Incipient Faults In Power Transformers Research Articles

Detection of Incipient Faults in Power Transformers using Fuzzy Logic and Decision Tree Models Based on Dissolved Gas Analysis

This paper proposes an integrated approach utilizing Fuzzy Logic and Decision Tree algorithms to diagnose early-stage faults in power transformers based on Dissolved Gas Analysis (DGA) test results of transformer insulation oil. Overcoming limitations in conventional methods such as Duval Triangle, Key Gas Analysis, Rogers Ratio, IEC Ratio, and Doernenburg Ratio, our Fuzzy Logic and Decision Tree models address issues like inaccurate diagnosis, inconsistent diagnosis, lack of decisions or out-of-code results, and time-intensive manual calculations for large DGA datasets. The Decision Tree algorithm, a machine learning technique is applied to categorize faults into thermal and electrical types. Trained with over 300 DGA samples from transformers with known faults, the models exhibit robust performance during testing with different datasets. Notably, the Duval Triangle decision tree model attains the highest accuracy among the ten developed models, achieving a 98% accuracy rate when tested with 50 samples with known faults. Moreover, Decision Tree models for KGA, Doernenburg, Rogers, and IEC also demonstrate substantial prediction accuracy at 92%, 86%, 92%, and 90% respectively underscoring the efficacy of artificial intelligence methods over traditional approaches.

Graphical Shape in Cartesian Plane Based on Dissolved Gas Analysis for Power Transformer Incipient Faults Discrimination

Graphical Shape in Cartesian Plane Based on Dissolved Gas Analysis for Power Transformer Incipient Faults Discrimination

Power Transformer Fault Diagnosis Using Neural Network Optimization Techniques

Artificial Intelligence (AI) techniques are considered the most advanced approaches for diagnosing faults in power transformers. Dissolved Gas Analysis (DGA) is the conventional approach widely adopted for diagnosing incipient faults in power transformers. The IEC-599 standard Ratio Method is an accurate method that evaluates the DGA. All the classical approaches have limitations because they cannot diagnose all faults accurately. Precisely diagnosing defects in power transformers is a significant challenge due to their extensive quantity and dispersed placement within the power network. To deal with this concern and to improve the reliability and precision of fault diagnosis, different Artificial Intelligence techniques are presented. In this manuscript, an artificial neural network (ANN) is implemented to enhance the accuracy of the Rogers Ratio Method. On the other hand, it should be noted that the complexity of an ANN demands a large amount of storage and computing power. In order to address this issue, an optimization technique is implemented with the objective of maximizing the accuracy and minimizing the architectural complexity of an ANN. All the procedures are simulated using the MATLAB R2023a software. Firstly, the authors choose the most effective classification model by automatically training five classifiers in the Classification Learner app (CLA). After selecting the artificial neural network (ANN) as the sufficient classification model, we trained 30 ANNs with different parameters and determined the 5 models with the best accuracy. We then tested these five ANNs using the Experiment Manager app and ultimately selected the ANN with the best performance. The network structure is determined to consist of three layers, taking into consideration both diagnostic accuracy and computing efficiency. Ultimately, a (100-50-5) layered ANN was selected to optimize its hyperparameters. As a result, following the implementation of the optimization techniques, the suggested ANN exhibited a high level of accuracy, up to 90.7%. The conclusion of the proposed model indicates that the optimization of hyperparameters and the increase in the number of data samples enhance the accuracy while minimizing the complexity of the ANN. The optimized ANN is simulated and tested in MATLAB R2023a—Deep Network Designer, resulting in an accuracy of almost 90%. Moreover, compared to the Rogers Ratio Method, which exhibits an accuracy rate of just 63.3%, this approach successfully addresses the constraints associated with the conventional Rogers Ratio Method. So, the ANN has evolved a supremacy diagnostic method in the realm of power transformer fault diagnosis.

A Review of Methods for Diagnosing Incipient Faults in Power Transformers using Dissolved Gas Analysis

Power transformers play an important role in ensuring efficient and reliable transmission and distribution of electrical power. Their failure can lead to significant disruptions in power supply and safety concerns. Transformer incipient faults may occur due to various reasons, including electrical, mechanical or environmental factors. They develop slowly and can lead to serious damage if not diagnosed and corrected in time. In this review, various techniques for diagnosing this category of faults are highlighted before focusing on Dissolved Gas Analysis (DGA). Discussion on DGA centres on the traditional methods for fault diagnosis as well as various stand-alone and hybrid artificial intelligence techniques developed to improve the fault diagnostic ability of the traditional methods.

