Transformer monitoring system and its testing

Transformer monitoring system and its testing | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Transformer monitoring system and its testing Sergey korobeynikov, Alexander Dvortcevoi, Irina Yakovina, Olesya Borush, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5888207/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 12 Aug, 2025 Read the published version in Electrical Engineering → Version 1 posted 12 You are reading this latest preprint version Abstract This study presents a transformer monitoring system designed for remote measurement of dissolved combustible gases, particularly hydrogen, as well as partial discharges (PD) and oil moisture content. The system was implemented on a 110 kV transformer in late December 2023. The behavior of oil moisture was analyzed, and based on the intensity of PD and the increase in combustible gas concentrations, an unscheduled chromatographic analysis was conducted. This analysis led to the decision to disconnect the transformer. Subsequent internal inspection revealed a defect in the input conductor of one of the phases. Partial discharge transformer Oil monitoring hydrogen water Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 1 Introduction Large oil-filled power transformers are critical and costly components of power systems across various voltage levels. Transformer oil, a key element of their insulation system, not only serves as an insulating medium but also acts as a diagnostic tool for predicting potential failures. Transformer malfunctions can result in widespread power outages, transformer explosions, fires, and significant financial losses [ 1 , 2 ]. Consequently, real-time monitoring, early defect detection, and timely intervention are essential for ensuring the stable operation of power systems [ 3 – 8 ]. The importance of timely diagnostics is further underscored by the aging infrastructure and the slow pace of transformer fleet renewal, which increases the risk of equipment failure. High-quality diagnostics and subsequent repairs can extend the residual service life of transformers and mitigate the risk of catastrophic failures [ 9 – 11 ]. Internal defects in power transformers are often characterized by increased gas formation, partial discharges, and changes in moisture content due to the degradation of paper insulation [ 12 – 15 ]. Overheating and partial discharges frequently result in the formation of combustible gases within the oil [ 16 – 18 ]. Specific gases are associated with particular fault types: partial discharges predominantly generate hydrogen and, to a lesser extent, methane; low-temperature overheating produces ethylene; and arcing generates acetylene. Each of these gases is flammable [ 19 – 21 ]. Monitoring the concentrations of hydrogen, combustible gases, and partial discharges is therefore a critical task for the early detection of defects and the prevention of transformer failures. This study aims to develop and test a comprehensive transformer monitoring system. 2 Monitoring systems Most transformer monitoring systems are based on the detection of partial discharges (PD). Recent advancements in PD diagnostics for oil-filled high-voltage transformers have leveraged machine learning (ML), deep learning (DL), artificial neural networks (ANN), support vector machines (SVM), decision trees, and other algorithms [ 22 – 28 ]. These methods enable efficient diagnostics of transformer conditions and the identification of insulation defects. For instance, ML algorithms such as K-nearest neighbors (KNN), Random Forest Classifier (RFC), Linear Discriminant Analysis (LDA), and Gaussian Naive Bayes (GNB) have been developed to classify five types of PD. Some studies have integrated ultrasonic sensors and antennas to localize PD occurrences with high accuracy, achieving localization errors as low as 1.5 cm in transformer tank models [ 29 , 30 ]. The influence of the layered structure of paper-oil insulation on PD detection has also been investigated [ 31 ]. Additionally, fuzzy logic has been employed to analyze transformer conditions by considering multiple factors [ 32 ]. In the past decade, significant attention has been devoted to modern diagnostic and monitoring methods for gases in oil transformers. These approaches often emphasize data processing rather than the analysis of oil-based processes. For example, hypersphere multi-class support vector machines (HMSVM) and improved Dempster-Shafer (D-S) theory have been used to diagnose transformer faults with an accuracy of up to 94.1% [ 33 ]. These methods rely on data collected from transformers and the application of ML algorithms to identify faults. Real-time dissolved gas analysis (DGA) has also proven crucial for fault detection. For instance, the installation of the Vaisala Optimus OPT100 model on a transformer in 2017 enabled the detection of a sharp increase in gas concentrations in October 2021, prompting the timely shutdown and replacement of the transformer [ 34 ]. Other studies have combined DGA measurements with the Degree of Polymerization (DP) and applied the Analytic Hierarchy Process (AHP) and Variable Weight Principle (VWP) to construct a Life Correlation Index (LCI) for transformers. This approach identified a transformer with an estimated residual service life of only 0.07 years, which subsequently experienced a short-circuit fault immediately after maintenance [ 35 ]. The use of Internet of Things (IoT)-based systems for real-time monitoring of dissolved gases has also been proposed. These systems employ artificial intelligence to calculate a transformer health index (HI) and detect potential issues promptly [ 36 ]. Digital twin-based monitoring systems have been developed to analyze parameters such as transformation ratio, angular displacement, relative capacitance, and grounding current. Probabilistic neural networks (PNN) trained on simulations of various operating modes are used to assess transformer conditions by comparing real-time data with digital models [ 37 , 38 ]. While many studies focus on physical or mathematical modeling of defects, real-world monitoring systems are less common. Notable examples include the systems described in [ 34 , 35 ], which successfully identified real defects, and the TDM (Transformer Diagnostics Monitor) system developed by DIMRUS [ 37 ]. The TDM system includes hardware and software for diagnostics and condition assessment, utilizing up to 80 sensors to measure parameters such as temperature, humidity, phase currents, and partial discharges. Despite the extensive research on transformer monitoring, a comprehensive analysis that accounts for the chemical and physical processes within transformers remains underdeveloped. Most monitoring systems rely on remote detection of PD, oil moisture, and dissolved gases. The monitoring system developed in this study focuses on these parameters, with an emphasis on hydrogen concentration, combustible gases, and oil moisture. 3 Our monitoring system The architecture of the proposed remote data collection and processing system is designed to monitor transformer conditions and detect anomalies. The system comprises the following components: 1. Measuring Equipment: The system measures partial discharges, concentrations of combustible gases (including hydrogen), and oil moisture. The minimum measurement parameters are: Apparent charge: 20 pC Combustible gas concentration: 0.002 ppm H₂O concentration: 0.002 ppm Partial discharges are recorded at intervals of 0.01 seconds, and transformer oil analysis cycles are conducted every 3 hours. The measurement error does not exceed ± 20%. 2. Local Information System: The iSMS system maintains a local archive of monitoring data. 3. Data Transmission Module: This module connects to the local control system and transmits data to a remote server via an encrypted channel. 4. Data Collection and Processing Server: The server processes and analyzes the collected data. It includes a data collection service, a historical data archive, and a data visualization service. 4. Monitoring results and discussion The monitoring system was installed on the 110 kV transformer in December 2023 and has been in continuous operation up until the present time (January 2025).The results are discussed in terms of oil moisture behaviour, combustible gas concentrations, and hydrogen levels. 4.1. Behavior of water Seasonal fluctuations in water content within transformer insulation systems are primarily influenced by the differing solubility of water in oil and paper, which constitute the main components of the electrical insulation system. The solubility of water in oil increases with rising temperature, whereas its solubility in paper decreases under the same conditions. Consequently, during the period from April to July, as ambient temperatures rise, water migrates partially from the paper into the oil, leading to an observable increase in water concentration in the oil. Conversely, from July to September, as temperatures decline, a portion of the water transitions back into the paper. These diffusion processes are consistent with established physical principles. This behaviour is illustrated in Fig. 2 , excluding abrupt concentration spikes. The data clearly indicate that as the average temperature increases, the moisture concentration in the oil also rises. Daily fluctuations in oil temperature and water content are even more pronounced, as shown in Fig. 3 . The temperature within the transformer tank exhibits a lag relative to changes in external temperature, resulting in a corresponding lag in the water concentration curve. For instance, the lowest external temperature is recorded at approximately 5 a.m., while the minimum water concentration in the oil is observed at 7 a.m. Anomalous behavior in water content is observed between 7 p.m. and 10 p.m. This deviation is likely attributable to increased transformer load during this period, which leads to elevated energy dissipation and a subsequent localized rise in the tank's internal temperature. To confirm this hypothesis, the installation of additional temperature sensors would be necessary for more precise measurements. However, the sharp fluctuations in water content recorded between mid-June and mid-August (see Fig. 7 ) cannot be