Browsing by Author "Awodele, Kehinde"

Now showing 1 - 13 of 13
  • Thumbnail Image
    Item
    Open Access
    An investigation into 88 KV surge arrester failures in the Eskom east grid traction network
    (2022) Mzulwini, Mduduzi Comfort; Awodele, Kehinde; Gaunt, Trevor
    The Eskom East Grid Traction Network (EGTN) supplying traction loads and distribution networks has experienced at least one surge arrester failure over the past ten years. These failures results in poor network reliability and customer dissatisfactions which are often overlooked. This is because reliability indices used in the reliability evaluation of transmission and distribution networks are different. It is suspected that fast transient faults in this network initiate system faults leading to surge arrester design parameter exceedances and poor network insulation coordination. Preliminary investigations in network suggest that transient studies were not done during network planning and design stages. This may have resulted in the lack of surge arrester parameter evaluations under transient conditions leading to improper surge arresters being selected and installed in this network resulting in surge arrester failures that are now evident. These failures may also have been exacerbated by the dynamic nature of traction loads as they are highly unbalanced, have poor power factors and emit high voltage distortions. Poor in-service conditions such as defects, insulation partial discharges and overheating, bolted faults in the network and quality of supply emissions can also contribute to surge arrester failures. To address problems arising with different reliability indices in these networks the reliability of the EGTN is evaluated. In this work the reliability evaluation of the EGTN is done by computing common distribution reliability indices using analytic and simulation methods. This is done by applying the analytic method in the EGTN by assessing network failure modes and effects analysis (FMEA) when the surge arrester fails in this network. The simulation method is applied by applying and modifying the MATLAB code proposed by Shavuka et al. [1]. These reliability indices are then compared with transmission reliability indices over the same period. This attempts to standardize reliability evaluations in these networks. To assess the impact of transient faults in the surge arrester parameter evaluation the EGTN is modelled and simulated by initiating transient faults sequentially in the network at different nodes and under different loading conditions. This is done by using Power System Blockset (PSB), Power System Analysis Toolbox (PSAT) and Alternate Transient Program (ATP) simulation tools and computing important surge arrester parameters i.e. continuous operating voltage, rated voltage, discharge current and energy absorption capability (EAC). These parameters are assessed by in the EGTN by evaluating computed surge arrester parameters against parameters provided by manufacturers, the Eskom 88 kV surge arrester specification and those parameters recommended in IEC 60099-4. To assess the impact and contribution of in-service conditions, faults and quality of supply emissions in surge arrester failures these contributing factors are investigated by assessing infra-red scans, fault analysis reports, results of the sampled faulted surge arrester in this network and quality of supply parameters around the time of failures. This study found that Eskom transmission and distribution network reliability indices can be standardized as distribution reliability indices i.e. SAIDI, SAIFI, CAIDI, ASAI and ASUI indices are similar to Eskom transmission indices i.e. SM, NOI, circuit availability index and circuit unavailability index respectively. Transient simulations in this study showed that certain surge arresters in the EGTN had their rated surge arrester parameters exceeded under certain transient conditions and loading conditions. These surge arresters failed as their discharge currents and EACs were exceeded under heavy and light network loading conditions. This study concluded that surge arresters whose discharge currents and EACs exceeded were improperly evaluated and selected prior to their installations in the EGTN. This study found the EAC to be the most import parameter in surge arrester performance evaluations. The Eskom 88 kV surge arrester specification was found to be inadequate, inaccurate and ambiguous as a number of inconsistencies in the usage of IEEE and IEC classified systems terminology were found. It was concluded that these inconsistencies may have led to confusions for manufacturers during surge arrester designs and selections in the EGTN. The evaluation of fault reports showed that two surge arrester failures in this network were caused by hardware failures such as conductor failure and poor network operating as the line was continuously closed onto a fault. There was no evidence that poor in-service and quality of supply emissions contributed to surge arrester failures in this network. PSB, PSAT and ATP simulation tools were found adequate in modelling and simulating the EGTN. However the PSB tool was found to be slow as the network expanded and the PSAT required user defined surge arrester models requiring detailed manufacture data sheets which are not readily available. ATP was found to be superior in terms of speed and accuracy in comparison to the PSB and PSAT tools. The MATLAB code proposed by Shavuka et al. [1] was found to be suitable and accurate in assessing transmission networks as EGTN's reliability indices computed from this code were comparable to benchmarked Eskom distribution reliability indices. The work carried out in this research will assist in improving surge arrester performance evaluations, the current surge arrester specification and surge arrester selections. Simulation tools utilized in this work show great potential in achieving this. Reliability studies conducted in this work will assist in standardizing reliability indices between Eskom's transmission and distribution divisions. In-service condition assessment carried out in this work will improve surge arrester condition monitoring and preventive maintenance practices.
