Authenticating ‘Black Box’ Data
(Room Windsor Complex)
28 Feb 18
10:00 AM
-
10:30 AM
Tracks:
Quality Assurance
Wind farm operators are increasingly faced with data and reports based on untraceable ‘black box’ calculations. It is critical to be able to validate these results, especially when they relate to large payments – such as curtailment compensation, availability guarantees or performance upgrades. This presentation will focus on three examples, each illustrated by a detailed case study: 1. Yield-based availability calculations are performed by algorithms based on complex contractual definitions. The result is often a simple set of numbers, but how can an operator validate that the value is correctly calculated and reflects actual performance? 2. Nacelle transfer functions, applied by manufacturers to correct nacelle wind speeds (themselves based on calibration factors). How accurate are they? How sensitive are they to changing conditions? What are the limits to nacelle power curve analysis? 3. Estimates of foregone energy lost during downtime (e.g. grid curtailment or following lightning strikes). Since no production data is available during this period, the estimate is typically based on secondary data such as met mast or modeled wind speeds, and a complex ‘black box’ calculation method. How can these estimates be validated?