explained solely by diffusion processes. A more detailed analysis of the water content measurement methodology is required. In the gas analyzer produced by INTERA JSC, water content is determined using the dielectric metric method. This technique relies on detecting subtle changes in the dielectric constant as water enters the sensor. The sensor operates within a flow chamber, where its capacitance changes due to the presence of medium components with high dielectric constants. A critical question arises: does the sensor respond uniformly to different states of water in oil? As reported earlier [ 40 , 41 ], water in oil can exist in several states, the most common being dissolved and emulsified. To address this, it is necessary to examine the changes in sensor capacitance under both conditions. This analysis highlights the need for further investigation into the measurement process and the potential influence of water's physical state on sensor readings. For water in the form of microdroplets, a well-known Maxwell expression can be applied, which is simplified taking into account that the dielectric constant of water is much greater than the dielectric constant of oil. $$\:\frac{{}_{\varvec{s}}}{{}_{\varvec{o}}}=1+3\varvec{\phi\:}$$ 1 Where \(\:{}_{\varvec{s}}\) is the dielectric constant of the medium (oil with droplets), \(\:{}_{\varvec{o}}\) is the dielectric constant of clean oil, φ, is the volume fraction of inclusions. For dissolved water, one can use the considerations of [ 25 ]. The total induction D in the medium can be written as D = ε 0 ε s E = ε 0 E + P o + ΔP = ε o E + ΔP=(ε 0 + \(\:\) )E For dissolved water, one can use the considerations of [ 25 ]. The total induction D in the medium can be written as... ΔP = \(\:\frac{{\varvec{p}}^{2}\varvec{n}}{3\varvec{k}\varvec{T}}\varvec{E}\) (2) Where k is the Boltzmann constant, T is the temperature. Hence the change in ε will be $$\:=\:\left({\varvec{p}}^{2}\:\varvec{n}\right)/3\varvec{k}\varvec{T}{}_{0}\:$$ 3 In this expression, p represents the dipole moment of a water molecule, with p = 1.8×10 − 30 K⋅m. To compare the given expressions, they must be reformulated into a common form. For this purpose, we express n in terms of ϕ. First, the average volume of a water molecule, v, is determined. Considering Avogadro's number (N = 6×10 23 ) and the fact that 18 cm 3 of water contains N molecules, the volume of a single molecule in cubic centimeters is given by v = 18/N. Consequently, the number of molecules can be expressed as n = ϕN/18. In the SI unit system, the concentration value must be scaled by a factor of 10 6 . Finally, expression (3) is reformulated into the form of expression (1): $$\:=\:\:\frac{{\varvec{p}}^{2}\varvec{N}\:{10}^{6}}{3\varvec{k}\varvec{T}{{}_{0}18}^{}}\:$$ 4 Next, we estimate the dimensionless coefficient associated with ϕ. Substituting numerical values, it follows that Δε ≈ ϕ. This indicates that water in the form of an emulsion increases the permittivity approximately three times more significantly than dissolved water. It should be noted, however, that there is a minor uncertainty related to the distinction between the external field and the local field acting on the water dipoles. In the case of a non-polar medium, such as transformer oil, this discrepancy does not exceed approximately 30%. To summarize this part of the discussion, dissolved water has a 2–3 times smaller effect on the permittivity compared to emulsified water. Furthermore, it is evident that the dynamics of the average concentration are governed by diffusion processes within the oil-paper system. What, then, causes the observed concentration jumps during measurement? To address this, the role of microdroplets must be evaluated. If a microdroplet is present in the oil, which is almost always the case, it may enter the measuring cell. It can be demonstrated that a change in capacitance on the order of several ppm occurs when a single droplet of approximately 100 µm in diameter, or several smaller droplets of equivalent total volume, enters the cell. It is also important to note that, in addition to water, other compounds formed in the oil during operation, such as aldehydes, ketones, carboxylic acids, and hydroxy acids [ 42 ], can have a similar effect on the dielectric constant. 4.2. Behavior of combustible gases and hydrogen The monitoring results indicate that the concentration of gases increases from April to July (Fig. 4 ). Following July, a decline in concentration is observed. It is important to highlight that the onset of this decline coincides with the shutdown of the transformer. What could explain this phenomenon? In our view, this is likely associated with the gradual leakage of gases from the transformer oil through the filter designed to facilitate the "breathing" process of the transformer. Since the transformer is not hermetically sealed, gases are able to escape freely from the transformer tank during this process. Simultaneously, a decrease in humidity is observed, as water is retained within the filter-dryer. It should be noted that the total concentration of combustible gases and the concentration of hydrogen are nearly identical (Fig. 5 ). This suggests that hydrogen is the primary gas being generated. As reported in the IEEE Guide for the Interpretation of Gases Generated in Oil-Immersed Transformers [ 43 ], hydrogen is predominantly formed during partial discharges. Therefore, the observed data directly indicate the occurrence of partial discharges. Measurements of partial discharges revealed that the apparent charge in phases A and B was approximately 1 nanocoulomb (nC), whereas in phase C, the apparent charge ranged from 6 to 8 nC (Fig. 6 ). For clarity, Fig. 6 simultaneously shows both the PD and hydrogen concentration signals. The most significant apparent discharges in phase C were recorded at the beginning of April, reaching approximately 15 nC. This level is considered critical for partial discharges. However, prior to this point, the hydrogen concentration remained negligible. Starting in April, the hydrogen concentration began to increase steadily. In response to this trend, an unplanned urgent chromatographic analysis of dissolved gases was conducted (Table 1 ). The analysis revealed that the hydrogen concentration exceeded the critical threshold of 0.01 vol.% by a factor of seven. This finding was a key factor in the decision to disconnect and inspect the transformer. Table 1 Extracts from the DGA protocol 02.07.2024 Gas Concentration, vol.% Limit value, vol.% Hydrogen 0.071 0.01 Methane 0.0001 0.01 Acetylene 0 0.0010 Ethylene 0.007 0.01 Ethane 0.0004 0.005 4.3. Decommissioning and transformer dissection results Figure 7 illustrates the condition of the conductor from the bushing to the phase C of the winding. It is evident that the upper section of the conductor is in good condition, with no visible damage to the insulation. In contrast, the lower section exhibits damaged insulation, accompanied by "fraying" of the multi-core wire. This damage leads to localized intensifications of the electric field around individual strands of the conductor, which act as sources of partial discharges (PD). An important question arises: why did the apparent charge of the partial discharge reach its maximum approximately three months prior to the transformer shutdown, subsequently stabilizing at a value approximately half of the maximum, while the gas content continued to increase during this period? In our view, the explanation is as follows. Initially, when the defect occurred, the partial discharge likely developed as an incomplete streamer discharge. In this scenario, the gas generated by the collapse of the streamer dissolves in the transformer oil. As the partial discharge progresses, the local gas concentration increases, with some of the gas remaining in the form of bubbles. Partial discharges occurring within these bubbles contribute to further gas formation; however, the apparent charge associated with such discharges is significantly lower compared to that of streamer discharges [44. Consequently, over time, the gas content continues to rise, while the apparent charge of the partial discharge remains relatively low. The decision to disconnect the transformer was made based on a significant (~ 7-fold) increase in hydrogen concentration relative to the permissible threshold. It is noteworthy that after cleaning and replacing the oil, as well as reconnecting the transformer to the load, no partial discharges, moisture, or combustible gases were detected, with all measured values remaining below the threshold limits. 5 Conclusions The transformer condition monitoring system, developed and implemented at an operational facility, included the measurement of the apparent charge of partial discharges in each phase, the concentration of dissolved combustible gases (with hydrogen measured separately), and the moisture content in the oil. The collected data were transmitted to a remote server for analysis via an encrypted communication channel. Monitorng over the course of a year revealed the onset and progression of adverse processes within the transformer, characterzed by an increase in the concentration of combustble gases, particularly hydrogen, as a result of high intensity partial discharges. Subsequent inspection of the transfomer confirmed these findings, identifying the source of the partial discharges as damage to the insulation and structural integrity of the phase C conductor. It can be argued that the correct use of monitoring results prevented the accident. Additionally, the monitoring system provided clear evidence of moisture diffusion processes within the transformer, driven by seasonal and daily temperature fluctuations, which caused moisture to migrate between the paper insulation and the oil. This study highlights the effectiveness of the monitoring system in detecting and diagnosing critical transformer defects, enabling timely maintenance actions to prevent further damage and ensure operational reliability. Declarations Competing interests The authors declare that we have no competing interests as defined by Springer, or other interests that might be perceived to influence the results and/or discussion reported in this paper. Ethical Approval The manuscript is not submitted to more than