Maintenance expenditure on wind turbines is predicted to reach $17 billion by 2020, according to the Globaldata's 2015 Global Wind Turbine Operations and Maintenance Market report. That is almost double the $9.25 billion spend in 2014. So, developing simple yet reliable, real-time monitoring technology that can be implemented cost effectively is crucial to avoid downtime and extend operating life.
Estimates of maintenance of old turbines are around 3% of the lifetime cost, and incorporating the latest technology can cut that cost in half, assuming a working life of 20 years. Yet, the wind industry has not wholly embraced the adoption of condition monitoring. An obvious and understandable reason is that real-time monitoring adds to overall capital and operating costs. As a result, a compromise often has to be found between frequency of monitoring, and the need to gather accurate data that enables an effective maintenance regime to be carried out.
No room for error
But compromises can be dangerous. Studies carried out over long periods on the causes of turbine failure, in both Europe and North America, highlight gearboxes as causing the highest levels of downtime, with failure rates rising significantly towards the end of the anticipated turbine life.
And while failure rates have reduced in recent years - through advances in technology and better reporting of issues - they still occur, not being picked up by the reactive maintenance procedures and poor analysis of the data gathered from sensors.
But experience in this industry and in others shows that if you stop using outdated reactive and fixed time-period maintenance programmes, you can get to the fault before it happens, meaning the issues of machine wear, imbalance and energy inefficiency will have less of an impact on output and operating costs.
In the marine sector, for example, where operating margins are under growing pressure, predictive maintenance techniques are playing a critical role in keeping ships at sea for far longer than had previously been the case. Vessel owners have for some time been continuously monitoring components such as bearings in pumps, motors, fans and generators, and assessing the rate of wear, to be able to plan exactly what maintenance is required for the brief period when a vessel is in port.
Mining and quarrying industries do it, as does the manufacturing sector, where factory-wide systems improve the efficiency of all production assets as part of long-term maintenance and productivity programmes. At the heart of many of these programmes lies vibration monitoring, often allied to other techniques, such as temperature and pressure sensing.
Although there are obvious differences between wind turbine monitoring and factory or marine equipment - speed of rotation and remote locations to name just two - the technology is easily transferable.
A number of original equipment manufacturers and operators have already embraced predictive maintenance as a critical condition monitoring technique. But there needs to be more take-up.
As the number of turbines in offshore or remote areas increases, the industry as a whole needs to take a far more proactive approach to predictive maintenance if we are to ensure that ever tougher targets for power generation and turbine availability are to be met while simultaneously protecting operating margins.
Chris Hansford is managing director at Hansford Sensors