Transformer Condition Assessment Research Articles

First Measurement Campaign by a Multi-Sensor Robot for the Lifecycle Monitoring of Transformers

Transformers are a very important asset in the electrical transmission grid, and they can suffer from destructive events—e.g., rare transformer fires. Unfortunately, destructive events often lead to a lack of data available for investigators during post-event forensics and failure analysis. This fact has motivated our design and implementation of a robotic multi-sensor platform and cloud backend solution for the lifecycle monitoring, inspection, diagnostics, and condition assessment of transformers. The robotic platform collects data from specific viewpoints around the transformer during operation and at specific relevant lifecycle milestones of the transformer (e.g., at the factory acceptance test) in an automated, repetitive, precise, and reliable manner. The acquired data are stored in the cloud backend, which also provides computing resources and data access to relevant in- and off-premises services (e.g., respectively, SCADA systems, and weather reports). In this paper, we present the results of our first measurement campaign to showcase the value of our solution for transformer lifecycle monitoring, for anomaly detection, and as a crucial tool for post-event forensics in the case of destructive events.

Integrated transformer condition assessment based on SOM neural network evidence cloud element model

Integrated transformer condition assessment based on SOM neural network evidence cloud element model

Insulation aging condition assessment of transformer in the visual domain based on SE-CNN

The aging of transformer oil–paper insulation (OPI) contributes significantly to transformer failure and reduced service life. Therefore, reliable and precise evaluation of the aging status of transformer insulation condition is crucial in averting transformer failure. This article proposes a novel deep learning-based aging condition assessment technique of transformers by combining FDS (Frequency domain spectroscopy) and squeeze-and-excitation-enabled convolutional neural network (SE-CNN). An intrinsic product-based transformation using a Gramian angular field is employed to dimensionally expand the 1-D FDS signals acquired from OPI samples into 2D images. Each feature extraction channel in the SE-CNN-based aging diagnosis framework is assigned a specific weight, enabling the model to concentrate on the most relevant features while discarding the rest. Furthermore, the effect of moisture variation on the model is also considered under three different conditions to ensure the consequence of compensation for the effect of moisture on FDS. A comparison with traditional machine learning models has also been made to ensure the viability of the proposed methodology. Additionally, the model is interpreted by creating explanation weight plots of the feature coefficients using a local interpretable model-agnostic explanations (LIME) method in Python. The findings demonstrate that the novel aging diagnosis framework visually assesses the transformer’s OPI system with the highest accuracy of 99.12% in the test set. Furthermore, it facilitates the automatic extraction of deep aging features, thus providing better generalization and convergence capability.

Development of fuzzy classifier for technical condition ranking of power transformer

The work aim is to develop a fuzzy classifier for technical condition ranking of power transformer under condition of vagueness and ambiguity diagnostic information. Methodology. The fuzzy classifier developing for technical condition ranking of power transformer was based on approach of using fuzzy set theory and optimization methods. The proposed approach for power transformer rank assessment by using a classifier was developed on the basis of Takagi-Sugeno fuzzy inference system. The input indicators choice is justified and their efficiency for classifier is evaluated by expert evaluation method. This makes it possible to formalize expert assessments regarding the development of power transformer defects. Results. The formalization of technical condition assessment of power transformer in knowledge base form, which implemented in expert system prototype for technical condition assessment, was carried out. The complex technical condition assessment for each functional unit of power transformer was determined based on expert evaluations with using the test and measurement parameters results. Originality. The considered approach to formalization of uncertainty regarding technical condition of power transformer allows building a deterministic decision-making scheme for further maintenance strategy, in which the ranking and decommissioning procedures for specific objects are implemented on the basis of objective criteria. Practical value. The proposed fuzzy classifier allows determination with a high probability degree of technical condition assessment of power transformer based on the test and measurement parameters results. Thus, an applied aspect of using the obtained scientific result is the possibility to objectively rank of power transformers park based on the identified possible defects and their development degree. This constitutes the prerequisites for determining the failure probability evaluation of power transformer at nearest observation period and emergency risk assessment in integrated electric power systems under power transformer failures.

