ABSTRACT. This paper explores the application of two wide-band excitation signals, the pseudo-random impulse sequence (PRIS) and band-limited white noise (BLWN) to measure frequency responses of a 16 kVA, 22 kV/240 V distribution transformer. The performance of these two signals is first compared through simulations against mathematical transfer functions that describe a simplified wideband transformer model. Frequency responses of the voltage transformation ratio and the primary and secondary input impedances are analyzed. Simulation results are evaluated both graphically and quantitatively using statistical metrics. Subsequently, practical measurements using PRIS and BLWN are performed and compared with traditional sweep frequency response analysis (SFRA) measurements obtained via the FRANEO 800. The responses measured from the three excitation methods are assessed in terms of accuracy, non-linearities of the device under test and measurement speed. The results demonstrate that impulse-based and white noise excitation signals achieve comparable accuracy with significantly faster measurement times compared to conventional methods, highlighting their potential for efficient transformer frequency response assessment.

ABSTRACT. Over the last few decades, the integration of renewable energy technologies within power systems has become increasingly prevalent due to the global shift towards environmental consciousness.  Power system stability is increasingly important in these hybrid systems where introduced renewable energy technologies such as wind and solar power, are intermittent in nature and thus contribute to large mismatch between the supply and demand of energy. The control of system frequencies requires advanced and complex techniques in comparison to traditional methods. This paper provides a comprehensive view of load frequency control (LFC) from early techniques such as the proportional–integral–derivative (PID) based controller control to more advanced techniques such as optimally tuned, robust and adaptive control. An analysis of the current challenges and proposed research developments to improve LFC in hybrid, decentralized networks are presented.  The paper highlights the challenges and opportunities associated with the integration of renewable energy sources into the power grid.

ABSTRACT. In agriculture, vegetable farming is the most dominating type of farming as it contributes 2.45% to the African GDP growth. It is central to fostering growth reducing poverty and improving food security in south Africa. With more than 70% of rural population depending on agriculture for their livelihoods. However, farmers are facing ongoing challenges over monitoring of farm produce that requires extensive physical interaction within the farm activities. Another challenge, theft and killing cases are increasing in the farming sector due to poor security system involved. The current study proposes the use of GSM sim800 module to communicate with the user. Inconsistent power supply and high energy has greatly affected the agricultural sector. In this paper a cheaper prototype that uses an Arduino microcontroller to control and monitor irrigation by eliminating water waste, and provide maximum security is proposed, it is remotely controlled.

ABSTRACT. Researchers have explored methods to maximize energy output from PV systems, with tilt and azimuth optimization being a significant area of focus. While some studies have proposed standard guidelines for tilt and azimuth based on regional differences, this project utilizes intelligent optimization algorithms to achieve optimal setting of these variables based on known mathematical model, to maximize energy production. Specifically, the study explores two nature-inspired algorithms, the Genetic Algorithm (GA) and Simulated Annealing (SA). Real-world data was obtained from South African University Radiometric Network (SAURAN) and residential PV system from Pretoria, Gauteng Province for the four seasons of the years 2023 to 2024. Performance was measured using metrics such as prediction accuracy, standard deviation, percentage difference, Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE). Although both algorithms obtain tilt and azimuth settings that yield better energy production than actual production obtained from the recommended settings used in practice, the optimal settings produced by SA are more realistic (13-40º) compared to GA (30-80º). In terms of standard deviation, both algorithms exhibit high precision (low variability) across all seasons. GA exhibited a lower MAPE, indicating performance that is closer to what is already obtained in the system being studied. MAE values for both algorithms were relatively similar. Finally, sensitivity heat maps demonstrated that energy output is more stable to variations in tilt and azimuth with SA compared to GA. Overall, the study establishes the value of our proposed approach in optimizing PV system utility.

ABSTRACT. This paper presents the design and implementation of an experimental stand-alone microgrid platform, specifically addressing the operational dynamics of low-inertia power systems. The platform consists of a DC machine coupled to a synchronous generator, regulated by a Variable Speed Drive for precise control of system parameters, including voltage, frequency, and synthetic inertia. Through real-time digital adjustment of inertia, the system replicates operational conditions characteristic of renewable energy-dominated grids. Experimental results highlight the critical role of system inertia in maintaining frequency stability in low inertia microgrids. The ability to dynamically adjust inertia demonstrated a significant reduction in the Rate of Change of Frequency during disturbances, thereby improving system resilience. This lab-scale platform bridges the gap between theoretical models and practical deployment, offering a valuable tool for advancing research into frequency control and stability solutions for low-inertia systems.

ABSTRACT. In the area of light, it is essential that brightness levels be adjusted in order to optimize energy consumption and create appropriate lighting conditions for a variety of environments. Microcontrollers are one of the technologies that are currently being developed, and many of them are used in everyday instruments. It is practical and easy to use microcontrollers for a variety of purposes because they can be programmed according to needs. The need for electricity is also growing, along with technological developments. Therefore, there is a need to limit the unnecessary use of electricity. Efficient lighting control systems help to reduce energy consumption, save costs, and improve the user experience. Users can adjust the brightness level so that it is tailored to each task, time of day, or mood thus reducing energy consumption and enhancing comfort while increasing productivity. This paper focuses on designing an AC lighting brightness adjusting circuit and modeling using the simulation software Proteus to model and simulate the circuit under different conditions. This paper also focuses on implementing safety measures. This system increases user convenience and user efficiency.

ABSTRACT. The water-energy-food (WEF) nexus framework serves as a roadmap for the clear understanding of the complex relationship and interdependency among the water, energy, and food sources which are basic human needs. This paper reviews the current research on the application of microgrid control systems for sustainable WEF nexus management in rural communities, considering the integration of renewable energy sources. This becomes especially important in rural communities, as the increase in population coupled with climate change is placing an increasing strain on these supplies. Microgrids are being considered as a promising localized solution that can balance out the demand for such resources and reduce vulnerabilities while enhancing food productivity. In this work, the existing frameworks on WEF nexus, microgrid energy management models for rural communities, and optimization techniques are critically reviewed. The study further identifies the knowledge gaps that exist in terms of integrated microgrid control strategies for sustainable WEF management and practical implementation within rural communities, showing the potential to develop a holistic framework with ability to realize sustainable resource management and resilience in underserved communities.

ABSTRACT. This study provides an in-depth examination of the efficiency and applications of photovoltaic (PV)-electrolyser hybrid systems, with a particular focus on their role in producing hydrogen and oxygen through water electrolysis and the broader implications for renewable energy systems. These systems leverage the integration of PV technology with electrolysers to generate green hydrogen, which has significant potential for energy storage and sustainability. Under optimal conditions, the electrolysis of one Liter of water produces approximately 111.12 grams of hydrogen and 888.96 grams of oxygen while consuming around 4.18 kWh of energy. A 1 kW electrolyser operating at 60% efficiency can complete the electrolysis of one liter of water in approximately 7.3 hours. Under these conditions, the hydrogen production-to-fuel cell consumption ratio is about 2.3:1, demonstrating the system's capability for efficient hydrogen storage and reusability in energy applications. This study highlights the synergistic relationship between PV systems and electrolysers, showcasing their capacity to produce hydrogen in a sustainable manner. By converting solar energy into storable hydrogen fuel, these hybrid systems contribute to the advancement of energy storage technologies and support the development of more resilient and sustainable energy grids. Through their ability to provide a renewable and green hydrogen solution, PV-electrolyser systems play a critical role in addressing the challenges of integrating renewable energy sources into modern energy infrastructure.

