Effect of cable shield grounding on CTs for KWH meters

[hidalgoe]

Our local utility has told us that all high voltage cable shields must be grounded before the CTs. Ostensibly if the shield is grounded after the meter an error will be introduced.

We are not interested in revenue; all we want are ballpark KW demand and KVA measurements. If the amount of error is small, then so be it. The cost of rearranging the cable shield ground will very high.

Comments please.

 

[electricpete]

It sounds like the cable will be grounded in 2 places. One at the utility supply transformer, and one at the customer near the metering ct's (utility side or customer side). When shield is grounded in 2 places a ground loop is formed which can conduct a circulating current. If it is a 1/c cable grounded twice than circulating current can get quite high... in theory approaching a current equal and opposite to the main current (total current through the ct would read zero in that case).

One alternative MIGHT be to simply lift the ground at your end, rather than relocating it. In that case the cable will still be grounded and no circulating current will flow, but you'll have to be careful to understand whether dangerous votlages might appear on your ungrounded end of the shield during fault scenario's.

 

[gordonl]

I think the question we are talking about here is the location of the CT. The CT should be below the termination, I haven't seen an installation where the shield is striped below the CT, and would be worried. To prevent inaccuracy due to shield currents flowing through the CT the shield ground wire should pass back through the CT so the two cancel out. Pete's point of the magnitude of the circulating current should be observed as well, but for short distances it isn't an issue. Southwire has some guidelines on this

http://www.southwire.com

Is there any other issue concerning the presence of the shield and ground lead in the window?

 

[busbar]

A significant warning to not ‘double back’ the shield connection is because it may desensitize overcurrent/differential protection. Seems that with a bit of caution shield current wouldn't be difficalt to measure.

 

[electricpete]

Gordon's answer was a good one. Simply sending the shield ground wire back through the ct in opposite direction would cancel the current, and seems like a very practical solution. Also I noted the Buff book page 379 item 9.3.6.2(3) says that cable shield should be routed through the ct in this manner. The context there is ground protection but the principle is the same.

I'm not sure I understand busbar's comment about desensitization.

 

[busbar]

To clarify-- if a shielded (MV) cable is bonded to ground on opposite sides of a CT, a parallel/shared path may exist, allowing only part and not all of the fault current to be sensed by the CT. Does that make sense?

 

[gordonl]

The shield wouldn't be bonded to ground on both sides of the CT, only downstream of the CT with the bond wire passing through the window to the shield. If the was a ground fault to the cable shield downstream, it would still be sensed, because it would be passing the ground current through three times, once in the conductor, once in the shield, and once at the bond wire. Even with both end of the shield grounded the ground current source end would still see full ground fault current because it would flow in the conductor, but the shield and bond current would cancel to leave only 1pu ground current.

Hope I didn't ramblr too much.