My research interests include Finite Element Method modelling, condition monitoring of transformers, Renewable Energy, Data Analysis and Mathematical modelling
Evaluating Wireless Network Technologies (3G, 4G, 5G) and Their Infrastructure: A Systematic Review, 2024
Wireless network technologies, including 3G, 4G, and 5G, are transforming telecommunications infr... more Wireless network technologies, including 3G, 4G, and 5G, are transforming telecommunications infrastructure globally. However, the adoption and effectiveness of these technologies vary significantly across regions and industries, posing unique challenges and opportunities for Small and Medium Enterprises (SMEs). Understanding the critical factors influencing network deployment and optimization in different contexts is essential for telecom companies and business leaders. This systematic review aims to evaluate the infrastructure, performance, and strategic implications of wireless network technologies (3G, 4G, and 5G) across multiple industries and geographic regions, providing insights for SMEs and telecom companies on adopting these technologies to enhance operational efficiency and competitiveness. A comprehensive search of academic databases, including Google Scholar, Web of Science, and SCOPUS, was conducted using keywords such as "wireless network," "3G," "4G," "5G," "evaluation," and "infrastructure." Studies were selected based on pre-established eligibility criteria, and a risk of bias assessment was performed using the Newcastle-Ottawa Scale. Statistical synthesis and sensitivity analyses were conducted to identify key trends and challenges. A total of 121 studies were included, with the majority focusing on 5G technology (42%) and its infrastructure. Key findings highlight the importance of network densification, high-speed connectivity, and lowlatency applications, particularly in urban regions. The analysis also revealed significant regional disparities in infrastructure deployment, with developing countries facing challenges in expanding coverage and integrating advanced technologies. Industry-specific customization of wireless networks is essential for sectors such as manufacturing, healthcare, and retail. Wireless network technologies present vast opportunities for SMEs, but their successful implementation requires addressing regional infrastructure gaps and tailoring solutions to industry-specific needs. Telecom companies must prioritize strategic investments in network densification, scalability, and security to fully leverage the benefits of 5G. The findings of this review provide actionable insights for business leaders and policymakers aiming to optimize wireless technology deployments for enhanced performance and competitiveness.
This article is an open access article distributed under the terms and conditions of the Creative... more This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY
Electric Transformers make up some of the most critical components of a power system. Their relia... more Electric Transformers make up some of the most critical components of a power system. Their reliableoperation is critical for an efficient power grid as well as economic viability for the transformer operators andowners. Power transformers, like other High Voltage (HV) electrical equipment experience aging andinsulation degradation due to chemical, mechanical and electrical forces during their operation. PartialDischargers (PD) are localized electrical discharges that develop within insulation systems of HV electricalequipment such as transformers. In transformers, PD occurs in different forms and various locations andinsulation both internal and external of the transformer. PD originates as small pulses but tends to increase insize and intensity which can result in complete insulation failure. Monitoring partial discharges has proven toprovide valuable information on the state of the insulation systems of the power transformer, allowingtransformer operators to make calculated decisions for the maintenance and life of plant plans. This systematicliterature review aims to systematically examine the use of machine learning techniques in classifying PD intransformers to present a complete indicator to allow for future research in the field. The systematic reviewsurveyed a total of 81 research literature published from 2010 to 2023 that fulfilled a specific methodologywhich was developed as part of this study. The results revealed that supervised learning has been the mostwidely used Artificial Intelligence (AI) algorithm utilized, primarily in the form of Support Vector Machine(SVM). Regarding PD, the survey revealed that most researchers tend to use numerous types of PD andcompare them to one another. Furthermore, the use of artificial PD defect models to simulate the occurrence ofPD is widely used versus the use of actual power transformers. Most of the literature tends to not specify thephysical characteristics of PD, such as the magnitude of PD, PD inception voltage and PD extinction voltage.
