Prediction of Wind Power Density for Electricity Generation at Makambako, Tanzania Using Auto-Regression Integrated Moving Average (ARIMA) Model

Abstract

Wind speed data used in this study were from Makambako and were recorded at a height of 10 m. In the analysis, the data were extrapolated to 80 m above ground level and wind speed from July to November were found to have an average value of 17.58 m/s. November was found to be the windiest month for the whole duration while the calm month was found to be March. The wind power density was found to be highest in the month of November for the whole study duration. The months of July to November have higher wind power density ranging from 1,739.00 W/m²/month to 12,244.35 W/m²/month at 80 m above ground level while the remaining months have values below the mentioned range. The findings of this study have shown that among five predictions, three showed acceptable predictions of 90.40%, 96.38% and 73.42%, one showing a fair prediction of 65.90% while the fifth one giving unacceptable prediction of 28.83%. It is proposed that ARIMA model be used to predict wind speed for other months apart from March 2009 and November for the years 2009, 2010, 2013 and 2016 that were predicted in this study.

Keywords:  ARIMA model; average wind speed; prediction; wind power density;