Repository logo
 

Masters Degrees (Electrical Engineering)

Permanent URI for this collection

Browse

Recent Submissions

Now showing 1 - 20 of 192
  • Item
    Damping subsynchronous resonance using HVDC supplementary controls.
    (2021) Tambwe, Christophe Basila.; Pillay, Rudiren Carpanen.
    ABSTRACT: In power transmission lines, series capacitors are used to increase power transfer capability. However, series capacitors are the primary source of subsynchronous torsional oscillations of nearby turbogenerator shafts. These electronic components in transmission lines expose the turbogenerator system's torsional dynamics to a resonance phenomenon. Typically, this phenomenon appears in the form of supersynchronous and subsynchronous resonance. Subsynchronous resonance is often the most dangerous for its harmful effects, leading to shaft failure and power system instability. It differs from supersynchronous resonance, which does not adversely affect system stability. A turbogenerator set commonly consists of natural modes with frequencies smaller than the power system rated frequency. When the torsional frequencies resulting from the transmission line's series capacitors affect the turbine-generator shaft's operation by coinciding with one of its natural frequencies, they induce the phenomenon known as subsynchronous resonance. Supersynchronous resonance refers to torsional oscillations whose frequencies exceed the nominal frequency of the electrical network. Researchers have used many approaches to solve the subsynchronous resonance phenomenon in power systems. This research focuses on exploiting supplementary controls of a parallel HVDC system to address the subsynchronous resonance problem. However, owing mainly to the converter's current regulators and many other parameters, the HVDC system can also potentially induce subsynchronous oscillation on nearby turbine-generator shafts. This work focuses on a power system comprising a modified IEEE First Benchmark Model to incorporate a parallel HVDC link. This power system is subjected to torsional instability from either the resonant line AC line, HVDC controls, or both. Thus, this study contrasts the effectiveness of single-mode and multimodal damping controllers when the turbogenerator shaft's torsional dynamics undergo instability from the parallel AC-DC system. Furthermore, this thesis utilizes the time domain-based test signal and phase correction methods to design the supplementary subsynchronous damping controllers. These two approaches allow obtaining the optimal parameters of the damping controllers. The controller consists of a Power System Stabilizer to control the inertial mode and Synchronous Damping Controller to control other unstable torsional modes to achieve positive damping. This research uses the FFT analysis to assess the performance of supplementary HVDC damping controllers on the torsional dynamics of the turbogenerator shaft in the parallel AC-DC system.
  • Item
    Solution of combined heat and power economic dispatch problem using genetic algorithm.
    (2022) Ohaegbuchi, Dedacus Nnadozie.; Saha, Akshay Kumar.
    The combination of heat and power constitutes a system that provides electricity and thermal energy concurrently. Its high efficiency and significant emission reduction makes it an outstanding prospect in the future of energy production and transmission. The broad application of combined heat and power units requires the joint dispatch of power and heating systems, in which the modelling of combined heat and power units plays a vital role. This research paper employed genetic algorithm, artificial bee colony, differential evolution, particle swarm optimization and direct solution algorithms to evaluate the cost function as well as output decision variables of heat and power in a system that has simple cycle cogeneration unit with quadratic cost function. The system was first modeled to determine the various parameters of combined heat and power units in order to solve the economic dispatch problem with direct solution algorithm. In order for modelling to be done, a general structure of combined heat and power must be defined. The system considered in this research consists of four test units, i.e. two conventional power units, one combined heat and power unit and one heat-only unit. These algorithms were applied to on the research data set to determine the required decision variables while taking into account the power and heat units, operation bound of power and heat-only units as well as feasible operation region of the cogeneration unit. Power and heat output decision variables plus cost functions from Genetic Algorithm, differential evolution, Particle Swarm Optimization and artificial bee colony were determined using codes. Also, the decision variables and cost function value were obtained by calculations using direct solution algorithm. The findings of the research paper show that there are different ways in which combined heat and power economic dispatch variables can be determined, which include genetic algorithm, differential evolution, artificial bee colony, particle swarm optimization and direct solution algorithms. However, each solution method allows for different combined heat and power output decision variables to be found, with some of the methods (particle swarm optimization and artificial bee colony) having setbacks such as: large objective function values, slower convergence and large number solution. The analysis revealed that the differential evolution algorithm is a viable alternative to solving combined heat and power problems. This is due in most part to its faster convergence, minimum cost function value, and high quality solution which are diverse and widespread, more as a result of its effective search capability than genetic algorithm, particle swarm optimization, direct solution and artificial bee colony algorithms. The methods investigated in this research paper can be used and expanded on to create useful and accurate technique of solving combined heat and power problems.
  • Item
    Sub transmission insulator and conductor ageing in coastal environments.
    (2022) Ramsoomar, Ashwin.; Swanson, Andrew Graham.
