Optimization of nitrogen removal through an intelligent automated operational strategy based on real-time process...

In this study, an intelligent automated operational strategy (IAOS) was developed and evaluated to enhance nitrogen removal efficiency in an anaerobic-anoxic-oxic (A2O) membrane bioreactor (MBR). To effectively respond to fluctuations in inflow load, an intelligent algorithm was developed to calculate optimal operating parameters in real-time. These parameters include the internal return (IR) ratio, anoxic hydraulic retention time (AnHRT), aerobic hydraulic retention time (AHRT), and mixed liquor suspended solids (MLSS) in the bioreactor. The calculations were performed using standard formulas, and their suitability was assessed using the activated sludge model (ASM 2d). A data-driven model (machine learning) was introduced to address potential inaccuracies that may arise when applying the steady-state outputs of ASM 2d to dynamic operating conditions in the field, thereby enhancing the accuracy of the simulation. When the developed IAOS was applied to the pilot plant, the automatic control of the IR ratio alone resulted in a 5.7 % improvement in nitrogen treatment efficiency compared to manual operation. Furthermore, with the installation of a swing reactor and the automatic control of all calculated parameters, nitrogen removal efficiency improved by 11.3 % compared to manual operation. These results demonstrate that automating the entire process is more effective in enhancing nitrogen removal efficiency than automating individual unit processes, given the complex characteristics of biological reactions.