Medium-Range, Weekly and Seasonal Peak Demand Forecasting by Probability Methods

Statistical procedures have been developed, computer programs prepared and documented, and techniques tested for the preparation of weekly and seasonal peak demand forecasts with lead times up to seven years. Probability information in the form of both standard deviation and 99% confidence interval gives a comprehensive basis for system planning but this can be supplemented by a tabulation of the forecast's probability distribution in cases where more detailed information is needed. The practicality of the forecasting procedure has been demonstrated by tests with data from four electric utility companies having different types of service territory and varied load composition. Results are included for three of these companies.