United States Nuclear Regulatory Commission - Protecting People and the Environment

NRC: Generic Environmental Impact Statement for License Renewal of Nuclear Plants (NUREG-1437 Supplement 2, Part 10)

5.0 Environmental Impacts of Postulated Accidents



[ View in Frames | Document Outline ]

Section Contents

Environmental issues associated with postulated accidents were discussed in the Generic Environmental Impact Statement for License Renewal of Nuclear Plants (GEIS), NUREG-1437 (NRC 1996). The GEIS included a determination of whether the analysis of the environmental issue could be applied to all plants and whether additional mitigation measures would be warranted. Issues were then assigned a Category 1 or a Category 2 designation. As set forth in the GEIS, Category 1 issues are those that meet all of the following criteria:

(1) the environmental impacts associated with the issue have been determined to apply either to all plants or, for some issues, to plants having a specific type of cooling system or other specified plant or site characteristics

(2) a single significance level (i.e., SMALL, MODERATE, or LARGE) has been assigned to the impacts (except for collective offsite radiological impacts from the fuel cycle and from HLW and spent fuel disposal)

(3) mitigation of adverse impacts associated with the issue has been considered in the analysis, and it has been determined that additional plant-specific mitigation measures are likely not to be sufficiently beneficial to warrant implementation.

For issues that meet the three Category 1 criteria, no additional plant-specific analysis is required unless new and significant information is identified.

Category 2 issues are those that do not meet one or more of the criteria of Category 1, and therefore, additional plant-specific review for these issues is required.

This chapter describes the environmental impacts from postulated accidents that might occur during the license renewal term.

5.1 Postulated Plant Accidents

[ Next | Top of file ]

A Category 1 issue in 10 CFR Part 51, Subpart A, Appendix B, Table B-1, is applicable to ONS postulated accidents and is listed in Table 5-1. Duke stated in its Environmental Report (ER) (Duke 1998a) that it is not aware of any new and significant information associated with the renewal of the Oconee operating licenses. No significant new information has been identified by the staff in the review process and in the staff's independent review. Therefore, the staff concludes that there are no impacts related to this issue beyond those discussed in the GEIS. For this issue, the GEIS concluded that the impacts are SMALL, and plant-specific mitigation measures are not likely to be sufficiently beneficial to be warranted.

Table 5-1. Category 1 Issue Applicable to Postulated Accidents During the Renewal Term

ISSUE--10 CFR Part 51, Subpart A, Appendix B, Table B-1 GEIS Sections

Postulated Accidents

Design-Basis Accidents (DBAs)

5.3.2; 5.5.1

A brief description of the staff's review and the GEIS conclusions, as codified in Table B-1, follows.

Design-Basis Accidents (DBAs): Based on information in the GEIS, the Commission found "The NRC staff has concluded that the environmental impacts of design basis accidents are of small significance for all plants." The staff has not identified any significant new information during its independent review of the Duke ER, the staff's site visit, the scoping process, its review of public comments on the draft SEIS, or its evaluation of other available information. Therefore, the staff concludes that there are no impacts of DBAs beyond those discussed in the GEIS.

A Category 2 issue related to postulated accidents that is applicable to ONS is discussed in Table 5-2.

Severe Accidents: Based on information in the GEIS, the Commission found that "The probability weighted consequences of atmospheric releases fallout onto open bodies of water, releases to groundwater, and societal and economic impacts from severe accidents are small for all plants. However, alternatives to mitigate severe accidents must be considered for all plants that have not considered such alternatives."

The staff has not identified any significant new information with regard to the consequences from severe accidents during its independent review of the Duke ER, the Duke Final Safety Analysis Report (FSAR) (Duke 1998b), the staff's site visit, the scoping process, the review of public comments on the draft SEIS, or its evaluation of other available information. Therefore, the staff concludes that there are no impacts of severe accidents beyond those discussed in the GEIS. However, in accordance with 10 CFR 51.53(c)(3)(ii)(L), the staff has reviewed severe accident mitigation alternatives (SAMAs) for ONS. The results of its review are discussed in Section 5.2.

5.2 Severe Accident Mitigation Alternatives

[ Prev | Next | Top of file ]

It is required in 10 CFR 51.53(c)(3)(ii)(L) that license renewal applicants provide a consideration of alternatives to mitigate severe accidents if the staff has not previously considered SAMAs for the applicant's plant in an EIS or related supplement or in an environmental assessment. The purpose of

Table 5-2. Category 2 Issue Applicable to Postulated Accidents During the Renewal Term

ISSUE--10 CFR Part 51, Subpart A, Appendix B, Table B-1 GEIS Sections 10 CFR 51.53(c)(3)(ii) Subparagraph SEIS Section

Postulated Accidents

Severe Accidents 5.3.3; 5.3.3.2; 5.3.3.3; 5.3.3.4; 5.3.3.5; 5.4; 5.5.2 L 5.2

this consideration is to ensure that plant design changes with the potential for improving severe accident safety performance are identified and evaluated. SAMAs have not been previously considered for ONS; therefore, the remainder of Chapter 5 addresses those alternatives.

5.2.1 Introduction

Duke submitted an assessment of SAMAs for ONS as part of the ER (Duke 1998a). This assessment was based on Revision 2 of the ONS Probabilistic Risk Assessment (Duke 1997a). Revision 2 constitutes a full-scope Level 3 Probabilistic Risk Assessment (PRA) with the analysis of both internal and external events; the internal events analysis is an updated version of the Individual Plant Examination (IPE) model (Duke 1990), whereas the external events analysis is the same as the Individual Plant Examination for External Events (IPEEE) model (Duke 1995). In identifying and evaluating potential SAMAs, Duke took into consideration the insights and recommendations from earlier risk studies as well as several more recent risk studies. Duke concluded that none of the candidate SAMAs evaluated were cost effective for ONS.

