Contributed by John Coultate, VP Advanced Sensing, ONYX Insight
Many wind turbine owners/operators now reap the benefits of data analytics and condition monitoring across their fleets. The rate of technological change means that these methods deliver a deeper understanding of their wind assets than ever before.
However, there remains a vast proportion who do not take a proactive approach to maintenance. In fact, across the wind industry, 65% of O&M costs are unplanned, leading to severe impacts on time, supply chain availability, and scheduled maintenance challenges.
For context, 80% out of the above 65% is typically attributed to blade and drivetrain failures, with generators and other components making up the rest. Major corrective spending is therefore on the rise and is expected to reach $4bn by 2029.
So, what can we do about it?
Condition monitoring has long been synonymous with drivetrain monitoring, but as the industry has expanded, so too have the monitoring solutions. Today, the industry is moving beyond drivetrain monitoring to include sensors on other turbine components.
As technology evolves, and adoption of these technologies follows suit, innovation in the sensor technology market is maturing. As a result, data management systems are far more powerful which means owner/operators can extract increasingly more value from condition monitoring.
The key differences between blade and drivetrain failures
Drivetrain components fail far more regularly than other parts of a turbine.
The consequences of a catastrophic failure in the drivetrain are, well, catastrophic. Perhaps your gearbox develops a bearing fault, leading to total failure of the gearbox – the cost of replacement part can be in the region of $150k-$300k to replace. Not only does the part need replacing, but hiring a crane, the labor to install the gearbox, and of course, the associated downtime become cost factors. And if a turbine is offshore, these costs swell significantly.
Compare a full replacement with the relatively very small costs of catching the fault early on with a CMS tool. You’re likely looking at $10-20k for an up-tower replacement of a high-speed bearing or generator bearing.
And so, because of this, the cost of installing an advanced CMS solution likely pays for itself within the first six months. The ability to detect faults of components with high failure rates significantly simplifies the planning and management of drivetrain O&M.
Blades are a slightly different proposition. Failure rates are far lower, but the consequences of a blade failure come with serious consequences. Not only are they expensive to replace, at around $80k-$300k per blade, but blade availability can also be a big problem.
In the case of some older turbines, the molds no longer exist to manufacture new blades. This poses a serious problem, leaving wind asset owners with two choices; either keep their existing blades running for as long as possible or try to source new blades from the finite stock in the market.
While both drivetrain and blades can experience similar challenges in the sense that unexpected failures are expensive and can be catastrophic, there is established technology to support the maintenance of drivetrains. Blades however remain a challenge with no clear-cut solution to their maintenance, with some operators, for example, using drones, and others using ground-based inspections using lenses.
There are two main types of blade failure – external failure modes and internal structural failures.
External failure modes are relatively easy to inspect and repair. They are typically detectable by visual or drone inspection, with up tower repairs often feasible. Inspections are typically done annually but can be more or less frequently depending on the site and owner’s strategy.
However, the objective assessment of damage progression remains a challenge. These failures typically develop slowly and predictably, e.g., leading edge erosion occurring due to particles and rain. Other failure modes include delamination, trailing edge cracks, and in some parts of the world, lightning damage.
Turbines with these types of failures can often continue to run for a period, prior to an in-situ repair. Internal structural failures are not well managed today, however.
An internal crack in the main spar, shear web, or internal bond failures can develop very quickly, growing from a few centimeters to 1-2m in just a few weeks or even days. And because the internal structure is load bearing, unlike external failure modes, the blade cannot keep running safely for a long period.
Internal faults can lead to catastrophic failure of the blade, and even the turbine, and currently blade monitoring technologies have not been widely accepted by the market.
Internal inspection methods can be used, such as visual inspections, internal drones, and crawlers – however, such inspections are periodic (e.g., annual) so the chance of finding a crack at an early stage is very small. And more regular internal inspection of every turbine is likely not feasible.
According to the 2020 Wood Mackenzie Global Onshore Wind Power O&M report, leading and trailing edge repair drive most of the blade repair spend growth up to 2029, with blade repairs alone costing the onshore wind industry nearly $2 billion by the end of the decade.
What can blade monitoring learn from drivetrain monitoring?
The key point here is this – if you’re able to catch a structural blade fault early enough, you can perform a low-cost on-site repair (which may be in the region of $20k), preventing a very expensive blade replacement, or even failure of the blade and turbine.
As it stands, there are a number of different blade monitoring technologies in the market, from acoustic emission and microphones to vibration accelerometers and strain gauges. Of course, this can be confusing for the buyer.
And with rotors getting larger every year, as the demand for wind-generated electricity increases, new blade defects will begin to develop over time, so blade predictive maintenance technologies that are working today, might not be adequate tomorrow.
There is, therefore, a strong need for more advanced sensing solutions for blades.
One of the limiting factors today is technology. Internal blade sensing has not yet been widely accepted by the market, but all indications are that adoption is growing. After all, drivetrain condition monitoring was not widely implemented 10 years ago, and few owners used drones for blade inspections until around 2017.
Since then, drones have been slowly commoditized, with the cost of inspections continuing to drop over recent years. Of course, this trend won’t be sustainable for many drone inspection providers, and will likely lead to increased pricing, which will in turn becomes less attractive to operators.
As a result, we may see the adoption of smaller, portable drones, thereby saving the costs associated with bringing a drone pilot onsite, losing production for inspection, and HSE implications, amongst others.
By utilizing advanced sensing and online measurements, which give an in-depth analysis of the health of blades, alongside low-cost inspections using portable drones, operators may achieve the best value in terms of maintenance costs across their operations when it comes to blades. Additionally, they will be able to create O&M strategies around the data they receive and plan ahead for anticipated downtime, rather than reacting to unexpected failures, which can result in huge savings across their fleets.
It is therefore expected that an integrated combination of drone inspections and online monitoring will deliver the greatest benefits and value to asset owners and operators. Both technologies work in harmony, each delivering unique advantages depending on the type of failure mode and symptoms presented – in a similar way to how faults are managed in the drivetrain today.
About the author
John Coultate is VP Advanced Sensing at ONYX Insight. ONYX Insight combines advanced sensing, analytics, and engineering with extensive experience across global multi-brand fleets, working with six out of the top ten IPPS in the industry.
This post appeared first on Power Engineering.