    Globally, power grids are gradually unbundling to encourage multiple sources of small to medium scale renewable generation, aiding in relieving demanding power constraints [1]. The additional generation product requires an effective and efficient power evacuation system; however, the constrained transmission and sub transmission systems are burdened with the higher power transfer limits coupled with aged infrastructure. Operations and maintenance strategies assist in alleviating the temporary system load increases; however, an asset management strategy is required to ensure that the power system operates at a peak without encouraging a redesign or replacement of the entire power system [2]. The guiding principles of an asset management strategy is understanding the effective utilization of sub systems and components. Effective utilization of components can only be achieved, once a useful lifespan is determined, replacing, refurbishing, or instituting life extension measures once the component end of life is neared. To avert underutilizing or overutilizing the asset, the asset age base needs to be determined incorporating in-situ conditions to provide a reliable and generic prediction model. In sub transmission and transmission systems, maintenance funding, is allocated to critical components such as insulators, ACSR phase conductor and galvanized steel shield wire. These high failure-impact components are susceptible to a higher rate of corrosion related failure, attributed to environmental conditions such as high humidity and high air contamination found in coastal environments. The corrosion models guided by ISO 9223 [3], makes use of empirical degeneration rates for Aluminium, Steel and Zinc. These models understate the initiation of the corrosion reaction for the first year only, requiring the derivation of an acceleration model to represent the degeneration of the composite conductor over its lifespan. The resulting model predicts the useful life of the conductor by measuring the loss in tensile strength attributed to corrosion. Age prediction modelling of the silicone composite insulator is achieved by predicting the loss of creepage or resistive layer on the insulator surface. The coastal environmental conditions, contribute significantly to degenerating the resistive layer leading to a flashover at line voltage. The validation of the insulator and conductor ageing models are achieved by comparing the calculated lifespans to measured field-samples tests. The results showed a standard deviation of 4 years between the measured and calculated values.
  • Item
    Renewables and energy storage to optimise the amount of electrical energy consumed in a building.
    (2022) Naidoo, Ulika Devina.; Swanson, Andrew Graham.
    The rising cost of electricity and fuel, along with the looming threat of load shedding has frustrated not only the business owner but the homeowner as well. The need to reduce costs, and the growing pressure for companies and individuals to become more environmentally friendly, is becoming more apparent. To reduce costs and the effects of load shedding, and to become more sustainable, the integration of renewable energy is a clear solution. The solution has led to the investigation of a hybrid system that uses grid-supplied power and renewable energy supplied power which will achieve an effective and efficient optimization of cost. This dissertation is centred on minimising the total cost of ownership over twenty years. This is done by comparing different optimisation algorithms and identifying a cost-effective way of integrating a source of renewable energy, specifically solar energy, with an existing grid-supplied building. The zoning of buildings was found to have an impact on the total cost of ownership as the tariffs were different. By developing a function, the efficiency of a system was quantified based on the load, and what type of building it was. The load has a direct impact on the total cost of ownership. The electrical energy used in a building, and the property type, whether industrial, commercial or residential zone, affects the optimisation algorithm that is used. To minimise the total cost of ownership over twenty years, consideration was given to trade-offs between the available solar, oversizing the PV installation, the cost of electricity at different hours and the use of a storage system. To ensure that the total cost of ownership was correct, financial equations for growing annuity and the prescribed rates for assets, maintenance, and electricity were used. Further to this, South African energy tariffs, actual prices of inverters, solar panels, batteries and solar data of South Africa was used. MATLAB was the application of choice of software due to its optimisation capabilities. Examples of each type of building were analysed to find the optimisation that returned the lowest TCO. Particle Swarm Optimisation, when used for industrial buildings produced the lowest TCO, while smaller loads from commercial buildings and a residential housing, showed that the lowest TCO came from Teaching-Learning Based Optimisation. In each case, the fastest and slowest optimisation technique was Pattern Search and Firefly Optimisation respectively.
  • Item
    Optimizing the protection of an auto-recloser in a DG integrated distribution network.
    (2022) Gumede, Siyabonga Brian.; Saha, Akshay Kumar.
    The integration of distributed generation into distribution networks is growing as most of the distributed generators have a sustainable power supply and can be used to improve the voltage profile. However, the type of a distributed generator and location in the distribution network can determine how a voltage profile behaves in a distribution feeder. They also contribute fault current in a new or same direction as the fault current from the utility. With this change in the fault current, the existing protection scheme may maloperate since the protection scheme was designed for fault current from the utility generator. One of the protection devices that can mal-operate is the auto-recloser. This is a device used for the self-remediation of the distribution network when there is a temporary fault. The IEEE and IEC standard for the international use of auto-reclosers in voltages between 1000 V and 38 kV states that the minimum tripping current shall be stated by the manufacturer with a tolerance not exceeding +/- 10% or 3 A, and the preferred operating sequence for auto-reclosers shall be; open – time delay of 0.5 seconds - close and open-second time delay 2 seconds - close and open - third-time delay of 5 seconds - close and open then lock out. However, these parameters can be violated when distributed generators are introduced into the distribution network. The change in the fault current may vary the operating time of the auto-recloser and it may not operate in this manner. The inverse time-current characteristics of the auto-recloser relay cause this. However, the operating time problem can be optimized. The inverse time-current characteristic of the auto-recloser relay can be used to formulate the auto-recloser operating time problem. The settings can be optimized to reduce the time and mitigate mal-operations such as protection blinding, fuse and auto-recloser losing coordination, and sympathetic tripping. To optimize the settings, optimization algorithms can be applied. In this research, the development of a single-shot auto-recloser is conducted. The IEEE 13-node and 34- node radial distribution feeders are used as a passive distribution network. The Wind Turbine and Solar Photovoltaic systems are distributed generators. MATLAB/Simulink is used for simulations, and the results obtained show that the integration of the distributed generators into a passive distribution network causes mal-operations in the auto-recloser when there is a fault. The factors that contribute to these mal-operations is the fault location, fault type, distributed generator type, distributed generator penetration and location. However, the auto-recloser shows improvement when the settings are optimized in these conditions.
