Probabilistic Safety Assessment Of A High Flux Engineering Test Reactor

Probabilistic Safety Assessment(PSA) has got more and more attention because of its ability to quantify the risk to public, with more people concern about the nuclear safety issues. It has proved to be a valuable and essential tool for improving the safety and reliability of nuclear reactors. Meanwhile, PSA methodology has become a proven technology as its decades application in such aspects as design, operation, maintenance and decommission of nuclear power reactors。Furthermore, as the birthplace and pioneer of PSA, the United States has taken it as one of the requirements for applying for licenses at the level of laws and regulations; PSA for nuclear power plants is also the requirement in our nuclear safety regulations.PSA may also be applied advantageously to research reactors under design, construction, or in operation, particularly those with long operating histories and therefore subject to ageing. Research reactors are usually arc simpler than nuclear power plants and the fuel mass, operating power level and conditions are also significant different from that of nuclear power plants. Furthermore, it is important to recognize that the core configurations in research reactor may be changed very often and more human actions are involved in various operating conditions due to experimental requirements. Therefore human factors play a more important safety role in research reactors.This paper represents the level 1 PSA program conducted for High Flux Engineering Test Reactor(HFETR) in power level through reference to general PSA method for nuclear power plants and results of similar research reactors PSA work, so as to estimate the total core damage frequency(CDF) and dominant initiating events and sequences to evaluate the safety of HFETR. The scope and strategy of this process is internal initiating events as well as loss of offsite power. A combination of several standard method arc used at the same time in the identification and selection of initiating events to give reasonable assurance of completeness of initiating events list, and some state-of-the-art generic data and frequencies of similar research reactors are used to quantify these initiating events. Event trees have been used to study the response of the installation to various initiating events whereas fault trees have been used in the modeling of safety system failures. ASEP and SPAR-H methods arc applied respectively to evaluate the pre-accident and post-accident human errors. After quantifying the PSA model with generic data, the total CDF caused by internal events in HFETR is 6.01E-7/yr, and the initiating events which have significant contribution to the CDF are loss of coolant water and loss of offsite power.