Improved GWO-MCSVM algorithm based on nonlinear convergence factor and tent chaotic mapping and its application in transformer condition assessment

The continuous and reliable operation of the transformer is the basis to ensure the normal operation of the power system. Relevant departments collect multi-dimensional and multi-source heterogeneous parameter data during the operation, maintenance and repair of transformers. The effective information contained in the parameter data can directly reflect the current operating status of the transformer. On the basis of support vector machine and grey wolf algorithm, an improved GWO-MCSVM algorithm based on nonlinear convergence factor and Tent chaotic mapping is proposed. The algorithm parameters are optimized through training samples, and the results are evaluated and verified in the algorithm itself, so as to improve the accuracy of the status assessment results. Finally, the accuracy of assessment results of the algorithm proposed in this paper, existing genetic algorithms and particle swarm optimization algorithms are compared by evaluating multiple sets of measured samples. By comparison, the effectiveness of the algorithm proposed in this paper for transformer condition assessment has been verified.

Identification and Calibration Method of Deviation of Main Transformer Online Monitoring Date Groups

Dissolved gas in transformer oil (DGA) online monitoring data can reflect equipment insulation performance, which provides an important basis for transformer condition assessment. The accuracy of online monitoring data directly affects the correctness of transformer condition assessment, and the health condition of the online monitoring data device also affects the effectiveness and reasonableness of online monitoring data. With the accumulation of operation time, the DGA online monitoring device faces many problems, resulting in the overall deviation of this indicator data from the actual value, and the transformer online monitoring device cannot play its monitoring effect. To solve these problems, this paper firstly linearizes the data of the DGA online monitoring device by segmentation, extracts the characteristics of the line segment curve, constructs a segmented correlation mining model, and mines the correlation of different indicators of DGA. Then, according to the change of correlation strength and weakness between online monitoring data in time sequence, the data deviation is tracked. Finally, a multi‐indicator back propagation (BP) neural network algorithm optimized by the genetic simulated annealing algorithm is constructed to calibrate the deviated data. The analysis shows that the method can identify the population deviation of online monitoring data and calibrate the deviated data effectively. © 2023 Institute of Electrical Engineer of Japan and Wiley Periodicals LLC.

Reproducing Transformers’ Frequency Response from Finite Element Method (FEM) Simulation and Parameters Optimization

Frequency response analysis (FRA) is being employed worldwide as one of the main methods for the internal condition assessment of transformers due to its capability of detecting mechanical changes. Nonetheless, the objective interpretation of FRA measurements is still a challenge for the industry. This is mainly attributable to the lack of complete data from the same or similar units. A large database of FRA measurements can contribute to improving classification algorithms and lead to a more objective interpretation. Due to their destructive nature, mechanical deformations cannot be performed on real transformers to collect data from different scenarios. The use of simulation and laboratory transformer models is necessary. This research contribution is based on a new method using Finite Element Method simulation and a lumped element circuit to obtain FRA traces from a laboratory model at healthy and faulty states, along with an optimization method to improve capacitive parameters from estimated values. The results show that measured and simulated FRA traces are in good agreement. Furthermore, the faulty FRA traces were analyzed to obtain the characterization of faults based on the variation of the lumped element’s parameters. This supports the use of the proposed method in the generation of faulty frequency response traces and its further use in classifying and localizing faults in the transformer windings. The proposed approach is therefore tailored for generating a larger and unique database of FRA traces with industrial importance and academic significance.