  • Item
    Investigation of traction motor control systems for electric vehicle applications.
    (2022) De Klerk, Matthew Liam.; Saha, Akshay Kumar.
    Electric vehicles are a promising solution to the current pollution and greenhouse gas issues faced by the transport sector. As such, the traction motor control system of an electric vehicle is worthy of investigation. Direct torque and indirect field-oriented control are commonly applied control techniques, enabling advanced control of the induction and permanent magnet synchronous motors currently used in most electric vehicles being produced. Various improvements have been made to current traction motor control schemes to reduce ripple, improve parameter insensitivity, and increase powertrain efficiency. Consequently, the objective of the research conducted is to contribute to the field of electric vehicle powertrains through comprehensive investigations into the suitability and performance of direct torque and indirect field-oriented control in the traction motor control system of an electric vehicle. A four-stage simulation-based investigation was undertaken, with five motor control techniques initially assessed, which were conventional direct torque and field-oriented control, two space vector modulation-based direct torque control systems and fuzzy logic-based direct torque control. Results from the first stage of the simulation-based study highlighted expected issues with conventional direct torque control and showed that fuzzy logic-based direct torque control and space vector modulational-based direct torque control with closed-loop torque and flux control present promising solutions for use in the traction motor control system of an electric vehicle. Extensions of the simulation-based investigation in stages two and three included the integration and assessment of field-weakening control and sensorless speed estimation. Furthermore, stage four concluded the investigation with an essential analysis of a complete control mechanism in realistic urban and highway driving conditions. The fourth stage utilised sections of the New York City Cycle and Highway Fuel Economy Test cycle, with a simulated vehicle load. The complete study indicated that space vector modulation-based direct torque control with closed-loop torque and flux control performs suitably for electric vehicle applications, providing favourable speed, torque, current and stator flux results with a faster computation time than some comparable control options. The comprehensive investigation extends current literature and forms a basis for further investigation in the field of traction motor control systems for electric vehicle applications.
  • Item
    Implementation of Iterative Learning Control on a Pneumatic Actuator.
    (2022) Rwafa, James.; Ghayoor, Najafabadi Farzad.
    Pneumatic systems play a pivotal role in many industrial applications, such as in petrochemical industries, steel manufacturing, car manufacturing and food industries. Besides industrial applications, pneumatic systems have also been used in many robotic systems. Nevertheless, a pneumatic system contains different nonlinear and uncertain behaviour due to gas compression, gas leakage, attenuation of the air in pipes and frictional forces in mechanical parts, which increase the system’s dynamic orders. Therefore, modelling a pneumatic system tends to be complicated and challenges the design of the controller for such a system. As a result, employing an effective control mechanism to precisely control a pneumatic system for achieving the required performance is essential. A desirable controller for a pneumatic system should be capable of learning the dynamics of the system and adjusting the control signal accordingly. In this study, a learning control scheme to overcome the highlighted nonlinearity problems is suggested. Many industrial processes are repetitive, and it is reasonable to make use of previously acquired data to improve a controller’s convergence and robustness. An Iterative Learning Control (ILC) algorithm uses information from previous repetitions to learn about the system’s dynamics. The ILC algorithm characteristics are beneficial in real-time control given its short time requirements for responding to input changes. Cylinder-piston actuators are the most common pneumatic systems, which translate the air pressure force into a linear mechanical motion. In industrial automation and robotics, linear pneumatic actuators have a wide range of applications, from load positioning to pneumatic muscles in robots. Therefore, the aim of this research is to study the performance of ILC techniques in position control of the rod in a pneumatic position-cylinder system. Based on theoretical analysis, the design of an ILC is discussed, showing that the controller can satisfactorily overcome nonlinearities and uncertainties in the system without needing any prior knowledge of the system’s model. The controller has been designed in such a way to even work on non-iterative processes. The performance of the ILC-controlled system is compared with a well-tuned PID controller, showing a faster and more accurate response.
  • Item
    Frequency response analysis of a current limiting reactor.
    (2022) Kuppan, Levashen.; Swanson, Andrew Graham.; Jarvis, Alan Lawrence Leigh.
