ML15237A334

Attachment 1 - A.15-02-023 Prepared Testimony of John Geesman

ML15237A334
Person / Time
Site:	Diablo Canyon
Issue date:	07/14/2015
From:	Alliance for Nuclear Responsibility
To:	Japan Lessons-Learned Division
DiFrancesco N, NRR/JLD, 415-1115
Shared Package
ML15237A311	List: ML15237A313 ML15237A323 ML15237A334 ML15237A335
References
A.15-02-023 A4NR-1
Download: ML15237A334 (384)

Text

Case No: A.15 023 Exhibit No: A4NR-1 Witness: John Geesman Application of Pacific Gas and Electric ) Company for Compliance Review of Utility ) Owned Generation Operations, Electric Energy ) Resource Recovery Account Entries, Contract ) Application 15 023 Administration, Economic Dispatch of Electric ) (Filed February 27, 2015)

Resources, Utility Retained Generation Fuel ) Procurement, and Other Activities for the Period ) January 1 through December 31, 2014. ) (U 39 E) ) __________________________________________)

PREPARED TESTIMONY OF JOHN GEESMAN ON BEHALF OF THE ALLIANCE FOR NUCLEAR RESPONSIBILITY BEFORE THE PUBLIC UTILITIES COMMISSION OF THE STATE OF CALIFORNIA JULY 14, 2015

1 PREPARED TESTIMONY OF JOHN GEESMAN ON BEHALF OF THE ALLIANCE FOR NUCLEAR RESPONSIBILITY ("A4NR") Q 01: Please state your name and business address for the record.

A 01: My name is John Geesman, and my business address is: Dickson Geesman LLP, 1999 Harrison Street, Suite 2000, Oakland, CA 94612.

Q 02: Are your professional qualifications included in your testimony?

A 02: Yes, my professional qualifications are contained in Appendix A to my testimony.

Q 03: Was your testimony prepared by you or under your direction?

A 03: Yes, it was.

Q 04: Insofar as your testimony contains material that is factual in nature, do you believe it to be correct?

A 04: Yes, I do.

Q 05: Insofar as your testimony contains matters of opinion or judgment, does it represent your best judgment?

A 05: Yes, it does.

Q 06: Does this written submittal complete your prepared testimony and professional qualifications?

A 06: Yes, it does.

2 Q 07: What is the purpose of your testimony?

A 07: The purpose of my testimony is to elaborate upon A4NR's Protest, which I incorporate by reference, using PG&E responses to A4NR data requests as well as other documents in the public domain.

Q 08: What do PG&E's data responses indicate about senior management's approach to the publication of the AB 1632 Seismic Studies Report, which PG&E also calls the Central Coastal California Seismic Imaging Project ("CCCSIP") report?

A 08: The primary indication comes from Ed Halpin, Senior Vice President and Chief Nuclear Officer, who focuses most on how the CCCSIP report will be communicated to the public and often lacks a firm grasp of the report's key findings. A June 25, 2014 email (when publication was scheduled for July) from Jeff Summy, a Halpin lieutenant, to PG&E Geosciences Department Director Richard Klimczak captures this:

... he [Mr. Halpin] reviewed the Executive Summary that was presented in the recent webcast. I think John 1 really alerted him to it but it is important for all of us to understand. He specifically asked about the last set of ground motion curves in the presentation where it shows us using the 86 th percentile for a deterministic evaluation that links Hosgris

[sic], San Simeon, and Shoreline and shows exceedance both at low frequency and high frequency. He is asking why are we even including that. I said I thought it was because state of the art would require us to link these so the IPRP would expect this combination. I also said the NRC would not worry about this (since they actually told us they will not expect us to link faults until we do the Seismic PRA) but he said (and he is likely right) that our opponents will challenge us on this. I responded that

Nozar 2 could demonstrate this is not really a safety challenge for the plant especially for how the NRC is treating eastern and Midwest plants. However his challenge is why do we need to include it at all?

3 1 I believe this reference is to John Conway, Mr. Halpin's predecessor as Chief Nuclear Officer.

2 I believe this reference is to Nozar Jahangir, 3 A4NR-00348, included in Appendix B as Ex. 1.

3 Several weeks later, on July 3, 2014, Mr. Halpin sends government relations director Tom Jones a stream

-of-consciousness summary of his understanding of the AB 1632 Seismic Studies, lauding four ocean

-bottom seismometers installed in 2013 seemingly unaware that they had stopped functioning soon after deployment, and hailing Diablo Canyon's site response data despite past criticism from the NRC and IPRP for relying upon only two earthquakes. Apparently uninformed of imaging problems in the shallow subsurface, Mr. Halpin extolls PG&E's 3D tomographic imaging at the plant site as "unprecedented

", "customized", and "(l)ike a cat scan

".4 His conclusions were unhesitant: PG&E has met our commitment to fully understand the seismicity surrounding DCPP in accordance with AB 1632 following the events of Fukushima.

Our efforts as a part of our LTSP and this commitment have brought greater clarity, reduced uncertainty and confirmed our conclusions that DCPP is seismically bounded, safe and a reliable asset that can serve California for decades to come.

.. We believe we have answered all of the questions posed by the IPRP of the PUC but will work with them to answer additional questions throughout the summer once the report is issued.

5 Twelve days later, on July 15, 2014, Dr. Stephen Thompson, Principal Geologist for PG&E's external coordinator of its SSHAC seismic source characterization ("SSC") project, Lettis Consultants International, Inc., communicates a radically different assessment to the SSC SSHAC technical integration team lead, Dr. William Lettis: Things are going horribly.

• • • AB 1632 continues to be a major major distraction for me

... I am being asked to review all AB 1632 reports late this week and next ... From what I have read ... or have learned

4 A4NR-00999, included in Appendix B as Ex. 2.

5 Id.

4 from others ... these reports are in bad shape and will be a blight on PG&E if they go out as is.

I brought up the subject of requesting a delay with Kent; 6 it was immediately rejected. My impression is that you have to take the lead in that type of discussion, and perhaps get Norm 7 on board first.

We need a major shift in focus and priorities if we are going to complete this project.

8 Nine minutes later, Dr. Lettis emails Dr. Abrahamson: Hi Norm I need to have an urgent discussion with you early next week. Please see email from Steve below.

... Steve (in particular) and Hans have been tied up almost full time with the AB 1632 study, which is in very sad state (this is another issue that I would like to discuss with you, in terms of PG&E Geosciences credibility).

At this point in time, I think it would be wise for PG&E to talk with the NRC about a 3 to 6 month delay in submitting the 50.54f response. We can offer the NRC many reasons, but the fundamental reason is the completion of ongoing new studies (AB 1632, USGS CRADA, etc.) that we would like to incorporate into the model.

9 Q 09: Do PG&E's data responses indicate whether any such delay was sought from the NRC for the SSHAC submittal?

A 09: No. The SSHAC submittal took place as originally scheduled on March 11, 2015, the four

-year anniversary of the Fukushima catastrophe.10 The CCCSIP publication date slipped slightly, 6 I believe this reference is to Kent Ferre.

7 I believe this reference is to Dr. Norman Abrahamson.

8 A4NR-1000, included in Appendix B as Ex. 3.

9 Id. 10 The evening before submittal of the 50.54(f) response, Jearl Strickland, Mr. Summy's replacement, assured Mr. Halpin and his deputy, Barry Allen, that "Based on our earlier discussion, (t)he process used to manage the SSHAC process will be evaluated in detail and a lessons learned developed with specific actions to improve the performance of the Geosciences department." A4NR-02215, included in Appendix B as Ex. 4.

5 from late July to September 10, 2015, but PG&E's documents show that all of the chapters except one (addressing the concerns raised on behalf of A4NR by Dr. Douglas Hamilton) were available for final review by July 17, 201411 and that the conclusions in the Executive Summary had not changed after July 2, 2014.12 Q 10: What was PG&E's reaction to the finished CCCSIP draft before its simultaneous public release and transmittal to the IPRP?

A 10: PG&E's pre

-release reaction is best characterized by three emails. The first is the endorsement from Geosciences Department Director Klimczak of comments attributed to PG&E's media representative in an Associated Press story about the recommendation by Diablo Canyon's former NRC inspector, Dr. Michael Peck, that the plant be at least temporarily shut down due to violations of the seismic design requirements of its license. "I agree with your statements in the article," Mr. Klimczak wri te s, citing the following: "PG&E spokesman Blair Jones said the NRC has exhaustively analyzed earthquake threats for Diablo Canyon and demonstrated that it 'is seismically safe.' Jones said in an email that the core issue involving earthquake ground motions was resolved in the late 1970s with seismic retrofitting of the plant."13 The second is from Mr. Halpin:

>>>Rich-as I retread the executive summary section of the report and go to the l ast page that summarizes old vs new data/assumptions, it seems the majority of the data has worsened and not improved.

The optics look bad. If I was to color code the summary sheet and show all data in red that's worse in regard to assumptions it would not look

11 A4NR-00263, included in Appendix B as Ex. 5.

12 A4NR-00264, included in Appendix B as Ex. 6.

13 A4NR-00331, included in Appendix B as Ex. 7.

6 good. I wonder how do I express that those areas that have worsened are low contributors to overall impact? How are they weighted in a deterministic calc and where is that weighting explained?

