Voltage rise phenomenon from shunt capacitors

[zackbanks]

I am a switchman at a transmission utility. I am having a hard time finding someone here who can give a basic explaination of the phenomenon of voltage rise/decay from inserting shunt caps/reactors. I have completed my theory classes to include solving complex RCL circuits. I will give it a try: By creating a tank circuit, the line current will drop by the amount of reactive current inserted. Also by adding another parallel branch to the circuit, overall impedance should also drop. So with a basic look at Ohms law, if current and impedance go down, the result would be a rise in voltage? (inserting Xc.) That being said, I have also read that in a tank circuit at resonance, impedance is virtually infinate. And, this explaination doesn't work for a voltage rise from inserting Xl. It would be great if I could prepare a breif explaination of this for myself and my peers.

This website is great!

 

[davidbeach]

Forget tank circuits and all that, the power system frequency is too low. In simple explanation, look at it this way:

1. The line consumes reactive power just by being energized.

2. Many loads can require a considerable amount of reactive power as well.

3. The line impedance is mostly inductive (reactive).

4. The reactive current that feeds the reactive power creates a voltage drop across the reactive portion of the line impedance.

5. Caps are a source of reactive current.

6. So the amount of reactive current sourced by the caps is no longer flowing from the normal source to the caps and the voltage drop to that point no longer happens. Voltage rise in the form of reduced voltage drop.

7. If you put in so much capacitance that the reactive current starts flowing from the caps back toward the source then the voltage at the caps can be higher than at the source. What you are seeing in that case is actually a voltage drop from the caps to the normal source. The caps are now (at least locally) the dominant source of reactive current and your voltage drop/rise calculations have to consider that.

 

[Skogsgurra]

You are both correct. But I see no reason to forget about tank circuits. Resonances occur at all frequencies, low or high, and a situation where the capacitor delivers exactly the same amount of reactive current that is consumed at the load end of the line is, in fact, a complex RLC circuit at resonance - although with a very low Q.

So, if that makes life and understanding easier, use the tank circuit model. But remember that Q is low. The 'antenna' (the energy spender) is in the circuit - not outside it.


Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[m3ntosan]

1. The line consumes reactive power just by being energized.

?

May you grow up to be righteous, may you grow up to be true...

 

[davidbeach]

Yep, it's true. Check out any of the Power System Analysis references in faq238-1287. Generally ignored at 230kV and below, but present for all AC circuits.

 

[zackbanks]

How does this sound?

Adding shunt caps and creating a reactive current "loop" or tank circuit between the shunt caps and the inductive load actually reduces the current through the line from the source. This lower current has a lower voltage drop across the impedance of the line.

When inserting a shunt reactor, you increase the current on the line (no tank circuit and an added parallel impedance to the load) thus increasing the voltage drop across it.

I hope those are correct.

 

[Skogsgurra]

Still like the step-by-step wording by davidbeach better. Why do you need to condense the wording? Some things are better dealt with without simplification.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[m3ntosan]

davidbeach,

Could you provide me with a more direct reference?

That's a revelation for me, there's no need for system operators to take lightly loaded lines out of the system during the night or weekend to keep the volts down, because an energized line actually consumes reactive power! What about SIL/ natural load?

May you grow up to be righteous, may you grow up to be true...

 

[zackbanks]

Skogsgurra-

Being a non-engineer in a group of the same, it helps to simplify an idea so the masses (including myself) can grasp it. We have (some not all) system operators who can only describe vars simply with the foam on the beer explaination which only covers capacity on the line. They understand caps in, voltage up. Reactor in voltage down. Peel banana, eat. I don't want to be a monkey, but I'm also in no position to pretend I'm an engineer. You'll be happy to know I just operate, I don't design any parts of the system, or protection.

Thanks for all the help people.

 

[m3ntosan]

That’s the disease of the new wave. Every carpenter is now a system operator. Too many ... just follow procedures and press the right button in the EMS, not too much engineering judgement involved.
Before...the system operator was considered a top position, in order to qualify for a system operation position you needed to have very good experience in power system protection, switchgear maintenance, switching, planning and have distribution network operator experience.

On the other hand, reactive power is not an easy subject to explain, there are some very good analogies and IMHO the one with the beer and foam is useless, nothing to do with reactive power.


May you grow up to be righteous, may you grow up to be true...

 

[davidbeach]

I left out the special issues of EHV lines to avoid confusing the issue. But since you insist:

8. Every line also has its own distributed shunt capacitance, both between phases and between each phase and ground. For a variety of reasons this can be ignored at distribution voltages and is generally ignored at transmission voltages at transmission voltages through about 230kV, but in EHV circuits it can have real impact. The line itself can supply the first increment of excitation current necessary. Under the right conditions it can supply more reactive current than the line needs for excitation and supplies reactive power to the rest of the system. Depending on when and where, this can either be a benefit or a detriment to system operation.

 

[m3ntosan]

davidbeach,

My understanding was that zackbanks is working at transmission level and also I’ve never seen shunt reactors below transmission level.

For HV lines is always the case to produce reactive power like a capacitor when energized and when loaded below natural load/SIL.
A practical example a 220 kV OHL with 1x 450 mm2 conductor it will produce roughly 16 MVAR /100 km and for a 400 kV rated OHL functioning at 220 kV with 2x 450 mm2 you’ll get close to 30 MVAR/100 km which will affect more or less the volts on the system depending on the fault level.


May you grow up to be righteous, may you grow up to be true...

 

[jghrist]

It might help to draw out the vector diagram of voltage drop through the line impedance. The attached diagram shows the sending end voltage VR and receiving end voltage VS with an inductive load (negative angle ø). Note that the magnitude of VR is less that VS, there is a voltage drop. If the load were made capacitive by the addition of a capacitor bank, ø would become positive and the IR and IX vectors would rotate counter-clockwise. At some value of ø, VR will become larger than VS, a voltage rise.

 

[dpc]

m3ntosan,

FWIW, we do see shunt reactors on extremely long underground distribution circuits (12.47 kV and 25 kV) on occasion.

David Castor

http://www.cvoes.com