    With the demand for electricity continuously increasing, power systems are required to increase capacity to meet such demands which can entail integrating renewable energy resources to the grid. This increase in capacity would mean a likewise increase in fault levels in the network which can result in costly damage to components such as circuit breakers, transformers and cables. Air-core reactors are commonly employed to prevent such damages from occurring, however, the increase in fault levels must also be accounted for in the design of reactors as they are also subject to transients. This dissertation documents the development of models to accurately represent an Air-core reactor in order to gain a better understanding of the design considerations required. Two models are developed for two desktop reactors using different methods as a form of cross-validation. The first model is developed in MATLAB r2020a and utilises an analytical approach through an equivalent circuit method (ECM). Equations are used to compute the inductive, capacitive and resistive components which are then used to guide the development of the FEM models. The second model is developed using COMSOL Multiphysics software which is based on the Finite Element Method (FEM) approach. A 2D-axisymmetrical model is constructed and simulated using COMSOL’s Magnetic and Electric field physics in the frequency domain from which a frequency response is obtained as well as values for the inductive, resistive and capacitive components. Final validation of the FEM models is done through comparisons to measured results of the two desktop reactors. FEM simulated RLC components showed fairly good agreement to the measured values, particularly the inductance having a difference of 3.4 μH and a capacitance difference of 1 pF for Reactor 1. The FEM simulated frequency response of 1.5 MHz differed by 0.4 MHz when compared to the measured frequency response for Reactor 2 of 1.9 MHz. A sensitivity analysis is conducted for the FEM model in order to obtain an understanding of the design considerations required for the air-core reactor. Simulations are performed on the FEM model with changes to geometry, permittivity of the insulation medium and resistivity of the copper coil. The effects of these changes on the RLC parameters and resonance frequencies are documented. The FEM model is then scaled to a full-scaled reactor which showed good agreement between the expected inductance of 2.24 mH and the simulated inductance of 2.28 MH. The resultant resonant frequency was observed to occur at 380 kHz. The aim of this is to develop an understanding of parameters and equations that should be considered in the design process of reactors which will then be employed in the development of a superconducting fault current limiter (SFCL).
  • Item
    Optimization of Hybrid Renewable Electrical Energy Systems for a Port.
    (2022) Nkosi, Mbonisi Shephaerd.; Swanson, Andrew Graham.; Carpanen, Rudiren Pillay.; Ghayoor, Najafabadi Farzadna.
    The port terminals in South Africa currently face significant challenges linked to electricity constraints and security. South Africa is predominantly reliant on coal-generated electricity from the local utility for the provision of base-load electricity to power operations, and municipal distribution systems in many areas. A reduction in power in the latter part of 2015 has resulted in the curtailment of load shedding that is continuing to date, 2019. This risk is not mitigated, it is ongoing. Port terminals in South Africa rely on the throughput of the vessels to ensure that freight is moved speedily. The aspects of port operations that are generally considered in measuring port performance and efficiency are berth productivity measured in moves or volumes per ship working hour also known as across the shipping rate, cargo dwell times, crane moves per hour, ship turnaround time with less focus or no energy management. The port operations are dependent on the throughput of the operation. The energy consumed for port operations is thus independent and is difficult to control in any sort of energy management plan. Coupling this with the time of use tariffs and demand penalties, the need for an adaptable hybrid system can alleviate some stresses and contribute positively to an energy management plan. This study aims to determine that the distributed generation and storage performed by a variety of small, grid-connected referred to as distributed energy resources mitigate the energy cost issues in the port environment and reduction of Green House Gases (GHG) emissions. Distributed resources can help reduce the capacity problems to which an aging or overstressed grid is liable. The study is contributing to reducing dependence on major power plants supply and redirecting the source of supply to renewable technologies. This results in eliminating the need to erect new big power generation and deferral of new capacity. It also demonstrates that PV Solar and Wind Turbine Generation can reduce environmental impacts and gas greenhouse gas emissions. Compared to the coal power plants, distributed generators units are renewable or low emission generator-based sources. The Homer Grid simulation tool suggest distributed generation to be the solution for the port i.e., the hybrid supply that includes the integration of photovoltaic generation, wind power generation, battery energy storage system (BESS) into the electricity Municipal Grid. Both Matlab and HOMER Grid optimizations tools are used to outline different options for reducing electricity costs. These tools compare the costs and savings and uses a powerful optimization to find the system that maximizes savings. We used these tools to analyze the distributed generation grid potential, peak renewables penetration, ratio of renewable sources to total energy, and grid stability. The tools present different study cases with optimal results. The outcome shows a decrease in the cost of energy in the long term and can contribute towards the better energy management of the port.
  • Item
    Study the effect of topology on the performance of an advanced metering infrastructure network.
    (2021) Ngcobo, Thobekile Joyce.; Ghayoor, Najafabadi Farzadna.
