Dealing with unscheduled turbine failure is invariably expensive, requiring emergency crew, possible delays in replacement parts and lost generation while the turbine is out of action, explains Emily White of asset-management firm Entap.
Monitoring services look for telltale signs that something is not as it should be, and detailed detective works pays when a clue is found and a plan can be put in place to remediate the situation before it leads to failure.
Many third-party management services include remote condition monitoring of clients' assets using the Scada interface for data acquisition. Important parameters should be checked regularly, including monthly checks of the temperature of drivetrain components such as the gearbox and generator.
Down not up
During routine checks of a client's 3MW turbines, we discovered that one particular gearbox was running somewhat cooler than the others on that site. We benchmarked the temperature readings against turbines of the same type on different sites to confirm that this gearbox was the exception.
Our concern was that by running cooler than normal, there would be a reduced flow of lubricant to some of the bearings that would cause damage to the gearbox. We were also concerned that the manufacturer's checks would not pick up this potential problem because it only monitored high temperature limits.
Many of the common failure modes cause temperature rises, and manufacturers are usually only concerned with temperatures that exceed the design limits for a particular component and that lead to rapid degradation of the lubricant. However, components such as bearings have a history of failing before they reach the design limits for temperature.
We raised our concerns with the manufacturer, which carried out further tests and concluded there was nothing wrong with the gearbox.
We continued to monitor the unit closely, and when an oil sample was analysed it showed unusually high levels of copper and zinc, which are usually associated with damage to brass bearing cages.
Because manufacturers are reluctant to disclose their "normal" value for oil sample contaminants, we derive our own expected "trigger" levels using in-house data. In this case, the levels for copper were 15 times our baseline expected levels, and zinc levels were 2.5 times higher than at the last service.
The quality of the lubricant also showed signs of degradation, with a significant drop in phosphorous content, which is part of the additive package.
We presented these significant changes in trigger levels as further evidence to the manufacturer that there was a clear problem developing in the gearbox, and that it should be investigated.
Prompting manufacturer to act
This time, the manufacturer responded to our concerns, thoroughly checking the gearbox temperature sensors and the thermostat controlling the gearbox cooling. Replacing the thermostat addressed the problem, and the gearbox oil temperatures rose to a range that was consistent with the other machines on site. Due to the high levels of wear debris seen in the oil samples, the manufacturer proposed to flush and refill the gearbox oil, based on our recommendations.
Downtime for the oil change was two days and but it has extended the life of the gearbox. If it had failed and required a complete replacement, the downtime would have been between 10 and 15 days, depending on the availability of parts, manpower, cranes and a reasonable weather window. The difference in cost to the owner, based on generating at full power, could have been in the order of £550,000 (€726,000).
This example demonstrates the value of experience when it comes to asset management, where access to detailed data informs decision making, and questions are asked of the turbine makers — often persistently — in order to represent the interests of the wind farm operator.
Emily White is managing director of Entap, a European renewable-energy asset management company