Predictive Condition Assessment of Distribution Transformers: Using Load, Thermal, Power-Quality, and Maintenance Indicators

References

[1] Abu-Elanien, A.E.B. and Salama, M.M.A. (2010) ‘Asset management techniques for transformers’, Electric Power Systems Research, 80(4), pp. 456-464. doi: 10.1016/j.epsr.2009.10.008. ScienceDirect

[2] Arias Velasquez, R.M. and Mejia, J.M. (2018) ‘Health Index for Transformer Condition Assessment’, IEEE Latin America Transactions, 16(12), pp. 2843-2849.

[3] Breiman, L. (2001) ‘Random forests’, Machine Learning, 45, pp. 5-32. doi: 10.1023/A:1010933404324. Springer

[4] Cazacu, E., Ionita, V. and Petrescu, L. (2018) ‘Thermal Aging of Power Distribution Transformers Operating under Nonlinear and Balanced Load Conditions’, Advances in Electrical and Electronic Engineering, 16(1). doi: 10.15598/aeee.v16i1.2701. AEEE

[5] Chandola, V., Banerjee, A. and Kumar, V. (2009) ‘Anomaly detection: A survey’, ACM Computing Surveys, 41(3), Article 15. doi: 10.1145/1541880.1541882. ACM

[6] Chatterjee, A. and Ahmed, B.S. (2022) ‘IoT anomaly detection methods and applications: A survey’, Internet of Things, 19, 100568. ScienceDirect

[7] Choga, C.R. (2026) Integrating DNA and RNA Multi-Omics for Improved Survival Prediction in Lung Cancer. Master’s thesis. Yeshiva University, New York.

[8] Foros, J. and Istad, M. (2020) ‘Health Index, Risk and Remaining Lifetime Estimation of Power Transformers’, IEEE Transactions on Power Delivery, 35. doi: 10.1109/TPWRD.2020.2972976. IEEE

[9] Friedman, J.H. (2001) ‘Greedy function approximation: A gradient boosting machine’, Annals of Statistics, 29(5), pp. 1189-1232. doi: 10.1214/aos/1013203451. Project Euclid

[10] Ganyani, L., Nare, S., Danda, C., Chingezi, L. and Mupa, M.N. (2026) ‘From Tax Mismatches to Revenue Protection: A Data-Driven Framework for Audit Prioritization, Voluntary Compliance, and SME Compliance Education’, Iconic Research and Engineering Journals, 10(1), pp. 517-530. doi: 10.64388/IREV10I1-1719533. IRE Journals

[11] Ghoneim, S.S.M. and Taha, I.B.M. (2016) ‘A new approach of DGA interpretation technique for transformer fault diagnosis’, International Journal of Electrical Power & Energy Systems, 81, pp. 265-274. doi: 10.1016/j.ijepes.2016.02.018. ScienceDirect

[12] Guo, H. and Guo, L. (2022) ‘Health index for power transformer condition assessment based on operation history and test data’, Energy Reports, 8, pp. 9038-9045. doi: 10.1016/j.egyr.2022.07.041. ScienceDirect

[13] Hagan, E., Akakpo, L., Kasuwa, T.L., Nhemachena, T.L., Taanisa, T., Osman, E.W., Tsveta, T. and Mupa, M.N. (2026a) ‘Climate-centric data governance for timber traceability: A GRC framework for decarbonized and audit-ready supply chains’, World Journal of Advanced Research and Reviews, 30(2), pp. 353-361. doi: 10.30574/wjarr.2026.30.2.1196. WJARR

[14] Hagan, E., Akakpo, L., Kasuwa, T.L., Nhemachena, T.L., Taanisa, T., Osman, E.W., Tsveta, T. and Mupa, M.N. (2026b) ‘Predictive analytics for climate-resilient timber supply chains: Integrating anomaly detection with carbon sequestration and risk scoring’, World Journal of Advanced Research and Reviews, 30(2). doi: 10.30574/wjarr.2026.30.2.1197. WJARR

[15] Huang, Y.-C. (2012) ‘Data mining for oil-insulated power transformers: an advanced literature survey’, WIREs Data Mining and Knowledge Discovery, 2. doi: 10.1002/widm.1043. Wiley

[16] IEEE (2011) IEEE Std C57.91-2011: IEEE Guide for Loading Mineral-Oil-Immersed Transformers and Step-Voltage Regulators. New York: IEEE. IEEE

[17] IEEE (2022) IEEE Std 519-2022: IEEE Standard for Harmonic Control in Electric Power Systems. New York: IEEE. IEEE

[18] International Electrotechnical Commission (IEC) (2018) IEC 60076-7: Power transformers - Part 7: Loading guide for mineral-oil-immersed power transformers. Geneva: IEC. IEC

