Water treatment facilities in Rwanda face challenges related to yield efficiency, leading to potential underutilization of available resources and infrastructure. A Bayesian hierarchical model was developed to analyse yield data from multiple water treatment facilities. The model accounts for spatial and temporal variations using generalized linear mixed models (GLMMs), incorporating prior knowledge about facility characteristics and environmental conditions. The analysis revealed significant variability in system yields across different regions, with an estimated average yield improvement of 15% when considering site-specific factors such as climate and local water usage patterns. Bayesian hierarchical modelling provided a robust framework for understanding yield dynamics and highlighted the importance of localized data integration for improving water treatment facility performance. Investment in regional monitoring networks and targeted interventions based on model predictions could lead to substantial improvements in Rwanda’s water supply efficiency. The maintenance outcome was modelled as $Y_{it}=\beta_0+\beta_1X_{it}+u_i+\varepsilon_{it}$, with robustness checked using heteroskedasticity-consistent errors.