The current implementation of the HTC model in MARS does not include the dependency of permeability on w/c (water to cement ratio) and s/c (silica fume to cement ratio), according to Equation 33 in the paper: Hygro-thermo-chemical modeling of high performance concrete. I: Theory, G. Di Luzio and G. Cusatis, Cement \& Concrete Composites 31 (2009) 301-308. Hence, the MARS permeability input must be $D_0$ and $D_1$ as opposed to $\tilde{D}_0$ and $\tilde{D}_1$.
Also the numerical values for $\tilde{D}_0$ and $\tilde{D}_1$ reported in Table 2 of Hygro-thermo-chemical modeling of high performance concrete. II: Numerical implementation, calibration, and validation, G. Di Luzio and G. Cusatis, Cement \& Concrete Composites 31 (2009) 309-324, are actually the values for $\tilde{D}_0/c$ and $\tilde{D}_1/c$, $c$= cement content, and the units are mm$^2$/h.
Finally, the original implementation of the HTC model (fortran code) required as input $D_0/c$ and $D_1/c$ in mm$^2$/h.