Forecasting of electricity demand to reduce the inventory cost of imported coal

Abstract

The purpose of this paper is to apply forecasting methods to forecast the electricity demand in Thailand. Demand forecasts will be used to estimate the imported coal order quantity in order to decrease the inventory cost of the imported coal. The monthly electricity demand data from January 2010 to December 2014 are used to forecast the monthly electricity demand from January 2015 to December 2015. The forecasting models are additive and multiplicative decomposition models and additive and multiplicative Holt-Winters models. The forecasting accuracies are measured by mean absolute percentage error and compared by randomized complete block design. The results of the study show that all forecasting accuracies are not significantly different so the multiplicative decomposition model is chosen because of its simplicity. The proposed imported coal order quantity is equal to 5.95 percent of the electricity demand forecasts. The inventory cost in 2015 decreased by 3,721.82 million baht or 14.84 percent compared to the inventory cost under the current order quantity.