Mark Evans, Obelisk Energy, UK
There’s an old saying “If it isn’t broke, don’t fix it”. However in the renewable world this philosophy leaves too much to chance. Failure is not an option and if it does occur it can be very costly for the operator.
Unfortunately, decisions over maintenance for wind turbines can be challenging, as stochastic loading due to hourly and seasonal variations in wind speed and direction makes it hard to predict their degradation or failure. So how can you put failsafe O&M procedures in place to minimize the risks of things going wrong? And what are the latest trends in proactive maintenance for onshore and offshore wind farms?
There are broadly three approaches to maintenance:
- Reactive maintenance (run to failure)
- Preventive maintenance (time-based)
- Predictive maintenance (condition-based)
Currently, the wind energy industry uses only reactive maintenance (that is, fix it when it breaks) and preventive maintenance (by following the servicing procedures set out in the OEM service manual). However, the industry is not yet well versed in predictive maintenance, which uses a range of high tech condition monitoring technologies to provide a more efficient O&M service.
One must bear in mind that the bulk of a breakdown repair cost can be attributed to waiting time for parts, delays in getting equipment such as cranes to the site, and the waiting time for experienced staff. This is before you take into account lost energy revenue – all costs the wind farm owner has to bear.
Experience has shown that wind turbine components will routinely fail, to varying degrees, significantly before their design life is reached. This poses a major problem for the wind industry as it results in significant unplanned and unscheduled repairs, which take the turbines off-line costing the wind farm owner lost energy revenues.
Preventive – or time-based – maintenance activities should, as a minimum, be carried out in accordance with the OEM manual for that turbine. This should be supplemented by additional time-based items as dictated by real-world experience of the maintenance contractor.
These activities would normally include a time-based turbine visit every six months and comprise mechanical checks of fluid levels, greasing, bolt torque checks, filter changes and inspection of blades and brake pads.
In addition there are routine electrical checks such as inspection of cable connections, fuse checks, voltage level checks, battery inspections, trip tests and electrical cable inspections.
While all of these activities are essential, they can be greatly enhanced by the feedback produced from coordinated predictive maintenance activities. It is even possible that some time-based activities might be reduced in frequency and scheduled only when the test results from predictive maintenance activities dictate that they are required.
One of the problems with conventional maintenance contracts is that maintenance teams on site may be left idle on “wind out” days Source: Obelisk Energy
This case study illustrates how being vigilant of minor changes can be of considerable significance. By monitoring the parameters such as temperature data you can reduce the risk. Take, for instance, a wind farm in the UK. The customary monitoring of the oil temperatures was being supervised closely. The temperature traces over a month period on one wind turbine indicated an explicit trend in increased oil temperatures, despite no reduction in power output.
The servicing personnel were sent in during a low wind period to check on this turbine. Further investigation found a small oil leak and in spite of a malfunction of the low oil level alarm, the problem was repaired with negligible downtime and almost no loss in production. It is apparent that the outcome could have been rather worse resulting in catastrophic failure of the gearbox.
So what is predictive or condition-based maintenance? It can be broadly described as consisting of a batch of equipment tests based on the use of on-line and off-line sensors and monitors.
The objective of these predictive maintenance activities is to produce an equipment condition assessment. This allows the plant owner/operator to plan in advance for ordering parts, scheduling work and forward planning for repair-refurbish activities.
A well-organized predictive maintenance strategy combined with state-of-the-art sensors and tests will minimize unplanned turbine outrages and maximize energy output and revenue generation. It is also vital to monitor the wind farm weather conditions, specifically wind energy and weather forecasts, for maintenance work scheduling. Obelisk Energy has adopted what it calls a “reactive 24-hour breakdown support” strategy which defines best practice for predictive maintenance.
This has a number of elements including:
- 24-hour surveillance of the wind farm through private networks linked to a central control centre;
- Analysis of fault alarms or turbine stoppages, leading to removal or on-line restarts if possible. OEM surveillance is also contacted (if under warranty);
- If a turbine cannot be restarted remotely, a maintenance team is sent to the site;
- For larger breakdowns a large stock of spare parts is available for replace and repair at strategic sites.
It is vital that the maintenance teams are supplemented by personnel with OEM experience who can advise on more complex problems with specific turbine types.
Obelisk Energy has also developed “Reactive/Planned Maintenance”. One of the problems with conventional maintenance contracts is that maintenance teams on site may be left idle on “wind out” days. Instead Reactive/Planned Maintenance monitors the weather two weeks ahead and maintenance teams are scheduled to undertake work when wind speeds are predicted to be low.
The impact on man-hours of high wind periods on site can cause unnecessary costs for the operator/owner when maintenance staff are winded out.
While carrying out planned maintenance on Quinn Wind Farm in Ireland, Obelisk Energy field staff found the RoCoF (Rate of Change of Frequency) relay was set in a way that might cause unwanted trip of the mains circuit breaker, causing interruption in power delivery and increasing the standby time for the whole wind farm.
Planned maintenance also picked up a second circuit in a substation, as one connector was too loose, resulting in an impact on the terminal, which in turn could cause the temperature to rise to a critical level, risking a fire in the substation. This was repaired immediately on site by our trained electrical maintenance staff.
In conclusion, close monitoring of the wind farm using sophisticated sensor technology and tests combined with a flexible approach to maintenance scheduling can pay dividends for the wind farm owner.
About the author
Mark Evans is head of Operations and Maintenance at Obelisk Energy Group, a leading provider of O&M services to the wind energy industry.
Power Engineering International Archives
View Power Generation Articles on PennEnergy.com