Prognostics and Health Management (PHM) technology is important for the safety and economy of energy storage station (ESS), and traditional manual maintenance is gradually shifting to data-driven maintenance. With the increasing installed capacity of ESSs and the transformation of dispatching conditions from stable single conditions to intricate coupled conditions, the difficulties caused by the large number of equipment, low data quality, and many types of faults have increased. Although there are many theoretically feasible fault diagnosis algorithms, it is difficult for the existing scattered and fragmented research to be applied in engineering. To this end, a fault diagnosis methodology is proposed, including a standardized diagnostic process and an algorithm library configured for each step. The requirements of three main users were analyzed, the software function architecture was established, and the Safety Assessment and Fault Diagnosis system (SAFDS) for ESS was developed. Finally, an application case is presented, which includes the entire process of early warning, problem identification, location, fault diagnosis and handling suggestions.