An improved wall heat flux partitioning model at the heated surface was developed by Yeoh et al. This model, coupled with a three-dimensional two-fluid model and Multiple Size Group model, has led to satisfactory agreement being achieved between the model predictions and experimental measurements. Nevertheless, one shortcoming is the reliance on empirical correlations for the active nucleation site density in the wall heat flux partitioning model. This discrepancy brings about uncertainties, especially in appropriately evaluating the vapor generation rate, which greatly influences the model prediction on the axial and radial void fraction profiles within the bulk fluid flow. By considering the fractal model with the aforementioned subcooled boiling flow model in the absence of empirical correlations for the active nucleation site density, a comprehensive mechanistic model to predict vertically oriented subcooled boiling flows is developed. The proposed model is assessed against the experimental data of axial measurements of Zeitoun and Shoukri and the radial measurements of Yun et al. and Lee et al. for vertical subcooled boiling flows within annular channels. Improved model predictions are obtained when the model is compared against typically applied empirical correlations for active nucleation site density. Discussions on the agreement of other two-phase flow parameters are also presented.