacturers apparently do not find that an Appendix B OA program is in their economic interest for this relatively small market." (p. 91) because ".... a broad base group of vendors and suppliers with Appendix B programs does not exist." (p. 95). These statements appear to be a bit too absolute. For there are currently 300

• nuclear
• vendors with Appendix B OA programs being audited on behalf of NUPIC. Many of these vendors are original equipment manufacturers (OEMs) and other vendors who remain committed to the nuclear industry and 10 CFR 50, Appendix B.

In addition, Yankee utilizes another 40 to 50

• nuclear
• vendors whose programs include commitments to 10 CFR 50, Appendix B; we are probably typical of most utilities. We do agree, however, that large numbers of vendors have fled the " nuclear market
• for what they perceive as high hassle and high cost to them for what is a small segment of their overall customer base. C76\\340 I

~ Efforts, now, to restore reason and remove intrusive hassle may be too little too late but, undoubtedly, should be pursued. Section 2.3.17 attempts to define the term

• marginal" for use in a threshold to safety context (P.131).

It proposes a definition of marginal as a 10% increase in calculated risk. The basis for this proposal is the potential negative perception that could attach itself to a change which is larger in absolute terms (e.g., 50%) even though such a change may be insignificant in actuality. Perhaps a fully-developed plant PRA analysis which includes a complete uncertainty analysis would be able to accommodate evaluations at a 10% level in most instances. However, as the report explains, most PRA analyses cannot accommodate making distinctions at the 10% level. An attempt to draw an arbitrary *line" in this fashion probably will not yield a beneficial result. The report concludes, rightfully, that the marginal to safety effort should not focus on a specific risk number below which effort is deemed

• marginal to safety
• but should focus on qualitative and performance-based criteria. We agree that this seems like a much better approach although, in some cases, quantitative PRA results provide valuable input.

The more significant issue of this section is the utter inappropriateness of NRC staff to set a level of acceptability based upon their expectation about public perception. As a matter of policy, the NRC staff should be constrained to base decision criteria and all decisions on sound technical bases.only. Section 4.4.5 describes graded QA implementation as " defining different categories of implementation for the SSC's commensurate with their relative importance.' It explains a hypothetical diagram of the graded approach process for classifying relatively important and relatively non-important SSCs. It ends with the phrase *any application should not violate the defense-in-depth philosophy." Apparently defense in-depth is perceived as not being an element of the significance weighting which determines " relative importance." Such a perception is entirely invalid. Further, ano most significantly, this line appears to virtually invalidate the entire discussion, indeed, the entire concept by permitting an escape. Introduction of an " escape clause" such as this into the approach (technically invalid though it may be) undermines the entire thrust of Volume IV. If the intent of the report is to establish a graded approach to determination of adequacy of safety measures based upon relative risk significance, then the report should project that concept as clearly as possible. To do less, as in this case, severely wounds the initiative! 2-C76\\340 i

f i Recommendations for Imorovement of the NRC Inspection Program Introduction j The purpose of this Enclosure is to offer recommendations to improve the NRC's program for the inspection of power reactor licensees. We believe that licensees generally recognize that the inspection program is a necessary element of the regulatory environment; however, there are a number of potential reforms of this program which would reduce the burden imposed on licensees and provide for fairer and more effective utilization of the NRC's l resources. The choice of Vermont Yankee and Maine Yankee for parts of this evaluation was dictated solely by the availability of detailed inspection information on these two particular plants. Aside from the fact that both are single units and located in the same region, these are essentially random selections. Recommendation Number 1: The NRC Should Establish More Realistic Criteria and Controls for Major Team Inspection NUREG-1395,

• Industry Perceptions of the Impact of the U.S. Nuclear Regulatory Commission on Nuclear Power Plant Activities.* reported the results of a survey conducted by the NRC staff in the Fall of 1989 at 13 utilities. The stated purpose of the survey was to obtain feedback from utility personnel on the potential safety impact that NRC licensing and inspection activities were l

having on plant operations. One of the many problems identified in this reportwasthg)significanttimeandresourceburdenimposedbyNRCteam inspections At the 1991 NRC Regulatory Information Conference, it was indicated that as a result of the impact survey, one of the major areas for improvement would the " scheduling and control of inspections, especially team inspections.* Igg Subsequently, this policy change was formalized in the following addition to the guidance in the NRC's Inspection Manual for coordination of NRC visits to reactor sites: f "In light of the potentially significant impact to licensees caused by major inspections and non-inspection activities the regional office shall ensure that no more than four major activities are conducted during any SALP cycle without approval of the Deputy Regional Administrator following coordination with the Associate Director for Projects, NRR. Major activities are II)NUREG-1395 (March 1990), pp. A42-A43. (2)NRC Regulatory Information Conference, (May 7-8, 1991), Abstracts, p. 65. . 0604? i

