Prevention is the new approach to cutting down your wind turbine maintenance.
One of the issues facing wind turbine servicing companies is that many of the wind turbines that have been installed were not mass produced. Many identical models use different components.
There is relatively little operational data to hand. While many control systems are still way behind the technological curve compared to components like the blades and generators.
But in the last eight years, many operators have focused on reducing the running costs of wind farm maintenance. They understand prevention is better than running the machine into the ground.
Minimise wind turbine maintenance costs
For wind farm servicing technicians, there is a need to focus on two areas of operations and maintenance (O&M).
Scheduled and unscheduled maintenance.
Unscheduled stoppages lead to a loss of sales. Thus it stands to reason every effort should be made to reduce them. The less money a project makes the more uneconomical it is to run.
So how do you breakdown O&M costs?
The traditional way is to assume that the total annual charges represent a percentage of the installed cost, often quoted between 3% and 5%.
One system is to mark up the annual maintenance budget at around 3%-5% of the installation costs. Frequently, this is worked out as $1 million per MW.
If you want to go into more detail, break down the cost of each component using the same system.
Windmill servicing costs are hard to estimate
The inexactitude of the science of calculating O&M costs begins with the variety of ingredients that make up the machine. On top of that scheduled and unscheduled maintenance.
Yet there are so many unknowns when it comes to estimating. The 3-5% rule might give you an idea but there is more to it than that.
Back in 2008, it was estimated that windmill maintenance costs varied from €15-26 per MWh. While the International Energy Agency (IEA) rates this at around €7-26/MWh.
At the same time, there is evidence to suggest costs per MW can fall when you use larger wind turbines. Moreover, larger projects can also reduce the per MW average for operational costs.
There is one major sticking point though. The IEA has pointed out that data on wind turbine reliability is still hard to come by. It said:
There is a strong demand for making better use of operational field data to improve O&M as well as for other applications from design optimization to risk management, but there is no common understanding which data to collect.
Material modelling and testing
From a wind project owners perspective, one of the few areas you can control is when you visit, check, and fix the generator.
What is the best way?
Is it lots of visits to check on performance and components? Or few visits, but these are major overhauls?
The answer is probably somewhere between the two. Too little maintenance and you have a big risk of failure but lower costs, or too much and its the other way around.
These days, thanks to material modelling from companies such as Sentient Science, even the most basic component (like a bearing) can be ground down to its most basic component.
Algorithms can then determine the components lifespan depending on the conditions, turbine, and other environmental and technological factors.
Failure can be predicted and the component replaced before it goes wrong and damages other parts.
Turbine failures and their probabilities
It is worth remembering a large percentage of the turbines currently installed onshore are around or over 10-15 years.
To gauge the performance of these machines it is worth looking at a German government-funded study between 1997-2006. It monitored 1,500 wind generators and compiled a comprehensive picture of the probabilities surrounding wind turbine failure.
Despite its age, the study is still a crucial window into the turbine downtime and how it is affected by component failure.
Interestingly, electrical failure is the most common type of malfunction.
Although this is quite easy to resolve, it is worth remembering that electrical systems are an area that has seen less development than other areas of the windmill design.
Surprisingly, with all the moving parts, gearboxes are the source of far fewer problems (around one day a year on average). But it takes longer to fix and so costs more. Hydraulics is the most problematic part of the turbine.
Bigger turbines change everything
But no study is perfect. The wind industry has changed so much since the early 2000s. Since 2010 we have been effectively been looking at different machines.
Similar studies to the German one have found that gearbox failures are more prevalent in larger turbines.
Probably one area where O&M is more crucial is offshore wind. Despite the continual upward curve in offshore wind turbine power ratings, the sea is as unrelenting as ever when it comes to access.
And helicopter access remains as expensive as ever.
Many wind turbine manufacturers are looking at ways they can automate the wind turbine maintenance or do it remotely from shore.
But another concern is if one component fails it can quickly damage another if not spotted quickly enough.
Data is all-important
Sentient’s head of industrial internet solutions Adrijan Ribaric said the company uses a high-fidelity process to study components including the bearings.
Ribaric said “We buy bearings, chop them up and put them on the profilometer and measure them. Even further, we take material samples.”
He added: “The advantage of this is that we get a good view of a specific asset and what is the weakest component in it.
“That component could be different for different turbines even though they use the same components because they operate at different locations or they have the same model of the bearing or gear but are a different match. And only material-based models can provide that.”
Sentient uses sensitivity tools to handle different properties. These include surface roughness or lubrication.
It can also examine different bearings from different suppliers. From there an assessment can be made into whether keeping the bearing cost effective.
Or whether it would be better value to replace them all with a different component.
Yet there are plenty of pitfalls with data.
Maintenance Partners (ME) used a combination of meteorological and turbine data to refine a wind turbine’s power curve.
ME project engineer Philippe Mol said there are still many variables to overcome during analysis and this is not always down to the complexity of the data.
Cut O&M costs but know where you are cutting
The wind industry has made enormous strides in cutting the cost of energy. Reductions in operational expenditure has played a key part of this.
Yet if these reductions are not thought through, they could end up costing more in the long-term.
We’re driven by the need to reduce operational expenditure. It’s ok to cut costs but you have to know what you’re doing. But it’s a mistake when you’re not in control because you don’t know the consequences.
There has been a tendency to surmise that if the generator performance can’t be improved its operating costs can be reduced.
Condition monitoring systems at the forefront of prevention
There is nothing new about content monitoring systems. Indeed some believe the systems themselves have not been able to keep up with development in other areas of the wind turbine.
But data analysis is now all-important.
Vibration is an indicator of potential issues. If it drifts towards a critical level this can trigger alarms and may even require a visit.
Yet there are studies afoot that may reduce the need to go into the nacelle. Recent research from DNV GL found there was likely to be a role for artificial intelligence in terms of wind farm maintenance.
In offshore this could include the use of drones to supply monitor external components or bring components from the mainland.
When the problems hit the generator
Overall, it is worth remembering many windmill faults tend to come in the bedding in period following installation. Generally, this levels off until the back end of its lifespan, generally around 10-15 years after commissioning.
Overall, costs will continue to fall. Reduction in O&M costs is a crucial element of bringing down the cost of energy.
It seems as new software applications develop to handle data, better predictive systems will come to the fore in making this happen. This can only help prevent the need for wind turbine maintenance.