enhancement: energy_loss update #11
Conversation
|
Resolves #10 |
There was a problem hiding this comment.
This is an excellent start @dmulash, thanks for updating all the impacted methods!
I left a lot of comments, but they're mostly relating to the use of an electrical losses method, rather than an input. I have a couple of questions in there as well that'd be great to get your input on, especially while we're modifying the code. I also made a handful of smaller changes myself just to reduce some of the back and forth.
|
Hi @RHammond2, Apart from your suggested changes (eliminate Feel free to adjust as needed! |
| @@ -1424,14 +1445,12 @@ def energy_production( | |||
| by: str = "windfarm", | |||
| units: str = "gw", | |||
| per_capacity: str | None = None, | |||
| with_losses: bool = False, | |||
| loss_ratio: float | None = None, | |||
| environmental_loss_ratio: float | None = None, | |||
There was a problem hiding this comment.
It looks like the energy losses weren't applied in this method, which should be done just before the AEP conditional.
There was a problem hiding this comment.
Basically the section you removed should be replaced with something like:
losses = self.energy_losses(...) # pass the parameters that were used for the energy potential basis
production = potential - losses
| @@ -1624,6 +1631,228 @@ def energy_losses( | |||
|
|
|||
| return losses / self.capacity(per_capacity) | |||
|
|
|||
| def wake_losses( | |||
There was a problem hiding this comment.
I realized there was an error in how this was done, which was mostly my fault, so I fixed that. I would double check the remaining losses calculations, and replicate the process in this method to ensure that both the control flow and the calculations are correct.
| @@ -1592,19 +1610,25 @@ def energy_losses( | |||
| by="turbine", | |||
| units="kw", | |||
| ) | |||
|
|
|||
There was a problem hiding this comment.
This section should be roughly in the form of:
if not losses_breakdown:
production = ... # the code that's there already
losses = potentail.copy()
losses["total_losses"] = potential.values - production.values # can use a different column name
else:
# raise an error for if by == "turbine" because detailed breakdowns are only for the wind farm
environmental_lr = self.environmental_loss_ratio()
availability_lr = self.availability(which="energy", by="windfarm", frequency="month-year")
... # the remaining loss ratios
losses = potential.copy
losses = losses.assign(
environmental_losses=environmental_lr,
availability_losses=availability_lr,
... # the remaining lossses
).rename(columns={}) # final step if you think we should have them displayed more nicely
# convert to desired units
# aggregate to apporpriate time scale ("annual", "project")
# apply aep
# apply per_capacityThere was a problem hiding this comment.
From our conversation: only calculate the loss ratio, not the actual losses. (removed from the example above)
waves/project.py
Outdated
|
|
||
| return total_loss_ratio | ||
|
|
||
| def losses_breakdown( |
There was a problem hiding this comment.
This should really be incorporated into the energy_losses method using a new keyword argument called breakdown because the loss amounts may need to be broken down across time as well.
waves/project.py
Outdated
| if self.floris_results_type == "wind_rose": | ||
| power = pd.DataFrame(0.0, dtype=float, index=availability.index, columns=["drop"]) | ||
| power = power.merge( | ||
| energy_gwh = pd.DataFrame(0.0, dtype=float, index=availability.index, columns=["drop"]) |
There was a problem hiding this comment.
@RHammond2 fix this to get potential as the base energy, so losses can be subtracted.
…loss_ratio and descriptive note for total_loss_ratio method
|
Summary of where we left this enhancement: The main objective of this enhancement is to be able to output a loss breakdown with the following losses. (please keep in mind that the total losses are calculated as a product of efficiencies as in the following table, so it will not be possible to output the losses in MWh for each category).
The branch today can output a losses breakdown. The problem is that the net energy production does not match when we use that loss ratio, compared to when we calculate it by turbine and by month. See screenshots below to see that numbers do not match. Rob and Daniel explored potential reasons for the mismatch on values and concluded that it might stem from the accumulation of shortened decimal values by month and by turbine. Rob recommended an update to WOMBAT to address this. |
Update Losses Methods (Rob's updates)
|
This looks good to me, but please note the following caveat: The total loss ratio cannot be applied with greater granularity than the "project" level (e.g., "month-year" or "annual") or the "windfarm" level (e.g., "turbine"). This is because the availability loss ratio calculation relies on the base potential energy production from WOMBAT, rather than the energy_potential used in WAVES for the wake loss ratio and net AEP calculations. |
No description provided.