ML17102A524
| ML17102A524 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 03/27/2017 |
| From: | Lochbaum D Union of Concerned Scientists |
| To: | Bill Dean Office of Nuclear Reactor Regulation |
| Watford M | |
| Shared Package | |
| ML16196A283 | List: |
| References | |
| 2.206, OEDO-16-00436 | |
| Download: ML17102A524 (5) | |
Text
March 27, 2017 William M. Dean, Director Office of Nuclear Reactor Regulation U.S. Nuclear Regulatory Commission Washington, DC 20555-0001
SUBJECT:
Comments on Proposed Directors Decision for 10 CFR 2.206 Petition re:
PG&Es Inaccurate and Incomplete Information Submittals
Dear Mr. Dean:
Your letter dated March 15, 2017, provided the proposed directors decision for the petition pursuant to 10 CFR 2.206 that I submitted July 14, 2016. You offered me the opportunity of commenting on the proposed directors decision, as long as I did so by March 29, 2017.
First, let me convey my sheer and total lack of appreciation for how the NRC is handling this matter. You offered me 14 days to review and comment on the proposed directors decision. This comment period seemed scant compared to the 30-day comment periods typically afforded licensees to respond to NRCs missives. However, this period is consistent with the guidance provided in Management Directive 8.11 (ML041770328), Review Process for 10 CFR 2.206 Petitions, specifically the final paragraph on page 26 of the handbook which states The letters will request a response within a set period of time, nominally 2 weeks. So, while 14 days seems like an unusually short turnaround time, it is consistent with Management Directive 8.11.
But it seems to be one of the few provisions of Management Directive 8.11 that the NRC bothered to follow. For example, the final paragraph on page 20 of the handbook states The first goal is to issue the proposed directors decision for comment within 120 days after issuing the acknowledgement letter. The acknowledgement letter was dated November 9, 2016, so the NRC issued the proposed directors decision 126 days laterI guess thats close enough for government work. Note that I submitted the petition on July 14, 2016. The NRC acknowledged it by letter dated November 9, 2016. So, the NRC took 116 days before even starting the 120-day clock in Management Directive 8.11. I wonder how the NRC would react to my taking 12 or 13 days or however long I pleased to acknowledge the proposed directors decision before starting my 14-day comment period clock. I wont find out because Im not stupid enough to trust the NRC not employing a double standard to dismiss my comments as being too late while using a home-field clock for its own glacial pace.
Returning to Management Directive 8.11 (because its amusing rather than instructive), page 26 of the handbook states that The letters [issuing the proposed directors decision for comments],
with the enclosure, will be made available to the public through the Agencywide Documents Access and Management System (ADAMS). Oh yeah? As of 9:00 am March 27, 2017, your March 15 letter to me is not publicly available in ADAMS. The enclosure to your March 15 letter states that the proposed directors decision has ML17031A266. Maybe it does, but no record by that accession number was publicly available in ADAMS as of 9:00 am March 27, 2017.
Management Directive 3.4 (ML080310417), Release of Information to the Public, states on page 14 of its handbook that Documents generated by the NRC are to be released to the public by the 6th working day after the date of the document. Thus, per Management Directive 3.11, your March 15 letter to me and its enclosure should have been made publicly available in ADAMS before 9:00 am March 27. But they were not. The NRC holds me and other members of the public to its internal procedure requirements, but scoffs at doing so itself. Such antics are wearing very, very, very, very thin.
I point these facts out not to be snide and snarkyokay, not just to be snide and snarkybut to provide the foundation for my comments about the proposed directors decision itself. The NRC has a very lax, casual, and informal approach to following its internal procedures. The NRC followed some parts of Management Directive 8.11 in this case while not following many other parts. The NRC did not follow Management 3.4 in this case. This spotty compliance is consistent with my experience dealing with the NRC during my 20 years with UCS. Its so random and chaotic that I dont know whether the occasional compliance is by intent or by accident, like a blind squirrel sometimes finding an acorn rather than yet another pebble.
