Failing Sheath Test - Reason for Concern? 1

[FEinTX]

Background: I'm working w/ a 34.5 KV 3-phase collection system consisting of multiple circuits each containing 12-13 2 MW turbines. The cable varies from 3/0 at the ends of circuits to 1250 kcmil at the substation, and it is direct buried in a sand bedding material.

Problem: Before the trenches can be backfilled, our EPC requires their subcontractor to conduct a sheath test to prove that the cable jacket was not damaged during the installation process. Initial tests of 350 kcmil cable went well, but when the first 3/0 run was tested, 2 of the 3 cables failed. A second megger was brought in, but it agreed, and additional runs of 3/O also showed 1-2 failed cables in each case. At this point, the cable manufacturer was brought in. They had a piece of equipment that put a pulse on the neutral and then they walked the trench setting up 2 probes about 12' apart and took readings which could detect current leaks. Three areas were dug up, but no obvious problems were found. Some nicks were present, but none seemed deep enough to reach the neutral conductors.

The megger test is showing typically 1-5 Mohm resistance on the failed cables while passing cables will read 10 Gohm or more (tested up to 10 KV). These nicks seem pretty small and the fault is starting to point toward the cable manufacturer. Now, the cable manufacturer is saying that the leakage must be coming through these nicks. They are also saying 10 KV is too high for a sheath test and that this may be damaging the cables. After all this, they go on to say that there is no reason to worry and that this sheath test really isn't necessary.

I asked about possible water penetration and its long-term effects. They said that they didn't expect there was any water penetration from such nicks, and even if there were , it would not affect the performance or life of the cable. They seemed willing to put such a statement in writing, but I doubt they'd offer an extended warranty.

Questions:
(1) Is 10 KV too high for a megger test on this neutral? If so, what would be a safe level?

(2) Would you be concerned about water penetration to the neutrals?

(3) Can small nicks create leakage points during these tests, or should I insist they look for an area where the neutral is clearly exposed? [Note - we have found numerous jacket nicks that have had to be patched due to poor installation practices, and I fear some were missed prior to burial.]

(4) Any other thoughts/suggestions?

Thanks for the help!

FEinTX

 

[davidbeach]

I don't have any standards at hand to check real values, but I would be surprised if the neutral is insulated to more than 600. And, I've never heard of testing 600V insulation at 10kV. 5kV seem a more likely value.

 

[ausphil]

Firstly, the main standard that talks about tests on the oversheaths is IEC 60229. It states that the test voltage shall be 4kV per mm of oversheath thickness, up to a maximum of 10kV. The test duration is 1 minute. Not knowing the thickness of the oversheath that you specified, we can't really state whether the test voltage is too high or not.

I am assuming that you are talking about your 34.5kV cable and the metallic sheath of this cable when you start talking about pulsing the "neutral" of the cable. Fault location of this type is common, and you probably do have a puncture through the oversheath at that point (even if it was only a small nick initially) because during pulsing, you will probably have punctured it though anyway. Have you exposed the underside of the cable at the fault point to see if there is any damage there?

I would be wary of nicks in this oversheath, because water will get in, and you will have a higher probability of failure because of this. If you can do repairs on it, do so (and since you have already dug it up, you might as well).

The cause of the problems is probably going to be hard to prove, unless the nicks were identified prior to laying the cable. The manufacturer will probably say that it was due to the laying process, and he would be well within his rights to do so, unless evidence to the contrary is shown.

This problem shows the true value of testing.

Also, I assume that you must have some cover over the cable to do the sheath test (ie not fully backfilled, but not uncovered).

 

[FEinTX]

Ausphil,

Yes, when I speak of the "neutral", I am talking about the concentric copper wire sheath. And yes, the cable is covered during this testing--it has 9" or more of the sand padding material so problem areas are relatively easy to dig up and inspect.

FEinTX

 

[stevenal]

What you are calling a sheath test sounds more like a jacket test. You are meggering from neutral to ground across the jacket? Cable manufacturer is right, the jacket is not designed to insulate. It is there to physically protect the concentric and to keep water from corroding the concentric. Your testing may have compromised these functions.

 

[oldfieldguy]

I checked the NETA Acceptance Test Specification for new cables and found no reference to the sheath (jacket) test, although I have seen clients who required it.

old field guy

 

[FEinTX]

Update - According to the IEC standard ausphil mentioned (thanks!) this particular cable w/ its 45 mil (minimum) jacket should have been field tested at 4.6 KV (not 10 kV). It's interesting that pre-installed testing is made at a much higher 12 kV (8000 V/mm using nominal jacket thickness), so the standard is clearly allowing for a certain amount of nicks and cuts during the installation process.

Further inspections have shown numerous burn marks immanating from small nicks on the cable. The testing was not done in a systematic manner, so we don't know if these breeches in the jacket are due to the use of 10 kV during the test or if the areas failed at a lower voltage level. I suspect that we would have had some failures even at 4.6 kV, but likely many less.

I'm going to start a new thread addressing the importance or non-importance of having a water-tight jacket in these power cables. Thanks to all who replied!

FEinTX