ML093100574
| ML093100574 | |
| Person / Time | |
|---|---|
| Site: | Harris  |
| Issue date: | 11/03/2009 |
| From: | NRC/OCM |
| To: | |
| References | |
| M091103B | |
| Download: ML093100574 (125) | |
Text
1 1
2 UNITED STATES NUCLEAR REGULATORY COMMISSION 3 BRIEFING ON FIRE PROTECTION LESSONS LEARNED FROM 4 SHEARON HARRIS 5
+ + + + +
6 TUESDAY 7
NOVEMBER 3, 2009 8 + + + + +
9 The Commission convened at 9:30 a.m., the Honorable 10 Gregory B. Jaczko, Chairman, presiding.
11 12 NUCLEAR REGULATORY COMMISSION 13 GREGORY B. JACZKO, CHAIRMAN 14 DALE E. KLEIN, COMMISSIONER 15 KRISTINE L. SVINICKI, COMMISSIONER 16 17 18 19 20 21 22
2 1 PANEL 1: INDUSTRY REPRESENTATIVES 2 JOSEPH W. DONAHUE, Vice President Nuclear Engineering & Services, 3 Progress Energy Carolinas, Inc.
4 5 DAN PACE, Senior Vice President, Engineering, FENOC 6
7 KEN CANAVAN, EPRI 8
9 PAUL GUNTER, Director of Reactor Oversight, Beyond Nuclear 10 11 12 13 14 15 16 17 18 19 20
3 1 PANEL 2: NRC STAFF 2 BILL BORCHARDT, EDO 3 JACK GROBE, NRR 4 ALEX KLEIN, NRR 5 DONNIE HARRISON, NRR 6
MARK SALLEY, RES 7
8 9
10 11 12 13 14 15 16 17 18 19 20 21
4 1 P-R-O-C-E-E-D-I-N-G-S 2
3 CHAIRMAN JACZKO: We will begin then, we 4 ask our guests to come to the table. We have today 5 I think -- this is one in a series of meetings that 6 the Commission has held on the issues of fire 7 protection. Last year -- a little over a year ago --
8 we held a meeting and I think one of the items that 9 came out of that was a recognition and a direction 10 to the staff from the Commission to go back and look 11 at and to start cataloging and get a better 12 understanding of what we were considering to be a 13 baseline for where we are in fire protection and how 14 we would be clear that we were making progress in 15 fulfilling our responsibilities to ensure that 16 plants address this issue in a safe and appropriate 17 way.
18 I think today we certainly are at a significant 19 point in our efforts to address this issue.
20 There are effectively I think three or four 21 significant issues: the multiple spurious 22 actuations, the use of operator manual actions, the
5 1 long term use of interim compensatory measures, and 2 other types of unapproved exemptions. Those are 3 really the challenges I think that we have in the 4 fire protection area.
5 As we stand today I think we are on the cusp of 6 making significant progress in a variety of those 7 issues. The staff recently issued the Reg Guide on 8 multiple spurious actuations that will begin a 9 six-month clock for licensees to identify or develop 10 their approaches for resolving that issue.
11 The staff has at this point we should have all the 12 responses on the operator manual actions from 13 licensees. The challenge now will be to verify the 14 issues that have been corrected in that area have 15 been corrected appropriately.
16 Then of course we have and will hear today 17 discussions about our progress in approving license 18 amendments or reviewing license amendments for the 19 two pilot plants that are transitioning to NFPA 805.
20 As I've said many times I continue to believe that 21 that's the right solution for everyone and I think 22 today I hope to hear from all of you about what your
6 1 thoughts are and whether that's the right approach 2 and what the challenges are and if there are 3 weaknesses with that approach how we can improve it 4 so I look forward to hearing from stakeholders first 5 and then after that we will hear from the staff. Any 6 comments from my fellow Commissioners?
7 COMMISSIONER KLEIN: I think that the NFPA 8 805 is probably been challenging both for the 9 industry and for our staff and I'd like to 10 compliment the industry and the staff for both 11 working hard to trying to resolve this.
12 It turns out it's more technically challenging than 13 we thought when it begin and I think from my 14 perspective this may be an area where one size 15 doesn't fit all.
16 I think different plants will have different 17 requirements and I look forward to your 18 presentations.
19 CHAIRMAN JACZKO: Okay, then I think 20 we will begin with Mr. Donahue.
21 MR. DONAHUE: Hello, my name is Joe 22 Donohue, I'm presently the Vice President of Nuclear
7 1 Oversight and I recently changed positions and I am 2 pleased to discuss with the Commissioners the status 3 of the Harris pilot.
4 What I plan to cover today on my first slide is 5 basically I plan to talk about the Harris fire 6 protection status. I'll talk a little bit about the 7 transition results and then some insights going 8 forward.
9 I think you'll hear today as the panel from the 10 utilities talk, I think we did underestimate what it 11 was going to take to implement both the fire PRA, 12 the licensing actions, the NFPA 805 and really you've 13 got to lump all of those together and we sort of 14 lump them in discussions we'll call NFPA 805 but 15 it's all of those.
16 The transition costs from our initial estimates have 17 doubled, in the case of my utility, we estimated it 18 with modifications to be near $40 million and it 19 will be near $80 million when we're done and more 20 important is not so much the money is the pure 21 man-hours that's going in to the transition.
22 The PRA cost itself -- one of the things when people
8 1 think about NFPA 805 we not only piloted the 805 2 transition we also piloted the Reg Guide 1.200, Fire 3 Reg Guide 6850, Fire PRA, and we also piloted --
4 well soon to be piloted -- the transition so we'll 5 talk a little bit about that.
6 Next slide, please.
7 The Harris NFPA 805 status we have responded to the 8 RIA responses and Supplement 3 was submitted in 9 October 2009. We had over 140 PRAs, we had lots of 10 engagement with the staff, we had lots of engagement 11 with the staff during the pilot period.
12 Obviously as the Reg Guide 1.205 and other what we 13 exactly we wanted to see in LAR and exactly to see 14 in the PAR documentation did cause us quite a bit of 15 re-work.
16 We are on track for program implementation in 17 mid-2010. We expect the SER from the staff in first 18 quarter of 2010 we will then implement the process 19 with license conditions to complete modifications 20 sometime in several months after receiving the SER 21 so mid to the summer of 2010.
22 We are committed on other actions to complete our
9 1 modifications at the Harris project by 2010, by the 2 end of 2010.
3 We are on track and do plan to commit those 4 modifications and then we will be in a 100%
5 compliance of the implementation of the rule, 6 implementation of the LER and the modifications that 7 go with those.
8 Next slide, please.
9 Progress of our fleet status, one significant thing 10 we do have to keep in mind as I have said we piloted 11 the development of an LER, the transition, the 12 methodology with both NEI, the industry and the 13 staff, we piloted the submittal process getting the 14 Fire PRA through various peer reviews in the 15 documentation of those.
16 The last remaining piece that we cannot 17 underestimate is the actual transition of 805 at the 18 site, transition from the staff to the Regions for 19 inspections in ensuring we minimize subjectiveness 20 in the inspection process and then piloting the 21 incorporation of the 805 process into the STP 22 process so that it all comes together and those are
10 1 a phase that we need to keep focus on.
2 I do believe we have a pretty good understanding 3 now.
4 The licensing process we have to drop it all the way 5 through to completion.
6 We are incorporating lessons learned in our fleet 7 response we have three other plants that are being 8 done in parallel.
9 We are incorporating the LERs and we're 10 incorporating the Rev 1 to RG 1.205.
11 It will cost us six to nine months to lay in our 12 next unit.
13 We are reassessing our schedules as a company, we 14 are committed to the 805 transition.
15 But we have to reassess our schedules as we 16 pull things together and incorporate the lessons 17 learned into our next non-pilot plants and we'll 18 also incorporate what we find as we move into the 19 inspection process which we realize we are the pilot 20 for that also.
21 Next slide please.
22 Quickly the Shearon Harris Fire Protection
11 1 improvements over 5,000 or 4,000 scenarios are now 2 evaluated.
3 This is by far the most complicated computer 4 modeling and fire modeling that I've ever been 5 involved in.
6 We originally had an internal events 40 7 scenarios, we're modeling near 4,000 for the detail 8 that we added to our model.
9 We will reduce the reliance on manual actions.
10 Many of our manual actions go to defense-in-depth, 11 they're still there but we do not need to rely on 12 them anymore so they're in the procedures as defense 13 in-depth and those ones which the few that we do 14 have are part of the LAR and will be 15 approved as part of the SER as we accept that.
16 The NFPA-805 transition and the modifications have resulted 17 in overall plant risk reduction at Harris and that 18 may be something that is plant unique and condition 19 unique to each of the individual plants. In our case 20 we've upgraded and can adequately take credit 21 for the Hemyc and fire wrap barriers.
22 We've upgraded those, in addition many of our mods
12 1 are in that area. More importantly, we're adding 2 incipient fire detection to bring on the onset of a 3 potential fire in cabinets which were not there.
4 That is not something we had at one plant. At Harris 5 we did have the already available, at Robinson or 6 more importantly we're adding a diesel generator 7 alternate seal injection which has huge internal 8 events risk safety as well as fire protection 9 safety.
10 So again for Harris we have improved safety.
11 28 total mods have been completed to date, 45 is the 12 total number which will be done by the end of 2010, 13 seven of which need to be tied in during the outage 14 and that's why it's the end of 2010.
15 So 45 is a total mods we're doing.
16 Next slide, please.
17 One key point as you actually implement 805 is now the 18 decision-making process during the post-transition 19 and that's again a very important part of 805 is the 20 ability to allow us as times change and as the 21 plants go on that we can have a process to make 22 changes in the fire protection process.
13 1
As we do it we will ensure fire defense in depth.
2 This is not a one size, one item just NFPA 805 you 3 really integrate classical fire protection safe 4 shutdown analysis Fire PRA which we always sort of 5 focus around the Fire PRA in other aspects.
6 So we'll also use multiple input sources as we move 7 forward.
8 Now all of those things bring some potential for 9 subjectivity and that's the part we have to work 10 with as we integrate all three of those but that is 11 an important part of NFPA 805 transition. And risk 12 informed post-transition fire change processes are 13 a very important part, it's part of the code its part of the regulations 14 and we have several FAQs in that area and that will 15 be another area that we have to demonstrate jointly 16 between the staff and the utilities as we move 17 forward.
18 The key here is putting a focus on plant safety based 19 on very specific plant information on ignition 20 source information for that plant.
21 Next slide, please.
22 There has been some improvements recently over the
14 1 last 6 to 8 months as we've worked through the NRC 2 guidance documents, in particular the original 3 versions of Reg Guide 1.205.
4 We had some concern with the utilities, we are 5 closer to allowing the peer review process of Reg 6 Guide 1.200 there have been improvements in the 7 latest rev of that.
8 The industry appreciates that.
9 Still some work has to be done but that's an area we 10 have improved.
11 We have had improvements in some movement in some 12 of the FAQs, I think you'll hear, they're some areas 13 that we need to continue to look at even some of the 14 open FAQs, some of the FAQs that.
15 are being perceived as being closed 16 which are really more in an interim status.
17 But in particular the FAQ 0608 which has to do 18 the engineering equivalency evaluations which is the change 19 process post-805, there has been some clarification 20 in those that are beneficial.
21 There was clarification in the documentation needed 22 for recovery actions.
15 1
Basically going from shutdown outside the control 2 room in Appendix R space to the same shutdown 3 outside the control room to NFPA 805 -- how much 4 documentation that you had a show on there so there 5 has been some improvements with those.
6 Next page please.
7 Going forward, you know, we really have to continue 8 to strive on the stable regulatory environment and I 9 think you'll hear that from several of my peers. And 10 again as I mentioned before we need to pilot the 11 inspection process, we need to minimize the 12 subjectiveness as we take a look at both the 13 inspections and as we look at 6850 and the change 14 processes as we go forward and how that's all 15 incorporated in other regulatory processes including 16 the SDP process.
17 We have to also keep in mind that we need to have, and 18 the key there is a consistent application of 6850 as that document evolves 19 as it is today incorporating FAQs in anything that 20 we do. Really its around many of the issues that are 21 in the 6850 arena.
22 Next page please.
16 1
A fire PRA conservatism realism you can call it 2 either of those. I really like the word realism.
3 The key area for us is continuing to focus on fire 4 growth and the heat release rates in electrical 5 cabinets that really is our priority. It is where 6 we're putting incipient fire protection and it is an 7 area that several others will talk about that we 8 need to continue research on that 9 minimizes some of the realism in the model.
10 And also how we treat internal events which is a 11 very mature over many years of gathering data 12 typifying the way we operate our plants today with 13 the fire PRA which is a little bit less mature and 14 how we integrate those and how we add those to or 15 don't add those to or treat those as fire hazards.
16 With that, I'm available for questions.
17 CHAIRMAN JACZKO: Well, thank you, we 18 appreciate your perspectives and we'll now hear from 19 Mr. Pace who will talk about a plant that is not 20 currently a pilot plant but is thinking about 21 transitioning to NFPA 805 so we'll hear about his thoughts.
22 MR. PACE: Thank you Chairman.
17 1 I was asked to come in and talk to you a little bit 2 about those that follow the pilot and so my 3 comments today are going to be really as I view some 4 of the generic concerns with those post-pilot 5 plants more than with any unique specific item for 6 one of my plants.
7 Next slide.
8 A little bit about who we are FirstEnergy.I have 9 four units, two relatively new as far as our 10 industry goes and two mid-seventies plants, two 11 Beaver Valley units which they are sort of in the 12 order that I'm approaching 805. Davis-13 Besse and Perry which is a relatively new 14 BWR. Next slide.
15 So why did we decide to transition to 805?
16 I was sort of the advocate for the post-transition 17 plants and in the job I have now I look at a few 18 items to force my decision. One is I looked at it as 19 an opportunity to improve safety through a risk 20 informed approach.
