ML24337A224

Corrected Mitman Expert Report - Redacted

ML24337A224
Person / Time
Site:	Oconee
Issue date:	05/15/2024
From:	Jeffrey Mitman - No Known Affiliation, Beyond Nuclear, Sierra Club
To:	NRC/SECY
References
RAS 57025, 50-269-SLR-2, 50-270-SLR-2, 50-287-SLR-2
Download: ML24337A224 (1)

Text

i NRC Relicensing Crisis at Oconee Nuclear Station:

Stop Duke from Sending Safety Over the Jocassee Dam Updated Analysis of Neglected Safety, Environmental and Climate Change Risks1 Jeffrey T. Mitman April 2024 Corrected May 15, 2024 Submitted to U.S. Nuclear Regulatory Commission on Behalf of Beyond Nuclear, Inc.

and The Sierra Club, Inc.

In Subsequent License Renewal Proceeding for Oconee Nuclear Power Plant, Units 1, 2, and 3 1 Authors note: This report updates and revises my September 2021 report, NRC Relicensing Crisis at Oconee Nuclear Station: Stop Duke from Sending Safety over the Jocassee Dam (U.S. Nuclear Regulatory Commission (NRC) Agencywide Data Access and Management System (ADAMS) Accession No. ML21270A250). This report updates the information and analyses provided in my 2021 report and adds my evaluation of the NRCs accident analysis in the NRCs Draft NUREG-1437, Site-Specific Environmental Impact Statement for License Renewal of Nuclear Plants Supplement 2, Second Renewal Regarding SLR for Oconee Nuclear Station (February 2024) (ML24033A298).

ii Contents

1.

INTRODUCTION..................................................................................................................................... 1

2.

BACKGROUND....................................................................................................................................... 3 2.1 Integrated Design and Operation of Oconee Nuclear Reactors and Upstream Dams................. 3 2.2 Jocassee and Keowee Dam Characteristics.................................................................................. 4 2.3 Oconee Nuclear Plant Design and Construction........................................................................... 4 2.3.1 Design and Construction of Oconee Units 1, 2 and 3........................................................... 5 2.3.2 Post-construction addition of Safe Shutdown Facility.......................................................... 5 2.4 Flood Risk Studies......................................................................................................................... 6 2.4.1 1983 Flood Study for FERC and Construction of Wall around SSF........................................ 6 2.4.2 NSAC-60 Probabilistic Risk Assessment................................................................................ 6 2.4.3 IPE/IPEEE for Severe Accident Vulnerabilities...................................................................... 7 2.4.4 1992 Flood Study for FERC.................................................................................................... 8 2.5 Initial Oconee License Renewal and Severe Accident Mitigation Alternatives Analysis.............. 9 2.6 Updated Dam Failure and Flood Routing Evaluations and Related Regulatory Actions............ 11 2.6.1 NRC 2006 Significance Determination Process on Oconee Flooding Issue........................ 11 2.6.2 2008 NRC 50.54(f) Letter Regarding a Jocassee Dam Failure............................................. 11 2.6.3 Dukes Response to 2008 50.54(f) Letter Regarding a Jocassee Dam Failure.................... 14 2.6.4 2011 NRC Safety Evaluation................................................................................................ 15 2.7 Fukushima - Lessons Learned 2012 50.54(f) Letter and Staff Assessment................................ 18 2.8 Dukes 2021 Subsequent License Renewal Application and SAMA Analysis.............................. 20 2.9 Duke Supplemental Environmental Report................................................................................ 21 2.10 Draft SEIS..................................................................................................................................... 21 2.10.1 Power Uprate Information (Section F.3.4 of 2024 Draft ONS SEIS).................................... 22 2.10.2 Higher Fuel Burnup Information (Section F.3.5 of 2024 Draft ONS SEIS)........................... 22 2.10.3 Additional Sensitivity as it Relates to Population Dose Risk and the Jocassee Dam SAMA (Section F.4.1 of 2024 Draft ONS SEIS)............................................................................................... 22 2.10.4 Summary and Conclusions (Section F.4.2 of 2024 Draft ONS SEIS).................................... 23

3.

ANALYSIS............................................................................................................................................. 23 3.1 Failure to Ensure Adequate Protection from Failure of the Jocassee Dam or to Adequately Evaluate Environmental Flooding Risks.................................................................................................. 23 3.1.1 Mischaracterization of the scope of the environmental review......................................... 24 3.1.2 Inadequate consideration of flooding risks from Jocassee Dam Failure............................ 25 3.1.3 Important conclusions to be drawn from the flooding risk analyses for Oconee.............. 32

iii 3.2 Other Deficiencies in the Draft SEISs Risk Analysis.................................................................... 34 3.2.1 PWR All Hazards CDF Comparison...................................................................................... 34 3.2.2 Fire Events........................................................................................................................... 34 3.2.3 Seismic Events..................................................................................................................... 35 3.2.4 Underestimating Risk by Failing to Aggregate Changes in Risk.......................................... 35 3.2.5 Invalid assumption that studies of BWR and Westinghouse PWR is applicable to Oconee reactors. 37 3.3 Failure to Address Uncertainties................................................................................................. 37 3.4 Inadequate Discussion Effects of Climate Change on Accident Risk.......................................... 39

4.

CONCLUSION....................................................................................................................................... 45

iv LIST OF ACRONYMS AC Alternating Current AEC Atomic Energy Commission BWR Boiling Water Reactor B&W Babcock & Wilcox CCDP Conditional Core Damage Probability CCFP Conditional Containment Failure Probability CDF Core Damage Frequency CFR Code of Federal Regulations DEIS Draft Environmental Impact Statement ECCS Emergency Core Cooling System EAP Emergency Action Plan EDG Emergency Diesel Generator EIS Environmental Impact Statement EPRI Electric Power Research Institute FERC Federal Energy Regulatory Commission FOIA Freedom of Information Act GDC General Design Criterion GEIS Generic Environmental Impact Statement GL Generic Letter HEC-RAS U.S. Army Corps of Engineers River Analysis System IEF Initiating Event Frequency IPE Individual Plant Examination IPEEE Individual Plant Examination for External Events kv kilovolt LERF Large Early Release Frequency LOCA Loss Of Coolant Accident LPI Low Pressure Injection MSL Mean Sea Level

v MSIV Main Steam Isolation Valve NRC U.S. Nuclear Regulatory Commission NSAC Nuclear Safety Analysis Center NTTF Near Term Task Force ONS Oconee Nuclear Station ORNL Oak Ridge National Laboratory PMP Probable Maximum Precipitation PRA Probabilistic Risk Assessment PWR Pressurized Water Reactor RFI Request for Information ROP Reactor Oversight Process SAMA Severe Accident Mitigation Alternatives SDP Significance Determination Process SG Steam Generator SEIS Supplemental Environmental Impact Statement SLR Subsequent License Renewal SSF Safe Shutdown Facility UCB Upper Confidence Bound

1

1. INTRODUCTION This purpose of this report is to explain and provide the basis for my expert opinion, as a nuclear engineer and risk analyst, regarding the safety and environmental impacts of Duke Energy Corporations (Dukes) current operation of Oconee Nuclear Station (ONS) Units 1, 2 and 3, and its proposal to the U.S. Nuclear Regulatory Commission (NRC) to extend the reactors operating license terms by 20 years until 2053 (Units 1 and 2) and 2054 (Unit 3).

This report updates my September 2021 report, NRC Relicensing Crisis at Oconee Nuclear Station: Stop Duke from Sending Safety over the Jocassee Dam (U.S. Nuclear Regulatory Commission (NRC) Agencywide Data Access and Management System (ADAMS) Accession No. ML21270A250). This report updates my analysis of Dukes and the NRCs failure to demonstrate or even assert that lowering the flood height for the failure of the Jocassee Dam from a previously established limit will provide adequate protection to public health and safety. This updated report also supplies my evaluation of the NRCs accident analysis in the NRCs Draft Supplemental Environmental Impact Statement (Draft SEIS) for the proposed subsequent renewal of the operating licenses, as well as Dukes Environmental Report and Severe Accident Mitigation Alternatives (SAMA) Analysis that the NRC used to prepare the Draft SEIS.2 Finally, this updated report addresses the NRCs failure to address the potential effects of Climate Change on accident risk in the Draft SEIS.

The report is based in significant part on my experience as a nuclear engineer and safety regulator with the NRC, including evaluation of Oconees safety in relation to potential failure of the upstream Jocassee Dam.

In my expert opinion, and as discussed in more detail below, Oconees current operation, and proposed operation under an additional twenty-year subsequent license renewal (SLR) term, pose an unacceptable risk to public health and safety and the environment, due to Dukes failure to protect ONS from the flood identified by the NRC in its 2011 Safety Evaluation.3 The NRC deemed those flood protection measures necessary to provide adequate protection 2 NUREG-1437, Site-Specific Environmental Impact Statement for License Renewal of Nuclear Plants Supplement 2, Second Renewal Regarding SLR for ONS Draft Report for Comment (February 2024)

(ML24033A298) (Draft SEIS); 2021.06.07, Oconee Nuclear Station Units 1, 2, and 3 Application for Subsequent License Renewal, Appendix E (ML21158A196) (2021 Environmental Report and 2021 SAMA Analysis); 2022.11.07 Supplemental Environmental Report (ML22311A036) (Supplemental Environmental Report).

3 2011.01.28, Safety Evaluation on Confirmatory Action Letter to Address External Flooding Concerns, (ML110280153) (2011 NRC Safety Evaluation Letter). As of 2024.04.26, this document is not in public ADAMS. It appears to have been removed sometime between a Beyond Nuclear FOIA request (FOIA 2022-000210) involving this document on 2022.07.28 and 2024.04.26. All quotes from this Safety Evaluation in this report were made in my previous report (ML21270A250) dated 2021.09.27 which is still publicly available in ADAMS.

2 against a core melt accident in the event the Oconee site becomes inundated by failure of the Jocassee Dam.

While the NRC has not enforced the flood height established by the 2011 Safety Evaluation or forced Duke to implement the flood protection measures it deemed necessary for adequate protect of the Oconee reactors against flooding, neither has it withdrawn or repudiated the 2011 Safety Evaluation in which it found those measures were necessary to provide adequate protection to public health and safety. Instead, the NRC has lowered the flood height and changed the measures required to address the new flood height, without making any finding that the Oconee reactors are adequately protected against flooding risks. This raises a serious concern because the adequate protection standard, as set forth in Section 182a of the Atomic Energy Act, (42 U.S.C. § 2232(a), sets a floor for minimum safety requirements. A nuclear reactor for which the NRC is unable or unwilling to make adequate protection findings does not provide the basic level of safety required by the Act.

In addition, my report discusses the concern that NRC has failed consider the impact on Oconee from Climate Change. While NRC has looked extensively on Oconees impact on Climate Change, it has refused to consider Climate Changes impact on safe operation on Oconee. Given that Climate Change is now generally considered to be reasonably foreseeable, and in light of recent criticisms by the Government Accountability Office (GAO), the Draft SEIS failure to address climate change is a serious deficiency. 4 Finally, the accident risk analysis in the Draft SEIS is seriously defective, and thereby fails to support the NRC Staffs conclusion that the environmental impacts of re-licensing the Oconee reactors are insignificant or SMALL.5 These deficiencies include inconsistent and unsupported estimates of core damage frequency (CDF) and lack of uncertainty analysis.

Now that Dukes SLR application has come before the NRC and the NRC has issued its Draft SEIS, it is time for the agency to break its silence and address the significant safety, environmental and Climate Change issues raised by Dukes bid for another 20 years of unprotected operation. The NRC should not accept Dukes erroneous and outdated risk assessment. Instead, it should require Duke to provide a thorough and accurate estimate of the core melt risk. In addition, the NRC should require Duke to implement the flood protection measures required thirteen years ago by the NRC.

A note about secrecy: A significant portion of the information relied on in this report was not available publicly until members of the public forced NRC to release it by requesting it under the Freedom of Information Act (FOIA). I am grateful to Jim Riccio for FOIA Request FOIA/PA-2012-0325 (submitted on behalf of Greenpeace) and Dave Lochbaum for FOIA Request 4 2024.04, Government Accountability Office, NRC Should Take Actions to Fully Consider the Potential Effects of Climate Change, GAO-24-106326 (GAO-24-106326), Page 34.

5 Draft SEIS Page F-4 Line 39, F-9 Line 4.

3 FOIA/PA-2018-0010 (submitted on behalf of the Union of Concerned Scientists), which generated some of the key information relied on this report. The NRC never attempted to justify withholding this critical, damming, and now-public safety information from the public eye, nor is any justification evident.

While Duke and the NRC have continued to withhold some information relevant to this report, and has even withdrawn several documents that formerly were released under FOIA (see note 4), the information now in the public record is more than sufficient to show that Duke has failed to provide the public with an accurate, up-to-date, and thorough risk analysis of the potential for a serious core melt accident at Oconee Units 1, 2, and 3 during the second license renewal term. In addition, publicly available information is more than sufficient to show that for the past ten years, the NRC has considered the risk of a core melt accident caused by Jocassee Dam failure to implicate the adequacy of protection to public health and safety and require significant measures to prevent catastrophe. By assembling this information into a single document, the author seeks to ensure a measure of accountability by Duke and the NRC that they previously eluded through secrecy.

Finally, while some nonpublic documents are cited in the footnotes to this report, the report does not rely directly on the content of any of those nonpublic documents. Citations of those documents are provided for completeness of the record, not for their content. When the content of nonpublic document is described in this report, that description is taken from descriptions in publicly available documents.

2. BACKGROUND 2.1 Integrated Design and Operation of Oconee Nuclear Reactors and Upstream Dams Duke Energy Corp.s three-unit Oconee Nuclear Station is located in the mountains of northwestern South Carolina, at the confluence of the Keowee and Little Rivers. Licensed by the NRC in 1973 and 1974, Oconee is uniquely designed as part of a pumped storage facility: at the same time the reactors were built, Duke also built two upstream dams, for the purpose of generating additional hydro-powered electricity. When demand for electricity from the reactors was low, the plant could be used to pump water into Jocassee Lake behind the Jocassee Dam.