  • Thumbnail Image
    Item
    Open Access
    Comparison of three power system software packages for small-signal stability analysis
    (2009) Mudau, Dovhani Selby; Folly, Komla A; Awodele, Kehinde
    Many power system simulation tools exist for small-signal stability analysis. This is due to the rapid development of computer systems, higher industrial growth and the need for reliable power system simulation tools for efficient planning and control of electric power systems. Three power system small-signal stability simulation tools have been selected for comparison and these are: PSAT 2.1.2, MatNetEig and PacDyn 8.1.1. These combine both open and closed source code industrial-grade power system analysis tools. The objective of this thesis is to compare three simulation tools on power system small-signal stability analysis. Input formats, data output flexibility, dynamic components and synchronous machine saturation modelling in all three simulation tools were amongst other features investigated for comparative studies.
  • No Thumbnail Available
    Item
    Open Access
    Distribution network performance improvement using load compensation
    (2023) Ojo, Alaba; Awodele, Kehinde
    The phenomenal growth in electricity demand due to population growth, industrialization, and the inadequacy of generation, transmission and distribution capacities have affected energy demand and the overall performance of the entire power system network as witnessed in South Africa since 2008. Even if additional power generating sources are provided, the power generated has to be transferred to the consumers through transmission and distribution networks which would also require additional investment for capacity expansion and this would amount to huge investment and burden for the utilities. To reduce this investment, an option could be, performance enhancement of the existing transmission and distribution system. Since the overall performance of power system networks depends on the loads connected to it, the distribution system is the backbone by which electricity supply is distributed to the customers, therefore, ensuring the economical operation of the distribution network within power quality standards is imperative. The presence of substantial single-phase loads, and large industrial loads such as arc welding, electric furnace, lightings contribute to the system unbalance and this may lead to the pollution of the supply voltages and excessive neutral current in three-phase four-wire networks. In addition, in the recent years, the use of power electronic-based devices/equipment in both industrial, commercial and residential sectors due to advancements in technology and the need for energy efficiency measures has increased the power quality problem. Such loads include Adjustable speed drives (ASD), High voltage Direct current (HVDC) system, traction systems, electric arc furnace, flexible AC transmission systems (FACTS), heating, ventilation and Air conditioning (HVAC) and loads with switch-mode power supplies. These power electronics-based devices draw nonsinusoidal currents which increased the reactive power demand from the source and other effects such as current and voltage distortion, errors in metering and malfunctioning of protective equipment, increased heating in conductors and power transformers, poor voltage regulation due in the supply's distribution system due to resonance condition. Therefore, at the beginning of this study, a power quality survey of commonly used nonlinear loads and a typical commercial building has been conducted to evaluate the existing level of power quality issues in the distribution system. Various voltage and current waveforms of some of the commonly used loads and their harmonic spectrum have been recorded using a power quality analyser. In mitigating the above problems due to poor power quality, an effective load compensation is required to improve the performance of the system. This dissertation, therefore, investigates the power quality issues, causes, effects, monitoring and evaluation. It also proposes effective load compensation techniques for performance improvement of distribution network considering unbalance, poor power factor and harmonic distortion using passive and active components. An innovative load compensation model has been proposed and developed for a three-phase, four-wire system through simulations using MATLAB/Simulink. An analysis of the distribution network with and without load compensation shows an improved power factor of about 38% and an appreciable reduction in the kVA demand of the network while the current harmonic of the load is minimize to an acceptable standards. Based on the investigations reported in this dissertation, it may be concluded that various options for load compensations in the distribution system for improving the performance of distribution networks are successfully explored.
