Experimental design for evaluating WWTP data by linear mass balances

Area of Science:

• Environmental Engineering
• Process Systems Engineering

Background:

• Wastewater treatment plants (WWTPs) require reliable data for effective operation and management.
• Identifiability of key process variables is crucial for accurate process assessment but often hindered by limited measurements.
• Linear mass balances provide constraints but may not be sufficient for full variable identifiability.

Purpose of the Study:

• To develop a stepwise experimental design procedure for WWTPs to ensure reliable data acquisition.
• To determine optimal sets of additional measurements that guarantee the identifiability of key process variables.
• To balance the accuracy of identified variables with the cost of additional measurements.

Main Methods:

• Developed a procedure to identify necessary additional measurements based on linear mass balance constraints.
• Employed multi-objective optimization to find solutions balancing variable accuracy and measurement cost.
• Utilized a Pareto-optimal front to represent trade-offs between accuracy and cost.
• Applied the procedure to a full-scale WWTP, analyzing measurement interdependencies and mass balance groups.

Main Results:

• Identified groups of overlapping mass balances, indicating which measurements are interdependent.
• Demonstrated that adding measurements only benefits identifiability within specific mass balance groups.
• Showcased significant reduction in computational effort by applying the procedure to individual mass balance groups.
• Validated the straightforward applicability of the procedure to diverse WWTPs.

Conclusions:

• The developed procedure effectively guarantees the identifiability of key process variables in WWTPs.
• Optimizing experimental design based on mass balance analysis improves data reliability and cost-efficiency.
• The method simplifies planning for monitoring campaigns and reduces computational load.
• This approach is adaptable for various WWTPs, regardless of existing data availability.