"Weird" Application of Line Differential Protection

[gordonl]

Let me first describe the application:
35kv shielded cable, direct buried approx 1km, 6-1C-500MCM. Line differential relays are 3 element, dual fiber channel, with integrated overcurrent and distance protection. My system has a fault level of about 20KA symetrical, and the system is resistively grounded with two sources of ground fault current on the system rated at 400A each. At one end of the line I only have 1 phase - phase PT. The line differential relays are using 1200/5A CT's and the pickup is .2pu, or 240A, and also has a 50/5A ZSCT at each end, connected to the ground current input of the relay.

My problem is I would like to increase the sensitivity of 87 relay for in zone ground faults to decrease damage to the cable for a cable fault, but the relay only compares the three individual phase currents, and not the ground current input. One of my two ground fault current sources is a signifigant distance from the cable installation (10km) so the amount ground current it would supply is reduced.

The "weird" idea I had was to use one of the phase inputs of the relays for the ZSCT's to provide me with 87P and 87G protection. Phase would be connected to the ZSCT at each end, and phase B & C CT's would be tied to their respective phase current inputs at each end. I would then have diff protection for in zone ground faults, and diff protection for any possible 2 or 3 phase faults. I've calculated the charging current at 7A, so that shouldn't be an issue. I would then increase the pick up from .2pu to .5pu so my 87G setting would 25A, and my 87P would be 600A. I'm not using overcurrent or distance protection in the relay.

Can any body think of any potential pitfalls with this arrangement? (I realize the reading's for the ground current would be off because the ratio is entered at 1200/5A, but I don't think this will be an issue.)

Also I used 6*charging current for my calculation for 6 conductor, should I have?

Thank You,
Gord

 

[jbartos]

Suggestion: 35kV line 7A charging current appears to be on very low side. What kind of system grounding approach do you use?
6*Icharging for 6 conductors is o.k. since the capacitance/phase of each conductor is supposed to be added together for the parallel 6-conductor run.

 

[busbar]

I'll bet there are some relay-manufacturer applications-support folks that may, if it's possible with their product, enjoy the challenge and give setup adivce as needed.

Is there a communications channel between the line ends?

 

[jbartos]

Suggestion: The dual fiber channel mentioned in the original posting implies some sort of communication that is redundant.

 

[gordonl]

jbartos,

I think my charging current calculation was right, I used capacitance tables from the Kerite website, if you think there is a problem I'll review. (Note length is only 1km) (Also note 35kV capacitance is much lower than 5 or 15kV as insulation thickness increases)

As for 6*Icharging, I was pondering the effect that the phase shift in the three ground currents would have on total Icharging seen at the core balance CT, will they cancel out?

 

[gordonl]

I implemented this system yesterday without complication, I beleive this will give me much better cable protection. But I have one observation which I would appreciate some thought on:

After being put into service the differential current on the ZSCT was zero with about 1.75A of through current. Why didn't I see the charging current which I calculated at 7A.
I suspect because the charging currents are balanced on the three phases they cancel in the core balance CT, otherwise I don't know.

 

[busbar]

As for not seeing shield current reflected in the zero-sequence CT, it may be that shield-bond conductors are (properly) routed "back through" the CT core.

On less-than-calculated charging-current readings, are cable shields grounded at both ends? It's likely that tables intend to list worst-case conditions.

 

[gordonl]

I haven't inspected this installation, but we've been quite diligent in keeping shields returning through the ZSCT, but they are all shorted together before returning through the ZSCT to ground.

The shield is connected to ground at one end only

Why should charging currents look like ground current to the ZSCT, in essence is it not a balanced three phase capacitive load whose summed currents equal zero?