Environmental Engineering Reference
In-Depth Information
10.4 Mathematical process modelling and optimisation
in practice
10.4.1 Mathematical modelling: a second option for process
understanding and optimisation
Mathematical modelling of AD can also help control the process and avoid
significant instabilities during fermentation. Such models can improve the
understanding of biological processes. The correct evaluation and applica-
tion of models has to follow particular stages in order to become an efficient
tool. The main objectives of using a mathematical model are (Donoso-
Bravo et al., 2011)
.
understanding the system's behaviour and the role of each particular
component
quantitatively verifying stated hypotheses
.
predicting system behaviour in the future.
.
The model should also be balanced between describing the important
characteristic of the systems and keeping a reasonable level of complexity
(Nopens et al., 2009). Several factors cause inhibition and failure of the
process, for example overloading, underloading, inhibitory compounds,
sudden temperature rise/drop, etc. (Mata-A
´
lvarez et al., 2000; Angelidaki,
2002). Hence, it is difficult to optimise the design and operation of the
process in order to achieve the maximum performance. Pilot testing requires
a long period of time and is costly, which is why application of mathematical
models for predicting the performance of the process is of great interest
(Parker, 2005). Models can be useful for understanding the numerous
processes and microorganisms involved in AD. One needs to remember that
a model first has to be calibrated but afterwards it should be able to predict
reactor behaviour under changing conditions (Koch et al., 2010). Over the
years a range of models has been developed. The first attempts dealt with the
steady state and assumed a rate-limiting step (Lawrence, 1971) but more
complex models have since been presented (e.g. Lyberatos and Skiadas,
1999). The advantages of energy-based models are illustrated by Rodriguez
et al. (2009) where the proposed modelling approach was based on a single
metabolic network (as a representation of anaerobic microbial ecosystem)
and on using a maximum-energy-yield selective force to define the reaction
fluxes as a function of environmental factors.
Search WWH ::
Custom Search