Accuracy of calculated equilibria

Only two factors are involved in the accuracy of calculated equilibria, i.e. the closeness of the calculated equilibrium temperature at a given pressure to the (accurately) measured temperature or temperature bracket. They are the accuracy of the thermochemical data employed, and the adequacy of the calculation procedure. (We will assume that both the thermochemical and experimental data refer to the same mineral compositions, state of ordering, and so on.) 

Calculated equilibria are in general extremely sensitive to inaccuracies in free energy or enthalpy data, and to a much lesser extent to those in entropy data. Molar volumes of minerals are generally very well known and errors in them have a negligible effect. At 2 Kb for example the temperature of the muscovite-sanidine-corundum-water equilibrium is shifted by 5.2 C and the brucite-periclase-water equilibrium by 7.1°C per 100 calories in 

---