Benchmarking techno-economic performance of bioenergy generation and utilization by water utilities: an efficiency analysis tree approach

The wastewater sector faces a growing need to transition towards energy-neutral operations. Despite this context, no dedicated benchmarking framework exists for assessing the techno-economic efficiency of bioenergy generation in water utilities. This study addresses that gap by introducing the Bioenergy Efficiency Index (BioEI), a novel synthetic indicator estimated using the Efficiency Analysis Tree (EAT) method. It is a non-parametric machine learning approach that integrates classification and regression tree algorithms with production frontier analysis, avoiding the overfitting limitations of non-parametric methods. The BioEI is applied to a panel of 50 observations covering the ten water and sewerage companies in England and Wales over the period 2020–21 to 2024–25. The input is total bioresources cost; the desirable outputs are energy generated and used internally and energy exported; the undesirable output is energy generated but unused. Results reveal substantial and persistent heterogeneity in bioenergy performance, with a sector-average BioEI of 0.555 and with no discernible sector-wide improvement over the study period, which coincides with the PR19 bioresources price-control period. The EAT identifies sample-specific expenditure partitions that define different cost-conditional bioenergy benchmarks. Unused energy averages approximately 60,741 MWh per utility-year. Sensitivity analysis using free disposal hull and data envelopment analysis confirms substantial performance heterogeneity and the absence of sustained sector-wide improvement, but reveals considerable estimator dependence in absolute efficiency scores and individual utility rankings.