Details of a complete methodology for porosimetry of virtual concrete have been described in a series of publications. This paper shortly sketches the main characteristics, whereupon it is indicated how a pore network model is designed based on data extracted from the virtual cement paste. This involves smoothing the main trunks as the main pore connections between outer surfaces. Network estimates for water transport through concrete is presented and shown exceeding commonly used MIP estimates. For quantitative image analysis results the gap will be much smaller, though. Moreover, experimental data are biased by continued hydration (self-healing effects). Some simple examples are presented to improve the insight into the methodology. Moreover, they can additionally serve as validation by comparing estimates with analytical results.