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

OPM Flow Not Writing SWATINIT and Endpoints Arrays to INIT File (2018-10) #1738

Open OPMUSER opened 5 years ago

OPMUSER commented 5 years ago

OPM Flow version 2018-10 is not writing out the SWATINIT array, or the endpoint scaling arrays to the INIT file even if:

--
--       ACTIVATES SATURATION END-POINT EXPORT TO THE INIT FILE
--
FILLEPS

is in the input deck.

Is this a known issue?

bska commented 5 years ago

not writing out the SWATINIT array, or the endpoint scaling arrays to the INIT file [...] Is this a known issue?

Yes, I'm afraid so. OPM Flow's INIT file is missing a lot of features that would be present in ECLIPSE's result files.

OPMUSER commented 5 years ago

Okay, it would be really nice if we can write the SWATINIT and endpoint scaling arrays out, as many users use the endpoint scaling option with the SWATINIT array, including the Norne model. I'm not a C++ programmer, but if we can write the PORO and other arrays, it should be straight forward to write out any input array data. But like I said I'm not a C++ programmer.

Inconsistent endpoint scaling data can cause numerical issues and are difficult to identify without examining the arrays and resulting the de-normalized relative permeability curves (see #1725)

bska commented 5 years ago

Quick follow-up here: We're still missing the SWATINIT array, and I do believe we don't quite handle TOLCRIT which affects the *CR end-points, but as of PR OPM/opm-common#849 OPM Flow does at least output the scaled end-points to its .INIT file. This will hopefully make diagnosing these issues easier since we now also output enough information in the LOGIHEAD vector to allow ResInsight's relative permeability plot window to distinguish between the scaled and the unscaled saturation function curves in individual cells.