Development and calibration of bio-kinetic models for organic carbon and nitrogen biodegradation in an aerobic activated sludge system

摘要

Substrate removal mechanisms in a mixed culture of an activated sludge system are still a mystery that researchers have been trying to unravel by proposing anddeveloping models to interpret the observed experimental data. Activated Sludge Model Number 1 (ASM1) was first introduced to better understand the biochemicalmechanisms during carbon and nitrogen oxidation and was based on the assumption that the external substrate is consumed only for biomass growth. This model ignoredthe formation of intracellular polymers (storage products) in the biomass cell, though several researchers observed the phenomenon of storage of carbon sources and thesignificant role it played in the carbon removal process. As a consequence, the Activated Sludge Model Number 3 (ASM3) was formulated assuming that all the easily biodegradable substrate is first stored internally during the feast phase before being used for growth during famine conditions. However, experimental observations proved that both storage and growth occur simultaneously during thefeast phase. Consequently, the simultaneous storage and growth (SSAG) model was developed as an extension to ASM3 specifically to gain an in-depth understanding ofthe carbon oxidation process.While considerable investigation has been focused on model calibration using respirometric measurements of the oxygen uptake rate (OUR) during the aerobic biodegradation of substrate, model-based interpretation of titrimetric measurements has been limited. In biological systems, the oxygen consumption and correspondingpH changes occur simultaneously and can be easily monitored using respirometric and titrimetric experimental observations respectively. Attempts were, therefore,made to calibrate ASM1 using titrimetric measurements. During the subsequent successful development, consideration was given to the pH effects of carbon uptake,ammonia uptake for growth, CO2 production from carbon metabolism and the nonlinear carbon dioxide transfer rate (CTR) due to stripping. While respirometry was successfully used to calibrate every proposed new model, the interpretation of titrimetric measurements, however, was always based on ASM1, using calibration substrates such as acetate and ammonium. The SSAG model, developed in the year2005, proposed an improved kinetic expression for the degradation of storage products under famine conditions. It was successfully calibrated using on-line respirometric measurements and validated using off-line storage productsmeasurements with acetate used as a calibration substrate. However it failed to explain the titrimetric behavior of the substrate biodegradation process. While in most cases the above models were calibrated using simple synthetic substrates, the application of these models for complex substrates was not investigated.Therefore, in this dissertation, the biodegradation kinetics for different substrates like acetate (simple carbon source), sodium dodecyl sulfate (relatively complex carbon source), ammonium (simple nitrogen source), urea (relatively complex nitrogen source) and glutamic acid (combination of carbon and nitrogen) were investigated. Atitrimetric respirometer, established in the Water-waste laboratory in the Faculty of Engineering and Surveying, University of Southern Queensland, Australia, was usedto conduct batch experiments in order to monitor substrate biodegradation process in an aerobic activated sludge system. Both the dissolved oxygen and the pH controldata were logged with the Labview software package. A spreadsheet program was used to calculate the oxygen uptake rate and the proton production/consumption ratefrom the raw measurements.During batch experiments, the biodegradation of all five test substrates showed unique respirometric and titrimetric behaviors indicating that each of these compounds is biodegraded using distinctive mechanisms with the involvement and coordination of different bacterial populations. The pattern of OURs were observedvarying from substrate to substrate describing the characteristics of test compounds. The titrimetric profiles were also different for different substrates biodegradationreflecting the on-going biochemical reaction of respective substrate in activated sludge system. While acetate biodegradation, for example, caused protonconsumption (at pH 7.8) in the liquid medium, proton production was noted under feast conditions when either sodium dodecyl sulfate or glutamic acid was used as atest substrate.An in-depth revision of the existing activated sludge models was completed and the models were assessed using experimental observations. An improved bio-kineticmodel which includes both the oxygen and proton balances for the biodegradation of each of the test substrates was then developed. In addition, the substratesbiodegradation pathway and the non-linear CO2 transfer process were considered during the modeling. For proper model evaluation, the proposed model was calibrated using varying initial substrate concentrations and pH levels. In addition, three different calibration approaches: using respirometric measurements alone, using titrimetric measurements alone and using combined respirometric-titrimetric measurements, were applied during the study. The estimation of model parameters was undertaken using non-linear techniques utilizing the algorithms in theoptimisation toolbox (MATLAB).For the biodegradation of each test substrate, the proposed model was successfully calibrated using both the respirometric and titrimetric behaviors in the activatedsludge system. The estimated model parameters showed consistent results for all three calibration approaches thereby confirming the precision of the proposed model.The parameter estimation errors (calculated for 95% confidence intervals) as well as the mean squared errors for the different calibration approaches were quitereasonable and confirmed the statistical soundness of the proposed model. In addition, the proposed model was validated using off-line measurements.This dissertation presents an in-depth explanation of how the proposed models interpret the biodegradation processes and how the model parameters vary for different substrates in an activated sludge system. The results will be helpful in further refining current models that can contribute to the optimization of the design operation and enhanced performance of full scale wastewater treatment plants.