Substantial capital investments in vital assets, particularly power transformers, necessitate the application of precise diagnostics. These diagnostics are crucial for assessing performance, identifying potential issues, and ensuring these assets' long-term operational and maintenance efficiency. The primary objective is proactively mitigating asset failure risk and the subsequent need for costly replacements. In recent years, significant progress has been made in developing AI models for fault classification, primarily leveraging machine learning methodologies. However, a notable characteristic of many machine learning approaches is their inherent black-box nature, which limits their interpretability. The opacity of these models necessitates adopting Explainable Artificial Intelligence (XAI) techniques to elucidate their decision-making processes. In this study, we have explored the application of various machine learning algorithms, including Support Vector Machines (SVM), k-nearest Neighbors (KNN), Random Forest (RF), eXtreme Gradient Boosting (XGBoost), and Artificial Neural Networks (ANN), for fault classification. Among these models, the Random Forest algorithm yielded the most promising results. We applied XAI approaches to enhance our understanding of its decision-making mechanisms and facilitate better-informed decision-making.