The value of vibration-based condition monitoring systems (CMS) is well established within the wind industry. At ONYX InSight, we know from our global experience of wind farm monitoring that vibration data is a critical source of information regarding turbine health, especially when matched with engineering expertise. But we also know that combining multiple streams of data provides owners and operators with even more valuable insights.
One such combination is merging data from ‘online’ oil sensors with information from vibration-based CMS, to gain a deeper understanding of overall turbine condition and oil health.
ONYX InSight recently delivered a multi-phase study, designed to explore the full potential of combining oil data with vibration data for wind turbine health monitoring. Laboratory investigation has been carried out into oil sensors using wind turbine gearbox oils, simulating known defects such as oil aging, water in oil and additive depletion.
Phase two of this project, which recently concluded, has seen an extensive programme of oil sensor installations in a commercial wind farm. So far, our case studies have demonstrated the benefits of combining a package of oil sensors with vibration CMS. The trials are helping ONYX InSight to refine new approaches for an advanced, integrated CMS by combining data from oil condition, vibration sensors, SCADA systems and other operational wind farm data.
So why did we choose oil data in our investigation?
Most wind farm operators are already utilising offline oil sampling. However, due to the infrequency and inconsistency of manual collection, this is unable to provide the high-quality data required for accurate predictions. Even the process by which technicians take an oil sample can vary – creating a data set that is very sparse, sporadic and subject to human error.
Online oil sensors deliver significantly more consistent data compared to the offline methodology, enabling operators to prioritise offline laboratory analysis for machines with known poor health. This reduces cost, as well as reducing time spent in the nacelle and with potential safety benefits of reduced turbine climbing. Generally speaking, oil sensors fall into two main categories – those that monitor machinery health, such as particle sensors, and those that monitor oil health, such as ‘oil quality’ or ‘oil condition’ sensors. There is overlap between these categories, and some sensors offer additional channels such as oil temperature and measurement of water-in-oil (e.g. relative humidity).
For example, a sensor that observes a change in oil condition could in fact be detecting a symptom of poor machinery health, which, if undiagnosed, might lead to a future failure of the machine. The ability to draw upon multiple data streams is a great asset to any engineer looking to identify the source of a problem with a high degree of confidence. Vibration data and oil data provide valuable insights into overall turbine health when combined, which generates significant savings for wind turbine owners and operators.
The pioneering study comes in response to the industry’s need to reduce O&M costs and lower the levelized cost of energy. ONYX InSight believes that combining multiple data streams with vibration data offers significant advantages that improves maintenance planning, particularly when deployed together with machine learning and engineering expertise. The knowledge we’ve gained by testing oil sensors both in the lab and in real world turbines will ultimately lead to big savings in O&M costs for our customers and more uptime generating clean electricity.
As wind farms find their profits increasingly squeezed in the post-subsidy market, it is vital to enable O&M teams to drive maximum operational efficiencies using the latest technology. To continue to innovate and improve predictive maintenance best practice, we need to understand how oil and vibration data can be combined to provide the most accurate insights to enable us to best support our customers’ O&M teams.
Bruce Hall, CEO, ONYX InSight
Find out more about the outcomes of the oil sensor study and the benefits of combining vibration and oil data streams.
Speak to our engineers today: firstname.lastname@example.org
Meet with our representatives at the upcoming wind shows in:
JAPAN: WindEXPO Japan (Feb 26-28) – UK Pavilion
USA: AWEA Wind Project O&M and Safety Conference (Feb 26 – 27) – Booth # 233
EU: Wind O&M EU 2020 (Mar 5-6)