ML21174A294
| ML21174A294 | |
| Person / Time | |
|---|---|
| Issue date: | 05/19/2021 |
| From: | Advisory Committee on Reactor Safeguards |
| To: | |
| Wang, W, ACRS | |
| References | |
| NRC-1518 | |
| Download: ML21174A294 (315) | |
Text
Official Transcript of Proceedings NUCLEAR REGULATORY COMMISSION
Title:
Advisory Committee on Reactor Safeguards Thermal-Hydraulic Phenomena and Reliability and Probabilistic Risk Assessment Docket Number: (n/a)
Location: teleconference Date: Wednesday, May 19, 2021 Work Order No.: NRC-1518 Pages 1-189 NEAL R. GROSS AND CO., INC.
Court Reporters and Transcribers 1323 Rhode Island Avenue, N.W.
Washington, D.C. 20005 (202) 234-4433
1 1
2 3
4 DISCLAIMER 5
6 7 UNITED STATES NUCLEAR REGULATORY COMMISSIONS 8 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 9
10 11 The contents of this transcript of the 12 proceeding of the United States Nuclear Regulatory 13 Commission Advisory Committee on Reactor Safeguards, 14 as reported herein, is a record of the discussions 15 recorded at the meeting.
16 17 This transcript has not been reviewed, 18 corrected, and edited, and it may contain 19 inaccuracies.
20 21 22 23 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 www.nealrgross.com
1 1 UNITED STATES OF AMERICA 2 NUCLEAR REGULATORY COMMISSION 3 + + + + +
4 ADVISORY COMMITTEE ON REACTOR SAFEGUARDS 5 (ACRS) 6 + + + + +
7 THERMAL-HYDRAULIC PHENOMENA AND RELIABILITY AND 8 PROBABILISTIC RISK ASSESSMENT JOINT SUBCOMMITTEE 9 + + + + +
10 WEDNESDAY 11 MAY 19, 2021 12 + + + + +
13 The Joint Subcommittee met via 14 Videoconference, at 1:30 p.m. EDT, Joy Rempe, Co-15 Chair, presiding.
16 17 COMMITTEE MEMBERS:
18 JOY L. REMPE, Co-Chair 19 VESNA B. DIMITRIJEVIC, Co-Chair 20 RONALD G. BALLINGER, Member 21 VICKI M. BIER, Member 22 DENNIS BLEY, Member 23 CHARLES H. BROWN, JR. Member 24 GREGORY H. HALNON, Member 25 WALTER L. KIRCHNER, Member NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
2 1 JOSE MARCH-LEUBA, Member 2 DAVID A. PETTI, Member 3 MATTHEW W. SUNSERI, Member 4
5 ACRS CONSULTANT:
6 STEPHEN SCHULTZ 7
8 DESIGNATED FEDERAL OFFICIAL:
9 WEIDONG WANG 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
3 1 CONTENTS 2 Subcommittee Chairman Introductory Remarks . . . 4 3 Southern Nuclear Operating Company, Inc. (SNC) 4 Overview of Vogtle LAR and Exemption Request 5 for a Risk-Informed Resolution to GSI-191 . . . . 7 6 Summary of Differences Between Vogtle and STP 7 Risk-Informed Approach . . . . . . . . . . . . . 83 8 U.S. Nuclear Regulatory Commission (NRC) Staff Review 9 of Vogtle LAR and Exemption Request 10 for a Risk-Informed Resolution to GSI-191 . . . 102 11 12 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
4 1 P R O C E E D I N G S 2 1:30 p.m.
3 CO-CHAIR REMPE: This meeting will now 4 come to order, this is a joint Subcommittee meeting of 5 the Accident Analysis and Thermal-hydraulic Phenomena 6 and Reliability and Probabilistic Risk Assessment 7 Subcommittees for the Advisory Committee on Reactor 8 Safeguards.
9 I'm Joy Rempe, Chairman of today's 10 Subcommittee meeting, Member Vesna Dimitrijevic is 11 the Co-Chairman. ACRS members in attendance include 12 Charles Brown, Dennis Bley, Jose March-Leuba, Dave 13 Petty, Matthew Sunseri, Ron Ballinger, Walt Kirchner, 14 Vicki Bier, and Greg Halnon.
15 ACRS consultant Steven Schultz is also 16 present. Weidong Wang of the ACRS Staff is the 17 designated federal official for this meeting.
18 During today's meeting, the joint 19 Subcommittee will review the Staff's safety evaluation 20 on Southern Nuclear Operating Company, SNC, exemption 21 and license amendment request for our risk-informed 22 resolution to GSI-191 for local electric generating 23 plants.
24 The joint Subcommittee will hear 25 presentations and hold discussions with NRC Staff, the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
5 1 SNC representatives, and other interested parties 2 regarding this meeting.
3 Part of today's discussion may be closed 4 in order to discuss information that's proprietary to 5 the licensee and its contractors pursuant to 5-USC-6 552-BC4.
7 Attendance during the meeting that deals 8 with such information will be limited to the NRC Staff 9 and its consultants, SNC, and those individuals and 10 organizations that have entered into appropriate 11 confidentiality agreements with them.
12 Consequently, we need to confirm that we 13 only have eligible observers and participants in the 14 closed part of this meeting. The rules for 15 participation in all ACRS meetings were announced in 16 the Federal Register on June 13, 2019.
17 The ACRS Section of the U.S. NRC public 18 website provides our charter, bylaws, agendas, letter 19 reports and full transcripts of all full and 20 Subcommittee meetings, including the slides that will 21 be presented therein.
22 The notice and agenda for this meeting 23 were posted there, we received no written comments or 24 requests to make an oral statement to the public. The 25 joint Subcommittee today will gather information, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
6 1 analyze relevant issues and facts, and formulate 2 proposed positions and actions as appropriate for 3 deliberation by the full Committee if we recommend at 4 the end of this meeting to have a subsequent full 5 Committee meeting and issue a letter of report on this 6 topic.
7 A transcript of the meeting is being kept 8 and will be made available, as stated in the Federal 9 Register notice. Due to the COVID-19 pandemic, 10 today's meeting is being held over Microsoft Teams for 11 ACRS, NRC Staff, and licensee participation.
12 There's also a telephone bridge line 13 allowing participation by the public. When addressing 14 the Subcommittee, participants should first identify 15 themselves and speak with sufficient clarity and 16 volume so that they may be readily heard.
17 When not speaking, we request that 18 participants mute your computer microphone or phone, 19 and I'm going to emphasize again today that this is 20 especially important for members of the public on the 21 bridge line.
22 If there's noise from the public bridge 23 line it will be disconnected and you're going to have 24 to rely on receiving the transcript after this 25 meeting. We're now going to proceed with this NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
7 1 meeting.
2 The agenda indicates that we're going to 3 be starting with SNC presenting but I would like to 4 first offer the opportunity to NRR Staff if they would 5 like to make any opening remarks?
6 Does anyone from NRR want to make any 7 opening remarks? I'm not hearing anyone so I will ask 8 is it Ryan Joyce to begin from SNC?
9 MR. JOYCE: Yes, thank you and Cheryl 10 Gayheart will actually kick us off.
11 CO-CHAIR REMPE: Ryan, I can't hear you, 12 can you turn your volume up?
13 MR. JOYCE: Yes, is that better?
14 CO-CHAIR REMPE: Yes, thank you.
15 MR. JOYCE: Cheryl Gayheart will be 16 kicking us off.
17 MS. GAYHEART: Good afternoon, my name is 18 Cheryl Gayheart and I'm the Southern Nuclear Fleet Reg 19 Affairs Director.
20 Southern Nuclear appreciates the 21 opportunity to address the ACRS Subcommittee regarding 22 the Vogtle 12 risk-informed amendment for resolution 23 of Generic Letter 2004-02.
24 As you know, significant NRC and utility 25 resources have gone into ensuring that we have the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
8 1 right approach to addressing the generic letter.
2 The risk-informed approach to be described 3 today coupled with the plant actions and modifications 4 already in place and the NRC Staff safety evaluation 5 provides the rationale for closure for this generic 6 letter for Vogtle 12. You will hear from the NRC 7 Staff later today.
8 CO-CHAIR REMPE: Could I interrupt just to 9 start off the ball rolling? Because that's what ACRS 10 Members do, we interrupt and ask questions.
11 Could you be very specific in telling me, 12 because if we do write a letter on this I want to make 13 sure I draft the letter appropriately, is there one 14 request or two requests?
15 You have two units and then there's 16 something with respect to the exemption versus the 17 actual approach to GS-191. So, is it a single request 18 or is it multiple requests?
19 MR. JOYCE: This is Ryan Joyce. We 20 submitted one letter last August, that letter 21 contained both an exemption request and a license 22 amendment request.
23 The license amendment request was to 24 implement our risk-informed methodology as described 25 in our technical report, which we will be discussing, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
9 1 and also to cement a list of travel regarding the 2 containment tech spec.
3 So, there is an exemption request and a 4 license amendment request, of which the license 5 amendment request is a two-part request.
6 CO-CHAIR REMPE: Thank you very much, this 7 helps.
8 MR. LAMB: This is John Lamb with the NRC, 9 I just wanted to let you know it would be one letter 10 request if ACRS feels they need to write one. The NRC 11 Staff does not need a letter request but we're leaving 12 it up to ACRS whether they wish to have one or not.
13 Yes, there's two amendments because it's 14 one for each unit but it's one action, and that action 15 of the amendment will cover the exemption because the 16 risk-informed methodology is being approved in the 17 amendment and then would carry to the exemption.
18 CO-CHAIR REMPE: So, at the end of this 19 meeting I will ask the Members their opinion on 20 whether we need a letter. And then I will do my best 21 to appropriately characterize how it should be raised.
22 (Simultaneous Speaking.)
23 -- do it specifically, advise it should be 24 done. Thank you.
25 MR. LAMB: It would be one letter.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
10 1 MS. GAYHEART: For the meeting agenda, the 2 Southern Nuclear team who will lead today's 3 presentation includes Matt Horn, the Vogtle 12 Reg 4 Affairs Manager, and Ryan Joyce, the Fleet Lead 5 Licensing Engineer for the GSI-191 submittal.
6 Other Southern Nuclear and Entercon team 7 members are present today who may be called upon to 8 address any questions that may come up during the 9 presentation and we will ask that they introduce 10 themselves when called upon.
11 Again, thank you in advance for your 12 insights and feedback regarding today's presentation.
13 With this, I will turn it over to the Southern Nuclear 14 team for our presentation, and that's Matt Horn.
15 MR. HORN: Thank you, Cheryl, and good 16 afternoon, my name is Matt Horn, I am the Vogtle 12 17 Regulatory Affairs Manager.
18 I will provide some background information 19 regarding Vogtle, followed by a timeline for the 20 Vogtle as well as the South Texas project, 21 risk-informed resolution for the Night Lab 200402.
22 Afterwards, my colleague, Ryan Joyce, will present the 23 rest of the items listed on the agenda.
24 So, now the background information, 25 Vogtle, 1 and 2 that is, is a Westinghouse floor water NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
11 1 reactors with large dry containment buildings. Each 2 unit has two trains of emergency core cooling and two 3 trains of containment spray.
4 Each train of emergency core cooling has 5 three pumps, residual heat removal for low-head 6 injection, safety injection pumps for medium-head 7 injection, and trickle recharging pumps for high-head 8 injection.
9 Initially, each ECCS pump as well as the 10 containment spray pumps will take suction from their 11 cooling water storage tank, and when the plant 12 transitions to the recirculation phase for the 13 emergency cooling system, the safety injection and 14 trickle recharging pumps will take suctions from the 15 RHR pump discharge.
16 In other words, during the recirculation 17 phase, the low-head system will provide suction to the 18 high and medium-head systems. As you can see, the 19 design flow rates of the RHR containment spray pumps 20 are 3700 and 2600 gallons per minute respectively.
21 In addition to containment spray, the 22 containment building has two trains of air cooling as 23 well.
24 CO-CHAIR REMPE: So, I had a question 25 about Units 1 and 2. It's my understanding that there NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
12 1 were some differences, for example, there's just one 2 boron injection tank and I believe that's tied to 1, 3 so correct me if I'm wrong.
4 But are there any other differences 5 between the two units that might be relevant to GSI-6 191? Are the coatings on all the pipings similar, for 7 example?
8 MR. HORN: I can address the boron 9 injection tank. So, you have one that does have an 10 installed boron injection tank, however, that tank is 11 strictly just, basically, a large section of pipe in 12 the discharge flow path that no longer contains boron 13 that's injected.
14 All the borated water comes from a 15 refilling water storage tank, so there in the design 16 and construction of Unit 2 we decided to just remove 17 that or not install that tank at all.
18 So, essentially, the Unit 1 boron 19 injection tank just serves as a piece of pipe so there 20 really is no difference as far as the emergency core 21 cooling system as far as how they operate during a 22 LOCA or loss of cooling accident.
23 CO-CHAIR REMPE: Thank you, what about the 24 question with the coatings?
25 MR. JOYCE: This is Ryan Joyce, we'll be NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
13 1 discussing differences between the two units as 2 pertains to GSI-191 in a few slides from now.
3 CO-CHAIR REMPE: Okay, great.
4 MR. HORN: All right, so each turning of 5 the RHR zero heat remove and containment spray has its 6 own independent containment suction strainer for a 7 total of four containment suction strainers per unit 8 inside the containment.
9 An example of the design of each of the 10 four is pictured here in the slide with the relative 11 locations shown in blue. In 2008, all suction 12 strainers were replaced with a stacked disc design 13 strainer as shown.
14 The RHR strainers are 4.9 feet tall with 15 765 square feet of surface area. The containment 16 spray strainers are 4 feet tall with 590 square feet 17 of surface area. Each strainer plate has design holes 18 of 3/37 of an inch.
19 The first change to align with our 20 risk-informed approach will modify the RHR substance 21 strainer to remove the top two plates and give an 22 overall height of 4.4 feet and 678 square feet of 23 surface area.
24 And the logic behind the proposed 25 modification in reducing the HRH strainer will be NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
14 1 covered later in the presentation.
2 So, I'll provide some background 3 information on plant response to loss of coolant 4 accident. So, the plant responses to the RCS LOCA, 5 initially, the ECCS accumulator will injury big or 6 small-break LOCAs.
7 As stated earlier, all three of the ECCS 8 pumps will inject the substance initiated in the 9 heating storage tank to the RCS cold legs.
10 Containment spray is also initiated from the re-11 cooling water storage tank.
12 At the low-level alarm set point, each 13 train of HRH has a motor operator valve that will open 14 to provide suction from the containment sumps.
15 The license operator will then take the 16 actions to align the high-head and medium-head 17 suctions from the recooling water storage tank to the 18 RHR pump discharges.
19 At the recooling water storage tank empty 20 alarm set point, operators will close the path from 21 the recooling water storage tank and RHR and all of 22 the ECCS will now be supplied only from the 23 containment sumps.
24 Also at this time, the R dose T empty set 25 point, the operators will swap containment spray over NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
15 1 from the recooling water storage tank to its 2 associated containment sump as well. The tank spray 3 will be operated for a minimum of two hours, one and 4 a half hours of which will be in a recirculation 5 phase, as required by procedure.
6 And at seven and a half hours into the 7 event, operators will align the RHR pumps from cold 8 leg recirculation over to hot leg recirculation.
9 So, why a risk-informed approach? May 10 2013, Vogtle submitted a letter to inform the NRC of 11 the choice to use the risk-informed approach.
12 And at that time, we've taken several 13 actions to address some concerns such as- it's all in 14 the new containment suction strainers we discussed 15 earlier, to increase the surface area of strainers.
16 Also, install new orifice plates to allow 17 the ECCS throttle valves to open further and mitigate 18 ex-vessel downstream effects. All balance per NEI O2-19 01 were completed along with debris sampling and 20 characterization.
21 We removed some installation that could 22 become debris sources and established programming and 23 procedure changes to protect the new licensing basis.
24 This left Vogtle with some significant 25 challenges as we continued down the deterministic NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
16 1 path. So, therefore we changed to go down to a 2 risk-informed approach. Some of the other challenges 3 that still remain was maximum strainer head loss from 4 the deterministic approach which succeeded the section 5 head limits.
6 Testing and analysis showed that fiber 7 penetration and quantity exceeded 15 grams per fuel 8 sampling. Also, meeting the 10 C.F.R. 5046 emergency 9 core cooling system requirements deterministically 10 would require new bounding analyses and significant 11 installation removal or replacement.
12 For perspective, we can utilize the South 13 Texas analysis due to their very similar Westinghouse 14 four-loop design and estimate between 4000 and 5000 15 cubic feet of installation would have to be removed or 16 replaced.
17 And that's per each reactor. The total 18 estimated dose for that alone would be 100 REM per 19 each unit and 300 REM total. So, significant dosing 20 challenges continue deterministically in the approach.
21 Now, I just want to present a timeline of 22 the South Texas and Vogtle risk-informed resolution 23 approach. These next two slides will cover the 24 milestones to the Generic Letter 2004-02.
25 I'm going to cover the items specific to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
17 1 Vogtle, you can also see some other industry 2 milestones on the blue timeline. In 2005, Southern 3 Nuclear submitted a combined response for plants in 4 Farley and Vogtle.
5 In 2008, Vogtle replaced our sump 6 strainers with the large surface areas and submitted 7 our generic letter submittal using a deterministic 8 approach. In 2007 and 2008, the NRC published 9 guidance in different areas for GSI-191 testing and 10 analysis, and we completed strainer head loss testing 11 in 2009.
12 In 2010 the Staff requirement memorandum, 13 the NRC Staff presented three options to closing GSI-14 191 including developing a risk-informed 15 implementation guide in SECY-10-113.
16 South Texas started a pilot program using 17 the risk-informed approach for their generic letter 18 submittal, which culminated in a 2013 submittal for 19 their initial LAR.
20 That same year, 2013, Southern Nuclear 21 formulated for the NRC the risk-informed approach that 22 would be used for Vogtle, and after an ACRS meeting on 23 the South Texas project submittal and receiving REIs 24 from the NRC, South Texas decided to switch over to 25 the RoverD Method in late 2014.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
18 1 In 2015, South Texas revised the LAR 2 submittal using the RoverD Method. That same year, 3 Vogtle performed a new fire and penetration testing at 4 Alion Labs. In 2017, the NRC issued the final safety 5 evaluation on the South Texas LAR submittal.
6 Vogtle had also made some significant 7 progress during this period and in 2016 the NRC 8 performed audits of Entercom, NARWHAL, and BADGER 9 software. The Vogtle analyses were used as examples 10 in the NRC audit.
11 In 2017, Vogtle submitted a technical 12 report summarizing the risk-informed approach. This 13 did not include any licensing change request at the 14 time because the in-vessel methodology was still being 15 reviewed by the NRC.
16 In 2017 and 2018 the NRC performed audits 17 on the Vogtle technical report as well as the software 18 used by Vogtle. We responded to a few rounds of REIs 19 from the NRC.
20 July of 2018, Vogtle submitted an updated 21 version of the technical report which incorporated a 22 response to the RAIs as well as audit questions.
23 In 2019 the NRC issued a Staff evaluation 24 of the technical report, and in mid-2020 Vogtle 25 submitted a LAR to adopt the risk-informed approach NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
19 1 for Generic Letter resolution as well as TISKA 567.
2 Late last year, Vogtle responded to the 3 NRC RAIs for the LAR submittal and that takes us to 4 our current time.
5 As you can see, obviously the resolution 6 has taken several years due to a few different 7 factors, such as updated guidance for resolutions and 8 GSI-191, the development of models and methods for the 9 software analyses and the complex and technical 10 reports required for evaluation, as well as additional 11 requests for information.
12 MR. JOYCE: Thank you, Matt, and this is 13 Ryan Joyce, I will be discussing and going through the 14 remainder of the presentation. So, I'd first like to 15 give an overview of our risk-informed approach.
16 So, the slide presents a high-level 17 summary of the Vogtle's risk-informed approach.
18 Additional details will be provided in the 19 remainder of the presentation so the high-level 20 summary, as will be discussed, our risk-informed 21 approach is similar to STP, similarities and 22 differences will be discussed later in the 23 presentation.
24 For the evaluation of risk attributable to 25 debris, we used physical models that have been used in NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
20 1 the past. We'll be discussing our BADGER software to 2 postulate multiple breaks for the various wells.
3 We'll be discussing our NARWHAL software, used to 4 determine any kind of strain and/or in-vessel failures 5 due to defection debris.
6 Our NARWHAL software also calculates, as 7 we'll be discussing, conditional failure probabilities 8 and we'll use inputs to our PRA to determine the 9 impacts of the delta core damage frequency and delta 10 large early release frequency.
11 And as will be shown, the results from 12 this assessment show our risk associated with LOCA-13 generated debris is very small and within the Reg 14 Guide 1174 Region 3.
15 So, this slide presents a review of the 16 Vogtle risk-informed approach. This visualizes the 17 process for the risk-informed evaluation for the 18 effects of debris that were described in the previous 19 slide.
20 The four charts show various analyses, 21 tested programs, and submittals all linked together.
22 A Vogtle containment CAD model is first used for the 23 debris generation analysis.
24 This debris generation analysis along with 25 the subvolume pool volume calculation provide input to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
21 1 the debris transport and chemical effects analysis.
2 The analysis will then inform the strainer head loss 3 and fiber penetration testing.
4 Next, the analysis and testing provide 5 inputs to NARWHAL, which evaluates each break against 6 various failure criteria as shown in the diamond. The 7 failure criteria was derived from different sources 8 such as strainer testing, NRC regulatory guidance, and 9 the Westinghouse WCAD reports.
10 The condition of failure of probabilities 11 determined by NARWHAL increases the risk 12 quantification to calculate change, delta CDF and 13 delta LAR due to the debris effects.
14 And as will be discussed, a few founding 15 analyses were performed outside of NARWHAL for all 16 breaks. The following slide describes acceptance 17 criteria evaluated either in NARWHAL or outside of 18 NARWHAL using bounding evaluations.
19 As shown in the previous slide, for break-20 specific analyses, NARWHAL evaluated each break 21 against various failure criteria, a determined break 22 would result in any strainer or reactor core failures.
23 The failure criteria for break-specific 24 analyses are listed in the first four rows of the 25 table and I'll relay the strainer head loss NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
22 1 degasification to flashing and vessel fiber loading.
2 The last four rows in the table show the 3 failure criteria that were analyzed in the bounding 4 analyses. As previously mentioned, these were 5 analyzed outside of NARWHAL and these failure criteria 6 will be discussed in more detail in a later slide.
7 Next, I'd like to discuss plant 8 modifications, testing and analyses performed for a 9 risk-informed approach.
10 So, based on some of the risk insights, 11 Vogtle decided to make two additional plant 12 modifications based on the risk assessment results, 13 PRA insights, fiber penetration sensitivity results, 14 and other constraints.
15 The first change is the emergency 16 operating procedure was modified to continue the RWST 17 drain-down by the RHR pumps for the RWST empty level 18 set points.
19 In addition, the RHR strainer height will 20 be reduced by removing the top-two desks. As shown on 21 the slide, these modifications or plans for the 22 upcoming Unit 1 and Unit 2 outages.
23 Implementing these modifications results 24 in more breaks with fully submerged strainers, which 25 reduces failures due to strainers being partially NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
23 1 submerged. As a result, this lowers overall risk and 2 provide additional safety margin.
3 CO-CHAIR REMPE: I have a question about 4 after you complete these modifications in fall of 2021 5 and spring of 2022. What's the process that occurs 6 between SNC and the NRC to verify that they were 7 completed as described in your submittal?
8 MR. JOYCE: For one, as far as our license 9 amendment request, we have a implementation time which 10 we specified in our request that we would implement.
11 Maybe the NRC could help me out with 12 exactly the verbiage we use but we meant within a 13 short timeframe after these modifications were 14 performed.
15 So, in essence, the NRC safety evaluation, 16 should we receive an NRC safety evaluation, it would 17 require us to make these modifications and we have to 18 get back to the NRC should they not have been made by 19 this time.
20 MR. LAMB: This is John Lamb at the NRC.
21 As part of the license, there would be a condition of 22 the implementation that would require them to 23 implement that.
24 Otherwise, the amendment would become null 25 and void if they failed to do that.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
24 1 CO-CHAIR REMPE: Does anybody check it?
2 Is it something that goes to a regional inspector so 3 he knows or she knows to go and make sure they did it?
4 Or does SNC send a letter saying we did what we 5 promised to NRC and that's it.
6 MR. LAMB: That would be a follow-up 7 inspection by the resident inspector and they would 8 put it in their report. They're aware of this so if 9 it gets approved then they would become aware of this 10 and they will put it on their schedule.
11 CO-CHAIR REMPE: Thank you.
12 DR. SCHULTZ: Ryan, this is Steve Shultz.
13 Are you going to describe later the quantification of 14 the benefit that's achieved with these modifications?
15 MR. LAMB: I have a back up slide that I 16 can show for that.
17 DR. SCHULTZ: That would be great, thank 18 you.
19 CO-CHAIR DIMITRIJEVIC: This is Vesna, 20 there is also some drawbacks of these modifications.
21 For example, in the PRA switch over to decoration of 22 human action, would this affect changing the level of 23 situation for a job?
24 Would that affect the time variable for 25 that situation?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
25 1 MR. JOYCE: This is regarding the operator 2 action change?
3 CO-CHAIR DIMITRIJEVIC: Yes, so currently 4 in your PRA you have operators switch over to 5 matriculation for LOCAs, it's a very common and 6 important action in every PRA.
7 My question is does it change or switching 8 that agenda on other BSC empty level set point, does 9 that affect the reliability of this action?
10 MR. JOYCE: Yes, that was looked into.
11 Tim Sande from Entercom, are you able to speak right 12 now?
13 MR. SANDE: Hi, can you hear me?
14 MR. JOYCE: Yes.
15 MR. SANDE: This Tim Sande, I don't 16 actually know the answer to your question, Vesna.
17 Let's see, I think Haifeng Li had looked into that.
18 Haifeng, are you available to answer the question?
19 MR. LI: Yes, this is Haifeng Li with 20 Entercom representing SNC. As part of the 21 modification, the operator action time was looked 22 into.
23 The minimum operator action time was 24 determined to be about three and a half minutes, and 25 this time has been incorporated into the site time-NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
26 1 critical operator action program. And this program 2 will validate and ensure the operator action can be 3 performed really in the timeframe allowed.
4 CO-CHAIR DIMITRIJEVIC: Did you change 5 your AGPs in the PRA to reflect this?
6 We are discussing that positive and 7 negative effects of this, so this is one of the 8 negative effects and I was just wondering, did you 9 update the PRA to reflect this? That's my question.
10 MR. LI: I do not have an answer to that 11 question.
12 MR. GILBREATH: This is Alex Gilbreath 13 from SNC, risk-informed engineering. The Vogtle base 14 model does have this operator action modeled and the 15 timing for the cue reflects the updated set point.
16 CO-CHAIR DIMITRIJEVIC: That was my 17 question. So, did you see any increase in your CDS 18 through that?
19 MR. GILBREATH: I'm not sure of the answer 20 to that question.
21 CO-CHAIR DIMITRIJEVIC: Maybe you can get 22 to ask later.
23 MR. GILBREATH: Sure.
24 CO-CHAIR REMPE: Let me explore this a 25 little more. You modeled this in the PRA and then you NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
27 1 guys report change in core damage frequency and change 2 in worth. So, you're small, we already know the 3 punchline.
4 CO-CHAIR DIMITRIJEVIC: This is just 5 related to debris, this change is only related to 6 these incorporated changes, it's not connected with 7 the GSI-191. It's completely independent.
8 CO-CHAIR REMPE: You're right, thank you, 9 Vesna, for clarifying that. Yes, it would be 10 interesting to know that answer.
11 MR. POURNIA: Yeah. This is Faramarz 12 Pournia. I'm the Risk Informed Engineering Director 13 here at Southern Nuclear.
14 Response to that question, that issue 15 about the operator actions time from the TTOA and 16 other elements that we're going to talk through these 17 presentations, we are having processes in house for 18 change management review impacting TRA.
19 So, questions like that, it was a very 20 question that was asked. And other questions that are 21 going to come later on.
22 They're all being controlled to our change 23 management process that we do by screening and 24 tracking to a model off change, and then picking the 25 changes up in our PRA model during the upbeat cycle.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
28 1 Just genetically to identify that there's 2 some question similar to this. Thank you.
3 MR. JOYCE: Does that address the question 4 regarding the PRA impacts of the procedure change?
5 CO-CHAIR DIMITRIJEVIC: Well, I didn't 6 exactly get what this last response told me. It says, 7 what -- what did you say that you track these change 8 -- how did that go?
9 MR. POURNIA: Yeah. What I was going to 10 say, first of all, you just don't make one change and 11 so the results of the PRA response or many changes 12 occur to the models before you see the change in the 13 CDF and LERF.
14 But also, response that whenever the MOD 15 process taking place, as was mentioned, and the 16 changes that are occurring to the procedures, then we 17 do the impact review for PRA.
18 And we pick those changes in the PRA 19 model.
20 CO-CHAIR DIMITRIJEVIC: And I understand, 21 you have your regular lab dates and everything. I 22 just want to say this in my experience, this is one of 23 the most important actions in PWRs.
24 And you have the fact that you are here 25 violently by changing the level of actuation. So, I'm NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
29 1 just curious, I mean, you know, how the -- that the 2 fact that you're -- there could be other results.
3 That was my question. I assume, I don't 4 know, did you get the, your regular lab date or in the 5 email, what schedule you are on.
6 But, I was curious, did you see change in 7 the PRA results do that? So.
8 MR. GILBREATH: Yeah. And I don't -- for 9 that one specific change, I don't know the specific 10 impact of that one change taken independently of other 11 changes that were made.
12 But, I can look into it. But, I will say 13 that this allows the operator more time to carry out 14 the operator action.
15 CO-CHAIR DIMITRIJEVIC: I thought it 16 allowed him less time? If your set point is to the 17 empty levels section. I mean, that's what I thought.
18 But, you know, because you -- okay. All 19 right. Let's just forget. Let's move out. We were 20 discussing GSI 191. Okay.
21 MR. JOYCE: And to clarify, this was a 22 change made about, I think Matt Horn could probably 23 correct me, but two or three years ago.
24 But, it was amended again a few years ago.
25 And so the operator actions using 159 process, and NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
30 1 showing that -- that this non-criteria could be met 2 for any kind of revised operator actions.
3 MR. HORN: Yeah. Ryan, that's correct.
4 This is Matt Horn at Vogtle. So, we've had this 5 action to swap, or actually, it's just to close off 6 the arch and all to the refueling storage tank at the 7 empty level alarm.
8 All the other operator actions are 9 completed well ahead of that time. So, there will be 10 no challenge to the completing of the actions from an 11 operator's standpoint.
12 I was formerly last at Vogtle, so, I did 13 this in the simulator many times.
14 CO-CHAIR DIMITRIJEVIC: Okay. And thank 15 you.
16 DR. SCHULTZ: And then to summarize the 17 change on the strainer, this is just -- this is 18 something that was identified as an impro -- a safety 19 improvement that you discovered while doing the PRA, 20 related to the GSI 191.
21 You're going to make the modification 22 independent of the overall evaluation of the debris.
23 Correct?
24 MR. JOYCE: Well, the plan is to get NRC 25 approval for the risk informed methodology. And then NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
31 1 make the modifications.