Recommendations for Imorovement of the NRC Insocction Program (Continued) defined as visits to commercial reactor sites by a group of four or more personnel for five or more days for the purpose of inspection, conducting research information visits, conducting licensing audits, or other activities requiring significant licensee input or interactions. This restriction does not apply to operator licensing or requalification examination visits, emergency preparedness visits, or reactive team inspections. In addition, visits by senior NRC subjecttotheserestrictions."gagementofficialsshallnotbe Although the promised improvements in scheduling and control of inspections could contribute significantly to reducing the regulatory burden, there is evidence to suggest that implementation of this policy has not been particularly effective. There appear to be at least two elements contributing to this lack of effective implementation: first, the accuracy of NRC management's data collection and analysis is suspect, and second, some inspectors appear to be circumventing both the spirit and the letter of the stated policy. Clearly, if NRC management is to provide proper oversight, they must have a reasonably accurate understanding of the activities of their personnel. In the last three Regulatory Information Conferences, slides were presented which were intended to show the trend in the number of team inspections. (See Attachments 1. II. and Ill.) When one compares the information presented in these slides it is immediately obvious that the NRC's own data are not ~ consistent from one year to the next. Note the variation in the number of team inspections for the period April 1990 through September 1990 reported at the 1991 Regulatory Information Conference and the 1992 Conference and likewise, the variation in the number of team inspections for the period October 1991 through March 1992 reported at the 1992 Conference and the 1993. Conference. In addition to this lack of. consistency in the NRC data, the total number of reported team inspections appears to be implausibly low. In order to assess whether the NRC's numbers are reasonable, we have undertaken a detailed review of e recent inspection histories of two plants. Maine Yankee and Vermont Yankee ( The results of this review are provided in Attachments IV and V. Note that the indicated team size does not include participation by the resident inspectors at these sites. ' Attachment VI provides a comparison of the number of team inspections reported by the NRC staff for the total population of plants i.e.,115. to the number of team inspections for the two sample plants during the same reporting periods. In (3)NRC Inspection Manual. Chapter 0301, Subsection 06.04 (9/30/92). MISee Introduction. . two r

Recommendations for Imorovement of the NRC InSDection Program f (Continued) assessing the number of maior team inspections we referred to the criteria l stated in the NRC Inspection Manual. Although the Inspection Manual indicated that operator licensing or requalification examination visits and emergency f preparedness' visits are not to be counted in the maior team inspection category, they have been counted in the summary of all team inspections for Maine Yankee and Vermont Yankee. Also, reactive team inspections were not included in any of the team inspection summaries for Maine Yankee and Vermont Yankee presented in Attachment VI. If we look at the team inspection data for the total population of plants for the last three six month periods presented in Attachment II, i.e., October 1990 through March 1992, there were 180 team inspections of which 79 were i categorized as large (or major). During this same period, Maine Yankee and Vermont Yankee had 20 team inspections of which 5 were.large (or major) inspections. If NRC's data is correct, then this analysis indicates that 1.7 percent of the total plant population has borne the burden of 6.3 percent of all large team inspections and 11.1 percent of all team inspections of every category. Likewise, during their most recently completed SALP cycles both Vermont Yankee and Maine Yankee were the subject of four major team inspections, the upper limit specified by the NRC Inspection Manual guidance. If one assumes the average SALP cycle duration is approximately.18 months and uses NRC's inspection data for the last 18 months, on average one would expect less than one major team inspection per SALP cycle, i.e., 79/115 -0.69. We' can find no reasonable explanation for why Maine Yankee and Vermont Yankee both should be such apparent outliers with respect to the NRC team inspection burdens. In addition to the questionable NRC data concerning trends in team inspection burden, certain NRC staff practices have the effect of imposing a major team inspection burden without being acknowledged as-such. For example, during Maine Yankee's last SALP cycle a team of five inspectors were on-site for four days for a safeguards inspection. Because this does not meet the NRC Inspection Manual minimum criteria of four inspectors for five days, it does-not count as a major team inspection. However, in terms of licensee burden, there clearly is no substantive difference between dealing with five inspectors for four days rather than four inspectors for.five days. To cite another example, one' inspector appeared at the Rowe site for seven straight weeks for the stated purpose of closing out issues raised in the maintenance team inspection for this unit. In our judgement, the activities by this inspector certainly had the same impact on plant resources as a major team inspection. (It should also be noted that no deficiencies in Rowe's maintenance program were ever identified). In recent industry meetings, other utility representatives have expressed similar concerns about NRC team inspection practices. These practices have ~ 3-C76\\34

