This paper presents a preliminary analysis of two dry-type distribution transformers recovered after a major flood in Rio Grande do Sul, Brazil, in 2024. Insulation resistance was measured before and after an emergency thermal-drying procedure, showing increases between one and three orders of magnitude depending on the dielectric path. To qualitatively represent the variability of post-recovery insulation, we apply a simple exploratory Monte Carlo simulation assuming a lognormal distribution centered on the measured reference value. The approach illustrates how dispersion around nominal insulation levels can be incorporated into qualitative risk discussions. Although limited to a small set of field cases and a simplified stochastic model, the study combines empirical measurements from a real flood event with probabilistic analysis to support discussions on risk and resilience in urban distribution systems exposed to extreme weather conditions.