ABSTRACT. The increasing demand for renewable energy sources, electric vehicles, AI tools and electronic devices has been hinged on the extraction and processing of lithium ores. This paper presents investigations into efficient processing techniques of lithium bearing minerals, which promote environmental sustainability whilst conforming to the stringent lithium quality control requirements. With a focus on bio-flotation, a review of the use of microorganisms in the concentration of spodumene will be displayed, systematically evaluating the advances made on enhancing bio-floatation processes. Further discussions on the approaches employed to separate lithium from spodumene concentrate are tackled, with specific emphasis on the impact of microwave technology on the purity requirement of lithium for use in energy storage devices. The subsequent conversion of lithium to lithium compounds such as hydroxides and carbonates will be assessed particularly in the manufacture of batteries and other technologies that use lithium. The paper seeks to edify the literature and knowledge around the sustainable and efficient extraction and processing of lithium, a major component in global transformation.

ABSTRACT. The global battery industry is experiencing significant growth, and this growth is predicted to continue and accelerate in the future. The cost of lithium (Li) and cobalt (Co) resources is rising as a result of growing applications and demand. Therefore, the abundance of sodium (Na) resources and their global distribution drive us to research Na-ion (Na+) batteries for immobile energy storage systems. The advancements of Na+ -batteries are reported in this paper, primarily presenting earlier and current studies in contrast to those of Li-ion (Li+) battery energy storage systems. Despite the increasing global use of Li+ -battery systems, academic research has largely overlooked sodium-ion Na+ -battery technologies. This study explores and details the most promising applications for Na+-stationary battery systems. The approach consists of two steps. First, it involves a comprehensive review of existing literature focusing on the applications, profitability, and use cases. Second, the study provides an in-depth analysis of these use cases, emphasizing the key factors driving their adoption, the sources of value they offer, and the associated risks.

ABSTRACT. The transition to renewable energy in South Africa faces significant challenges due to the inherent variability of wind power. This paper presents novel applications of advanced machine learning approaches for wind power forecasting, evaluating Physics-Informed Neural Networks (PINN), Long Short-Term Memory and Convolutional Neural Network hybrids, Transformer models, Extreme Gradient Boosting, and Temporal Fusion Transformer models. The study considers operational data from the Sere wind farm in South Africa. The PINN model demonstrates the highest forecast accuracy using turbine physics with data-driven learning. The architecture shows particular strength during power ramp events and transitional periods, where physical constraints guide forecasts. A comparative analysis in multiple operating conditions demonstrates the improved accuracy and computational efficiency of the PINN model. Statistical analysis reveals consistent improvements in various evaluation metrics, with improved performance in challenging weather conditions. The implementation framework provides practical pathways for operational deployment to meet renewable energy targets in South Africa, establishing a robust foundation for enhanced grid stability.

ABSTRACT. Inverters need to evacuate clean power by following ments, however, they are usually located far from the grid require main users. This electricity needs to be sent through transmission lines where faults can occur affecting the stability of the inverter. The paper investigates the control and stability of inverters during faults on different strength grids. A 2.3 MW inverter with a synchronous reference frame phase locked loop (SRF-PLL) and current controller was designed and simulated in Simulink. Various voltage phase and magnitude disturbances were tested on the inverter first using industry-standard testing, then trying to simulate a transmission line and creating faults with a different X/R ratio impedance. The inverter was stable for most faults, however, severe voltage drops caused the inverter to become unstable due to the PLL losing synchronization. When switching impedanceses into the network to simulate faults, those with a lower X/R ratio behaved like the phase shift test and faults with higer X/R ratios behaved more like a voltage drop test.

ABSTRACT. In this paper, a new method is proposed for the design optimization of slip permanent magnet couplers (S-PMCs) for wind turbine systems, where various S-PMC topologies are evaluated. Previous methods used to optimise S-PMCs introduced bias, as the pullout slip was preselected, thus resulting in a sub-optimal design at rated conditions. Furthermore, the torque versus slip curve and pullout slip are not necessarily similar for all S-PMC topologies. Therefore, an alternative method is proposed, where the pullout slip and torque are determined without any prejudice, where the optimization considers rated and pullout conditions. The proposed optimization strategy in this paper adequately shows the difference between the various S-PMC topologies, based on their pullout torque capabilities for the same rated torque and rated slip conditions.

ABSTRACT. In this paper, the design and performance of four permanent magnet wind generators in the sub-5-kW power level are compared. The considered wind energy system consists of a 4.2-kW wind turbine at 12 m/s and a direct drive wind generator that is directly connected to a battery storage via a three-phase diode rectifier. Ferrite and rare-earth magnets are considered in the four designs. The results show that the ferrite permanent magnet generator can be competitive in terms of energy harvesting and cost in low wind speed regions. A new method is proposed whereby the ferrite generator can also be competitive with the rare-earth permanent magnet generators for strong wind regions.

ABSTRACT. This paper presents the detection of broken rotor bars (BRB) in induction motors (IM) using three supervised machine learning algorithms (SMLA), including Decision Three Classification (DTC), Artificial Neural Network (ANN), and Support Vector Machine (SVM). The three SMLAs are trained to detect BRB features from measured steady-state line current signatures. The training data were collected in the time domain from laboratory experiments and transformed to the frequency domain using the Discrete Fourier Transform (DFT). A confusion matrix was used to validate the models’ performance using accuracy, precision, recall, and f1-scores. The results evidence that the DTC has better accuracy and precision for both half and full-load operations of the squirrel cage IM when compared with the ANN and SVM algorithms. The DTC obtained the best F1 score, accuracy, precision, and recall, followed by the SVM.

ABSTRACT. This paper analyzes the performance of a Self-excited Novel Synchronous Reluctance Generator (NSynRG). The rotor of the proposed NSynRM has a slitted rotor core to improve selected vital performance parameters. The slitted-rotor core design has been optimized for optimal performance using the Genetic Algorithm (GA) technique for minimal torque ripple and maximum average torque. The electromagnetic performance of the proposed self-excited NSynRG with an unoptimized and optimized slitted-rotor core has been analyzed using a two-dimensional (2D) Finite Element Analysis (FEA) for AC magnetic transient solutions. The FEA results evidenced that after optimization, the torque ripples of the NSynRG with slitted rotor-core are reduced by 24.51%, 29.72%, and 13.13% when feeding 8A to a load with unity, lagging, and leading power factors, respectively. Furthermore, both FEA and Experimental (EXP) results reveal that the voltage regulation has increasing characteristics when the NSynRG operates with leading and unity power factor loads and a falling characteristic when operating with lagging power factor loads.

ABSTRACT. In this paper, the performance of various rotor pole configurations in a novel -shaped stator wound-field flux switching machine (WFFSM) is investigated to address concerns with cost and environmental concerns associated with permanent magnet machines. By employing the finite element analysis (FEA) method, the study focuses on the no-load, on-load and overload performance characteristics of the proposed machines based on four different rotor poles – 13P, 14P, 16P and 17P. The 14P rotor configuration emerges as superior machine, showcasing highest torque, lowest torque ripple and best overload capability. The future study will cover full optimization of all four machines with experimental validation.

ABSTRACT. Double stator wound field flux switching motor (DWFFSM) has the potential to replace synchronous homopolar motors and induction motors in mining trucks traction and other heavy-duty off-road traction applications due its high torque density, efficiency, and relatively simple structure suitable for operation in any harsh environment. But the inherent nature of the back emf wave form of DWFFSM has not been studied before. Therefore, in this paper different winding configurations were considered, and the nature of their back emf wave forms were analyzed and validated in FEA. The FEA results show that all pole would winding configuration performs better in terms of lower total harmonic distortion.

ABSTRACT. Recent advancements in the electric power sector, including the reformation and reorganization of the electric power supply chain, the growth in distributed generation (DG), and developments in information and digital technologies, have substantially impacted the method employed in the planning, design, and operation of distribution networks. Hence, there is a need to create a realistic and practical control, automation and optimization technique for demand-side resources to ensure the smooth management, efficiency and affordable economic operation of an active distribution network. In this work, a Mountain Gazelle Optimization (MGO) approach is proposed to solve the optimization function globally in a deterministic manner to ensure a reliable and efficient energy management system. The results obtained from this work show that the intelligent employment of MGO in a deterministic energy management scheme results in an optimal solution and an improved operating cost.

