Economic Model Predictive Control for optimal struvite recovery

Abstract

Resource recovery from municipal wastewater has been a prime focus for a decade. Although several recovery processes already exist in the market today, the high cost of material, inherent disturbance in the influent quality, lack of real time monitoring of critical parameters, and lack of a robust automation system may result in suboptimal performance. This work attempts to construct a model based predictive control for optimal operation of a struvite recovery unit in a full scale WRRF. A multi-parameter based predictive control has been developed by implementing an Economic Model Predictive Controller (EMPC) for optimal dosing of magnesium hydroxide in a struvite recovery unit. The EMPC used customized objective function for real-time optimization of performance and economical parameters of the crystallization unit. The effectiveness of the proposed EMPC controller is verified through tests conducted on the Benchmark Simulation Model No. 2 (BSM2d.). The results obtained from the simulator-based evaluation of EMPC demonstrate a significant improvement in resource recovery at reduced operational costs. The economic advantages of implementing an EMPC compared to proportional and constant magnesium dosage has also been enumerated.