21 Having lived through the power PRAs it was clear 22 that we gained some insights that weren't at all
18 1 intuitive coming up in a deterministic world that 2 the model showed us that this is something that 3 you need to focus your attention on.
4 Secondly, I've had a long career in fire protection 5 and I don't like to admit to everybody starting 6 early in my career as a fire engineer in startup and 7 dealing with that.
8 I was design manager at the first Appendix R pilot.
9 I was a director at the first fire protection and functional 10 inspections in the nineties so I've followed these 11 issues for a long, long time and I looked at it as 12 putting it into an analytical tool maybe gave us our 13 best opportunity to resolve the long-term legacy 14 issues in the fire protection arena.
15 I looked at it from practical standpoint.
16 The fleet I was working in the fleet now had a very 17 diverse approach to fire protection at each of the 18 sites and I looked at it as an opportunity to bring 19 the fleet into alignment and consistent method of 20 handling fire protection and we didn't have very 21 strong PRA resources at the time.
22 I know this is an opportunity to leverage to new
19 1 people coming into the business and to grow the PRA 2 capability of the company.
3 So those are sort of the things that said we will 4 embark down this path.
5 Next slide.
6 We did an initial feasibility study for each site.
7 It felt like something we could accomplish.
8 I was aware from my perspective that if we approach 9 this in a pragmatic way we ought to be able to do 10 a fire protection code reconciliation including a PRA 11 model for around a million dollars.
12 NEI had said 3 million which I thought was a 13 ridiculous figure.
14 But we'll see more to come.
15 We sent a letter of intent in, in December of 2005 16 and it implemented a fleet approach.
17 We said we would pilot the Beaver Valley Unit One 18 followed with Beaver Valley Unit Two, followed then 19 with Davis-Besse and then with Perry.
20 Based on their age and potential significance of their 21 issues so we would look at the pilot, implement it 22 at Beaver then Besse then Perry.
20 1
Next slide.
2 So at Beaver Valley our initial estimates after we did 3 the project reports and things were about $7 million 4 for two units, so three and a half-million dollars a unit, 5 closer to the NEI estimate, that includes no NRC 6 review fees or modifications, that's just to build 7 the program.
8 As of this year we're currently estimating a cost of 9 $15.4 million also including no NRC review fees and 10 no modifications.
11 That's an uncertain number because as you'll see 12 some -- we still believe that there are a lot of 13 unknowns that we have to deal with.
14 Our unit one screening model was done some time ago 15 and we embarked on fire area modeling trying to 16 stay just behind the pilots. In hindsight that 17 may not have been a wise approach because that's 18 caused a great deal of recycling of resources. As the 19 pilot adjusted we became the tail on the dog that 20 was whipping us around.
21 As a matter of fact one of my contractors said to me 22 one day, "We'll keep doing this until you guys run out of
21 1 time or money", to give you a mental image of what 2 we're going through this, it's recycling.
3 Unit two we're about 30% complete. We're really 4 trying to watch unit one and the pilots to see where 5 we launch next with unit two, but probably quite 6 capable of pulling Beaver Valley both units across 7 within the projections of the industry.
8 Next slide.
9 The Davis-Besse transition started a little later.
10 We had submitted in '05 assuming a three-year 11 transition for each plant so we should have 12 been through with Beaver in early '08.
13 We would have had Besse done in the fall of '09 14 about this time and be starting on Perry.
15 As it turns out it's really sort of stalled the 16 Besse activities because we're throwing good money 17 after bad in some cases to see where the pilots land 18 and then what sort of experience we have in 19 transitioning our Beaver Valley unit.
20 After that Perry will go after Beaver and Besse 21 has been submitted and then as I said it will be in 22 a much better condition from the deterministic fire
22 1 protection standpoint.
2 Next slide.
3 There are four issues I'd like to discuss with you. One is 4 the methods used in selecting the modeling inputs, 5 we think is a concern; schedule overlap between 6 pilots, and post-pilot transition plants; 7 our solution as an industry and a regulator, I 8 believe, has introduced new challenges to us; and 9 cost-benefit of the 805 transition.
10 Next slide.
11 In the methods arena some of the inputs arguably are 12 deterministic based and when you put deterministic 13 inputs into a probabilistic model you get ridiculous results.
14 That's some of the issues you see the staff working 15 back and forth with the industry on, its trying to 16 bring some reconciliation to what's the appropriate 17 input and parameters to run in the model.
18 The model is a fairly straight forward tool. It's a 19 mathematical model but poor inputs in will result 20 in poor outputs.
21 This would yield unrealistic modeling outputs. It's 22 akin to the engineers when they do calculations and
23 1 they come back and tell me the car was going 2 337 miles an hour when it hit the wall which is 3 an impossibility.
4 Therefore the results arent comparable to the power 5 models.
6 Since they were built with different inputs 7 comparing the fire models that we're building today 8 with the power models that we're running today is 9 akin to comparing tomatoes to apples, they're not 10 even in the same variety.
11 And then some of the existing standards that we've 12 used in power models don't translate well to what 13 we're trying to deal with in the fire modeling arena.
14 Next slide, please.
15 As far as the scheduling overlap since the pilot 16 plants had such a difficult time in defining the 17 course we're going to follow, their schedule has 18 slipped. As I said earlier we should have been 19 through Beaver in '08, Besse in '09, Perry next 20 year.
21 We're struggling with Beaver because we're still 22 trying to understand the outputs of the pilots.
24 1
Since we laid out a fleet approach you can imagine 2 we've got a bough wave going and now all three plants have 3 fallen on top of each other which is a physical 4 impossibility to manage, so we're we considering those 5 schedules.
6 It has resulted in substantial rework.
7 These models are very complex and the mental picture 8 I have is building one of these childs models of dominoes 9 and filling the room with them and just about the 10 time you're through with the Domino kick over model 11 we decide we'd like to put an extra dot on each of 12 those dominoes.
13 So we have to sweep them all down and redot them and 14 rebuild it, that's akin to the cycling that we're trying to 15 build in these very complex models so we lost the 16 fleet benefit and it's challenging the limited resources 17 which were limited even with the schedule we laid out.
18 Next slide, please.
19 As I said I believe the solution has possibly 20 introduced new challenges.
21 We, the industry, and the regulatory body set out to 22 solve manual operator actions and circuit analysis
25 1 issues that you led out with Mr. Chairman at the 2 beginning.
3 And we're coming to realize we have to solve the 4 circuit analysis issues to solve the fire PRA 5 issues.
6 So the deterministic solution of those issues has 7 been working in parallel. In addition we've added 8 more complexities in the modeling confusion so we start 9 out with two issues and arguably now have eight or 10 ten to solve as an industry.
11 Next slide.
12 So in the cost-benefit arena obviously our original 13 assumptions were no longer valid the deterministic 14 approach is gaining certainty now.
15 We know more about how to do circuit analysis now 16 which once again I say has to be solved to do the 17 PRA approach.
18 So we may well have been better served to put the 19 15, 18, 20 million dollars into doing the 20 deterministic approach and making modifications as 21 necessary to comply rather than building a more complex 22 analytical model.
26 1
Analytical models do not build margins at nuclear power 2 plants.
3 Next slide.
4 In summary I believe our risk informed approach 5 was good intention, we all set up to try to frame 6 an analytical disagreement into an analytical model 7 we could agree on.
8 We still need to resolve some of these open 9 technical issues on what modeling inputs we need to 10 assume and make it more akin possibly to the power 11 models.
12 I think the front end transition plants will 13 complete. Much like Beaver Valley, we're in this 14 poker game and all the chips are in.
15 The follow-on plants will much more likely have 16 to re-evaluate their position.
17 Thank you.
18 CHAIRMAN JACZKO: Thank you I appreciate 19 those insights and I will now turn to Mr. Canavan 20 who will talk a little bit about the research 21 activities and the work that EPRI has been doing in 22 fire PRA and other issues.
27 1 MR. CANAVAN: Thank you, Mr. Chairman, 2 Commissioners.
3 I'm the Manager of the Risk and Safety Program at 4 the Electric Power Research Institute and I want 5 to thank you for the opportunity to brief you and 6 other interested parties on where we are or lessons 7 learned realized during the performance of the PRAs in 8 support of the NFPA 805 transition.
9 We've been at several briefings and I titled my 10 presentation as lessons learned after Joe Donahue's 11 presentation because I think there's some important 12 lessons that we can learn but before we get to that 13 I thought I would give a brief history and when I 14 say brief I mean brief history on Fire PRA methods.
15 I'd like to correct my initial bullet I wrote 20 16 years and then last night as I was thinking about it 17 it's probably not quite 20 years ago where we 18 started the wide use of fire PRA methods maybe even 19 more like ten.
20 But that was in the performance of the individual 21 plant examinations for external event vulnerabilities 22 or the IPEEE and that was really the first time that
28 1 we wide-spread used fire PRA methods.
2 Since that time those methods have been modified by 3 NRC RES and EPRI and they are published in NUREG 4 CR 6850 and EPRI 1011989 and the EPRI number is too 5 complicated to say so we usually refer to it a 6850 6 and guess we should re-number our report.
7 But as we mentioned in other Commission briefings we 8 discussed that the 6850 methods were not fully 9 trialed prior to their implementations in the pilots 10 that our transitioning to 805.
11 And as expected with any complex methodology we 12 expected and did indeed run into some issues with 13 the methods.
14 You've heard some of them over the last few 15 Commission briefings.
16 And over the last 18 months we worked with the staff 17 and other stakeholders including utilities, owners 18 groups, consultants and NEI to improve those 19 methods.
20 We've had some successes and some setbacks.
21 Today's presentation isn't going to focus on 22 particular issues or their successors or their
29 1 failures, but rather try to focus on answering the 2 questions: how can we do this a little better, how 3 can we refine the methods more efficiently and more 4 effectively and I hope my next couple of slides 5 address those questions.
6 Next slide, please.
7 Prior to the start of the NFPA 805 pilots and 8 probably in the beginning of those transitions as 9 well at least for a short duration, the PRA methods 10 developments and handling of the issues were 11 performed by a relatively small group of researchers 12 who were the authors of much of the material that is 13 in 6850.
14 This group addressed those changes in a relatively 15 lengthy response window and as you heard from my 16 colleagues as the pilots progress that lengthy 17 response window changed and became a little bit 18 shorter and also had quite a few additional 19 stakeholders. They included the NRC reviewers, 20 utility project managers, utility PRA personnel, 21 consultants, senior managers all became increasingly 22 involved in the identification and the resolution of the
30 1 frequently asked questions or the fire PRA methods 2 issues.
3 The evolution of the process significantly expanded 4 a number of people directly involved in the 5 development of consensus approaches and they added 6 an element to time pressure as well due to 7 the tight implementation schedules associated 8 with 805.
9 The group dynamics shifted to a higher stress 10 environment but more importantly than the higher 11 stress was that you had a larger group of 12 stakeholders with different and sometimes divergent 13 goals such as the need to preserve the ease of 14 review or schedule adherence with the desire to get 15 a guaranteed regulatory accepted method and others.
16 The group dynamics were an extremely inefficient use of 17 the resources accompanied with some unrealistic 18 expectations that these competing goals could be 19 met simultaneously.
20 And another common result that we had was the 21 compromise method which I'll talk a little bit on 22 the later slide.
31 1
Next slide, please.
2 A second lesson that we learned is that the method 3 shouldn't be developed in abstract and many of our 4 PRA issues came from a single plant.
5 When developing resolution to the issue inevitably 6 the question "but what if" arises.
7 If I could count the number of times we did the "but 8 what if" it would be a very large number.
9 The questions usually asked and we have a reasonable 10 solution to a problem and that's proposed and a 11 well-intentioned participant or stakeholder asks in 12 the interest of expanding the applicability of the 13 solution "but what if".
14 The goal was to be a little bit more comprehensive 15 which is an honorable goal.
16 But this is our classic snatching defeat from the 17 jaws of victory.
18 Where the "but what if" turns into a situation where 19 we address the significantly number of hypothetical 20 situations that may or may not have occurred in 21 experience and may or may not be applicable to the 22 next series of fire PRAs we do.
32 1
Next slide, please.
2 All these lessons are related and in less than three 3 the larger group of stakeholders with the best of 4 intentions try to ensure that their goals are met in 5 the resolution and what tends to happen is what I've 6 termed a compromise method.
7 Compromise methods are usually conservative and 8 bounding inputs because they're trying to bound the 9 goals of all the folks involved in developing that 10 solution.
11 They also have significant elements of the "but what 12 if" philosophy.
13 In the past after that conservative method is 14 developed and bounding inputs are applied then significant 15 resources are expended in the form of data 16 collection, engineering analysis to attempt to 17 increase the realism of the models and to date we've 18 had sort of a spectrum of success here and some of 19 the interim methods that are produced still don't 20 comport with data and experience.
21 Several of the interim solutions produced will 22 need to be revisited or refined as the they're
33 1 applied to the wider range of plants.
2 These are three lessons that I've listed, other 3 lessons could be provided in including some of those 4 that are very positive.
5 These lessons were chosen because these are 6 the lessons that we can get the most benefit at 7 improving the process so you heard about all the 8 process problems now we're focusing in on what are 9 the things we can do to improve.
10 There are a number of positive outcomes that weren't 11 discussed in this presentation and given the fact 12 that we probably will proceed with fire PRAs 13 whether or not they're for NFPA 805 or just for risk 14 informed performance-based regulation in other areas, 15 we probably will continue to develop these methods 16 so it's important that we address these lessons 17 learned and mature the methods as quickly as we can 18 In the area of the future of fire PRAs methods 19 we should use the best available information and 20 analysis.
21 And in that vein we should look at all the evidence 22 and analysis which would be judged on its merits and
34 1 included when it has reasonable basis and provides 2 the best information and we can monitor it.