Recommendations for Imorovement of the NRC Inspection Program 4 (Continued) included characterizing multi-inspector activities as " audits" cr " follow-ups" in denying that the activity can be described as a major team inspection. Also, there is a sense that the use of three-person teams is more prevalent to i avoid the Inspection Manual's criteria for major inspection activity. To address the issues discussed above we offer the following specific changes: 1. The criteria for determining what constitutes a major team inspection should be redefined to more clearly identify the real impact of a multi-inspector activity. We suggest that a major activity be defined i as two or more inspectors on-site for 160 or more person-hrs. In Recommendation Number 6 of this paper, we recommend that each inspection report include a summary of the total NRC resources utilized in the inspection. This summary would be the logical place for the NRC to also indicate if it is believed that the subject inspection is, in fact, a major activity. 2. The NRC should establish criteria for a minimum elapsed time between major team inspections. Current NRC practices seem to concentrate major team inspections in the final months of the SALP review cycle. (See-Attachments IV and V.) Overlapping team inspections or team inspections occurring in rapid succession almost invariably impose a [ disproportionate burden on licensees. Better utilization of the NRC's Master Inspection Planning System as recommended later would be another means of dealing with this problem. Recommendation Number 2: The NRC Should Perform Regular Reviews of the Effectiveness of Special Issue Major Team Inspections Because of the very significant impact of major team inspection programs on the resources of both the NRC and licensees, we believe that the results of 'i these inspections should be reviewed on a regular basis to assure that there is a safety benefit commensurate with the resource expenditure. We are aware that the staff is currently planning a program effectiveness f review of the Service Water System Operational Performance Inspections (SWSOPIs) and applaud this undertaking. However, we are concerned that the timing of this projected review is such that its value will be minimized. It was indicated at the May 1993 Regulatory Information Conference, that the i SWSOPI review would occur when approximately one half of the planned inspections have been completed. By the time this review is completed and any recommendations implemented, only a minority of the licensee population will be able to benefit from the review. , tuon s f

Recommendations for Improvement of the NRC In5Dection Procram (Continued) In the future, we recommend that the staff undertake the effectiveness review s much earlier in the program. We believe that a sample of ten to fifteen inspections should provide sufficient data to determine if continuing the program would be productive and, if the program is to be continued, what ~ changes in scope are appropriate. This would also allow the NRC to quickly communicate significant findings to the remaining plant population. If after t this first review it is determined that continuation of the program is warra.nted, we recommend that another review take place after the next ten plants are inspected. With a data base of twenty to twenty-five inspections, it should be possible to identify all generic issues and evaluate the program effectiveness with a high degree of confidence. We offer the following specific changes to increase the effectiveness and the efficiency of this program review-1. The staff should be encouraged to expedite issuance of the team l inspection reports. Our experience is that issuance of the team inspection reports takes approximately two months. The licensee who has been inspected can quickly obtain an indication of any potential problems during the inspection exit meeting. However, other interested parties e.g.. INPO, NUMARC, other licensees, and other NRC staff, typically rely on the written report for communication of important inspection results. 2. Licensees and the public should be encouraged to comment on the results of the program effectiveness review. Providing a comment opportunity need not delay implementation of any program changes recommended as a result of the effectiveness review. 3. The staff should be encouraged to expedite the pilot program for auditing selected licensee assessments as a substitute for a direct NRC team inspection for special inspections such as the SWOPIs. We strongly endorse this initiative and suggest that the report of the results of the program effectiveness review could include specific guidance for scoping licensee self-assessments to assure that the goals of the direct team inspection are met. Recommendation Number-3: The NRC Should Improve Management Oversight of Inspection Activities to Provide Assurance That Individual Inspectors Do Not Impose Inappropriate Regulatory Actions or Positions The present NRC inspection process offers NRC Inspectors considerable latitude and opportunities for individual interpretations of NRC regulations. As noted in Recommendation Number S. even the NRC's Inspection Manual which is to " ensure quality, uniformity and effectiveness of inspections and to present a -S-C7 N 42