2 And as far as, you know, what's discussed, 3 we did show a benefit to the PRA, in the PRA space, 4 based on these two modifications about increasing the 5 RWST drain down and these trainer modifications.
6 Scoping studies show that the -- these two 7 changes reduced the delta CDF by about a factor of 8 three. So again, we believe these are, you know, 9 definitely quantifiable benefits based on these 10 changes.
11 And also, I'd like to point out also, back 12 in 2008, we made a big strainer modification. The 13 strainer was originally only 54 cubic -- 54 square 14 feet.
15 DR. SCHULTZ: Um-hum.
16 MR. JOYCE: And with the modifications 17 made, the strainer will have a surface area of 678.
18 So again, I think, I believe 12 times.
19 Once the modification (inaudible) 20 originally licensed to be.
21 DR. SCHULTZ: Thanks for that detail.
22 Appreciate it.
23 MR. JOYCE: All right. Well, I believe 24 we've covered this slide. Any other questions on this 25 slide?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
32 1 (No response) 2 MR. JOYCE: All right. So, as previously 3 discussed, we had -- we developed a CAD model based on 4 the Unit One site, or based on the Unit One 5 containments.
6 However, is asked about the similarities 7 between the Units. First off, the concrete and major 8 equipment are identical between the two Units.
9 Some small bore piping maybe slightly 10 different. However, these differences will not change 11 the overall eval -- results of the overall evaluation.
12 The CAD model that was done was validated 13 by comparison of later scans. And then was used for 14 various calculations, as will be discussed, 15 specifically regarding the debris generation and 16 debris transport.
17 Any other questions on the similarities 18 and differences between the two models? The two 19 Units?
20 (No response) 21 MR. JOYCE: So, this slide talks about the 22 debris generation methodology. The debris generation 23 methodology is consistent with deterministic analyses 24 that have been extensively reviewed and approved by 25 the NRC.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
33 1 The debris generation from insulation, 2 fire barrier, and qualified coatings, was analyzed in 3 software developed by Enercom, the BADGER software, 4 used in the CAD model.
5 This model analyzed breaks from each class 6 when weld, including double-ended guillotine breaks, 7 and partial breaks.
8 The zone of influence, size and shape in 9 BADGER are based on the guidance in NEI 0407. This is 10 illustrated graphically on the next slide.
11 Unqualified coatings, latent debris, and 12 miscellaneous debris is not break specific. And was 13 determined via coating logs and walk down.
14 This figure shows the weld locations where 15 breaks were postulated. There are 413 welds inside 16 the first isolation valve shown on -- shown as red 17 dots on the slide.
18 A range of break sizes and orientations 19 was postulated for each break, resulting in a total of 20 over 18,500 breaks analyzed.
21 MEMBER HALNON: Okay. And just before you 22 go, could you go back to that last slide? This is 23 Greg Halnon. I have a couple of questions just real 24 quick.
25 The quality assurance program, QA program NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
34 1 for Enercom, did you guys do a review of the 2 corrective action program at Enercom and ensure there 3 were no discrepancies or deficiencies in the BADGER 4 before you used it that could have affected your 5 specific application?
6 MR. JOYCE: Ted, can you address the QA 7 review of Enercom software?
8 MR. SANDE: Sure. Yes, this is Tim Sande 9 with Enercom. The -- so the team that was preparing 10 this calculation was the same team that developed the 11 software.
12 And it was being used specifically for GSI 13 191 regeneration calculations. So, all -- any 14 corrective actions associated with the software were 15 items that the team developed in the calculation, 16 we're intimately familiar with.
17 MEMBER HALNON: Okay. So, there's --
18 there's no issues there. The second question, the 19 last bullet, the coating debris was assumed to be 20 particulate.
21 Is that the most conservative, I guess, or 22 non-conservative, however you want to put it, is that 23 the worst case configuration of that?
24 It just seems like a sheet of coating 25 would be more detrimental than a bunch of small NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
35 1 particulate. Just help me with that understanding.
2 MR. SANDE: All right. In general, 3 coating particular is more conservative, as long as 4 you have a fiber debris bed on the strainer.
5 If you have no fiber debris bed then 6 failure of coatings and as chips, it wouldn't be able 7 to transport as one large sheet. But, if it fails as 8 chips, then those could accumulate on the strainer and 9 cause some head loss.
10 In general, what has been found is that on 11 head loss associated with chips is not all that 12 significant. Head loss is much higher if you have a 13 fiber bed that forms a filter on the strainer that is 14 able to capture a particulate.
15 And then the particulate basically fills 16 in the gaps and creates much higher head loss than 17 chips would.
18 MEMBER HALNON: Okay. So, yeah. That 19 makes sense.
20 MR. SANDE: Okay.
21 MEMBER HALNON: Thank you.
22 MR. SANDE: You're welcome.
23 CO-CHAIR REMPE: So, this is Joy. I guess 24 I'd like to bring this up now since Greg broached the 25 issue of quality assurance.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
36 1 In the initial technical report you 2 submitted, there were some licensing condi --
3 limitations and conditions that the staff imposed when 4 they accepted it.
5 And one of them was about the need to 6 correct an error in your estimated, your wash down of 7 the coating volumes or something, as I recall. Could 8 you talk about this and why it happened?
9 I know it was corrected in the subsequent 10 LAR. But, I'd like to hear you talk a little bit more 11 about it.
12 And I didn't see that in subsequent 13 slides, other than a brief mention that you fixed it.
14 MR. JOYCE: Tim, can you address that, 15 please?
16 MR. SANDE: Yeah. I think, Haifeng, do 17 you remember the details on that? You may be able to 18 speak to that better than I can.
19 MR. LI: Yes. This is Haifeng from 20 Enercom representing S&C. The error was not 21 associated with the software itself.
22 It's a use error that happened when we 23 incorporate the wash down fraction for the uncrossed 24 by coating in the narrow model. Incorrect values were 25 used.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
37 1 That's how the error happened.
2 CO-CHAIR REMPE: So, but this happened, it 3 was not detected in your initial sub -- technical 4 report submittal.
5 So, would -- is your -- did you revise the 6 quality assurance or something so that you won't have 7 similar user errors in the future?
8 MR. LI: Yeah. We followed our corrective 9 action program under our QA program, and issued a 10 corrective action report.
11 And it did a full cause analysis, you 12 know, regarding how the error happened, and what can 13 be done differently to avoid the error.
14 One of the lessons learned was that when 15 doing the modeling, using NARWHAL, the preparer used 16 to save screenshots in the preparation process. And 17 a reviewer, when used to review those screenshots 18 making sure that corrected inputs was actually placed 19 in the model rounds.
20 CO-CHAIR REMPE: Great. Thank you very 21 much.
22 MR. LI: Um-hum. You're welcome.
23 MR. JOYCE: Any other questions on this 24 slide?
25 (No response)
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
38 1 MR. JOYCE: All right. So, this slide 2 here illustrates the break zones for -- break zones of 3 influence for the debris generation analysis.
4 So, the figure on the left illustrates a 5 spherical zone of influence for a proposed double line 6 guillotine break.
7 The figure on the top right illustrates a 8 hemispherical zone of influence for partial break at 9 a given orientation.
10 The figure on the bottom right illustrates 11 the same partial break, size, and location. But 12 includes eight analyzed break locations at the same 13 weld.
14 The zones of influence are illustrated as 15 overlapping hemispheres. The zone of influence size 16 -- the zone of influence size varies with break size 17 and debris type.
18 For the containment sump water level, the 19 sump pool that's evaluated on a break specific basis 20 using conservative inputs that minimize water level.
21 The evaluation considers break location 22 and size for the contributions of RWST, reactor 23 coolant system, and SI accumulators at full volume.
24 So, as we discussed -- as we'll discuss at 25 the start of the switch over to recirculation, the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
39 1 modified strainers with the modifications made, are 2 fully submerged small breaks except for large reactor 3 cavity breaks to actuate containment sprays.
4 And by the time the switch over -- the 5 switch over to sump recirculation is complete, the 6 strainers are fully submerged for all breaks.
7 The next slide illustrates our debris 8 transports. The debris transport methodology is 9 essentially identical to deterministic analyses that 10 have been extensively reviewed and accepted by the 11 NRC.
12 The main difference in the Vogtle 13 transport analysis is the wider range of break 14 locations and different equipment configurations were 15 included to support the risk informed evaluation.
16 So the debris used logic tree approach is 17 consistent with NEI 0407 guidance. The brought on 18 transports based on containment volumes and things 19 such as flow past through gratings.
20 The washdown transport is based on whether 21 sprays are initiated in items such as the flow pass 22 through the gratings. The full pool transport based 23 on the pool volumes and inactive in sump cavity 24 volumes.
25 Recirculation transport is based on the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
40 1 CDF analyses. And the erosion of small and large 2 pieces of fiberglass debris is based on Alion testing.
3 For strainer head loss testing, so as 4 debris transports to and accumulates on the strainer, 5 it can result in flow resistance/head loss. The 6 debris head loss was based on tank tests performed at 7 Alion in 2009.
8 The test protocol is consistent with the 9 NRC March 2008 guidance. The testing used a strainer, 10 strainer arrangement and debris types prototypical to 11 Vogtle plant strainers.
12 New con insulation was shredded and boiled 13 before being mixed by a paint mixer into finds.
14 Mixing was maintained inside the test tank to prevent 15 -- to prevent the induced debris from settling.
16 The testing also considered chemical 17 effects using pre-made chemical participants following 18 WCAP-16530 approach, similar to approach used by many 19 of the licensees.
20 When applying the test results, the 21 measured -- the measured head loss were extrapolated 22 for 30 days as necessary. And were also corrected for 23 temperature and approach velocity.
24 This is consistent with the NRC guidance.
25 MEMBER HALNON: On this test, this is Greg NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
41 1 again. I was curious, clearly the -- I say clearly.
2 But, this description says that it was pretty much a 3 homogeneous solution that was being developed in that 4 tank to inject.
5 Was there any discussion, or does the WCAP 6 talk about the potential for the homogeneous versus a, 7 pretty much near random solution mix?
8 In other words, you know, could the 9 randomness of a containment mix be different in, you 10 know, such that it could affect any of the results 11 rather than the tank, which was pretty homogeneous?
12 MR. JOYCE: Haifeng, can you please 13 address that?
14 MR. LI: Yeah. This is Haifeng Li from 15 Enercom again. The main goal for the keeping -- well, 16 I think to keep in mind is the debris amount we added 17 to the test tank was already determined to be 18 transportable in the debris transport analysis.
19 One goal that the NRC guidance, and also 20 our, how we do the testing, was to making sure that 21 introduced debris will load onto the strainer.
22 That's why, you know, we were using the 23 various measures, you know, spot juries, you know, 24 making a mechanical mixture, and the manual agitation 25 as needed, to make sure that the introduced debris NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
42 1 will load onto the strainer and, you know, demonstrate 2 the head loss impact.
3 That's the main reason why we're mixing 4 the debris during testing. I have not seen the 5 guidance regarding, you know, random combination of 6 debris types.
7 MEMBER HALNON: Okay. I think I 8 understand. I mean, clearly the loading is the -- is 9 the result of -- in the debris, not necessarily how 10 the debris entered it. Or whether it was chunks or, 11 you know, over time.
12 MR. LI: Right.
13 MEMBER HALNON: But yeah, I can understand 14 that. The second question, when you extrapolated the 15 head loss, is how was that done?
16 Was that -- was there an algorithm? Or a 17 certain test curve that was used? Or how was that 18 extrapolated?
19 MR. LI: Yeah. I can talk to that. So, 20 each head loss test took about three to four days.
21 After we introduced all the debris, typically was 22 around for half a day roughly.
23 You know, the -- during testing we do have 24 a head loss stabilization criteria. We want the head 25 loss to, you know, change by less than 1 percent over NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
43 1 a one hour period of time.
2 And also, we require that the rate of head 3 loss increase, if they do increase, is decreasing over 4 time.
5 What we did for the extrapolation is we 6 take the -- we take the test data recorded, which is 7 the head loss versus time. And then feed that into a 8 logarithm function.
9 And then we shift that curve up to making 10 sure that all of the fluctuation from the recorded 11 head loss was founded. And then we calculate how that 12 head loss will be at the end of the 30 day mission 13 time for the strainer.
14 We did that exercise for all tests, for 15 all individual tests. And the greater, the greatest 16 extrapolation constant was used for -- in NARWHAL.
17 MEMBER HALNON: So, could -- but given 18 that, could you say that the conservatisms you placed 19 in it cause it to be, you know, clearly bounding for 20 the worst case head loses?
21 MR. LI: That's right. And like I said, 22 we actually shifted the curve up, making sure of that.
23 You know, when we record that head loss continues over 24 time, we will have the exact, you know, random 25 fluctuation in the head loss.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
44 1 We made sure most of those peaks were 2 bounded by the curve peak.
3 MEMBER HALNON: Okay. So, even the peaks 4 were below the curve then.
5 MR. LI: That's right.
6 MEMBER HALNON: Okay. Thanks.
7 MR. LI: Um-hum.
8 MR. JOYCE: Any other questions on our 9 head loss?
10 (No response) 11 MR. JOYCE: For fiber penetration testing, 12 the evaluation of in-vessel dose from effect is the 13 results from fiber only penetration tests.
14 Nine different tests were performed at 15 Alden in 2014. And used various combinations of 16 strainer approach velocity, number of strainer discs, 17 and boron/buffer concentration.
18 The overall approach is similar to other 19 tests performed by Alden, and observed by the NRC 20 staff.
21 Fake Nukon was cut and pressure washed 22 into fines following the NEI guidance. Prepared fiber 23 was batched into test tanks with increasing batch 24 sides.
25 Fine micron filter bags were used to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
45 1 collect fiber penetration downstream of the strainer.
2 The test data was used to develop a curve fix, which 3 was applied to evaluate fiber penetration of the plant 4 strainers.
5 And as previously discussed, we performed 6 bounding evaluation for certain criteria, certain 7 failure criteria.
8 For upstream blockage, it was qualitative 9 -- qualitatively evaluated multiple passageways 10 between the containment annulus and inside secondary 11 shif -- and inside secondary shield wall and refueling 12 canal drain. All the flow paths were shown not to be 13 blocked.
14 For ex-vessel downstream effects 15 calculation, the methodology and acceptance criteria 16 of WCAP-16406 -- 16406, and the NRC safety evaluation 17 was used with conservative debris concentrations.
18 This showed acceptable wear, and no 19 blockage of piping or components on this ECCS or 20 containment spray flow paths, such as pumps, valves, 21 heat exchangers, orifices, or spray nozzles.
22 For flow testing, the flow testing 23 evaluation was based on testing performed during 2009 24 strainer head loss testing.
25 No error in training vortexes were NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
46 1 observed during testing, with clean or debris laden 2 strainer, and conservatively lower strainer 3 submergences and higher approach velocities.
4 Next, I would like to discuss the risk 5 quantification for our risk informed approach. So, 6 the earlier slide I just showed summarized the 7 bounding analyses performed, they're applicable for 8 all breaks.
9 These bounding analyses from the previous 10 slide, were performed outside of NARWHAL. The 11 submodels required for break specific evaluations, as 12 listed on this page, are integrated to NARWHAL, such 13 as chemical effects, NPSH margin, and degasification.
14 The NARWHAL software was developed using 15 Enercom's own -- Enercom's QA program. And the Vogtle 16 specific analysis was also developed under Enercom's 17 QA program.
18 So, for the key elements of the NARWHAL 19 analysis, the chemical effects' methodology was 20 consistent with WCAP-6530 with refinements for 21 phosphate catalyzation of aluminum surfaces based on 22 University of New Mexico testing.
23 Break specific chemical for precipitate 24 quantities were calculated using NARWHAL. The 25 aluminum solubility limits based on argon in that NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
47 1 equation, were used to determine break specific 2 precipitation timing.
3 When the concentration exceeded solubility 4 limits, 100 percent of aluminate solution was assumed 5 to precipitate a sodium aluminum silicate. And if 6 solubility limits were not exceeded before 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, 7 100 percent of aluminate solution was seen to 8 precipitate a sodium aluminum silicate in 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />.
9 For strainer head loss testing, I would 10 like to highlight that a failure was recorded if any 11 of the tested debris qualities were exceeded.
12 The total strainer head loss for each time 13 step was calculated by combining the clean strainer 14 head loss, debris head loss, and extrapolated 15 constants as necessary. And a rule-based approach was 16 used to calculate debris head loss based on testing 17 results.
18 For our NPSH margin, there was no 19 containment accident pressure credited. And the 20 impact on pump NPSH required by void fraction, was 21 addressed per methodology of Reg Guide, Regulatory 22 Guide 1.82.
23 For flashing and degasification, a 24 flashing failure was recorded if the pressure to stop 25 the strainer at the top elevation was less than the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
48 1 vapor pressure at the corresponding sump temperature.
2 The void fraction due to degasification 3 was evaluated at mid height of the strainer. A 4 degasification failure was recorded if the void 5 fraction at the pump suction was greater than 2 6 percent.
7 Air bubbles that formed downstream of the 8 strainer were assumed to stay intact as they traveled 9 to the pump suction without creating any bubble 10 compression.
11 And to preclude that flashing and 12 degasification, the analysis created a small amount of 13 accident pressure, three and a half psi for a short 14 period of time during the first two and a half hours 15 of the event.
16 For the fiber penetration, the fiber 17 penetration curve fit from testing was used with a 18 time dependent analysis to quantify fiber debris 19 reaching the reactor core.
20 This used WCAP-1778 revision one 21 methodology and acceptance criteria in accordance with 22 the NRC 2019 guidance on in-vessel effects. And in 23 accordance with the NRC guidance, no alternate flow 24 paths and reactor barrel baffle region were credited.
25 So, this slide illustrates the tracking of NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
49 1 -- the tracking of movement of flow and debris. In 2 the NARWHAL model, the key components required for the 3 accident mitigation are modeled and arranged 4 consistent with the site emergency operating 5 procedures.
6 Different plant states, such as injection 7 for the RWST, and different stages to recirculation, 8 and full recirculation, are also modeled in NARWHAL.
9 As shown in the flow diagram here, NARWHAL 10 tracks flow through the system, as well as transport 11 and accumulation of debris. So, as shown in this 12 slide, it starts with the RWST and from there it goes 13 to ECCS or containment spray pumps.
14 And it tracks -- it tracks the flow of 15 this flow and debris through the reactor vessel, 16 through the break, spray nozzles, into the sump pool.
17 And then once recirculation starts, through the ECCS 18 and containment spray strainers.
19 So, for the analysis of the condition of 20 failure probability, NARWHAL analyzes each partial new 21 break in a time dependent manner against all failure 22 criteria to determine if a break would have resulted 23 at any strainer and/or reactor core failures due to 24 the impact of LOCA generated debris.
25 Some different equipment line ups for NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
50 1 different pump failures were analyzed. The overall 2 frequen -- LOCA frequencies were allocated across the 3 analyzed breaks using a top down method, which 4 essentially distributes the frequency equally across 5 breaks of a given size.
6 The key outputs from NARWHAL are the 7 conditional probabilities, which are reported for 8 different equipment line ups based on its 9 classification, for example, small, medium, large 10 LOCAs, and their own failure mechanisms, for example, 11 strainer failure or for a blockage.
12 NARWHAL also outputs detailed break 13 specific and time dependent information that can be 14 used to check -- that can be used to check and verify 15 the calculations, identify potential refinements and 16 others, other actions.
17 For secondary side breaks, the risk 18 quantification considers secondary side breaks and 19 side containments.
20 Secondary side breaks and side 21 containments could result in a consequential LOCA due 22 to a stuck open power operated relief valve, or loss 23 of auxiliary feed water, which would require a feed 24 and bleed operation.
25 The ECCS flow rates to recirculation were NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
51 1 based on PRA being released to capacity. Which is 2 much lower than the flow rate for large -- for large 3 breaks on the primary side.
4 So, for these breaks, they're evaluated in 5 a manner consistent -- with a manner similar to 6 primary side breaks with the following exceptions.
7 The breaks postulated at five foot 8 intervals are main steam and feed water piping. All 9 breaks assumed to be double and engaging breaks, 10 smaller zone of influence sizes, these reduce pressure 11 and temperature in accordance with the NEI-0407 12 guidance, lower ECCS flow rates, and it was 13 conservatively assumed that both trains of containment 14 spray would initiate for all breaks.
15 The main stream line breaks and feed water 16 line breaks were evaluated separately in NARWHAL for 17 five different equipment configurations. None of the 18 feed water breaks produced a sufficient quantity of 19 debris to fail.
20 And main steam line breaks resulted in 21 some failures for single train operation only if both 22 containment spray pumps fails -- pumps fail.
23 For the risk quantification, the Vogtle-24 based PRA model was modified to incorporate events for 25 GSI-191 sump strainer and core blockage failure, with NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
52 1 the associated LOCA initiating events and equipment 2 configurations.
3 Sump strainer failures were added into the 4 sump recirculation lodge gates, and the core blockage 5 failures were at or under the CDF top gates.
6 The conditions of failure probabilities 7 are up to the PRA models to calculate delta CDF and 8 delta LERF associated with a high likelihood 9 configuration.
10 Delta CDF and delta LERF were also 11 quantified using bounding or conservative methods for 12 unlikely equipment configurations, secondary disc 13 LOCAs and secondary side breaks.
14 And the risk quantification results for 15 the base CDF and LERF values as show compared to the 16 Reg Guide 1174 acceptance criteria are shown in this 17 slide.
18 As can be seen on this slide, delta CDF 19 from the two units is about 2.8, 10 to the minus 8.
20 The delta LERF is about 3.2, 10 to the minus 10.
21 Both the delta CDF and delta LERF values, 22 as I mentioned are orders of magnitude within the 23 region three risk guidance, region three.
24 Next, I'd like to discuss the sensitive 25 and uncertainty analysis performed.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
53 1 CO-CHAIR DIMITRIJEVIC: I have a couple of 2 questions, Bob, on the previous slide. Sorry it took 3 me to get to my microphone. So you modify your PRA 4 model to incorporate this condition of fiber 5 probabilities from NARWHAL for different, for example, 6 LOCA sizes, right?
7 MR. JOYCE: Yes, that would be correct.
8 CO-CHAIR DIMITRIJEVIC: Okay. So how many 9 LOCA sizes you have added to your PRA model, or are 10 you just failing the same basic, you know, small, 11 medium, large?
12 MR. JOYCE: Tim, can you address that 13 question, please?
14 MR. SANDE: Sure. This is Tim Sande.
15 Yes, that's correct. We just left the three LOCA 16 sizes in the model, and we did not add additional LOCA 17 sizes for the GSI-191 application.
18 CO-CHAIR DIMITRIJEVIC: So then you have 19 large LOCA in your PRA model, that just implies larger 20 than six inches?
21 MR. SANDE: Yes, that's correct.
22 CO-CHAIR DIMITRIJEVIC: And you didn't 23 analyze this because, if I remember, you just show us 24 that, for larger than 13 inches, it's assumed 25 guarantee failure?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
54 1 MR. SANDE: No, we are not assuming all 2 breaks larger than 13 inches fail. What we're doing 3 is we're calculating conditional failure probabilities 4 where the condition is an equipment configuration, for 5 example all pumps running or single train failure, and 6 the LOCA size. So we say conditioned on a large LOCA 7 occurring and let's say single train operation, what 8 is the probability that you would fail due to the 9 effects of debris?
10 And so that is calculated using NARWHAL.
11 We're analyzing a full range of break sizes and, 12 basically, summarizing those results in terms of a 13 conditional failure probability that can be plugged in 14 as a basic event to the PRA model.
15 CO-CHAIR DIMITRIJEVIC: Well, okay. Let 16 me just understand this now. Obviously, your 17 probability of the failure is different for the 18 different LOCA sizes, right?
19 MR. SANDE: Yes, that's correct.
20 CO-CHAIR DIMITRIJEVIC: So where is this 21 average calculated? In NARWHAL or in, I mean -- so 22 NARWHAL gives you the condition of failure probability 23 for different break sizes, break location, all of 24 these inputs, right? How do you average that when you 25 input it to the PRA?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
55 1 MR. SANDE: Ryan, could you please bring 2 up the backup slide number 4, CFP results?
3 MR. JOYCE: Yes.
4 MR. SANDE: Yes, that's a very good 5 question, Vesna. There is a very large difference in 6 terms of breaks on the order of 6 to 10 inches versus 7 breaks on the order of 20 to 30 inches in terms of 8 both the LOCA frequency and the conditional failure or 9 the probability of failing. So the way that we do 10 this averaging to develop the overall conditional 11 failure probability for large LOCA is illustrated on 12 this slide where we break the large LOCA category into 13 multiple size ranges. For the base case evaluation, 14 we used three size ranges, the first one from 6 to 15 15 inches, then from 15 to 25 inches, and then the final 16 from 25 and larger.
17 And so based on the LOCA frequency from 18 essentially equivalent to NUREG-1829, we actually use 19 a frequency out of the PRA model, but it was based, in 20 part on 1829 data. So we used that frequency curve 21 to, basically, weight each of those size ranges, and 22 then we calculated the failure probability or the 23 percent of breaks that would fail within each size 24 range from the NARWHAL results and then essentially 25 added the weighted values together to determine the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
56 1 overall conditional failure probability for a large 2 LOCA.
3 CO-CHAIR DIMITRIJEVIC: Okay. So I read 4 somewhere the CFP for the breaks bigger than 13 inches 5 was one, so I read that wrongly?
6 MR. SANDE: Yes. I don't recall where 7 that would have shown up, but we did not assume that 8 all breaks larger than 13 inches automatically fail.
9 CO-CHAIR DIMITRIJEVIC: And is there some 10 break size where you got CFPs for one?
11 MR. SANDE: I'm trying to recall if there 12 were some equipment configurations where the largest 13 break sizes all failed. I don't recall any of those 14 off the top of my head. This particular graph --
15 let's see. I believe this was for all pumps running, 16 and so the bars would shift depending on, like single 17 train failure is generally worse in terms of strainer 18 failures than if you have both RHR trains available 19 because the debris would accumulate on one strainer if 20 you only have one strainer in operation.
21 So, in general, you would see more 22 failures for that scenario, but I don't recall any 23 specific cases where we saw that, for a given size, 24 all breaks were failing.
25 CO-CHAIR DIMITRIJEVIC: Okay, all right.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
57 1 So I have to go to check where did I get this 2 information because this all was puzzling me. I 3 didn't see these graphs. So, basically, here you have 4 a frequency of, you know, 2 e minus 8 or one 2 minus 5 8 and a probability of failure 6. So that CDF would 6 be something, you know, what I did there, like 2 e 7 minus 8 or something.
8 Okay, all right. I was puzzled by this, 9 actually, how was it incorporated. Is this 10 incorporated in your basic PRA model now? I mean, you 11 sort of have that, or this was just outside stuff?
12 MR. SANDE: Well, what is shown on this 13 graph is all done as part of the post-processing of 14 the NARWHAL results, so it's calculated using the 15 NARWHAL software. And that final number, the 16 conditional failure probability for a large LOCA 17 equals 1.2 percent, that's, basically, the equivalent 18 to a basic event value for the PRA model. And, again, 19 this is specific to a given equipment configuration.
20 So in the PRA model, for the GSI-191 21 application, we modified the PRA to include basic 22 events associated with different pump configurations 23 and for small, medium, and large LOCAs. And so what 24 we would do in the PRA to calculate delta CDF and 25 delta LERF is we would take these conditional failure NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
58 1 probabilities for the various equipment configurations 2 and we would enter them into the PRA model as basic 3 events and then quantify the model to determine the 4 delta CDF values.
5 CO-CHAIR DIMITRIJEVIC: And so basic event 6 for your calculation failure, right? That's where you 7 put it?
8 MR. SANDE: Yes, that's right. We also 9 have a --
10 CO-CHAIR DIMITRIJEVIC: And so now my 11 previous question was simpler than this was. Is it 12 part of your base PRA model today, or this was done as 13 a separate study, that's my question, and it's not 14 incorporated in your base PRA?
15 MR. SANDE: When we did the GSI-191 16 application, it was not part of the model of record.
17 I'll let Alex or Faramarz speak to whether it's in the 18 current model of record.
19 CO-CHAIR DIMITRIJEVIC: Because you've 20 committed to maintain it, I just wonder do you have to 21 do these separate studies when you track the, you 22 know, what is the, you know, when you are updating 23 your risk-informed applications.
24 MR. SANDE: Sure. Alex, can you address 25 that question, please?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
59 1 MR. GILBREATH: We developed an 2 application-specific model to incorporate these basic 3 events.
4 CO-CHAIR DIMITRIJEVIC: So it's not part 5 of your base PRA model which you maintain, right?
6 It's a separate model.
7 MR. GILBREATH: Yes. It was developed 8 separately.
9 CO-CHAIR DIMITRIJEVIC: Okay. Well, your 10 PRA maintains updates, blah, blah, blah, procedures.
11 I was just wondering does this model automatically 12 also update when you update your base model? I mean, 13 that's what I was wondering. I mean, it's not really, 14 because you did commit to, you know, maintain this 15 application, so I was just wondering how easy that 16 maintenance would be. Are you directly connected with 17 your base model which has been maintained?
18 MR. POURNIA: This is Faramarz Pournia, 19 the RA director here. Basically, the gates that were 20 identified in the discussion we had earlier, all of 21 those are going to be pulled into our base model 22 because we will have processes in place that, in the 23 future, when there are any changes and we review for 24 impact, as I mentioned earlier, then we can go back 25 into our model and update it if there are any changes NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
60 1 to it.
2 So the way Alex explained that the way we 3 do this, we create what are called application-4 specific model to get through the process of our 5 licensings before we bring that to our base model to 6 make it a part of our control. And then, once they 7 become part of that, we maintain that because it's 8 part of the as-built, as-operated. In that phase, we 9 take the issues related to these application-specific 10 models. Does that answer your question?
11 CO-CHAIR DIMITRIJEVIC: Well, basically, 12 you have to make then two models, right? This --
13 MR. POURNIA: No, ma'am. This is Faramarz 14 Pournia again. Once the license get approved and 15 become part of as-built, as-operated, then we put that 16 into our main model, our base model. But until then 17 --
18 CO-CHAIR DIMITRIJEVIC: That was my 19 question, are you going to put it in there. So you're 20 planning to integrate that into your main model?
21 MR. POURNIA: Yes, ma'am.
22 CO-CHAIR DIMITRIJEVIC: So your main model 23 will have, actually, this associated with strainer 24 blockage as outcome?
25 MR. SANDE: Yes, ma'am.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
61 1 CO-CHAIR DIMITRIJEVIC: Okay. And now, 2 while you were talking, when I was looking at this 3 figure, I notice that, actually, your condition of 4 failure probability is zero for the below 20 inches; 5 is that true? Or is it's less than 10 to minus 9? Is 6 that what these graphs show me?
7 MR. SANDE: It's showing you that, for 8 breaks smaller than 20 inches, we're not exceeding any 9 of the acceptance criteria for this particular 10 equipment configuration. So, yes, that would be 11 equivalent to, if you're looking at breaks from a half 12 inch to 19 inches, the conditional failure probability 13 for this equipment configuration would be zero.
