Abstract: To effectively reduce the cost of comprehensive energy system capacity allocation, a double-layer optimal allocation algorithm considering reliability constraints was proposed. Firstly, based on the fundamental architecture of the integrated energy system, a double-layer capacity allocation optimization model was carried out, of which the upper layer was for capacity allocation and the lower layer was for optimal operation. Secondly, a Markov two-state fault transition model using sequential Monte Carlo method to extract equipment faults was added to this double-layer capacity allocation model to evaluate the reliability of the system, and the obtained results were taken as the reliability constraints of the system. Ultimately, the equipment capacity configuration of the comprehensive energy system satisfying the reliability constraint interval was output. The outcome demonstrates that the comprehensive energy system with multiple energy equipment not only facilitates the utilization of new energy units and reduces the use of fossil energy, but also improves the reliability of the system. Synchronously, the study case compares the economy of system capacity allocation under different energy supply reliability constraints and verifies the effectiveness of the proposed model.