When demand was high, Duke would then allow flow through hydroelectric generators in the dam generating power.

The Jocassee Dams tailwaters were dammed by the Keowee Dam, below which Duke built the Oconee reactor complex. The crest of the Keowee Dam is at 815 feet mean sea level (MSL). Plant grade at ONS is 796 feet MSL. Thus, ONS plant grade is two feet below normal Lake Keowee level. Two hydroelectric generators, built into the side of the Keowee Dam, were designed to provide the nuclear plant with an emergency power supply in the event of a loss of offsite power. The Oconee design did not and does not include diesel-powered emergency generators, which are at every other U.S. nuclear power plants.

4 Thus, the Jocassee Dam and the Keowee Dam, as well as the lakes behind them, constitute an integral part of the Oconee nuclear power plant, including its backup emergency power supply.

2.2 Jocassee and Keowee Dam Characteristics Completed in 1971 and licensed by the Federal Energy Regulatory Commission (FERC), the Keowee Dam is a 170 foot-high rock-filled earthen dam about 3,500 feet in length. The Oconee nuclear power plant complex is built into the side of the dam, which contains two hydroelectric generators with a combined output of about 150 MW.6 These hydroelectric generators provide emergency power to Oconee.

The Keowee Dam lies about 14 miles downstream of the Jocassee Dam. It impounds about one million acre-feet of water and has a surface area of about 18,000 acres. The top of the dam is at 815 feet above MSL. Full pond or normal operating level of Keowee Lake is at 800 ft.

Construction of the Keowee Dam was completed in 1971.

Completed in 19751 and also licensed by FERC, Jocassee Dam is a rock-filled earthen dam 385 feet high and about 1,000 feet long. It also impounds about a million acre-feet of water in the Jocassee Lake at normal lake operating level, with an area of 7,565 acres. The lakes pumped storage capability is supplied by four hydroelectric turbines that can be reversed to pump water from below the Jocassee Dam to above the dam.

The top of the Jocassee dam is at 1,125 ft. full pond operating level of Jocassee Lake (i.e.,

normal operating level) is 1,110 ft.

2.3 Oconee Nuclear Plant Design and Construction NRC Safety Requirements for Nuclear Plant Design and Construction All nuclear power plants constructed after 1973 are required to meet 10 Code of Federal Regulations (CFR) Part 50 Appendix A General Design Criteria for Nuclear Power Plants, including Criterion 2 - Design Bases for Protection Against Natural Phenomena. General Design Criterion (GDC) 2 states in part:

Structures, systems, and components important to safety shall be designed to withstand the effects of natural phenomena such as earthquakes, tornadoes, hurricanes, floods, tsunami, and seiches without loss of capability to perform their safety functions....

6 https://www.duke-energy.com/community/lakes/hydroelectric-relicensing/keowee-toxaway/keowee-toxaway-project.

5 Oconee was built prior to 1973 and therefore was not required to meet GDC 2; but it was required to meet a similar draft version of the criterion. The pre-GDC 2 version for Oconee provided that:

Those systems and components of reactor facilities which are essential to the prevention of accidents which could affect the public health and safety or to mitigation of their consequences shall be designed, fabricated, and erected to performance standards that will enable the facility to withstand, without loss of the capability to protect the public, the additional forces that might be imposed by natural phenomena such as earthquakes, tornadoes, flooding conditions, winds, ice, and other local site effects. The design bases so established shall reflect: (a) appropriate consideration of the most severe of these natural phenomena that have been recorded for the site and the surrounding areas and (b) an appropriate margin for withstanding forces greater than those recorded to reflect uncertainties about the historical data and their suitability as a basis for design.7 It is understood that the intent of GDC 2 and its predecessors was to ensure that the emergency core cooling systems and the associated electrical power systems were protected against credible external hazards.

2.3.1 Design and Construction of Oconee Units 1, 2 and 3 During initial Oconee licensing, Duke convinced the Atomic Energy Commission (predecessor to the NRC) that a Jocassee Dam failure was not credible. Duke has repeatedly stated that they believe a Jocassee Dam failure is not credible.8 Thus, at the earliest point of design and construction, the NRC did not require Duke to protect Oconee from a Jocassee Dam failure. For instance, the turbine building, located at a grade of 796 feet mean sea level MSL), houses portions of the emergency core cooling system (ECCS) and other safety related and important to safety systems, including the service water systems and the 4kv emergency buses. But the NRC did not require Duke to build the turbine building as a watertight structure. As a result, 2.3.2 Post-construction addition of Safe Shutdown Facility Sometime prior to 1983, in order to address other Oconee design weakness not related to flooding, Duke decided to install additional equipment to improve Oconees safety. Duke 7 2021.12.31, Duke Energy Company Oconee Nuclear Station Updated Final Safety Analysis Report, (ML22180A123) at Section 3.1.2 Criterion 2 - Performance standards (Category A) Page 3.1-8.

8 2008.09.26, Duke Response to 50.54(f) Request (ML082750106), (2008 Duke 50.54(f) Response Letter) Attachment 2, Page 6. 2009.04.30, NRC Letter to Duke Evaluation of Duke Responses to NRC Letter Dated August 15, 2008, Related to External Flooding at Oconee (ML090570779), Page 2.

6 completed the installation of the safe shutdown facility (SSF) prior to 1983. The SSF is designed to address events including fire, sabotage, turbine building floods, station blackouts and tornado missile events. It contains a single diesel generator capable of supplying sufficient power only for the SSF equipment. It contains pumps capable of supply water to the steam generators and to the reactor coolant systems of all three units and a service water pump only capable of cooling the SSF loads. None of this equipment is safety related, single failure proof or redundant. It is manually controlled and operated locally from the SSF itself. The SSF is at a grade of 796 ft.9 because the SSF was not intended to be used for external floods, it was not protected from them. However, the SSF diesel and several of its pumps are below grade.

2.4 Flood Risk Studies 2.4.1 1983 Flood Study for FERC and Construction of Wall around SSF In a 1983 hydrological analysis, Duke evaluated the impacts of a postulated sunny-day failure of the Jocassee Dam and determined as follows:

The results of the study indicated an estimated peak flood elevation of

) MSL at Keowee Dam, and a resulting ONS powerblock flood depth of

). In order to reduce the risk of flooding, the licensee erected walls around the entrances to the Standby Shutdown Facility (SSF) with average wall height of

(

. The construction of the walls was not part of the design-basis.10 Thus, by 1983 Duke recognized that external flooding was possible and that if Oconee experienced a flood above grade, the flood would incapacitate the ECCS. In that event, Duke would have no way to mitigate the flood.

2.4.2 NSAC-60 Probabilistic Risk Assessment In the late 1970s and early 1980s Duke initiated one of the first industry-conducted nuclear power plant probabilistic analyses (PRA). The study was prepared by the Nuclear Safety Analysis Center 11 and was called NSAC-60. NSAC-60 was a full-scope PRA, meaning it included both internal hazards such as loss of coolant accidents (LOCA) and external events such as earthquakes. It included an analysis of core damage frequency (referred to as a Level 1 9 2018.06.18 NRC Staff Assessment Related to Focused Evaluation for Oconee, Page 3 (ML18141A755).

10 2016.04.14 NRC Staff Assessment by the Office of NRR Related to flooding Hazard Reevaluation Report NTTF Recommendation 2.1 Oconee (ML16273A128).

11 NSAC initially was a separate legal entity, collocated with the Electric Power Research Institute (EPRI).

About 1990 it was folded into EPRI.

7 analysis), containment failure frequency (referred to as a Level 2 analysis), and impacts on the surrounding population (referred to as a Level 3 analysis).

The NSAC-60 analysis included contributions to core melt frequency by failures of the Jocassee and Keowee Dams.12 As described by Duke, the study determined the failure frequency for the Jocassee Dam by compiling data for dams with similar attributes. It considered three time periods and derived three median annual failure frequencies for causes other than earthquakes and overtopping:

1900 to 1981 2.3x10-5 per year 1940 to 1981 1.6x10-5 per year 1960 to 1981 1.4x10-5 per year13 2.4.3 IPE/IPEEE for Severe Accident Vulnerabilities In 1988 the NRC issued Generic Letter (GL) 88-20, requesting all reactor licensees submit a systematic examination in order to identify any plant-specific vulnerabilities to severe accidents and report the results to the Commission.14 Initially, GL-88-20 requested licensees to analyze only internal events such as loss of coolant accidents (LOCA) and transients. The NRC subsequently issued 5 revisions. Among other changes, the revisions, expanded the scope to include external events such as tornados, seismic events and external floods.

In response, in December 1990, Duke submitted an Individual Plant Examination (IPE) that evaluated internal events.15 In December 1995, Duke submitted an Individual Plant Examination for External Events (IPEEE) that evaluated external events.16 In 1997, in a nonpublic document, Duke updated the IPE and IPEEE and resubmitted the results.17 In the 1995 IPEEE, Duke considered whether and how to evaluate the risks of external flooding at Oconee. First, Duke considered evaluating the risk of a probable maximum precipitation 12 Nuclear Safety Analysis Center, NSAC-60, A Probabilistic Risk Assessment of Oconee Unit 3, June 1984.

13 US NRC Information Notice 2012-02, Potentially Nonconservative Screening Value for Dam Failure Frequency in PRA, March 5, 2012, Page 2, (ML090510269).

14 1988.11.23, NRC Generic Letter 88-20, Individual Plant Examinations for Severe Accident Vulnerabilities.

15 1990.12, Duke IPE (nonpublic). As discussed above in my Note on Secrecy, the IPEE is cited here for purposes of identification. This report does not rely directly on the content of the IPEEE, or any other nonpublic document. When the content of the IPEEE or any other nonpublic document is described in this report, it is taken from descriptions in publicly available documents.

16 1995.12.21, Duke IPEEE (nonpublic).

17 1996.12, Duke Oconee Nuclear Station PRA Revision 2 Summary Report (ML080780111) (nonpublic).

8 (PMP) event at the Oconee site, i.e., a large storm in the direct vicinity of the plant. But Duke screened out a PMP event based on the size of the reservoirs above the Keowee and Jocassee Dams.

Duke also considered whether to evaluate a Jocassee Dam failure in the IPEEE. In making this evaluation, Duke focused on three types of dam failures: seismic dam failure, random (i.e.,

sunny day) dam failure, and a dam failure caused by a PMP above the Jocassee Dam that overtopped the dam (i.e., a dam breach caused by water flowing over the top of the dam).

The IPEEE found that a seismic failure of Jocassee Dam was a dominant contributor to the total Oconee CDF, and calculated the contribution to core damage frequency from a seismic failure of Jocassee at 7.2E-6 per year (i.e., 20% of the total seismic CDF of 3.6E-5).18 In evaluating a random or sunny day failure, the IPEEE found a CDF of 7.0E-6 per year.19 In making this estimate, Duke used a dam failure frequency of 1.3E-5 per year, an insignificant decrease from the values derived and used in NSAC-60. 20 With respect to a PMP-induced Jocassee Dam failure, Duke concluded that such a failure was not credible.21 Therefore, Duke did not evaluate a PMP-induced Jocassee Dam failure.

2.4.4 1992 Flood Study for FERC In 1992, Duke performed an inundation study to meet a FERC requirement for formulating an emergency action plan in the event that the Jocassee Dam failed. This study showed that approximately of water would inundate the yard area surrounding the SSF (i.e., the SSF yard is at elevation 796 ft MSL as previously discussed), thereby rendering inoperable Oconees all systems necessary to shut down and maintain all three units in a safe and stable condition.22 18 1995.12.05, Oconee IPEEE Submittal Report (nonpublic). See also 2008 Duke 50.54(f) Response Letter; 1996.07.08, NRC Letter: Draft Reports Related to the Keowee Hydro Station Emergency Electrical System Supply to Oconee (ML15118A442). Total seismic CDF is 3.6E-5 per year (see Page 106) while 20% of this is from a Jocassee Dam failure (Page 107), i.e., 3.6E-5 x 0.2 = 7.2E-6 per year.

19 2000.03.15, NRC Letter: Oconee Review of IPEEE (ML003694349), Staff Evaluation Page 2.

20 FOIA Response 2012-0325 Page 17 of 308, (ML15156A702) (FOIA Response 2012-0325). See also 1996.07.08, NRC Letter: Draft Reports Related to the Keowee Hydro Station Emergency Electrical System Supply to Oconee (ML15118A442), Page 110.

21 1996.07.08, NRC Letter: Draft Reports Related to the Keowee Hydro Station Emergency Electrical System Supply to Oconee (ML15118A442) Section 6.4.1, Page 110. See also FOIA Response 2012-0325.

22 While the inundation study is not a public document, the NRC described it in its 2011 NRC Safety Evaluation Letter (ML110280153 Page 1).

9 2.5 Initial Oconee License Renewal and Severe Accident Mitigation Alternatives Analysis In July 1998, Duke submitted a license renewal application to NRC, requesting an extension of the Oconee reactors licenses terms by 20 years. The NRC renewed Dukes licenses in May 2000.23 Dukes Environmental Report for the license renewal application included a Severe Accident Mitigation Alternatives (SAMA) analysis, containing a review of potential design alternatives along with any procedural, non-hardware, alternatives. 24 For its risk estimates, the SAMA analysis relied on the IPE/IPEEE risk analyses, as well as a non-public revised IPE/IPEEE submitted in December of 1997, also referred to as Oconee PRA Revision 2 and Oconee PRA/IPE Revision 2.25 The SAMA analysis started with the total core damage frequency from the IPE/IPEEE of 8.9E-5 per year, with 2.6E-5 per year (29%) from internal events and 6.3E-5 per year (71%) from external events. The external events were broken down as follows:

CDF from External Events26 Frequency (per reactor-year)

Initiating Events Seismic 3.9E-05 Tornado 1.4E-05 External Flood 5.9E-06 Fire 4.5E-06 Total External 6.3E-05 The following Table 1 shows these values in comparison to the 2021 Environmental Report and the Draft SEIS:

23 2000.05.23, NRC Renews License of Oconee for an Additional 20 Years (ML003718834).