14 However, we're not calculating conditional failure 15 probability for that specific size range. We're 16 calculating it for large LOCAs, which is six inches 17 and larger. And so when we report a conditional 18 failure probability, it's conditioned on a break 19 greater than six inches so it can be plugged into the 20 PRA model directly as a large LOCA basic event.
21 CO-CHAIR DIMITRIJEVIC: I understood that, 22 but, basically, I'm surprised here because how you 23 discuss the only breaks larger than 20 inches actually 24 show any condition of failure probability. Okay, all 25 right. I will keep that in mind as we move. Okay.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
62 1 Thanks.
2 CO-CHAIR REMPE: This is Joy, and I just 3 wanted to request that the backup slides that you 4 present to this meeting be sent to Weidong after this 5 meeting today, please.
6 MR. JOYCE: Yes, we will do that. Thank 7 you. Who should I send it to again?
8 CO-CHAIR REMPE: Either NRR or Weidong 9 Wang. It's up to you guys how you transmit things.
10 MR. JOYCE: Yes, we will do that 11 immediately after this meeting.
12 MR. WANG: Yes, please send to NRR staff 13 first, and then send to me. Thanks.
14 MR. JOYCE: Okay. So I'll send it to 15 John, I'll send it to John Lamb, and then --
16 MR. WANG: Yes, please.
17 MR. JOYCE: Okay, thank you.
18 DR. SCHULTZ: Ryan, Steve Schultz. Before 19 you get into the sensitivity evaluation, at the 20 beginning of this section you talked about the NARWHAL 21 and the evaluations done were performed under the 22 Enercom quality assurance program, and my question is 23 does Southern Company, have you audited that program 24 as a part of your evaluations in this project?
25 MR. JOYCE: Tim or Haifeng, can you speak NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
63 1 to any audits that were performed?
2 MR. SANDE: Your question was -- I'm 3 sorry. This is Tim Sande. Your question was whether 4 SNC has audited Enercom's QA program; is that correct?
5 DR. SCHULTZ: Yes, since it's a large part 6 of what you've done here.
7 MR. SANDE: Yes, absolutely. So Enercom 8 is regularly audited by NUPIC, and so, basically, my 9 understanding is that SNC is a part of NUPIC and 10 doesn't necessarily audit Enercom directly but does 11 through the NUPIC process.
12 DR. SCHULTZ: Through NUPIC. I was just 13 curious whether you ever participated in a NUPIC audit 14 since the result of this project. It's, obviously, a 15 pretty big one for you.
16 MR. SANDE: Yes, that's right. NUPIC did 17 specifically look at our software development because 18 we developed both BADGER and NARWHAL software for the 19 risk-informed GSI-191 application, and so that was 20 something that they had specifically looked at. I 21 think it was back in 2016.
22 DR. SCHULTZ: Good. Thank you.
23 MR. SANDE: You're welcome.
24 MR. JOYCE: Any other questions about our 25 risk quantification?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
64 1 (No audible response.)
2 All right. Next, I'd like to discuss the 3 sensitivity and uncertainty analyses performed for our 4 risk-informed approach.
5 So for the sensitivity analyses, we 6 performed parametric sensitivity analysis to identify 7 which parameters have the most impact on the risk 8 quantification results. To avoid skewing the results, 9 it's important not to arbitrarily pick a very wide 10 range for one parameter and a very narrow range for 11 another parameter. As the parameter where a wide 12 range input was used would tend to show a wider range 13 of delta CDF results. But the model may be actually 14 more sensitive to the parameter where a very narrow 15 range was used for the sensitivity analysis.
16 Selecting a consistent range for input 17 parameters is challenging because the inputs used are 18 very different. For example, RWST level has 19 procedural and physical limits for min and max volume 20 first in the sub-temperature profile, which is 21 analytically derived and has significant time-22 dependent variations based on the thermal hydraulic 23 inputs used. As discussed in this slide, SNC's set of 24 rules to objectively define appropriate ranges for 25 each input parameter. Sensitivity analyses showed NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
65 1 that the hot leg break specified in the limits, LOCA 2 frequency, and RHR pump flow rate values have the most 3 impact on the risk quantification results.
4 So this slide discusses the uncertainty 5 quantification. It's important to note the difference 6 between uncertainty quantification and the parametric 7 sensitivity covered in the previous slide. Parametric 8 sensitivity analysis was performed to identify which 9 individual parameters have the most impact on the 10 results while the uncertainty quantification is used 11 to determine the level of confidence in the base case 12 results. This includes parametric uncertainty where 13 multiple parameters are changed in the conservative 14 direction similar to a deterministic analysis and 15 model uncertainty where alternatives are evaluated for 16 models where no consensus exists.
17 The Reg Guide 1.174 risk region for an 18 application is determined based on the mean delta CDF 19 and delta LERF values, and the uncertainty analysis is 20 used to determine the confidence level in those 21 results. As described in NUREG-1855, three types of 22 uncertainty were addressed: parametric uncertainty, 23 model uncertainty, and completeness uncertainty.
24 Parametric uncertainty analysis 25 quantitatively evaluated cases where all input NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
66 1 parameters were shifted to the conservative bounding 2 conditions. For these, multiple runs were performed 3 to address min and max values for parameters that have 4 competing conservatisms.
5 Model uncertainty analysis quantitatively 6 evaluated alternatives for models where no consensus 7 exists. The completeness uncertainty was 8 qualitatively addressed. The results from all 9 parametric and model uncertainty cases remained within 10 Reg Guide 1.174, Region III and would provide high 11 confidence in the risk associated with the effects of 12 debris is very small.
13 MEMBER HALNON: This is Greg. I just 14 continue poking at the question of bounding numbers.
15 The RHR flow rate you said was a significant 16 contributor to the results. Did you guys use the 17 minimum optimal flow rates for that by your Section 1 18 program?
19 MR. JOYCE: I believe that's addressed in 20 the -- is that addressed in backup slide 7?
21 MEMBER HALNON: Was it 7? Okay. Let me 22 go back.
23 MR. JOYCE: Tim, Tim Sande. Should I 24 address --
25 MR. SANDE: Yes, this is Tim. I don't NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
67 1 recall. Haifeng, you, I think, maybe could answer 2 flow rate questions better than I can.
3 MR. LI: I think earlier in the 4 presentation we only mentioned about 3,700 gpm flow 5 rate. I'm looking back into the submittal to see if 6 we get any, what flow rate we use in uncertainty 7 quantification. Tim, I will report back on that.
8 MEMBER HALNON: Again, I'm just poking 9 around the bounds of your analyses and making sure I 10 understand what values you use and whether or not 11 they're bounding or not. It seems like the bounding 12 value of the RHR flow rate is sensitive to the results 13 that you'd use the amount, the flow rate that is 14 minimally acceptable for operability such that, you 15 know, the action the operators would take would be low 16 bounded.
17 So that's the question. You might be able 18 to answer it in a couple different ways. But if you 19 could get back to me, I appreciate it.
20 MR. LI: Yes, this is Haifeng Li again.
21 Tim, are you saying something? Sorry.
22 MR. SANDE: Oh, go ahead. I can add 23 something, but go ahead.
24 MR. LI: Yes. This is Haifeng Li again.
25 Overall, for GSI-191 analyses, a higher flow rate is NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
68 1 generally conservative, especially for the strainer 2 failures. So for in-vessel, we do have to look at min 3 and max flow rates and a best estimate of flow rate.
4 That's why I want to look back through our uncertainty 5 analyses and see what flow rates we used there. Tim, 6 going back to you.
7 MEMBER KIRCHNER: This is Walt Kirchner.
8 On the question of flow rates, I'm thinking back to 9 earlier in the presentation. You said you assumed, 10 took no credit, essentially, for containment 11 atmosphere pressure or you put a nominal value on it, 12 but you didn't have cavitation in the suction line.
13 Did you explore that in any detail? It seems to me if 14 you realistically put the containment pressure in, you 15 would likely, in most scenarios, have higher flow 16 rates. That's part observation, part question.
17 MR. SANDE: This is Tim Sande. So that's 18 a good question. Going back to the flow rate in terms 19 of the value that was used, the 3,700 gallons per 20 minute, I don't recall the details of how that flow 21 rate was determined, but it was essentially equivalent 22 to the flow rate that would have been used for 23 deterministic analysis. So we weren't trying to come 24 up with a best estimate flow rate for the risk-25 informed application. Instead, we were using what NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
69 1 would be, essentially, a maximum flow rate that would 2 be consistent with a large LOCA where you have, 3 basically, minimal back pressure from the RCS.
4 I believe, to address the other question 5 that was brought up, if you have higher containment 6 pressure, that would, it would potentially provide 7 additional pressure for the pumps. But because we're 8 injecting back into containment, I believe that it, 9 basically, cancels out in terms of overall pressure 10 effects on the pumps themselves.
11 MEMBER KIRCHNER: Yes. So I was just 12 thinking of maintaining a good net positive suction 13 head on the, into the pumps. You're right. Across 14 the entire system, it's kind of a wash at those low 15 pressures.
16 MR. SANDE: Yes. And we did -- so for net 17 positive suction head, that's one of the failure 18 criteria that we looked at, do we lose net positive 19 suction head due to the head loss across the strainer; 20 and, for the purpose of that evaluation, we did not 21 take credit for any accident pressure. For the 22 purpose of the degasification and flashing 23 calculations, which also use containment pressure as 24 an input, we did take credit for a small amount of 25 accident pressure for a short period of time. And NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
70 1 Ryan has a slide later that discusses that.
2 MEMBER KIRCHNER: Okay, thank you.
3 CO-CHAIR DIMITRIJEVIC: I have a question 4 about your approach here. You have analyzed, I don't 5 know, 20-something parametric uncertainties, you know, 6 which include this flow rate for multi adjusted 7 (phonetic) pump, other BST (phonetic) level of a lot 8 of frequency values, but that combination only for, 9 that combination between level and the average flow 10 rate or level and duration of containment spray, you 11 did not add the combinations of this parametric 12 values. And, actually, one of those combinations show 13 your delta CDF in value minus 7. So how did you 14 select those combinations, and why didn't you run more 15 combinations of this parametric? That's my first 16 question.
17 My second question, I also noticed in 18 model uncertainty that you're very sensitive to the, 19 as you said yourselves, to LOCA frequencies that use 20 geometric versus certimetric (phonetic) mean, and with 21 telemetric mean, actually, your delta CDF increased 22 over ten times. So, singularly, with that, for that 23 part, was the CS activated for different break sizes?
24 For smaller break sizes, you have that, your delta CDF 25 increases to close to the, you know, Reg Guide 1.174 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
71 1 limit.
2 So my general question is do you run 3 combinations of this parametrical uncertainties and 4 the model uncertainties other than the ones, the eight 5 combinations which you show in the report. And the 6 second question is how did you determine confidence 7 level, and what is your 95 confidence level on this?
8 Okay.
9 MR. SANDE: Yes, Ryan, would you like me 10 to address those questions?
11 MR. JOYCE: Yes, please. You want any 12 backup slides, Tim?
13 MR. SANDE: Yes, if you could bring up the 14 backup slides that have the results of the uncertainty 15 quantification, that would be helpful. Thank you. So 16 this is Tim Sande again. A lot of very good questions 17 there. If I miss some, then please let me know.
18 So to address your question on the 19 parametric uncertainty, we focused on the water volume 20 and containment spray duration when we were looking at 21 potential strainer failures. And then we focused on 22 water volume and the RHR flow rate when we looked at 23 the core failures. And the reason was that, in 24 general, most of the input parameters that we're using 25 either are already conservative or bounding parameters NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
72 1 that are, basically, consistent with how a 2 deterministic evaluation would be done. And so, based 3 on that, we didn't really need to say how much worse 4 could it get because we're already using, essentially, 5 the worst-case scenario.
6 And then there are other parameters that 7 we didn't necessarily use the worst-case scenario.
8 For example, we used geometric mean LOCA frequencies 9 to calculate both the additional failure probability 10 and the delta CDF. And since that's a mean value, 11 what we did for the parametric uncertainty was we 12 increased that to the 95th percentile value. And 13 then, once we did that, there were only a handful of 14 parameters left where there was some uncertainty in 15 terms of would a minimum value be worse than a maximum 16 value because of competing conservatisms in the 17 analysis. For example, a minimum water volume is 18 usually more conservative for a number of areas of the 19 evaluation, but a maximum water volume has the 20 potential to be conservative specifically for chemical 21 effects or some things related to the timing 22 associated with debris transport.
23 So in order to address both ends of the 24 uncertainty there, we used both a minimum and a 25 maximum volume, and we ran separate parametric NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
73 1 uncertainty cases to explore all of those different 2 possibilities. But the goal really for the parametric 3 uncertainty quantification was to find the worst-case 4 scenario of all the combination of parameters that 5 would give you the worst-case scenario and see how 6 high that delta CDF value was. And our conclusion, as 7 shown on this slide, was the worst-case was still 8 within Reg Guide 1.174, Region III, which indicates 9 that the uncertainty is not high enough to challenge 10 our confidence that the actual risk is within Region 11 III.
12 CO-CHAIR DIMITRIJEVIC: So this is your 13 combination based on the water volume and the flow 14 rate in which you explain why you choose this 15 combination, but what if you added some other 16 parameter switches independent of that, like LOCA 17 frequencies or things like that? There's so many ways 18 you can combine your parametrics, your parametric or 19 your model uncertainty. I mean, I'm not sure did you 20 choose the worst case. I cannot be sure from what you 21 just told me. I could probably think about the worst-22 case combinations, you know, that --
23 MR. SANDE: Right. So that's an excellent 24 question, and we did go through methodically and 25 looked at each input parameter for the NARWHAL NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
74 1 evaluation. I could double-check the submittal. I 2 believe we had included something in the submittal, 3 but I would have to go back and look. But for each 4 parameter, we identified is this already a 5 conservative input value.
6 For example, debris quantities generated.
7 We're already using all of the conservative guidance 8 from NEIO 407 to calculate, essentially, the maximum 9 quantity of debris generated for any given break 10 that's postulated. So we're already using 11 conservative input there, and it wasn't necessary to 12 change that input.
13 And we did that for each input value, and 14 some inputs, as I mentioned, the geometric mean LOCA 15 frequency, is, essentially, a best-estimate input.
16 It's not a conservative input. And so we increased 17 that for the parametric uncertainty to the 95th 18 percentile value, the geometric 95th percentile, and 19 then we went through and there were only a couple of 20 parameters, like the water volume and CS duration, 21 that we said we don't know which direction is 22 conservative so we need to run a couple of cases.
23 But everything else was, basically, set at 24 that conservative value, so we're not just limiting it 25 to the water volume and containment spray duration.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
75 1 We're looking at every single parameter.
2 CO-CHAIR DIMITRIJEVIC: So when you get 3 events from NARWHAL, it doesn't give you, it cannot 4 run, it doesn't give you CFPs with the uncertainty 5 ranges, right? Is that, am I right to conclude that?
6 MR. SANDE: Yes, that's right. NARWHAL 7 actually does have some uncertainty tools built into 8 it where you could forget that, but we didn't use it 9 for Vogtle. What we did was we simply ran a 10 sensitivity case where we changed all the parameters 11 based on our analysis of what the conservative 12 direction would be.
13 So we're not coming up with a probability 14 distribution that shows uncertainty. What we're doing 15 is we're kind of looking at what would be the worst 16 case, if we changed all the parameters in the worst-17 case direction, what would the resulting delta CDF 18 values be. And if those delta CDF values were in 19 Region I, that would say, hey, maybe there's some 20 significant uncertainty associated with our conclusion 21 that the risk is in Region III. But the conclusion 22 from that evaluation was that all of the delta CDF 23 values from those parametric uncertainty results were 24 still well within Region III, which gives us very high 25 confidence that the risk truly is in Region III.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
76 1 CO-CHAIR DIMITRIJEVIC: But that's --
2 well, how can you say that, you did not run all 3 parameters on the most (unintelligible) and none of 4 them went over, you know, the 10 to minus 6, but many 5 of them went over 10 to minus 7, some of them went 6 over 9 minus 7, some of them went 5 e minus 7.
7 So if you did not even consider 8 combination, how do you determine the confidence?
9 Where did you calculate that? What is actually 10 confidence? What is your 95 value and how did you 11 calculate it?
12 MR. SANDE: So, again, we didn't develop 13 a probability distribution as part of the uncertainty 14 quantification, so we don't have a 95th confidence 15 interval that we can report to you. However, I think 16 what you are looking at when you're talking about 17 changing individual parameters, that was part of the 18 parametric sensitivity analysis, which is not the same 19 thing as a parametric uncertainty quantification.
20 For the parametric sensitivity, we were 21 attempting to change the individual parameters to look 22 at how much each parameters affects the results, but 23 we weren't staying within the bounds of, basically, 24 what you would do for deterministic analysis. For 25 example, we're already using the RWST minimum level NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
77 1 based on the tech spec minimum as our input for the 2 base case. So that's already a conservative value.
3 For the parametric sensitivity, we 4 actually decreased it below that tech spec minimum, 5 which is not realistic but it gives us an idea of, 6 hey, if, for some reason, the water level was lower, 7 how would that affect the results. But for the 8 parametric uncertainty, we left that as the tech spec 9 minimum because that is the minimum water level that 10 you would have in the RWST.
11 And the other thing that we did for the 12 parametric uncertainty is we didn't just change one 13 parameter at a time. We changed all parameters to the 14 conservative direction so that, if you have competing 15 effects between parameters, basically, all parameters 16 are conservative values and it gives you a good idea 17 of what's the upward ceiling on the risk 18 quantification as a function of all those worst-case 19 parameters.
20 MR. LI: This is Haifeng Li with Enercom.
21 Tim, going back to what you mentioned about we 22 screened all the parameters and determined the 23 bounding direction, we do have a table in the 24 submittal that demonstrates that. I'm not sure if 25 that will help with the discussion if we can somehow NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
78 1 --
2 CO-CHAIR DIMITRIJEVIC: I saw this table, 3 but that only shows me parameter by parameter, your 4 delta CDF minus and plus (phonetic). I saw this 5 table, and that's why I'm sort of curious now when you 6 say that parametric sensitivity is not the same as 7 uncertainty. I agree completely. So where do I see 8 parametric uncertainty results? Because these tables 9 in your -- and I apologize. Because of an issue with 10 my NRC computer, I could not see your preview 11 submittal. I only saw your August submittal and 12 looking at all results there. But I didn't see this 13 uncertainty results, I just saw sensitivity results in 14 the tables and I saw only four combinations which came 15 out of, like, I don't know, a thousand possibilities, 16 and I didn't see any confidence levels. So where can 17 I find those uncertainty results and your confidence 18 level?
19 MR. SANDE: Haifeng, do you have the page 20 number in the submittal that --
21 MR. LI: Yes. So if you look at the July 22 2018 technical report which has a section number 23 ML18193B165. That's the July 2018 technical report.
24 If you look at page E3-72, it has Table 3-13 that 25 shows we reviewed each parameter with respect to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
79 1 strainer failure and a core failure and which 2 direction is the bounding direction.
3 For the parameters that we do have a clear 4 bounding direction, we used the extreme value, either 5 maximum or minimum value. In that table, it also 6 showed that for two parameters for each scenario 7 there's not a clear bounding direction. That's why we 8 used either minimum or maximum. That's where those 9 four cases came from.
10 So if you could review the table and see 11 if that's the information you are looking for. So 12 Table 3-13 was for the strainer failure, and the next 13 table, 3-14, was for the core failure. Am I clear on 14 the page numbers and table numbers?
15 CO-CHAIR DIMITRIJEVIC: Sorry. I didn't 16 realize my microphone -- yes, you're clear. I mean, 17 I can look in the Tables 12 and 14 and 15. I will 18 just look in those. And I saw your combination, and 19 it almost looks to me it was related to sensitivity 20 cases, but I will look in your July submittal and 21 check this out during the break, all right?
22 MR. LI: Yes, thank you. Going back to 23 the earlier question about the pump flow rates, as our 24 team was saying earlier, the design flow rate for the 25 RHR pump was 3,000 gpm, and the deterministic NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
80 1 evaluation used 3,700 gpm to analyze pump MPSH.
2 That's the flow rate we were using in NARWHAL for the 3 base case analyses.
4 In the sensitivity, we varied the RHR pump 5 flow rates within a range of 3,700 minus 25 percent as 6 minimum and a 4,500 as maximum. In the uncertainty 7 quantification, for the strainer failure case, we were 8 using the maximum RHR, the run-out flow rate, 4,500 9 gpm for that uncertainty case. For the core failure 10 uncertainty case, we tried both the minimum and a 11 maximum RHR flow rate, which were 2,775 gpm and 4,500 12 gpm. So that's responding to the earlier question 13 regarding the RHR flow rate.
14 MEMBER KIRCHNER: While we're on 15 uncertainties, could you just describe briefly how you 16 did your completeness uncertainty quantification? You 17 mentioned that it was actually qualitative.
18 MR. SANDE: Sir, this is Tim Sande. Yes, 19 because completeness uncertainty addresses unknowns, 20 which can't be quantified, we did do a qualitative 21 evaluation. And the qualitative evaluation, 22 essentially, focused on the fact that strainer 23 performance has been evaluated for over 40 years now.
24 USIA 43 first became an issue back in 1979, and there 25 were tests performed at that time focused on head loss NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
81 1 due to fiber and the potential for vortexing and air 2 ingestion. Then in the early 90s with the Barsebck 3 event, the issue was opened up for the BWRs, and there 4 was a lot of testing and analysis done at that point.
5 I believe GSI-191 was first opened in 1996 and then in 6 2004 the NEIO 47 guidance came out and also GL-2004-02 7 was issued. So the U.S. PWRs have done a tremendous 8 amount of work evaluating this issue. There's been a 9 lot of work done by the industry, as well as by the 10 NRC, and that's just in the U.S. There's also been a 11 lot of work done by industry and regulators 12 internationally.
13 And so that very large body of knowledge 14 has evaluated just about every aspect of strainer 15 performance and gives us confidence that there aren't 16 additional things that might be lurking out there that 17 would significantly affect risk.
18 MEMBER KIRCHNER: Yes. Tim, I would agree 19 with that in general. There's quite a bit of 20 literature and experience and history in this 21 particular issue. Did your completeness review send 22 you back to your parametric analyses and add or change 23 any of your scenarios that you ran?
24 MR. SANDE: No, it did not. As I 25 mentioned before, we took a very methodical approach NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
82 1 to identifying both the parametric sensitivity and the 2 parametric uncertainty and looking at every parameter 3 and the model uncertainty, as well. And so we didn't 4 discover anything new as part of that completeness 5 uncertainty evaluation.
6 MEMBER KIRCHNER: Thank you.
7 MR. JOYCE: Any other discussion on the 8 uncertainty quantification before we move on to the 9 technical report submittal?
10 (No audible response.)
11 Next, I'd like to discuss our technical 12 report submittal that was previously discussed. So by 13 letter dated February 14th, 2017, the NRC agreed to 14 review a technical report, an SNC-specific technical 15 report, for addressing GSI-191 that does not rely on 16 WCAP-17788. This report, WCAP-17788, at the time was 17 still under NRC review.
18 The purpose of the report is to receive 19 the NRC review and approval of our supplemental 20 Generic Letter 2004-02 response using risk-informed 21 approach. And on April 21st, 2017, SNC submitted the 22 technical report to the NRC.
23 The technical report summarized risk-24 informed approach to resolve Generic Letter 2004-02 25 and superseded previous responses. Enclosure 1 was a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
83 1 high-level summary of the submittal and was organized 2 in accordance with Draft Reg Guide 1129, Section C.
3 Enclosure 2 described plant conditions, analyses, and 4 testing that informed the resolution following the NRC 5 Content Guide for Generic Letter 2004-02 Responses, 6 including project information. This also addressed 7 NRC RAIs from our previous Generic Letter 2004-02 8 submittals.
9 Enclosure 3 summarized methodology and 10 results, risk quantification, and sensitivity analyses 11 and uncertainty using NARWHAL and the Vogtle PRA 12 model. This was organized in accordance with Draft 13 Reg Guide 1129, Appendix A.
14 Enclosure 4 described defense-in-depth 15 measures and safety margin. Enclosure 5 was a non-16 proprietary version of Enclosure 2.
17 So in 2017 and in 2018, the NRC performed 18 audits and issued RAIs based on their review of the 19 technical reports. SNC responded to the RAIs and 20 questions raised in the audits, incorporated these 21 into an updated version that was submitted in July 22 2018. The NRC issued a staff evaluation in September 23 2019 and included that the technical report contains 24 sufficient information to address Generic Letter 2004-25 02 for Vogtle, except for in-vessel downstream NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
84 1 effects, and was acceptable for use in future 2 licensing applications subject to the conditions 3 limitations discussed.
4 SNC discussed responses to the conditions 5 limitations and the proposed LAR submittal with the 6 NRC during public meetings November 2019 and April 7 2020. In August 2020, SNC submitted our submittal to 8 the NRC. And between the technical reports, July 2018 9 technical report and the August 2020 submittal, all 10 five key principles of Reg Guide 1.174 were addressed.
11 So this shows the five key principles of 12 Reg Guide 1.174. The principles in the second and 13 third bullet were addressed in the technical report 14 submittal. The principles in the fourth bullet were 15 addressed by a combination of the technical reports 16 and the 2020 LAR submittal. And the first and fifth 17 bullet were addressed in the 2020 LAR submittal.
18 Next, I'll give a review of our License 19 Amendment Request submittal. So as previously 20 discussed, this LAR submittal contained both an 21 exemption from 10 CFR 50.46(a)(1) to support our risk-22 informed approach and also requested License Amendment 23 Requests to implement the risk-informed approach, as 24 was described in our technical reports, our 2018, 2019 25 technical reports and to adopt TSTF Traveler 567 to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
85 1 adopt a new tech spec for containment sump.
2 The LAR submittal was arranged in four 3 different enclosures. Enclosure 1 was the request for 4 an exemption. For the License Amendment Request, due 5 to the two kind of distinct aspects of it, we broke it 6 up into three different enclosures, Enclosure 2 being 7 specific to the risk-informed approach, Enclosure 3 8 being specific to our request to adopt TSTF-567, and 9 Enclosure 4 being a combined regulatory evaluation.
10 So Enclosure 2, as I said, was requesting 11 to use our risk-informed approach, as described in our 12 technical reports. This follows what was considered, 13 what is a considered a Type 1, a plant-specific LAR 14 formats for NEI 0602, and included attachments for the 15 proposed SR (phonetic) markups, guidance for 16 supporting the operability evaluations, and the 17 evaluation for in-vessel effects of coating.
18 Enclosure 3 to adopt TSTF-567 would be 19 considered a Type 3 LAR formats, which is to adopt a 20 TSTF Traveler, a CLIP, Consolidated Line Item 21 Improvement Product, CLIP Traveler, and this included 22 proposed tech spec changes in accordance with this NEI 23 602 format for adopting TSTF Travelers, proposed tech 24 spec changes, revised technical specification pages 25 and proposed technical specification bases changes NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
86 1 marked up.
2 All right. Next, I'd like to discuss the 3 exemption requests, Enclosure 1 of our submittal. So 4 the exemption was requested from certain requirements 5 of 10 CFR 50.46(a)(1), other properties, relating to 6 using deterministic methodology to evaluate the 7 effects of debris on long-term core cooling.
8 This request for exemption supports the 9 risk-informed approach in Generic Letter 2004-02 10 enclosure as an alternative to the current 11 deterministic evaluation required by 10 CFR 12 50.46(a)(1). The exemption will apply only to the 13 effects of debris, as described in Enclosures 2 and 3 14 of the July 2018 submittal and will apply to any that 15 can generate and transport debris exceeding the Vogtle 16 analyzed limits, provide the delta CDF and delta LERF 17 values remain within Reg Guide 1.174, Region III.
18 The exemption was evaluated against the 19 conditions in 10 CFR 50.12(a). It was determined that 20 the request meets the requirements for granting an 21 exemption to the regulation and that special 22 circumstances, as described in 10 CFR 50.12(a)(2)(2) 23 and (3) are present. The exemption was reviewed for 24 environmental impacts, was determined to be eligible 25 for categorical exclusion per 10 CFR 51.22(c)(9).
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
87 1 The risk associated with post-accident 2 debris effects is within Reg Guide 1.174, Region III, 3 for very small change. And the requested exemption is 4 consistent with Reg Guide 1.174, Key Principle 1, and 5 that the proposed licensing change meets the current 6 regulations, unless explicitly related to a requested 7 exemption.
8 Next, I'd like to discuss our License 9 Amendment Request submittal or our implementation of 10 the risk-informed approach, Enclosure 2 of this 11 submittal. So as previously mentioned, the proposed 12 changes in this submittal are based on the July 2018 13 SNC technical reports and the NRC staff evaluation 14 with no deviations. Approval of this change will 15 allow Vogtle to use risk-informed approach to address 16 the effects of accident-generated debris on 17 containment sumps.
18 The technical evaluation using the risk-19 informed approach was described in the July 2018 20 submittal. As described in the submittal, evaluate 21 LOCAs up to double-ended guillotine breaks on large 22 RCS piping, use break-specific analyses to identify 23 break scenarios that fail any GS1-191 criteria, 24 quantify delta CDF and delta LERF using the plant PRA 25 model, and it shows that the risk associated with NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
88 1 post-accident debris is very small, as defined in 2 Region III of Reg Guide 1.174.
3 So for the engineering analysis, approval 4 of this License Amendment Request does not change the 5 design licensing basis descriptions of accidents 6 requiring ECCS and containment spray operating, 7 including analysis method, assumption, results 8 provided in SSCR Chapters 6 and 15. The functionality 9 of ECCS and containment spray system during design-10 basis accidents were still confirmed. Safety margin 11 and defense are maintained with high probability and, 12 based on this, the calculated risk is conservative and 13 the actual risk is much lower.
14 Our submittal also addressed the nine 15 conditions limitations identified in the NRC staff 16 evaluation. Specifically, some of these address Reg 17 Guide 1.174, Key Principles 1 and 5; confirm the 18 applicability of technical reports; responded to in-19 vessel downstream effects following the NRC review 20 guides, WCAPs and the NRC WCAP 1778 and associated NRC 21 guidance on that report; and identified key elements 22 of the risk-informed analysis; described controls to 23 ensure relevant elements are periodically updated; and 24 corrected the error in the evaluation of coating 25 debris.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
89 1 MEMBER KIRCHNER: May I ask a question on 2 the bullet safety margin and defense-in-depth are 3 maintained with high probability? How did you define 4 defense-in-depth, and then how did you determine it 5 was maintained with high probability? I generally 6 agree with the statement, but I would be interested in 7 any quantitative or method to actually reach that 8 conclusion.
9 MR. JOYCE: Haifeng, do you want to 10 discuss defense-in-depth and safety margin analyses 11 performed.
12 MR. LI: Yes. This is Haifeng Li with 13 Enercom. The defense-in-depth and safety margin were 14 addressed in Enclosure 4 of the submittal. As for the 15 defense-in-depth, we followed the Reg Guide 1.174 16 philosophy and looked into, followed that guidance --
17 I'm looking at Enclosure 4 and trying to figure out a 18 way, the best way to summarize it in a short 19 paragraph.
20 We looked into the evaluation of the Reg 21 Guide 1.174 philosophy and also detecting and 22 mitigating the adverse conditions, looked into plant 23 programs and instrumentation available to do those.
