Index | Site Map | FAQ | Facility Info | Reading Rm | New | Help | Glossary | Contact Us |
POLICY ISSUE SECY-02-0041 March 8, 2002
To inform the Commission of the status of the Accident Sequence Precursor (ASP) Program and the development of the Standardized Plant Analysis Risk (SPAR) models that support this programs. In an April 24, 1992, memorandum to the Chairman, the staff committed to report periodically to the Commission on the status of efforts to improve the ASP Program. In SECY-94-268, dated October 31, 1994, two changes were made to the report. First, the staff committed to provide the report annually, and second, the staff began providing annual quantitative ASP results. Since the last status report, SECY-00-034, dated March 1, 2001, the staff has:
The discussion below provides a brief background of the ASP and SPAR Model Development Programs and their uses. ASP Program. The ASP Program was established by the NRC in 1979 in response to the Risk Assessment Review Group report (see NUREG/CR-0400, September 1978). The primary objective of the ASP Program is to systematically evaluate U.S. nuclear plant operating experience to identify, document, and rank operating events most likely to lead to inadequate core cooling and severe core damage (precursors), if additional failures had occurred. The secondary objectives of the ASP Program are
The program is also used to monitor the agency's performance against the following Strategic Plan performance goals in the Reactor Safety arena for maintaining safety (see Volume 2, Part 1, of NUREG-1614, September 2000):
Events and conditions from licensee event reports, inspection reports, and special requests from NRC staff are reviewed for potential precursors. These potential precursors are analyzed, and a conditional core damage probability (CCDP) is calculated by mapping failures observed during the event onto accident sequences in risk models. An event with a CCDP or a condition with a change in core damage probability (CDP or importance) greater than or equal to 1.0 x 10-6 is considered a precursor in the ASP Program. Current ASP analyses do not explicitly include uncertainties for calculating CCDPs. Revision 3 of the SPAR models includes uncertainty analysis capability for future ASP analyses. Sensitivity analyses are used to assess the impact of uncertainties in key parameters that could influence the characterization of an event as a precursor (CCDP or CDP > 1.0 x 10-6), important precursor (CCDP or CDP > 1.0 x 10-4) or significant precursor (CCDP or CDP > 1.0 x 10-3). SPAR Model Development Program. SPAR models are the analysis tool used by staff analysts in many regulatory activities, including the ASP Program and Phase 3 of the Significance Determination Process (SDP). The SPAR models have evolved from two sets of simplified event trees that were used initially to perform precursor analyses in the early 1980s. One set of event trees was used for boiling-water reactors (BWRs) and one set for pressurized-water reactors (PWRs) to model plant response to the same set of initiating events. In 1985, the models were improved to better reflect the operating plant population by using eight plant classes of event trees that reflected differences among the plant designs. In 1992, the staff conducted a workshop of internal and external stakeholders to discuss the methods, models, and data needs of the program to better meet the program's objectives. The results and recommendations generated at this meeting were documented in NUREG/CP-0124 (June 1992). In response to the recommendations, the staff prepared the Integrated ASP Program Plan (Enclosure 1 to SECY-94-076, March 22, 1994) for improving the models used for ASP analyses. Among the improvements identified was the development of Level 1, Revision 1 SPAR models, which included plant-specific fault trees and other improvements to the existing event tree-based models. Another improvement was the development of feasibility studies for specialized models, such as low power or shutdown, external events, and Level 2 models. In 1998, Revision "2QA" SPAR models were issued that more closely reflected plant-specific risk based on comments generated by peer reviews of precursor analyses and models. In 1999, the SPAR Model Users Group (SMUG) assumed coordination of model development efforts that support the ASP Program and other risk-informed regulatory processes. This group is comprised of representatives from the Offices of Nuclear Reactor Regulation and Nuclear Regulatory Research, and regional offices that use reactor plant risk models in regulatory activities. In August 2000, the SMUG completed the SPAR Model Development Plan, which conforms to the modeling needs identified by SMUG members and their management for performing risk-informed regulatory activities. The following models are addressed in this plan:
Uses. The NRC staff uses the ASP methodology, SPAR models, and results of ASP analyses to perform the following risk-informed regulatory activities:
