ML21225A705

Transcript for T11

ML21225A705
Person / Time
Issue date:	03/09/2021
From:	Office of Nuclear Reactor Regulation
To:
References
NRC-1420
Download: ML21225A705 (63)

Text

Official Transcript of Proceedings NUCLEAR REGULATORY COMMISSION

Title:

33rd Regulatory Information Conference Technical Session - T11 Docket Number: (n/a)

Location: teleconference Date: Tuesday, March 9, 2021 Work Order No.: NRC-1420 Pages 1-57 NEAL R. GROSS AND CO., INC.

Court Reporters and Transcribers 1323 Rhode Island Avenue, N.W.

Washington, D.C. 20005 (202) 234-4433

1 UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION

+ + + + +

33RD REGULATORY INFORMATION CONFERENCE (RIC)

+ + + + +

TECHNICAL SESSION - T11 EVOLUTION OF EXTERNAL HAZARD RISKS FOR U.S. PLANTS

+ + + + +

TUESDAY, MARCH 9, 2021

+ + + + +

The RIC session convened via Video Teleconference, at 1:30 p.m. EST, Stacey Rosenberg, Chief, PRA Licensing Branch C, presiding.

PRESENT:

STACEY ROSENBERG, Chief, PRA Licensing Branch C, Division of Risk Assessment, NRR/NRC ALISSA NEUHAUSEN, Reliability and Risk Analyst, PRA Licensing Branch C, Division of Risk Assessment, NRR/NRC MILTON VALENTIN, Reliability and Risk Analyst, PRA Licensing Branch C, Division of Risk Assessment, NRR/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

2 ROBERT RISHEL, Director, Nuclear Energy, PRA, Duke Energy MARC LEVITAN, Lead Research Engineer, National Windstorm Impact Reduction Program, National Institute of Standards and Technology 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 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 P R O C E E D I N G S 1:30 p.m.

MS. ROSENBERG: Okay. All right, so --

good afternoon. Good afternoon, everyone. Welcome to Session T11, Evolution of External Hazard Risks for U.S. Plants. I hope you all can hear me. I am so glad you're here today and I am happy to be here with you. We have a distinguished panel of speakers who are going to be -- to share some of the latest information on these topics. My name is Stacey Rosenberg and I will be the chair for this session.

A little -- just a little background about me. I am a branch chief of the -- one of the PRA Licensing branches in the Division of Risk Assessment in the NRC's Office of Nuclear Reactor Regulation. And I've served as a branch chief in this division for over five years. I've been with the NRC for over 25 years and have served in many different divisions and different offices prior to my present position.

Our panelists are going to present the most current information on the following topics.

Can we have the next slide, please? Thank you.

Alissa Neuhausen from the NRC will share interesting 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 insights about external hazard risks for advanced light-water reactor reviews. And we have Milton Valentin, also from the NRC, who will talk about seismic and external flood hazard risk insights from post-Fukushima evaluations.

We have Bob Rishel from Duke Energy who will talk about insights and lessons learned from external hazard risk models at Duke Energy Nuclear Sites. And finally we have Marc Levitan from the National Institute of Standards and Technology, NIST, who will discuss engineering design using New Tornado Risk Maps. So during this session please remember that we will be capturing questions via the chat.

And if you have a question, you can submit it via the chat at any time. You don't need to wait until the end. We will be discussing these questions at the end of the session and we look forward to having a really interesting discussion at that time.

So now I would -- I would like to welcome our first speaker, Alissa Neuhausen. Ms. Neuhausen is a Reliability and Risk Analyst at the Nuclear Regulatory Commission. She joined the NRC in 2014 as a structural engineer in the Office of New Reactors. In her current position, Ms. Neuhausen 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 reviews external hazard risk analysis, including PRA submittals, related to certification, licensing, and

-- and risk-informed applications. Ms. Neuhausen received a Bachelor in Civil Engineering from the University of California, Berkeley. And a Master's in Structural Engineering and a Master's in Public Affairs from the University of Texas at Austin. Ms.

Neuhausen is a registered professional engineer in the District of Columbia. Alissa, welcome. The stage is yours.

(Pause.)

MS. NEUHAUSEN: Thank you, Stacy. And good afternoon, everybody. I am happy to present today to talk about external hazard risk analysis for advanced light-water reactor reviews. My intent is to walk through the NRC staff's experience and discuss some successes, challenges, and lessons learned. Since this title can be interpreted in multiple ways, I am going to break it down a little bit. For the purpose of this presentation, advanced light-water reactors will refer to light-water reactor designs that have come to the NRC for licensing under Part 52. These reactors include the advanced boiling-water reactor, economic simplified 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 boiling water reactor, AP100, APR1400 and NuScale designs. Examples and experience to be discussed come from the NRC staff's more recent experience.

For the remainder of this presentation I will refer to these reactors as new reactors. Non-light-water reactors are not a part of this presentation.

External hazard risk analysis refers to insights derived from both probabilistic risk analysis or other types of risk analysis. And those successes, challenges, and lessons learned are derived from the U.S. experience. Next slide, please.

The takeaways that I hope to leave you with today are an understanding of the U.S. NRC's external hazard risk analysis with a focus on the design certification and combined license reviews for new reactors, and broad observations, insights, and lessons learned from these reviews. Next slide, please.

This is an -- this is an analogy that I like to use to think about the progression of a PRA.

The PRA can be thought of as an in-progress puzzle with different PRA hazard groups described as various sections of the puzzle. Each piece must fit into its 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 section, as well as the puzzle as a whole.

Especially at the design and combined license phases, each section of the puzzle has a varied level of completeness based on the available information.

Traditional PRA analysis is performed for some aspects of that puzzle. For example, most internal events. Each puzzle section can then be broken down into smaller parts and pieces, some of which are well defined during the design such as initiating events, accident sequences, and success criteria. Other pieces, like operating data, can't be placed until operating experience is gained.

One section, and the focus of this presentation, is external events. Within external events, roughly represented by the pink section of the graph that's shown, there are individual hazard types. For instance, seismic hazards and external flood hazards. For those hazards that require a risk evaluation, common information that is typically unavailable at the design stage, is represented by the floating pieces. Site-specific information and lockdown information, for example, are generally unavailable.

This information becomes available at 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 later stages of design, construction, or operation and allows the puzzle, or PRA, to be completed. The partially completed puzzle also demonstrates the big picture regarding the successes and challenges that I will cover in later slides. The successes are usually derived from the sections of the puzzle that are complete -- that is, incorporating PRA earlier in the design helps reduce vulnerabilities, and early availability of the PRA results allows staff to better plan and focus resources on the most risk-significant items. On the other hand, the challenges are related to interpreting results that rely on assumptions, or that have greater uncertainty. Next slide, please.