l Recommendations for Improvement of the NRC Insoettion Program (Continued) r well-defined base from which evaluations of licensee activities and overall performance can be made," is not subject to review and approval by the NRC's Committee for Review of Generic Requirements. As a result, there are numerous opportunities in the implementation of the Inspection Program for the imposition of inappropriate regulatory actions or positions. The Backfit Rule,10CFR50.109, provides a means for licensees to formally 1 appeal such actions. However, since the adoption of the Backfit Rule in 1985. NRC management has recognized that staff retaliation for a licensee's raising a backfitting concern could be a problem. See, e.g., NRC Backfit Workshop, Regions IV and V, April 29. 1986, at Tr. 136-139 (statement of Mr. Eisenhut). Concerns about perceived retaliation were also expressed during the Regulatory Impact Survey. We believe that formal backfit appeals by affected licensees should not be the sole means of redressing such situations: the NRC should provide a proactive oversight function to identify and minimize such inappropriate aspects of the inspection process. We believe that this could best be accomplished by the Office of the Inspector General preparing an annual report assessing the compliance of the Commission's Offices and employees with the provisions of 10CFR50.109. This report, which would be available to the public, would be expected to clearly document the scope of the audit activities and the bases for concluding whether or not full compliance with 10CFR50.109 is being achieved. Also, it would be expected that the report would recommend corrective actions and program enhancements to improve compliance, where appropriate. Recommendation Number 4: There Should Be a Strong Correlation Between Licensee Performance and the Inspection Burden Given that available inspection resources are finite, one would expect that those units that are performing better than average would be subject to less inspection than those units with perceived problems. However, a review of NRC SALP scores and the distribution of inspection hours suggests that no such correlation exists in present inspection practices. Furthermore, there are significant regional differences that are difficult to understand. In Attachment VII we provide a comparison of the NRC inspection effort by reactor unit for the period January 1992 to December 1992 and the units' SALP scores for the preceding assessment periods most closely matching this time frame. The inspection hours information is taken from the Proceedings of the May 1993 Regulatory Information Conference. Only data for operating units are included. , C76\\342

Recommendations for Improvement of the NRC InsDettion Program (Continued) We note that Diablo Canyon 1 and 2, units which are identified by the NRC as having a " sustained high level of safety performance,* were inspected more intensively in 1992 than Dresden 2 and 3. units included on the NRC's ' problem plant list.* Likewise, Grand Gulf and Summer, two more of the plants identified as superior performers, had greater inspection burdens than " problem plant

• units at Dresden and South Texas. These are some of the more obvious inequities.

We believe that the lack of correlation between performance and inspection burden on an overall basis is shown in Attachment VIII and Figure 1. From the information provided in Attachment VII, we have calculated the mean inspection burden for peer plants (i.e., the same SALP ratings) and listed the highest and lowest inspection burdens for each peer group. Although the mean inspection burden does show an upward trend as performance decreases, the range of inspection burden in each peer group (excepting the two St. Lucie units with a SALP 1 average) is extreme. We further note that the distribution of total inspection hours about the mean for each peer group is very broad and the indicated highs and lows for each peer group are definitely not isolated anomalies. We also believe that the extreme differences in inspection burden for similar performing plants are not an isolated, one year phenomena. We reviewed the inspe RegulatoryInformationConference{gonburdendatapresentedattheprevious and found that the extremes within the peer groups persist, i.e., the difference in burden for individual peer plants did not tend to average out. Another curious anomaly is the variation in the average number of inspection hours per plant when compared on a regional basis. From Attachment VII it can be seen that in 1992, the average inspection burden for a Region 11 plant was 3670.4 hours, whereas for a Region IV plant the average burden was 5066.2 hours. Even if Regions IV and V are excluded from this comparison because of the fewer number of units and the geographical scatter, the average burdens for Region I and Region III plants were respectively 22% and 18% higher than the average Region 11 plant burden. The Region II average plant inspection burden was also consistently lower than the other regional averages for the preceding year. In order to provide for a more intelligible and equitable distribution of the inspection burden we recommend that the NRC undertake a conscious effort to link licensee performance and inspection burden and implement this policy through better utilization of the NRC's Master Inspection Planning System (MIPS). We believe that the staff should be able to develop a

• target inspection budget
• for each refueling cycle of a plant based on licensee (5)NRC Regulatory Information Cor.ference. (July 21 and 22,1992), Abstracts, p.19 7

C MUit

L Recommendations for Improvement of the NRC InsDection Prcoram (Continued) performance. The

• target inspection budget
• concept would be one means of attempting to minimize the present significant differences in burden between performance peers and the above noted regional variations. MIPS, in turn, would serve as the detailed planning and scheduling tool for implementing the target budget. We acknowledge that scheduling as much' as a refueling cycle in advance could create a burden on the NRC due to the need to respond to unanticipated events and issues.

