Discussion Forum

Dear Vikrant,
Your first error is related to the fact that your functions evaluate to 0 on the boundaries of your model. A permeability of zero is not possible in COMSOL. You can solve this by setting the extrapolation to "constant" in the 3 imported functions.
When this is done, you get an error on initialization. This has to do with the fact that your initial conditions are not well defined. I would use a step function to ramp up the in/outlet velocities from 0 at t=0. Next the in and out flow is not equal in your case, if you want to model that, you need a storage model instead of the "fluid and matrix properties" model (add by right click on Darcy's law). Now you need to define the compressibility of the matrix and the fluid to be able to compute the model.
Note that the pressure rises very quickly due to the inflow velocity.

Best regards,
Frank

Many many thanks for your reply Frank.
I try to correct my model as per your instruction, but again my model does not converge.
I am attaching my corrected model here. Please give me valuable instruction.

Regards,
Vikrant

Dear Vikrant,

in your case the inflow does not match the outflow, this is not possible if there is not some sort of storage in your model, eg. the compressibility of the fluid or the rock matrix.
The standard "fluid and matrix porperties" domain condition does not allow for this, but the "storage" model does.
Attached you can find a file with a compressibility in both fluid and rock, this has no issues in the calculation. But you can see that the pressure in the rock slowly decreases.

Best regards,
Frank

Dear Frank sir,
Thanks for knowledge and support.
I have some questions.....
Upper and lower limits in pressure plot(legend) are same why?
I also wanted to add "Solute Transport" model to study the dispersion behavior in the reservoir.
In solute transport model, pore volume factor and bulk density of porous media are required as inputs, but for my case these are not constant because of heterogeneous reservoir, so what value should i put?
Another question is that i have diffusivity (m) value not the diffusivity co-efficient(m^2/s), so what value should i put?
I want to see the Solute concentration vs. Time at outlet i.e. Breakthrough curve and cumulative breakthrough curve at outlet. How to plot these two curve at outlet?

I am attaching my model after adding "Solute Transport" model.

Again many many thanks to continuous help and support me !!!

Best regard,
Vikrant

Dear Vikrant,

Your pressure distribution is not uniform, but the spread is not very big. This lets me look at the permeability values which you have used. COMSOL uses SI units, thus for permeability that is m^2. It might be that you filled in permeability values in mD, but COMSOL now uses them as m^2. You can modify that in the function permx (replace m^2 with mD).

This will show that the velocities you prescribed are not possible, as the pressure will drop dramatically in the first 100 seconds.

Regarding your diffusivity in [m], are you sure that is not the dispersivity?

For the pore volume factor and bulk density, both are dependent on the porosity, so you should be able to write an equation dealing with this.

Good luck with your model.

Best regards,
Frank

Hi,
I am sure about diffusivity because am using another software where i am assigning spatially varying diffusivity.
Do you have any idea about any formula to calculate pore volume factor and bulk density?
I also try to correct my model according to your suggestion and am attaching here after correction.

Dear Vikrant,

You can use the following equations:
:: Wet Bulk density = rho_r*(1-phi)+rho_w*phi
:: pore volume fraction = porosity

For rho_rock you can probably use something like 2650 kg/m^3, depending on the matrix (for pure quartz sandstone 2650 is a fine value).

Regarding your diffusion, note that the dispersion is usually dominant, and is modeled as a velocity dependent diffusion. It might be that the other software uses a different jargon for this. The fact that it is in meters leads me to believe that we deal with dispersion instead. As diffusion is typically very small, you might be able to neglect that as a whole (choose a small value).

Please note that your pumping velocities are still quite high, and your pressure drops well below vacuum (-1200 bar) in your production well. I suggest using a pressure on the wells instead of a velocity.

Good luck with your model.
Best regards,
Frank

Hi sir,
I try to correct my model, but am wondering about concentration plot and break through curve(BTC) i.e. amount of solute concentration at outlet in water flux.
You mention a formula to calculate wet Bulk density, but what are the meaning of 'rho_r and rho_w' in the formula.
Again i am attaching my corrected model here.

Best regards,
Vikrant

Dear Vikrant,

for the BTC, you evaluated the concentrations at the four boundaries of the outflow circle. This is why the plot looks strange. This can be fixed e.g. by creating a boundary probe to integrate c over the circumference.

About your second question: Wet bulk density = (mass of soil plus liquids)/ (volume as a whole).

Good luck with your model.


--
Best Regards
Phillip

Hi sir,
I don't know, how to fix the BTC problem. Could you please do for me?

Best regards,
Vikrant

Hi sir,
I don't know, how to fix the BTC problem. Could you please do for me?

Best regards,
Vikrant

I recommend you to read the help entries about (boundary) probes or to search for "Boundary Probe" in the COMSOL reference manual. Be aware about the fact that you have to take into account the complete concentration distribution around the pumping "circle".
If you understand the principle, you will find out quickly how to solve your problem.


--
Best Regards
Phillip

Hi sir,
I don't know, how to fix the BTC problem. Could you please do for me?

Best regards,
Vikrant

I recommend you to read the help entries about (boundary) probes or to search for "Boundary Probe" in the COMSOL reference manual. Be aware about the fact that you have to take into account the complete concentration distribution around the pumping "circle".
If you understand the principle, you will find out quickly how to solve your problem.


--
Best Regards
Phillip

How to create fractured vuggy media model
One injection and one production well of water flooding.
Best regards,
Zafar Abro

Sir Philip,
I am new user
Best regards,
Zafar abro