This section provides summaries of historical trends and insights, documentation and issuance of ASP results, and status of SPAR model development. Historical Trends and Insights A review of the ASP analyses(2) for FYs 1999, 2000, and 2001 and a comparison with analyses from previous years for insights and trends are summarized in the Attachment. The ASP results used to monitor the agency's performance against the two Strategic Plan goals are as follows: Significant Precursors. The Strategic Plan performance goal is "No more than one event per year identified as a significant precursor of a nuclear accident." A significant precursor is defined in the Strategic Plan as an event that has a 1/1000 (10-3) or greater probability of leading to a reactor accident. No significant precursors were identified during FYs 1999, 2000, and 2001(3) with a CCDP or CDP > 1.0 × 10-3. Significant precursors have occurred, on the average, about once every 4 to 5 years. The events in the significant precursor group involve differing failure modes, causes, and systems. Industry Trends. The occurrence rate of precursors has exhibited a decreasing trend that is almost statistically significant during the 1993-2000 period (see Figure 1 of the Attachment). The occurrence rate of precursors has decreased over the period by about a factor of 2. Documentation and Issuance of ASP Results There are 17 reports in the NUREG/CR series documenting the results of the ASP Program covering the years 1969-1998. The precursor report for FY 1999 was completed and sent to publications in September 2001. However, the report was subsequently withdrawn from distribution because of the agency's heightened awareness of the release of sensitive information to the public following the September 11th terrorist attacks. In addition, preliminary and final precursor analyses are not yet being issued to licensees or released to the public pending Agency guidance (in response to the Staff Requirements Memorandum, "Staff Requirements - COMSECY-01-0030 - Guidance to the Staff on Release of Information to the Public," dated January 25, 2002) and office procedures for classifying and transmitting sensitive information involving results of risk analyses. Staff is working to resolve this issue as soon as possible. Trends in ASP events are also displayed on the NRC web page maintained by NRR for the industry trends program. The ASP trends on this page will be updated after this status report is issued. SPAR Model Development The SPAR Model Users Group (SMUG) is comprised of representatives from each of the organizations within the agency's program and regional offices that use risk models in regulatory activities. The SMUG meets on a regular basis to provide technical guidance for the SPAR Model Development Program consistent with the approved Integrated SPAR Model Development Plan. This plan conforms to the modeling needs identified by the SMUG members and their management for performing risk-informed regulatory activities. In the last year, the staff completed the following activities in model and methods development: Level 1: Internal events during full power operation
Level 1: Internal events during low-power and shutdown operations
Level 2: Large early release frequency (LERF)
Results, Trends, and Insights from the ASP Program This section describes the status of Accident Sequence Precursor (ASP) analysis of Fiscal Year (FY) 1999, 2000, and 2001 events, the analysis of historical Accident Sequence Precursor (ASP) trends, and the evaluation of insights. 1.0 ASP Event Analyses The status of analyses is provided in Table 1. The results are summarized below. FY 1999 analyses. The final analyses of events in FY 1999 were completed in September 2001. Peer review comments were received from the respective licensee and NRC staff and incorporated into the final analysis. Seven of the initiating events and conditions that occurred during FY 1999 had a conditional core damage probability (CCDP) or a change in core damage probability (CDP or importance) >1 × 10-6. Of these seven precursors, all were at-power precursors. The results of final ASP analyses for FY 1999 are presented in Table 2 for precursors involving initiating events and Table 3 for precursors involving conditions. FY 2000 analyses. The screening, review, and preliminary analysis of events in FY 2000 were completed in February 2002. The final analysis of four precursors were completed. The issuance of preliminary analyses for licensee peer review was delayed because of the agency's heightened awareness of the release of sensitive information to the public following the September 11th terrorist attacks. This will delay the completion of the remaining final analyses. The ASP analyses identified 14 precursors for FY 2000 (4 precursors based on final analyses, 10 precursors based on preliminary analyses). Of these 14 precursors, all were at-power precursors. The results of ASP analyses for FY 2000 are presented in Table 4 for precursors involving initiating events and Table 5 for precursors involving conditions. FY 2001 analyses. The screening and review of licensee event reports (LERs) of events in FY 2001 were completed in January 2002. The preliminary analyses are ongoing. A combination of LERs and daily event notification reports (10 CFR 50.72) were reviewed to identify potential significant precursors (CCDP or CDP > 1 × 10-3). No such events were identified for FY 2001. 2.0 Overall Industry Trends This section provides the results of trending analyses for all precursors and for precursors grouped by the order of magnitude of their CCDP or CDP (called CCDP bins). Note: The trending results cover the FY 1993-2000 period. Although most of the results for FY 2000 are preliminary, all preliminary analyses have undergone in-house review. However, FY 2000 results are subject to change when they become finalized. Results for FY 2001 are not included in the trends because the preliminary analyses of all events are not complete. 