  • No Thumbnail Available
    Item
    Open Access
    Distribution network performance improvement using load compensation
    (2023) Ojo, Alaba; Awodele, Kehinde
    The phenomenal growth in electricity demand due to population growth, industrialization, and the inadequacy of generation, transmission and distribution capacities have affected energy demand and the overall performance of the entire power system network as witnessed in South Africa since 2008. Even if additional power generating sources are provided, the power generated has to be transferred to the consumers through transmission and distribution networks which would also require additional investment for capacity expansion and this would amount to huge investment and burden for the utilities. To reduce this investment, an option could be, performance enhancement of the existing transmission and distribution system. Since the overall performance of power system networks depends on the loads connected to it, the distribution system is the backbone by which electricity supply is distributed to the customers, therefore, ensuring the economical operation of the distribution network within power quality standards is imperative. The presence of substantial single-phase loads, and large industrial loads such as arc welding, electric furnace, lightings contribute to the system unbalance and this may lead to the pollution of the supply voltages and excessive neutral current in three-phase four-wire networks. In addition, in the recent years, the use of power electronic-based devices/equipment in both industrial, commercial and residential sectors due to advancements in technology and the need for energy efficiency measures has increased the power quality problem. Such loads include Adjustable speed drives (ASD), High voltage Direct current (HVDC) system, traction systems, electric arc furnace, flexible AC transmission systems (FACTS), heating, ventilation and Air conditioning (HVAC) and loads with switch-mode power supplies. These power electronics-based devices draw nonsinusoidal currents which increased the reactive power demand from the source and other effects such as current and voltage distortion, errors in metering and malfunctioning of protective equipment, increased heating in conductors and power transformers, poor voltage regulation due in the supply's distribution system due to resonance condition. Therefore, at the beginning of this study, a power quality survey of commonly used nonlinear loads and a typical commercial building has been conducted to evaluate the existing level of power quality issues in the distribution system. Various voltage and current waveforms of some of the commonly used loads and their harmonic spectrum have been recorded using a power quality analyser. In mitigating the above problems due to poor power quality, an effective load compensation is required to improve the performance of the system. This dissertation, therefore, investigates the power quality issues, causes, effects, monitoring and evaluation. It also proposes effective load compensation techniques for performance improvement of distribution network considering unbalance, poor power factor and harmonic distortion using passive and active components. An innovative load compensation model has been proposed and developed for a three-phase, four-wire system through simulations using MATLAB/Simulink. An analysis of the distribution network with and without load compensation shows an improved power factor of about 38% and an appreciable reduction in the kVA demand of the network while the current harmonic of the load is minimize to an acceptable standards. Based on the investigations reported in this dissertation, it may be concluded that various options for load compensations in the distribution system for improving the performance of distribution networks are successfully explored.
