RESEARCH ON FIRE RISK ANALYSIS METHOD FOR SAFETY-LEVEL DCS SYSTEM IN NUCLEAR POWER PLANTS

Abstract

The Distributed Control System (DCS) of the nuclear power plant plays an important role as the operation center that realizes the process control as well as the centralized monitoring and management under all different operating conditions for the entire nuclear power plant. Safe, reliable and effective operation of DCS is considered to be a pivotal guarantee for the design concepts of nuclear safety, e.g., defense in depth of the nuclear power plant, which is very critical to the safety and economy of nuclear power plants. However, the centralized management of DCS leads to a great increase in the possibility that the multilayer preset barriers are destroyed by one local fire in the nuclear power plant, eventually threatening the safety of the power plant. Different from the other systems in the nuclear power plant, the DCS is mainly composed of different components like computers and servers with sophisticated structure, which involves transmission of complex logic signals. These make the fire risk assessment of DCS quite challenging. In this paper, by considering the DCS in the Chinese 2nd-generation improved unit of the pressurized water reactor (PWR), we propose a method of fire risk analysis for DCS, based on probability theory and system reliability evaluation technology. By establishing the logical relationship between the spatial location of fire in the plants and DCS racks, cabinets, optical cables, signals and related equipment, an analysis method is proposed for the possible initiating events after a fire occurs in the DCS-related cabinets. Then a modeling strategy is developed for the impacts of the DCS fire failure on the accident mitigation function. The proposed method of fire risk analysis for DCS is capable of achieving quantitative assessment under all different operating conditions of the entire plant, which, thus, can support comprehensive PSA analysis of the internal fire.