Based on a review of the SAMA assessment, NRC issued a request for additional information (RAI) to Duke by letter dated December 29, 1998 (NRC 1998). Major issues concerned the process used by the license renewal applicant to identify potential SAMAs, the implementation status of numerous enhancements identified in previous studies, and the inclusion of averted onsite costs (AOSC) in Duke's value impact analysis. Duke submitted additional information by letter dated March 4, 1999 (Duke 1999), clarifying the SAMA identification process, the disposition of previously identified design enhancements, and the impact of AOSC on the cost-benefit analysis. This response provided additional clarification regarding the staff's concerns and reaffirmed that none of the SAMAs would be cost-beneficial even when averted onsite costs are included.

The staff's assessment of SAMAs for ONS is provided in Section 5.2.3.2.

5.2.2 Estimate of Risk for ONS

Duke's estimates of the offsite risk at ONS are summarized below. The summary is followed by the staff's review of Duke's risk estimates.

5.2.2.1 Duke Risk Estimates

The ONS PRA model, which forms the basis for the SAMA analysis, is a Level 3 risk analysis; i.e., it includes the treatment of core damage frequency, containment performance, and offsite consequences. The model, which Duke refers to as PRA, Revision 2, consists of an internal events portion, based on an updated version of the IPE (Duke 1990) and an external events portion, based on the current version of the IPEEE (Duke 1995). The calculated total core damage frequency for internal and external events in Revision 2 is 8.9E-5 per year.

Since the ONS PRA is a "living" PRA, the original version of the IPE is being continuously updated to reflect various design and procedural changes, such as those related to the improvements identified in the IPE, to incorporate comments from the "peer review certification" and to reflect up-to-date operational experience. A comparison of risk profiles between the original IPE PRA (which was reviewed by the staff) and the current version (internal events portion of PRA, Revision 2) indicated that there are no significant differences that could change the results of the SAMA analysis by impacting the approach used to identify potential SAMAs or the assessed risk reductions.

Since the issuance of the ONS PRA, Revision 2, report, the total core damage frequency has been recalculated. An IPEEE supplemental report (Duke 1997b) further evaluated the relay chatter issue and updated the seismic core damage frequency (CDF) to be 3.5E-5 per year. A high pressure injection (HPI) reliability study performed in response to an operational event (Duke 1997c) resulted in an updated core damage frequency of 4.3E-5 per year for all events, excluding seismic. Thus, by removing conservative assumptions related to the original seismic analysis and the HPI system, the net effect of these two studies would be to reduce the total CDF for ONS from 8.9E-5 per year to 7.8E-5 per year. Despite the availability of these later studies, the results of the ONS PRA, Revision 2, were used as the basis for the SAMA analysis since the later studies did not include Level 2 and Level 3 calculations and because the net impact of the changes was a small decrease in CDF.

Since Duke's PRA is based on ONS Unit 3, the licensee performed an analysis to determine the applicability of the PRA results to Units 1 and 2 and submitted the analysis as part of the IPE. This analysis concluded that inter-unit differences do not have a significant impact on the PRA results. Most mechanical and electrical systems of Units 1 and 2 are redundant and diverse from those of Unit 3. Those systems and structures that are shared affect all three units in a similar fashion during a severe accident scenario. Because civil structures of Units 1 and 2 are similar to those of Unit 3, external events impact structures and components similarly for each unit. Therefore, the results and insights of the ONS PRA are applicable to all three units.

The Level 2 (also called containment performance) portion of the ONS PRA model, Revision 2, including the plant damage state descriptors, the Containment Event Tree, and the source term binning and containment release categories, is essentially the same as the IPE Level 2 analysis. The offsite (or Level 3) consequence analyses were carried out using the NRC-developed Calculations of Reactor Accident Consequences Version 2 (CRAC2) code, and site-specific data for meteorology, population, and evacuation modeling.

Duke estimated the total CDF for internally and externally initiated events to be 8.9E-5 per year based on Revision 2 of the ONS PRA. The breakdown of the CDF is provided in Table 5-3. External event initiators represent about 71 percent of the total CDF and are dominated by seismic (44 percent of total CDF) and tornado initiators (16 percent of total CDF). External flood and fire initiators together account for about 11 percent of the total CDF. Internal event initiators represent about 29 percent of the total CDF and are dominated by internal flood (11 percent of total CDF), transient (9 percent of total CDF), and loss of coolant accident initiators (8 percent of total CDF). Remaining contributors together account for less than 2 percent of total CDF.

Duke estimated the dose to the population within 80 km (50 mi) of the ONS site from all initiators (internal and external) to be 0.0492 person-sievert (person-Sv) (4.92 person-rem) per year (Duke 1999). The breakdown of the total population dose by containment end-state is summarized in Table 5-4. Of the total risk from all initiators, about 80 percent is due to external events. Interfacing system loss-of-coolant accident (LOCA), containment isolation failure, and late containment failure dominate external event risk (Column 3 of Table 5-4) and total risk from all initiators (Column 4 of Table 5-4) with nearly equal contributions from each. Early containment failure accounts for approximately 10 percent of the total risk from all initiators, with the majority of this contribution coming from external events. Only about 20 percent of the total risk from all initiators is due to internal events, with the majority of this risk from late containment failure (Column 2 of Table 5-4). All other internal event contributors combined account for less than 10 percent of the total risk from all initiators.

