wind farm interconnection study

[ahmed318]

Hi ,

I work for a utility in the mid western united states.Lately, I have been doing short circuit studies for a wind farm interconnection.Instead of modeling each wind turbine in aspen ,is there a way to model just one turbine with a step up trafo instead of modeling 50s of turbines.What I think is to parallel all the generators connected to 565/660V bus and parallel all the transformers connected to the 34.5kV bus which means z/n where n is the number of turbines and z is the impedance .Please let me if this could be done for generator's sub transient ,transient and sync reactances too

 

[magoo2]

I don't see why not. I would treat it like the way we estimate peak demand. Apply a diversity factor.

With wind, you might have 20 or 30% of them on at one time, so take that into account when you make your equivalent generator and GSU.

 

[SparkyLarks]

With wind, you are likely to have all of the turbines running at the same time.

the speed of the turbine depends on the wind, so once the windspeed across the windfarm is above a cetrain threshold all of the turbines will be on max output.

you would need to assume all generators are on full load.

 

[cranky108]

The question would be, what type of turbines are they and how do you model them?

If it is possible the model the whole farm, then it is also possible to reduce that model to an equivlence.

Aspen should be able to reduce the model. But knowing how to model the machines is an issue the industry needs to work on.

 

[ahmed318]

Thanks all for your answers .They help.

The turbines are being modelled as doubly fed type 3 induction machines where aspen requires to keep sub transient ,positive ,and negative sequence impedances equal to the sub transient reactance with some very high zero sequence reactance.If I have to a model a 100MW wind farm with 2.5MW suzlon wind turbines ,then I would model 40 type 3 generators in aspen each with a pad mounted trafo.I have figured out that if I just model a single generator and a trafo with impedances divided by 40,it gives almost same fault currents as modeling all 40 of them .I would appreciate your comments on this technique.

 

[cranky108]

It would be better to model the whole farm in a small model, and take an equivlence to your general model.

The approch of using 40 equivlent machines is that it does not account for interconnect cables between the towers, or other things they may add. Things like capacitors, zig-zag's, etc.

This approch also helps with figguring short circuit in the wind farm. If required.