基于抽样统计的CPR1000全失流事故分析  

Study of the CPR1000 Complete Loss of Coolant Flow Accident Based on Sampling Statistical Method

作　　者：王可[1] 蔡杰进[2] 任志豪 朱元兵 王婷 厉井钢 WANG Ke;CAI Jiejin;REN Zhihao;ZHU Yuanbing;WANG Ting;LI Jinggang(SUN YAT-SEN University,Guangzhou of Guangdong Prov.510275,China;South China University of Technology,Guangzhou of Guangdong Prov.510275,China;China Nuclear Power Technology Research Institute,Shenzhen of Guangdong Prov.518035,China)

机构地区：[1]中山大学,广东广州510275 [2]华南理工大学,广东广州510640 [3]中广核研究院有限公司,广东深圳518035

出　　处：《核科学与工程》2019年第4期581-587,共7页Nuclear Science and Engineering

基　　金：广东省科技项目(No.2014A010106012)

摘　　要：CPR1000反应堆发生全失流事故后,冷却剂流量迅速下降,一回路温度和压力升高,可能导致堆芯发生偏离泡核沸腾(DNB)的危险.本文使用中广核自主开发的系统程序GINKGO和子通道程序LINDEN分别对CPR1000的电厂系统和堆芯部件进行了建模,再使用GINKGO进行系统分析、LINDEN根据系统分析结果进行子通道分析后确定偏离泡核沸腾比(DNBR)达到最小值的时刻,并在该时刻使用抽样统计方法对部分相关输入参数进行不确定性分析,同时考虑系统程序、子通道程序以及CHF关系式的不确定性,最终得到满足双95下的最小偏离泡核沸腾比(MDNBR).结果表明:CPR1000全失流事故进程中,并未发生DNB,且具有较大裕量,同时也证明将抽样统计应用到CPR1000全失流事故分析中是可行的.During the complete loss of coolant flow accident (LOFA) in CPR1000 plant,the coolant flow rate would be decreased,the temperature and pressure would be increased,which could cause the departure from the nucleate boiling (DNB).This paper uses the system code GINKGO and the sub-channel code LINDEN,both developed by China General Nuclear Power Group,to establish the model of the CPR1000 plant and to calculate the transients DNBR of the complement loss of flow accident.And the uncertainties of some key-parameters are analyzed by the sampling statistical method when DNBR reaches the minimum value.The transient code uncertainty,the sub-channel code uncertainty and the CHF correlation uncertainty are also concerned.The results show that there is no DNB phenomenon but much DNBR margin during the complete loss of coolant flow accident in CPR1000 plant,and it is acceptable to introduce the statistical method to complete of flow accident analysis of CPR1000.

关 键 词：全失流事故 系统程序GINKGO 子通道程序LINDEN MDNBR 抽样统计 

分 类 号：TL329[核科学技术—核技术及应用]