Power Planning in Hydropower Plants Using Artificial Intelligence: A Case Study of Nigeria

Abstract

Power generation in Nigeria has faced significant challenges, prompting the urgent need for improved efficiency and performance in existing hydropower plants. This research aims to address this need by developing Artificial Intelligence (AI) models to facilitate power planning in three operational hydropower plants in Nigeria. These AI models, specifically artificial neural network (ANN) models trained on historical data from the power plants, were utilized to forecast inflows and energy generation. The results demonstrate the effectiveness of the AI models, with the best inflow forecast model achieving correlation coefficients (r) of 0.8825, 0.8785, and 0.8712 for Jebba, Kainji, and Shiroro hydropower plants respectively. Similarly, the energy generation forecast model achieved high r values of 0.9690, 0.9732, and 0.9643 for the respective plants. These models offer invaluable support to power plants in planning and scheduling operations. However, the study encountered challenges, primarily related to the quality and availability of data. Specifically, the absence of weather data posed a limitation in developing more accurate prediction models for inflow. Despite these challenges, the AI models present a significant advancement in enhancing the efficiency and performance of hydropower plants in Nigeria.