Hi everybody,

I've a little but tricky question for a assignment. It's about designing a RLC single tuned harmonic filter to insert in a three-phase circuit at the point of common coupling in order to cut off the harmonics due to the presence of non linear components. The question is:

How the designed filters should theorically be connected (star, delta, with or without an earthed neutral) and why? Which are the benefits for that configuration instead of the other ones?

Thank you in advance, any suggestion would be helpful.

 

Hi,

Filters often have to go in line with something, or at least partially in line.
Maybe you can draw a schematic so we can see what you really want to do.

 

Certain current harmonics of the fundamental will (possibly) be in phase. What issues might then arise for a three wire star topology?

 

Hi,

Filters often have to go in line with something, or at least partially in line.
Maybe you can draw a schematic so we can see what you really want to do.

Filters are simply series R-L-C. So, this is the circuit:

The harmonics are introduced by the diode bridge rectifier. As you can see the filters F at the PCC are grounded Y-connected. This should be probably the best configuration but I didn't get the reason. Why do I prefer this configuration to a Delta connection, for istance?

 

.....Why do I prefer this configuration to a Delta connection, for istance?

In post #3 I proposed a question which might lead you to think about such a choice. Did you give it any thought or are you simply hoping to have your answers provided without any effort on your part. After all, this is homework help - not do my homework for me.
You provided a schematic - without comment as to which harmonic would be targeted by the filter. What are your thoughts?

 

In post #3 I proposed a question which might lead you to think about such a choice. Did you give it any thought or are you simply hoping to have your answers provided without any effort on your part. After all, this is homework help - not do my homework for me.
You provided a schematic - without comment as to which harmonic would be targeted by the filter. What are your thoughts?

I'm sorry, I've just not seen your reply. The filter is targeting the odd harmonics except for the triplen, in particular they are the 5th, the 7th, 11th. My originally thought was that in the presence of an earthed neutral the current which produces the harmonics would be drawn out of the network, while with a delta connection it cannot (since this configuration don't give any access to a neutral point). But as far as I understood you are trying to point out the issues related to the star topology, which means the opposite of what I supposed.

 

Hello again,

Most definitely with a three phase or even single phase rectifier the addition of a series (or more) inductor cuts harmonics because the diodes then conduct less abruptly.
It's not that hard to see the difference if you look at a very low impedance source and rectifier/capacitor set and then compare that to the same circuit but with an inductor in series with either the transformer secondary or the output of the (4 to 6 ) rectifiers.
So just the addition of an inductor in series with each phase would show a marked decrease in harmonic content.
Calculations are not that difficult, especially in the single phase case.
You can also do simulations with and without the inductor to see the difference.
The current peak goes down, and so the current waveshape smooths out.
Many higher current rectifier bridges will use an inductor with the ability to 'swing' it's inductance. As the current goes up the inductance comes down which helps improve efficiency at increased load while still improving power factor.

 

I'm sorry, I've just not seen your reply. The filter is targeting the odd harmonics except for the triplen, in particular they are the 5th, the 7th, 11th. My originally thought was that in the presence of an earthed neutral the current which produces the harmonics would be drawn out of the network, while with a delta connection it cannot (since this configuration don't give any access to a neutral point). But as far as I understood you are trying to point out the issues related to the star topology, which means the opposite of what I supposed.

Actually it wasn't my intention to discredit the star topology. If you refer back, I was specifically thinking about harmonics which were effectively in phase and had no conducting path to a system neutral - as would be the case with a three wire system.
On another point are you looking at reducing all of the the 5th, 7th & 11th harmonics? One would need more than a single shunt RLC branch at the pcc to achieve an effective reduction of all the aforementioned harmonics.

 

Why do I prefer this configuration to a Delta connection, for istance?

I believe it doesn't matter what topology you use. The goal with a simple RLC passive filter is to reduce the harmonic impedance at or near the harmonic frequency of interest. It boils down to the appropriate selection of the RLC values for the particular topology which can be either delta or star (with or without neutral available). I would suggest you look at a specific design for just one (harmonic) frequency - such as the 5th harmonic. A web search of this very topic (5th order harmonic reduction) may well prove fruitful.