The theoretical basis and quantitative evaluation of a new approach for modeling biofilm growth are presented here. Soluble components (e.g., substrates) are represented in a continuous field, whereas discrete mapping is used for solid components (e.g., biomass). The spatial distribution of substrate is calculated by applying relaxation methods to the reaction-diffusion mass balance. A biomass density map is determined from direct integration in each grid cell of a substrate-limited growth equation. Spreading and distribution of biomass is modeled by a discrete cellular automaton algorithm. The ability of this model to represent diffusion-reaction-microbial growth systems was tested for a well-characterized system: immobilized cells growing in spherical gel beads. Good quantitative agreement with data for global oxygen consumption rate was found. The calculated concentration profiles of substrate and biomass in gel beads corresponded to those measured. Moreover, it was possible, using the discrete spreading algorithm, to predict the spatial two- and three-dimensional distribution of microorganisms in relation to, for example, substrate flux and inoculation density. The new technique looks promising for modeling diffusion-reaction-microbial growth processes in heterogeneous systems as they occur in biofilms.
Copyright 1998 John Wiley & Sons, Inc.