24 1998.04, Environmental Report, Application for Renewed Operating Licenses, Oconee Nuclear Station, Units 1, 2, and 3, Attachment K, Page 1 (1998 SAMA Analysis).

(https://www.nrc.gov/reactors/operating/licensing/renewal/applications/oconee/exhibitd.pdf) 25 1998 SAMA Analysis Pages 4, 9, 10.

26 1998 SAMA Analysis, Page 10.

10 Table 1 Core Damage Frequency (CDF) for External Events CDF for Internal & External Events (per reactor year)

Initiating Events 1998 SAMA27 SEIS 199828 License Amendment 2021 SAMA29 2024 Draft SEIS30 Internal Events 6.3E-5 2.6E-5 2.4E-5 2.4E-5 Internal Flood 9.5E-631 9.5E-6 1.9E-6 1.6E-6 Seismic 3.9E-05 3.9E-5 5.7E-532 3.3E-5 3.3E-5 Tornado 1.4E-05 1.4E-5 1.7E-5 3.3E-5 External Flood 5.9E-06 5.6E-6 2.5E-7 2.5E-7 Fire 4.5E-06 4.5E-6 6.0E-533 4.6E-5 5.1E-5 Total External 6.3E-05 6.3E-5 9.7E-5 9.7E-5 Total Internal & External 8.9E-5 8.9E-5 1.2E-4 1.3E-4 The SAMA analysis considered flooding hazards from a Jocassee Dam failure, apparently in reliance on the NSAC-60 and IPEEE studies.34 The discussion about a Jocassee Dam failure describes it in the context of random failures.35 Based on this statement, it is reasonable to assume that Duke only considered random sunny-day dam failures, ignoring seismic and overtopping, failures. This approach of excluding seismic and overtopping-related dam failures was consistent with the IPEEE.

But the 1998 SAMA analysis differed from the IPEEE in the respect that it estimated the external flooding contribution at 5.9E-6 per year, whereas the IPEEE estimated the external flooding contribution at 7E-6. The 1998 SAMA analysis did not address or explain this difference. The 1998 SAMA analysis evaluated two alternatives that would impact Jocassee Dam failure consequences. The first alternative was to staff the SSF continuously with a trained operator, and the second was to increase the height of the wall protecting the SSF from 27 1998 SAMA Analysis Page 10.

28 1999.12, Generic Environmental Impact Statement for License Renewal of Nuclear Plants Supplement 2, Regarding Oconee, NUREG-1437, at Table 5-3 Page 5-6.

29 2021.06.07, Oconee Nuclear Station Units 1, 2, and 3 Application for Subsequent License Renewal, Appendix E - Applicants Environmental Report, (ML21158A196) (2021 Environmental Report). CDF values come from Table 4.15-2 Pages 4-89 to 4-109. These are the averages of the Unit 1, 2 & 3 values.

30 2024 Draft SEIS from Table F-5 Page F-15.

31 The internal flood value of 9.5E-6 from the 1998 SAMA is already included in the total value of 6.3E-5 for Total Internal Events in the previous row for 1998 values only.

32 2024 Draft SEIS from Table F-8 Page F-22.

33 2024 Draft SEIS from Table F-7 Page F-18.

34 1998 SAMA Analysis Pages 7, 15, 19.

35 1998 SAMA Analysis Page 15.

11 floods to

.36 But Duke determined these alternatives were not cost-effective.37 Duke also identified a third alternative: strengthening the Jocassee Dam and thus lowering the random failure frequency. But Duke rejected this alternative without evaluating it, on the ground that the cost would far exceed the benefit of core damage frequency reduction.38 The NRC reviewed the SAMA analysis and concluded: Based on its review of SAMAs for ONS (Oconee Nuclear Station), the staff concurs that none of the candidate SAMAs are cost beneficial.39 This included the two evaluated alternatives addressing a Jocassee Dam failure.

2.6 Updated Dam Failure and Flood Routing Evaluations and Related Regulatory Actions 2.6.1 NRC 2006 Significance Determination Process on Oconee Flooding Issue In November 2006, the NRC completed a Significance Determination Process (SDP) evaluation related to a performance deficiency involving a missing covering in the wall protecting the SSF.40 NRC characterized the missing flood barrier as a violation and determined its significance as a White finding.41 After Dukes appeals of the finding, the NRC affirmed the finding.42 Dukes repeated appeals prompted the NRC to re-evaluate the flooding risk at Oconee from a Jocassee Dam failure. While Duke had previously estimated the dam failure rate in the range of 2.3E-5 to 1.4E-5 per year (NSAC-60) and had revised it to 1.3E-5 per year (IPEEE), the NRC found these estimates of failure frequency of the Jocassee dam were too low. In the SDP appeal process the NRC calculated a Jocassee Dam failure rate of 1.8E-4 per year.43 2.6.2 2008 NRC 50.54(f) Letter Regarding a Jocassee Dam Failure In 2008, in light of its new understanding from the previously discussed SDP that the Jocassee Dam failure frequency was significantly larger than what Duke had previously represented, NRC 36 1998 SAMA Analysis Page 16.

37 1998 SAMA Analysis Page 28.

38 1998 SAMA Analysis Page 15.

39 1999.12, Generic Environmental Impact Statement for License Renewal of Nuclear Plants Supplement 2, Regarding Oconee, NUREG-1437, Page 5-19.

40 2006.11.22, NRC Final Significance Determination for White Finding and Violation (ML063260282)

(2006 NRC White Finding). The SDP is part of the NRCs reactor oversight process (ROP). The ROP is the NRCs program to inspect, measure and assess the safety performance of operating plants. The SDP is the NRCs process for assessing the significance of findings identified in the ROP.

41 2006 White Finding, Page 1.

42 2007.11.20, NRC Reconsideration of Final Significance Determination Associated with SSF Flood Barrier White Finding, (ML073241045) (2007 NRC Reconsideration of Significance Determination).

43 2007 NRC Reconsideration of Significance Determination, Page 1.

12 issued Duke a 10CFR50.54(f) letter requesting additional information.44 First, the 50.54(f) letter laid out the regulatory requirements applicable to Oconee, and described the status of Dukes flood protection measures:

Section 3.1.2 of the UFSAR, Criterion 2 - Performance Standards (Category A), states, Those systems and components of reactor facilities which are essential to the prevention of accidents which could affect public health and safety or to mitigation of their consequences shall be designed, fabricated and erected to performance standards that will enable the facility to withstand, without loss of the capability to protect the public, the additional forces that might be imposed by natural phenomena such as earthquakes, tornadoes, flooding conditions, winds, ice, and other local site effects. The current UFSAR discusses walls that are used for flood protection at the SSF.

However, it does not include the effects of a Jocassee Dam failure, nor does it include the flood protection features to mitigate the consequences of such an event. We further note that in the mid-1990s, the UFSAR was revised by removing the reference to the Jocassee Dam failure and postulated wave height of in the yard at the Oconee site.45 The letter also references the flood heights calculated from the 1992 FERC analysis. This letter characterizes the 1992 FERC analysis as

,46 while the previous discussion of the FERC analysis characterized the same analysis as having a flood height.47 In addition, the NRCs letter requested Duke to address three specific issues:

1) Explain the bounding external flood hazard at Oconee and the basis for excluding consideration of other external flood hazards, such as those described in the Inundation Study, as the bounding case.

2) Provide your assessment of the Inundation Study (the 1992 study conducted for FERC) and why it does or does not represent the expected flood height following a Jocassee Dam failure.

3) Describe in detail the nuclear safety implications of floods that render unavailable the SSF and associated support equipment with a concurrent loss of all Alternating Current power.48 In subsequent discussions with Duke, the NRC compared the Jocassee Dam hazard with other hazards considered in the design and licensing basis. It observed that a Jocassee Dam failure frequency of about 2E-4 per year was less than the hazard from general transients, losses of 44 2008.08.15, NRC letter to Duke: Information Request Pursuant to 10CFR50.54(f) Related to External Flooding Including Failure of the Jocassee Dam at Oconee (ML081640244) (2008 NRC 50.54(f) Letter).

45 2008 NRC 50.54(f) Letter, Page 1, 9 (emphasis added).

46 2008 NRC 50.54(f) Letter, Page 2.

47 2011 Safety Evaluation Letter, Safety Evaluation, Page 1.

48 2008 NRC 50.54(f) Letter, Page 2.

13 offsite power, etc., but greater than the hazard from medium and large break LOCAs (see Figure 1 below).49 It should be noted here that for all of the other hazards listed, Oconee -- as well as every other US nuclear power plant -- is required to have safety grade, fully redundant, single failure proof ECCS capable of responding. For the Jocassee Dam failure, Oconee had the SSF which is non-safety grade, has no redundancy, is not single failure proof and is not part of the ECCS. Even if the original Jocassee Dam failure rate of 1.3E-5 was correct, this is still an order of magnitude greater than the large LOCA rate of 2E-6 per year which is in the design basis and requires the ECCS to protect the public. At this point in time (2008) the SSF was protected from a Jocassee Dam failure by a -

wall that Duke from its previous analysis knew was inadequate because the 1992 analysis for FERC showed that there was a potential for over of water at the SSF.

Figure 1 Oconee Hazard (or Initiating Event) Frequency Comparison Credible Events 50 49 2008.08.28, NRC Presentation Oconee Flood Protection and Jocassee Dam Hazard (2008 NRC Presentation Oconee), Slide 8 (ML082550290).

50 2008 NRC Presentation Oconee), Slide 8 (strike through text in original).

14 2.6.3 Dukes Response to 2008 50.54(f) Letter Regarding a Jocassee Dam Failure Duke responded to NRCs 50.54(f) letter that: Duke considers a random sunny day failure of the Jocassee dam not credible because of the nature of its design, its construction, the inspections conducted during its construction, and those periodic inspections that have occurred, and continue to occur, since its construction.51 Duke further argued that the higher flood elevations posited by NRC in the 50.54(f) letter were not applicable to Oconee, because they came from the 1992 study Duke had conducted for FERC to establish an Emergency Action Plan (EAP) for the population downstream of Jocassee, and thus was intended to provide a worst case analysis rather than credible flood levels.52 After considerable discussion with Duke, the NRC sent a letter in the spring of 2009. This letter states in part: The NRC staffs position is that a Jocassee Dam failure is a credible event and needs to be addressed deterministically.53 The letter clearly articulates that the NRC is concerned about adequate protection. For example, it states: When the inundation study and sensitivity analyses are completed, the NRC staff will evaluate the results to determine whether further regulatory actions are necessary to ensure there is adequate protection against external flooding at Oconee.54 Finally, the NRC states its expectation of receiving analyses which would establish an adequate licensing basis for external flooding...55 In response to the NRCs concerns, and after further analysis, Duke decided to raise the wall height protecting the SSF by to a total height of

. It completed this work in February of 2009.56 Duke also responded to the NRCs inquiries by performing an additional hydrological analysis of the failure of Jocassee Dam and propagating the resulting flood onto the Keowee Lake and Dam and then onto Oconee. Building on the model in the 1992 study for FERC, Duke modified it and increased the level of detail. Duke reported its preliminary results to the NRC in a presentation 51 2008.09.26, Duke Letter in Response to 10CFR50.54(f) Request, Attachment 2 Page 3 (ML082750106)

(2008 Duke 50.54(f) Response Letter).

52 2008 Duke 50.54(f) Response Letter, Attachment 2 Page 3.

53 2009.04.09, NRC letter to Duke Evaluation of Duke September 26, 2008, Response to NRC Letter Dated August 15, 2008, Related to External Flooding at Oconee (ML090570779), Page 2 (2009 NRC External Flooding Letter).

54 2009 NRC External Flooding Letter, Page 3.

55 2009 NRC External Flooding Letter, Page 3.

56 2009.05.11, Duke Presentation on Oconee External Flood (ML091380424).

15 on October 28, 2009.57 Duke had expected the flood heights to decrease by using the new software and model. However, flood heights increased. The new model, using a conservative but not worst-case scenario, predicted a flood height at the SSF of about

.58 To resolve this adequate protection issue, NRC required Duke to re-perform the Jocassee Dam failure analysis using conservative input parameters (i.e., assumptions) and methods.59 Using a conservative approach would supply margin and account for uncertainty, as is the norm for design basis and licensing basis issues -- which this adequate protection issue had become.

NRC issued a Confirmatory Action Letter (CAL) on June 22, 2010. This letter confirms commitments made by Duke Energy Carolinas, LLC (the licensee) in your June 3, 2010, letter.

Specifically, the June 3, 2010, letter listed compensatory measures the licensee will implement at the Oconee Site and Jocassee Dam to mitigate potential external flooding hazards resulting from a potential failure of the Jocassee Dam.60 In response, Duke revised its 1D and 2D analysis. And Duke committed to protecting the SSF based on results from its revised analysis.61 Protective measures would include increasing the height of the flood barriers protecting the site, protecting an offsite power line from the expected flood conditions and other improvements.62 2.6.4 2011 NRC Safety Evaluation In January of 2011, the NRC transmitted to Duke a Safety Evaluation confirming Dukes approach to the issue resolution. This safety evaluation concluded:

The NRC staff evaluated the information provided by Duke in their August 2, 2010, letter. The unmitigated Case 2 dam breach parameters that were used in the flooding models, provided by Duke for the ONS site, demonstrated that the licensee has included conservatisms of the parameters utilized in the dam breach scenario. These conservatisms provide the staff with additional assurance that the above Case 2 scenario will bound the inundation at ONS, therefore providing reasonable assurance for the overall flooding scenario at the site. This new flooding 57 2009.10.28, Duke Presentation on Oconee External Flood with Initial HEC-RAS Results (ML093080034)

(2009 Duke Presentation with Initial HEC-RAS Results).

