Electricity price forecasting with confidence-interval estimation through an extended ARIMA approach

Abstract

Accurate electricity price forecasting is a crucial issue concerned by market participants either for developing bidding strategies or for making investment decisions. Due to the complicated factors affecting electricity prices, accurate price forecasting turns out to be very difficult. The autoregressive integrated moving average (ARIMA) approach has been extended to make hourly market clearing price (MCP) forecasting in electricity spot markets with error correction and confidence interval estimation. The ARIMA model used for forecasting price and the method to implement price forecasting are presented first. Then the ARIMA approach is extended to include error correction for improving accuracy of price forecasting. Moreover, the confidence interval of the forecasted prices is estimated assuming the residual errors are in gaussian or uniform distribution. Hourly MCP forecasting of the Californian Power Market is used as a computer example, and the comparison with conventional ARIMA approach is given. Computer test results show clearly that the suggested extended ARIMA approach for spot price forecasting is very effective with satisfactory accuracy. It can work under very worse market conditions with high price volatility.