ABSTRACT. This study addresses the high operational costs and inefficiencies associated with electric boilers used for sanitary hot water in South African university residences. With electricity tariffs rising due to the increasing cost of coal-based energy, traditional electric boilers remain energy-intensive and unsustainable. This research proposes retrofitting a 2000L (24kW) electric boiler with a Hybrid Solar-Assisted Air Source Heat Pump (HSAASHP) system, combining solar collectors and air source heat pump technologies to improve energy efficiency and reduce costs. The methodology involved designing and installing an HSAASHP system in the Mannheim Ladies’ Residence at the Central University of Technology, Free State, and deploying a Data Acquisition System (DAS) to monitor its performance. Comparative analysis of the baseline electric boiler and the retrofitted system revealed significant energy savings: annual electricity consumption dropped from 151,128.9 kWh to 41,036.8 kWh, with an Energy Factor (EF) ranging from 3.0 to 5.2, higher in summer. Techno-economic analysis demonstrated the system's viability, with a payback period of 1.25 years, a break-even point of nearly 2 years, and a 15-year net present value of R10,503,958. The investment yielded a Return on Investment (ROI) of 2493.5% and a profit margin of 96%. The HSAASHP system offers a cost-effective and energy-efficient solution, addressing the pressing need for sustainable water heating in resource-constrained environments. This approach not only enhances energy efficiency but also provides a scalable model for reducing energy costs and environmental impact in similar settings.

ABSTRACT. Abstract— The power system is a critical subsystem of CubeSats, directly impacting their ability to operate autonomously and achieve mission objectives. Despite advancements in power generation, storage and distribution technologies, CubeSats face persistent challenges due to their compact size, limited for solar panels and exposure to harsh space conditions. This study aims to enhance CubeSats power systems by introducing a novel approach using light Dependent Resistors ( LDR ) for solar tracking. These Resistors are lightweight, cost effective and consume minimal power, making them ideal for the constrained size and resource requirements of CubeSats. LDR- based system dynamically detects sunlight intensity and direction, enabling precise alignment of deployable solar panels to maximize energy capture. By addressing issues such as shading and misalignment, the proposed solution ensures continuous power optimization during varying orbital conditions, including eclipse periods. The system is integrated with existing power management technologies, such as Maximum Power Point Tracking ( MPPT ) and efficient DC-DC converters, to further improve energy efficiency. This approach not only mitigates power shortage but also extends the operational lifespan of CubeSats, supporting more ambitious scientific, commercial and explanatory missions. The study provides a scalable and sustainable solution to CubeSats energy constraints, laying a foundation for the next generation satellite applications.

ABSTRACT. The objective of this research is to investigate whether rooftop solar PV power can be used as a feasible solution for multistorey apartment buildings, where the ratio of rooftop space relative to floor space is much less than for typical homes. The goal of this research is to strategically control and allocate photovoltaic (PV) power generation as a feasible solution for a multi-storey apartment building. Results are obtained for a multi-storey building with 10 apartments. For this study, energy storage in the form of electrical hot-water geysers is investigated. This essential load is specifically selected due to its storage capability and high electrical energy consumption, which typically contributes 30-50 % of the total household electrical consumption. With this research an additional study is presented, which demonstrates the feasibility of Nort-facing wall-mounted PV modules, in addition to roof PV. Interestingly, results indicate this does not present a more favourable financial case. The analysis indicates that adding façade PV to roof PV does not improve financial outcomes, with both scenarios yielding similar levelized costs of energy (LCOE) and returns on investment (ROI), achieving a payback period of four years.

ABSTRACT. By classifying loads into shiftable and non-shiftable groups and rearranging a distribution system's load demand model, demand-side management (DSM) reduces operating expenses. This is made possible by shifting variable loads when utility costs per unit are lower. This work adopts a bi-level optimization technique to reduce the operational costs of a low-voltage microgrid (MG) system that incorporates plug-in hybrid electric vehicles (PHEVs), renewable energy sources, and fossil fuel generators while operating in grid-connected mode. The load model is reconfigured at the first optimization stage according to the DSM participation level. Then, the restructured load demand models are considered, and recommendations for distributed generator scheduling are developed to reduce the producing costs of the second-level microgrid system. The study's optimization technique used the popular differential evolution (DE) method, which has been applied to several power system optimization issues. When DSM, various grid participation and pricing schemes, and other factors were considered, the cost of producing active power reduced. To ascertain their effect on the MG system's generation cost, scenarios involving different PHEV state of charge (SOC) levels and a taxable rate during V2G were also investigated.

ABSTRACT. This project investigates the frequency of Electric Field Mills in predicting true thunderstorms through an analysis of alarm types, including Effective Alarms, Non-Effective Alarms, False Alarms, and their respective ratios. The study was conducted over multiple seasons, focusing on the relationship between electric field strength thresholds, lightning strikes occurrence and detection radii. Findings indicate that while False alarms peak at 100 V/m for any radii, the most effective detection threshold is 750 V/m at 20 km, achieving a combined Effective Alarm Ratio and Non-Effective Ratio of 22.69%. The Probability of Detection, and Failure to Warn Ratio were both found to be 50%, highlighting a need for enhancement in alarm accuracy.

ABSTRACT. The Dynamic Voltage Restorer (DVR) an effective advanced power custom device is proffered to enable Sub-Saharan distribution systems operate optimally. For a system to perform optimally it requires standard acceptable voltage loss, acceptable voltage supply, reliability of supply, phase current and voltage are of standards and distribution transformers (DTs) and cables are not in way overloaded. This study involves the effective mitigation of power quality disturbance in Sub-Saharan distribution networks because of poor voltage profile, voltage variation and voltage imbalance using a very effective power electronics-based custom power controller known as DVR. A DVR is connected between the Sub-Saharan distribution transformer and the customer load along a feeder with a radial arrangement. An innovative new design-model of the DVR has been proposed and developed using VSI-PWM based on dq0 controller. Model simulations were carried out using MATLAB/Simulink in Sim Power System toolbox. Results demonstrate that utilizing the proposed method reduces the Sub-Saharan distribution systems power quality disturbances to standards and acceptable limits

ABSTRACT. Abstract—The large-scale adoption of behind-the-meter rooftop solar PV generation in South Africa has added operational complexity to distribution utilities. In most cases, these utilities do not have control or visibility of such installations. Therefore, a prudent operator's first step is quantifying this local resource. A number of different estimation approaches exist to this end. This paper demonstrates two especially promising methods, using a case study that focuses on the residential suburb of Dalsig, located in the South African town of Stellenbosch. Dalsig is an affluent community with over 300 households. Method A is based on satellite imagery in the public domain. Method B is deducted from the changes in the electricity profile supplied by the utility; this data is typically confidential.  Method A uses satellite imagery and the Mask Region-based Convolutional Neural Network to detect and assess the surface area of rooftop solar panels, ultimately calculating the installed capacity of these solar PV systems. On the other hand, Method B assesses fluctuations in electricity demand on different days with significant changes in cloud cover and minimal temperature variations to infer behind-the-meter solar PV output. Using 2024 satellite imagery, Method A estimated a capacity of approximately 500 kWp, which closely aligns with the manually identified capacity of around 570 kWp. Electricity supply data was only available for 2020-2021; thus, Method B could only be deployed over this period. It underestimated the embedded capacity at 110 kWp compared to the manually identified 165 kWp. The paper finds that Methods A and B provided reliable yet conservative insights. The paper aims to raise awareness amongst distribution utility operators and decision-makers that reasonably accurate methods exist to, on a regular basis, estimate the installed capacity of behind-the-meter rooftop solar photovoltaics in their local supply areas. Future work includes extending the assessment to larger areas with a more diverse consumer profile.

