Batch Reactor

90 views
Skip to first unread message

reisi...@gmail.com

unread,
Oct 30, 2024, 4:58:17 PM10/30/24
to pflotran-users
Hello,

I’m working on modeling a batch reactor filled with CO₂-saturated Brine, in which a core sample is submerged. I was wondering if there are any PFORTRAN examples or guidance that could assist with setting up this model? I'm a bit unsure about the appropriate boundary conditions.

Thank you,  
Fatemeh

Fatemeh

unread,
Nov 3, 2024, 12:08:22 PM11/3/24
to pflotra...@googlegroups.com

Just a kind reminder for this.

Thank you 

On Oct 30, 2024, at 18:28, reisi...@gmail.com wrote:

Hello,


I’m working on modeling a batch reactor filled with CO₂-saturated Brine, in which a core sample is submerged. I was wondering if there are any PFORTRAN examples or guidance that could assist with setting up this model? I'm a bit unsure about the appropriate boundary conditions.

Thank you,  
Fatemeh

--
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/e1cf47fc-e07c-46e4-9c39-8c9a21ef1676n%40googlegroups.com.

Nole, Michael A

unread,
Nov 4, 2024, 11:24:58 AM11/4/24
to pflotra...@googlegroups.com

Fatemeh,

 

You could try out SCO2 Mode. Take a look at regression_tests/co2 for some example models with boundary conditions.

 

Michael

 

From: pflotra...@googlegroups.com <pflotra...@googlegroups.com> on behalf of Fatemeh <reisi...@gmail.com>
Date: Sunday, November 3, 2024 at 9:08 AM
To: pflotra...@googlegroups.com <pflotra...@googlegroups.com>
Subject: Re: [pflotran-users: 8095] Batch Reactor

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

 

Fatemeh

unread,
Nov 4, 2024, 12:16:36 PM11/4/24
to pflotra...@googlegroups.com

Hi Michael

Thank you kindly for your response. My issue is that I’m unsure how to define the brine and rock when I have a core surrounded by brine in a vessel. I know I should remove the flow part since there is no flow and I how to define the chemistry and reactions and…., I am confused in the physical definition of rock, carbonated brine and vessel.

Bests,

On Nov 4, 2024, at 12:54, 'Nole, Michael A' via pflotran-users <pflotra...@googlegroups.com> wrote:



Hammond, Glenn E

unread,
Nov 13, 2024, 11:56:41 AM11/13/24
to pflotra...@googlegroups.com

Fatemeh,

 

This all depends on whether you plan to include flow process model in addition to reactive transport. From the below, I will assume that you are not doing so. In that case, set up the back reactor with volume fractions of select minerals through an initial condition based on a transport condition -> transport constraint that defines the minerology and aqueous concentrations. Please let me know if this does not answer your question.

 

Glenn

 

Fatemeh

unread,
Nov 13, 2024, 2:24:35 PM11/13/24
to pflotra...@googlegroups.com

Dear Glenn

Thank you for your reply. 
I sent my input file and detailed explanations to your email, can you please look at it? I think logic behind my input file has problems.

Bests,

Fatemeh 

Fatemeh

unread,
Nov 14, 2024, 3:03:25 PM11/14/24
to pflotra...@googlegroups.com

Hi Glenn

I think you might not have had time to review my file yet. My question is: do I need to define separate regions within my cylinder that contain brine, or should I consider the entire system as the core? In that case, would the brine only occupy the pores and boundaries?

This is a brief explanation of my experiment which I am trying to model:

I want to analyze changes in rock composition after a year). My setup involves a cylinder filled with CO₂-saturated brine, with a cement core sample positioned at the bottom center. The core has a diameter of 2.54 cm and a height of 5 cm, while the cylinder dimensions are 7.62 cm in diameter and 9 cm in height.

Hammond, Glenn E

unread,
Nov 15, 2024, 5:12:17 PM11/15/24
to pflotra...@googlegroups.com

Fatemeh,

 

PFLOTRAN considers the entire domain to be porous media. You could create separate regions for the core and the surrounding brine and set the porosity of the surrounding brine cells to be 1, but the code will only differentiate on the porosity. Minerals can still precipitate-dissolve in the surrounding brine if it becomes supersaturated with respect to a mineral.

