Skip to content

Add SPE 2 dataset#7

Open
Mechtanium wants to merge 1 commit into
OPM:masterfrom
Mechtanium:Add-SPE-2-dataset
Open

Add SPE 2 dataset#7
Mechtanium wants to merge 1 commit into
OPM:masterfrom
Mechtanium:Add-SPE-2-dataset

Conversation

@Mechtanium

Copy link
Copy Markdown

In this update,

I've added the long missing SPE 2 dataset: tested on the latest OPM Flow simulator (release 2025.10), SPE-2 adds a critical piece to the conversation of simulator design around the subjects of cusping and coning in near-well regions which will find new use in the new age of Neural Network inverse reservoir simulation studies.

It's my hope that many researchers find this as useful as I found it, necessitating the work to modernize and restore it.

@alfbr

alfbr commented Jun 22, 2026

Copy link
Copy Markdown
Member

Thanks for the contribution! This is highly appreciated and I will do the review. We do need a copyright header with license on the DATA and INC file, just look at one of the other data sets for example. Please also post the pull request to opm-tests repository instead, as that is where the activity on the data sets are today.

@atgeirr

atgeirr commented Jun 24, 2026

Copy link
Copy Markdown
Member

Please also post the pull request to opm-tests repository instead

... or in addition to here! The opm-data repo is unencumbered by test result data and is therefore leaner and smaller, which is useful. I agree it would be good to have also in opm-tests though!

@bska

bska commented Jun 24, 2026

Copy link
Copy Markdown
Member

This is very welcome and we should get it in. It represents a bit of a dilemma, however, because OPM doesn't really support radial grids (i.e., keyword RADIAL in the RUNSPEC section) and this means, as you've discovered, that the simulator does not know how to calculate the correct transmissibilities. I appreciate the effort to supply transmissibilities through the TRANX and TRANZ keywords in the EDIT section, but it also means we're limited to this exact setup until the simulator gains the ability to support radial grids.

To reiterate though, we definitely should get this in, but to @alfbr and @atgeirr: I think opm-data is the best location right now. We should postpone adding this into opm-tests until we actually support radial grids.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants