In semi-arid regions such as Morocco, sustainable water management requires greater energy efficiency in treatment infrastructure. Bacterial beds, which utilise natural ventilation, are a low-energy technology well-suited to these contexts. Volumetric energy intensity (kWh/m3) is the standard benchmark, but its interpretation can be skewed in plants operating below capacity. This study assesses this limitation at the Khénifra wastewater treatment plant (WWTP), where the biological stage relies on natural ventilation and electricity demand is concentrated in pumping and auxiliary systems. The results reveal a marked decoupling between the hydraulic load and electricity demand. While the flow rate varied by 16.7 % (CV) in 2023, daily electricity consumption remained virtually constant at around 910 kWh/day (CV 2.5 %), showing no positive correlation with the volume treated. The wastewater treatment plant is fully compliant with standards, achieving removal rates of 89.8 % for BOD₅ and an energy intensity of just 0.123 kWh/m³. However, between 2015 and 2023, a paradoxical discrepancy was observed: the volumetric ratio improved by 14 %, whilst consumption per unit of pollutant removed (kWh/kg BOD5) increased by 45 %. This divergence is consistent with operation at 26 % of nominal organic capacity on a population-equivalent basis and with progressive dilution of the influent. The results indicate that the kWh/m3 ratio alone may give an incomplete and potentially misleading picture of performance in underloaded bacterial beds, since it responds to dilution and throughput as well as to treatment effort. For reliable benchmarking, it is essential to combine volumetric indicators with indicators normalised by the pollution removed (kWh/kg) and to systematically specify the capacity utilisation rate.