ABSTRACT. This study investigates the relationship between the number of applied switching surges and the clamping voltage response of Metal Oxide Varistors (MOVs) using bivariate statistical analysis. MOV samples were subjected to varying numbers of surge events, and their reference voltage was measured before and after each test. Clamping voltage was also recorded at the beginning and end of the testing period. Scatter plot analysis revealed a slight increase in the clamping voltage ratio with an increasing number of surges. However, the data points were clustered within a narrow range of 1.1 to 1.5 p.u., indicating a weak correlation. The Pearson correlation coefficient of 0.22 further confirmed this weak relationship. Additionally, linear regression analysis yielded a non-significant positive slope of 0.0008 and a low R-squared value of 0.049, suggesting that only a small portion of the variance in clamping voltage can be attributed to the number of surges. Based on these results, while a slight increase in clamping voltage may be observed with repeated surge exposure, it is not a reliable indicator of degradation level compared to changes in reference voltage. The varying number of applied surges appears to have a statistically insignificant impact on the clamping voltage response.

ABSTRACT. This paper investigates the impact of lightning mast height on the frequency of lightning flashes to a residential structure. By applying Eriksson’s Collection Volume Method and the SANS standard for lightning risk assessment, we analyze the expected annual occurrences of hazardous events for various mast heights: 0 metres (no rod), 18 metres, and 24 metres. Counterintuitively, our findings suggest that taller lightning masts may increase the likelihood of lightning flashes. While the expected annual strike frequency without a mast is 0.02, it rises to 0.08 for an 18-metre 0.13 for a 24-metre rod. This unexpected trend highlights the complex interplay between lightning activity, environmental factors, peak lightning current, and flash density. Our results emphasise the importance of a holistic approach to lightning protection design, highlighting the need for careful consideration of mast height to optimise protection and mitigate lightning hazards for residential structures.

ABSTRACT. The paper details the design of corona-free line hardware for an Eskom ±1000 kV DC test facility. Two items are covered, namely adjustable line termination assemblies and adjustable rigid conductor spacers It is crucial that the facility is corona free such that the device under test can be evaluated for corona performance even under adverse environmental conditions. Notably, the field intensification when the thunderstorm ambient electric field increases conductor surface gradients significantly and has largely been neglected in corona studies. The hardware is benchmarked against other international UHV projects.

ABSTRACT. The paper details the design of corona-free line hardware for an Eskom ±1000 kV DC test facility. Two items are covered, namely filed control at post insulator connection flanges and bus support grading ring assembly. It is crucial that the facility is corona free such that device under test can be evaluated for corona performance even under adverse environmental conditions. Notably, the field intensification when the thunderstorm ambient electric field increases conductor surface gradients significantly and has largely been neglected in corona studies. The hardware is benchmarked against other international UHV projects.

ABSTRACT. The economic downturn in Western Australia's oil, gas, and mining sectors has led to a focus on optimizing existing equipment and maintenance activities. This shift prompted research into remote maintenance testing of Intelligent Electronic Devices (IEDs) to reduce site mobilization costs. This study identified suitable technologies for designing a Remote Verification Platform. A prototype system was developed and tested in a laboratory to execute maintenance sequences on two IEDs. The study demonstrated successful remote communication, access, and maintenance testing of IEDs, with minor issues like preventing PC hibernation from being easily resolved. Practical tests confirmed that a remote operator could set up communications, interrogate, and control both the IED and the Omicron Test Set, completing the maintenance sequence without local assistance beyond the initial connection setup. Comparative tests showed that remote methods produced results within allowed tolerances, comparable to conventional methods, for basic Overcurrent and Earth Fault (OCEF) protection IED installations. Further research and design efforts are necessary for more complex IEDs and protection schemes.

ABSTRACT. This paper presents a comprehensive economic analysis and cash flow assessment of a hybrid microgrid system integrating photovoltaic (PV) panels, advanced energy storage solutions, and combined heat and power (CHP) units. The project is designed to provide a sustainable, reliable energy solution with a focus on maximizing financial viability over a 15-year period. The analysis evaluates key financial parameters, including capital investment, operational costs, tax incentives, and utility pricing. The results highlight the economic feasibility of the microgrid, demonstrating its potential for significant returns on investment. Visual representations and data models further illustrate the robustness of the proposed energy infrastructure, underscoring its long-term benefits for stakeholders and investors.

ABSTRACT. The assessment of hybrid energy systems for generating electricity and green hydrogen for cooking and heating is crucial for advancing green technology integration. The annual increases in electricity tariffs have prompted individuals to revert to traditional methods of cooking and heating. Research conducted by climatologists indicates that these conventional practices are environmentally harmful and contribute to adverse climate change. This study is aimed at rural communities of Limpopo province to find new ways to bring affordable, reliable electricity and green hydrogen energy for cooking and heating purposes. The study focuses on implementing population size-dependent hybrid energy systems to produce green hydrogen. Four communities in Limpopo province, Ga-Masekwa, Ka-Dzingidzingi, Duthuni, and Mookgopong non-urban, have been randomly selected for this assessment. The Herman-Beta method, in conjunction with Homer Pro software, is employed to estimate maximum loads in selected communities. This approach facilitates the simulation of various hybrid energy system configurations, allowing for a comprehensive analysis of energy generation and consumption dynamics. Configurations comprised of PV/H2/Grid and PV/BES/H2/Grid are evaluated to determine the best hybrid energy system for each community based on economic performance. Results obtained indicate that the PV/H2/Grid configuration is the most cost-effective, with the lowest NPC and LCOE, offering a high return on investment for potential investors. The NPC of this configuration in all communities is R3.01M, R88.3M, R162M, and R299M, respectively. The corresponding LCOEs are 1.02 R/kWh, 1.2 R/kWh, 1.18 R/kWh, and 1.12 R/kWh. However, it is noted that green hydrogen production requires a significant amount of energy, making the hybrid system run more expensive than it would be otherwise. Successful green hydrogen use in rural communities can lead to economic growth, sustainable cities, reduced emissions, and a transition to renewable energy.

ABSTRACT. Communication plays a significant role in an electrical grid that includes Distributed Energy Resources (DERs). Modern control and automation systems used by these systems are predominantly distributed in nature using domain standards such as IEC 61131-3. The IEC 61131-3 standard, however, does not directly address the communication aspects. The IEC 61499 standard, in contrast, introduces the use of the Service Interface Function Block (SIFB) concept and high-level communication patterns to achieve hardware-independent access to communication services. This paper proposes the use of the IEC 61499 standard for DER control and automation and the use of SIFB to facilitate the ease of system integration in a DER plant.

ABSTRACT. There is an increasing demand for affordable, decentralised and distributed electricity, which is one of the key motivating factors that has incentivised this review article's writing. Industrial 4.0 has been a conduit through which the conventional grids shift towards a self-autonomous system, and blockchain platforms are emerging as viable solutions to facilitate that paradigm. With the advent of peer-to-peer (P2P) energy trading, community-based energy exchanges have significantly contributed to reduced grid dependency. Machine learning (ML) is being implemented in predictive analytics for energy forecasting (EF), thereby aiding both distribution network operators (DNOs) and aggregators to precisely and predictably monitor, optimise and stabilise grid operations through smart agents. The main aim of this review paper is to provide a detailed analysis of the existing research and implementation approaches on blockchain-based P2P energy trading and a comparative overview of how ML applications aid predictive analytics in community-based energy exchanges.

ABSTRACT. This paper evaluates energy graph-based visualisation (EGBV) data structures for fault classification using machine learning (ML). EGBV transforms process variable (PV) data into node signature matrices (NSMs) and cost matrices. Tests with k-nearest neighbour (KNN), logistic regression, and feedforward neural networks (FNNs) show that transforming data into the NSM format achieves comparable performance to the PV format, indicating that EGBV offers similar classification characteristics. Future work could apply deep learning to the NSMs and cost matrices to leverage their structural features.