Induction motors are ever-present in various commercial and industrial processes owing to their h... more Induction motors are ever-present in various commercial and industrial processes owing to their high-power capacity, dependability, and low manufacturing costs. However, during their operation their susceptible to heat, mechanical stress, electrical stress, and corrosion. These stresses do not affect the operation of induction motor, however, in a long run it may develop into a major fault which can cause additional maintenance costs and unscheduled downtime, resulting in overall production loss, high financial loss, and sometimes serious human injuries. This work investigates the performance of an induction motor based on various faults conditions. A MATLAB/Simulink platform is used to develop an induction motor model. The performance of induction motor is evaluated on the three-phase line connected to it since implementing a fault in a line affects the performance induction motor. The faults that were tested were no fault condition and faults between AB, AC, BC, AG, CG, BG as well as ABCG. Training dataset was extracted from the sinusoidal waveforms of this various fault conditions. The data is given to the classification learner app in MATLAB for learning. Furthermore, this data set imported to workspace for the training purposes in the classification learner app. Different SVM learning algorithms are trained to deduce the highest learning algorithm in prediction time. The highest SVM learning algorithm according to the results is cubic and Fine Gaussian SVM learning algorithms which shows 100% accuracy on prediction time. The confusion matrix is then used to compare the true predicted and the false predicted classes.
2023 31st Southern African Universities Power Engineering Conference (SAUPEC), Jan 24, 2023
Induction motors are ever-present in various commercial and industrial processes owing to their h... more Induction motors are ever-present in various commercial and industrial processes owing to their high-power capacity, dependability, and low manufacturing costs. However, during their operation their susceptible to heat, mechanical stress, electrical stress, and corrosion. These stresses do not affect the operation of induction motor, however, in a long run it may develop into a major fault which can cause additional maintenance costs and unscheduled downtime, resulting in overall production loss, high financial loss, and sometimes serious human injuries. This work investigates the performance of an induction motor based on various faults conditions. A MATLAB/Simulink platform is used to develop an induction motor model. The performance of induction motor is evaluated on the three-phase line connected to it since implementing a fault in a line affects the performance induction motor. The faults that were tested were no fault condition and faults between AB, AC, BC, AG, CG, BG as well as ABCG. Training dataset was extracted from the sinusoidal waveforms of this various fault conditions. The data is given to the classification learner app in MATLAB for learning. Furthermore, this data set imported to workspace for the training purposes in the classification learner app. Different SVM learning algorithms are trained to deduce the highest learning algorithm in prediction time. The highest SVM learning algorithm according to the results is cubic and Fine Gaussian SVM learning algorithms which shows 100% accuracy on prediction time. The confusion matrix is then used to compare the true predicted and the false predicted classes.
In the Southern African Development Community (SADC), transmission lines facilitate power utiliti... more In the Southern African Development Community (SADC), transmission lines facilitate power utilities in distributing a considerable amount of electricity from generation stations to end users. Long conductors used in transmission lines can cover great distances. However, they are susceptible to various environmental and weather conditions which can have adverse effects such as power outages, and equipment malfunctions, posing a serious threat to the reliability of the system. Hence constant monitoring of the performance characteristics of this system is crucial. The main objective of this work is to propose an Artificial Intelligent (AI) based technique, which is the Artificial Neural Network (ANN) to detect and classify faults on a transmission line. In contrast to the most common approach viz. the protective relay system, a five-layer feed-forward-back propagation neural network architecture is proposed in this work to detect and classify faults on a 230kV transmission line system. A set of 12 fault conditions have been predefined viz. No fault, AB, AC, BC, ABC, AG, BG, CG, ABG, ACG, BCG and ABCG conditions. The results indicate that the proposed ANN approach with Levenberg- Marquardt (LM) algorithm, 5-Layers and TANSIG transfer function yield an output of 0.8927, 0.8882, 0.905 and 0.8938 in the training, validation, testing and overall accuracy respectively. To corroborate these results, a comparative study of the proposed network and other neural networks was also carried out.
2021 International Conference on Electrical, Computer and Energy Technologies (ICECET), 2021
The presence of harmonic currents due to the switching action of inverters has spread the realiza... more The presence of harmonic currents due to the switching action of inverters has spread the realization of the prospective rapid depletion of the transformer's intended lifetime owing to increased losses and temperature rise. With the prospect of reduced transformer service lifetime under harmonic conditions, there is a growing interest by manufacturers to explore design procedures that enables transformers to operate reliably under harmonic conditions. The current work examines K-rated transformers with regards to harmonic indexes in particularly with respect to K-Factor, Harmonic loss factor (HLF) and Factor K for derating the transformer. The transformer considered in this study has a rated 630kVA. Based on a supplied harmonic current spectrum, a comparative study is carried out to calculate the additional winding Eddy losses using the K-Factor and HLF methods. The study reveals that when delivering this harmonic spectrum, the investigated unit must be de-rated to 86.6% of its nominal power rating.