3 The methods inputs and results should comport with existing 4 experience and current operating experience to provide assurance that the 5 model is a reasonably accurate representation of the as built, 6 as operated plant.
7 This in turn leads to increased acceptance of the model which is important if 8 we are going to be applying this in risk informed performance-based regulation.
9 So basically the previous history that we are aware of should comport well with 10 what we see today.
11 Conservative in the process should be limited to the 12 decision-making guidance and criteria. And the 13 methods and input should be sufficiently dynamic to 14 allow us to improve them over time and encourage good 15 behavior by accurately reflecting safety.
16 improvements in a corresponding risk reduction.
17 Next slide, please.
18 The fire PRA action plan is our solution to these 19 issues. The fire PRA action plan consists of a matrix of 20 fire PRA method issues including a characterization of the issue, the issue 21 owner, a schedule and resources. Along with the matrix a new approaches 22 is proposed which addresses lessons learned in an
35 1 attempt to improve the process.
2 So the new approach begins with small teams of 3 knowledgeable experts that develop a solution which is then 4 applied to several plants.
5 The team develops these solutions not necessarily 6 for the method itself to be approved but rather for 7 it to be approved in context in the application.
8 This isn't to say that the methods can't be reviewed 9 or not available for review or they are not stand alone 10 but this is to say that their best reviewed when 11 they're done in context rather than the abstract as 12 we've seen what can happen with the "but what ifs".
13 Reviews and refinements of the methods are 14 initially provided by the application plants 15 so when they are being applied we can look at that 16 application and refine the methods if necessary.
17 They're also then reviewed and input received via 18 independent reviews so there is an independent peer review 19 required. It's part of the fire PRA process.
20 Another opportunity for review and approval method 21 would be during a request for additional information 22 process and then finally the approval in context
36 1 would be provided in the NRC safety evaluation 2 report.
3 That provides sort of a high level outline of how we 4 could improve the process of addressing these 5 findings.
6 Next slide, please.
7 I don't intend to review this slide in any detail 8 but I did want to mention that these are the 9 activities that are listed on the current fire PRA 10 action matrix.
11 The items listed in red are those interim solutions 12 that will need additional analysis for complete 13 closure or to provide applicability to a wider range 14 of plants.
15 The remaining issues that are items that we have 16 identified but we don't yet have detailed approaches 17 for.
18 That completes my prepared remarks. Thank you.
19 CHAIRMAN JACZKO: Thank you, I appreciate the 20 comments. Im sure the Commissioners will have 21 lots of questions for you and the 22 others.
37 1
We'll turn now to our last speaker Mr. Gunter with 2 Beyond Nuclear.
3 MR. GUNTER: Thank you Commissioner.
4 My name is Paul Gunter Director of Oversight Projects for Beyond 5 Nuclear and we really appreciate the opportunity to participate as 6 public stakeholders.
7 The Browns Ferry fire in 1975 demonstrated as 8 reality that a significant fire can occur at a 9 nuclear power station and that a fire can 10 significantly challenge the safe shut down 11 capability of the reactor.
12 The Browns Ferry fire further demonstrated that even an 13 incalculably improbable source of ignition can lead 14 to a significant fire in reactors operating today.
15 The near catastrophic experience proved so harrowing 16 that the Nuclear Regulatory Commission responded by 17 dramatically amending and expanding its fire 18 protection philosophy to include the development of 19 General Design Criteria 3, Branch Technical Position 20 9.5.1 and the promulgation of law under Code of 21 Federal Regulation for minimum fire protection 22 requirements to conservatively ensure that a level
38 1 of compliance exists at all nuclear power plants.
2 Unfortunately, as witnessed through my personal 3 experience since 1991 before the Commission, one critical 4 analyzed area of these fire safety requirements in 5 nuclear power plants was not properly implemented 6 nor subsequently enforced.
7 Namely for the large number of Appendix R III G.2 8 fire areas requiring qualified physical and passive fire 9 protection features for control room power, control 10 and implementation electrical circuits to reasonably 11 assure that the redundancy for reactor safe shutdown 12 equipment cannot be destroyed by a single fire.
13 Apparently after 29 years of effort, such regulatory 14 assurance appears to be overly burdensome and no 15 longer considered reasonable, attainable by industry 16 nor enforceable by the Federal agency without a 17 large number of exemptions.
18 Given the widespread level and duration of 19 non-compliance, the infrequency of serious fires at 20 nuclear power plants is at the same time a blessing 21 and a curse.
22 A blessing in that, to date, more significant fires
39 1 have not challenged nuclear power stations safe 2 shutdown operations, a curse in that the lack of 3 such experience leaves many broad areas of 4 uncertainty in an aging industry.
5 The expanse of this uncertainty includes not only a 6 lack of experiential knowledge base but introduces 7 questions and disputes involving variability, 8 randomness, indeterminacy, judgment, approximation, 9 linguistic imprecision, error, and the unreliability 10 of human behavior and the significance surrounding 11 fire safety issues.
12 These broad uncertainties play a major role in our 13 discussion and our concerns today regarding the 14 public's confidence in the proposed transition from 15 the ongoing failure to achieve compliance with a 16 prescriptive fire code to the optional NFPA 805 17 Performance-based Standard for Fire Protection for 18 Light Water Reactor Electric Generating Plants.
19 Because of these uncertainties we remain skeptical 20 of the outcome of the NFPA 805 transition and the 21 implementation process.
22 I would like to focus my presentation on the issue
40 1 of fire modeling.
2 Verified and validated fire models used to predict 3 the extent of fire damage from a range of fire 4 sources are held up as an integral, indeed essential 5 part of the transition to NFPA 805 in determining 6 the survivability of reactor safe shutdown equipment 7 in lieu of protecting that same equipment through 8 compliance with Appendix R III.G.2 through the qualified 9 fire barriers.
10 Given the potential high safety consequence arising 11 from a fire that knocks out the control room 12 operation and maintenance of reactor safe shut down, 13 accurately capturing all of the proper fire 14 scenarios becomes crucial to public safety.
15 We argue that fire modeling remains a significant 16 limitation to NFPA 805 and fire safety analysis and 17 design for power reactors.
18 Published literature continues to warn that fire 19 modeling is still in its developmental stages with 20 its associated uncertainties.
21 In our view, that remains a significant stumbling 22 block to a reasonableness assurance and a continued
41 1 impasse to effective enforcement policy for future fire 2 safety issues arising in NFPA 805 nuclear power 3 plants.
4 It remains very difficult to employ a computer 5 generated fire model with a high level of confidence 6 so that it makes a valuable contribution to 7 real-world decision-making as opposed to leading to 8 inaccurate and inappropriate interpretations that 9 can leave power reactors vulnerable to fire.
10 The European experience in fire modeling further 11 suggests that different fire model users can produce 12 very different results even when using the same 13 probabilistic model and applying it in the same 14 case where risk estimates 15 can differ by several orders of 16 magnitude and our crucially based on the users' 17 knowledge and experience or lack thereof.
18 A number of identified error sources and gray areas 19 in fire modeling include: the lack of reality of the 20 theoretical and numerical assumptions used in fire 21 models. The assumptions used 22 in field models are approximations
42 1 to real-world experience from a 2 particular fire; lack of fidelity of various 3 numerical solution procedures; direct errors 4 in computer software where the software will not be 5 an accurate representation of the model and numerical 6 solutions procedures; faults in computer hardware 7 where a fault can exist as a result of the mistakes 8 in microprocessors; significant and undetected 9 mistakes in fire model applications while inputting 10 into the model.
11 These potential error sources can remain significant 12 challenges to both industry and regulator that 13 cloud, complicate and further prolong the 14 development of a fire safety resolution path and 15 improved enforcement policy.
16 Given the troubled history of NRC's official policy 17 of non-enforcement which spans decades old fire 18 protection violations it begs the question if a 19 transition to NFPA 805 helps or further hinders the 20 institution of NRC enforcement policy on fire 21 protection.
22 The failure of the NRC to effectively take
43 1 enforcement action on the violation of inspectable 2 prescriptive requirements, widespread abandonment of 3 subsequent corrective action programs and failure to 4 follow through with fire safety confirmatory action 5 orders does not lend to building public confidence 6 that the agency can effectively address violations 7 of an arguably more nebulous and difficult to 8 inspect performance-based standard, potentially 9 involving disputes between staff, industry and 10 public over any number of areas of uncertainty 11 identified.
12 Finally, there is the concern that malevolent acts 13 are beyond the scope of NFPA 805.
14 The risks and consequences associated with sabotage 15 cannot be accurately analyzed by probabilities nor 16 can they be modeled.
17 As we have raised to the staff, we see a significant 18 fire safety disconnect in a shift to 19 performance-based risk informed fire protection 20 regulation that does not address security concerns 21 when coupled with ongoing industry wide noncompliance 22 with prescriptive fire codes for Appendix R
44 1 III.G.2 fire areas.
2 The same nuclear power stations have long been 3 identified by national laboratory study to have been 4 inadequately evaluated in their design and 5 construction for the effects of explosion and fire 6 resulting from the impact of aircraft.
7 These same nuclear power stations have been further 8 exempted from any further mandatory aircraft impact 9 hazards analysis.
10 The security veil then falls to obscure from public 11 view how the risks of deliberate destruction of 12 reactor safety systems by fire are or are not being 13 addressed.
14 As a result the question remains in the public 15 interest community, is the Federal regulator 16 pursuing a compliance strategy to douse the flames 17 of a fire protection controversy or is it at long 18 last prioritizing the establishment and enforcement 19 of fire safety regulations to maximize public safety 20 margins during post-fire safe shutdown.
21 CHAIRMAN JACZKO: Thank you Mr. Gunter. I 22 appreciate your insights and we will now begin our
45 1 questioning with Commissioner Klein.
2 COMMISSIONER KLEIN: Thanks.
3 I will start in the order of the presenters so Joe 4 I'll start with you.
5 What were your most significant plant modifications 6 in Shearon Harris?
7 MR. DONAHUE: I think I would probably 8 categorize them in three areas.
9 One is we were able to by actual fire testing on 10 the fire barriers to determine for the exact plant 11 conditions and the exact configurations at Harris 12 what the barriers and how long that plant 13 would last and maintain the barrier. Then we have gone 14 in and made appropriate modifications and then took 15 credit for that real-life, real testing 16 conditions in the fire PRA models, so quite a few of our 17 models dealt with that.
18 Second, I believe the incipient fire detectors in 19 our cabinets will bring on the onset of any 20 instrumentation and control potential of fires and 21 no matter how we end up in the conservatisms and 22 realism of fire promulgation in the cabinets we will
46 1 have implemented those and I think third, probably the one 2 that I am most proud of with the staff, is the diesel 3 generator backup steel injection system which 4 addresses internal events safety while at the same 5 time fire safety for giving makeup capability during 6 a shutdown.
7 COMMISSIONER KLEIN: Could you tell me 8 again what your cost estimates were when you started 9 and what you've spent to date?
10 MR. DONAHUE: For the fleet for all four of 11 our plants with modifications was going to be near 12 $40 million and right now were estimating about 13 $82 million.
14 COMMISSIONER KLEIN: So about doubled from 15 what you said.
16 When you look at a specific plant and you look at 17 what you spent are you able to quantify the risk 18 reduction?
19 MR. DONAHUE: In the case of Harris which 20 we've got the peer review fire model the internal 21 events CDF has gone down by 20% predominately by the 22 seal injection fire. Our other units,
47 1 the expected modifications were seeing, I would 2 call them more compliance related, we do 3 not have an actual number for our next plant which 4 is further along is CR3. But CR3 with its additional diesel generator 5 capability that it has is a relatively low profile 6 plant from a risk perspective so I don't expect to 7 see the gain that we got at the Harris project.
8 COMMISSIONER KLEIN: And what is the status 9 of the piloting the inspection process, where do you 10 and the staff currently lie?
11 MR. DONAHUE: I think that's probably a 12 question to best -- better ask the staff -- I do not 13 believe we have seen the draft inspection 14 guidelines.
15 I know there's been discussions going on with the 16 staff and the Regions and I think again that's where 17 as we get the SER we plan to deploy and get 18 into the implementation of 805 and welcome 19 piloting that part of the process, that's the third 20 leg of the stool that needs to be done successfully 21 for the rest of the industry and obviously for 22 ourselves.
48 1 COMMISSIONER KLEIN: So you haven't seen a 2 draft plan or anything?
3 MR. DONAHUE: I do not believe we've seen a 4 draft to date.
5 COMMISSIONER KLEIN: Dan you talked a little 6 bit about the different types of plants that you had, some 7 that were in the seventies and some in the 8 eighties, fundamentally is there a difference in 9 which ones might lend themselves more or less to NFPA 10 805 compared to a deterministic approach.
11 MR. PACE: There is a difference in how 12 much remediation needs to be done to solve the 13 original issues, those circuit analysis in the 14 manual actions required to comply with redundant safe 15 shutdown cabling area is differing on a mid-16 seventies plant than it is on a late eighties plant 17 and I believe it's different between some of the 18 design AEs, NSSS firms.
19 I have a great deal of experience on the later 20 boilers and they're quite well-designed on 21 separation and require minimal impact to comply 22 with almost any reasonable deterministic role
49 1 on circuit analysis and have little or no manual actions.
2 Where as some of the PWRs in the mid-seventies.
3 they required significant manual actions and had 4 more circuit interaction issues.
5 Given that you can resolve some of these ambiguities in 6 these inputs and propagation parameters on a fire PRA and 7 make it a practical tool to use, it obviously would 8 lend itself better to a mid-seventies plant than a 9 more modern plant.
10 COMMISSIONER KLEIN: So it looks like you 11 started about four years to go to transition to NFPA 12 805, would your decision process be different 13 today versus four years ago?
14 MR. PACE: I have that decision to face.
15 We've got a basic fire PRA model on our Davis-Besse 16 unit, it would be what we envision a fire PRA model 17 to be four years ago and I get to watch and see how 18 the pilot comes down and what impact that has on my 19 Beaver Valley plant before I embark on Davis-Besse.
