Analysing the economic benefit of electricity price forecast in industrial load scheduling

The current trend of electricity market deregulation ushers in increasingly dynamic electricity pricing schemes. The cost-optimal scheduling of industrial loads with accurate price forecasts is therefore important. However, results in the current literature suggest that mean absolute percentage error (MAPE) is poor at indicating the economic benefit of a forecast. This paper presents the economic benefit analysis of electricity price forecast on the day-ahead scheduling of load-shifting industrial plants. A coal-conveying system with storage is used as a case study. The research uses three price forecasting methods on the PJM's market prices over a period of two years. Rank correlation (RC) between the predicted price and the actual price is proposed as an indicator of economic benefit. The results show that RC is a better indicator of economic benefit than root mean square error (RMSE) and MAPE. They also show that potential economic benefit obtainable from forecasts depends on price volatility and not mean price. An artificial forecast is used to validate the superiority of RC over MAPE and RMSE. It is observed that the predictability of a forecast's economic benefit is largely dependent on how responsive the load is to electricity price changes.