24 We also, in terms of detection of strainer blockage 25 and mitigation with strainer blockage, all those NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
90 1 measures that have been in place. Tim, do you have 2 anything to add?
3 MR. SANDE: Yes. I think the discussion 4 at the beginning of that section is helpful where we 5 define what is meant by defense-in-depth and safety 6 margin. So for the purpose of this submittal, 7 defense-in-depth is defined as the response to the 8 question of what happens if the analysis is wrong 9 about a successful end state and that end state 10 actually turns out to be a failure. So defense-in-11 depth includes mitigative design features and actions 12 that address protection of the public from radiation 13 due to sequences that go to failure, as well as --
14 which includes things like containment integrity, the 15 emergency plans, operator actions that aren't credited 16 in the GSI-191 evaluation, and use of flex.
17 MEMBER KIRCHNER: Okay, thank you. I'll 18 go back and look at Enclosure 4. That's Enclosure 4 19 of the TR.
20 MR. SANDE: That's correct.
21 MR. LI: Yes. For the safety margin --
22 this is Haifeng Li again. For the safety margin, we 23 were looking at, responding to the question of, you 24 know, what aspects of these analyses will increase the 25 confidence that the greater success is indeed a NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
91 1 success. Essentially, we summarized various ability 2 and conservatisms in our analyses that would increase 3 our confidence data, you know, a scenario that were 4 not resulting in failure, that's indeed the case. We 5 had a table to summarize those conservatisms in the 6 analyses.
7 MEMBER KIRCHNER: Well, I would have 8 answered the first half, the safety margins, based on 9 your case analyses which showed that you were in the 10 Region III very small effects on the delta CDF and 11 LERF. It was more the defense-in-depth, how you --
12 were you just looking at it in terms of the 13 performance of the ECCS and CS systems, not just 14 during design-basis accidents, but for the spectrum of 15 possibilities that you analyzed? It seems to me you 16 demonstrated there that the systems had sufficient 17 defense-in-depth as measured by safety margins.
18 MR. SANDE: This is Tim Sande. Yes, the 19 points that you're making I agree with. Really the 20 reason why we wanted to clearly define both -- what we 21 meant by both defense-in-depth and safety margin in 22 the submittal is because those are somewhat confusing 23 terms and I don't believe the Commission has endorsed 24 a specific definition for defense-in-depth. And so we 25 want to define what we mean by defense-in-depth so NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
92 1 that it's clear how we're addressing that requirement 2 from Reg Guide 1.174.
3 MEMBER KIRCHNER: You're correct. That's 4 why -- it was a leading question. I was just 5 exploring what your approach was to demonstrating 6 that. Thank you.
7 MR. SANDE: Sure. You're welcome.
8 MR. JOYCE: Any other questions on this 9 slide?
10 (No audible response.)
11 MR. JOYCE: All right. Next I'd like to 12 discuss Enclosure 3 of our technical report of our 13 license amendment request.
14 So the changes were consistent with the 15 NRC safety evaluation for tech spec 567. This added 16 a new tech spec for containment sump and added an 17 action to address the effects the condition of 18 containment sump being operable due to containment 19 accident debris exceeding the analyzed limits. And 20 for the Vogtle change the containment sump debris 21 limits are provided in the tech spec bases.
22 And then lastly for your license amendment 23 request submittal I'd like to discuss the regulatory 24 evaluation and environmental considerations. For the 25 regulatory evaluation the proposed amendment does not NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
93 1 involve a significant hazards consideration under the 2 standards of 50.92(c). License basis changes are 3 shown to meet the five key principles of Reg Guide 4 1.174 and the PRA model used for the risk-informed 5 evaluation complies with Reg Guide 1.200, Rev 2 and 6 can be applied in regulatory decision making.
7 For precedence, as previously discussed in 8 earlier slides, the proposed changes are similar to 9 license amendment requests in a 10 CFR 50.46 exemption 10 granted to South Texas (audio interference) risk-11 informed approach. Some of the key similarities 12 include use of the Reg Guide 1.174 acceptance 13 guidelines and key principles, the identification of 14 key methods and approaches in the risk-informed 15 methodology for evaluating potential departures for 16 method evaluation. (Audio interference) request for an 17 exemption from the 10 CFR 50.46(a)(1) (audio 18 interference) properties and the tech spec change 19 provides additional time to address the effects of 20 debris on ECCS and containment spray system 21 operability.
22 Some of the key differences, the first two 23 of which will be discussed in subsequent slides: The 24 software evaluation used in risk analysis, SNC's 25 NARWHAL versus STP's CASA Grande. The methodology NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
94 1 used. SNC used the conditional failure probability 2 approach versus South Texas' RoverD approach. STP --
3 in addition to 10 CFR 50.46(a)(1), STP requested 4 exemptions from GDCs 35, 38 and 41. And the probable 5 tech spec change follow the format of TSTF-567, which 6 was not available at the time for South Texas.
7 For the environmental consideration, 8 pursuant to 10 CFR 51.22(b) no Environmental Impact 9 Statement or environmental assessment needs to be 10 prepared for the proposed LAR because it meets the 11 eligibility criterion for categorical exclusion in 10 12 CFR 51.22(c)(9). The proposed amendment does not 13 involve a significant hazards consideration. The 14 proposed amendment does not propose a significant 15 change in the types or a significant increase in the 16 amounts of effluents that may be released off site.
17 The proposed amendment does not result in a 18 significant increase in individual or cumulative 19 occupational radiation exposure.
20 And for the remaining actions, as 21 previously discussed, we will do -- we have 22 modifications to the RHR sump strainers planned during 23 the upcoming outages for Unit 1 and Unit 2. Once 24 these modifications are complete we will implement --
25 incorporate tech spec changes to add the containment NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
95 1 sump tech spec and the risk-informed methodology into 2 our licensing basis. Instead of a staggered 3 implementation because we have one technical 4 specification and one FSER for both units we requested 5 to implement the risk-informed methodology and tech 6 spec changes after the modifications for both units 7 are complete.
8 Next I'd like to discuss the differences 9 between the Vogtle and South Texas risk-informed 10 evaluations. So the first difference to discuss is 11 the framework for quantifying risk. Typically the STP 12 RoverD versus the Vogtle conditional failure 13 probability approach.
14 So for STP's original submittal they used 15 an approach that was essentially equivalent to 16 Vogtle's conditional failure probability approach.
17 South Texas did a lot of great work in piloting the 18 risk-informed approach and pushed the boundaries on 19 many different aspects of the GSI-191 resolution.
20 This work clearly demonstrated that the GSI-191 risk 21 was a much lower risk than previously thought, however 22 the NRC had numerous RAIs that has to be resolved.
23 So rather than addressing all the RAIs 24 explicitly, STP backed off many of the refinements and 25 used a much more simplified approach; for example, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
96 1 using existing data from NRC head loss tests rather 2 than head loss correlation. At the same time South 3 Texas switched from a conditional failure probability 4 framework to a new method called RoverD. Vogtle was 5 already working on the risk quantification at this 6 time. The lessons learned: Simplification models 7 were incorporated, but there was no significant 8 benefit in switching to the RoverD framework.
9 For the evaluation software differences 10 STP and Vogtle used different software tools for the 11 analyses. So for example, STP used CASA Grande to 12 evaluate effects of debris in transport and FIDOE and 13 RUFF for in-vessel effects, whereas SNC used BADGER 14 for debris generation and NARWHAL for the integrated 15 effects of -- integrated evaluation of strainer in-16 vessel effects.
17 One main difference in the software tools 18 used by South Texas and Vogtle is how the sub-models 19 are integrated. South Texas used different software 20 to evaluate strainer failures and core failures; i.e.,
21 in-vessel effects. This made it difficult to use 22 self-consistent inputs and generally increased the 23 level of conservatisms. Vogtle used the NARWHAL 24 software, which integrates the evaluation of strainer 25 and in-vessel effects. As a result self-consistent NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
97 1 inputs can be used to analyze two phenomena.
2 For the equipment configurations South 3 Texas has three ECCS containment spray trains and 4 could accommodate significant conservatisms. The 5 South Texas base case strainer evaluation was 6 equivalent to the design-basis single train failure; 7 i.e., for South Texas they assumed only two trains 8 available. This is a very conservative assumption 9 since the likelihood of failing a single train is 10 relatively low and the probability of three related 11 strainer failures would significantly decrease if 12 spray is split between all three strainers.
13 A variety of equipment configurations 14 were investigated for in-vessel effects. Vogtle 15 performed an integrated evaluation of strainer and in-16 vessel failures, evaluated multiple equipment 17 configurations including all pumps available, single 18 train failures and multiple failures. The random 19 equipment failure probabilities were used along with 20 condition of all strainer/core failure probabilities 21 to calculate change in risk.
22 From the strainer acceptance criteria 23 South Texas simplified the analysis by reducing all 24 strainer acceptance criteria down to simple fiber--
25 dependent debris limits. Vogtle used multiple debris NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
98 1 limits along with other relevant acceptance criteria 2 to avoid unnecessary conservatism and to more clearly 3 identify which aspects of the analysis resulted in 4 failures.
5 For time-dependent effects South Texas did 6 not explicitly model time-dependent effects. Vogtle 7 formed a time-dependent evaluation for the strainer 8 and in-vessel evaluation. Including time-dependent 9 effects provides a more realistic assessment. In 10 other words, it avoids unnecessary conservatism and 11 provides better insights on risk-significant 12 contributors.
13 For in-vessel effects WCAP 17788 was not 14 available at the time South Texas submitted, therefore 15 they used thermal hydraulic analysis modeling to 16 address core blockage issues. Vogtle was able to use 17 the methodology debris limits from WCAP-17788 to 18 analyze the in-vessel effects on the latest NRC in-19 vessel review guidance.
20 For the risk contribution for secondary 21 side breaks STP performed a bounding evaluation for 22 secondary side breaks. Initially Vogtle also 23 performed a bounding evaluation assuming the strainers 24 would fail for all secondary side breaks that require 25 ECCS for circulation. Vogtle's bounding analysis NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
99 1 indicated that secondary side breaks were more risk-2 significant than primary side breaks, which we knew 3 was not realistic. So to provide better insights on 4 risk significance of secondary side breaks a more 5 detailed evaluation was performed. From this the 6 ratio of showed secondary side breaks are less risk-7 significant than primary side breaks.
8 For safety margin there were conservatisms 9 used in many aspects of the GSI-191 evaluation. Some 10 of these conservatisms are included to provide 11 operating margin and others can be considered safety 12 margin, however because NRC approval is required to 13 reduce safety margin, SNC felt it was important to 14 clearly identify the conservatisms with their credited 15 safety margin to differentiate other conservatisms 16 that were used as operating margin.
17 And for uncertainty and for sensitivity in 18 earlier submittals STP had done extensive work on 19 sensitivity analysis and uncertainty quantifications.
20 SNC built on lessons learned from STP's work and 21 performed a detailed quantification of parametric 22 model uncertainties. A significant advantage of 23 having an integrated NARWHAL model was that it was 24 very easy to quantify the uncertainties by running 25 sensitivities cases.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
100 1 And for the quality assurance differences 2 South Texas addressed QA requirements after the fact 3 for some aspects of the evaluation in response to NRC 4 comments. Two of the more important aspects of the 5 quality assurance had to be addressed were with the 6 CASA Grande software and the thermal hydraulics 7 modeling. For Vogtle most of the analysis and testing 8 was performed as safety-related under a vendor QA 9 program compliant with 10 CFR 50, Appendix B. The 10 NARWHAL and BADGER software packages were developed 11 and are maintained by Enercon as safety-related items 12 in accordance with Enercon's QA program.
13 And this middle summarizes a few aspects 14 of the evaluation that were not performed under a 15 formal QA program along with the basis for why it's 16 acceptable. As an example, the fiber penetration 17 equations used the NARWHAL conditional failure 18 probability were developed through testing at 19 Alden research labs. Alden has a 10 CFR 50, Appendix 20 B program. Although testing was not officially 21 conducted under the Alden QA program, it's formed 22 using most of the same processes, reviews and 23 procedures as a QA program.
24 Okay. To recap, so far we've discussed 25 the background information. We discussed the timeline NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
101 1 of the South Texas and Vogtle risk-informed 2 resolution. We discussed an overview of the Vogtle 3 risk-informed approach including plant modifications, 4 testing and analyses, risk quantification and the 5 sensitivity uncertainty analyses. We've discussed a 6 summary of Vogtle's technical reports, a summary of 7 the Vogtle LAR submittal, and a summary of the Vogtle 8 -- of the differences between Vogtle and South Texas' 9 risk-informed methodology.
10 With that I'd like to thank the ACRS for 11 their time and ask if there's any additional questions 12 for us to address.
13 CO-CHAIR REMPE: So this is Joy, and thank 14 you for a very comprehensive presentation on this 15 topic.
16 Members, do you have any questions you'd 17 like to bring up at this time?
18 (No audible response.)
19 CO-CHAIR REMPE: I'm giving the -- we have 20 a 10-second rule.
21 (Pause.)
22 CO-CHAIR REMPE: We're about 30 minutes 23 late, but I'm thinking we're not going to have to have 24 the time required for the closed session, so I think 25 that's great.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
102 1 I would like us to go ahead and take a 15-2 minutes break. Even though we're running late I think 3 that this won't be a problem. And so how about we 4 come back at 4:10 after the hour. Does that sound 5 good to everybody?
6 (No audible response.)
7 CO-CHAIR REMPE: Not hearing any response 8 from anyone to complain, I'm going to do that. And so 9 we will take a -- oh, I see, Greg, your hand's up.
10 I'll give you the floor.
11 MEMBER HALNON: No, I was trying to un-12 mute to say that I'm good with that just to give you 13 some affirmation.
14 CO-CHAIR REMPE: Okay. That sounds good.
15 Okay. So let's come back at 4:10 D.C.
16 time, 2:10 my time. Okay? Thank you.
17 (Whereupon, the above-entitled matter went 18 off the record at 3:55 p.m. and resumed at 4:10 p.m.)
19 CO-CHAIR REMPE: Okay. So this is Joy 20 Rempe and I have that it's 4:10, and so we're ready to 21 resume this meeting and I'll ask the NRR staff to take 22 over the floor.
23 MS. CARUSONE: Thank you, Chairman Rempe.
24 My name is Caroline Carusone. Nice to see 25 everybody virtually. I'm the Deputy Director of the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
103 1 Division of Operating Reactor Licensing in the Office 2 of Nuclear Reactor Regulation.
3 As you know, there's decades of history 4 related to Generic Safety Issue 191 and Generic Letter 5 2004-02. Matt Horn of SNC provided a detailed 6 timeline as part of SNC's presentation before the 7 break, and by way of introduction into NRC's 8 presentation I'm going to briefly highlight and recap 9 a few key activities related to this effort.
10 In September of 2019 the NRC issued a 11 staff evaluation of Vogtle's systematic risk-informed 12 assessment of debris technical report. This provided 13 the basis for us to consider the use of the technical 14 report for Vogtle in future licensing applications.
15 The NRC staff concluded that the technical report 16 contained information to address the NRC Generic 17 Letter 2004-02, except downstream effects.
18 In August of 2020 Southern Nuclear 19 submitted a license amendment request and an exemption 20 to revise their licensing basis to allow for the use 21 of a risk-informed approach to address safety issues 22 discussed in GSI-191 that you just heard about. This 23 was supplemented in December of 2020 and again in 24 February of 2021.
25 In addition, the proposed license NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
104 1 amendment request would add a new technical 2 specification action to address the condition of a 3 containment sump made inoperable solely due to the 4 loss-of-coolant-accident-generated debris exceeding 5 the analyzed limits.
6 The proposed exemption would be from 7 certain requirements of 10 CFR 50.46, requirements 8 associated with the emergency core cooling system 9 following a postulated loss of coolant accident.
10 This review is different from typical 11 reviews in that we already reviewed the majority of 12 the technical information in the September 2019 NRC 13 staff evaluation. In NRC's staff evaluation we 14 identified areas that still needed to be addressed.
15 The licensee's August 2020 license amendment request 16 included specific responses to the open issues in 17 addition to other licensing requests to include the 18 technical specification changes and exemptions.
19 Chairman, as you mentioned in your opening 20 remarks we look forward to hearing the ACRS' opinion 21 on this matter. The NRC does not need a letter report 22 from ACRS to continue our review if the members deem 23 one is not needed, however the NRC staff also welcomes 24 a letter report should the ACRS members decide to send 25 one.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
105 1 We really appreciate your interest and 2 feedback on this topic, and with that quick intro I'll 3 turn it over to Steve Smith of the Technical 4 Specifications Branch in the Division of Safety 5 Systems in NRR.
6 MR. SMITH: Okay. Thanks, Caroline.
7 As Caroline said, this is Steve Smith and 8 I'll be starting off our presentation and then when we 9 get somewhat into it Odenayo Ayegbusi is going to --
10 from DRA will be presenting the slides on the 11 systematic risk assessment. We also had help -- DRA 12 and DSS had help from other divisions in NRR, 13 particularly DNRL. Other folks in DSS and DEX when we 14 did -- in doing this review.
15 If you have questions -- I think it's been 16 going fine. If you use the chat or raise your hand, 17 I may not see that comment, so just go ahead and speak 18 up like you've been doing. I think that was working 19 good.
20 And also we developed this presentation 21 without knowing what was in the Vogtle presentation, 22 so this may get a little bit repetitive, and if it 23 seems like it's too repetitive, just let me know and 24 I'll move along. I'll skip slides or whatever seems 25 good to the Committee.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
106 1 And thanks to SNC for providing a thorough 2 discussion of the evaluation.
3 And, Andrea, I'm ready for the next slide.
4 This slide just -- it's an outline of what 5 we're going to present today. We're going to give 6 some background on the overall GSI-191 issue. We're 7 going to describe actions that industry is taking, do 8 a comparison between the Vogtle and STP methodologies, 9 which I think Vogtle did a very thorough job of that 10 already. So we might save some time there. And we 11 will discuss the staff evaluation of the Vogtle 12 submittals which includes the technical report and the 13 LAR, and the LAR also included some technical areas.
14 All right. Next slide. So a little bit 15 of background on the ECCS debris effects on sump 16 recirculation of long-term core cooling. This issue 17 has been around for a while, as I'm sure ACRS is well 18 aware. The bottom line that we went to get to; and 19 we're getting pretty close now, is to have PABRs 20 demonstrate compliance with 10 CFR 50.46 considering 21 the potential effects of debris on ECCS recirculation 22 and long-term core cooling. GSI-191 did not identify 23 in-vessel or chemical effects as issues. Those came 24 up later and those are being addressed by sites in 25 their plant responses to Generic Letter 2004-02.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
107 1 Next slide. This slide provides the major 2 actions that have been taken by licensees to address 3 the issue. All plants installed larger strainers and 4 made administrative changes to reduce debris sources 5 in containment, and that includes controls to ensure 6 that their design-basis is maintained. What I mean by 7 that is they've done two basic things: First, they've 8 implemented controls to remove and keep debris sources 9 out of containment. And second, they have implemented 10 measures and modification procedures that anything 11 that they install in containment gets evaluated to see 12 if it could be a debris source.
13 Other actions that were taken by some 14 plants; not all plants, but some plants changes 15 insulation to less problematic materials. Some made 16 insulation systems more robust. Some improved their 17 chemical effects attributes by changing their sump 18 buffers or removing calcium and aluminum sources. And 19 that's pretty -- those are the major things that were 20 done.
21 So we're ready for the next slide. Okay.
22 The overall status of compliance. SECY-12-0093 23 provided three options for closeout of the issue, and 24 the SRM, the Staff Requirements Memorandum, approved 25 those three options. Nineteen units so far have NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
108 1 closed out the issue using option 1, and the remainder 2 of the plants chose option 2. And the only plant that 3 has closed out under option 2 so far is South Texas, 4 and they were the pilot project. They were the pilot 5 plant for the risk-informed methodology. So all the 6 other option 2 plants are still working on this.
7 Ready for the next slide. So the option 8 2 plants, I will say -- let me -- I'll just say right 9 now no one's using option 3. So we basically had 10 option 1 and option 2. Option 2 got split up into two 11 different categories: 2A and 2B. 2B are the risk-12 informed plants. So this slide shows to the best of 13 the staff's knowledge what -- how plants are planning 14 to close this out. It's not likely that many plants 15 will change the methodology they're planning, but it 16 is possible. For example, Point Beach just switched 17 to option 2B.
18 So this slide basically just shows that 19 most plants are not going to use a risk-informed 20 approach. There's only three plants left after 21 Vogtle. STP's complete. Vogtle is the one we're 22 discussing today. And then we have the other three:
23 Wolf Creek, Point Beach and Callaway.
24 Next slide. Okay. This just gives us a 25 break from having to look at a lot of words; get a few NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
109 1 pictures. This lists the strainer vendors and it 2 shows some of the designs. And all the designs are 3 intended to provide a large surface area and a given 4 volume. On the upper right is a PCI strainer 5 installed in a plant. The lower left is also a PCI 6 strainer. The center, the lower center is an Enercon 7 top hat strainer after testing. And the lower right 8 is a CCI pocket strainer after testing.
9 And the pictures you see here, they're 10 small portions of the strainers installed in the 11 plant. These are relatively large arrays that are 12 installed in the plant. Both the top hat and the 13 stacked disc strainers can be installed either 14 vertically or horizontally depending on the plant-15 specific application. And under non-uniform flow you 16 see AECL also has an asterisk next to it, and that is 17 because they make both uniform and non-uniform flow 18 strainers.
19 Okay. Next slide. All right. This slide 20 discusses the status of plant responses for the in-21 vessel downstream effects. Nineteen option 1 units 22 closed out in-vessel. These are the plants that have 23 low fiber amounts and they demonstrated that they can 24 meet conservative in-vessel debris acceptance limits 25 from WCAP-16793. South Texas also closed out in-NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
110 1 vessel, but they used a plant-specific analysis to 2 address the in-vessel issue. And the remaining plants 3 plan to use the staff review guidance for in-vessel, 4 which is what Vogtle is doing. And they discussed 5 that earlier today.
6 Next slide. Okay. This slide provides a 7 summary of the purpose, scope and the staff use of the 8 WCAP-17788, which was the -- it is the large program 9 that was undertaken by the PWR Owner's Group to 10 address in-vessel fiber limits for the plants that 11 couldn't meet the very conservative 16793 limits.
12 The other thing is that boric acid 13 precipitation was not addressed in WCAP-16973, but it 14 was considered in 17788. And the WCAP was a 15 significant effort. I'd say it was a very significant 16 effort. It involved a lot of testing and analytical 17 work. And the WCAP was submitted in six volumes, each 18 for a different part of the evaluation.
19 In the long run the NRC staff did not 20 approve the WCAP, but we did conclude that it provides 21 a significant understanding of the phenomena and the 22 issues and it provides good predictions of how the 23 plant's going to behave. And we performed (audio 24 interference) analyses to gain understanding and to 25 confirm that the industry analyses were relatively NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
111 1 accurate.
2 Next slide. Okay. This is a comparison 3 of Vogtle and the South Texas project. I know that 4 Vogtle said that they didn't use a RoverD-type method, 5 but we'll talk about that on the next slide. I think 6 that we can say that they were both RoverD-type 7 methods and the next slide will kind of illustrate 8 that.
9 Both used CAD models to map debris sources 10 in containment and calculate debris generation. CASA 11 Grande and NARWHAL were the programs that calculated 12 some of these aspects of debris transport, debris 13 generation. BADGER did combine with NARWHAL in order 14 to get the debris generation for the Vogtle-type 15 analysis.
16 I would say that the NARWHAL 17 implementation for scenario evaluations were more 18 refined as Vogtle did discuss. STP simply used a 19 limiting fiber value and assumed that all breaks that 20 generated and transported this amount of fiber or more 21 to the strainer contributed to delta CDF, whereas 22 Vogtle's implementation performed time-based 23 calculations for each scenario and compared several 24 parameters against the appropriate acceptance criteria 25 at each time step. And if at any time step the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
112 1 parameter exceeded the acceptance criteria, then it 2 was considered to contribute to changing core damage 3 frequency.
4 The NARWHAL method Model 8 break 5 orientations (audio interference) CASA Grande Model 6 360 break orientations for partial breaks have -- we 7 can discuss this later in another slide if we want to 8 get into that.
9 In-vessel treatment. STP used a plant-10 specific thermal hydraulic analysis and Vogtle's using 11 the staff guidance for the acceptance criteria for in-12 vessel. And both sites used deterministic methods to 13 evaluate upstream effects, downstream ex-vessel 14 effects, latent debris amounts and structural limits, 15 and probably a couple other things. There was a lot 16 of deterministic methods mixed in with the risk-17 informed methodology. And that is it for that slide.
18 So this is just talking about RoverD, and 19 I guess my view of RoverD is a lot more simplified.
20 Vogtle's presentations stated that they used a 21 conditional failure probability approach. So I think 22 the difference between Vogtle and STP would be the 23 diamond where it says debris calculated less than or 24 equal to testing limit. For STP that was true. For 25 Vogtle it's true. There's a lot more that goes into NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
113 1 that box. So I think that's the easiest way to say 2 it, and we've kind of talked about that.
3 So any scenario that -- the scenarios that 4 pass all deterministic criteria and don't contribute 5 to risk or increase in core damage frequency go to the 6 blue box and any scenarios that fail -- all scenarios 7 that fail any of the deterministic acceptance 8 criteria, they do contribute to change in core damage 9 frequency and they go to the brown box. And that's 10 where you get the green box, R over D, or RoverD. And 11 that is it for that slide.
12 Okay. This is talking about the staff 13 evaluation, the technical report, a high-level slide.
14 We wrote a Safety Evaluation Report to document our 15 findings on the technical report. Our Safety 16 Evaluation Report identified areas that required 17 additional information or input from the licensee and 18 identified these as limitations and conditions. And 19 in our evaluation of the technical report we evaluated 20 the five key principles of risk-informed regulation.
21 We have a few slides that talk in more detail about 22 our evaluation. And the items that were identified 23 that needed further evaluation as far as limitations 24 and conditions that were addressed in the licensee's 25 LAR.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
114 1 Next slide. Okay. This talks about a 2 high-level evaluation of the LAR. The LAR was more 3 about licensing issues and had a little bit of follow 4 up on a few open technical issues. The major areas 5 evaluated in the staff evaluation in the LAR are the 6 tech spec changes, the exemption request and then the 7 LNCs from the staff SER on the technical report, which 8 were mostly administrative, but there were a couple of 9 technical things that were addressed in the LNCs. The 10 staff evaluation of the LAR will also be discussed in 11 more detail in later slides.
12 Next slide. We just wanted to list the 13 major references that were used by the NRC staff and 14 by Vogtle in the evaluation just so that we would have 15 most of them in one place.
16 Next slide. Okay. The five key 17 principles of risk-informed regulation: These were 18 discussed a little bit by Vogtle, but we'll go over 19 them again kind of discussing which groups were 20 responsible for evaluating each one. Now there is 21 overlap in staff responsibilities for the review of 22 each area. This is just kind of a high level of which 23 division worked on which principle.
24 So Key Principle 1 is that the change is 25 consistent with regulation or request an exemption.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
115 1 And DORL was mainly responsible for Key Principle 1.
2 And as you know, Vogtle requested an exemption to 10 3 CFR 50.46. And DORL is mostly in charge of it and 4 then the staff of course is doing the technical 5 evaluation as to whether that exemption is acceptable 6 or not.
7 Key Principles 2 and 3 are that the change 8 is consistent with the defense-in-depth philosophy and 9 that it also maintains sufficient safety margins. And 10 DSS and DNRL did the majority of those evaluations, 11 all those safety margins and DID are addressed in 12 accordance with Reg Guide 1.174. They are addressed 13 using deterministic evaluations and engineering 14 judgment.
15 Key Principle 4 is that any increase in 16 risk is small, and Principle 4 is addressed using both 17 deterministic and risk-informed methods. As we 18 discussed on the RoverD slide only scenarios that fail 19 the acceptance criteria contribute to risk. And most 20 of the technical groups helped with the evaluation of 21 Key Principle 4, but DRA did the actual risk 22 evaluations.
23 Key Principle 5 is that the impact of the 24 changes monitor and use performance measurement 25 strategies, and DRA's evaluation made sure that NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
116 1 adequate performance measurement strategies are 2 incorporated by the licensee. And the following 3 slides discuss each of the key principles and provide 4 additional details.
5 Okay. This slide talks about meeting the 6 regulation or requesting an exemption, Key Principle 7 1. And I'm not going to talk too much specifically 8 about the exemption request, but I'll give a little 9 bit of background on 50.46.
10 We've been working -- the staff has been 11 working on a rulemaking for 10 CFR 50.46(c) for 12 several years. In the process of its review of GSI-13 191 and Generic Letter '04-02 issues the Commission 14 directed that staff include a section that allows 15 licensees to use risk-informed methods to evaluate the 16 effects of debris on long-term core cooling. The 17 Commission request was in addition to the original 18 rulemaking, 50.46(c), that was intended to deal with 19 short-term LOCA effects on the fuel.
20 So we did write the risk-informed part of 21 the rule. And the rule was sent to the Commission for 22 a vote a couple years ago and it's still under 23 Commission consideration as of this time. Only one 24 vote has been submitted.
25 Since 50.46(c) has not been approved yet, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
117 1 exemptions are still required for the risk-informed 2 evaluations and it's our feeling that it's unlikely 3 that 50.46(c) will be implemented before all of the 4 operating PWRs close Generic Letter '04-02. As we 5 discussed before, there's only three more plants after 6 Vogtle that we think are going to use a risk-informed 7 methodology.
8 Just as a footnote, if the rule does get 9 approved and it maintains the risk-informed portion 10 for evaluating debris for long-term core cooling, it's 11 not intended to allow new plants to use risk-informed 12 methods. We expect the new plants to come in 13 basically as clean plants, so they don't have to be 14 concerned with the effects of debris on recirculation.
15 MEMBER BLEY: Excuse me.
16 MR. SMITH: Yes.
17 MEMBER BLEY: This is Dennis Bley. Do you 18 have any hints about what's hanging up with the 19 Commission voting on this rule? Are there real issues 20 that have arisen?
21 MR. SMITH: I don't really -- I don't have 22 any insights into why it has taken so long. It would 23 be pure speculation on my part.
24 MEMBER BLEY: I won't ask you to do that.
25 Thanks.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
118 1 MR. SMITH: Okay. Let's go to the next 2 slide. Okay. We discussed safety margin and defense-3 in-depth, or you did and Vogtle did earlier today.
4 The treatment of DID and safety margins is discussed 5 and evaluated in our SER. We had input from several 6 technical branches on these including DRNL, DSS, DRA 7 and DEX. And we don't intend to discuss these areas 8 in detail, but if there are any questions about 9 defense-in-depth and safety margins, we can try to 10 answer those.
11 MEMBER HALNON: Yes, this is Greg. I was 12 just curious if there was any one or two design 13 features that was different than other plants that we 14 were able to get what you call significant safety 15 margins in this plant. I know that the industry has 16 been struggling with this for a decade or more, but 17 why are we able to get significant safety margins and 18 defense-in-depth here as opposed to other plants that 19 have struggled with this thing for so many years?