20 If we get pragmatic solutions that are comparable to 21 power PRAs out of the firm PRA modeling efforts 22 that are going on now then well likely proceed
50 1 with the fire PRA model at the other units as well 2 as other risk informed applications.
3 If we continue to have unrealistic outcomes to this 4 fire PRA modeling exercise then we will have to 5 reevaluate that.
6 COMMISSIONER KLEIN: If you look at the 7 two pilots being finished, and one earlier in 2010 8 and then another one may be in the spring of 2010, 9 how much time do you think it would really take 10 utilities once you have that better determination to 11 really do a NFPA 805 decision?
12 MR. PACE: Well if we were truly operating 13 pilots like we do processes and procedures 14 at our fleet plants we let the pilot run its course and 15 determine the lessons learned and then proceed with 16 certainty on our other plants.
17 So if this had progressed according to plan the 18 pilot would've completed in '07-'08 and then we had 19 a two to three year implementation because it takes 20 two to three years to implement the fire modeling on 21 a plant.
22 Where we are today with a plant like Beaver Valley,
51 1 once some certainty is brought to the rules, were probably 2 10 to 12 months from having a completed product.
3 COMMISSIONER KLEIN: You had talked about 4 your cost differences have increased from what you 5 initially expected to what they are now.
6 What would you attribute to primary cost escalation?
7 MR. PACE: Primarily in recycling the model, we had a 8 as the architects like say a substantially complete 9 model two years ago that we've re-performed with 10 different inputs for initiation frequency, re-preformed 11 with different propagation parameters, re-performed 12 for different fire areas for the last 24 months to the point where on 13 Unit Two and on Besse we have just stopped until some of the uncertainty 14 settles out on these decisions.
15 COMMISSIONER KLEIN: Well, Ken, you listed 16 on your slide eight several areas that needed 17 additional analysis and development.
18 What's EPRI doing for those?
19 MR. CANAVAN: As I mentioned during the 20 presentation and I can expand on it a little bit 21 here.
22 Those are the items that are in the fire action
52 1 matrix. Currently on each one of these items a plan is being 2 developed that includes the owner of the issue which 3 may not be EPRI it may get assigned to owners groups, consultants, 4 vendors. That issue will then be worked by a small 5 team.
6 In the matrix, it's getting a detailed action plan with resources 7 and schedules for completion and those are going to be 8 realistic schedules on what we think we can do given the 9 resources and the ability of folks to work on those 10 issues given that the same people who support 11 implementation and development of the fire PRAs are the ones 12 who also do a lot of the methods development..So each one of these issues 13 will then be worked in accordance to that plan.
14 There may be a few that may be added as a result of 15 follow-ups plans as well.
16 COMMISSIONER KLEIN: So looks like you had 17 eight that were red in color?
18 MR. CANAVAN: Yes.
19 COMMISSIONER KLEIN: About how much is EPRI 20 spending a year to address these?
21 MR. CANAVAN: Costs vary, but we been about 22 in the million and a half to 2 million range over
53 1 the last couple of years and we expect in the future 2 to get significantly larger.
3 COMMISSIONER KLEIN: And of those eight 4 which is the top two?
5 MR. PACE: The top two are, actually Joe 6 had mentioned them in his presentation, the first 7 one is heat release rates and the second would be 8 the experiential data getting all the fire events database, 9 collecting and analyzing that data.
10 COMMISSIONER KLEIN: Thanks. Paul you 11 indicated a blessing and a curse regarding the 12 fires.
13 Do you think that the attention the industry and 14 the NRC is paying to the fire that that might be one of the reasons 15 that there has been a lack of fires, like a Browns Ferry?
16 MR. GUNTER: Certainly, I think that it's 17 clear that because of the risk and consequence of fires 18 it is a chief concern in the industry and should be for the 19 regulator as well.
20 So I do understand that and I've toured power plants 21 as well to the extent to see the introduction of 22 combustibles, for example, is attended to in great
54 1 detail.
2 But nevertheless as I said history has demonstrated 3 that these plants are vulnerable and its the curse of 4 course that we're worried about because the 5 consequences could be unacceptable.
6 COMMISSIONER KLEIN: Thank you. No further questions.
7 CHAIRMAN JACZKO: Thank you. Commissioner 8 Svinicki.
9 COMMISSIONER SVINICKI: Thank you. Thank you for your 10 presentations. Mr. Donahue, Im going to return to a topic that you raised that 11 I've given much thought to.
12 There's plenty to think about right now in terms of 13 the pilots but you were kind of looking over your 14 horizon, which I appreciate. You said something along these lines, 15 after driving through the review of the LAR a next area to focus on would be the 16 inspection process and you talked a little bit about that. You talked about 17 minimizing subjectivity. Can you help me understand with a little greater 18 specificity? Why is it that at post transition to NFPA 805 it would be more 19 challenging for an inspector, is it just simply because a deterministic approach 20 is I have a set of very prescriptive things that I inspect to 21 versus the more nuanced understanding that's 22 necessary under the risk informed PRA process is
55 1 that kind of the kernel of the issue there?
2 MR. DONAHUE: Yes that's the kernel if you think of one in the 3 deterministic being very rule based, and in the case of the fire PRA when you 4 get into 6850 you are going to get into what assumptions were made for 5 the fire propagation, what assumptions were made for that particular cabinet, 6 that particular raceway and then you can get into 7 discussions of interpretations of a very prescriptive document 6850 how 8 was that applied in the computer modeling. So it adds a 9 different level of subjectivity.
10 Then also part of the 805 process is the engineering 11 evaluation change process which allows us going 12 forward in the future to make changes to the plant, 13 use the risk models and say that this particular 14 change we made, meets a threshold 15 that we can go ahead and move forward 16 and make the change without NRC approval.
17 That could come under scrutiny several years later 18 on a tri-annual when they come back in and said we 19 believe you should not have made that change and we 20 don't believe in the basis. So it adds a little bit of subjectivity 21 that we have to work through and to be able to have some certainty as to 22 how that works and we have to get in some real-life examples of
56 1 changes because I can think of quite a few changes which 2 absolutely dont have any impact. It's going to be can you find 3 some of those in the real-life world that are on that 4 right at the threshold and how do we communicate 5 those rather than wait three years.
6 COMMISSIONER SVINICKI: And then hearing 7 that, Mr. Gunter would you say that looking at that 8 same issue from a slightly different angle, that the 9 scrutability of this in all of these determinations and 10 interpretation that needs to be done when we've 11 moved away from the deterministic approach that it 12 may make it more challenging for public transparency 13 in the scrutability of looking into that 14 implementation and inspection process, is that 15 accurate or something that you think from the 16 public's perspective of kind of understanding this 17 transition away from Appendix R prescriptive 18 requirements, is that at least one aspect of the 19 challenge for the public?
20 MR. GUNTER: It certainly is, but I would also raise that issue 21 for inspectors of the agency itself.
22 There is this concern that were adding a new wing to
57 1 the labyrinth here.
2 COMMISSIONER SVINICKI: Thank you at least 3 two of you, I think, refer to the frequently asked 4 question process and it is not unusual for the staff 5 to use that when working through a development process on 6 construction inspection. I know that we have a very 7 fulsome FAQ process but it came up here and it sounded 8 like it's been a little bit challenging but also a 9 way maybe to expose issues and then move forward to 10 give, I think Mr. Donohue, you mentioned you've got 11 to get some sense of understanding of what the 12 staff's interpretation or maybe it was Mr. Pace, you said 13 acceptable interpretations from the staff is there 14 anything again since a couple of you mentioned the 15 frequently asked questions process.
16 How is that benefited in NFPA 805 and is there 17 anything about it that's been challenging or do you 18 say it's kind of the use of the FAQ is an absolute 19 good or has it been challenging here at all?
20 MR. DONAHUE: I'll start again, and again it 21 picks up with what my colleague Ken mentioned.
22 Being the lead plant many of the FAQs in the aid are a direct offshoot of
58 1 the particular circumstances we have Harris and Oconee, 2 so we were able to ask very specific questions of some of the FAQs say around.
3 the fire initiation and cabinets, which are very particular around our case.
4 Some of that could be research-based, some FAQs are 5 more interpretation of words written down in draft 6 or regulation or the NEI document, those we more or 7 less worked through, some are around, theres a particular 8 number in 6850 that was created as Mr. Canavan 9 mentioned then how does it apply to this particular 10 case.
11 I think we're a little bit slow as a overall body of 12 the staff working with the industry and addressing 13 some of those ones which may have been a little bit 14 more research-based or number based or realistic 15 based.
16 COMMISSIONER SVINICKI: Is it also true 17 that for the follow on transition plants, the FAQs 18 may or may not be a very good relevance to their 19 circumstance?
20 Were they truly tailored to the pilots?
21 MR. PACE: Thats my understanding, but Mr. Canavan 22 has a much more crisp view of the FAQ process
59 1 MR. CANAVAN: I think that I can offer two perspectives.
2 One is that the FAQs that I am familiar with from web sites and other locations 3 are traditionally answered in a paragraph or two not 16 or 25 or 50 page 4 dissertation or report that contains additional research, interpretations --
5 COMMISSIONER SVINICKI: Is that what is happening here?
6 MR. CANAVAN: The Frequently Asked Questions range from a 7 paragraph response on a single interpretation, where it would appear that an 8 FAQ is very appropriate all the way to the point 9 where they are indeed research activities 10 that in my opinion don't belong in a FAQ process.
11 And as for culpability, you'll find 12 that probably 80% or maybe higher are applicable to 13 all the plants as we move forward especially to the 14 smaller interpretation ones, but the 10% or 20% that are 15 not applicable are usually the lengthier 16 dissertations on them.
17 COMMISSIONER SVINICKI: Again I'd know it's 18 a little bit of a side but I'm always curious if 19 we're looking at the best ways to communicate as a 20 agency and FAQs is something that I've been hearing 21 about lately so I just wanted to pull that thread a 22 little bit.
60 1
Mr. Pace, it's interesting because on your slide 2 three you talked about your decision framework for 3 transitioning to NFPA 805, I took from that you list 4 four factors but something that appeared to be 5 important was the risk informed enhancements to 6 safety and a standard approach across your fleet.
7 If we get to a point where some who have indicated 8 their intent to transition might be looking at that 9 and I guess I am now getting you to your slide 11 10 where you talk about some of the original 11 assumptions might have - what you assumed there might not be 12 valid, and not just speaking for yourself but maybe 13 for the larger wave of folks who might transition to 14 NFPA 805, they may relook at their intent to do 15 that. You made a statement and I'll paraphrase 16 although I tried to write it down, "It may have been 17 better to put 15 to 18 million into solving some of 18 the analysis questions and then making the mods to 19 comply", you also talked about or maybe it was 20 Mr. Canavan that a number of applicants -- there will 21 be use of fire PRAs whether or not folks are 22 transitioning to NFPA 805 so I guess from where we sit now
61 1 and the pilots aren't done, if a number of folks 2 are relooking or do end up looking at their 3 intention to transition but yet they do a fire PRA and 4 fire modeling and they make modifications I guess 5 I'm not sure in terms of kind of the overall 6 enhancement to fire safety or consistency of 7 approach, what's the difference for an individual applicant?
8 Would you make different physical modifications if 9 you are transitioning to NFPA 805 versus just doing 10 a fire PRA in making some physical mods and how 11 would they differ?
12 MR. PACE: Well, we'll never know maybe, going back on some of 13 these plants --
14 COMMISSIONER SVINICKI: Thats way Im asking you to speculate.
15 MR. PACE: One might say that on a later model BWR, like I am familiar 16 with, we may do the circuit evaluations that are now out in the NEI documents 17 and the NRC research documents that say here are the rules of 18 engagement for circuit analysis, which also has to be 19 applied to the PRA modeling by the way, and say I can 20 solve these four circuit issues by completely 21 re-routing these circuits out of the building even.
22 That may cost me $2 million.
62 1
And now I'm in full compliance with the Appendix R under 2 the deterministic rules of engagement and I don't 3 have too, I haven't spent the other 17 on fire 4 modeling.
5 COMMISSIONER SVINICKI: Okay, Okay.
6 MR. PACE: Some licensees are committing and 7 building fire models but not engaging in the fire 8 modeling dialogue that's going on with the 805 9 transition. They are building fire models more along the 10 lines that the power models were built. Once again 11 Mr. Canavan can probably tell you some of the nuances and 12 differences in those techniques and approaches.
13 COMMISSIONER SVINICKI: You're reminding me 14 in the response to Dr. Klein's question that you did 15 indicate a lot of your investment had gone in the 16 recycling and the iterating on some of the modeling 17 and analysis so its sounding like, although I agree 18 with you we'll never know and I'm asking a strange 19 question here, but it may be that under that 20 framework of not transitioning to NFPA 805 an applicant would have 21 less investment, or licensee would have less investment, in the 22 analysis and the modeling and maybe more in mods to
63 1 meet the deterministic requirements. Thats what might look different about it..
2 I would say that you said one of the benefits for 3 you was a fleet wide approach and I would say it 4 would seem like you'd lose that particularly if you 5 transitioned part of a fleet and not the rest of 6 it.
7 MR. PACE: We may very well, hypothetical, we 8 very well complete the Beaver Valley modeling exercise 9 including NRC reviews and have invested $20 million 10 in a very sophisticated fire PRA model and eliminate 11 the proponents of the manual actions and evaluate 12 all the circuits and no modifications be required.
13 Success, right?
14 I haven't modified the plant so have I fundamentally 15 changed the plants ability to respond to a fire 16 event?
17 No.
18 What I have done is vindicated those that argued 19 that it was safe under the previous rules of 20 engagement by absolutely proving that through 21 analytical methods using agreed-upon rules of 22 engagement that these were not realistic scenarios
64 1 in the beginning.