58 2009 Duke Presentation with Initial HEC-RAS Results, Slide 26.

59 2010.01.29, NRC letter Evaluation of Duke Response to Related to External Flooding at Oconee (ML100271591), See Enclosure.

60 2010.06.22, NRC Letter to Duke, CAL - ONS Commitments to Address External Flooding Concerns, (ML12363A086), Page 1.

61 2009 Duke Presentation with Initial HEC-RAS Results. Duke had presented examples of these preliminary results in its previous meeting with NRC.

62 2010.11.29, Duke Letter: Oconee Response to CAL, Page 2 (ML103490330).

16 scenario is based on a random sunny-day failure of the Jocassee Dam. This Case 2 scenario will be the new flooding basis for the site.63 The NRCs Safety Evaluation required Duke to protect the Oconee site from random sunny day failures of the Jocassee Dam to a flood depth of in order to ensure adequate protection.64 The requirement was based on conservative deterministic analysis.65 However, the Safety Evaluation was silent with respect to other relevant Jocassee Dam failure mechanisms including seismic and overtopping even though these mechanisms had been constantly discussed both internally within the NRC and with Duke. As the text indicates, this also established a new flooding basis for the site.

Dukes initial response to the CAL is contained in an April 29, 2011 letter to the NRC. As stated in that letter, Dukes purpose in writing it was to respond to the NRCs request, as noted in the Confirmatory Action Letter dated June 22, 2010... for a list of all modifications necessary to adequately protect the Oconee site from the impact of a postulated failure of the Jocassee Dam.66 Those proposed changes were summarized in a table (unnumbered by Duke).67 The table is reproduced below. For purposes of this report, I have numbered it Table 2. As can be seen from Table 2, proposed adequate protection measures included substantial additional flood walls on the North, East and South sides of the ONS.

63 2011 NRC Safety Evaluation Letter (emphasis added).

64 2011 NRC Safety Evaluation Letter, Page 12.

65 2011 NRC Safety Evaluation Letter.

66 2011.04.29, Duke Letter: Oconee Response to Confirmatory Action Letter (CAL) 2-10-003 (ML111460063), Page 2.

67 2011.04.29, Duke Letter: Oconee Response to Confirmatory Action Letter (CAL) 2-10-003 (ML111460063), Attachment Page 2.

17 Table 2 Modifications Necessary to Adequately Protect Oconee In 2010, NRC finalized its own generic dam failure frequency analysis for the Jocassee Dam.68 The staff estimated generic dam failure rates for large rock-fill dams, which it considers applicable to the Jocassee Dam, as 2.8E-4 per year.69 The authors of that analysis and other members of the NRC Staff subsequently performed additional analyses exploring and confirming those results.70 68 2010.03.15, Generic Failure Rate Evaluation for Jocassee Dam (ML13039A084). The NRCs generic analysis was published internally and subsequently released via a Freedom of Information Act request.

(2010 Generic Failure Rate Evaluation for Jocassee Dam). (I am a coauthor of this report.)

69 2010 Generic Failure Rate Evaluation for Jocassee Dam, Page 6.

70 2013.07.17, Uncertainty Analysis for Large Dam Failure Frequencies Based on Historical Data (ML13198A170); Ferrante, et al., An Assessment of Large Dam Failure Frequencies Based on Us Historical Data ANS PSA 2011 International Topical Meeting on Probabilistic Safety Assessment and Analysis, March 13-17, 2011, Wilmington, NC, USA. (I am a coauthor of this report.)

18 In 2012, because the demonstrably erroneous NSAC-60 dam failure rate was widely referenced and used throughout the nuclear industry at that time, the NRC issued an information notice warning of the inadequacies in the dam failure rate found in the NSAC-60 report.71 According to Information Notice 2012-02, NSAC-60 provide(d) an insufficient basis for estimating site-specific dam failure frequency.72 In 2019, in a more detailed study commissioned by the NRC, Oak Ridge National Laboratory (ORNL) stated:

NRC has estimated the likelihood of failure of Jocassee dam, upstream of the Oconee Nuclear Station in South Carolina, at approximately 2.8x10-4 per year. This estimate aligns with historical dam failure rates found in literature.73 If the Jocassee Dam fails but the flood does not inundate the SSF, the SSF is the first of two lines of defense (the second is the FLEX equipment). It should be noted that Dukes own risk analysis calculated that the SSF had a failure probability of about 0.27 or 27%.74 This is a very high failure probability, orders of magnitude greater than the failure probability estimated by Duke for safety related equipment found in the ECCS.

Thus, the outcome of the multi-year NRC safety evaluation was to increase the flood protection from a Jocassee Dam failure from approximately to a new licensing basis height of about Thirteen years later, the 2011 Safety Evaluation and the safety requirements it imposed remains the only NRC safety evaluation on the issue. Duke has not appealed the 2011 Safety Evaluation, nor has the NRC retracted or repudiated it. Yet, Duke has not completed the modifications described in the table above to protect the plant to a flood depth of Nor has the NRC sought to ensure their completion.

2.7 Fukushima - Lessons Learned 2012 50.54(f) Letter and Staff Assessment In 2011, the Fukushima Dai-chi disaster occurred, with waves as high as 45 feet, leading to core damage and containment failures at three of the six nuclear power plants on the site. A year later, in 2012, the NRC issued 10CFR50.54(f) letters to all licensees including Duke for ONS, requesting them to reevaluate the flooding hazards at their sites against present-day 71 NRC Information Notice 2012-2.

72 NRC Information Notice 2012-02, Page 4.

73 2019.12.14, Current State-of-Practice in Dam Safety Risk Assessment, https://www.osti.gov/servlets/purl/1592163/.

74 FOIA Response 2012-0325 Pages 110, 115 of 308.

19 regulatory guidance and methodologies being used for early site permits and combined license reviews.75 One of the tasks requested by these letters was to perform a Flood Hazard Reevaluation Report (FHRR). Duke submitted an FHRR in 201376 and a revised response in 2015.77, 78 The flood heights at the SSF appear to be in redacted material in the public version of the revised FHRR.79 At the same time that Duke submitted their revised FHRR, they submitted a second letter also supplying what appears to be a condensed version of the FHRR Revision results. While some material is also redacted from this document, the flood height at the SSF from a Jocassee Dam failure is not redacted. The flood height from a Jocassee Dam failure is given as MSL or a flood depth of

.80 Presumably, this is the flood height identified in redacted material in the FHRR Rev. 1.

The NRC evaluated the ONS FHRR Rev. 1 and issued a Staff Assessment accepting the FHRR Rev. 1 results.81 Many of the flooding parameters including the flood heights from a Jocassee Dam failure are also redacted from this document. Presumably, the redacted flood heights in the Staff Assessment agree with both the FHRR Rev.1 and the Duke letter supplying Supplemental Information re External flooding, i.e., a flood height of at the SSF.

By titling the document a Staff Assessment rather than a Safety Evaluation, the NRC Staff indicated that the document did not have the regulatory equivalence of safety findings. Safety 75 2012.03.12 Letter from NRC to all Power Reactor Licensees and Construction Permit Holders re:

REQUEST FOR INFORMATION PURSUANT TO TITLE 10 OF THE CODE OF FEDERAL REGULATIONS 50.54(f)

REGARDING RECOMMENDATIONS 2.1.2.3, AND 9.3, OF THE NEAR-TERM TASK FORCE REVIEW OF INSIGHTS FROM THE FUKUSHIMA DAI-ICHI ACCIDENT (ML12053A340), Enclosure 2, Page 1 (NRC Post-Fukushima 50.54(f) Letter).

76 2013.03.12 Flooding Hazard Reevaluation Report Enclosure 1 (ML13240A016). As of 2024.04.26 this document is not in ADAMS. It appears to have been removed sometime between a Beyond Nuclear FOIA request (FOIA 2022-000210) involving this document on 2022.07.28 and 2024.04.26.

77 2015.0.06 Duke letter re Revised Flood Hazard Reevaluation Report per NRCs Request for Additional Information, (ML15072A106).

78 2015.003.06, Enclosure 1 Revision 1 to Flood Hazard Reevaluation Report Oconee Nuclear Station (ML16272A217) (FHRR Rev. 1).

79 FHRR Rev. 1 Pages 48, 52 - 54.

80 2015.03.06, Duke letter re Supplemental Information Regarding NRC 2008 and 2012 Requests for Information Pursuant to 10 CFR 50.54(f) Pertaining to External Flooding at ONS, (ML16272A219), Table 4 Page 7 of 13 of the letters enclosure (Duke letter Supplemental Information re External Flooding).

81 2016.04.14, NRC letter re Oconee Staff Assessment of Response to Request for Information Pursuant to 50.54(f) Flood-Causing Mechanisms Reevaluation (ML15352A207), enclosing NRC Staff Assessment by the Office of NRR Related to flooding Hazard Reevaluation Report NTTF Recommendation 2.1 (M16273A128) (2016 NRC Letter re 50.54(f) Response). A redacted version of this document was released in Interim Response 3 to FOIA-2018-0010 on October 26, 2017.

20 Evaluations are one NRC means of documenting Atomic Energy Act and the Administrative Procedure Acts basis for its actions and the associated required findings of reasonable assurance that operation of the facility can be conducted without endangering the health and safety of public and will not be inimical to the health and safety of the public (e.g., 10CFR 50.57(a).82 To my knowledge NRC Staff Assessments are not procedurally controlled and do not make the required findings.

And indeed, the conclusions of the Staff Assessment do not measure Dukes submittal against the NRC safety standard of reasonable assurance of adequate protection or no undue risk.

Instead, the Staff measured Dukes submittal against an undefined reasonableness standard.

The Staff, for instance, found that [s]eismically-induced failure of the Jocassee Dam is not a reasonable mode of failure based on current information, present-day methodologies and regulatory guidance.83 Similarly, the Staff found that [o]vertopping-induced failure of the Jocassee Dam is not reasonable model of failure based on current information, present-day methodologies and regulatory guidance. 84 The NRC also approved Dukes conclusion that a random sunny-day failure was an unlikely, although reasonable, failure mode.85 These documents do not explain what criteria the NRC used to judge reasonableness. It could be just about anything. Thus, the standard appears to be weaker than the Atomic Energy Act-based regulatory standard of reasonable assurance that the licensee has achieved a minimum level of protection that is adequate to protect health and safety.

Because the Staff Assessment did not repudiate, replace or supersede the 2011 Safety Evaluations conclusions, because it applied the distinctly different and weaker (albeit undefined) standard of reasonableness rather than reasonable assurance of adequate protection, and because it did not even purport to be a Safety Evaluation, the Staff Assessment cannot be compared to the 2011 Safety Evaluation or presumed to override it in any way. The 2011 Safety Evaluation remains on the docket as the only NRC safety determination addressing the acceptability of flooding from a Jocassee Dam failure for purposes of establish compliance with the Atomic Energy Act.

2.8 Dukes 2021 Subsequent License Renewal Application and SAMA Analysis In June 2021, Duke submitted a subsequent license renewal application to NRC, requesting an extension of each of the three Oconee reactors operating licenses terms by an additional 20 years. Like Dukes initial license renewal application in 1998, the Environmental Report in Dukes SLR application relied on its PRA to look for insights into whether there was new and 82 2020.08.03 LIC-101 License Amendment Review Procedures (ML19248C539) at Appendix B Page 15.

83 2016 NRC Letter re 50.54 Response, Enclosure 2, Page 3.

84 2016 NRC Letter re 50.54 Response, Enclosure 2, Page 3.

85 2016 NRC Letter re 50.54 Response, Enclosure 2, Page 3.

21 significant information that would provide a significantly different picture of the impacts from severe accidents during the second license renewal period.86 2.9 Duke Supplemental Environmental Report On November 7, 2022, following a decision by the NRC requiring new analyses for subsequent license renewal, Duke submitted a supplemental Environmental Report for subsequent license renewal of the Oconee operating licenses. The Environmental Report asserted that continued operation of the Oconee reactors did not pose significant environmental impacts, i.e., that they were SMALL.87 The Environmental Report also stated:

On November 17, 2020 (NRC 2020), the NRC completed its review of [updated external hazards information for all operating power reactors (as ordered by the Commission following the Fukushima Dai-Ichi accident] and concluded that no further regulatory actions were needed to ensure adequate protection or compliance with regulatory requirements, including site-specific external hazards information, re-confirming the acceptability of ONSs design basis.88 2.10 Draft SEIS In February 2024, the NRC issued the Draft SEIS, repeating Dukes assertion in the Environmental Report that the environmental impacts of accidents at the Oconee reactors are insignificant or SMALL.89 In addition, like the Environmental Report, the Draft SEIS stated:

On November 17, 2020, the NRC staff completed its review for Oconee Station and concluded that no further regulatory actions were needed to ensure adequate protection or compliance with regulatory requirements, including site-specific external hazards information, re-confirming the acceptability of Oconee Stations design basis. (NRC 2020-TN8995).90 Thus, both the Environmental Report and the Draft SEIS each stated that the environmental impacts of accidents during continued operation of the Oconee reactors would be insignificant and that the reactors were adequately protected from accidents caused by external events such as flooding and earthquakes.

86 2021 Environmental Report Page 4-75.

87 Id. Page 100.

88 2022 ER Appendix E Supplement 2 Page 101. NRC 2020 is a letter from R.J. Bernardo, NRC to J.E.

Burchfield, Jr., Duke re: Oconee Nuclear Station Units 1, 2, and 3 - Documentation of the Completion of Required Actions Taken in Response to the Lessons Learned from the Fukushima Dai-Ichi Accident (Nov.

