Monitoring, modeling, and control of nutrient removal in the activated sludge process

Abstract

The activated sludge process has been one of the most widely used biological processes for treating wastewater containing inorganic and organic pollutants. However, increased process complexity in the activated sludge system makes it more vulnerable to external disturbances, such as large variations in flow. Under these circumstances, analysis in mathematical form stands out with the potential benefits of improving understanding of process performance under dynamic conditions, and optimizing operation to treat greater volumes of wastewater, to deal with higher variability in influent load, and meet even more stringent discharge standards.

The objectives of this research are accordingly (i) to develop an understanding of the dynamic behavior of the biological nutrient removal processes in the activated sludge process and (ii) to determine efficient control strategies to achieve more reliable plant operation.

The dissertation begins with an extensive review of (i) dynamic behavior of the process; (ii) dynamic models of the activated sludge processes; and (iii) control schemes applied to the system. Collection of high-quality data at Athens Wastewater Treatment Facility No. 2 for an extended period of time is then described. The University of Georgia's Environmental Process Control Laboratory (EPCL) demonstrates enormous potential in retrieving high-frequency, high-quality field data, which is a prerequisite for success in the subsequent development of process models and process control strategies for these systems. A high-order model was then built up to simulate the nutrient removal processes, including especially a modified means of characterizing transport and mixing of both solute and particulate matter in the activated sludge system. Subsequently, the calibrated model was validated to test its robustness of under conditions different from those of calibration. The model is considered to be acceptable in its performance and thus adequate for its subsequent application in studies of process control. Using the model thus identified, a detailed assessment of control strategies for the activated sludge process has been conducted. This is focused on storm event control. The manipulated control variables are confined to those routinely used, i.e., recycle rate, wastage rate, step-feed, and step-sludge. Even so, the operational flexibility of the activated sludge process can thus be fully exploited.