  • No Thumbnail Available
    Item
    Open Access
    Distribution network performance improvement using load compensation
    (2023) Ojo, Alaba; Awodele, Kehinde
    The phenomenal growth in electricity demand due to population growth, industrialization, and the inadequacy of generation, transmission and distribution capacities have affected energy demand and the overall performance of the entire power system network as witnessed in South Africa since 2008. Even if additional power generating sources are provided, the power generated has to be transferred to the consumers through transmission and distribution networks which would also require additional investment for capacity expansion and this would amount to huge investment and burden for the utilities. To reduce this investment, an option could be, performance enhancement of the existing transmission and distribution system. Since the overall performance of power system networks depends on the loads connected to it, the distribution system is the backbone by which electricity supply is distributed to the customers, therefore, ensuring the economical operation of the distribution network within power quality standards is imperative. The presence of substantial single-phase loads, and large industrial loads such as arc welding, electric furnace, lightings contribute to the system unbalance and this may lead to the pollution of the supply voltages and excessive neutral current in three-phase four-wire networks. In addition, in the recent years, the use of power electronic-based devices/equipment in both industrial, commercial and residential sectors due to advancements in technology and the need for energy efficiency measures has increased the power quality problem. Such loads include Adjustable speed drives (ASD), High voltage Direct current (HVDC) system, traction systems, electric arc furnace, flexible AC transmission systems (FACTS), heating, ventilation and Air conditioning (HVAC) and loads with switch-mode power supplies. These power electronics-based devices draw nonsinusoidal currents which increased the reactive power demand from the source and other effects such as current and voltage distortion, errors in metering and malfunctioning of protective equipment, increased heating in conductors and power transformers, poor voltage regulation due in the supply's distribution system due to resonance condition. Therefore, at the beginning of this study, a power quality survey of commonly used nonlinear loads and a typical commercial building has been conducted to evaluate the existing level of power quality issues in the distribution system. Various voltage and current waveforms of some of the commonly used loads and their harmonic spectrum have been recorded using a power quality analyser. In mitigating the above problems due to poor power quality, an effective load compensation is required to improve the performance of the system. This dissertation, therefore, investigates the power quality issues, causes, effects, monitoring and evaluation. It also proposes effective load compensation techniques for performance improvement of distribution network considering unbalance, poor power factor and harmonic distortion using passive and active components. An innovative load compensation model has been proposed and developed for a three-phase, four-wire system through simulations using MATLAB/Simulink. An analysis of the distribution network with and without load compensation shows an improved power factor of about 38% and an appreciable reduction in the kVA demand of the network while the current harmonic of the load is minimize to an acceptable standards. Based on the investigations reported in this dissertation, it may be concluded that various options for load compensations in the distribution system for improving the performance of distribution networks are successfully explored.
  • Thumbnail Image
    Item
    Open Access
    Improving the reliability performance of medium voltage networks
    (2015) Khan, Zahir Hoosain; Awodele, Kehinde; Gaunt, C Trevor
    The aim of this dissertation is to investigate alternative, more reliable and cost effective ways of improving the reliability performance of medium voltage networks. Customers are mainly affected by faults on the distribution MV network, to which, consequently, we have to pay particular attention. A major requirement on electricity supply systems is high supply reliability for the customer which is mainly determined by the distribution networks. Power system reliability is an essential factor in the quality of supply and is directly related to the number and duration of outages. By analysing the power system properly, the weaknesses will then be identified and improvements can be introduced to minimise the occurrence of outages. A decrease in the outage rate will result in an improvement in reliability and quality of supply of the distribution MV network. The dissertation focuses on improving the network management by increasing the level of network automation and control which improves the operating efficiency of medium voltage distribution networks. Steps are shown how to equip the network according to progressive investment capability, from Fault Path Indicators (FPIs) and remote control Pulseclosing technologies to automatic FuseSavers and Tripsavers used in a feeder automation scheme to minimise the number of disturbances and the outage durations experienced when they occur. The results of a study analysing the impact of different intelligent automation solutions on the reliability performance of Medium Voltage distribution networks are presented in the dissertation. The respective system topologies are modelled and the resulting system reliability performance is determined by reliability calculations such as the SAIDI and SAIFI values. The results show that the distribution automation technologies can have a very significant impact on both the SAIDI and SAIFI performance of the systems. Further, selected details related to the implementation of such intelligent automation schemes are presented in this dissertation.