one journal for simultaneous consideration. The submitted work is original and has not been published elsewhere in any form or language (partially or in full), unless the new work concerns an expansion of the previous work. A single study is not split up into several parts to increase the quantity of submissions and submitted to various journals or to one journal over time (i.e., “salami-slicing/publishing”). Results are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation (including image-based manipulation). No data, text, or theories by others are presented as if they were the author’s own (“plagiarism”). Proper acknowledgments to other works are given. Funding No funding. Author Contribution S.K. wrote the main manuscript text, O.B. and V.S. prepared review of current monitoring systems, I.Ya. and A.D. organized a system for monitoring and analyzing data and wrote the corresponding part. All authors reviewed the manuscript. Acknowledgement The authors are grateful to EMA LLC for the equipment provided, Unisoft LLC for the software provided. Availability of data and materials All data and materials as well as software applications support our published claims and comply with field standards. References Fischer, M., Tenbohlen, S., Schafer, M., & Haug, R. (2011). Determining power transformers' sequence of service in power grids. IEEE Transactions on dielectrics and electrical insulation, 18(5), 1789-1798. https://doi.org/10.1109/TDEI.2011.6032851 Khaparde, S. A. (2017). Transformer engineering: design, technology, and diagnostics. Crc Press. Abbasi, A. R. (2022). Fault detection and diagnosis in power transformers: a comprehensive review and classification of publications and methods. Electric Power Systems Research, 209, 107990. https://doi.org/10.1016/j.epsr.2022.107990 Murugan, R., & Raju, R. (2021). Evaluation of in-service power transformer health condition for Inspection, Repair, and Replacement (IRR) maintenance planning in electric utilities. International Journal of System Assurance Engineering and Management, 12(2), 318-336. https://doi.org/10.1007/s13198-021-01083-1 Perepechenyi, V. (2023). Worn Power Transformers Influence on the Rural Electrical Networks Electrical Supply Reliability. Lighting Engineering & Power Engineering, 62(3), 71-78. https://doi.org/10.33042/2079-424X.2023.62.3.01 Tenbohlen, S., Coenen, S., Djamali, M., Müller, A., Samimi, M. H., & Siegel, M. (2016). Diagnostic measurements for power transformers. Energies, 9(5), 347. https://doi.org/10.3390/en9050347 Chakravorti, S., Dey, D., & Chatterjee, B. (2013). Recent trends in the condition monitoring of transformers. Power Systems. https://doi.org/10.1007/978-1-4471-5550-8 El-Bassiouny, A., El-Shimy, M., & Hammouda, R. (2017, December). Impact of power transformer failures on customer interruptions costs using customer damage functions. In 2017 Nineteenth International Middle East Power Systems Conference (MEPCON) (pp. 565-570). IEEE. https://doi.org/10.1109/MEPCON.2017.8301237 Christina, A., Salam, M. A., Rahman, Q. M., Wen, F., Ang, S. P., & Voon, W. (2018). Causes of transformer failures and diagnostic methods–A review. Renewable and Sustainable Energy Reviews, 82, 1442-1456. https://doi.org/10.1016/j.rser.2017.05.165 Tamma, W. R., Prasojo, R. A., & Suwarno. (2021). High voltage power transformer condition assessment considering the health index value and its decreasing rate. High voltage, 6(2), 314-327. https://doi.org/10.1049/hve2.12074 Foros, J., & Istad, M. (2020). Health index, risk and remaining lifetime estimation of power transformers. IEEE Transactions on Power Delivery, 35(6), 2612-2620. https://doi.org/10.1109/TPWRD.2020.2972976 Wang, M. V. A. J., Vandermaar, A. J., & Srivastava, K. D. (2002). Review of condition assessment of power transformers in service. IEEE Electrical insulation magazine, 18(6), 12-25. https://doi.org/10.1109/MEI.2002.1161455 Velásquez, R. M. A., & Lara, J. V. M. (2020). Root cause analysis improved with machine learning for failure analysis in power transformers. Engineering Failure Analysis, 115, 104684. https://doi.org/10.1016/j.engfailanal.2020.104684 Islam, M. M., Lee, G., & Hettiwatte, S. N. (2018). A review of condition monitoring techniques and diagnostic tests for lifetime estimation of power transformers. Electrical Engineering, 100, 581-605. https://doi.org/10.1007/s00202-017-0532-4 Soni, R., & Mehta, B. (2022). Evaluation of power transformer health analysis by internal fault criticalities to prevent premature failure using statistical data analytics approach. Engineering Failure Analysis, 136, 106213. https://doi.org/10.1016/j.engfailanal.2022.106213 Ivanov, G., Spasova, A., Mateev, V., & Marinova, I. (2023). Applied complex diagnostics and monitoring of special power transformers. Energies, 16(5), 2142. https://doi.org/10.3390/en16052142 Nanfak, A., Samuel, E., Fofana, I., Meghnefi, F., Ngaleu, M. G., & Hubert Kom, C. (2024). Traditional fault diagnosis methods for mineral oil‐immersed power transformer based on dissolved gas analysis: Past, present and future. IET Nanodielectrics. https://doi.org/10.1049/nde2.12082 Sun, C., Ohodnicki, P. R., & Stewart, E. M. (2017). Chemical sensing strategies for real-time monitoring of transformer oil: A review. IEEE Sensors Journal, 17(18), 5786-5806. https://doi.org/10.1109/JSEN.2017.2735193 Mharakurwa, E. T. (2022). In‐Service Power Transformer Life Time Prospects: Review and Prospects. Journal of Electrical and Computer Engineering, 2022(1), 9519032. https://doi.org/10.1155/2022/9519032 Aciu, A. M., Enache, S., & Nițu, M. C. (2024). A Reviewed Turn at of Methods for Determining the Type of Fault in Power Transformers Based on Dissolved Gas Analysis. Energies, 17(10), 2331. https://doi.org/10.3390/en17102331 Ali, M. S., Omar, A., Jaafar, A. S. A., & Mohamed, S. H. (2023). Conventional methods of dissolved gas analysis using oil-immersed power transformer for fault diagnosis: A review. Electric Power Systems Research, 216, 109064. https://doi.org/10.1016/j.epsr.2022.109064 Lu, S., Chai, H., Sahoo, A., & Phung, B. T. (2020). Condition monitoring based on partial discharge diagnostics using machine learning methods: A comprehensive state-of-the-art review. IEEE Transactions on Dielectrics and Electrical Insulation, 27(6), 1861-1888. https://doi.org/10.1109/TDEI.2020.009070 Kumar, K. R., & Haque, N. (2022, May). A Comparative Analysis of Different Machine Learning Algorithms for Classification of Partial Discharge Signals under HVDC. In 2022 3rd International Conference for Emerging Technology (INCET) (pp. 1-5). https://doi.org/IEEE. 10.1109/INCET54531.2022.9824127 Jariwala, J., Pradeep, L., Biswas, S., & Haque, N. (2023, June). A method for Localization of Single source Partial Discharges using Acoustic Sensors and Random Forest Algorithm. In 2023 5th International Conference on Energy, Power and Environment: Towards Flexible Green Energy Technologies (ICEPE) (pp. 1-5). https://doi.org/IEEE. 10.1109/ICEPE57949.2023.10201588 Sharma, A., & Shanker, T. B. (2023, October). Comparative performance analysis of machine learning algorithms in partial discharge source classification. In 2023 7th International Conference on Computer Applications in Electrical Engineering-Recent Advances (CERA) (pp. 1-6). IEEE. https://doi.org/10.1109/CERA59325.2023.10455529 Kumar, H., Shafiq, M., Hussain, G. A., & Kauhaniemi, K. (2021, October). Comparison of machine learning algorithms for classification of partial discharge signals in medium voltage components. In 2021 IEEE PES Innovative Smart Grid Technologies Europe (ISGT Europe) (pp. 1-6). IEEE. https://doi.org/10.1109/ISGTEurope52324.2021.9639923 Vigneshwaran, B., Iruthayarajan, M. W., & Maheswari, R. V. (2022). Enhanced particle swarm optimization-based convolution neural network hyperparameters tuning for transformer failure diagnosis under complex data sources. Electrical Engineering, 104(4), 2621-2636. https://doi.org/10.1007/s00202-022-01501-y Bezerra, F. E., Di Santo, S. G., Junior, G. M., Tiferes, R. R., da Rocha Albertini, A., & Santos, G. R. (2024). A forecasting methodology based on second-generation wavelets and neural networks: application to power transformer oil temperature. Electrical Engineering, 106(3), 3017-3029. https://doi.org/10.1007/s00202-023-02134-5 Sikorski, W. (2018). Active dielectric window: A new concept of combined acoustic emission and electromagnetic partial discharge detector for power transformers. Energies, 12(1), 115. https://doi.org/10.3390/en12010115 Ren, S., Hao, D., Zhao, X., He, Y., Yang, C., Xin, L., ... & Zhao, B. (2024). An ultrasonic signal propagation model of transformer partial discharge considering nonlinear attenuation in the temperature gradient field. Electrical Engineering, 106(1), 53-62. https://doi.org/10.1007/s00202-023-01965-6 Beura, C. P., Beltle, M., Tenbohlen, S., & Siegel, M. (2019). Quantitative analysis of the sensitivity of UHF sensor positions on a 420 kV power transformer based on electromagnetic simulation. Energies, 13(1), 3. https://doi.org/10.3390/en13010003 Atalar, F., Önhon, N. Ö., & Uğur, M. (2023). Investigation of the effect of layered structure on partial discharges in transformer pressboard insulator. Electrical Engineering, 105(6), 3459-3467. https://doi.org/10.1007/s00202-023-01903-6 Manoj, T., Ranga, C., & Mohan Rao, U. (2022). Multi-criterion fuzzy model for power transformer insulation monitoring. Electrical Engineering, 104(5), 3535-3546. https://doi.org/10.1007/s00202-022-01566-9 Leivo, S., Sattler, D., Duval, M., & Chiarella, C. (2023, June). A Case Study: Transformer Failure Detected with Real-time DGA Monitoring. In 2023 IEEE Electrical Insulation Conference (EIC) (pp. 1-4). IEEE. https://doi.org/10.1109/EIC55835.2023.10177343 Zhang, W., Jiang, J., Li, B., Shen, S., Lu, Y., Tao, Y., & Fan, L. (2023). Residual lifetime evaluation of power transformers based on data fusion and wiener model. IEEE Transactions on Power Delivery, 38(6), 4189-4201. https://doi.org/10.1109/TPWRD.2023.3305732 Hashemi, M. H., & Dikmen, S. (2023, May). Real-Time Condition Monitoring of Power Transformers Using the Internet of Things (IoT) and Dissolved Gas Analysis. In 2023 XIX International Scientific Technical Conference Alternating Current Electric Drives (ACED) (pp. 1-4). IEEE. https://doi.org/10.1109/ACED57798.2023.10143439 