Use, or abuse, of NRCs internal procedures is relevant to my comments about the proposed directors decision. Pages 2 through 4 of the proposed directors decision describes how LIC-109, Revision 21 (ML16144A521), Acceptance Review Procedures, governs how the NRC handles licensing submittals like the license amendment request that is the subject of our petition. Page 4 of the proposed directors decision describes how LIC-101, Revision 5 2 (ML16061A451), License Amendment Review Process, governs submittals like the PG&E ones we contested. A careful reading of LIC-109 and LIC-101 will not find a single reference to 10 CFR 50.9 (hereafter 50.9), will not find a single use of the word inaccurate, and will find all uses of the word incomplete limited to the NRCs authority under 10 CFR 2.101 to return a submittal deemed to be incomplete. Careless readings reach the same outcome, only faster.
Likewise, both LIC-109 and LIC-101 refer to requests for additional information (RAIs) often, but never in the context of possibly indicating a 50.9 violation.
1 UCS notes that LIC-109 Revision 2 became effective on January 16, 2017, months after our petition was submitted and acknowledged by the NRC. LIC-109 Revision 1 (ML091810088) was actually the controlling procedure at the time of the license amendment request and our petition. But since our point is that the NRC does not follow its internal procedures, it doesnt really matter which procedure they are not following.
2 UCS notes that LIC-101 Revision 5 became effective on January 16, 2017, months after our petition was submitted and acknowledged by the NRC. LIC-101 Revision 4 (ML113200053) was actually the controlling procedure at the time of the license amendment request and our petition. But since our point is that the NRC does not concern itself much with details, this detail likely doesnt concern them either.
March 27, 2017 Page 2
Page 6 of the proposed directors decision describes how LIC-111 (ML082900195), Regulatory Audits, can more efficiently enable the NRC to access licensees programs and processes. The inference I drew from this text is that these audits can gauge whether programs to ensure accurate and complete submittals are made to the NRC are adequate, or not. Oddly enough, LIC-111 mentions RAIs only in the context of obtaining additional information considered necessary for the audit report, not as potential indications of inadequate submittal assurance programs by licensees.
You may recall that I left UCS in February 2009 to go to work for the NRC. For about a year, I was a reactor technology instructor for the NRC at the Technical Training Center here in Chattanooga. That work required me to first become qualified as an instructor and then support other NRC employees obtain their initial qualifications and subsequent recertifications. I have some awareness therefore of the NRCs training and qualification programs. I am not aware of classroom, computer-based, or on-the-job training specifically intended to help NRC staffers determine when a licensee has violated 50.9, or even to identify evidence of potential violations.
For example, Inspection Manual Chapter 1245 Appendix A (ML15181A320), Basic-Level Training and Qualification Journal, is part of the training and qualification program required for NRC inspectors. 50.9 is explicitly referenced in an Individual Study Activity (ISA-5) dealing with allegations. It is one of ten references for this study topic, which is estimated to required 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> to complete. The topic contains 10 evaluation criteria, none of which touch upon, even in passing, the complete and accurate requirements in 50.9. Similarly, Inspection Manual Chapter 1248 Appendix F (ML15266A113), Training Requirements and Qualification Journal for Decommissioning Inspectors, also explicitly references 50.9, but only in context of an individual study activity about allegations.