2.1 Occurrence Rate of Precursors The 1993-2000 occurrence rate of all ASP precursors are trended by fiscal year. The trend plot is shown in Figure 1. The mean occurrence rate of precursors has exhibited a decreasing trend that is almost statistically significant. The occurrence rate of precursors has decreased over the period by a factor of 2. 2.2 Occurrence Rate of Precursors by CCDP Bins In addition to the occurrence rate of all precursors, the data were analyzed to determine whether trends exist in the occurrence rate of precursors with CCDP of different orders of magnitude. The analysis method is based on a staff technical paper reported in Reference 1. The objective of this analysis was to determine whether a trend exists in the ASP precursor bin data. The results of the trending analysis of the four probability bins (>10-3, 10-4, 10-5, 10-6) for the FY 1993-2000 period are as follows:
A histogram of the occurrence rate as a function of fiscal year for each probability bin is provided in Figures 2a-d. The trend line of the mean occurrence rate (with the 90% confidence band) is shown in the two figures with statistically significant trends. No trend line is shown when a statistically significant trend was not detected. 3.0 Insights The number of significant precursors presented in Section 3.1 cover the FY 1993-2001 period. The result for FY 2001 is based on the screening and review of a combination of LERs and daily event notification reports (10 CFR 50.72). The insights presented in the remaining sections cover the FY 1993-2000 period. Although most of the results for FY 2000 are preliminary, all preliminary analyses have undergone in-house review. However, FY 2000 results are subject to change when they become finalized. 3.1 Significant Precursors The ASP Program is used to monitor the agency's performance against the following Strategic Plan performance goal: "No more than one event per year identified as a significant precursor of a nuclear accident." A significant precursor is defined in the Strategic Plan as an event that has a 1/1000 (10-3) or greater probability of leading to a reactor accident (Ref. 2). No significant precursors were identified during FYs 1999, 2000, and 2001 with a CCDP or CDP > 1.0 × 10-3. Significant precursors have occurred, on the average, about once every 4 to 5 years. The events in this group involve differing failure modes, causes, and systems. Two precursors with a CCDP > 1.0 × 10-3 have occurred since 1991--the Wolf Creek event involving a reactor coolant system draindown to the refueling water storage tank during hot shutdown (1994) and the Catawba 2 event involving an extended plant-centered loss of offsite power with an emergency diesel generator out of service for maintenance (1996). 3.2 Important Precursors Precursors with a CCDP or CDP > 1.0 × 10-4 are considered important in the ASP Program. There were no important precursors in FY 1999. Two important precursors were identified in FY 2000--the Cook 1 and 2 conditions involving potential high-energy line breaks. The review of the ASP data reveals the following:
3.3 Initiating Events vs. Conditions A precursor can be the result of an operational event involving an actual initiating event (e.g., loss of offsite power) or a precursor can be a condition found during a test, inspection, or engineering evaluation. A condition involves a reduction in safety system reliability or function for a specific duration (and no initiator actually occurred during this time). Five of the seven precursors in FY 1999 involved conditions (unavailability of equipment); two involved initiating events. In FY 2000, 12 of the 14 precursors involved conditions; two involved initiating events. A review of the data reveals the following:
3.4 Precursors Involving Loss of Offsite Power Initiating Events One precursor involving a loss of offsite power (LOOP) initiating event occurred in both FY 1999 and FY 2000--the Indian Point 2 loss of offsite power to safety-related buses following a reactor trip (1999) and the Diablo Canyon 1 extended loss of offsite power (2000). The review of the ASP data reveals the following:
3.5 Precursors at BWRs vs. PWRs None of the precursors in FY 1999 and FY 2000 occurred at a boiling-water reactor (BWR); there has been only one such precursor since 1996. A review of the data for BWRs and pressurized-water reactors (PWRs) during the FY 1993-2000 period reveals the following:
The slight increase in the number of precursors in FY 2000 (14 precursors) over the historical average was attributed in part to conditions related to design-basis issues affecting multi-unit plants--one condition affecting all three units at Oconee and two conditions affecting both units at Cook. According to the staff's review of individual plant examinations (NUREG-1560, Ref. 3), the core damage frequencies estimated in the individual plant examinations were generally lower for BWRs than for PWRs. NUREG-1560 attributed the difference to the larger number of injection systems in the BWR design along with the ability to rapidly depressurize to allow the use of low-pressure injection systems. This may explain, in part, the lower number of precursors at BWRs. 3.6 Precursors Caused by Unavailability of Safety-Related Equipment Most precursors involve the unavailability of safety-related equipment. These events occur during periods of extended unavailability of equipment without a reactor trip, or in combination with a reactor trip in which a risk-important component is unable to perform its safety function due to an unavailability condition. A review of the ASP data during the FY 1993-2000 time period produced the following insights about the unavailability of safety-related equipment. Equipment unavailabilities in BWRs
