ML011980485
| ML011980485 | |
| Person / Time | |
|---|---|
| Issue date: | 02/20/2002 |
| From: | Meserve R NRC/Chairman |
| To: | Colvin J Nuclear Energy Institute |
| Whitney, L E, NRR/DIPM/IIPB, 415-3081 | |
| References | |
| Download: ML011980485 (4) | |
Text
February 20, 2002 Mr. Joe F. Colvin, President and Chief Executive Officer Nuclear Energy Institute 1776 I Street NW, Suite 400 Washington, D.C. 20006
Dear Mr. Colvin:
I am responding on behalf of the U.S. Nuclear Regulatory Commission (NRC) to advise you of the results of the joint NRC/industry effort to develop and test replacement performance indicators (PIs) for the existing unplanned scrams per 7,000 critical hours and unplanned scrams with loss of normal heat removal PIs. This effort was undertaken in response to industry concerns regarding the potential for the existing scram PIs to result in unintended consequences, as described in your letter of May 19, 2000.
As described in my letter to you of June 22, 2000, the Commission took industry concerns very seriously and directed the staff to explore alternatives that would mitigate the adverse effects of the scram PIs. As a result, the NRC took four actions to address your concerns: working with industry and other stakeholders to define proposed replacement PIs, conducting a six-month pilot test at selected reactor sites, evaluating the test data against specific criteria, and notifying all stakeholders of the results in accordance with the NRCs formal PI change process (Inspection Manual Chapter 0608).
The pilot program was completed in March 2001. The NRC discussed the pilot test results with industry and other external stakeholders at several joint NRC/industry Reactor Oversight Process (ROP) Working Group Meetings. Additionally, we have taken into consideration your letter of September 10, 2001, which indicates that the NEI Executive Committee preferred the proposed alternative indicators, but had confidence in the NRCs established procedure for processing changes to the ROP.
Based on the ROP Working Groups review of the pilot test results against the pre-established criteria, the NRC has decided to retain the existing scram PIs in the ROP for the time being until another suitable alternative is developed. In reaching this conclusion, the NRC noted that the proposed alternative PIs would have missed nearly 15 percent of all reactor scrams, which makes them unacceptable alternatives. A summary of the pilot test results is enclosed. This summary will form the basis for a pending Regulatory Information Summary announcing this decision.
I thank you for your interest and efforts regarding the scram PIs. As you know, the Commission is eager to identify ways to improve the oversight process to enhance its efficiency and effectiveness. In this vein, the staff plans to engage the industry at an upcoming ROP Working Group meeting concerning another possible alternative to the current scram PIs. In light of the human factors concerns regarding the current versions of the scram PIs, we remain open to the consideration of appropriate alternatives and will work with industry to mitigate the potential for unintended consequences.
If you have any further concerns, please feel free to contact me.
Sincerely,
/RA/
Richard A. Meserve
Enclosure:
Summary of Replacement Scram Performance Indicator Pilot Results
SUMMARY
OF REPLACEMENT SCRAM PERFORMANCE INDICATOR PILOT RESULTS In March 2000, nuclear industry representatives expressed concern with including the phrase manual scram in two scram-related performance indicators (PIs), because this terminology could possibly send conflicting messages to licensee managers and operations personnel, and potentially result in non-conservative decision-making during a plant event for which a manual scram may be warranted. The NRC agreed to pilot test proposed replacement scram PIs in accordance with its formal process for addressing questions and feedback from internal and external stakeholders. The purpose of the pilot test was to determine if the candidate replacement PIs were as effective as the existing PIs at providing an indication of performance in the initiating events cornerstone, while at the same time reducing the potential for unintended consequences.
As described in Regulatory Information Summary (RIS) 2000-21 of October 31, 2000, Changes to the Unplanned Scram and Unplanned Scram with Loss of Normal Heat Removal Performance Indicators, the NRC conducted a six month pilot test of Unplanned Reactor Shutdowns per 7,000 Critical Hours and Unplanned Reactor Shutdowns with Loss of Normal Heat Removal candidate replacement PIs for the two existing scram PIs.