2021 Southern African Universities Power Engineering Conference/Robotics and Mechatronics/Pattern Recognition Association of South Africa (SAUPEC/RobMech/PRASA), 2021
The modern electricity society has seen an increase in the integration of renewable energy source... more The modern electricity society has seen an increase in the integration of renewable energy sources into the power grid. This has significantly improved the electricity supply and sustainability. However, there are technical challenges related to the power quality (PQ) analysis for grid integrated systems. In this paper, we propose a PQ detection technique using the discrete wavelet transform (DWT) and support vector machine (SVM). The DWT analysis technique is used to extract statistical features, which are used as input to train the SVM classifier. The parameters of the SVM are optimised using the harris hawks optimisation (HHO) algorithm. Various scenarios of cases which may affect the quality of network performance are investigated. These events include the voltage sag, voltage swell, notch, transient fault conditions and sudden load increment. The proposed method is validated using the modified practical Eskom network. The presented results show that the proposed scheme correctly classified different cases.
With the evolution of solar photovoltaics, the capacity of solar photovoltaic transformers is inc... more With the evolution of solar photovoltaics, the capacity of solar photovoltaic transformers is increasing. The winding conductor succumb to larger dimensions, to be able to withstand the extreme operational condition of solar photovoltaic application. The magnetic flux leakage responsible to produce Eddy currents in windings and other metallic components over and above that also increase. At the same time, the conventional method of estimating the winding Eddy losses result in significant errors when employed for large conductor dimensions at high harmonic orders.This paper conduct an investigation into the computation of the winding Eddy losses of a 300kVA solar photovoltaic transformer. The assumption made in the C57.110-2018 standard about the winding Eddy losses to be increasing in accordance with the non-sinusoidal load current with the exponent 2 is realized to be unsuitable for solar photovoltaic application. There is a shortfall in the C57.110-2018 standard method to take into consideration the tendency of a high- frequency alternating current flowing through only the outer surface of the winding conductor (skin effect) when the oil- immersed transformer is supplying non-sinusoidal current. The frequency response of the winding conductors under non- sinusoidal load current condition is examined by evaluating the dispersal of the magnetic flux leakage in the conductors, from which the skin effect and proximity effect are taken into account. In order to authenticate the soundness of the emendation method, a 2D Finite Element Method model of the transformer under study is developed for the computation of the winding Eddy losses. This method reflect on geometrical properties of the winding conductor and it is revealed some benefit for both the transformer manufacturer and the power utility.
The transformer cellulose insulating material is a key benchmark for evaluating the remnant trans... more The transformer cellulose insulating material is a key benchmark for evaluating the remnant transformer service lifetime. The decomposition of the cellulose paper may be assessed by using the degree of polymerization (DP) which necessitates that paper samples be extracted from the unit in- service. Whereas, extracting the cellulose paper is particularly complex, alternative indicators have been employed to indirectly divulge the level of DP in the cellulose paper. For case in a point, the measurement of the Furan compounds in the oil samples extracted from the unit in-service. Notwithstanding, the existing DP models have proven to be unreliable in the assessment of the DP value in the cellulose paper on account that for the same furan concentration, these models lead to varying values of the DP. This is attributed to the fact that the statistical survey of these models studies different transformers with varying blends of insulating oil and application. In this work, new formulae for estimating the DP value of cellulose insulation for transforms particularly in the South African electrical grid are formulated. Measured DP values are used as a benchmark for the new formulae and further compared with the exiting DP models. The results yield an error of estimate of less than 3% against the measured data. The study further evaluates the remnant life of the studied transformers and yields an error of estimate of 2% against the measured data.
