Abstract: Conventional reliability evaluation models focus on equipment operating under ideal steady-state conditions, failing to fully characterize the dynamic behaviors of high-pressure LNG boil-off gas (BOG) compressor systems throughout the complete "shock-response-recovery" cycle. To address this limitation, this study develops an integrated operational reliability evaluation method covering both the vulnerable and recovery stages of the system. Utilizing the theoretical framework of Markov repairable processes and multi-state systems, an evaluation dimension for system recovery characteristics is introduced. The state evolution rules of the system following external disturbances are analyzed, and an operational reliability evaluation index system spanning four dimensions-resistance capacity, absorption capacity, vulnerability, and recoverability-is established on the basis of calculus and cumulative probability theory. Numerical case analysis is performed on a high-pressure BOG compressor system with one operational unit and one standby unit at a large LNG terminal in China. The results reveal that the operation mode with fixed main and standby units delivers higher resistance capacity compared to the random switching mode. Furthermore, natural degradation of the standby unit (imperfect standby state) drastically weakens the absorption capacity and recoverability of the overall system. The proposed evaluation model effectively quantifies the dynamic operational reliability performance of the LNG export compression system under external disturbances. This study demonstrates that, given identical maintenance resources, maintaining standby units in sound operating conditions and optimizing main-standby operation strategies are critical to improving overall system reliability and resilience.