Multidimensional efficiency evaluation of Africa's energy system, accessibility, and low-carbon performance

Africa's substantial potential for future CO2 emissions makes a comprehensive energy-efficiency evaluation essential for identifying utilization gaps and inform sustainable development. We build a three-dimensional framework—energy system, energy accessibility, and low-carbon performance—and, as a first step, implement a dual clustering strategy: K-means to group countries and R-type clustering to screen indicators within each dimension. We then estimate efficiencies for 32 countries (2000–2022) using a super-efficiency slacks-based measure Data Envelopment Analysis (SBM-DEA) model aligned with the three dimensions. And leveraging the country clusters, we use Monte Carlo weighting to allocate dimension weights and compute a comprehensive efficiency. To interpret the determinants of efficiency, we paired the DEA results with eXtreme Gradient Boosting (XGBoost) and SHapley Additive exPlanations (SHAP) analyses. Comprehensive efficiency remained low and largely flat over 2000–2022. System efficiency is relatively stable, with Cluster 1 outperforming the others. Accessibility efficiency has risen gradually; Cluster 1 is high while Cluster 3 remains low. Low-carbon efficiency is low in Cluster 1 but higher in Cluster 3, though it has declined in recent years. The interpretability results indicate that energy exports and grid infrastructure drive Cluster 1's system performance; improvements in electricity access are the primary driver of Cluster 2's accessibility; and despite infrastructure constraints, Cluster 3's low emissions and renewable energy shape its low-carbon efficiency. This study provided a three-dimensional evaluation and an interpretable driver attribution framework to support differentiated energy-policy design and targeted transition pathways across African countries.