Understood, but windmills obviously spin at various speeds. How are they controlling the frequency?

You don't have to worry about the frequency as the windmill output is rectified to DC. Then the DC is fed to an inverter that produces the AC.
That is the way it is usually done. Directly feeding the mains from a generator will only work well for constant shaft speed, like from an engine or water wheel.

 

Bringing this from the dead.

I am reading a lot of windmills are directly producing AC current now. I understand it is more efficient this way because you are doing away with the inverter.

Can anyone explain to me how they are accomplishing this?

 

You don't have to worry about the frequency as the windmill output is rectified to DC. Then the DC is fed to an inverter that produces the AC.
That is the way it is usually done. Directly feeding the mains from a generator will only work well for constant shaft speed, like from an engine or water wheel.

With the above in mind I have heard of some wind turbines which generate an AC voltage, convert it to DC and then convert that DC back to AC so they can control voltage and frequency for a grid tie. Before I retired I watched This project being built behind the facility I worked in. Lincoln Electric installs wind turbine made by Kenersys Group of Germany it was a pretty cool build. Anyway this unit is a DC 2.5 Mega Watt unit. The numbers look a little like this:

• 2.5: Maximum output of electricity in megawatts (2,500,000 watts) at 690 volts direct current, converted to 4,160 volts AC on the ground for use in Lincoln's shops

The 4160 is a very common industrial sub station voltage in US industrial parks at 60 Hertz. So while this beast is actually a DC generator an AC generator could be used, converting to DC and then back to AC to achieve the desired voltage and frequency. Even using a constant speed drive I can't see how they can do away with the inverter especially if grid tie is involved. Do you have any links to this newer method? I would like to read them.

Ron

 

Isn't all that a bit much? Why not just make the turbine blades maximally efficient permanently (the lowest resistance to turn in the wind) and then convert to DC and then invert to AC?? much easier than physical gearing, pitch adjustment, feathering, and general mechanical claptrappery, no?

 

Isn't all that a bit much? Why not just make the turbine blades maximally efficient permanently (the lowest resistance to turn in the wind) and then convert to DC and then invert to AC?? much easier than physical gearing, pitch adjustment, feathering, and general mechanical claptrappery, no?

This is an old thread that perhaps no one cares about anymore, but since you asked, there is no such thing as blades that are maximally efficient. Blades have to work in a wide range of wind speeds and will have to spin at different rpm. There is no one solution that produces the perfect foil - the wings have to have variable pitch to achieve the greatest efficiency.

Every step in transferring energy thru the power train, from the wind itself passing by, to the electricity loaded up to the grid, needs to be optimized so that the transfer is as efficient as possible. Generators need to spin within an rpm range, the output voltage needs to be in the right range and frequency and so on. It's the great challenge of wind power – it varies. If the machines could handle it, a windmill might make more power in a single day than in the rest of the year. But you can't design a machine for that one day, you have to design it for every day. Very challenging.

 

This is an old thread that perhaps no one cares about anymore, but since you asked, there is no such thing as blades that are maximally efficient. Blades have to work in a wide range of wind speeds and will have to spin at different rpm. There is no one solution that produces the perfect foil - the wings have to have variable pitch to achieve the greatest efficiency.

Every step in transferring energy thru the power train, from the wind itself passing by, to the electricity loaded up to the grid, needs to be optimized so that the transfer is as efficient as possible. Generators need to spin within an rpm range, the output voltage needs to be in the right range and frequency and so on. It's the great challenge of wind power – it varies. If the machines could handle it, a windmill might make more power in a single day than in the rest of the year. But you can't design a machine for that one day, you have to design it for every day. Very challenging.

So, if i understand your point, the wind conditions are SO diverse, that the blade has to adapt it's airfoil. In other words the adaptation is strictly one of aerodynamic optima over orders of magnitude differences in wind-speed and torques etc. But am i correct, at least for electrical, that the blades need not concern themselves with frequency regulation per se, and that at the least frequency considerations can be decoupled from this orders of magnitude varying input wind characteristic, using a rectify-to-dc-then-inverter set up?

 

So, if i understand your point, the wind conditions are SO diverse, that the blade has to adapt it's airfoil. In other words the adaptation is strictly one of aerodynamic optima over orders of magnitude differences in wind-speed and torques etc. But am i correct, at least for electrical, that the blades need not concern themselves with frequency regulation per se, and that at the least frequency considerations can be decoupled from this orders of magnitude varying input wind characteristic, using a rectify-to-dc-then-inverter set up?

That would be correct. You want the most efficient blade pitch for the given wind conditions and the load on the blades.

Ron

 

And it's very difficult. If there was always one speed of wind, you could design the system end-to-end and each component could be chosen to optimize each step. Even then it's challenging, because a choice at the load end affects the upstream components. You might get better generator efficiency with one choice, but at the expense of inverter performance, or whatever. Lots of interaction. But with steady wind you might even be able to eliminate some components.

Now consider the design when every component has to tolerate a 10X range of operation. Optimizing end-to-end performance is really tough with so many interacting variables.