The persistent underperformance of the national power grid necessitates rigorous, data-driven methodologies for evaluating infrastructure investment. Current assessments of distribution equipment often lack the analytical depth to account for hierarchical cost structures and longitudinal performance degradation. This case study aims to develop and apply a novel multilevel modelling framework to evaluate the cost-effectiveness of power-distribution equipment systems. The objective is to quantify the influence of equipment type, environmental stressors, and maintenance regimes on long-term lifecycle costs. A longitudinal dataset of operational and financial metrics for transformers, switchgear, and cabling was analysed. A three-level linear mixed model was specified: y₈₉ = ₀ + ₁x₈₉ + u₉ + e₈₉, where u₉ represents random intercepts for geographical regions. Robust standard errors were used for inference on fixed effects. The analysis revealed a significant negative association between proactive maintenance expenditure and total lifecycle cost, with a £1 increase in scheduled maintenance correlated with a £3. 20 reduction in long-term cost (95% CI: £2. 85 to £3. 55). Environmental corrosion was the dominant predictor of cost escalation for coastal installations. The methodological framework provides a superior tool for capital planning, demonstrating that strategic maintenance investment is a primary driver of cost-effectiveness in power-distribution networks. Utilities should adopt multilevel regression for asset investment appraisal. Budget allocations must prioritise preventative maintenance programmes, particularly for equipment in corrosive environments. asset management, lifecycle cost, linear mixed model, power infrastructure, predictive maintenance This study introduces a novel application of multilevel regression modelling for infrastructure cost-effectiveness analysis, providing utilities with a statistically robust decision-support tool.
Lebo van der Merwe (Thu,) studied this question.