ABSTRACT. The research focuses on the integration of solar photovoltaic (PV) systems, electric vehicles (EVs), and battery energy storage systems (BESS) into the Cape Peninsula University of Technology (CPUT) distribution network to enhance power quality and stability. The study evaluates four configurations to assess these technologies' technical and economic impacts on the distribution network. Results show that while EV integration leads to increased power losses and voltage deviations, including BESS mitigates these adverse effects, improving overall system performance. By strategically placing BESS within the network, power losses are reduced, and voltage profiles are stabilized, especially during peak load periods. The findings underscore the importance of renewable energy integration and storage solutions in achieving grid stability and environmental sustainability

ABSTRACT. This article proposes a machine-learning algorithm for the automatic classification of single-source partial discharge (PD) in power transformers. PD testing provides valuable information of the state and deterioration of the insulation systems of transformer windings and core. A PD testing setup based on the IEC 60270 International standard is used to test PD from transformers of varying size and operational age. PD is recorded at different voltage levels applied to the transformer under test. Where the majority of PD classification literature has focused on laboratory developed artificial PD models, this work uses practical power transformers as a basis for testing and collecting the PD database. The PD database collected from this testing is utilized for the training, validation and testing of the machine learning algorithm. In this article, a comparative analysis of various trained machine learning algorithms for classifying PD is performed. The results of the classification show very pleasing performance from the tested classifier algorithms, with Bilayered Neural Network achieving a 96.97% validation accuracy of and a test accuracy of 97%.

ABSTRACT. Medium Voltage Direct Current (MVDC) systems have traditionally been used in specialized applications such as shipboard power systems, railway networks, and more recently, DC links between AC networks. Recently, the growing global emphasis on decarbonization, driven by the adoption of sustainable energy sources, and advancements in power electronic systems have renewed interests in expanding MVDC applications to broader distribution systems such as in the electrification of high-power DC loads through renewable energy sources. Despite these growing interests, research that focuses on utilizing MVDC systems for long distance electrical power distribution has been limited. This paper presents a study of a proposed MVDC distribution system that interconnects electric vehicle (EV) ultrafast charging stations (UFCS) and a photovoltaic (PV) solar system supported with a battery energy storage system (BESS). The study aims to investigate the voltage behavior on the MV line at UFCS connection points in the MV line. The study is undertaken through modelling and simulation in MATLAB software. The simulation results indicate a voltage drop of up to 5.75% and a power loss of 3.28% in the MVDC line. These performance figures provide validation for the suitability of MVDC systems for the interconnection of high-power DC-based power sources and loads over long distances.

ABSTRACT. Power transformers are critical infrastructure in the energy distribution sector. Thus, constant monitoring of the condition and health of the system is crucial. Digital technologies such as machine learning have shown promising applications in fault diagnosis and condition monitoring of smart power systems. In this study, seven machine learning models were developed and trained on dissolved gas analysis (DGA) data to determine the condition of the power transformers. Techniques explored include logistic regression (LR), decision tree (DT), random forest (RF), gradient boosting (GB), k-nearest neighbors (KNN), support vector machines (SVM), and multilayer perceptron (MLP). The performance of the models was studied with and without correction for data imbalance via oversampling, as well as feature scaling techniques. For hyperparameter tuning, the GridSearch technique with 5-fold cross-validation was applied.

ABSTRACT. "Electric vehicles for grid flexibility" examines the impact of electric vehicle (EV) charging on the South African electricity grid and explores the potential for managed charging strategies to enhance grid flexibility. As EV adoption grows globally, increased demand for electricity poses challenges and provides opportunities for grid operators like Eskom. Managed charging, which involves controlling the rate and timing of EV charging based on grid conditions, could help balance loads, regulate frequency, and integrate renewable energy. The research assesses various charging strategies, including timed charging, dynamic smart charging, and Vehicle-to-Grid (V2G) technologies. While managed charging offers significant benefits, including potential savings and improved grid stability, it also presents challenges, such as high implementation costs and the need for extensive cooperation among all stakeholders. The research recommends that Eskom and its partners develop a managed charging strategy to prepare for future EV uptake, although immediate implementation is not required at present due to the current low penetration of battery EVs in South Africa. A strategy should be formulated within the next 2-3 years and should be aligned with market growth and technological advancements to optimise the role of EVs in supporting grid operations.

ABSTRACT. Traffic congestion poses significant challenges for cities around the world, affecting travel efficiency, environmental sustainability, and economic productivity. This paper presents an adaptive IoT-based traffic control system designed to dynamically adjust signal timings at high-density urban intersections. Using real-time simulated traffic data, fuzzy logic and reinforcement learning algorithms implemented within the SUMO (Simulation of Urban Mobility) environment, the system optimises traffic flow by responding to changing conditions. Key findings demonstrate that the adaptive control system significantly reduces average waiting times and queue lengths, while improving vehicle throughput and average speed. This study contributes to the field of intelligent traffic management by offering a scalable and IoT-enabled solution to reduce congestion in complex urban environments, thus improving traffic efficiency and reducing pollution.

ABSTRACT. The increasing penetration of electric vehicles (EVs) and renewable energy sources present a significant opportunity for enhancing the power grid in South Africa, provided these resources are effectively coordinated. For instance, charging EVs during periods of high renewable energy generation and utilizing vehicle-to-grid (V2G) technology to discharge electricity back to the grid when supply is insufficient can optimize energy utilization. However, incentivizing EV owners to engage in V2G programs poses a challenge, particularly in the absence of financial rewards. Additionally, there is a notable lack of economic feasibility analyses regarding V2G in South Africa. Therefore, assessing the economic viability of V2G within this context is essential. This paper evaluates the economic benefits for four distinct types of EV owners participating in the V2G program. The findings indicate that the V2G initiative can provide economic benefits for both power system operators and EV owners, allowing the latter to earn between US$ 62 and US$ 250 annually

ABSTRACT. Electric vehicles (EVs) are increasingly becoming pivotal in modern day transportation hence worth considering in research. This paper proposes an energy-efficient simplistic model of a slip-based EV braking controller prototyped using a quarter car powered by a simple DC motor. Input Output Feedback Linearization is applied to the prototype's state equations to linearize the non-linear system so that a linear slip-based controller can be designed. Two new linear states are formed, and the third hidden state is analyzed to evaluate the stability of the system internal dynamics. Henceforth, a negative linear state gain feedback controller is designed using pole-placement to regulate slip and achieve quick braking from 25 m/s to rest only in 2.843 seconds. These prototype results can be extrapolated to real EVs, offering a demonstrable energy-efficient braking approach and a mathematically simpler linear controller design to the growing industry of EVs.

ABSTRACT. Solid Waste Management in Non-Mediterranean Africa has proved to be a growing challenge, predicted to soar to the end of the twenty-first century. Coupled with the recent ambitions to retrofit old and already manufactured internal combustion engine vehicles to new energy vehicles, if mismanaged, metallic wastes generated as by-products can exacerbate the challenge. This paper explored a sustainable beneficiation, valorization, and re-purposing circular economy strategy centered on homologation, as a combination of hydrometallurgical and pyrometallurgical metal extraction, metal processing and product fabrication approaches are proposed. The strategy recovers the metals into bars, ready for further valorization using advanced manufacturing techniques to make products that can be viable in the automotive industry. The value of this approach is translatable to a viable business model, practical for Non-Mediterranean Africa.

ABSTRACT. Abstract— The growing adoption of renewable energy primary plants has been seen globally. South Africa has seen a major development in utility-scale photovoltaic (PV) farms in recent years due to the alleviated regulations stipulated in the latest Electricity Regulation Amendment Act. Connecting the PV plants to the power grid has also undergone a technical evolution by incorporating modern embedded computers known as Intelligent Electronic Devices (IEDs). The International Electrotechnical Committee (IEC) 61850 standard defines a suite of protocols used in modern power systems. This study aims to compare the IEC 61850 Manufacturing Messaging Specification (MMS) with the well-known MODBUS protocol. The comparison is achieved by simulating a solar farm's Power Plant Controller (PPC).

