Collaborative project E2Fuels – Simulation and integration of OME-synthesis
Oxymethylene ethers (short: OME) are synthetic diesel fuels of the general structure CH3O(CH2O)nCH3 which have a very clean combustion and can be produced sustainably on a syngas basis. Recently, our group has developed an innovative production process for producing OME from methanol and formaldhyde. In this project, this novel process is integrated in a sustainable process chain starting from CO2 activation up to the final use in vehicles. We tackle this task with several partners from industry and academia.
Collaborative project NAMOSYN – Sustainable mobility with synthetic fuels
In collaborative research project with more than 30 partners from industry and academia, the production and use of synthetic fuels is studied in holistic fashion. We contribute by errecting a mini-plant for the production of oxymethylene ethers (short: OME), diesel fuels with a very clean combustion and produced sustainably, e.g. from renewable resources. The mini-plant is designed to follow an innovative process concept that was developed in our group within the last five years.
Modelling poorly specified mixtures with perturbation schemes
Poorly specified mixtures appear regularly in biotechnological and chemical processes. They contain unknown components or known components of unknown concentration. Goal of our work is to model the thermodynamic properties of such mixtures with an innovative approach that can be used in simulations during process design and optimization. Usually good thermo models of parts of the mixtures are available, e.g. the behaviour of selected key components in pure water. Our approach adopts these models and adds generalized perturbation schemes to reflect the influence of the unknown components on the activities of the key components. The developed perturbation schemes have a minmum number of parameters and can be combined with any thermo model (gE-models, EoS). In the particular project, the perturbation scheme is applied to model aqueous solutions in municipal waste water plants.
Machine Learning in conceptual process design
The conceptual process design is a highly complex task which requires the combination of several disciplines ranging from physico-chemical properties, process unit concepts, simulation techniques, optimization, and not least creativity. Computers serve as pure tools in this process. The process synthesis is led by the creativity of the human engineer. Several attempts have been undertaken to teach computers how to come up with a novel designs. These attempts have been mostly algorithmic so far using superstructures of heuristics. In this project, we explore techniques that allow computers to derive their own heuristics based on automated simulation experiments.