http://hdl.handle.net/2268.2/22265 Thu, 20 Aug 2026 16:10:16 GMT 2026-08-20T16:10:16Z http://hdl.handle.net/2268.2/26206 Title: Master's thesis and Internship : Integration of a supercritical carbon dioxide cycle with a solid oxide electrolyser Abstract: The goal of this thesis is to assess the benefits of integrating a bottoming cycle with a solid oxide electrolyser. Two models of a supercritical carbon dioxide Brayton cycle are developed and subsequently integrated into the hydrogen production process. The integration is evaluated using both exergy and primary energy analyses. Finally, a scenario in which the net power output of the cycle is maximised is investigated. Sun, 28 Jun 2026 22:00:00 GMT http://hdl.handle.net/2268.2/26206 2026-06-28T22:00:00Z http://hdl.handle.net/2268.2/26205 Title: Master's thesis and Internship : - Resonant planar transformer modelling methodology in the application of an X-ray generator- Serel Abstract: Serel enterprise is a major sensor company which utilizes an X-rays generator and receptor (TSP), to measure the density of any textile material driven between both devices. In other words, rough textile roller is led towards the sensors, where X-rays run through it and are then received by the TSP. The TSP computes the X-ray composition after running through the textile and measure its density at a precise point. Allowing to monitor the motor responsible of aligning the textile fibers, in order to have a homogeneous fabric with a constant density. To this end, the generator requires high voltage in order to generate X-ray. Therefore, high voltage must be precisely generated, to do so, a conventional transformer with high transformation ratio is utilized, combined with a multiplier. Noticing how high the voltage is required to be, the resonance frequency is experimentally determined after its manufacture and is exploited. In this instance, conventional transformers are hardly accurate due to many factors such as the specific spire ratio and spacing, making it difficult to predict their intrinsic resonance frequencies. In this regard, planar transformers are a promising technology allowing their intrinsic characteristics to be rigorously predicted. To predict the fundamental electro-magnetic planar transformer’s behaviour, analytical and FEM analysis are concurrently undertaken, under a same simplified T model. Indeed, an analytical analysis details the geometrical limitations such as the amount of PCB layers and copper track dimensioning, leading to a first approximation of physical parameters such as track resistance, magnetizing inductance, core loss and intrinsic capacitive effects. FEM analysis is then performed with GetDP solver and Gmsh software to enhance the accuracy of the computed parameters and to compute the more complex ones such as the leakage inductance. Combining both analysis, results in a high precision resonance resolution and specifies the required modifications to perform in order to match given nominal requirements. Eventually, this only input of this approach is the nominal conditions such as transformation ratio, input voltage and magnetic core specifications, and outputs a fully designed planar transformer. The strength of this methodology is the planar transformer itself, making transformers precisely reproducible in large-scale manufacturing. Therefore, this thesis marks a significant accomplishment in the design of resonant planar transformers allowing to optimize the compactness and performance for any kind of application and more specifically for high voltage. This paper details a simplified methodology which combines efficiency and resonance monitoring for a given nominal frequency. Allowing to build any kind of planar transformers which intrinsic characteristics are tailor-made for a precise purpose. Sun, 28 Jun 2026 22:00:00 GMT http://hdl.handle.net/2268.2/26205 2026-06-28T22:00:00Z http://hdl.handle.net/2268.2/26192 Title: Master's thesis and Internship : - Analysis and evaluation of maximum power point tracking algorithms based on machine learning techniques- CE+T Power Abstract: Maximum Power Point Tracking (MPPT) and Global Maximum Power Point Tracking (GMPPT) consist of techniques that are employed to keep a photovoltaic (PV) array operating at its optimal point. This report first investigates the implementation of two Machine Learning (ML) algorithms, namely Support Vector Regression (SVR) and K-Nearest Neighbors (KNN), under homogeneous irradiance conditions. The performance of these two methods is analysed and compared with that of an improved Perturb and Observed (P&O) process. Several scenarios involving varying solar irradiance and temperature conditions were simulated to evaluate the tracking efficiency of the different methods. The obtained numerical results show that ML methods provide a significant advantage when a temperature measurement is included as input to the model, achieving tracking efficiencies higher than 99%, even in the presence of measurement noise. The case of partial shading is also investigated, highlighting the limitations of conventional MPPT techniques and the need for adapted strategies, such as a Q-Learning approach, to identify the global maximum power point. The results indicate that a key challenge lies in resetting the Q-table, which stores previously encountered states, in order