Between apparent conflicts like this and the oldest modern wind turbines approaching the end of their design life, the industry is starting to face some big decisions about length of service. And yet, how much can we really know about how long our wind turbines will survive structurally?
Options for end-of-life strategies include:
- Decommissioning even if there is remaining life in the structure;
- Continuing to operate turbines blindly; or
- A middle road: taking steps to understand the risks, costs and opportunities, and making intentional, site-specific decisions that maximise returns while maintaining safe operations.
Other industries have faced the question of structural longevity, including offshore structures (such as oil rigs), bridges, dams and airplanes. Historically, these industries have approached life extension decisions through the use of conservative safety factors and deterministic analysis. But more recently they have tended towards probabilistic risk analysis and use of both prescribed (fixed interval) and risk-based inspections to calibrate and reduce uncertainty in risk analyses.
The wind industry can leverage the progress made by these industries in areas such as structural-reliability analysis and risk-based inspections.
Structural-reliability analysis (SRA) looks at the probability of structural failure at any given time in the turbine's operating life - considering the site-specific loading, operating history and component strength. Inherently uncertain when forecasting for a specific structure, the SRA is complemented by risk-based inspections (RBI). RBI planning considers the structure's probability of failure, costs of repairs, inspections, and the consequences of failure. The result is an RBI plan that keeps the risk of structural failure below a target level.
Testing at 25 years
For example, take a hypothetical 25-year-old wind farm using risk-based inspections to facilitate a safe operating-life extension. The turbines have been well maintained and operated over the past 25 years. After considering the consequences of tower failure, the project stakeholders agree on a target reliability of one in 1,000. The site Scada data and meteorological data are analysed in an SRA, showing the probability of failure is approaching the target reliability level in 2015, at which stage the risk of failure was then considered too high.
The project owners respond by inspecting a random sampling of 50% of towers using either visual inspections or non-destructive testing (NDT) technology that has a well understood probability of detecting a crack of a given size. Visual inspections would cost less, but the crack size that could be detected would be smaller with NDT. The resulting reduction in probability of failure will depend on which form of inspections is chosen.
No cracks were found, and the structural-reliability analysis was updated with the additional knowledge that the towers survived 25 years without cracks, reducing the uncertainty in the analysis. The updated probability of failure is now well below one in 1,000.
The turbines can operate for another five years before the probability of failure once again approaches the target level and another round of tower inspections is required. If cracks had been found, the project owners would have made a decision to repair or decommission. Finding and repairing any structural damage also improves knowledge of the tower's structural reliability and could be incorporated into the analysis.
An alternative plan could have been to inspect fewer towers, but the total reduction in risk would be less, and the next set of inspections would be required sooner.
Project owners and operators undertake this same approach for all structural components, maximising the period over which they can safely operate their turbines.
Structural survival is certainly not the only question on the minds of stakeholders: the financial profitability of aging assets is challenging to predict. How can accurate forecasts be made past year 20 for energy production and operations-and-maintenance costs with such uncertainty around structural reliability? How can inspection plans be developed to mitigate the risk of catastrophic failure? Are new maintenance procedures needed to assure personnel and equipment safety? These and other pressing questions will be on the minds of wind-industry owners, operators and manufacturers more and more in the years to come.
Alex Byrne is senior engineer for turbine technology at DNV GL