Implement IEER calculation through interpolation for multistage units

[wanhanlong1130]

Add new IEER calculation and tests.

[lymereJ]

Thanks! I think we need to make a few changes.

[lymereJ]

Should it be plural? compressor_stages?

[wanhanlong1130]

Sure. Updated.

[lymereJ]

I'm thinking that we probably either need to 1) check that they are provided in increasing order, or 2) sort them ourselves. For instance if a user provide [0.5, 0.3, 0.8] we would probably need it to be [0.3, 0.5, 0.8] for the rest of the code execution.

[wanhanlong1130]

Sure! It's not difficult to handle. I will update.

[lymereJ]

What if we have a three-stage unit and stage 2 (out of 3) is = self.compressor_stage_input?

You probably need something like that, just an example:

interpolation = False
for stage_id, capacity_ratio  in enumerate(self.compressor_stage):
  if stage_id + 1 < len(self.compressor_stage):
    if self.compressor_stage[stage_id + 1] > reduced_plr[red_cap_num] > capacity_ratio:
      interpolation = True
    end
  end
end
    
  

[wanhanlong1130]

did you mean we need to change > to >=? Not quite understand the logic difference between your approaches here with the one above.

[lymereJ]

In the current proposed logic you seem ton compare the required part load ratio to the minimum and maximum compressor ratio. That works if you only have two stages, but if a particular case has three stages you need to know if you need to interpolate between stage 1 and 2, or between 2 and 3. What I proposed above should do that, at least that's the idea.

[lymereJ]

We need a better docstring here: we need to document the arguments, provide a description, and document what is returned by the function. The formatting needs to be consistent with the other functions for the documentation to be generated.

[wanhanlong1130]

Will add! I thought it was inner function so didn't write it.

[lymereJ]

It might be good to explicitly define the variable name. Maybe something like lower_stage_load, and current_stage_load?

[wanhanlong1130]

Sure I will revise the names for 'values' to be more detailed.

[lymereJ]

If you look at example G4.3 in AHRI 340/360, the interpolation is based on the actual percent load, not just capacity ratio. I think that they're different.

If you have two stages: 100% and 50% of the rated (net) capacity. When you calculate the 75% EER at 81.5F OA DB, I think that you need to 1) calculate the actual net capacity of the unit for the larger stage and at the lower stage, 2) calculate and actual percent load where for the larger stage it would be the net capacity calculated in 1) divided by the rated net capacity, and for the lower capacity it would be the net capacity calculated in 1) divided by the rated net capacity * 0.5 (since the second stage is 50% of the rated net capacity. These would be the number to use in the interpolation.

[wanhanlong1130]

Did you mean using this load factor, not the EER_reduced at upper/lower stage?

We can discuss more during the meeting.

[lymereJ]

Yes, see the Actual Percent Load column:

[lymereJ]

To make sure that we're following the right approach, it would be nice to reproduce example G4.3 in this test. This would involve using dummy performance curves to get the same "test" results that they show in the example.

[wanhanlong1130]

Sure. I know that. It's just for current testing.

But, my question is, if we want to follow G4.3, do we need to add the curve files first? Should we use Table G12 or something else?

[lymereJ]

You could just create curve objects on-the-fly like here: https://github.com/pnnl/copper/blob/develop/tests/test_unitarydirectexpansion.py#L86-L94