Abstract: The failure events in natural gas pipeline network will seriously afect the security of natural gas supply and reduce gas supply reliability and user satisfaction. Therefore, in order to meet the natural gas demand to the greatest extent and minimize the impact of failure events on gas supply reliability and user satisfaction, a method was proposed for gas flow allocation in natural gas pipeline network under accident conditions coupling with the unit failure and demand sideimpact. In this method, firstly, the demand prediction model suitable for diferent types of natural gas users and the user classification mechanism were established. Secondly, the hydraulic-thermal simulation of natural gas pipeline network was used to quantify the impact of failure event on the pipeline transmission capacity. Finally, the optimization model for flow allocation in the natural gas pipeline network was established based on the mixed-integer linear programming. Meanwhile, the optimal flow allocation plan under accident conditions was determined with consideration to the impact of user demand side and failure events, as well as the gas inflow at the source and the flow, pressure and hydraulic constraints of pipeline network. Furthermore, the proposed method was applied to a real natural gas pipeline network to identify the key nodes and links of gas supply in the pipeline network. The gas flow allocation plans under diferent accident conditionswere determined. The results show that the application of the proposed flow allocation method for the pipeline network could give priority to supply natural gas to the users with higher importance, thus minimizing the impact of accidents on downstream users and providing technical support for the precise regulation and control of the pipeline network.