2 Does that make sense?
3 COMMISSIONER SVINICKI: Yes, again you are 4 indulging me. Im peering a little bit into hypothetical decision frameworks for 5 various folks that might again take the whatever 6 comes out the pilot and relook at whatever they had 7 planed before.
8 MR. PACE: I'm familiar with this I've done 9 this exercise. So the deterministic approach would be 10 I don't agree with you that the circuit would get 11 involved in flames in this room, I don't believe 12 that there's a fire credible there and even if it 13 was it wouldn't propagate to the cable on this side 14 of the room. But under the old rules I assume the 15 room burns up and I will just move the cable, 16 right?
17 You win.
18 I'll move the cable and I'll spend the money on the mod 19 necessary to move the cable and I don't have to 20 argue about whether it's probable that it will be consumed in 21 fire or not.
22 That's kind of the point I was making.
65 1
So, the ones not committing to convert are 2 obviously believing they can take that approach and 3 I believe there are some plants, to Commissioner's 4 Klein question, that are much more capable of taking 5 that approach than others.
6 A late model boiling water reactor, which I'm familiar with could likely 7 take that approach for very little money.
8 COMMISSIONER SVINICKI: Thank you. Thank you Mr. Chairman.
9 CHAIRMAN JACZKO: I think there's a very 10 good discussion on perhaps provide a bit of context 11 for discussion before I ask questions.
12 I think it's important to remember I think -- I 13 don't have the 29 experience, I have four years of 14 experience -- in dealing with these issues and we do 15 seem to have a challenge with resolving issues.
16 If I would walk away, I would walk away thinking 17 that it's been extremely trivial to resolve the deterministic 18 challenges that really motivated NFPA 805.
19 I don't think that's the case, if that were the case 20 we should've been done and we shouldn't have to 21 worry about where we are.
22 We have the operator manual actions which were never
66 1 approved, never authorized, that have existed for a 2 long time and perhaps Mr. Gunter you would say 3 that's an enforcement challenge.
4 I may not disagree with that but that's a little bit of 5 the background so the approach that was taken was 6 for the industry to develop a performance-based risk 7 informed standard in NFPA, hat stands for the National Fire Protection 8 Association. That was an industry developed standard, an 9 industry developed approach.
10 Then the NRC then when it adopted that and on a 11 voluntary basis allowing the industry to decide and 12 choose whether or not that was the path that they wanted 13 to follow.
14 Following that, ultimately the PRA standards came out 15 and an ANS/ASME/joint industry standard was developed for the PRA quality.
16 EPRI working with NRC developed 6850.
17 The bulk of the work and the bulk of the concerns of 18 what I'm hearing about is in the PRA work in the fire modeling by and 19 large most of which is developed and worked through 20 by the industry.
21 So the frustration that I have and perhaps the staff 22 has and maybe we'll hear about it, is that we have
67 1 now concerns and complaints about those issues which are fair 2 and thats not to say that is not the situation we want to be in, Mr. Pace as you 3 indicated we wanted to be done with this a lot sooner. I want to be done 4 a lot sooner. Im sure Mr. Gunter would prefer that it were done a lot sooner, or 5 perhaps not at all. I should let you all speak for yourselves. But I think the point 6 is the challenges that exist in PRA, the new ANS/ASME standard requires 7 fire PRA as an initiating event. That is something that if the industry wants use 8 PRA models it's going to have to incorporate and 9 include.
10 So some of these challenges are challenges we are going to have 11 to address one way or another.
12 NFPA 805 has helped us because it has brought these 13 issues to the forefront and allowed us to address 14 them and allowed us to make what I think actually is 15 progress.
16 I do fundamentally believe that this is the right 17 route because it makes the plant safer and I think 18 we haven't necessarily talked about that but I think 19 Mr. Pace you hinted at that.
20 It gives us a better understanding of we right now 21 don't have as good of a tool to say if a fire 22 barrier comes in and as we've learned with Hemyc in
68 1 some cases in some applications it cannot meet that one 2 hour2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> rating. We don't have a good analytical method 3 right now to grant an exemption for that.
4 We do it as you said Mr. Pace we look at it 5 we say think that there are metal combustibles in 6 the room so we think that it may be acceptable to do 7 that.
8 Thats not, I think the best basis, I think the 9 analytical tools within NFPA 805 give us the 10 ability to say that this is a better, much better basis.
11 That having been said I think it's a little bit -- I 12 think these are very good points and it's 13 important to understand where we are in the 14 challenges we have to moving forward and I think 15 it's important to understand that the other 16 path was not as simple perhaps for many of these plants, and 17 so unfortunately we find ourselves in a difficult 18 situation and trying to work through what is 19 ultimately an important safety issue but also a 20 difficult one to resolve.
21 Having said that I thought I would explore a little 22 but the issues that we're dealing with in the
69 1 uncertainties in the modeling and perhaps trying to 2 better characterize where the models are. I think a 3 big philosophy that the NRC has incorporated with 4 its use of PRA models is that they need to be sophisticated 5 enough, and sufficient fidelity for the application at hand.
6 Would you say right now that the models don't meet 7 that threshold and that standard that the 8 uncertainties are large to the point that they don't 9 allow use of the models in this particular 10 application or anyone who wanted to comment on this 11 point.
12 MR. CANAVAN: I don't think it's so much 13 the uncertainties.
14 You brought up a lot of different questions. I 15 guess I think that our challenge is right now we 16 know the uncertainties are bounded, if you were to ask 17 me, or significantly bounded.
18 CHAIRMAN JACZKO: So we dont have a concern 19 that there is under conservatism in the modeling, it a concern that there is too 20 much of a conservatism.
21 MR. CANAVAN: That is the natural 22 progression of the process so when the 6850 and EPRI authors
70 1 developed the methodology when they hit a tough 2 issue, one that was difficult to get data or 3 difficult to analyze, the natural tendency of those 4 developers was to say let's use this number because 5 it bounds everything and when the pilots are done 6 we'll refine that number and we'll do that analysis.
7 And that's where we are now. We/re at the tough spots 8 of doing that analysis.
9 CHAIRMAN JACZKO: So you wouldn't say that there's anything - I 10 wouldnt want people to walk away and think, and Mr. Gunther perhaps you 11 would want to comment on this too, that the models wouldnt provide for 12 adequate safety in this case.
13 MR. CANAVAN: The current set of models generally over predict when 14 compared to experience. So if we do a model of a turbine building, for 15 example, we do a very detailed model of the turbine building. We will find that 16 the numbers produced are something on the order of several turbine building 17 fires per year when we do the math of multiplying the number of plants. So 18 simple math tells us our models over predict, which was a great first start. Now 19 we need to refine those models to have them comport with experience because 20 in order to have trust in the model to use it for decision making we should at 21 least be able to reflect our previous experience at the appropriate 22 levels.
71 1
So this is not to preclude a large fire from occurring or propagating its way 2 through the model but it's to say that for the smaller events they're accurately 3 reflected in a reasonable range of their occurrence and experience.
4 CHAIRMAN JACZKO: But right now its not 5 under conservative in any way --
6 MR. CANAVAN -- not in any area that I'm 7 aware of.
8 MR. DONAHUE: Let me talk a little bit 9 about Harris if I might and I absolutely agree with 10 you.
11 If I take a look at the areas that are ire PRA 12 model when we go to a level of monitoring every 13 single -- the majority of the cables -- and that's how we 14 got to 4000 scenarios, those scenarios did bring 15 focus on three or four areas that my gut as an 16 operator and a licensed operator would say those are 17 interesting areas so they called those out.
18 How we deal with those -- the one area that seemed a 19 tad over conservative for us was the propagation of 20 fires in low-voltage I&C cabinets which are adjacent say 21 to our control room which our operators are right 22 there.
72 1
Really where the discussion gets into is how much 2 credit, how fast can an operator open the door and go 3 see if there's a fire.
4 So we're really around how quick mitigation strategy 5 and does the fire happen that the cabinet just 6 burns up in a minute or is it going to get other 7 indications that you can put an operator in there to 8 get it. So it's really around some of the mitigation 9 strategies and really around, Ill call in our case, 10 cabinets.
11 The other areas, they were of interesting areas, we are 12 addressing those areas directly. What happens is in 13 the order of magnitude in CVF risk which may not, 14 there's where can't add fire plus internals 15 directly.
16 The numbers are both acceptable but I truly believe 17 the internal events number with the years behind it 18 is a very valid number and I think we're 19 over predicting it to add in fire, so when you add them up 20 if you think of these as hazard groups which is 21 really the way the code does, you're almost giving 22 too much credit for the hazard group of fire
73 1 compared to other hazards.
2 I think the maturity will come there as we address 3 some of these issues and we've addressed them by our 4 modification strategies that we have done at our 5 particular plant but they are not under estimating 6 based on the knowledge, nothing overly surprised me.
7 CHAIRMAN JACZKO: That's good to hear and I 8 guess I would take your comment to mean giving a 9 choice having to make a modification that would 10 effect an internal event, flooding hazard perhaps, or something like that 11 versus a fire event, the models right now wouldn't 12 necessarily tell you which one is where you would 13 spend your dollars.
14 But within the fire realm it gives you a relative 15 ranking among the cabinets versus the fire wrap 16 somewhere or something like that.
17 Mr. GUNTER: Obviously, thank you, one of 18 the concerns that we have is, how do you model 19 accurately reliable human behavior?
20 And that seems to be certainly one area of 21 uncertainty that I don't think you can 22 underestimate, how people can just simply by error
74 1 or by ignorance influence a consequence. I mean I 2 would just challenge that I think that it's a 3 mistake to think that we're -- that these models can 4 be overly conservative when you begin to introduce 5 things like human behavior into it.
6 CHAIRMAN JACZKO: I have a person on my staff who 7 had some experience in the area of human 8 reliability and the work that's done in the models.
9 It's certainly a very fair point.
10 I don't know if any of the -- Mr. Canavan -- want to 11 comment on that or anyone on how those issues are 12 incorporated into the PRA models and how that aspect 13 is addressed.
14 MR. CANAVAN: I don't think I disagree that 15 it's a source of uncertainty.
16 I think I would disagree that you can't be over 17 conservative in characterizing that performance.
18 The fire scenarios are trained on, they're 19 rehearsed, they're practiced, there is reasonable 20 assurance that they can actually be performed, that there 21 were walkthroughs, it's proceduralized.
22 This is not just someone deciding one day this is
75 1 how they are going to react.
2 These are actions that are practiced, trained, 3 proceduralized. A note of caution about 4 conservatism is your ability -- if you're conservative 5 in one part of the model your ability to rank within 6 the hazard group or your 7 ability to say that these are my fire issues 8 one through ten can be limited. As a matter of 9 fact one of the outcomes of non-PRA pilots that 10 they're finding out now is that the staff at the 11 plants intuitively know the fire areas that are of 12 concern because they walk and they see all the cables.
13 They say this is the place that we are concerned 14 about.
15 Then they go off and do a Fire PRA in accordance 16 with the rule set that they currently have and that 17 areas ranked number five.
18 So there's immediate -- this PRA can't be accurate.
19 What we need to do is if that area is suppose to be 20 number one and its number five because we haven't 21 appropriately credited things or we haven't modeled 22 them correctly, we need to fix the model.
76 1 CHAIRMAN JACZKO: I appreciate that and I 2 think this has been a very good discussion and I 3 appreciate you all coming in and sharing your 4 experiences. What we want to have is a 5 process that works and in the end we'll hear from 6 the staff and I'm sure they have some insight to 7 share on their take and how we're going to come to 8 resolution on these issues and this meeting has 9 focused on the NFPA 805 but I think as I said at 10 the outset we are providing alternatives and 11 solution paths in some of these other areas but my 12 experience in the fire area has been that new issues 13 crop up all the time and I think the NFPA 805 14 provides a framework to deal with those new issues 15 in a way that one off approach doesn't necessarily 16 satisfy in the deterministic areas. I think 17 that's one of the added benefits as we go 18 forward.
19 I appreciate your being here and would continue to, those who are 20 transitioning, encourage you to continue to transition and those that are 21 thinking about it, encourage you to continue to do that and continue to share 22 your experiences so we can make the process better.
77 1
Thank you.
2 3 CHAIRMAN JACZKO: I think we had very good 4 input and insight from our stakeholders. Bill I 5 will turn it over to you to start the staff 6 presentation.
7 MR. BORCHARDT: Thank you, Chairman, good morning.
8 The fire at Brown's Ferry over thirty years ago 9 was a pivotal event.
10 We heard about it earlier in the presentation this 11 morning. It brought about some fundamental changes to 12 the way the NRC regulated and the industry worked to 13 address the fire issues.
14 In 1980, the NRC issued Appendix R that back-fit 15 certain requirements. Then in addition the rule 16 provided an exemption process that allowed licensees 17 to seek alternatives to meeting the regulations.
18 That set the stage for creating a wide range of 19 approaches and licensing bases throughout the fleet.
20 As a result of those numerous guidelines, back-fit 21 of regulations and the varying dates of plant 22 operation, each plant licensing basis is unique with
78 1 multiple supporting documents and this can make all 2 of our jobs more challenging.
3 In the 1980s, NFPA initiated work to develop the 4 national consensus for performance-based fire 5 regulation and in 2004 the NRC issued the 6 regulations which was an alternative approach to 7 Appendix R.
8 Addressing fire protection has been a long, 9 difficult, and resource intensive activity for both 10 the industry and the NRC.
11 Progress has been made though to improve fire 12 protection and to enhance knowledge about fire 13 hazards.
14 While there's certainly more work to be done we 15 shouldn't lose sight of the fact that we are in a 16 far better place today than we were 25 years ago.
17 Slide two, please.
18 This is the agenda and the presenters for this 19 morning's briefing by the staff and I'll turn the 20 meeting over to Jack.