>>> >>>Thanks >>> >>>Ed14 And the third is Dr. Abrahamson's response to Mr. Halpin, sent through Mr. Klimczak:

... the increase in the maximum magnitudes will probably be the focus for the public.

A key issue to communicate is that for a site a [sic] close distances to the earthquakes, the high frequency ground motions do not increase much with increasing magnitude once the magnitude reaches M6.5...So the level of the ground motion is not sensitive to these changes in the magnitudes. The maximum magnitude remains an [sic] topic of ongoing scientific study, but the seismic hazard as [sic] DCPP is not sensitive to this uncertainty. This is good...15 Q 11: In 2014, did the IPRP communicate to PG&E a desire to review the CCCSIP Report and/or any inputs to the Report prior to its completion? A 11: PG&E's data responses indicate that the IPRP expressed such intent early in the year. After attempts on March 12,16 April 2,17 and the morning of April 318 to establish an agreed meeting proved unsuccessful (other than to trigger an inquiry from Dr. Stuart Nishenko to PG&E legal counsel Jennifer Post as to how to respond, and her "

What is the agenda for the meeting?" rejoinder19), CPUC staff Eric Greene beco mes more emphatic. At 2:49 p.m. on the afternoon of April 3, he wri tes to Dr. Nishenko:

14 A4NR-00115, included in Appendix B as Ex. 8.

15 Id. 16 A4NR-00529, included in Appendix B as Ex. 9.

17 A4NR-00520, included in Appendix B as Ex. 10.

18 Id. 19 A4NR-00721, included in Appendix B as Ex. 11.

7 I'm hearing that PG&E plans on submitting a report of its results and findings to the NRC in June 2014. The IPRP would like to see a draft before then ... the sooner the better ...

so that any comments or suggestions we might have could be incorporated into your report.20 (ellipses in original)

Mr. Greene reiterates that concern to Dr. Nishenko the following day ("The IPRP would like the opportunity to review any data, analyses, drafts before PG&E submits to the NRC so that our comments and suggests might be incorporated."21), but by then - aware that the IPRP planned to meet with visiting NRC Commissioner William Magwood

-- PG&E is already strategizing how to explain its dealings with the IPRP to NRC Commissioner Magwood before h e ho lds any meeting with the IPRP.22 As Chief Nuclear Officer Halpin later explains, "he [Magwood] is going to want to understand the advanced studies as well as the role of the IPRP. He will want to be horsed up on their concerns and how we have, or intend to address their concerns."23 Q 12: Did PG&E share drafts of any of the inputs to the CCCSIP Report with the IPRP? A 12: No. PG&E's data responses indicate the choice to withhold drafts from the IPRP as a source of concern to Regulatory Relations staff Valerie Winn and Mark Krausse. On March 19, 2014 Ms. Winn emails Mr. Krausse:

See slide 3 from Feb 2013 presentation that says we're sending draft technical reports on Irish Hills to the IPRP in February 2014. I've rec'd nothing to send them... but clearly we indicated we were going to share drafts.24 Mr. Krausse responds:

20 A4NR-00517, included in Appendix B as Ex. 12.

21 Id. 22 A4NR-00719, included in Appendix B as Ex. 13.

23 A4NR-00772, included in Appendix B as Ex. 14.

24 A4NR-00798, included in Appendix B as Ex. 15.

8 Why don't you just remind the Dream Team of that representation and see if they have something for you to share. They'll likely say they will get what they get at SSHAC next week, but it's worth a try.25 And Ms. Winn makes the attempt:

>All, > >In Monday's briefing with John Conway, it was indicated that we were not planning to share the draft technical report with the IPRP before it became public. Just wanted to make folks aware that in a presentation to the IPRP last year, we indicated that we were sending draft technical reports on Irish Hills 2D/3D onshore to the IPRP in February 2014 (see page 3)

>

>Not sure where that leaves us with respect to sharing or not sharing drafts, but wanted to make you aware of earlier communications with the IPRP on the topic.

>

>Thanks! >

>Valerie26 And Dr. Nishenko responds the next day, March 20, with:

Valerie et al Basically, the report is not finished at this time and is now scheduled for release in June Stu27 But a March 19, 2014 "Seismic Path Forward" meeting inside PG&E had already apparently resolved the question of sharing drafts. A summary of the meeting written by Phillippe Soenen, Supervisor of Licensing Regulatory Services at DCPP, acknowledging that "Not 25 Id. 26 A4NR-00729, included in Appendix B as Ex. 16.

27 Id.

9 all the required stakeholders were invited to the meeting" and addressed to Mr. Halpin, Mr. Summy, Mr. Klimczak, and Ms. Post, notes, It was agreed upon that once the seismic report is finalized that it will be sent out enclosed to two different letters. One letter will be sent to the State with a copy to the

ASLB and the other letter will be sent to the NRC. The cover letters will be written to specifically address the commitments to and concerns of the receiving party.28 Of the three Action Items identified in Mr. Soenen's summary of the March 19 meeting, onl y the one assigned to Ms. Post ("Develop a draft agenda for the IPRP follow up meeting and include required stakeholders"

) is labeled "COMPLETE".29 Consequently, Dr. Nishenko rebuffs Mr. Greene's request for drafts again on April 15, 2014:

While we are not going to present our findings at this time, our presentation of the report outline at the May Workshop will provide opportunity for dialogue about the report, our schedule to issue the final results report to the IPRP in June, coordination for a public meeting later in July, and the IPRP review schedule.30

And Mr. Greene's plea on behalf of the IPRP 31 minutes later is to no avail:

If we do meet with you at PG&E, it would be a closed technical session. Would be preferable if you could provide some our [sic] your findings, however preliminary. We recognize that the information is still in draft form and very preliminary, but any insight into the direction you are taking should be beneficial to the IPRP and PG&E.31 28 A4NR-00516, included in Appendix B as Ex. 17.

29 Id. 30 A4NR-00108, included in Appendix B as Ex. 18.

31 Id.

10 Q 13: Hadn't the IPRP been planning to hold public meetings at least quarterly?

A 13: Yes, that's what CPUC Energy Division Director Ed Randolph had indicated at the February 25, 2013 IPRP meeting. Thirteen months later, on March 21, 2014 (two days after the "Seismic Path Forward" meeting) PG&E's internal compliance program generated an automatic alert to Geosciences Department Director Klimczak, with a follow

-up email from Conor Doyle in Regulatory Relations: "Rich, Remind me, when is the next meeting?"32 Mr. Klimczak's reply:

No meeting is currently scheduled. We are issuing a final report June 30 th. I expect meetings after we issue the report based on when the CPUC/IPRP requests them. If a meeting is scheduled for any reason, I will let you know.

Rich33 When Mr. Klimczak presents "a timeline for required reviewers"34 of the CCCSIP Report as one of his Action Item assignments from the "Seismic Path Forward" meeting, even Mr. Halpin stumbles over the illogic of post

-submittal IPRP review:

What about the post June 30 th submittal date in regard to IPRP feedback? In other words is the report final once submitted? What is the purpose of the 30 day review by the IPRP?35 Mr. Summy explains to Mr. Halpin: Ed, We discussed that to a certain extent in a call we had Friday to discuss a request from IPRP to have a meeting with us before they meet with Magwood while he is here. Our sense is they will not be happy not getting an advanced review before we issue but we really have no choice at this point. I suspect if they come back with substantive

32 A4NR-00404, included in Appendix B as Ex. 19.

33 Id. 34 A4NR-00377, included in Appendix B as Ex. 20.

35 Id.

11 comments that we agree with we could see a way to revise but not if they are just providing comments that express their general disagreement with our approach or things to that affect [sic].

Rich/Stu would you agree with that?

Thanks Jeff S36 Mr. Summy's email did not explain why, as of April 6, 2015, "we really have no choice at this point" regarding IPRP "review before we issue." PG&E's data responses do not include any replies from Mr. Klimczak or Dr. Nishenko. Mr. Greene's earlier

-mentioned April 15, 2014 request37 for drafts from PG&E prior to the IPRP meeting with NRC Commissioner Magwood may have proven futile, but it did stir concern in Mr. Klimczak. He calendars a teleconference for April 17 with Mr. Strickland, Mr. Summy, Dr. Nishenko, Dr. Abrahamson, and Ms. Post with the following notation: "Discuss CPUC/IPRP response (below) to our request to move the meeting. Notice reference to SSHAC...."38 In planning a June 30, 2014 submittal date, with a 30

-day IPRP review starting on July 1, PG&E's "2014 Rollout Plan" acknowledges a need for further internal discussions in light of IPRP requests for more extended review.39 Mr. Klimczak emails Mr. Summy on May 29 with the news: Since our meeting with the IPRP on 5/21 we learned that they want about ~ 6 months to respond to our report and to hold ~4 public meetings. We did not agree with this, but

36 Id. 37 A4NR-00108, included in Appendix B as Ex. 18.

38 A4NR-00142, included in Appendix B as Ex. 21.

39 A4NR-00773, included in Appendix B as Ex. 22. PG&E's proposed schedule also contains a notation concerning a possible mid

-July public meeting of the IPRP, "Prior to meeting, Geosciences to establish an agreement with the IPRP on how to address public comments during the public presentation."

