Optimization of generation investments under uncertainty in restructured power markets

This paper presents a model for optimisation of investments in new power generation within restructured power markets. We use stochastic dynamic programming to identify an optimal investment strategy when the objective is to maximize the expected profit over a long time horizon. This formulation of the investment problem takes into account the value of postponing the investment decision to wait for new information to be unfolded, for instance about uncertainties in load growth, fuel prices, new market regulations etc. The model can be used to identify optimal entry thresholds for investments in new power plants. The optimal investment strategy tends to be more restrictive than what the traditional net present value evaluation suggests, and this can have severe long-term consequences for the balance between supply and demand in the power system. A selection of results from an evaluation of a realistic power plant project in Norway is presented to illustrate potential use of the model.