Stochastic programming approach for unit availability consideration in multi-area generation expansion planning

This paper proposes a solution approach for the optimal capacity planning problem in multi-area power systems in the presence of randomness in the availability of additional units. The problem is formulated as a two-stage recourse model in the mixed-integer stochastic programming framework. The uncertainties in area generation, transmission fines, and load are incorporated in the model and characterized by their discrete probability distributions. Reliability index used in this analysis is expected unserved energy as this index integrates duration and magnitude of load loss. The objective function is to minimize expansion cost in the first stage and, at the same time, to minimize operation and expected unserved energy costs in the second stage. The first stage decision variables on the number of additional units are represented by binary variables to allow individual unit availability considerations. The solution algorithm utilizes the L-shaped method.