21 MR. GROBE: Thank you, Bill.
22 The majority of our presentation today will be on
79 1 the progress we have made and the lessons we've 2 learned in the transition to NFPA 805 the risk 3 informed performance based fir protection requirements.
4 I'll start the presentation with a brief summary of the 5 progress we've made in implementing our closure 6 plans stabilizing the regulatory infrastructure.
7 Next Alex Cline on my right.
8 Alex is the Chief of the Fire Protection branch, NRR 9 and Donnie Harrison on Bill's left.
10 Donnie is the chief of the PRA licensing branch, 11 NRR.
12 They will present our experiences with the 13 transition to NFPA 805 and then Mark Salley on the 14 far left will summarize.
15 Mark's the Fire Research Branch Chief in the office 16 of Nuclear Regulatory Research.
17 He'll give you some insights on our focus area for 18 fire research.
19 Next slide, please.
20 In November 2008 the staff prepared a fire 21 protection closure plan.
22 That plan included eight tasks intended to stabilize
80 1 the regulatory infrastructure for fire protection.
2 Just this week we forwarded to the Commission our 3 second semi annual progress report, three of those 4 eight tasks concerning electrical raceway fire protection barriers, regulatory 5 effectiveness assessments and past fire protection identification resolution have 6 been resolved to date. Substantial progress has been made on the remaining 7 five tasks.
8 Regarding those facilities not transitioning to NFPA 9 805, the last remaining instability in the 10 infrastructure was resolved on Monday this week.
11 The two most challenging issues for those plants has 12 been operator manual actions and multiple cable 13 faults resulting in spurious operation of equipment.
14 The regulatory guidance in both of those areas has 15 been clarified in collaboration with industry.
16 Enforcement discretion has told on operator manual 17 actions and the industry has addressed those issues.
18 Enforcement discretion on circuits allow six months 19 from Monday to identify noncompliances, to utilize 20 the guidance in identifying noncompliance situations 21 and then 30 months to resolve those.
22 At this point I'd like to turn it over to Alex Klein
81 1 who will begin our discussion on at NFPA 805 2 transition issues.
3 MR. KLEIN: Thank you, Jack.
4 Slide four, please.
5 What I'd like to do is give you an overview of the 6 NFPA 805 activities.
7 I'd like to start off with where we are with the 8 infrastructure development for transitioning to NFPA 9 805.
10 You've heard a lot of the history in terms of when 11 the rule was issued back in 2004.
12 A year later, issued a Regulatory Guide 1.205 Rev 0 13 back in 2005.
14 Around that same time frame we also received 15 interest from a number of licensees to 16 transition to NFPA 805 and that occurred later on 17 in 2005 and that's how we obtained the two pilot 18 plant Oconee and Harris.
19 The NFPA 805, I think you heard that also, the standard itself, and I want to 20 stress that, is a national consensus standard and I 21 think that there were some talk previous about the 22 fact that the industry was involved and I wanted to
82 1 stress the fact that this was a national consensus 2 standard developed by the NFPA at the request of the 3 staff back in the 1990s timeframe.
4 Theres been a long history in terms of developing the 5 standard and understanding what's in NFPA 805.
6 At the same time the staff continued its public 7 collaboration process through the pilot plant process from 2005 8 onward to refine and make adjustments to the Regulatory Guide 1.205.
9 We've done that.
10 There's draft Revision 1 that the staff has worked 11 on. It's a draft that we will be going to ACRS with 12 within the next two weeks or so.
13 The ACRS subcommittee we hope to have the revision 14 to Reg Guide 1.205 issued in January 2010.
15 Along with that the staff has also worked to put 16 together a new Standard Review Plan for NFPA 805 17 plants.
18 We worked that in parallel with the revision to 19 Reg Guide 1.205.
20 We've done that in the open collaborative process.
21 We've issued the SRP for public comment earlier this 22 year. We've received comments and incorporated those
83 1 comments so that the publication date for the SRP is 2 in parallel with the revision to Reg Guide 1.205.
3 The staff is also working on a safety evaluation 4 template and that is an effort to put together a 5 document such that we would be consistent and 6 effective in our license amendment request reviews.
7 from the licensees, and of course we're informing 8 ourselves through the pilot plant process when we 9 developed this.
10 We aim to get this done sometime the second quarter 11 of calendar year 2010.
12 You heard some talk about inspections also.
13 The staff has been preparing for inspections. As a 14 matter of fact we have a final draft of a post if 15 you will, NFPA 805 plant that the staff has worked 16 on with the Regions.
17 We've incorporated all the Regional comments.
18 We've incorporated the lessons learned from the 19 pilot plants.
20 Currently the staff's plans are we're evaluating how we 21 can pilot this inspection, you mentioned that one of 22 the pilot plants indicating an interest so we're
84 1 continuing the dialogue with that pilot plant to 2 implement the pilot inspection plan and of course 3 once we've done the pilot inspection plan we will 4 revise the inspection procedure accordingly.
5 Next slide please.
6 With respect to the actual pilot plant license 7 amendment request, we received the two pilot plant 8 license amendment requests in May of 2008.
9 We were informed back then that the license 10 amendment requests were incomplete.
11 We recognize that it was part of the pilot plant 12 process. This license amendment request, it's a first 13 of a kind, integrated approach with using the 14 risk informed, performance based methodology. The staff 15 has continued to work with these two pilot plants to 16 develop their license amendment request in an open 17 and transparent way.
18 We've performed on-site regulatory audits at each of 19 the two pilot plants this year.
20 The Harris plant has completely revised their 21 license amendment request. The staff received the 22 application from the pilot plant about two weeks ago.
85 1 We expect the same type of submittal from Oconee near 2 the end of January 2010.
3 The staff is currently writing safety evaluation 4 report sections for the Harris SER and we expect to 5 issue the Harris SE in the first quarter of 2010 and 6 the Oconee SE in the second quarter of 2010.
7 Next slide, please.
8 With respect to some of the lessons learned with 9 NFPA 805, you heard some of the other prior speakers 10 talk about plant modifications, what I want to I 11 guess mention to you is that when licensees transition 12 to NFPA 805 they reanalyze their fire protection 13 programs.
14 They trace their cables, they look at their 15 procedures, they do a number of activities but, 16 however, it's not just a paper exercise is not just 17 an analytical method that licensees go through. Our 18 experience with the two pilots indicates that the 19 two pilot plants have a better understanding of 20 their risk especially the fire risk in their plants.
21 As a result of that they're making what we believe 22 are substantive safety enhancements in their plants.
86 1 They're not just making procedural changes they're 2 actually making hardware modifications in their plant, 3 and I think you heard some examples from one 4 of the pilot earlier this morning.
5 The other item I want to mention to you and I label 6 it communication, but I want to stress to you what 7 we believe is a success of the frequently asked 8 question process.
9 We implemented that process several years ago when 10 the pilots were first going through their transition 11 to NFPA 805.
12 We've addressed technical and regulatory issues 13 throughout this process, we've done it in a very 14 open and collaborative way.
15 We hold monthly public meetings to discuss these 16 frequently asked questions. Just to give you a flavor 17 we've had somewhere in the order of 47 frequently 18 asked questions and right now there are six that are 19 still open.
20 We've closed the rest.
21 You've heard some of the issues associated with 22 NUREG CR 6850.
87 1
The staff recognized that some of the issues 2 associated with that, the technical issues that were 3 identified by the pilots and the non-pilot plants, 4 were not achieving timely resolution so we 5 implemented a modified frequently asked question 6 process for the 6850 related technical issues 7 earlier this year.
8 As of this point right now we have closed out all of 9 the identified 6850 related facts that have been 10 identified by the industry.
11 There's some follow on paperwork that's still 12 necessary through the issuance of closure memos but 13 the staff is on a path to closure for those.
14 With respect to planning, what I want to say about 15 that is when licensees transition to NFPA 805 16 you hear a lot about fire protection that seems to 17 be the operative word but I think as one speaker 18 indicated and rightly so identified that the 19 resources and expertise needed to transition to NFPA 20 805 is not just limited to classical fire 21 protection. There are many other disciplines 22 involved, for example, fire modeling expertise is
88 1 necessary, post-fire safe shutdown expertise, 2 circuit analysis expertise and not to forget 3 expertise in probabilistic risk assessment is 4 necessary.
5 So an integrated team approach is necessary 6 certainly to transition to NFPA 805.
7 It's not simply a one discipline approach.
8 The team approach is not only on the licensee 9 side, it's also on the staff side.
10 Next slide, please.
11 12 With respect to the non-pilot plants coming in for 13 the license amendment request as I indicated to you 14 we've had many public interactions and I think 15 something on the order of 50 or more public meetings 16 have been held with the industry over the last 17 several years while the pilots transitioned.
18 I think that the non-pilots have had an 19 opportunity to learn and understand the issues that 20 the non-pilots have been facing and the challenges 21 so we've continued this open and collaborative 22 process.
89 1
We believe that the issues are well known that our 2 guidance at this point is well-known.
3 That our guidance is stable, and that licensees can 4 move forward to NFPA 805.
5 We believe that some licensees may submit license 6 amendment requests in the early to mid 2010 calendar 7 year time frame with the majority of the license 8 amendment requests expected in the later part of 9 2010 in the fall.
10 I now hand it over to my colleague, Donnie Harrison.
11 MR. HARRISON: Thank you Alex.
12 Slide eight: I just want to start by saying that we 13 have established the infrastructure for being able 14 to perform quality PRAs, fire PRAs that support 15 decision-making not just for NFPA 805 but for other 16 risk informed applications; that's an important 17 point.
18 We've done this through an open, collaborative 19 environment. We've had numerous public meetings.
20 We've gone through the various guidance to develop 21 that and create an infrastructure that's also stable 22 so the industry knows what to expect and how to
90 1 perform the analysis.
2 That being said we also recognize that NFPA 805 3 applications are the first time some of this 4 guidance has been implemented and therefore you are going 5 learn things.
6 We're feeding those lessons back into revising the 7 guidance, refining the methods and moving forward 8 with again stable guidance for the industry to use.
9 Next slide.
10 As Alex mentioned there is this frequently asked 11 question process that we follow to address issues 12 that were raised during the NFPA 805 implementation 13 modeling as issues came up.
14 At this time of the dozen or so fire PRA related 15 6850 related FAQs, all of them have been resolved.
16 We have no new issues that have been implemented 17 that are submitted to the process for us to address 18 and again we're just finishing out the paperwork on 19 the last couple of items.
20 From my perspective that process has worked well to 21 come to a staff position on these issues, so that it 22 enables the industry to move forward.
91 1
At the same time, we also recognize that there are 2 some issues, some fire modeling issues that if you 3 want to refine the model you need to do additional 4 research and to support that, Mark Salley is going 5 to talk about more about that in a minute. Again 6 the issue there is if you want to refine the PRA you 7 need to have the technical basis to make those 8 refinements and that's where in some areas we need 9 additional research to collect the data and do the 10 analysis. With that I'll turn it over to Mark to 11 talk about the research.
12 MR. SALLEY: Thank you Donny, Chairman, 13 Commissioners.
14 I would like to talk about the Office of Research 15 and the part where we are supporting NRR here. Firstly is how do we do this 16 work. We work off of user need requests.
17 This is where we meet with our colleagues in NRR and 18 they give us their priorities and the things they 19 would like us to do the research on.
20 So that kind of sets our precedence as to how we'll 21 work the research.
22 The second thing is we like to work in a
92 1 collaborative nature, we don't like to go it alone. So 2 where possible we'll look for partners. One you have 3 heard from EPRI this morning where they have similar 4 interests that are of a technical nature where we can 5 pool our resources, we'll work with EPRI and put out 6 joint reports.
7 We do a number of that.
8 Second thing is other experts in the government.
9 In the 1960s and 70s there was a fire problem in 10 America that brought a Presidential look at it and 11 the Report America Burning came from that. In 1975, the same year we were 12 established, they established a Center for Fire Research out at NIST.
13 So are very tight working with NIST on their fire modeling.
14 They have almost 35 years of fire modeling 15 experience that they help us do our jobs with.
16 We also have a very strong national laboratory 17 program.
18 Chairman, you and Commissioner Svinicki have both 19 been to Sandia. You have seen the work that we have done and how we are.
20 doing our cable functionality testing out at Sandia. Next slide, please.
21 Briefly I'd like to just touch on the key areas that 22 we're working in Research right now.
93 1
I've heard the term modeling thrown around a lot and 2 its got me confused whether we were talking about 3 fire modeling or PRA modeling, or human reliability, so I want to try to 4 keep it a little tighter into the bounds.
5 Fire modeling is not unique to NFPA 805 6 we've been at this for quite a while if you think 7 back into the 1990s when we decided to use the risk 8 information and the STP process was developed we had 9 a change there where we had to teach our inspectors 10 on how to understand fire dynamics if they're going 11 to use the significant determination process. When 12 they had have a finding one of the first things they do is 13 postulate the fire so they need to understand some 14 of the basic fire dynamics.
15 We ran a three-year program, quarterly workshops 16 with our inspectors and we brought all those lessons 17 together and we issued NRC NUREG 1805 which is our 18 basic fundamentals of fire dynamics for our inspectors.
19 That was in 2004.
20 We continued on and we had a joint program with EPRI 21 to V&V fire models.
22 That came together as a report we looked at not one
94 1 but five different fire models.
2 We brought a lot of partners together, EPRI, NIST, 3 NRC and we did five fire models. We followed the national 4 standards, ASTME, 1355.
5 We looked at 26 different experiments, 13 key 6 parameters.
7 Basically, it gives you the War and Peace of fire 8 model V&V and the fact that its seven volumes long 9 and it's quite in depth.
10 It addressed a lot of uncertainty.
11 It went through peer reviews.
12 It went through ACRS so it was quite a rigorous 13 piece of work.
14 We followed that up with an expert elicitation 15 where we bought fire modeling experts together and said 16 how good are we and where do we go in the future.