12 asked them to revisit this topic after they get the report for review. We will need to discuss and agree upon our position on IPRP review time and public meetings.40 Mr. Summy's response

"As we have discussed in the past the IPRP will be what it will be and we will have to deal with it as it occurs. It would be nice if it was clearly defined and played by some fixed set of rules but it doesn't."41 Q 14: As the envisioned submittal date slipped past June 30, 2014, did PG&E make any explicit plans for the IPRP post-submittal review to be reflected in the models used for the SSHAC process? A 14: The PG&E data responses indicate an acute awareness of the time pressure on incorporating any IPRP review comments on the CCCSIP Report into the SSHAC process. On July 7, 2014, Geosciences Manager Kent Ferre responded to a schedule inquiry from media spokesperson Blair Jones:

I have a call into Norm but have not heard back. I spoke to Steve Thompson at LCI regarding the seismic source model, and he feels September 1 is when the model needs to be completed. He did leave the door open for subsequent minor tweaks. It is really Norm to establish the final date since he does much of the downstream work. If Norm calls back I will give you an update.42 Forty-two minutes later, Mr. Ferre elaborates:

Blair, I just spoke to Norm, and he said that for the site response and sensitivity studies, they need final models no later than October 15. Working backwards, the TI team would therefore need the IPRP input no later than October 1 to make final tweaks to the source model. This establishes a last date for the AB1632 final report transmittal to the IPRP at

40 A4NR-00364, included in Appendix B as Ex. 23.

41 Id. 42 A4NR-00279, included in Appendix B as Ex. 24.

13 mid to late August, depending on the amount of time allowed for their review and comment.43 Notwithstanding the CCCSIP Report concerns voiced by Dr. Thompson on July 15 ("these reports are in bad shape and will be a blight on PG&E if they go out as is"44) or Dr. Lettis' request that same day for extra time on the SSHAC report ("I think it would be wise for PG&E to talk with the NRC about a 3 to 6 month delay in submitting the 50.54f response "45), PG&E allowed the IPRP review opportunity to compress as the CCCSIP Report release slid through August and the SSHAC deadline remained intact. Ironically, the day before the Thompson/Lettis emails, Mr. Ferre informed Mr. Klimczak that as of June 30, 2014, PG&E's 2014 budgeted amount for the IPRP contained a $720.1k unspent balance, 52% of the 2014 total unspent amount for the AB 1632 Seismic Studies.46 Q 15: What happened to the IPRP review window as the CCCSIP Report's submittal date slipped to late August?

A 15: A series of PG&E emails on August 20 and 21, 2014 describe the squeeze:

First, from Regulatory Director Thomas Jones to Mr. Klimczak and Ms. Post: The Sr. leadership of Corporate Affairs has requested a simple timeline that shows the drop dead dates in sequence to understand the required vs. desired timing for release of the report to address concerns raised on related business matters based on our conversations and emails of yesterday and today. The goal is to understand the due dates for ground motion to the NRC in March, and how the intermediate steps are driven by the release on August 28. I need this tomorrow and will help support this in any way you need. I suggest a simple timeline and a reference below for each milestone and its weighted importance to satisfy the March 2015 NRC order for ground motion value.47 43 Id. 44 A4NR-1000, included in Appendix B as Ex. 3.

45 I d. 46 A4NR-00231, included in Appendix B as Ex. 25.

47 A4NR-00402, included in Appendix B as Ex. 26.

14 Then Mr. Klimczak's request for assistance from Geosciences Manager Ferre:

Kent I will need your help putting the requested timeline together today has [sic] I am traveling to Washington D.C. Put the timeline together based on your SSHAC schedule and send it to me P&C. Show the 30 day IPRP review in the timeline. I will call to talk when I am able.

Thanks, Rich48 Then Mr. Summy's input:

Rich/Kent, We need to be clear on the potential impact to IPRP's ability to provide input to SSHAC if we delay past August 28.

Thanks Jeff S49 And Mr. Ferre's work product:

Rich, Jeff:

Attached is a high level SSC SSHAC schedule with the IPRP review and SSC TI team evaluation and integration (of IPRP comments) included (tasks 8c3 and 8c4).50 Finally, Mr. Klimczak's summation:

As we discussed, to meet March 2015 NRC deadline we need:

1) SSC logic trees with documentation due 10/31/14 to PPRP.51 To meet this date it requires published AB1632 results to be used as input by the SSC TI team for the logic trees development. It also requires time to allow 30 days for IPRP review and for them to provide comments to the SSC TI team. The SSC TI team needs time to consider integration of IPRP input and still meet the 10/31/14 due date.52 48 Id. 49 Id. 50 Id. 51 I believe this reference is to the Participatory Peer Review Panel, an NRC

-required aspect of the SSHAC process.

52 A4NR-00402, included in Appendix B as Ex. 26.

15 Then, on August 28, the very threshold warned of by Mr. Summy one week earlier

("We need to be clear on the potential impact to IPRP's ability to provide input to SSHAC if we delay past August 28"53), another burst of emails:

First from CPUC staff Eric Greene , on behalf of the IPRP , to Dr. Nishenko : Is PG&E still on schedule to submit to us your seismic studies report for DCPP by the end of August? If not, do you have a realistic revised delivery date for the report? The IPRP is anxious to begin our review without further delay. Thanks

.54 Then from Dr. Nishenko to Mr. Klimczak, and Regulatory Relations staff Valerie Winn and Thomas Jones, with a copy to Ms. Post:

To follow up on my email earlier today - please advise as to response.55 And Ms. Winn's response:

Stu, let me give Tom a call to discuss. I would like Eric to poll the IPRP as to what they need in terms of hard copies, CDs, etc [sic] and being able to communicate a date (e.g., in [sic] or before September 15) would be helpful. It would also be nice, to manage IPRP expectations, to work with Eric to set a date for meeting with the IPRP to discuss. That might help to get them into our court or at least minimize any blow back from the delayed issuance.56 Ms. Winn, whose full title is "Chief, State Agency Relations," may have been particularly attuned to PG&E's need to "at least minimize any blow back from the delayed issuance" in light of her earlier communications with the IPRP:

On June 11, 2014 she writes to CPUC staff Eric Greene, with a copy to IPRP Chair Chris Wills: Hi Eric (and Chris),

53 Id. 54 A4NR-00707, included in Appendix B as Ex. 27.

55 Id. 56 A4NR-00584, included in Appendix B as Ex. 28.

16 I wanted to provide this month's report to the ASLB on various DCPP issues. Please note that for the item "Issue Final Seismic Report to the NRC and ASLB", we had previously indicated submittal in June 2014. That date has now slipped to July 2014.

While I don't have a precise date today for when the report will be submitted to the NRC (and the IPRP) in July, I expect the submittal will now be made in late July. I'll be in touch as I know more.

Please don't hesitate to call if you have any questions.57 Eight weeks later, on August 5, 2014 Mr. Wills inquires:

Hi Valerie

Since I haven't heard anything, I assume that the reports for NRC and ASLB are still not quite done. Can you give me a date when we might expect to receive them? I need to plan for staff time to review them. I also need to know the dates when PG&E expects to finalize their seismic hazard model inputs, so we can try to provide our review comments before then.

Chris58 Ms. Winn responds ten minutes later:

Hi Chris and Eric, Sorry for the delay in communicating to you. As I understand the current schedule submitted to the ASLB, it looks like a late August release date. I'll check with the team on other dates for finalizing the seismic hazard model inputs and let you know.

Thanks! Valerie59 Thirty days later, on September 5, 2014 (eight days past Mr. Summy's forewarned threshold), Mr. Wills inquires again:

Hi Valerie Any update on the schedule?

57 A4NR-00800, included in Appendix B as Ex. 29.

58 Id. 59 Id.

17 Chris60 This time Ms. Winn takes several hours to respond:

Hi Chris and Eric:

I just received word that the report will be submitted to the NRC on Wednesday, September 10, at which time the report will become public. While we're working to finalize the report and prepare hard copies ad CDs (no easy feat, it's a big report) and other logistics, we'd be happy to do a pre

-brief of the report's findings with the IPRP on Tuesday afternoon, if that is possible. We could do the briefing by phone if you'd like.

Please let me know your availability and I'll work to get this set up.

Eric, I know the IPRP is also interested in having a meeting or two once the report is submitted to the NRC. Now that we know the date for the submittal, could you please let me know when/where the IPRP may wish to meet to discuss the report in greater depth? I can then get this on my team's calendars.

Best regards, Valerie Winn61 Q 16: Do the PG&E data responses identify any specific accommodations of the IPRP's review role as publication of the CCCSIP Report slipped across the August 28, 2014 threshold?

A 16: An 8:26 a.m.