ABSTRACT. This paper presents an optimal power flow dispatching for a grid-connected photovoltaic-battery energy storage system under grid-scheduled load-shedding to explore solar energy sufficiently and to benefit the electric vehicle battery swapping station at the charging demand side. The proposed system comprises a solar photovoltaic system, a battery energy storage system, and an electric vehicle battery swapping station. The optimization problem is formulated as a multi-objective optimization problem in a discrete-time domain to minimize the operational costs associated with the power flow drawn from the utility grid and the wearing cost of the hybrid system due to the frequent charging and discharging of the battery energy storage system when charging the depleted EV battery. A linear programming method determines the optimal power flow in the proposed system to charge the depleted battery for electric vehicles. Simulation results show the effectiveness of the developed model by providing the optimal dispatch power flow at the electric vehicle battery swapping station at the charging demand side.

ABSTRACT. This investigation explores the performance metrics of a Peltier-based water distillation system, intended for the purification of greywater. The system aims to leverage the Peltier effect to simultaneously evaporate the liquid greywater on the hot-side of the module and condense the water vapour on the cold-side; the distillate being pure water. Several experimental setups were explored to try and quantify the production rate, with the aim of bettering that from a conventional solar still; these included increasing the supply voltage (and hence input power), increasing the number of Peltier cells (and hence heating / cooling capacity) and adding a circulation fan into the condensation chamber – from the results, the latter two offered the best improvements and offered almost double the production rate of a comparable conventional solar still.

ABSTRACT. This paper investigates the strategic placement of Battery Energy Storage Systems (BESS) in a modified 16-bus Witzenberg distribution network with a renewable energy source, such as a photovoltaic (PV) system to improve voltage profile regulation and reduce power losses. The integration of renewable energy sources (RES) into distribution networks has introduced challenges in maintaining voltage stability and minimizing power losses. To address these issues, the optimal placement of BESS is achieved by minimizing the costs associated with voltage profile deviations and power losses in the distribution system, thereby improving the performance of the 16-bus Witzenberg distribution network. The methodology adopted involves the use of Particle Swarm Optimization (PSO) and Genetic Algorithm (GA) to solve the optimization problem, and the results obtained from both algorithms are compared. The optimization objective function considers both voltage deviation costs and power loss reduction costs. The results demonstrate that after optimal placement of BESS, both achieved a 58.75% reduction in active power losses (from 8 MW to 3.3 MW) and improved the voltage deviation index by 56.26% (from 11.98% to 5.24%). These results suggest that strategic BESS placement can significantly enhance distribution network performance and support increased RES integration.

ABSTRACT. The transition to a low-carbon economy necessitates a paradigm shift in energy policy, prioritizing community-led renewable energy initiatives. This paper examines the role of energy policy in facilitating community renewable energy power sharing, enabling decentralized energy generation, and promoting energy democracy. We analyze existing energy policies and identify key drivers for community renewable energy adoption, including net metering laws, tax incentives, and grid access regulations. Our research highlights the potential for energy policy to unlock community-led renewable energy projects, fostering energy sharing and cooperation among households, businesses, and institutions. By empowering communities to generate and share renewable energy, energy policy can play a crucial role in reducing greenhouse gas emissions, enhancing energy security, and promoting sustainable development. Our findings could inform policymakers, regulators, and community stakeholders seeking to harness the benefits of community renewable energy power sharing, ultimately contributing to a more equitable and sustainable energy future and supporting the Sustainable Development Goals.

ABSTRACT. Data centers consume a lot of electricity for continued operation and pose a significant challenge to realizing energy security. This has led to the need for data center operators to develop their own power sources and systems. The development of their own power systems has the challenges of high data center acquisition costs, and increased overhead. The research being presented proposes the virtual ownership of renewable energy farms accessible via a supergrid. The supergrid context is realized via the integration of multiple renewable energy farms. The data center operator accesses energy from these energy farms in a virtual ownership model. This reduces the total ownership costs and the operational overhead. Performance evaluation shows that the use of the proposed approach reduces the total ownership costs by a minimum and maximum of 10% and 30% on average, respectively.

ABSTRACT. Wind and solar energy are among the most utilized renewable energy sources in the world. These sources, however, are subject to fluctuations due to local weather conditions, bathymetry and shoreline features. Wave energy on the other hand, offers a more predictable and consistent alternative with a high degree of consistency compared to renewable market leaders, solar and wind. Existing literature, however, does not fully explore this upper hand in terms of its reliability and predictability in the context of South African ocean environments. In this literature review, the current status quo of wave energy converters (WECs) is examined including market shifts toward smaller WEC technologies tailored for lower sea environments. The review further emphasizes the need for improved wave energy resource assessments and then concludes by assessing the perceived risks & deployment strategies. Recommendations for future research are suggested for the above-mentioned topics to advance the business case for WEC development in South Africa.

ABSTRACT. Integrating renewable energy sources (RES) into power systems has created challenges in frequency stability and Automatic Generation Control (AGC). Traditional AGC methods often fail to cope with the variability of RES, highlighting the need for adaptable approaches. This paper reviews model-free AGC control strategies, such as fuzzy logic, artificial neural networks (ANN), deep reinforcement learning (DRL), and hybrid methods, which provide viable solutions for grid stability in RES-dominated power systems. Model-free techniques do not depend on precise system models, allowing for real-time adaptation to changing grid conditions. A comparative analysis based on robustness, computational requirements, and adaptability reveals the strengths and limitations of each method. Fuzzy logic is robust and cost-effective but may struggle with complexity, whereas ANNs and DRL provide high adaptability but require significant computational resources. We discovered there is a chance for improvement for model-base techniques if we hybridize them with model-free techniques and that they should provide better results than model-based strategies. The review discusses key challenges and future directions, stressing the necessity for hybrid models and enhanced computational efficiency to optimize AGC in renewable-dominated grids. This review emphasizes the vital importance of hybrid AGC strategies in fostering resilient and sustainable power systems.

ABSTRACT. The United Nations approved the Sustainable Development Goals in 2015 to help countries around the world work toward a more sustainable future. The SDG7 goal is to ensure that everyone has access to cheap, dependable, clean, renewable, and sustainable energy. In view of the shifting energy sources, modernizing the power systems is imperative to satisfy daily demands. Complex energy networks are made simpler by the power system's integration of smart grids. Furthermore, Distributed Generation (DG) is integrated with Renewable Energy Sources (RESs) such as wind and photovoltaic (PV) systems to provide converter autonomy and smart grids. However, frequency changes caused by imbalances in load and generation, together with fault outbreaks, continue to pose a challenge to smart grids. While many academics concentrate on using machine learning (ML) in steady state models, this research aims to analyze robust performance optimization and discuss potential uses of ML in the future to help with transient behavior during fault outbreaks. This research builds a power system model for a smart grid to contribute attaining SDG7. The key contribution is the tuning of the gains by the meta-heuristic optimization algorithms. The machine learning based on improved spider wasp optimizer (SWO) is used in this study, while considering the limitations of the operating system. The simulations were carried on MATLAB. Using SWO the system produced satisfactory results.

ABSTRACT. Global aggressive policies in control of Carbon-dioxide ($CO_2$) emissions from several combustion mechanisms have paved way for alternative/ clean energy sources. Green hydrogen is a promising means of clean energy sources for commercial power generation stations and other crucial systems such as hydrogen fuel cell vehicles (HFCVs), motors, drives, backup power for buildings, and also serving as a source of electrical energy for remote locations. This paper examines the economic and environmental advantages of incorporating green hydrogen into the energy mix of SSA.

