A novel methodology is developed for predicting the thermal impact of fouling in Steam Generators (SG). The originality of this methodology is to resort to fractal and statistical theories to depict the porous structure of the deposits. The proposed Statistical Fractal methodology (SF) accounts for the heat transfer driven by the liquid-vapor phase change inside the deposits. It simulates the complex intricate networks of sinuous open pores of different scales, with liquid inflows (capillaries) and vapor outflows (steam-chimneys). The multi-layered representation of fouling deposits allows to mimic aging mechanisms such as densification which occur during SG operation. The SF predictions are consistent with experimental data. The deposit thickness and the profile of porosity are found to be the most influential fouling properties on the heat exchange. The methodology is capable to simulate the experimentally observed heat transfer enhancement for thin and porous deposit as well as the heat exchange decline for thick and dense deposit.