Studies Of Severe Accident Management Optimization For Pressurized Water Reactor

Nuclear safety is a prerequisite for the development of nuclear energy.After the Fukushima Daiichi nuclear accident in Japan,studies on the Severe Accident Management Guideline(SAMG)of nuclear power plants and the station blackout(SBO)have attracted people’s attention.When an accident occurs in a nuclear power plant,mitigation strategies must be adopted to avoid the external release of radioactive materials.At the same time,according to the process of accident,it is important to predict the time of entering SAMG more timely and accurate,which can improve the feasibility and effectiveness of the mitigation strategies.Therefore,finding the optimal mitigation strategy and predicting the SAMG entry time more timely and accurately in the event of a serious accident is the research focus in the field of severe accident management.In this paper,the optimal mitigation strategy of PWR nuclear power plant can be analyzed by the Modular Accident Analysis Program 4(MAAP4).Firstly,the basic process of the SBO accident sequence and the influence of depressurization and injection strategy are analyzed.It is found that different depressurization strategies have different influences on accident mitigation.Secondly,an optimization platform is developed by Downhill Simplex Method and MAAP4,which realize the automatic search of the optimized mitigation strategy.Finally,the optimization platform is used to search for the optimized depressurization under the accident sequence of highpressure,and the enumeration method is used to verify it.It is shown that the optimization platform can not only get correct optimized depressurization strategy,but also improve the efficiency by about 10 times,compared with the traditional enumeration algorithm.In addition,the effect of severe accident management strategy is related to the execution time.Therefore,the efficiency of executing the mitigation strategy can be improved by the accurate prediction of the entry time of severe accident.The maximum temperature of cladding is more related to the condition of core,but it is difficult to obtain it through monitoring.In this study,the neural network is used to predict the maximum temperature of cladding by other measurable parameters in the nuclear power plant.And the maximum temperature of cladding reaching 800℃ is selected as the SAMG entry condition to analyze the severe accident.The neural network model can also predict the maximum temperature of the cladding even though the type of accident is unknown.It is shown that selecting the maximum temperature of cladding as the entry condition of SAMG can delay the reactor pressure vessel failure time more efficiently and reduce the release of radioactive materials,when different entry conditions of SAMG are adopted.