17 What experiments do we need to do and where do we need to 18 work with NIST to improve the fire models. We've completed that.
19 Currently today a joint team between EPRI and NRC 20 and NIST is putting together a fire models users 21 guide, if you will, a guide to help the plant people 22 and the consultants and our inspectors when they
95 1 look at the fire models to make sure that we've 2 adapted to the nuclear environment. That has just 3 finished the peer review.
4 The guys are finishing the comments up now and hopefully 5 by the end of the year that will be out for public 6 comment.
7 We have a routine where we do this. We use a 8 lot of peer review and we would like to go for a 9 good sixty day public comments so that we can get 10 everybodys concerns about our product.
11 That's where fire modeling is and with that I think well have a stable 12 base to work on and to go to the next level with 13 fire modeling which we're thinking about now.
14 The next form of modeling is the fire PRA.
15 You've heard a lot about 6850, it came about in the 16 year 2005 and again it was a joint project between 17 EPRI and the NRC.
18 The keys were here to get the methods, the tools and 19 the data needed to do a fire PRA.
20 At the time it was the state-of-the-art so in 2005 it 21 was a very good state-of-the-art document.
22 We're now in 2009.
96 1
We now have some experience in using it.
2 We conducted other research and other experiments 3 and the state-of-the-art is moving forward.
4 We're seeing that.
5 We're seeing that where 6850 is being refined.
6 We've looked at all the different elements of it and 7 the biggest question for me right now is when do 8 we do the revisions because the state of the art will 9 advance. Were seeing that.
10 That's why we're tracking that 11 right now and we expect to do a revision again 12 hopefully with EPRI as a joint document.
13 The third activity is experiments.
14 To get the numbers and the data a lot of times it 15 takes an experiment to answer one of the FAQs and 16 experiments don't happen overnight.
17 Chairman, you just saw some of the work we were 18 doing on cable flame spread and ignition.
19 That's a multiyear program that we're working with 20 NIST and yes, it will answer an FAQ and yes, it will 21 improve the understanding but it will take a little time 22 to do.
97 1
The cable functionality testing that's one that we 2 been with a few years and in 2001 NEI and EPRI 3 conducted a series of test and came up with a bunch 4 of numbers and different correlations of how the 5 cables would fail when exposed to fire. We 6 followed that in 2003 with a facilitated workshop.
7 We brought the best minds together and said what 8 have we learned about this and how can we predict 9 this cable response to fire.
10 From that some unanswered questions came, we 11 followed it up with a cable program. We went to 12 Sandia. In 2008 we completed that and we got the 13 answers to those questions so we have a pretty good 14 handle on the AC response, AC circuit response to fire.
15 As a byproduct we also developed the improved fire 16 modeling tool where we can do the one dimensional 17 heat transfer to get a better, more accurate 18 prediction to cable failure.
19 Finally, the third piece is going on right now and 20 that's been a change from an AC circuit to a DC 21 circuit, how does that affect it?
22 This testing will complete at the end of the year.
98 1 Earlier in the year we will come up with a report and then 2 we'll move into another facilitated workshop 3 approach to see how we've advanced.
4 The final thing is knowledge management, we've got 5 35 years of information of history we're starting to 6 collect that.
7 You've seen the Browns Ferry NUREG brochure this 8 year as well as one on fire research.
9 With that I will turn it back over to Jack.
10 MR. GROBE: Thanks Mark.
11 Just a brief summary. In the past several decades 12 there has been very significant improvements in fire 13 safety at nuclear power plants.
14 For the non-805 plants the staff has worked very closely 15 with the industry and our external stakeholders and 16 the regulatory infrastructure for those plants is 17 stable and predictable.
18 Those issues have been resolved.
19 We've collaborated closely with our external 20 stakeholders to stabilize the infrastructure for 805 21 transition. The pilots are nearly complete, we 22 believe that that's sufficiently stable to move
99 1 forward with 805.
2 As we move forward for any additional questions that 3 come up we have an effective process, the frequently 4 asked questions process and a robust relationship 5 that Mark just described with the industry to 6 resolve any necessary research issues that are 7 important to answering frequently asked questions.
8 As with most of our activities since we've 9 implemented the closure plan, we completed one 10 ahead of schedule.
11 That completes the staff presentation, we're ready 12 to answer any questions.
13 CHAIRMAN JACZKO: Thank you for sharing 14 your thoughts and insights we'll start with 15 Dr. Klein.
16 COMMISSIONER KLEIN: I think the staff 17 should be complimented on all of their progress on 18 the fire protection all the way to the research 19 that's always a fun part to do and to the results of 20 doing all the regulations and inspections. I guess, 21 Jack, just to start with you in terms of the 22 technical issues in regarding the deterministic
100 1 process, are there any technical issues that need to 2 be resolved for the deterministic approach?
3 Mr. GROBE: We're not currently aware of 4 any.
5 The final issue was the multiple cable faults 6 spurious operation issue. That resolution methodology 7 is well understood by both the staff and the 8 industry and has been piloted at one plant.
9 We're not aware of any technical issues with respect 10 to the deterministic approach.
11 Mark highlighted a very interesting issue with 12 respect to DC circuit failures.
13 We're staying very closely connected with what 14 research is learning in that area.
15 That might be something that precipitates some 16 further action in the deterministic area but as of 17 right now there is none that we're aware of.
18 COMMISSIONER KLEIN: Thanks.
19 Alex, you heard from the first panel, there is some 20 interest in the inspection program for the NFPA 805 21 and you talked about that you're working with a 22 plant, how are you communicating that to the
101 1 industry; is there good communication channels on 2 how do you intend to inspect; and then how do you 3 take lessons learned?
4 MR. KLEIN: As I indicated Commissioner, 5 the inspection procedure is draft final, right now, 6 we have not shared it yet, with the public, so 7 therefore the non-pilots are not aware.
8 We are evaluating what our plan should be what the 9 next step should be with respect to that inspection 10 procedure and piloting it with the pilot 11 plant. We certainly do intend to share it with the 12 public and with the non-pilots.
13 We will do the this entire process in a very open 14 and collaborative manner with the industry when the 15 staff has made the determination that we can release 16 the inspection procedure to the public.
17 COMMISSIONER KLEIN: Thanks.
18 Well I had visited Shearon Harris several months ago 19 and was very impressed with the thoroughness of what 20 the staff has done and the industry in regard to the 21 fire protection in their activities to NFPA 805.
22 Are you able to quantify the safety values that they
102 1 have achieved in terms of actions they've taken in 2 other words we always look in numbers, as a 3 regulator so when you look at all of the things that 4 they've done, are you able to quantify that?
5 MR. KLEIN: I'll ask Donnie to chime in, 6 in a minute because I'm more of a classical fire 7 protection engineer if you will and if I could 8 answer from that point of view.
9 From my understanding from the modifications that 10 the Harris plant has made, one of the ones that I 11 would bring forth that Mr. Donohue did not mention 12 was their work in the application of the 13 material called mega cable which provides a 14 three-hour fire barrier in their plant without using 15 any fire wrap material.
16 My understanding is that the Harris plant has 17 implemented this modification and put in hundreds of 18 feet of this material if not thousands and my 19 perspective is that when a licensee puts in hardware 20 modifications such as incipient fire detection to 21 me that is a benefit to plant fire safety. In terms 22 of quantification, Donnie could probably better
103 1 answer that question.
2 MR. HARRISON: I'll start by quoting the 3 Harris representative they said they had a 20%
4 reduction in their fire core damage frequency 5 so there's that quantitative amount, but the 6 one thing I would stress is that synergistic benefit 7 like the reactor coolant pump seal injection, the alternate seal injection 8 capability that they're adding. It's not just improving the fire risk it's improving 9 the overall plant risk and that is a big benefit. One last 10 thought would be, you can't quantify the benefit 11 associated with the enhanced knowledge. They've 12 learned a lot about their plant about where the risks are, 13 where there are areas that they need to focus attention on.
14 The one good thing with a fire PRA, it's an integrated 15 tool and it brings that clarity to focus even if 16 you're arguing that it's somewhat conservative it 17 still focuses in on those areas that are important 18 to look at and you can't devalue that knowledge.
19 It's an important aspect that were gaining through 20 NFPA 805.
21 COMMISSIONER KLEIN: Well this may be a 22 question either for Bill or for Jack there are I
104 1 believe over 50 plants that indicated that they wanted to 2 look at transitioning to NFPA 805, do you think 3 they'll all proceed down that path?
4 MR. BORCHARDT: I wouldn't hazard a guess 5 myself. I think that we are hearing varying degrees 6 of feedback from licensees.
7 Theyre still interested in the pilots ultimately 8 play out and I think there's a lot of individual 9 decisions to be made.
10 COMMISSSIONER KLEIN: Any staffing issues 11 related as we move forward?
12 MR. BORCHARDT: For NRC?
13 COMMISSIONER KLEIN: Yes.
14 MR. BORCHARDT: None that I know of.
15 MR. GROBE: No, none.
16 The Commission adopted budgets for 2010-2011 that 17 included additional resources.
18 One of the reasons that the staff generated the 19 Standard Review Plan and is generating a standardized 20 Safety Evaluation Report is anticipating significant 21 workload in the fall.
22 We're going to be effectively integrating contractor
105 1 resources along with our staff resources to get that 2 work done, so we're in good shape.
3 COMMISSIONER KLEIN: Are you going to stage 4 the applications that we've received and then the 5 review of those or how are you going to handle the workload?
6 MR. KLEIN: The SECY Paper that we had 7 submitted to the Commission about a year or so ago 8 Jack mentioned, the additional resources that were 9 given to the staff.
10 We did mention an approach where we would then take 11 a look at -- if we did receive 15 to 20 of these license 12 amendment requests all within a span of a week or so, 13 or what have you.
14 There is a process the staff would go through to 15 determine which ones we look at first and we used 16 examples such as where that licensee may stand with 17 respect to the quality of their submittal, where 18 they stand in line with respect of the next tri-annual 19 fire protection inspection and any other issues 20 that the staff may be aware of at that particular 21 licensee.
22 COMMISSIONER KLEIN: Thanks, well moving
106 1 onto the research aspect how much does the NRC 2 typically spend a year on fire research?
3 MR. SALLEY: That varies, right now we have 4 approximately ten FTE that are in the fire research branch.
5 We have ten individuals that are working all the aspects of the 6 fire modeling, the fire PRA, and the experiments.
7 The budget numbers vary depending on if we have 8 experiments or not.
9 We typically run in the 4 to 6 million. Obviously, 10 when we're doing a lot of experimental work that 11 costs a lot more money.
12 COMMISSIONER KLEIN: In terms of the 13 research activities, what would you list as number 14 one on what we need to do in research in the fire 15 protection area?
16 MR. SALLEY: They're all number one with 17 me, but the fire PRA seems to be the big challenge when 18 you look at the different communities the fire 19 modeling group tends to work pretty well together. There is a fire HRA group 20 that is also going, but the larger fire PRA is the one we need to focus in 21 on right now.
22 COMMISSIONER KLEIN: Great, thanks, no
107 1 further questions.
2 COMMISSIONER SVINICKI: Thank you I would 3 return to a topic I was looking at the transcript from last year's meeting 4 and I had heard about some training on the broad issue of the human capital 5 challenges in fire PRA and making sure that if there was going to be a 6 concentrated industry focus that they would have folks trained and if the 7 transcript is accurate I was informed that there was a course 8 jointly offered by NRC and EPRI and I think it was 9 for people that were already somewhat expert in PRA 10 and they could do a module on fire PRA. Do we still participate in that and 11 could any of you react on any of the human capital challenges. Have we 12 trained folks and made visible progress from last years meeting.
13 MR. KLEIN: If I could respond first and 14 I'll ask Mark to please respond 15 In terms of staff training, the staff does 16 participate in this training. When we 17 receive new staff members into the branch we develop 18 a training plan with them and identify the 19 training needs for that individual based upon prior 20 knowledge and expertise of the individual and we do 21 send those individuals to the training that you 22 speak of Commissioner, with the, under the joint
108 1 effort with EPRI and Research and that's why I wanted 2 Mark to perhaps make some remarks about that effort.
3 MR. SALLEY: Along with EPRI when 6850 was 4 published we recognized that you couldn't just 5 publish it and walk away from it, the training 6 aspect needed to take place.
7 Under the same EPRI and NRC Research MOU twice a year we 8 hold the training.
9 We alternate the years, one year EPRI will take the lead, 10 the second year we will take it.
11 It's held for free because they're working with us 12 so it's kind of a public meeting type of atmosphere.
13 But the training is fairly involved, it's a full week 14 and there's three separate classes one on the PRA 15 HRA aspect, one on the fire and fire dynamics aspect 16 and the third thing is the electrical engineering, the 17 systems aspect.
18 We just finished up this past year, it was EPRI's 19 turn to host it, they did one in Palo Alto, second 20 one was done in Richmond.
21 Next year will be the NRC's turn and it will be 22 holding them probably up in this area two times.
109 1 It's well received.
2 We're typically getting over a hundred people both 3 from the regulator and from the industry and we're 4 also seeing international involvement, typically nine 5 to 13 people will come from different countries to 6 attend this.
7 COMMISSIONER SVINICKI: Thank you for the 8 fuller description. You mentioned my visit to Sandia 9 and I did have the opportunity to see NRC's fire 10 research there with Brian Sheron who heads our 11 Office of Research.
12 I don't want anyone at the table to take this the 13 wrong way but people who spend their entire careers 14 studying fire have some interesting 15 personalities. What it left me with on a 16 more serious note is, there's a lot of phenomenology 17 and a lot of behavior of fire. If it's not your 18 field I think you kind of take for granted that it's 19 just kind of a chaotic phenomenon and that how much 20 have you studied it for a long time could you really 21 come to a good data set about behavior and actually 22 there is a lot we do know it isn't as if there's
110 1 ignition and then it goes wherever it wants to.
