Dynamic fuel switching #121

[urwahah]

For #121:

• Added a new fuel switching input "Use optimal heating fuel" under the backup heating equipment selection
• Added new scenarios and scenario group to compare fuel switching
• Added new columns to the loads_to_site dataframe:
  • FUEL_SWITCHING, HHW_REM_W_NG_MODE, BOILER_HHW_W_NG_MODE, GAS_BOILER_WH_NG_MODE, intermediate columns to store the "natural gas mode" load/energy use that it reverts to if gas emissions < electric emissions
  • BOILER_EQ_CALC, BOILER_CAP_W to store the scenario's boiler sizing information as it's used in the emissions function
• Added fuel switching calculation in site_to_source
• Moved the boiler sizing calculation to site_to_source

Here is a link to the change request to update docs for this.

A few minor changes unrelated to the fuel switching:

• Updated the AWHP sizing logic to use the remaining HHW/CHW loads instead of the original load profile. For heat recovery scenarios this will allow the AWHP to be downsized. (This is the change causing the merge conflicts, I think all the new changes should be accepted but wasn't sure if me doing that would also merge the PR)
• Added a 60% efficiency boiler

[t-kramer]

Hi @urwahah, great work as always.

I added some minor comments/questions that should be pretty easy handle. Let me know if you have questions. When these are resolved, I'll go ahead with the merge and handle the merge conflicts.

[t-kramer]

For this block, I guess AWHP_COP_H can be 0 for hours when AWHP isn't operating? That would produce inf and inf > emissions_intensity_NG is True. Wouldn't that incorrectly trigger NG mode even when the AWHP wasn't running? Please double check, I don't know if I am right on this one. You could use AWHP_HHW_W > 0 is a condition.

[urwahah]

AWHP_COP is never 0, _capacity_constraints overrides the capacity to 0 for OATs outside the operating constraints, which prevents it from operating - but the COP is just interpolated once (_per_unit_heating_cop) and returns the min/max value when OAT is outside the curve. also even if COP were 0 and returned an infinite value, we would want the boiler to operate to serve the loads

[t-kramer]

Can you double-check that the change you made in this line is correct? I just want to make sure we don't sum up total_emissions_kg across all scenarios.

[urwahah]

yep - that's what it was doing before - final_df is the results from the scenario currently being processed, and is appended to results iteratively. this will still print total emissions across all equipment scenarios, but i don't think it's needed to print those separately for a quick check

[t-kramer]

This and previous version are mathematically equivalent only if the boiler always serves exactly the remaining load, which should be true, right? Can you check?

[urwahah]

yes, have spot checked some results and they are equivalent. the reason to change this is that my understanding is there should never be a scenario in which we reach this stage and the boiler doesn't serve exactly all the remaining HHW load. there is a logger warning in case both gas and electric resistance operate (line 904) - but that warning doesn't really do anything until we make this change because currently the remaining load is manually set to 0 before that phase.

[urwahah]

@t-kramer - updated based on your review, i left open a few comments with responses.
@praftery - could you take a look at the new scenario group before we merge? looking back on it i'm not 100% sure i have reasonable ones based on what we discussed a few weeks ago.