    A smart grid operates based on the integration of various renewable energy sources, distributed generators and storage units in order to deliver an uninterrupted energy supply to consumers. Such a complex grid requires a network of intelligent sensors and an effective communication infrastructure to provide bi-directional flows of information between different grid entities for monitoring and control purposes. A crucial part of the smart grid communication network is the advanced metering infrastructure connecting a utility company to end-users to support telemetry and remote-control applications. Although different technologies and standards for smart metering systems exist, the G3-PLC standard, which uses the power-line communication (PLC) technology, is the accepted standard in South Africa for connecting smart meters to data concentrators. Studying the topology of an AMI network can help improve the network’s Quality-of-Service to support more advanced applications. The analytical analysis is usually considered a viable method for studying the topological effect on the performance of PLC-based AMI networks, as physically altering such networks can become very costly. Therefore, in this research, such methods have been used to investigate the effect of topology on the performance of the G3-PLC AMI network. To better understand the system, an overview of the G3-PLC standard for smart metering application has been covered. This includes covering the DLMS/COSEM protocol at the application layer and its relation to the G3-PLC. This follows by providing the mathematical model for the G3-PLC AMI network to study the effect of topology on its performance. Based on the provided method, first, the distances between data concentrators and smart meters are identified. Then the graph theory has been used to calculate the transfer function between every node in the system for obtaining the system’s total capacity. It was shown that the performance of the system decreases as longer branches are added to the network.
  • Item
    Design of a novel axial-flux induction machine for traction applications.
    (2021) Mushid, Cinyama.; Dorrell, David Goerge.
    Induction motors are an important element in the industrial world; they are used in many applications, such as electric fans, elevators, pumps, conveyor belts, compressors and now even traction motors. Electric motors consume about 70 % of all industrial power consumption. Induction machines are also the source of the power generation such as in wind turbines. In recent years, the increase in price and supply-chain issues of rare earth magnets, which are currently an important material in brushless permanent machines, which are the most popular vehicular drive motor, has led to a focus on non-permanent magnet machine replacements, such as the induction machine. The induction machine is still undergoing design development and being used in an increasing number of applications. They can be used in fixed speed (grid-connected) or variable speed (variable-frequency inverter-connected) depending on the application. Loss reduction, weight, size, as well as minimizing the cost of raw materials for manufacturing, are some of the issues in design improvement. In view of this, it is important to develop innovative methods for producing electrical machines that will reduce losses and minimizing cost of production. The aim of this research work is to develop an appropriate analytical design procedure for designing an axial-flux induction machine and to evaluate the performance of the designed machine under various conditions. The machine must be robust and cheaper. ANSYS Maxwell software is used for 3D finite element modelling and simulation of the proposed axial-flux induction machine AFIM). For fast calculation, a simple sizing exercise is done using a pre-defined stator core. Then a radial-flux machine representation is developed in Siemens SPEED motor design software for fast assessment. The electromagnetic motor model is further tested to take into account the variations in rotor design. A proof-of-concept prototype was constructed for initial validation that the machine works and this design was modelled. The result of the simulation and the measurements from the laboratory design prove the possibility of the proposed AFIM for use in automotive application. Further design was carried out to improve the prototype using more substantial windings and a longer rotor. This design was tested with ANSYS Maxwell and SPEED. The designed machine offers a cost effective solution for future drive systems in automotive applications.
  • Item
    Modelling of streamer breakdown under positive polarity HVDC in subtropical conditions.
    (2021) Moodley, Keshlan.; Swanson, Andrew Graham.
    Atmospheric air is an essential component in high voltage power systems as it is used to provide electrical insulation. Therefore, it is required that research is extensively performed on the gaseous insulator to ensure reliability of the system. This dissertation presents the research that has been performed on the streamer breakdown of needle-plane air gaps that were less than 15 cm in length. An experiment was conducted to investigate the streamer breakdown in air gaps. The investigation also included determining the influence of humidity and pressure on the breakdown voltage when HVDC is applied to the high voltage electrode in subtropical conditions. The results indicated that humidity did not influence the breakdown voltage for small gap lengths but both humidity and pressure have a directly proportional relationship with the breakdown voltage of air gaps longer in length. It was observed in the results that the breakdown voltage reduced in magnitude when the humidity increased, and this could have been a result of the breakdown developing out of the glow discharge without impulses when the HVDC was applied. It was observed that the atmospheric pressure did not influence the breakdown voltage for the smaller gap lengths, but the breakdown voltage was directly proportional to the pressure for larger gap lengths. High-speed photography was implemented in the experiment to capture the streamer breakdown mechanism that occurred in these air gaps. An analysis of the results obtained from the experiment showed that the needle-plane electrode combination created single, straight streamers whose stem increased in length when the gap length increased. Branches would also occur from the streamer when the gap length was greater than or equal to 12 cm and the streamer would occasionally take a bent path for gap lengths greater than 10 cm when propagating towards the cathode. Streamers would take this bent path because of space charge being present along the normal path when propagating towards the cathode. This caused the streamer to propagate around the space charge as it is unable to propagate through regions of high concentration. The dissertation also consisted of a numerical simulation on COMSOL Multiphysics which modelled the initiation of a cathode directed streamer and the propagation of it in an air gap. The necessary steps that were taken for the implementation and simulation of the electrical streamer in a 0.1 cm needle-plane gap is also included. The results of the numerical simulation were presented and analysed. The model was used to investigate the influence of both humidity and pressure on the breakdown voltage. The results indicated that the breakdown voltage of an air gap increased when humidity increased in the system as the atmospheric pressure was controlled. The breakdown voltage is also directly proportional to the atmospheric pressure when the humidity in the system was controlled. The experimental results and the numerical results obtained indicate that the breakdown voltage of an air gap is dependent on the humidity and pressure of the system and that space charge influences the path a streamer takes when propagating to the cathode.
