"background": "Water treatment systems in sub-Saharan regions often operate below design capacity due to operational and maintenance challenges. Quantifying the impact of specific interventions on system yield is complex, as external factors like raw water quality and seasonal demand confound direct before-and-after comparisons. ", "purpose and objectives": "This case study aimed to develop and apply a robust quasi-experimental methodology to isolate and measure the true effect of a major technical upgrade programme on the volumetric yield of drinking water treatment facilities. ", "methodology": "A difference-in-differences (DiD) model was employed, analysing panel data from 12 treatment plants. Six received upgrades (treatment group), while six operated under baseline conditions (control group). The core statistical model is Y{it = \0 + \1 + \2 + \ (\) + \₈ₓ, where \ captures the causal effect. Inference was based on cluster-robust standard errors at the plant level. ", "findings": "The DiD estimator, \, was +0. 18 (95% CI: 0. 12 to 0. 24), indicating the intervention caused an 18 percentage point increase in average daily yield ratio. This effect was statistically significant (p < 0. 01) and robust to several specification checks. The control group showed no significant trend. ", "conclusion": "The difference-in-differences approach successfully isolated the causal impact of the upgrades from background trends, providing credible evidence of substantial yield improvement. The methodology offers a rigorous framework for performance evaluation in operational engineering contexts. ", "recommendations": "Adopt quasi-experimental evaluation designs, like DiD, for infrastructure programme assessment. Utilities should maintain consistent operational data logs for both treatment and control facilities to enable such analyses. Future upgrades should prioritise the components most linked to the identified yield gains. ",
Diop et al. (Tue,) studied this question.