To increase the operational efficiency of power systems with high penetration of renewable energy sources, a two stage economic dispatch (ED) is proposed based on two stage chance constrained programing (TSCCP). In the first stage, the set-points and reserve capacities are optimized to minimize the operational cost. In the second stage, the real-time outputs of generators are adjusted to minimize the recourse cost. In addition, a recovery problem is proposed when the second stage problem is infeasible. By adopting the sample average approximation method, the TSCCP is reformulated to a convex mixed integer quadratic programming, which is solved by commercial software packages. Simulation results very the effectiveness of the proposed method. By introducing the recovery cost in the second stage ED, the operational efficiency can be improved.