September 2, 2014 email from Dr. Thompson to PG&E Geosciences Manager Ferre states, "Bill Lettis and I are coming to 245 Market today to meet with Norm. Are you free to talk briefly about implications of the new IPRP scope?"62 At the end of that same day, at 5:29 p.m., Dr. Thompson sends an email to the SSC TI team (with copies to Mr. Ferre, Mr. Klimczak, and Dr. Nishenko) with a proposal

Subject:

DCPP SSHAC: Possible meeting with IPRP to discuss implementation of AB 163 2 in SSC model

60 Id. 61 Id. 62 A4NR-00256, included in Appendix B as Ex. 30.

18 SSC TI Team plus Kent, Rich, and Stu:

We understand that the IPRP - a review board working on behalf of the California PUC to review the AB1632 program - has new scope to explore how the results of the AB1632 program are being used in the SSHAC for the SSC model. As such, we are just beginning to figure out how we can accommodate this change. A proposed approach for interacting with the IPRP to help them achieve their goal is as follows:

1) AB1632 reports are delivered to the IPRP and NRC (this is coming soon)

2) The IPRP perform an initial review of the AB1632 reports (approx. 3 weeks)

3) The SSHAC TI Team and PG&E Geosciences hold a 1

-day meeting with the IPRP (+/- members of the PPRP?) to discuss:

a. The SSHAC process and the Role of the AB1632 program in the process

b. The SSHAC SSC Model (update from Workshop 3; almost final)

c. How the AB1632 reports have been reviewed, considered, and integrated by the TI Team for the SSC Model

d. Any initial comments or questions the IPRP have about how the AB1632 results have been or will be incorporated into the SSC model

4) The IPRP complete their review of the AB1632 reports and draft a letter recommending how AB1632 information should be incorporated into the SSC Model 5) The SSHAC TI Team and PPRP review the IPRP letter. The SSHAC TI Team considers the letter, makes modifications to the SSC as deemed appropriate, and addresses the IPRP letter in the SSHAC SSC Final Report.

The main action for the TI Team now is to understand this may be coming up, and to know whether you are available for a one

-day meeting during the OCTOBER 1-3 WINDOW.

For Rich, Kent, and Stu, please let me know your general thoughts on the plan outlined above, and your availability the Oct. 1

-3 window. Bill Lettis and N orm Abrahamson have indicated their availability those days.

Thanks, -steve63 The PG&E data responses do not indicate what "general thoughts" Mr. Klimczak, Mr. Ferre, Dr. Nishenko, or anyone else at PG&E may have conveyed in response to Dr. Thompson's

63 A4NR-00255, included in Appendix B as Ex. 31.

19 September 2 proposal, but the conference call Dr. Thompson scheduled the morning of September 4, 2014 between the TI Team and Mr. Ferre includes an agenda item "

-IPRP"64 and Ms. Post calendared a conference call with Mr. Ferre, Mr. Klimczak, and PG&E legal counsel William Manheim to discuss "Implementing IPRP Provisions in GRC Decision" for 3 p.m. on September 9, 2014.65 At 3: 38 p.m. on September 9, 2014, Dr. Thompson emails Mr. Ferre, Mr. Klimczak, Dr. Nishenko and the TI Team:

Hi Kent, Rich, and Stu, I just spoke with Kent about meeting as a group at PG&E to discuss the process of the IPRP review of the AB1632 studies and their scope to review how the reports are being implemented in the SSHAC model. The agenda would include:

1. Update on understanding between PG&E and IPRP concerning scope of review

2. Discuss possible scope and schedule of review

3. Discuss current SSHAC TI Team observations, interpretations, and strategies for incorporating AB1632 report content

4. Outline next steps.66 On September 11, 2014 Dr. Lettis sends the following email from his mobile phone to Mr. Ferre and Dr. Thompson:

Hi Kent and Steve I just attended the CGS67 mapping committee meeting in Sacramento chaired by Chris Wills. In a side bar hallway conversation, Chris indicated that he, Gordon Seitz, and Tim Dawson would like to meet informally 'any time next week' to discuss how we are using AB 1632 in our SSHAC deliberations, if possible. Chris has some personal issues that will prevent him from meeting later until later in October.

Chris indicated that their committee received the ab 1632 reports yesterday, and it would be good to have an idea of how we are using the data so they can consider this in their review.

64 A4NR-00170, included in Appendix B as Ex. 32.

65 A4NR-00252, included in Appendix B as Ex. 33.

66 A4NR-00249, included in Appendix B as Ex. 34.

67 I believe this reference is to the California Geological Survey.

20 The meeting would be informal non public.

Is this possible? Anyway, this will be good to discuss when we meet on Monday (or Friday? Has a meeting time been set up?).

Thanks Bill68 The following Monday, September 15, 2014, Mr. Wills engages CPUC staff Eric Greene and PG&E in an effort to obtain a missing reference document relied upon by the CCCSIP Report: Hi Eric In skimming through the report, I've found that the studies for slip

-rate on the Hosgri fault refer to a previous report by PG&E that has not been provided. Since that report describes the way PG&E has determined the ages of sediments, it's critical to the evaluation of slip

-rate. If it's OK, I'll send an email to PG&E asking for that report.

Chris69 Mr. Greene responds:

Sure. Please include your e

-mail below so PG&E will know I'm aware of the request and would also want to see PG&E's response. It is possible that it is something internal that PG&E does not want released. Thanks.

Eric70 And Mr. Wills transmits his request to Ms. Winn and Dr. Nishenko:

Hi Valerie and Stu

In beginning our review of the reports last week, I've noted that discussion of the ages of sediments and channels offset by the Hosgri and Shoreline faults refer to a 2013 report

68 A4NR-00246, included in Appendix B as Ex. 35.

69 A4NR-00786, included in Appendix B as Ex. 36.

70 Id.

21 by PG&E. I do not find that report with the reports released last week and do not think we've been provided a copy previously. Since the age of the offset features is a critical aspect of any determination of slip rate, we need that report to review the slip

-rate estimates. Please send a copy of the 'Stratigraphic Framework' report referenced below to me as soon as possible.

Thank you Chris71 At the end of the day on September 15, 2014, Dr. Thompson sends his follow

-up to the prior week's hallway conversation between Mr. Wills and Dr. Lettis:

Dear Chris,

Following up on our conversation earlier, we invite you, Tim, and Gordon over to LCI

Friday to discuss the current SSHAC SSC model and our progress to date incorporating the results from the AB1632 studies. As mentioned, Friday is the earliest day we can meet; we understand Gordon cannot attend, but perhaps we can mitigate this another way. The goal is to keep this to a high

-level discussion of the SSC model, parts of the SSC model relevant to the AB1632 program, and our approach and findings for incorporating the data and interpretations into our logic trees. It is unlikely we will have time to delve into the seismic data volumes, but perhaps we will be able to show you and Tim an example or two.72 On September 17, 2014 Dr. Lettis invites PPRP Chair Kevin Coppersmith to join the meeting with CGS personnel:

I know that it is short notice, but we just found out this schedule and talked with PG&E. No other IPRP member will attend, and this meeting is informal (I.e., not public). If you are available, it would be great to have you attend - especially to provide support for our approach in utilizing the data and proponent interpretations from AB 1632, as well

71 Id. 72 A4NR-00576, included in Appendix B as Ex. 37.

22 as the 3role2 [sic] that the IPRP has within the SSHAC process (I.e., no formal role but we will consider their comments in our assessment process).73 A slide presentation dated September 19, 2014, the day Dr. Thompson had scheduled the CGS visit to LCI, purports to reflect prior communication with Mr. Wills. The initial slide is entitled: SSHAC Implementation of AB1632 Results SSHAC SSC TI Team Presenting to CGS/IPRP reviewers 9/19/2014 SSC working meeting 3574 The 61 st slide contains the following:

Wrap-up Notes SSHAC Deadline - end of October is drop

-dead Approx. Oct. 10 deadline to get input from C. Wills Next IPRP/C. Wills Plan Focus on Hosgri slip rate - Gordon/Dawson lead Know Estero Bay study is key; initial question is about eastern channel uncertainty Goal - have done right studies to constrain Hosgri slip rate Secondary goals - ibid on Shoreline slip rate, other items (e.g., Gibson meetings) Regarding Hosgri, Shoreline slip rate Intent to give verbal feedback by early October from CGS/IPRP staff.

Working backwards, arrange meetings with interpreters prior to Early October.

Chris gone October 8 th. Wills - comment that initial IPRP report (#3) appears to have been fulfilled

73 A4NR-00244, included in Appendix B as Ex. 38.

74 A4NR-00657, included in Appendix B as Ex. 39.

23 Cannot see any hazard

-significant parameter that PG&E has not investigated Cannot foresee conclusion that SSHAC should not have considered AB1632 along with other data Anticipates IPRP stating that AB1632 has fulfilled recommendations and performed what they set out to do.75 And the 62 nd slide states:

Add'l notes IPRP/AB1632 review schedule Chris anticipates schedule faster than stated 6

-8 months October - meeting on slip rates November - meeting on tectonic models December - meeting on Site conditions/site amplification Three reports. Does not anticipate a lot of depth.

Goal is to have first two reports by end of this year.

Hope its relevant to the SSHAC process as it's being done.

Once SSHAC report is done, anticipates CPUC wanting IPRP to look over the SSHAC process and report.76 Dr. Thompson's later characterization of his several interactions with CGS is consistent with the IPRP role envisioned in the September 19 slides. On October 9, 2014, he writes the TI Team, with copies to Mr. Ferre, Dr. Abrahamson, and the members of the PPRP: Incorporation of AB1632 work and IPRP review: We have had two meetings with Chris Wills and his staff (Tim, Gordon) to discuss the AB1632 reports and how the SSHAC TI

Te am is evaluating and incorporating the information. In a phone call Tuesday, Chris gave us an update on their progress and thoughts. One possible action item is that Tim and/or Gordon may be calling to set up another meeting to review data at LCI's office prior to the Oct. 23th [sic] IPRP/AB1632 meeting at the CPUC. This activity is not central

75 Id. 76 Id.

24 to the SSHAC work and schedule, but communicating with the CGS has been helpful to understand their initial thoughts and logic in digesting the AB1632 results and their relevance to the SSC.77 By the time the IPRP turned its attention to the onshore aspects of the CCCSIP Report, PG&E's Chief of State Agency Relations, Valerie Winn, is discouraging attendance by Coastal Commission staff

... you may want to touch base with Chris Wills at CA Geological Survey on whether you should participate or perhaps Chris can provide a report to your agency after the 11/17 meeting.... You are always welcome to attend, although I expect with Mark [Johnsson]

out of the office, you have a number of issues you are trying to cover!78 And neglecting altogether to invite the California Energy Commission's IPRP representatives:

Hi Valerie

Hey, Danielle and I got dropped from your email list.

