My philosophy on the implementation of the retract phase, and management input power variation and transients has always been to employ energy buffering/damping behind the machine side convertor by means of supercapacitors, batteries or independent fly wheels working most likely from a common dc bus. I feel that using a flywheel in front of the machine side convertor using direct mechanical recovery would potentially be more efficient but more difficult to control; you wouldn’t want to rely on it for your retract either. So I would say that there is potential efficiency benefit to be had but from a practical side energy buffering behind the machine side convertor will likely still be required. I suppose also the benefits of the flywheel may be relative to the length of the generating cycle employed; a long generate cycle would likely max out the flywheel (depending on its size of course) but it will always have significantly more energy than is needed for retract, I believe in the transport application it useful for absorbing negative power spikes from short vehicle braking period; in a yo-yo system we are braking for the majority of the time and looking to export that energy. This leads me to conclude that its best application would be to absorb transients (from gusting/turbulence) and smooth the energy outflow for what happens next. I guess that lead to the question how much more efficient would this be than just throwing everything at super capacitor bank (or equivalent) though the motor drive and tapping into it when you need to retract, while exporting the rest at a steady rate.
I looked into MAGSLIT previously as it seems to be a excellent technology to deliver a high torque low speed generating phase followed by a low torque high speed retract with only a single drive unit, however magnomatics suggested the PDD option instead; this could have been for any number of reasons.
All top of my head stuff on the flywheel though.
I see how the magnetic gearing works. It is basically the same idea as most hybrid linear stepper motors work with permanent magnets attracting a iron core with many “teeth” and a electromagnet forcing the magnetic field to change so it will move to the next position (but in the case of this gearbox in stead of a electric coil shifting the field another rotating permanent magnet with few poles does this).
But when it comes to the full PDD implementation I fall off. It is a normal permanent magnet motor with a LARGE air gap and lots of magnetic things (the magnetic gearing) between the rotor and the stator.
- 1: Don’t makers of electric motors try to make the air gap as small as possible to make the motor more efficient and powerful? In that case they have just made a really inefficient motor.
- 2: I can not imagine that all of the iron bars and permanent magnets in the “gearing” part between the stator and the rotor does not have a negative effect on the magnetic field between the EM in the stator and the PM in the rotor.
And as a side note: Don’t we want to get away from PM motors and generators and over to alternatives like induction motors and electrically excited generators (like in car alternators with a coil in the rotor) to minimize our use of the limited rear earth materials used in neodymium magnets for PM motors/generators?
I would agree that the stability of the price of rare earth magnets is a risk for the commercialization of this technology, the CO2 cost of producing rare earth magnets is not attractive either; in terms of CO2 for kg to produce they will likely be the most costly components of all within HAWT or AWES. I guess this is similar to the lithium produced to be used in the batteries of electric vehicles, lithium mining is dirty.
I think the subtleties of why its efficient and the technical detail subtleties of the design are certainly beyond my level of understanding of EM machines.
I get that they want to keep the unit compact and put everything in one unit in stead of a separate gearbox unit and motor/generator unit (that is then a normal motor/generator and therefore not something they can patent). But if the airgap and the interferance of the gearing stuff makes the motor/generator less efficient i think their best bet is just to go for the pure magnetic gearbox unit and connect it to a off the shelf motor/generator. But then maybe their advantage compared to a normal direct drive generator with a large diameter and many poles disappears?