Meggar value for 4160 volt cable 4

[gsjhand]

Hello Everybody

Recently we had yearly shut down in the plant and as planned we got our substation checked. This was done by world wide reputed company and it involved checking of some underground 4160 volts cables. While checking they found one of the 4160 volt cables giving 2 Mega ohms meggar value. Cables were checked at 2.2 KV. When they went to 5 KV they found meggar value of same cable around 60 Mega ohms. Then they decided to check it on 7.5 KV and there was further improvement in the value.
When I questioned them for checking these cables at 2.2 KV they told me that this is their standard for high voltages cables.
Next no body was able to explain why meggar value improved on high voltage. These tests were done repeatedly with same results.
I would appreciate if somebody can explain to me that why meggar value improved by going to high voltage (I was expecting totally opposite results) &
Why H.V. cables are checked at 2.2 KV as I am under impression that cables should be checked at least at their operating voltage?

 

[flexoprinting]

Hello gsjhand

Those readings indicate good insulation, your megger is calculating the resistance based on higher voltage increments I think these reading would vary on different units even of the same model. It is when the ohms reading lowers you should be concerned, this would indicate a problem.

The Fluke web site has information on using a meg ohmmeter, also years back Avo had a good down loadable PDF on this subject.

I am no expert on this someone else will most likely follow with a better explanation.

 

[oldfieldguy]

There is a high voltage megohmmeter that requires the use of multipliers to calculate readings. I had issues with technicians who couldn't grasp the concept of having to multiply a meter reading by two different multipliers based on voltage level and scale. Questionable results often appeared on test reports due to this "complexity".

A reading of two megohms on a 4160-volt cable would cause me to seriously question its serviceability. The other two readings are more in line with what one would expect. The whole set, though, causes me to question the ability of the technician performing the test.

As for the "world-wide reputed company", I have some doubts. I was formerly a senior field engineer for one of those companies, and I can tell you that I lost sleep worrying about the capabilities of some of the "technicians" we had on our roster.

I recently had to have several commissioning tests redone because the quality of the data on the test reports by a nationally-known testing company could not be interpreted to indicate that my brand-new equipment was fit for service.

As for the low reading followed by two satisfactory subsequent tests, I'd likely suspect poor preparation of the test specimen before the first test.

old field guy

 

[electricpete]

2 megaohms sounds low to me also for typical cable run lengths. Of course, the longer the cable is, the lower the insulation resistance. This might be normal for some very very very long run of cable (several thousand yards?)

Also of course, sometimes it is not the cable but some connected equipment that is the culprit.

NETA Maintenance Testing Standard identifies a single test voltage 2500 vdc. There is no 2-voltage P.I. test for cable like there is for motors. Personally, I wouldn't expect the resistance to go down as voltage increases, but I wouldn't worry about it either. I would worry about the relatively low reading.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[itsmoked]

Perhaps a semiconductor suppression device?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Zogzog]

A MEgger test aloone wont tell you much about the dielectric strength of the cable and any "World wide reputed company" should know this and have also performer a VLF, Tan Delta or PD test on your cable also.

OFG made a great point, dont rely on the companies reputation, I also have worked as a SFE for one of these world wide companies and some of the techs were pretty weak. Power system testing is a small world and most of the best techs prefer to work for the independants, better pay, better benifits, more freedom, so most of the big guys get the bottom of the barrel, dont get me wrong they have some great techs too but you cant count on it. Also the big guys use random job generation, in other words the next job goes to the next tech, regardless of his area of expertise, you might get a relay guy assigned to your cable testing job, the smaller companies send thier cable guys to every cable job, relay guy to every relay job, etc....

What specs was he testing to? Obviously not NETA.

 

[oldfieldguy]

Zog--

You and I, we need to sit down and do a lunch sometime...

old field guy

 

[Zogzog]

Yep, what part of the country are you in? Will you be at Powertest (NETA) in San Antonio in March?

 

[oldfieldguy]

I'm down here in southwest Louisiana. I may make that NETA conference.

old field guy

 

[rockman7892]

Why does the insulation resistance decrease with long cable runs as mentioned in an earlier post. Does it have something to do with capacitive coupling?

 

[davidbeach]

Think about having the insulation of a foot of cable in parallel with the insulation of the next foot of cable in parallel with the insulation of the next foot of cable and so forth and so on.

If (totally made up numbers) you have 1[μ]A of leakage current at a test voltage of 1000V, for a 100 foot chunk of cable, you have 1000M[Ω] of resistance. Add another 100 feet of cable, which will add another 1[μ]A of leakage current and now your resistance is 500M[Ω]. take it up to 1000 feet of cable and now you have 10[μ]A of leakage current and your resistance is 100M[Ω].