 

Glenn

 

Fatemeh

unread,
Nov 15, 2024, 5:29:25 PM11/15/24
to pflotra...@googlegroups.com
Hi Glenn,  

Thank you for your response. I created a new model where I considered the core as the entire system, with brine only at the boundaries. Do you think this approach better replicates the experimental results? I've attached my input file but please note that the values are placeholders and do not match those from my experiment..could you please review it and let me know if my definitions are correct?  

Additionally, I have a question regarding the constraints for the core. For ions that are not initially present in my brine or are present in both the brine and the core's minerals—such as Ca and Al+++ in my example—what concentration should I assign to them?  

Best regards,  

Fatemeh

cement3.in

Hammond, Glenn E

unread,
Nov 19, 2024, 12:09:30 PM11/19/24
to pflotra...@googlegroups.com

Fatemeh,

 

The physical domain appears correct. Regarding geochemistry, I cannot say whether many of the parameters (e.g., rate constants) are correct (or realistic), but I provide a couple comments:

 

  • Your Na+ and Cl- concentrations are very high and may require the use of a Pitzer ion activity model, which PFLOTRAN lacks.
  • Why are the initial concentrations (core constrain) set to 1.d-10? Are you flooding the core with DI water?
  • I would not adjust the Newton solver tolerances (NUMERICAL_METHODS -> NEWTON_SOLVER) unless you are confident that those settings will improve convergence.
  • Set your initial time step size to be 1 second and allow the code to increase the time step as the Newton solver converges.

 

Glenn

 

Fatemeh

unread,
Nov 19, 2024, 12:59:24 PM11/19/24
to pflotra...@googlegroups.com
Dear Glenn,

Thank you for your valuable feedback. 

Could you clarify which specific values seem unrealistic to you? (I am uncertain about the appropriate value for specific surface area, as I haven’t measured it and the literature reports a wide range of values. What would you recommend in this case?)

I am not performing core flooding; instead, the core is submerged in CO₂-saturated brine. My main confusion lies here: what should I specify as the concentration constraint for the core? Should it match exactly what I have defined for the brine in the concentration section?

Additionally, I still have this question: For primary species that are either not initially present in the brine or exist in both the brine and the core’s minerals—such as Ca and Al+++ in my case—what concentration should I assign? For instance, since Ca++ initially is present in my brine and also exists in calcite, when setting the constraint for the core,1) should I assign a very small initial concentration for Ca++, assuming there is no brine in the core at the beginning? 2) Or should I use the Ca++ concentration from the brine? 3) Or, should I add the concentration from the brine to account for the presence of Ca++ in calcite?

Bests, 
Fatemeh

Peter Lichtner

unread,
Nov 19, 2024, 2:20:34 PM11/19/24
to pflotran-users
Fatemeh: you might want to consider the dual continuum model with one continuum referring to the brine as primary continuum and the other the core as secondary continuum, with diffusive mass transfer between the two. Not sure this will work in the presence of gases, however.

Alternatively you could grind up the core and create a mixture with the brine as a single continuum.

Some figures might help here to define the initial configuration of your experiment.
Peter

Fatemeh

unread,
Nov 19, 2024, 2:42:24 PM11/19/24
to pflotra...@googlegroups.com
Dear Peter,  

Thank you very much for your response.  

I was wondering if you could kindly explain how I might go about implementing this. Is there an example that could help clarify? Glenn mentioned that the entire system is treated as porous media in PFLOTRAN.

Do you think the input file I provided is not suitable for this experiment?
  
I would also greatly appreciate it if you could share your thoughts regarding my questions on constraints and chemistry.  

I have attached a picture that shows a schematic of my experiment just consider one core instead of four.


Best regards,  
Fatemeh

image0.png


On Nov 19, 2024, at 15:50, Peter Lichtner <peter.l...@gmail.com> wrote:

Fatemeh: you might want to consider the dual continuum model with one continuum referring to the brine as primary continuum and the other the core as secondary continuum, with diffusive mass transfer between the two. Not sure this will work in the presence of gases, however.