Review of Transformer Health Index from the Perspective of Survivability and Condition Assessment

As a critical indicator for assessing the survivability and condition of transformers in a fleet, the transformer health index has attracted attention from both asset owners and international organizations like CIGRE and IEEE DEIS/PES. To provide a systematic and comprehensive review for further study or to guide transformer asset management, this paper summarizes the state-of-the-art of the transformer health index, from the early proposed weighted-score-sum approaches to the more recently proposed artificial intelligence algorithm-based methods. Firstly, different methods for determining the transformer health index are reviewed. Each of these is specified as belonging to a certain type on the basis of its formulation and composition schematic. Subsequently, the steps to determine each type of health index are summarized, and examples derived from literature are provided for further illustration. Comparisons are finally carried out in order to better understand the pros and cons of different types of transformer health index, and the future development trends for transformer health indexes are also discussed. This work can serve as a valuable reference for the survivability and condition assessment of transformers in the power industry.

Modelling the lumped parameter network of a power transformer using sweep frequency response analysis

ABSTRACT Accurate modelling of power transformers is beneficial for fault diagnosis and the health condition assessment of the power transformers. However, measuring the internal parameters of power transformers while they are in operation is not feasible due to the costs involved in interrupting the power supply and dismantling the components. To address this issue, this paper proposes a new method that uses artificial neural networks and genetic algorithms to synthesise the power transformer lumped parameter network based on the frequency response. The proposed methodology can be utilised to model power transformers and obtain lumped parameter values with minimal cost and time. A physically realisable transfer function derived using the sweep frequency response test data of a power transformer is the input of the proposed model. The combined approach of an artificial neural network and genetic algorithms generates the output as resistive, inductive, and capacitive components of the lumped parameter circuit of power transformers. Furthermore, the entire approach is explained using a case study and validated by means of error analysis. The proposed method is an easy-to-implement process to develop in a computer program that can be used effectively for accurate fault identification in power transformers and health assessments.

Transformer Winding Fault Classification and Condition Assessment Based on Random Forest Using FRA

At present, the condition assessment of transformer winding based on frequency response analysis (FRA) measurements demands skilled personnel. Despite many research efforts in the last decade, there is still no definitive methodology for the interpretation and condition assessment of transformer winding based on FRA results, and this is a major challenge for the industrial application of the FRA method. To overcome this challenge, this paper proposes a transformer condition assessment (TCA) algorithm, which is based on numerical indices, and a supervised machine learning technique to develop a method for the automatic interpretation of FRA results. For this purpose, random forest (RF) classifiers were developed for the first time to identify the condition of transformer winding and classify different faults in the transformer windings. Mainly, six common states of the transformer were classified in this research, i.e., healthy transformer, healthy transformer with saturated core, mechanically damaged winding, short-circuited winding, open-circuited winding, and repeatability issues. In this research, the data from 139 FRA measurements performed in more than 80 power transformers were used. The database belongs to the transformers having different ratings, sizes, designs, and manufacturers. The results reveal that the proposed TCA algorithm can effectively assess the transformer winding condition with up to 93% accuracy without much human intervention.

Comparative analysis of methanol generation mechanism in different oil-cellulose insulation based on ReaxFF MD

Currently, methanol is used as an aging indicator to evaluate the aging condition of transformer insulating paper. Nevertheless, the generation mechanism of methanol in vegetable oil-paper insulation is still unclear, which is not conducive to further research on methanol analysis. In this study, three types of oil-paper insulation models of palm fatty acid ester, FR3 natural ester, and mineral oil are established by reaction molecular dynamics (ReaxFF MD) simulation. Firstly, the pyrolysis products of the three oil-paper insulation systems are described. Then the methanol generation pathways are analyzed using the elemental tracing method. There are two identical elementary reactions for the production of methanol in the three types of oil-paper insulation systems. One is that the methyl produced by pyrolysis combines with the hydroxyl group separated from the cellulose to generate methanol, and the other is that the hydroxymethyl separated from the cellulose combines with free hydrogen ion to form methanol. Based on experimental and simulation results, it is demonstrated that vegetable oil produced more methanol than mineral oil as aging progressed. In addition, it is further proved that methanol is co-generated from insulating oil and cellulose-insulating paper during the pyrolysis process. In other words, the methanol produced by the oil-paper insulation system is much higher than that of a single paper/oil insulation. Furthermore, the mineral oil-paper system produced more methanol than the other two vegetable oil-paper systems are observed. Therefore, this study provides a theoretical basis for the employment of methanol in the condition assessment of transformers with different insulating oils.

