Two new articles were published on gearbox and bearing failures in wind power turbines in the last couple of weeks. One in Power Transmission Magazine, March 2014 and one in Windpower Engineering & Development magazine, February 2014. Two completely different groups of people, both examining the same phenomenon.
I find this to be an interesting coincidence. Both groups were describing the same causation, torque reversal stress leading to material fatigue in the rolling elements of the wind turbine bearings.
At this point I would like to mention my favorite “stump the band” question when conducting a class on Motion Control. The question is; How do we calculate the torque loading (stress) on the gear tooth of a planetary gearbox when the input is reversed? There is not a simple answer but it can be approximated as 2 times the applied torque divided by the time (duration) of the reversal. You may know of a better estimate, and that’s perfectly OK. The point is to understand the question at hand.
The torque of a sudden reversal in a gear reducer system is gigantic. Due to the time-variant nature of the force, it can be an order of magnitude beyond the design limit of the system. Imagine driving your car at 70 miles an hour down the freeway and jamming the transmission into reverse. Of course, you can’t actually do that, car makers know that it would destroy the drive train completely. Again, that’s the point.
So there are a couple of troubling ideas here. Inertial forces on the rotor increase exponentially with speed and diameter. The failures should increase with diameter. So why do industry experts argue for larger turbines as a way to make wind power viable? This makes no sense.
The second thing I find troubling is one author’s finding that failures can take place in one to two years. Wind turbines are supposed to be a capital asset with a 30 year life expectancy and a very long payback period. The more they fail, the higher the maintenance cost, the worse the payback gets.
The solution would seem to me to be a compliant coupling that can slow down the reversal period and reduce or eliminate reversal stress. Cycloidal reducers and rolling pinion mechanisms such as produced by Nexen might offer a solution. These mechanism have significant history, albeit at lower power levels, in coupling to high loads. They have very simple geometries that would appear to be scalable to wind power.
In order for “new tech” to survive, it has to be profitable and it should be lower cost than the technology it seeks to replace. If the electric utility can produce power at 4.5 cents per kilowatt hour 24/7/365, than a candidate technology to replace should be required to produce the same cost and reliability. Which, so far, has not been the case.