In the recent multifaceted transformation dawn of low-carbon power generation markets and escalat... more In the recent multifaceted transformation dawn of low-carbon power generation markets and escalated growth of the renewable energy technologies in South Africa, a comprehensive analysis of the technical and economic performance of transformers has been under-explored as a result of insufficient knowledge in this study area. The Transformer Total Ownership Cost (TTOC) is an economic evaluation pre-planned to give the transformer purchasers and owners the intended service-lifetime costs basis of their transformer's investment. The procedure for evaluating the transformer TTOC depend upon the concept of service-lifetime loss evaluation of transformers. Meticulously, this techno-economic concept considers the arithmetic sum of the presented discounted value (PDV) of a unit kilowatt loss of the transformer during its intended service-lifetime. In a competitive bidding process, the TTOC is employed as a criterion to select the most suitable option among competing offers by transformer manufactures and hence to substantiate the purchasing decision of the units capable of achieving maximum productivity at minimum expense. This work provides transformers' service-lifetime loss evaluation method which fulfills some of the operational requirements that have been introduced from the recent deployment of solar photovoltaic plants into the modern South African energy mixed electrical network. The proposed method herein takes into account the intermittent nature of the solar PV, annualized energy cost and services losses. The results yield a TTOC optimized for the operational requirements of modern solar PV plant in a South African context.
This article presents an ultramodern modelling algorithm for predicting the remnant cellulose lif... more This article presents an ultramodern modelling algorithm for predicting the remnant cellulose life cycle for oil-submerged power transformers based on the adaptive neuro-fuzzy interference system (ANFIS). The polymer characteristics, degree of polymerization (DP), and 2-furaldehyde (2FAL) of 100 power transformers were measured and collated, which were apportioned into 70 training databanks and 30 as testing datasets. The remnant cellulose life cycle of the transformer was predicted using the proposed ANFIS model characterized by polymer characteristics, DP and 2FAL as inputs. The proposed approach returns 98.23% training and 99.86% testing reliability. The proposed model was applied to 10 transformer case studies in predicting their remnant cellulose life cycle. To corroborate the proposed ANFIS, a comparative study was carried out by employing existing approaches in predicting the remnant life cycle of the case studies, and significant error margins were observed. At large, the re...
Evaluating Wireless Network Technologies (3G, 4G, 5G) and Their Infrastructure: A Systematic Review, 2024
Wireless network technologies, including 3G, 4G, and 5G, are transforming telecommunications infr... more Wireless network technologies, including 3G, 4G, and 5G, are transforming telecommunications infrastructure globally. However, the adoption and effectiveness of these technologies vary significantly across regions and industries, posing unique challenges and opportunities for Small and Medium Enterprises (SMEs). Understanding the critical factors influencing network deployment and optimization in different contexts is essential for telecom companies and business leaders. This systematic review aims to evaluate the infrastructure, performance, and strategic implications of wireless network technologies (3G, 4G, and 5G) across multiple industries and geographic regions, providing insights for SMEs and telecom companies on adopting these technologies to enhance operational efficiency and competitiveness. A comprehensive search of academic databases, including Google Scholar, Web of Science, and SCOPUS, was conducted using keywords such as "wireless network," "3G," "4G," "5G," "evaluation," and "infrastructure." Studies were selected based on pre-established eligibility criteria, and a risk of bias assessment was performed using the Newcastle-Ottawa Scale. Statistical synthesis and sensitivity analyses were conducted to identify key trends and challenges. A total of 121 studies were included, with the majority focusing on 5G technology (42%) and its infrastructure. Key findings highlight the importance of network densification, high-speed connectivity, and lowlatency applications, particularly in urban regions. The analysis also revealed significant regional disparities in infrastructure deployment, with developing countries facing challenges in expanding coverage and integrating advanced technologies. Industry-specific customization of wireless networks is essential for sectors such as manufacturing, healthcare, and retail. Wireless network technologies present vast opportunities for SMEs, but their successful implementation requires addressing regional infrastructure gaps and tailoring solutions to industry-specific needs. Telecom companies must prioritize strategic investments in network densification, scalability, and security to fully leverage the benefits of 5G. The findings of this review provide actionable insights for business leaders and policymakers aiming to optimize wireless technology deployments for enhanced performance and competitiveness.
