Unexpected Pressure Drop at the Top of the Domain When Chemistry Module Is Enabled in CO₂ Two-Phase Simulation

[chenyao su]

Dear all,

I am using PFLOTRAN to simulate CO₂ injection, and I encountered a strange phenomenon.

When I run the case without the chemistry module (i.e., pure two-phase flow), the pressure field behaves normally. However, when I enable the chemistry module, an unexpected pressure drop appears at the top of the computational domain. This behavior does not occur in the pure two-phase flow case.

Some additional details:

• I did not enable porosity–permeability relationships.

• The relative permeability functions are identical in both simulations.

• The case is based on Case 5 from the PFLOTRAN v7 paper, but I change the computational domain and chemistry.(Modeling Supercritical CO2 Flow and Mineralization in Reactive  Host Rocks with PFLOTRAN v7.0)

• I found that this pressure-drop phenomenon only occurs when the domain is sufficiently wide (for example, radius larger than about 5000 m).

My question is:

Why does enabling the chemistry module cause this pressure decrease at the top of the domain?

Do I need to explicitly impose boundary conditions at the top and bottom of the domain, such as:

• LIQUID_PRESSURE DIRICHLET

• LIQUID_FLUX NEUMANN

• GAS_FLUX NEUMANN

in order to control or eliminate this behavior?

  For reference, the input file and a video of the simulation results are attached. Any suggestions or explanations would be greatly appreciated.  (pflotran version:V7.0)

  Thank you in advance. 

[Hammond, Glenn E]

Chenyao,

When supercritical CO₂ and chemistry are combined in your simulation, mineral precipitation can act as a sink for CO₂, creating feedback between chemistry and flow. To test this effect, please add the card "MINERAL_RATE_SCALING_FACTOR 0." to the CHEMISTRY block—at the same level as MINERAL_KINETICS, but not inside MINERAL_KINETICS. This will set all mineral reaction rate constants to zero. After making this change, does the flow solution align with your expectations?

Adjusting the rate constants within this simulation can introduce instability, potentially leading to unexpected or unreliable results. To maintain consistency and better understand the model's sensitivity, it's advisable to retain the original rate constants and utilize the MINERAL_RATE_SCALING_FACTOR mentioned earlier. By scaling all rates simultaneously with this factor, you can more effectively observe how variations in the rate constant influence the model's behavior, allowing for a clearer assessment of its response.

Boundaries in the simulation are set to no flow or zero flux by default. Therefore, it's unnecessary to explicitly define boundary conditions at the top or bottom of the domain unless there is a thermal gradient.

Glenn

From: pflotra...@googlegroups.com <pflotra...@googlegroups.com> on behalf of chenyao su <suche...@gmail.com>
Date: Monday, March 16, 2026 at 7:50 AM
To: pflotran-users <pflotra...@googlegroups.com>
Subject: [pflotran-users: 8715] Unexpected Pressure Drop at the Top of the Domain When Chemistry Module Is Enabled in CO₂ Two-Phase Simulation

You don't often get email from suche...@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/37b70a2e-62dd-47be-a9fd-af43a0646c92n%40googlegroups.com.