Help with THC simulation setup for natural hydrogen transport

[sahil wani]

Hello everyone,

I am currently working on modelling natural hydrogen storage and transport in the Earth’s crust. Initially, I was using PFLOTRAN-OGS, but due to several constraints, I am planning to shift fully to PFLOTRAN.

I am particularly interested in setting up a thermo-hydro-chemical (THC) coupled simulation for natural hydrogen transport. As I am still new to PFLOTRAN, I would like to start from a simple and reliable base case.

Could you please suggest:
-Any basic example input file for THC simulations
-Relevant tutorials or benchmark problems that include gas transport and reactions
-Guidance on how to define hydrogen as a component in the simulation
-A minimal working input file would help me understand the structure and modify it for my case.

Thank you for your time and support.

Sahil

[Hammond, Glenn E]

Sahil,

To get started, I recommend reviewing the multiphase gas transport example found in the regression tests:

https://bitbucket.org/pflotran/pflotran/src/master/regression_tests/default/column/multiphase_transport_pulse.in 

Although this example does not include energy conservation, you can easily incorporate it as needed. Additionally, you may find useful resources in the anisothermal directory of the regression tests:

https://bitbucket.org/pflotran/pflotran/src/master/regression_tests/default/anisothermal/ 

It's important to keep in mind that PFLOTRAN’s THC framework generally separates the thermo-hydro processes from the chemical processes. With the exception of supercritical CO2, reactive transport does not feed back into the flow, meaning, for instance, that H2(g) production in the transport gas phase does not affect flow dynamics.

I'm interested to learn more about the specific challenges or limitations you encountered with PFLOTRAN-OGS that led you to transition away from it. Could you share some details about those constraints?

Thanks,

Glenn

From: pflotra...@googlegroups.com <pflotra...@googlegroups.com> on behalf of sahil wani <shl...@gmail.com>
Date: Friday, April 3, 2026 at 8:08 AM
To: pflotran-users <pflotra...@googlegroups.com>
Subject: [pflotran-users: 8738] Help with THC simulation setup for natural hydrogen transport

You don't often get email from shl...@gmail.com. Learn why this is important

Check twice before you click! This email originated from outside PNNL.

--
You received this message because you are subscribed to the Google Groups "pflotran-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pflotran-user...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/pflotran-users/8ce899e4-4c49-467b-a570-53caefb61d8bn%40googlegroups.com.

[sahil wani]

Dear Glenn,

Thank you for the pointers to the multiphase transport and anisothermal tests.

I switched to GENERAL mode since SOURCE_SINK flux is not supported in PFLOTRAN-OGS for distributed boundary injection. Dirichlet BCs for gas saturation and pressure worked better for representing diffuse H2 sources.

I am running a case study with RKS gas density, the energy equation, and van Genuchten curves. Most cases converge, but those with high capillary entry pressure for caprock do not.

At the reservoir and caprock interface, cells oscillate between Liquid and 2-Phase states. As Sl approaches 1, the VG curve becomes very steep, which makes the Jacobian ill-conditioned and leads to persistent Newton oscillations. The RESTRICT_STATE_CHANGE is already active, but it has not improved convergence. 

Can you suggest something so I can run those simulation at high capillary pressure. 

Thank you again for your guidance.