This article is an open access article distributed under the terms and conditions of the Creative... more This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY
Electric Transformers make up some of the most critical components of a power system. Their relia... more Electric Transformers make up some of the most critical components of a power system. Their reliableoperation is critical for an efficient power grid as well as economic viability for the transformer operators andowners. Power transformers, like other High Voltage (HV) electrical equipment experience aging andinsulation degradation due to chemical, mechanical and electrical forces during their operation. PartialDischargers (PD) are localized electrical discharges that develop within insulation systems of HV electricalequipment such as transformers. In transformers, PD occurs in different forms and various locations andinsulation both internal and external of the transformer. PD originates as small pulses but tends to increase insize and intensity which can result in complete insulation failure. Monitoring partial discharges has proven toprovide valuable information on the state of the insulation systems of the power transformer, allowingtransformer operators to make calculated decisions for the maintenance and life of plant plans. This systematicliterature review aims to systematically examine the use of machine learning techniques in classifying PD intransformers to present a complete indicator to allow for future research in the field. The systematic reviewsurveyed a total of 81 research literature published from 2010 to 2023 that fulfilled a specific methodologywhich was developed as part of this study. The results revealed that supervised learning has been the mostwidely used Artificial Intelligence (AI) algorithm utilized, primarily in the form of Support Vector Machine(SVM). Regarding PD, the survey revealed that most researchers tend to use numerous types of PD andcompare them to one another. Furthermore, the use of artificial PD defect models to simulate the occurrence ofPD is widely used versus the use of actual power transformers. Most of the literature tends to not specify thephysical characteristics of PD, such as the magnitude of PD, PD inception voltage and PD extinction voltage.
Induction motors are ever-present in various commercial and industrial processes owing to their h... more Induction motors are ever-present in various commercial and industrial processes owing to their high-power capacity, dependability, and low manufacturing costs. However, during their operation their susceptible to heat, mechanical stress, electrical stress, and corrosion. These stresses do not affect the operation of induction motor, however, in a long run it may develop into a major fault which can cause additional maintenance costs and unscheduled downtime, resulting in overall production loss, high financial loss, and sometimes serious human injuries. This work investigates the performance of an induction motor based on various faults conditions. A MATLAB/Simulink platform is used to develop an induction motor model. The performance of induction motor is evaluated on the three-phase line connected to it since implementing a fault in a line affects the performance induction motor. The faults that were tested were no fault condition and faults between AB, AC, BC, AG, CG, BG as well as ABCG. Training dataset was extracted from the sinusoidal waveforms of this various fault conditions. The data is given to the classification learner app in MATLAB for learning. Furthermore, this data set imported to workspace for the training purposes in the classification learner app. Different SVM learning algorithms are trained to deduce the highest learning algorithm in prediction time. The highest SVM learning algorithm according to the results is cubic and Fine Gaussian SVM learning algorithms which shows 100% accuracy on prediction time. The confusion matrix is then used to compare the true predicted and the false predicted classes.
2023 31st Southern African Universities Power Engineering Conference (SAUPEC), Jan 24, 2023
Induction motors are ever-present in various commercial and industrial processes owing to their h... more Induction motors are ever-present in various commercial and industrial processes owing to their high-power capacity, dependability, and low manufacturing costs. However, during their operation their susceptible to heat, mechanical stress, electrical stress, and corrosion. These stresses do not affect the operation of induction motor, however, in a long run it may develop into a major fault which can cause additional maintenance costs and unscheduled downtime, resulting in overall production loss, high financial loss, and sometimes serious human injuries. This work investigates the performance of an induction motor based on various faults conditions. A MATLAB/Simulink platform is used to develop an induction motor model. The performance of induction motor is evaluated on the three-phase line connected to it since implementing a fault in a line affects the performance induction motor. The faults that were tested were no fault condition and faults between AB, AC, BC, AG, CG, BG as well as ABCG. Training dataset was extracted from the sinusoidal waveforms of this various fault conditions. The data is given to the classification learner app in MATLAB for learning. Furthermore, this data set imported to workspace for the training purposes in the classification learner app. Different SVM learning algorithms are trained to deduce the highest learning algorithm in prediction time. The highest SVM learning algorithm according to the results is cubic and Fine Gaussian SVM learning algorithms which shows 100% accuracy on prediction time. The confusion matrix is then used to compare the true predicted and the false predicted classes.