I just received notification of tomorrow's meeting from Bruce Gibson and noticed Danielle and I were not on the list. Please add us back on.

Thanks, Casey79 Q 17: After the concerns expressed in IPRP Report No. 6 and NRC Research Information Letter 12-01 ("RIL 12

-01") about PG&E's reliance on only two local earthquakes in its Diablo Canyon

77 A4NR-00230, included in Appendix B as Ex. 40.

78 A4NR-00799, included in Appendix B as Ex. 41.

79 A4NR-00780, included in Appendix B as Ex. 42.

25 ground motion assessments, do PG&E's data responses identify any 2014 change in the number of local earthquakes used in PG&E's evaluation?

A 17: No. As indicated in the minutes of a December 9, 2014 meeting between PG&E and its Fugro consultants regarding the 3D velocity model, "San Simeon (2006) and Parkfield (2004) nothing else triggered the system at DCPP."80 The minutes of a follow

-up meeting on December 18, 2014 attribute the following observation to University of Texas civil engineering professor Ellen Rathje: "3D Vs model was developed solely from geophysics, are there boring data available for ground trothing?"81 The minutes record Dr. Abrahamson's response as, "No geotechnical information available."82 The limitations in recorded earthquake data used to establish Diablo Canyon site

-specific terms for the CCCSIP Report are described in Calculation Document GEO.DCPP.14.03 prepared by Dr. Abrahamson, independently verified by two other Geosciences employees, and approved by Mr. Klimczak. Of the two free

-field recording sites at Diablo Canyon, ESTA27 (with two profiles, A1200 and B1200) and ESTA28 (with one profile, A100), only ESTA27 was installed at the time of the San Simeon earthquake. Despite this longer history and the presence of two profiles rather than one, PG&E selected ESTA28 as the reference free

-field station for analytic purposes because "(t)he deeper part of the velocity profile at Station ESTA28 is more consistent with deeper parts of the velocity profile for the power block and turbine building than station ESTA27."83 Consequently, an "amplification factor" is applied to data from ESTA27, with the expected difference between ESTA27 and ESTA28 computed using the NGA

-W2 GMPEs for a

80 A4NR-00996, included in Appendix B as Ex. 43. The San Simeon earthquake took place in 2003.

81 A4NR-00997, included in Appendix B as Ex. 44.

82 Id. 83 A4NR-00831, included in Appendix B as Ex. 45.

26 magnitude 6 strike

-slip earthquake at a distance of 30 km - "This distance and magnitude is selected to capture the linear site amplification and be in the center of the data range where the GMPEs are most accurate."84 Although the velocity profile for ESTA28 is similar at depth to that used in NRC RIL 12

-01, Dr. Abrahamson acknowledges that "it has deeper soil ... which will tend to affect the low frequency amplification"85 and so "(a)n additional factor is developed to account for this difference in soil depth."86 Finally, in "smoothing" the Diablo Canyon site term derived from the two earthquakes, Dr. Abrahamson notes the contrast between similar within

-event residuals at high frequencies and "a wider range" at low frequencies, "indicating that path effect differences are being seen in addition to site term."87 He concludes, "This i s consistent with the observation that the low frequency ground motion from the San Simeon earthquake is controlled by late arriving surface waves"88 without repeating that San Simeon was not recorded at ESTA28.

Q 18: Couldn't PG&E's ocean

-bottom seismometers, approved by the CPUC in 2010 for the AB 1632 Seismic Studies program, have provided useful data on small

-magnitude earthquakes closer to Diablo Canyon than San Simeon (35 km) or Parkfield (85 km)?

A 18: Apparently not. According to the final CCCSIP Report, PG&E installed an array of four three

-component broadband ocean bottom seismometers and accelerometers in the region offshore of the DCPP in 2013. The objective of the OBS array is to improve earthquake detection capability and location accuracy for earthquakes on the continental shelf adjacent to the Hosgri and Shoreline fault zones as well as constrain the path effects from these offshore events to the DCPP.

Data are streamed in real time to the PG&E Central Coast Seismic Network for

84 Id. 85 Id. 86 Id. 87 Id. 88 Id.

27 distribution to the U.S. Geological Survey (USGS) and the California Integrated Seismic Network.89 However, PG&E's data responses offer a different perspective. An undated "Request for Major Project Contingency Release" states, The OBS system, soon after deployment in late 2013, stopped functioning due to underwater cable damage. To mitigate this problem, four temporary OBS units were deployed on November 4, 2014. The cost for the temporary units is covered under warranty. The longer term mitigation is to redesign the OBS system using an enhanced (thicker and more armored) cable to withstand sea floor abrasions. The preliminary costs for the enhanced OBS system will be shared by the manufacturer of the OBS system, Guralp (under warranty) and PG&E from this contingency request.90 The "Lessons Learned/Corrective Action" identified in the "Request for Major Project

Contingency Release" are: Do not rely on designs by others/establish a design review process for third party designs. Guralp is providing various options for the proposed cable with associated costs-benefits. A formal design review will be conducted using outside experts, PG&E personnel, and Guralp to select the optimum product considering various criteria such as cost, ruggedness, flexibility, ease of deployment, etc.91 A January 22, 2015 internal email from PG&E's Environmental Management provides

additional detail:

The existing OBS is inoperable due to a failed power/data cable. PG&E needs to replace the power/data cable and re

-calibrate the four long

-term units. Environmental Mgmt worked with the regulatory agencies to identify that repair of the system would require

amendments to existing permits and require a minimum six

-month process. Environmental Mgmt obtained necessary approvals to allow deployment of temporary

89 CCCSIP Report, Technical Summary, p. 3.

90 A4NR-0168, included in Appendix B as Ex. 46. 91 Id.

28 units so that data could be collected until the 'fix' was deployed. Temp OBS units were effectively deployed on November 3, 2014. Environmental Mgmt Vetted proposed 'fix' with NGOs and the permitting agencies. Targeting submittal of the first application by January 30, 2015 with all permits obtained in time to accommodate a September/October 2015 deployment of the 'fix'.92 Q 19: How did PG&E's CCCSIP Report address the criticisms in IPRP Report No. 6 of the utility's shear wave velocity estimates for the Diablo Canyon site?

A 19: An October 20, 2014 email from Mr. Ferre to media spokesperson Blair Jones (with copies to Dr. Abrahamson, Dr. Nishenko, and Mr. Summy) suggesting responses to newspaper comments attributed to former State Senator Sam Blakeslee emphasizes Chapter 10 of the report: Blair: Below are my proposed responses.

• • • In his own early assessment, Blakeslee said he has concerns about PG&E's conclusions, specifically about the way in which PG&E interpreted ground

-motion levels. His concerns fell on many of the same issues the IPRP discussed in its last report, issued

about one year before PG&E went public with its final conclusions. (emphasis in original) PG&E Response:

Since the issuance of IPRP Report #6, Site shear wave velocity at Diablo Canyon, PG&E developed a comprehensive 3

-D velocity model using thousands of data points obtained during the onshore seismic studies (summarized in CCCSIP Report #10).

This extensive data set was used to estimate ground motions in

-lieu of the generic shear wave values proposed in the IPRP Report #6, or shear wave values from the on

-site borehole data used in the 2011 Shoreline Report...93 While PG&E would ultimately revert to a "generic" approach in its March 11, 2015 SSHAC submittal to the NRC, Mr. Ferre's suggested response does not include the following disclaimers identified in Chapter 10 (aka "CCCSIP Report #10"):

92 A4NR-02287, included in Appendix B as Ex. 47. 93 A4NR-00611, included in Appendix B as Ex. 48.

29 This seismic

-reflection data

-acquisition geometry is not optimal to estimate shallow geotechnical 3D velocity structure, particularly velocities to depths of approximately 33 feet (~10m) or less below surface topography.94 Thus, the 3D tomography does not constrain shallow (20

-40 feet, 6

-12 m) 3D velocities within this central portion of the DCPP foundation.95 The irregular source

-receiver geometry within the DCPP area required to accommodate large structures, security zones, and other restricted entry areas produces data gaps in source-receiver offset. These gaps cause large zones where first breaks cannot be picked; the automated first

-break-detecting algorithm is designed for regular binned

-offset (conventional) 2D and 3D seismic

-reflection acquisition geometries

.96 Some areas of the DCPP contain substantial thicknesses of foundation concrete that produce fast, near

-offset first breaks. In these areas, the first breaks reflect the as-built conditions, not conditions on natural ground.97 The Vp-depth conversion compensates for the extensive distribution of higher velocity construction materials (e.g., pavement, concrete blocks, piping) that were encountered at the surface and to depths of up to approximately 30 feet during the 2012 data acquisition within the DCPP foundation area, and in areas where receiver offsets were too large to resolve Vp at depths less than 30 feet

.98 It is difficult to find a set of parameters that both produces near zero velocity bias across all six sites and also minimizes the maximum range of bias observed at any one site.99 An additional parameter to represent a horizontal gradient in Vp/Vs was evaluated but there is not a sufficient distribution of sites to realistically constrain this additional parameter.100 Q 20: Has PG&E incorporated the hypothesis offered by Dr. Douglas Hamilton regarding a San Luis Range/Inferred Offshore Fault into the three fault geometry tectonic models used in its 50.54(f) submittal?