Accurate Identification of Transformer Faults From Dissolved Gas Data Using Recursive Feature Elimination Method

Dissolved gas analysis (DGA) of insulating oils is one of the most popular methods to detect incipient faults in power transformers. However, appropriate feature selection is crucial for accurately detecting incipient faults using DGA data. Another issue is the unavailability of a balanced DGA dataset, which can hamper the fault classification accuracy. Considering these two issues, this article proposes a novel and accurate fault classification framework using gas ratios as features obtained from the DGA data of power transformers. The obtained unbalanced DGA data was initially balanced using the synthetic minority oversampling technique (SMOTE) in the data pre-processing stage. Following this, an efficient feature selection algorithm, namely, recursive feature elimination (RFE) was used to select the best possible features prior to the fault classification using three benchmark machine learning (ML) classifiers, namely, <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${k}$ </tex-math></inline-formula> -nearest neighbor (KNN), multiclass support vector machines (SVMs), and extreme gradient boost (XGBoost). The proposed classification model was tested on the DGA data obtained from the local power utility and on the benchmark IEC TC-10 database. Investigations revealed that the proposed classification model delivered detection accuracy of 98.84% and 97.43%, respectively. The proposed method may be reliably used to diagnose incipient faults in power transformers.

Diagnosing Faults in Power Transformers With Variational Autoencoder, Genetic Programming, and Neural Network

This work presents a new approach for the diagnosis of incipient faults in power transformers by considering dissolved gas analysis (DGA). A multilayer perceptron (MLP) neural network was trained to diagnose the type of transformer fault. For training and testing of the classifier, data were used from in-service transformers obtained from the IEC TC 10 database and other data obtained from the literature. To address the imbalance of the data from the database adopted and thus improve the generalization power of the classifier, a data augmentation technique based on a variational autoencoder neural network was used. For the selection and extraction of characteristics from the inputs to the classifier, a technique based on genetic programming (GP) is proposed, which allows the creation of a new n-dimensional space of characteristics, providing a greater ability to increase interclass distances and intraclass compaction. For the performance analysis of the proposed classifier, comparisons were made using the classification results obtained through the IEC 60599 conventional fault diagnosis method and other trained MLPs without the use of data augmentation and the proposed characteristics extractor. The results obtained demonstrate the applicability of the proposed methodology for fault diagnosis, with the proposed system obtaining an accuracy of 95.18% in the test basis, which is higher than the results achieved by the other methods used to perform a comparison and analysis of results.

Real-Time Monitoring of Oil Temperature in Distribution Power Transformer by Using Internet of Things

In Malaysia, on-site technical personnel manually inspect power transformers. Some vital condition indicators, such as oil and winding temperatures, are not monitored in real-time. This condition can be hazardous if the transformer gets overheated. Overheating can cause mechanical deformation and insulation degradation if not monitored regularly. Thus, an online monitoring system that meets industry standards is needed to enhance power transformer monitoring and troubleshooting. In this research, the Internet of Things (IoT) based data acquisition (DAQ) system was deployed for real-time oil temperature monitoring and inspection to detect incipient faults in power transformers early. This IoT-based DAQ system was connected to the substation remote terminal unit (RTU) to update real-time data on each power transformer. The long-range (LoRa) technology is proposed for the system to transmit temperature, current, and voltage from the power transformers. The data transmission from the oil temperature indicator (OTI), network server, and database was monitored and compared. It is observed that the temperature data was transferred from the network server to the database without any transmission delay. The average deviation from the two experiments was 0.006 and 0.003, respectively, compared to the manual reading from the OTI scale meter with a digital reading by the proposed DAQ system. For testing purposes, the alert module in this system would notify technical personnel if the temperature exceeded +40°C in the power transformers. The proposed system can be used to assist with the upgrade and maintenance of the existing power transformer.

A Proposal for the Diagnosis of Incipient Faults in Power Transformers Using Fuzzy Logic Techniques

The availability of power transformers is essential for the safety and continuity of electrical service. Today's fault diagnosis methods use intelligent techniques such as neural networks, support machines, hybrid techniques, among others. Although they present good results, these techniques find restrictions in the ability to determine the precise moment in the event of multiple and small-magnitude faults. The proposal includes a new algorithm based on fuzzy rules that incorporates the daily increase of dissolved gases in the transformer oil that improves the classification of incipient faults. With reliable samples of gas dissolved in oil, the method proposed in the research can obtain a total precision rate of 91.4%. In contrast, this degree of precision is lower in other conventional methods reported in the bibliography. In addition, its performance in the classification of multiple failures is 97.5%. The method uses fuzzy logic tools to suggest actions aimed at preventive maintenance by monitoring the total of combustible gases dissolved in the oil. The proposal is a simple and easy solution to implement in practice that allows determining the status of the transformer in service without affecting the continuity of the electricity supply.