5.2.2.2 Review of Duke's Risk Estimates

Duke's estimate of offsite risk at ONS is based on Revision 2 of the ONS PRA. For purposes of this review, the staff considered the ONS study in terms of the following major elements:

  • the Level 1 and 2 risk models that form the bases for the November 1990 IPE submittal (Duke 1990)
  • the major modifications to the IPE model that have been incorporated in Revision 2 of the PRA (Duke 1997b)

Table 5-3. ONS Core Damage Frequencies

Initiating Event Frequency (per year) % of Total CDF (Int+Ext)
External Initiators
Seismic 3.9E-5 44
Tornado 1.4E-5 16
External Flood 5.9E-6 6
Fire 4.5E-6 5
Total External 6.3E-5 71
Internal Initiators
Internal Flood 9.5E-6 11
Transients 8.2E-6 9
LOCAs (small, medium, large) 6.8E-6 8
RPV Rupture 1.0E-6 1
Steam Generator Tube Rupture 4.1E-7 <1
ATWS 1.7E-7 <1
Interfacing systems LOCA 6.9E-9 <1
Total Internal 2.6E-5 29
Total CDF (Internal + External) 8.9E-5 100
  • the external event models that form the basis for the December 1995 IPEEE submittal (Duke 1995)
  • the analyses performed to translate fission product release frequencies from the Level 2 PRA model into offsite consequence measures.

The staff reviewed each of these analyses to determine the acceptability of Duke's risk estimates for the SAMA analysis, as summarized below.

The staff's review of the ONS IPE is described in an evaluation report dated April 1, 1993 (NRC 1993). In that review, the staff evaluated the methodology, models, data, and assumptions used to estimate CDF and characterize containment performance and fission product releases. The staff concluded that Duke's analysis met the intent of Generic Letter 88-20 (NRC 1988); that is, the IPE was of adequate quality to be used to look for design or operational vulnerabilities. Although the staff reviewed certain aspects of the IPE in more detail than others, the review primarily focused on the licensee's ability to examine ONS for severe accident vulnerabilities and not specifically on the detailed findings or

Table 5-4. Breakdown of Population Dose by Containment End-State

(Total Dose = 4.92 person-rem per year)

Containment End-State

% of Total Dose Internal Initiators

% of Total Dose External Initiators

% of Total Dose All Initiators

Steam Generator Tube Rupture 2.7 <0.1 2.8
Interfacing System LOCA 0.8 24.4 5.2
Containment Isolation Failure 0.5 22.0 22.5
Early Containment Failure 3.7 6.5 10.2
Late Containment Failure 9.4 22.8 32.2
Basemat Melt Through 2.2 4.6 6.8
No Containment Failure <0.1 0.2 0.3
Total 19.3 80.7 100

quantification estimates. However, ONS's risk profile and important IPE findings compare well to those of other Babcock & Wilcox plants (NUREG-1560) (NRC 1997a), and any differences are well understood. Overall, the staff believes that the ONS PRA is of adequate quality to be used as a tool in searching for areas with high potential for risk reduction and to assess such risk reductions, especially when the PRA models are used in conjunction with insights, such as those from risk importance, sensitivity, and uncertainty analyses.

The staff's review of the applicant's IPEEE is currently underway. The preliminary results did not identify any significant shortcomings or deficiencies. A limited review of the Duke submittal finds that the overall method, scope, and level of detail are generally comprehensive. The staff also notes that the Duke IPEEE has been subjected to both internal and external peer reviews. Based on these findings, the staff concludes that the external events portion of the ONS PRA provides an acceptable platform for identifying potential SAMAs and for assessing risk reductions.

The staff reviewed the process used by Duke to extend the containment performance (Level 2) portion of the IPE to the offsite consequence (Level 3) assessment. This included consideration of the source terms used to characterize fission product releases for each containment release category and the major input assumptions used in the offsite consequence analyses. This information is provided in Section 6.3 of Duke's IPE submittal. Duke used the Modular Accident Analysis Program code to analyze postulated accidents and develop radiological source terms for each of 35 containment release categories used to represent the containment end-states identified in Table 5-4. These source terms were incorporated as input to the CRAC2 analysis. The staff reviewed Duke's source term estimates for the major release categories and found these predictions to be in reasonable agreement with estimates of NUREG-1150 (NRC 1990a) for the closest corresponding release scenarios. The staff concludes that the assignment of source terms is acceptable.

The CRAC2 code has been superceded by the Melcor Accident Consequence Code System (MACCS), which, among other advancements, incorporates more recent models for calculating health effects (e.g., latent cancers). Although MACCS represents a significant improvement over CRAC2, both codes use a straight line Gaussian plume dispersion and transport model and, for the same input assumptions, provide comparable estimates of population dose (person-rem). Thus, the CRAC2 code is considered acceptable for purposes of estimating population dose for a severe accident.

The CRAC2 input in PRA, Revision 2, used site-specific meteorological data processed from measurements taken during the mid-1970s. To assess the impact that data from two different time periods may have on offsite dose, Duke obtained more recent data from the ONS site for the period January 1, 1997, through December 31, 1997. Re-analysis of the Level 3 portion of the PRA using the 1997 meteorological data (Duke 1999) shows that the risk results are only slightly impacted (reduced by about 2 percent). The staff therefore considers the meteorological data in PRA, Revision 2, to be representative of the climate for the site.

The population distribution used in Revision 2 of the PRA is based on 1990 census data. The impact of population increases was not included in Revision 2 since the purpose of the PRA was to understand the risk associated with current operation of the plant. Based on information contained in NUREG-1437 (NRC 1996), the population within an 80-km (50-mi) radius of the ONS site is projected to increase by about 33 percent between the years 1990 and 2030. Since the population dose is roughly proportional to the total population, use of the increased population value would result in an increase in the total risk from all initiators of approximately 1.6 person-rem per year. This increase is small in absolute terms and does not have a significant impact on the conclusions of the SAMA analysis, as discussed later.