94 CCCSIP Report, Chapter 10, p. 10 of 33.

95 Id., p. 11 of 33.

96 Id., p. 12 of 33.

97 Id. 98 Id., p. 20 of 33.

99 Id., p. 21 of 33.

100 Id.

30 A 20: Yes, although the language used in the CCCSIP Report to suggest this is a bit grudging:

Whereas the specific SLRF interpretation by Hamilton is not well supported by the available data, and by no means can be held up as a unique or preferred interpretation, the general solution of a primary, north- or north-northeast-dipping fault beneath the Irish Hills is consistent with several observations, and is a possible fault model that should be considered for seismic hazard analysis to the DCPP. We note that the interpretations b y Hamilton (2012a, 2012c) are being considered for evaluation and integration with other available data following the SSHAC Level 3 process. The SSHAC program for the DCPP, which is being performed under regulatory review by the NRC, is creating a new SSC model.101 Three and a half months previously, on May 20, 2014, the assigned PG&E author (William Page) began preparing this section of the CCCSIP Report and emails Dr. Abrahamson (with copies to Mr. Klimczak and Dr. Nishenko) with his conclusion already established:

Norm As you know from the meeting you had last Thursday with Stu and Rich, I am preparing the response to Douglas Hamilton's Testimony for 1632. I need your help in preparing a Report Section that shows that the potential ground motions calculated using his projected San Luis Range fault beneath the DCPP are enveloped by the ground motions used for the plant.

Draft conclusions are Using both the San Luis Bay fault and Hamilton's postulated San Luis Range thrust fault as a potential seismic sources [sic], the calculated ground motions are enveloped by the DCPP ground motions, and hence do not present an issue from seismic ground motions.

Let me know what you need to complete this section of the report. I'm Bill102 101 CCCSIP Report, Chapter 12, p. 74 of 82.

102 A4NR-00065, included in Appendix B as Ex. 49.

31 On June 12, 2014, Dr. Thompson emails Mr. Page, Mr. Klimczak, Dr. Nishenko, and Dr. Abrahamson (with a copy to Ms. Post):

PROPOSED RESPONSE STRATEGY

I am working on text for the response report. The approach will include what we've written in the executive summary, namely we collected 2D and 3D seismic data using the best available sources, receivers, and processing (fulfilling the commitment PG&E made at the hearing, correct?).

we don't see a moderately NE

-dipping fault in the 2D or 3D data beneath the Irish Hills.

However, the data have insufficient depth penetration to evaluate Hamilton's model as stated in his testimony and workshop presentation (we should check this one more time with ONSIP teams) his general solution of a moderately north to northeast dipping ramp explaining uplift of the Irish Hills and San Luis Range is being considered as part of the SSHAC process, as was presented at Workshop 3.

• • • To wrap up, I propose to state that the appropriate arena for incorporating Hamilton's SLRF model into hazard is through the SSHAC:

The efforts of the SSHAC are giving full consideration to the specific elements of Hamilton's SLRF model (e.g., exact length, dip, and location as indicated in his testimony) As well as general elements of Hamilton's model (consistency with seismicity, moderate dip, primary uplift rate boundary for the Irish Hills on a northwest to north-dipping fault as opposed to a south or southwest

-dipping fault).

The SSHAC process and PSHA approach allow for alternative models to be evaluated, integrated, and weighed based on their relative merits and consistency with all the available data.

The SLBF model proposed by Hamilton cannot be considered a unique or preferred solution based on the discussion above, and thus it is inappropriate to consider it in a deterministic hazard assessment.

CLOSURE: I will continue to work on edits and language for this draft report with Bill Page. In the meantime, I welcome comments and thoughts about this strategy. I think the wrong strategy is to go down a road of performing additional analyses now to try and answer an RAI103that has not materialized. It absolutely does seem worthwhile to discuss and

103 I believe this reference is to a Request for Additional Information, a procedural tool used by the NRC.

32 create an internal list of additional analyses and their relative merits that we can perform if an RAI on the topic does arise.

Thanks, -steve104 Mr. Klimczak quickly emails his concurrence105 and, several hours later, Mr. Page emails draft language that includes the following:

In addition the ongoing seismic source characterization efforts for the SSHAC are considering a moderately north

-to northeast dipping reverse fault beneath the southwestern margin of the Irish Hills similar to the geometry being proposed by Dr.

Hamilton as an alternative fault geometry that may explain the current tectonic uplift beneath the DCPP.106 Six days later, Mr. Page and Dr. Thompson exchange emails:

Steve Based in discussion with Stu, Norm and Rich yesterday, I'm working on beefing up the seismicity section (with Marcia) and adding the LCI interpretation of the geophysics that helps constrain an east dipping fault near the DCPP. The section on deterministic is deleted. Will have a draft of this late today or first thing tomorrow. Will be adding several figures.

How goes your parts?

Let's talk this afternoon.

Bill

Hi Bill, I had Matt evaluate where Hamilton's San Luis Range thrust would intersect the ONSIP

104 A4NR-00054, included in Appendix as Ex. 50. 105 Id. 106 A4NR-00743, included in Appendix as Ex. 51.

33 2D line 112

-140 - the line from San Luis Hill to Tidewater well that crosses the San Luis Bay fault.

The attached figure shows the result, and it's not promising. The upper map shows the seismic line in red, and two reference dip lines (purple and brown) oriented perpendicular to the Shoreline fault. The lower cross section has two black stars plotted in the lower left corner. These stars show where a 35 degree dipping fault that strikes parallel to the Shoreline fault and impinges on it at 1 km depth would intersect the seismic line. Only the more southerly of the two points plots on depth extent shown by the ONSIP t eam, but you can imagine a line connecting the two stars that would traverse the lower-left corner of the profile. The deep corners of 2D seismic profiles are commonly inferred to be of the poorest interpretability based on low fold. Thus, I don't think there is a basis to say we 'don't interpret' Hamilton's San Luis Range fault on this 2D seismic line. The more accurate statement is that the data are not of high enough quality at the depths proposed by Hamilton to warrant an evaluation.

Thanks,

-steve107 Q 21: Did the IPRP express a view on Dr. Hamilton's hypothesis?

A 21: Yes, IPRP Report No. 8 contains an extended discussion, including:

The model explaining the tectonic uplift of the Irish Hills hypothesized by Dr. Hamilton consists of a low

-angle northeast

-dipping thrust fault, the SLRF (Figure 6

-12), underlying the Irish Hills with a postulated surface trace almost entirely offshore. This inferred fault, would have a length of 60

-80 km extending from an intersection with the Hosgri fault, about 8 km south of Point Estero in the north, to the onshore mapped Wilmar Ave fault to the south (Figure 6

-21). The SLRF proposed by Dr. Hamilton appears to be a variation of the Inferred Offshore Fault of Nitchman and Slemmons (1994), Figure 6

-23 in Chapter 12. Along the central portion this inferred fault is coincident with the mapped Shoreline fault. The SLRF is interpreted by Dr. Hamilton to be a thrust fault dipping to the northeast that intersects the Shoreline fault at a depth of 1 to 2 km. He hypothesizes that this is the main structure accommodating regional northeast to southwest compression, which ultimately results in uplift of the Irish Hills. *** Although surface faults recognized to date appear to be consistent with strike

-slip faulting on the Shoreline fault, rather than thrusting on the SLRF, the possibility of thrust

107 A4NR-00551, included in Appendix B as Ex. 52.

34 faults in the subsurface is not ruled out by on

-land seismic survey data. The interpretation of the ONSIP data is far from unique and allows one to interpret a low angle reverse fault at the proposed location, contrary to what is stated in the CCCSIP report (p.70 Figure 6

-54). The CCCSIP interpretation criteria are not clearly defined and do not appear consistent in terms of selections made when seismic reflections are truncated.

• • • The assertion by Dr. Hamilton that seismicity beneath the Irish Hills shows an alignment that indicates the SLRF location and activity at depth is not confirmed by the more rigorous seismicity analysis performed by Hardebeck (2010, 2013, 2014a, 2014b). Hardebeck has shown convincingly that these data do not allow a unique interpretation and clearly do not strongly favor any Irish Hills uplift model. However, as previously implied, the interpretation of microseismicity has clear limitations in mapping faults and in this case also cannot be used to rule out the existence of the proposed SLRF.