Zhang, J., Jiang, L., Zhang, H., Zhao, S., Lin, Y., & Chen, J. (2022, October). Research on On-line State Monitoring Method of Voltage Transformer based on Digital Twins. In 2022 4th International Conference on Artificial Intelligence and Advanced Manufacturing (AIAM) (pp. 243-248). IEEE. https://doi.org/10.1109/AIAM57466.2022.00053 Xiong, J., Ye, H., Pei, W., Kong, L., Huo, Q., & Han, Y. (2022). A monitoring and diagnostics method based on FPGA-digital twin for power electronic transformer. Electric Power Systems Research, 210, 108111. https://doi.org/10.1016/j.epsr.2022.108111 Rusov V.A. (2022) Partial discharges in power transformers, diagnostic equipment used, expected results. DIMRUS. (in Russian). https//dimrus.com/texts/trans_pd.html. Accessed 20 January 2025. Lyutikova, M. (2020). An improved electrochemical method for moisture determination in mineral oil. IEEE Transactions on Dielectrics and Electrical Insulation, 27(6), 2172-2178. https://doi.org/10.1109/TDEI.2020.009016 Lyutikova, M. N., Korobeynikov, S. M., Ridel, A. V., Vagin, D. V., & Konovalov, A. A. (2022). The breakdown voltage of moistened mineral oil during its cooling. IEEE Transactions on Dielectrics and Electrical Insulation, 29(6), 2181-2189. https://doi.org/10.1109/TDEI.2022.3203658 Li, Z., Shi, X., Feng, W., Luo, Z., & Zheng, Q. (2025). Study on Online Monitoring Sensor for Transformer Oil AN Based on Dielectric Constant Measurement. In Frontier Academic Forum of Electrical Engineering (pp. 575-582). Springer, Singapore. https://doi.org/10.1007/978-981-97-8832-3_58 IEEE-C57.104-–-2019.pdf Ovsyannikov, A. G., Korobeynikov, S. M., & Vagin, D. V. (2017). Simulation of apparent and true charges of partial discharges. IEEE Transactions on Dielectrics and Electrical Insulation, 24(6), 3687-3693. https://doi.org/10.1109/TDEI.2017.006635 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 12 Aug, 2025 Read the published version in Electrical Engineering → Version 1 posted Editorial decision: Revision requested 19 Mar, 2025 Reviews received at journal 18 Mar, 2025 Reviews received at journal 18 Mar, 2025 Reviews received at journal 14 Mar, 2025 Reviewers agreed at journal 19 Feb, 2025 Reviewers agreed at journal 18 Feb, 2025 Reviewers agreed at journal 18 Feb, 2025 Reviewers agreed at journal 18 Feb, 2025 Reviewers invited by journal 17 Feb, 2025 Editor assigned by journal 23 Jan, 2025 Submission checks completed at journal 23 Jan, 2025 First submitted to journal 23 Jan, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5888207","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":406230812,"identity":"0c23cfaa-c61f-4581-829d-832a4aa1e8ec","order_by":0,"name":"Sergey korobeynikov","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/ElEQVRIiWNgGAWjYBACCSBmBjHYeIDEB5K1MM6ACR8gRgsDUAszDzFaJNt7Hz4uqDicx8dz+Ji0TY1NvjkD7+HPH9sY8g1w6JPmOW5sPOPM4WI23rY06ZxjaZY7G/jSJA62MVhuwKFFTiKNTZq37XBiGz+PsXFuw2EDgwM8ZgxALQa4bIFo+QfVYgnRYvwBnxZpsJYGoBbeHsPHjBAtBhL4tEj2HGM2nnEsPbGN51jiw55jaQaWzTxmEmfOSRhI4tAicbyN8XFBjXXi/J7kAwd+1NgYmLP3GH+oKLMx4MOhBQqaEUwDSDRJ4FUPBHVIWgipHQWjYBSMghEHAMsHU2OHOVOiAAAAAElFTkSuQmCC","orcid":"","institution":"Novosibirsk State Technical University","correspondingAuthor":true,"prefix":"","firstName":"Sergey","middleName":"","lastName":"korobeynikov","suffix":""},{"id":406230813,"identity":"100344eb-af10-4ddd-a4f7-9584c263e16c","order_by":1,"name":"Alexander Dvortcevoi","email":"","orcid":"","institution":"Novosibirsk State Technical University","correspondingAuthor":false,"prefix":"","firstName":"Alexander","middleName":"","lastName":"Dvortcevoi","suffix":""},{"id":406230814,"identity":"2dc817e3-f854-410b-a201-57f62baea34b","order_by":2,"name":"Irina Yakovina","email":"","orcid":"","institution":"Novosibirsk State Technical University","correspondingAuthor":false,"prefix":"","firstName":"Irina","middleName":"","lastName":"Yakovina","suffix":""},{"id":406230815,"identity":"9746fd4d-7006-4c54-9c38-ee94138db3a6","order_by":3,"name":"Olesya Borush","email":"","orcid":"","institution":"Novosibirsk State Technical University","correspondingAuthor":false,"prefix":"","firstName":"Olesya","middleName":"","lastName":"Borush","suffix":""},{"id":406230816,"identity":"fb38c0c2-ef4f-47d1-b2fe-756cd4dab79f","order_by":4,"name":"Vladimir Shevchenko","email":"","orcid":"","institution":"Novosibirsk State Technical University","correspondingAuthor":false,"prefix":"","firstName":"Vladimir","middleName":"","lastName":"Shevchenko","suffix":""}],"badges":[],"createdAt":"2025-01-23 12:08:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5888207/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5888207/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s00202-025-03292-4","type":"published","date":"2025-08-12T15:57:55+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":74893791,"identity":"9fc0a9e1-c395-4ea1-a3c5-07dd19b6d70f","added_by":"auto","created_at":"2025-01-28 05:43:39","extension":"jpg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":32971,"visible":true,"origin":"","legend":"\u003cp\u003eSystem architecture\u003c/p\u003e","description":"","filename":"1.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5888207/v1/c295e1fb1789838ddaaea8fd.jpg"},{"id":74893788,"identity":"59fd1f4c-d672-4bf5-88cf-12bd9f16f090","added_by":"auto","created_at":"2025-01-28 05:43:39","extension":"jpg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":37074,"visible":true,"origin":"","legend":"\u003cp\u003eSeasonal variations in average external temperature and water content in oil (ppm).\u003c/p\u003e","description":"","filename":"2.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5888207/v1/c30d867cd7866e51f713fb12.jpg"},{"id":74894549,"identity":"8d89443e-e2cb-4002-9e5e-dccc33827295","added_by":"auto","created_at":"2025-01-28 05:59:39","extension":"jpg","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":32173,"visible":true,"origin":"","legend":"\u003cp\u003eTypical daily variations in external temperature (triangles) and measured water concentration (round), ppm.\u003c/p\u003e","description":"","filename":"3.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5888207/v1/7a017d7e1072d75656f432dc.jpg"},{"id":74894550,"identity":"7d688854-eaa8-47af-9c40-76faf56d550e","added_by":"auto","created_at":"2025-01-28 05:59:39","extension":"jpg","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":31511,"visible":true,"origin":"","legend":"\u003cp\u003eAverage concentrations of combustible gases and hydrogen.\u003c/p\u003e","description":"","filename":"4.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5888207/v1/bfe9ed5b5aaf4b386bfc2be2.jpg"},{"id":74893797,"identity":"22d8155a-9274-4f1f-abac-54cd73fa9c9d","added_by":"auto","created_at":"2025-01-28 05:43:39","extension":"jpg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":22971,"visible":true,"origin":"","legend":"\u003cp\u003eCorrelation between the concentrations of combustible gases and hydrogen.\u003c/p\u003e","description":"","filename":"5.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5888207/v1/6d042e03ce34fb7a6919de40.jpg"},{"id":74894552,"identity":"e38d61a5-24db-4196-901b-d69cb2d8044a","added_by":"auto","created_at":"2025-01-28 05:59:39","extension":"jpg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":36722,"visible":true,"origin":"","legend":"\u003cp\u003eApparent charge of the partial discharge (in pC) in phase C. Blue intermittent line is combustible gas (and hydrogen) concentration.\u003c/p\u003e","description":"","filename":"6.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5888207/v1/349d04fb8ed83eb5ba73820e.jpg"},{"id":74893804,"identity":"8578881c-80e1-420e-ad02-e890b8725bf2","added_by":"auto","created_at":"2025-01-28 05:43:39","extension":"jpg","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":28570,"visible":true,"origin":"","legend":"\u003cp\u003eFlexible connection. Phase C current supply.\u003c/p\u003e","description":"","filename":"7.jpg","url":"https://assets-eu.researchsquare.com/files/rs-5888207/v1/d1d6699b9f690f85ca10e106.jpg"},{"id":89310585,"identity":"5ea0b2bf-d105-4530-a757-8a010619ccff","added_by":"auto","created_at":"2025-08-18 16:08:31","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":813496,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5888207/v1/e5328ba1-f321-4d3c-9b85-7d20611c7ee7.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Transformer monitoring system and its testing","fulltext":[{"header":"1 Introduction","content":" \u003cp\u003eLarge oil-filled power transformers are critical and costly components of power systems across various voltage levels. Transformer oil, a key element of their insulation system, not only serves as an insulating medium but also acts as a diagnostic tool for predicting potential failures. Transformer malfunctions can result in widespread power outages, transformer explosions, fires, and significant financial losses [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Consequently, real-time monitoring, early defect detection, and timely intervention are essential for ensuring the stable operation of power systems [\u003cspan additionalcitationids=\"CR4 CR5 CR6 CR7\" citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe importance of timely diagnostics is further underscored by the aging infrastructure and the slow pace of transformer fleet renewal, which increases the risk of equipment failure. High-quality diagnostics and subsequent repairs can extend the residual service life of transformers and mitigate the risk of catastrophic failures [\u003cspan additionalcitationids=\"CR10\" citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eInternal defects in power transformers are often characterized by increased gas formation, partial discharges, and changes in moisture content due to the degradation of paper insulation [\u003cspan additionalcitationids=\"CR13 CR14\" citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. Overheating and partial discharges frequently result in the formation of combustible gases within the oil [\u003cspan additionalcitationids=\"CR17\" citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. Specific gases are associated with particular fault types: partial discharges predominantly generate hydrogen and, to a lesser extent, methane; low-temperature overheating produces ethylene; and arcing generates acetylene. Each of these gases is flammable [\u003cspan additionalcitationids=\"CR20\" citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eMonitoring the concentrations of hydrogen, combustible gases, and partial discharges is therefore a critical task for the early detection of defects and the prevention of transformer failures. This study aims to develop and test a comprehensive transformer monitoring system.