Evacuation modeling is based on site-specific evacuation studies carried out by Duke. It was assumed that only 95 percent of the people within the emergency planning zone (determined by the plume exposure pathway) would participate in the evacuation. The remaining 5 percent would delay evacuation for 24 hours. This assumption is conservative relative to the NUREG-1150 (NRC 1990a) study, which assumed evacuation of 99.5 percent of the population within the emergency planning zone.

Site-specific economic data were used in the CRAC2 code. However, as discussed later, the applicant based their assessment of offsite costs on generic cost estimates rather than CRAC2 code calculations.

The staff concludes that the methodology used by Duke to estimate the CDF and offsite consequences for ONS provides an acceptable basis from which to proceed with an assessment of risk reduction potential for candidate SAMAs. Accordingly, the staff based its assessment of offsite risk on the CDF and offsite doses reported by Duke.

5.2.3 Potential Design Improvements

This section discusses the process for identifying potential design improvements, the staff's evaluation of this process, and the design improvements evaluated in detail by Duke.

5.2.3.1 Process for Identifying Potential Design Improvements

Duke's process for identifying potential plant improvements consisted of the following three elements:

  • The core damage cutsets from Revision 2 of the ONS PRA were reviewed to identify potential SAMAs that could reduce CDF.
  • The Fussell-Vesely (F-V) importance measures were evaluated for the basic events (including initiating events, random failure events, human error events, and maintenance/testing unavailabilities), and the importance ranking was examined to identify any events of significant F-V importance.
  • Potential enhancements to reduce containment failure modes of concern for ONS (including early containment failure, containment isolation failure, and containment bypass), were reviewed for possible implementation.

This included a review of recommendations from the ONS IPE and IPEEE (those that had not been implemented), results of other plant-specific SAMA analyses, and insights from the staff's report on the individual plant examination (NRC 1997a) for possible inclusion of these concepts as additional SAMAs.

As a starting point for the core damage cutset review, Duke developed a listing of the top 100 cutsets (severe accident sequences) based on internal initiators and the top 100 cutsets for external initiators. These 200 sequences include all potential core damage sequences with at least a 0.06-percent contribution to the total CDF. Duke reviewed the cutsets to identify potential SAMAs that could reduce CDF. Cutoff values of 4.5E-7 per year and 8.5E-7 per year were used to screen internal and external events, respectively. To account for the cumulative effect of cutsets below these cutoff values, the basic events importance measure was also used to identify potential enhancements, as discussed below.

For each seismic initiator cutset, Duke calculated the associated offsite risk based on the person-rem risk and CDF for the plant damage states (PDSs) attributable to the seismic initiator. Duke conservatively assumed that the implementation of plant enhancements for seismic events would completely eliminate the seismic risk and calculated the present worth of the averted risk based on a $2000 per person-rem conversion factor, a discount factor of 7 percent, and a 20-year license renewal period. This process was repeated for each of the remaining seismic initiator cutsets above the cutoff frequency. The present worth of averted risk for all of the seismic cutsets combined was estimated to be about $51,000. Duke cited sensitivity studies performed previously as part of the IPEEE analysis, which show that most of the seismic upgrades to plant components would result in only a small reduction in CDF (less than 5E-6 per year). On the basis of the small risk reduction achievable and the large costs associated with substantial seismic upgrades, Duke eliminated seismic SAMAs from further consideration.

Duke reviewed the F-V Basic Event Importance Ranking presented in the ONS PRA report, Revision 2, and identified the top 30 basic events for further consideration. These included seismic-related events, initiating events, equipment failures, and human-error events. Seismic-related events were not evaluated further for reasons discussed above. Duke judged that all but one of the initiating events, such as tornado, dam failure, and fire events, could not be significantly impacted by SAMAs and that the remaining initiator (reactor trip initiator) is adequately addressed by their current ORAM-SENTINEL configuration management system. Based on a review of the remainder, Duke identified nine events/sequences and a potential plant enhancement to address each event. The list of the potential enhancements to reduce CDF are presented in Table 5-5.

Duke also considered potential alternatives to reduce containment failure modes of concern for ONS. These alternatives included nine containment-related improvements evaluated as part of the staff's assessment of severe accident mitigation design alternatives for Watts Bar (NRC 1995a) and five containment-related improvements derived from the staff's report on the individual plant examination program (NRC 1997a). Duke eliminated those alternatives that are either (1) not applicable to ONS (e.g., containment air return fans used only in ice condenser containments), (2) related to control of hydrogen combustion (since the Level 2 PRA shows the ONS containment is capable of withstanding large hydrogen burns), or (3) already implemented at ONS, e.g., by inclusion either in emergency operating procedures, severe accident management guidelines, or the operator training program. Based on the screening, Duke designated seven of the containment related SAMAs for further study. The list of the potential enhancements to improve containment performance is presented in Table 5-6.

