Open mauriciolmatos opened 5 years ago
joakim-hove
commented
5 years ago From flow you can only get compressed natural order - is it possible to request both forms from Eclipse?
Given the compressed firm it is quite simple to create a full version with some programming/scripting - what is your use case?
mauriciolmatos
commented
5 years ago The ECLIPSE reference manual says there is a way to generate these files in "uncompressed natural order" using the keyword PETOPTS-NOCOMPIN. But I did not try. I do not have ECLIPSE300. I'm working with data from Norne field. It has a total of 113344 cells with only 44441 active. I would like the results of the .FINIT(PORO, NTG, SATNUM) and .FUNRST(PRESSURE, SWAT, SGAS, Rs...) arrays for all cells.
OPMUSER
commented
5 years ago No there is no way to do this in OPM Flow and the commercial simulator; however, there is a work around.
You can load the model into OPM ResInsight, then select the property to display, then right click the Cell Result (under Project Tree), and then export the property. This will give you the data you want and looks like this:
-- Exported from ResInsight
PRESSURE
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
268.897 268.39 268.268 268.204 268.145
268.094 268.078 268.064 268.026 267.985
267.95 267.915 267.879 267.839 267.771
267.692 267.634 267.588 267.549 267.531
267.51 267.449 267.33 267.207 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 269.196 268.449 268.325 268.252
268.192 268.132 268.116 268.122 268.081
268.015 267.976 267.946 267.921 267.897
267.826 267.724 267.653 267.601 267.562
267.552 267.524 267.446 267.324 267.201
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0 This works for all arrays.
Unfortunately, this could be time consuming if you need to do this for many arrays and for several runs. However, OPM ResInsight has an interface to OCTAVE (an open source version of MATHLAB) so you can code this functionality in an OCTAVE script in OPM ResInsight. If you are going to manipulate the array data then you can do all this in OCTAVE script in OPM ResInsight and then display any new arrays you may have created.
joakim-hove
commented
5 years ago I would like the results of the .FINIT(PORO, NTG, SATNUM) and .FUNRST(PRESSURE, SWAT, SGAS, Rs...) arrays for all cells.
yes - I understand that; my question was more why you wanted that - and exactly what you wanted to achieve. I do not know how you are working practically - but you should script/program this in some way. As mentioned in the comment above you can use Resinsight and the Octave plugins, personally I would use the Python wrappers from libecl. If you head over to: https://github.com/equinor/libecl and describe what you want to achieve - I or someone else can probably give you a small Python script to get you going on what it is you want.
When it comes to output with corner values for the grid - my answer is again the same, use a Python script - load the EclGrid instance - then you can ask for corner coordinates.OPM does not output the .GRID file (only .EGRID).
If you use Python wrappers as suggested - or the Resinsight + Octave approach you can just use the binary files .EGRID / .INIT and forget about the ASCII formatted versions.
mauriciolmatos
commented
5 years ago Thanks for the suggestion OPMUSER. This is a good suggestion for a simple simulation. Unfortunately for many time steps it will take a lot of time. I'll learn about Octave-ResInsight.
@joakim-hove thanks for the support. We want to use the OPM simulation outputs as inputs in petrophysical simulations. Therefore, we need the properties in all cells. If you can give us a script that can help us, we will be very grateful. Anyway, I'll be working on a script to help us.
About the format of the output files (formatted or binary), I noticed that the .FEGRID file is in binary when it should be "formatted" (.FINIT and .FUNRST are "formatted"). Will there be a new version of OPM this month? Can you tell me if this bug will be corrected?
Thank you very much.
OPMUSER
commented
5 years ago Maurício,
You can use the the command line function of ResInsight to export property data instead of using an Octave script - see the attached script file for an example. This will produce one file per property per time step and the output looks like this.
-- Exported from ResInsight SOIL
0.895 2.23517e-08 2.23517e-08 2.23517e-08 2.23517e-08
2.23517e-08 2.23517e-08 -1.49012e-08 -1.49012e-08 -1.49012e-08
-1.49012e-08 -1.49012e-08 -1.49012e-08 -1.49012e-08 2.98023e-08
2.98023e-08 2.98023e-08 2.98023e-08 2.98023e-08 2.98023e-08
2.98023e-08 2.98023e-08 2.98023e-08 2.98023e-08 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0 0 0 0
0 0.895 2.23517e-08 2.23517e-08 2.23517e-08
2.23517e-08 2.23517e-08 2.23517e-08 -1.49012e-08 -1.49012e-08
-1.49012e-08 -1.49012e-08 -1.49012e-08 -1.49012e-08 2.98023e-08
2.98023e-08 2.98023e-08 2.98023e-08 2.98023e-08 2.98023e-08
2.98023e-08 2.98023e-08 2.98023e-08 2.98023e-08 2.98023e-08 So this will give you what you want.
If you wish to manipulate the property data you can do this using an Octave script to create new arrays etc. Once you have the script working you can call the Octave script using the runOctaveScript command line command in your batch script. So you can virtually automate this process through the command line script.
Yes the developer's are aware of the *.FEGRID bug and it is fixed in 2019-04-RC1. The plan is to release a new version this month.
DOS-ResInsight-Export-SOIL.zip
OPM User
joakim-hove
commented
5 years ago Therefore, we need the properties in all cells. If you can give us a script that can help us, we will be very grateful
Hmmm - I must admit I am bit skeptical; if you really need the numerical value for all cells you are out of luck - the solution data is only calculated for the active cells - there is no such thing as the value of the pressure in an inactive cell. The solutions proposed in this thread - i.e. Resinsight w/o Octave and also the libecl/Python solution are based on "filling up" the vector with a default value for inactive cells.
OPMUSER
commented
5 years ago @joakim-hove he just wants to manipulate the arrays and needs the arrays to be the same size (DIMENS). None active cells will have zero PORV so one can just skip those cells. It is the dimension of the arrays that was his issue.
mauriciolmatos
commented
5 years ago Thanks for all the suggestions. For now, I have chosen to insert ("filling up") default values of the inactive cells into the arrays of the active cells. Waiting for the 2019.4 release. Thank you.
Good morning everyone.
I'm working with the FINIT and FUNRST files and would like their arrays to all cells in the model (
NX*NY*NZ), called "uncompressed natural order" by ECLIPSE. Currently, the files are presented only to the active cells of the model (NACTIV), called "compressed natural order" by ECLIPSE.I would like to know if there is a way to generate the FINIT AND FUNRST files in the "uncompressed natural order" mode.
Thanks.