2 It was really fascinating to spend that time and 3 understand about flame and flame behavior and just 4 all of the components here but it also 5 leaves me with a deeper understanding of the 6 complexity of these issues so with the previous 7 panel or as you talk about resolving issues.
8 Mark, you'd mentioned something that I had heard 9 about in a few years but I think maybe is sometimes 10 underutilized which is the expert elicitation 11 process. I think it can be really valuable in a 12 field like this in directing research in the most 13 productive -- again you mentioned the user needs 14 process and we do need to be driven by what the 15 needs are we just can't explore issues because of 16 their interest to us as a regulator we need to have 17 an application for what we fund but I think expert 18 elicitation and I think you talk about a facilitated 19 workshop which would be a great environment to come 20 up with the plans and constantly re-examine what are 21 the issues that we need to resolve.
22 We heard a little bit from EPRI in the last panel.
111 1 They had a very extensive list so when I compare 2 that, Mark, against your statement of being 3 responsive in the Office of Research to the user needs 4 process, EPRI has this larger universe of 5 things that need to be resolved and looked at with 6 that.
7 So I suspect within that there are things that are 8 appropriate for us to cooperate on and other things 9 probably that the industry should more directly fund 10 and might be near-term issues that it would benefit 11 them to resolve.
12 So I don't really think there's a question in there 13 anywhere but I appreciate, again I spent some time 14 recently on the research aspects of this question 15 and appreciate that you're driven by the processes 16 the user needs process and that it's just a very 17 complex question.
18 So I don't know if there's anything more generally 19 about moving forward in the future that you would 20 like to add to that, again I saw that cable testing, 21 saw some of the large-scale facilities that they have at 22 the national lab to do fire testing and they of
112 1 course do a lot of work to the military as well on 2 related questions.
3 Is there anything going forward, Dr. Klein asked you 4 about staffing and resourcing in Research but over 5 the longer term, do you think we kind of have the 6 size and scope of program that we need?
7 MR. SALLEY: The Office of Research pays a 8 lot of attention to the fire research and I think 9 we're where we need to be..Theres going to be work, theres going to be 10 refinement. I think the low hanging fruit we've already picked 11 so we now need to get into the more refining of 12 the process and that's where we're focusing now.
13 Case in point is the flame spread.
14 There is a method in 6850 and when people do use it 15 yes the fire does seem to move a little too fast and 16 it's a first order principal type of approach so it 17 is conservative as people would say, we need to 18 study the phenomena of the cable trays burning and 19 get more research done on that so that we can refine 20 that and feel comfortable with it.
21 Those are the kinds of things we're working with 22 now.
113 1
Fire modeling, too, we look to advance it, we want 2 to push the state-of-the-art.
3 Hopefully we will get into a cycle of five year where 4 we're refining this and we'll continue to refine our 5 skills, that's our goal.
6 COMMISSIONER SVINICKI: Thank you.
7 Thank you, Mr. Chairman.
8 CHAIRMAN JACZKO: Thank you Commissioner 9 Svinicki.
10 Jack maybe Ill ask you or Alex either one of you want to answer this, you 11 did send us up earlier today, or yesterday, the update on the closure plan. One 12 of the big changes of course is the movement of the completion 13 date of the two pilots further out in the future and 14 certainly there are a lot of concerns expressed this 15 morning and I think fair points about some of the 16 challenges and uncertainties about the work that 17 we're still doing.
18 What should make the Commission believe now that the dates that we 19 now see in the closure plan, are going to be 20 met this time?
21 MR. GROBE: I think the most important 22 thing is that with respect to Harris all the
114 1 questions have been answered.
2 The licensee has incorporated all the effort from 3 the past year and a halfs work into a revision of the 4 license amendment request and we're actually writing 5 the safety evaluation report.
6 It's frequent that in the course of writing safety 7 evaluation report the reviewer will scratch his head 8 and say I still got this one thing that I need to 9 get nailed down.
10 So it wouldn't surprise me if there's a few 11 additional questions but essentially the work is 12 done.
13 Now it's just writing a report and that the lengthy 14 process. The licensee would then once we complete the 15 report have 30 days to review it for accuracy and 16 proprietary information then it goes to tech editing 17 so just the issuance of that takes 8 to 12 18 weeks.
19 So we're done essentially with Harris except for the 20 paperwork.
21 The same thing would apply to Oconee but starting roughly in 22 January.
115 1
They anticipate finishing the last of the answers 2 and resubmitting a license amendment request that 3 incorporates all the information in one place and 4 then well be writing the Safety Evaluation Report.
5 So we're certainly rounding third and heading home.
6 CHAIRMAN JACZKO: It's good to hear.
7 I think as I said earlier this will be a good 8 enhancement to safety in having these first license amendment 9 requests completed and certainly decisions on those will 10 perhaps give us a the way to document that.
11 Mark, this is a follow on a little bit of what 12 Commissioner Svinicki was asking and Dr. Klein had 13 asked as well.
14 EPRI did have a list of items, I think Commissioner 15 Svinicki hinted at this, some of those may be appropriate for Entergy, some of 16 them may be activities to do in collaboration. Do you 17 have a sense of that particular list, is that 18 something that you worked with EPRI before, is it a 19 new set of information or do you have a sense of how 20 our research activities will coordinate with those?
21 MR. SALLEY: I had not seen that list 22 before that was an EPRI list, to me it looks like a
116 1 bunch of opportunities.
2 CHAIRMAN JACZKO: I should've never ask an open 3 question like that to researcher.
4 MR. SALLEY: We also have a list, and what we 5 will do under the MOU quarterly or at least 6 biannually is what we're trying to get at, we'll sit 7 down with EPRI and Ken and I, for example, will compare notes and we'll 8 sit down and say what are you doing, what are we doing?
9 Which ones can we do together to get the maximum 10 effect and which ones do you need to do separate and we need to do 11 separate, so there's a lot of communication.
12 The DC testing was one, for example, that we started 13 alone and we started moving and getting it in 14 process and talking with Ken, NEI, and the industry 15 and say that we'd like to contribute, we'd like to 16 give you some of the cables to test.
17 We'd like to take part of it and we amended the MOU 18 and we brought them on as a partner.
19 There's also the intellectual part, there's some very 20 good consultants that work for them that we can also bring 21 in.
22 Doing a joint project is always harder than going it
117 1 alone because you have the two groups so it's much 2 more of a challenge but I think at the end of the 3 day for the end-user it's better that we work 4 together when we can and get a single product and 5 not have competing methods where we fight about the 6 third decimal point, we look at the larger safety issue.
7 CHAIRMAN JACZKO: Thank you.
8 We heard a lot this morning and the title of 9 this meeting was lessons learned from the Harris 10 review and we certainly heard a lot I think this 11 morning from Shearon Harris themselves about some of 12 the lessons and the challenges, where would you say 13 we are at this point in incorporating those lessons, and the 14 staff you talked about the new Standard Review 15 Plan, you talked about Reg Guide 1.205.
16 Is that where the staff would say right now we've 17 documented those lessons learned or is that still 18 have to be processed a little bit?
19 MR. GROBE: I think there's two locations 20 where the lessons exist today, one is in the results 21 of the FAQs (the frequently asked questions) and the 22 second is the revision to Reg Guide 1.205.
118 1
That revision is essentially complete and its 2 been a collaborative process. Both the industry and NRC 3 well understand the changes in that revision.
4 We're meeting with the chairman of the ACRS subcommittee 5 to make sure that there's no outstanding questions 6 that they have that we're unaware of.
7 We anticipate -- Alex said January -- I'm still 8 pushing for late December to have that on a street 9 as a final document.
10 It has gone through extensive industry comment so 11 the two sources of lessons learned are the facts and the Reg Guide..
12 CHAIRMAN JACZKO: I think the reasons for 13 the lessons learned is to make it more predictable 14 and I think easier for the subsequent applicants to 15 know exactly what information they need to include 16 in their applications as they go forward. As this 17 process improves, how long does the staff think it 18 will take to do perhaps the nth review for a 19 license amendment request in this area, perhaps 20 we're there already, maybe we need a few more 21 iterations of reviews to better understand what the 22 challenges are?
119 1 MR. KLEIN: The staff's estimation right 2 now for the initial license amendment requests that 3 are coming in sometime next year is anywhere from 4 one to two years.
5 Of course as we learn, as we progress, as we refine, 6 as we better understand things we think that by the 7 time we get through the nth plant license amendment 8 request that the timeframe would be something less 9 than one year.
10 But that's a prediction on my part, looking at my 11 crystal ball but I think given the number of 12 license amendment requests that are coming in initially and 13 given the staffing that we have and the resources 14 available that we'd be able to turn around a license 15 amendment request for the nth plant on something on 16 order of one year or less.
17 CHAIRMAN JACZKO: I think the last question 18 and this Donnie might be a question that you can 19 answer.
20 One of these issues that came up earlier is the 21 comparability of the fire PRAs to the power PRAs, the other internal 22 event PRAs we have.
120 1 In your sense, when do you think that 2 will get all of those to a common or to a level 3 which within a particular plant that the fire PRAs 4 will be comparable to the other internal event PRAs 5 and be able to be used in a comprehensive way or 6 consistent manner.
7 MR. HARRISON: I guess I would answer that 8 in two different directions, one would be you have to 9 take the application to the context of what the 10 decision you're making.
11 We've heard a lot of comments about conservatism and 12 bounding approaches.
13 It's not like there isn't conservatism in internal 14 events PRAs, there are, however they usually don't 15 drive the decisions therefore you can live with them 16 and so you have to look at the decision you're 17 making and if you can use the guidance that 18 we put out, work collaborative with the industry to 19 develop and you are able to come up with risk 20 results that support making decisions going forward 21 than you can live with those models the way they 22 are.
121 1 So that's one part of the answer. The other thing is when we 2 wrote Reg Guide 1174, which is the risk informed decision making guidance 3 document, the acceptance guidelines in there require that you use 4 a total CDF or a total large early release frequency value so it was 5 recognized in the late nineties that you needed to 6 be able to reflect on that.
7 If you didnt have a PRA you were supposed to have to 8 look to see if your IPEEE had vulnerabilities that 9 would make you question if you were higher in a risk 10 area than you were. With the development of fire PRA 11 methodology refinements and enhancements we'll have 12 actually a clearer answer to that question.
13 We won't have to rely on a subjective evaluation 14 of the IPEEEs, we'll actually have results that we can look 15 at.
16 That will actually make the decision making process 17 smoother.
18 So I guess that's how I would answer your question.
19 MR. GROBE: Donnie, I'm glad you brought 20 that up.
21 805 is just another in a long line of enhancements, 22 PRA enhancements, to our regulatory processes.
122 1
There's two initiatives on the table today.
2 It's the risk informed applications for allowed 3 outage times in the tec specs as well as the surveillance 4 frequencies.
5 Under Reg Guide 1.200 Revision 2, that's the Reg 6 Guide for doing PRA, when licensees come in for 7 those they're going to have to have an integrated 8 internal and external events PRA which not only includes 9 internal events and fire but also includes seismic.
10 Fire and seismic are the two biggest challenges 11 today. So those will all have to be melded together 12 and I anticipate in the next year to two years that 13 we're going to be getting a number of applications 14 for those tec spec enhancements which will integrate fire, 15 seismic and internal events.
16 CHAIRMAN JACZKO: I think it is one of the 17 side effects of the NFPA 805 transition that should 18 help those plants move along more quickly on having 19 their fire PRA.
20 The last question I have, and again this is for anyone who 21 wants to answer.
22 This is arguably a more complex way to address fire
123 1 protection. I'm always reminded that Appendix R is 2 somewhat simple. I can remember the three basic requirements, 3 the deterministic requirements in the regulations. Every time I read 4 NFPA 805 I learn something new.
5 How are we going to make sure that we are able to 6 communicate these results to the public?
7 I think that there were comments earlier about that 8 and I think that's a fair point and make sure that 9 the public can understand the decisions that were 10 making and understand how we're fulfilling our responsibilities 11 there with a much more complicated framework.
12 I don't know if anyone has any thoughts on that.
13 MR. GROBE: Mr. Gunter raised two very 14 important issues. One is making sure that you don't apply 15 the models outside of their capability and we're 16 very focused on that and the second is the more 17 complex your analytical approach gets the more 18 difficult it is to explain it to the public.
19 Fire is one of the areas where we have one of the 20 most advanced websites.
21 There's a multitier website with high level 22 questions and then you can drill down in any
124 1 particular area to more detail or more detail.
2 That's been a collaborative effort between, NRR, 3 Research and our Office of Public Affairs to make 4 sure that is a user-friendly approach to the public 5 gaining information about how we do fire protection 6 regulation.
7 So it's one of, I think, our most successful 8 initiatives from the standpoint of making the most 9 effective use of our web resources.
10 CHAIRMAN JACZKO: I think that is certainly 11 one tool and I encourage the staff to continue to find ways to make sure that 12 we can communicate this because I think is a very 13 fair point that as these things get more complicated 14 they get harder to explain, and perhaps internally as 15 well.
16 Any other comments or questions?
17 COMMISSIONER KLEIN: Again I would like to 18 summarize and compliment the industry and the staff 19 for all the efforts that have been put forth on fire 20 protection.
21 I think while NFPA 805 was probably more challenging 22 both for the staff and industry than you expected
125 1 when you started down the path, I think the bottom line 2 is that the plants are safer, we know more and the 3 public is better protected by the actions that you 4 all have taken.
5 Thanks for the hard work.
6 CHAIRMAN JACZKO: Thank you very much that 7 concludes our meeting and we certainly look forward 8 to the staff continuing to make progress and 9 continuing to work on engaging, 10 improving the areas where we have some uncertainties 11 and continue to refine the models and the tools that 12 we have.
13 I appreciate the staff effort.
14 Thank you.
15 16 17 18 19 20 21 22