5.2.3.2 Staff Evaluation

Duke's effort to identify potential SAMAs focused on areas found to be risk-significant in the ONS PRA. The list of SAMAs generally coincide with accident categories that are dominant CDF contributors or with issues that tend to have a large impact on a number of accident sequences at ONS. Duke made

Table 5-5. Value-Impact Results for Potentially Cost-Beneficial SAMAs that Prevent Core Damage

SAMA Sequences/Failures Addressed Percent Reduction Present Worth Cost of Enhancement
CDF(a) P-Rem(b)
Strengthen east and west penetration rooms and BWST(c) to withstand tornado winds Tornado strikes that damage penetration room and BWST 2.1 14.6 74,000 >$1M
Man SSF(d) 24 hours a day with a trained operator Operator failure to align SSF RCM(e) system in events with turbine building fire or failure of Jocassee Dam 1.4 10.0 49,500 >$5M
Install an automatic backup system to refill elevated water storage tank for HPI(f) cooling Operator failure to refill elevated water storage tank during turbine building flood 6.5 8.7 230,000 >$1M
Install automatic swap of HPI to spent fuel pool Operator failure to swap HPI to spent fuel pool during a flood 3.3 4.3 117,000 >$1M
Increase the height of the SSF flood barrier Failures of the Jocassee Dam that result in flood levels exceeding 5-ft flood barrier 2.9 1.6 103,000 $500K
Install protective barrier around upper surge tanks Tornado strikes that cause a LOCA(g) with failure of all power and upper surge tanks 6.0 8.1 212,000 >$1M
Upgrade 4160 switchgear in turbine building to withstand F4 intensity tornadoes Tornado strikes that cause a LOCA with failure of all power and upper surge tanks 6.0 8.1 212,000 >$1M
Install automatic swap from injection to high pressure recirculation Operator failure to initiate high pressure recirculation during LOCAs 4.6 6.3 163,000 >$1M
Replace reactor pressure vessel Spontaneous failure of the reactor vessel 1.1 <0.1 37,100 >$1M
(a) Total CDF = 8.9E-5/year.
(b) Total offsite dose = 4.92 person-rem/year.
(c) BWST = borated water storage tank.
(d) SSF = standby shutdown facility.
(e) RCM = reactor coolant makeup.
(f) HPI = high pressure injection.
(g) LOCA = loss of coolant accident.

Table 5-6. Value-Impact Results for Potentially Cost-Beneficial SAMAs that Improve Containment Performance

SAMA Sequences/Failures Addressed Percent Reduction Present Worth Cost of Enhancement
CDF (a) P-Rem (b)
Install independent containment spray systems Late containment failure from over-temperature or steam over-pressure NA 43.7 46,200 >$1M
Install filtered containment vent system Late containment failure from over-pressure NA 43.7 46,200 >$1M
Install additional containment bypass instrumentation Inter-system LOCAs(c) that could be mitigated through improved detection capabilities NA 25.2 27,300 >$1M
Add independent source of feedwater to reduce induced SGTR(d) Induced steam generator tube failures in high pressure core melt sequences NA 2.8 3100 >$1M
Install reactor depressurization system Direct containment heating and induced steam generator tube failures in high pressure core melt sequences NA 10.9 14,300 >$1M
Install reactor cavity flooding system Basemat failure due to core-concrete interactions NA 6.7 7300 >$1M
Install core retention device Basemat failure due to core-concrete interactions NA 6.7 7300 >$1M
(a) Total CDF = 8.9E-5/year.
(b) Total offsite dose = 4.92 person-rem/year.
(c) LOCA = loss of coolant accident.
(d) SGTR = steam generator tube rupture.

a reasonable effort to use the ONS PRA to search for potential SAMAs and to review insights from other plant-specific risk studies and previous SAMA analyses for potential applicability to ONS. The staff notes that Duke identified a number of recommendations for reducing risk as a result of the ONS IPE and IPEEE, and that many of these plant improvements have been implemented or are planned and being tracked for resolution (Duke 1998c; Duke 1999). For those recommendations that were not implemented, Duke provided justification as to why these improvements are not warranted.

The staff reviewed the set of potential enhancements considered in Duke's SAMA identification process. These include improvements oriented toward reducing the CDF and risk from major contributors specific to ONS, improvements identified as part of the NRC containment performance improvement program, accident management strategies identified by NRC in Generic Letter 88-20, Supplement 2 (NRC 1990b), and improvements identified in the previous severe accident mitigation design alternative review for Watts Bar (NRC 1995a) that would be applicable to ONS. The SAMAs also include a filtered containment vent and a bed-core retention device for flooded rubble, which are cited specifically in NUREG-0660 (NRC 1980) for evaluation as part of Three Mile Island Task Action Plan Item II.B.8.

The staff notes that most of the SAMAs involve major modifications and significant costs and that less expensive design improvements and procedure changes could conceivably provide similar levels of risk reduction. However, lower cost improvements are not expected to offer significant risk reduction, given that external events account for the majority (80 percent) of the risk. Much of this risk is due to postulated earthquakes with ground accelerations significantly greater than the ONS design-basis earthquake. As such, SAMAs that would significantly reduce overall risk would involve substantial upgrades in the seismic ruggedness of the plant and would be very costly.

It should be noted that Duke has made extensive use of PRA methods to gain insights regarding severe accidents at ONS. Risk insights from various ONS risk assessments, such as the ONS IPE, the ONS IPEEE, the Keowee PRA, and ONS HPI reliability study, have been identified and implemented to improve both the design and operation of the plant. For example, using the IPE process, Duke identified and implemented modifications to procedures to (1) isolate the high pressure service water (HPSW) to the condenser circulating water (CCW) pumps during a turbine building flooding event to extend the time the elevated water storage tank (EWST) inventory would last, (2) power the SSF from the Unit 2 main feeder bus, (3) terminate containment sprays to conserve the BWST inventory to enhance long-term HPI cooling following a flooding event in the turbine building, and (4) cope with common cause failure of both HPI suction valves. Examples of plant improvements that resulted from IPEEE findings and whose implementation is being planned by Duke are (1) the mounting of the combustible storage locker near the SSF diesel to prevent combustible materials from being spilled around the diesel during a seismic event or knocked over by personnel, and (2) the replacement of the deluge (open head) sprinklers in the Cable and Equipment Rooms with closed head sprinklers to reduce water damage to equipment important to safety during a fire. The implementation

of such improvements reduced the risk associated with the major contributors identified by the ONS PRA and contributed to the reduced number of candidate SAMAs identified as part of Duke's application for license renewal.