[19] Islam, N., Khan, R., Das, S.K., Sarker, S.K., Islam, M.M., Akter, M. and Muyeen, S.M. (2023) ‘Power transformer health condition evaluation: A deep generative model aided intelligent framework’, Electric Power Systems Research, 218, 109201. doi: 10.1016/j.epsr.2023.109201. ScienceDirect

[20] Jahromi, A.N., Piercy, R., Cress, S., Service, J.R.R. and Fan, W. (2009) ‘An approach to power transformer asset management using health index’, IEEE Electrical Insulation Magazine, 25(2), pp. 20-34. doi: 10.1109/MEI.2009.4802595. IEEE

[21] Jin, L., Kim, D. and Abu-Siada, A. (2023) ‘State-of-the-art review on asset management methodologies for oil-immersed power transformers’, Electric Power Systems Research, 218, 109194. doi: 10.1016/j.epsr.2023.109194. ScienceDirect

[22] Kabera, S.T., Mupa, M.N., Gezah, F.Y., Nhevera, P.N., Mtemeli, P.S. and Hove, T.L. (2026) ‘Leveraging Digital Twin Technology and Predictive Maintenance for Optimizing Solar PV and Battery Storage Systems’, World Journal of Advanced Research and Reviews, 29(2), pp. 1051-1059. doi: 10.30574/wjarr.2026.29.2.0335. WJARR

[23] Liu, F.T., Ting, K.M. and Zhou, Z.-H. (2008) ‘Isolation Forest’, Proceedings of the 8th IEEE International Conference on Data Mining, pp. 413-422. doi: 10.1109/ICDM.2008.17. IEEE

[24] Machekera, G.N., Mtemeli, P.S., Dongo, M.M., Sajanga, G.A. and Mupa, M.N. (2026) ‘An Explainable Deep Learning Framework for Energy-Efficient Water and Sewer Reticulation Infrastructure: Predicting Project Risk, Pumping Demand, and Sustainable Construction Performance’, Iconic Research and Engineering Journals, 9(12), pp. 3197-3209. doi: 10.64388/IREV9I12-1719256. IRE Journals

[25] Mirowski, P. and LeCun, Y. (2012) ‘Statistical machine learning and dissolved gas analysis: a review’, IEEE Transactions on Power Delivery, 27(4), pp. 1791-1799. doi: 10.1109/TPWRD.2012.2197868. IEEE

[26] Mitiche, I., McGrail, T., Boreham, P., Nesbitt, A. and Morison, G. (2021) ‘Data-driven anomaly detection in high-voltage transformer bushings with LSTM auto-encoder’, Sensors, 21, 7426. MDPI

[27] Murugan, R. and Ramasamy, R. (2019) ‘Understanding the power transformer component failures for health index-based maintenance planning in electric utilities’, Engineering Failure Analysis, 96, pp. 274-288. doi: 10.1016/j.engfailanal.2018.10.011. ScienceDirect

[28] Okebu, D.T.C. and Mupa, M.N. (2026) ‘Predictive training analytics for competency-based pilot instruction: Early Detection of Checkride Risk, Procedural Noncompliance and Safety-Relevant Learning Gaps’, World Journal of Advanced Research and Reviews, 30(3), pp. 1928-1942. doi: 10.30574/wjarr.2026.30.3.1774. WJARR

[29] Putchala, S. (2020) Distributed Transformer Monitoring [Dataset]. Kaggle. Available at: https://www.kaggle.com/datasets/sreshta140/ai-transformer-monitoring (Accessed: 26 August 2026). Kaggle

[30] Ramesh, J., Shahriar, S., Al-Ali, A.R., Osman, A. and Shaaban, M.F. (2022) ‘Machine Learning Approach for Smart Distribution Transformers Load Monitoring and Management System’, Energies, 15(21), 7981. doi: 10.3390/en15217981. MDPI

[31] Rudin, C. (2019) ‘Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead’, Nature Machine Intelligence, 1, pp. 206-215. doi: 10.1038/s42256-019-0048-x. Nature

[32] Saito, T. and Rehmsmeier, M. (2015) ‘The precision-recall plot is more informative than the ROC plot when evaluating binary classifiers on imbalanced datasets’, PLOS ONE, 10(3), e0118432. doi: 10.1371/journal.pone.0118432. PLOS

[33] Singh, S. and Bandyopadhyay, M.N. (2010) ‘Dissolved gas analysis technique for incipient fault diagnosis in power transformers: a bibliographic survey’, IEEE Electrical Insulation Magazine, 26(6), pp. 41-46. doi: 10.1109/MEI.2010.5599978. IEEE

[34] Sun, H.-C., Huang, Y.-C. and Huang, C.-M. (2012) ‘A Review of Dissolved Gas Analysis in Power Transformers’, Energy Procedia, 14, pp. 1220-1225. doi: 10.1016/j.egypro.2011.12.1079. ScienceDirect

