Publications

MODELLING AUTOREGRESSIVE INTEGRATED MOVING AVERAGE OF ELECTRICITY GENERATED AND CONSUMED



Olatunde John Kuranga. on 06/01/2017

The study is based on modelling Autoregressive Integrated Moving Average (ARIMA) to the electricity generated by Ibadan Electricity Distribution Company, and also on electricity consumption in three zones of Kwara state (Challenge hub, Baboko and Omu-Aran business unit). It is aimed at constructing time plots, testing the presence of stationary, Autocorrelation and Partial Autocorrelation Function and also forecasting for the future electricity generated, and electricity consumption in three zones of Kwara state. With the help of Akaike Information Criteria (AIC), the best model was selected. The series of electricity generated by IBEDC in Challenge, Baboko and Omu-Aran, shows that there is stationary in the data while the consumptions are non-stationary, the first differencing give a stationary data. Autocorrelation and Partial Autocorrelation Function (ACF and PACF) plot help in the order of Autoregressive Integrated Moving Average (ARIMA) model, used in this study. The forecast shows an increase in both generated and consumption of electricity in Challenge hub, Baboko and Omu-Aran business unit. We also have high rate of electricity generated in December. We recommend Enlightenment campaign on the value of electricity and of the use ARIMA (1, 1, 1) for electricity generated to Challenge and Baboko while ARIMA (2, 1, 1) is recommended for that of Omu-Aran. But for electricity consumption we recommended ARIMA(2,1,2), for Baboko unit and Omu-Aran while challenge hub, ARIMA(1,1,2) is recommended.

Keywords: ARIMA, Stationary data, Electricity Generated, ACF and PACF, Consumption,