• • • As presented by Dr. Steve Thompson at the November 17, 2014 IPRP meeting, the SSHAC process is considering an alternative model that includes northeast

-dipping thrust faults to explain the uplift of the Irish Hills which largely encompasses the hazard implications of the SLRF model

.108 Q 22: Is there any indication in the PG&E data responses as to how the company responded to Dr. Thompson's discussion of the fault geometry models at the November 17, 2014 IPRP meeting? A 22: Yes. That same afternoon, Dr. Abrahamson emailed Mr. Ferre - "Kent: I have been listening to the IPRP comments. Their basic comments seem to be that seismic data allow for a wide range of models beyond the range that were presented int he

[sic] [CCCSIP] report. I think that these are reasonable comments ..." Mr. Ferre responded:

Norm:

108 IPRP Report No. 8, pp. 9 - 10.

35 Steve presented the three fault geometry models in the morning...and Chris Wills asked him if he did the hazard sensitivity by giving full weight to each model. Steve answered in the affirmative and that Nick presented these at SSC SSHAC WS#3. We may want to expand or repeat some of Nick's presentation at the next IPRP meeting.109 A month later, Mr. Ferre transmitted IPRP Report No. 8 to Mr. Summy (enthusing, perhaps without reading it closely: "Overall a very positive report for PG&E."110) with the observation, They do however recognize the challenges of imaging the complex geology in the Irish Hills and the difficulty of constraining geometry of the faults at depth. But they do give credit to the 3 tectonic models, developed in the SSHAC seismic source characterization study, as capturing the range of values and kinematics to explain uplift in the Irish Hills.111 Q 23: Was PG&E committed to affording equal weight to each of the three models?

A 23: This question featured prominently in a May 16, 2014 written exchange between the SSC TI Team and the SSC PPRP after SSC Workshop No. 3. The PPRP's formal written comments to the TI Team identified the following as the final item on a list of eight action items for the TI Team: Documentation of Hazard Sensitivity to Fault Models. The three primary fault models could represent different hazard levels at DCPP. It will be important to document the hazard sensitivity of these various models to establish whether one model represents a

substantially higher level of hazard than others, and to be sure that all assumptions are well-founded and applicable uncertainties incorporated.112

This evoked the following response from Dr. Lettis on behalf of the SSC TI Team:

109 A4NR-00289, included in Appendix B as Ex. 53. 110 A4NR-01444, included in Appendix B as Ex. 54. 111 Id. 112 A4NR-00070, included in Appendix B as Ex. 55.

36 The documentation will include hazard sensitivity analyses comparing each alternative tectonic model, and elements within each model. The assumptions and technical bases for each model will be described in the SSC report. The epistemic weighting given to each model, and elements within each model, however, will be based solely on the technical assessments and not on the level of hazard that each model represents.113 Some nine months later, although it is not clear from the PG&E data responses what prompted the exchange (it may have been Mr. Halpin's planned presentation to the PG&E Board), the question of weighting the tectonic models prompted a chain of emails stretching from Mr. Halpin through Mr. Strickland and Mr. Ferre to Dr. Nishenko. It starts with Mr. Strickland on February 16, 2015 at 5:30 p.m.:

The IPRP does not see a strong reason to favor the single tectonic model presented in the CCCSIP report over the two alternative models presented by Dr. Thompson at the IPRP meeting on November 17, 2014. The Dr. Thompson approach is used in the SSHAC.114 Mr. Halpin responds at 7:12 a.m. the following morning, February 17, with a subject line reading "Re: Board Meeting", to Mr. Strickland: "I will need to really understand this one:"115 Mr. Strickland forwards Mr. Halpin's email to Mr. Ferre at 8:23 a.m. on February 17, with the request, "Can you give a one paragraph explanation that I can us [sic] in a call at 0900?"116 Mr. Ferre responds to Mr. Strickland at 8:39 a.m. on February 17, with a copy to Dr. Nishenko:

Jearl:

Three faults [sic] models were developed by the TI team to explain the uplift of the Irish Hills: 1) outward vergent (most similar to the LTSP model) 2) southeast vergent (shallow dipping San Luis bay) and 3) northeast vergent (shallow dipping los osos); th is is 113 Id. 114 A4NR-01582, included in Appendix B as Ex. 56. 115 Id. 116 Id.

37 intended to capture the center, body, and range. There is no preferred model based on available data.

Kent117 And Dr. Nishenko responds to both Mr. Halpin and Mr. Strickland at 8:55 a.m. on February 17:

Ed Here's my take -

The three SSHAC models (outward vergent, SE vergent, and NE vergent) are all variations on the same theme. The basic Outward Vergent models [sic] contains both SE and NE vergent faults and the CCSIP [sic] Report further refined the subsurface mapping and identification of the Los Osos and San Luis Bay faults. The role or activity that each one of these individual faults plays varies according to the model (e.g., SE vergent - San Luis Bay fault is dominant while in the NE vergent - the Los Osos is dominant). As Kent said, at this point there is no preferred model based on available data.

Stu118 By the morning of Saturday, February 28, 2015 Mr. Halpin's concern has shifted to understanding why the annual hazard contribution from the Irish Hills he had told the PG&E Board was 1 to 2 percent less than a fortnight earlier had suddenly climbed to 10 to 15 percent.

He emails his deputy, Barry Allen, and Mr. Strickland: "I'm not picking on anyone. I just need to be able to explain it. By Monday please."119 By 6:05 p.m. that evening, Mr. Ferre has an explanation for Mr. Strickland:

Jearl: The GMRS120 presented a couple of weeks ago was preliminary and did not have the final seismic source characterization model as an input. The biggest change between the GMRS shown a couple of weeks ago and now is the greater contribution from the

117 Id. 118 Id. 119 A4NR-01559, included in Appendix B as Ex. 57. 120 I believe this reference is to ground motion response spectra.

38 background source in the Irish Hills. This background source now contributes ~5

-15%, depending on the frequency range and the hazard level. For example, as shown in Figure 2.2.2-1 of the SPID121 report (Reference Rock Hazard by Source for 1 Hz Spectral Acceleration) the background source contributes about 8% to the total hazard at a 10

-4 annual hazard. In Figure 2.2.2

-2, in a similar curve except that it is for 10Hz, the background contribution contributes ~15% at 10

-4 annual hazard. The latest GMRS exceeds the Hosgri at 1.3Hz (by a few percent). At 10 Hz, Hosgri bounds the GMRS.

The Technical Integrator team used the best scientific method to characterize the background source. Their method was presented and accepted by the PPRP. It is a technically defensible approach. The background source model from the Shoreline study resulted in a lesser contribution to hazard, but the model did not reflect the current technical approach. Since the AB1632 report is based on a deterministic comparison, a background source is not applicable.

Keep in mind, the GMRS in the SPID is not final. Norm will rerun the hazard with the final WAACY Magnitude

-Distribution model. Norm has said that based on his preliminary runs, the WAACY will result is [sic] a slight increase in the high frequency range (~1

-2% increase) but a negligible change in the low frequency. The final GMRS is expected to be completed mid next week.

Kent122 Monday morning, March 2, 2015, at 6:41 a.m., Mr. Strickland provides a response to Mr. Halpin, augmenting Mr. Ferre's explanation somewhat:

he deletes the examples drawn from the SPID report.

he deletes the reference to background sources being inapplicable to a deterministic comparison like the CCCSIP report. he deletes the reference to the WAACY Magnitude

-Distribution model, and the prediction of a ~1

-2% increase in the high frequency range with negligible change in the low frequency.

he revises the final GMRS completion from "mid next week" to "tomorrow."

121 I believe this reference is to PG&E's Seismic Hazard Screening Report, the primary 50.54(f) document submitted to the NRC on March 11, 2015, whose name derives from an EPRI document entitled "Seismic Evaluation Guidance: Screening, Prioritization, and Implementation Details (SPID) for the Resolution of Fukushima Near

-Term Task Force Recommendation 2.1: Seismic."

122 A4NR-01559, included in Appendix B as Ex. 57.

39 he adds the explanation that "Norm used the contribution for the Irish hills that he used in previous hazard evaluations. He revised his model when he received the final input from the Seismic Source team." he adds the observation that "Completion of the SSHAC seismic source should have been completed late last year to avoid last minute changes like this."

and he expands the description of the background source approach to include, "It consider [sic] three postulated (different) models for the Irish Hills while the AB 1632 studies only consider one. The CPUC IPRP noted in their comments that there was no strong evidence to prefer the model that was used in the AB 1632 studies and that the uncertainty of the three models should be considered."123 And the tectonic models for the Irish Hills play a prominent role some two weeks later, on March 19, eight days after submittal of PG&E's SSHAC report, when media spokesperson Blair Jones finalizes talking points for a SLO Tribune article. Mr. Jones has assembled a number of rebuttal arguments to claims he says former State Senator San Blakeslee and IPRP member Bruce Gibson have made regarding the content of the SSHAC report.124 Soliciting input from Mr. Ferre, Mr. Strickland, Dr. Nishenko, Regulatory Services Manager Thomas Jones, and Mr.

Halpin, the final question posed is whether the SSHAC seismic hazard re

-evaluation was "at all altered based on feedback from the IPRP?"

The carefully crafted answer:

... An example of how IPRP feedback was incorporated had to do with the Irish Hills. The [AB 1632] seismic studies identified a single preferred fault model for the Irish Hills. The IPRP, after receiving additional insight from the SSHAC team, supported that multiple models be utilized. The SSHAC accepted the IPRP's input and utilized three separate, equally weighted models to characterize faults in the Irish Hills...125 Q 24: So, the SSHAC report submitted to the NRC uses three equally weighted models?

123 A4NR-01881, included in Appendix B as Ex. 58. 124 A4NR-02004, included in Appendix B as Ex. 59. 125 A4NR-01870, included in Appendix B as Ex. 60.

40 A 24: No. As explained in Chapter 7 of the SSC SSHAC report, the Outward Vergent and the Southwest Vergent models are each weighted at 0.4, while the Northeast Vergent model is weighted at 0.2.126 Q 25: What is the hazard sensitivity of that unequal weighting?

