非能动系统物理过程失效分析方法研究

发布时间：2018-03-29 02:26

本文选题：AP1000　切入点：非能动系统　出处：《华北电力大学(北京)》2016年硕士论文

【摘要】：在二代、二代加、三代AP1000和华龙一号的压水堆中均已引入非能动安全系统,以保障事故条件下的安全。但是,非能动系统也有相应的失效模式。非能动系统的失效应包含常规的部件失效以及物理过程的失效,目前,针对部件失效的研究较多,而针对物理过程失效的研究还比较欠缺,所以针对物理过程失效的研究具有重要的理论意义和应用价值。首先,引入拉丁超立方抽样(LHS),并结合蒙特卡洛模拟,建立功能可靠性分析(FRPS)方法。利用其研究了AP1000非能动余热排出系统在丧失正常给水事故时,不同的初始参数组合对于输出参数的影响。其主要的计算流程为：系统功能分析及关键参数选取；关键参数分布量化；抽样得到输入向量；运行Relap5得到输出；结合输入输出以及失效准则建立响应面；最后用蒙特卡洛方法进行概率计算。FRPS方法计算得到的失效概率,在106次、107次以及5×107次蒙卡计算中,得到的物理过程失效概率分别为9.4304E-7、9.7762E-7和9.7875E-7,可以认为FRPS方法计算的AP1000非能动余热排出系统物理过程失效概率在9E-7量级。同时,利用遗传神经网络方法(GNN)分析了各个输入参数对输出参数的重要度,得到了相应的重要度排序。其次,引入根本原因分析方法,用于运行参数偏离其设计值的根本原因分析。将运行参数偏离其设计值导致的非能动系统物理过程失效归结为具体部件的失效,并通过RiskSpectrum软件建立系统故障树,计算运行参数偏离导致的物理过程失效概率。根本原因分析方法的计算结果表明,其物理过程失效的概率为2.73E-6,与功能可靠性分析(FRPS)方法结果在同一数量级。最后,基于AP1000非能动余热排出系统的结构布置,借助RiskSpectrum分析软件建立非能动余热排出系统(PRHRS)的故障树,计算得到了PRHRS的部件失效概率。AP1000非能动余热排出系统部件失效的故障树分析表明,其部件失效概率为9.559E-6,与物理过程失效概率在同样的数量级上,说明了物理失效研究的重要性。综上,引入拉丁超立方抽样(LHS),并结合蒙特卡洛模拟建立功能可靠性方法,并通过根本原因分析方法,对非能动系统物理过程失效进行评估。同时通过传统的PSA方法,计算得到常规的部件失效概率。对比物理过程失效概率与部件失效概率,为非能动系统物理过程失效研究提供一种可靠的方法。
[Abstract]:In the second generation, the second generation plus, the third generation AP1000 and the Hualong 1 pressurized water reactor, the passive safety system has been introduced to ensure the safety under the accident conditions. The failure of inactive systems should include conventional failure of components and failure of physical processes. At present, there are more researches on failure of components, but less on failure of physical processes. Therefore, the study of physical process failure has important theoretical significance and application value. First of all, the Latin hypercube sampling (LHS) is introduced and combined with Monte Carlo simulation. A functional reliability analysis (FRPS) method was established to study the loss of normal water supply in AP1000 inactive residual heat discharge system. The main calculation flow is as follows: system function analysis and key parameter selection; key parameter distribution quantization; sampling to get input vector; running Relap5 to get output; The response surface is established by combining input and output and failure criterion. Finally, the failure probability calculated by Monte Carlo method. FRPS method is used to calculate the failure probability of 106 times 107 times and 5 脳 107 times of Monka calculation. The failure probabilities of the physical processes obtained are 9.4304E-7 9.7762E-7 and 9.7875E-7 respectively. It can be considered that the failure probability of the AP1000 inactive residual heat removal system calculated by the FRPS method is in the order of 9E-7. The importance of each input parameter to the output parameter is analyzed by using the genetic neural network method (GNN), and the corresponding ranking of importance degree is obtained. Secondly, the fundamental cause analysis method is introduced. It is used to analyze the root cause of running parameter deviating from its design value. The physical process failure of inactive system caused by running parameter deviating from its design value is reduced to the failure of specific parts, and the system fault tree is established by RiskSpectrum software. The calculation results of the fundamental cause analysis method show that the failure probability of the physical process is 2.73E-6, which is the same order of magnitude as that of the FRPS method. Based on the structural arrangement of the passive residual heat removal system (AP1000), the fault tree of the passive residual heat exhaust system (PRHRS) is established by means of the RiskSpectrum analysis software. The failure probability of components of PRHRS. AP1000 is calculated. The failure tree analysis shows that the failure probability of components is 9.559E-6, which is the same order of magnitude as the failure probability of physical process. In this paper, the importance of physical failure research is explained. In summary, the Latin hypercube sampling (LHS) is introduced, and the functional reliability method is established by Monte Carlo simulation, and the fundamental cause analysis method is used. This paper evaluates the failure of the physical process of the inactive system, calculates the failure probability of the conventional components by the traditional PSA method, compares the failure probability of the physical process with the failure probability of the component, and compares the failure probability of the physical process with the failure probability of the component. It provides a reliable method for the study of the failure of non-active system physical process.
【学位授予单位】：华北电力大学(北京)
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TL364

【相似文献】