20 MR. SMITH: Okay. So for Reg Guide 1.174, 21 which is only being used by the plants that are 22 implementing the risk-informed method -- so STP, 23 Vogtle, Wolf Creek, Callaway and Point Beach, the ones 24 we know of now anyway -- STP and Vogtle are the only 25 ones who have come in and discussed these things.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
119 1 So the other plants, since they're not 2 using a risk-informed method, they're not required to 3 address safety margins and defense-in-depth. They 4 have done a lot of that just because in the interim 5 before they are able to show that they are compliant 6 with 50.46 the Commission has asked them to take 7 mitigative measures. And one thing, they've all put 8 the new strainers in and things like that, but they've 9 also done a lot of things with operational -- changing 10 operational practices whereas if they do have -- get 11 into an EOP situation where they have a LOCA they 12 monitor the strainers and they take action if they see 13 the pumps are not working. Those are the kind of 14 defense-in-depth things.
15 But anyway, we define safety margins as 16 things that if you have -- say for structural strength 17 of the strainer a safety margin is that you're using 18 a code which has margin built into it when you 19 evaluate that. Or another thing that provides margin 20 is our approved methods for doing the evaluation.
21 Those have conservatism built into them because we 22 want to be sure that the absolute worst conditions are 23 considered when they do these analyses.
24 So those are where we come up with safety 25 margins. I don't think I'm really answering your NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
120 1 question. And I don't think that there was anything 2 special about Vogtle that gives them more safety 3 margins than any other plant would have.
4 MEMBER HALNON: Okay.
5 MR. SMITH: They just documented it in 6 their submittal and we reviewed it.
7 MEMBER HALNON: So just let me frame it a 8 different way. You would expect similar safety 9 margins at other plants if they've taken this route 10 given the present technology?
11 MR. SMITH: Yes, I think if you look at 12 the safety margins and defense-in-depth that were 13 referenced by South Texas, it would be quite similar 14 to what Vogtle has.
15 MEMBER HALNON: Okay. So that just kind 16 of gives a testament to the conservative nature of the 17 GSI-191 issue in whole, don't you think? Is that 18 fair?
19 MR. SMITH: I think that's fair and I 20 think it's also -- as a regulatory agency most of our 21 acceptance criteria are based on somewhat conservative 22 codes and models and things like that, so it all kind 23 of feeds into safety margins.
24 MEMBER HALNON: Yes, well, and it actually 25 speaks to the length of time it's been out there that NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
121 1 it hasn't really been a percolating safety issue to 2 the extent that it sounds like given the fact that it 3 looks like there's significant safety margins exist in 4 the industry. Just some of them aren't just as well 5 defined as these risk-informed plants.
6 MR. SMITH: Right. And speaking directly 7 to that, in one of the SRMs the Commission's I believe 8 -- I'm not going to give their exact words, but they 9 said there is adequate defense-in-depth that they can 10 take, that the plants can take time to show that they 11 meet 50.46.
12 MEMBER HALNON: Okay. All right. Thanks.
13 MEMBER KIRCHNER: And, Steve, this is Walt 14 Kirchner. Your third bullet there, this looks like 15 Mary Drouin's Knowledge Management 009 NUREG. How do 16 you in practice use that?
17 MR. SMITH: The third bullet for defense-18 in-depth includes actions?
19 MEMBER KIRCHNER: Yes.
20 MR. SMITH: Okay. The balance is 21 maintained among prevention mitigation. Redundancy is 22 -- I'm not sure I understand the question.
23 MEMBER KIRCHNER: The second sentence 24 almost looks like a quote from Mary Drouin's Knowledge 25 Management 009 NUREG report, but in practice when we NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
122 1 get to later in your presentation how -- are you going 2 to have a slide that addresses how you assessed 3 adequate defense-in-depth?
4 MR. SMITH: I don't have one, but we can 5 talk about that.
6 MEMBER KIRCHNER: Yes, if you would. It's 7 a leading question again. I asked it also of -- from 8 the Vogtle presentation.
9 MR. SMITH: So as far as defense-in-depth 10 is concerned, it's probably just as well to talk about 11 it now.
12 So some of these -- I did say that we used 13 engineering judgment for some of these, and I think 14 it's one of the things where you would say you'd have 15 to use engineering judgment. It says a reasonable 16 balance is preserved, right, between the prevention of 17 core damage, containment failure and consequence 18 mitigation.
19 So what we did is we looked at what Vogtle 20 has done. They've performed physical modifications, 21 they've done procedure changes, they put the bigger 22 strainers in, they increased their RWST inventory to 23 increase the sump pool level. And then they have 24 trained their operators to actually look for issues 25 that might occur if you did have a LOCA. So they're NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
123 1 trying to look at this in -- I think it's like a 2 holistic manner. And so they're doing that.
3 And then the other thing for that one is 4 prevention of containment failure. And for that they 5 also have containment air coolers which even if you 6 lost containment spray, you would have these air 7 coolers which would still function to reduce the 8 pressure in containment.
9 So these are some of the things we looked 10 at when we're talking about balance preserved among 11 core damage, prevention of containment failure and 12 consequence mitigation.
13 MR. VASAVADA: Steve?
14 MR. SMITH: Yes?
15 MR. VASAVADA: So this is Shilp Vasavada.
16 I'm with the NRC. If I can just add to that?
17 What you're seeing in the third bullet, it 18 says et cetera, but it actually comes from Reg Guide 19 1.174. So 1.174 provides like seven considerations 20 that together are looked at by the staff to see if the 21 proposed change maintains adequate defense-in-depth.
22 So Steve talked about the first one, reasonable 23 balance, but there are others.
24 For example, is there an over-reliance on 25 compensatory measures as opposed to engineering NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
124 1 features? And in case of this particular case, for 2 example, there are no new operator actions that are 3 being introduced by this change. In fact, they have 4 not even taken credit for operator actions in their 5 evaluation.
6 Another consideration would be are there 7 new common-cause failure mechanisms? This one doesn't 8 introduce new common-cause failure mechanisms and the 9 common-cause failure of the strainer is in fact 10 explicitly considered in the risk assessment.
11 So there are seven considerations in 1.174 12 which were looked at as part of the review and our 13 safety evaluation provides our findings on each and 14 every one of those seven considerations to, again as 15 Steve mentioned, holistically make the case that the 16 methodology would maintain appropriate defense-in-17 depth.
18 MEMBER KIRCHNER: Yes, thank you. So but 19 it still remains largely qualitative in terms of 20 checking the box so to speak, whereas the safety 21 margins one is more amenable, at least for LWRs, to 22 the variation in the delta CDF or LERF.
23 (Simultaneous speaking.)
24 MR. VASAVADA: This is Shilp again.
25 MR. SMITH: Yes, go ahead, Shilp.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
125 1 MR. VASAVADA: No, I was going to say for 2 the most part you are right that it would be a 3 qualitative case for certain items. For example, the 4 common-cause failure that we talked about. It is 5 specifically considered through the conditional 6 failure probability, so that you can say is addressed 7 and quantified. But there are other elements which 8 are not amenable to easy quantification.
9 MEMBER KIRCHNER: Thank you.
10 CO-CHAIR REMPE: So this is Joy and I 11 guess I want to follow up on Steve's response to Greg.
12 Your SER, as well as the licensee's submittals, talk 13 about the fact that each of the pumps have their own 14 strainer. And I would have thought you might have 15 mentioned that, that might have increased safety 16 margin or it did not? I mean, it gives them more 17 flexibility, but it did not impact the safety margin?
18 MR. SMITH: Well, I there's advantages and 19 disadvantages to having separate or single strainers.
20 The advantage of having separate strainers is that if 21 one fails, then if the other ones fail, you still have 22 another one that's there to operate, whereas if you 23 have one large strainer, if it fails, you're in 24 trouble. However, if only one RHR pump is running and 25 it has a relatively small strainer on it, it's a lot NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
126 1 more likely to get clogged with debris than if you 2 have a large strainer that feeds all the pumps with 3 only RHR pump running. So there's various ways you 4 have to look at that.
5 But, yes, I think it does improve 6 redundancy to have more than one strainer, but some 7 plants have two. Like STP was in really good shape 8 because they had three and they have three trains of 9 ECCS. So each plant is different with that and 10 there's pluses and minuses to each configuration.
11 CO-CHAIR REMPE: Okay. Thank you.
12 MEMBER HALNON: So sorry, just one other 13 question. I don't know if it's the right place for it 14 or not, but Joy got my brain working.
15 But my question really on the low flow for 16 the pumps was -- I didn't get a chance to fully vet 17 that out, but I was thinking through the containment 18 spray nozzles and the debris blockage of those. Was 19 that considered in this, quote, safety margin 20 discussion whether or not there are other downstream 21 effects? I know we place a lot of interest in the 22 vessel, but what about the spray nozzles themselves?
23 Are they a concern at all?
24 MR. SMITH: Yes. Well, the spray nozzles 25 are of concern, but the spray nozzles are way -- the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
127 1 holes in the spray nozzles are way bigger than the 2 perforations in the strainer. There is a downstream 3 ex-vessel effects evaluation that's done by all the 4 plants. It's deterministic. It's not risk-informed.
5 It's one of the deterministic things that both STP and 6 Vogtle did. And it looks at other things that are 7 more likely to have issues than the spray nozzles.
8 For example, STP mentioned that they had to install 9 some orifices in their HPCI injection line so that 10 they could open the injection valves larger so that 11 they -- or rather so that they would have larger 12 openings and they wouldn't become blocked. But they 13 also look at pump seals and pump wear and things like 14 that.
15 MEMBER HALNON: Okay. I recall from my 16 early days in operating plants that the nozzles, at 17 least the plants I was involved with, were about a 18 quarter inch.
19 MR. SMITH: Yes, or three-eighths of an 20 inch. That's right. And the perforations are around 21 a tenth of an inch.
22 MEMBER HALNON: Okay. I didn't realize 23 that. Thanks.
24 MR. SMITH: Okay. This slide we aren't 25 going to discuss much. We're ready to go to the next NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
128 1 one. All right. So this -- now we're starting to get 2 into where we have a little bit more detailed 3 discussion. I'll try not to take too long on these.
4 This discusses the risk-informed methodology and the 5 deterministic inputs in the staff evaluation. In this 6 one we're starting out with debris generation.
7 Vogtle performed evaluations for thousands 8 of potential debris generation and transport scenarios 9 to determine the predicted fiber and particular 10 amounts that would arrive at the strainer. The breaks 11 were assumed to occur at welds and other non-weld 12 potential break locations were evaluated, although not 13 explicitly. They were more evaluated using 14 engineering judgment.
15 Debris amounts that reached the strainer 16 based on each scenario were compared against the 17 tested amounts. So you know that they ran tests.
18 We'll talk about the tests a little bit later, but 19 they were -- for each scenario they figured out how 20 much debris got to the strainer and they compared it 21 against tests that were done with various amounts of 22 debris.
23 This is different from typical non-risk-24 informed evaluations because what those normally do is 25 they identify one or two limiting break locations and NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
129 1 they evaluate the debris generation and transport for 2 those. And sometimes they may combine two or more 3 limiting cases just to be sure that they have the 4 potential largest amount of debris in there including 5 various different kinds of debris in their tests and 6 they will develop a head loss program based on that.
7 Anyway, by automating the debris 8 generation and transport calculations using a CAD 9 model the plants like STP and Vogtle, they're able to 10 do debris generation and transport calculations for 11 thousands of breaks and partial breaks and double-12 ended guillotine breaks, whereas these other plants, 13 they just assumed -- they just went and said, okay, 14 this is where most of our debris is; we're going to 15 look here. So it's quite a bit different from a 16 typical non-risk-informed debris generation 17 evaluation.
18 The debris generation and transport models 19 that were used in the NARWHAL program were developed 20 using staff-approved guidance and assumptions. And we 21 audited and evaluated the use of the models, although 22 there are some -- and we also looked at -- because --
23 I guess it's probably unfortunate for NARWHAL, but the 24 CASA Grande, which STP used -- they did 360 break 25 orientations and NARWHAL used 8. So we did an NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
130 1 evaluation of that and they did some parametric 2 studies and showed that there might be a small 3 increase in debris generation if you only used eight 4 orientations. It's not a significant effect.
5 Let's see what else do we have? Yes, 6 okay. Vogtle, they postulated partial breaks and 7 double-ended guillotine breaks at each weld. And one 8 thing that we did that really helped us out with this 9 review is we got Southwest Research to verify the 10 calculations that were done in the NARWHAL and BADGER 11 programs, and they used a combination of independent 12 calculations and examination of the outputs for 13 consistency. And I think that's about it for this 14 slide.
15 The next -- hopefully the next slide --
16 yes, not so many words. This slide and the next slide 17 help to illustrate why the risk-informed methodology 18 works. This particular slide shows new kind debris 19 generation; this is for Vogtle, amounts is affected by 20 break size. And these are all partial break amounts.
21 This slide shows your minimum, maximum and averages 22 for each one of these breaks sizes.
23 Only very large breaks can generate enough 24 debris to cause a strainer failure. And there are 25 fewer large welds than small welds, and also the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
131 1 larger welds are less likely to fail than the smaller 2 welds. So the probability of failure is really down 3 in these smaller breaks that are towards the bottom of 4 the chart here.
5 Just for a little bit of -- I guess a 6 little bit of background, if 114 cubic feet of fiber 7 reaches the RHR strainer, it would be considered to 8 fail. Usually there are two RHS strainers in service, 9 so you would normally have to have double that amount 10 going only to the RHR. And then if containment spray 11 is running, the debris would be spread over those 12 strainers as well.
13 The amounts you see in these graphs are 14 the generated amounts, so they get reduced 15 significantly because some of the debris gets held up 16 during the transport, probably half or more. It 17 depends on the situation. Usually large and intact 18 pieces of fiber won't transport to the strainer and 19 some small pieces will transport, but all of the fine 20 debris is assumed to transport. So most 20-inch break 21 scenarios would not result in a failure due to 22 exceeding fibrous debris limits. Now I think that 23 actually the only breaks that will fail a fibrous 24 debris limit are those on the main loop. So other 25 attached piping would not fail a fibrous debris limit NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
132 1 for Vogtle.
2 If we look at the next slide, this shows 3 similar information for the guillotine breaks. The 4 raw debris amounts listed here are reduced during 5 transport and split up between operating strainers.
6 The double-ended breaks of course have the maximum 7 amounts of debris. They're going to have more debris 8 than a partial break. The very large breaks are 9 unlikely to occur and the smaller more likely breaks 10 -- more likely break -- debris loads can be handled by 11 the strainer.
12 Next slide. Okay. This slide discusses 13 debris transport. Transport to the strainer was 14 calculated implementing NRC approved methods, and 15 those were implemented in NARWHAL. Basically logic 16 trees and CFD were used as Vogtle showed earlier.
17 They used conservative values from testing as inputs 18 to the in-vessel fiber calculations and they 19 calculated the amount of fiber that could arrive at 20 the core inlet based on various pump combinations.
21 Hot let breaks resulted in limiting fiber amounts to 22 the core.
23 In general for in-vessel the fewer pumps 24 that are running the more fiber will reach the core 25 because less flow along with debris that would be NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
133 1 carried in the flow goes out the break and the 2 containment spray nozzle and gets returned back to the 3 strainer where it could be captured. And we also had 4 Southwest Research perform independent calculations to 5 validate results of both the strainer and the in-6 vessel transport models.
7 Next slide. Okay. This talks about the 8 head loss caused by the debris. Vogtle performed 9 strainer head loss tests with known amounts of debris.
10 So the test provided head losses that were associated 11 with various amounts of particulate, fibrous and 12 chemical debris on the strainer and those debris 13 amounts and head losses were used as acceptance 14 criteria in NARWHAL. If a scenario transported more 15 than the amount of debris that was included in the 16 test, it was automatically considered to contribute to 17 any core damage frequency. If it did not, then the 18 head losses from the tests, the lead loss that was 19 associated with the tests, those debris values were 20 assessed against different things like NPSH strainer 21 structural abilities, flashing, things like that.
22 So they used a rule-based approach to 23 determine what the head loss at the strainer was. The 24 first thing they did is if there's any flow through 25 the strainer, the clean strainer head loss was NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
134 1 included as a strainer head loss. If there's any case 2 with greater than zero but less than the full load 3 debris amount, then they would include the thin bed 4 head loss. The full load head loss was added for any 5 case where you had greater than the thin bed amount, 6 which is 0.57 inches of fiber on the strainer. And 7 then calcium phosphate head loss would be added if the 8 amount of fiber on the strainer is greater than 0.45 9 inches. And sodium aluminum silicate would be added 10 if its saturated limit is predicted to be reached or 11 at 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, whichever is first and you had greater 12 than 0.45 inches of fiber on the strainer 13 Vogtle -- and what they did evaluate was 14 NPSH margin, structural margins, flashing and 15 deaeration at the dP associated with the tested debris 16 loading. This was kind of confusing for the staff.
17 We had some RAIs about this and finally got it figured 18 out. We agreed that the way that they implemented 19 their head losses was good.
20 Next slide. Okay. Then this talks about 21 the impact of debris on the vessel. The in-vessel 22 evaluation was performed following the recent staff 23 guidance. They used the same methods and acceptance 24 criteria that the non-risk-informed plants are using 25 and we found that this area was adequately addressed NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
135 1 and all scenarios passed the in-vessel acceptance 2 criteria. There's no increase in risk associated with 3 any in-vessel debris effects.
4 At this point Ayo's going to be presenting 5 the slides on the systematic risk assessment and then 6 I will come back and wrap things up. And this might 7 be a good time -- I'll ask if there's any questions on 8 what we've heard up to now before Ayo takes over.
9 (No audible response.)
10 MR. SMITH: All right. Ayo, I guess it's 11 all yours.
12 MR. AYEGBUSI: All right. Thanks, Steve.
13 So good afternoon to the ACRS Subcommittee. My name 14 is Odunayo Ayegbusi. Please let me know if you can 15 hear me or if there are any issues.
16 Previously, I was a risk and reliability 17 analyst in DRA up until last month before I moved over 18 to DRO, the Division of Risk Oversight. I'm sorry, 19 Reactor Oversight.
20 So my first two slides here are just going 21 to be a summary of Vogtle's systematic risk 22 assessment, most of which you've already heard from 23 Vogtle's presentation. And then the slides after that 24 will just be the staff's review of the assessment.
25 So as you've probably heard, as you heard NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
136 1 earlier, Vogtle was looking to respond to Principle 4 2 of Reg Guide 1.174, and so in that case they used the 3 existing acceptance guidelines of that reg guide.
4 They determined that only large break 5 LOCAs contribute to strainer failure or core damage.
6 Here I have -- I put 12 inches, only to just indicate 7 that, you know, we typically see large break LOCAs of 8 12 inches are greater. This is not to say that they 9 start to see failures at 12 inches.
10 Lastly on this slide, they determined that 11 delta CDF and delta LERF are --
12 CO-CHAIR DIMITRIJEVIC: Just a second.
13 Isn't the large break LOCA about six inches? I mean, 14 you just said it's usually about 12 inches, but that's 15 not what it usually is.
16 MR. AYEGBUSI: You are correct, Vesna.
17 I'm sorry. So the large break LOCA is above six 18 inches. In this case, this 12 inches refers to --
19 Steve spoke earlier about NRC's consultants and the 20 evaluation that they did. And in that case, they used 21 12 inches, right, for their evaluation, right?
22 So what I meant to say was the 12 inches 23 was from the NRC's confirmatory calculations, not 24 Vogtle's calculations.
25 CO-CHAIR DIMITRIJEVIC: Does that mean NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
137 1 they introduce new event in the PRA model? I mean, 2 they have event within six inches and 12 inches, and 3 then it was 12 inches.
4 MR. AYEGBUSI: No, there isn't. So what 5 I'm saying is, you're correct, typically -- I 6 misspoke. Typically, large break LOCAs are greater 7 than six inches, right? What I wanted to indicate 8 here was Vogtle determined that, you know, they find 9 failures for larger breaks of LOCAs, right?
10 Some of the minimum breaks were 12 inches 11 and above, right, which is what the NRC used in our 12 non-confirmatory calculations.
13 CO-CHAIR DIMITRIJEVIC: Okay. Because 14 based on figure, what they show us, there are no --
15 only large break LOCA larger than 20 inches 16 contribute. And so --
17 MR. AYEGBUSI: So that --
18 CO-CHAIR DIMITRIJEVIC: -- now I'm 19 confused by this 12 inches, but okay.
20 MR. AYEGBUSI: I listened to that 21 presentation, so what he was pointing out is that --
22 I'll call it a graph or -- that graph was showing you 23 one of I believe seven configurations on pump 24 configurations, right? I think that was in the 25 configuration where there were -- no pumps failed, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
138 1 right?
2 There are periods around 12 inches for I 3 believe the configuration when you have one RHR pump 4 and one continuous spray pump fill. So there are 5 multiple high likelihood configurations.
6 CO-CHAIR DIMITRIJEVIC: Okay. So the 7 thing about they show up is for the -- the 8 configurations which are worse than that. That's what 9 -- it was mostly -- is that what you are telling me, 10 that their configuration --
11 MR. AYEGBUSI: No. What I'm telling you 12 is the opposite, right? The figure they showed is for 13 -- for a configuration is better than in the case 14 we're talking about.
15 CO-CHAIR DIMITRIJEVIC: Yeah. That's what 16 I meant is -- is the configuration where the CFP will 17 be different than zero for smaller -- all right.
18 Okay. I understand what you're saying.
19 MR. AYEGBUSI: Yeah. Sorry for the 20 misspeak. I probably shouldn't have used -- put the 21 acronym "large break LOCA" here.
22 All right. So lastly on this slide, so 23 they developed a CDF with debris, and then they have 24 the -- the base dry model with the CDF there, and 25 determined the delta, and then compared the results of NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
139 1 that to the acceptance guidelines in Reg Guide 1.174, 2 specifically Region III of that -- of the schematic in 3 the reg guide.
4 The next slide, please.
5 Again, going on with a summary of the 6 systematic risk assessment, they perform sensitivity 7 uncertainty analyses, you know, trying to be 8 consistent with NUREG-1855 and considering parametric 9 model and completeness analysis.
10 All of the results fell in Region III of 11 Reg Guide 1.174, acceptance guidelines meaning it's a 12 very small risk increase. The acceptance guidelines 13 there is just like, you know, a delta CDF of, you 14 know, less than 10-6 and LERF 10-7, and obviously 15 that's based on what the current plant's CDF and LERF 16 are, the base CDF and base LERF.
17 The last bullet here speaks to the 18 subsequent LAR that was submitted that is currently 19 under review after we had reviewed the technical 20 evaluation report. And basically one of the things 21 they had to take care of was correct debris loading.
22 23 There are some numbers that were not 24 correct in the initial submittal, and so in the LAR 25 they corrected it, and vis--vis Vogtle was saying in NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
140 1 the LAR that it did not change their conclusions as 2 far as meeting the requirement for a very small risk 3 increase.
4 All right. Next slide, please.
5 Okay. So this -- sorry, I just had an 6 error code.
7 All right. This slide and the slides 8 after this talk about the staff's review. So the 9 staff sought to determine the acceptability of 10 Vogtle's base PRA model by reviewing scope, level of 11 detail, and technical adequacy of that model.
12 So, in essence, we are here to see if they 13 considered the different hazards such as internal 14 events, seismic activity, and things of that nature, 15 right, that are pertinent to this risk-informed 16 approach. And then we needed to see if the model had 17 to -- had a PRA review, if there were any findings, 18 and how those findings were resolved.
19 And then, lastly, how any key assumptions 20 were addressed in the PRA.
21 The next bullet, the staff sought to 22 determine the acceptability of an approach to focus on 23 -- of Vogtle's approach to focus on ISI welds in 24 unaccessible portions of Class 1 pressure boundary 25 piping; and, second, identify breaks in containment.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
141 1 In this case, we did a review to see if 2 there were piping with welds that were outside of the 3 unaccessible portion that could also lead to -- that 4 could also lead to debris appearing in containment and 5 back in the core.
6 In that case, what we found was most of 7 the piping were of small diameter and wouldn't be --
8 wouldn't be a significant issue.
9 On the third bullet, the staff sought to 10 determine the acceptability of Vogtle's focus on high 11 likelihood configurations of ECCS pumps. When I say 12 "focus" there, really it's, you know, they focused on 13 I think several configurations that -- that they put 14 through the analysis, right?
15 So they focused on accidents that need 16 recirc through the ECCS strainers. Then they had a 17 certain number of high likelihood equipment 18 configurations. But for likely -- low likelihood 19 configurations, they assumed that they can be bounded 20 by the high likelihood configurations.
21 For this -- in this case, what we're 22 looking at was the reasonability and acceptability of 23 these high likelihood configurations and how to 24 determine the, you know, functional strainer 25 probabilities of each configuration.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
142 1 And then, if it looks reasonable for --
2 what will they use these high likelihood 3 configurations in their assessments.
4 And on the last bullet, the staff sought 5 to determine the acceptability of using the 6 conditional failure probability, which they -- which 7 the Vogtle individual spoke about earlier.
8 So using a CFP for breaks that impact --
9 that impact the strainers, there are certain, you know 10 -- for breaks that didn't fill the strainer. In this 11 case, what we're looking for was -- I think Steve 12 earlier went into a lot of detail about, you know, 13 STP's process in resolving this GSI-191 issue, and 14 then how Vogtle is trying to address this issue.
15 But, you know, what he found was that this 16 -- using the CFP approach is -- was developed in a 17 systematic and acceptable manner, and it's pretty 18 similar to typical risk-informed approaches that we 19 see when using Reg Guide 1.174.
20 Next slide, please.
21 Okay. So, again, the staff sought to 22 determine Vogtle's disposition of key assumptions in 23 the PRA and sources of uncertainty. The staff -- the 24 staff focused on the four bullets here, looking at the 25 geometric arithmetic mean, aggregation schemes, the NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
143 1 LOCA frequency, how the LOCA frequency were allocated 2 to break sizes, partial and, you know, double-ended 3 breaks, and then the discretization of the large break 4 LOCA frequency.
5 So here the staff was looking to see if 6 Vogtle had not just used one -- not just used one 7 methodology in their analysis, but -- in the 8 uncertainty and sensitivity analysis, that they 9 considered other approaches that may be more 10 conservative than the approach they used to present 11 the results.
12 And the last bullet, which is where there 13 was a lot of work done between the staff and the staff 14 consultants, was confirmatory calculations. So for 15 that calculation, the calculation used the minimum 16 transition break size, which seemed to be 12 inches.
17 And as we talked a few minutes earlier, 18 the 12 inches came from pretty much the smallest break 19 -- the minimum break size that Vogtle determined for 20 one of the -- for one of the -- I believe it's seven 21 high likelihood configurations, right?
22 And so staff took the 12 inches and 23 assumed that is -- it's in our calculation as our 24 minimum break size, and then -- and then that was 25 conservative. We assume that any breaks larger than NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
144 1 12 inches would lead to failure of the strainers and 2 core damage.
3 CO-CHAIR DIMITRIJEVIC: So this is where 4 -- I'm sorry to interrupt you, but this is where I saw 5 the CFP, which I asked when we are talking with 6 Southern Nuclear, and it equals to one.
7 So what you want to say, you assume the 8 CFP is equal to one just for the specific 9 configuration but not for high likely configuration.
10 Is that what you are saying?
11 MR. AYEGBUSI: I don't think I understand.
12 You said for a specific configuration. So here what 13 we're saying is for the staff -- so the staff's 14 confirmatory calc was way more conservative than what 15 Vogtle did, right? We only focused on high likelihood 16 configurations, and for all -- for every 17 configuration, whenever there was a break size, 18 whenever there was a break size of 12 inches or more, 19 we just assumed that there was failure.
20 CO-CHAIR DIMITRIJEVIC: Okay. So let me 21 ask you this, because that's something which I have a 22 tough time with. Basically, if you get LOCA which is 23 higher than 12 inches, you say the same probability of 24 the plant response is equal to one, because the 25 circulation will fail guaranteed, right?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
145 1 So every LOCA higher than 12 inch leads to 2 CDF, right? So if you are showing -- so basically if 3 you are still satisfying Reg Guide 1.174, that means 4 that your frequency of LOCA above 12 inches is smaller 5 than 1E-6.
6 MR. AYEGBUSI: I don't think I would say 7 that because you have big -- a big break that releases 8 CDF. Because, again, the CFP is a -- the CFP is input 9 into the PRA model where you put it -- where you 10 insert whatever -- well, first of all, the CFP is a 11 ratio of failure you see when you -- of failures you 12 see when you -- within a certain break size range, 13 right, over all of the potential failures you have, 14 right? Oh, sorry, over all of the potential break 15 sizes that you have within that range, right? So 16 that's one, too. That's input into the PRA model, 17 right, as a probability, right?
18 So I guess -- I guess what story you'll 19 hear is that the CFP equals one, right? We probably 20 shouldn't have put that there. What we're trying to 21 communicate is that the -- that we didn't -- that the 22 staff confirmatory calculation didn't go -- go in and 23 look for -- what's the word now?
24 The staff calculation didn't just for each 25 likelihood -- let me put it this way, for each NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
146 1 likelihood, the staff calculation didn't look for 2 every failure. Once you have a failure above 12 3 inches, they're all assumed to be a failure for the --
4 for the scenario or the configuration.
5 MR. VASAVADA: This is Shilp Vasavada from 6 the NRC staff. If I may just add something maybe that 7 can help Dr. Dimitrijevic understand better, because 8 her question is, if you have a CFP of one, which we 9 did, then basically the initiating event frequency has 10 to be less than 1E-6.
11 In addition to that is that there is a --
12 there is a probability for that particular 13 configuration to occur. So it is essentially the 14 initiating frequency times the probability that a 15 particular configuration is in play, and CF -- sorry, 16 conditional failure probability of one, that together 17 gives the risk for a particular configuration, and 18 then you have the different configurations that Ayo 19 talked about.
20 CO-CHAIR DIMITRIJEVIC: Well, I asked 21 about configuration, but they said that that was most 22 likely configuration, so -- which is close to one. I 23 mean, so I -- the thing is, it could be true that this 24 initiating event -- you know, if you are analyzing all 25 large LOCAs above six inches, and then you are NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
147 1 averaging, you may not see this.
2 But if you split in this case, when you 3 are running sensitivity case, large LOCA between six 4 to 12 and then 12 to whatever, 44 inches, then -- then 5 you can see that this frequency above 12 inches is the 6 one that will impact results the most. And this still 7 may be less than 1E-6. I don't know because you have 8 these dramatic means. I have to go back to that 9 curve.
10 So I just was pointing out that this is 11 where I got the CFP, the one, and I don't really need 12 any more explanation. I will just look at this 13 myself. Thanks.
14 MEMBER KIRCHNER: Vesna, this is Walt.
15 Just to confirm, if it's a break larger than 12, 16 you're assuming failure of the strainer to do its 17 function because of debris loading. The 18 configuration, as you say, is effectively some number 19 between zero and one. Let's assume it's one.
20 Then this all boils down to just the 21 probability of a large break LOCA greater than 12 22 inches and integrating over that range of 23 probabilities, right? And that --
24 CO-CHAIR DIMITRIJEVIC: That's right. But 25 that comes down to --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
148 1 MEMBER KIRCHNER: That becomes the delta.
2 CO-CHAIR DIMITRIJEVIC: -- frequency of 3 LOCA larger than 12 inches, which is --
4 MEMBER KIRCHNER: Yeah. Exactly. It just 5 boils down to that.
6 CO-CHAIR DIMITRIJEVIC: But this has been 7 discussed on calculation. So they took these 8 conservative assumptions, because as we saw in this 9 figure presented by Southern, and we saw it here, 10 actually they never assumed the CFP is equal to one.
