DSpace Repository

Welcome to the NWU Repository, the open access Institutional Repository of the North-West University (NWU-IR). This is a digital archive that collects, preserves and distributes research material created by members of NWU. The aim of the NWU-IR is to increase the visibility, availability and impact of the research output of the North-West University through Open Access, search engine indexing and harvesting by several initiatives.

Communities in NWU Repository

Select a community to browse its collections.

Now showing 1 - 5 of 18

Recent Submissions

Induction heating for catalytic dehydrogenation of liquid organic hydrogen carriers
(North-West University, 2025) Ford, DA; Modisha,PM; Krieg, HM; Bessarabov, DG
Hydrogen storage using liquid organic hydrogen carriers (LOHCs) is a promising solution due to their high hydrogen storage density, safe handling under ambient conditions and compatibility with existing crude oil infrastructure. However, large-scale implementation is limited by the high energy demands and uneven temperature profiles of the endothermic dehydrogenation reaction. Current heating approaches electrical heaters and hydrogen burners—suffer from heat losses, reducing the overall energy efficiency. Electromagnetic inductive heating offers a compelling alternative; it delivers rapid, volumetric heating by generating heat directly within inductively active materials and transferring it efficiently to the catalyst and surrounding fluid. This study explores inductive heating for LOHC dehydrogenation using perhydrobenzyltoluene (H12-BT) as a model compound and Pt/-Al₂O₃ catalysts synthesised via wet impregnation (0.3–8 wt.% Pt). Catalyst loading was confirmed using Inductively coupled plasma–optical emission spectroscopy (ICP-OES), and characterisation was performed using Brunauer–Emmett–Teller (BET), Carbon monoxide (CO) pulse chemisorption, scanning electron microscopy coupled with energy dispersive X-ray spectroscopy (SEM-EDS), Transmission electron microscopy (TEM), X-ray powder diffraction (XRD) and Hydrogen temperature programmed reduction (H₂-TPR). Four catalyst configurations were tested in an inductively heated reactor: (1) uncoated stainless steel (SS) pellets, (2) a mixture of SS and Pt/-Al₂O₃ pellets, (3) Pt/-Al₂O₃ coated SS pellets, and (4) a mixture of catalyst pellets and SS. The Pt/-Al₂O₃ coated SS pellets with 0.11 mol% Pt gave the highest performance: 0.35 gH₂/gPt/min productivity and 97.82% H12-BT conversion. Inductive and conventional electrical heating were compared using 5 wt.% Pt/Al₂O₃ coated SS pellets. Although electrical heating reached a higher maximum hydrogen flow rate (62.56 mL/min) and cumulative H₂ volume (2.03 L), inductive heating delivered a more stable profile (max flow rate: 27.78 mL/min, 1.65 L cumulative H₂), showing better thermal regulation. Among the catalyst loadings, 5 wt.% Pt gave optimal performance. Higher loadings (8 wt.%) led to decreased metal dispersion, larger particle sizes and pore blockage due to agglomeration. The influences of polymer binders, Polyvinyl alcohol (PVA) and Polyvinylpyrrolidone (PVP), and coating thickness on catalyst performance were also assessed. Binder molecular weight and concentration influenced coating uniformity, adhesion, and diffusion. PVP (MW 55,000) at 0.5 wt.% offered the best results (0.26 gH₂/gPt/min productivity, 65.39% conversion, 1.64 L H₂). Coating thickness effects were explored using eggshell Pt/-Al₂O₃ coatings on SS pellets. A thickness of 81 µm (5 wt.% Pt) maximised H₂ yield and minimised by-products, outperforming both the thinner (62 µm) and thicker (>100 µm) coatings. Gas chromatography–mass spectrometry (GC-MS) confirmed the dehydrogenation pathway via partially hydrogenated BT (H6-BT) to benzyltoluene (HBT). Using GC-MS with single quadrupole detection (GC-SQMS) and advanced gas analysis, no by-products were detected at conversions >90%. Catalyst stability over four cycles showed signs of deactivation: productivity declined from 0.302 to 0.120 gH₂/gPt/min and Degree of dehydrogenation (DOD) from 27% to 3%. Despite this, byproduct formation declined, indicating reduced catalytic activity-limited side reactions. Kinetic analysis showed that the 82 µm coating gave the highest reaction rate constant and turnover frequency value (28.12 min⁻¹), suggesting an optimal balance between surface area and mass transfer. Thorough characterization confirmed the synthesis and dehydrogenation efficacy of Pt/γ-Al₂O₃. Results from ICP-OES revealed that Pt loadings exceeded the intended levels, ranging from +1.67% to +17%. This suggests a minor overloading that could enhance the active sites, while also posing a risk for particle agglomeration, potentially affecting catalyst performance. SEM-EDS showed a transition from well-dispersed nanoparticles (NP’s) at 0.3 wt.% Pt to aggregated clusters at 8 wt.%. BET analysis indicated a reduction in surface area from 186.8m²/g to 169.3 m²/g, while TEM demonstrated NP growth from 1.11nm to 4.53 nm prereaction, increasing to 2.69nm and 6.91 nm post-reaction. XRD revealed sharper Pt peaks and enhanced crystallinity with increasing loadings, and CO pulse chemisorption indicated a decline in dispersion from 65.64% to 35.34%. Post-reaction HR-TEM indicated a contraction in d-spacing from 0.23 to 0.20 nm. TGA exhibited initial weight loss at ~200 °C (moisture removal) and primary decomposition between 300–450 °C, with PVP (MW 55,000) showing the most significant mass loss, while higher molecular weight binders exhibited slower degradation, suggesting enhanced thermal stability.
Predicting epiliptic seizures using machine learning applied to EEG data
(North-West University, 2025) Coetzer, RM
