Building a Price Model for Electricity Markets

Abstract

In the previous chapters we have outlined the dynamic constraints which restrict the evolution of market price, based on arbitrage theory, replication strategies, market agents’ utility functions, and physical constraints on production and storage. In this chapter we attempt to build a price model for the electricity industry. Based on the lessons learned from other industries, we will use the following criteria in evaluating the effectiveness of the model:

• 1. The model must reflect reality in a dynamic and probabilistic setting. This includes properly reflecting the probability distributions and covariance matrices of future electricity prices.

• 2. The model must have a form conducive to solving the stochastic optimization problems facing the electricity industry, including valuating generation assets, hedging long term contracts, and managing locational risk. This involves addressing the unique properties of how electricity is produced, transmitted and consumed. Specifically it must address the inter-temporal relationships of prices for a non-storable commodity, the locational price differences in complex transmission networks, and the inter-commodity dynamics of electricity, fuel and emissions rights markets, which influence the cost of production for generators. Designing the model so that it can easily be expanded in the temporal, special and inter-commodity domain is therefore crucial to its performance.