Prognostics and Data Modeling that Extend the Remaining Useful Life of Wind Turbines

Wind energy continues to grow in popularity as a source of renewable energy. In 2012, 42% of all new energy generating capacity is attributed to wind turbines making it the #1 source of new capacity. Managing the use of wind turbines has been exceptionally challenging because their operating lifetime is only for five to thirteen years which is much lower than the expectation that they will last between 20 and 30 years.

The main source of premature failures is the gearbox but the source of these failures has not yet been determined. As a result, wind farm operators must evaluate the asset risk of their turbines because the average cost of fixing a wind turbine is between $ 300,000 and $ 700,000 per failure. This means wind farm operators can readily incur costs ranging in the millions of dollars to deal with turbine failures.

Currently, the wind industry uses a diagnostics-only technology known as CBM to monitor turbines. But, this approach can be problematic because one to two years of data are needed to better understand the source of failures. In addition, the sensors used to collect the data can generate false positives and also not operate properly. Wind farm operators may need to use an extensive amount of monitoring equipment and hire additional personnel to monitor and evaluate the data leading to additional costs with an uncertain ROI.

Author: Sentient Science
Co-Authored By: Firstwind
Publish Date: September 30, 2013