Modelling and forecasting the demand for electric energy in New Zealand

This thesis studies the demand for electric energy and ways of forecasting it, as an aid to the economical design and operation of electric power systems. An examination of the nature of consumers demands leads to a two part model of the demand. Long term growth of demand is shown to be determined by the way the numbers of appliances owned by consumers increases, while the short term daily, weekly and seasonal demand fluctuations result from the way consumers use their appliances. A number of forecasting methods utilizing this model are studied. The accuracy of a demand forecast influences the amount of reserve capacity needed to satisfy a given level of demand with a specified degree of reliability. A criterion is presented which determines when a forecast may be considered sufficiently accurate. Sufficient accuracy is defined in terms of minimizing the costs of providing the reserve capacity and of improving the accuracy of the forecasts