17, 2020) (ML20304A369).

89 Draft SEIS Page F-4 Line 39, F-9 Line 4.

90 Draft SEIS Page F-4 Lines 10-14. NRC 2020-TN8995 is the same NRC letter cited in footnote 3 above.

22 2.10.1 Power Uprate Information (Section F.3.4 of 2024 Draft ONS SEIS)

This section hypothesizes the impact of a potential power uprate on risk. It uses and explores a power uprate up to 30% and an average uprate of 10% and concludes that this uprate would not impact the conclusions of the 1996 or 2013 license renews GEISs.

2.10.2 Higher Fuel Burnup Information (Section F.3.5 of 2024 Draft ONS SEIS)

This section discusses increasing PWR fuel burnup to 42 to 75 gigawatt days per metric ton of uranium and that could increase population dose risk by 38 percent. But concludes that this dose risk would be bounded by the 95% UCB values in the 1996 GEIS.

2.10.3 Additional Sensitivity as it Relates to Population Dose Risk and the Jocassee Dam SAMA (Section F.4.1 of 2024 Draft ONS SEIS)

This section responds to a scoping comment on the impact of a SAMA regarding raising the height of the flood barrier wall around the SSF. It states in part:

The Oconee Station SAMA evaluated a SAMA potential improvement to increase the height of the Safe Shutdown Facility flood barrier to address the PRA sequence relating to a random failure of Jocassee Dam exceeding the

) Safe Shutdown Facility (SSF) flood barrier.91 It concludes that the overall effect of an increase by 30 times of the total population dose risk during the SLR period of extended operation does not result in significant environmental impacts.

In fact, Duke raised the height of the SSF from in circa 2010.92 This basic fact is discussed in many of the document relating to the Jocassee Dam failure and was at least at one point in time well understood by the NRC. It appears that the NRC no longer understands the basic of the issue. In any case, it is unclear what this NRC sensitivity case is attempting to illustrate. In any case, the comments author (this writer) was attempting to show the risk benefit of protecting Oconee from a Jocassee Dam failure. I was not proposing a wall around the SSF. This sensitivity analysis did not address the concern.

91 2024 Draft SEIS at Page F-29 Lines 40-42.

92 FHRR Rev. 1, Enclosure 1 at Page 7.

23 2.10.4 Summary and Conclusions (Section F.4.2 of 2024 Draft ONS SEIS)

This section makes the following conclusion: No new and significant information regarding Oconee Station was identified that was above the values previously evaluated in the 1996 LR GEIS.93

3. ANALYSIS In Sections 3.1, 3.2, 3.3, and 3.4 below, I will provide my analysis of the safety and environmental risks posed by continued operation of the Oconee reactors for an additional twenty years past their expiration dates of 2033 (Units 1 and 2) and 2034 (Unit 3). In Section 3.1, I will address the accident risk posed by failure of the Jocassee Dam. In Section 3.2, I will address other deficiencies in the Draft SEIS accident risk analysis. In Section 3.3, I will address the lack of uncertainty analysis in the Draft SEIS. And in Section 3.4, I will address the Draft SEIS failure to consider the environmental impacts of Climate Change.

3.1 Failure to Ensure Adequate Protection from Failure of the Jocassee Dam or to Adequately Evaluate Environmental Flooding Risks.

In my expert opinion as a nuclear engineer and risk analyst, Duke is now operating Oconee at an unacceptable risk to public health and safety, due to flooding risks identified by the regulatory process leading up to the 2011 Safety Evaluation. The NRC deemed those flood protection measures necessary to protect against a core melt accident with subsequent containment failure in the event the Oconee site becomes inundated by failure of the Jocassee Dam. The NRC has not sought to force Duke to implement those measures, and it has lowered the flood level to which safety equipment must be protected. Yet, the NRC has not withdrawn or repudiated the 2011 Safety Evaluation in which it found those measures were necessary to provide adequate protection to public health and safety. Nor has the NRC made any finding that the flood height and mitigation measures that were determined to be appropriate in the NRCs post-Fukushima review are adequate to protect public health and safety.

In my years as a NRC senior risk analyst, this is one of the most serious safety issues I have encountered. Yet, the NRC regulations for license renewal exclude it from the scope of safety issues that may be reviewed, because it does not relate to the aging of Oconees safety equipment. However, the NRC must also review Dukes SLR application under the National Environmental Policy Act (NEPA), which requires NRC to fully evaluate the environmental impacts of its proposed actions, including the environmental impacts of reasonably foreseeable accidents. NEPA also requires Duke to evaluate the relative costs and benefits of Severe Accident Mitigation Alternatives (SAMAs). I have applied my knowledge as a risk analyst to evaluate whether Duke has taken into account all relevant data regarding the likelihood and consequences of a core melt accident caused by failure of the Jocassee Dam. My analysis also 93 2024 Draft SEIS at Page F-30 Lines 28-30.

24 includes Dukes most recent revision to its Environmental Report and SAMA Analysis as well as the Draft SEIS.

3.1.1 Mischaracterization of the scope of the environmental review.

The Draft SEIS makes a bold statement that as part of its post-Fukushima review the staff re-confirm[ed] the acceptability of Oconee Stations design basis."94 But I cannot find any support for this assertion in the Draft SEIS or in the post-Fukushima documentation. According to the Draft SEIS, the post-Fukushima analysis focused on several areas including but not limited to: 1) Beyond design basis external events under Order (EA-12-049), and 2) Requests for information (RFI) (under 10 CFR 50.54(f)) of seismic and flooding hazards.

Order EA-12-049 is an investigation into beyond design basis external events and is by simple logic not about the design basis and therefore cannot confirm it. And the review of seismic and flooding hazards (as requested by the RFI) does not constitute a review of the entire design basis. In addition, the staff made no Atomic Energy Act assessments what-so-every regarding the seismic and flood hazards at Oconee or any other plant in regards to design basis hazards. It is accurate that the NRC did not change the Oconee (or any other licensed reactors) design basis because of these reviews but that is not the same as reviewing the entire design basis.

Hopefully, the staff in preparation of this analysis, has a better understanding of what has and has not been done, then illustrated in this claim.

Thus, the NRCs characterization of the NRCs post-Fukushima analysis is incorrect. The primary focus of the NTTF analysis was on two external hazards, i.e., flooding and seismic. While the mitigating strategies orders required protection against other beyond design basis external hazards, e.g., high winds, it in no way confirmed the existing design basis. Thus, any statement claiming that the NRC re-confirmed the ONSs design basis is wrong. The NRC did write a Safety Evaluation but as the subject of the letter and the Safety Evaluation indicate they addressed only the acceptability of the referenced orders responses.95 94 2024 Draft ONS SEIS Page F-4 Lines 13 - 14. Duke makes a similar statement in the Supplemental Environmental Report:

On November 17, 2020 (NRC 2020), the NRC completed its review of such information as to ONS and concluded that no further regulatory actions were needed to ensure adequate protection or compliance with regulatory requirements, including site-specific external hazards information, re-confirming the acceptability of ONSs design basis.

2022 ER Appendix E Supplement 2Page 101.

95 2017.08.30, NRC letter ONS Safety Evaluation Regarding Implementation of Mitigating Strategies and Reliable Spent Fuel Pool Instrumentation Related to Orders EA-12-049 and EA-12-051, (ML17202U791).

25 3.1.2 Inadequate consideration of flooding risks from Jocassee Dam Failure.

I have also found that neither Duke nor the NRC has adequately addressed the risk of flood-related accidents at the Oconee reactors. A significant problem is that Duke supplied no flooding hazard information in its 2021 Environmental Report or SAMA analysis. It simply referred back to the 1998 SAMA, which in turn refers back to the IPEEE. (And as the history shows, the NRC found via the RFI process initiated in 200896 inadequate.) And as discussed below, the IPEEE does not supply an adequate flooding analysis. And the 2022 Supplemental Environmental Report and Draft SEIS do not supply the missing information.

I will begin with Dukes risk analysis in its Environmental Report of 2021, as supplemented in 2022.

Level 1 PRA is used to evaluate the frequency of severe accidents while Level 2 and 3 PRA are used to evaluate the consequences. To perform the Level 1 analysis the basic PRA Equation is used:

CDF (/yr.)

=

IEF (/yr.)

X CCDP 97

[Eq. 1]

Where CDF is the core damage frequency (in events per year), IEF is the initiating event frequency (in events per year) and CCDP is the conditional core damage probability (all probabilities are unit-less). PRA is always intended to be a best estimate analysis.

Typical PRA start with evaluation of IEF. In the case of external flooding, a thorough analysis would include flooding from all sources. Each hazard (e.g., local intense storms (LIP), dam failures, etc.) would be characterized with a hazard curve that supplies a range of intensities (e.g., flood height and flood inundation timing) and the corresponding frequency (in some reports it is characterized as annual exceedance probabilities). An example (not Oconee) of a detailed flooding hazard curve is shown in Figure 2. The horizontal axis shows the annual exceedance probability or how often will a flood occur. The vertical axis shows the flood depth.

For example, the 1E-2 (or once in a hundred years) exceedance flood would reach a flood depth of about 510 feet. The dashed lines indicated the 90% confidence range. Again, for the 1E-2 flood the 90% confidence ranges from 507.5 feet to 511 feet. The graph gives the frequency of a flood for the range of floods between a frequency of about once every 100 years (1E-2) to once every million years (1E-6) with the uncertainty for each flood frequency. Finally, the red dashed lines indicate flood elevations where cliff edge effects. In this example a flood exceeding a height of 517 feet which is ground elevation. Again, this graph does not apply to Oconee but is an example of thorough flood hazard analysis conveying critical information regarding 96 2008 NRC 50.54(f) Letter 97 Oconee Nuclear Site Adequate Protection Backfit Documented Evaluation (circa 2010), Page 6 (ML14058A015).

26 flooding. For a plant as vulnerable to flooding as Oconee is, this is the type of analysis that could and should have been performed.

Figure 2 Best Estimate and Approximate 90% Uncertainty Bounds of Peak River Level on the Kankakee River at the Nuclear Plant Site 98 As discussed above, Duke supplied no flooding hazard information in its 2021 Environmental Report. It simply referred back to the 1998 SAMA, which in turn refers back to the IPEEE. The 1998 SAMA, however, supplied a single value, in contrast to more detailed example hazard curve illustrated in Figure 2. The single value supplied is for a Jocassee Dam failure with a rate of 1.3E-5 per year.99 That is the only information supplied by Duke about flooding initiating events. But this one data point is insufficient information to obtain any insights from the likelihood of dam failure events.

Equally important, the limited initiating event information provided in Dukes SAMA analysis is wrong. While Duke presents a Jocassee Dam failure rate of 1.3E-5 per year, NRC calculated a best generic failure rate for Jocassee of 2.8E-4 per year - more than twenty times greater.100 This information is well-known to Duke, because NRC shared its conclusions with Duke in 2008 98 2014.08 EPRI Riverine Probabilistic Flooding Hazard Analysis, Figure 8-10, Page 8-9 (3002003013).

99 FOIA Response 2012-0325 Page 17 of 308.

100 Generic Failure Rate Evaluation for Jocassee Dam.

28 injection (LPI). It would cover all the combinations that would fail both trains. A few examples combinations that would fail both trains of LPI are:

Table 4 Example Equipment Failures causing Train Failures Combination LPI Train 1 Fails LPI Train 2 Fails

1.

Train 1 pump fails Train 2 injection valve fails

2.

Train 1 pump motor fails Train 2 pump suction valve fails

3.

Train 1 power fails Train 2 injection valve fails If Duke had evaluated the scenarios described in Table 3 above, it would have derived CCDPs for each scenario. For the first scenario, with Jocassee flooding below grade, Duke could have evaluated the failure probability of the ECCS, the SSF and any other equipment that might be available. For the middle scenario where the flood waters come above grade but not to the top of the SSF wall, the ECCS fails (and is given a failure probability of 1.0), but the SSF would not be incapacitated by the flood and thus would be assigned a failure probability based on historical data. In the final scenario where the flood water come above the SSF wall, the SSF also fails and it would also be given a failure probability of 1.0.

Neither the SAMA Analysis in Dukes 2021 Environmental Report nor its 1998 SAMA analysis supplied any information about mitigating equipment failure probabilities. In fact, neither SAMA analysis supplies any CCDP information at all.

However, a minimal amount of CCDP information can be extracted from the limited amount of information that Duke supplied. Equation 1 from above (reproduced below) can be used as a starting point to extract the composite CCDP.

CDF (/yr.)

=

IEF (/yr.)

X CCDP

[Eq. 1]

Solving for the CCDP gives us Equation 2:

CCDP

= CDF (/yr.)

/

IEF (/yr.)

[Eq. 2]

From the 1998 SAMA analysis, we know that Duke used a flooding external event IEF value of 1.3E-5103 per year. The corresponding external event flooding CDF is also supplied by the 1998 SAMA analysis in the table reproduced below (this is a duplicate of Table 1 above):

Plugging the external flooding IEF and CDF (from the 1998 SAMA in Table 1) into Equation 2 allows us to find the associated CCDP:

CCDP

= CDF (/yr.)

/

IEF (/yr.)

[Eq. 3]

103 FOIA Response 2012-0325 Page 17 of 308.

29 4.5E-1

=

5.9E-6

/

1.3E-5 Thus, Dukes CCDP for external flooding is 4.5E-1 (based on the 1998 results).

If we assume this composite CCDP is correct, we can calculate a corrected best estimate CDF for external flooding events using this CCDP and the NRCs best estimate IEF of 2.8E-4 per year and Equation 1.104 CDF (/yr.)

=

IEF (/yr.)

X CCDP

[Eq. 4]

1.3E-4

=

2.8E-4 X

4.5E-1 Thus, a corrected external flooding event CDF has value of 1.3E-4 per year (again based on the 1998 SAMA), which is more than 20 times higher than Dukes wrong value of 5.9E-6 per year. It should be noted that the data used as input into the NRCs generic Jocassee Dam failure rate calculation does include failures from seismic and overtopping. Thus, my calculation includes seismic and overtopping contributions.

