search

OPM / opm-simulators

Simulator programs and utilities for automatic differentiation.
http://www.opm-project.org
GNU General Public License v3.0
111 stars 121 forks source link

FPR comparison between OPM and Eclipse #1663

Open ghost opened 5 years ago

ghost commented 5 years ago

Hello. I'm running the same simulation deck on Flow (2018.10) and Eclipse (2016.2) to compare/verify the results. Both runs predict same the FOPT, FLPT, FWIT, WBHP and so on, expect for average pressure for the field. Is this because of the method used by simulators for averaging the pressures is different ?

fopt flpt fwit wbhp_producer wbhp_injector fpr

joakim-hove commented 5 years ago

Could you also try adding the block pressure in some cells?

ghost commented 5 years ago

I'm trying to load the EGRID files to ResInsight but it keeps crashing when loading the case run with Flow. I only have cell pressure vs time for Eclipse case. I can add them but dont think it will help without anything to compare against.

joakim-hove commented 5 years ago

I'm trying to load the EGRID files to ResInsight but it keeps crashing when loading the case run with Flow.

Hmmm - that was not good; care to share the EGRID file?

ghost commented 5 years ago

Unfortunately, the model is confidential.

atgeirr commented 5 years ago

Could you add block pressure output to the SUMMARY section and rerun the model to get that output? Even if you can just run a few timesteps we can quickly tell if there is a difference, or if the difference is due to a bug in the averaging.

We had some discussion earlier about pore-volume weighted average pressure vs. hydrocarbon-volume weighted average pressure. Perhaps we have made a mistake here. Does the model have a large water zone and/or high connate water saturation?

joakim-hove commented 5 years ago

Unfortunately, the model is confidential.

OK - that was unfortunate; but maybe there is a backtrace/error message from Resinsight?

ghost commented 5 years ago

Could you add block pressure output to the SUMMARY section and rerun the model to get that output?

Sure, I'll do that and report back.

Does the model have a large water zone and/or high connate water saturation?

SWL is set to 0.2 with no aquifer support

ghost commented 5 years ago

but maybe there is a backtrace/error message from Resinsight?

It crashes when Reading Result index and quits. No error msg.

ghost commented 5 years ago

Could you add block pressure output to the SUMMARY section and rerun the model to get that output?

Here is how sample cell pressure comparison looks like: bpr

atgeirr commented 5 years ago

Here is how sample cell pressure comparison looks like:

Thanks! While no exact match, it shows that the difference must be in the averaging process. I won't be able to address this until next week at least, perhaps someone else has time to fix this?

totto82 commented 5 years ago

Did #1861 fix your issue @subhis? If FPRP was used in combination with FPR, the FPR output was overwritten by FPRP before #1861.