A SCOPING STUDY: Development of Probabilistic Risk Assessment Models for Reactivity Insertion Accidents During Shutdown In U.S. Commercial Light Water Reactors

S. Khericha

关键词: ACCIDENTS;    BOILING;    CONTROL ELEMENTS;    CONTROL ROOMS;    NUCLEAR FUELS;    NUCLEAR POWER PLANTS;    REACTIVITY INSERTIONS;    RECOMMENDATIONS;    RISK ASSESSMENT;    SHUTDOWN;    TREES;    VALVES;    WATER Shutdown;    PRA;    Reactivity Insertion;    Boron Dilutio;   

DOI  :  10.2172/1027939
RP-ID  :  INL/EXT-11-22523
PID  :  OSTI ID: 1027939
Others  :  TRN: US1105906

美国|英语

来源: SciTech Connect

【 摘 要 】

This report documents the scoping study of developing generic simplified fuel damage risk models for quantitative analysis from inadvertent reactivity insertion events during shutdown (SD) in light water pressurized and boiling water reactors. In the past, nuclear fuel reactivity accidents have been analyzed both mainly deterministically and probabilistically for at-power and SD operations of nuclear power plants (NPPs). Since then, many NPPs had power up-rates and longer refueling intervals, which resulted in fuel configurations that may potentially respond differently (in an undesirable way) to reactivity accidents. Also, as shown in a recent event, several inadvertent operator actions caused potential nuclear fuel reactivity insertion accident during SD operations. The set inadvertent operator actions are likely to be plant- and operation-state specific and could lead to accident sequences. This study is an outcome of the concern which arose after the inadvertent withdrawal of control rods at Dresden Unit 3 in 2008 due to operator actions in the plant inadvertently three control rods were withdrawn from the reactor without knowledge of the main control room operator. The purpose of this Standardized Plant Analysis Risk (SPAR) Model development project is to develop simplified SPAR Models that can be used by staff analysts to perform risk analyses of operating events and/or conditions occurring during SD operation. These types of accident scenarios are dominated by the operator actions, (e.g., misalignment of valves, failure to follow procedures and errors of commissions). Human error probabilities specific to this model were assessed using the methodology developed for SPAR model human error evaluations. The event trees, fault trees, basic event data and data sources for the model are provided in the report. The end state is defined as the reactor becomes critical. The scoping study includes a brief literature search/review of historical events, developments of a small set of comprehensive event trees and fault trees and recommendation for future work.