Solvent model in 2 phase run does not run

OPM / opm-simulators

OPM Flow and experimental simulators, including components such as well models etc.

http://www.opm-project.org

GNU General Public License v3.0

121 stars 121 forks source link

Solvent model in 2 phase run does not run #4548

Open EdmundStephens opened 1 year ago

EdmundStephens commented 1 year ago

[2022.10] The attached run is modelling CCS in a depleted gas using the black oil SOLVENT method. (It is a 41 x 1 x 50 layer 2D secton based on an outcrop we are working with). I was able to test it in Eclipse and runs nicely albeit with tight tolerances and 1 day timesteps. But Flow immediately stops. Can you check what exactly is going wrong?

Covesea1C_DG_CCS.zip Covesea1C_DG_CCS-F.zip

Thank you!

totto82 commented 1 year ago

Thanks for reporting the issue. The files you shared does not contain the include files which makes it hard to debug locally. Are you able to reproduce the issue on a model you can share with us?

EdmundStephens commented 1 year ago

Thanks, Attached this example with includes. Covesea1C_DG_CCS (flat).zip Covesea1C_DG_CCS-F (flat).zip

EdmundStephens commented 1 year ago

Also for your consideration, the same model with a dummy oil phase .... this again runs in Eclipse but not in Flow.

Covesea1C_DG_CCS3 (flat).zip Covesea1C_DG_CCS3-F (flat).zip

The output error:

===============Saturation Functions Diagnostics===============

System:  Solvent model.
Relative permeability input format: Saturation Family II.
Number of saturation regions: 5

Program threw an exception: TableContainer - invalid tableNumber

Quite likely here I have got one of the table sets wrongly enumerated, as there are 5 facies (NTSFUNC) ... but I cannot see what. Eclipse can often be more lenient with defaults etc. which in my opinion can be quite dangerous in making wrong input apparently work!

totto82 commented 1 year ago

Thanks for the cases. Flow does not support gas+water+solvent at the moment, but it has been on my TODO list for some time. In principle it shouldn't be too hard to add, but will require some debugging to find places where 3p is hardcoded. I will report back when I have a working version. I haven't looked at the dummy-oil version yet. Maybe you could try to run a case with only 1 region first?

EdmundStephens commented 1 year ago

I can confirm the dummy oil phase version works when NTSFUN=1.

DG_CCS3-V1_F: NTSFUN=5 with 5 SATNUMs assigned --> fails
DG_CCS3-V2_F: NTSFUN=1 with 1 SATNUM assigned --> success
DG_CCS3-V3_F: NTSFUN=5 with 1 SATNUM assigned --> fails

Models & includes: DG_CCS3-V1_F.zip DG_CCS3-V2_F.zip DG_CCS3-V3_F.zip Covesea1_includes.zip