At Zentus, we are on a mission to squeeze more wind energy out of intelligent data utilization for wind farm owner/operators to accelerate the clean energy transition. As I’ve written about previously, first-hand-yes-really-talking-to-other-humans research is requisite to defining your target problem to solve and person who needs a’solvin. I have conducted close to 100 interviews with stakeholders across the wind energy sector, in addition to devouring white papers written by well-respected agencies in the industry, and one urgent area of concern that I hear time and time again is that Blades are Falling Off (BFO). In this post I will summarize what I’ve learned from real people on the subject. Disclaimer – you will find no citations here, a) to protect the anonymity of the people I have spoken to, and b) to underscore that these findings are heavily influenced by the perception and opinions of key stakeholders in the industry, not by hard statistics. But problems are personal, so perception matters!
The Blade Length Race
When you double the blade length of a single wind turbine, you generate four times as much power, which, curtailment/downtime/unruly winds notwithstanding, scales with the annual energy production (AEP) of the asset. Hence, the blades are the favorite child of wind turbine original equipment manufacturers (OEMs). The blade is designed first, and every other component must be designed around it. Despite lacking information on the structural health of the blade while it’s in operation, OEMs do not want to cover the blades in sensors for fear of hurting its aerodynamics. They also, are not particularly motivated to uncover critical defects for fear of tarnishing their reputation. Despite wind turbines being critical infrastructure, they have been manufactured and sold as commodity items over the past few years, with costs being driven down at the expense of ever-tighter design margins (i.e. less wiggle room between what loads they are designed to withstand, and what loads they might withstand in the field) and declining quality assurance on the factory floor. The diameter of the rotor (the rotating part of the turbine, including the blades and hub to which the blades connect) of commercial offshore wind turbines has grown by almost 100m in 5 years, from 164m in 2018 to 260m in 2023, but the weight of the blades has not scaled at the same rate. Engineers at OEMs are thus tasked with producing ever-longer blades while minimizing the amount of material used, resulting in more flexible components with lower design margins.
Higher Infant Fatality
Previously OEMs offered warranties of 10 to 20 years, during which they were obligated (except for certain spontaneous causes of damage such as storms and lightning) to repair turbine damages. These repairs ain’t cheap, especially for offshore sites, where a vessel needs to be rented for ~$250K/day, several $1000/day can be lost on turbine downtime, and that’s before the cost of repairing or replacing the blade itself. Wind turbines are generally marketed to have an expected lifetime of ~20 years. And yet, In recent years, the warranty durations offered have reduced to as low 2-3 years. Whereas with shorter blades in years gone by, fatigue cracks started to propagate at year ~5 and peak at year ~10, blade reliability engineers are seeing cracks start at year 1-3 with more severe categorization (sidenote – there still doesn’t seem to be a standard for categorizing blade defects, their rate of development, their severity, the repair required etc. – it is up to the expertise and experience of individual reliability engineers to make their best judgement based on the data). Bi-annual/quarterly drone inspections have been normalized in the industry and various continuous monitoring sensor solutions (e.g. vibration, acoustic, strain gauge) are have been tested and implemented, but it remains to be seen what combination of these technologies will provide the necessary predictive capabilities to reliably alert the operator that one of their blades is starting to hurt, and avoid catastrophic failures.
Manufacturing Defects
Part of the reason for the limitations of these monitoring solutions is that the blade may well have been doomed from the factory floor. The consensus among many experts I have spoken to is that manufacturing flaws are the leading cause of blade failures because a) blade manufacturing remains a highly manual process, and given the pressure that these teams are under to construct several blades a day, it is an error-prone process supported by quality assurance programs that are falling short; and b) as a consequence of the ‘blade length race’ described above, longer and more flexible blades are now designed with a diminished safety factor to allow for these errors, and given that the “as-designed” components generated by engineers will never exactly match “as-manufactured” turbines leaving the factor floor, these safety factors can easily be surpassed.
Root Cause Analysis (RCA)
After a catastrophic failure, the RCAs intended to uncover and resolve the root causes of this failure allegedly take 8-12 months for an OEM to complete. They may be dragging their feet intentionally, if their findings implicate a critical issue across all assets they have sold of a particular model, and they are trying to manage their public image in parallel. Or it may take that long to collect and synthesize data across the supply chain, from the original factory to the supply location. Since this is such an arduous process, it seems that it is only conducted for significant and recurring failures in the field.
Lightning
I admittedly, have never driven (or sailed?) my 2014 Prius through a wind farm during a storm, but I’m told that turbines are hit by lightning all the time. This is expected and lightning protection technology that mitigates the damage incurred is now industry-standard. However, with newer blades it is suspected that electromagnetic fields are interacting with the carbon fibers to cause a new class of “flashover” damages. Furthermore, condensation and manufacturing debris have been leaking into the drain holes at the blade tip, such that when lightning strikes, the water trapped by this debris can vaporize and cause the blade to split – another manufacturing defect.
Thanks for reading! If you work(ed) for an OEM or owner/operator and you’re mad as hell about bottlenecks in your workflows or some of the issues I have discussed, please do reach out and rant to us about it – we are cheaper than therapy and will probably care more than your spouse about your challenges (the wind energy ones, we do charge an hourly rate to talk about the other stuff)!
Leave a Reply