The most common external hazards reviewed for new reactors include external floods, high winds, seismic hazards, and other external hazards. These are consistent with the table of contents in the currently endorsed PRA standard. Other hazards may be considered at the COL stage, such as pipeline accidents; or screened out, such as volcano hazards for most sites. Next slide, please.

This slide covers the regulations --

specifically, an overview of how the regulations 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 change as the design evolves. As the plan progresses from the design certification application to the combined license application, to fuel load, and then operation -- the relevant regulations governing the PRA also evolve. For the DCA and COLA, the regulations listed state that a final safety analysis report must include a description of the design-specific PRA and its results. For the DCA applicants are generally performed on PRA-based, seismic margins analysis, in selected bounding site parameters for other external hazards. For the COLA, assumptions should be verified and site features, such as dams and any exceedances of site parameters included in the DCA are expected to be incorporated in the PRA.

For fuel load, no later than the scheduled date or for initial loading of the fuel, a COL holder must develop a level one and level two PRA. The PRA must cover those initiating events and moments for which NRC-endorsed consensus standards on PRA exist one year prior to the scheduled date for initial loading of fuel. At this point, a seismic PRA is required by the regulations.

For an operating plant, each holder of a COL must maintain and upgrade the PRA. The PRA must 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 be upgraded every four years. Also, for a renewed license, the PRA must be upgraded to cover all components and initiating events. Further, based on the current commission policy statement, the new Part 50 plan is also expected to have a PRA and to utilize risk insights. Next slide, please.

My experience, and the bulk of NRC staff's experience to date, is that the design certification and combined license stages. At the DCA stage, the application is generic. It is expected to be used at multiple sites or for multiple units. The lack of site information is especially important to external hazards, whereas lack of operating experience is important for internal and external hazards.

For seismic hazards -- because site seismic hazard curves are unavailable, the PRA-based seismic margins assessment is performed. Aspects of the plant layout, such as cable routing, are unavailable and assumptions are made and documented in the design certification document. At the COLA stage, these assumptions must be verified. The PRA-based SMA includes site and plant-specific updates.

As-built information is still unavailable since 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

12 plant is not yet built, and some plant -- plant operating experience for new designs may be unavailable. To the extent possible, data from operating units is reviewed for applicability for a new design, and this is included in the DCD. Next slide.

On this slide what's important is that there's overlap. The graphic illustrates common assumptions that apply to most or all external hazards for the same application, as represented by the central blue oval. These are items, like bounding site parameters and site interface requirements. Conservative assumptions and simplifications are made. Lockdowns are required before fuel load.

Individual hazards also have hazard-specific assumptions that may be common between different applications, and provide valuable risk insights. For example, it may be reasonably assumed the components inside robust structures are protected from high winds. Next slide, please.

The information on this slide is the most important to the staff review of the DCA. It provides at a high level what the PRA can achieve, 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 and what types of conclusions NRC staff can draw from the PRA. At the design certification stage, staff reviews the PRA description and results. The full PRA is generally available for audit. The review focuses on risk insights and vulnerabilities, input to operational programs, commission goals, and subsidiary goals. The staff does not certify PRA numbers. The staff ensures that an acceptable PRA was performed which identified appropriate risk insights and vulnerabilities, provided appropriate inputs, and that -- depending on the assumptions made

-- the quantitative results meet or show margin to the commissions goals and subsidiary goals.

Quantitative results are used for future risk-informed decision making and must be demonstrated to be applicable for a particular application. Next slide, please. Successful reviews of external hazard risk assessments were completed for several designs. COL applicants confirmed that assumptions in the DC PRA were applicable for selected sites and incorporated site-specific information. For DCAs and COLAs, PRA-based seismic margin analysis was used successfully to determine risk insights and vulnerabilities. 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

14 risk estimates are lower for new reactors than for operating reactors -- consistent with the commission policy statement expectation. Next slide, please.

The first observation is really important for understanding the risk profile. As internal events risk is reduced, external hazards contribute more to the risk profile. Considering PRA earlier in the design process has led to decreased internal events risk, and an overall reduction in total risk.

The scope of information left to COL applicants varies among DCs. For example, in seismic fragility evaluations, for applicants that use more generic fragilities, it is expected that a COL applicant would need to perform additional work to confirm the validity of assumptions that were used as input to the evaluation. Early access to PRA results allows the NRC to better focus resources on risk-significant SSCs during the review. Next slide, please.

The challenges are a reflection of the maturity of the design of the DC stage. As I noted earlier, the NRC's experience has been with designs prior to fuel load, mostly at the DC stage.

Therefore, identifying assumptions is critical for determining the appropriate risk insights. Also 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

15 DCAs, practical experience has shown that it's important to understand and evaluate uncertainties, especially for new and novel design features. The confidence staff has in the quantitative analysis results improves with additional information for both internal and external events.

Risk-informed applications that rely on external hazard results are challenging. And only internal -- only international operating data is available at this time. Last slide, please.

Lessons learned for plans under construction that will be especially helpful are looking at evolutions in the PRA between the design stage and the fuel load; and for the seismic hazard, comparing the risk insights from a PRA-based seismic margins assessment performed during design, and from the fuel load seismic PRA. External events are comparably more significant for new reactors. This is due to lower internal events risk for new reactors, and is a result of using PRA insights and results early in the design.

Alternate criteria for importance measures have been used to assess the risk significance of structures, systems and components.

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 Thus far, staff does a review of this criteria on a case by case basis, and has shown flexibility and adaptability in performing these reviews. The seismic risk may limit the lower-bound CDF. As other risks are designed away, seismic ends up dominating the risk profile. Thank you everyone for listening.

I will turn this presentation back over to Stacey.

(Pause.)

MS. ROSENBERG: Okay. All right.

Okay, thank you Alissa. Our next speaker is Mr.

Milton Valentin who is a Reliability and Risk Analyst for the Office of Nuclear Reactor Regulation. Mr.

Valentin is the Agency's point of contact for the seismic reviews in response to the request for information associated with the Fukushima Near Term Task Force Recommendations. Mr. Valentin completed reviews of external flooding evaluations submittals, and the evaluation of flex programs against new evaluated hazards. Before that, he served as a structural engineer with the Office of Nuclear Regulatory Research, and the Office of New Reactors

-- has also served as a consultant for engineering services.

Mr. Valentin received a Master's 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

17 Engineering from the University of Maryland and a Bachelor's in Civil Engineering from the University of Puerto Rico, and is a registered professional engineer in the State of Maryland. Welcome, Milton.

