Forecasting Electricity Demand Per Capita Using Generalized Linear Models

Abstract

This paper presents static models based on Generalized Linear Models (GLM) to forecast electricity demand per capita in Nigeria. It considered that socioeconomic variables such as gross domestic product, population, temperature and consumer price index are factors that directly or indirectly influence electricity demand per capita. The research employs data from 1971 to 2010 using different variants of GLMs to predict the electricity demand per capita in Nigeria from 2011 to 2050. The GLM Gaussian family power 2 link function model was found to be the most preferred model to estimate electricity demand per capita among different models investigated in this study with a forecast results that shows the growth of electricity demand per capita at the annual rate of 38.9% from 120.4859 kWh per capita in 2011 to 167.3775 kWh per capita in 2050.