Megger testing feeders

[wayni]

From the utility transformer secondary, 480Y/277, we pulled 600 kcmil conductors 350 ft, underground in PVC to our main distribution panel. We used the Ideal Megger set at 500V and checked each conductor. A, B, and Neutral readings were 999 Mohms - maxed the resolution of the megger. C read anywhere from 65-115 Mohms.

Tomorrow, I want to hit C phase again at 500V and 1000V. It's my understanding that I want to get simular readings at both voltages. If my 1000V reading is significantly lower than my 500V this would indicate a deteriorating situation and the replacement would be necessary.

However, I find a lot of varying information on what an acceptable reading should be. Some info I have read indicates 0-50 Mohms, replace. Some say anything under 100 Mohms replace. I have yet to check the job specs to see if there are any minimums listed, which I will do tomorrow.

Assuming I get the readings I am after with the step voltage testing and there is nothing in the specs to say otherwise, go or no go? Is my testing methodology sound and is there anything else I should or could do?

Thank you.

 

[mdshunk]

Somthing's not making a whole lot of sense to me here. If you pulled in PVC, and you say just the "C" conductor tests lower, what were you checking it in reference to? That is to say, if you pulled in metallic conduit, I can see being able to pin down a single lower conductor. With PVC, you're only checking conductors with reference to each other, and I'm not real sure how you pin down one single low conductor.

As far as your go/no-go thing, many guys have lots of personal opinions on that. I'm not sure what the NETA guidelines are, but Megger has the so-called "One Megohm Rule". I think that's way too low, but your readings don't strike me as bad either.

Might read "A Stitch In Time" also. Good free downloadable book on the topic:

http://www.biddlemegger.com/biddle/A Stitch In Time.pdf

 

[don_resqcapt19]

While, I don't see any of the readings as "bad", they should be much closer if they all have the same insulation type. The first thing I would do is re-clean the insulation at the ends of C phase and megger it again. Often wire pulling lubricant left on the cable will cause readings like this.
Don

 

[wayni]

Somthing's not making a whole lot of sense to me here. If you pulled in PVC, and you say just the "C" conductor tests lower, what were you checking it in reference to? That is to say, if you pulled in metallic conduit, I can see being able to pin down a single lower conductor. With PVC, you're only checking conductors with reference to each other, and I'm not real sure how you pin down one single low conductor.

As far as your go/no-go thing, many guys have lots of personal opinions on that. I'm not sure what the NETA guidelines are, but Megger has the so-called "One Megohm Rule". I think that's way too low, but your readings don't strike me as bad either.

Might read "A Stitch In Time" also. Good free downloadable book on the topic:

http://www.biddlemegger.com/biddle/A Stitch In Time.pdf

Sorry, we tested each conductor to ground. Specifically, we clamped onto the H0X0 terminal at the tranformer. Also, the building GEC was used.

Thanks for the link. I've read through it as well. Done a whole lot of surfing on megger testing info since yesterday.

 

[wayni]

While, I don't see any of the readings as "bad", they should be much closer if they all have the same insulation type. The first thing I would do is re-clean the insulation at the ends of C phase and megger it again. Often wire pulling lubricant left on the cable will cause readings like this.
Don

Thanks. I'll do that.

 

[iwire]

Sorry, we tested each conductor to ground.

How is that of any use with a PVC conduit?

 

[mdshunk]

Sorry, we tested each conductor to ground. Specifically, we clamped onto the H0X0 terminal at the tranformer. Also, the building GEC was used.

That still makes no sense. In that case, if the C conductor is "leaky" it would have to be at one of the exposed ends where it's near the metal of the panel or the metal of the transformer. Sounds like contamination to me too, in that case.

 

[wayni]

I think I'm misunderstanding the whole PVC/Metal conduit angle.

We have everything pulled, nothing has been landed. However, hyplugs an spade terminals have been crimped onto the ends of the wire. Also, there is a groundbox we've pulled through. We lopped C phase at this point and took more readings. From the transformer to the box, 999. From the panel to the box, 65-100.

The way we have been testing is by clamping one lead to ground and testing each conductor individually with the other lead for 15-20 seconds.

Is this correct? Should I be testing phase to phase?

Also, thanks for the reference to NETA. Thier standards cost $$ but I was able to google out a reference from their 2005 standard which states 100 Mohm per 1000V. Since each individual conductor feeds 277V my NETA minimum would be 27.7 Mohms, correct? Or is it wiser to go with 480V?

 

[mdshunk]

I think I'm misunderstanding the whole PVC/Metal conduit angle.

If you're testing phase to ground, how is that even the least bit helpful if you have a nonmetallic raceway? Think about that.

Yes, you should be testing from each conductor to each other conductor when you have nonmetallic raceway. You're trying to see if there are any "leaks" in the conductor's insulation that will jump a spark from one conductor to another. If you have a nonmetallic raceway, and you try to jump a spark from a phase conductor to ground, where's it going to jump to? Trying to simplify this a bit for clarity.

 

[wayni]

If you're testing phase to ground, how is that even the least bit helpful if you have a nonmetallic raceway? Think about that.

Yes, you should be testing from each conductor to each other conductor when you have nonmetallic raceway. You're trying to see if there are any "leaks" in the conductor's insulation that will jump a spark from one conductor to another. If you have a nonmetallic raceway, and you try to jump a spark from a phase conductor to ground, where's it going to jump to? Trying to simplify this a bit for clarity.

