Research On Reliability Assessment Of Bulk Power Systems Based On Sensitivity Analysis And A Corrective Model

Bulk power system reliability assessment(BPSRA) can provide important information and, therefore, is a significant auxiliary tool hi power system planning, design and operation, at the same time it is still an important topic that needs to be deeply studied.The systematical studies on theory and application of the models and algorithms of BRSRA are thoroughly conducted. In this dissertation, there are many remarkable works on constructing a sensitivity index system used to assess bulk power system reliability, deducing the analytical expressions of reliability indices as functions of component reliability parameters in the system considered, implementing the optimal placement of various FACTS devices based on reliability, and reducing the computing complexity of reliability assessment(RA) to alleviate "calculation catastrophe".The main results hi detail are included hi the following:(l)Based on theory analysis in depth of bulk power system reliability, a set of sensitivity formulas such as loss of load probability(LOLP),loss of load frequency(LOLF), and expected demand not supplied(EDNS), with respect to component parameters such as component availability and unavailability, component failure and repair rates are deduced. Taking full advantage of the Lagrange multiplies of equality and inequality constraints in the Linear Programming based optimal load shedding model, the sensitivity formula of EDNS with respect to component capacity is creatively presented. These sensitivity indices show the impact of variations in component parameters on system reliability and reflect the importance degree of each equipment. This sensitivity information is very useful to effectually identify system "bottlenecks" and to provide guide information for power system planning and operation, furthermore, it prepares a theoretic basis for developing a fast evaluation algorithm of bulk power system reliability with sensitivity indices. The results tested on the RBTS and IEEE-RTS79 systems show the effectiveness, correctness and practicability of the sensitivity analysis method.(2)Based on rigorous theoretical proof, the analytical expressions of reliability indices such as LOLP, LOLF and EDNS, with respect to component reliability parameters are firstly presented in this paper, and then the physical meanings of the coefficients K1~K4 and D1D2 are stated in detail. Utilizing these analytical formulas,we can easily calculate the system reliability indices in an analytical mode when component reliability parameters vary. The proposed method can effectively alleviate "calculation catastrophe" and provide an advanced technique for promoting the engineering application of BPSRA. The analytical formulas can also provide more detailed and valuable information to planners and designers than just sensitivity analysis does, as well as provide important guide to component maintenance strategies.(3)The configuration characteristics, power flow and reliability models of the Static Var Compensators(SVCs), the Thyristor Controlled Series Compensators(TCSCs) and the Thyristor Controlled Phase Shifting Transformers(TCPSTs) are studied in detail. The optimal load shedding model considering the action of SVCs, TCSCs and TCPSTs and their allocation schemes are developed. This dissertation, firstly, from the point of view on reliability, presents a new algorithm to determine the optimal locations of the FACTS devices. Taking full advantage of valuable information from the Lagrange multiplies in the optimal load curtailment model, the creative sensitivity indices of the system expected energy not supplied(EENS) with respect to the control parameters of the SVCs, TCSCs and TCPSTs are deduced. These sensitivity indices reflect the potential improvement hi system reliability when installing some FACTS devices in the corresponding sites of the bulk power systems, we can, therefore, determine their optimal locations, hi the thesis, the study of the BPSRA considering multi-type FACTS devices is completed. The proposed method is tested on the RBTS and th...