Emergency core cooling systems (ECCS)
Most of these unavailabilities were caused by failures in the ECCS (33%), emergency power source (27%), and design-basis issues of other systems/structures impacting the ECCS or an ECCS support system (25%). - Sixteen precursors involved a conditional unavailability found during testing, inspection, or engineering reviews. This indicates that conditions are being identified prior to unplanned demand of the degraded ECCS. - Eleven precursors involved a condition affecting sump recirculation during postulated loss-of-coolant accidents of varying break sizes. Auxiliary/emergency feedwater systems (AFW/EFW)
Most of these unavailabilities were caused by failures in the AFW/EFW system (29%), emergency power source (43%), and design-basis issues of other systems/structures impacting AFW/EFW or an AFW/EFW support system (26%). - Five of the train failures occurred following a reactor trip. - Eight of the precursors involved the unavailability of the turbine-driven AFW/EFW pump train. Emergency power sources in PWRs
Most of these unavailabilities were caused by failures in the emergency power system (75%). 3.7 Annual ASP Index The annual ASP Index is derived for order of magnitude comparisons with industry average core damage frequency (CDF) estimates derived from probabilistic risk assessments (PRAs) and individual plant examinations (IPEs). The index for a fiscal year is the sum of the conditional core damage probabilities (CCDPs) divided by the number of reactor years (RYs). Limitations. Using CCDPs from ASP results to estimate CDF is difficult because (1) the mathematical relationship requires a great level of detail, (2) statistics for frequency of occurrence of specific precursor events are sparse, and (3) events and conditions that did meet ASP precursor criteria also need to be accounted for in the assessment. The ASP models and process do not explicitly cover all core damage frequency scenarios, such as fires, flooding, and external events, and are therefore incomplete for estimating total CDF. Also, using CCDP to estimate CDF can overestimate the frequency because of double counting. Because of these and other limitations, the CCDPs have been used primarily as a relative trending indication. Despite these limitations, ASP results can be linked to CDF by using an Annual ASP Index. The IPEs also give incomplete estimates of total CDF, but IPEs are reasonably similar in scope to the current ASP Program. Results. For the last eight fiscal years, the annual ASP index is as follows:
The estimated CDFs in the IPEs range from 1.2 x 10-6/RY to 3.7 x 10-4/RY, with an average value of 6.2 x 10-5/RY. On an order of magnitude basis, the ASP Index over the last eight fiscal years is consistent with the order of magnitude of estimates of CDFs from the IPEs. However, because of the limitations discussed above, the ASP results are not sufficient to verify the IPE results. 3.8 Consistency with PRAs/IPEs Most of the precursor events are consistent with failure combinations identified in probabilistic risk assessments (PRAs) and individual plant examinations (IPEs). A review of the precursor events for the period 1994-2000 shows that 19% of the precursors involved event initiators or conditional availability of equipment not typically modeled in PRAs or IPEs. Precursors not typically modeled in PRAs and IPEs are listed in Table 6. 4.0 References 1. Rasmuson, D.M., and P.D. O'Reilly. "Analysis of Annual Accident Sequence Precursor Occurrence Rates for 1984-94," in Proceeding of the International Topical Meeting on Probabilistic Safety Assessment. American Nuclear Society (ANS), Park City, Utah. 29 September-3 October 1996. Vol. III, pp. 1645-1652. ANS: LaGrange Park, Illinois. 1994. 2. U.S. Nuclear Regulatory Commission. NUREG-1614, Vol. 2, Part 1, "Strategic Plan, Fiscal Year 2000 - Fiscal Year 2005." NRC: Washington, D.C. September 2001. 3. U.S. Nuclear Regulatory Commission. NUREG-1560, "Individual Plant Examination Program: Perspectives on Reactor Safety and Plant Performance, Summary Report." NRC: Washington, D.C. October 1996. Table 1. Status of ASP analysis as of 02/02/02.
Table 2. FY 1999 at-power precursors involving initiating events.
Table 3. FY 1999 at-power precursors involving conditional unavailabilities.
Table 4. FY 2000 at-power precursors involving initiating events as of 02/02/02. (Preliminary results are subject to change.)
Table 5. FY 2000 at-power precursors involving conditional unavailabilities as of 02/02/02. (Preliminary results are subject to change.)
Table 6. Precursors not typically modeled in PRAs or IPEs.
1. One industry trend indicator used in this performance measure was provided by the ASP Program for FY 2000. The Industry Trends Program is discussed in SECY-01-0111, "Development of an Industry Trends Program for Operating Power Reactors." 2. Results for FY 1999 are based on final analyses. Results for FY 2000 are based on a combination of final and preliminary analyses. The preliminary analysis of events in FY 2001 is ongoing. 3. Although, preliminary analyses of events in FY 2001 are ongoing, the screening and review of FY 2001 events for potential significant precursors were completed.
|
Privacy Policy |
Site Disclaimer |