What does globalCompFraction mean in GEOSX? #1457

wdmhouston · 2021-06-10T15:12:19Z

wdmhouston
Jun 10, 2021

Hi expert,

Please find a small CO2 injection test  example:

geosx_co2mass.zip

CO2 was injected at specified surface condition(same as initial reservoir condition, 14432381.756 Pa and 368.15Kelvin) and well is under rate control(0.001m3/s, 5e6 seconds). I'm trying to calculate the total CO2 dissolved in water. I remember Francois mentioned that CompFraction isn't exported to vtk file due to some technical issue of 4d array, but I find globalCompFraction with values. What does this globalCompFraction mean? Can I use this to calculate total dissolved CO2 mass?

In test example, I tried another way to calculate:
Total CO2 injected:
Injection rate* time* density = 0.001(m3/s)* 5e6(s)* 331.17(kg/m3)= 1655850 kg
CO2 mass calculated from paraview:
ElementVolume* ReferencePorosoity* PhaseVolumeFraction_X* Fluid_PhaseMassDensity_X= 978267kg
So the dissolved CO2 in water: = 1655850kg - 978267kg = 677583kg
Is this correct? If no, can you advise me the right method to calculate the dissolved CO2 in GEOSX?

Thanks!

francoishamon · 2021-06-15T18:13:09Z

francoishamon
Jun 15, 2021
Collaborator

Hello Deming,
The simplest method to do what you are trying to do is the following:

Checkout branch bugfix/hamon/4DfieldsInVTK
Recompile
Run your test
In Paraview, you should be able to visualize PhaseCompFraction_1, which stores the phase component fractions (for CO2 and H2O components) in the brine phase. You can select the first component of PhaseCompFraction_1, and that will give you the amount of CO2 component dissolved in brine. Then you can compute the mass of CO2 dissolved in brine as volume x porosity x phaseVolumeFraction_1 x phaseCompFraction_1_0 x phaseMassDensity_1

14 replies

francoishamon Jun 30, 2021
Collaborator

Ok great. If you checkout the latest version of develop, you will be able to output phaseCompFraction to vtk, which may be useful to compute the CO2 mass in the reservoir.

wdmhouston Jun 30, 2021
Author

Will try. Thanks!

wdmhouston Jun 30, 2021
Author

Ok great. If you checkout the latest version of develop, you will be able to output phaseCompFraction to vtk, which may be useful to compute the CO2 mass in the reservoir.

Francois,
In the latest develop version, do I have to change temperature in co2flash, pvtgas, pvtliquid from C to F?

francoishamon Jun 30, 2021
Collaborator

Yes, there was an inconsistency in the units so now the temperature is expected in Kelvin in co2flash, pvtgas, and pvtliquid.

In the base case you posted above, you have defined a temperature of 316.15 K in the solver

<CompositionalMultiphaseFVM discretization="fluidTPFA"
		                            fluidNames="{ fluid }"
		                            logLevel="2"
		                            maxCompFractionChange="0.2"
		                            name="compositionalMultiphaseFlow"
		                            relPermNames="{ relperm }"
		                            solidNames="{ rock }"
		                            targetRegions="{ Reservoir }"
		                            temperature="316.15"
		                            useMass="1"/>

so I suggest to use values consistent with that in the pvt files. For instance, for co2flash.txt you would have:

FlashModel CO2Solubility  1e6 5e7 5e4 315 317 1 0.4278

(for now the temperature is constant, we are are sure that is it will remain equal to 316.16 K everywhere during the entire simulation)

wdmhouston Jul 27, 2021
Author

Francois,

Today, I re-tested the same case with temperature corrected to K unit in the latest develop branch version(see that attached test case files)

CO2MassSimple.zip

and here are the results of CO2 mass calculated from paraview and from injection rate.
calculated from output properties in paraview:
DissolvedCO2 in brine: 9972.36kg
NonDissolvedCO2: 30690kg
Total: 9972.36 + 30690 = 40662.36 kg

calculated from injection rate:
        646.9535807985kg/m3 * 0.00002m3/s * 2592000s = 33538.07 kg.

As we can see, the total CO2 mass doesn't match(no even close).  I've seen other cases with similar trend(over injected). Can you take a look and find out where the issue might be? If it's a manual calculation issue, please let me know which part is wrong.

Thanks.

