Application of a Markov process to evaluate the reliability of the power supply systems utilizing storage batteries

A method applying a Markov process to compute the reliability of power systems that use storage batteries in facilities such as data centers and communications buildings is proposed. In conventional computations using series-parallel reliability blocks, each element is evaluated in terms of one of two exclusive states: normal or in failure. With the proposed Markov-process method, the supply failure frequency can also be computed, which was difficult with earlier methods. State changes due to control operations in various types of power supply system can also be evaluated, and in the future, it should be possible to perform further study, such as building repair models that incorporate many weaknesses and other changes in each state.