Ghana faces significant challenges in water treatment efficiency due to varying facility designs and operational conditions. A Bayesian hierarchical model was applied to assess the performance of water treatment facilities, accounting for variability between different sites and within each site over time. The model revealed that average treatment efficacy varied by up to 20% across different systems, with some facilities showing improvement through targeted interventions. Bayesian hierarchical modelling provided nuanced insights into the operational efficiency of water treatment systems in Ghana, highlighting areas needing further attention. Further research should focus on implementing and validating these models for broader application in similar contexts. The maintenance outcome was modelled as $Y_{it}=\beta_0+\beta_1X_{it}+u_i+\varepsilon_{it}$, with robustness checked using heteroskedasticity-consistent errors.