Fix RE tests

[dguittet]

Addresses issue:

removes test_h2_valve_opening since the valve is no longer being used in the RE models

Legal Acknowledgement

By contributing to this software project, I agree to the following terms and conditions for my contribution:

1. I agree my contributions are submitted under the license terms described in the LICENSE.md and COPYRIGHT.md file at the top level of this directory.
2. I represent I am authorized to make the contributions and grant the license. If my employer has rights to intellectual property that includes these contributions, I represent that I have received permission to make contributions and grant the required license on behalf of that employer.

[radhakrishnatg]

@dguittet Could you try adding TransformationFactory("network.expand_arcs").apply_to(m) before line 91 and see if this test passes?

[dguittet]

@radhakrishnatg It won't pass when the Transformation is done. There are infeasible constraints. I'm not sure of the value of maintaining this Valve model test since it's no longer being used, and a working valve coefficient is finicky to identify, but I can try playing around with it to see if I can get some values that will solve.

[radhakrishnatg]

@dguittet I think, TransformationFactory is definitely required for completeness. Otherwise, the required equality constraints which ensure that the outlet of the tank and the inlet of the valve the same will not be imposed.
But I agree with your point. It may not be worthwhile maintaining the test, since we are not using the valve model in any of our case studies.

[dguittet]

@ksbeattie @lbianchi-lbl fixes tests that are failing in #177

[lbianchi-lbl]

@dguittet Thanks for the fix. I've noticed that there's still one RE flowsheet test that's failing (from the looks of it, a solver error?):

I'm not sure if this is connected with the changes being introduced in this PR, of if it's a different, unrelated issue that happens to be triggered in the same test_RE_flowsheet.py test file.

[nareshsusarla]

4 of the 5 failing tests are from the FE case study files and have been addressed in #179

[dguittet]

Thanks @lbianchi-lbl and @nareshsusarla. I am looking into that SolverResults error in Github Actions, but can't reproduce it on my Mac or the HPC:

results = {'Problem': [{'Lower bound': -inf, 'Upper bound': inf, 'Number of objectives': 1, 'Number of constraints': 34847, 'Num...Phase Failed.', 'Termination condition': 'internalSolverError', 'Id': 501, 'Error rc': 0, 'Time': 128.07175731658936}]}

https://github.com/gmlc-dispatches/dispatches/actions/runs/4537252316/jobs/7994898724?pr=179

[lbianchi-lbl]

@dguittet thanks for looking into this. For the sake of getting out of the quasi-deadlock of test failures we're currently experiencing, I propose the following:

• Open an issue to track the SolverResults error in test_wind_battery_pem_tank_turb_optimize_simple()
• "Recycle" the infrastructure in Add functionality to xfail all tests affected by known issue #178 #186 to instead xfail test_wind_battery_pem_tank_turb_optimize_simple() instead of the FE USC tests tracked in Failures in USC FE case study tests with IDAES 2.0.0/Pyomo 6.5.0 #178
• Merge Add functionality to xfail all tests affected by known issue #178 #186
• Start merging the other fixes (updates to the FE case models with new Pyomo #179, Fix RE Notebook Check #181, Fix RE tests #185)

At that point, if you manage to fix the test_wind_battery[...] failure, we can then remove the xfail logic.

[dguittet]

@lbianchi-lbl I restored this branch so that it's fixed at least one of the failing tests. To figure out the remaining one, I'm on another branch #188

[dguittet]

@lbianchi-lbl I seem to be making progress on the remaining test. Perhaps I will get it fixed within the hour

[dguittet]

@radhakrishnatg @bknueven @lbianchi-lbl I've resolved both the RE tests failures. Please re-review?

[radhakrishnatg]

Looks good to me @dguittet

[lbianchi-lbl]

@dguittet this is awesome, thanks! I'll go ahead and merge it right away. Once that's done, merging main into #179 (and possibly #181) should hopefully be enough to address all of the failures.