The staff concludes that Duke has used a systematic process for identifying potential design improvements for ONS and that the set of potential design improvements identified by Duke is reasonably comprehensive and, therefore, acceptable.

5.2.4 Risk Reduction Potential of Design Improvements

Section 4.3 of the ER describes the process used by Duke to determine the risk reduction potential for each enhancement.

For each seismic initiator cutset, Duke calculated the associated offsite risk based on the person-rem risk and CDF for the PDSs attributable to the seismic initiator. Implementation of the plant enhancement was assumed to completely eliminate the seismic risk associated with the cutset. For each (non-seismic) sequence/enhancement, Duke assigned a PDS based on the type of plant damage and potential containment release characteristics. In general, where an alternative impacted more than one PDS, Duke used the PDS with the highest conditional person-rem risk to characterize the associated risk and assumed that implementation of the alternative would completely eliminate the risk. For each containment-related improvement, Duke assumed that all of the person-rem risk associated with the release categories impacted by the SAMA would be eliminated. For those alternatives that benefit more than one containment failure mode (i.e., independent containment spray system, reactor depressurization system, and filtered containment vent), the total person-rem dose for all affected failure modes was assumed to be completely eliminated by implementing the alternative.

The staff notes that Duke evaluated the risk reduction potential for each SAMA in a bounding fashion; i.e., each SAMA was assumed to completely eliminate all sequences that the specific enhancement was intended to address. As a result, the benefits are generally over estimated and conservative. Accordingly, the staff based its estimates of averted risk for the various SAMAs on Duke's risk reduction estimates.

5.2.5 Cost Impacts of Candidate Design Improvements

Duke's estimated costs for each potential design enhancement are provided in Tables 4-2 and 5-1 of Attachment K to the ER. For most of the SAMAs, Duke estimated the cost of implementation to be greater than $1 million based on cost estimates developed in previous industry studies. For three SAMAs, Duke developed plant-specific cost estimates because there was no readily available information on the estimated cost to implement similar alternatives and because the basic events associated with these alternatives were found to have a high importance in the ONS PRA. These SAMAs involve (1) increasing the height of the SSF flood barrier, (2) manning the SSF 24 hours a day with trained operators, and (3) installing a protective barrier for the upper surge tanks or upgrading the 4160 volt switchgear to withstand tornado winds. The costs to implement these SAMAs were estimated to be on the order of $500,000, $5 million, and $1 million, respectively. Because the safety benefits of the potential SAMAs were significantly less than their estimated implementation costs (by about a factor of five), none of the cost estimates were further refined.

The staff compared Duke's cost estimates with estimates developed elsewhere for similar improvements, including estimates developed as part of the evaluation of severe accident mitigation design alternatives for operating reactors and advanced LWRs. The staff notes that Duke's estimated implementation costs of $1 million dollars or greater are consistent with the values reported in previous analyses for changes of similar scope and are not unreasonable for the SAMAs under consideration, given that these enhancements involve major hardware changes and impact safety-related systems.

Although the applicant did not provide the underlying bases for its cost estimates, the staff views their cost estimates as reasonable for evaluating the SAMAs because the estimates are consistent with those developed by others and because the spread between the estimated costs and benefits is significant. Accordingly, the staff adopted Duke's cost estimates for the various candidate improvements.

5.2.6 Cost-Benefit Comparison

The following sections describe Duke's cost-benefit comparison and the staff's evaluation of the cost-benefit analysis.

5.2.6.1 Duke Evaluation

In the analysis provided in the ER, Duke did not include the following factors in its cost-benefit evaluation: averted onsite cleanup and decontamination cost, replacement power cost, and averted offsite property damage cost. In view of the significant impact of these averted costs on the estimated benefit for a SAMA, the staff requested that Duke include these factors in their cost-benefit analysis for each affected SAMA. In their response to the request for additional information, Duke updated the benefit estimates to include these factors for all SAMAs that reduce CDF. The methodology used by Duke was based primarily on NRC's guidance for performing cost-benefit analysis, i.e., NUREG/BR-0184, Regulatory Analysis Technical Evaluation Handbook (NRC 1997b), and NUREG/BR-0058, Regulatory Analysis Guidelines of the U.S. Nuclear Regulatory Commission (NRC 1995b). The guidance involves determining the net value for each SAMA according to the following formula:

Net Value = (APE + AOC + AOE + AOSC) - COE

where APE = present value of averted public exposure ($)
AOC = present value of averted offsite property damage costs ($)
AOE = present value of averted occupational exposure ($)
AOSC = present value of averted onsite costs ($)
COE = cost of enhancement ($)

If the net value of a SAMA is negative, the cost of implementing the SAMA is larger than the benefit associated with the SAMA and is not considered beneficial. Duke's derivation of each of the associated costs is summarized below.

Averted Public Exposure (APE)

Averted public exposure costs were calculated using the following formula:

APE = Annual reduction in public exposure (person-rem/reactor-year)
x monetary equivalent of unit dose
x present value conversion factor

Duke estimated the annual reduction in public exposure for each SAMA as discussed previously. The reduction in public exposure (person-rem per year) was converted to a monetary equivalent by applying NRC's conversion factor of $2000 per person-rem and then discounting the monetary equivalent to present value. A 20-year period for the license renewal period and a 7-percent real discount rate was assumed, resulting in a present value conversion factor of 10.76.

