Query on High-Impedance Bus Differential Relay Operation

[GRAEE]

Dear All,

Good day!!! Hope everyone doing well now.

I have a query regarding the effect of power cable shield ground which is covered by the high-impedance bus differential relay with a window type CT as current analog input.

Our bus is having 6 distribution feeders each with breaker and power cable length to about 10meters as take off and distributed through overhead lines. The zone of the bus protection is covered by window type CTs which is partially extended to the power cable length. The power cable shield is being grounded only at one, near the bus, and the other end (towards overhead line) left hang.

My question is, does will affect the operation of the relay when external faults (fault at overhead line) occur? Considering our power cable shield is grounded that way.

Please advise.

Thanks & Regards,

GRAEE

 

[davidbeach]

If the shield is only grounded once, there is no current through the shield; with no current through the shield it can't have much impact on the differential.

 

[Marmite]

The cable shield ground needs to be connected below the CT, or if it is above the CT then the connection needs to be brought back out through the window of the CT so that the CT only measures the summated currents in the phase conductors. Otherwise a fault out on the network with stray current returning back via some of the cable sheaths could cause operation of the busbar protection.
Regards
Marmite

 

[GRAEE]

Dear davidbeach & Marmite,

Thanks for the enlightenment.

The cable shield ground is connected above the CT but only in one end, the other end is not grounded. As explained by davidbeach, we expect no return current for external fault.

Regards,

GRAEE

 

[Marmite]

As long as your cable metallic shield remains isolated from the ground surrounding it then you what you say is true. I don't know what type of cable you are using or the nature of the installation whether ducted or direct buried, but if there is a possibility of the shield being in direct contact with ground due to a damaged oversheath etc then there is a possibility of stray currents returning via the sheath to the substation, which are not measured and cancelled out by the CT on the same circuit, and hence operate the bus bar protection. An example might be a fault on an adjacent circuit such as a downed conductor.
Regards
Marmite

 

[GRAEE]

Dear Marmite,

We're using single core unarmoured XLPE Insulated PVC Oversheathed Cable and this cable is lay down on the trench.

What's the best practice?..Is it to connect one end of cable shield above the CT and brought back out through the window type CT and connect to ground? Or we left both ends of cable shield floating and ungrounded?

Regards,

GRAEE

 

[stevenal]

Will the jacket withstand possible transient shield to earth voltages during through faults? As far as I know, sheathing is for physical protection and is not tested for its dielectric strength. I would bring the shield back through the window.

 

[rcw retired EE]

Bring the shield back through the window to ground. If a cable fault occurs, the fault current will flow in the shield to the ground connection. If it the connection is above the CT, the current will trip the differential. This may be OK if the intent is to include the cables in the differential zone of protection.

If you don't want to trip the bus differential on a cable fault, route the shield ground back through the CT.

 

[GRAEE]

Ok, we have to route back out the shield to the CT window.

I have another query on bus differential relay performance where it covers the whole length of the cable. The shield is within/between the zone of the bus differential and one end of the shield is grounded. Would that also affect the operation of the relay?

Thanks

 

[jghrist]

You would still route the cable shield ground back through the CT. Consider a ground fault between the conductor and the shield, with all of the return current flowing through the shield. If you grounded the shield on the source side of the CT, then no net current would flow through the CT and the differential would not operate.