ML20236P736
| ML20236P736 | |
| Person / Time | |
|---|---|
| Site: | Millstone  |
| Issue date: | 06/01/1998 |
| From: | Mahoney S DEVONRUE, LTD. |
| To: | Collins T NRC (Affiliation Not Assigned) |
| References | |
| NUDOCS 9807170119 | |
| Download: ML20236P736 (4) | |
Text
,
'(
?..
t/,
DEYONRUE J June 1,1998 :
38 Nurth street Sad' 200 DLNRC 9806-01/SM Hingham. Masscuhusetts 02043 wke: 781-7401445 FAX: 781740-1465 Mr. Timothy E. Collins websis<: wwscw-rue.<w=
Office of Nuclear Reactor Regulations NRR/DSSA/SRXB, Mail Stop OWFN/8E23 United States Nuclear Regulatory Commission f
l:
Washington, District of Columbia 20555 ~
b l
Subject:
Indications Plants Licensed Using Siemens LOCA Methodology May Be Operating in Violation l
of 10 CFR 50.46
Dear Mr. Collins:
Thank you for calling regarding my May 14,1998 letter. I understand you are prepanng a response I'd like to offer some additional input for your consideration.
' Summary of Additional Considerations for Your Reolv to My May 14.1998 Letter Clearty, NRC needs complete and accurate information in order to do its job. I presume your staff is seeking that information from Siemens right now. I certainly don't have to tell you this is a very serious matter, it deserves careful considershon, and you may not have all information you would like to have.
At the same time, it is evident that the cunent Siemens models are not accurate, and that in at least one case, Siemens admits its nort conservatism lead to PCT exceeding Federal limits. ' Until Siemens can develop a model satisfactory to NRC, there is no basis to assume plants other than Millstone 2 are operating safely, it is also important the public be given ample opportunity to fully observe and partiopste in the technical dialog.
l-The to::hnical information may well find itself the subject of one or another license amendment applications.
i
. Now is not the time to hide technical matters behind shallow claims of proprietary information.
i Since I first contacted you, there has been at least one meeting between Semens and the Staff. I also understand a mooting with Siemens and the HB Robmson staff may cover some ground related to the LOCA model. This dialog is very relevant to your reply. For now, let me focus on last week's meeting with Siemens Comments on the Limited Information Made Public by Siemens I obtained a non-propnetary version of the Siemens handout used in the May 21,1998 meeting with NRC Staff.I I conclude from the information an even gfsater imperative exists for prompt action to ensure i
compliance with 10 CFR 50.46.
c Among other things, Siemens describes a major effort to correct a host of long-standing ermrs and nonconservatisms in its accident model. These proposed corrections will be presented to NRC in a topical j
l.
report, and Siemens anticipates Staff will issue a safety evaluation report (SER). I presume your staff knows d
i j
. hen to expect this topical. Unfortunately, Siemens does not provide a schedule in its handout.
w Once the topical is presented, Staff will need several months before acquiring sufficient familiarity to question Siemens on one or another matter. Since Siemens withheld many technical details from its handout, it is difficult for me to fully identify or rank-order the issues you are likely to encounter or their impact.
' Revised EXEM/PWR Evaluation Modet," S. E. Jensen. D. J. Denver, and J. S. Holm, Semens Power Corporation, May 21, 1996 u*~
9807170119 990601 PDR ADOCK 05000336 P
)
L__.
=y DLNRC 9806-01/SM, Juns 1,1998
, e As a physicist and licensed engineer experienced in simulation, I feel somewhat comfortable hypothesizing some issues that could arise. For example, it seems evident one important issue is whether heat transfer coefficients should be changed the way Siemens proposes. Siemens acknowledges the obvious on Page 21 of its presentation: removing the ".. non-conservatism from Dougall-Rohsenow correlation will significantly increase PCT. 2 How much safety margin should be applied to the address the Dougall-Rohsenow correlation issue? How will the CE-EPRI pump degradation model be incorporated and in what manner? Do the 16 data points provide information of sufficient quality and quantity to model creep in RODEX27 How should uncertainties be treated?
Some of these issues may stand while others fall. Still other issues will no doubt arise as technical details are revealed. The point is, Siemens must put its cards on the table. Otherwise, the public may be playing yet another game of " hide the peanut," if not "Three-card Monte."
Identifying and resolving questions such as these can take months, and may well lead to more conservatism than Siemens is currently prepared to assume. As such conservatism are added back into the methodology Siemens is proposing, that which Siemens presumes is safe may well be unsafe.
At the bottom line, it is by no means a " slam dunk" Siemens' current presumptions will stand the test of time.
Moreover, if presumptions give way, today's bland assurances that "all is well" could become tomorrow's
" famous last words."
Siemens promises the current fleet of operating reactors will not suffer PCT margin erosion, and confidently reassures Staff only Millstone 2 will be affected by the methodological changes proposed. Perhaps they're right, but that conclusion rests on analyses that have yet to be validated or approved by NRC. Until Staff approves the methodology and inspects individual licensee results, what do we do? Do we assume plants are safe until shown otherwise, or should prudent safety regulation place the burden squarely on those with the technical problem and responsibility for precipitating the current crisis - Siemens and its licensees?
