Probabilistic Safety Risk Assessment Of Frequency Stability For Bulk Power Systems Based On Frequency Dynamic Characteristic

This dissertation is supported by National Natural Science Foundation of China (No.50977094, No.50607021).The modern power systems have showed the main features of large units and large capacity remote transmission, the ability to maintain frequency stability is facing serious challenge while the system are subjected to random serious power shortage, in order to objectively reflect the risk level of frequency stability, introducing the risk assessment to the study of frequency stability has become an urgent task. Therefore, study on appropriate theories, models and algorithms of probabilistic risk evaluation for frequency stability to provide the risk level of frequency stability for the planning and operation personnel have become an important research topic which needs to solve urgently.This dissertation studies the frequency stability problem based on frequency dynamic characteristics. To overcome the frequency stability assessment in consideration of the lack of spinning reserve and reduce the approximate calculation error of the action time in the under frequency load shedding, this dissertation derives the dynamic frequency characteristics which accounting the impact of spinning reserve, and the analysis model of frequency stability is built for under frequency load shedding. To achieve probabilistic risk assessment for frequency stability, the stochastic models of generating dispatch and load demand are established, and the expected indices which reflect the severity of the frequency instability and the probability distribution indices which reflect the risk level are constructed.Based on the probabilistic risk assessment model and the indices, this dissertation proposes an algorithm of probabilistic risk evaluation of frequency stability which considers the uncertainty of generating system's operational state and the effect of under frequency load shedding, in this algorithm the outage models of components are built, then the outage probabilities are calculated in order to consider the randomness of the system, non-sequential Monte Carlo simulation is utilized to sample system states. Taking into account the role of automatic under frequency load shedding, according to non existence and existence of spinning reserve, frequency stability analysis is made based on the frequency dynamic characteristic for every contingency, finally formatting the cumulative risk index. The probability distribution of system lowest frequency, steady state frequency and frequency deviation safety margin which reflect the distribution of frequency stability risk are analyzed, and the time of frequency instability are compared, moreover the influence of spinning reserve capacity diversification and generator partial failure mode on frequency stability are discussed as well. Finally the annual indices of the multi-level load model including uncertainty and excluding uncertainty in load are calculated, and the analysis and comparison of the various load levels on the contribution of the annual indices are made. The calculation and analysis of IEEE-RTS 24 are showed to verify the validity of the proposed algorithm, the results reflect the probabilistic risk level of frequency stability.