Loss attribution

Loss attribution starts from an ideal - the energy available in the wind that actually blew across the site - and works down to metered production, assigning every step to a cause. Wake losses, availability losses, curtailment, electrical and transformer losses, environmental losses such as icing, and finally the residual: energy the machines should have produced and did not.

The residual is the interesting part, because it is the only category that is both unexplained and potentially recoverable. Everything above it is either a design consequence, a commercial decision, or a physical constraint. What remains is underperformance, and it is where a control offset, a degraded blade, or a failing component is hiding.

Doing this properly is harder than it sounds because the categories overlap and the order of subtraction changes the answer. An hour that was both waked and curtailed can be assigned to either bucket; a turbine that was derated for a temperature limit may be counted as available. Serious energy accounting therefore fixes a convention and applies it consistently, and the convention is stated rather than assumed.

It is worth the effort because it converts a disagreement into an engineering question. A shortfall against P50 is a source of argument between owner, operator, and supplier until it is broken into named components, at which point each part has an owner and a remedy - and the recoverable share becomes a work list rather than a dispute.