Medium-term forecasts of half-hourly system demand: development of an interactive demand estimation coefficient model

This paper describes a model which predicts half-hourly system demand for representative days in the year using detailed load research data and electricity sales forecasts. The approach is to construct daily load curves for individual consumer categories, expressed in the form of demand estimation coefficients, and aggregate these into a total system demand curve by applying the appropriate sales projections. The nature of the model permits a full interaction with the disaggretated electricity sales forecasts which are produced as part of the electricity supply industry´s regular review of energy prospects. Results of validation tests are reported, and applications of the model to demand forecasting and to more general policy issues are discussed.