ABSTRACT. The transformation of traditional power grid systems into Smart Grids (SGs) allows for the optimal management of energy generation, transmission, distribution, and consumption. While this advancement is highly beneficial, it also introduces security risks, highlighting the need for a secure and scalable network infrastructure to ensure the efficient and reliable automation of energy This paper focuses on privacy, confidentiality, and security in SG applications and services, with a particular emphasis on Auto-mated Meter Reading (AMR). At the semantic level, key privacy and authentication mechanisms are crucial to safeguard data from unauthorized access. We explore multipath routing for specific hid-den services and applications as a method to enhance resistance against unauthorized traffic analysis attacks, making group authentication an area of investigation. AMR is one of the essential SG applications, significantly improving the efficiency of meter reading within the SG network. In our study, we establish security rules to protect the security and privacy of the AMR application. We propose a comprehensive protocol analysis based on security and privacy requirements. Through simulations and analysis, the proposed Grp-AKA protocol demonstrates improved computational speed, reduced signaling overhead, and better bandwidth usage compared to existing protocols.

ABSTRACT. Agriculture makes up one of the three main economic sectors that contribute to the GDP of Africa. Thus, in order to sustain the competitiveness and growth of the GDP, farms must have a steady and adequate supply of electricity. Hence, any of the available renewable energy sources can be used to create a supplement standalone micro-grid. Biogas energy technology is unique because it can coexist with the surrounding farms' operations and maintenance as well as the rural lifestyle of residents. Moreover, this technology may be safely explored to provide dependable and reasonable electricity cost by employing the organic waste material from the farms. With this paper, the authors seek to investigate the policy landscape of biogas technology and its economic benefits and environmental implications in the South African context. The study highlights how rural livelihoods can be enhanced, how sustainable developmental goals can be supported and the reduction of greenhouse emissions. Through the findings, the authors intend to inform stakeholders and policymakers on contributions on the broader energy security and improved energy access in rural areas by optimising biogas usage.

ABSTRACT. the objective of this study is to increase the use of renewable energy sources and reduce reliance on diesel generators by analysing and designing the optimal hybrid renewable energy power system for a remote island. The island's energy consumption was assessed using load profile analysis, and the feasibility of renewable energy sources including solar and wind was investigated, By utilizing the HOMER modelling tool to analyse several configurations of solar PV, wind turbines, battery storage, and diesel backup, the most cost-effective and reliable energy mix identified was using the configuration solar PV, wind turbines, battery storage and a generator as a backup. The results shows that the LCOE decreased due to the configuration to R1.90/kwh and the carbon emission by 50% when compared to other configurations.

ABSTRACT. This study evaluates the performance of machine learning and deep learning models for predicting voltage swell peak amplitude in grid-tied photovoltaic (PV) systems, with the goal of improving power quality management. The dataset, collected over 24 months (January 2022 to December 2023) from a 3.3 kWp grid-tied PV system at the Council for Scientific and Industrial Research (CSIR) in Pretoria, South Africa, includes power and weather data recorded at intervals of 10 to 30 seconds. The time series data is hourly averaged, capturing PV system and weather measurements between 5 am and 6 pm. The system is connected to the building's low-voltage network and consists of 10 monocrystalline modules, each with a capacity of 327 Wp. Five models—Artificial Neural Network (ANN), Support Vector Regression (SVR), Random Forest (RF), K-Nearest Neighbors (KNN), and Long Short-Term Memory (LSTM)—were trained to predict voltage swell peak amplitude. Among these models, the Random Forest model closely matched the actual peak voltage, followed by ANN and SVR, both performing well. LSTM exhibited moderate accuracy, while KNN showed the largest deviations from the actual values. Model performance was assessed using Mean Squared Error (MSE) and Root Mean Squared Error (RMSE), with Random Forest achieving the lowest MSE (0.01V²) and RMSE (0.02V), followed by ANN. KNN performed the worst overall. The study concludes that the Random Forest model provides the most accurate predictions of voltage swell peak amplitude, offering valuable insights for improving power quality in grid-tied PV systems.

ABSTRACT. Manufacturing and utilization of electric vehicles are expanding quickly due to their imperative advantages over traditional vehicles, which include superior efficiency and environmental friendliness. However, a safe and reliable electric vehicle charging infrastructure operation is crucial to ensure seamless operation, user satisfaction and the development of electric cars. As a result, this study provides a comprehensive review of reliability evaluation method for electric vehicle charging infrastructure. Reliability evaluation is deemed as one of the acceptable and understandable tools used to assess the operation of a power system. There are two existing crucial techniques to assess the electric vehicles charging infrastructure reliability, which includes mean time to first failure (MTTFF) and mean time to failure (MTTF). This study show that reliability indices are significant for improving user experience, operational efficiency, grid stability, and cost-effectiveness of EV charging infrastructure, ultimately contributing to the widespread adoption of EVs.

ABSTRACT. This paper discusses a state-of-the-art hybrid system capable of performing two crucial tasks in farming operations, viz., fruit disease detection and fruit grading or sorting operations. Agriculture is a major contributor to the global economy, especially in developing regions such as the continent of Africa. On top of this, the global food demand is heightening on an annual basis and the agricultural sector is under great pressure to keep up. There are many opportunities to help the agricultural sector keep up with the demand of high quantity and quality yields and this paper has taken advantage of one. In this paper, a Hybrid Fruit Disease Detection-Sorting model was conceptualized and modeled in the environment of MATLAB to sort oranges and apples into 6 classes, viz., healthy oranges, healthy fruits, damaged oranges, damaged apples, black rot-affected oranges, and botch-affected apples. This model had to further sort these fruits as per the mentioned classes. The model achieved 100% classification and sorting accuracy as it managed to classify and sort all the test fruits correctly.

ABSTRACT. This paper presents the design and implementation of an automated food dryer controlled by fuzzy logic. The food dryer aims to improve traditional drying methods by automating the drying process and allowing specific conditions for different types of foods. Fuzzy logic is utilized to control the internal environment, enabling the dryer to respond to real-time changes in temperature and humidity. The system features a microcontroller (ESP32) that monitors sensor inputs, calculates drying progress and provides feedback through an MQTT-based interface. Experiments conducted on biltong and apple slices demonstrated the operation of the system, with the dryer successfully reaching target moisture contents.

ABSTRACT. Nonlinearity and other existing uncertainties, such as time delay, are commonly found in systems involving realistic operations. Lately, the time-delay estimation (TDE) mechanism has been gaining popularity in various control theory-related fields concerning the compensation of uncertainties present in control systems. This paper examines the effects of nonlinear backlash-like hysteresis and system time delay using a proportional-integral-derivative (PID) controller, followed by introducing a TDE technique to a PID-controlled time-delay system. Experimental studies using simulation results are performed to validate the effectiveness of the proposed TDE-PID control method by emphasizing the estimation and attenuation of the nonlinearity effect present in a time-delay system. The results indicate an improved and robust transient system response under the impact of uncertainties like hysteresis and system disturbance present in a time-delay system. Further, the proposed TDE-PID control method is analysed for systems with varying time delays, and the simulation results further substantiate that the proposed method is effective. By considering various uncertainties, the proposed method can be used in different industrial applications, such as the frequency and voltage control in power systems.

ABSTRACT. Electricity theft continues to be a major problem for utility companies, especially in developing areas, causing financial losses, technical inefficiencies, and safety risks. This research investigates public knowledge, understanding and opinions on electricity theft using a survey method, along with additional information from utility company audits. The purpose of the study is to evaluate the extent of public awareness regarding electricity theft, its social and economic impacts, and the efficiency of audits in preventing illegal consumption of electricity. The study was conducted with different demographic groups to investigate their knowledge of electricity theft, opinions on its legality, and readiness to be involved in prevention measures. At the same time, data collected from utility companies was examined to assess how common electricity theft is, how it is detected, and the obstacles auditors encounter. Results reveal a high level of awareness about the issue but mixed perceptions on its causes and consequences. The audit insights further validate the extent of electricity theft in the community, underscoring the need for enhanced public education and stricter enforcement mechanisms. The research highlights the critical role of community engagement and robust monitoring systems in combating electricity theft. This research highlights the need for stricter enforcement measures and advanced technological interventions to curb electricity theft.