Even if Siemens is right, must the public hold its breath for months going on a year or longer while an untold number of reactors continue operating as if they were in full compliance? This is whistling past the graveyard.
Let's face it, Siemens will not fully validate their model until August 31,1998. - that's on top of the 3+ months i
that's elapsed since the Part 21 was filed. Actual plant reanalyses have yet to be scheduled, so it may be more months. Finally, some number of license amendments must still be submitted. A year since the Part 217 Longer than that?
)
ECCS is Currently inoperable No matter how much smoke-screen is generated, or how things are dressed up, or how many reassurances are put forward, it is bayond dispute that emergency core cooling sy*ms (ECCS) for Siemens' c~es are currently
" inoperable," as defined by NRC Generic Letter 91-18. ArDuably, one or another LCOs should have beer, triggered by now. The Generic Letter says action must be taken to assure safety.
Federal law, as provided in 10 CFR 50.46, is more emphatic, and directs licensees:
".. propose immediate steps to demonstrate compliance or bring plant design or operation into compliance with $50.46 requirements." (emphasis added)
I am unaware of any such immediate steps to bring plants into compliance.
Until licensees propose such steps to bring desirin or operations into compliance, "The Director of Nuclear Reactor Regulations may impose restrictions on reactor operation if it is found that the evaluations of ECCS coolin consistent with paragraphs (a)(1)(i) and (ii) of this section."'g performance submitted a (emphasis added) 8 Jensen et al., Page 21 2
y DLNRC 9806-01/SM, Juns 1,1998 Specific Points Recuested in Your Reolv to My May 14.1998 Letter in preparing your reply, I would appreciate if you would specifically address my May 14,1998 request for the following information at your earliest convenience, and as each piece becomes available:
(1)
A listing of all nuclear facilities with a license amendment relying upon the Siemens' methodology.
(2)
A copy of the Siemens' Part 21 report and other reports and correspondence filed by Siemens and licensees conceming the PCT errors and issues associated with ECCS performance.
(3)
Justification provided by NRC, Siemens, or the affected licensees for continued power operations while ECCS is inoperable for LOCA and other design basis accidents.
(4)
An explanation how the errors identified by Siemens could have remained undiscovered in US reactors allthese years.
(5)
An assessment of the likelihood that comparable errors exist among other safety analyses, including those of other vendors.
(6)
What NRC intends to do to ensure the cunent Siemens PCT problem does not recur.
(7)
The actions under 10 CFR 50.46 your staff recommends, or your schedule for pmviding such recommendations, to the Director of Nuclear Reactor Regulations.
You might also consider posting this information in a special place on NRC's vuosite to facilitate pMic access.
Additional Reauest Con:ernina Proprietary Treatment of Experimental Data It is possible that the more important aspeds of the May 21 presentation have been improper 1y kept out of the public domain by Siemens' claims of proprietary information. Siemens certainly had little to say on the record at that meeting. Clearty, the technical sper,ifics available were that meeting's " peanut" and was well hid.
It is not at all obvious why the information Siemens blanked out is truly a trade secret of competitive value -
the test for proprietary protection.
I don't believe we are discussing experimental data developed by Siemens. On the contrary, I must infer Siemens'so-called trade secrets were primarily statistical correlations of public data, rather than truly proprietary data. Moreover, the proposed conelation still requires NRC approval of the safety margins and methods. In fact, the only competitive value in this dialog is the margin the Staff ultimately imposes. The compet"ive value of that information to a Westinghouse et al. may be negligible.
As things now stand, even if th. so-called protected information is made available to Siemens' competitors, there is no obvious benefit to them in terms of market share or income.
Trade secret protection is normally reserved for data or processes. Data may be so protected when it is developed at a company's expense, and is of commemial value to a vendor's competitor. Processes may be so protected when it meets a similar test. However, it is the process that is protected, not the product.
Siemens might be able to protect its correlation methods, but not the product of those methods - the heat transfer coefficients themselves.
In any case, Siemens must do more than simply slap a
- proprietary" label on a handout. The company carries the burden to demonstrate why the information withheld from the May 21 meeting meets the test for a trade secret. Otherwise, it is just as easy to envision a scenario whereby Siemens classified as proprietary a
" smoking gun" for its LOCA model problems - heat transfer coefficients. If that is the case, regardless of s
10 CFR s0 46(a)(3)(1) io CFR s0 46(a)(2) 3
(
l
- ~
e
' DLNRC 9806-01/SM, June 1,1998 e
,- i intent, the practical impact is to exclude the public from meaningful comment and participation in the on-going safety dialog. Sierpens would have successfully "hid the peanut."
i must assume the company has a legitimate reason for hiding information. The burden for a showing is also theirs. Absent such a showing from Siemens why the subject information warrants special protodion, I request NRC reclassify the correlation's statistical parameters as the pure science it seems to be so it may be available to the public.
I appreciate your consideration of this matter, and look forward to your reply Given Dr. Jackson's special interest in verbatim compliance with Title 10, and operability issues involving safety systems, I am also forwarding a copy to her office.
Si ly 1
g EPHEN MALONEY Cc:
Dr. Sh ey Jackson, Chairman, US Nuclear Regulatory Commission Via:
Facsimile US Mail l
l l
l l
4 L___-______--___-____-__.__-_-