Application of discrete lumping methods in modeling complex systems is a practice that dates back nearly half a century. This technique is presently used to give researchers a better understanding of experimental data or to model industrial scale processes. In complex chemical systems where many reactions occur between a high number of species, it is often necessary to use some form of lumping in order to obtain a simpler kinetic model whose parameters can actually be determined from the available data. Designing a lumped reaction network, however, involves dealing with great degrees of freedom. For instance, one has to define the actual pseudocomponents, the reactions that take place in the system and finally, reliable kinetic parameters have to be identified for these reactions. More often than not, this results in high uncertainties regarding these kinetic models as well as in their limited usability in carrying out experimental design or other process engineering tasks. This point of interest is more often than not ignored in various related literatures.
In some cases, such as: interim measurements or preliminary experimental design procedures, the complex methods are not applicable simply because there are no detailed measurement data available. This limitation hinders realization of the optimum utilization of the model. To address this, a team of researchers from the University of Pannonia in Hungary: Zoltán Till (PhD candidate), Dr. Tamás Varga, János Sója (PhD candidate),, Dr. Norbert Miskolczi and Dr. Tibor Chován conducted a study where they applied carefully selected set of tools and methods to establish significantly more reliable lumped kinetic models. Their work is currently published in the research journal, Chemical Engineering Journal.
In their approach, the research team applied and evaluated five different global sensitivity analysis methods based on their performance, using the case studies of real plastic waste pyrolysis and vacuum gas oil hydrocracking. The details of their methodology involved the adoption of global sensitivity analysis methods and their application on two kinetic models with discrete lumps as case studies. Lastly, based on the two case studies, the researchers estimated the parameter confidence of their approach.
The authors reported that a lower uncertainty of the underlying kinetic model was critical to achieve proper reactor design or operation. Moreover, they highlighted that their proposed methods could contribute significantly towards the realization of this objective. In addition, the researchers further reported that global sensitivity analysis was an effective tool in carrying out the model reduction step as it required minimal information about the kinetic parameters; hence, it could be implemented before the actual parameter identification step.
In summary, the key idea behind this research was that the uncertainty of lumped kinetic models could be significantly diminished by reducing the size of the reaction network while at the same time preserving the ability of the model to describe the observed behavior of the given system. Remarkably, the authors here reported that the evaluated five global sensitivity analysis methods could be effectively used to construct lumped reaction networks with fewer parameters to be estimated with narrower confidence intervals (i.e. lower model uncertainty). In a statement to Advances in Engineering, Dr. Tamás Varga pointed out that their work presented the inaugural step towards the construction of reliable lumped kinetic models with a gradually increasing number of species included, while keeping the uncertainties in the model as low as possible.
Zoltán Till, Tamás Varga, János Sója, Norbert Miskolczi, Tibor Chován. Reduction of lumped reaction networks based on global sensitivity analysis. Chemical Engineering Journal, volume 375 (2019) 121920.