A 25: As explained in Chapter 14 of the SSC SSHAC report, at the 5 Hz spectral frequency the different models "correspond to noticeable differences in hazard" the Outward Vergent resulting in ground motion ratios slightly less than unity, the Southwest Vergent yielding ground motion ratios near unity, and the Northeast Vergent yielding ground motion ratios slightly in excess of unity. There is less difference at the 0.5 Hz spectral frequency.127 Q 26: Do the PG&E data responses indicate any obvious omissions from the CCCSIP Report? A 26. Yes, two in particular do. One is Mr. Klimczak's July 18, 2014 email to Dr. Abrahamson, Dr. Nishenko, and Ms. Post providing his suggested response to the CEC's original 2008 recommendation concerning an earthquake directly beneath the plant. First he frames the issue: PG&E should assess the implications of a San Simeon

-type earthquake beneath Diablo Canyon. This assessment should include expected ground motions and vulnerability assessments for safety

-related and non

-safety related plant systems and components that might be sensitive to long period motions in the near field of an earthquake rupture.128 Then he provides the desired response:

126 SSC SSHAC Report, Rev. A, pp. 7 7-25. 127 Id., p. 14-5. 128 A4NR-00546, included in Appendix B as Ex. 61.

41 The Shoreline fault report (2011) included a San Simeon

-type earthquake beneath the Irish Hills and DCPP where the San Luis Bay fault (50° - 80° north) and the Los Osos fault (45° to 75° Southwest) intersect at depth. The SSC SSHAC logic trees will consider various fault models to explain the uplift of the Irish Hills, including a San Simeon

-type earthquake model.129 Both of Mr. Klimczak's paragraphs made it into the CCCSIP Report essentially intact.130 By misstating the original CEC request, which specifically discounted probabilistic assessments, PG&E's answer is substantively unresponsive.

The 2008 CEC report emphasized "an earthquake directly beneath the plant" and not simply one beneath the Irish Hills:

Another potential seismic hazard at Diablo Canyon occurs from the possibility of an earthquake directly beneath the plant. Based on seismologic interpretations and conclusions from investigations of the 2003 San Simeon earthquake (magnitude 6.5) that occurred approximately 35 miles north of the Diablo Canyon site, the tectonic (geologic plate) setting where this earthquake occurred appears similar to the local tectonic setting of Diablo Canyon. The deep geometry of faults that bound the San Luis

-Pismo structural block, where Diablo Canyon sits, is not understood sufficiently to rule out a San Simeon

-type earthquake directly beneath the plant.131 And the CEC

-prescribed analysis was also clear:

PG&E has considered a San Simeon

-type earthquake scenario within probabilistic seismic hazard assessments for Diablo Canyon. However, further studies that consider such an earthquake from a deterministic basis (i.e., using a probability of 1) are recommended to evaluate the full implications of this earthquake, particularly for non

-safety related plant components and reliability.132 129 Id. 130 CCCSIP Report, Chapter 14, p. 3 of 5, makes minor typographical changes to the second paragraph.

131 CEC, An Assessment of California's Nuclear Power Plants: AB 1632 Report, November 2008, p. 5.

132 Id., p. 7, footnote 6.

42 The second PG&E data response which points to a significant omission in the CCCSIP Report is the "DCPP Seismic Hazard Update Summary" which is identified as "Results from a 12/6/13 CNO Seismic Update meeting."133 This meeting occurred during the mid

-2013 to mid

-2014 period when there was virtually no communication between the IPRP and PG&E and the document describes the anticipated content of a final report then planned for completion in June 2014: The June 2014 report will also have deterministic ground motion spectra plots for the Hosgri, Los Osos, San Luis Bay and Shoreline faults based on multiple faults linkage.134 i. Hosgri will be linked to faults up to Mendocino Triple Junction (located offshore N. California) ii. Los Osos linked to Hosgri iii. San Luis Bay linked to Hosgri iv. Shoreline linked to Hosgri

v. Linkage of faults allows for larger magnitude earthquake possibilities that have low probability of occurrence 1. Deterministic plots are not based on earthquake recurrence rates

2. A hybrid approach, deterministic criteria with probabilistic earthquake magnitude recurrence consideration, will be taken in that our plots will be based on magnitudes at 10

-6 annual recurrence rate. This approach addresses fault linkage that is acknowledged by the technical community as possible based on the analysis of earthquake data without overly penalizing ourselves with large magnitude earthquake possibilities with very low probability of occurrence.

a. Justification of the use of 10

-6 would be it is a reasonable cutoff for deterministic analysis, less than 1 in a million chance of exceeding the selected magnitude of an earthquake on each fault.

PRA group (Nathan) has provided justification for the use of 10

-6 as follows:

133 A4NR-00660, included in Appendix B as Ex. 62. 134 An exchange one month later between Mr. Klimczak and Dr. Carola DiAlessandro, Project Manager for the SWUS GMC SSHAC process, identifies the degree to which the severe potential of these joint rupture cases is diminished by probabilistic analyses when run at 10

-4. Dr. DiAlessandro: "The relative change in predicted spectral ordinates in

[sic] around 150%! That's why we refer at [sic] the change as LARGE." Mr. Klimczak: "I understand that the spectral accelerations significantly increase from a rupture on the SLB or LO when they are linked to the Hosgri/San Sim but, when the contribution of each fault and the low rate of occurrence for these linked ruptures are accounted for in the development of the hazard the impact is small." A4NR-00098, included in Appendix B as Ex. 63

.

43 A recent reference that defines design basis accident scope can be found in INL/EXT 19521 "Next Generation Plant Licensing Basis Event Selection White Paper". This paper describes the frequency based categories for licensing basis events (LBEs). These categories are anticipated operational occurrences (AOOs), infrequent design basis events (DBEs) and beyond design basis events (BDBEs).

AOOs - greater than 10

-2 per plant-year DBEs - < 10-2 and > 10-4 per plant-year BBDEs - < 10-4 For seismic events, definition of a lower limit on BDBE's can be aligned with the concept of the 'maximum credible earthquake' or MCE (NUREG

-0800 Rev. 2). The MCE represents the level of ground motion for which a nuclear power plant must be designed to safely shut down and is defined as the 'largest earthquake that can be reasonably expected to occur on a geologic structure in the current tectonic region'. An accepted quantitative measure of the credibility or reasonableness of severe accident consequences has been established as a frequency of core damage of less than 10-6 per year (Regulatory Guide 1.174). This criterion can be conservatively applied to the frequency of earthquake occurrence for the purposes of identifying the characteristics of the MCE. Also supporting the determination of a 10-6 per year frequency of occurrence criterion for the MCE is PRA scoping guidance from Regulatory Guide 1.200 which states that an external event can be screened from a plant probabilistic risk assessment (PRA) if a conservative analysis shows that the event's contribution to CDF is less than 10

-6 per year. Therefore, a seismic hazard frequency of 10-6 per year conservatively meets the RG 1.200 external event screening criterion.

b. Plots at magnitudes based on 10

-7 annual recurrence rate will also be developed but, not included in the report

c. The report will also note that there is a probabilistic seismic hazard update in process that considers all possible magnitudes of earthquakes at any annual recurrence rate.

The "Results from a 12/6/13 CNO Seismic Update meeting" document also identifies the logical inconsistency of confining PG&E's linkage

-based update to the NRC's RIL 12

-01 to only the greater capability of the Shoreline Fault:

44 The RIL letter requires an update to the NRC if the Shoreline Fault is found to be more capable. Linking the Shoreline Fault (SLF) to the Hosgri will allow for a higher magnitude earthquake on the SLF and thus be more capable. The other three faults in the SLF report will also be more capable due to linkage. Need to decide how this information is to be passed on to the NRC with Regulatory Services and the Law department.135 Finally, the "Results from a 12/6/13 CNO Seismic Update Meeting" document suggests a deterministic evaluation requested by the previous Chief Nuclear Officer, John Conway, will also be kept out of the CCCSIP report:

J. Conway's M8 deterministic request:

i. Norm has provided plots of M8 earthquakes on the Hosgri, SLB, Los Osos and Shoreline.

ii. It was decided to only evaluate the 'critical' SSCs in any frequencies of exceedance ranges to show they can perform their safety function. Action by Seismic Projects.

iii. This is an internal evaluation only.136 Q27: Is A4NR prepared to make a ratemaking recommendation with regard to the DCSSBA?

A27: Not at this point. PG&E has not yet responded to significant A4NR data requests, and

our specific proposals will have to be deferred to the briefing stage of this proceeding.

135 Id. 136 Id.

Appendix A

Professional Qualifications of John Geesman

PROFESSIONAL QUALIFICATIONS OF JOHN GEESMAN John Geesman is an attorney with the Oakland law firm, Dickson Geesman LLP.

He was a member of the California Energy Commission from 2002 to 2008, and its Executive Director from 1979 to 1983. Between his two tours at the Commission, Mr. Geesman spent 19 years as an investment banker focused on the US bond markets. He has previously served as Co-Chair of the American Council on Renewable Energy, Chairman of the California Power Exchange, President of the Board of Directors of TURN, a Board Member of the California ISO, and Chairman of the California Managed Risk Medical Insurance Board.

He is a graduate of Yale College and the UC Berkeley School of Law.

Appendix B

PG&E Data Responses