Merge changes from upstream

[tgilon]

Closes # (if applicable).

Changes proposed in this Pull Request

This PR brings the latest changes from PyPSA/PyPSA-Eur. Among other things, this PR introduces the following:

• A FAQ section in the documentation (doc: Add FAQ section for certificate errors PyPSA/pypsa-eur#2228): It is suggested to move this section to Background on PyPSA-Eur to avoid confusion with the Open-TYNDP FAQ.
• A Linopy version pin (override exclude-newer for pypsa/linopy/atlite PyPSA/pypsa-eur#2237)

Merge conflicts:

• doc/release_notes.md
• envs/default_linux-64.pin.txt
• envs/default_osx-64.pin.txt
• envs/default_osx-arm64.pin.txt
• envs/default_win-64.pin.txt
• pixi.lock
• pixi.toml

Tasks

• Integrate PyPSA-Eur FAQ in the documentation
• Decide the Linopy version to pin (see override exclude-newer for pypsa/linopy/atlite PyPSA/pypsa-eur#2237 (comment))

Workflow

Open issues

Notes

Checklist

Required:

• Security scans show no high-severity bugs, critical vulnerabilities, or exposed secrets.
• Changes are tested locally and behave as expected.
• Code and workflow changes are documented.
• A release note entry is added to doc/release_notes.md.
• The description is human-written and any AI-generated content is marked.

If applicable:

• Changes in configuration options are reflected in scripts/lib/validation.
• Changes in configuration options are added to config/test/*.yaml.
• Multiple climate years test passes locally (pixi run -e open-tyndp tyndp-cyears-test).
• For new data sources or versions, these instructions have been followed.
• Open-TYNDP SPDX license header is added to all touched files.
• Module docstrings are added to new Python scripts.
• New rules are documented in the appropriate doc/*.md files.
• Major features are documented in doc/index.md.

[daniel-rdt]

@tgilon
cc: @measrainsey @lisazeyen
I ran a high-resolution test on this PR branch and compared it with the previous run NT-1H-20260703-pr-sort-links from our previous upstream-merge (see details for the config).

Details

NT-cy2009-20260727-upstream-merge:
  tyndp_scenario: NT
  clustering:
    temporal:
      resolution_sector: 1H

  solving:
    options:
      assign_all_duals: true
    solver:
      name: gurobi
      options: gurobi-default
      solver_options:
        gurobi-default:
          BarConvTol: 1.0e-06
          FeasibilityTol: 1.0e-06
          OptimalityTol: 1.0e-06
    partition: big
    mem_mb: 60000
    runtime: 1h

  cba:
    projects: [t4]

    msv_extraction:
      solving:
        options:
          include_objective_constant: false
        solver:
          name: gurobi
          options: gurobi-default
        solver_options:
          gurobi-default:
            BarConvTol: 1.0e-06
            FeasibilityTol: 1.0e-05
            OptimalityTol: 1.0e-06
        partition: big
        mem_mb: 60000
        runtime: 1h

    # Rolling horizon solving settings
    solving:
      options:
        include_objective_constant: false
      partition: big
      mem_mb: 60000
      runtime: 1h

      solver:
        name: highs
        options: highs-simplex

      solver_options:
        highs-simplex:
          primal_feasibility_tolerance: 1.0e-06
          dual_feasibility_tolerance: 1.0e-06

Here are some system-level comparisons first:

Expanded capacity

Link comparisons

Only a few links differ in marginal cost:

Total Opex and objective

Load Shedding

My main feeling is that the results all look plausible and very very similar to before. But tightening numerical tolerances for the new run creates some visible shifts in the results when we look into the details.

Analogue to the previous investigation, I created some figures to visualise the deviations >1% and >1 GWh for the solved SB networks. In the details, you'll also find a scatter plot with all the data points.

Energy balance - AC

Details

Energy balance - H2

Details

Transmission - AC

Details

Transmission - H2

Details

Opex - AC

Details

Opex - H2

Details

[tgilon]

@tgilon cc: @measrainsey @lisazeyen I ran a high-resolution test on this PR branch and compared it with the previous run NT-1H-20260703-pr-sort-links from our previous upstream-merge (see details for the config).

Thank you for the detailed comparisons. I agree with you that the results are reasonable and that the deviations are minor. This is sufficient for me and I'm happy to keep the current versions (PyPSA / Linopy).

[daniel-rdt]

@tgilon cc: @measrainsey @lisazeyen I ran a high-resolution test on this PR branch and compared it with the previous run NT-1H-20260703-pr-sort-links from our previous upstream-merge (see details for the config).

Thank you for the detailed comparisons. I agree with you that the results are reasonable and that the deviations are minor. This is sufficient for me and I'm happy to keep the current versions (PyPSA / Linopy).

Great, then we can move on :) I think we should open a follow up PR though to specify the tolerances for SB and CBA we figured out along the way, don't you think?

[tgilon]

Great, then we can move on :) I think we should open a follow up PR though to specify the tolerances for SB and CBA we figured out along the way, don't you think?

That makes a lot of sense. I'll let you do it in your solver updates PR. Anything left to add to this PR before merging?

[daniel-rdt]

Anything left to add to this PR before merging?

Don't think so. Not many other changes in this merge. Looking good to me.