feat: scale hydro stub branches

[danielolsen]

Purpose

Never again have to manually scale stub branches to deal with hydro. This is in conjunction with #162, to get accurate Pmax values for hydro plants into the base grid.

What is the code doing

One line change, adding 'hydro' to the list of renewable plant types to perform renewable stub scaling on.

Time to review

5 minutes.

[rouille]

Indeed.

[danielolsen]

I relaxed the Pmax = 0 check because it was complaining about a few hydro plants with 0 capacity in Montana Eastern. Here are the new warnings:

hydro plant 10378 at bus 69975 has 0 Pmax!
hydro plant 10379 at bus 69976 has 0 Pmax!
hydro plant 10380 at bus 69978 has 0 Pmax!
hydro plant 10381 at bus 69978 has 0 Pmax!
hydro plant 10382 at bus 69979 has 0 Pmax!

[danielolsen]

@BainanXia, I refactored the warning logic a bit in this last commit, now it should only warn if scaling for a given plant == 1 when considering both plant_id scaling and zone_id scaling. Does this make the logic more clear? Are there other places where you think I should update the documentation?

[BainanXia]

Reading through the code to understand the logic. Is there a specific reason that we are not using if/else instead of try, assert, except AssertionError?

[danielolsen]

...only because it was an assert before, and this was a quick fix. I agree, your proposal is cleaner, I will change it.

[danielolsen]

Done

[BainanXia]

I'm trying to understand whether this is a common situation or not, in other words, how often is a plant of type 'r' not being scaled neither via 'zone_id' nor 'plant_id', given you are looping through all the plants of such type.

[danielolsen]

I guess that depends on the scenario in question. In our Eastern 2030 Independent OB1, when we are loading capacities from a change table and then turning off offshore_wind, we have scaling factors for all plants/types. But I can imagine other scenarios which would not have scale factors for all types, where we would maybe still run renewable stub scaling.

[BainanXia]

Then in such scenario we are going to have a lot of such warnings for each plant of 'r'+'zone_id'.

[danielolsen]

Maybe it makes more sense to turn verbose off by default instead of on? The user can always specify from the change_table.scale_renewable_stubs() call whether they want these warnings or not.

[BainanXia]

Agree. Otherwise user like me would be anxious in seeing so many warnings and suspecting something seriously is going wrong.

[danielolsen]

Done, with an update to the docstring as well.

[BainanXia]

Let me confirm with you regarding the logic

I relaxed the Pmax = 0 check because it was complaining about a few hydro plants with 0 capacity in Montana Eastern. Here are the new warnings:

Does this happen after we reset the Pmax of hydro generators according to the maximum of the corresponding profiles, which implies we have a bunch of 'dummy' hydro plants with zero capacities and all zero profiles?

[danielolsen]

Does this happen after we reset the Pmax of hydro generators according to the maximum of the corresponding profiles?

Unfortunately, yes. I think all these generators were scaled down to zero when we rebased from our Eastern cost curve tuning process, so the Pmax is 0 and the profiles are [0, 0, 0, ...].

[BainanXia]

@danielolsen I don't remember we ever turned any hydro plants from TAMU network down before. Is it because they shouldn't there according to the external references and we turn them off by scaling capacities of those plants down to zero? Or, they come with zero capacity generators from TAMU network originally.

[danielolsen]

@BainanXia when we were running cost curve tests, we scaled plants based on excel spreadsheets, and the entry for hydro in Montana was 0. Results/Scenarios/ScenarioRuns/EasternCostCurveTests/Eastern_generator_capacity_2016_322.xlsx

Then we rebased the grid based on one of those scenarios.

[BainanXia]

@danielolsen I see. RIP for those hydro plants. It is what it is. Probably in the future when we are cleaning up the grid, we could erase them together with those dummy branches.

[BainanXia]

Thanks for walking through the logic.