Prediction and Simulation of Kainji Hydropower Reservoir Operation in Nigeria

Abstract

Kainji hydropower reservoir streamflow was forecast from 2017 to 2050 using historical streamflow data and a Markov model. The model was  evaluated with statistical parameters. Various percentages of water stored in the reservoir as Ecological Flow Release (EFR) were used to simulate  future energy generation subject to operational constraints on storage and turbine releases. It was observed that the model forecast the sreamflow  adequately with coefficient of determination (R2), Mean Absolute Error (MAE), Relative Bias (RB) and correlation coefficient (r) of: 0.98, 0.10, 0.002 and  0.99 respectively. Total simulated energy generation reduced as EFR percentage increased. The energy generation simulated was greater than the  observed generated energy for the scenarios. The outcomes of this study can be applied to the Kainji hydropower operational management.