to cope with dynamically evolving outdoor conditions. Finally, the concept of Flexible Power Point Tracking is briefly discussed as a potential extension of MPPT strategies. Sun, 28 Jun 2026 22:00:00 GMT http://hdl.handle.net/2268.2/26192 2026-06-28T22:00:00Z http://hdl.handle.net/2268.2/26189 Title: Master's thesis and Internship : - Modelling and Integration of a Hydrocracking Unit for Fischer-Tropsch Syncrude in a Power-to-Kerosene Process- Chemical Engineering (Université de Liège) Abstract: Aviation is among the hardest sectors to decarbonise: long-haul flight depends on high-energy-density liquid fuels for which direct electrification offers no realistic substitute. The Power-to-Liquids route, in which captured CO2 is combined with electrolytic hydrogen to synthesise drop-in jet fuel, represents a relevant pathway for reducing aviation’s dependence on fossil kerosene. In the low-temperature Fischer-Tropsch route over a cobalt-based catalyst, syngas is converted into a broad hydrocarbon distribution that is, however, weighted towards the heavy C17+ paraffinic wax fraction rather than the C8-C16 kerosene cut targeted by the process. Converting this heavy fraction into kerosene-range products therefore requires a dedicated upgrading step, and hydrocracking over a bifunctional catalyst is the technology retained for this purpose. The present work develops a hydrocracking reactor model and integrates it downstream of the Fischer-Tropsch (FT) unit of the Power-to-Kerosene process under development at the University of Liège (ULiège) within the Neutral-Kero-Lime project, completing the upstream chain previously modelled by A. Rouxhet and A. Morales. Three modelling approaches were evaluated successively. A stoichiometric model and a thermodynamic-equilibrium model were both shown to be structurally unable to reproduce the experimental hydrocracking behaviour, confirming that the reaction is kinetically rather than thermodynamically controlled. The Langmuir-Hinshelwood-Hougen-Watson kinetic model of Pellegrini et al. (2007) was consequently retained, because it resolves the product stream at the component level, a resolution required to track the C8-C16 cut without committing to a fixed lump definition, under the simplifying assumption of a single vapour phase combined with midpoint cracking. Two pieces of information omitted from the original publication had to be recovered in Python before the model could be applied: the inlet distribution, identified through hypothesis testing as corresponding to the distribution later reported by Pellegrini et al. (2008), and the final optimised kinetic parameters, reconstructed by regularised parameter estimation against the published model curves. The validated model was then applied to the ULiège FT hydrocarbon stream after olefin-hydrogenation pre-treatment, with the objective of identifying the temperature, pressure, weight hourly space velocity (WHSV) and hydrogen-to-wax ratio that maximise the C8-C16 mass fraction at the reactor outlet. Within the Pellegrini calibration domain, the procedure converges at T = 369.3°C, P = 60 bar, WHSV = 3.00 kg_n-C,in kg_cat^-1 h^-1 and H2/wax = 0.150 kg/kg, yielding 52.0 wt-% of C8-C16 at the outlet for a C17+ mass conversion of 55.9 %. At the pilot-scale throughput of the ULiège installation, applying standard plug-flow design criteria to this operating point yields a preliminary trickle-bed reactor geometry of approximately 26 mm internal diameter and 132 mm bed height, holding close to 38 g of bifunctional Pt/SiO2-Al2O3 catalyst. Three of the four coordinates lie on the upper bound of the calibration window, indicating that the model response would continue to climb beyond the calibrated range. Dimensional sensitivity analyses around this point expose two limitations of the model: the H2/wax ratio acts on the model exclusively through the hydrogen partial pressure, so the genuine two-phase character of the system is not captured; and the midpoint cracking stoichiometry underestimates the iso-paraffin yield relative to the experimental data. A comparison of the recommended conditions across feed selection (full C1-C70+ stream versus C17+ fraction only), exploration domain (author calibration versus extended literature range) and optimality criterion (strict versus relaxed economic ranking) further isolates the contribution of each modelling choice. One caveat conditions the reading: the kinetic parameters, calibrated on the Pellegrini C4-C70 feedstock and not refitted to the ULiège stream (for which no experimental data exist), make the reported yields and conditions order-of-magnitude indications and a transferable methodology, not definitive design values, though a check on the calibration feed itself suggests the trends carry over between feeds. Several perspectives follow naturally. At the reactor scale, replacing the single-phase partial-pressure assumption with a proper vapour-liquid description and the midpoint cracking with a breakage-probability stoichiometry would address the two structural limitations identified by the sensitivity analyses. At the process scale, integrating the model into the full Power-to-Kerosene flowsheet in Aspen Plus would enable a techno-economic assessment of the recommended operating point. Sun, 28 Jun 2026 22:00:00 GMT http://hdl.handle.net/2268.2/26189 2026-06-28T22:00:00Z