The platform is yours.

MR. VALENTIN: Thank you, Stacey and good afternoon everyone. My name is Milton Valentin. I am a Risk and Reliability Analyst for the NRC and today I will be talking about the NRC efforts to consider seismic and external flood hazard risk insights. Next slide, please.

Since the working PRA started, the Agency has documented ways to assess external hazards for nuclear power plants, and this timeline includes only some examples of those activities. Among the most important they developed an examination of external events, or IPEEE, enhanced the understanding of severe accident sequences, and help identify the plan modifications that could prevent those accident sequences.

Most recently, the work done here in the U.S. following the Fukushima accident enabled the staff to assess the need for regulatory action to ensure protections against those reevaluated hazards.

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 The same information was also used to close some of the generic issues. For examples, we closed a GI 199, generic issue on probabilistic seismic hazard estimates in Central and Eastern U.S., and GI 204 for flooding of nuclear power plant sites following upstream dam failure. Next slide, please.

How did we do it? In all our reviews we prioritize safety. We use the latest methods and information available, like the recent versions of the SRP and other recent guidance. The staff consider operational experience and all information available to improve the state of knowledge. We also work together with the industry to endorse guidance for consistency during these reviews. What did we do with all that information? Well, we confirmed safety for the licensing basis events which are those hazards that nuclear power plants are licensed for.

For those sites where the new hazards exceeded the licensing bases, we worked with the licensees to enhance protection and mitigation capabilities. And I am going to describe that in my next slides. Also, we improved the state of knowledge across the board thanks to all the information that was gathered. Next 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

19 Here we start talking about external flooding. And in -- in this figure with all the bubbles, we are trying to explain the percentage of sites that address flooding mechanisms. For example, starting at the top, out of the 61 sites that had to address external flooding hazards, 89 percent had to address local intense precipitation. This is the highest number because most licensees didn't have to consider local intense precipitation for design. To go counter-clockwise, there is a 30 percent under storm surge and that means that 30 percent of those sites have to consider storm surge and so on.

Some of the flooding mechanisms were more complex to assess than others. But all licensees were able to assess the likelihood of -- and potential effects of these hazards by following the guidance in NEI 16-05, in particular Appendix B to NEI 16-05 --

which is a guidance for the nuclear industry to do -

- it provides guidance for estimating the likelihood of flooding events in terms of annual exceedance probability, which would be used to dictate the course of action for protection or mitigation of these effects. And there's some figures there to make the difference between the actions 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

20 protection and mitigation. Next slide, please.

This is my favorite slide. We see two arrows -- one pointing down, the other one pointing up. We're going to start talking about the events represented by the arrow pointing down. These are the flood events that have high likelihood. And those would be the ones that we saw in lower water elevation. For sites that had those high likelihood events, licensees would rely on passive site protection. And those would be either the site or floor elevation, drainage, and all those basic protections to maintain safety. The blue line is intended to represent the consequential flood height.

The event accident for probability for this consequential flood height would be in the order of one in every thousand years -- rough margin -- our one in -- once every 10,000 years. And those are the same figures that I -- I had in my previous slide.

For the arrow pointing up, which would represent the low likelihood event -- the -- that can produce the high-water elevation, licensees would rely on mitigating strategies developed for those specific areas. Sites that could have those rare events already have procedures and equipment in place 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 to deploy these strategies if needed. I would like to mention also that, nowadays, weather forecasts allow sites to prepare for these events on time.

The Fukushima flooding evaluations produced a big number of enhancements. For example, there were enhancements on plant response and mitigation procedures. There were new flood barriers installed. There were improved flood barriers installed. There were improved drainage, enhancement on existing protection features. But overall, we improved our understanding for anticipation -- for anticipation and response. Just like done for the IPEEEs, the recent flood reevaluation produced most of the safety enhancements out of all of the Fukushima activities. Next slide, please.

Now we're talking seismic. In this graph we're showing the contribution to core damage frequency from a number of sources -- as explained in the legend -- green represents seismic, red represents fire, yellow is internal events, and blue is intended to show other -- other events, other sources -- and it's mostly wind. But the takeaway from this slide is that seismic risk can be non-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 trivial. There is a lot of green in this graph and all over the board for some sites. But because of the situation, many sites proposed plant modifications and performed detailed evaluations to provide assurance that the reevaluated seismic hazard would not compromise their ability to maintain safety. Next slide.

Here is some data from the seismic, probabilistic risk assessment reports. And in this slide, we're showing frequent risk contributors based on appearance. This is not based on actual risk contribution. The actual risk contribution would look very different and it would vary for each site.

The SPRAs provided the risk-significance for each contributor. So based on that significance, the licensees would assess potential modifications that could reduce risk. When reviewing these reports, the NRC staff performed an independent screening evaluation to assess the need for further regulatory action. However, thanks to the solutions proposed by the licensees, the staff did not pursue additional regulatory actions.

One observation is that SPRA results also validated findings from the IPEEE. For example, 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

23 saw that off-site power and on-site electrical equipment were once again dominant risk contributors.

Also, building and structural failures were once again an important contributor. Next slide.

Here -- next slide, please. Thank you.

Here are some examples of plant modifications proposed by sites and their associated risk reduction in terms of delta seismic core damage frequency and seismic large early release frequency. We take one example -- in the middle of the slide -- one licensee provided alternate power to their hydrogen ignition system and that modification has the capability to reduce the seismic large early release frequency in half.

So in summary, we understand that seismic PRAs enable licensees to identify additional enhancements over IPEEE and to reduce the associated risk at certain sites. The risk contribution is something very site specific, and SPRAs enabled licensees to act on these potential vulnerabilities.

I'd like to mention also that other safety enhancements came from interim evaluations completed before the FPRAs. Interim evaluations were completed to provide assurance that sites could cope 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

24 beyond-design basis earthquakes while the SPRA reports were completed. Next slide.

So in conclusion, we understand that U.S.

nuclear plants are better prepared against external hazards. We have a better understanding of the hazards, and our tools to assess the risks are much better. The plant protection and mitigation are much improved, and they are more commensurate to our understanding of risk. Our understanding of the current design basis robustness for all side has also improved tremendously. We're a learning organization and we're transforming the remaining ongoing activities to further enhance the state of knowledge. For example, one of the activities is the process for ongoing assessment of natural hazard information. We also -- just last month -- we have the probabilistic flooding hazard assessment workshops. And those were great. I would like to give a big shout-out to Tom Aird, one of our session coordinators in the Office of Research for hosting those. They were excellent. And we will continue to have periodic guidance and tool updates because we will continue to use these tools.