11 So I'm just sort of curious on the -- on 12 this -- you know, because it is really one of the 13 major things which I am here both fascinated and 14 questioning is that this -- these calculations about 15 the risk basically concludes the risk associated in 16 GSI-191 is much smaller than originally thought.
17 So basically this delta, this calculation 18 shows the risk associated with GSI-191 is very small.
19 And if it's really so small, then we have to 20 concentrate on those sensitivities and uncertainties 21 just to understand that this can be conclude with the 22 acceptable level of confidence.
23 So that's my answer of like trying to --
24 to fashion this. All right?
25 MEMBER KIRCHNER: Well, it seems to me --
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
149 1 I'm being very simplistic about this -- it seems to me 2 that if through whatever deterministic means you 3 determine that a break greater than 12 inches would 4 load so much debris on the -- on the filters as to 5 render them inoperable, so that you get to a failure 6 of one, then it all hangs just on how much confidence 7 you have in the probabilistic estimates of the 8 spectrum of large break LOCAs for a specific plant 9 with a specific amount of potential debris sources.
10 I'm saying -- and it's a different way of 11 saying what you just said.
12 CO-CHAIR DIMITRIJEVIC: Yeah, yeah.
13 Walter, I think you are right saying this. It's only 14 if you remember, this additional slide we saw in 15 previous presentation, the additional slide shows the 16 only -- but I don't know what configuration that is.
17 I assume that that's also high likely. So the only 18 LOCAs higher than 20 inches without the environment, 19 it shows that even for the largest LOCA size it's 20 still only .6 CFP.
21 So this is -- this is some confirmatory 22 calculations we assume was in the opinion of 23 conservative. So, you know, that's only -- but you 24 are completely right. What is written here is true.
25 It just boils down now to what is the LOCA frequency NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
150 1 or probability of the LOCA being larger than 12 2 inches.
3 MR. AYEGBUSI: So can I provide just one 4 more thing on this? It just a -- or it's to add on to 5 what Shilp said, right? for the confirmatory 6 calculation. There were five equipment configurations 7 that were evaluated, right? One of which was no 8 equipment failure, right? Another was one train, 9 which one -- one RHR and one CS pump failed, right?
10 For the no equipment failure, as Shilp 11 said, you know, what you take is your -- whatever 12 breaks you have bigger than 12 inches, you take that 13 and multiply that with the functional failure 14 probability for each equipment configuration, right?
15 So for the no equipment failure, right, 16 the functional probability -- failure probability was 17 I think .1 -- 0.915, right? But for like -- for a one 18 train failure, the functional failure probability was 19 3.9E-3.
20 So I think what Shilp was saying, and what 21 I'm trying to reiterate, is that it's not just once 22 you have failure you -- you just -- you end up with 23 just a LOCA frequency as -- as your CDF, right? There 24 is a functional failure probability factor there.
25 MEMBER KIRCHNER: Yes, agreed. I was just NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
151 1 simplifying and just taking what I would call a 2 deterministic mixed with a probability of the large 3 break kind of simple bounding estimate. You're right.
4 If you look at equipment failures, then the 5 probability goes further down.
6 MR. AYEGBUSI: Okay. Any other questions?
7 All right. So I'll just continue on the 8 -- all right. So I'll continue with the last bullet, 9 right? So the results of the confirmatory 10 calculations overall did not challenge Region III of 11 Reg Guide 1.174. And there are some specific details 12 to discuss there that we will talk about I believe on 13 the next slide.
14 So next slide, please.
15 Okay. So as I mentioned already, Vogtle's 16 results, and even the NRC's confirmatory calculations, 17 didn't challenge Region III, but you might look at 18 this graph and say, well, the maximum -- the staff's 19 confirmatory calculation is above Region III.
20 There are a couple of things there. One 21 is the -- as we discussed, you know, the staff's 22 calculation was way more conservative than what we 23 did. And then this -- this blue star is -- really 24 indicates when the staff used the arithmetic mean to 25 aggregate the LOCA frequency.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
152 1 That's when we get to I think a delta CDF 2 of about 2E-6 -- oh, sorry, 1E-6 to 2E-6, which if 3 looking at Reg Guide 1.174, when it's really close to 4 the denotation between Region III and Region II, you 5 know, we don't treat that as -- we treat that as -- we 6 don't treat -- we typically don't treat that as a 7 significant place in Region II, right?
8 And so this was -- but the arithmetic mean 9 evaluation was done as a sensitivity study on the 10 confirmatory calculation, right? So that's why we say 11 the results fall within Region III with no concerns.
12 Next slide, please.
13 So on the -- specifically on the 14 systematic risk assessment, the staff's conclusion is, 15 you know, the licensee used the PRA of appropriate 16 scope, level of detail, and technical adequacy. The 17 approach addressed the effects of debris on long-term 18 core cooling was acceptable, you know, all of the 19 different alternative assumptions that we have kind of 20 discussed about the risk assumptions.
21 There were considered sensitivities for 22 each assumption, and these are for the ones where they 23 employed non-consensus approaches.
24 And the last two we already discussed 25 earlier.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
153 1 That's all I have. I will turn it back to 2 Steve.
3 MR. SMITH: Okay. Yeah. This is Steve 4 Smith again. I'll talk about Principle 5 a little 5 bit. The performance monitoring was addressed in the 6 lower review, and basically Vogtle identified programs 7 and processes that they had already existing in the 8 plant that they would use to make sure that the key 9 inputs and assumptions remain valid, and that they 10 would take corrective actions if something got outside 11 of the bounds of -- of their analysis.
12 So our review of Principle 5 determined 13 that they adequately addressed using -- the use of 14 performance measurement strategies.
15 Next slide?
16 Okay. This talks about the review of the 17 LAR a little bit. This is less -- you know, there is 18 a lot less to the LAR that's technical. It's mostly 19 regulatory.
20 We evaluated the tech spec changes. Those 21 were made per TSTF-567, which is a new sump tech spec 22 that all -- all PWRs can adopt, whether they use a 23 risk-informed evaluation for Generic Letter 0402 or 24 not.
25 We reviewed the licensee's FSAR changes.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
154 1 We reviewed their exemption request, and we reviewed 2 the licensee's responses to the limitations and 3 conditions that we had established when we did our 4 safety evaluation report on the technical report.
5 Next page?
6 Overall summary of the issue is that we 7 concluded that Vogtle's evaluation is acceptable.
8 Most scenarios are known to be mitigated using 9 conservative deterministic methods and changing risk 10 due to any scenarios that did not meet the 11 deterministic methods. It is very small.
12 We are -- the staff is interested in 13 receiving feedback from the ACRS. But as -- as 14 Caroline said, we don't need a letter from the ACRS to 15 proceed in issuing this amendment.
16 And before we move on to questions -- and 17 I don't want to incite any really difficult ones -- I 18 know that some of the ACRS members have ties to the 19 Boston area, and I just have one thing to say: Go, 20 Caps.
21 (Laughter.)
22 CO-CHAIR REMPE: I don't know if you're 23 allowed to give those type of opinions. But do any of 24 the members have any additional questions for the 25 staff?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
155 1 MEMBER HALNON: This is Greg. I've got 2 just a couple. The variations from the TSTF that were 3 used, did you guys look through those to make sure 4 they were just minor and not fringing on causing a 5 problem with the TSTF SER?
6 MR. SMITH: Yeah. We always evaluate any 7 variations, I mean, even including just numbering 8 changes. And we address each one of those when we do 9 our safety evaluation.
10 So, yeah, and I think Vogtle -- they did 11 some -- they did simplify things a little bit. For 12 one thing, the notes for the required action, I think 13 it's B1. I don't remember which required action it 14 is. Anyway, the required action, if you do have a 15 strainer that -- or a strainer that is out of service, 16 they simplify it and they remove the notes that tell 17 you to go to a -- go to either the ECCS or the 18 containment spray tech spec. And they just have a 19 direct -- you know, go straight to the tech spec.
20 That's the required action.
21 So they were able to do that because they 22 have a separate strainer for each pump. So actually 23 their implementation and the variations kind of 24 simplify things from what they were, and it's just as 25 effective.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
156 1 MEMBER HALNON: Okay. That's kind of the 2 way I read it was that give it back to the statistics 3 group; maybe they can improve that statistically.
4 Maybe no one else will use it, I don't know, but --
5 MR. SMITH: No. They will use it, and I 6 guess, you know, I'm -- I'm in the Tech Spec Branch 7 now. I was -- we had a separate branch before for 8 this, but the workload went down, so the Tech Spec 9 Branch absorbed us -- absorbed the GSI-191 Branch.
10 And so I'm sure Vic is the one that called 11 two of my branch chiefs, so, I mean, we can -- we can 12 think about this. And maybe for plants that have a 13 similar plant configuration we could talk to the TSTF, 14 you know, and make a recommendation that they use this 15 kind of an approach.
16 MEMBER HALNON: Okay. On the L&C number 17 7, it said that -- or something in the SCRO said that 18 there was a set of inputs and attributes that if 19 altered would impact the risk-informed approach.
20 Is there any -- did you look into how the 21 licensee is going to be determining whether or not any 22 of those sets of inputs or attributes are changed? I 23 mean, is this going to be -- have they incorporated 24 into their 50.59 program or some other program that 25 tracks those things to ensure that they're not NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
157 1 changing something that could alter the final 2 conclusion?
3 MR. SMITH: Okay. I think what we're 4 talking about here is the periodic update of the -- of 5 the model or periodic review. I'm going to ask if Ayo 6 or Shilp could address this one. I think that this 7 one was done by -- you know, looked at by DRA when 8 they -- when they owned -- when did the evaluation of 9 the L&Cs. If not, I can -- I have our SE ,and I could 10 see --
11 MEMBER HALNON: Yeah. I didn't see -- I 12 didn't see anything in this either beyond what I just 13 quoted. And I know that the updates are like four 14 years apart.
15 MR. SMITH: That's right.
16 MEMBER HALNON: It could be up to three 17 outages if you place them right on top of each other.
18 So I was just curious if there was some program onsite 19 or if there is any monitoring of those things.
20 What I didn't want to do is come four 21 years later find out that, oops, we altered something 22 three years ago and didn't track it.
23 I guess -- go ahead.
24 MR. AYEGBUSI: This is Ayo. Yeah, I think 25 I'm with Steve. I think I will have to go -- I will NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
158 1 have to go back and look at the SE specifically. I 2 can't remember off the top of my head --
3 MEMBER HALNON: Okay.
4 MR. AYEGBUSI: -- what -- how often the --
5 they would report the information to the NRC. Oh, 6 actually, you know what? Sorry about that.
7 I think the -- this -- I know this L&C may 8 be the one where -- this is -- this is only if or when 9 they determine that there needs to be some change to 10 the evaluation, right? I believe this was -- this was 11 it.
12 And I don't want to misspeak, so I think 13 I'll --
14 MEMBER HALNON: That's fine. I read 15 essentially as good a quote as I could read out of the 16 SE, which is set of inputs and alternative -- or 17 attributes that if altered would impact the risk-18 informed approach.
19 And I guess my question is is how do we 20 know whether or not these inputs and attributes are 21 changed. And I guess the question is, you guys wrote 22 that in the SE. I assume that there is some 23 verification and validation somewhere that either 24 we're so far away from those inputs and attributes we 25 don't have to worry about them, or there is some NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
159 1 program outside that checks it.
2 And it may be the 50.59 program. It may 3 be something -- something else, maybe a corrective 4 action program. I'm not sure. But it seems like 5 there should be something tagged to it.
6 MR. VASAVADA: This is Shilp Vasavada from 7 the NRC staff, if I might just add. So I think that 8 there was a list of the inputs that can impact the 9 evaluation.
10 And there were -- I expected the 11 performance monitoring piece of it, which provided the 12 different existing procedures, would kind of like 13 track that. As you said, maybe a corrective action, 14 there's also a design change process. Every site has 15 its own design change control process to make sure 16 that any design change would -- is fed through 17 different organizations to make sure that the impact 18 is -- is addressed or captured.
19 And then the way the LAR is, I think that 20 if the -- if the change resulting in using this risk-21 informed approach, if the change in the import is in 22 exceedance of Region III of 1.174, the licensee has to 23 take action, and that can take different -- on 24 different parts, but they do have to do something to 25 maintain the risk within Region III.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
160 1 MEMBER HALNON: Okay. One way you can 2 tell -- make sure is just to ensure that that level of 3 detail is in the FSAR updates that go in. And that 4 will tie it into the 50.59 program as well.
5 So, you know, that was a relatively I 6 guess new old plan. The FSAR has probably got a lot 7 of detail in it, so it's probably okay, but there was 8 also some very old plants that that level of detail 9 may not be in the FSAR. So --
10 MR. VASAVADA: Again, this is Shilp. I 11 believe that -- and Steve and Ayo can correct me if 12 I'm wrong, but I believe that the -- the license 13 amendment request that was presented included markups 14 to the new FSAR to indicate essentially what I 15 mentioned but in a summary manner that it -- 1.174 16 Region III criteria acceptance crackdowns are 17 exceeded, then the license has to take certain 18 actions.
19 MEMBER HALNON: Okay.
20 MR. SMITH: This is Steve Smith. On that, 21 and some of the limitations and conditions address, 22 you know, what was in the FSAR. We had some comments 23 on that, and they actually made some changes to the 24 FSAR based on, you know, comments that we had. So it 25 had a more detailed description of what the key NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
161 1 elements of this evaluation are.
2 MEMBER HALNON: Okay. Yeah. If it's 3 captured in the FSAR, then like I said, it will -- it 4 will be part of a program. Yeah.
5 MR. SMITH: Just limitation and condition 6 5 on -- as we addressed it in the SE, it tells, you 7 know, what they consider to be the key elements, and 8 those are the ones that we felt were the ones that 9 should be --
10 MEMBER HALNON: Okay. Yeah, that's good.
11 If I had seen that, I may not have asked the question.
12 I'll give it back to you, Joy.
13 CO-CHAIR REMPE: Okay. Thank you. While 14 I will see if other members have any final questions 15 to the staff or the licensee, I'd like to ask Thomas 16 to open up the public line because we will call for 17 comments from the public at this time.
18 So do any other members have any questions 19 or comments?
20 Okay. Thomas, let me know when the public 21 line is open, so we can ask members of the public if 22 they would like to make any comments.
23 MR. DASHIELL: The public line is open to 24 the public.
25 CO-CHAIR REMPE: Great. So at this time, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
162 1 do any members of the public have any questions or 2 comments? And, if so, please state your name first 3 before you provide that comment.
4 So I heard a noise, but I didn't hear any 5 -- so maybe someone is trying to say something, but 6 you need to really speak up. I'm not getting 7 anything, so I guess we will close the public line 8 again.
9 Yes, Thomas, just go ahead and close the 10 public line, then, I think at this time.
11 So, members, at this time I'd like to ask 12 you to tell me your thoughts about whether we should 13 have a subsequent full committee meeting and issue a 14 letter on this. I'll let you know that we actually 15 have been working on a draft letter, and by "we" I 16 mean my co-chair, Vesna, as well as the co-chair of 17 the Accident Analysis and Thermal-Hydraulic 18 Subcommittee. Jose, I have shared a draft with him, 19 and he has provided comments back. But what we heard 20 today from SNC, as well as the staff, will make me at 21 least want to update that draft.
22 But what are your thoughts about whether 23 you'd like to have a letter? And do you have any 24 final comments you'd like to give?
25 And I guess I'll start with Charlie. I NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
163 1 know Charlie said he had to be away, but if you're 2 still there, Charlie, do you have any --
3 MEMBER BROWN: No, I don't have any. I've 4 got to leave that judgment up to you all who are more 5 talented on this overall subject of the GSI-191 than 6 I am. It sounded to me, based on listening, that the 7 staff had done a thorough job of reviewing it, and 8 that it appeared to be satisfactory to me. But some 9 of the issues I can't -- I can't comment on responses 10 to a lot of Vesna's questions. So I'll leave it up to 11 some other more competent individuals.
12 CO-CHAIR REMPE: Okay. Ron Ballinger, do 13 you have any thoughts on this?
14 MEMBER BALLINGER: Yeah. I'm a little bit 15 conflicted. On the one hand, I think I agree with 16 Charlie that the staff did a great job of doing the 17 review. And if they think it's okay, then I think 18 it's good. On the other hand, this is precedential, 19 not presidential. It's a precedent.
20 And the applicant has done a lot of work 21 to finish off, if you look -- if you want to use that 22 word -- using the risk-informed approach for GSI-191.
23 I'm not sure that anybody else is going to do it, but 24 they have done a job and they've had to wade through 25 a lot of technical issues which to me seem to think NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
164 1 that we ought to provide some feedback related to 2 that.
3 But I guess I'm a little -- you can see my 4 ambivalence, but I guess I would weigh 51 percent to 5 have a letter.
6 CO-CHAIR REMPE: Okay. Vicki Bier, what 7 are your thoughts on this issue?
8 MEMBER BIER: Hi. I feel like I am kind 9 of still coming up to speed, not having tracked the 10 whole development of the issue historically. I kind 11 of agree that overall it looks like staff has done a 12 pretty good job.
13 But listening to some of Vesna's detailed 14 questions I also feel like there is areas maybe that 15 I need to dig into to understand better before I have 16 a really reliable opinion on that. So I don't have 17 strong feelings right now.
18 CO-CHAIR REMPE: Okay. Dennis, what are 19 your thoughts?
20 MEMBER BLEY: Given the committee's role 21 in getting all of this underway and looking over it 22 for years, I think we ought to do a letter. And I --
23 I appreciated the presentations and the responses 24 given today.
25 CO-CHAIR REMPE: Thank you.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
165 1 Greg Halnon, what are your thoughts on 2 this?
3 MEMBER HALNON: Yeah. I kind of -- I 4 agree with Dennis. I, too, think that, you know, both 5 the staff and the licensee did thorough jobs, and I 6 think that there are some good lessons learned in this 7 project to pull forward to other risk-informed 8 applications.
9 And, you know, we are really going to be 10 doing that more and more in the future.
11 Notwithstanding Vesna's questions, which I don't 12 profess to understand totally the initials and 13 everything else, so I -- I don't know if I can comment 14 on that.
15 But I think that given the length of time 16 that the issue, the GSI, has been out there, the --
17 anything we can do to light a fire under getting this 18 thing closed and getting additional attention in the 19 industry and the public that it is getting closed I 20 think is a good thing.
21 So I think I would agree with Dennis that 22 we should probably write a letter on this.
23 CO-CHAIR REMPE: Okay. Thank you.
24 Walt Kirchner, what are your thoughts?
25 MEMBER KIRCHNER: Sorry. Trouble finding NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
166 1 the unmute. I agree with Dennis and Greg. Given the 2 significance of this issue in the past, the amount of 3 effort that has gone into it by the industry and the 4 staff, and the improvements demonstrated here by the 5 applicant in using risk-informed approaches in a more 6 holistic manner, I weigh in on behalf of writing a 7 letter.
8 CO-CHAIR REMPE: Okay. Thank you.
9 Dave Petti, your thoughts?
10 MEMBER PETTI: I'm on the fence. You 11 know, the staff doesn't feel like they need one, so I 12 kind of agree with Ron. Initially, I -- I thought, 13 well, you know, given South Texas had done this 14 before, but then, you know, there were significant 15 differences in what they did, and that's really valid 16 in terms of, you know, the methodology that he used, 17 and the like.
18 So I'm, you know, kind of ambivalent at 19 this point.
20 CO-CHAIR REMPE: Okay. Matt Sunseri?
21 MEMBER SUNSERI: Well, I'm neutral as 22 well. I think, as I recall, we wrote a letter on the 23 South Texas handling of this.
24 CO-CHAIR REMPE: Yes, we did. Yeah, we 25 did, just to answer that question. Is that -- did you NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
167 1 want to take that into consideration in your comments?
2 MEMBER SUNSERI: Well, yeah. So I think, 3 you know, at least in my view, I think that letter 4 probably covers us. I didn't hear anything all that 5 substantially different, I don't think, as far as 6 application of risk-informed. They use different 7 tools and things of that nature, but so will other 8 people.
9 So, like I said, I'm neutral, but it --
10 and I'm neutral because I think the South Texas letter 11 probably covers this already. That's all I have.
12 CO-CHAIR REMPE: Okay. Jose, what are 13 your thoughts?
14 MEMBER MARCH-LEUBA: Oh. I thought you 15 had skipped me and forgotten me.
16 CO-CHAIR REMPE: No. I'm kind of leaving 17 you and then Vesna for the end, along with me, since 18 we've already seen the draft letter. I didn't -- I 19 wasn't going in alphabetical order this time.
20 MEMBER MARCH-LEUBA: Oh, thank you. You 21 think I'm special.
22 CO-CHAIR REMPE: I'd never forget you.
23 MEMBER MARCH-LEUBA: Well, I have to say 24 that I am humbled and awed by the amount of work the 25 licensee and the staff have performed to try to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
168 1 resolve this issue. This implies they have taken it 2 seriously, which is something -- well, I won't say 3 that, but they have taken seriously and we should 4 issue a letter, a positive letter, just to reinforce 5 the fact that they have done their homework.
6 So, yes, I -- I support having the full 7 committee -- I would prefer, if bringing the full 8 committee, we only have a presentation from the staff 9 with 10 slides, with a summary of -- of the main 10 topics and the conclusions, because the full committee 11 is for the benefit of the public. We have already 12 heard the information, so please don't repeat 13 everything again.
14 And then I think we should write a 15 positive letter.
16 CO-CHAIR REMPE: Okay.
17 MEMBER MARCH-LEUBA: Because they have 18 done a good job.
19 CO-CHAIR REMPE: Okay. Thank you.
20 Vesna, do you have any comments about the 21 need for a letter, as well as are there some of your 22 questions that you would like to have some additional 23 answers for before the next meeting? Because I think 24 that -- we have only heard people that are neutral or 25 people saying, yes, do a letter, so I think we're NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
169 1 going to have a subsequent meeting and have a letter.
2 But do you have any additional thoughts on 3 that topic?
4 CO-CHAIR DIMITRIJEVIC: I have gone and 5 checked these tables, which I have -- let me put it 6 this way. My question is on the uncertainty and since 7 -- how the uncertainty configuration was informed were 8 not really answered to the -- my satisfaction. But 9 that's no surprise. There are some many open 10 questions, how do you address uncertainty, especially 11 when you start talking about such small numbers as 12 10-8.
13 So basically, do I think it's important, 14 it would change your conclusion? The answer is no.
15 I mean, you know, they have to go four orders of 16 magnitude to get in the region, which is not 17 acceptable based on the Reg Guide 1.174.
18 So I think this is likely to -- to satisfy 19 the risk-informed application requirement. So from 20 that perspective, we can write a positive letter, if 21 we want to. I mean, I don't really -- my main thing 22 is that I want to make sure that they -- all of the 23 members realize that this 2E-8 delta CDF is not some 24 delta CDF -- some delta. It's basically CDF 25 associated with bad risk creation.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
170 1 So this application, which is very true, 2 this is extremely complex thing which requires so much 3 work, exactly showing the risk associated with -- with 4 191 is really small. So the question is, of this work 5 which was invested in this to show manageable risk 6 associated with this, this is what somehow I have --
7 I have, you know, the -- you know, especially because 8 we're going to address the regulatory approach for 9 them.
10 We are going to, in the next full 11 committee, discuss regulatory approach for not safety 12 significant issue. This is -- this report proved that 13 this is not safety significant issue. That's -- and 14 in order for me to be 100 percent sure that that 15 proves that, I would like to see a little more -- you 16 know, more confirmatory uncertainty analysis set to 17 this -- but that's all.
18 I don't really have -- you know, Joy, you 19 invested much -- most effort to this -- and, Jose, you 20 -- and I was waiting because I was also 50/50 on the 21 letter. So that's how I still feel.
22 MEMBER BROWN: Joy?
23 CO-CHAIR REMPE: Yes, Charlie?
24 MEMBER BROWN: I just wanted to ask a 25 question relative to Vesna's comment. I guess I NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
171 1 didn't get this out of her comment. She said her 2 conclusion, even though some of her questions weren't 3 necessarily answered, you know, but this was not a 4 safety significant issue.
5 Does that mean GSI -- this is a neophyte 6 asking this question. Does that mean GSI-191 is an 7 insignificant issue?
8 CO-CHAIR DIMITRIJEVIC: Basically, what I 9 said, the risk associated with GSI-191, as this 10 application show, is 3E-8.
11 MEMBER BROWN: Yeah. That's what I --
12 CO-CHAIR DIMITRIJEVIC: So that's not 13 really significant. It's comparing to the regular 14 CDF, which is 5E-5. It's totally never mind, so --
15 well, that's what this posed.
16 MEMBER BROWN: Is that -- the other plants 17 that have used -- based on the slide that the staff 18 provided, they talk about most -- most people use this 19 option 2, which was the risk significant approach and 20 two breakdowns in terms of how they were done. Were 21 they all coming out with significance in this range of 22 10-8 or 10-7 or something like that? Nobody talked 23 about that relative to how -- how small the numbers 24 were for the other plants. They only addressed this 25 plant itself.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
172 1 CO-CHAIR REMPE: Well, Charlie, most of 2 the plants, except South Texas Project, use a 3 deterministic approach, right?
4 MEMBER BROWN: No. That's not what their 5 -- their staff said that they -- 19 units used 6 existing guidance. Option 3 was deterministically 7 evaluating it. They didn't say how many used that, 8 but all of the -- or maybe it was none. And then all 9 the remainder of the plants used option 2, which it 10 sounds like that's what -- that's what Vogtle did. So 11 --
12 MEMBER BIER: Charlie, this is Vicki.
13 There was option 2A and 2B and --
14 MEMBER BROWN: Yeah.
15 MEMBER BIER: -- 2A, I'm not sure I 16 understood it properly, but it's not a full PRA 17 quantification. But in any case, I don't think we can 18 really comment from what we saw today on what number 19 another plant would get because they all may have 20 different design considerations, and sometimes some 21 pretty small physical differences can make a big 22 difference in the risk number. So --
23 CO-CHAIR REMPE: And I guess I misspoke, 24 but they used option 1, and it's my understanding that 25 with option 1 they do not have to give a risk number, NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
173 1 right? And the staff -- Steve, you can correct me, 2 but that is my understanding. They did not have to 3 present a number.
4 So we only have another number from South 5 Texas Project at this time. Right, Steve?
6 MEMBER BALLINGER: Yeah. And South Texas 7 used the R over D approach, which is different.
8 CO-CHAIR REMPE: Right. A number --
9 MEMBER KIRCHNER: Joy, this is Walt.
10 CO-CHAIR REMPE: Yeah.
11 MEMBER KIRCHNER: Let's just ask the staff 12 to address this. My notes would say that option 2A, 13 which the staff labels, quote/unquote, 14 "Deterministic," was used by 29 units. And 2B, risk-15 informed approach, was used by South Texas Project as 16 we heard and wrote. And now Vogtle, and three more 17 are in the pipeline to use 2B.
18 But perhaps for the record and your 19 letter, it would be useful to have the staff just 20 confirm.
21 CO-CHAIR REMPE: So, Victor, you have your 22 hand up?
23 MEMBER KIRCHNER: And the reason why.
24 There is an underlying reason, which the staff I think 25 could explain.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
174 1 CO-CHAIR REMPE: Right. So, Victor, you 2 have your hand up. Did you want to respond?
3 MR. CUSUMANO: Sure. This is Vic 4 Cusumano. I'm the Branch Chief who used to have the 5 GSI-191 Resolution Branch, and now I have the branch 6 that it's rolled into.
7 So, yeah, option 1 was deterministic.
8 Option 2 alpha is deterministic. We don't use PRA to 9 reach our conclusions there. The only one where we're 10 relying on probabilistic risk assessment is option 11 2 bravo. And as we discussed, the only one to have 12 completely gone through that so far was South Texas, 13 and Vogtle is next.
14 MEMBER BALLINGER: Yeah. This is Ron.
15 That's why I suggested that this -- this analysis was 16 a precedent, the first one, really.
17 MR. CUSUMANO: Well, this would be the 18 second after South Texas, just to be --
19 MEMBER BALLINGER: But South Texas kind of 20 defaulted back to a different method, right?
21 MR. CUSUMANO: No, they did -- they did a 22 risk assessment, a risk-informed analysis just like 23 Vogtle, but slightly different with some different 24 software and different configuration.
25 MEMBER BALLINGER: But I thought there NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
175 1 were some significant simplifications.
2 MR. CUSUMANO: What they both did was 3 cover deterministic, which the way I always thought 4 about that was because risk analyses are not quick, 5 simple, or inexpensive, they tried to minimize the use 6 of them, so they did deterministic where they could, 7 and they did a more robust risk analysis --
8 quantitative -- where they had to.
9 MEMBER BALLINGER: Thank you.
10 MEMBER KIRCHNER: Victor, could you 11 elaborate -- this is Walt Kirchner again. I think 12 there is an underlying reason why 29 plants were able 13 to use the deterministic approach. But I don't want 14 to speculate, but I would imagine it's the loading of 15 -- insulation loading, the debris source, that 16 probably allowed those others to use a deterministic 17 approach and demonstrate to you, the staff, that they 18 would not have failure of their ECCS and containment 19 spray systems.
20 MR. CUSUMANO: Yeah. And I'm not sitting 21 next to Steve, so he can't physically kick me under 22 the table, but, yeah, essentially that's correct. The 23 first 19 had such low amounts of LOCA-generated debris 24 that they met the WCAP-16793 limit, which was very, 25 very conservative. And the option 2A plants that are NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
176 1 deterministic were still able to do a deterministic 2 analysis because they had less debris.
3 South Texas, for example, had a lot more 4 things in containment insulated. And that's a little 5 bit of an oversimplification, because the type of 6 insulation matters as well, but you're absolutely 7 right, yeah. The plants with the most insulation are 8 the ones who have to deal with it.
9 MEMBER KIRCHNER: So, Joy, for the layer, 10 an important conclusion for me is that this risk-11 informed approach allowed applicants to avoid 12 unnecessary exposure and expense to address a problem 13 that had low safety significance. And somehow I think 14 the letter should address that.
15 CO-CHAIR REMPE: Okay. So --
16 MEMBER BROWN: Joy? That's an interesting 17 point. If the -- because I'm not -- based on what was 18 said during the meeting by the staff, those taking the 19 2A, you know, mostly the deterministic approach, is 20 that deemed to be more intrusive, more difficult, and 21 more time-consuming and expensive? And this risk 22 approach that was used by Vogtle demonstrates that the 23 benefits of the risk approach -- informed approach to 24 doing these?
25 MEMBER MARCH-LEUBA: No, no. Absolutely NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
177 1 not. They both have to spend a lot more time and 2 effort to --
3 MEMBER BROWN: That's what it seemed like 4 -- that's what it seemed like to me.
5 MEMBER MARCH-LEUBA: Yeah. But --
6 MEMBER BROWN: Looking at all of the 7 documentation.
8 MEMBER MARCH-LEUBA: -- going back, 9 Charlie, to your original question, was GSI-191 an 10 important -- did we waste our time for 20 years 11 following 191? And the answer is absolutely not. It 12 was a serious problem. All of those plants are able 13 to survive now this transient because they changed the 14 strainer.
