Power-distribution systems in many developing nations face significant efficiency challenges, leading to substantial technical and commercial losses. In Uganda, the performance of these systems is under-documented, with a lack of robust methodological frameworks for evaluating equipment yield at a national scale. This study aims to methodologically evaluate the performance of power-distribution equipment and to quantify the determinants of system yield improvement using a multilevel modelling approach. A hierarchical dataset was constructed from operational records of transformers, conductors, and switchgear across multiple regions. A three-level mixed-effects regression model was employed, specified as Y₈₉₊ = ₀ + X₈₉₊ + u₉ + v₊ + ₈₉₊, where u₉ and v₊ are random intercepts for district and supplier, respectively. Inference was based on robust standard errors. The analysis indicates that conductor upgrades and transformer load management explain 34% of the variance in yield improvement. A one-standard-deviation increase in preventive maintenance frequency was associated with a 7. 2% yield increase (95% CI: 5. 1% to 9. 3%). Regional random effects were significant, highlighting infrastructural heterogeneity. System yield is predominantly influenced by targeted equipment interventions and maintenance regimes, with substantial variation attributable to geographical and procurement tiers. Utilities should prioritise conductor modernisation and structured maintenance programmes. Policy should support standardised data collection to facilitate future hierarchical analyses. power distribution, multilevel regression, system yield, technical losses, infrastructure, Uganda This paper provides a novel hierarchical modelling framework for disaggregating yield determinants in power networks, offering a replicable methodology for other developing economies.
Kato et al. (Fri,) studied this question.