Data-Driven Wind Turbine Power Generation Performance Assessment Using NI LabVIEW’s Watchdog® Agent Toolkit

Power generation performance is a fundamental metric that all wind farm operators use to determine whether expected power throughput is actually being met. IEC 61400-12-1 has been drafted as an exhaustive power performance measurement scheme for wind turbines. The primary weakness of such a standard is the required level of depth of the associated performance tests, which is more than sufficient for operators to use to run daily wind farm activities. In addition, since this IEC test is not really meant for frequent evaluation, it also fails to capture any loss in power generation performance over time. This paper addresses the aforementioned weaknesses of the IEC standard by the application of data-driven approach to model a wind turbine’s power curve. A set of measurements during a known good condition is utilized to setup a baseline model. Regular power curve measurements are then compared while taking into account the multi-regime dynamics of the turbine. The approach was implemented using NI LabVIEW’s Watchdog Agent® Toolkit and was successfully validated using actual SCADA data collected from an on-shore wind turbine.