--

Yes, there was an inconsistency in the units so now the temperature is expected in Kelvin in co2flash, pvtgas, and pvtliquid.

In the base case you posted above, you have defined a temperature of 316.15 K in the solver
<CompositionalMultiphaseFVM discretization="fluidTPFA"
		                            fluidNames="{ fluid }"
		                            logLevel="2"
		                            maxCompFractionChange="0.2"
		                            name="compositionalMultiphaseFlow"
		                            relPermNames="{ relperm }"
		                            solidNames="{ rock }"
		                            targetRegions="{ Reservoir }"
		                            temperature="316.15"
		                            useMass="1"/>
so I suggest to use values consistent with that in the pvt files. For instance, for co2flash.txt you would have:
FlashModel CO2Solubility  1e6 5e7 5e4 315 317 1 0.4278
(for now the temperature is constant, we are are sure that is it will remain equal to 316.16 K everywhere during the entire simulation)

francoishamon · 2021-08-16T15:13:29Z

francoishamon
Aug 16, 2021
Collaborator

Hello @wdmhouston,
Thanks for the catch, there was indeed a problem in the injection rate set a surface condition.
Now the problem should be fixed once #1536 is merged.

I attached your updated file for which I verified that the injection rate matches the amount of CO2 in the reservoir.
deming.zip

3 replies

wdmhouston Aug 17, 2021
Author

Francois,

Thanks. I tested in this simple case and I got about 32000kg. The calculated injection total is 33538.07. It's very close. Did you get similar results?

I noticed there are some changes. One of the changes is :

This is rock permeability and later it will be overwritten by permxFunc. Is my understanding correct?

Hello @wdmhouston,
Thanks for the catch, there was indeed a problem in the injection rate set a surface condition.
Now the problem should be fixed once #1536 is merged.

I attached your updated file for which I verified that the injection rate matches the amount of CO2 in the reservoir.
deming.zip

francoishamon Aug 17, 2021
Collaborator

Ok, let me try again, I am a little bit surprised by your result. I think I was getting a much better match than that. Are you sure that you are using the final vtk file of the simulation?

francoishamon Aug 17, 2021
Collaborator

@wdmhouston Here is what the simulator actually computes, checked by outputting values to the terminal:

Total co2 mass injected = co2Density * injectionVolRate * totalTime
= 647.006 kg.m^-3 * 0.00002 m3.s^-1 * 2,592,000 s = 33,541 kg.
Total co2 mass in the reservoir = co2Saturation * co2PhaseDensity + co2InBrinePhaseCompFraction * brineSaturation * brinePhaseDensity = 33,541 kg.

(note: co2InCo2PhaseCompFraction = 1 in our PVT model)

So with the very tight Newton tolerance that I put in the xml I sent you (10^-9), we obtain a very good match.

The problem, if any, must be in the output. Can you please explain again how you grab the values to make your calculations? From the VTK file? At what time? Do you use the hdf5 output at all (I can show you how)?

wdmhouston · 2021-08-17T17:38:52Z

wdmhouston
Aug 17, 2021
Author

Francois, That is what I expected. Let me try again. Regards, Deming

…

On Tue, Aug 17, 2021 at 12:27 PM Francois Hamon ***@***.***> wrote: @wdmhouston <https://github.com/wdmhouston> Here is what the simulator actually computes, checked by outputting values to the terminal: - Total co2 mass injected = co2Density * injectionVolRate * totalTime = 647.006 kg.m^-3 * 0.00002 m3.s^-1 * 2592000 s = 33,541 kg. - Total co2 mass in the reservoir = co2Saturation * co2PhaseDensity + co2InBrinePhaseCompFraction * brineSaturation * brinePhaseDensity = = 33,541 kg. (note: co2InCo2PhaseCompFraction = 1 in our PVT model) So with the very tight Newton tolerance that I put in the xml I sent you (10^-9), we obtain a very good match. The problem, if any, must be in the output. Can you please explain again how you grab the values to make your calculations? From the VTK file? At what time? Do you use the hdf5 output at all (I can show you how)? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#1457 (reply in thread)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AETQFWABQF4QPAWHLKIHDYTT5KLZJANCNFSM46OXKYRA> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email> .

