On Modeling the Dynamics of Fixed Biofilm Reactors. With focus on nitrifying trickling filters

Doktorsavhandling, 1999

Fixed biofilms, which are matrix-enclosed populations of organisms attached to solid surfaces, are increasingly used in environmental biotechnology processes, such as the treatment of potable water and wastewater. This thesis aims to give some insight into the dynamic modeling of biofilms and fixed biofilm reactors, with a focus on nitrifying trickling filters (NTFs) for the treatment of wastewater. However, most of the presented methods are applicable to fixed biofilm reactors in general. Some of them are also applicable to mathematically related areas, such as the modeling of porous catalytic reactors and immobilized cells. Various models are derived for different purposes and time-scales using a physically based multi-species model of biofilms, which takes into consideration the transportation phenomena inside the biofilm as well as the bacterial processes causing the population dynamics. Different operating modes of a biofilm reactor can be deduced from the origin of the dynamics. The slow dynamics are mainly caused by changes in biology, which can take days or weeks to change, and the fast dynamics are mainly caused by reactor hydraulics and transients of the dissolved components in the biofilm. The fast transients typically settle in less than a few hours after a change in operating conditions. Based on this division, simplified models and numerical solution methods are presented for (i) a steady state, (ii) a quasi-steady state, where the fast and slow modes are in steady states corresponding to two different operating conditions, (iii) fast dynamics, where the biofilm composition is constant and (iv) slow dynamics, where only the fast modes are in a steady state. A wide range of biofilm reactors can conveniently be modeled by units called continuously stirred biofilm reactors (CSBR), which are tanks with a homogeneous gas phase and bulk liquid, and a biofilm that varies with the distance from the substratum. Specifically, it is shown that the dynamics of NTFs can be modeled by cascaded CSBRs. A method to determine physically based rational transfer functions of low order, describing the fast dynamics of CSBRs, is derived. With these transfer functions an appropriate number of CSBRs, as well as other biofilm and reactor parameters, can be determined from residence time distributions. From comparisons between simulations and experiments carried out on different pilot scale NTFs, physical phenomena, such as adsorption and desorption of ammonium, denitrification, oxygen mass transfer, changes in bulk volume, are observed and quantified. Simulations of operating strategies indicate that the nitrifying capacity can be improved by regularly inversing the order of cascaded NTFs and by varying the individual flow through NTFs operating in parallel.