 

[flexoprinting]

Hello Davidbeach

"If (totally made up numbers) you have 1?A of leakage current at a test voltage of 1000V"

What if there is no leakage current, does the resistance reading decrease with length of cable?

 

[flexoprinting]

I wish there was an edit button!

The original question by the O.P. was
"While checking they found one of the 4160 volt cables giving 2 Mega ohms meggar value. Cables were checked at 2.2 KV. When they went to 5 KV they found meggar value of same cable around 60 Mega ohms. Then they decided to check it on 7.5 KV and there was further improvement in the value."

These are the readings I get with a Fluke 1587 under good conditions, or so I thought until now!

Electricpete stated above "2 megaohms sounds low to me also for typical cable run lengths. Of course, the longer the cable is, the lower the insulation resistance. This might be normal for some very very very long run of cable (several thousand yards?)"

I am a motor guy and have no experience with the big wire, however find this statement confusing.

Rockman7892 questioned it in relation to capacitive coupling, this condition also would indicate a leakage or improper procedure I believe.

Will post later, gotta go!

 

[edison123]

gsjhand

The disparity in cable IR values with different test voltage levels could be due to

1. Different meggers were used and there were instrumental errors.

2. The IR values were noted at different times during these tests. Since cable insulation currents are mainly capacitive (i.e. they tend to decrease rapidly initially and then slowly over a period of time), you need to measure the IR values at the same time (e.g. 1 min)for all the test voltage levels.

3. Also, you need to discharge the cables to ground for at least 30 minutes before you reapply the test (even if it is at the same test voltage).

 

[Zogzog]

Edison 123 has it nailed, I think 2 ro 3 is your culprit.

 

[rockman7892]

"If (totally made up numbers) you have 1?A of leakage current at a test voltage of 1000V, for a 100 foot chunk of cable, you have 1000M? of resistance. Add another 100 feet of cable, which will add another 1?A of leakage current and now your resistance is 500M?. take it up to 1000 feet of cable and now you have 10?A of leakage current and your resistance is 100M?."

I like davidbeach's example above!

Following this example it would show that indeed the resistance values would increase with using a higher voltage. If 5000V was used in the example above with 1uA of leakage the resistance in the first 100ft would be 5000M?. I believe this would help answwer the question in the OP.

The one question that I have about this example however is the resistances being in parallel. As mentioned by others this leakage current is represented by a capacitance, and capacitors in parallel are added directly not inversely? With that said I thought 1000M? + 1000M? would equal 2000M??

 

[oldfieldguy]

I'm going to add one more caveat to the disparity in the readings: Cable preparation.

A typical scenario: Client tells the test technician that the cable is ready for testing. The tech,assuming (there's that word) that the client has properly cleaned the stress cones, etc., conducts a test. The readings don't look good. A second test is conducted. This time, before starting the test, the stress cones are carefully wiped off and dried. The readings are an order of magnitude better.

I'm just saying that a good test technician would have made sure of the validity of his test results before scaring the client.

old field guy

 

[edison123]

rockman

You have turned around the ohm's law. The cable leakage current would increase proportionately with the increasing voltage so that R (i.e. IR) is constant. Generally, all the good insulations tend to have a constant IR (at a particular temperature) meaning that applied voltage level (within reason) does not change it. A bad insulation will tend to have more than proportionate increase in current with increasing voltages leading to lower IR values.

In OP's case, he says that the IR increases with increasing voltage. Something wrong with the procedure or the instruments.

 

[davidbeach]

It's not just capacitance. Even if it were just capacitance you'd find that the longer the cable the more charging current which is all driven by capacitance.

It might help to think of the cable insulation as providing a certain admittance (inverse of impedance) per unit length. In this cable the admittance will be mostly conductance (inverse of resistance) and not much susceptance (inverse of reactance). Capacitive susceptance is a function of voltage and the voltages in question are much too low for this to be a major factor. More cable means more admittance; more admittance means less impedance.

You'll find that using DC as the test voltage you get leakage current through the insulation. But with DC there is no capacitive current, only resistance or conductance.

In my made up example the leakage current would be expected to increase with increasing voltage. With a test voltage of 2000V, I'd expect something much closer to 2[μ]A of leakage current (it isn't a straight proportionality) rather than the leakage remaining at 1[μ]A.

 

[Zogzog]

OFG, did you catch Don Genuitis article in the Summer NETA World on corona supression? Good stuff.