To calibrate or not to calibrate, that is the question

Area of Science:

• Environmental Engineering
• Process Control
• Measurement Science

Background:

• Sensors are crucial for efficient operations in water resource recovery facilities.
• Current sensor calibration lacks theoretical understanding of measurement uncertainty's impact on control.
• Wastewater treatment processes often rely on sensors with unknown accuracy.

Purpose of the Study:

• To investigate how sensor accuracy is affected by calibration practices, including varying calibration thresholds.
• To develop a theoretical model for analyzing sensor calibration errors.
• To provide insights into optimizing sensor calibration for improved wastewater treatment control.

Main Methods:

• Development of a simple theoretical model with closed-form expressions.
• Modeling based on variance and bias in sensor and laboratory measurements.
• Simulation of the theoretical model to analyze calibration effects.

Main Results:

• Calibration thresholds showed no practical benefit, except when calibration is time-consuming.
• Consistent sensor calibration, rather than threshold-based, yielded the best accuracy.
• Sensor calibration errors directly transfer to the process, impacting control accuracy.

Conclusions:

• Minimizing laboratory measurement uncertainties is critical for accurate sensor calibration and process control.
• A shift from sequential detection-correction to an estimation approach for sensor bias is recommended.
• Accurate sensor measurements are essential for efficient and reliable water resource recovery operations.