\u003c/p\u003e"},{"header":"2 Monitoring systems","content":"\u003cp\u003eMost transformer monitoring systems are based on the detection of partial discharges (PD). Recent advancements in PD diagnostics for oil-filled high-voltage transformers have leveraged machine learning (ML), deep learning (DL), artificial neural networks (ANN), support vector machines (SVM), decision trees, and other algorithms [\u003cspan additionalcitationids=\"CR23 CR24 CR25 CR26 CR27\" citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. These methods enable efficient diagnostics of transformer conditions and the identification of insulation defects. For instance, ML algorithms such as K-nearest neighbors (KNN), Random Forest Classifier (RFC), Linear Discriminant Analysis (LDA), and Gaussian Naive Bayes (GNB) have been developed to classify five types of PD.\u003c/p\u003e \u003cp\u003eSome studies have integrated ultrasonic sensors and antennas to localize PD occurrences with high accuracy, achieving localization errors as low as 1.5 cm in transformer tank models [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. The influence of the layered structure of paper-oil insulation on PD detection has also been investigated [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. Additionally, fuzzy logic has been employed to analyze transformer conditions by considering multiple factors [\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eIn the past decade, significant attention has been devoted to modern diagnostic and monitoring methods for gases in oil transformers. These approaches often emphasize data processing rather than the analysis of oil-based processes. For example, hypersphere multi-class support vector machines (HMSVM) and improved Dempster-Shafer (D-S) theory have been used to diagnose transformer faults with an accuracy of up to 94.1% [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]. These methods rely on data collected from transformers and the application of ML algorithms to identify faults.\u003c/p\u003e \u003cp\u003eReal-time dissolved gas analysis (DGA) has also proven crucial for fault detection. For instance, the installation of the Vaisala Optimus OPT100 model on a transformer in 2017 enabled the detection of a sharp increase in gas concentrations in October 2021, prompting the timely shutdown and replacement of the transformer [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eOther studies have combined DGA measurements with the Degree of Polymerization (DP) and applied the Analytic Hierarchy Process (AHP) and Variable Weight Principle (VWP) to construct a Life Correlation Index (LCI) for transformers. This approach identified a transformer with an estimated residual service life of only 0.07 years, which subsequently experienced a short-circuit fault immediately after maintenance [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe use of Internet of Things (IoT)-based systems for real-time monitoring of dissolved gases has also been proposed. These systems employ artificial intelligence to calculate a transformer health index (HI) and detect potential issues promptly [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eDigital twin-based monitoring systems have been developed to analyze parameters such as transformation ratio, angular displacement, relative capacitance, and grounding current. Probabilistic neural networks (PNN) trained on simulations of various operating modes are used to assess transformer conditions by comparing real-time data with digital models [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e, \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eWhile many studies focus on physical or mathematical modeling of defects, real-world monitoring systems are less common. Notable examples include the systems described in [\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e], which successfully identified real defects, and the TDM (Transformer Diagnostics Monitor) system developed by DIMRUS [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]. The TDM system includes hardware and software for diagnostics and condition assessment, utilizing up to 80 sensors to measure parameters such as temperature, humidity, phase currents, and partial discharges.\u003c/p\u003e \u003cp\u003eDespite the extensive research on transformer monitoring, a comprehensive analysis that accounts for the chemical and physical processes within transformers remains underdeveloped. Most monitoring systems rely on remote detection of PD, oil moisture, and dissolved gases. The monitoring system developed in this study focuses on these parameters, with an emphasis on hydrogen concentration, combustible gases, and oil moisture.\u003c/p\u003e"},{"header":"3 Our monitoring system","content":"\u003cp\u003eThe architecture of the proposed remote data collection and processing system is designed to monitor transformer conditions and detect anomalies. The system comprises the following components:\u003c/p\u003e\n\u003ch3\u003e1. Measuring Equipment:\u003c/h3\u003e\n\u003cp\u003eThe system measures partial discharges, concentrations of combustible gases (including hydrogen), and oil moisture. The minimum measurement parameters are:\u003c/p\u003e \u003cp\u003eApparent charge: 20 pC\u003c/p\u003e \u003cp\u003eCombustible gas concentration: 0.002 ppm\u003c/p\u003e \u003cp\u003eH₂O concentration: 0.002 ppm\u003c/p\u003e \u003cp\u003ePartial discharges are recorded at intervals of 0.01 seconds, and transformer oil analysis cycles are conducted every 3 hours. The measurement error does not exceed\u0026thinsp;\u0026plusmn;\u0026thinsp;20%.\u003c/p\u003e\n\u003ch3\u003e2. Local Information System:\u003c/h3\u003e\n\u003cp\u003eThe iSMS system maintains a local archive of monitoring data.\u003c/p\u003e\n\u003ch3\u003e3. Data Transmission Module:\u003c/h3\u003e\n\u003cp\u003eThis module connects to the local control system and transmits data to a remote server via an encrypted channel.\u003c/p\u003e\n\u003ch3\u003e4. Data Collection and Processing Server:\u003c/h3\u003e\n\u003cp\u003eThe server processes and analyzes the collected data. It includes a data collection service, a historical data archive, and a data visualization service.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"4. Monitoring results and discussion","content":"\u003cp\u003eThe monitoring system was installed on the 110 kV transformer in December 2023 and has been in continuous operation up until the present time (January 2025).The results are discussed in terms of oil moisture behaviour, combustible gas concentrations, and hydrogen levels.\u003c/p\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e4.1. Behavior of water\u003c/h2\u003e \u003cp\u003eSeasonal fluctuations in water content within transformer insulation systems are primarily influenced by the differing solubility of water in oil and paper, which constitute the main components of the electrical insulation system. The solubility of water in oil increases with rising temperature, whereas its solubility in paper decreases under the same conditions. Consequently, during the period from April to July, as ambient temperatures rise, water migrates partially from the paper into the oil, leading to an observable increase in water concentration in the oil. Conversely, from July to September, as temperatures decline, a portion of the water transitions back into the paper. These diffusion processes are consistent with established physical principles. This behaviour is illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, excluding abrupt concentration spikes. The data clearly indicate that as the average temperature increases, the moisture concentration in the oil also rises.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eDaily fluctuations in oil temperature and water content are even more pronounced, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e. The temperature within the transformer tank exhibits a lag relative to changes in external temperature, resulting in a corresponding lag in the water concentration curve. For instance, the lowest external temperature is recorded at approximately 5 a.m., while the minimum water concentration in the oil is observed at 7 a.m.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAnomalous behavior in water content is observed between 7 p.m. and 10 p.m. This deviation is likely attributable to increased transformer load during this period, which leads to elevated energy dissipation and a subsequent localized rise in the tank's internal temperature. To confirm this hypothesis, the installation of additional temperature sensors would be necessary for more precise measurements.\u003c/p\u003e \u003cp\u003eHowever, the sharp fluctuations in water content recorded between mid-June and mid-August (see Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e) cannot be explained solely by diffusion processes. A more detailed analysis of the water content measurement methodology is required. In the gas analyzer produced by INTERA JSC, water content is determined using the dielectric metric method. This technique relies on detecting subtle changes in the dielectric constant as water enters the sensor. The sensor operates within a flow chamber, where its capacitance changes due to the presence of medium components with high dielectric constants.\u003c/p\u003e \u003cp\u003eA critical question arises: does the sensor respond uniformly to different states of water in oil? As reported earlier [\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e], water in oil can exist in several states, the most common being dissolved and emulsified. To address this, it is necessary to examine the changes in sensor capacitance under both conditions.\u003c/p\u003e \u003cp\u003eThis analysis highlights the need for further investigation into the measurement process and the potential influence of water's physical state on sensor readings.