A simulation model of a real electricity supply undertaking was used to provide a financial performance measure for growth curve forecasting models. The impact on financial performance was determined when changes were made in (1) the method of estimating the model parameters, (2) the period between re-estimations, (3) the growth curve fitted and (4) the amount of smoothing of the demand time-series. The response to variation of the parameter review period was found to behave surprisingly, in that it exhibited different signs for two different estimation methods. Changes in re-estimation period explained somewhat more of the variation in performance than did a change in growth curve. Correcting the demand series for conditions which were known to be abnormal improved performance.
KEY WORDS Growth curves Simulation Forecasting performance Logistic
Surveys conducted by and Naylor and Schauland (1976) confirm the popularity of fitting growth curves as a forecasting method. has reviewed a group of applications in which growth curves were used to forecast demand across a broad range of industrial sectors. Among the areas to which this approach has been applied is that of forecasting the demand for electricity Bossert, 1977;.
In a previous study, examined the impact on the financial performance of a large electrical supply undertaking of varying the method of forecasting peak demand used in the capacity acquisition process. A group of four methods were examined: autoregressive, Box-Jenkins, Holt and fitting a Logistic curve. When the interval between reestimation of the forecasting models was changed from annually to five-yearly it was observed that the financial performance improved for the first three methods but deteriorated for the logistic fitting. Subsequent additional analyses of variance of these results showed that for each of the four methods considered individually, the update or frequency of re-estimation main effect was significant at the 0.05 level.