However, we believe that the benefits of planning on a longer term basis outweigh any perceived disadvantages and should be attempted. An important element of any management system is the close tracking of actual performance against planned budgets and schedules. We would suggest that MIPS be updated every three to four months and reviewed with the licensee along with an evaluation of how closely the actual inspection hours are tracking the target inspection budget. (We believe that Recommendation Number 6 would greatly assist in this analysis.) We fully recognize that the NRC ca inspect wherever and whenever it chooses. The systet we are proposing is simply intended to assist both the NRC and the i licensee in planning their respective activities and to gain a better understanding of how the inspection process is functioning. Recommendation Number 5: Proposed Rulemaking Packages and Changes in Regulatory Guidance Should Also Include the Related Additions or Changes to the NRC Inspection Manual The information provided by the staff to the NRC's Committee for Review of Generic Requirements (CRGR) to explain and justify proposed rules and new guidance (e.g., Generic Letters, Regulatory Guides, NUREGs etc.) should include the related proposed changes to the NRC Inspection Manual. NRC Inspection Manual modules often include significant direction, guidance, and interpretations of the NRC rules, Generic Letters, Regulatory Guides, etc. However, at the present time these modules are not subject to CRGR review thereby creating the possibility of inconsistencies and inappropriate expansion of the scope of the approved regulatory action. Inclusion of the proposed inspection module (s) in the information provided to CRGR in support of new rules and guidance should contribute to greater coherence and consistency in the implementation of the changes. Recommendation Number 6: Resource Expenditures for Inspections Should be Clearly Documented in Each Inspection Report We recommend that each inspection report include an accurate summary of the i total NRC resources utilized in the inspection. We envision this summary including both the on-site hours and any hours expended off-site, e.g.. .g. C(042 l m_.

) 9 Recommendations for improvement of the NRC Inspection Program (Continued) management review, in the preparation of the report. A distribution of hours by SALP assessment category would also be informative. An excerpt from a NRC inspection report which is provided in Attachment IX to this paper indicates that some inspection reports do attempt to quantify the expenditure of resources. However, this example is the exception rather than the rule, and such information is not provided on a consistent basis. The purpose of this recommendation is to assist both licensees and the NRC management in establishing a comprehensible correlation between resource expenditures and product. The number of inspection hours expended neither confirms or questions the quality or validity of an inspection. However, if one is able to readily identify anomalies such as significant differences in the hours required for a specific type of inspection, significant differences in the number of hours devoted to the different SALP assessment areas, wide regional variations, etc., potential inequities and misapplication of resources can be identified, evaluated and, where appropriate, corrected in a i more timely manner. I 'I l 4 i I 9 tuuo e i

FIGURE 1 1992 NRC INSPECTION BURDEN VS. 1 (Hours) 9 10500 I I l 9 10000 2 m l-N 8000 S P M E 5 E C 6000 s T l I m a g u O u = e a N 4000-l' l E u e m R E E B U i ; g a R 2000 D E 1000 N 1.0 1.3 1.5. 1.8 2.0 2.3 AVERAGE SALP RATING EJE!7"

TlEND N TlJVBER OF TEAV I\\SPECTIONS_- 1 APRIL 1990 OCTOBER 1990 THRU THRU-SEPTEMB'ER -99'O MARCH 1991 i = .LARGE TEAM 56 29 m ~ INSPECTIONS 2;" g at 20 j: 4 As, DALL TEAM .115 5 4.. Er i. o n D: ~ l'NSPECTIONS g 8 i E J 1 v., -.v.,...-.=.-,.w-,,- ,,,-.y .y... .~ mv..v.~,. *...., - -.wn, .....,...,e5 sw. .w e ..,.-.,,w.vd- ...,-.,-+n.#,...* s' -,Ec-v <. .-=,e----w-*-

• -,..,,--e--+-i,r

.4---, '._.--...rm._,m.* .=,4,-,.-

{ '" 4-t. v TREND IN FUMBER OF TEAM INSPEC c r Apr 1990 Oct 1990 Apr 1991 Oct 1991 thru thru thru thru Sep 1990 Mar 1991 Sep 1991 Mar 1992-53 29 25 25 h Large Team - Inspections is an All Team 109 54 60 66 L[ I0, 1y Inspections e r i,l a o Y ,ee,_s,-#,w. ..-..-.-,ru,o-.n.-... -e w s-. <-.s,..-v,.-,. ..w.,.,+-- ev.,, .-.-,,n.-.,-, ..-,.,...+,ew-ev.. ,.r,..ww. ,,-,-,-...w .,n..-- c,,,, ..,,.e...mw-.www,--,-c.-ww.--,n. ,v.

