Dynamic models and expert systems for the activated sludge process

The objective of this research was to develop dynamic models and expert systems to assist the design, optimization, operation, and control of the activated sludge process. The research effort focuses on two types of activated sludge processes--the high-purity oxygen activated sludge and petroleum refinery activated sludge.; The new high-purity oxygen activated sludge process model developed in this research, HIPURE2, is the first such model that uses the popular Activated Sludge Model No. 1, which is the activated sludge model developed by the International Association on Water Quality Task Group on Mathematical Modeling for Design and Operation of Biological Wastewater Treatment. HIPURE2 fits the data from a pilot plant and a full-scale plant very well using a single set of stoichiometric and kinetic coefficients. The comparison of HIPURE2 with a previous model, HIPURE, reveals that both of them are capable of describing the oxygen utilization patterns realistically though oxygen uptake rates are modeled very differently in these two models. A model calibration algorithm was developed and evaluated. The algorithm was capable of estimating model parameters reliably and accurately. Algorithm evaluation provided valuable information on the influences of a number of factors on model calibration.; An activated sludge model was developed for a petroleum refinery activated sludge process which performs nitrification as well as model carbon oxidation. The model predicts effluent COD, ammonia, nitrite, nitrate, sulfide, and sulfate concentrations as well as oxygen uptake rate, air flow rate (aeration), and sodium hydroxide dosage (pH control).; An expert system was developed to improve the operation and control of an activated sludge wastewater treatment process in an integrated petroleum refinery. The expert system emphasizes the integration and interactions among process schematics, data management, report generation, data trending, and process diagnosis. The expert system indicates the location of a detected process failure through alarms. The expert system allows direct entry of monitoring data into a spreadsheet like template with checking for implausible values, data management and trending with a user friendly interface. The knowledge base is generic and structured which enables easy modification and expansion.; The concept of integrating process model and expert system was developed for the effluent treatment plant in the model refinery. It is illustrated that integrated expert system is a powerful tool to improve operation and control. It is shown that an integrated expert system can help find operation problems, diagnose causes, and provide both qualitative and quantitative control advice to the operators. Process efficiency and reliability can be enhanced substantially.