A two mile underground 69kV XLPE tr

[stevenal]

A two mile underground 69kV XLPE transmission line is located in a coastal location. Terminators and arresters are located at both ends. Last month following a dry spell and when the fog rolled in, the line experienced an A to ground fault. Burn marks were noted at the ground end of the A phase terminator at the load end of the line. Cable tested okay, so the terminators were cleaned and the line was reenergized. Conclusion was that the fog wetted the accumulated salt causing a flash-over, although we don't recall a similar event for similar conditions.

Last Friday we tripped on A to G again, also during fog. We noted fresh looking burn marks on the switch adjacent to the load end A phase terminator. Reenergization was unsuccessful, line tripped right out. Using the thumper, we found a bad spot conveniently located adjacent to splice inside a vault about a thousand feet from the load end termination.

The question is: Did the flash-over cause the cable fault, or did the cable fault cause flash-over? The relay provides no fault data other than targets. Thanks.

 

[MikeHalloran]

Remember Hypalon? DuPont's chorosulfonated polyethylene? They sold the hell out of it for, among other things, spark plug wires. The stuff seemed to have a very finite life; it would work okay for a while, then it just suddenly got funky, and the car wouldn't run for crap. After a couple of sets on the same car, I started paying extra for silicone insulated wire. ... as did the car manufacturers.

I see that XLPE has become quite popular for HV transmission cable insulation.
I wonder if history is repeating itself.

How do you feel about replacing two miles of cable?


Mike Halloran
Pembroke Pines, FL, USA

 

[MKFPE]

Based on the limited information my money is on the flashovers stressing the cable in the area of the splice and causing the fault there.
What is the fault current?
Size of cable?
Is the 69kV system effectively grounded?
Ratings of arresters?
Two miles of 69kV cable could cause some interesting things to happen.
Maybe a look with ATP or similar transient analysis program is indicated.

 

[stevenal]

Fault current gauged from remote PQ metering was about 3900 during the Aug and Friday events, 3400 when the cable was tested by reenergizing.

Arresters are station class: 42kV MCOV

69kV radial system is single point grounded at the source.

Cable is 500 kCM AL.

 

[hdp]

Is there a possibility that the burn marks at the grounded end of the termination were due to voltage rise of the shielding under fault conditions?

 

[jghrist]

How was the cable tested after finding the initial terminator failure? Latest recommendations are to not

 

[stevenal]

Not what?

We used a meg-ohmmeter.

Shield is grounded both ends. Shield should rise with GPR.

 

[7anoter4]

In my opinion there is no connection between these phenomena. The shield induced voltage -if the cables are laid parallel without any rotation
at the far end [2 miles] will be around 1kV at 3700 A -not enough to produce an insulation break down .
By the way, Mike, I don't remember using Hypalon for insulation. But as jacket it was an excellent material-we have Kerite EPR insulated
and Hypalon jacketed 2000 MCM copper cable in operation for more than 30 years.
Also XLPE TR is a very good material. It depends, naturally, on execution and installation.

 

[jghrist]

Sorry - posting problem.

The question is: Did the flash-over cause the cable fault, or did the cable fault cause flash-over?

How was the cable tested? It is no longer recommended to hi-pot test aged cable. It can cause incipient faults.

 

[stevenal]

5kV meg-ohmmeter and thumper shouldn't stress a 69 kV cable much.

 

[Zogzog]

I would not think a 5kV megger would find a fault in many cases either. I suggest a VLF - Tan Delta test or PD.

 

[MKFPE]

I didn't express my thoughts very well earlier.
I don't think induced voltage on the shield is involved, rather I am wondering if the through fault current of the two faults at the termination stressed the area of the splice mechanically or thermally.
Perhaps the splice had unknown issues. I have seen splices fail after though faults before.
So additional things to consider would include normal load current, if it is small a poor connection can survive a long time; and fault clearing time. What type of protective relaying is used?
And, what is the definition of "adjacent"?
As an aside we use EPR only. I like the fact that it is lossy at higher frequencies and helps to dissipate fast transients in addition to no treeing. We have a lot of it (miles) installed at 12kV with no problems.
Finally, years ago there was a stamping plant owned by one of the "big three" auto makers in our area that followed a service company's recommendation to Hi-Pot their 15kV cables as "preventive maintenance". I think they cranked up the voltage to the level recommended for new cable. They ended up with the plant shut down for a couple of weeks to repair the damage.

 

[stevenal]

I think you may be correct regarding the stress on the cable near the splice. See the video above.
Normal load current is small, since cable feeds a single 10 MVA transformer that rarely sees loads at half that capacity.
Relay tripped on instantaneous ground overcurrent.
Adjacent means the switch in question is connected to the termination with only an angle adapter between.
We've also gone to EPR for distribution.
Relay tripped on instantaneous ground overcurrent.