I. l 1 TREND IN NUMBER OF TEAM INSPECTIONS 1 l Apr 199 Oct 1990 Apr 1991 Oct 1991 Apr 1992 Oct 1992 ' thru thru thru thru thru thru Sep 1990 Mar 1991 Sep 1991 Mar 1992 Sep 1992 Mar 1993 - Major Team 53 29 25 32 21 18 ' Inspections (NRR Developed or led) All Team 109 54' 60 53 48 40 Inspections (As defined in MC 2901) _ i3 S2 l5 5 ii 5 4 g 5 !r 4 I.5 2 C 8

~ / f s-I l I Attachment IV [ t Maine Yankee - Recent Team Inspection History Date Team Size Days on Site Comments / Notes 1 11/5/90 2 5 Safety Issues (Bulletin 88-04) l 11/13/90 3 4 Safeguards (One team j member characterized ~as j observer) 1/28/91 2 4 Low Level Waste Transport 1/28/91 2 5 Eddy Current Testing i Start of Most Recent SALP-f 3/1/91 Period 3/4/91 4 5 Operator Requalification 4/9/91 2 2 Special-Human Factors Review l 4/29/91 AIT for Main Transformer l FIRS t 5/27/91 3 5 Post AIT to Observe Plant Restart 6/10/91 4 10 Partial EDSFI ] 9/16/91 2-5 Operator Requalification 10/28/91 5 4 Safeguards (Including US I Army Special Forces) l 4/6/92 5 10 EDSFI 4/6/92 2 5 Effluent: Environmental Monitoring 1 4/13/92 4 5 Engineering / Technical Support 6/22/92 7 4 Operational / Safeguards j Response Evaluation .(Includes US Army Special Forces Personnel) j -6/22/92 3 ~4 Emergency Preparedness 6/27/92 End of Most Recent SALP l Period ~. i

~. a l Attachment V Vermont Yankee - Recent Team Inspection History i.! f Date Team Site Days on Site Comments / Notes -{ 10/1/90 2 5 Radiological Controls 12/12/90 2 3 Fitness for Duty. 1/28/91 3 5 Security 'l 2/25/91 2 5 Rad. Protection 3/11/91 2 5 RG 1.97 1 Start.of Most Recent SALP j 3/17/92 Period 4/23/91 6 AIT for Loss of Normal Power Event 5/20/91 4 5 Generic' Letter 89-10 ? 9/3/91 2 4 Radwaste Program l l 9/23/91 2 5 Radwaste Systems (Mobile j NRC Chem Lab) 10/8/91 6 4 Operational' Safeguards Response Evaluation-f (Included U.S. Army l Special. Forces Personnel). j 10/21/91 6 5 Training Audit j 2/24/92 5 5 Emergency Preparedness 6/1/92 3 5 Engineering / Technical i Support !:I 6/29/92 3 4 Emergency Preparedness .7/6/92 2 3 Pre-EDSFI Information Gathering 7/6/92 2 5 ' Effluent / Environmental. Monitoring t 7/22/92 8 10' EDSFI End of Most Recent SALP ~f 8/15/92 Period -ii I i l' 'f 1 i f i

e i Attachment VI c 4 Comparison of Team Inspection Data NRC DATA FOR TOTAL MAINE VERMONT PLANT POPULATION YANKEE YANKEE October 1990 - March 1991 i Large Team Inspection 29 0 0 All Team Inspection 54 5 5 April 1991 - September 1991 Large Team Inspection 25 1 1 All Team Inspection 60 3 3 October 1991 - March 1992 Large Team Inspection 25 1 2 All Team Inspection 66 1 3 April 1992 - August 1992 Large Team Inspection 3 1 All Team Inspection 5 4

• Data not provided 6

9 k r

Attachment VII (Pago 1 of 4) REGION I 1992 PLANT PERIOD SALP AVG INSPECTION EFFORT (HRS) Beaver Valley-1 1/91-6/92 1.43 3169.0 j Beaver Valley-2 1/91-6/92 1.43 2471.4 Calvert Cliffs-1 4/91-3/92 1.43 4153.8 Calvert Cliffs-2 4/91-3/92 1.43 3442.3 Connecticut Yankee 2/90-7/91 1.14 5844.6 Fitzpatrick 2/91-4/92 2.29 9237.0 Ginna 10/90-1/92 1.57 4089.6 Hope Creek 8/90-12/91 1.29 4850.3 Indian Point-2 5/91-9/92 1.43 4296.0 Indian Point-3 6/91-8/92 2.00 8949.8 Limerick-1 10/90-3/92 1.14 4224.3 Limerick-2 10/90-3/92 1.14 2635.0 Maine Yankee 3/91-6/92 1.71 4782.6 Millstone-1 12/90-2/92 2.14 4023.2 Millstone-2 12/90-2/92 2.14 4310.1 Millstone-3 12/90-2/92 2.14 3504.5 Nine Mile Point-1 4/91-5/92 1.71 3639.2 Nine Mile Point-2 4/91-5/92 1.71 3784.1 Oyster Creek 4/91-7/92 1.86 5540.7 Peach Bottom-2 6/90-8/91 1.71 5240.9 Peach Bottom-3 6/90-8/91 1.71 4627.3 Pilgrim 8/90-9/91 1.57 6660.3 Salem-1 8/90-12/91 1.71 2335.9 Salem-2 8/90-12/91 1.71 2705.9 Seabrook 11/90-2/92 1.71 4887.5 Susquehanna-1 12/90-4/92 1.14 2941.3 Susquehanna-2 12/90-4/92 1.14 2586.0 TMI-1 5/90-11/91 1.14 4211.8 Vermont Yankee 3/91-8/92 1.71 6458.6 Regional Average 1.60 4469.1 c:\\gla\\regionI

