PROBABILITY DISTRIBUTION FUNCTIONS FOR THE PREDICTION OF MEAN RESERVOIR INFLOW AT HYDROPOWER DAMS IN NIGERIA

Abstract

Four methods of probability distribution analysis were evaluated for the prediction of mean reservoir inflow at hydropower dams in Nigeria; the hydropower dams include Kainji, Shiroro and Jebba dams. The reservoir inflow data were subjected to probability distribution analysis such as Gumbel (EVI), Normal (N), Log-Pearson type III (LP₃) and Log-Normal (LN), the selection of the appropriate probability distribution model for each hydropower dam was based on the result from the goodness of fit tests performed on the various probability distribution models under consideration. The goodness of fit tests considered include: Chi-square, Correlation coefficient, Coefficient of determination and Standard error of estimate. However, for both Kainji and Shiroro hydropower dams the model evaluated to the best fit was Log-Pearson type III, this is an indication that the reservoir inflow at both dams are skewed. While at the Jebba hydropower dam the best fit model was evaluated to be Log-Normal