Research on transformer condition assessment based on improved D-S evidence fusion theory

In view of the trend of the transformation of transformer maintenance mode to condition-based maintenance in power grid companies and the shortcomings of D-S evidence theory in transformer operation state assessment, a transformer operation state assessment method based on improved D-S evidence fusion theory is proposed. Pre-process the transformer state quantity data to obtain the corresponding weight of the state value of each state quantity index; The improved D-S theory is used to introduce the uncertainty, and the modified basic reliability allocation is obtained. Then, the judgment results of transformer operation state level are obtained by fusing all evidence bodies. Finally, the feasibility of this method is verified by the actual operation data of a power transformer.

State-of-Art Review on Chemical Indicators for Monitoring the Aging Status of Oil-Immersed Transformer Paper Insulation

Chemical compounds dissolved in insulating oil, as indicators can excellently monitor the paper aging condition, which has attracted increasing interest in areas of transformer condition monitoring and fault diagnosis. Because of their outstanding features, such as good correlation with the degree of polymerization of cellulose paper and the aid of non-destructive online monitoring, chemical indicators have been effectively used for transformer condition assessment. In this study, a comprehensive, in-depth insight into the indicators of the aging of insulating paper from aging characteristics, physico-chemical characteristics, shortcomings of various compounds, generation pathways and mechanisms, and monitoring technologies are provided. It is expected that these chemical indicators can provide better guidance for the evaluation of paper insulation performance and transformer aging. In addition, the latest research progress, as well as current challenges and future prospects are also outlined. This study provides a theoretical basis and reference for chemical indicators in the fields of microscopic formation mechanism, diffusion equilibrium phenomenon, and insulation aging state assessment.

Condition Assessment of Power Transformers through DGA Measurements Evaluation Using Adaptive Algorithms and Deep Learning

Condition assessment for critical infrastructure is a key factor for the wellbeing of the modern human. Especially for the electricity network, specific components such as oil-immersed power transformers need to be monitored for their operating condition. Classic approaches for the condition assessment of oil-immersed power transformers have been proposed in the past, such as the dissolved gases analysis and their respective concentration measurements for insulating oils. However, these approaches cannot always correctly (and in many cases not at all) classify the problems in power transformers. In the last two decades, novel approaches are implemented so as to address this problem, including artificial intelligence with neural networks being one form of algorithm. This paper focuses on the implementation of an adaptive number of layers and neural networks, aiming to increase the accuracy of the operating condition of oil-immersed power transformers. This paper also compares the use of various activation functions and different transfer functions other than the neural network implemented. The comparison incorporates the accuracy and total structure size of the neural network.

Condition assessment of transformers in wind farm based on modified one-dim residual neural network

The working environment of transformers in the wind farm is more complex than others, which brings the difference in condition assessment. Moreover, many condition assessment methods based on characteristics or machine learning have difficulty in recognition in cases of multiple transformers, conditions and measuring points. To assess conditions, this paper establishes a condition classification model of the transformer with a modified one-dim residual neural network and uses vibration signal, current and voltage as inputs. The built network mode has faster convergence speed and classification accuracy in transformer condition assessment and is more suitable for transformer condition assessment than the original one.

Synchronous Salient Poles Fault Localization by SFRA and Fault Diagram Method

Sweep Frequency Response Analysis (SFRA) is an extended and non-invasive technique used for transformer condition assessment. In the last years, its proven sensitivity and ease of testing have generated interest for rotating machine applications. Since SFRA measurements are curves plotted on Bode diagrams, the failures can be detected by comparison between a reference curve (performed at healthy conditions), and a curve obtained after an incident. However, the interpretation of results is a hard task because the usual way of interpreting the results still being by visual inspection, then, extensive knowledge and appropriate experience are required. To solve it, in previous publications, the authors developed a method called “Fault Diagram” to analyze SFRA results based on fitting the measurement with an equivalent three-variable circuit. This paper presents several improvements in the method based on changing the equivalent circuit, the optimization algorithm and the frequency range analyzed. Furthermore, the paper provides experimental results of the method's application in a 40 MVA synchronous machine pole. SFRA results presentation into a 3D plot provide information about the fault type (ground or inter-turn fault), location, number of faulty turns and fault resistance with more defined tendencies than in previous research.

