A novel dynamic economic dispatch model considering network security constraints and customer interruption costs is proposed， and its corresponding algorithm is given. In the proposed model， the outputs of generators in the non-fault operation and fault operation（including stochastic generation and transmission component faults） states， and the necessary load shedding are treated as independent variables for decision. The optimization objective is the minimum of generation costs and customer interruption costs. Generation jointly shift distribution factor（GJSDF） is used here to calculate the power flow through transmission lines， by limiting the transmitted power in branches ensures the dispatch and re-dispatch results to meet network security constraints. The constructed model is formed as a quadratic programming model， and primal-dual interior point method is used to solve this model. To treat with the large-scale problem including multi-periods and multi-states variables and constraints， the optimizing process makes full use of the weak coupling between different time intervals and between the non-fault and fault operation states. Decomposed necessary condition for time intervals was used to reduce the number of forward-looking time intervals. Subsequently， a Newton correction equation is formed from special form of primal-dual interior point method KKT conditions. A decomposed calculation method was derived to solve this Newton correction equation to improve its computational efficiency. Case study on the IEEE 30-bus system demonstrates the efficiency of the proposed model and algorithm.