ABSTRACT. High industries motor-driven systems energy consumption has been growing significantly at an alarming rate, owing to higher cooling system demand. This contribute to a large portion of building total energy usage and an appropriate system operation control strategy promises large energy saving with embedded energy efficiency enhancement. The paper proposes cooling system chilled water temperature and water flow rate using energy efficiency input–output model-based PSO, while minimizing energy consumption. This work investigates energy consumption using simulated data from data in brief and a statistical model with a variable frequency drive to control a HVAC fixed-speed cooling water pump. This work investigates energy consumption using simulated data from data in brief and a statistical model with a variable frequency drive (VFD) to control a HVAC fixed-speed cooling water pump. The model energy costs vs material utilization in mining through load testing using measured data every 15 seconds for 20 days with 32164 datasets. The methodology is based on the system delivering cold air at a HVAC system simulation result when the cooling water flow rate been varied between -3,573446907 m^3/h and 50 m^3/h, while the evaporation outlet temperature has been kept constant at 2.1°C. In this scenario, chiller power consumption is constant while pump power varies with the cooling water flow rate. Optimizing cooling systems chilled water flow rate on reduced chilled water pump energy savings by 41.9%. Sustainable industry sector energy usage with a particular focus on South African goals and policies on energy efficiency opportunities in the industry sector value chain is envisaged.

ABSTRACT. The strong demand for emissions control and improved energy efficiency in diesel engines has opened new interest in improved waste heat recovery systems. This study reviews integrating Response surface methodology (RSM) with Organic Rankine Cycle (ORC) systems as a novel approach to emission control. RSM, a strong optimum statistical tool, helps determine the multiple engine parameters such as the injection timing, pressure ratios, and fuel-air mixture. When integrated with ORC it recovers and transforms waste heat into usable energy the use of the two-combine methodology provided a chance to improve thermal effectiveness and decrease emissions in diesel engines. Previous researchers have addressed these approaches in isolation; however, this study focuses on the possibility of combining RSM and ORC. In the ORC system, organic fluids are used to recover low-grade heat across different engine loads, decreasing NOx, PM, and GHG emissions. This work also shows that integrating model-based techniques can solve various difficulties occurring with engines and environment consequently improving engine characteristics and decreasing emissions. Thus, this strategy provides an opportunity to solve the issues of environmentally friendly and sustainable operations of diesel engines, which are of interest worldwide.

ABSTRACT. the increasing demand for energy and the need for sustainability have driven the adoption of smart buildings equipped with advanced technologies to optimize energy consumption. Machine learning (ML) algorithms play a pivotal role in enhancing the efficiency of energy use in smart buildings by predicting energy consumption patterns, detecting anomalies, and automating control systems. This review paper provides a comprehensive overview of the state-of-the-art machine learning algorithms applied in the context of energy optimization in smart buildings. This paper reviewed various ML techniques, their applications, benefits, and limitations, and highlight recent advancements and future research directions.

ABSTRACT. This work focussed on a comparative energy and economic analysis of a Hyundai Kona EV and Toyota Corolla 2008 model. The study involved a controlled driving experiment from the NADDC solar charging station to Nru junction, revealing specific energy consumption values. Results indicate that the Hyundai Kona consumed approximately 4.5 MJ without AC and 10.6 MJ with AC on low mode, while the Corolla expended about 28 MJ for the same journey. Further analysis revealed that the Hyundai Kona's range, showed its capability to travel approximately 420 km without AC and 410 km with AC at Low Mode on a fully charged battery. Additionally, assessing the sustainability of the Hyundai Kona in relation to the capacity and environment of the NADDC Solar charging station, it was observed that the station can support only one EV sustainably. However, accommodating two EVs necessitates implementing a staggered charging schedule to optimize resource utilization and maintain sustainability.

ABSTRACT. Formula SAE is a challenging engineering competition where students from universities will design, construct and compete in an electrically powered race car. This project aims to design and develop the Drag Reduction System (DRS) to fit onto the rear wing of the formula race car that the university’s engineering team are currently working on. This project also aims to design and develop the mounting system for the rear wing to attach to the race car, while staying within the geometrical requirements as set by the rules and regulations of the competition. The mountings would need to satisfy the structural integrity limits as well to allow a maximum force of 200N to be applied to it without deflecting more than 10mm. The workings of the system would be to incorporate a push button like switch to activate the system, while having a sensor to deactivate the system when the brake pedal is depressed. Once the system is in operation the servo and its connecting arm, which are parts of the DRS, would rotate the wing on its axis by tilting it away from having a maximum angle of attack (AoA) to being set to a 0° angle, creating a bigger space between the wing profile in order to allow air to flow more freely over it. This would create less aerodynamic downforce on the race car and provide assistance to the driver while overtaking another car ahead of them. All of these aspects being designed and developed should be done while keeping stiffness and being lightweight in mind.

ABSTRACT. This paper presents the design and investigation of an IoT-enabled smart collar system that aims to mitigate cattle theft and improve livestock monitoring in rural areas. The proposed system integrates GPS for real-time location tracking, an accelerometer for motion detection, and temperature sensors for monitoring animal health, all connected via cellular communication networks. The collar continuously monitors cattle movement and health indicators, generating alerts when unusual behavior is detected, potentially indicating theft or distress. Prototype testing demonstrated satisfactory performance in GPS accuracy (± 3 meters), accelerometer sensitivity, and alert response times (under 60 seconds). However, power consumption tests revealed a 5-hour operation limit without solar support on cloudy days, underscoring the need for improved energy management strategies. This paper concludes with recommendations for further development, including system scalability and improved power strategies to enhance performance in large-scale deployments. The results suggest that the smart collar offers a feasible and cost-effective solution for cattle theft detection and remote livestock management, with significant potential for adoption in regions prone to cattle theft.

ABSTRACT. City centres, shopping malls, and airports provide adequate vehicle charging infrastructure for daily requirements. However, long-distance travelling such as intercities lacks adequate charging stations. Therefore, this paper gives the optimal placement of charging stations for electric vehicles to help intercity travel. Three optimisation algorithms including particle swarm optimisation (PSO), genetic algorithm (GA) and grey wolf optimisation were evaluated for optimal charging station placement in South Africa (SA). The results of this study are tested and validated using MATLAB software. The outcomes show that PSO achieved the lowest fitness score of 2311.64, GA produced a higher fitness score of 116416.52, and GWO achieved a fitness score of 2334.06. In addition, PSO exhibited the shortest execution time of 0.055 s compared to GA and GWO at 0.079 s and 0.09 s. Therefore, PSO identified charging station placements with the most efficient coverage of travel routes and minimal penalties.

ABSTRACT. The use of data science and artificial intelligence (AI) in power and energy systems to do predictive modelling for power system planning and operation have become more critical during this time in South Africa and other African countries where loadshedding and power interruptions have become household phenomena. The applications are wide-ranging and diverse. The paper will at first focus on factors affecting integrity of data, a requirement to do advanced data analytics or projections using Artificial Intelligence (AI), secondly it will give a glimpse of models used in the field as well as a dashboard that was developed to track transformer health. The dashboard reports data to support the planned condition-based maintenance model with the aim of eventually managing unplanned (emergency or reactive) maintenance through the prediction of potential failures in transformers that could lead to interruption of power or premature ageing of the fleet. Proactive maintenance is derived through the modelling of combinations of preventative maintenance data (essential to ensure efficiency and reliability through time or condition intervals) and predictive maintenance data (current status through data collection, analysis and performance). Ultimately the solution to reduce unplanned maintenance is to do planned maintenance, based on predictive modelling that relies on accurate current and historical data.