[35] Susa, D., Lehtonen, M. and Nordman, H. (2005a) ‘Dynamic Thermal Modelling of Power Transformers’, IEEE Transactions on Power Delivery, 20(1), pp. 197-204. doi: 10.1109/TPWRD.2004.835255. IEEE

[36] Susa, D., Lehtonen, M. and Nordman, H. (2005b) ‘Dynamic thermal modeling of distribution transformers’, IEEE Transactions on Power Delivery, 20(3), pp. 1919-1929. doi: 10.1109/TPWRD.2005.848675. IEEE

[37] Taha, I.B.M. (2023) ‘Power Transformers Health Index Enhancement Based on Convolutional Neural Network after Applying Imbalanced-Data Oversampling’, Electronics, 12(11), 2405. doi: 10.3390/electronics12112405. MDPI

[38] Taha, I.B.M. et al. (2024) ‘Power Transformer Health Index Using Cost-Sensitive Learning to Consider the Impact of Misclassification’, IEEE Access, 12, pp. 191790-191807. doi: 10.1109/ACCESS.2024.3516785.

[39] Zahra, S.T., Imdad, S.K., Khan, S., Khalid, S. and Baig, N.A. (2025) ‘Power transformer health index and life span assessment: A comprehensive review of conventional and machine learning based approaches’, Engineering Applications of Artificial Intelligence, 139, 109474. doi: 10.1016/j.engappai.2024.109474. ScienceDirect

[40] Zemouri, R. (2025) ‘Power Transformer Prognostics and Health Management Using Machine Learning: A Review and Future Directions’, Machines, 13(2), 125. doi: 10.3390/machines13020125. MDPI

@article{1723216,
      author = {Munashe Cosmas Kabuya, Reign Tongai Mutendahama, Geoff Mangwiro, Liberty Munyaradzi Kanonjera, Munashe Naphtali Mupa},
      title = {Predictive Condition Assessment of Distribution Transformers: Using Load, Thermal, Power-Quality, and Maintenance Indicators},
      journal = {Iconic Research And Engineering Journals},
      year = {2026},
      volume = {10},
      number = {3},
      pages = {2081-2103},
      issn = {2456-8880},
      url = {https://www.irejournals.com/formatedpaper/1723216.pdf},
      abstract = {Distribution transformers occupy a critical position between bulk-power infrastructure and end users, yet many units are still managed through periodic inspection, fixed alarms, and reactive maintenance. This study develops an interpretable predictive condition-assessment framework that integrates loading, thermal behavior, power quality, and maintenance history into a single engineer-facing reliability health score while retaining statistical and machine-learning challengers. The public-data foundation is the Kaggle Distributed Transformer Monitoring dataset, which contains five IoT data files and 54 variables collected at approximately 15-minute intervals between 25 June 2019 and 14 April 2020, including three-phase current and voltage, oil and winding temperatures, oil level, power, power factor, frequency, and harmonic-distortion measures (Putchala, 2020; Ramesh et al., 2022). Because that public release does not contain adjudicated maintenance histories or a complete failure-outcome record suitable for causal maintenance inference, the empirical evaluation deliberately separates data provenance from scenario validation. A 19,200-observation benchmark-calibrated engineering panel reproduces the source signal families and adds explicitly simulated maintenance-age, defect-backlog, and prior-alarm variables. A severe multi-indicator condition onset within the subsequent six hours is used as a controlled prospective target; it is a method-validation event, not a claim of actual transformer failure prevalence. The study compares a fixed threshold rule, an interpretable 0-100 condition-stress score, robust multivariate distance, Isolation Forest, class-weighted logistic regression, random forest, and gradient boosting under a chronological 60/20/20 development-validation-test design. On the locked temporal test, logistic regression achieved the strongest discrimination (AUROC 0.763; 95% bootstrap interval 0.744-0.780; AUPRC 0.405), while the interpretable condition score achieved AUROC 0.647 and AUPRC 0.315 and produced a monotonic risk gradient: the condition-event rate rose from 7.8% in the very-healthy band to 70.0% in the critical band. Feature-layer ablation showed that adding thermal variables to load indicators increased AUROC from 0.716 to 0.742, while adding maintenance evidence increased AUPRC from 0.366 to 0.405. Static thresholds were precise for some severe states but captured fewer emerging events, whereas anomaly methods were sensitive to temporal threshold transport. The results support a hybrid maintenance architecture in which interpretable engineering scores and hard limits provide transparent triage, statistical models provide an independent challenger, and engineers retain authority over inspection, testing, derating, repair, and replacement decisions. The contribution is therefore not a claim that machine learning replaces transformer engineering; it is a reproducible decision framework for converting heterogeneous condition evidence into prioritized, auditable maintenance action.},
      keywords = {distribution transformer; condition assessment; predictive maintenance; health index; thermal aging; power quality; anomaly detection; Isolation Forest; maintenance prioritization; reliability engineering.},
      month = {September},
  }