thread238-153501: Testing of medium voltage solid dielectric cable.

I read through this old forum post but it did not quite answer my question. I was asked last week to find out what the standard is from NETA or IEEE or some other reliable source even a VLF equipment manufacturer referencing a paper that in some way says how often VLF testing should be conducted. So far I have come up empty handed. I was also asked about oil-filled transformers and figuring out a standard testing on those to which I did find many recommendations (even in here) about testing the oil annually or more frequently if a problem is suspected and going by either a 3 or 5 year shut down check.

Can anyone tell me and/or reference a link/paper to the frequency of VLF testing and what is supposed to be included in the scheduled shut down testing on the transformers?


Thanks and if I am not clear on something in here bare with me I will try to clarify if needed/if I find out more.

also I guess the title should be Cable testing with VLF (frequency of)

Sorry about that.

Have you read the NETA specs? They will tell you what tests should be done. NETA also has recommended intervals for testing.

• http://files.engineering.com/getfile.aspx?folder=7ec86c19-239b-4a53-85d9-c1

Thank you sir. That is precisely what I was looking for! I had another NETA form that gave testing information but nothing on the time frames.

Glad to help, that matrix is a free download on the NETA site.

The other thing you need to clarify is the actual VLF test you are using, and the usefulness of this test for the failure modes that occur in the cable you are looking at.

Is it just a high voltage withstand test? If you are just doing a HV withstand test, I would question why you would want to take out a cable at MV that is performing, and subject it to an overvoltage, which may cause the cable to fail, which may be a hard to find fault because it only had the capacity of the test set cause the damage (as opposed to having the power from the system break down the fault and make it much easier to find). Also, passing a HV withstand test doesn't necessarily give you any idea of the condition of the cable for the medium or long term in service, it really gives you the confidence that it will not fail on energisation or soon thereafter ( a condition that you have only subject the cable to just to perform your test).

Are you measuring power factor or tan delta with the HV test (to try to determine water tree activity)? This measurement gives an average of all the problematic and good sections of the cable. If you assume that your cable has grown its problems evenly, then it may be helpful to determine if you want to replace it or not, but it cannot determine the location of a poor 50m of cable in a run of 5km, it may not actually even detect it.

Are you measuring PD in conjunction with the HV test? If you measure PD in your cable during an overvoltage test, this may be of use, but you then must be able to locate it to make it of value. Some estimations of PD tree growth with VLF show that PD would lead to a failure in XLPE within the test period at a multiple of the system voltage. The problem can be, if I have multiple location of PD activity, I will have grown all trees at all locations, but only one of them will have grown to failure. Once this one is repaired, I must retest, and I will grow the next longest tree to failure, and I will keep doing this until I have many joints in the cable - has this increased reliability, or deacreased it with many more points for moisture entry and human error. Also, the PD may only be present at high overvoltages, and not be present at normal system voltages. So, in actual fact, you may need to have many years of switching and lightning overvoltages causing PD activity for very short times before the cable actually fails.

You must work out what failure mode you are trying to detect, then determine which test is useful in detecting this failure mode, and then work out what you are going to do with the answers.

At MV, I would only be pulling a cable out of service to do any sort of test on it if it is absolutely system critical. Otherwise you may actually be unnecesarily shortening the life of the existing insulation by this "intrusive" test.

Ausphil

Good points Ausphil but do you have a reference for this statement? "Some estimations of PD tree growth with VLF show that PD would lead to a failure in XLPE within the test period at a multiple of the system voltage"

I have seen evidence that DC testing can degrade XLPE cables but this is the first I have heard of VLF testing being an issue.

I do agree that a VLF test should be followed by a TD test, in fact many VLF test sets have the TD module built in to the test set.

Zogzog,

There are a number of references that show this. The main 2 that come to mind are shown below.

Table 1 in IEEE400:2001 gives the growth rates using different test sources - 50Hz, 0.1Hz sin and cosine supplies.

For example

At a test voltage factor of 2 (ie test voltage / system voltage = 2) the tree growth rates are:
- at 50Hz - 1.7-2.4mm/hour
- at 0.1Hz - 2.3mm/hour

At a test voltage factor of 3 (ie test voltage / system voltage = 3) the tree growth rates are:
- at 50Hz - 2.2-5.9mm/hour
- at 0.1Hz - 10.9-12.6mm/hour

So you can see that by going up in volts, you can really get trees growing fast.

Also attached is a presentation by Ben Lanz who has been involved in a lot of cable testing and standards writing. It has a well used graph on page 20 from work done by TU in Berlin showing tree growth rates using different test supplies.

There are a lot of other papers that have information of this sort, but these are probably the best summary.

Hopefully this is of help

Ausphil

• http://files.engineering.com/getfile.aspx?folder=451ea5e1-3e0e-42ab-9ace-e8

Thanks, great stuff (As usual from Ben). I have been out of cable testing for about 5 years now (Just a switcgear guy now) and I see MV cable testing is still changing as fast as it was back then.

To my experience & knowledge, VLF is too strong for cable joints, especially for aged cables. If you only want to test the PD in cable but not a HV withstand test, maybe the OWTS testing is the better option. Even for HV withstand test, I recommend AC resonant testing with PD testing. The only advantage of VLF is the price of test set is much cheaper than the other two kinds of test sets.

Another VLF advantage is the power required for the test set. Most sets I have seen operate from 120V, 15A source.