Optimal working path algorithm for mitigation tasks under nuclear accidents
[Objective]In an event of a severe nuclear accident,the reactor core may be at risk of meltdown.In such case,radioactive materials may leak into the reactor building from the containment,exerting a radiological impact on individuals who move in the reactor building to perform mitigative missions.It is thus important to estimate the potential radiological consequences of the mission and identify optimized working paths before taking actions to minimize radiation dose of people working in the reactor building and to avoid putting them in harm's way unintentionally.[Methods]A self-developed simulation software,REM AD A,is used to simulate the distribution of radioactive materials in CPR1000 reactor building during small break loss of coolant accident(SBLOCA),main steam line break(MSLB),and station blackout(SBO).Based on the dose estimation module of REMADA,an optimal path algorithm based on depth-first search is established to search for all valid paths and determine the optimal path for individuals involved in different mitigative missions within the reactor building.[Results]The Modular Accident Analysis Program MAAP5 is used to simulate the leakage rates of three different radioactive materials(including inert gas,cesium iodide,and metallic oxide)from the containment under SBLOCA,MSLB,and SBO in the study.By establishing a mathematical model of the CPR1000 reactor building and selecting typical atmospheric boundary conditions,REMAD A simulates the airflow inside the reactor building.Based on aerosol models,the air and ground distributions of radioactive materials in different compartments are calculated,and the total radioactivity in the fields of all compartments after decay is estimated.According to severe accident management guidance,this study assumes three mitigative tasks for SBLOCA,MSLB,and SBO,respectively,assuming departure time,starting locations,destinations,and time limit for all tasks.Depth-first search algorithm is applied and optimized by pruning unqualified cases based on restrictive conditions,which results in obtaining all possible working paths for timely completing mitigation tasks.As a result,the algorithm has found 4 463 possible paths for SBLOCA,9 763 possible paths for MSLB,and 4 283 possible paths for SBO.Radiation doses received by people working on each path for mitigation tasks are estimated using the dose evaluation module of REMADA.The optimal working paths for all tasks are obtained through dual ordering of time consuming and radiation doses.The results show that the optimal paths found by the algorithm can effectively reduce the radiation doses for people.Compared with conventional paths which take less time for traveling,the optimal paths can almost eliminate radiation doses in SBLOCA and reduce radiation doses by about 50％and 55％for MSLB and SBO,respectively.The running time of the algorithm in each case is less than 3 s,proving high efficiency to meet the requirements of emergency decision making in nuclear accidents.[Conclusion]This study demonstrates the feasibility of dynamic dose evaluation of people under nuclear accidents and presents a method for searching minimum doses paths for individuals carrying out mitigative tasks based on a depth-first search.Based on the highly extensible modules of REMADA,the algorithm established in this study can be applied to various practical scenarios,including different nuclear power plants and nuclear accidents.The calculation results of the algorithm indicate that the identified optimal paths can effectively reduce the radiation dose damage to people working in the reactor building.The algorithm is efficient enough to meet the requirements of practical application for emergency decision making during severe accidents.In conclusion,the algorithm can effectively aid in minimizing radioactive risks for individuals and facilitate the formulation of rational emergency decisions.
nuclear accidentemergency decision makingradioactive materials dispersionradiation dose impactoptimal working path
国家科技期刊平台
NSTL
万方数据
维普
