Real Time Process Control Algorithms Combining Process Modeling and Instrumentation
- Joe Rohrbacher - Hazen and Sawyer
The purpose of this paper is to outline comprehensive process monitoring algorithms that combine process modeling with instrumentation for all of the following processes:
- Comparing real-time aerobic SRT to required SRT as a function of pH, DO, and temperature
- Optimized denitrification and biological phosphorus uptake in real-time through pacing of internal recycle flows to nitrate leaving the first anoxic zone
- Pacing chemical phosphorus addition to real-time orthophosphate concentrations
- Pacing supplemental carbon addition to real-time nitrogen load entering the anoxic zone
- Providing guidance systems to manage the number of clarifiers online as a function of influent flow, RAS flow, SVI, and MLSS concentration
- Utilizing equalization volume to provide a constant mass of ammonia to the BNR process
1. Aerobic SRT Monitoring for Nitrification
Assuming uninhibited growth and proper DO control, three key parameters will determine the required aerobic SRT to maintain full nitrification—pH, DO and temperature. The modeling/operations link that would most benefit nitrification in real-time is an ongoing comparison of actual and minimum required aerobic SRT for complete nitrification.
Real-time aerobic SRT can be calculated using solids analyzers to measure at least one of either waste activated sludge or the mixed liquor suspended solids (MLSS) concentrations. The other value can be a daily average from a laboratory. DO concentrations, pH, and temperature can be obtained from DO and/or pH probes.
2. Solids Separation
The solids separation process is a function of the following independent variables: mixed liquor suspended solids concentration, X; hindered settling coefficient, K; settling velocity, Vo; forward flow, Q; return activated sludge (RAS) flow, Qras; and the number of clarifiers in service, total surface area, SA. State point analysis (SPA) is commonly used to provide a simplified determination of whether or not a clarifier will fail under a given set of conditions. Applying this tool to real-time process control involves solving systems of equations with advanced mathematics and software tools. While that is a long-term objective, a simplified empirical approach was applied on a recent project. A matrix of potential Q, Qras, X, and SVI values was modeled using SPA. The number of clarifiers in service was varied to cover the range of options. Next, each SPA was solved by adjusting X manually until the clarifier was just at the brink of failure but not failing.
A multivariable linear regression (MVLR) equation was developed to represent the required clarifier surface area as a function of variables the plant could measure in real-time: MLSS, Q, Qras, and sludge volume index (SVI), which is measured daily as is a function of Vo and K. Site specific testing was conducted at the facility to determine Vo and K.
This equation can easily be integrated into the SCADA system and provide real-time information to operators about the number of clarifiers they ought to have in service. This equation will also be used to provide prompts to the user about how an increase in RAS capacity could improve settling since insufficient RAS flow can limit capacity during specific events.
For more information or a copy of the full paper, please contact the author at email@example.com.
Hear about new publications with our email newsletter
Horizons showcases significant water, wastewater, reuse, and stormwater projects and innovations that help our clients to achieve their goals, and can help you achieve yours. Articles are written by top engineers and process group leaders, demonstrating and explaining the beneficial application of a variety of technologies and tools.