ABSTRACT
In Nigeria, there is a problem of inadequate electricity supply to the populace. A recognized reason for the energy poverty in the country is the poor handling of the electricity statistics provided by the National Electricity Regulatory Commission (Ajayi, 2006). Absence of good forecast is a dominant reason for the NERC's inability to manage the supply chain of electricity or carry out effective electricity demand and production planning. In this project, specific data has been obtained on electricity production (in kWh), electricity consumption (in kWh), electricity consumption per capita (in kWh), electricity transmission losses (in kWh) and percentage electricity transmission losses from the statistical bulletin of the National Bureau of Statistics between the year 1991 and 2011. This data reflects electricity demand and consumption pattern over the last two decades to help carry out a time-series analysis of the data and plot a time-series graph of the data to show the pattern of electricity demand and consumption using the Data Analysis Package, Time Series Modeller and Sequence Chart Analyser of the IBM SPSS 21 application. A Time series forecast of the data obtained was done using the Auto Regressive Integrated Moving Average model. The forecast predicted that if electricity consumption per capita remains static and unchanged at 145.146 kW per person per annum, the annual electricity consumption in 2016 will stand at 30.319 billion KWh thereby surpassing the total amount of electricity produced, 30.251 billion KWh if all other parameters that affect the production and consumption such as the amount of electricity lost in transmission continues in the same pattern as predicted by the model. Therefore the percentage of Nigerians that have access to electricity supply will also reduce from 48% in 2011 to about 41.83% in 2016. Error tests were carried out to ascertain the efficiency of the forecast and the model used. Forecasts are deemed to be accurate and authoritative enough if the MAPE value is below 10 and a forecast whose MAPE is between 1 and 5 is considered authoritative and accurate (Bozarth, 2011). The forecast was able to achieve a MAPE of 1.207 for the forecasts below the 10th Percentile of all forecasts and a MAPE of 1.714 for the forecasts between the 10th and 25th Percentile of all forecasts, which is the forecast between 2012 and 2017. This value increased to 25.037 as the model generated more forecasts. The results obtained by this forecast is further validated by the actual 2013 value of average annual electricity production which stands at 29.166 billion kWh according to (Nnodim, 2013), a 2.89% error difference from the forecasted 28.321 billion kWh. Hence, the forecasts between 2012 and 2017 are relatively accurate and authoritative.