Abstract
{ "background": "Rwanda's power-distribution infrastructure faces significant reliability challenges due to ageing assets and increasing demand. Effective policy for infrastructure investment requires robust, forward-looking risk assessments to prioritise interventions and allocate resources efficiently.", "purpose and objectives": "This policy analysis evaluates a novel forecasting model designed to quantify risk reduction within the national power-distribution network. The objective is to provide a methodological framework for evidence-based policy planning and capital expenditure prioritisation.", "methodology": "A time-series forecasting model was developed, integrating historical failure data, maintenance records, and load growth projections. The core statistical model is an ARIMAX formulation: $yt = \\mu + \\sum{i=1}^{p}\\phii y{t-i} + \\sum{i=1}^{q}\\thetai \\epsilon{t-i} + \\sum{i=1}^{r}\\betai X{t,i} + \\epsilont$, where $Xt$ represents exogenous policy and engineering variables. Model parameters were estimated using maximum likelihood, with inference based on heteroskedasticity-robust standard errors.", "findings": "The analysis projects that a targeted policy of strategic asset replacement, informed by the model, could reduce the aggregate risk of catastrophic failure across the network by approximately 22% over the forecast horizon. The 95% confidence interval for this reduction is between 18% and 26%, indicating a statistically significant effect.", "conclusion": "The forecasting model provides a technically sound and transparent tool for linking engineering data to infrastructure policy outcomes. It demonstrates that data-driven policy can substantially mitigate systemic risk in power distribution.", "recommendations": "Policymakers should institutionalise the use of predictive, engineering-based risk models in national infrastructure planning cycles. A dedicated fund for data collection and model updating should be established to ensure long-term efficacy.", "key words": "infrastructure risk, predictive maintenance, ARIMAX modelling, capital planning, electrical grids", "contribution statement": "