  • Thumbnail Image
    Item
    Open Access
    Incorporating time and statistical variations in load modelling for reliability and customer interruption costs evaluations
    (2013) Ip Cho, Nicolas Foumg Sian; Awodele, Kehinde
    This study was carried out with two main objectives. Firstly to investigate different types of load modelling in reliability or CIC [customer interruption costs] evaluations to obtain an understanding on how these evaluations are performed and identifying the benefits or limitations of existing models. Secondly to develop several load modelling approaches, including a time dependent beta PDF load model, which are simulated using an appropriate simulation technique and software. A comparison using historical load data based on South African residential customers and shops (used as commercial customers) with the different load models is performed and their impact on the calculated reliability or cost indices are analysed. The analysis shows that the reliability and cost indices resulted in different values of varying degree as compared to the base case (average load model) when using reconfiguration with sufficient spare capacity.
  • Thumbnail Image
    Item
    Open Access
    Investigation and analysis of the causes of 11/0.4 kV distribution transformers' high failure rate : case study - Swaziland Electricity Company
    (2016) Maziya, Mphumuzi Thembinkosi; Awodele, Kehinde
    The distribution network of the Swaziland Electricity Company consists of a number of components like distribution lines, wooden poles, insulators, conductors, pole mounted transformers and metering, to mention but a few. The most expensive of these components per unit cost is the pole mounted transformer and it takes longer to install. The distribution network is the less reliable compared to the transmission network, and its components fail regularly, and the pole mounted transformers are among the components which have a significant contribution to the failures. Some interventions in the past have been tried, but the failure rate of transformers has not reduced to reasonable limits, thereby compromising the reliability of the distribution network. This research therefore tries to identify the causes of the high failure rate of transformers, the pattern of failure, areas where the failures are prevalent and what can be done to counter the root causes. Statistical data on the transformer failures was gathered over a period of four years, where on a daily basis, the number of failed transformers, weather conditions and time of day when the failures occurred were recorded. The weather conditions and time of day when the failures occurred made it easier to identify the possible causes of the failures. For instance, if there was a thunderstorm in a certain area and there were some failed units which were discovered immediately after the storm, then the failures of those units was attributed to lightning. Similarly, when there were no thunderstorms, but a transformer has failed windings, the likely cause of failure is overloading. After replacing that failed unit, its load was monitored to confirm overload and then take a corrective action. It is a standard practice that a failed transformer is replaced and the failed one taken for repairs and then later kept as a spare at the Swaziland Electricity Company. Basic tests on the failed transformers and visual inspections were carried out to determine the nature of damage on the failed unit, for instance, winding damage and bushing damage. Footing resistances were also recorded at the transformer structures and if there was a need, they were reduced to 10 Ω or less using the crowfoot earthing method or a combination of the crowfoot earthing method and conductive cement.
  • Thumbnail Image
    Item
    Open Access
    Non intrusive load monitoring & identification for energy management system using computational intelligence approach
    (2015) Aladesanmi, Ereola Johnson; Folly, Komla A; Awodele, Kehinde
    Electrical energy is the life line to every nation’s or continent development and economic progress. Referable to the recent growth in the demand for electricity and shortage in production, it is indispensable to develop strategies for effective energy management and system delivery. Load monitoring such as intrusive load monitoring, non-intrusive load monitoring, and identification of domestic electrical appliances is proposed especially at the residential level since it is the major energy consumer. The intrusive load monitoring provides accurate results and would allow each individual appliance's energy consumption to be transmitted to a central hub. Nevertheless, there are many practical disadvantages to this method that have motivated the introduction of non-intrusive load monitoring system. The fiscal cost of manufacturing and installing enough monitoring devices to match the number of domestic appliances is considered to be a disadvantage. In addition, the installation of one meter per household appliances would lead to congestion in the house and thus cause inconvenience to the occupants of the house, therefore, non-intrusive load monitoring technique was developed to alleviate the aforementioned challenges of intrusive load monitoring. Non-intrusive load monitoring (NILM) is the process of disaggregating a household’s total energy consumption into its contributing appliances. The total household load is monitored via a single monitoring device such as smart meter (SM). NILM provides cost effective and convenient means of load monitoring and identification. Several nonintrusive load monitoring and identification techniques are reviewed. However, the literature lacks a comprehensive system that can identify appliances with small energy consumption, appliances with overlapping energy consumption and a group of appliance ranges at once. This has been the major setback to most of the adopted techniques. In this dissertation, we propose techniques that overcome these setbacks by combining artificial neural networks (ANN) with a developed algorithm to identify appliances ranges that contribute to the energy consumption within a given period of time usually an hour interval.