Epilepsy is a neurological condition that affects roughly 50 million people world wide. Although the majority of seizures themselves are not life threatening, the loss of motor control and consciousness has a major impact on the safety of a patient. While all-cause mortality decreased 16.4% from 8756.34 per million to 7319.17 per million between 1999 and 2017, the death rate for epilepsy climbed 98.8% in the USA, from 5.83 per million in 1999 to 11.59 per million [1]. If there was a method for accurately predicting when a person will have a seizure, this mortality rate would drastically decrease, since the person would abstain from driving or doing any other dangerous activity during a time period that has been predicted to have a higher chance of seizure onset. The aim of this dissertation is to use machine learning methods to analyse EEG signals from patients with epilepsy, and be able to predict with high sensitivity and specificity whether or not the person has a high chance of seizure onset during a certain time period. Based on previous research conducted on the topic, it is evident that leveraging machine learning techniques to identify and predict seizures holds considerable promise. This approach has been widely explored and documented in the literature, showcasing its validity and potential. One crucial aspect that stands out from the collective body of work is the significance of preprocessing and feature extraction in the overall process. Virtually all studies included in the literature review acknowledged the importance of these steps in enhancing the effectiveness and efficiency of their models. By carefully processing and extracting relevant features from the input data, researchers have consistently achieved improved performance and accuracy in seizure detection and prediction. Moreover, among the various machine learning algorithms investigated for this dissertation, convolutional neural networks (CNNs) have emerged as the most popular choice and the literature survey reveals a strong prevalence of CNNs in seizure detection studies. However it is worth noting that no study included in the analysis directly compares the performance of a CNN to other models using the same dataset, thereby revealing a notable research gap. Consequently, addressing this gap becomes a crucial next step in advancing the field of seizure prediction. Conducting comprehensive comparative studies that systematically evaluate the performance of different machine learning algorithms, including CNNs, on identical datasets would contribute significantly to our understanding of their relative merits. Such investigations would enable researchers to make informed decisions regarding the most suitable algorithmic approaches for seizure detection and prediction, thereby potentially enhancing the overall accuracy and reliability of these models. In this dissertation, EEG data is generated using a recurrent variational auto encoder, and the benefits of generated EEG data are discussed. A variational auto encoder is then used to compress EEG data to a latent space representation without losing any of the important features of the data. Next, a random forest ensemble model and an extreme gradient boost model are compared to see which model can most accurately classify EEG into two categories: ictal (seizure) or pre-ictal (non-seizure). Finally, three models are compared to find out which model can most accurately predict epileptic seizures. It was found that a combination of an LSTM and a CNN model was the best at predicting seizures, with the ability to predict forecast data 50 seconds into the future with a mean absolute error of 0.18.
Current switching device development to characterise and monitor a proton exchange membrane water electrolyser
(North-West University, 2025) Bornman, M.L.; Gouws, R; Kruger, GL; Martinson, CA; Bessarabov, DG
The development and of key hydrogen production techniques play an essential role in the evolution of renewable energy solutions. Numerous studies have been carried out in recent years to improve and enhance these techniques, including the proton exchange membrane water electrolysis (PEMWE) process, which has produced encouraging technological outcomes that will enable the efficient and sustainable generation of hydrogen. Furthermore, because of its high current density, higher energy efficiency compared to alkaline water electrolysis, high hydrogen discharge pressures, and smaller gas crossover, the PEMWE device is an ideal high purityhydrogen generation approach. In contrast to the alkaline water electrolysis technique, the PEMWE technology is still relatively new and has a number of intricate components and interfaces that call for particular characterisation techniques in order to get a deeper basic understanding. However, alternative cell characterisation techniques are still being investigated as to efficiently monitor and control the cell with minimal resources and response times. The research presented in this dissertation focussed on the current interrupt (CI) method, entailing the development of a current switching (CS) approach to characterise a PEMWE cell. The CS method includes the design of a CS development board responsible for producing an attenuated and controlled perturbation signal generated via an STM32 microcontroller. This signal is responsible for the control of the cell supply current administered within a frequency range of 0.8 Hz – 20 kHz. Furthermore, the voltage and current signals measured through each experiment were recorded and analysed using Matlab® and Simulink®. However, to characterise the PEMWE cell, an equivalent electrical circuit (EEC) was