MS#06.4 Turbine Performance Monitoring

P. BRADSTOCK¹, Y. DING², C.PLUMLEY³
¹ Bitbloom|² Georgia Tech|³ Nuveen Infrastructure

Reliability, monitoring and sensing, O&M

There has been an increasing focus on monitoring the performance of operational wind turbines in recent years as wind farm owners and operators try to detect and reduce causes of lost yield and thereby maximising revenues and profit margins. As turbines produce ever more data and are ever more connected, the application of live automated analytics to monitor turbine performance has become more common. However, due to the complexity of environmental conditions and inaccuracy of anemometry installed on most turbines, evaluating performance is not a trivial task. This has led to a wide variety of interesting algorithms and approaches to evaluate turbine performance and they continue to evolve rapidly.

These include, most notably (but not limited to), the detection and analysis of:

• Static yaw misalignment
• Long-term blade degradation
• Rotor mass or aerodynamic imbalance
• General power performance tracking (e.g. power curve analysis or neighbouring turbine power comparisons)
• Analysis of controller characteristics

Submissions to the this mini-symposium should focus on solutions to monitoring any aspects of performance from live or regular data streams, and may be machine learning models, but also statistical approaches or consisting of other deployable algorithms.