Long term infrastructure investments under uncertainty in the electric power sector using approximate dynamic programming techniques

Abstract

A computer model was developed to find optimal long-term investment strategies for the electric power sector under uncertainty with respect to future regulatory regimes and market conditions. The model is based on a multi-stage problem formulation and uses approximate dynamic programming techniques to find an optimal solution. The model was tested under various scenarios. The model results were analyzed with regards to the optimal first-stage investment decision, the final technology mix, total costs, the cost of ignoring uncertainty and the cost of regulatory uncertainty.