designed to represent the different operational features within the cell from which a simulation model was designed within Simulink®. Numerous EEC designs were created and compared using RelaxIs®. The closest fit to the theoretical cell response resulted in the optimal EEC circuit utilised for cell characterisation. Thus, by incorporating the optimal EEC in the Simulink® simulation model, the voltage and current signalsmeasured during each experiment could be added to the model. The measured voltage was used to generate a simulated current from the EEC, which was compared to the measured current. Utilising Response Optimizer®, the EEC parameters were calculated based on the model results by minimising the error between the simulated and measured current. The dissertation verifies and validates that the CI method is a suitable characterisation method delivering EEC results that are closely correlated with electrochemical impedance spectroscopy (EIS) performed under similar experimental conditions.
Exploring the digital readiness of first-year rural students studying at a South African further education institution
(North-West University (South Africa)., 2026) Mofokeng, Cascious; Mofokeng, K.N
This study examines the digital readiness of rural first-year students of a South African Further Education Institution, namely Vuselela TVET College, located in North-West Province. Based on the Technology Acceptance Model (TAM) and the Appropriation Theory of the Digital Divide (ATDD), this research explored ways in which rural high school learners are making the transition to digitally based learning environments and the obstacles faced along with its impact on academic achievement. A total of 9 first-year students from rural high schools were purposively selected and virtually interviewed via Zoom using semi-structured interviews with open-ended questions, where they shared their lived experiences regarding the phenomenon. The study employed a qualitative, phenomenological methodology and the interpretivist paradigm. The results indicated that most of the participants entered further education with insufficient knowledge about digital technologies, low levels of computer literacy, and lack of confidence in operating online platforms. The major obstacles to digital readiness were low ICT infrastructure in schools in the rural areas, low Internet connectivity, load shedding, and lack of institutional support. These issues were compounded by cultural attitudes and other digital pedagogy among teachers. However, the students showed resilience by developing self-learning behaviours, peer learning, and mobile-based learning as learning coping strategies. The study concludes that digital readiness goes beyond technology availability; it includes expertise, motivation and institutionalisation to facilitate meaningful digital interaction. It suggests that ICT skills should be included in basic education syllabus, that connectivity in rural and areas should be enhanced and that specific digital literacy programmes should be introduced to the first-year TVET students. The findings can contribute to the discussion on digital equity, the National Development Plan (NDP) 2030, and Sustainable Development Goal 4 of South Africa, which advocates for inclusive and quality education in a digitally transforming world.
Exploring the incorporation of gender-inclusive indicators in the practices and policies of selected Johannesburg Stock Exchange(JSE)- listed corporations
(North-West University, 2025) Ndaba, CN; Jackson, LTB
Gender mainstreaming is a critical aspect of fostering equality and inclusion in the workplace, yet its implementation varies significantly across sectors and organisational sizes. This study examines the extent to which South African corporations incorporate gender-inclusive indicators into their practices and policies. By employing a qualitative approach, the research analysed data from interviews, and annual reports to identify common themes and disparities in gender mainstreaming efforts. Key themes explored include strategic intent, financial commitment, structural frameworks, gender governance, governance structures, and cultural transformation. Findings indicate that strategic intent, supported by leadership accountability, is a driving force behind successful gender mainstreaming efforts. Organizations with measurable gender equity goals embedded in leadership performance metrics, such as recruitment targets and pay equity audits, showcased higher levels of inclusivity. In addition, financial investment in mentorship programs, Science Technology Engineering and Mathematics (STEM) initiatives, and gender-sensitive benefits emerged as critical enablers of progress. The research “exploring the incorporation of gender-inclusive indicators in the practices and policies of selected Johannesburg Stock Exchange(JSE) -listed corporations also highlights the importance of robust governance structures and transparent reporting mechanisms to track progress and ensure accountability in implementing gender equity policies. This study contributes to the discourse on gender equality by offering actionable recommendations for organizations and policymakers such as strengthening policy frameworks, allocating resources for gender-specific programs, and fostering inclusive workplace cultures through diversity training and leadership development and statutory continuous progress monitoring and reporting Future research should focus on intersectional analysis and longitudinal studies to assess the long-term impact of gender mainstreaming efforts and address the unique challenges faced by smaller organisations and marginalised groups