adapt emissions reporting to fixing and restructuring of industry feedstock equations

[fschreyer]

These are the reporting changes required for remindmodel/remind#1829.

[fschreyer]

Uargh, this was again more complicated and annoying than I expected. I receive no more summation errors with the GDXs of my test runs and carbon accounting in feedstocks should be correctly represented.

This should ideally be merged tmrw, I think, because I will be gone after Friday and I still have to do some adjustments to the project mappings after this. My proposal forward would be

1.)

• I will run a last test setting cm_nonPlasticFeedstockEmiShare = 0.6 as proposed.
• @JakobBD: could you merge my REMIND changes into restructure chemical feedstock emissions balancing remindmodel/remind#1829 and check everything again that it is ready to be merged. You would need to install this remind2 branch.
• @fbenke-pik: We will need to merge this PR together with the REMIND PR. Could you replace the test GDX for remind2 with one I am going to provide to you (or let me know how to change it).

2.)

Once the REMIND PR is ready to merge -> we replace the remind2 test GDX -> we merge this PR -> we adapt the REMIND PR to require the new remind2 version -> we merge the REMIND PR

3.)

I will prepare a PR to the mappings in piaminterfaces.

[fschreyer]

also tagging @amerfort to make the CarMa team aware

[fschreyer]

I also wrote some more general explanation on the changes here.

What changes in terms of CDR / carbon management variables?

• non-fossil waste incineration CCS is now also accounted under Emi|CO2|CDR|BECCS and Emi|CO2|CDR|Synthetic Fuels CCS

• non-fossil non-plastics landfilling is now accounted under Emi|CO2|CDR|Materials|+|Non-Plastics

• I also added a tree of Carbon Management|Feedstocks|Emitted/Stored|+|Fossil/Biomass ... variables to track carbon input and output of the feedstocks. This was mainly for diagnostics to verify that we account it correctly and is not necessary to report anywhere. I tested that all waste emissions correspond to the sum emitted fossil and stored negative non-fossil carbon:

Emi|CO2|+|Waste
+ Emi|CO2|Energy|Waste
=
Carbon Management|Feedstocks|Emitted|+|Fossil
- Carbon Management|Feedstocks|Stored|+|Biomass
- Carbon Management|Feedstocks|Stored|+|Synthetic

[JakobBD]

@JakobBD: could you merge my REMIND changes into remindmodel/remind#1829 and check everything again that it is ready to be merged. You would need to install this remind2 branch.

I did. I also merged the current develop branch. Had to rebase, strangely.
Ran with your remind2 branch.
Got errors that seem to be unrelated to the changes here, but someone else messing up REMIND. Will try again in a few days, I guess? Runs are here: /p/tmp/jakobdu/remind_dev2/output/SSP2-NPi_2025-01-28_15.52.56.

[fschreyer]

Got errors that seem to be unrelated to the changes here, but someone else messing up REMIND

Update: Found the issue. A demScen was by accident removed today.
remindmodel/remind#1965
Hope this is fixed tomorrow.

[amerfort]

I didn't spot any flaws and I appreciate all your work.
As we discussed the changes earlier today, I fully support it conceptually. When all tests succeed, I gladly approve the PR.

[JakobBD]

make test and H12 NPi both working now. From my side, ready to do buildLibrary and merge!
See /p/tmp/jakobdu/remind_dev2/output/SSP2-NPi_2025-01-29_12.59.37.
Thanks for all the effort.

[fschreyer]

ok, cool. @fbenke-pik can you upload /p/tmp/jakobdu/remind_dev2/output/SSP2-NPi_2025-01-29_12.59.37 to the RSE repo for the remind2 tests such that I can run buildlibrary?

[fbenke-pik]

ok, cool. @fbenke-pik can you upload /p/tmp/jakobdu/remind_dev2/output/SSP2-NPi_2025-01-29_12.59.37 to the RSE repo for the remind2 tests such that I can run buildlibrary?

Seems like we need two gdx files in a peculiar format.
https://github.com/pik-piam/remind2/blob/master/tests/testthat/test-convGDX2mif.R#L17-L18

@orichters can you let us know what the procedure is to generate test gdxes for the remind2 test case?

Looks like we need to start two different runs to get specific gdxes?

[fschreyer]

Ok, so the tests require a H12 Npi run and a Policy EU21 run. I remember that we introduced this to cover different cases of GDXs that arrive. Generally, that is useful. However, we have to finish this by tmrw and I fear we won't get there if we wait for these runs to finish. So I'd propose a shortcut:

Could we set the tests only to this Npi GDX for now?

@JakobBD Can you set up a EU21 Npi + PkBudg cascade? I would take the fulldata.gdx of the 2nd or 3rd iteration after an hour or so and test against the reporting. If this works, I would built and merge.

We should merge everything today so that I have a chance to adapt the mappings tmrw and respond to potential issues.

Also, Falk could you maybe write down a procedure in remind2 for future people if they want to introduce changes to the library that need new test GDXs?

[orichters]

I'm afraid this is not how the tests are supposed to work They are meant that you have to make sure your changes in the remind2 code are backwards compatible and do not break if someone comes with a gdx made with the latest remind release. So if your code fails the tests (and not only shows some summation warnings which does not let buildLibrary fail), your code should not be merged in my opinion. Sorry for these bad news

…

-- Oliver Richters Research Department III e-mail: ***@***.*** Potsdam Institute for Climate Impact Research Telegraphenberg A 31 P.O. Box 60 12 03 D-14412 Potsdam www.pik-potsdam.de

[fschreyer]

Due to time pressure, I agreed with @fbenke-pik that I will merge now and he will take care of a (technical) backwards compatibility to the release version GDXs. Thanks!