Biomedical Engineering Reference
In-Depth Information
The difficulties encountered when applying biofilm models have been further discussed
by
Morgenroth et al.
[36]:
• Biofilm models are too detailed but leave out important factors. Biofilm models have
become more and more complex, taking into account an increasing amount of details
at the micro-scale. Practitioners are not interested in micro-scale details unless the
detailed information becomes directly important for the macro-scale performance of
the plant. Thus, for practical applications simpler biofilm models have to be derived
from the more complex models. On the other hand, the available models often
disregard important processes like attachment and detachment.
• The purpose of mathematical modelling is unclear and many practitioners do not see
the need for using mathematical models.
• Too many biofilm models are available. A variety of modelling approaches are
available and even biofilm modellers are sometimes in doubt what model to apply for
what purpose.
• Model calibration is difficult. A large number of input parameters is required by most
models but parameters are often very difficult to determine. Guidance for model
calibration is often not provided.
Since biofilm models have found little application in engineering practice so far, certain
needs have emerged in the biofilm modelling community [36]:
• Models are needed for predictions of dynamic responses to influent variations.
• Operators of wastewater treatment plants need models for trouble shooting and plant
optimization.
• Models are needed that integrate the multiple processes, e.g. particle removal, carbon
oxidation, nitrification, denitrification, and biological phosphorous removal, thereby
helping to understand the complex interactions between these processes.
• Models are needed for reactor design and testing of reactor configurations. These
models could be used to evaluate data from pilot-scale plants and to predict the
performance of planned full-scale plants.
There is a trend towards simplified biofilm models with less complexity, which can
overcome the difficulties and which can satisfy the needs described above. In the past this
trend has pointed from three-dimensional (3D) towards one-dimensional (1D) biofilm
models, although it was known, that 1D biofilm models with purely diffusional mass
transport may paint an inadequate picture of biofilm structure and performance, since
biofilms have a complex 3D structure with pores and channels, which might allow convective
mass transport into the biofilm (see chapter 3). In this context the performance of 1D and 3D
biofilm models has been evaluated [37]. It was found that a 1D model was suited to
approximate average concentration profiles of 3D simulations. This points towards 1D
models for engineering applications, because simplified models are needed for this purpose.
In line with the trend towards biofilm models with less complexity a zero-dimensional
(0D) biofilm model has recently been developed for dynamic simulation of moving bed
bioreactor (MBBR) systems [38-41]. The model was successfully applied to a pilot-scale
