| With the rapid development of world economy and increasing demand for energy,the proportion of nuclear power in power generation system has gradually increased due to its low operating cost and small greenhouse effect.After nuclear power unit is connected to the power grid,there will be serious interaction between them.Frequency and voltage are the main ways of power grid affecting the operation status of nuclear power unit.Because of the large capacity of nuclear power unit,high safety of nuclear energy and sensitivity of nuclear island to power grid disturbance,normal or abnormal operation of nuclear power unit will directly affect the safety and stability of power grid,therefore,it is of great significance to evaluate the reliability of nuclear power unit.However,the traditional outage probability model of nuclear power unit only gives its average outage probability based on historical statistical data,and does not consider the influence of grid operating parameters’variation on the outage of nuclear power unit.At the same time,the number of nuclear power units in operation is far less than that of conventional thermal power units,and due to confidentiality and other reasons,nuclear power unit lacks or even has no outage statistics,the classical reliability evaluation approach of power equipment based on the law of large numbers in small sample cases is no longer applicable.Therefore,it is very urgent to study a reliability evaluation approach for nuclear power unit that can take into account both the variation of operating parameters of power grid and the limitation of outage samples.This paper firstly introduces the theories of precise probability and imprecise probability(IP),emphasizes that imprecise probability is an effective method to describe small sample events,the interval probability is used instead of the precise single-valued probability to deal with the uncertainty of fault or outage probability of power equipment caused by insufficient samples.Then it introduces the probability estimation method based on imprecise Dirichlet model(IDM),classical reliability model of power equipment,and estimates the imprecise fault probabilities of overhead transmission lines(OTLs),transformers and generators by IDM.When disturbances or equipment faults occur in the power grid,the frequency and voltage of the power grid will change,which may cause the outage of nuclear power unit.In this regard,this paper proposes an imprecise outage probability model for nuclear power unit that takes into account grid operating parameters.According to the frequency and voltage protection characteristics of nuclear power unit,the variation curves of outage probability of nuclear power unit related to frequency and terminal voltage are obtained respectively.By simulating N-1 fault scenarios,considering imprecise fault probabilities of power equipments and the resulted outage probability of unit,adding the outage probability caused by unit’s fault,the imprecise outage probability of nuclear power unit is finally obtained.The approach of this paper comprehensively considers the factors that cause the outage of nuclear power unit such as the variation of operating parameters of the power grid,aging of the unit’s components,manufacturing defects and so on,at the same time,the interval form of outage probability can effectively deal with the situation of insufficient outage samples for nuclear power unit,which provides a new approach for reliability evaluation of power equipment under the condition of insufficient outage samples.To verify the effectiveness and practicability of this approach,the improved IEEE 9-node system and actual 500kV power grid in a province of China are analyzed by case studies.By sequentially simulating N-1 faults of OTLs,transformers and generators,analyzing the impact of fault scenarios on the operation status of nuclear power unit,comprehensively considering various internal and external factors that cause the outage of nuclear power unit,the more conservative and reasonable imprecise outage probability of nuclear power unit is estimated. |