Increase in global energy demand has brought to a massive rise in greenhouse gases emissions. In particular, the continuously increasing atmospheric concentration of carbon dioxide is a major concern among the scientific community. This paved the way for an intense research of CO2 emissions mitigating technologies. Those ones including carbon capture, storage and utilisation are seen as part of the solution for this global problem. In this framework, Zero Emission Fuels (ZEF), a startup company located in Delft (The Netherlands) set the ambitious goal of developing a micro process plant producing methanol from ambient air and sunlight only. Part of the process involves direct capture of carbon dioxide from the atmosphere (direct air capture) with the use of a pure tetraethylenepentamine (TEPA) absorbent. ZEF sets itself aside from the rest of the industry developing an absorber which collects CO2 with no support structure by continuously flowing TEPA inside open channels. Previous research shows how this approach brings limitations connected to slowly diffusing CO2 molecules in the absorbent liquid film in proximity of the gas-liquid interface. In this framework, circulation of TEPA particles from the gas-liquid interface into the bulk of the flowing absorbent is seen as a viable solution to improve the process. This thesis focuses on investigating this hypothesis by inducing mixing on the liquid side. That is done by building two experimental setups investigating both passive and active mixing. Experimental results show that active mixing can be used to improve the rate of CO2 absorption, while passive mixing does not bring significant advantages. Results from the passive mixing experiments are further investigated by modelling the fluid dynamics of the process through a Direct Numerical Simulation of the particular Stokes’ flow in the engineered absorption channel. The process is characterised with the definition of a theoretical framework describing mass transfer in the liquid side. Following an analogy with ice formation on top of a frozen lake, this theory, also known as the ”Ice-Sheet” theory, shines some light on the way this diffusion limitations are happening at a molecular level. In particular, a highly viscous, heavily loaded layer of sorbent on the gas-liquid interface is believed to be the cause for observed CO2 diffusion limitations. This theory is backed-up by defining its mathematical equivalent in the form of a mass transfer model. Comparing the results of the model with experimental results, a very good agreement is observed. That is believed to add credibility to the proposed theory. Moreover, that is also found to be in line with the latest knowledge available in literature about CO2 absorption in TEPA films. Finally, the experimental results and the developed models are used to engineer a new iteration of ZEF’s absorber on a cost reduction basis.

In recent years, the energy transition has led to increasingly higher installed capacities of intermittent renewable energy sources. This has sparked the interest in power-to-fuel technology to store excess renewable energy in the chemical bonds of synthetic fuels. As a highly reactive, turn-key process, plasma-chemical conversion offers a promising approach to power-to-fuel technology. However, recent studies have shown limitations in either energy efficiency, conversion or flow rate. The novel Spinning Arc Plasma Reactor (SAPR) developed by J. van Kranendonk stabilises the plasma by mechanical rotation with the potential to contain a liquid film for in-situ product removal. The aim of this thesis is to characterise this reactor by studying the effect of rotational confinement on the reactor performance. To this end, the prototype initially built by J. van Kranendonk was further developed. A wealth of recent studies on the plasma-chemical conversion of CO2 as a step in the formation of synthetic fuels serves as a benchmark to compare reactor performance of the SAPR concept against. A series of experiments with argon were performed to establish the relationship between the system variables (pressure, power, flow rate and rotational speed) and the plasma parameters (electron temperature and electron density). Based on this knowledge, a series of experiments with CO2 were performed to establish the relationship between the reactor performance measured by the conversion and energy efficiency and the system variables: pressure, power and flow rate. The maximum conversion of about 30% to 55% was obtained at an energy efficiency of approximately 0.25% and the maximum energy efficiency of about 3% was achieved at a conversion of approximately 8.5% conversion. Similarly to other studies, improvements to either one came at the expense of the other. The performance of the SAPR proved comparable to dielectric barrier and other glow discharges but was outperformed by microwave and gliding arc discharges. Especially energy efficiency was inferior to these configurations. Further analysis of the discharge emission spectra suggested that CO2 dissociation mainly occurred through electron impact excitation. This explains the large difference in energy efficiency as microwave and gliding arc configurations are able to exploit more energy efficient dissociation channels. Nonetheless, this study was able to further develop the concept and has laid the ground work to explore the potential effect of a liquid film and in-situ product removal on the reactor performance.

Vapor-liquid phase equilibrium (flash) calculations largely contribute to the total computation time of many process simulation models. As a result, process simulations, especially dynamic cases, are limited in the amount of detail that can be included due to time restrictions. In addition, under certain conditions flash calculations can fail to provide acceptable results. In this work, artificial neural networks were investigated as a potentially faster and more robust alternative to conventional flash calculation methods. Classification neural networks were used to determine the phase stability of a given mixture of fluids, while regression networks were used to make predictions of thermodynamic property values. In addition to conventional flash types such as the constant pressure, constant temperature (PT), and constant pressure, constant entropy (PS) flash, neural networks are used to develop two concept flash types: a constant entropy, constant volume (SV), and a constant enthalpy, constant volume (HV) flash. All neural networks were trained on, and compared to, data generated using the PT-flash algorithm from the Thermodynamics for Engineering Applications (TEA) property calculator. Data was generated for mixtures of water and methanol over a wide range of pressures and temperatures. The artificial neural networks showed speed improvements over TEA of up to 35 times for phase classification and 15 times for property predictions. Overall phase classification accuracy scores of around 97% were achieved. Average property value prediction errors range between 0.5% and 7% when compared to the spread in magnitude of the test data, and R2 scores were in the general order of 0.95 and higher, although classification and property prediction in the two-phase region showed markedly higher errors than properties in the pure liquid or vapor regions. Moreover, thermodynamic consistency and the stability of a system consisting of multiple neural network flash types both still require considerable improvement. Finally, this work shows that artificial neural networks can be used to create unconventional flash types such a the SV- and HV-flash.