Page 10 of the proposed directors decision states that While the NRC staff seldom issues violations for information submitted in the initial phase of the licensing process, it can and has taken enforcement actions (including escalated enforcement) for cases where licensees provide inadequate information that the NRC used or considered in reaching a decision. I downloaded the NRCs significant enforcement action dataset3 on March 27, 2017, to double-check this statement. Indeed, enforcement actions EA-12-021 against TVAs Watts Bar Unit 2; EA-12-075 againstironicallyPG&Es Diablo Canyon nuclear plant; EA-10-090, EA-10-248, and EA-11-106 against Entergys FitzPatrick nuclear plant; EA-11-252 against TVAs Browns Ferry nuclear plant; EA-10-205 against Dukes Robinson nuclear plant; EA-10-094 against Dukes Oconee nuclear plant; EA-09-193 against Xcels Prairie Island nuclear plant; EA-09-012 against Xcels Point Beach nuclear plant; and EA-09-010 against Xcels Monticello nuclear plant clearly demonstrate that the NRC has taken enforcement actions for 50.9 violations. They also suggest that the NRC seldom takes such actions, given that they account for fewer than 5% of the 231 significant enforcement actions in the dataset.
3 Online at https://www.nrc.gov/about-nrc/regulatory/enforcement/significant-enforcement-action.xls March 27, 2017 Page 3
I searched ADAMS for the number of publicly available records issued by the NRC and having request for additional information in the title since 2008. The annual tallies:
Number of Records in ADAMS Issued by NRC with request for additional information in the Title Year Number of Records 2017 (through 03-26) 63 2016 443 2015 527 2014 752 2013 802 2012 846 2011 835 2010 1,000-plus4 2009 1,000-plus4 Thus, there were over 6,268 requests for additional information (RAIs) issued by the NRC between 2009 and 2017 to date; the same period in which the NRC issued a whopping total of 11 significant enforcement actions for 50.9 violationsless than two-tenths of one percent of the number of RAIs. This is not to suggest or imply that every RAI constituted a violation of 50.9.
Or that even half of them did. But it boggles the mind to suspect, yet alone believe, that less than 0.18 percent of the RAIs involved 50.9 violations.
On the contrary, because NRCs procedures do not formally have staffers ask whether 50.9 violations may have caused the need for RAIs and because NRCs training and qualifications programs do not help staffers properly answers such questions should they inadvertently be asked, its miraculous that any 50.9 violations are ever issued by the NRC. (On a somewhat related note, a Google search to see whether the odds of a blind squirrel finding an acorn are better or worse than 0.18 percent was inconclusive.)
The rarity of NRC issuing a 50.9 violation coupled with the non-existent training for and procedural governance of potential 50.9 violations suggests that process alone does not account for the occasional violations. Instead, it seems more likely that violations are issued for other motivations, such as some NRC staffer being irked by a licensee and using the enforcement tool to retaliate. Because theres neither criteria within NRCs procedures nor training within NRCs qualifications programs differentiating between RAIs resulting from benign or malignant licensee behavior, it is impossible to understand why the NRC sanctioned Plant Z for a 50.9 violation yet didnt sanction Plants A-Y for literally thousands of RAIs. But its easy to get that somebody at Plant Z may have irked some NRC staffer and invited a benign RAI to magically morph into a malignant 50.9 violation. I cant say whether 50.9 violations are arbitrary and capricious, but they cannot be convicted of being objective even if so accused.
4 ADAMS limits the hits returned to 1,000 per search. There were more than 1,000 records in 2009 and 2010.
March 27, 2017 Page 4
The lack of criteria and training also makes it impossible for the NRC to provide a cogent explanation in the proposed directors decision for not granting our requested actions. If the NRC had criteria or training, it could relate them to explain why the Diablo Canyon RAIs fell short of those thresholds. Instead, the proposed directors decision had to rely exclusively on non-relevant information and gobblygook that I quite simply did not understand.
I would recommend that the final directors decision eliminate the non-relevant information and gooblygook and replace it with the Rhett Butler doctrinethe NRC just doesnt give a damn whether licensees comply with 50.9. That I would understand. I might not agree, but Id understand.
Sincerely, David A. Lochbaum Director, Nuclear Safety Project Union of Concerned Scientists PO Box 15316 Chattanooga, TN 37415 423-468-9272, office dlochbaum@ucsusa.org March 27, 2017 Page 5