Overview

User of Level 2 Results to Refine Large Early Release Frequency (LERF) for San Onofre Nuclear Generation Station (SONGS) Integrated Leak Rate Test (ILRT) Extension

RSC Partnered with SONGS to do an Extended Test Interval for the ILRT for Unit 3 for the next 16 years. SONGS is a nuclear generating station in Southern California that generates approximately 2,254 megawatts of power, enough to serve 2.75 million average homes.

Challenge

SONGS requested RSC to do an Extended Test Interval for the Integrated Leak Rate Test for Unit 3 for 16 years. A risk evaluation was required using the NEI Guidance and to compare the results to risk metric defined in Regulations Guide 1.174.

• increases of core damage frequency (CDF) below 1E-6/yr
• increases in LERF below 1E-7/yr

Since the ILRT does not impact CDF, the relevant metric is LERF. Calculating the increase in LERF requires determining the impact of the ILRT interval on the leakage probability.

Initial Results

The Change Exceeded the Limits Defined by the Regulation Guide. This was expected due to comments from NEI that indicated:

• Plants with CDF in excess of 1.0E-5/yr may have difficulty demonstrating a change in LERF less than 1.0E-7/yr.
• It further states that the analysis as embodied in the NEI approach is conservative and provides additional guidance with respect to refining the initial analysis.

Process

RSC refined the LERF frequency using NEI guidance. Other causes that had an effect on the LERF were defined. The NEI approach utilized total CDF. Scrubbed releases were defined. Then using the CET logic, these contributors were removed from consideration and the analysis.

The results were still above the target, making further evaluation needed. As result, an additional evaluation of the contributors associated with this LERF contribution was performed. The controlling accident sequence was defined as a Small LOCA using the CET information. A significant portion of this contribution dealt with the failure of the sump suction due to a debris clogging. This information was used to further refine the contribution to CDF.

The accident sequence progression was determined with regard to timing. The time at which the core is uncovered was examined. Typically, early releases involve fairly rapid voiding of the core and results in a significant quantity of radionuclides being present in the containment and available for early release. Early release timing is typically associated with a time sufficiently short that there is an impaired ability to evacuate individuals near the plant such that the fatality could be possible. For this assessment, the breakpoint is chosen as 4 hours.

Final Results

• The removal of both release categories reduced the contribution to releases.
• The total increase from the baseline (3 year) to the proposed (16 year) interval met the definition for an acceptably small change in risk.

Summary

Conservative grouping can result in over-estimation of LERF contribution which could influence results.

A combined approach that addresses frequency, source term and timing provides the best means for refinement.

This approach could be employed to other analyses and to the Level 2 model to better estimate the baseline LERF contribution.

Return to Menu