As stated in NUREG/BR-0184 (NRC 1997b), it is important to note that the monetary value of the public health risk after discounting does not represent the expected reduction in public health risk due to a single accident. Rather, it is the present value of a stream of potential losses extending over the remaining lifetime (in this case, the renewal period) of the facility. Thus, it reflects the expected annual loss due to a single accident, the possibility that such an accident could occur at any time over the renewal period, and the effect of discounting these potential future losses to present value.

Averted Offsite Property Damage Costs

Averted offsite property damage costs were calculated using the following formula:

AOC = Annual CDF reduction
x offsite economic costs associated with a severe accident (on a per event basis)
x present value conversion factor

Duke determined the offsite economic costs for a severe accident based on the weighted costs for offsite property damage for the five NUREG-1150 plants (reported in Table 5.6 of NUREG/BR-0184). These costs were inflated to year 2000 dollars based on a 4-percent inflation rate, yielding a value of $364 million. Calculated values for offsite economic costs were discounted to present value in the same manner as for averted public exposure.

Averted Occupational Exposure

Averted occupational exposure was calculated using the following formula:

AOE = Annual CDF reduction
x occupational exposure per core-damage event
x present value conversion factor

Duke derived the values for averted occupational exposure based on information provided in Section 5.7.3 of the regulatory analysis handbook (NRC 1997b). Best estimate values provided for immediate occupational dose (3,300 person-rem) and long-term occupational dose (20,000 person-rem over a 10-year cleanup period) were used. The present value of these doses was calculated using equations provided in the handbook in conjunction with a monetary equivalent of unit dose of $2000 per person-rem, a real discount rate of 7 percent, and a time period of 20 years to represent the license renewal period.

Averted Onsite Costs

AOSC includes averted cleanup and decontamination costs and averted power replacement costs. Duke derived the values for AOSC based on information provided in Section 5.7.6 of the regulatory analysis handbook (NRC 1997b).

Averted cleanup costs are calculated using the following formula:

ACC = Annual CDF reduction
x present value of cleanup costs per core-damage event
x present value conversion factor

The net present value for cleanup and decontamination of a severe accident (discounted over 10 years) is given as $1.1 billion in the handbook (NRC 1997b). Use of a discount factor of 10.76 to account for the 20-year license renewal period yields an integrated cleanup cost of $12 billion. This value was multiplied by the annual reduction in core damage frequency to obtain the averted cleanup costs portion of the AOSC.

Long-term replacement power costs (URP) are calculated as

URP = Annual CDF reduction
x present value of replacement power for a single event
x factor to account for remaining service years for which replacement power is required

In accordance with guidance provided in Section 5.7.6.2 of the handbook (NRC 1997b), Duke estimated the net present value of replacement power for a single event to be $1.23 billion, based on a replacement power cost for each ONS unit of $152 million (year 2000 dollars), a real discount rate of 7 percent, and a 20-year license renewal period. This value was multiplied by a factor of 8.1 to obtain a summation of the single-event costs over the entire license renewal period, yielding a replacement power cost of $10.0 billion. This value was multiplied by the annual reduction in core damage frequency to obtain the averted replacement costs portion of the AOSC.

The value-impact results for the 16 SAMAs are presented in Tables 5-5 and 5-6. All of the SAMAs have a negative net value, even when bounding risk reduction benefits are assumed, and AOSC is included. Duke concluded that implementation of SAMAs is not justified since the cost of implementation far exceeds the benefit of these SAMAs. As such, Duke has decided not to pursue any of these SAMAs further.

5.2.6.2 Staff Evaluation

The updated cost-benefit analysis provided by Duke (Duke 1999) was based primarily on NRC's Regulatory Analysis Technical Evaluation Handbook (NRC 1997b). The only noted deviation from the regulatory guidance was the omission of the averted offsite property damage cost component for those SAMAs that impact only containment performance. (A reduction in offsite consequences results in both averted public exposure and averted offsite property damage. Duke appropriately considered averted offsite property damage costs for the SAMAs that prevent core damage, but failed to include these averted costs for the SAMAs that improve containment performance.) The staff has evaluated the averted offsite property damage cost component for these SAMAs and found it to be small (less than $100,000 for the most effective mitigative SAMA identified) and well below the cost of the enhancements. Thus, the total present worth benefit for any of the containment-related SAMAs would be less than $150,000.

The staff concludes that the cost of implementing any of the 16 SAMAs would far exceed the estimated benefit, with a margin of about a factor of five. Based on its review, the staff notes the following:

  • Averted onsite costs are the single most important factor in the cost-benefit analysis. However, no SAMAs are cost-beneficial when these costs are included in the analysis in accordance with NRC's regulatory analysis guidance.
  • Use of a 3-percent discount rate in place of the 7-percent discount rate used in the base case analysis increases net values, but does not lead to identification of any cost-beneficial SAMAs.
  • The effect of implementing the SAMA in the near term rather than delaying implementation until the start of the license renewal period (i.e., use of a 35-year rather than a 20-year period in the value impact analysis) is bounded by the sensitivity study, which assumed a 3-percent discount rate, and does not lead to identification of any cost-beneficial SAMAs.

5.2.7 Conclusions

Duke completed a comprehensive effort to identify and evaluate potential cost-beneficial plant enhancements to reduce the risk associated with severe accidents at ONS. As a result of this assessment, Duke concluded that no additional mitigation alternatives are cost-beneficial and warrant implementation at ONS.

