Univariate Time Series Models for the Electricity Consumption in Khartoum and Algazira States in Sudan

Abstract

The main objective of this study is to apply an autoregressive integrated moving average (ARIMA) models for analyzing the monthly electricity consumption data in Khartoum and Algazira states. The methodology used is the Univariate Box-Jenkins models which is based on combining the autoregressive (AR) with moving average (MA). The independent-samples t test (two-sample t test) procedure is also used for comparison between two means of monthly electricity consumption. According to the objectives of this study, the data was obtained from the NEC as a monthly time series in kilowatts. Data from 1999 to 2005 are used for the process of identification, estimation and diagnostic checking. And data from 2006 to 2008 are used to generate forecasts of the series. The statistical packages used in this study- namely, E-views software is used to perform the time series analysis and forecasting, SPSS software is also used to perform the independent-samples t test (two-samples t test). After identifying and estimating the parameters in the three different ARIMA models. ARIMAs would be selected as the best model. The study had come-out with the following findings: 1- No presence of autocorrelation among monthly electricity consumption data in Khartoum and Algazira states. This indicates that, the electricity consumption is independent each month. 2- There is significance difference between two means of monthly electricity consumption in Khartoum and Algazira states. Based on those findings, the study recommends the following: 1-The government of Sudan and NEC must complete the electrification programmes currently being planned to provide electricity to not less than 80% of the population. 2- In order to avoid loss of information and to provide accurate data, NEC has to improve her database on regular basis.