  • Thumbnail Image
    Item
    Open Access
    Optimal allocation of distributed generation for power loss reduction and voltage profile improvement
    (2016) Oluwole, Osaloni Oluwafunso; Awodele, Kehinde; Folly, Komla A
    Distributed generation (DG) integration in a distribution system has increased to high penetration levels. There is a need to improve technical benefits of DG integration by optimal allocation in a power system network. These benefits include electrical power losses reduction and voltage profile improvement. Optimal DG location and sizing in a power system distribution network with the aim of reducing system power losses and improving the voltage profile still remain a major problem. Though much research has been done on optimal DG location and sizing in a power system distribution network with the aim of reducing system power losses and improving the voltage profile, most of the existing works in the literature use several techniques such as computation, artificial intelligence and an analytical approach, but they still suffer from several drawbacks. As a result, much can still be done in coming up with new algorithms to improve the already existing ones so as to address this important issue more efficiently and effectively. The majority of the proposed algorithms emphasize real power losses only in their formulations. They ignore the reactive power losses which are the key to the operation of the power systems. Hence, there is an urgent need for an approach that will incorporate reactive power and voltage profile in the optimization process, such that the effect of high power losses and poor voltage profile can be mitigated. This research used Genetic Algorithm and Improved Particle Swarm Optimization (GA-IPSO) for optimal placement and sizing of DG for power loss reduction and improvement of voltage profile. GA-IPSO is used to optimize DG location and size while considering both real and reactive power losses. The real and reactive power as well as power loss sensitivity factors were utilized in identifying the candidate buses for DG allocation. The GA-IPSO algorithm was programmed in Matlab. This algorithm reduces the search space for the search process, increases its rate of convergence and also eliminates the possibility of being trapped in local minima. Also, the new approach will help in reducing power loss and improve the voltage profile via placement and sizing.
  • Thumbnail Image
    Item
    Open Access
    Power consumption and costing of crop sensing systems for monitoring common Western Cape crops growth
    (2022) Damilare, Dunmoye Isaac; Winberg, Simon; Awodele, Kehinde
    In recent years, agricultural practices have been influenced to an ever-increasing extent by Industry 4.0 trends. While there is much fear of what this type of industrialization implies, these fears are often misplaced: full automation, where robots take over, is just one form of industrialization and is not likely to happen on a wide scale for farming contexts anytime soon. However, there is a certainty of the increasingly widespread use of methods such as accurate and large-scale sensing, tracking of production and applying data science to circumvent problems of crops and livestock health problems. This project focuses on the investigation of power consumption and costs of the farm sensing system, which is designed around these new approaches to production to propose a cost-effective solution for monitoring and controlling agricultural production, providing a comprehensive analysis of the contextual complications, the mechanism needed to realize this system and the cost and the anticipated power consumption of the system, to deliver an advisory system for farmers. The power consumed by the temperature sensor, pH sensor, and soil moisture during different stages of growth were investigated. Among the considered crops, spinach growth monitoring consumed the least amount of power during monitoring. The highest amount of power was consumed during garlic growth monitoring. Considering the time of the crop growth, spinach took just two months to be matured and requires less monitoring. The cost for monitoring spinach was $ 34.38, $ 1.13, and $ 0.44 using nickel-cadmium, solar (PV), and electric grid, respectively. The overall cost and power consumed increased with the period of germination to maturity. The highest power consumed was by garlic which took up to six months. The highest energy was consumed by the carrot's pH sensor, followed by the onion's pH sensor. It shows how important the range of carrot and onions pH should be kept at 5.0 – 6.0 and 5.5 – 6.5 respectively throughout the growing process. Hence the soil pH is most prominent in carrots, onions, and fresh green pepper. The soil moisture is slightly more prominent than the temperature in garlic, onions, spinach, and carrots while the temperature is slightly more prominent than soil moisture in sweet corn and fresh green chili pepper. With the investigated power consumed by sensors monitoring crop growth, and the cost associated with the sources of energy considered, assistive technologies can be provided which assist farmers with the existing practices.