But the CDF of 1.3E-4 per year assumes that the CCDP of 4.5E-1 derived from the Duke analysis is appropriate. However, in 2008, Duke told the NRC that based on the 1992 inundation study, if the dam fails:

[T]he predicted flood would reach ONS in approximately

, at which time the SSF walls are overtopped. The SSF is assumed to fail, with no time delay, following the flood level exceeding the height of the SSF wall. The failure scenario results are predicted such that core damage occurs in about following the dam break and containment failure in about When containment failure occurs, significant dose to the public would result.105 Hidden in this statement is the fact that even Duke believes that if the SSF walls are overtopped, all mitigation fails, including the SSF -- thus resulting in core damage and containment failure. In other words, Duke is saying that the conditional core damage probability (CCDP) given a Jocassee Dam failure which overtops the SSF wall is a given, or has a value of 1.0, not the value of 4.5E-1. If we use this CCDP, i.e., a value of 1.0 then the CDF from a Jocassee Dam failure is equal to the Jocassee Dam failure rate or from Equation 1:

CDF (/yr.)

=

IEF (/yr.)

X CCDP

[Eq. 5]

2.8E-4

=

2.8E-4 X

1.0 104 Generic Failure Rate Evaluation for Jocassee Dam.

105 2008 Duke 50.54(f) Response Letter, Attachment 2.

30 Revisiting the Jocassee Dam failure rate, we can compare it to other initiating events. The NRC calculated a Jocassee Dam failure rate of 2.8E-4 per year. This value is in the range of LOCAs.

Even the Duke value of 1.3E-5 per year is larger than the value for large LOCA (see Figure 1 above).

Therefore, a reasonable best estimate CDF from a Jocassee Dam failure is 2.8E-4 per year based on the available PRA information, i.e., information supplied by Duke and NRC that is on the docket. This CDF is larger than the total CDF from all Oconee internal and external events of 8.9E-5 per year reported to NRC in 1999.106 It is also larger than the total CDF values reported by both Duke and NRC in their 2021 and 2024 reports respectively (see Table 1). Dukes 2008 Duke 50.54(f) Response letter supplied the following insights regarding LERF:

(T)he predicted flood would reach ONS in approximately at which time the SSF walls are overtopped. The SSF is assumed to fail, with no time delay, following the flood level exceeding the height of the SSF wall. The failure scenario results are predicted such that core damage occurs in about following the dam break and containment failure in about

. When containment failure occurs, significant dose to the public would result.107 This Duke statement is telling the NRC that given a flood from a Jocassee Dam failure in which the flood heights overtop the SSF walls containment failure is inevitable. Duke did not say the containment might fail, nor did it estimate the probability of containment failure. Duke told the NRC that the Conditional Containment Failure Probability (CCFP) given a flood induced core damage event was 1.0. This is PRA language for a LERF multiplier of 1.0. Multiplying the CDF by the LERF multiplier gives us the LERF. With a LERF multiplier of 1.0, the LERF is equal to the CDF.108 Thus, not only is the CDF from an external flooding event 2.8E-4 per year but the LERF from an external flooding event is 2.8E-4 per year.

All of the preceding impact discussion is based on Dukes 2008 conclusion of core damage in and a flood height at the SSF between which comes from the 1992 inundation analysis performed for FERC.109 However, the NRC required Duke to perform a new dam failure and flood routing analyses. Dukes new analysis increased the flood height at the SSF to about 110 The PRA values in Dukes 2021 Environmental Report and SAMA updates substantially changed, in a manner that calls into question the technical validity of Dukes CDF values for flooding.

106 1998 SAMA Analysis Page 10 (2.6E-5 + 6.3E-5 = 8.9E-5).

107 2008 NRC 50.54(f) Response Letter (emphasis added).

108 2013.09.23 NRC letter, NMP1 Integrated Inspection Report and Preliminary Greater than Green.

Finding, Page A-8 (ML13266A237).

109 2008 NRC 50.54(f) Letter, Page 1.

110 2011 NRC Safety Evaluation Letter, Page 12.

31 Table 1 supplies both the 1998 and 2021 SAMA results (see the table for all of the relevant references). One result that is not explained within the SAMA update is a ten-fold increase in the fire results from 4.5E-6 to 4.6E-5. This one hazard contributes more than all of the internal events combined.

Of more relevance to this discussion is the change in external flooding risk. Duke believes that this hazard risk drops to a value of 2.5E-7. According to the 1998 SAMA external flooding comprised 9% of the total external event risk (= 5.9E-6 / 6.3E-5). The 2021 SAMA shows a decrease to less than 1% of total external event risk (= 2.5E-7 / 9.7E-5). It should be noted that total external event risk has increased by 54% (= 1 - 9.7E-5 / 6.3E-5) primarily from the afore mentioned ten-fold increase in fire risk. Based on my search of the records, I find no documentation of a peer reviewed analysis or a licensing application that justifies these flood calculational reductions.

Large early release frequency (LERF) information can be extracted from the 2021 SAMA. Those extracted values are shown in Table 5 below.

Table 5 LERF and CCFP from 2021 SAMA111 LERF Values (per year)

(From Table 4.15-2)

Condional Containment Failure Probabilies (CCFP)1 (Calculated from Table 4.15-2)

Hazard Unit 1 Unit 2 Unit 3 Average Unit 1 Unit 2 Unit 3 Average Internal Events 4.8E-7 4.8E-7 4.8E-7 4.8E-07 2.0E-2 2.0E-2 2.0E-2 2.0E-2 High winds 2.9E-7 3.2E-7 2.9E-7 3.0E-07 1.8E-2 1.7E-2 1.7E-2 1.7E-2 External "ood

<1E-11

<1E-11

<1E-11

<1E-11

<4.1E-5

<4.1E-5

<4.1E-5

<4.1E-5 Fire 4.5E-6 4.3E-6 2.8E-6 3.9E-06 8.8E-2 7.9E-2 8.4E-2 8.4E-2 Seismic 1.4E-5 1.4E-5 1.4E-5 1.4E-05 4.2E-1 4.2E-1 4.2E-1 4.2E-1 Total external events 1.8E-5 1.8E-5 1.7E-5 1.8E-05 1.8E-1 1.7E-1 2.0E-1 1.9E-1 Total Base 1.9E-5 1.9E-5 1.7E-5 1.8E-05 1.5E-1 1.4E-1 1.6E-1 1.5E-1 Table Note 1: The CCFPs are calculated by dividing the LERF by the corresponding CDF value (CCDF = LERF / CDF). The CDF values come from Table 1 (which are also from 2021 SAMA).

In the above discussion regarding the 2008 Duke 50.54(f) Response Letter (Attachment 2), Duke indicated that in the postulate flood scenario LERF was equal to CDF or as previously discussed, the CCFP was equal to 1.0. I previously estimated CDF at 2.8E-4 and thus LERF is also equal to 2.8E-4. While the 2008 scenario was based on a Jocassee Dam failure, we dont know what flood scenarios are contained in the 2021 SAMA flood analysis. None the less, the decrease is astounding. LERF drops to <1E-11. This LERF value is over four orders of magnitude lower than any other internal or external event LERF! All other LERF values are in a range of 2.9E-7 to 111 2021 Environmental Report, Table 4.15-2, Pages 4-89 to 4-109.

32 1.4E-5. I have derived Conditional Containment Failure Probabilities (CCFP) from the data in Dukes 2021 SAMA Table 4.15-2 by manipulating Equation 6.

LERF (/yr.)

=

CDF (/yr.)

X CCFP

[Eq. 6]

As we know the values for each hazards CDF and LERF from the 2021 SAMA (see Tables 1 & 5 respectively). CCFP can be calculated by equation 7.

CCFP

=

LERF (/yr.)

/

CDF (/yr.)

[Eq. 7]

The derived CCFP values are also shown in Table 5 above. The CCFP values for internal events and high winds are typical of large dry pressurized water reactor (PWR) containments like Oconee. The fire value is substantially elevated and the value for seismic is larger than the value for boiling water reactors (BWRs), which are understood to have substantially weaker (i.e.,

more failure prone) containments. But the derived value for Oconee flooding is more than three orders of magnitude better. This is astounding and not correct. It illustrates that the entire flooding PRA is questionable at best.

The Draft SEIS parrots Dukes flooding CDF values (see Table 1) without supplying any explanation. It cites no audit reports or internal verification calculations, it simply repeats Dukes values.

3.1.3 Important conclusions to be drawn from the flooding risk analyses for Oconee.

As discussed above, the 1998 SAMA analysis considered flooding hazards from a Jocassee Dam failure, apparently in reliance on the NSAC-60 and IPEEE studies.112 Duke in its 2021 SAMA radically revised downward the earlier estimates and the NRCs 2024 Draft SEIS apparently accepted these new estimates without a basis. It should be noted that the discussion about a Jocassee Dam failure describes it in the context of random failures.113 Based on this statement, it is reasonable to assume that Duke only considered random sunny-day dam failures, ignoring seismic and overtopping, failures. This approach of excluding seismic and overtopping-related dam failures was consistent with the IPEEE. The omission is significant, with potentially huge implications for flooding risk at Oconee.

It is helpful to put these flooding results into perspective. Dukes August 2010 analysis indicated a peak flow across the Keowee Dam and significantly onto the Oconee site, of between 2.3 and 2.8 million cubic feet per second (cfs) and a peak flow across the Oconee intake canal structure 112 1998 SAMA Analysis Pages 7, 15, 19.

113 1998 SAMA Analysis Page 15.

33 of between 0.7 and 0.8 million cfs.114 As a point of reference, the average flow of the Mississippi River at New Orleans is approximately 0.6 million cfs.115 The 2010 Duke analysis also tells us that the flood height at the Keowee Dam is to an elevation between MSL.116 Bear in mind that the top of the Keowee Dam and the intake dike are at 815 feet MSL, thus the dam is overtopped by some

.117 This is a lot of water on the Oconee site, a site that was never designed to handle any water on site. Instead, Oconee was designed as a dry site, i.e., a plant that would expect no water on site.

These significantly higher CDFs and LERFs indicate a significantly higher risks to the public and the environment than Duke and the NRC acknowledge. Yet, there is no evidence that Dukes 2021 Environmental Report has considered this new and significant flooding hazard information, the information from the more current dam failure and flood routing study that concluded with the flood depth or how this would impact the corresponding CDFs or LERFs. Nor has it considered the significant uncertainty on the timing, flood heights and flows, which should be part of any thorough risk assessment. Section F.4.1 of the Draft SEIS attempts to address the issue with a sensitivity case. It does not review the risk results from a Jocassee Dam failure nor address why the results should or should not be incorporated into the NEPA analysis.

As previously discussed, Dukes Environmental Report does not resolve or adjudicate the extensive work done in the Jocassee Dam failure and flood routing analysis, even though this work has supplied significant insights into possible additional severe accident mitigating strategies. For instance, although the NRC required significant flood control measures in the 2011 Safety Evaluation, Duke does not mention them at all - either to take credit for them or, if they have not been installed, to explain why not. Nor does the NRCs 2024 Draft SEIS. Duke has also failed to mention some other obvious ways to reduce the flood hazard from Oconee, such as preemptively shutting down the reactors when reservoir water levels get too high, lowering the water levels in the lake behind the Jocassee and Keowee Dams, or lowering the crest elevation of some of the surround Jocassee Dam earthworks such that they overtop before the Jocassee Dam proper, thus lowering the flood impacts at ONS. PRA is a valuable tool for identifying vulnerabilities (and suggesting associated corrective measures), evaluating the costs and benefits of these measures, and also prioritizing them for their effectiveness.

Unfortunately, the public has not benefited from a thorough and comprehensive external events flooding PRA.

114 2010.08.02 Duke letter Oconee Response to CAL, Attachment 1, Table 1, Page 4 (ML102170006)

(2010.08.02 Duke Oconee Response to CAL).

115 National Park Service, Mississippi River Facts, https://www.nps.gov/miss/riverfacts.htm 116 2010.08.02 Duke Oconee Response to CAL, Attachment 1, Table 2, Page 9.

117 2011 NRC Safety Evaluation Letter, Page 12.

34 Another significant shortcoming of Dukes Environmental Report and NRCs Draft SEIS is their failure to consider other Jocassee Dam failure mechanisms besides random sunny-day failures.

Both Duke and the NRC ignore seismic failures and overtopping failures, although they are both comparable contributors to public and environmental risk. Seismic failure could cause the dam to fail faster and overtopping failures would include additional water volumes behind the Jocassee Dam and potentially the Keowee Dam both scenarios could increase the flood volumes and heights at Oconee. The NRC should update its Draft SEIS to consider these significant contributors to accident risk. And ask discussed in below, that discussion should include the effects of increased frequency and intensity of flooding on overtopping risks.

3.2 Other Deficiencies in the Draft SEISs Risk Analysis.

In addition to the deficiencies described above, the Draft SEIS risk analysis is deficient in other significant respects.

3.2.1 PWR All Hazards CDF Comparison.

In Section F.3.2 of the Draft SEIS, Table F-4 presents a PWR All Hazards (Full Power) CDF Comparison. The accompanying text states the Oconee all hazards CDF is less than the highest estimated Internal Events CDF from the 1996 LR GEIS (Indian Point 2) at a value of 8.90E-5. But this value is not the known current all hazard value for Oconee -- which is 1.3E-4 per year.

The discussion goes on to state:

Although the Combined CDF (All Hazards) increased to 1.26 x 10-4 per reactor-year, the Oconee Station All Hazards CDF is still less than the highest estimated internal events CDF (Indian Point 2 is 3.5 x 10-4 per reactor-year) used in the 1996 LR GEIS.118 But again, this comparison is not based on the latest available information. The Draft SEIS shows that the Oconee all hazards CDF of 1.3E-4 is about twice as high as the mean PWR &

BWR all hazards values of 6.1E-5 and 6.6E-5 mean and median values.119 In this respect, the Draft SEIS significantly understates accident risks.

