geothermal model produces T lower than injection temperature and unrealistically large heat power after years

[Yulongggggg]

Hi PorousFlow/MOOSE team,

I am running a fully-saturated ThermoHydro (TH) PorousFlow simulation for a geothermal injection/production setup on a 2D fracture network (lower-D fractures). I believe my parameters are physically reasonable, and I am scaling a 2D slice to represent a 3D reservoir using an out-of-plane length Lz.

However, even after ~3 years of simulation time, I observe:
1. Large areas where temperature T becomes lower than the injection temperature (T < T_inj).
2. The reported thermal power (from produced_heat / enthalpy extraction) becomes unrealistically large.

[Yulongggggg]

temperature is more realistic after i used [PorousFlowPointEnthalpySourceFromPostprocessor

[josebastiase]

Hi,
When you say “more realistic”, do you mean that the issue is now resolved, or just that the results look qualitatively better?

Without a minimal input file it’s unfortunately very difficult to diagnose this properly, but a few general comments:

If you are currently injecting energy via enthalpy, you might try instead fixing the temperature at the injection nodes using a Dirichlet BC on T.

I would carefully re-check the boundary conditions for heat vs. mass. A common issue is to allow fluid to leave the domain (e.g. via pressure or Peaceman wells) while unintentionally restricting heat transport, which can lead to unphysical cooling or inflated produced heat.

That said, this is hard to evaluate meaningfully without a minimal, reproducible example. If you can share a stripped-down input that still shows the behavior, it would make it much easier to pinpoint the cause.