Current-Limiting Reactors

[CoryW]

System voltage = 4160V

Just wondering if anyone could help me out with Air-Core current-limiting series reactors in series with a 3-phase capacitor bank.

Field guys want to use the existing conductors to coil up and create series reactors for the capacitor bank.

They are going to bunch them together and tie-wrap them.
Does the A and B phase have to be reversed in terms of coil orientation in order for magnetic fluxes to interact properly?

Does magnetic flux interaction between phases cause undesired effects on the coil inductance?

Some of the guys have seen current limiting reactor coils bunched together like this before.

 

[wareagle]

I've seen utilities do just what you are indicating. They used a fiberglass frame to coil the conductor on. You need to caculate the inductance using the diameter of the frame
and the conductor size. It did not take many turns to do the job.

 

[scottf]

Depending on what kind of inductance your looking for, I would not recommend doing what you listed.

A few reasons....

I assume you want to use the reactors to limit the inrush current.

One reason why simply wrapping conductor around a form is not a good idea is mechanical strength during the inrush current. Reactors have a rating called mechanical current, which is the peak current the reactor can withstand without mechanical damage. Simply wrapping conductor around a form will have a pretty low mechanical current rating and thus could be prone to failure during inrush current, which, of course is when you need it most.

Secondly, if the reactor is not to switched out of the circuit after inrush, the wrapping of conductor around a form could have substantially higher losses than a well designed reactor.

Also, air-core reactors, particularly for in-rush applications are relatively cheap. Certainly a lot less costly than a damaged cap switcher/switch/breaker or cap bank.

In the interest of full disclosure, I work for a reactor manufacturer.

 

[Skogsgurra]

Hi,

Provided you can make that thing strong enough - and I actually think that you can if you use the standard cables - you will have to make the A and B phases "co-operate". That means you have to change current direction of either phase A or B (what do you do with C?).

The normal way of doing it is to use three separate air cored glass fibre/epoxy fortified reactors. One for each phase.

(I do not work for a reactor manufacturer ;-) )

Gunnar Englund

http://www.gke.org

 

[swgrmfg]

One more thing...

if you get too carried away with 'turns' without credible calculations, you might end up with a resonnant circuit. Ideally, LC should be tuned to a fractional harmonic that does not exist (like 150Hz on a 60Hz system).

 

[Skogsgurra]

The resonance risk is not a problem with current limiting reactors for PF capacitors. The resultant inductivity will never get close to what is needed to get into resonance. It is only when you add an iron core that you need to be careful not to have resonance.

Gunnar Englund

http://www.gke.org

 

[Warpspeed]

One way to construct the coils is as a flat spiral "pancake" construction. The strong repulsive forces generated between the turns then largely cancel. Only the inner and outermost turns try to explode radially away from the other turns.

If the coils cannot be separated much physically, mount them in three different planes to avoid magnetic interaction as much as possible. It will look rather ugly, but it is practical.

I have never worked for an inductor manufacturer either, but I have been known to wind up my own high current air cored inductors during a full moon.

 

[jghrist]

skogsgurra,
What is the meaning of

make the A and B phases "co-operate"

If the idea is to add inductance to limit high frequency inrush currents, the inductance of a coil is the same no matter which direction it is wound.

 

[Skogsgurra]

jghrist,

The OP said: "They are going to bunch them together and tie-wrap them. Does the A and B phase have to be reversed in terms of coil orientation in order for magnetic fluxes to interact properly?"

That means (as I interpret it) that he will be coiling up the three-phase cable in one coil, not three separate coils. That will work, but it is necessary to reverse one of the current directions or the three fields will cancel.

This arrangement is not ideal, since the C phase will be left "on its own". But it will increase the HF impedance in the circuit and reduce the normally very high transient when closing the breaker/contactor and will usually avoid welding of the contacts.


Gunnar Englund

http://www.gke.org

 

[scottf]

You only need to worry about reverse winding the center phase in a 3-phase stacked configuration, not in a single-phase mounting arrangement.

 

[jghrist]

I hadn't thought about wrapping a 3ø cable to make an air core reactor, but I guess that's what the OP means. I don't think this would be practicable. Even reversing the current in one phase leaves you with oddball mutual inductances to worry about. If I were going to try to make my own air-core reactor, I'd make 1ø ones.

 

[CoryW]

Thanks for all the info people, but there's a little miscommunication. The plan was to make three seperate 1-phase reactors and tie-wrap them all together for space restrictions.

I was wondering if the close proximity of the reactors to one another would cause problems with the inductance of the reactors.

Also, does B-phae have to be reversed?

 

[Skogsgurra]

Yes, that would be very similar to using a three-pase cable coiled up (if the three coils are close enough). If you reverse current in the middle coil, you will have a not-so-perfect but working reactor.

Gunnar Englund

http://www.gke.org

 

[Warpspeed]

Are you sure coiled three phase cable would work ?

It would then be rather like a three wire common mode choke, but would have little or no differential mode inductance. All the inrush load currents would cancel!

The three inductors definitely need to be far enough apart to work independently as three separate single phase inductors. As I mentioned earlier, they could be physically close if mounted at right angles to each other in three different planes (X,Y,&Z).

 

[Skogsgurra]

Yes, it works.

But you have to make one of the currents go "the wrong way" so the fields do not cancel (the common mode effect you are mentioning). That is what the whole thread has been about. OP says "Does the A and B phase have to be reversed in terms of coil orientation in order for magnetic fluxes to interact properly?"

As said several times, it works, but it is not an ideal solution.


Gunnar Englund

http://www.gke.org

 

[scottf]

Cory-

Depending on the inductance of the coils you want to make and the expected in-rush current, you shouldn't simply tie-wrap them all together.

You need to consider the mechanical forces that one coil places on the other during a fault/in-rush. Also, if the continuous current is of a decent level, you need to be cautious of the heating effects of the reactor and if in a confined space, the reactors may induce a current in any nearby closed loops, thus creating the chance for touch potentials.