Model-based energy optimisation of a small-scale decentralised membrane bioreactor for urban reuse.

The energy consumption of a small-scale membrane bioreactor, treating high strength domestic wastewater for community level wastewater recycling, has been optimised using a dynamic model of the plant. ASM2d was chosen as biological process model to account for the presence of phosphate accumulating organisms. A tracer test was carried out to determine the hydraulic behaviour of the plant. To realistically simulate the aeration demand, a dedicated aeration model was used incorporating the dependency of the oxygen transfer on the mixed liquor concentration and allowing differentiation between coarse and fine bubble aeration, both typically present in MBRs. A steady state and dynamic calibration was performed, and the calibrated model was able to predict effluent nutrient concentrations and MLSS concentrations accurately. A scenario analysis (SCA) was carried out using the calibrated model to simulate the effect of varying SRT, recirculation ratio and DO set point on effluent quality, MLSS concentrations and aeration demand. Linking the model output with empirically derived correlations for energy consumption allowed an accurate prediction of the energy consumption. The SCA results showed that decreasing membrane aeration and SRT were most beneficial towards total energy consumption, while increasing the recirculation flow led to improved TN removal but at the same time also deterioration in TP removal. A validation of the model was performed by effectively applying better operational parameters to the plant. This resulted in a reduction in energy consumption by 23% without compromising effluent quality, as was accurately predicted by the model. This modelling approach thus allows the operating envelope to be reliably identified for meeting criteria based on energy demand and specific water quality determinants.