Induction of Decision Trees to Diagnose Incipient Faults in Power Transformers

This article describes the use of a decision tree (DT) approach based on computational intelligence (CI) for the analysis and diagnosis of incipient faults in power transformers. The method relies on measuring the combustible gas concentration in parts per million (ppm) from samples of the insulating oil. Currently, the Duval triangle method is one of the most traditional techniques used for the dissolved gas analysis (DGA), but it has limited accuracy. To overcome such conventional performance problems, CI techniques as artificial neural networks, fuzzy systems, and more recently DTs have been proposed as possible solutions. In this context, by using the gain ratio as a metric for attribute selection, this work demonstrates that the DT algorithm is capable of extracting as much information as possible from each class. It can also provide a solution for unsolved cases by using the traditional diagnosis method. Overall, it is reasonable to state that DT plays an important role in the improvement of DGA performance, this being a promising tool that can be combined with other traditional techniques. Another important aspect is that in the end of training the tree generates clear and ease-of-use rules. The proposed technique results in fast and accurate solutions for diagnosing faults in power transformers. It can be effectively implemented in corrective maintenance to avoid permanent burning and damage of equipment.

Reliable Estimation for Health Index of Transformer Oil Based on Novel Combined Predictive Maintenance Techniques

Transformer oil insulation condition may be deteriorated due to electrical and thermal faults, which may lead to transformer failure and system outage. In this regard, the first part of this paper presents comprehensive maintenance for power transformers aiming to diagnose transformer faults more accurately. Specifically, it aims to identify incipient faults in power transformers using what is known as dissolved gas analysis (DGA) with a new proposed integrated method. This proposed method for DGA is implemented based on the integration among the results of five different DGA techniques; 1) conditional probability, 2) clustering, 3) Duval triangle, 4) Roger’s four ratios refined, and 5) artificial neural network. Accordingly, this proposed integrated DGA method could improve the overall accuracy by 93.6% compared to the existing DGA techniques. In addition, the second part used for predictive maintenance is based on determining the health index for five new transformers and an aged power transformer (66/11 kV, 40 MVA) filled with NYTRO 10XN oil by evaluating the breakdown voltage, DGA, moisture content, and acidity for the oil. In the breakdown voltage test, two practical types of transformer oil; NYTRO 10XN and HyVolt III alongside their mixtures are estimated and compared. In addition, aged oil samples extracted from a real case study in-service transformer during operation with different aged durations; 9, 10, 11, 12, and 13 years, are tested for breakdown voltage, and then compared with fresh oil samples. For DGA, a temperature rise test is performed on the five new transformers with a comparison between dissolved gases before and after the temperature rise. In addition, winding resistance is measured after the temperature rise. Also, acidity and moisture are measured for oils extracted from the new five transformers and from the 13-year in-service transformer for studying their health index. The health index of the transformer insulation system is examined using only DGA and DGA plus breakdown voltage (BDV), moisture, and acidity. The results show that by using DGA plus BDV, moisture, and acidity, the health index provides reliable estimation results compared to using only DGA.

Incipient fault diagnosis in power transformers by data-driven models with over-sampled dataset

Early diagnosis of incipient faults in power transformers enables their predictive maintenance and guarantees their proper operation. Recently, machine learning (ML) techniques have played special role in fault diagnosis in power transformers; however, the application of such data-driven methods has been hampered by the lack of quality data to support their learning process. Since the collection of dissolved gas analysis (DGA) data depends on equipment failures, the obtaining of large labeled datasets that characterize incipient faults is a difficult task. The use of over-sampling techniques can overcome this challenge by providing a synthetic dataset with balanced classes for the ML method’s learning process. This paper addresses a novel application of a deep neural network classifier for the diagnosis of a dataset enriched by the Borderline synthetic minority over-sampling method. The performance of the model was compared with those of traditional DGA interpretation methods, traditional multilayer percetron networks (MLP) and a DNN working with the original dataset. The results indicate the superiority of the approach, and a noise-resilience analysis conducted revealed its ability to deal with corrupted data. The methodology is of simple implementation, highly accurate, and capable of correctly classifying over 84% of the test samples.

Application of Logistic Regression Algorithm in the Interpretation of Dissolved Gas Analysis for Power Transformers

Dissolved gas analysis (DGA) is one of the regular routine tests accepted by worldwide utilities to detect power transformer incipient faults. While the DGA measurement has fully matured since the development of offline and online sensors, interpretation of the DGA results still calls for advanced approaches to automate and standardize the process. Current industry practice relies on various interpretation techniques that are reported to be inconsistent and, in some cases, unreliable. This paper presents a new application for the advanced logistic regression algorithm to improve the reliability of the DGA interpretation process. In this regard, regularized logistic regression is used to improve the accuracy of the DGA interpretation process. Results reveal the superior features of the proposed logistic regression approach over the conventional and artificial intelligence techniques presented in the literature.

