The 'wall heat flux partitioning' (WHFP) model in conjunction with an Eulerian-Eulerian Multiphase Flow (EEMF) method is found to be an apt tool, to simulate the physics of subcooled flow boiling phenomena. However, the empiricism of the constituent model limits its applicability to predominantly low pressure operating conditions. Since pressurized heavy water reactors (PHWRs) are mostly operated at higher pressures, their reliable usage becomes highly questionable. To this end, we extensively assess and validate the WHFP model in the coupled EEMF-WHFP framework to simulate subcooled flow boiling conditions at high pressures. Based on the simulations, an improved mechanistic correlation combination for use in the WHFP model is recommended in place of the standard combination that is popularly used. This would enable reliable application to nuclear systems at higher pressures. To start with, a suitable high-pressure flow boiling experiment was identified and the coupled EEMF WHFP method typically used in commercial solvers was assessed. Following this, several new nonstandard correlation combinations were examined, in order to understand the dynamics and parametric sensitivities of the WHFP model in the coupled EEMF-WHFP framework. The entire focus is on finding out, the structure of the formulation with a sound physical basis and best possible prediction capability. To this end, a comprehensive literature review of the bubble parameters such as, diameter (D), number density (N) and frequency (f) based correlations, was performed. Popular models were identified, and systematically categorized based on the physical mechanisms behind their modeling. Based on an extensive performance evaluation, it was found that the coupled WHFP model converges and generally predicts physically relevant values across all of the modeling combinations and this robustness is found to be primarily due to the N-T-sup interdependence. It is also shown that the models, which are formulated in terms of the active cavity radius and the bubble growth modulus span very well across various pressures. Based on this study, we recommend a new non-standard correlation combination with a good physical basis that enables excellent predictions for the model, over a wide range of pressures. (C) 2016 Elsevier Ltd. All rights reserved.