The development of analytical tools for the assessment of power transformer reliability is aimed at improving the performance of power systems. This article contributes to the body of knowledge about power transformer outages resulting from defects or failures, highlighting the essential components for decision making. These include the application of the Pareto Diagram (in Phase I); mathematical modeling using reliability functions for different voltage classes; proposal of preventive maintenance schedules based on statistical data from real case studies at a Brazilian company operating in the national power grid; and the application of numerical programming (in Phase II). The originality of this article lies in combining the analysis of the two phases as a complete tool that can be widely applied in power supply companies to support maintenance strategies. In Phase I, the main causes of outages were determined for two time periods, indicating which components should be prioritized. In addition, in Phase II, exponential and two-parameter Weibull distributions pertaining to failures of power transformers with nominal voltages of 34.5 kV, 69 kV and 138 kV. The approach used in Phase II was the least squares method, calculating the Pearson correlation coefficient, and computationally, the 𝒍 𝟐 vector norm, and determining the numerical error. The results were satisfactory, leading us to conclude that the tools complement each other and provide efficient support in decision making.