Parameter sensitivity analysis for activated sludge models No. 1 and 3 combined with one-dimensional settling model.

Area of Science:

• Environmental Engineering
• Wastewater Treatment Technologies
• Biochemical Modeling

Background:

• Activated sludge models (ASM1 and ASM3) are crucial for simulating wastewater treatment processes.
• Parameter uncertainty and influent composition can significantly impact model accuracy and calibration efforts.
• Developing robust sensitivity analysis techniques is essential for reliable effluent quality prediction.

Purpose of the Study:

• To propose a sensitivity analysis technique for activated sludge models (ASM1 and ASM3) that minimizes calibration needs.
• To evaluate techniques for their ability to predict effluent quality despite influent variations and parameter uncertainty.
• To identify a reliable and practical method for parameter sensitivity analysis in wastewater treatment modeling.

Main Methods:

• Compared three sensitivity analysis techniques: SVM-Slope, RVM-SlopeMA, and RVM-AreaCRF.
• Analyzed parameter sensitivities for ASM1 and ASM3, including settling model parameters.
• Estimated sensitive parameters using a genetic algorithm and evaluated simulation results using deltaEQ.

Main Results:

• SVM-Slope demonstrated consistent sensitive parameter identification and smaller deltaEQ for ASM1 across tested conditions.
• For ASM3, while no single technique consistently identified parameters, SVM-Slope still provided reliable deltaEQ.
• SVM-Slope was identified as the simplest and easiest methodology among the tested techniques.

Conclusions:

• The SVM-Slope technique is a suitable approach for sensitivity analysis in both ASM1 and ASM3.
• This method offers reliable effluent quality prediction and simplifies calibration for activated sludge models.
• The simplicity and effectiveness of SVM-Slope make it a practical choice for wastewater treatment modeling.