On-line estimation and detection of abnormal substrate concentrations in WWTPs using a software sensor: a benchmark...

Area of Science:

• Environmental Engineering
• Wastewater Treatment Technology
• Process Monitoring and Control

Background:

• Wastewater treatment plants (WWTPs) face challenges monitoring substrate concentrations (readily biodegradable S(s) and slowly biodegradable X(s)).
• Direct measurements of S(s) and X(s) are often unavailable in real-time WWTP operations.
• Influent variations, such as toxic or shock loads, can disrupt treatment processes.

Purpose of the Study:

• To propose a novel on-line monitoring approach for detecting abnormal S(s) and X(s) concentrations.
• To develop a software sensor that utilizes dissolved oxygen (DO) measurements for alarm activation.
• To address the challenge of unavailable direct S(s) and X(s) measurements in WWTPs.

Main Methods:

• Implementation of a software sensor based on an extended Kalman filter observer.
• Modeling disturbances using Fast Fourier Transform (FFT) and spectrum analyses.
• Utilizing dissolved oxygen measurements for estimating S(s) and X(s) concentrations.

Main Results:

• The software sensor achieved response times of approximately 60 minutes for S(s) and 90 minutes for X(s).
• The extended Kalman filter demonstrated fast and accurate convergence within 2 hours in case studies.
• Performance evaluation highlighted the sensor's capability even without direct S(s) and X(s) measurements.

Conclusions:

• The proposed software sensor offers a viable solution for on-line monitoring of substrate concentrations in WWTPs.
• It provides timely alarms for process disturbances, enhancing operational stability.
• Further discussion addresses estimation challenges and potential applications in broader wastewater monitoring scenarios.