The scientific location of the VPSS is of great significance to improve the efficiency of emergency repair and reduce the loss of users. To reflect the impact of the load-loss level on the selected location of the power supply, this paper analyzes the service characteristics of the VPSS and compares whether the corresponding repair distance and cost change when the two conditions of the load-loss level are considered to establish the power substation emergency repair multi-objective model. After that, the pareto-dominated multi-objective particle swarm optimization algorithm is adopted to solve the model and the crowd distance sorting is utilized to update the non-inferior solution set and determine the final optimization scheme. Finally, the calculation example shows that the weighted distance and the comprehensive cost are better than the case of not considering the load level when the load level is taken into account. The results show that the multi-objective site selection model for emergency repairs constructed and the proposed solution method have certain guiding significance for the emergency repair strategies of the VPSS.