Risk Analysis Of Power System Cascading Blackouts

Since 2003, large blackouts happened frequently, affecting millions of people worldwide. In these blackouts, sequences of failures happened successively and weak the power systems gradually along the procedure. The current power system is not as reliable as we thought, and there is a critical need for the reexamination of the reliability of the bulk system and figure out the risk of cascading blackouts in power system.The thesis analyze the risk of power system cascading blackouts by using of sensitivity method and probability, statistical method, integrating the basic idea of self-organized criticality. Main thesis has two parts. In the first part, a model is proposed, focusing on power system's evolving to a critical state and the key factors that affect the risk of cascading blackouts. The second part emphasises on the application of branching process model in real power system. Branching process model uses a bulk statistical approach to quantify system's stress level by using of historical disturbance data. The model proposed in the first part can also provide important sample data for the research in the second part due to the difficulty of acquiring disturbance data from industries.In the first part of the thesis, a better power system cascading blackout model is proposed based on the idea of self-organized criticality. Various forces that shape the power system are modeled in a long-term time frame. Within the long-term simulation, fast dynamics by random component failure and the cascading tripping thereafter are embedded. The output of the fast dynamics function as the feedback, affecting system's long-term evoving. Power system is constantly evolving toward a critical state under those opposing forces, under which the risk of cascading blackouts is greatly increased. Indices, including average line loading, average line flow limit per MW served, line investment ratio are proposed to quantify the stress level of the system, while the probability distribution function (pdf) of fractional load shed and total failure components, cascading failure frequency, and risk are used for the risk analysis of cascading blakcouts.IEEE-118 test system is taken for the risk sensitivity analysis of power system cascading blackouts. The thesis proposed that economic forces and reliability criteria are two key factors shaping the evolving of the power system and affecting the stress level of the power system. In the second part of the thesis, a practical validity of branching process model for power system's risk analysis is given for the first time. The cascading transmission line outages recorded over 9 years in a regional electric power grid with approximate 200 lines are studied by describing their bulk statistical behavior. The average amount of propagation of the line outages is estimated from the data. The distribution of the total number of line outages is predicted from the propagation and the initial outages using a Galton-Watson branching process model of cascading failure. Test results and analysis show that branching process model is a novel and practical way to compute the consequences of propagation of outages in cascading failure.