Relating onshore wind turbine reliability to offshore application

Abstract

With the award of the latest Round 3 offshore wind farm sites around the UK coast the wind industry is moving from the operation of near inshore to truly offshore wind farms. This has two major implications, the first being that wind turbines are now being specifically designed for offshore deployment, a key feature being that the new wind turbines are likely to be two to four times the size of the largest current onshore machines. The second is that due to the limitations of access to offshore wind turbines, their availability needs to be in the order of 98% or greater if reasonable costs of energy are to be achieved. The distance of the wind turbines from shore means that more attention needs to be given to the availability, reliability and maintainability of these offshore wind turbines.

The research discussed in this report set out to examine these factors in more depth, using the reliability data of Clipper Windpower’s onshore 2.5 MW Liberty machine as the practical evidence for doing so. In analysing the data the primary aim was to build a picture of typical fault type and duration and more specifically alarm type, distribution and alarm quantity. These results were then compared with an external data source to identify common trends or major divergences and the findings used to identify potential improvements in availability, reliability and maintainability for the design of Clipper Windpower’s offshore Britannia 10 MW machine.

The key conclusions of the research are that:
 The Britannia wind turbine pitch system needs dramatic improvement on that of the Liberty wind turbine and this requires further detailed investigation.
 The ability to access the wind farms quickly and cost effectively will be critical to maintaining the required levels of wind turbine availability.
 The Britannia wind turbine needs to be designed for reliability and availability not simply for keeping the wind turbine in a safe mode.
 The number and classification of alarms built into the wind turbine monitoring system needs to be critically reviewed with the aim of reducing and rationalising responses where possible.