-- Deming Wang ***@***.***

0 replies

wdmhouston · 2021-08-17T21:11:42Z

wdmhouston
Aug 17, 2021
Author

Francois, I found an error in my old calculation(misunderstanding about co2 solubility unit in brine), and corrected it. Now the value is very close. I used that 1E-9 converge criteria and calculated in paraview this way: Calculator: DissolvedCO2 = elementVolume*referencePorosity*phaseVolumeFraction_water*fluid_phaseCompFraction_water/co2*fluid_phaseMassDensity_water Python Calculator: DissolvedCO2Total = sum(DissolvedCO2) = 7949.6 kg CO2: Calculator: CO2 = elementVolume*referencePorosity*phaseVolumeFraction_gas*fluid_phaseCompFraction_gas/co2*fluid_phaseMassDensity_gas Python Calculator: CO2Total = sum(NonDissolvedCO2) = 25403.8 kg Total CO2 in reservoir = 7949.6 kg + 25403.8 kg = 33353.4 kg. It's very close to 33,541 kg. You mentioned you calculated using hdf5 with 100% matched, can you show me how to do it? Regards, Deming

…

On Tue, Aug 17, 2021 at 12:38 PM Deming Wang ***@***.***> wrote: Francois, That is what I expected. Let me try again. Regards, Deming On Tue, Aug 17, 2021 at 12:27 PM Francois Hamon ***@***.***> wrote: > @wdmhouston <https://github.com/wdmhouston> Here is what the simulator > actually computes, checked by outputting values to the terminal: > > - > > Total co2 mass injected = co2Density * injectionVolRate * totalTime > = 647.006 kg.m^-3 * 0.00002 m3.s^-1 * 2592000 s = 33,541 kg. > - > > Total co2 mass in the reservoir = co2Saturation * co2PhaseDensity + > co2InBrinePhaseCompFraction * brineSaturation * brinePhaseDensity = = > 33,541 kg. > > (note: co2InCo2PhaseCompFraction = 1 in our PVT model) > > So with the very tight Newton tolerance that I put in the xml I sent you > (10^-9), we obtain a very good match. > > The problem, if any, must be in the output. Can you please explain again > how you grab the values to make your calculations? From the VTK file? At > what time? Do you use the hdf5 output at all (I can show you how)? > > — > You are receiving this because you were mentioned. > Reply to this email directly, view it on GitHub > <#1457 (reply in thread)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/AETQFWABQF4QPAWHLKIHDYTT5KLZJANCNFSM46OXKYRA> > . > Triage notifications on the go with GitHub Mobile for iOS > <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> > or Android > <https://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email> > . > -- Deming Wang ***@***.***

-- Deming Wang ***@***.***

0 replies

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

What does globalCompFraction mean in GEOSX? #1457

{{title}}

{{editor}}'s edit

{{editor}}'s edit

Replies: 4 comments 17 replies

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

{{title}}

{{title}}

{{editor}}'s edit

{{editor}}'s edit

Select a reply

What does globalCompFraction mean in GEOSX? #1457

wdmhouston Jun 10, 2021

Replies: 4 comments · 17 replies

francoishamon Jun 15, 2021 Collaborator

francoishamon Jun 30, 2021 Collaborator

wdmhouston Jun 30, 2021 Author

wdmhouston Jun 30, 2021 Author

francoishamon Jun 30, 2021 Collaborator

wdmhouston Jul 27, 2021 Author

francoishamon Aug 16, 2021 Collaborator

wdmhouston Aug 17, 2021 Author

francoishamon Aug 17, 2021 Collaborator

francoishamon Aug 17, 2021 Collaborator

wdmhouston Aug 17, 2021 Author

wdmhouston Aug 17, 2021 Author

wdmhouston
Jun 10, 2021

Replies: 4 comments 17 replies

francoishamon
Jun 15, 2021
Collaborator

francoishamon Jun 30, 2021
Collaborator

wdmhouston Jun 30, 2021
Author

wdmhouston Jun 30, 2021
Author

francoishamon Jun 30, 2021
Collaborator

wdmhouston Jul 27, 2021
Author

francoishamon
Aug 16, 2021
Collaborator

wdmhouston Aug 17, 2021
Author

francoishamon Aug 17, 2021
Collaborator

francoishamon Aug 17, 2021
Collaborator

wdmhouston
Aug 17, 2021
Author

wdmhouston
Aug 17, 2021
Author