So in summary, we look forward 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

25 continue enhancing the state of knowledge to maintain safety. And next slide is -- next slide is just acronyms because I used too many during my presentation. So this is for your reference. Thank you for your time. And I will turn it back to Stacey.

MS. ROSENBERG: Thank you, Milton. Our next speaker is Mr. Bob Rishel, who is the Director of Nuclear Engineering PRA for Duke Energy, a position held since 2012. Mr. Rishel is also the chairperson of the Boiling Water Reactor Owners Group Committee for Integrated Risk-Informed Regulation.

His PRA staff support all risk-informed applications for all Duke Energy nuclear plants, contributing and developing almost all tools for both internal and external event PRAs. Prior to his current position, he held various positions with First Energy, including nine years as a senior reactor operator in the Shift Manager position. Following college, Mr.

Rishel was commissioned in the U.S. Navy where he served for nine years on nuclear -- U.S. nuclear submarines, as well as on the staff of the Commander Submarines U.S. Atlantic Fleet.

Mr. Rishel graduated from the University of Wisconsin with a Bachelor's Degree in Nuclear 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 Engineering. Please join me in welcoming Bob. Bob, the platform is yours.

MR. RISHEL: Thank you, Stacey. Good afternoon, everyone. You know, I'd like to start with some overall conclusions first. So as indicated by Milton, in looking to results and processes, it was clear that the U.S. fleet, as built, is a robust and capable withstanding events well beyond their design basis. The nuclear numerical values we calculate is reflection of the state of knowledge and practice. Costs and the capability to increase that state of knowledge is the limit. Simplification of complex scenarios also contributes to those values.

And additionally, there's a definite tilt towards worst case when there is uncertainty. Next slide, please.

Keep going, next slide. So I'd like to start off with fire -- as indicated. And typically fire is one of the largest contributors to plant risk.

Realism has been improving over the years.

Cooperation with the NRC Research and EPRI. It's used in almost every risk application question about fire needs to be answered. And the fire PRAs provide a method to answer those questions.

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 Fire PRAs do contain some known conservatism -- and I listed those on my slide here.

The point I'd like to make is that the more you use it, the more you learn. And you find limiting scenarios, we investigate them, do more detailed analysis, and it pushes us into even more realism.

Next slide, please.

Seismic PRAs talked about -- we -- Duke Energy started and completed some seismic PRAs for the Fukushima near-term Task Force requirements. It does provide insights into the design features that were not fully appreciated before we took this effort. Some of those insights were good, and some provided areas for improvement. These insights did allow cost-effective modifications, as Milton talked about, that can significantly reduce the seismic risks. Most of these modifications were not expensive, which goes back to the robustness of the designs.

Current applications for Duke include use of 50.69 and using seismic for the risk-informed tech spec completion times, which I believe using the seismic allow wider scope of the SSCs to be considered for these applications while avoiding a potential 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

28 necessary large seismic penalty without the seismic PRA. Next slide, please.

Duke Energy developed a number of high wind PRAs. High wind, hurricane, missile -- in that effort, a number of plant design features were identified which, when credited, would reduce the overall risk of high winds and missiles. We also identified a number of areas where further analysis could move the results closer to realism. In some cases those analysis were more complex than we expected. But we do have them.

We found that when applying the high wind PRA, for most of our sites, the consequences of those high wind events are really enveloped by the internal events loss -- wind-induced loss of off-site power.

And I -- Milton also indicated to that. As a result, we have worked at screening out high winds when we can meet the requirements of the PRA standard for screening for risk applications. We have applied the PRA -- high wind PRA for some applications when the high wind provides particular insights or need, such as a changing in design basis for the high wind tornado missile. The other place -- other uses when the -- validates the tornado missile risk evaluation 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

-- TMRE process -- Duke used their high wind PRA.

The development of these high wind models has allowed us better understanding of the high wind issues, vulnerabilities, which in turn have helped us focus more on plant modifications that add value and eliminate some of the proposed mods that were less valuable. Next slide, please.

External flooding -- external flood in the PRAs has the largest uncertainty. We've been able to screen out external flooding from most of our risk applications. But it's been our experience in some cases it's more cost effective to increase the robustness of the plant design via modifications for the worst case, rather than going under the expense of developing a external flooding model. External flooding models, we have found, have -- can be very expensive efforts and developing them may not provide any cost benefit.

External flooding PRAs do allow a risk-informed process to do things like inspection or barriers and doors and seals, we also intend to use it for 50.69 screening and the PRA standard will not allow us to screen a hazard out. We have used --

I'll say many external flood PRAs for our efforts 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

30 significant termination process when a flood design issue has been found to be compromised. The external flooding PRAs talk about the uncertainties is a return frequency in the hazard curve. Developing those curves is difficult and limited data -- we have 100 years of flood data and we're trying to extrapolate that out to once in a million.

Some sites may have -- do have unique issues where a flooding PRA (audio interference) to evaluate the risks and identify the solutions. The Fukushima Response Evaluation -- talked about earlier with Milton -- does provide some assistance in external flooding PRA. But the results are extreme and may not provide a path or support for the development of the hazard curve, which is -- you have to know the initiating frequency of the -- the PRA.

Next slide, please.

Talked about lessons learned here. So if you can screen the hazard using the PRA standard for your risk submittals, you know, that -- that's the best approach. The second best approach is if you can increase your plant design robustness and remove the hazard that -- that may be more beneficial.

But -- and I would add that the new PRA standard as 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 currently written is going to make screening external hazards a bit more challenging.

It would seem that it would be more beneficial for a simpler external hazard PRA model development process to be developed in the current standard, going through capability category 2. This

-- the simplified process could be used to identify issues and correct them, avoiding the costly capability category 2 effort. Examination of -- so in our effort, especially with external events, there are hidden assumptions that we need to explore and find out. Anything to drive conservatism. And these assumptions sometimes are not something that's the typical PR practitioner would understand. They're frequently caused by the individuals doing the hazard analysis who sometimes do not fully embrace the best estimate approach and tend to fail conservative direction when faced with some uncertainty.

If you have these assumptions and uncertainties or issues, say -- don't be afraid to go get a second opinion. It may cost more money, it may involve more analysis, but it could be worth it --

especially if it is driving new results. This is --

can cause something, what I call dueling PhDs. 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

32 that's something that -- we have to deal with when it occurs. But it could be worth it. Next slide, please.

So I've already -- already discussed the difficulty with developing hazard curves with a large bias toward conservatism. In high wind PRAs, the largest risk, typically, is in the lower-end wind speeds -- less than 110 miles per hour. With more -

- with adding more bins, you get more resolution and that can improve your results to be more realistic.