15 They understood they had a problem, and 16 they fixed it. They put bigger strainers, and now 17 everybody can survive the problem. With the old 18 strainers, it was very hard to justify that you would 19 survive it.
20 So the issue is, there was a problem, we 21 identify it, we fix it. Now, there are two plants 22 that --
23 MEMBER BROWN: That point was not made 24 clearly. That -- I mean, I've been listening to the 25 GSI-191 discussions now for 13 years, and that is NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
178 1 being --
2 MEMBER MARCH-LEUBA: Since 1996.
3 MEMBER BROWN: Well, I've been on the 4 committee since 2008, so I just --
5 MEMBER HALNON: Charlie, keep in mind 6 that, you know, it wasn't just -- the strainers are 7 the biggest part, but there is also a lot of 8 modifications done to remove fiber, other things that 9 --
10 CO-CHAIR REMPE: Right.
11 MEMBER HALNON: -- containment. And those 12 things take a couple of years to design, and then 13 you've got to put it into an outage. You know, it 14 could literally be six years before you could get an 15 idea and translate it into a physical modification.
16 So, yeah, 13 years is a long time, but to 17 get something as big as removing full insulation and 18 replacing it with mirror insulation or even putting in 19 a bigger sump where you have to -- in an office sump, 20 you have to clean it out, you have to do all of the 21 radiological issues, I mean, it can literally be six 22 years before you can get something done in --
23 MEMBER MARCH-LEUBA: Greg, I'm not 24 complaining about the length of time it took. I'm 25 telling the story that this is a success story.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
179 1 MEMBER HALNON: Absolutely, Jose.
2 MEMBER MARCH-LEUBA: If we find a problem, 3 we fix it.
4 MEMBER HALNON: I completely agree with 5 you. I just wanted to make -- put more context around 6 that.
7 MEMBER MARCH-LEUBA: I think we need to 8 emphasize that point. I mean, there was a problem.
9 Sure, there was a problem. We identified it; we fixed 10 it.
11 MEMBER HALNON: Right. And part of the 12 modification issues, like you said, is what caused the 13 risk to go way down?
14 CO-CHAIR REMPE: And in some cases, 15 without a risk-based approach, my impression is that 16 they would have difficulty justifying that they fixed 17 it. And so that's why you see some of these plants 18 going this option, right?
19 MEMBER HALNON: Yes. I think you're 20 right. And some of the fiber couldn't be removed, 21 some of the vessel insulation and other things that 22 you can't literally get to, but --
23 MEMBER BROWN: You're talking like --
24 plants like Wolf Creek and Point Beach, the ones 25 they've listed, and Callaway?
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
180 1 MEMBER HALNON: Yeah.
2 MEMBER BROWN: The ones they listed in 3 their slides?
4 CO-CHAIR REMPE: What was the beginning of 5 your question, Charlie? I got the names of the 6 plants, but what was --
7 MEMBER BROWN: I think that was Greg that 8 made the comment about some of them can't do it at 9 all, and that's why the full risk-informed is an 10 approach he has been taking. It sounds -- and they 11 listed Wolf Creek, Point Beach, and Callaway in their 12 -- in their slides, so --
13 MEMBER HALNON: Yeah. I don't know the 14 specific configurations, but obviously some cost-15 benefit has been done to show that this is a better 16 way to go.
17 MEMBER BROWN: Yeah. Okay.
18 CO-CHAIR REMPE: Well, I'm going to 19 interrupt here, and I know, Paul Klein, you've got 20 your hand up, but we've never given Stephen Schultz, 21 our consultant, an opportunity to make any comments.
22 Steve, do you have any comments? And, 23 Steve, I see you are active, but I don't -- ah, there 24 you go. Now you're unmuted.
25 DR. SCHULTZ: I'm here. Yes, I agree with NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
181 1 the comments of the committee and the most recent 2 evaluation of the issue, the GSI-191 issue. It 3 certainly was a significant issue, and clearly what 4 was done here was to demonstrate that the risk is low 5 and that gave the licensee the opportunity to save 6 dose to -- dose to the workers and substantial 7 expenditures from modifications to the plant. So it 8 is a real success story.
9 With regard to today's presentation, but 10 more importantly all the work that has gone into this 11 project with Southern Nuclear and with the staff, I 12 found it very impressive. It is in addition to what 13 South Texas has done, so I think it's worthy of 14 additional recognition.
15 I was very pleased in the staff's 16 evaluation, as well as the licensee's work, that this 17 comes to -- to us or to the end of the process with no 18 open items by the staff. And this represented a lot 19 of work over the last year or two to get to that 20 point.
21 Southern Nuclear has addressed a number of 22 very important comments and questions that the staff 23 had had, and they have been able to close that within 24 the last six months to come forward with a very clean 25 presentation for the staff to consider and finalize NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
182 1 and for us to consider today.
2 The presentations today were great. They 3 were very detailed and hit the key points of the 4 evaluation very nicely. But I think it's worthwhile 5 to follow up in detail by the committee because it is 6 an augmentation of the risk-informed approach that 7 South Texas presented to us.
8 CO-CHAIR REMPE: Okay. Dennis, I see your 9 hand is up, but Paul Klein I think had a comment that 10 he wanted to add earlier, and so I'm going to let him 11 have that opportunity. Paul?
12 MR. KLEIN: Yeah. I was just going to 13 provide a long-term perspective from the staff, since 14 I've been working on this issue since the 2004 15 timeframe. I would say it is low risk now, but back 16 in the early days of GSI-191 plants did have the much 17 smaller strainers.
18 And when we realized the amount of fiber 19 that could be created during a LOCA, and then the 20 chemical effects at that time were very -- was a very 21 new issue, and we knew that they could potentially 22 cause clogging of the sump or -- or the core inlet 23 perhaps.
24 The ACRS agreed with the staff at that 25 time that the logical step was to install large NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
183 1 strainers and then move forward with testing to better 2 characterize the issue. And so there were a few bumps 3 along the way, but over time the knowledge base became 4 much better, and so now I think the staff is 5 comfortable with the resolution, and the risk-informed 6 approach is just a logical final step for the plants 7 that have the most fiber because when the staff tried 8 to do an integrated review team and trade 9 conservatisms early on in the process, it was very 10 difficult to try to do that, to compare a chemical 11 effect, conservatism versus sump debris or debris 12 transport conservatism.
13 So I think it is a success story over the 14 long run, and we're in a much better place where it is 15 a low-risk issue now.
16 CO-CHAIR REMPE: Okay. Thank you.
17 Dennis?
18 MEMBER BLEY: Yeah. I just wanted to get 19 in on a little bit of this. There is a little 20 forgetting that has gone on about South Texas. They 21 were originally doing a very broad and thorough 22 probabilistic risk assessment approach, and they got 23 into discussions with the staff that they eventually 24 figured out it was easier to change their approach and 25 have this hybrid approach rather than sign to get NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
184 1 agreement with their work.
2 So they were really doing something a lot 3 larger originally, and then they changed toward the 4 end.
5 If you go back to VARSAVEC, we were many 6 years after VARSAVEC before we really did anything.
7 I knew a lot of plants with a lot of fiber, and that 8 has really changed. So, Charlie, I'm not sure where 9 you got the impression that all of this is for not 10 much, but it's -- that isn't right. It's for a lot.
11 That's all.
12 CO-CHAIR REMPE: Thank you. So I don't 13 see any more hands up. And I guess I'd like to talk 14 just briefly about what should be presented at the 15 upcoming full committee meetings. We will have 16 limited time, as Jose said.
17 All but one of our members are present 18 today, so it doesn't need to be an elaborate 19 presentation, but I really liked some of the 20 visualization that was provided by Southern on their 21 presentation. So I would not say that we don't need 22 necessarily a presentation from Southern or -- but 23 perhaps some of the images the staff could plagiarize 24 and include in theirs.
25 But I would leave it to you guys to NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
185 1 decide, but I wouldn't think you'd need a total 2 presentation more than a half hour. But, again, we 3 can discuss that, and we'll see how much time is 4 allocated in the July meeting. Okay?
5 Any more comments or questions from any 6 committee members?
7 MEMBER PETTI: Yeah. Joy, I just -- I 8 wonder if the staff ought to include a slide or 9 something related to -- to the bigger picture here, 10 because it's clear that our letter is going to talk 11 about things in a broader context. And so a little 12 bit of the history might be useful, so it's on the 13 record in an integrated, coherent picture for the 14 public that will be listening in, and sort of setting 15 the stage. And it makes our letter have more context 16 I think.
17 CO-CHAIR REMPE: So I guess, you know, 18 historically when we did things on other topics, we 19 don't usually go to the big picture of where all of 20 the other plants are. We're commenting on a safety 21 evaluation and work completed by the staff on this 22 application.
23 So I know -- I heard what Walt said, and 24 I guess I'm hearing that from you. But I'm not sure 25 that's a wise idea.
NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
186 1 MEMBER KIRCHNER: It's a rather limited 2 view of the world, Joy.
3 CO-CHAIR REMPE: Well, it's a committee 4 letter. It's up to everybody, and I guess I'd like to 5 hear from some other members. I hear where you're at, 6 Walt. Other members have thoughts on that?
7 MEMBER BIER: Yeah. This is Vicki. I 8 feel like without going into the detail of here's what 9 every plant has done, it might make sense to give kind 10 of a summary at the level of what Jose said of -- is 11 historically was a big issue, plants have made a lot 12 of changes, here's one example of how this was 13 successfully handled or something, and provide some 14 historical context for people reading the letter who 15 may not all be up to speed on the details of the whole 16 issue.
17 CO-CHAIR REMPE: Okay. Any other members?
18 I see Dennis' hand is up again, or is it from the 19 earlier time, Dennis? Dennis, you're on mute, I see.
20 Your hand is up.
21 MEMBER BLEY: It's hard to get off mute 22 sometimes, Joy.
23 CO-CHAIR REMPE: Okay.
24 MEMBER BLEY: Try to do a hand when you 25 raise it sometimes. But I kind of like what Vicki was NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
187 1 saying. I think something that pulled together a 2 quick bit of history and where we've come. Because 3 this has been going on so long, this would probably 4 put a nice cap on it to have that summary, you know.
5 CO-CHAIR REMPE: Okay. Any other members?
6 So, Walt, will that -- is that going to take care of 7 your -- will that scratch your itch, as we often ask 8 our members?
9 MEMBER KIRCHNER: Yeah. It's not an itch.
10 I think it's a -- it's a useful summary to go into the 11 background of the letter and acknowledge that there is 12 a fair amount of history, tremendous amount of 13 industry work on this, and staff work, and a 14 tremendous amount of progress. And it can be 15 concisely captured in a background section to the 16 letter.
17 CO-CHAIR REMPE: Okay. This all sounds 18 good. I'm glad we're having this discussion. It 19 helps us in drafting a letter. Anything else, 20 anybody? Or are we ready to close this meeting?
21 MEMBER BROWN: Well, let me -- this is 22 Charlie again. In terms of addressing it, a 23 combination of Jose's comments about we found a 24 problem and it got fixed, Greg had a comment, kind of 25 another comment relative to that. It was a little bit NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
188 1 different.
2 But this risk-informed approach really 3 allowed those who maybe would have had to do such 4 extensive plant modifications that it would have been 5 impractical to start with, but that allowed them to 6 analyze it and found it was not a big deal.
7 So somehow that context I think ought to 8 be part of the letter. That's my only thought. I 9 thought his point was very good, and coupled with 10 Jose's observation that it's a big deal, they fixed 11 them, put bigger strainers in, and this is another 12 approach that keeps it within a cost effective venue 13 anyway.
14 That's it.
15 CO-CHAIR REMPE: Okay. Anyone else?
16 MS. CARUSONE: Yes, this is Caroline 17 Carusone with NRR. I just wanted to confirm for the 18 full committee here meeting, then. So our focus will 19 be -- it will be an NRC staff presentation. We'll 20 incorporate some of the items that were of interest to 21 the subcommittee as part of that presentation.
22 And then, you know, without going into 23 sort of the big history of what other plants are 24 doing, we will provide some context around the 25 timeline and some of the major activities associated NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
189 1 with this issue to sort of open up the story to the 2 deeper discussion on this particular review.
3 Does that sound like -- I just want to 4 make sure I capture correctly what the focus is for 5 the full committee?
6 CO-CHAIR REMPE: That works for me. I 7 really like the Southern timeline that they showed.
8 It is -- do other members have some requests?
9 MEMBER PETTI: I think she captured it 10 quite well.
11 CO-CHAIR REMPE: Okay. So with that, I 12 want to thank everybody for their participation. All 13 of the members, all of the great conversations we have 14 had, and the responses from Southern as well as the 15 staff. I also want to thank the public that kept 16 their phones on mute. We have had a lot of trouble 17 with that lately, so I appreciate that.
18 And so with that, I guess I'd like to bang 19 the virtual gavel and close this meeting.
20 (Whereupon, the above-entitled matter went 21 off the record at 6:11 p.m.)
22 23 24 25 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.
(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433
Vogtle Risk-Informed GL 2004-02 Resolution (ACRS Thermal-Hydraulic and Risk-Informed Subcommittee Meeting)
May 19, 2021
Meeting Purpose
- Provide overview of risk-informed Generic Letter (GL) 2004-02 resolution history for Vogtle
- Provide high-level summary of SNCs technical approach
- Describe content of Vogtle license amendment request (LAR) submittal
Meeting Agenda
- Background Information
- Timeline of STP and Vogtle risk-informed resolution
- Overview of Vogtle risk-informed approach
- Plant modifications
- Testing and analyses
- Risk quantification
- Sensitivity and uncertainty analyses
- Summary of Vogtle technical report
- Summary of Vogtle LAR submittal
- Summary of differences between Vogtle and STP risk-informed approach 3
Vogtle Background Information Background Information - Vogtle Plant Layout
- Westinghouse 4-loop pressurized water reactor (PWR) with large dry containment
- Two redundant emergency core cooling system (ECCS) and containment spray (CS) trains
- Each train has a residual heat removal (RHR) pump, a high head charging pump, an intermediate head safety injection (SI) pump, and a CS pump
- Maximum pump design flow rates:
- RHR 3,700 gpm/pump
- CS 2,600 gpm/pump
- Two independent and redundant containment air cooling trains 5
Background Information - Strainer arrangement
- Replaced sump screen with stack-disk strainers
- Each GE strainer is similar with four stacks of disks
- RHR strainer (current):
18-disk (4.9 ft) tall, 765 ft2
- RHR strainer (modified):
16-disk (4.4 ft) tall, 678 ft2
Background Information - Plant Response to Loss-of-Coolant Accident (LOCA)
- Accumulators inject except for small break LOCA (SBLOCA)
- ECCS injection initiated from refueling water storage tank (RWST) to cold legs via RHR, SI, and charging pumps
- Containment spray initiated from RWST via CS pumps
- Valve to RWST closes at empty level alarm
- CS pumps secured no earlier than 1.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> after start of recirculation, and before approximately 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br /> depending on pressure and dose rate
- RHR pumps switched to hot leg recirculation at 7.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> 7
Why Risk-Informed Approach?
- Vogtle chose the risk-informed approach in a letter to the NRC in May 2013
- At the time, Vogtle had taken compensatory measures to address the concerns:
- Installed new orifice plates to allow ECCS throttle valves to open further and mitigate ex-vessel downstream effects
- Completed containment debris source walkdowns per NEI 02-01
- Performed latent debris sampling and characterization
- Removed Min-K insulation that could become debris sources
- Established programmatic and procedural changes to protect the new design and licensing basis
- The risk-informed approach was chosen to resolve the following challenges:
- Maximum strainer head loss from the deterministic approach exceeded pump NPSH limits
- Earlier testing and analysis showed fiber penetration quantity exceeded 15 grams per fuel assembly (g/FA)
- Meeting the 10 CFR 50.46 ECCS requirements deterministically would require bounding analyses and significant insulation removal or replacement 8
Timeline of STP and Vogtle Risk-Informed Resolution
Timeline for STP and Vogtle Risk-Informed GL 2004-02 Resolution Vogtle and Farley 90-Day Vogtle Sump Response to GL 2004-02 Strainer Upgrade Vogtle chose risk-informed approach Combined Vogtle and Updated Vogtle GL Farley GL 2004-02 2004-02 Submittal Submittal Vogtle Strainer Head Loss Testing STP switched STP Initial STP First LAR to RoverD Quantification Submittal method 2004 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2014 STP ACRS NEI 04-07 Revised content guide for SECY-10-113 Subcommittee Meeting GL 2004-02 submittal RG 1.174, Revision 2 GL 2004-02 WCAP-16530-NP-A WCAP-16793-NP-A, Revision 2 NRC SE of NEI 04-07 NUREG-1829 RG 1.82, Revision 4 WCAP-16406-P-A, Revision 1 SECY-12-0093 Proposed 3 RG 1.200, Revision 2 options of resolving GSI-191 10
Timeline for STP and Vogtle Risk-Informed GL 2004-02 Resolution (Contd) .
NRC agreed to review a Technical Report by Vogtle Vogtle Technical Report for risk-informed resolution NRC Staff Evaluation of Vogtle Technical Report NRC Audit Vogtle Fiber Penetration Testing Vogtle LAR Submittal Updated Vogtle NRC Technical Report Software NRC RAIs on Audit LAR Submittal STP revised NRC Audit LAR with NRC SE on STP LAR RoverD method Submittal Vogtle Responses to RAIs 2015 2015 2016 2017 2018 2019 2020 2021 2021 STP ACRS Subcommittee NRC technical evaluation of in-RG 1.174, Revision 3 vessel effects Meeting STP ACRS Subcommittee Meeting NRC review guidance on in-vessel effects STP ACRS Subcommittee Meeting NRC SE on TSTF-567 WCAP-17788-P, Revision 1 ACRS Letter on STP Submittal PWROG-16073-P, PWROG guidance on in-vessel effects and TSTF-567 RG 1.200, Revision 3 11
Vogtle Risk-Informed Approach Overview
Summary of Vogtle Risk-Informed Approach
- Vogtles risk-informed approach for GL 2004-02 resolution is similar to STP
- Overall evaluation of risk attributable to debris is based heavily on physical models that have been used in the past and accepted by the NRC for GSl-191 resolution
- Multiple breaks postulated at each Class 1 weld and analyzed for debris generation using the BADGER software
- The NARWHAL software evaluated each break against failure criteria to determine if it would result in strainer and/or in-vessel failures due to effects of debris
- Conditional failure probabilities (CFPs) calculated by NARWHAL were used as inputs to the Vogtle probabilistic risk assessment (PRA) model to determine the debris related:
- Change in core damage frequency (CDF)
- Change in large early release frequency (LERF)
- The results showed that the risk associated with LOCA-generated debris is very small and within Regulatory Guide (RG) 1.174 Region III 13
Overview of Vogtle Risk-Informed Approach Post Accident Fiber Penetration Sump Volume Testing Debris Transport Analysis NARWHAL A break passes all failure criteria Chemical Effects Debris Analysis Generation Strainer Head Analysis Loss Testing No Yes Break fails Break passes CAD model of Various Bounding Analyses Vogtle Upstream effects Containment Ex-vessel wear and blockage Vortexing for bounding conditions LOCADM analysis Risk Quantification Technical and Calculate delta core License Amendment damage frequency and Request Submittals delta large early release frequency This flow chart is adapted from Figure 1-1 in the July 2018 Vogtle Technical Report.
14
Failure Criteria Used in Vogtle Analysis Failure Criteria Method for Addressing Strainer head loss exceeds pump NPSH Break-specific analysis based on tested debris margin or strainer structural margin limits* and maximum tested head losses Strainer head loss exceeds half strainer Break-specific analysis based on tested head submergence for partially submerged strainer losses and strainer submergence Gas voids from degasification or flashing Break-specific analysis based on head loss, exceed pump limits pool temperature, etc.
Penetrated debris exceeds in-vessel fuel Break-specific analysis based on penetration blockage and boron precipitation limits testing and flow splits Upstream blockage prevents water from Bounding analysis reaching sump Pumps fail due to air intrusion from vortexing Bounding analysis Penetrated debris exceeds ex-vessel wear and Bounding analysis blockage limits Debris accumulation on cladding prevents Bounding analysis adequate heat transfer (LOCADM) (No longer required per NRC guidance)
- Breaks that generate and transport more debris of any one type than tested were assumed to fail. 15
Vogtle Risk-Informed Approach Plant Modifications, Testing, and Analyses
Plant Modifications
- Decision on additional plant modifications considered risk assessment results, PRA insights, fiber penetration sensitivity, and other constraints
- Changed emergency operating procedure to continue RWST drain down by the RHR pumps to the RWST empty level setpoint
- Will reduce height of RHR strainers by ~6 inches by removing 2 top disks
- Fall 2021 outage for Unit 1
- Spring 2022 outage for Unit 2
- Combination of these modifications results in strainers being fully submerged for most break scenarios and lowers overall risk 17
Containment CAD Model
- Developed in Autodesk Inventor using plant drawings for Unit 1
- Applicable for both units due to similarities between the two units
- Validated by comparison with laser scans
- Used as input for various calculations, as shown in the next few slides
- Debris generation (insulation and coatings zone of influence (ZOI) volumes)
- Debris transport (Computational Fluid Dynamic (CFD) model geometry, blowdown and pool fill volumes, spray distribution areas)
Vogtle Containment CAD Model 18
Debris Generation
- Insulation, fire barrier and qualified coatings
- Calculated quantity and size distribution for each type of debris
- Breaks postulated at all Class 1 welds
- Double-ended guillotine breaks (DEGBs)
- Partial breaks from 1/2 inch up to 31 inches in 45° increments around each weld
- ZOI size and shape consistent with NEI 04-07
- Unqualified coatings and latent debris quantities are identical for all breaks
- Coatings debris was assumed to fail as 100%
particulate 19
Break Zone of Influence for Debris Generation Analysis Hemispherical ZOIs used for partial breaks and scaled based on postulated break size Eight orientations (every 45°)
analyzed for partial breaks for each weld Spherical ZOIs used for DEGBs and scaled based on pipe diameter 20
Containment Sump Water Level
- Sump pool depth is evaluated on a break-specific basis using conservative inputs to minimize water level
- The evaluation considers break location and size for the appropriate contribution of RWST, reactor coolant system (RCS), and SI accumulators to pool volume
- At start of switchover to sump recirculation, strainers (as modified) are fully submerged for all breaks except for large reactor cavity breaks that actuate containment sprays
- Strainers are fully submerged for all breaks before the switchover to sump recirculation is complete 21
Debris Transport
- Debris transport used logic tree approach consistent with NEI 04-07 guidance
- Blowdown transport based on containment volumes and flow paths through grating, etc.
- Washdown transport based on whether sprays are initiated and flow paths through grating, etc.
- Pool fill-up transport based on pool volume and inactive and sump cavity volumes
- Recirculation transport based on CFD analysis using the Flow-3D software
- Erosion of small and large pieces of fiberglass debris based on Alion testing Streamlines showing flow to each strainer LBLOCA on Loop 4 Steam Generator Compartment 22
Strainer Head Loss Testing
- Conventional (i.e., fiber and particulate) and chemical debris head loss based on tank tests
- Head loss testing performed by Alion in 2009
- Consistent with the NRC March 2008 Guidance
- Test strainer: prototypical 7-disk module (65.6 ft2)
- Walls and suction pipe arranged consistent with plant strainer
- Nukon fines prepared by mixing shredded and boiled Nukon with a paint mixer
- Used sparger, mechanical mixers and manual agitation to keep debris in suspension
- Pre-prepared precipitates (using WCAP-16530 recipe) were used for chemical effects
- Head losses were extrapolated to 30 days as necessary and corrected for temperature and approach velocity following March 2008 guidance 23
Fiber Penetration Testing
- Fiber penetration based on tank tests with a prototype strainer module
- Performed nine tests at Alden in 2014
- Approach similar to other tests observed by staff
- Varied strainer approach velocity, number of strainer disks and boron / buffer concentrations
- Prepared Nukon fines per NEI Guidance
- Fiber debris added in batches with increasing batch size
- Collected fiber penetration using 5-m filter bags
(>90% capture efficiency)
- Collected time-dependent fiber penetration data
- Filter bags swapped before adding a fiber batch
- Additional filters used for long-term shedding
- A curve fit of the test data was used to evaluate fiber penetration for plant strainers 24
Bounding Evaluation of Certain Failure Criteria
- Upstream Blockage
- Qualitatively evaluated multiple passageways between containment annulus and inside secondary shield wall, and refueling canal drain
- All flow paths were shown not to be blocked
- Ex-Vessel Downstream Effects Calculation
- Uses methodology and acceptance criteria of WCAP-16406-P-A (Rev. 1) and NRC Safety Evaluation (SE) with conservative debris concentrations
- Showed acceptable wear and no blockage of piping and components on ECCS or CS flow path (e.g., pumps, values, heat exchangers, orifices, spray nozzles)
- Vortexing
- Based on vortex testing performed during 2009 strainer head loss testing
- No air-entraining vortices observed during testing with clean or debris-laden strainer at conservatively lower strainer submergences and higher approach velocities 25
Vogtle Risk-Informed Approach Risk Quantification
Sub-Model Integration in NARWHAL
- Sub-models for evaluating debris impact are integrated using NARWHAL
- Post-accident conditions
- Debris generation quantities
- Debris transport fractions
- Chemical effects
- Time-dependent water/debris mass balance
- Strainer debris limits
- Strainer head loss
- NPSH margin
- Strainer structural limit
- Degasification/flashing
- Gas void limits
- Debris penetration
- Core debris limits
- NARWHAL software was developed under ENERCONs QA program
- Vogtle-specific NARWHAL model also developed under ENERCONs QA program 27
Key Elements of NARWHAL Analysis
- Chemical effects
- Chemical effects methodology consistent with WCAP-16530 with refinements for phosphate passivation of aluminum surfaces based on University of New Mexico (UNM) testing
- Break-specific chemical precipitate quantities calculated using NARWHAL
- Aluminum solubility limit (based on Argon National Lab (ANL) equation) was used to determine break-specific precipitation timing
- When concentration exceeds solubility limit, 100% of aluminum in solution assumed to precipitate as sodium aluminum silicate (SAS)
- If solubility limit not exceeded before 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />, 100% of aluminum in solution assumed to precipitate as SAS at 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />
- Strainer Head Loss
- A failure was recorded if any one of the tested debris quantities is exceeded
- Total strainer head loss for each time step was calculated by combining clean strainer head loss, debris head loss, and extrapolation constant (as necessary)
- A rule-based approach to calculate debris head loss based on testing results 28
Key Elements of NARWHAL Analysis
- NPSH margin
- No containment accident pressure was credited
- Impact on pump NPSH required by void fraction was addressed per the methodology of Regulatory Guide 1.82
- Flashing and Degasification
- A flashing failure was recorded if the pressure inside the strainer (at its top elevation) was less than the vapor pressure at corresponding sump temperature
- Void fraction due to degasification was evaluated at mid-height of the strainer
- A degasification failure was recorded if void fraction at pump suction was > 2%
- Air bubbles formed downstream of the strainer were assumed to stay intact as they travel to the pump suction without crediting bubble compression
- Credited small amount of accident pressure (3.5 psi) for a short period of time (first 2.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> of event) to preclude flashing and degasification failures 29
Key Elements of NARWHAL Analysis
- Fiber Penetration
- Fiber penetration curve fit from testing was used in time-dependent analysis to quantify fiber debris reaching the reactor core
- Used the WCAP-17788, Revision 1 methodology and acceptance criteria in accordance with the NRC 2019 guidance on in-vessel effects
- No alternate flow paths in the reactor barrel/baffle region were credited 30
Tracking of Movement of Flow and Debris Spray Nozzles CS Pumps Core Reactor Containment Break ECCS Vessel Compartments RWST Pumps ECCS and CS Sump Pool Strainers Acronyms:
- RWST: Refueling Water Storage Tank 31
NARWHAL Analysis of Conditional Failure Probability
- NARWHAL evaluated all postulated breaks in a holistic, time-dependent manner to determine which breaks pass and fail the strainer and core failure criteria
- Evaluation was performed for various equipment configurations
- Used a top-down method to allocate overall plant-wide LOCA frequencies from Vogtle PRA model to individual welds and break sizes
- Calculated conditional failure probability (CFP) for large breaks (> 6) based on NARWHAL calculated failures and LOCA frequency weight for various size ranges
- CFPs are reported as a function of equipment configuration, break size, and failure cause (sump strainer failure or core blockage)
- The analysis showed no small (< 2) or medium (2 - 6) break LOCAs that fail for any equipment configurations
- The analysis showed no reactor core failures 32
Secondary Side Breaks
- Secondary side breaks inside containment (SSBI) were evaluated in a manner similar to primary side breaks with the following exceptions:
- Breaks postulated in 5 ft intervals on main steam and feedwater piping
- All breaks assumed to be DEGBs
- Smaller ZOI sizes due to reduced pressure and temperature in accordance with NEI 04-07
- Lower ECCS flow rate
- Conservatively assumed both trains of containment spray would initiate for all breaks
- The main steam line breaks and feedwater line breaks were evaluated separately in NARWHAL for five different equipment configurations
- None of the feedwater line breaks produced a sufficient quantity of debris to fail
- Main steam line breaks resulted in some failures for single train operation or when both CS pumps fail 33
Risk Quantification
- Vogtle base PRA model modified to incorporate events for GSI-191 sump strainer and core blockage failures, with associated LOCA initiating events and equipment configurations (GSI-191 PRA model)
- Core blockage failures added under CDF top gate
- CFP values are input in PRA model to calculate CDF and LERF associated with high likelihood configurations
- CDF and LERF also quantified using bounding or conservative methods for unlikely equipment configurations, seismically induced LOCAs, and secondary side breaks 34
Risk Quantification Results
- Risk quantification results and base CDF and LERF values are compared to RG 1.174 acceptance guidelines and found to be in Region III 2.81x10-08 3.18x10-10 4.39x10-05 Unit 1 1.73x10-06 Unit 1 5.05x10-05 Unit 2 1.90x10-06 Unit 2 35
Vogtle Risk-Informed Approach Sensitivity and Uncertainty Analyses
Sensitivity Analysis
- Sensitivity analysis was done for 20 different input parameters and one parameter was varied at a time to determine its impact on the overall results
- Selected a range of values for each parameter based on its value used in base case NARWHAL model
- If base case value already skewed in conservative direction, min [max] value was assumed to be 10% lower [higher]
- If available, design limits of an input parameter were used
- Otherwise, the min [max] value assumed to be 25% lower [higher] than the base case value
- The parameters that have the most significant effect on CDF are:
- In-vessel fiber limit for hot leg breaks
- LOCA frequency values
- RHR pump flow rates 37
Uncertainty Quantification
- In accordance with RG 1.174 and NUREG-1855, the uncertainty evaluation included parametric, model, and completeness uncertainty
- Parametric uncertainty addressed by evaluating the worst-case combination of inputs for strainer failure and core failure
- For each failure scenario, four different cases were run using different combinations of inputs
- Even with the worst-case values for all inputs, risk is still in Region III
- Model uncertainty addressed by postulating alternate models for each model where no consensus exists
- Nine models were analyzed
- All model uncertainty results are within RG 1.174 Region III
- Completeness uncertainty was qualitatively determined to be low, given nearly four decades of industry and NRC research and analysis on strainer performance issues
- The results of the uncertainty analysis show that there is high confidence that the risk increase associated with LOCA-generated debris is very small 38
Technical Report Submittal Technical Report Submittal by Vogtle
- By letter dated February 14, 2017, the NRC agreed to review an SNC technical report (TR) that does not rely on WCAP-17788
- WCAP-17788 for in-vessel downstream effects was under review by the NRC
- The purpose of the TR is to receive NRC review and approval of the SNC supplemental GL 2004-02 response using a risk-informed approach
Technical Report Submittal by Vogtle
- The TR summarized the risk-informed approach to resolve GL 2004-02, and superseded Vogtles previous responses to GL 2004-02
- Enclosure 1 provided a high-level summary of the submittal and was organized in accordance with draft RG 1.229 Section C
- Enclosure 2 described plant conditions, analyses and testing that informed the resolution following NRC content guide for GL 2004-02 responses (including proprietary information)
- Addressed NRC RAIs from previous Vogtle GL 2004-02 submittals
- Enclosure 3 summarized the methodology and results of risk quantification, and sensitivity and uncertainty analyses using NARWHAL and Vogtle PRA model
- Organized according to draft RG 1.229 Appendix A
- Enclosure 4 described defense-in-depth measures and safety margin
- Enclosure 5 duplicated the content of Enclosure 2 with the proprietary information redacted 41
Technical Report Submittal by Vogtle
- SNC responded to the RAIs and questions raised during audits, incorporated these into an updated version of the TR and submitted it to the NRC in July 2018
- NRC issued a staff evaluation of the TR in September 2019 and concluded that:
- The TR contains sufficient information for SNC to address GL 2004-02 for Vogtle except for in-vessel downstream effects
- The TR was acceptable for use in a future Vogtle licensing application, subject to nine conditions and limitations
- SNC discussed responses to the conditions and limitations and the proposed LAR submittal with the NRC during public meetings in November 2019 and April 2020
Five Key Principles of RG 1.174
- The Vogtle 2020 LAR submittal included a request for exemption to specific criteria in 10 CFR 50.46(a)(1)
- The Vogtle TR included an enclosure, which shows that defense-in-depth will be maintained
- The Vogtle TR included an enclosure, which shows significant safety margins in the GSI-191 evaluation
- The Vogtle TR and LAR submittal showed that the risk associated with LOCA-generated debris is very low
- The Vogtle 2020 LAR submittal described the plant procedures and programs that monitor, control, and assess plant changes that could impact sump strainer performance 43
License Amendment Request Submittal Overview
Purpose of LAR Submittal
- Request for exemption from 10 CFR 50.46(a)(1) to support Vogtle risk-informed approach to GL 2004-02 closure as an alternative to the current deterministic evaluation required by 10 CFR 50.46(a)(1)
- Request for license amendment to:
- Implement a risk-informed approach for addressing GL 2004-02
- Adopt Technical Specification Task Force (TSTF) Traveler 567 to add a new Technical Specification (TS) for containment sump 45
Content of LAR Submittal
- Enclosure 1: Request for exemption from requirements of 10 CFR 50.46(a)(1)
- Proposed LAR includes three separate enclosures:
- Enclosure 2: Implementation of a Risk-Informed Approach for Addressing GSI-191
- Includes the following Attachments:
- Proposed FSAR Changes (Mark-Up) for Information Only
- Guidance for Supporting Operability Evaluations for Information Only
- Updated Evaluation for In-Vessel Effects and Coatings
- Enclosure 3: Proposed Changes to the TS for implementation of TSTF-567
- Includes the following Attachments:
- Proposed Technical Specification Changes (Mark-Up)
- Revised Technical Specification Pages
- Proposed Technical Specification Bases Changes (Mark-Up) for Information Only
- Enclosure 4: Regulatory Evaluation and Environmental Consideration 46
License Amendment Request Submittal Exemption Request (Enclosure 1)
- An exemption is requested from certain requirements of 10 CFR 50.46(a)(1), other properties, as it relates to using deterministic methodology to evaluate the effects of debris on long-term core cooling
- The request for exemption under 10 CFR 50.12 supports the Vogtle risk-informed approach to GL 2004-02 closure as an alternative to the current deterministic evaluation required by 10 CFR 50.46(a)(1)
- The exemption will apply only to the effects of debris as described in Enclosures 2 and 3 of the July 2018 submittal
- The exemption will apply to any breaks that can generate and transport debris exceeding the Vogtle analyzed limits, provided that the CDF and LERF remain in RG 1.174 Region III 48
Evaluation for Exemption Request
- The requested exemption was evaluated against the conditions in 10 CFR 50.12(a)
- The request meets the requirements in 10 CFR 50.12(a)(1) for granting an exemption from the regulation
- Special circumstances in 10 CFR 50.12(a)(2)(ii) and (iii) are present
- The exemption was reviewed for environmental impacts and was determined to be eligible for categorical exclusion per 10 CFR 51.22(c)(9)
- The risk associated with post-accident debris effects is within RG 1.174 Region Ill, Very Small Change
- The requested exemption is consistent with the RG 1.174 Key Principle #1
- The proposed licensing basis change meets the current regulations unless explicitly related to a requested exemption 49
License Amendment Request Submittal Implementation of a Risk-Informed Approach for Addressing GSI-191 (Enclosure 2)
Overview of Technical Evaluation
- The proposed licensing changes in this submittal are based on the July 2018 submittal and the NRC staff evaluation with no deviations
- Approval of the proposed change would allow Vogtle to use a risk-informed approach to address effects of accident-generated debris on containment sumps
- Vogtles technical evaluation using the risk-informed approach was described in the July 2018 submittal
- Used break-specific analysis to identify break scenarios that fail any GSI-191 acceptance criteria
- Showed that the risk associated with post-accident debris is "very small" as defined by Region Ill of RG 1.174 51
Overview of Engineering Analysis
- Approval of the LAR does not change the design and licensing basis descriptions of accidents requiring ECCS and CS system operation, including analysis methods, assumptions, and results provided in FSAR Chapters 6 and 15
- Safety margin and defense-in-depth are maintained with high probability
- The calculated risk is conservative and the actual risk is likely much lower
- The LAR submittal also addressed the 9 conditions and limitations identified in the NRC staff evaluation of the TR
- Addressed RG 1.174 Key Principles 1 and 5
- Conformed to the applicability of the TR
- Responded to in-vessel downstream effects following NRC review guidance
- Identified key elements of the risk-informed analysis
- Described controls to ensure relevant elements are periodically updated
- Corrected an error in the evaluation of coatings debris 52
License Amendment Request Submittal Proposed Changes to the Technical Specifications (Enclosure 3)
Adoption of TSTF-567
- Added an action to address the condition of containment sump made inoperable due to containment accident generated debris exceeding analyzed limits
- The containment sump debris limits are provided in TS Bases 54
License Amendment Request Submittal Regulatory Evaluation and Environmental Consideration (Enclosure 4)
Regulatory Evaluation
- The proposed license amendment does not involve a significant hazards consideration under the standards in 10 CFR 50.92(c)
- The licensing basis changes are shown to meet the five key principles of RG 1.174
- The PRA model used for the Vogtle risk-informed GSI-191 evaluation complies with RG 1.200 Revision 2 and can be applied in regulatory decision-making 56
Precedent
- The proposed licensing changes are similar to the license amendment and 10 CFR 50.46(a)(1) exemption granted to STP for implementing their risk-informed approach to address GSI-191 concerns
- Key similarities include:
- Use of RG 1.174 acceptance guidelines and key principles
- Identification of key methods and approaches in the risk-informed methodology for evaluating potential departure from a method of evaluation in FSAR
- Associated request for exemption from 10 CFR 50.46(a)(1) other properties
- TS changes that provide for additional time to address the effects of debris on ECCS and CS system operability
- Key differences include:
- Software used for risk analysis (NARWHAL vs. CASA Grande)
- Methodology used for risk quantification (CFP approach vs. RoverD approach)
Environmental Consideration
- Pursuant to 10 CFR 51.22(b), no environmental impact statement or environmental assessment needed to be prepared for the proposed LAR because it meets the eligibility criterion for categorical exclusion in 10 CFR 51.22(c)(9).
