Reliability Modeling of MMCs Considering Correlations of the Requisite and Redundant Submodules

The requisite and redundant submodules (SMs) in modular multilevel converters (MMCs) can ensure the continuous operation of MMC-HVDC systems, but to date, the correlations of the SMs in normal and abnormal operations have always been neglected in the reliability modeling of MMCs. To address the correlation issue, a novel MMC reliability modeling approach is proposed. It is implemented by using the Copula Gumbel Archimedean function that belongs to probability distribution functions. This approach is realized in three steps. First, the reliabilities of the semiconductor switches, capacitors and the grading resistors, the protective switche, and control system of individual SMs are obtained using failure-rate information present in the field statistic data. Second, the reliability of the requisite category SMs is calculated using the series probability model, and the parallel probability model is employed to calculate the reliability of the redundant category SMs. Finally, MMC reliability is calculated using the classic k/N(G) probability model. The effectiveness of the proposed approach is shown using different test cases.