Optimizing Reservoir Operations on Nepalese Hydropower Systems: A Comparative Study of Metaheuristic Algorithms and Environmental Flow Tradeoffs

Abstract

Optimization of reservoir operation is essential for efficient and sustainable water utilization and management. This plays a critical role in balancing energy generation with ecological sustainability, particularly in hydropower-dominant regions like Nepal. This study investigates the application of metaheuristic algorithms (MHAs) to optimize reservoir operations while addressing environmental flow requirements. The research focuses on two key hydropower projects in Nepal: the Budhigandaki Hydropower Project (BGHEP) and the Dudhkoshi Storage Hydropower Project (DKSHEP). The comparative analysis of seven MHAs is conducted in BGHEP, while the tradeoff analysis of different environmental flow methods is done in DKSHEP.

After optimization, there was a considerable increase in total energy production, better dry-season energy distribution, improved storage management, and enhanced performance indices when compared to the non-optimized case. Under the given objective function and constraints such as storage limits, turbine capacities, hybrid algorithms, particularly hybrid Genetic Algorithm-Particle Swarm Optimization (GAPSO) and hybrid Particle Swarm Optimization-Grey Wolf Optimization (PSOGWO), outperform standalone algorithms in achieving optimal solutions with higher computational efficiency. It was found that GAPSO leads in smaller dimensions while PSOGWO was more reliable in higher-dimensional problems. Additionally, noble approaches such as the Football Team Training Algorithm (FTTA) and less explored Moth Flame Optimization (MFO) provide reliable results and can be used in reservoir optimization.

Furthermore, the study highlights the tradeoffs of integrating environmental flow considerations into reservoir operation policies. It was found that while the DOED guideline is better for power-focused projects, the Minimum Monthly Flow Method can be a better substitute because it is a dynamic approach and provides similar results to DOED. However, the Annual Distribution Method (ADM) can be a more balanced and suitable approach for projects that need to protect the ecosystem.

This research contributes to the growing body of knowledge on sustainable hydropower development by demonstrating the potential of advanced optimization techniques in optimizing the release schedule. The findings underscore the need for further research into ecological impacts, climate change adaptation, and multi-objective optimization frameworks to refine reservoir management strategies.

Yogesh Sharma Neupane (079MSHPE024)

Masters in Hydropower Engineering Pulchowk Campus

Source Code

Working on Progress… will be updated soon