Bug: Results drift in docs example notebooks

[RHammond2]

Nonreplicable Results in Documentation Build

While working on #75, I was converting the Jupyter notebooks in the documentation to executable markdown files, I found that there was an uncaught behavioral change resulting in nonmatching results when it was fixed. Replacing the "finance_subgroup_default.LCOH" in lcoh = h2i_model.model.get_val("finance_subgroup_default.LCOH", units="USD/kg")[0] with "finance_subgroup_hydrogen.LCOH_produced_profast_model" allows the notebook to run without failure, but also changes the outputs signficantly . Specifically, in the results processing, I found the output figure in https://github.com/NREL/H2Integrate/blob/main/docs/user_guide/how_to_set_up_an_analysis.ipynb produces a horizontal line of scatter points, indicating no change in results as the inputs are varied.

How to reproduce

1. Checkout my branch: https://github.com/RHammond2/H2Integrate/blob/docs/convert-notebooks/
2. Create an environment with the develop dependency add-ons
3. Build the docs sh docs/build_book.sh
4. Open the locally built documentation home page that is presented at the end of the documentation build, and navigate to: <your-file-path>/H2Integrate/docs/_build/html/user_guide/how_to_set_up_an_analysis.html

Relevant output

The actual hydrogen produced differs in the section Modifying and rerunning the analysis: 56631148.67 kg/year vs. 51724448.11 kg/year. The resulting plots are shown below.

Original:

Updated:

System Information

• OS: macOS 12.4
• Python version: 3.13.3
• H2Integrate version: 0.5.1, branched off 8f40f29 (latest commit to develop)
  • Installed from source using an editable installation with developer tools: pip install -e .[all]
  • Commit hash: 4ba397b

Relevant library versions

Package	Version
hopp	3.4.0
numpy	2.4.1
Pyomo	6.9.5
scipy	1.17.0