  • Item
    Voltage control and stability analysis in a multi-machine power system with increasing penetration of intermittent renewable energy generation.
    (2020) Loji, Kabulo.; Tiako, Rémy.
    Among multiple distributed generation (DG) supply means, photovoltaic (PV) and wind technologies are the most important and widely used renewable energy sources (RES) throughout the world. However, solar intermittency and the stochastic nature of radiation on one hand and grid integration-related issues on the other are fundamental concerns in the development and smooth deployment of solar energy contribution to conventional power systems networks. In addition, given that they modify both the structure and the operation of the distribution networks, RES increase uncertainty in power system operations, thus affecting power systems variables such as the voltage profiles and direction of network power flows. It is also largely established that a high penetration of DGs at the distribution end is associated amongst others, with voltage rises at PV buses that may lead to the violation of grid codes, if not adequately mitigated. There is a need to investigate both the effect and the impact of increasing penetration of these intermittent RES on, particularly, voltage and frequency stability power systems and the utilization thereof of such sources to improve voltage stability margins and predict voltage stability conditions. This research work investigated voltage control and stability conditions at Solar PV buses through various case studies and scenarios simulated using the Power Factory® tool, both in static and dynamic analysis modes. A modified standard IEEE 9-Bus Sub-transmission system was used to assess the voltage profile, system loadability and system stability. The comparison and discussion of the results obtained from the integration of the Solar PV and FACTS devices under various scenarios revealed that their respective impacts and abilities to improve voltage stability differ. The results confirmed that under any operating conditions, reactive power control remains the most effective method to control voltage stability and power transfer capability, especially in the context where an increasing penetration of renewable and inertia-less generating sources is planned. The results further revealed that there is a specific location and a specific siting architecture for a given size of PV that produces the best results for voltage stability, as well as improved system stability and loadability conditions for a given load distribution profile in a particular network. Lastly, the results demonstrated the effectiveness of the use of a Battery Energy Storage System (BESS) in achieving voltage control and regulation in distribution networks highly penetrated by PV generation, subsequently enabling greater RE penetration.
  • Item
    Optimal sizing for a grid-connected hybrid renewable energy system.
    (2021) Sibanda, Hudson.; Ghayoor, Najafabadi Farzad.; Swanson, Andrew Graham.
    Hybrid renewable energy systems (HRESs) refer to power generating systems that integrate several sources of energy, including renewables, to provide electricity to consumers. HRESs can either work as standalone or grid-connected systems. Since wind and solar have complementary characteristics and are available in most areas, they are considered as suitable energy sources to be combined in an HRES. Moreover, the maturity of technologies needed for generating electricity from wind and solar has turned them into more economical options in many locations. Many countries, including South Africa, have introduced policies and incentives to increase their renewable energy capacities in order to address environmental concerns and reduce pollutant emissions into the atmosphere. In addition, consumers in South Africa have faced the ever-increasing price of electricity and unreliability of the grid since 2007 due to the lack of sufficient electricity production. As a result, employing HRESs has gained popularity among consumers in different sectors. This research is focused on grid-connected hybrid energy systems based on solar photovoltaic (PV) panels and wind turbines as a potential solution to reduce the dependency of residential sector consumers on the grid in Durban. The aim of the research is to identify the optimal sizing of such a HRES to be cost-effective for consumers over a certain period of time. Since the energy supplied by renewable sources are intermittent and dependent on the geographical location of the system, identifying optimal sizing becomes a challenging task in HRESs. In this research, Durban’s meteorological data and eThekwini municipality tariff rates have been considered. Moreover, two artificial intelligence methods have been used to obtain the optimal sizing for different types of available PV panels, wind turbines and inverters in the market. The results have shown that the combination of PV panels and battery storage (BS) can become a profitable option for Durban area. Moreover, the systems using higher rated power PV panels can start to become profitable in a shorter lifetime. Considering BS in a system can only become a cost-effective choice if we consider a long enough lifespan for the system.
  • Item
    Investigating lightning arrester behaviour in substations.
    (2020) Mzulwini, Buyisile Zwe.; Saha, Akshay Kumar.
    Abstract available in PDF.
  • Item
    Power demand and supply allocation using inherent structural theory of network systems and voltage stability index based on multi-bus reactive power loading.
    (2021) Mutuku, Peter Munyao.; Agee, John Terhile.; Tiako, Remy.