\u003c/p\u003e \u003cp\u003eFor water in the form of microdroplets, a well-known Maxwell expression can be applied, which is simplified taking into account that the dielectric constant of water is much greater than the dielectric constant of oil.\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:\\frac{{}_{\\varvec{s}}}{{}_{\\varvec{o}}}=1+3\\varvec{\\phi\\:}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eWhere \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{}_{\\varvec{s}}\\)\u003c/span\u003e\u003c/span\u003e is the dielectric constant of the medium (oil with droplets), \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{}_{\\varvec{o}}\\)\u003c/span\u003e\u003c/span\u003e is the dielectric constant of clean oil, φ, is the volume fraction of inclusions.\u003c/p\u003e \u003cp\u003eFor dissolved water, one can use the considerations of [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe total induction D in the medium can be written as\u003c/p\u003e \u003cp\u003eD\u0026thinsp;=\u0026thinsp;ε\u003csub\u003e0\u003c/sub\u003eε\u003csub\u003es\u003c/sub\u003eE\u0026thinsp;=\u0026thinsp;ε\u003csub\u003e0\u003c/sub\u003eE\u0026thinsp;+\u0026thinsp;P\u003csub\u003eo\u003c/sub\u003e\u0026thinsp;+\u0026thinsp;ΔP\u0026thinsp;=\u0026thinsp;ε\u003csub\u003eo\u003c/sub\u003eE\u0026thinsp;+\u0026thinsp;ΔP=(ε\u003csub\u003e0\u003c/sub\u003e+\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\)\u003c/span\u003e\u003c/span\u003e)E\u003c/p\u003e \u003cp\u003eFor dissolved water, one can use the considerations of [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThe total induction D in the medium can be written as...\u003c/p\u003e \u003cp\u003eΔP = \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\frac{{\\varvec{p}}^{2}\\varvec{n}}{3\\varvec{k}\\varvec{T}}\\varvec{E}\\)\u003c/span\u003e\u003c/span\u003e (2)\u003c/p\u003e \u003cp\u003eWhere k is the Boltzmann constant, T is the temperature. Hence the change in ε will be\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$$\\:=\\:\\left({\\varvec{p}}^{2}\\:\\varvec{n}\\right)/3\\varvec{k}\\varvec{T}{}_{0}\\:$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e3\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eIn this expression, p represents the dipole moment of a water molecule, with p\u0026thinsp;=\u0026thinsp;1.8\u0026times;10\u003csup\u003e\u0026minus;\u0026thinsp;30\u003c/sup\u003e K\u0026sdot;m. To compare the given expressions, they must be reformulated into a common form. For this purpose, we express n in terms of ϕ. First, the average volume of a water molecule, v, is determined. Considering Avogadro's number (N\u0026thinsp;=\u0026thinsp;6\u0026times;10\u003csup\u003e23\u003c/sup\u003e) and the fact that 18 cm\u003csup\u003e3\u003c/sup\u003e of water contains N molecules, the volume of a single molecule in cubic centimeters is given by v\u0026thinsp;=\u0026thinsp;18/N. Consequently, the number of molecules can be expressed as n\u0026thinsp;=\u0026thinsp;ϕN/18. In the SI unit system, the concentration value must be scaled by a factor of 10\u003csup\u003e6\u003c/sup\u003e. Finally, expression (3) is reformulated into the form of expression (1):\u003cdiv id=\"Equ3\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ3\" name=\"EquationSource\"\u003e\n$$\\:=\\:\\:\\frac{{\\varvec{p}}^{2}\\varvec{N}\\:{10}^{6}}{3\\varvec{k}\\varvec{T}{{}_{0}18}^{}}\\:$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e4\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e \u003cp\u003eNext, we estimate the dimensionless coefficient associated with ϕ. Substituting numerical values, it follows that Δε\u0026thinsp;\u0026asymp;\u0026thinsp;ϕ. This indicates that water in the form of an emulsion increases the permittivity approximately three times more significantly than dissolved water. It should be noted, however, that there is a minor uncertainty related to the distinction between the external field and the local field acting on the water dipoles. In the case of a non-polar medium, such as transformer oil, this discrepancy does not exceed approximately 30%.\u003c/p\u003e \u003cp\u003eTo summarize this part of the discussion, dissolved water has a 2\u0026ndash;3 times smaller effect on the permittivity compared to emulsified water. Furthermore, it is evident that the dynamics of the average concentration are governed by diffusion processes within the oil-paper system. What, then, causes the observed concentration jumps during measurement? To address this, the role of microdroplets must be evaluated. If a microdroplet is present in the oil, which is almost always the case, it may enter the measuring cell. It can be demonstrated that a change in capacitance on the order of several ppm occurs when a single droplet of approximately 100 \u0026micro;m in diameter, or several smaller droplets of equivalent total volume, enters the cell.\u003c/p\u003e \u003cp\u003eIt is also important to note that, in addition to water, other compounds formed in the oil during operation, such as aldehydes, ketones, carboxylic acids, and hydroxy acids [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e], can have a similar effect on the dielectric constant.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003e4.2. Behavior of combustible gases and hydrogen\u003c/h2\u003e \u003cp\u003eThe monitoring results indicate that the concentration of gases increases from April to July (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Following July, a decline in concentration is observed. It is important to highlight that the onset of this decline coincides with the shutdown of the transformer. What could explain this phenomenon? In our view, this is likely associated with the gradual leakage of gases from the transformer oil through the filter designed to facilitate the \"breathing\" process of the transformer. Since the transformer is not hermetically sealed, gases are able to escape freely from the transformer tank during this process. Simultaneously, a decrease in humidity is observed, as water is retained within the filter-dryer.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eIt should be noted that the total concentration of combustible gases and the concentration of hydrogen are nearly identical (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). This suggests that hydrogen is the primary gas being generated. As reported in the IEEE Guide for the Interpretation of Gases Generated in Oil-Immersed Transformers [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e], hydrogen is predominantly formed during partial discharges. Therefore, the observed data directly indicate the occurrence of partial discharges. Measurements of partial discharges revealed that the apparent charge in phases A and B was approximately 1 nanocoulomb (nC), whereas in phase C, the apparent charge ranged from 6 to 8 nC (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eFor clarity, Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e simultaneously shows both the PD and hydrogen concentration signals. The most significant apparent discharges in phase C were recorded at the beginning of April, reaching approximately 15 nC. This level is considered critical for partial discharges. However, prior to this point, the hydrogen concentration remained negligible. Starting in April, the hydrogen concentration began to increase steadily. In response to this trend, an unplanned urgent chromatographic analysis of dissolved gases was conducted (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The analysis revealed that the hydrogen concentration exceeded the critical threshold of 0.01 vol.% by a factor of seven. This finding was a key factor in the decision to disconnect and inspect the transformer.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eExtracts from the DGA protocol 02.07.2024\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"3\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGas\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eConcentration, vol.%\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eLimit value, vol.%\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHydrogen\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.071\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eMethane\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.0001\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAcetylene\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.0010\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEthylene\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.007\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.01\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eEthane\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.0004\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e \u003cp\u003e0.005\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003e4.3. Decommissioning and transformer dissection results\u003c/h2\u003e \u003cp\u003eFigure \u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003e illustrates the condition of the conductor from the bushing to the phase C of the winding. It is evident that the upper section of the conductor is in good condition, with no visible damage to the insulation. In contrast, the lower section exhibits damaged insulation, accompanied by \"fraying\" of the multi-core wire. This damage leads to localized intensifications of the electric field around individual strands of the conductor, which act as sources of partial discharges (PD).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAn important question arises: why did the apparent charge of the partial discharge reach its maximum approximately three months prior to the transformer shutdown, subsequently stabilizing at a value approximately half of the maximum, while the gas content continued to increase during this period? In our view, the explanation is as follows. Initially, when the defect occurred, the partial discharge likely developed as an incomplete streamer discharge. In this scenario, the gas generated by the collapse of the streamer dissolves in the transformer oil. As the partial discharge progresses, the local gas concentration increases, with some of the gas remaining in the form of bubbles. Partial discharges occurring within these bubbles contribute to further gas formation; however, the apparent charge associated with such discharges is significantly lower compared to that of streamer discharges [44. Consequently, over time, the gas content continues to rise, while the apparent charge of the partial discharge remains relatively low.