We are working on a project of designing a 10KW Battery Charger to be powered by a 12.5KW, 3 phase wind generator. The frequency of the generator output varies from about 24Hz at 120A AC to 90Hz at 440V AC for min. to max. wind speed.

While researching on the subject, I came across an article by Dr. Ray Ridley on implementation of a wide range 3 phase PFC circuit using three single phase units.

Whether anyone can throw more light on PFC with variable frequency input.

Perhaps you should research the Vienna Converter topology. It is a good fit for this power range. While the control method can be challenging, the efficiency is high. Alternatively, 3x single phase boost PFCs may be a more straight forward option if you are not concerned about current in the neutral line. I'm assuming the 440VAC is L-L measurement. The variable frequency and wide range of the AC should not limit the topology choices for your PFC. It should only complicate the method for current feedback/measurement.

Thanks for the input. To keep the control simple, I am seriously considering 3 single phase boost PFC stages with front end DC-DC Converter. Paralleling for higher power is proposed be done after the DC-DC Converter stage. There is no neutral available. 440V is L-L. I have looked at Vienna converter topology, but find that the control methods are too complex to handle. We would like to keep the system as simple as possible.

the three single phase approach that is described on our website had no neutral, and therefore no neutral current. each phase is run line to line.

it was implemented by IBM at about 20 kW.

adding the twist of making it a buck stage for part of the cycle does complicate the control, but it was implemented fairly simply with just analog parts.

the three single phase choice for ibm was driven by the need to continue running at full power and unity power factor even if one of the input lines went down. a common failure event with mains power.

That gives me lot of confidence. In fact your article is extremely well written & easy to understand. Since I did not know the application, I was not too clear of the use of SCRs as rectifiers cum switches. We do not have the issue of one line going down, since the source is a wind generator. So in case a phase goes down, we will shut down the system. I am more concerned about the performance with a variable frequency source, but not so much about wide input voltage range.

If wind is not available to deliver 100% power for most of the time, it may make sense to use a 3 phase rectifier followed by a single phase Boost PFC.
This will yield a PF>.95 which will not affect the output power availability for most of the time.

I did not understand that. How do you use a 3 phase rectifier with a single phase boost converter. It will not give sinusoidal current in the 3 phases. I am also unable to understand as how a multiplier in a single phase PFC chip will respond to a 3 phase rectified waveform.

Secondly in this case, the client is using some product from Finland, which already has an active rectifier, so the comparisons are bound to arise.

A correctly designed single phase Boost PFC at the output of the 3 phase rectifier emulates a resistive load.
The current in each phase will be quasi square and will flow for 120 degrees yielding a PF of ~.955 - same as if the rectifier's load is a resistor.

Since you said that the neutral is not available, I would suggest a 3-phase Space Vector Modulation (SVM). With this topology & control method your switching frequency and thus, your losses basically get cut in half.

For Vienna Rectifier as well as 3 phase space vector modulation, the controls become complex. We wish to keep the control method simple. As of now I think we will choose between Buck + Boost or 3 phase rectifier with single phase PFC.

If you have already constrained into using PFC, then the following comment is not relevant. If not, how amenable is your application to a multi-pulse auto-transformer front-end based solution. In this case, the wind generator is connected through a 12-pulse (or 18 pulse) autotransformer to a 12 pulse/18 pulse diode bridge to dc capacitor bus. The unregulated capacitor bus voltage is then used for battery charging using an appropriate dc-dc converter.