The CALPHAD approach takes experimental data and parameterises Gibbs energy curves with an optimal fit to this data, which can be phase boundary measurements or thermodynamic data. Thermo-Calc, and other phase diagram assessment codes, make no reference to the vacancy formation energy explicitly, leaving it only to be referred to implicitly within the Gibbs energy formulation . Nevertheless, vacancy formation energies must significantly affect the phase stability when compounds are not strictly stoichiometric.
In recent years, theoretical calculations of the vacancy formation energy (or other point defect formation energies), which are difficult or impossible to measure experimentally, have been produced. This data could be used to increase the physical information encoded within a thermodynamic database.
For the zirconium-carbon phase diagram, I show how data from DFT calculations of the vacancy formation energy in the zirconium carbide phase can introduced as a constraint on the parameterisation of the Gibbs energy, and how this changes the resulting phase diagram.
 J. Rogal et al. (2014), Phys. Status Solidi B, 251(1) 97-129.