\u003c/p\u003e \u003cp\u003eThe decision to disconnect the transformer was made based on a significant (~\u0026thinsp;7-fold) increase in hydrogen concentration relative to the permissible threshold. It is noteworthy that after cleaning and replacing the oil, as well as reconnecting the transformer to the load, no partial discharges, moisture, or combustible gases were detected, with all measured values remaining below the threshold limits.\u003c/p\u003e \u003c/div\u003e"},{"header":"5 Conclusions","content":"\u003cp\u003eThe transformer condition monitoring system, developed and implemented at an operational facility, included the measurement of the apparent charge of partial discharges in each phase, the concentration of dissolved combustible gases (with hydrogen measured separately), and the moisture content in the oil. The collected data were transmitted to a remote server for analysis via an encrypted communication channel.\u003c/p\u003e \u003cp\u003eMonitorng over the course of a year revealed the onset and progression of adverse processes within the transformer, characterzed by an increase in the concentration of combustble gases, particularly hydrogen, as a result of high intensity partial discharges. Subsequent inspection of the transfomer confirmed these findings, identifying the source of the partial discharges as damage to the insulation and structural integrity of the phase C conductor. It can be argued that the correct use of monitoring results prevented the accident.\u003c/p\u003e \u003cp\u003eAdditionally, the monitoring system provided clear evidence of moisture diffusion processes within the transformer, driven by seasonal and daily temperature fluctuations, which caused moisture to migrate between the paper insulation and the oil.\u003c/p\u003e \u003cp\u003eThis study highlights the effectiveness of the monitoring system in detecting and diagnosing critical transformer defects, enabling timely maintenance actions to prevent further damage and ensure operational reliability.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003cstrong\u003eCompeting interests\u003c/strong\u003e \u003cp\u003eThe authors declare that we have no competing interests as defined by Springer, or other interests that might be perceived to influence the results and/or discussion reported in this paper.\u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eEthical Approval\u003c/strong\u003e \u003cp\u003eThe manuscript is not submitted to more than one journal for simultaneous consideration. The submitted work is original and has not been published elsewhere in any form or language (partially or in full), unless the new work concerns an expansion of the previous work. A single study is not split up into several parts to increase the quantity of submissions and submitted to various journals or to one journal over time (i.e., \u0026ldquo;salami-slicing/publishing\u0026rdquo;). Results are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation (including image-based manipulation). No data, text, or theories by others are presented as if they were the author\u0026rsquo;s own (\u0026ldquo;plagiarism\u0026rdquo;). Proper acknowledgments to other works are given.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eNo funding.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eS.K. wrote the main manuscript text, O.B. and V.S. prepared review of current monitoring systems, I.Ya. and A.D. organized a system for monitoring and analyzing data and wrote the corresponding part. All authors reviewed the manuscript.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eThe authors are grateful to EMA LLC for the equipment provided, Unisoft LLC for the software provided.\u003c/p\u003e\u003ch2\u003eAvailability of data and materials\u003c/h2\u003e \u003cp\u003eAll data and materials as well as software applications support our published claims and comply with field standards.\u003c/p\u003e"},{"header":"References","content":"\u003col class=\"decimal_type\"\u003e\n\u003cli\u003eFischer, M., Tenbohlen, S., Schafer, M., \u0026amp; Haug, R. (2011). Determining power transformers\u0026apos; sequence of service in power grids. IEEE Transactions on dielectrics and electrical insulation, 18(5), 1789-1798. https://doi.org/10.1109/TDEI.2011.6032851\u003c/li\u003e\n\u003cli\u003eKhaparde, S. A. (2017). Transformer engineering: design, technology, and diagnostics. Crc Press.\u003c/li\u003e\n\u003cli\u003eAbbasi, A. R. (2022). Fault detection and diagnosis in power transformers: a comprehensive review and classification of publications and methods. Electric Power Systems Research, 209, 107990. https://doi.org/10.1016/j.epsr.2022.107990\u003c/li\u003e\n\u003cli\u003eMurugan, R., \u0026amp; Raju, R. (2021). Evaluation of in-service power transformer health condition for Inspection, Repair, and Replacement (IRR) maintenance planning in electric utilities. International Journal of System Assurance Engineering and Management, 12(2), 318-336. https://doi.org/10.1007/s13198-021-01083-1\u003c/li\u003e\n\u003cli\u003ePerepechenyi, V. (2023). Worn Power Transformers Influence on the Rural Electrical Networks Electrical Supply Reliability. Lighting Engineering \u0026amp; Power Engineering, 62(3), 71-78. https://doi.org/10.33042/2079-424X.2023.62.3.01\u003c/li\u003e\n\u003cli\u003eTenbohlen, S., Coenen, S., Djamali, M., M\u0026uuml;ller, A., Samimi, M. H., \u0026amp; Siegel, M. (2016). Diagnostic measurements for power transformers. Energies, 9(5), 347. https://doi.org/10.3390/en9050347\u003c/li\u003e\n\u003cli\u003eChakravorti, S., Dey, D., \u0026amp; Chatterjee, B. (2013). Recent trends in the condition monitoring of transformers. Power Systems. https://doi.org/10.1007/978-1-4471-5550-8\u003c/li\u003e\n\u003cli\u003eEl-Bassiouny, A., El-Shimy, M., \u0026amp; Hammouda, R. (2017, December). Impact of power transformer failures on customer interruptions costs using customer damage functions. In 2017 Nineteenth International Middle East Power Systems Conference (MEPCON) (pp. 565-570). IEEE. https://doi.org/10.1109/MEPCON.2017.8301237\u003c/li\u003e\n\u003cli\u003eChristina, A., Salam, M. A., Rahman, Q. M., Wen, F., Ang, S. P., \u0026amp; Voon, W. (2018). Causes of transformer failures and diagnostic methods\u0026ndash;A review. Renewable and Sustainable Energy Reviews, 82, 1442-1456. https://doi.org/10.1016/j.rser.2017.05.165\u003c/li\u003e\n\u003cli\u003eTamma, W. R., Prasojo, R. A., \u0026amp; Suwarno. (2021). High voltage power transformer condition assessment considering the health index value and its decreasing rate. High voltage, 6(2), 314-327. https://doi.org/10.1049/hve2.12074\u003c/li\u003e\n\u003cli\u003eForos, J., \u0026amp; Istad, M. (2020). Health index, risk and remaining lifetime estimation of power transformers. IEEE Transactions on Power Delivery, 35(6), 2612-2620. https://doi.org/10.1109/TPWRD.2020.2972976\u003c/li\u003e\n\u003cli\u003eWang, M. V. A. J., Vandermaar, A. J., \u0026amp; Srivastava, K. D. (2002). Review of condition assessment of power transformers in service. IEEE Electrical insulation magazine, 18(6), 12-25. https://doi.org/10.1109/MEI.2002.1161455\u003c/li\u003e\n\u003cli\u003eVel\u0026aacute;squez, R. M. A., \u0026amp; Lara, J. V. M. (2020). Root cause analysis improved with machine learning for failure analysis in power transformers. Engineering Failure Analysis, 115, 104684. https://doi.org/10.1016/j.engfailanal.2020.104684\u003c/li\u003e\n\u003cli\u003eIslam, M. M., Lee, G., \u0026amp; Hettiwatte, S. N. (2018). A review of condition monitoring techniques and diagnostic tests for lifetime estimation of power transformers. Electrical Engineering, 100, 581-605. https://doi.org/10.1007/s00202-017-0532-4\u003c/li\u003e\n\u003cli\u003eSoni, R., \u0026amp; Mehta, B. (2022). Evaluation of power transformer health analysis by internal fault criticalities to prevent premature failure using statistical data analytics approach. Engineering Failure Analysis, 136, 106213. https://doi.org/10.1016/j.engfailanal.2022.106213\u003c/li\u003e\n\u003cli\u003eIvanov, G., Spasova, A., Mateev, V., \u0026amp; Marinova, I. (2023). Applied complex diagnostics and monitoring of special power transformers. Energies, 16(5), 2142. https://doi.org/10.3390/en16052142\u003c/li\u003e\n\u003cli\u003eNanfak, A., Samuel, E., Fofana, I., Meghnefi, F., Ngaleu, M. G., \u0026amp; Hubert Kom, C. (2024). Traditional fault diagnosis methods for mineral oil‐immersed power transformer based on dissolved gas analysis: Past, present and future. IET Nanodielectrics. https://doi.org/10.1049/nde2.12082\u003c/li\u003e\n\u003cli\u003eSun, C., Ohodnicki, P. R., \u0026amp; Stewart, E. M. (2017). Chemical sensing strategies for real-time monitoring of transformer oil: A review. IEEE Sensors Journal, 17(18), 5786-5806. https://doi.org/10.1109/JSEN.2017.2735193\u003c/li\u003e\n\u003cli\u003eMharakurwa, E. T. (2022). In‐Service Power Transformer Life Time Prospects: Review and Prospects. Journal of Electrical and Computer Engineering, 2022(1), 9519032. https://doi.org/10.1155/2022/9519032\u003c/li\u003e\n\u003cli\u003eAciu, A. M., Enache, S., \u0026amp; Nițu, M. C. (2024). A Reviewed Turn at of Methods for Determining the Type of Fault in Power Transformers Based on Dissolved Gas Analysis. Energies, 17(10), 2331. https://doi.org/10.3390/en17102331\u003c/li\u003e\n\u003cli\u003eAli, M. S., Omar, A., Jaafar, A. S. A., \u0026amp; Mohamed, S. H. (2023). Conventional methods of dissolved gas analysis using oil-immersed power transformer for fault diagnosis: A review. Electric Power Systems Research, 216, 109064. https://doi.org/10.1016/j.epsr.2022.109064\u003c/li\u003e\n\u003cli\u003eLu, S., Chai, H., Sahoo, A., \u0026amp; Phung, B. T. (2020). Condition monitoring based on partial discharge diagnostics using machine learning methods: A comprehensive state-of-the-art review. IEEE Transactions on Dielectrics and Electrical Insulation, 27(6), 1861-1888. https://doi.org/10.1109/TDEI.2020.009070\u003c/li\u003e\n\u003cli\u003eKumar, K. R., \u0026amp; Haque, N. (2022, May). A