Reactor vs Resistor 1

[AggiesNM]

Hi
I've been working on a short circuit study for a collector substation for a wind farm site. I want to specify a neutral impedance in the secondary winding of the GSU auto-transformer which is rated at 99MVA (base @65C), 138kv-34.5kV (14.4kV tertiary). Some here prefer putting in a reactor while others (like me) prefer a resistor. Is there a reason why an air-core reactor would be useful at the collector sub?

Also, I have a go-by specification for the reactor. It states "2.65mH 1ohm, 600Amp (10sec?), 19920V, 200kV BIL, 60hz" for the reactor. I have been doing the simple V/I method to determine the reactance (19920/600 = 33.2ohms). How does one come up with 1ohm reactor that can limit fault current to 600Amp? I seem to be missing something here.

Any help is appreciated.

Regards.

 

[rcw retired EE]

I believe the reactor spec is for a 1 ohm, 60 Hz reactor that can carry 600A for 10 seconds. Z= jL(f x 2pi). 0.00265 Henrys x 60 Hz x 2pi = 1.0 ohms.

Your calculation negelcts the zero sequence impedance of the transformer and the system.

 

[odlanor]

calculate the impedance of the resistor or reactor.
Sometimes the resistor is too large to be unfeasible manufacturing. Consult the manufacturer.

 

[AggiesNM]

@rcwilson,

I believe I don't need to specify the zero sequence impedance of the transformer and the system to specify a reactor. The reactors impedance alone should limit the fault current to 600Amps. I guess that is what a manufacturer expects when I indicate that rating.

However, getting back to the topic, a 2.65mH inductance indeed amounts to 1ohm inductive reactance. How is this reactance sufficient to limit fault current magnitude to 600Amps?

@odlanor: That is a possible answer. thanks.

 

[Marmite]

Reactance earthing is typically reserved for applications where there is a desire to limit the earth fault duty to a magnitude that is relatively close to the magnitude of a three phase fault. Use of neutral earthing reactors to provide this fault limitation will often be found to be a less expensive application than use of earthing resistors if the desired current magnitude is several thousand Amps. Are there any single phase to neutral loads connected? This would be unusual in your situation but may influence the choice of a reactor over a resistor.
In a reactance earthed system, the available earth fault current should be at least 25% and preferably 60% of the three phase fault current to prevent serious transient overvoltages. This is considerably higher than the level of fault current desirable in a resistance earthed system, and therefore reactance earthing is usually not considered an alternative to low resistance earthing.
In your situation I can't see why you would go for a reactor over a resistor, and it would probably cost much more in any case.

Regards
Marmite

 

[LiteYear]

When a reactor is used for earth fault limitation it is often known as a Petersen coil. Sizing it is a bit different to sizing a resistor - instead of thinking of the single phase to earth voltage appearing across the neutral impedance, think of the reactor appearing in parallel with the distributed capacitance to earth of the other two phases. There are two paths for the fault current to flow once it passes through the fault and enters the earth - either it flows back through the reactor to the neutral point, or to flows via the distributed capacitance of the other two phases, back through those phases to the neutral point. The job of the reactor then is not to limit fault current via I=V/Z, but to match the complex impedance of the distributed capacitance so the two paths cancel each other out.

There's an extra complication because the faulted phase causes the voltage of the other two phases to increase by root 3, so the current through the two capacitive paths actually turns out to be 3 times that of the reactor current. For complete fault current cancellation then, you need L = 1 / (3*ω*ω*C). Where C is the distributed capacitance from phase to earth.

Now your L of 2.65mH is unlikely to be enough to completely cancel the fault current (that would suggest a distributed capacitance phase to earth of almost 1mF). Is it enough to restrict it to 600A? Hard to be sure without knowing that capacitance value.

To answer your first question - a couple of reasons to go reactor: it can push the power factor of the fault current towards unity, which helps with extinguishing the arc because the voltage goes to zero at the same time as the current; and as already mentioned, the amount of material required can be favourable, depending on the actual current limit.

I think from your description that the secondary will be on the grid. In that case I'd be checking with the utility involved and probably going with their recommendation. What do they do at that power/voltage level, for that distance of transmission, for that region of the grid?

 

[AggiesNM]

@Marmite
The GSU transformer is connected to the utility grid on the high side and to the wind turbines on the low side. There is no single phase load as far as I see. Are you concerned about the unbalanced loads?

I see the benefit in sticking with a resistor.

Regards.

 

[AggiesNM]

@liteyear
You are talking about high impedance reactor grounding (using a peterson coil). I would like to stick with the low impedance grounding scheme. I believe there are no complexities like matching reactance to distributed capacitance in this case. Also, the low side is connected to the windfarm.

 

[odlanor]

1 - I think the utility of 138kV system does not allow a source distributed generation (wind-farm) helps to increase the level of short-to-ground 138kV system; contribution current of wind-farm powerplant to the fault can not exceed 600A .

2 - NBI 200kV indicates that the 138kV transformer winding should have graduated 650kV or 550kV BIL at phase terminal and reduced to 200kV BIL at neutral terminal. This is a grounded system with middle impedance (600A) in the load terminal side (wind-farm)

3 - With BIL graduated you cannot use Peterson coil.