Biomedical Engineering Reference
In-Depth Information
Overall, the prediction and control of biofilm formation in porous media
remains dicult since the ability to observe biofilm development spatially
and temporally is severely limited even in laboratory systems. Mathemati-
cal modeling can aid in the development of biofilm-based technologies, but
the development of conceptual and mathematical models is not only limited
by a lack of highly resolved experimental data but also by computational
challenges.
5.5 A Few Notes on Modeling
The experimental observations summarized above clearly show that biomass is
often distributed heterogeneously in porous media. To properly model biofilm
processes in porous media, models should take into account the influence
of microscale heterogeneities and distributions. Unfortunately, it can become
computationally burdensome to model a large (meter to hundreds of meter)
scale system on the microscale. Hence, compromises have to be made with
respect to the desired accuracy at the microscale and the computational fea-
sibility for larger-scale systems.
5.5.1 Macroscopic versus Microscopic Models
Bulk-scale models using analytical or fast numerical solutions are compu-
tationally ecient and work well for cases where large volume averaging is
appropriate, that is, where microscale processes are negligible in relation to
the overall behavior of the system (Clement et al
.
1996). Such models are
capable of modeling bulk changes for parameters, such as porosity, specific
surface area, permeability, and dispersivity, in dependence of biofilm forma-
tion in porous media but in general they do not assume any specific pattern
for microbial growth but instead use macroscopic estimates of the average
biomass concentration.
Microscale models, which, in contrast, treat biofilm-affected porous media
as multidimensional on the pore scale, can potentially predict localized clog-
ging, which can have a significant effect on the overall hydrodynamics of a
system, but become computationally demanding if they are to predict the
effect of localized (i.e., pore scale) biofilm growth on the bulk properties of
the porous medium.
Over the past years, a number of papers have been published that address
the issues associated with bulk-scale modeling, while attempting to obtain
computationally ecient, appropriate descriptions of bioclogging processes in
porous media. The pore-network model approach allows for the simulation of
porosity and hydraulic conductivity changes without having to describe the
process of biofilm development on the microscale for every single pore (Dupin
Search WWH ::
Custom Search