An Advanced Fuzzy Logic Based Method for Power Transformers Assessment

Dissolved Gas Analysis is an effective method for detecting faulty power transformers in their early stages. However, technical interpretation of results can be complex and highly dependent on the experience of experts. This paper presents an attempt to detect power transformer incipient fault via gas concentrations obtained from oil sampling and Dissolved Gas Analysis. The proposed method uses a sophisticated fuzzy logic system to perform fault type classification. Ratios and relative percentages of 5 key gases (Hydrogen, Methane, Ethane, Ethylene, and Acetylene) are taken as input variables, then the fuzzy system will try to generate an output vector that indicates six basic fault types, including partial, low, and high energy discharges as well as three ranges of thermal fault. This method can be easily implemented in any environment that supports basic mathematical operators. To demonstrate how the proposed fuzzy logic method works, the authors developed an offline MATLAB script and an online web-based application that can provide multiple assessments by various methods simultaneously. The set of membership functions and fuzzy rules presented in this paper allows the detection of multiple faults at once. Performance tests on many actual data sets show that the proposed method achieves better accuracy than the traditional ratio codes, even on a par with state-of-the-art graphical-based tools such as the Duval triangle or pentagon.

Fuzzy reinforcement learning based intelligent classifier for power transformer faults

In this work a fuzzy reinforcement learning (RL) based intelligent classifier for power transformer incipient faults is proposed. Fault classifiers proposed till date have low identification accuracy and do not identify all types of transformer faults. Herein, an attempt has been made to design an adaptive, intelligent transformer fault classifier that progressively learns to identify faults on-line with high accuracy for all fault types. In the proposed approach, dissolved gas analysis (DGA) data of oil samples collected from real power transformers (and from credible sources) has been used, which serves as input to a fuzzy RL based classifier. Typically, classification accuracy is heavily dependent on the number of input variables chosen. This has been resolved by using the J48 algorithm to select 8 most appropriate input variables from the 24 variables obtained using DGA. Proposed fuzzy RL approach achieves a fault identification accuracy of 99.7%, which is significantly higher than other contemporary soft computing based identifiers. Experimental results and comparison with other state-of-the-art approaches, highlights superiority and efficacy of the proposed fuzzy RL technique for transformer fault classification.

Novel prediction‐reliability based graphical DGA technique using multi‐layer perceptron network &amp; gas ratio combination algorithm

Dissolved gas analysis (DGA) is among the most essential techniques for diagnosis of incipient faults in power transformers. Here, a novel graphical DGA technique is proposed in which fault zones are distinguished based on certainty of prediction. The Duval Pentagon 1 and gas ratio combination methods are two most recent techniques with high prediction accuracy. In the Duval Pentagon 1, the rigidly separated distinct fault zones reduce the flexibility of analysis because the fault distributions themselves are not that strictly separated. This also prevents the full utilisation of the information available from the distribution patterns of the graphical representation. This problem has been addressed by overlapping individual fault zones and overlapping them using a multi-layer perceptron (MLP) network with fuzzy class boundaries. Then, in the regions, where multiple fault zones overlap, a modified gas ratio combination method is applied. Finally, a fuzzy decision-making system is developed for predicting the fault using information from both graphical distribution and gas ratios. The combined accuracy of the regions of certainty has been found exceptionally high (98.36%) compared to the regions of uncertainty (58.97%), whereas the overall prediction accuracy of the proposed technique is found comparatively higher (83%) than both the existing methods.

Power Transformer Fault Severity Estimation Based on Dissolved Gas Analysis and Energy of Fault Formation Technique

Decision making on transformer insulation condition based on the evaluated incipient faults and aging stresses has been the norm for many asset managers. Despite being the extensively applied methodology in power transformer incipient fault detection, solely dissolved gas analysis (DGA) techniques cannot quantify the detected fault severity. Fault severity is the core property in transformer maintenance rankings. This paper presents a fuzzy logic methodology in determining transformer faults and severity through use of energy of fault formation of the evolved gasses during transformer faulting event. Additionally, the energy of fault formation is a temperature-dependent factor for all the associated evolved gases. Instead of using the energy-weighted DGA, the calculated total energy of related incipient fault is used for severity determination. Severity of faults detected by fuzzy logic-based key gas method is evaluated through the use of collected data from several in-service and faulty transformers. DGA results of oil samples drawn from transformers of different specifications and age are used to validate the model. Model results show that correctly detecting fault type and its severity determination based on total energy released during faults can enhance decision-making in prioritizing maintenance of faulty transformers.

