Scenario generation for price forecasting in restructured wholesale power markets

In current restructured wholesale power markets, the short length of time series for prices makes it difficult to use empirical price data to test existing price forecasting tools and to develop new price forecasting tools. This study therefore proposes a two-stage approach for generating simulated price scenarios based on the available price data. The first stage consists of an autoregressive moving average (ARMA) model for determining scenarios of cleared demands and scheduled generator outages (D&O), and a moment-matching method for reducing the number of D&O scenarios to a practical scale. In the second stage, polynomials are fitted between D&O and wholesale power prices in order to obtain price scenarios for a specified time frame. Time series data from the Midwest ISO (MISO) are used as a test system to validate the proposed approach. The simulation results indicate that the proposed approach is able to generate price scenarios for distinct seasons with empirically realistic characteristics.