Health Index for Power Transformer Condition Assessment: A Comparison of Three Different Techniques.

In practice, the condition state of Power Transformers (PT) is quantified by using Health Index (HI). This paperanalyzes and compares three different state-of-the-art algorithms to compute HI. The first one uses a Weighted SumModel (WSM), the second is based on a Fuzzy Inference System (FIS), and the third combines both techniques, i.e.,WSM and FIS. These three approaches are tested in a PT fleet composed of 30 units. Results show that eachapproach produces different HI values for the same PTs. Therefore, decision making regarding the PT fleet willdepend on the selected approach for HI calculation. This work proposes merging the knowledge involved in eachanalyzed approach by using a K-means clustering technique to overcome this drawback. This solution could help theasset manager to make adequate decisions regarding the maintenance scheduling of PT when there is uncertaintyabout the appropriate approach to be selected.

Health condition assessment of transformers based on cross message passing graph neural networks

A method is proposed to assess the health condition of transformers based on cross message passing graph neural networks (CMPGNNs) in this paper. In order to improve the accuracy of transformer health condition assessment, multiple indicators and their strong correlation are taken into account. The evaluation indicators are divided into four comprehensive state categories, and each category has several indicators. First, the correlation between indicators of a state category is extracted by the health index method and the importance of criteria is analyzed through the inter-criteria correlation (CRITIC) method, and the health index of comprehensive indicators is obtained. Then, a diagram of comprehensive indicators is constructed, and the correlation between indicators of different state categories is extracted. Finally, CMPGNN is constructed to achieve the health assessment. The experimental results show that the proposed method can improve the accuracy of transformer health condition assessment.

Failure mapping and critical measurements for the operating condition assessment of power transformers

The causes of failure of critical component of the electrical power distribution network like the power transformers, has always attracted the interest of the scientific and technical community, since the failures of electrical transformers almost always have a great impact on the consumers, as they can lead to a black out situation. A reliable and coherent maintenance schedule, that would predict the causes and prevent un-expectable shutdowns of power transformers, is of outmost importance nowadays, since electrification is expanding in terms of volume and power, as well as in terms of production sources, voltage and location. Such a reliable and coherent maintenance schedule depends on many factors and this article describes the parts of oil-immersed power transformers that are prone to failure. It also illustrates the relationship of the well known dissolved gasses that appear in the insulating oil as a result of specific operating faults, with specific measurements that should be done periodically, in order to prevent a major failure. Additionally, it describes the relationship between these dissolved gases with specific maintenance actions that should be implemented during the preventive maintenance process.

Health index for power transformer condition assessment based on operation history and test data

Transformers are key equipment in a power grid. The stability of the power system is depending more and more on the safety and reliability of transformers. Health Index (HI) is a practical tool that combines complex condition data into a single value as a comparative indication of the overall condition of a transformer. Most of the existing health index procedures are based on laboratory or on-site test data, with few considering the actual operating time. In this research, a comprehensive assessment method of transformer operating condition is proposed. Based on the HI model, this method the first considers the aging process of the transformer insulation as well as the load and operating environment to form the theoretical health index HI1. The aging process is characterized by the operating time of the transformer, the load rate, and the pollution level. The test health index HI2 is then formed based on on-site tests. The final health index HI is obtained with the combination of HI1 and HI2 to characterize the health condition of the transformer. To make sure the proposed method can be readily adapted in the field, only traditional parameters with well-established interpretations are required. This method has been applied to assess multiple 330kV transformers in service and provided a useful method to transformer asset management.