Achieving clean energy transitions while maintaining ecological integrity is essential for meeting sustainable development goals (SDGs). This study optimises the operation of a run-of-river small hydropower plant (SHP) in Indonesia, balancing energy production with river ecological sustainability. Using artificial neural networks (ANN) and response surface methodology (RSM), the study evaluates power generation under various environmental flow (e-flow) scenarios. Based on one year of hourly operational and hydrological data from an SHP with a 2 × 350 kW capacity, the Q25MAF scenario was found to offer the optimal balance, increasing e-flow discharge by 89.8% while reducing power generation by only 20.54%. This scenario supports continuous operation and contributes to clean energy, climate action, and aquatic ecosystem preservation. The findings offer a practical framework for sustainable SHP operations in regions with diverse hydrological and operational conditions. Copyright © 2026 Inderscience Enterprises Ltd.