With high wind you do need to consider concurrent events and the impact. And you know, the good example is rain with hurricanes. You don't get rain

-- hurricanes without rain. Next slide, please.

Seismic PRAs, like high wind, bends at the lower end. And I'll say between 0.2 and 0.5 G could help you improve your realism and -- and refine where you may have issues. I'd say also don't --

look outside the nuclear industry for seismic events.

And EPRI has been doing a good job of this -- looking at refineries and other similarly designed structures for the seismic response.

External flooding, I guess, one thing there is, you know, cliff edge effects. Make sure 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 you're aware of them. Look for them. And the other part is, again, Milton -- you know, my straight man there -- talked about preparation, warning time, weather. So when you're doing your external flooding operator action unit reliability analysis, that gives you time for preparation. It gives time for the sites to do oversight of the preparations, provide an assurance that the site is prepared for the incoming weather. Next slide, please.

Some final thoughts going forward. So the Fire PRA and realisms improving, I think the industry is on the right track. External PRA standard, as written, is -- is a large investment in dollars to get the capability Category 2, and being able to get your investment on that effort is a challenge. A simplification of the PRA standard should be considered for risk applications where most of the safety benefit and actions can be identified very early on in the process. And after that, I will take any questions.

(Pause.)

MS. ROSENBERG: Okay, thank you so much Bob. Now our final speaker is Dr. Marc Levitan. Dr.

Levitan is a lead research engineer for the National 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 Windstorm Impact Reduction Program at the National Institutes of Standards and Technology. He serves as -- he served as the lead investigator of the National Construction Safety Team Technical investigation of the 2011 EF-5 tornado in Joplin, Missouri; and the NIST study of the 2013 EF-5 tornado in Moore, Oklahoma which has led to improved tornado hazard characterizations and improved performance of buildings and shelters.

Dr. Levitan chairs several tornado-related standards committees for the American Society of Civil Engineers, and the International Code Council Committee. Dr. Levitan has his Bachelor's, Master's, and Ph.D. in Civil Engineering from Texas Tech University. Welcome, Marc. The platform is yours.

DR. LEVITAN: Thank you, Stacey. Glad to be here back at the RIC presentation. And one of the comments that Bob had on his slide, about the challenges of doing high-wing PRAs, particularly for tornados, and -- so hopefully we have some information here that's going to help with -- with some of that. So let me go ahead and acknowledge my co-authors. Dr. Larry Twisdale from Applied Research 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 Associates, who has developed an awful lot of the tornado research over the years that's used by the nuclear industry. And my colleague at NIST, Dr. Long Phan. Next slide, please.

Okay, so I am just going to go over a few slides on the background -- an introduction here.

I'll talk briefly about the summary of the map --

tornado map development methodology. And I'll introduce you to some of the -- new tornado hazard maps and what they look like and talk briefly about how those maps are being implemented -- the ASCE 7-22 standard. I'll just make a note here in the --

schedule to present last year at the RIC, and the slides from that presentation -- which have a lot more detail about the development methodology -- are available on the RIC website from last year. I will note that the hazard maps -- those were draft maps.

They've since been finalized. And those are -- so the maps have been updated. But that -- that previous presentation has more information and background on the methodology. Next slide, please.

So the impetus for this work came from NIST's Technical Investigation of the 2011 Joplin Tornado -- the single deadliest and most costly 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 tornado since the U.S. began keeping official records in 1950. We developed a series of recommendations from that study. It called for improving tornado hazard characterization, proving -- improving how we designed and construct buildings and shelters. Next slide, please.

Recommendation -- the implementation of the recommendation I am going to talk about today is

-- was recommendation number three. Recommend the development of tornado hazard maps for use in engineering and design of buildings and structures be developed considering spatially-based tests in the midst of tornado hazards instead of point-based estimates. And you'll notice that -- we had identified NIST as the lead agency and the series of interested parties, including the Nuclear Regulatory Commission.

Shortly after we completed the report in 2014 we contracted with Applied Research Associates and began what would turn out to be a six-year, or a multi-million-dollar study to develop these tornado hazard maps. Next slide, please.

Some of the background indeed -- in terms of what the existing hazard maps were -- 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

37 available at the time, there was some -- some of the challenges and limitations that there was no consideration of tornado reporting limitations and development of these existing maps, there was no treatment of target sizes, the use of judgment-based wind speeds, and the uncertainties were not systematically considered. Next slide, please.

So we worked early on, as soon as we completed our technical report, we engaged directly and regularly with the Nuclear Regulatory Commission as one of our -- our key stakeholders. They participated in a series -- we had a series of workshops for the -- the broader stakeholder community, as shown here on the right. We had also had a workshop specifically for federal agencies.

And at -- we also, as we were developing, about halfway through our process, the NRC provided some supplemental funding to assist us and to be able to handle the -- the consideration of the epistemic uncertainties. Next slide, please.

Our -- this slide presents an overview of the entire project. I could spend a few hours going through all of the -- the details of this. And as I said, the -- last year's presentation had a little 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 bit more of the information. But the major components, looking on the left-hand side -- and the grey boxes -- assessments of tornado climatology.

Looking at old tornado data. Particularly looking at the population bias and the known problems and reported problems is that we don't have as many tornados on the record in areas with lower population.

The pink box in the middle -- we did a lot of work on improving the tornado wind field modeling. At the top -- boxes on the top right, we worked to exclusively model in a better -- to better understand what the wind speeds were. When the tornado is just reported in the database as an EF2 or an F-3, what does that really mean? And so we actually did explicit modeling of the -- damage to houses which drive most of the -- the large EF scale number ratings. We consider the epistemic uncertainties, but certainties combine all this information and put hazard risk models to develop our hazard cures, and ultimately the tornado was wind speed maps. Next slide, please. Next slide please

-- there we go.

So we considered uncertainties in about 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 a dozen different areas, including -- broadly in terms of different tornado properties, in terms of climatology, the wind field, the damage modeling, et cetera. Next slide, please. So ultimately the --

in the top right here, we -- based on our climatology analysis we developed a regionalization scheme where, assuming that the tornadoes in each one of those regions shown has similar characteristics. And for each one of those regions we developed a hazard curve for different target sizes. As you'll see on the right, a point target and an area target is the black

-- small black and the larger black dot. You have a different risk -- you have a different strike probability, and that -- it turns out at the same return period that you have a different wind speed based on your target size. And what was shown right here is the -- the slide for -- is the hazard curves for -- for a point target. Next slide, please.