Attachment VII (Pago 2 of 4) REGION II 1992 l PLANT PERIOD SALP AVG INSPECTION EFFORT (HRS) j Browns Ferry-2 5/91-5/92 1.57 -5123.3 Brunswick-1 10/90-11/91 2.29 5732.4 i Brunswick-2 10/90-11/91 2.29 5363.4 Catawba-1 2/91-3/92 1.57 4400.0 i Catawba-2 2/91-3/92 1.57 3818.5 l Crystal River-3 4/91-8/92 1.57 7109.5 Farley-1 1/91-5/92 1.71 3954.5 Tarley-2 1/91-5/92 1.71 3842.3 [ Grand Gulf-1 2/91-8/93 1.14 5002.5 Harris 6/91-9/92 1.29 5660.7 Batch-1 3/91-5/92 1.29 2447.0 Batch-2 3/91-5/92 1.29 2706.3 l EcGuire-1 2/91-8/92 1.29 2215.5 i McGuire-2 2/91-8/92 1.29 2310.2 l i North Anna-1 9/90-11/91 1.29 3044.0 l North Anna-2 9/90-11/91 1.29 2689.6 f I Oconee-1 8/90-2/92 1.57 2307.6 i Oconee-2 8/90-2/92 1.57 3183.1 Oconee-3 8/90-2/92 1.57 1881.2 { Robinson-2 3/91-6/92 1.57 3977.7 Sequoyah-1 6/91-8/92 1.71 4119.2 j Sequoyah-2 6/91-8/92 1.71 3787.2 St. Lucie-1 11/90-5/92 1.00 2487.5 St. Lucie-2 11/90-5/92 1.00 2575.7 Surser 5/90-8/91 1.29 4920.7 Surrey-1 3/91-4/92 1.43 3165.0 Surrey-2 3/91-4/92 1.43 2721.3 l Turkey Point-3 8/90-9/91. 1.43 3349.0 i Turkey Point-4 8/90-9/91-1.43 3068.2 { } Vogtle-1 10/90-9/91 1.86 3378.3 l j Vogtle-2 10/90-9/91 1.86 3439.8 Fagional Average 1.51 3670.4 l c:\\gla\\regionI

Attachment VII (Page 3 of 4) REGION III 1992 PLANT PERIOD SALP AVG INSPECTION EFFORT (HRS 1 Duane Arnold 4/91-8/92 1.86 6582.5 Big Rock Point 5/90-9/91 1.86 6608.1 Braidwood-1 5/91-9/92 1.71 2584.1 i Braidwood-2 5/91-9/92 1.71 2180.3 Byron-1 4/90-8/91 1.14 2112.0 Byron-2 4/90-8/91 1.14 2449.7 Callaway 2/90-1/92 1.14 4988.1 Clinton 2/91-4/92 1.71 4493.5 Cook-1 9/90-12/91 1.71 3488.5 Cook-2 r/90-12/91 1.71 3382.5 Davis-Besse 7/90-11/91 1.71 4865.1 Dresden-2 8/91-7/92 2.29 4561.0 t Dresden-3 8/91-7/92 2.29 4063.8 Fermi-2 3/91-6/92 1.43 6065.5 Kewaunee 12/90-2/92 1.57 5947.3 LaSalle-1 10/90-12/91 1.57 3203.6 ( LaSalle-2 10/90-12/91 1.57 3348.0 Monticello 7/90-11/91 1.43 5071.8 Palisades 1/91-3/92 1.57 7036.3 Perry 8/90-10/91 1.71 6609.2 Point Beach-1 9/90-1/92 1.86 2713.0 Point Beach-2 9/90-1/92 1.86 2573.5 Prairie Island-1 5/89-4/91 1.29 3620.3 Prairie Island-2 5/89-4/91 1.29 4268.6 Quad Cities-1 3/91-5/92 1.71 4780.5 l Quad Cities-2 3/91-5/92 1.71 4211.7 Zion-1 11/90-10/91 2.14 5106.1 Zion-2 11/90-10/91 2.14 4603.0 Regional Average 1.67 4339.9 i I I c:\\gla\\regionI .i