    Power availability is a crucial factor in determining new load centers. There is a need for adequate power reserve and maximum load capacity allocation to ensure continued power demand and supply electrical systems. In modern interconnected power systems, a high peak load power demand is met by the contribution of the available generator units. There is an urgency to solve the challenges arising from interconnected network configurations such as the loss of generation, inadequate supply capacity to meet load demand during peak time, transmission losses and significant voltage drop at the heavily loaded buses. This dissertation investigates the influence of inter-connected load buses on the system’s voltage profile and the electrical proximity from generation sites to load centers as captured by the Y-admittance matrix. The inherent structural theory of networks was used in determining the required power reserve and load capacity allocation using the ideal generator contribution index. PowerWorld simulator, Dig SILENT Power Factory and MATLAB were used as simulation and presentation tools for the modified IEEE 14 bus system and the Southern Indian 10 bus system. From the analysis of the results, much of the load capacity needed for electrical load growth is feasible for the bus that is most electrically proximal to a high-rated power source. The use of the ideal generator contribution index exploits the structural properties of the network. That being the case, its advantages include minimum expansion of existing structures and minimal transmission active power losses. Also, in this dissertation, a V-Q curve characteristic approach was used to identify the weak load buses in an interconnected power system. This was done by simulating uniformly distributed multi-bus loading conditions and the conventional analysis of the sole bus loading method in a power network system up to the minimum acceptable per unit voltage point. This lead to the formulation of a novel V-Q curve-based index. The voltage critical multi-bus index is a variable state-based index. This index was compared with the self-sensitivity index of the reduced Jacobian matrix and the ‘load structural electrical attraction region’ index of the inherent structural theory of power networks, giving a deeper insight into the system characteristics under light and heavy loading states.
  • Item
    Application of optimization techniques to solve overcurrent relay coordination.
    (2021) Langazane, Sethembiso Nonjabulo.; Saha, Akshay Kumar.
    Distribution systems continues to grow and becoming more complex with increasing operational challenges such as protection miscoordination. Initially, conventional methods were favoured to optimize protection coordination; however, the implementation process is laborious and time-consuming. “Therefore, recent studies have adopted the utilisation of particle swarm optimization and genetic algorithms to solve overcurrent relay coordination problems and maximise system selectivity and operational speed. Particle swarm optimization and genetic algorithms are evolutionary algorithms that at times suffer from premature convergence due to poor selection of control parameters. Consequently, this thesis aims to present a comprehensive sensitivity analysis to evaluate the effect of the discrete control parameters on the performance of particle swarm optimizer and genetic algorithms, alternatively on the behaviour of overcurrent relays. The main objectives of this research work also include modelling and simulation of distribution system protection scheme, employment of evolutionary algorithms with control parameters that perform efficiently and effectively to maximise protection coordination between relays, optimize relay operating time and maintain the stipulate coordination time interval, and lastly, to outline future recommendations. The distribution network understudy was modelled and simulated on a real-time digital simulator to validate protection settings, and the verification of evolutionary algorithms performance was displayed on Matlab/Simulink. An extensive parametric sensitivity analysis was conducted to understand the impact of the individual control parameters and their respective influence on the performance of evolutionary algorithms. The findings indicate that particle swarm optimization is more sensitive to inertia weight and swarm size while the number of iterations has minimal effect. The results also depict that genetic algorithms’ performance is mostly influenced by crossover probability, mutation probability, and population size. Sensitivity analysis results were verified by comparing the performance of particle swarm optimizer with genetic algorithms, which demonstrated that particle swarm optimization performs efficiently and robustly in solving the considered problem, especially in terms of convergence speed. Furthermore, overcurrent relays were more sensitive, selective, and the operational speed was reduced for particle swarm optimizer compared to other algorithms. The optimal protection coordination achieved using particle swarm optimization showed superiority of the algorithm, its ability to circumvent premature convergence, consistency, and” efficiency.
  • Item
    Model predictive control of a doubly fed induction generator.
    (2020) Reddy, Kumeshan.; Saha, Akshay Kumar.
    The world is currently is energy despair. For years, the world has relied on fossil fuels as the main energy source to produce electricity. At the start, this worked well as there was an abundance. However, due to the increase in population, urbanisation and the birth of many industries, this fuel source has been put under strain. Furthermore, the harmful emissions from the use of fossil fuels has been a great contributor to the destruction of our precious ozone layer. This in turn has gradually increased the harmful effects of global warming on Earth. The need for clean, reliable sources of energy has increased over time, and in a few years, it is expected to be the only source of energy utilized in the production of electrical energy. The research undertaken in this project involves the control of the doubly fed induction generator, which is used in wind energy conversion systems. Commonly termed DFIG, this generator has gained worldwide popularity and is used in majority of wind energy conversion systems. It provides direct grid connection and uses only a partially rated converter. However, the conventional control methods used in the control of the DFIG are either difficult to implement or inefficient. Some require complex tuning of proportional-integral controllers while some produce distorted results. The aim of this research was to investigate and evaluate the application of model predictive control to the control of the DFIG. There exist various different control strategies for the control of the DFIG. This research involved implementing all of the different control strategies via conventional methods and then via the use of model predictive control. Despite there being various methods to implement model predictive control, due to its simplicity and strong suitability, finite control set model predictive control was used in this research. Each of the control strategies implemented both conventionally and via model predictive control were thoroughly analysed in terms of the steady state response, dynamic response and quality of stator current. A comparison between the corresponding control methods is also presented.
