IAEE Members and subscribers to The Energy Journal: Please log in to access the full text article or receive discounted pricing for this article.

Estimating Hourly Electric Load with Generalized Least Square Procedures

Although electricity demand receives much attention in the empiri-cal literature (see Taylor (1975) and EPRI (1982b) for excellent surveys on the topic), hourly load demand analysis has only recently begun. Notable contributions are a series of studies sponsored by the Electric Power Research Institute (EPRI (1977. 1979a, 1979b, 1981a, 19816. 1982a) and Platt (1983)). These studies estimate load curve models for regions of the United States. Unfortunately, from a utility perspective, the empirical results presented in these studies are not directly applicable. Further, because the data used in these studies are not generally available at the geographic level of a utility service area, applying their methodology is problematic. This paper presents a practical method for an electric utility to produce an hourly load curve model similar in overall framework to these studies. Our procedure is innovative in that it produces statistically efficient estimates, which the above papers do not. We also demonstrate a method that uses supplementary forecasts to enhance the forecasting performance of the hourly load model.

Purchase ( $25 )

Energy Specializations: Energy Modeling – Sectoral Energy Demand & Technology; Electricity – Markets and Prices

JEL Codes: C53: Forecasting Models; Simulation Methods, Q54: Climate; Natural Disasters and Their Management; Global Warming, C51: Model Construction and Estimation, Q42: Alternative Energy Sources, Q41: Energy: Demand and Supply; Prices

Keywords: Hourly electricity load, Generalized least square procedure, Forecasting

DOI: 10.5547/ISSN0195-6574-EJ-Vol7-No2-11

Published in Volume 7, Number 2 of the bi-monthly journal of the IAEE's Energy Economics Education Foundation.


© 2021 International Association for Energy Economics | Privacy Policy | Return Policy