Based on its review of SAMAs for ONS, the staff concurs that none of the candidate SAMAs are cost beneficial. This conclusion is consistent with the low residual level of risk indicated in the ONS PRA and the fact that Duke has already implemented many plant improvements identified from previous plant-specific risk studies. Both the conditional probability of an early release of fission products and the total offsite risk at ONS are already quite small (less than 4 percent and 5 person-rem per year, respectively). External events account for the majority (80 percent) of the risk, with much of this from postulated earthquakes with ground accelerations significantly greater than the ONS design-basis earthquake. Given the low level of residual risk and the large cost of seismic-related enhancements necessary to substantially reduce risk, cost-beneficial enhancements that can significantly reduce risk are unlikely and have not been identified. The margins in the analysis are considered ample to cover uncertainties in risk and cost estimates given that, in general, estimates for these factors were conservatively evaluated.

5.3 References

[ Prev | Top of file ]

10 CFR 51.53, "Postconstruction environmental reports."

10 CFR Part 51, Subpart A, Appendix B, "Environmental effect of renewing the operating license of a nuclear power plant."

Duke Energy Corporation. 1990. Letter from Tuckman, Duke Energy Corporation. Subject: Transmitting Individual Plant Examination. Dated November 30, 1990.

Duke Energy Corporation. 1995. Letter from Hampton, Duke Energy Corporation. Subject: Transmitting Individual Plant Examination for External Events. Dated December 28, 1995.

Duke Energy Corporation. 1997a. Letter from Hampton, Duke Energy Corporation. Subject: Transmitting Oconee Probabilistic Risk Assessment. Dated February 13, 1997.

Duke Energy Corporation. 1997b. Letter from McCollum, Duke Energy Corporation. Subject: Transmitting Oconee Supplemental Individual Plant Examination for External Events Submittal Report. Dated December 18, 1997.

Duke Energy Corporation. 1997c. Letter from McCollum, Duke Energy Corporation. Subject: Transmitting High Pressure Injection Reliability Study. Dated December 18, 1997.

Duke Energy Corporation. 1998a. Application for Renewed Operating Licenses, Oconee Nuclear Station-Units 1, 2, and 3. Volume IV-Environmental Report.

Duke Energy Corporation. 1998b. Final Safety Analysis Report (Oconee Updated FSAR). Charlotte, North Carolina.

Duke Energy Corporation. 1998c. Letter from McCollum, Duke Energy Corporation, Subject: SQUG Resolution of USI A-46 (Generic Letter 87-02) Expected Completion of SQUG Outliers. Oconee Supplemental IPEEE Submittal Report. Dated June 30, 1998.

Duke Energy Corporation. 1999. Letter from M. S. Tuckman, Duke Energy Corporation to U.S. Nuclear Regulatory Commission. Subject: License Renewal - Response to Requests for Additional Information, Oconee Nuclear Station. Dated March 4, 1999.

U.S. Nuclear Regulatory Commission (NRC). 1980. NRC Action Plan Developed As a Result of TMI-2 Accident. NUREG-0660. U.S. Nuclear Regulatory Commission, Washington, D.C.

U.S. Nuclear Regulatory Commission (NRC). 1988. Generic Letter 88-20, "Individual Plant Examination for Severe Accident Vulnerabilities." November 23, 1988.

U.S. Nuclear Regulatory Commission (NRC). 1990a. Severe Accident Risks - An Assessment for Five U.S. Nuclear Power Plants. NUREG-1150. U.S. Nuclear Regulatory Commission, Washington, D.C.

U.S. Nuclear Regulatory Commission (NRC). 1990b. Letter from J. G. Partlow, U.S. NRC, to All Holders of Operating Licenses and Construction Permits for Nuclear Power Reactor Facilities. April 4, 1990. Subject: Accident Management Strategies for Consideration in the Individual Plant Examination Process - Generic Letter 88-20, Supplement No. 2.

U.S. Nuclear Regulatory Commission (NRC). 1993. Weins (NRC) letter dated 4/1/93 transmitting Evaluation of the Oconee 1, 2, and 3 Individual Plant Examination (IPE) - Internal Events.

U.S. Nuclear Regulatory Commission (NRC). 1995a. Final Environmental Statement Related to the Operation of Watts Bar Nuclear Plant Units 1 and 2. NUREG-0498, Supplement 1. U.S. Nuclear Regulatory Commission, Washington, D.C.

U.S. Nuclear Regulatory Commission (NRC). 1995b. Regulatory Analysis Guidelines of the U.S. Nuclear Regulatory Commission. NUREG/BR-0058, Revision 2. U.S. Nuclear Regulatory Commission, Washington, D.C.

U.S. Nuclear Regulatory Commission (NRC). 1996. Generic Environmental Impact Statement for License Renewal of Nuclear Power Plants (GEIS), NUREG-1437. U.S. Nuclear Regulatory Commission, Washington, D.C.

U.S. Nuclear Regulatory Commission (NRC). 1997a. Individual Plant Examination Program: Perspectives on Reactor Safety and Plant Performance. NUREG-1560. U.S. Nuclear Regulatory Commission, Washington, D.C.

U.S. Nuclear Regulatory Commission (NRC). 1997b. Regulatory Analysis Technical Evaluation Handbook. NUREG/BR-0184. U.S. Nuclear Regulatory Commission, Washington, D.C.

U.S. Nuclear Regulatory Commission (NRC). 1998. Letter from NRC to Duke Energy Corporation. Subject: Request for Additional information for the Review of the Oconee Nuclear Station Unit Nos. 1, 2, & 3 Environmental Report Associated with License Renewal - Environmental. Dated December 29, 1998.

[ NRC Home Page | Top of file | Contents ]

Page Last Reviewed/Updated Thursday, March 29, 2012