Decide on how to structure the data models for uncertainty

[jonssonchristian]

Previously I had suggested dividing the uncertainties conceptually in the EYA DEF between the “wind resource uncertainties” and “plant performance uncertainties.” Based on our discussion, I concluded that the term “wind resource” can be ambiguous in this context, since aspects of wind resource can affect both gross energy and plant performance. What I was referring to was the elements of the wind resource feeding into the gross energy calculation. To make this clearer, I propose instead labelling it “gross energy uncertainties.” Then it should be obvious that captures the elements of the wind resource assessment that affect gross energy uncertainty.

The conceptualisation of a “gross energy” is of course an artificial one, but it is one that the industry has settled on and followed by the IEC 61400-15-2 reporting guidelines. So although physically the “real” wind resource relates to the gross energy, turbine interaction and other plant performance losses, we have to do the “accounting” within the conceptual frameworks that we have. For the purpose of the EYA DEF, it would be useful to group the gross energy assessment details and results together with its associated uncertainties, and similarly for the plant performance assessment.

Considering the current uncertainty quantification (UQ) model, the exception is the “plant performance” subcategory under the “project evaluation period variability” of the wind related uncertainty (see screenshot of annotated uncertainty model below). Whereas all other categories and subcategories as far as I can see relate to the uncertainty in the gross energy estimate, this one relates to the plant performance assessment. I would be tempted to move that uncertainty component from the wind-related uncertainties to the plant performance assessment. Then we could have a clear distinction between gross energy and plant performance in the uncertainty framework.

If we adopt this proposal, we will have all uncertainty categories classified into either gross energy uncertainty or plant performance assessment uncertainty.

[jonssonchristian]

Here is the annotated sceenshot that was missing in the original issue description.

[jonssonchristian]

I discussed with Mark and Steve yesterday. We were in agreement on the proposal to organise the uncertainty assessment into a gross energy uncertainty assessment and a plant performance uncertainty assessment, moving inter-annual variability of plant performance losses into the latter. We will bring this proposal to the IEC 61400-15-2 plenum to get consensus within the larger group.

We noted that a true calculation of gross energy uncertainty would have a different wind speed to energy sensitivity as compared to net energy, since the perturbation would then need to be calculated around gross energy. So the net energy uncertainty cannot be a simple combination of gross energy uncertainty and plant performance uncertainty (in energy terms). However, the actual absolute prediction of gross energy uncertainty is not likely to be of significant interest. The most important aspect is the conceptual division of the uncertainties into gross energy and plant performance, which should help making the data model clearer and easier to understand.

[jonssonchristian]

Considering this further, I think it makes most sense to keep the wind uncertainty assessment directly under the energy assessment object, rather than under the gross energy assessment object. Placing it under the gross energy assessment risks causing the misunderstanding that it predicts the uncertainty in the gross AEP. The wind uncertainty relates to the gross energy assessment insofar as it covers all the components that feed into that assessment, but it does not predict gross energy uncertainty. The wind uncertainties cover the various components of the wind resource assessment. The gross energy assessment combines the wind resource assessment results with the turbine power curve. The two are related in the sense that they are both based on the wind resource assessment. The results in energy terms are however not related. The gross energy is a kind of ideal energy based on the turbine power curve, without any deductions. The wind uncertainties are converted from relative wind speed terms to relative energy terms by a sensitivity factor that is calculated around the net energy estimate, not around the gross energy estimate.

[jonssonchristian]

Closed with #59