Forecasting Regional Electricity Demand for Turkey

Abstract

Electricity demand forecasting is important for planning in the different segments of the electricity sector. In this study, the main aim is to forecast regional electricity demand by using time series, panel data and spatial panel data models and compare their forecasting performance. As in 2004, 21 regional monopolistic distribution companies were established by the implementation of Strategy Paper, this study employs data on these regions between 1986 and 2013. The results show that forecasts obtained by using pooled panel data model are on average better than the ones obtained by using other models based on root mean squared error evaluation criterion. © 2017, Econjournals. All rights reserved.