  • Item
    Investigation of different graphite precursors for graphene oxide supercapacitors.
    (2021) Perumal, Solan.; Jarvis, Alan Lawrence Leigh.
    The impact of non-renewable energy sources has had adverse effects on the environment resulting in climate change. Many countries have undertaken a call for renewable energy sources as a cleaner and more sustainable alternative. With the change towards renewable energy sources, storing this energy has become a challenge due to its intermittent nature. An energy storage device that could help solve the above problem is a supercapacitor due to its high power density, long cycle life, and high rate capability, which are desirable characteristics for energy storage devices. Supercapacitors downfall is their low energy density. Improvement in the electrode material of the supercapacitor may help address the low energy density issue. A promising candidate is graphene oxide (GO). GO has shown notable potential as electrode material in past research due to the pseudocapacitance effect. The primary precursor for the synthesis of GO is graphite. Varying graphite precursors may yield GO with different properties. In this research, graphite precursors with different characteristics were investigated to determine the effect they have on GO supercapacitor energy storage capabilities. Eight graphite precursors were used to synthesise GO. The samples were characterised using High-Resolution Transmission Electron Microscopy (HRTEM), Scanning Electron Microscopy (SEM), Elemental Analysis, Fourier Transform Infrared (FTIR) Spectroscopy and Raman Spectroscopy. GO supercapacitors were fabricated using GO as an active electrode material, stainless steel plates as current collectors, and phosphoric acid (H3PO4) hydrogel polymer as electrolyte and separator. The electrochemical testing conducted on GO electrode material were Cyclic Voltammetry (CV) at different scan rates to determine specific capacitance and energy density. It was found that increasing flake size of natural graphite precursors produced GO with higher specific capacitance with an implicit limiting point. With the lack of peaks between the voltage limits of the CV curves for GO produced from natural graphite precursors, this indicated that the pseudocapacitance effect from oxygen functional groups is insignificant to the overall specific capacitance for these samples. These results led to further research into other possible factors that can be playing a role in GO’s high specific capacitance. It was observed that GO produced from the smallest flake size (0.045 mm) synthetic graphite precursor had the highest specific capacitance compared to GO produced from natural graphite precursor of all flake sizes investigated in this research. Firstly, the synthetic GO sample produced from smallest flake size had higher crystallinity compared to natural GO samples which were estimated using Raman spectroscopy. Secondly, high oxygen content shown in elemental analysis and peaks observed between voltage limits of CV curves provided a high pseudocapacitance effect that is significant to the overall capacitance. Thirdly, low amount of defects determined from low ID/IG (Intensity of D band/Intensity of G band) ratio in Raman spectrum may enable the ions from the electrolyte to move through the GO structure efficiently. It is the combination of these characteristics that attribute to GO produced from smallest flake size (0.045 mm) synthetic graphite precursor that improved energy storage capabilities and be an excellent electrode material to use in supercapacitors.
  • Item
    Optimal placement of statcom controllers with metaheuristic algorithms for network power loss reduction and voltage profile deviation minimization.
    (2020) Ogunwole, Emmanuel Idowu.; Saha, Akshay Kumar.
    Transmission system is a series of interconnected lines that enable the bulk movement of electrical power from a generating station to an electrical substation. This system suffers from unavoidable power losses and consequently voltage profile deviation which affects the overall efficiency of the system; hence the need to reduce these losses and voltage magnitude deviations. The existing methods of incorporation of static synchronous compensator (STATCOM) controllers to solve these problems suffer from incorrect location and sizing, which could bring about insignificant reduction in transmission network losses and voltage magnitude deviations. Hence, this research aims to reduce transmission network losses and voltage magnitude deviation in transmission network by suitable allocation of STATCOM controller using firefly algorithm (FA) and particle swarm optimization (PSO). A mathematical steady-state STATCOM power injection model was formulated from one voltage source representation to generate new set of equations, which was incorporated into the Newton-Raphson (NR) load flow solution algorithm and then optimized using PSO and FA. The approach was applied to IEEE 14-bus network and simulations were performed using MATLAB program. The results showed that the best STATCOM controller locations in the system after optimization were at bus 11 and 9 with the injection of shunt reactive power of 8.96 MVAr, and 9.54 MVAr with PSO and FA, respectively. The total active power loss for the network under consideration at steady state, with STATCOM only and STATCOM controller optimized using PSO and FA, were 6.251 MW, 6.075 MW, 5.819 MW and 5.581 MW, respectively. The corresponding reactive power were 14.256 MVAr, 13.857 MVAr, 12.954 MVAr and 12.156 MVAr, respectively. In addition, bus voltage profile improvement indicates the effectiveness of metaheuristic methods of STATCOM optimization. However, FA gave a better power loss and voltage magnitude deviations minimizations over PSO. The study concluded that FA is more effective as an optimization technique for suitably locating and sizing of STATCOM controller on a power transmission system.