RIS 2000-21 provided five criteria that the NRC planned to consider in reaching its determination regarding the efficacy of the proposed PIs:
1.
comparability of the data collected for the Unplanned Reactor Shutdowns per 7,000 Critical Hours and Unplanned Scrams per 7,000 Critical Hours PIs; 2.
comparability of the data reported for the Unplanned Reactor Shutdowns with Loss of Normal Heat Removal and the Unplanned Scrams with Loss of Normal Heat Removal PIs. Additionally, the rate of occurrence of Unplanned Reactor Shutdowns with Loss of Normal Heat Removal and the results presented in NUREG/CR-5750, Rates of Initiating Events at U.S. Nuclear Power Plants 1987 - 1995, Sections: Loss of Feedwater and Loss of Heat Sink Events, were compared to identify differences; 3.
the ability of licensees to report the requested data accurately and with minimal need for clarification; 4.
the ability of each alternative PI to reduce the potential for unintended consequences without introducing other unintended consequences; 5.
the impact of reporting burden on licensees.
No performance thresholds were applied to the data reported in the pilot test since the pilot plants continued to be assessed using the previously established scram PIs.
At the completion of the six-month trial, the joint NRC/Industry ROP Working Group evaluated the pilot program results against these five criteria and concluded that:
1.
there were no differences between the data collected in the pilot program for the Unplanned Reactor Shutdowns per 7,000 Critical Hours PI and Unplanned Scrams per 7,000 Critical Hours PI; the same eight scrams were reported by the thirteen pilot plants under each indicator. However, the NRC staff also reviewed the Licensee Event Report (LER) database for all scrams reported by the industry in the year 2000 to identify any events that would likely not have been reported under the alternative PIs had they occurred during the trial. Thirteen such scrams were found (about 14 percent of the 2.
92 scrams that year). As a result, the NRC judged that full implementation of the replacement PI as proposed would likely not have captured all reactor scrams. Of the thirteen scrams in the year 2000 that might have been missed, a relatively simple clarification to the guidance would resolve six of them. However, no simple and effective clarification could be developed for the other seven. Therefore, no alternative replacement PIs were suggested.
3.
from NUREG/CR-5750, the value for the total number of scrams with loss of normal heat removal expected to occur in the pilot plant population would be 2.42 events in six months. Two such events were reported under the pilot program (indicating good correlation), though none were reported under the existing ROP guidance. Those two events captured in the pilot program should also have been captured in the existing PI data. Early in the first year of implementation, and before the beginning of the pilot program, the NRC/Industry Working Group became aware of problems in the definition of loss of normal heat removal in the existing guidance. Consequently, a revised definition was developed and tested during the pilot program, along with a revised definition of scrams. The pilot results confirmed that the revised loss of normal heat removal guidance in the pilot program should be included in the guidance regardless of which PI is employed.
4.
the ability of licensees to report the replacement PI data accurately and with minimal need for clarification was not established during the pilot program. As described in paragraph 1 above, the staff concluded that the proposed replacement indicator would likely not provide the same data that the existing PIs provide.
5.
the proposed alternative PIs did not demonstrate the ability to reduce the potential for creating unintended consequences below those which might be postulated using the existing scram PIs. For example, it has been suggested that continued use of the current scram PIs may result in operators not initiating a manual scram when needed to avoid a PI hit, which would be counter to safety-conscious operation. However, the proposed replacement PIs do not resolve this concern.
6.
no additional regulatory burden was reported by the pilot plants in reporting the proposed replacement indicators. However, as described in paragraph 1 above, at least two clarifications would be needed to ensure that accurate data are reported. While one clarification would be straightforward, the other would not. Requiring more complex clarification could add to licensee reporting burden by necessitating more interaction with the staff to ensure data are reported correctly.
After evaluating the pilot program experience against the five criteria in RIS 2000-21, the NRC decided to retain the existing scram PIs in the ROP. This conclusion was reached largely because of the likelihood that a significant percentage of reactor scrams would be missed.