- The proposed amendment does not involve a significant hazards consideration
- The proposed amendment does not pose a significant change in the types or a significant increase in the amounts of any effluents that may be released offsite
- The proposed amendment does not result in a significant increase in individual or cumulative occupational radiation exposure 58
Remaining Actions
- Implement containment RHR sump strainer modifications during fall 2021 (Unit 1) and spring 2022 (Unit 2) outages
- Incorporate TS Changes and risk-informed methodology after modifications are complete
- Vogtle has one TS and one FSAR for both units
- Instead of staggered implementation, Vogtle would like to implement the change after the modifications for both units are complete 59
Differences between Vogtle and STP Risk-Informed Evaluations
Differences between Vogtle and STP Risk-Informed Approach
- Framework for Quantifying Risk
- Vogtle used CFP method with PRA model
- More traditional approach for a risk-informed evaluation
- Used deterministic tests and analyses (similar to STP) to quantify CFPs
- Evaluation Software
- STP used CASA Grande to evaluate debris generation and transport, and FIDOE and RUFF to evaluate in-vessel effects
- Vogtle used BADGER for debris generation and NARWHAL for integrated evaluation of strainer and in-vessel effects
- Sub-Model Integration
- STP used separate models to analyze phenomena associated with debris effects
- Vogtle used an integrated model to evaluate debris transport, chemical effects, strainer head loss, NPSH, degasification, flashing, and core blockage 61
Differences between Vogtle and STP Risk-Informed Approach (Contd)
- Equipment Configurations
- STP used single-train failure case as the base case for strainer evaluation and considered two-train failure case for limiting strainer performance
- STP considered a variety of equipment configurations for in-vessel effects analysis
- Vogtle evaluated seven equipment configurations with an integrated evaluation of strainer and core failures for each configuration
- Strainer Acceptance Criteria
- STP used a fiber debris limit as the only acceptance criterion
- Vogtle identified debris limits for fiber, particulate, and chemical precipitates and used head losses (based on test data) associated with varying debris loads to evaluate NPSH, strainer structural failure, degasification, and flashing 62
Differences between Vogtle and STP Risk-Informed Approach (Contd)
- Time-Dependent Effects
- STP did not explicitly model time-dependent effects (such as chemical precipitation) for strainer evaluation
- Vogtle used a time-dependent evaluation for integrated strainer and in-vessel evaluation using methodology that has been accepted for deterministic evaluations
- In-Vessel Effects Methodology
- STP used thermal-hydraulic modeling to address core blockage issues
- Assumed that all hot leg breaks greater than 16 inches would fail
- Closed boron precipitation in June 2020
- Vogtle used WCAP-17788 method in accordance with the NRC 2019 guidance on in-vessel effects for the assessment of core blockage and boron precipitation 63
Differences between Vogtle and STP Risk-Informed Approach (Contd)
- Risk Contribution from Secondary Side Breaks
- STP evaluated secondary side breaks with a bounding analysis (assuming strainer failure for all secondary side breaks requiring sump recirculation)
- Vogtle evaluated secondary side breaks with a conservative analysis (calculating CFPs for main steam and feedwater line breaks assuming all breaks are DEGBs)
- Presentation of Safety Margins
- STP provided a summary of conservatisms in various areas of the analysis
- Vogtle identified safety margins as conservatisms throughout the evaluation, but explicitly differentiated between safety margin and operating margin
- Uncertainty Quantification
- STP did not provide a formal evaluation of uncertainties in their final submittal, but provided sufficient information in their discussion of safety margins
- Vogtle performed formal uncertainty quantification
- Evaluated parametric uncertainty for all parameters that were not bounding
- Evaluated model uncertainty for all models that were not consensus models 64
Differences between Vogtle and STP Risk-Informed Approach (Contd)
- Quality Assurance
- Vogtle addressed QA requirements early in the process with most supporting documentation developed under an Appendix B QA program 65
Recap and Summary
- Background Information
- Timeline of STP and Vogtle risk-informed resolution
- Overview of Vogtle risk-informed approach
- Plant modifications
- Testing and analyses
- Risk quantification
- Sensitivity and uncertainty analyses
- Summary of Vogtle technical report
- Summary of Vogtle LAR submittal
- Summary of differences between Vogtle and STP risk-informed approach 66
Decision on Procedural Change and Strainer Modification
- Decision on reducing RHR strainer height considered risk assessment results, PRA insights, fiber penetration sensitivity, and other constraints
- Modification would result in more breaks with fully submerged strainers, reduce overall risk, and provide additional safety margin
- Thermal hydraulic analysis showed that many breaks may not actuate containment spray
- Changed emergency operating procedure (EOP) to continue RWST drain down by the RHR pumps to the RWST empty level setpoint
- Increasing RWST Technical Specification minimum water level setpoint to High Level setpoint would only increase the sump level by less than 3
- Fiber penetration testing showed insignificant impact on in-vessel effects with reduced strainer surface area 1
NARWHAL Sensitivity
- A sensitivity study was performed in August 2015 using NARWHAL to document the benefit of the EOP change and strainer modification in reducing overall risk
- The 2015 sensitivity study cannot be directly compared with the risk results in the 2018 Technical Report because of various input and methodology changes to the risk quantification made since 2015
- The sump screen from the original plant construction had a surface area of 54 ft2
- The shortened RHR strainer with a surface area of 678 ft2 significantly increases the strainer surface area 2
Conditional Failure Probability Results CFP Results for large pipe break LOCAs Equipment Configuration Core Strainer A Strainer A Strainer B and B Only Only No Equipment Failure 0 0.0118 0 0 RHR Pump B Failure 0 N/A 0.0679 N/A Charging Pump B Failure 0 0.0118 0 0 SI Pump B Failure 0 0.0118 0 0 Train B (ECCS and CS) Failure 0 N/A 0.1017 N/A CS Pump B Failure 0 0.0139 0 0 Both CS Pumps Failure 0 0.0177 0 0 1
Size Ranges for CFP Analysis 2
Sensitivity Analysis Results 1
Parametric Uncertainty Quantification Results 2
Model Uncertainty Quantification
- Model uncertainty addressed by postulating alternate models for each model where no consensus exists
- Break model (continuum vs. DEGB-only)
- LOCA frequencies (geometric mean vs. arithmetic mean)
- LOCA frequency allocation (top-down vs. hybrid)
- CS actuation (hot leg breaks larger than 15 inches vs. multiple options including no breaks and all breaks larger than 2, 6 or 15 inches)
- Aluminum metal release equation (UNM vs. WCAP-16530)
- Fiber bed thickness required for chemical head loss (0.45 inches vs. 0 inches)
- LBLOCA size range discretization
- Chemical effects
- NARWHAL time step size
- Similar to parametric uncertainty results, all model uncertainty results are within RG 1.174 Region III
- The results of the uncertainty analysis show that there is high confidence that the risk increase associated with LOCA-generated debris is very small 3
Model Uncertainty Quantification Results 4
Model Uncertainty Quantification Results (Contd) 5
Vogtle Unit 1 and 2 Risk-informed Resolution of Generic Letter 2004-02 ACRS Subcommittee Meeting May 19, 2021
Presentation Outline
- Background - Generic Safety Issue (GSI) - 191, Generic Letter (GL) 2004-02
- Actions: strainer and reactor vessel
- Staff Evaluation of Vogtle GL 2004-02 analysis
- Technical Report
- Limitations and Conditions
- License Amendment Request 2
ECCS Debris Effects Background
- In the early 1990s, NRC identified that sump blockage due to loss-of-coolant accident (LOCA) generated debris could challenge emergency core cooling system (ECCS) and containment spray (CS)
- GSI-191 for PWRs issued in 1996 to address strainer performance
- GL 2004-02 requested plant-specific evaluation of debris to demonstrate compliance with 10 CFR 50.46 3
General Licensee Actions to Address Debris Effects on Long Term Core Cooling (LTCC)
- Installed larger strainers
- Removed unnecessary debris sources
- Removed/replaced some problematic materials, or reinforced with banding and jacketing
- Implemented administrative controls
- Control materials in containment
- Implemented procedural/operational changes
- Improved chemical effects attributes 4
Overall Status of Compliance
- 3 Staff Requirements Memorandum options from SECY-12-0093
- Option 1 - Closeout using existing guidance
- Option 2 - Perform additional testing and analysis, including risk-informed resolution
- Option 3 - Treat strainer deterministically and in-vessel using risk analysis
- 19 units demonstrated compliance using Option 1
- Remainder of plants chose Option 2 5
Option 2 Plant Compliance
- 2A - Deterministic and Transition Break Size
- 29 Units likely to use typical deterministic methods
- 5 Units likely to use NEI 04-07, Section 6 evaluation (TBS)
- Staff currently reviewing those in-house and waiting on the remainder
- 2B - Full Risk-Informed
- STP has demonstrated compliance
- Vogtle review discussed in this meeting
- Wolf Creek and Point Beach have not yet submitted a LAR. A LAR was recently received from Callaway.
6
Types of strainers
- Uniform Flow PCI stack disk strainer
In-Vessel Downstream Effects Status (IVDE)
- 19 units resolved issue based on low fiber loading at the core
- Basis is WCAP-16793, Rev. 2 and staff SE
- Two STP units used plant-specific thermal-hydraulic (TH) analysis
- Remainder of licensees using recent staff guidance
- Staff review guidance (ML19228A011) based on
- Industry testing and analysis
- NRC analysis
- Safety Significance
- Discussed with ACRS in April and September 2019 8
IVDE - WCAP-17788 Summary
- Plant-specific evaluations with event timing considerations
- Higher debris limits
- Address NRC staff and ACRS comments from WCAP-16793
- Effects of debris, boric acid precipitation (BAP)
- Performed fuel assembly (FA) testing, TH analysis, chemical effects (autoclave/filtration) testing 9
Methodology Comparison - Vogtle and South Texas Project (STP)
- RoverD
- CAD for debris generation
- Casa Grande vs. NARWHAL
- Vogtle used more refined headloss model
- Very small change in risk (RG 1.174)
- Difference in break evaluation 360 vs 8
- In-vessel - STP used TH, Vogtle used staff guidance
- Deterministic staff approved methods for some aspects 10
Risk over Deterministic Methodology (RoverD)
Deterministic Test Data Debris Limit Established Risk Informed Analysis Calculate Debris for Individual Breaks R
over D
Scenario Deterministically Debris Acceptable No Calculated Testing Limit Yes 11
Evaluation of Technical Report
- Most review of technical areas was completed during review of the technical report (Staff Evaluation ML19120A469)
- Open technical issues were identified as limitations and conditions (L&Cs)
- Staff evaluated the 5 key principles of risk-informed regulation
- Used approved guidance to evaluate risk and deterministic aspects 12
Evaluation of LAR
- More focused on evaluation of licensing issues
- TS changes
- L&Cs from the staff evaluation of the licensees plant-specific Technical Report
- The staff evaluation of the Technical Report was for Vogtle-specific only and not for generic use
- Staff evaluated the 9 limitations and conditions in the technical report 13
References
- Staff guidance on coatings, chemical effects, and headloss
- Draft RG 1.229
- WCAP-16530 Chemical Effects
- WCAP-16406 Ex-vessel Downstream Effects
- In-vessel staff review guidance 14
Staff Methodology Five Key Principles of Risk-Informed Regulation
- 2. Change is consistent with defense-in-depth philosophy.
- 1. The proposed licensing basis change meets the current 3. Maintain sufficient regulations or an safety margins.
exemption is requested. Integrated Decisionmaking
- 5. Impact of proposed
- 4. Proposed increases change should be in risk should be small monitored using and consistent with the performance Commissions Safety measurement Goal Policy Statement.
strategies.
15
Principle 1 Meet Regulations or Request Exemption
- 10 CFR 50.46c rulemaking status
- Exemptions requested from use of deterministic analysis method
- Acceptance Criteria for ECCS
- Did not request exemption to GDC like STP
- GDC 35, 38, 41 16
Principles 2 and 3 Safety Margins and Defense-In-Depth
- Licensee met guidance of RG 1.174 and provided significant safety margins and Defense-in-depth (DiD)
- Safety margins include construction and inspection per industry codes and the use of licensing basis values when assigning strainer failure criteria.
- DiD includes actions identified that are taken in response to the loss of the normal ECCS function. DiD also includes verification that balance is maintained among prevention and mitigation, redundancy is maintained, barrier independence is maintained, etc.
17
Principle 4 Deterministic Inputs to Risk Analysis
- Debris Source Term
- Used NRC approved guidance for all areas
- Calculations performed in BADGER
- Differences from typical deterministic evaluations
- For partial breaks, all weld locations evaluated for multiple orientations instead of focusing on the limiting large break
- 8 orientations vs. 360 for STP
- Double-ended guillotine break (DEGB) source term uses the same method as typical deterministic calculations
- Source term and transport calculated for each break and compared against tested amount
- The most conservative orientation was selected for partial breaks at each weld location
- Assumptions and calculations verified independently by SwRI
- Extensive Review of BADGER/CAD debris generation software 18
31 inch 29 inch 20 inch 12 inch 6 inch 19
Minimum Average Maximum Main Loop, 56 Welds 10-16 inch, 39 welds 6-8 inch, 77 welds 2.5-4 inch, 122 welds
< /=2 inch, 119 welds 20
Principle 4 Deterministic Inputs to Risk Analysis
- Extensive review of NARWHAL that implemented transport, headloss, failure criteria, conditional failure probability, and sensitivity calculations
- Debris Transport - Strainer Evaluation
- Used NRC approved guidance implemented via NARWHAL
- Correction of coatings transport was the subject of an L&C that resulted in small changes in overall risk. Correction was documented in the LAR.
- Debris Transport - In-Vessel Effects
- Fiber penetration determined via testing
- Used conservative bypass values from testing
- Calculated fiber amounts arriving at the core for hot-leg breaks using conservative plant state (pump combinations)
- Evaluated fiber amount reaching the core against limits using NRC staff guidance 21
Principle 4 Deterministic Inputs to Risk Analysis
- Impact of Debris - Strainer
- Strainer evaluated at tested debris loads/dPs for effects on net positive suction head, structural, deaeration, partial submergence failure, vortexing, and flashing
- If a transported debris amount of any type exceeded the tested amount, the scenario was assumed to fail
- Testing and evaluations were performed using staff approved guidance
- Testing showed that increasing fiber amounts result in greater head losses
- Majority of breaks bounded by test results
- Some breaks generate much larger debris amounts 22
Principle 4 Deterministic Inputs to Risk Analysis
- Impact of Debris - In-vessel
- This area was not addressed in the technical report and was the subject of an L&C. It was addressed in the LAR.
- Debris amounts low enough to permit adequate cooling flow to the core based on WCAP-17788 findings and NRC Staff Review Guidance
- Evaluated core inlet fiber amount that was > WCAP limit, but < WCAP total in-vessel limit for Vogtle RCS design
- Non-uniform debris distribution at core inlet
- HLSO time < 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br />, Chemical effects time > 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br />
- Other WCAP parameters bound Vogtle values 23
Summary of Systematic Risk Assessment
- Used existing acceptance guidelines in RG 1.174 to address Principle 4
- Vogtle determined only LBLOCAs (> 12 inches) contribute to strainer failure
- CDF (LERF) = CDF with debris - CDF no debris, compared to RG 1.174, Region III acceptance guideline 24
Summary of Systematic Risk Assessment
- Sensitivity and Uncertainty Analyses (focused on CDF)
- Performed consistent with NUREG-1855 guidance
- Parametric sensitivity and uncertainty analysis
- All cases fall within Region III
- Model uncertainty analysis
- Key assumptions and sources of uncertainty
- Strainer testing used conservative NRC staff guidance
- Completeness uncertainty analysis
- Results fall in Region III acceptance guidelines in RG 1.174, for very small risk increase
- CDF < 10-6
- LERF < 10-7
- Corrected debris loading in LAR does not change conclusions 25
Staff Review of Systematic Risk Assessment
- Acceptability of base PRA Model
- Scope, level of detail, technical adequacy
- Acceptability of approach to focus on ISI welds in un-isolable portion of Class-I pressure boundary piping and SSBIs
- Acceptability of Vogtles focus on high likelihood configurations of ECCS pumps:
- Accidents that need recirc through ECCS strainers
- High likelihood equipment configurations
- Low likelihood configuration can be bounded
- Acceptability of using CFP from NARWHAL for breaks that impact strainer (CFP = 0, for breaks that didnt fail the strainer) 26
Staff Review of Systematic Risk Assessment
- Vogtles disposition of key assumptions and sources of uncertainty
- Geometric and arithmetic mean aggregation schemes
- LOCA frequency allocation (top down and hybrid)
- Partial (continuum) and complete (DEGB) break consideration
- Discretization of LBLOCA frequency
- NRC Staff Confirmatory Calculations
- Minimum transition break size assumed to be 12 inches
- Conservatively assumed failure for breaks larger than minimum break (CFP = 1)
- Used Vogtles high likelihood configuration
- Results do not challenge Region III of the RG 1.174 27
Summary of Systematic Risk Assessment
- Results fall in Region III acceptance guidelines in RG 1.174, for very small risk increase
- CDF < 10-6
- LERF < 10-7 Maximum of staffs confirmatory calculations Vogtles base assessment 28
Staff Review of Systematic Risk Assessment CONCLUSION:
- The licensee used a PRA of the appropriate scope, level of detail, and technical elements and plant representation.
- The risk-informed approach used by the licensee to address the effects of debris on LTCC is acceptable.
- Alternative assumptions were considered as sensitivities for each key assumption employing non-consensus approaches.
- The increase in risk meets the risk acceptance guidelines in RG 1.174, Revision 2.
- Principle 4 of integrated risk-informed decision-making is addressed.
29
Principle 5
- Performance monitoring addressed in LAR review
- Vogtle identified existing programs and processes
- Ensures key inputs and assumptions remain valid
- Ensures corrective action, if necessary
- Staff review determined Principle 5 of monitoring using performance measurement strategies is addressed.
30
Review of LAR
- Evaluated TS changes
- Reviewed FSAR changes
- Evaluated exemption request
- Evaluated each L&C from the staff evaluation of the licensees plant-specific Technical Report 31
Overall Summary
- Vogtle acceptably evaluated the impact of debris.
- Vogtle appropriately considered both risk and deterministic aspects in the submittal.
- Most break scenarios are addressed using conservative deterministic methods.
- Vogtles LTCC evaluation method and simulations are conservative and meet acceptance criteria.
- Vogtles debris analyses meet the key principles of risk-informed regulation.
- Vogtles PRA results show that the change in risk is very small.
32
Questions?
33
Primary References
- (1) ML18193B162 - SNC Technical Report Submittal dated July 10, 2018.
- (2) ML19120A469 - NRC Staff Evaluation of the Technical Report dated September 30, 2019.
- ML21068A109 - Amendment Package for Vogtle GSI-191.
- ML21071A050 - Exemption Package for Vogtle GSI-191.
34
5 Principles - RG 1.174
- (1) The proposed licensing basis change meets the current regulations unless it is explicitly related to a requested exemption (i.e., a specific exemption under 10 CFR 50.12).
- (2) The proposed licensing basis change is consistent with the defense-in-depth philosophy.
- (3) The proposed licensing basis change maintains sufficient safety margins.
35
5 Principles - RG 1.174
- (4) When proposed licensing basis changes result in an increase in risk, the increases should be small and consistent with the intent of the Commissions policy statement on safety goals for the operations of nuclear power plants.
- (5) The impact of the proposed licensing basis change should be monitored using performance measurement strategies.
36
Limitations & Conditions
- (1) The applicability of the NRC's acceptance is limited to the structures, systems, and components; plant configurations; and operations described in Enclosures 2, 3, and 4 of SNC's letter dated July 10, 2018 and the strainer design described in the Section entitled, "16-Disk ECCS Suction Strainer Summary," of Enclosure 2.
37
Limitations & Conditions
- (2) The applicability of the NRC's acceptance is limited to the Vogtle assessment of risk attributable to debris described in Enclosures 1 and 3 of SNC's letter dated July 10, 2018.
- (3) Describe in-vessel analysis, establish in-vessel acceptance criteria, and demonstrate the criteria are met.
38
Limitations & Conditions
- (4) Address Key Principle 1 (i.e., the proposed licensing basis change meets the current regulations unless it is explicitly related to a requested exemption) and Key Principle 5 (i.e.,
the impact of the proposed licensing basis change should be monitored using performance measurement strategies) in RG 1.17 4, Revision 3.
39
Limitations & Conditions
- (5) Identify key elements of the risk-informed analysis (e.g., methods, approaches, and data) that will be described in the Vogtle UFSAR.
- (6) Identify key elements of the risk-informed analysis and corresponding methods, approaches, and data that, if changed, would constitute a departure from the method used in the safety analysis as defined by 10 CFR 50.59.
40
Limitations & Conditions
- (7) Identify the relevant elements of the risk-informed assessment that may need to be periodically updated. The licensee must describe the program or controls that will be used to ensure relevant elements of the risk-informed assessment are periodically updated.
41
Limitations & Conditions
- (8) Describe a reporting and corrective action strategy for addressing situations in which an update to the risk-informed assessment reveals that the acceptance guidelines described in Section 2.4 of RG 1.17 4, Revision 3, have been exceeded.
42
Limitations & Conditions
- (9) Correct the error concerning the evaluation of transported coatings debris loads described in SNC's letter dated December 4, 2018. Specifically, provide corrected coating debris volumes and describe how coating debris loads on the strainers are determined.
43
Limitations & Conditions
- (9a) Verify that the use of the corrected coating debris volumes has a limited impact on strainer head loss and the head loss is acceptable. Also, the licensee must describe the method of verification.
44
Limitations & Conditions
- (9b) Verify that the use of the corrected coating debris volumes has a limited impact on CDF and does not result in exceeding the acceptance guidelines for very small change in risk, as described in Section 2.4 of RG 1.17 4, Revision 3. Also, the licensee must describe the method of verification.
45
Limitations & Conditions
- Modify RHR strainer height and maintain plant configuration consistent with the Technical Report
- Maintain evaluation consistent with TR
- Evaluate in-vessel effects
- Address Key Principle 1 - change meets regulations
- Address Key Principle 5 - monitoring of impact of change
- Define key elements of the analysis in the FSAR
- Define elements of the analysis that would require NRC review prior to change
- Approach to ensure periodic update of risk-informed assessment
- Reporting and corrective action strategy in case the acceptance criteria of RG 1.174 are exceeded
- Correct coatings transport error and evaluate effect on risk 46
Limitations and Conditions LAR RESOLUTION OF LIMITATION AND CONDITIONS:
- L&C 4b: Performance Monitoring
- Staff found Vogtle identified a set of existing programs and procedures to monitor the performance of the sump strainers in containment and assess the impact on the risk-informed assessment
- L&C 8: Condition Reporting and Corrective Action
- Staff found Vogtles proposed to use appropriate plant programs and procedures for its corrective action strategy
- Staff found Vogtles reporting strategy is consistent with the existing requirements in 10 CFR 50.72 and 10 CFR 50.73, as applicable
- L&C 9b: Debris Loading Error Results
- Staff found CDF and LERF increased slightly but remained well within the RG 1.174 range for very small changes (Region III) 47
Plant LOCA Response Spray nozzles Fuel Assemblies Tank RWST Heat exchanger Valve Strainers CSS SIS Sump Safety Injection Containment Spray System (CSS)
(SIS) 48