I am trying to model thermal condutive heating in unsaturated soil and I need the temperature to get above 500 degrees C after water dries out. There seems to be a temperature limit at 350 degrees C as my model stops at that point. I wonder if this can be solved or the code simply has a temperature limit.
TMVOC has an upper limit of T=350 C and P=1e8 Pa. It checks at the start of each time step whether T and P are within bounds, and it stops if not. You should get a message "CANNOT FIND PARAMETERS AT ELEMENT ". TMVOC is designed for near-surface problems, where typically T and P are low. Even if you turned off this T,P check, the code would likely not work well because the functions describing the thermophysical properties are not designed to go above these limits.
TMVOC in TOUGH2 uses the IFC-67 correlations for pure water properties. I guess TMVOC in TOUGH3 uses IAPWS-IF97.
In both cases, Region 1 correlations (liquid water) have an upper limit of T=350°C (even though they are accurate up to 365°C), while Region 2 vapor (sub-critical and super-critical steam) has correlations valid up to 800°C.
Because of that limit, TMVOC has an internal control to avoid running at T>350°C, as in all TOUGH2 modules, except EOS1sc in iTOUGH2 (Magnusdottir and Finsterle, 2015).
In practice, it would be possible to let the code work for single-gas conditions even at T above 350°C. This has been done within two modules called EOS1H and EOS2H (Battistelli et al., 2020, Geothermics), without implementing the correlations for Region 3 (supercritical).
TMVOC has another problem linked to the T limit for the thermodynamic approach followed to describe hydrocarbon mixtures.
In TMVOC it is assumed that the local T is lower than the critical temperature of all the VOCs simulated. For instance, methane cannot be modeled as a VOC, but only as an NCG. Thus, unless your VOCs have all a critical T above 500°C, it would not be possible to model T above 350°C and up to 500°C, even though modifications as those applied in EOS2H will be included.