3.2.2 Fire Events In Section F.3.2.1, the Draft SEIS discusses how the Duke used an external events multiplier to calculate the estimated population dose risk.120 It states that the external events multiplier is 118 2024 Draft ONS SEIS Page F-15 Lines 11 - 13.

119 2024.02, Generic EIS for License Renewal of Nuclear Plants Technical Appendices Volume 3 Revision 2 (ML23201A226) (2024 Draft GEIS) Table E.3-12 Page E-36.

120 Id., Page F-19 (starting at Line 4).

35 obtained by dividing the all-hazards CDF by the internal events CDF. But an examination of the Large Early Release Frequency (LERF) data supplied in Table 4.15-2 (and reproduced here in Table 5) shows that the LERF values for fire are disproportionally larger for fire then for internal events. This can be seen by deriving the Conditional Containment Failure Probability (CCFP). A discussion on CCFP derivation is given below. The CCFP for internal events is about 2E-2 while the value for fire is 8.4E-2 (see Table 5 below for these values). Thus, using an external events multiplier (based on a ratio of all hazards to internal events CDF) underestimates the impact of a significantly larger containment failure probability for fire. This process leads to unrepresentatively low population dose by a factor of four for fire and fire is the largest external event contributor to risk.

3.2.3 Seismic Events In Section F.3.2.2, the Draft SEIS states:

[G]iven the significant margin between the cumulative population dose risk results from the Oconee Station license renewal SAMA Analysis and the cumulative 95th percentile UCB [upper confidence bound] population dose risk results from the 1996 LR GEIS (factor of 266), the Oconee SLR ER [Environmental Report] FCDFs [fire core damage frequencies] do not challenge the 95th percentile estimates used in the 1996 LR GEIS.121 Presumably, this comparison relies upon the same external event multiplier discussed above.

But this same fallacy exists here as with the fire results: the seismic CDF values that the multiplier relies upon depend only on the CDF values and do not take into consideration the difference in the containment failure probabilities. In the seismic case the CCFP is 4.2E-1 or 42%

(see Table 5 below). Thus, it is far from clear that the assumed margin in the population dose exists.122 3.2.4 Underestimating Risk by Failing to Aggregate Changes in Risk The Draft SEIS evaluates changes in impact for various accident/risk scenarios (e.g., changes in fire CDF, changes in seismic CDF) to see how each affects the conclusions of the 1996 GEIS regarding accident risk. The Draft SEIS then compares these changes with the margin between Oconee SAMA and the 1996 LR GEIS. Thus, the Draft SEIS states:

Given the significant margin between the cumulative population dose risk results from the Oconee Station SAMA and the cumulative 95th percentile UCB population dose risk results (factor of 266) 121 Draft SEIS Page F-19, Lines 16 through 18.

122 Draft SEIS Page F-20 at lines 22-23 there is an editorial error when the text states: lie in the range of 10 x 10-4 per year to 10 x 10-4 per year

36 from the 1996 LR GEIS, the reevaluated Oconee Station SCDF does not 95th percentile estimates used in the 1996 LR GEIS.123 Each time that the Draft SEIS does this the NRC concludes that there are large margins between the risk results derived by the 95th-percentile UCB and the scenario being evaluated (see example above). But there is a fundamental problem with this approach, which causes the NRC to seriously underestimate the total risk increase from the aggregate of the scenario. The NRC is evaluating each scenario in isolation, without examining their compounding effects. The NRC thereby seriously underestimates the change in risk.

Instead of looking at each scenario in isolation, Ill aggregate the increased risk posed by the multiple scenarios which the NRC evaluates individually. I accept, only for the purposes of this illustration, the NRC risk impact calculations on each individual scenario.

Table 6 - Aggregation of Changes in Risk Scenario SAMA 1998 (CDF)

LAR (CDF)

Risk Factor Change Comment Fire124 4.5E-6 6.0E-5 13 factor change in CDF Seismic125 3.9E-5 5.7E-5 1.5 factor change in CDF Power Uprate126 1.3 factor change in LERF Higher Burnup Fuel127 1.38 factor change in population dose risk LPSD128 2

factor change in CDF Jocassee Dam Failure129 30 factor change in population dose risk Total Impact 2098 The total impact of this aggregation is a factor of over 2000 risk increase. This total is the simple product of the individual values supplied by the NRC and shown in the risk factor change column of the table. As can be seen from the table the supposed margin between the potential change in seismic risk and the 1996 SAMA is a factor of 266. But the risk aggregation of these 123 2024 Draft SEIS Page F-21 Lines 38 - 41 (emphasis added).

124 2024 Draft SEIS Table F-18 Page F-18.

125 2024 Draft SEIS Table F-8 Page 22.

126 2024 Draft SEIS Page F-24 Lines 41.

127 2024 Draft SEIS Page F-25 Lines 17.

128 2024 Draft SEIS Page F-25 Lines 41 - 42.

129 2024 Draft SEIS Page F-29 Line 46.

37 six issues has a factor of over 2,000. Thus, the risk aggregation swamps the seismic factor of 266.

In providing this estimate, I recognize that the mathematical aggregation of these individual risks in this table is inaccurate because they multiply relate to different elements of a risk analysis that may not be compared directly to achieve an accurate result. For instance, three of the results are changes in core damage frequency, two are in population dose risk and the final is in large early release frequency. However, it is legitimate to aggregate these numbers in order to illustrate the scale on which the NRC is underestimating the effect of individual CDF changes on overall risk. I would note that it is possible to conduct a mathematically correct aggregation, but the available data in the various sources (i.e., 1996 GEIS, 1998 Oconee SEIS, 1998 Oconee ER, 2013 GEIS Rev. 1, 2021 Oconee ER and Draft Oconee SEIS) are simply insufficient to the task. The NRC should provide the necessary data and conduct the analysis in order to provide a reasonable estimate of how changes in CDF estimates for multiple scenarios affect overall risk.

3.2.5 Invalid assumption that studies of BWR and Westinghouse PWR is applicable to Oconee reactors.

In Section F.3.2.2, the Draft SEIS also discusses the State-of-the-Art Reactor Consequence Analysis (SOARCA) work. As stated in the text, the SOARCA work was performed on a BWR and two Westinghouse PWRs.130 But Oconee, as a Babcock & Wilcox (B&W) PWR -- with once through steam generators (SG) in contrast to more common u-tube steam generators -- is significantly different than the SOARCA plants.

In addition to the differences in reactor design, Oconee has other unique features. It is the only plant in the nation without emergency diesel generators (EDG) as the required source of onsite emergency power, relying instead on the Keowee hydro units. SOARCA identified losses of offsite power (LOOP) as the dominant contributor to population dose. With a completely different approach to addressing LOOPs, it is unclear whether the SOARCA insights do or do not apply to Oconee absent the EDGs. Oconee also, does not have main steam isolation valves (MSIV) between the SG and the turbine. MSIV have an impact on population dose. Without additional analysis it is unclear how any useful insights obtained from the SOARCA work is relevant to Oconee.

3.3 Failure to Address Uncertainties In Section F.3.9, the Draft SEIS discusses uncertainty in the 1996 GEIS. It states that the 1996 GEIS uses the very conservative 95th-percentile, UCB estimates for environment impact.131 130 Draft SEIS page F-22.

131 2024 Draft ONS SEIS, Page F-29 Line 1.

38 These estimates are of the total population dose. But using the 95th percentile of the final results does not constitute comprehensive uncertainty analysis.

The NRCs PRA Policy Statement states: The Commission's safety goals for nuclear power plants and subsidiary numerical objectives are to be used with appropriate consideration of uncertainties in making regulatory judgments on the need of proposing and backfitting new generic requirements on nuclear power plant licensees.132 Consideration of uncertainties is an integral component of PRA133 and it is also an NRC-required component.134 The NRCs NUREG-1855 Rev. 1 supplies extensive guidance on how to perform uncertainty analysis and how to use uncertainty analysis in risk-informed decision making.135 This is because the uncertainties show the degree to which the NRC can have confidence in its predictions.136 With respect to PRAs, the NRC expects that appropriate consideration of uncertainties will be given in the analyses used to support the decision and the interpretation of the findings of those analyses.137 Because PRAs are integral to reactor risk analyses in Environmental Impact Statements (EIS),138 the requirement for uncertainty analysis is equally important to an environmental analysis as to a safety analysis.

To illustrate this point: NRC Regulatory Guide 1.174, Rev. 1, An Approach for Using Probabilistic Risk Assessment in Risk-Informed Decisions on Plant-Specific Changes to the Licensing Basis, states that if total CDF is considerably greater than 1E-4 per reactor year, the NRCs focus in considering licensing actions should be on finding ways to decrease rather than increase the risk.139 As Dukes own SAMA results indicate (see their results in Table 1) the CDF values at 1.2E-4 for Oconee are above the RG 1.174 thresholds. Likewise, the LERF values of 132 Final Policy Statement, Use of PRA Methods in Nuclear Regulatory Activities, 60 Fed. Reg. 42,622 (Aug. 16, 1995) (emphasis added).

133 American Society of Mechanical Engineers /American Nuclear Society Standard ASME/ANS RA-Sa-2009, Standard for Level 1/Large Early Release Frequency Probabilistic Risk Assessment for Nuclear Power Plant Applications, Addendum A to RA-S-2008, ASME, Feb. 2009.

134 NRC Regulatory Guide 1.200 Rev. 3 Dec. 2020 (ML120238B871). See Table 2 Summary of Technical Characteristics and Attributes of a Level 1, Internal Events PRA for the At-Power Operating Mode Page 17 and Section C.1.2.11 Technical Elements for the Interpretation of Results (Including Uncertainty Analysis) Page 34.

135 NUREG-1855, Rev. 1, Guidance on the Treatment of Uncertainties Associated with PRAs in Risk-Informed Decision Making, March 2017 (ML17062A466).

136 Id. at 1.

137 Id. at iii.

138 See Standard Review Plans for Environmental Reviews for Nuclear Power Plants, Supp. 1, Operating License Renewal at 5-3, 5-5, 5-7 (NUREG-1555, Supp. 1, Draft for Comment, Feb. 2023).

139 2018.01, An Approach for Using PRA In Risk-Informed Decisions on Plant-Specific Changes to the Licensing Basis, Revision 3 (ML17317A256) (RG 1.174) Page 28.

39 1.9E-5 (Oconee Units 1 & 2) and 1.7E-5 (for Unit 3) exceed the RG 1.174 thresholds of 1E-5.140 This is precisely the type of situation where uncertainty information should be taken into consideration. Based on my years of experience, it is reasonable to suspect that the 90 percent uncertainty confidence bands around the external hazards of fire, flood and seismic are around two or event three orders of magnitude. These deficiencies deserve serious consideration.

An adequate probabilistic risk analysis would include parametric uncertainty data on all input parameters and calculate the corresponding CDF and LERF with uncertainty bounds (e.g. a CDF or LERF of 1E-5 per year with a 90% confidence band of 1E-6 to 5E-5 per year). The analysis would then propagate those CDF and LERF values with their uncertainty bands through the Level 2 and Level 3 PRA evaluations ending with estimates of both prompt and latent cancer fatalities with uncertainty bands. Finaly, the analysis would compare that calculated values with their corresponding uncertainties against the decision thresholds, i.e., safety goals. But the Draft SEIS deficient and unacceptable as it never calculates probabilistic uncertainties.

Therefore, it does not have a basis for confidence in its risk estimates for purposes of assessing environmental risk or compliance with safety goals.

3.4 Inadequate Discussion Effects of Climate Change on Accident Risk All of the above discussion on risk, including flooding risks, are based on past assumptions regarding external hazards. The authors of those previous analysis (e.g., 1983 FERC, NSAC-60, IPE/IPEEE studies) were unaware of the significant changes in the likelihood and severity of extreme weather events that are occurring today and will inevitably increase as a result of Climate Change. Consistent with GDC 2, the discussion of risk in these documents are based on natural phenomena which are historically reported for sites and their surrounding areas.

While GDC 2 required the consideration of sufficient margin, there was never any consideration that future hazards could be worse than anticipated from the past ones.

However, the NRC knows that Climate Change is inevitable and that it will substantially impact both the frequency and severity of weather-related hazards including Local intense precipitation (LIP), probably maximum precipitation (PMP) and other severe weather effects.

In describing the local environment, the Draft SEIS states that nuclear power plant structures, systems, and components (SSCs) important to safety are designed to withstand the effects of natural phenomena, such as flooding, without loss of capability to perform safety functions.

And the Draft SEIS states that if new information about changing environmental conditions becomes available, the NRC will evaluate the new information to determine if any safety-related changes are needed.141 However, the Draft SEIS does not address the fact that the Oconee reactors were not designed to withstand a flood caused by failure of the Jocassee Dam, 140 RG 1.174 Page 28.

141 Draft SEIS Pages 3 3-36.

40 and it does not discuss the environmental risks caused by inevitable increases in the frequency and severity of severe storms as a result of Climate Change.

Instead, the NRC has arbitrarily decided:

The impacts of natural phenomena, including seismic hazards, on nuclear power plant systems, structures, and components are outside the scope of the NRCs license renewal environmental review.142 In making this statement, the NRC has ignored guidance from the Council on Environmental Quality which states:

[A]gencies should consider increased risks associated with development in floodplains, avoiding such development wherever there is a practicable alternative, as required by Executive Orders 11988 and 13690. Agencies also should consider the likelihood of increased temperatures and more frequent or severe storm events over the lifetime of the proposed action... For example, an agency considering a proposed development of transportation infrastructure on a coastal barrier island should consider Climate Change effects on the environment and, as applicable, consequences of rebuilding where sea level rise and more intense storms will shorten the projected life of the project and change its effects on the environment.143 Focusing on the NRC, a recent report by the Government Accountability Office states:

Nuclear power plants can be affected by natural hazardsincluding heat, drought, wildfires, flooding, hurricanes, sea level rise, and extreme cold weather eventssome of which are expected to be exacerbated by climate change144 However, the GAO concludes that:

NRCs actions to address risks to nuclear power plants from natural hazards in its licensing, license renewal, and inspection processes do not fully consider the potential increased risks from natural hazards that may be exacerbated by Climate Change.145 142 Draft SEIS Page 3-30, Lines 12 - 14.

