Statistical modelling of electricity consumption in Sri Lanka

Abstract

—Consumption of electricity is growing each year in Sri Lanka due to industrialization, urbanization, modernization. Forecasting electricity consumption is highly important when deciding on investment and construction. Therefore, this study is focused to develop suitable statistical models for forecasting monthly electricity consumption. The monthly electricity consumption data (in Gigawatt hours) of Sri Lanka were obtained from data library of central bank of Sri Lanka from January of 2010 to June of 2018.Two different approaches namely, by ARIMA approach and HoltWinters exponential smoothing approach were used to develop two statistical models for forecasting electricity consumption. Model fitting was done by the data from January of 2010 to February of 2018 whereas electricity consumption data from March of 2018 to June of 2018 were used to validate the models.