Mitigating risks of wind turbine blade failures: Matching detection methods to failure modes.

Introduction to blade failures in the wind industry

Wind turbine blade failures are prevalent in the industry, and high failure rates have been seen across many types of wind turbine. Blades failures are complex and blades can fail in many ways, some of which are more severe than others.

Broadly speaking, the consequence of a blade failure can vary by orders of magnitude:

• Low impact – failures leading to an up-tower repair costing $10,000s (USD)
• Medium impact – failures leading to a blade replacement costing $100,000s (USD)
• High impact – failures leading to a catastrophic failure of the turbine costing $1,000,000s (USD)

Therefore, a primary objective in monitoring and detecting blade failures is to enable more up-tower repairs – effectively turning what would have been ‘medium’ or ‘high’ impact events into ‘low’ impact events. The way to do this is to detect problems early, before they progress and while a low cost repair is still feasible. The financial motivation to do this is very strong, and is currently driving a great deal of innovation in the industry around technologies for blade monitoring and inspection.

Categorising blade failure modes

There are many ways in which a blade can fail. Some failures modes are less severe and can be observed developing over a long time period, whereas other faults can develop very quickly, over just a matter of weeks, and can lead to catastrophic failure of the blade. It is therefore useful to categorise the failures by separating a blade into three ‘zones’, as shown in the diagram and table below.

Table 1 – summary of blade regions, damage types and detection methods

Diverse Detection Methods for Varied Failure Modes

What is clear from Figure 1 and Table 1 above is that blade failure modes are very diverse, and no single detection method can be applied to all failures. Blade drone inspections, for example, deliver great value and, when performed annually, are proven to provide a strong return on investment (ROI) to the end user.

However, drones cannot detect all failure modes, and some types of failure are often detected too late. For example, a crack that originates inside the load bearing structure of the blade can be several metres in length by the time it becomes observable on the blade’s surface. By this point, an up-tower repair may be difficult, costly, or even impossible – leading to replacement of the entire blade at huge expense.

Therefore it is inevitable that drone inspections will co-exist with other detection methods, depending on the blade failure modes being targeted. The same principle applies to the drivetrain, where borescope inspections complement vibration CMS, oil analysis and SCADA analytics to provide a complete picture of a gearbox’s health.

Take, for example, failures of the connection between the blade root and the pitch bearing. This is a failure mode that typically cannot be detected using an external inspection. On turbines which are known to be at high risk of this failure mode, it is common practice to perform regular visual inspections and measurements, but these are labour-intensive and costly.

When this failure mode develops, failure of the blade root insert causes the blade to separate from the pitch bearing. If left unchecked, the blade will eventually detach, in some cases leading to catastrophic collapse of the turbine. Knock-on effects can be even more severe – with huge safety risks associated with blade loss, some governing authorities have threatened to totally halt operations at sites with a high occurrence of blade loss.

Consequently, many turbine owners have now taken the decision to deploy ONYX’s ecoPITCH to track the progression of this failure mode, and plan repairs and blade replacements accordingly. ONYX has installed permanent and temporary instrumentation on over 1000 turbines to date, creating a huge dataset. From this, accurate thresholds and analytics have been developed to assist turbine owners in planning the appropriate maintenance or blade replacement – in many cases with 6-12 month lead times, therefore delivering significant O&M cost savings and significant additional revenue generation.