Three Essays on the Economics of Competitive Electricity Market

Abstract

In this thesis we present three essays on the economics of competitive electricity. In chapter one we use a stochastic dynamic model to examine the economic impact of storing wind power in a hydroelectric reservoir. Using a stochastic dynamic model allows us to explicitly model the uncertainty of wind power in the objective function of the hydroelectric generator. We show that the amount of water released by the hydroelectric generator in any period is a decreasing function of the number of periods that the hydroelectric generator has to deplete its reservoir. Furthermore if the hydroelectric generator has a high enough number of periods to deplete the water in its reservoir it will be optimal for it not to release any water at the beginning. In chapter two we present a model of operating reserve in a competitive electricity market. Our paper departs from other papers in several respects. First, we concentrate on the possibility of generator failure and not demand uncertainty. Second, we allow demand to be highly inelastic rather than perfectly inelastic like other papers do. This allows price spikes to occur when there is a major generator failure. To moderate price spikes, options for operating reserves can be purchased and exercised when price spikes occur. Third, we model uncertainty on the supply side and not on the demand side. Fourth we adopt a two price approach where one price is used to reserve capacity and the other price is the strike price paid when the options are exercised unlike other models which use the spot price as the strike price. Finally we explicitly model the demand and supply in the market. Using the concept of rational expectation we develop and prove the existence and uniqueness of a rational expectations equilibrium and analyse its characteristics. Furthermore we show that the competitive electricity market will provide more operating reserve capacity than is socially optimal. In the chapter three we formalize a model of reliability for an electric grid when consumers’ preferences for electricity consumption are private information. In our model we design an optimal blackout strategy for the regulator. The model demonstrates that an ex post efficient social choice function is truthfully implementable in Bayesian Nash equilibrium. It also yields both the optimal level of generation capacity investment and the second-best blackout program.