Title: Univariate time-series forecasting of monthly peak demand of electricity in northern India

Authors: Sajal Ghosh

Addresses: Management Development Institute, Room No C–10, MDI, PO Box No. 60, Mehrauli Road, Sukhrali, Gurgaon 122001, India

Abstract: This study forecasts the monthly peak demand of electricity in the northern region of India using univariate time-series techniques namely Multiplicative Seasonal Autoregressive Integrated Moving Average (MSARIMA) and Holt-Winters Multiplicative Exponential Smoothing (ES) for seasonally unadjusted monthly data spanning from April 2000 to February 2007. In-sample forecasting reveals that the MSARIMA model outperforms the ES model in terms of lower root mean square error, mean absolute error and mean absolute percent error criteria. It has been found that ARIMA (2, 0, 0) (0, 1, 1)¹² is the best fitted model to explain the monthly peak demand of electricity, which has been used to forecast the monthly peak demand of electricity in northern India, 15 months ahead from February 2007. This will help Northern Regional Load Dispatch Centre to make necessary arrangements a priori to meet the future peak demand.

Keywords: ARIMA; electricity demand; exponential smoothing; India; monthly peak demand; univariate forecasting; time-series forecasting.

DOI: 10.1504/IJICBM.2008.018626

International Journal of Indian Culture and Business Management, 2008 Vol.1 No.4, pp.466 - 474

Published online: 02 Jun 2008 *

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