An integrated method of ANFIS and Dempster-Shafer theory for fault diagnosis of power transformer

Dissolved gas analysis (DGA) approach is extensively applied to detect incipient faults of power transformers. This paper presents a novel DGA method for power transformer incipient fault diagnosis based on integrated adaptive neuro fuzzy inference system (ANFIS) and Dempster-Shafer Theory (DST). Four out of seven common conventional methods which are studied and compared for better consistency and accuracy, are used to develop new ANFIS based fault diagnosis models. To promote fault diagnostic performance further and make fault decision process more reliable and reasonable, an improved DST is introduced to integrate outputs of each ANFIS based model, and to provide comprehensive and convincing fault diagnosis results. The fault diagnosis capability of the proposed method is validated by a reported fault dataset and 10-fold cross validation experiment. The performance of the proposed method is compared with conventional approaches and ANFIS based models which demonstrate that the proposed method is superior to other methods and is more effective and stable for power transformer fault diagnosis with high accuracy and consistency.

SOFC detector for portable gas chromatography: High-sensitivity detection of dissolved gases in transformer oil

Dissolved gas analysis (DGA) is widely used in diagnosing the incipient faults of power transformer. The detector is the key component for this method. Despite various sensors for DGA application, many of them suffer from poor stability and low sensitivity. A solid oxide fuel cell (SOFC) detector is designed in this paper to address these problems. The mechanism of the detector is briefly introduced, and it is fabricated in a mold. The test process is elucidated. For calibration purpose, two mathematical models of the detector, i.e., double-logarithmic model (DLM) and linearity model (LM), are compared with each other. Experimental results demonstrate that the SOFC detector exhibits good repeatability at given concentrations and excellent linearity with DLM. The results indicate that accurate and high-precision measurement of H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> , CH <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">4</sub> , CO, C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">4</sub> , C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">6</sub> , and C <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> can be achieved using the SOFC detector and the proposed DLM.

Study of approaches to incipient fault detection in power transformer by using dissolved gas analysis

To maintain the reliability of power transmission it is important to detect the incipient fault of power transformer as early as possible. If the fault of a power transformer is not detected promptly, it can evolve resulting in high repair costs or even failure of the power transformer and decreasing reliability of power transmission. The most commonly used method for power transformer fault detection is the dissolved gas analysis (DGA) of transformer oil. Various methods have been developed to interpret the data of dissolved gas analysis, but not many are applicable for the detection of the incipient fault. The detection of the incipient fault of a power transformer is included in both IEEE C57.104-2008 “Guide for the Interpretation of Gases Generated in Oil-Immersed Transformers” and Standard of Latvian Electrotechnical Committee LEK 118 “Transformer Oil Inspection Standards”. In both standards, the limits of dissolved gases in transformer oil are divided into levels, each corresponding to different technical conditions of the power transformer including the level that indicates the incipient fault. However, these approaches vary to a great degree – one approach mostly indicates that transformers are in good condition with several cases that must be additionally evaluated, but the second approach mostly results in warning about the incipient fault, which must be confirmed by additional evaluation. The objective of this paper is to determine the most suitable approach to detect the incipient fault of power transformers. A case study is provided, which includes analysis of DGA data of 48 power transformers installed in the transmission network in Latvia with both methodologies mentioned above for detecting the incipient fault.

Selection of Most Relevant Input Parameters Using Principle Component Analysis for Extreme Learning Machine Based Power Transformer Fault Diagnosis Model

The dissolved gas-in-oil analysis is a prevailing methodology being extensively utilized to diagnose incipient faults in oil-immersed power transformers. However distinct approaches have been implemented to find out dissolved gas analysis (DGA) results, they may sometimes fail to diagnose precisely. The incipient fault identification accuracy of various artificial intelligence (AI)-based methodology is assorted with change of input parameters. Thus, selection of input variable to an AI model is major research area. In this paper, principle component analysis algorithm using RapidMiner is applied to 360 experimental datasets, imitated in lab to identify most pertinent input variables for incipient fault classification. Thereafter, multi-class Extreme Learning Machine (ELM) technique is implemented to classify the incipient faults of power transformer and its performance is compared with artificial neural network, gene expression programming, fuzzy-logic, and support vector machine. The compared result shows that ELM provides better diagnosis results up to 100% accuracy at proposed input variable in short of time period which is helpful in on-line condition monitoring.