This slide shows the differences -- the curve based on those differences in target size with a couple of examples. The three red curves on the right are hazard curves for the area -- Region 4-B, so basically the center of the country. Our -- where we have our most intense tornadoes. The three blue 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 curves on the bottom are for the west of the Rockies where we have the lowest. Basically you -- you can see these -- the -- the red and blue arrows show the difference in the wind speeds at different return speeds. Between the farther left value is a point target, and the value at the far right is a very large target -- a 4,000,000 square foot target.

And so we can see this area affects --

this target size area effects vary depending upon return period and depending upon region of the country. But they can be 30, 40, 50 miles per hour difference between what target size that you're looking at. Next slide, please.

So once we have the hazard curves -- go through this slide -- walks through the map development process. So we start with the hazard curves in step one in the flow chart here. Then for

-- we interpolate those hazard curves for the particular return period of interest in step two.

And then from there, we use the values for each of the different regions to populate a grid map -- as shown, like in the bottom left -- and then that grid map, we provide a Gaussian smoothing because of course these are not hard boundaries, right? There's 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 a lot of uncertainty. And there's a transition --

these -- in between these different regions. So in

-- in order to -- and this is just one window into some of the -- how we handled some of the epistemic uncertainties is we went through -- and this was our best estimate of the -- the red lines present -- on the map on the top right -- the red lines present our best estimate of the boundaries between the different climate -- tornado climatology regions. But the black solid and black dotted lines around each one of these red boundary lines are the uncertainty based on different modeling and different assumptions that we came up with. And those typically range -- so those boundaries typically ranged over a few hundred miles.

And we use that uncertainty, then, to inform the size of the Gaussian smoothing window that we passed over the grid cell map to smooth over the wind speeds across these different regions.

Once so you -- just -- the Gaussian smoothing -- then we use kriging to develop contours.

Then we finally ended up -- we smoothed those contours with the peak smoothing method. And finally some --

and cleanups on -- on the maps. Next slide, please.

So we produced maps for a range of return 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 periods. Form 300 years all the way to 10 million years. So we're interested in applications from ASCE-7 for design of conventional buildings, as well as nuclear facility. And we produced maps for a range of -- eight different target sizes. From a geometrical point targets, all the way to about 4 million square feet. We -- in production mode we only looked at square targets. There is some sensitivity to target aspect ratio -- also to target orientation since tornados have preferred paths from sort of southwest to northeast, in general. However, the -- the -- those are -- so if -- those aren't too great -- if the dimensions aren't too far off form being square. We did not include -- as I said, certainly linear targets, like power lines or something, would be very different. These lower return periods were for a -- really for the ASCE-7 risk categories. Next slide, please.

So I'll now show you a sample of some maps. As you saw, we produced actually a large number of maps. But this -- the first three maps I'm going to show you for a 200-foot by 200-foot sized building, or a 40,000-square-foot target area. This first map is for 100,000-year return period. We can 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 see that we have a wind speed in the -- and all these maps have a -- some -- look generally the same, at least at the higher return periods where you have the

-- highest -- greatest tornado wind speeds, of course

-- and in the Midwest. And then it extends off into the -- into the Southeast as well. And then it tapers off as we move to New England -- to the Gulf Coast.

And tapers off much more rapidly as we move west of the Rockies. So we have a peak wind speed of about 180-or-so miles per hour in the middle of the country, here. If we go to the next slide, you'll see the map for the 1,000,000 year return period. Everything moves up to about 20 or 30 miles per hour as we move up a -- by a factor of ten in years. So now we're up to 220 or so miles per hour in the center of the country. But again, with the same general trends.

And if we look at the next slide, you'll see a million

-- a map for a million-year return period. Up to 268 miles per hour in the -- the middle of the country.

If we go to the next series of -- these three slides. The next one -- now we're going to look at a much larger target. So it's of two thousand

-- by 2,000-foot area. Going back to the 100,000,000 years. Now you see we're already up over 200 miles 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 per hour even if the -- the -- 211 miles per hour maximum at the 100,000 years. On the next slide, it's going to -- if we see it for a million years.

Gone up another 30 or 40 miles per hour. And if we go to 10,000,000 -- next slide -- 10,000,000 years.

We see where -- up in the -- in the high 200s now for our maximum -- our maximum wind speeds. Next slide, please.

Okay, so how does this compare with guidance that's already out there? So if we look at a selection of different cities in this table on the bottom right, and the -- the -- the grey -- the three grey-shaded bars are wind speeds for these cities for

-- this is for the 200 by 200 foot building -- so the 40,000-square-foot target area. We have the hundred-thousand, million, and ten-million year in the grey.

We can see for the -- the middle three columns in the beige color are the wind speeds based on the -- the new tornado hazard maps. And the third set of columns on the right are the differences between the new maps minus the values for those -- the same cities in -- in the new red. And as you can see, in most cases there's some increase. Maybe average in an order of 20 to 25 miles per hour. In some cases quite 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 a bit more. And in some cases there's actually small

-- a small decrease. But overall the -- the -- what we found is the -- the tornado hazard and the tornado wind speeds at those return periods for those locations is somewhat larger than what's been --

being designed for in the current NuRegs. Next slide, please.

So if we just sort of summarize the map development there. We've used an engineering modeling process, developed tornado wind speed maps in order to systematically produce speeds for a wide range of return periods. Our goal was the best estimate modeling consistent with ASCE wind hazard maps and nuclear power plant high wind PRA standards.

They used an engineering process for wind speed estimation and believe that the model tornado wind speed hazard can reasonably associate with return periods usable for engineering design. We've attempted to quantify epistemic uncertainties for key variables. Nevertheless, the resulting hazard curves and associated maps have large residual epistemic uncertainties. Next slide, please.

So now the -- the last few slides here, to wrap up, it's how we're implementing these maps 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

46 the ASCE-7 standards. So this building -- this standard is used for designing of conventional buildings. It's a minimum design loads and associate criteria -- for buildings and other structures. And this standard is adopted by reference into the International Building Code and other -- other --

other model building codes to provide the basis for load requirements. And so ASCE 7-16, the current version standard does not currently -- does not address tornado hazards. We have proposed --

included the tornado hazards, including these new maps for the 2022 edition. Next slide, please.