In the Southern African Development Community (SADC), transmission lines facilitate power utiliti... more In the Southern African Development Community (SADC), transmission lines facilitate power utilities in distributing a considerable amount of electricity from generation stations to end users. Long conductors used in transmission lines can cover great distances. However, they are susceptible to various environmental and weather conditions which can have adverse effects such as power outages, and equipment malfunctions, posing a serious threat to the reliability of the system. Hence constant monitoring of the performance characteristics of this system is crucial. The main objective of this work is to propose an Artificial Intelligent (AI) based technique, which is the Artificial Neural Network (ANN) to detect and classify faults on a transmission line. In contrast to the most common approach viz. the protective relay system, a five-layer feed-forward-back propagation neural network architecture is proposed in this work to detect and classify faults on a 230kV transmission line system. A set of 12 fault conditions have been predefined viz. No fault, AB, AC, BC, ABC, AG, BG, CG, ABG, ACG, BCG and ABCG conditions. The results indicate that the proposed ANN approach with Levenberg- Marquardt (LM) algorithm, 5-Layers and TANSIG transfer function yield an output of 0.8927, 0.8882, 0.905 and 0.8938 in the training, validation, testing and overall accuracy respectively. To corroborate these results, a comparative study of the proposed network and other neural networks was also carried out.
2021 International Conference on Electrical, Computer and Energy Technologies (ICECET), 2021
The presence of harmonic currents due to the switching action of inverters has spread the realiza... more The presence of harmonic currents due to the switching action of inverters has spread the realization of the prospective rapid depletion of the transformer's intended lifetime owing to increased losses and temperature rise. With the prospect of reduced transformer service lifetime under harmonic conditions, there is a growing interest by manufacturers to explore design procedures that enables transformers to operate reliably under harmonic conditions. The current work examines K-rated transformers with regards to harmonic indexes in particularly with respect to K-Factor, Harmonic loss factor (HLF) and Factor K for derating the transformer. The transformer considered in this study has a rated 630kVA. Based on a supplied harmonic current spectrum, a comparative study is carried out to calculate the additional winding Eddy losses using the K-Factor and HLF methods. The study reveals that when delivering this harmonic spectrum, the investigated unit must be de-rated to 86.6% of its nominal power rating.
2021 Southern African Universities Power Engineering Conference/Robotics and Mechatronics/Pattern Recognition Association of South Africa (SAUPEC/RobMech/PRASA), 2021
The modern electricity society has seen an increase in the integration of renewable energy source... more The modern electricity society has seen an increase in the integration of renewable energy sources into the power grid. This has significantly improved the electricity supply and sustainability. However, there are technical challenges related to the power quality (PQ) analysis for grid integrated systems. In this paper, we propose a PQ detection technique using the discrete wavelet transform (DWT) and support vector machine (SVM). The DWT analysis technique is used to extract statistical features, which are used as input to train the SVM classifier. The parameters of the SVM are optimised using the harris hawks optimisation (HHO) algorithm. Various scenarios of cases which may affect the quality of network performance are investigated. These events include the voltage sag, voltage swell, notch, transient fault conditions and sudden load increment. The proposed method is validated using the modified practical Eskom network. The presented results show that the proposed scheme correctly classified different cases.
With the evolution of solar photovoltaics, the capacity of solar photovoltaic transformers is inc... more With the evolution of solar photovoltaics, the capacity of solar photovoltaic transformers is increasing. The winding conductor succumb to larger dimensions, to be able to withstand the extreme operational condition of solar photovoltaic application. The magnetic flux leakage responsible to produce Eddy currents in windings and other metallic components over and above that also increase. At the same time, the conventional method of estimating the winding Eddy losses result in significant errors when employed for large conductor dimensions at high harmonic orders.This paper conduct an investigation into the computation of the winding Eddy losses of a 300kVA solar photovoltaic transformer. The assumption made in the C57.110-2018 standard about the winding Eddy losses to be increasing in accordance with the non-sinusoidal load current with the exponent 2 is realized to be unsuitable for solar photovoltaic application. There is a shortfall in the C57.110-2018 standard method to take into consideration the tendency of a high- frequency alternating current flowing through only the outer surface of the winding conductor (skin effect) when the oil- immersed transformer is supplying non-sinusoidal current. The frequency response of the winding conductors under non- sinusoidal load current condition is examined by evaluating the dispersal of the magnetic flux leakage in the conductors, from which the skin effect and proximity effect are taken into account. In order to authenticate the soundness of the emendation method, a 2D Finite Element Method model of the transformer under study is developed for the computation of the winding Eddy losses. This method reflect on geometrical properties of the winding conductor and it is revealed some benefit for both the transformer manufacturer and the power utility.