Comparative Analysis of Different Machine Learning Algorithms for Classification of Partial Discharge Signals under HVDC. In 2022 3rd International Conference for Emerging Technology (INCET) (pp. 1-5). https://doi.org/IEEE. 10.1109/INCET54531.2022.9824127\u003c/li\u003e\n\u003cli\u003eJariwala, J., Pradeep, L., Biswas, S., \u0026amp; Haque, N. (2023, June). A method for Localization of Single source Partial Discharges using Acoustic Sensors and Random Forest Algorithm. In 2023 5th International Conference on Energy, Power and Environment: Towards Flexible Green Energy Technologies (ICEPE) (pp. 1-5). https://doi.org/IEEE. 10.1109/ICEPE57949.2023.10201588\u003c/li\u003e\n\u003cli\u003eSharma, A., \u0026amp; Shanker, T. B. (2023, October). Comparative performance analysis of machine learning algorithms in partial discharge source classification. In 2023 7th International Conference on Computer Applications in Electrical Engineering-Recent Advances (CERA) (pp. 1-6). IEEE. https://doi.org/10.1109/CERA59325.2023.10455529\u003c/li\u003e\n\u003cli\u003eKumar, H., Shafiq, M., Hussain, G. A., \u0026amp; Kauhaniemi, K. (2021, October). Comparison of machine learning algorithms for classification of partial discharge signals in medium voltage components. In 2021 IEEE PES Innovative Smart Grid Technologies Europe (ISGT Europe) (pp. 1-6). IEEE. https://doi.org/10.1109/ISGTEurope52324.2021.9639923\u003c/li\u003e\n\u003cli\u003eVigneshwaran, B., Iruthayarajan, M. W., \u0026amp; Maheswari, R. V. (2022). Enhanced particle swarm optimization-based convolution neural network hyperparameters tuning for transformer failure diagnosis under complex data sources. Electrical Engineering, 104(4), 2621-2636. https://doi.org/10.1007/s00202-022-01501-y\u003c/li\u003e\n\u003cli\u003eBezerra, F. E., Di Santo, S. G., Junior, G. M., Tiferes, R. R., da Rocha Albertini, A., \u0026amp; Santos, G. R. (2024). A forecasting methodology based on second-generation wavelets and neural networks: application to power transformer oil temperature. Electrical Engineering, 106(3), 3017-3029. https://doi.org/10.1007/s00202-023-02134-5\u003c/li\u003e\n\u003cli\u003eSikorski, W. (2018). Active dielectric window: A new concept of combined acoustic emission and electromagnetic partial discharge detector for power transformers. Energies, 12(1), 115. https://doi.org/10.3390/en12010115\u003c/li\u003e\n\u003cli\u003eRen, S., Hao, D., Zhao, X., He, Y., Yang, C., Xin, L., ... \u0026amp; Zhao, B. (2024). An ultrasonic signal propagation model of transformer partial discharge considering nonlinear attenuation in the temperature gradient field. Electrical Engineering, 106(1), 53-62. https://doi.org/10.1007/s00202-023-01965-6\u003c/li\u003e\n\u003cli\u003eBeura, C. P., Beltle, M., Tenbohlen, S., \u0026amp; Siegel, M. (2019). Quantitative analysis of the sensitivity of UHF sensor positions on a 420 kV power transformer based on electromagnetic simulation. Energies, 13(1), 3. https://doi.org/10.3390/en13010003\u003c/li\u003e\n\u003cli\u003eAtalar, F., \u0026Ouml;nhon, N. \u0026Ouml;., \u0026amp; Uğur, M. (2023). Investigation of the effect of layered structure on partial discharges in transformer pressboard insulator. Electrical Engineering, 105(6), 3459-3467. https://doi.org/10.1007/s00202-023-01903-6\u003c/li\u003e\n\u003cli\u003eManoj, T., Ranga, C., \u0026amp; Mohan Rao, U. (2022). Multi-criterion fuzzy model for power transformer insulation monitoring. Electrical Engineering, 104(5), 3535-3546. https://doi.org/10.1007/s00202-022-01566-9\u003c/li\u003e\n\u003cli\u003eLeivo, S., Sattler, D., Duval, M., \u0026amp; Chiarella, C. (2023, June). A Case Study: Transformer Failure Detected with Real-time DGA Monitoring. In 2023 IEEE Electrical Insulation Conference (EIC) (pp. 1-4). IEEE. https://doi.org/10.1109/EIC55835.2023.10177343\u003c/li\u003e\n\u003cli\u003eZhang, W., Jiang, J., Li, B., Shen, S., Lu, Y., Tao, Y., \u0026amp; Fan, L. (2023). Residual lifetime evaluation of power transformers based on data fusion and wiener model. IEEE Transactions on Power Delivery, 38(6), 4189-4201. https://doi.org/10.1109/TPWRD.2023.3305732\u003c/li\u003e\n\u003cli\u003eHashemi, M. H., \u0026amp; Dikmen, S. (2023, May). Real-Time Condition Monitoring of Power Transformers Using the Internet of Things (IoT) and Dissolved Gas Analysis. In 2023 XIX International Scientific Technical Conference Alternating Current Electric Drives (ACED) (pp. 1-4). IEEE. https://doi.org/10.1109/ACED57798.2023.10143439\u003c/li\u003e\n\u003cli\u003eZhang, J., Jiang, L., Zhang, H., Zhao, S., Lin, Y., \u0026amp; Chen, J. (2022, October). Research on On-line State Monitoring Method of Voltage Transformer based on Digital Twins. In 2022 4th International Conference on Artificial Intelligence and Advanced Manufacturing (AIAM) (pp. 243-248). IEEE. https://doi.org/10.1109/AIAM57466.2022.00053\u003c/li\u003e\n\u003cli\u003eXiong, J., Ye, H., Pei, W., Kong, L., Huo, Q., \u0026amp; Han, Y. (2022). A monitoring and diagnostics method based on FPGA-digital twin for power electronic transformer. Electric Power Systems Research, 210, 108111. https://doi.org/10.1016/j.epsr.2022.108111\u003c/li\u003e\n\u003cli\u003eRusov V.A. (2022) Partial discharges in power transformers, diagnostic equipment used, expected results. DIMRUS. (in Russian). https//dimrus.com/texts/trans_pd.html. Accessed 20 January 2025.\u003c/li\u003e\n\u003cli\u003eLyutikova, M. (2020). An improved electrochemical method for moisture determination in mineral oil. IEEE Transactions on Dielectrics and Electrical Insulation, 27(6), 2172-2178. https://doi.org/10.1109/TDEI.2020.009016\u003c/li\u003e\n\u003cli\u003eLyutikova, M. N., Korobeynikov, S. M., Ridel, A. V., Vagin, D. V., \u0026amp; Konovalov, A. A. (2022). The breakdown voltage of moistened mineral oil during its cooling. IEEE Transactions on Dielectrics and Electrical Insulation, 29(6), 2181-2189. https://doi.org/10.1109/TDEI.2022.3203658\u003c/li\u003e\n\u003cli\u003eLi, Z., Shi, X., Feng, W., Luo, Z., \u0026amp; Zheng, Q. (2025). Study on Online Monitoring Sensor for Transformer Oil AN Based on Dielectric Constant Measurement. In Frontier Academic Forum of Electrical Engineering (pp. 575-582). Springer, Singapore. https://doi.org/10.1007/978-981-97-8832-3_58\u003c/li\u003e\n\u003cli\u003eIEEE-C57.104-\u0026ndash;-2019.pdf \u003c/li\u003e\n\u003cli\u003eOvsyannikov, A. G., Korobeynikov, S. M., \u0026amp; Vagin, D. V. (2017). Simulation of apparent and true charges of partial discharges. IEEE Transactions on Dielectrics and Electrical Insulation, 24(6), 3687-3693. https://doi.org/10.1109/TDEI.2017.006635\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"electrical-engineering","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"elen","sideBox":"Learn more about [Electrical Engineering](http://link.springer.com/journal/202)","snPcode":"202","submissionUrl":"https://submission.nature.com/new-submission/202/3","title":"Electrical Engineering","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Partial discharge, transformer, Oil, monitoring, hydrogen, water","lastPublishedDoi":"10.21203/rs.3.rs-5888207/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5888207/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis study presents a transformer monitoring system designed for remote measurement of dissolved combustible gases, particularly hydrogen, as well as partial discharges (PD) and oil moisture content. The system was implemented on a 110 kV transformer in late December 2023. The behavior of oil moisture was analyzed, and based on the intensity of PD and the increase in combustible gas concentrations, an unscheduled chromatographic analysis was conducted. This analysis led to the decision to disconnect the transformer. Subsequent internal inspection revealed a defect in the input conductor of one of the phases.\u003c/p\u003e","manuscriptTitle":"Transformer monitoring system and its testing","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-01-28 05:43:34","doi":"10.21203/rs.3.rs-5888207/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-03-19T12:18:39+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-03-18T22:43:56+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-03-18T19:55:04+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-03-14T10:09:54+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"330962698687833084881649892035424200227","date":"2025-02-19T20:08:50+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"271768207429266659929167190828647285626","date":"2025-02-18T12:33:36+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"57413081621011337669957459486964274311","date":"2025-02-18T09:14:43+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"25620818287423944002444165802419259395","date":"2025-02-18T07:22:45+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-02-17T19:44:19+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-01-23T14:59:28+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-01-23T14:22:04+00:00","index":"","fulltext":""},{"type":"submitted","content":"Electrical Engineering","date":"2025-01-23T11:58:17+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"electrical-engineering","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"elen","sideBox":"Learn more about [Electrical Engineering](http://link.springer.com/journal/202)","snPcode":"202","submissionUrl":"https://submission.nature.com/new-submission/202/3","title":"Electrical Engineering","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"8b03f77e-3b2d-4094-8ecf-af9bd2032949","owner":[],"postedDate":"January 28th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-08-18T16:02:46+00:00","versionOfRecord":{"articleIdentity":"rs-5888207","link":"https://doi.org/10.1007/s00202-025-03292-4","journal":{"identity":"electrical-engineering","isVorOnly":false,"title":"Electrical Engineering"},"publishedOn":"2025-08-12 15:57:55","publishedOnDateReadable":"August 12th, 2025"},"versionCreatedAt":"2025-01-28 05:43:34","video":"","vorDoi":"10.1007/s00202-025-03292-4","vorDoiUrl":"https://doi.org/10.1007/s00202-025-03292-4","workflowStages":[]},"version":"v1","identity":"rs-5888207","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5888207","identity":"rs-5888207","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}