Script for CO2 containment

[AudunSektnanNR]

Script for calculating the amount of CO2 inside and outside a given perimeter, and separates the result per formation and phase (gas/dissolved). Output is a table on CSV format.

Can take both a containment polygon and a "hazardous" polygon. Result is divided into the regions inside vs outside vs hazardous, where inside means inside the containment polygon.

Can calculate either CO2 mass or a representative CO2 volume (three different calculation methods for volume). CO2 mass is the most used and tested.

The resulting CSV-file can be used as data in the CO2Leakage webviz plugin.

See (old) issue: #502

[alifbe]

Please fix the order of import (using isort or vscode)

[AudunSektnanNR]

Have fixed some of the PR issues. Still some error from mypy-checks. Looking into that.

[codecov-commenter]

Codecov Report

Merging #568 (d601a66) into main (68957a2) will decrease coverage by 1.46%.
The diff coverage is 62.36%.

❗ Current head d601a66 differs from pull request most recent head 691a0ca. Consider uploading reports for the commit 691a0ca to get more accurate results

@@            Coverage Diff             @@
##             main     #568      +/-   ##
==========================================
- Coverage   86.16%   84.71%   -1.46%     
==========================================
  Files          49       52       +3     
  Lines        7035     7492     +457     
==========================================
+ Hits         6062     6347     +285     
- Misses        973     1145     +172     

Impacted Files	Coverage Δ
src/subscript/co2_containment/co2_containment.py	44.06% <44.06%> (ø)
src/subscript/co2_containment/co2_calculation.py	67.10% <67.10%> (ø)
src/subscript/co2_containment/calculate.py	82.85% <82.85%> (ø)

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[mferrera]

Looking good! 😄 Thanks a lot for adapting to the linters, a bit of a gauntlet to be sure.

My last concern is with the test coverage -- it becomes our third outlier at significantly below 80% coverage and more importantly we lack the CO2 domain knowledge that may prevent us from addressing issues ourselves. I guess normally coverage is the sort of thing that increases over time as users do unexpected user things, but because we are doing a change of hands it is a bit more of a worry.

Being aware that you have a time runway to work with, do you think there are any critical or highly domain specific methods that could potentially add to long-term robustness?

[AudunSektnanNR]

When it comes to test coverage, I think the best thing would be to have a synthetic data set (EGRID, UNRST, INIT files) with all the required properties used for the co2 calculations. Then it would be easier to test a larger part of the code. But this is probably a bit of work to get done, so realistically this will have to be added sometime in the future I think.

I added a few more tests, so the code coverage will hopefully increase slightly.

[mferrera]

Thanks very much for considering the comments!

We will need to have a quick consensus chat on our end before taking it in so it might be a short while longer -- early next week at the very latest.

I will add co2_containment as the script invocation unless you have some other preferred command line name (doesn't make sense to add it in this PR since it will be moved in #557)

[AudunSektnanNR]

Sounds good! Thank you for all the feedback and fast responses, I think it improved the code.

The proposed script invocation is fine!

[alifbe]

LGTM 👍