Fire probabilistic risk assessment applied to the Pebble Bed Modular Reactor

Abstract

A methodology for assessing the frequencies of fire occurrences in the PBMR nuclear power plant is presented. Fire is a concern because it may lead to loss of support systems, which could result in the failure of dependent systems. Ultimately this may lead to the release of radioactive material into the atmosphere. The methodology used is probabilistic in nature. The methodology will be applied to the Pebble Bed Modular Reactor to determine the impact of fire on safety of plant personnel and the public . The methodology described here utilises both a qualitative and quantitative assessment in order to quantify the risk posed by fires in the PBMR reactor . The qualitative steps identify areas where fire is likely to be started . Only fire events that may result in the loss of a safety function are of safety concern. Thus , this process is based on identification of critical locations where there is potential for fire initiation resulting in loss of safety function. The quantitative analysis utilizes an event tree approach to quantify the frequency of fire initiation . An event tree is constructed based on the mitigating actions protection system in each critical area. of the fire This study is meant to be a preliminary assessment of Fire PRA for the PBMR . This task is made difficult by the fact that the plant is still in the design phase. Traditional Fire PRA's have been done on existing power stations with operational histories. To demonstrate the methodology in the PBMR, data from Light Water Reactors (LWR) and Boiling Water Reactors (BWR) was used as a baseline for the Pebble Bed Modular Reactor (PBMR) since there is no historical information on the plant because one has not been built yet. When the plant is built and is in operation the assessment will need to be updated. A brief summary of the study, the conclusion and recommendations for this study are also presented in the last section of the document.