The transformer cellulose insulating material is a key benchmark for evaluating the remnant trans... more The transformer cellulose insulating material is a key benchmark for evaluating the remnant transformer service lifetime. The decomposition of the cellulose paper may be assessed by using the degree of polymerization (DP) which necessitates that paper samples be extracted from the unit in- service. Whereas, extracting the cellulose paper is particularly complex, alternative indicators have been employed to indirectly divulge the level of DP in the cellulose paper. For case in a point, the measurement of the Furan compounds in the oil samples extracted from the unit in-service. Notwithstanding, the existing DP models have proven to be unreliable in the assessment of the DP value in the cellulose paper on account that for the same furan concentration, these models lead to varying values of the DP. This is attributed to the fact that the statistical survey of these models studies different transformers with varying blends of insulating oil and application. In this work, new formulae for estimating the DP value of cellulose insulation for transforms particularly in the South African electrical grid are formulated. Measured DP values are used as a benchmark for the new formulae and further compared with the exiting DP models. The results yield an error of estimate of less than 3% against the measured data. The study further evaluates the remnant life of the studied transformers and yields an error of estimate of 2% against the measured data.
In the recent multifaceted transformation dawn of low-carbon power generation markets and escalat... more In the recent multifaceted transformation dawn of low-carbon power generation markets and escalated growth of the renewable energy technologies in South Africa, a comprehensive analysis of the technical and economic performance of transformers has been under-explored as a result of insufficient knowledge in this study area. The Transformer Total Ownership Cost (TTOC) is an economic evaluation pre-planned to give the transformer purchasers and owners the intended service-lifetime costs basis of their transformer's investment. The procedure for evaluating the transformer TTOC depend upon the concept of service-lifetime loss evaluation of transformers. Meticulously, this techno-economic concept considers the arithmetic sum of the presented discounted value (PDV) of a unit kilowatt loss of the transformer during its intended service-lifetime. In a competitive bidding process, the TTOC is employed as a criterion to select the most suitable option among competing offers by transformer manufactures and hence to substantiate the purchasing decision of the units capable of achieving maximum productivity at minimum expense. This work provides transformers' service-lifetime loss evaluation method which fulfills some of the operational requirements that have been introduced from the recent deployment of solar photovoltaic plants into the modern South African energy mixed electrical network. The proposed method herein takes into account the intermittent nature of the solar PV, annualized energy cost and services losses. The results yield a TTOC optimized for the operational requirements of modern solar PV plant in a South African context.
This article presents an ultramodern modelling algorithm for predicting the remnant cellulose lif... more This article presents an ultramodern modelling algorithm for predicting the remnant cellulose life cycle for oil-submerged power transformers based on the adaptive neuro-fuzzy interference system (ANFIS). The polymer characteristics, degree of polymerization (DP), and 2-furaldehyde (2FAL) of 100 power transformers were measured and collated, which were apportioned into 70 training databanks and 30 as testing datasets. The remnant cellulose life cycle of the transformer was predicted using the proposed ANFIS model characterized by polymer characteristics, DP and 2FAL as inputs. The proposed approach returns 98.23% training and 99.86% testing reliability. The proposed model was applied to 10 transformer case studies in predicting their remnant cellulose life cycle. To corroborate the proposed ANFIS, a comparative study was carried out by employing existing approaches in predicting the remnant life cycle of the case studies, and significant error margins were observed. At large, the re...
Uploads
Papers by Bonginkosi Thango