So NIST, working with Applied Research Associates and Tornado Task Committee of the ASCE 7 Committee developed the new tornado load provisions for ASCE 7. The comprehensive set of provisions include these new tornado hazard maps at the return periods consistent with the reliability provided by the non-tornadic wind provisions of ASCE 7. We've done a lot of work to better understand the vertical profile of the horizontal component of tornado winds and have a new profile parameters for that. We've -

- a lot of work to understand the effects of the --

around pressure coefficients, and particularly 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

47 effects of the vertical component of the wind. And that's handled through our modification on the external roof pressure coefficients -- where that --

where the vertical updrafts of the tornado increase the suction on the roof. And we've addressed that through adapting some wind tunnel modeling. And then we also have modified the internal pressure coefficient to also count the effects of atmospheric pressure change. And altogether these have developed procedures then to determine tornado loads to the main wind force resisting system, and the components in cladding. Next slide, please.

This is organized in a brand new Chapter 32 on tornado loads that includes these maps. So again, for Risk Category 3 and 4 building and structures. Tornado loads are not required for Risk Category 1 and 2. So at 300- and 700-year return periods, the -- only the -- the very, very largest sizes of 4-million-square-foot even had any risk at all. They're just -- you're just not -- just not going to get struck by a tornado. Very unlikely to get struck by a tornado at those low return periods.

So we're only requiring it for Risk Category 3 and 4.

But we also have proposed an appendix 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 that has a tornado maps -- hazard maps for long return periods. It has the ten thousand, hundred thousand, one million, and ten million years. And there's eight target sizes for each of those four return periods. And those potentially support the nuclear power industry, as well as performance-based design.

Next slide, please.

In terms of our current status. Right now, within ASCE 7, if you are in the shaded region of the country -- the -- the grey shaded region to -

- on the right. That's the area where we have a --

do have a mean tornado risk at those low return periods. And that's where you have to consider design for the -- what we call the Risk Category 3 and 4 structures. Next slide, please.

So in terms of the status, this is still working its way through the ASCE 7 approval process.

But we anticipate -- it's likely that it will be included in the ASCE 7-22, which will be reducing a public comment draft in the June time period -- for 45-day public comments. So those are available at all -- the community must consider and respond to all the comments through the ASCE 7 Main Committee. Next slide, 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

49 So NIST is preparing several publications. One a major report on the development of the maps themselves. We have another report on the development of the k factors an internal pressure coefficients used in ASCE 7-22. And a third report looking at the probabilistic analysis, and reliability analysis used to determine the appropriate return periods for consistency with ASCE

7. All three of those reports should be coming out from -- published from NIST in the next several months. Next slide, please. And with that, we're wrapped up. Thank you for your attention.

(Pause.)

MS. ROSENBERG: Okay. So, great. Thank you very much, Marc. I think we'd like to go right into our question and answer session. So with that, I'm going to start off with a question for Alissa.

The question is, is EPRI SMA -- which is Seismic Margins Analysis -- methodology acceptable, or just the U.S. NRC-based SMA acceptable?

MS. NEUHAUSEN: Thank you. I'll answer that with respect to new reactors, since that was the topic of my presentation. And I'll point to --

either in the Staff Guidance 20 -- RUSG 20 on 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

50 based margins analysis for new reactors. So we've endorsed that through the -- that ISG. And that --

that's just for new reactors. Our method has been used for different risk applications. Thank you.

MS. ROSENBERG: All right. Thank you, Alissa. Let's go on to our next question. This question is directed to Milton. Milton, you said that the NRC worked with licensees to address the risks for reevaluated hazards. To clarify, were there any cases where the NRC actually required licensees to make any changes?

MR. Valentin: Thank you for that question. The answer -- the short answer is -- it's no. We -- we did evaluate the safety significance of all of the plan modifications and the risk contribution. And we followed the basic guidance and

-- and short -- long story short, we -- that -- those changes didn't meet the backfit mark. So we -- we didn't have the -- the grounds, or the need -- because the licensees proposed solutions that were reasonable and that could mitigate the -- the risks. So we didn't have to. Thank you.

MS. ROSENBERG: Okay, great. Thank you, Milton. I have a question now for Bob Rishel. 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

51 the question is, there has been a significant increase in requirements on external events in the PRA standard. As you suggest, what exactly is needed for simplifying the standard to bring it back to a more rational approach. It seems some of the requirements are focused on expanding the state of the art rather than state of practice, for example.

Bob?

MR. RISHEL: So, yes. That's good --

good question, thank you. I think one of the things from the standards committees that they need is a lack of feedback on what -- what is the state of practice? What's been done and what are the results showing about what's important? And so as -- as we get the early efforts -- let's take seismic PRAs, another -- that's pretty -- that's pretty -- pretty new in the efforts. You go back to the standards and say, okay, here's -- here's what we learned out of this effort. Here's what was important in this standard that contributed to our understanding. And here's what was not important. And so the -- some efforts could be curtailed, or the requirements reduced. It -- and reduce the overall cost.

The number fragility analysis that 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

52 have to do per the standard, you know, could be excessive in -- in -- you know, fragility, in each one, is very expensive effort. And those -- and those individual fragilities drive the cost. That's just one aspect of the cost curves. And -- and we're not getting anything out of that effort really.

There's not a lot of new information that we're learning. You know, in this case it's just to comply with the requirement. So.

MS. ROSENBERG: Okay. All right, great.

Thank you for that. All right, I have a question here for Marc. Marc, with climate change increasing and the intensity -- the intensity and frequency of tornadoes, how often will these maps be updated so that -- this is a two-part question. So that's the first part. Do these maps change with El Nino, La Nina, effect in the Pacific Ocean and West Coast impacts?

DR. LEVITAN: Thank you for the question.

There is a lot of uncertainty within the -- the climate community on the effects of -- of tornadoes.

There's less certainty about that than most any other effect of climate change. When you look at the historical record for -- NOAA database from the 1950s 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 through 2016 when we're doing this work, there's so many problems with the -- challenges with the database in terms of, you know, the first 20 or so years were actually created for reconstruction through newspaper records and things. And so really, only since the mid '90s are we capturing a lot of the tornados. There's still a very large percentage of tornados -- particularly depending up on the region of the country that we don't even get recorded --

that we know don't get recorded in the -- in the database. So there's so many other challenges that we -- it's hard to tell if there's any climate signal in there. In terms of -- no, we don't explicitly consider El Nino, La Nina -- those other sorts of phasing. We did look at some -- some of the trends.

We weren't able to -- said some of these -- some of these other -- other reporting issues kind of washed out any trends with -- related to climate.

There has been some work in the literature showing that maybe in recent years some of the more intense tornados are shifting from the Midwest more to the Southeast. And -- again, with such a short record though -- is that climate change?

Is that part of a multidecadal type oscillation like 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 we see with hurricanes? With 20- or 30-year trend in the North Atlantic, for example.

And so I certainly anticipate that we would be updating these maps as our state of knowledge for tornados increases in future additions. ASCE 7 is on a six-year cycle. And so we may be producing new -- new maps if we have new knowledge at that point for -- on a six-year pace.

