Title: Using simulation to estimate the uncertainty in Union Electric Company's residential customer forecast

Authors: Peter J. Weigel, Reuven R. Levary

Addresses: Union Electric Company, P.O. Box 149, Mail Code 1400, 1901 Chouteau, St. Louis, MO 63166, USA. Department of Decision Sciences and MIS, School of Business and Administration, 3674 Lindell Blvd., St. Louis, MO 63108, USA

Abstract: An estimation of uncertainty for Union Electric Company|s residential customer forecast is made using Monte Carlo simulation. The methodology is an improvement over the deterministic simulation methodology the company used when it filed an integrated resource plan with the Missouri Public Service Commission in 1993. Two different sensitivity analysis methodologies are applied to the company|s 1994 residential customer forecast. One methodology involves testing each discreet combination of model inputs within the uncertainty range. The other methodology involves simulating the distribution of model inputs. Neither methodology produces adequate results by itself. A combination of the two methodologies is used to efficiently determine an acceptable conclusion about forecast sensitivity.

Keywords: forecast uncertainty; simulation; electric utilities.

DOI: 10.1504/IJCAT.1999.000207

International Journal of Computer Applications in Technology, 1999 Vol.12 No.2/3/4/5, pp.225-232

Published online: 13 Jul 2003 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article