Forecasting High Frequency Intra-Day Electricity Demand Using Temperature

Abstract

This paper introduces a Generalised Additive Model (GAM) to link high frequency intraday (5-minute) aggregate electricity demand in Australia to the time of the day and intra-day temperature. We show a superior model fit when using Daylight Saving Time (DST), or clock time, instead of the standard (solar) time. We also introduce the time weighted temperature model that relates instantaneous electricity demand sensitivity to temperature as a function of the daily activity cycle. The results on DST and time weighted temperature modelling are novel in the literature and are important innovations in high frequency electricity demand forecasting. The overall accuracy of the proposed GAM specification in predicting demand is comparable to the accuracy of the commercial demand forecasting model used by the Australian Energy Market Operator (AEMO). The parsimonious GAM model provides a solid foundation for the development of more elaborate models for forecasting high frequency electricity demand.