  • Thumbnail Image
    Item
    Open Access
    Protection-based Distributed Generation Penetration Limits on MV feeders - Using Machine Learning
    (2021) Nxumalo, Emmanuel; Awodele, Kehinde
    The rise of disruptive technologies and the rapid growth of innovative initiatives have led to a trend of decentralization, deregulation, and distribution of regulated/centralized services. As a result, there is an increasing number of requests for the connection of distributed generators to distribution networks and the need for power utilities to quickly assess the impacts of distributed generators (DGs) to keep up with these requests. Grid integration of DGs brings about protection issues. Current protection systems were not designed for bi-directional power flow, thus the protective devices in the network lose their ability to perform their main functions. To mitigate the impact of distributed generation (DG), some standards and policies constrain the number of DG that can be connected to the distribution network. The problem with these limits is that they are based only on overload and overvoltage, and do not adequately define the DG size/threshold before the occurrence of a protection issue (NRS 097-2-3). The other problem with distributed generation is the vast difference in the technology, location, size, connection sequence, and protection scheme requirements which results in future DG network planning inadequacies – The Network DG Planning Dilemma. To determine the amount of DG to connect to the network, a detailed analysis is required which often involves the use of a simulation tool such as DIgSILENT to model the entire network and perform load flow studies. Modelling networks on DIgSILENT is relatively easy for simple networks but becomes time-consuming for complex, large, and real networks. This brings about a limitation to this method, planning inadequacies, and longer connection approval periods. Thus, there is a need for a fast but accurate system-wide tool that can assess the amount of DG that can be connected to a network. This research aims to present a technique used for calculating protection-based DG penetration limits on MV networks and develop a model to determine medium voltage opportunity network maps. These maps indicate the maximum amount of DG that can be connected to a network without the need for major protection scheme changes in South Africa. The approach to determining protection-based penetration limits is based on supervised machine learning methods. The aim is to rely on protection features present in the distribution network data i.e. fault level, Inverse Definite Minimum Time (IDMT) curve, pick-up current settings, Time Multiplier Settings (TMS), calculated relay operating times and relay positions to see how the network responds at certain DG penetration levels (‘actual' relay operating times). The dataset represents carefully anonymized distribution networks with accepted protection philosophy applied. A supervised machine learning algorithm is applied after nontrivial data pre-processing through recommendation systems and shuffling. The planning dilemma is cast into three parts: the first part is an automated pattern classification (logistic regression for classification of protection miscoordination), the second part involves regression (predicting operating time after different levels of DG penetration), and the last part involves developing a recommendation system (where, when and how much photovoltaic (PV) DG will be connected). Gradient descent, which is an optimisation algorithm that iterates and finds optimal values of the parameters that correspond to the local or global minimum values of the cost function using calculus was used to measure the accuracy of each model's hypothesis function. The cost function (one half mean squared error) for the models that predict ‘actual' relay operating times before DG penetration, at 35%, 65%, and 75% DG penetration converged to values below 120, 20, 15, and 15 seconds2 , respectively, within the first 100 iterations. A high variance problem was observed (cross-validation error was high and training error was low) for the models that used all the network protection features as inputs. The cross-validation and training errors approached the desired performance of 0.3±0.1 for the models that had second-order polynomials added. A training accuracy of 91.30%, 73.91%, 82.61%, and a validation accuracy of 100%, 55.56%, 66.67% was achieved when classifying loss of coordination, loss of grading and desensitization, respectively. A high bias problem was observed (cross-validation error was high and training error was high) for the loss of grading classification (relay positions eliminated) model. When the models (horizontal network features) were applied to four MV distribution networks, loss of coordination was not predicted, the loss of grading model had one false positive and the de-sensitization model had one false negative. However, when the results were compared to the vertical analysis (comparing the operating times of upstream and downstream relays/reclosers), 28 points indicated a loss of coordination (2 at 35%, 1 at 65% and 25 at 75% DG penetration). Protection coordination reinforcements (against loss of grading and desensitization) were found to be a requirement for DG connections where the MV transformer circuit breaker TMS is between 0.5 and 1.1, and where the network fault level is between 650 and 800A. Distribution networks in affluent neighbourhoods similar to those around the Western CapeSomerset West area and Gauteng- Centurion area need to be reinforced to accommodate maximum DG penetration up to the limit of 75% of the After Diversity Maximum Demand (ADMD). For future work, the collection of more data points (results from detailed analytical studies on the impact of DG on MV feeders) to use as training data to solve the observed high variance problem is recommended. Also, modifying the model by adding upstream and downstream network features as inputs in the classification model to solve the high bias problem is recommended.
  • Thumbnail Image
    Item
    Open Access
    State Estimation in Active Distribution Systems: Comparison Between Weighted Least Squares and Extended Kalman Filter Algorithms
    (2022) Watitwa, Jeff Kimasere; Awodele, Kehinde
    Distributed Generation units (DGs) have been continuously deployed since the early 1990s in distribution power systems to mitigate greenhouse gas emissions and climate change. Therefore, distribution systems, which were designed initially as passive networks, with little monitoring, are evolving into active networks, forcing Distribution System Operators to implement newer control applications in Distribution Management Systems (DMS). Distribution System State Estimation (DSSE) is a crucial component in DMS. DSSE is the process of obtaining the nodal voltage magnitude and its respective phasor angle in real-time by utilizing available recorded measurements and the parameters from the network topology. The state estimator's outputs are then used to realize appropriate monitoring and control for the distribution networks. Potential applications include optimal voltage and VAR control, DERs dispatch, islanding operation, and fault detection and location. Transmission System State Estimation (TSSE) has been developed and applied since the late 1960s. However, TSSE methods are not directly transferable to distribution networks (DNs) since they differ in their design, topology and operation. DNs have distinctively shorter lines with higher Resistance to Reactance ratio, and single-phase and unbalanced threephase circuits and loads. Importantly, scarcity of measurements in DNs present tremendous challenges in DSSE. Despite these challenges, pioneering work in DSSE commenced in the 1990s. The motivation of this research is to extend this work by comparing the performance of the Weighted Least Square (WLS) and the Extended Kalman Filter (EKF) state estimation algorithms in active DNs. This dissertation develops and tests the performance of active distribution system state estimation WLS and EKF algorithms that consider the integration of DGs in standard IEEEbus test feeders. An important contribution of this dissertation consists in the validation of the theoretical state estimation findings via the use of real-life data. The ADRESCONCEPT project data were used to simulate real-time measurements, while the Particle Swarm Optimization algorithm was used to place the DGs and PMUs on the best possible nodes of the selected test system. This dissertation starts by reviewing the state-of-the-art in DSSE, and provides the measurement and process model of the WLS and the EKF algorithms. Then, it illustrates the analytical formulation of the two algorithms as a function of the input measurements. Finally, a case studies on modified IEEE-33 bus and IEEE-69 bus test feeders were carried out on MATLAB/OpenDSS software, and numerical evaluation and results are presented. The EKF algorithm out-performs the WLS algorithm with an average RMS error of 0.00020588 to 0.00025168. Similarly, EKF converges in 3 iterations, while WLS converges in 4. Preceding this dissertation some of the research findings were published and presented in the 2019 and 2020 Southern Africa Universities Power Engineering Conferences and the 2020 Power Africa Conference.