143 2023.01.09, Federal Register Notice, National Environmental Policy Act Guidance on Consideration of Greenhouse Gas Emissions and Climate Change, Page 1207 Section V (footnote omitted).

144 GAO-24-106326, Page 1.

145 2024.04, Government Accountability Office, NRC Should Take Actions to Fully Consider the Potential Effects of Climate Change, GAO-24-106326 (GAO-24-106326), Page 34.

41 In fact, this report goes on to briefly address license renewals. It also states:

Without incorporating the best available information into its licensing and oversight processes, it is unclear whether the safety margins for nuclear power plants established during the licensing periodin most cases over 40 years agoare adequate to address the risks that Climate Change poses to plants.146 I agree with the GAO that Climate Change will inevitably affect the safety of nuclear reactors, including Oconee -- and therefore should be considered by the NRC in the SEIS. If anything, the GAO Report is not strong enough because it missed the hazards associated with high winds and tornadoes and the impact of Climate Change on them. These hazards are incorporated into most if not all nuclear power plants PRAs including Oconees. At Oconee the calculated CDF from high winds is 1.6E-5 per year (see Table 1 above). This CDF is more than 50% of the CDF from all internal events. Climate Change will increase the frequency and severity of these events too. In any case, Im going to focus on Climate Changes impact on flooding.

I agree with the GAO report when it observes that climate change has driven increases in the frequency and severity of some extreme weather events.147 The NRC has focused on two types of flooding events at Oconee. The first is Probable Maximum Precipitation (PMP) falling on the watershed both upstream and adjacent to the plant. Filling the watershed beyond its capacity can cause flooding the plant and causing core damage and containment failure. The second is Local Intense Precipitation (LIP). LIP is essentially a PMP event immediately on top of the plant. The concern with LIP events is that they drop large quantities of rain directly on the plant proper without filling the watershed. If the plant is not designed to handle this event (i.e., sufficient site drainage capability), there is the potential again for water to enter the plant cause damage including core damage and containment failure. Keep in mind that Oconee was designed and built as a dry site, it was never intended and cannot deal with large quantities of water on site.

Dukes Flood Hazard Reevaluation Report address both of these hazards. Table 13 of this document identifies the flood heights associated with both. It should be noted that the flood height associated with dam failures is redacted in this version.148 The associated NTTF evaluations look at these flood hazards but none addressed Climate Change.

This omission constitutes a significant deficiency in the Draft SEIS because Climate Change demonstrably affects the frequency and intensity of some external events and therefore has the potential to significantly increase accident risks. Moreover, the frequency and intensity of 146 GAO-24-106326, Page 39.

147 GAO-24-106326, Page 13, Footnote 21.

148 FHRR Rev. 1 at Table 13 on Page 58.

42 Climate Change effects are increasing over time. Given that the NRC is proposing to rely on the Draft SEIS for decisions that could affect reactor safety decades from now, the Draft SEIS must address these changing effects over the entire licensed lifetime of reactors, which may end 4 decades from now.

Climate change has already started to increase the frequency and intensity of these events.149 As discussed above in, the Draft SEIS is already inadequate as a general matter for making broad generalizations about external event CDF based on extrapolations from internal event CDF values and limited actual plant-specific values for external event impact on population dose.

The NRC is well-aware of the issues of Climate Change and its impact on nuclear plant safety.

After the Fukushima meltdowns, the NRC Office of Research initiated a research program to develop tools to assist in probabilistic and deterministic assessments of external hazards including seismic, high winds and flooding with a consideration of Climate Change.150 In addition, Climate Change has been a topic of discussion at the NRCs Regulatory Information Conference (RIC) in recent years.151 The effects of Climate Change on accident risk are and will continue to be site-specific and not subject to generalization. For example, the three reactors at the Oconee plant -- for which the NRC is now considering an application for subsequent license renewal -- lie downstream of two large dams. The design of the dams includes consideration of the maximum probable flood induced by the maximum probable precipitation (i.e., storm). The assumption in all past FERC and NRC required studies, assumes that neither the Jocassee or Keowee Dams will be overtopped by a PMP event on their respective watershed. Climate change has the potential to significantly increase the amount of precipitation falling on watersheds above the dams. But 149 See, for example, Effects of climate change on probable maximum precipitation: A sensitivity study over the Alabama-Coosa-Tallapoosa River Basin, April 13, 2017, Journal of Geophysical Research:

Atmospheres, https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2016JD026001 (Effects of Climate Change on PMP); Climate change is probably increasing the intensity of tropical cyclones, March 31, 2021 NOAA, https://www.climate.gov/news-features/understanding-climate/climate-change-probably-increasing-intensity-tropical-cyclones; Climate Change Indicators: Weather and Climate, EPA, https://www.epa.gov/climate-indicators/weather-climate; Global Warming and Hurricanes, NOAA Geophysical Fluid Dynamics Laboratory, April 11, 2023, https://www.gfdl.noaa.gov/global-warming-and-hurricanes/.

150 See NRC Probabilistic Flood Hazard Assessment Research Program Overview, February 22 - 25, 2021 (ML21064A418) and Potential Impacts of Accelerated Climate Change, PNNL-24868, May 2016 (ML16208A282)).

151 See Climate Change Impact on the Safety of Nuclear Installations, March 8-10, 2022 (ML22140A312)) & Observations on Extreme Weather and Impacts on Nuclear Power Plants, EPRI ML22140A320, 2022).

43 with Climate Change will the respective reservoirs have sufficient capacity to handle a Climate Change exacerbate PMP? The NRC seems uninterested.

Climate change affects risk in two ways. First, it increases the likelihood or initiating event frequency of events. For example, increased storm frequency can lead to higher initiating event frequency for losses of offsite power (LOOPs). Second, Climate Change can increase the probability of failure of design features or mitigation equipment. A 2020 severe windstorm at the Duane Arnold plant (ML21139A091) illustrates this phenomenon. While the storm may or may not be directly attributable to Climate Change, it is a reasonable example of the type of severe weather effects that Climate Change can cause today and will cause in the future. In that case, a severe windstorm caused a loss of offsite power (LOOP). As a result of the LOOP, debris accumulated at the suction of the service water systems, which are necessary to cool the emergency diesel generators (EDGs) and the emergency core cooling system (ECCS) heat exchangers. The NRCs risk analysis of the event showed an increase in the failure probabilities of the service water system, the EDGs and the ECCS due to this climate-related external event.

Consideration of these risks in an EIS would provide important information regarding climate-related accident risk as well as identification of mitigation measures to address those risks.

A third way that Climate Change affects risk analysis, which is unique to flooding risk, is the cliff edge effect. With most hazards if the severity is increased slightly, the stress on the system is increased somewhat proportionately. However, with many flood-related issues, a small increase in the hazard can cause a dramatic and often overwhelming impact on a structure. For example, a small increase in wave height could raise the flood height sufficiently to overtop a floodwall inundating the equipment the floodwall is designed to protect. Risk analyses for climate change-related flooding must look carefully at this cliff-edge phenomenon.

PMP is a significant input into the design of critical infrastructure such as dam and reactor safety analysis directly and indirectly through its impact on probable maximum flood (PMF).

The National Academies under sponsorship of the National Oceanic and Atmospheric Administration (NOAA) has started a project to modernize the probable maximum precipitation (PMP) methodology.152 The NRC is well aware of this effort as they have already participated in at least one of the initial project workshops. PMP and PMF also impact reactor safety directly via their impact on local intense precipitation (LIP). This project will consider approaches for estimating PMP in a changing climate, with the goal of recommending an updated approach, appropriate for decision-maker needs. This project is clear evidence that the Federal Government and the NRC understands the significance and severity of Climate Change on critical infrastructure. Waiting for the project completion is unnecessary and inappropriate.

Climate change is here, the NRC and the licensee know it, steps should be taken now to protect the plant and the public from its effects.

152 See https://www.nationalacademies.org/our-work/modernizing-probable-maximum-precipitation-estimation#sectionSponsors).

44 As an example of what Climate Change has the potential to do let me use the example of Duane Arnold. The Duane Arnold plant in Iowa was prematurely and permanently shuttered after being hit with a Derecho with wind speeds exceeding 100mph. Climate change has been implicated in the severity of this extreme weather event (Hints of a derecho-climate change link, ten years after 2012 storm, Washington Post, June 29, 2022, https://www.washingtonpost.com/climate-environment/2022/06/29/derecho-climate-change-severe-storm/)

So how could a Climate Change exacerbated flood impact Oconee? If water levels rise above grade, systems need to protect the plant during transients and accident will begin to fail. The first cliff edge, when water rises essentially above grade it will flow unimpeded into the turbine building and its below grade levels. These levels contain safety related and other risk significant equipment including high pressure service water, low pressure service water and emergency feedwater. The loss of these systems will impact other emergency equipment dependent on them. If water level reaches approximately not already impacted by the loss of cooling water. Finaly (the third cliff edge), if water level overtops the wall protecting the SSF it too will fail. The SSF is the last line of permanently installed equipment potentially capable of dealing with the flood. If the SSF fails all that remains is the untried FLEX equipment.

The Jocassee Dam has about of freeboard during a PMP according to the currently evaluations. See Figure 11, Jocassee PMP hydrographs in FHRR Rev.1.153 This hydrograph shows a maximum flood elevation behind Jocassee Dam of

. MSL and the top of the dam is at MSL, leaving freeboard of approximately

. The PMP event causes the lakes elevation to rise from a level of in less than 7 hours8.101852e-5 days <br />0.00194 hours <br />1.157407e-5 weeks <br />2.6635e-6 months <br /> (the hydrograph indicates rising water level starting at 2000 minutes into the analysis and hitting its maximum at about 2400 minutes. Overtopping in the current analysis of record is prevented by only significant pre-flood events (lowering the water level based on warning starting at time 0 in the hydrograph by utilizing grid dependent hydro-generators) and during the event by utilizing the tainter gates and the hydro-generators. Each additional foot of elevation behind the Jocassee Dam contains about 1% extra capacity. Thus, the extra of freeboard is about 3% extra capacity. Will a Climate Change exacerbated PMP consume that 3% of freeboard? The previously reference paper on the Alabama-Coosa-Tallapoosa River Basin (slightly to west of the Jocassee and Keowee basin) finds that:

Our results showed that PMP driven by projected future climate forcings is higher than 1981-2010 baseline values by around 20% in the 2021-2050 near-future and 44% in the 2071-2100 far-future periods (emphasis added).154 153 FHRR Rev. 1 Figure 11, Page 31.

154 Effects of Climate Change on PMP, see Abstract on Page 4808 (emphasis added).

45 Maybe, the NRC should look into it.

As discussed above Climate Change will make this type of event more likely, increasing the associated CDF, LERF and population dose values. Climate Change will also increase the intensity of storms increasing the likelihood (and speed of such events) that if the event occurs plant equipment will be impacted.

In summary, in my professional opinion, the Draft SEIS does not reflect a complete or adequately rigorous evaluation of all external hazards, does not consider uncertainties and does not address the reasonably foreseeable effects of Climate Change on the risks of accidents at Oconee. It simply does not perform the NEPA required hard look at environmental impacts.

The GAO certainly agrees that the NRC has not looked hard enough at Climate Change. Given these serious deficiencies, the NRC cannot claim to have a reasonable basis for concluding that the environmental impacts of accidents during a license renewal term are SMALL.

4. CONCLUSION The history of the NRCs regulation of the Oconee reactors presents grave concerns in several significant respects.

First and foremost, from a regulatory perspective, it is unacceptable that the NRC has allowed Duke to operate for the past ten years without completing flood protection measures that NRC required ten years ago in 2011 to protect the public from the undue risk of a core melt accident caused by failure of the Jocassee Dam.

Second, the NRCs silence on this matter for the past ten years is inexcusable. The NRC should stand by its judgment, which it has never repudiated or withdrawn, that protection of public health and safety requires installation of substantial additional flood protection measures.

Third, the NRCs risk analysis in the Draft SEIS is seriously deficient in other respects, including incomplete and misleading evaluations of CDF and lack of uncertainty analysis.

Fourth, the Draft SEIS fails to consider the real threat of climate change on reactor safety and environmental risks. These impacts are inevitable and significant.

Finally, Duke has consistently downplayed the severity of the risk posed by the Jocassee Dam, to the point that it now seeks approval of a second license term for its three Oconee reactors, based on flood risk estimates that are demonstrably incorrect, incomplete, and poorly conducted. Duke has ignored data in its own possession showing that the risk of a core melt accident with subsequent containment failure caused by Jocassee Dam failure is significantly higher than Duke asserts. Duke has also ignored significant additional contributors to core damage frequency, including seismically induced dam failure, overtopping, and outages. Of course, Climate Change will only make the flood results and effects worse.

46 SLR Proceeding: a moment of crisis and opportunity: The NRCs SLR proceeding provides the agency with an opportunity to restore public confidence in its commitment to ensure public health and safety, by ending its silence regarding the crucially important 2011 Safety Evaluation, and by requiring Duke to complete the flood protection measures required ten years ago. The NRC should prepare a new environmental risk analysis that uses correct, complete, and up-to-date methods and data. Finally, Duke should account for its failure to implement measures required by the NRC ten years ago for adequate protection, and now ignored in Dukes SLR application.