Attachment VII (Paga 4 of 4) i REGION IV 1992 PLANT PERIOD EALP AVG INSPECTION EFFORT (HRS) t Arkansas Nuclear-1 12/90-2/92 1.57 2879.1 Arkansas Nuclear-2 12/90-2/92 1.57 3683.7 Cormanche Peak-1 2/91-1/92 1.29 5356.7 Cooper 7/90-1/92 1.71 5338.0 S Fort Calhoun 5/90-7/91 1.43 6081.1 River Bend 4/91-9/92 1.86 6619.3 South Texas-1 6/91-8/92 1.86 4671.5 South Texas-2 6/91-8/92 1.86 4191.2 Waterford 5/91-8/92 1.29 4528.4 Wolf Creek 10/91-10/92 ,1.86 7314.7 [ Regional Avg. 1.63 5066.4 f L Region V 1992 PLANT PERIOD SALP AVG INSPECTION EFFORT (HRS) Diablo Canyon-1 1/90-6/91 1.43 4458.5 I Diablo Canyon-2 1/90-6/91 1.43 4319.4 Palo Verde-1 12/90-2/92 1.71 4600.4 Palo Verde-2 12/90-2/92 1.71 3213.3 i Palo Verde-3 12/90-2/92 1.71 3706.0 San Onofre-2 2/90-7/91 1.57 2847.2 San Onofre-3 2/90-7/91 1.57 2835.2 Trojan 4/91-5/92 1.71 NA WNP-2 9/90-12/91 2.14 10231.1 Regional Avg. 1.66 4526.4 j 1 c:\\gla\\regionI

4 Attachment VIII INSPECTION BURDEN DATA FOR 1992 AVERAGE MAIIMUM MINIMUM SALP NUMBER INSPECTION INSPECTION INSPECTION MAI/ MIN AVG OF UNITS ERS HRS HRS DIFFERENTIAL 1.00 2 2531.6 2575.7 2487.5 88.2 1.14 10 3699.5 5844.6 2112.0 3732.6 1.29 13 3739.9 5660.7 2215.5 3445.2 1.43 14 3988.0 6081.1 2471.3 3609.8 1.57 18 4129.5 7109.5 1881.2 5228.3 1.71 26 4139.2 6609.2 2180.3 4428.9 1.86 11 4875.7 7314.7 2573.5 4741.2 2.00 1 8949.8 2.14 6 5296.3 10231.1 3504.5 6726.6 2.29 5 5791.5 9237.0 4063.8 5173.2 t c41eiteurden jds

ATTACHFIET IX + .o 12 Other mstances or a faulted approach to instrumen:ation were manifest during this inspection i period: the operators assumed an alarm condition of the service water RMS was due to erroneous instnamentation without having a solid basis; and the reactor engineer assumed core differential pressure indications were due to instrument failure. The above instances indicate that some Maine Yankee personnel may be predisposed to question the accuracy ofinstrument indications when the indication is not normal. Although this approach does not appear prevalent, it indicates a partial lack of a sufficiently questioning attitude. This consideration warrants immediate management attention. j i 7. ADMINISTRATIVE 7.1 Person Contacted During this report period,' interviews and discussions were conducted with various licensee - personnel, including plant operators; maintenance technicians and the licensee's management staff. 7.2 Summary of Facility Activities The plant operated at full power from the beginning of the inspection period until December 17, when a tube failure in Steam Generator SG-1 forced a shutdown for repair. On January 7,1991, inspection and repairs in all three steam generators were completed, and operators began a plant heatup. Criticality was achieved on January 8; on January 9, at 10:30 p.m., the plant was shut down from 19to power to replace MS-70, a failed main steam non-return bypass valve, 7.3 Interface with the State of Maine Periodically, the resident inspectors and the onsite representative of the State of Maine discussed. findings and activities of their cor esponding organizations.~ Issues discussed included the status 1 of state rule-making regarding Maine Yankee, the technical issues related to the tube leak in the Steam Generator SG-1 and the associated Emergency Technical Specification Change. The State i Nuclear Safety Inspector also accompanied the Senior Resident Inspector on a tour of the expanded radiological control area. A summary of the findings based on that tour are included in Detail 2.2. 7.4 Exit Meeting Meetings were periodically held with senior facility management to discuss the inspection scope and findings.' A summary of findings for the report period was also discussed at the conclusion of the inspection. The inspection involved 252 inspection hours, including 20 backshift and 28 deep backshift hours. i .}}