Peter Lichtner

unread,
Nov 20, 2024, 5:44:02 PM11/20/24
to pflotran-users
Greetings Fatemeh: First is this a well-stirred reactor? How many groups of concentrations are you planning on measuring at each time step: one for the brine and one for cement pore fluid? Or will you just measure one value? 

Minerals not initially present should be assigned a zero volume fraction. 

Aqueous species not initially present should be given a small value (e.g. 1.e-16 mol/L or smaller depending on the species).

Peter

On Nov 19, 2024, at 12:41, Fatemeh <reisi...@gmail.com> wrote:

Dear Peter,  

Thank you very much for your response.  

I was wondering if you could kindly explain how I might go about implementing this. Is there an example that could help clarify? Glenn mentioned that the entire system is treated as porous media in PFLOTRAN.

Do you think the input file I provided is not suitable for this experiment?
  
I would also greatly appreciate it if you could share your thoughts regarding my questions on constraints and chemistry.  

I have attached a picture that shows a schematic of my experiment just consider one core instead of four.


Best regards,  
Fatemeh

<image0.png>

Fatemeh

unread,
Nov 20, 2024, 6:06:07 PM11/20/24
to pflotra...@googlegroups.com
Dear Peter,

Thank you for your prompt reply.

Yes, that is well-stirred. In this experiment, I only measured the rock mineralogy (XRD), but in the next experiment, I will also measure the water composition. When defining the core constraints, I assigned small values to all the aqueous phases. Could you please confirm if this approach is correct?

Additionally, do you think the physics I defined in the uploaded input file is accurate? Glenn mentioned it seems correct, but you had suggested the Dual Continuum model.

The results from the code show a very low change in calcite volume fraction, while my experimental data indicates significant Portlandite dissolution and calcite precipitation. Could you advise on potential sources of error in my code?

Thank you again for your insights.

Fatemeh


To view this discussion visit https://groups.google.com/d/msgid/pflotran-users/7C98C19D-713B-45A2-A198-715688CDB9CD%40gmail.com.

Fatemeh

unread,
Nov 25, 2024, 8:59:46 AM11/25/24
to pflotra...@googlegroups.com
Dear Peter & Glenn

Just a kind reminder, I really appreciate your support and help.

Bests,

Fatemeh 

Hammond, Glenn E

unread,
Nov 27, 2024, 6:05:41 PM11/27/24
to pflotra...@googlegroups.com

See my comments below in red. Glenn

 

From: pflotra...@googlegroups.com <pflotra...@googlegroups.com> on behalf of Fatemeh <reisi...@gmail.com>
Date: Monday, November 25, 2024 at 5:59 AM
To: pflotra...@googlegroups.com <pflotra...@googlegroups.com>
Subject: Re: [pflotran-users: 8155] Batch Reactor

Dear Peter & Glenn

 

Just a kind reminder, I really appreciate your support and help.

 

Bests,

 

Fatemeh 

 

On Nov 20, 2024, at 23:05, Fatemeh <reisi...@gmail.com> wrote:



Dear Peter,

 

Thank you for your prompt reply.

Yes, that is well-stirred. In this experiment, I only measured the rock mineralogy (XRD), but in the next experiment, I will also measure the water composition. When defining the core constraints, I assigned small values to all the aqueous phases. Could you please confirm if this approach is correct?

I don’t know that there is an answer as to what is correct. Assigning low concentrations would be correct if you were using DI water.

Additionally, do you think the physics I defined in the uploaded input file is accurate? Glenn mentioned it seems correct, but you had suggested the Dual Continuum model.

I would avoid using the dual continuum model until after you have a single continuum running robustly.

The results from the code show a very low change in calcite volume fraction, while my experimental data indicates significant Portlandite dissolution and calcite precipitation. Could you advise on potential sources of error in my code?

There are many possible explanations for the discrepancies:

  • Incorrect kinetic rate constants for minerals (too low)
  • Incorrect specific surface areas for minerals
  • Missing aqueous and mineral species
Reply all
Reply to author
Forward
0 new messages