Modelling an SBR waste treatment system.

Abstract

The sequencing batch reactor (SBR), developed from batch reactor technology, has made significant progress in the past two decades with increasing use in wastewater treatment. However, there are still some important aspects of the system that need improvement. Modelling SBR system, developing lab methods for estimation of the model parameters, re-defining SBR design parameters and comparing SBR and continuous flow stirred tank reactor (CFSTR) systems were the four objectives of this thesis. Previous SBR models cannot predict an entire SBR operation cycle because they have some obvious deficiencies. No consideration of soluble microbial products (SMP) by the models is one of the deficiencies. These deficiencies were carefully reviewed and discussed before beginning the modelling work. A new mathematical model based on the standard Lawrence and McCarty model and IAWPRC model No. 1 has been developed for the SBR activated sludge system. In contrast with previous SBR models, the most important developments of the new model are involving expressions for SMP formed from substrate metabolism and released from activated sludge decay and classifying different basic chemical oxygen demand (COD) categories based on their biodegradabilities. In the new SBR model, total COD is divided into two parts: soluble and insoluble. The soluble part is further subdivided into easy-to-biodegrade substrate (EBS), difficult-to-biodegrade substrate (DBS) and biologically inert organic materials (IOM). The classification of COD makes it possible to describe the degradation process of organic substances more precisely. Quantitative batch test methods for the classification of the substrate have been devised to aid in characterizing and developing the coefficient values of the new SBR model. The batch test techniques were employed for measuring the three different (EBS, DBS and IOM) soluble COD fractions in the influent as well as the formation or release of these three types of substances from metabolism and sludge decay. The kinetic constants of the new model can also be estimated by the techniques. It was found that EBS degradation followed Monod kinetics and DBS degradation followed first-order kinetics. It was also found that the soluble COD released from substrate metabolism is closely related to the type of substrate degraded while the soluble COD released from activated sludge decay is not. (Abstract shortened by UMI.)