MR. ROSENBERG: Okay, great. Very interesting. All right, now I have a question here for Milton. Milton, some plants are, or would be, located downstream from dams. Which can potentially fail due to seismic or flooding events -- or other failure models. Do PRAs take these possibilities into account? If not, why not? If so, how is that done?

MR. VALENTIN: Thank you for that question. I would say that the upstream dam failure is looked at in different ways. In particular, that issue would be addressed by the integrated assessments that some sites had to complete. So the

-- that was looking, too, from -- basically two aspects. We have the integrated assessment. Then also we would -- if -- that was considered 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

55 initiating event for the PRAs as identified in their model, then we would look at it. But long story short, for upstream dam failure, we have the integrated assessments and all the sites that were prone to have that unlikely event take place at their sites, they did address the -- the possibility of having upstream dam failure. And for -- for PRAs, if it was identified as one of the initiating events.

It would be something that was looked at. But I don't recall from the top of my head any of the sites having that as initiating event. Thank you.

MS. ROSENBERG: All right. Thank you, Milton. All right, I have another question here for Bob. All right, Bob. This looks like a two-part question. Could you clarify how you could address a lack of knowledge of flooding event frequencies by increasing plant robustness? And how do you know what degree of robustness is actually needed under those conditions of uncertainty?

MR. RISHEL: So that -- so that's a good question. Thank you for that question. So talked about robustness -- so one of the areas there is --

really is external flooding, storm surges. And so we -- you have to do enough analysis to know how high 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 it's reasonable that the water can get. And -- so it -- so with that knowledge, then you can go on and see if -- see if it's reasonable to increase your flood barrier height, either permanent mods or -- or some temporary mods. So -- so it's -- you can't operate of course in a vacuum. You have to do enough work to figure out what -- how high is -- is reasonable. And -- and then -- and then of course, as you indicated with uncertainty, then you add a bit for -- for margin. And you know, and that's -- and that's a judgment call that's done by the -- with the design engineers and some PRA input about, you know, what -- what do they believe some of the uncertainties are and how far off it could be.

So -- so the answer is, you can't -- you do need to do enough work to figure out what's reasonable things to do.

MS. ROSENBERG: Okay, great. Thank you a lot, Bob. Great help. All right, now I have a multi-part question for Alissa. Alissa, during the IPE -- I'm sorry, during the IPEEE period -- the IP for external events, which is IPEEE. Most of the external events were qualitatively assessed as low-risk significant and excluded from detailed 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 assessment. Now we say external events are more risk significant than internal events. Have we missed something in our earlier risk assessments? Any lessons learned there? Are there current external event carrying model -- modeling methodologies in data? Are they up to the level they should be? Let me know if you need me to repeat any of the question.

MS. NEUHAUSEN: Thank you, Stacey. And I'll try to answer part of that question. It may be

-- some parts of it may be better for other panelist members to take on. The piece I can sort of address is saying that external hazards are more risk significant than internal events. And really what we've been saying is -- is that internal events risk

-- at least for new reactors, is -- is very low because we've been using the PRA earlier in the design process. So it's really a relative comparison to the

-- the profile. That's kind of the piece that I can address. But maybe some other panelists can take --

MR. VALENTIN: I can -- I can add something. At least -- and I agree with what you said, but I would -- I would add that nowadays we have more information and -- and more tools to quantify and consider the effects of external hazards 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 in -- in the PRA. Some of this information, when --

when the PRA started was just not available. The technology was not available. The -- the understanding was not there.

Nowadays we -- we continue to evolve. We are in a much better place to make a better judgment of what could happen if we have -- a stronger earthquake, now as we understand them, or a -- a higher, more severe flood event. And we are more capable of making that judgment today than where we were. And we will continue to do a better work as -

- as we continue to learn of -- of these methods and get more information.

MS. ROSENBERG: Thank you. You know, I guess I would like to add to that. I -- you know, I think for operating reactors, it is true that internal events are usually higher risk sometimes, depending upon the site, than the external events.

But for the new reactor designs, the advanced light-water reactor designs that we've been reviewing, the internal event risk has been a lot lower. And so the external event risk looks -- is higher in comparison.

From --

MR. VALENTIN: Stacey, if I may add --

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 also, that's a very site-specific issue.

(Simultaneous speaking.)

MS. ROSENBERG: Right.

MR. VALENTIN: -- we can not generalize that external events are going to be dominate because it -- it depends on -- the site-specific hazard. So one site may have a different risk profile than --

than others, so it -- it would be a very big generalization to make that external hazards are more dominant than internal events because it will depend on the -- on the -- on the site.

MS. ROSENBERG: Thank you, Milton. I think we have time for one more question. I want to be a little bit sensitive to the time. So this question is for Marc. And the question is, are any tornados spawned out of coastal hurricanes? How are hurricanes integrated into these maps?

DR. LEVITAN: Yes, great question. Yes, the -- the maps did consider hurricane-spawned tornados. We did do a little side study on that.

Hurricane-generated tornados have different characteristics. They're typically less intense and we don't see F-4s of F-5s tornados coming from them.

And those actually -- but they are very frequent.

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 They can be very frequent. And so, if you remember on one of the slides I showed kind of our regionalization and the different areas. And so we had kind of a -- a somewhat kind of a narrow strip along the coast that really captured some of those hurricane-spawned tornados. Although hurricane-spawned tornados can curve farther inland, a lot of them do occur at least fairly close to the coast. So those -- those are built into the maps.

MS. ROSENBERG: Great. Okay, thank you.

So I just want to go to the closing remarks slide, if we can. Great. So you know, as we discussed, the area of external hazard risk continues to evolve.

And we look forward to continuing to use this knowledge to maintain safe operation of our nuclear facilities.

So I would like to thank our distinguished panel members for their time today, and also I would really like to thank you for being with us. And I want to put a special thank you out there for the -- our session coordinators, Tom Aird and Milton Valentin. Next slide, please.

Okay, the -- that slide. So I just want to say, feel free to reach out to us if you have any 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 questions after this session. I know we didn't --

we weren't -- we didn't have time to answer every question. So you have our contact information. And please remember to provide your feedback on the session. So it's a beautiful day out there and I hope you have a great rest of your day. This session is adjourned.

(Whereupon, the above-entitled matter went off the record at 2:46 p.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

62 NEAL R. GROSS COURT REPORTERS AND TRANSCRIBERS 1323 RHODE ISLAND AVE., N.W.

(202) 234-4433 WASHINGTON, D.C. 20005-3701 (202) 234-4433