I'm following you now. The earlier posts on contamination and proximity to bonded metal make more sense as well. Could it be possible for water in the pipe to also affect my phase to ground reading?

Thanks for your wisdom, guys.

 

[don_resqcapt19]

Marc,

If you're testing phase to ground, how is that even the least bit helpful if you have a nonmetallic raceway? Think about that.

Are you saying that you wouldn't meg these conductors to ground? I would megger both phase to phase and phase to ground. There is still the possibility leakage to ground in a non-metallic raceway, especially one that it underground an may have water in it, or even the wire pulling lube as most of it is somewhat conductive. If the conduit is completely dry, then there should not me much of a leakage path, even with conductor damage.
Don

 

[nakulak]

When we meg conductors we also meg phase to phase, and phase to ground, but we make a point not to meg too far past the operating voltage that is going to be used since you run the risk of damaging the wire you just installed (unless it is absolutely necessary). If the c phase insulation was cracked or skinned as the cause, then the megging would have at least saved you from burning the other wire up when you energized it.

 

[wayni]

Marc,

Are you saying that you wouldn't meg these conductors to ground? I would megger both phase to phase and phase to ground. There is still the possibility leakage to ground in a non-metallic raceway, especially one that it underground an may have water in it, or even the wire pulling lube as most of it is somewhat conductive. If the conduit is completely dry, then there should not me much of a leakage path, even with conductor damage.
Don

No may about it. I'm certain there is water in there. Not to mention a couple buckets of pulling lube.

 

[brian john]

NA:

480 VAC cables can safely be tested at 1000 VDC,,,,This is industry standards and is NON-DESTRUCTIVE TESTING. now hit it with 50KV hi-pot, now that something else.

If there is water in the conduit or excessive soap ground readings are possible not to mention the EGC.

Generally we'd like to think phase to ground faults are minimized in PVC.

As others have noted the readings you are supplying are acceptable BUT if I was the building owner I would want MAX readings from all of my NEW conductors, skip the 500 VDC, test for a minimum of one minute at 1000 VDC or till the readings max out.

Is the reading at 100 meg a fluctuating reading or is the reading stable? Often when moisture is the culprit the reading fluctuates.

 

[wayni]

NA:

480 VAC cables can safely be tested at 1000 VDC,,,,This is industry standards and is NON-DESTRUCTIVE TESTING. now hit it with 50KV hi-pot, now that something else.

If there is water in the conduit or excessive soap ground readings are possible not to mention the EGC.

Generally we'd like to think phase to ground faults are minimized in PVC.

As others have noted the readings you are supplying are acceptable BUT if I was the building owner I would want MAX readings from all of my NEW conductors, skip the 500 VDC, test for a minimum of one minute at 1000 VDC or till the readings max out.

Is the reading at 100 meg a fluctuating reading or is the reading stable? Often when moisture is the culprit the reading fluctuates.

Thanks for your response, Brian. My thoughts are the same, if I were the owner I would want MAX readings as well. This is why I would like to get separate readings at 500V and 1000V as I had read somewhere that a discrepancy (meter algorithms being the same) would indicate a deterioration issue in the conductor's insulation. In the meantime I have found a NETA testing spec which says I should be testing 1 min. at 1000 VDC as you have pointed out.

I really don't want to suggest to my JW that we sacrifice quality. I also know the costs of pulling and replacing 250' of 600 kcmil copper will be huge. Not to mention the work involved and I have to protect my rabbit. :wink: No. When I see the megger maxed out on all but the last conductor I question whether the right move is to replace the conductor or not. Safety, above all, is my primary concern. I've gotten some great tips from you guys and I think it is worth it to do some more investigation. Ultimately, it will be up to others to make this decision.

Regarding the reading at 100 Mohm being stable: It is not. First test ran from 95-115 before stabilizing. The second reading was less. Finally, we recieved good and stable readings which where 65 Mohms. Since this was at 500VDC I am of thed opinion that testing at 1000V will give a clearer picture of the situation.

Thanks for your reply, Brian.

 

[mdshunk]

nljmonklm,.

 

[wayni]

It's sorta neat that you stumbled across that document, but that's a copyrighted thing that other people have to pay money for. It's not really supposed to be out there for anyone to get for free.

Well it's from 1999. It should be a simple thing to verify its applicabilty to the 2007 spec. Either the shop has paid for it already or a phone call to our commisioning agent will reconcile any questions, The important thing is that I have something to work from.

I don't endorse breaking the law in any way but I am an inquisitive sort... I'll use what I can find and then make sure it gets backed up solid before suggesting an alternative. It is Sunday, after all. :wink:

 

[wayni]

It's sorta neat that you stumbled across that document. It's not really supposed to be out there for anyone to get for free. While this lasts, since you already have the ATS standard, you can also get NETA's MTS standards here: http://nasagalaxie.larc.nasa.gov/neta_mts-2005.pdf

I get a message:

"NASA Galaxie is accessible to NASA employees and contractors only, from on-site workstations and approved off-site workstations. If you believe you have received this page in error, please contact your center's library. You will need to provide your workstation's IP address and/or domain name. If you are unable to contact your center library, you may email the Langley library IT team at library-it@larc.nasa.gov. "

Now I am sad. But thanks for the link! :smile:

 

[zog]

Either you misunderstood the NETA spec or it was misquoted. The minimum spec for 600V cable is 100M. The test should be conducted at 1000V for 1 minute. A megger test doesnt tell you much besides that it is safe to conduct an overpotential test. It looks like you have a damaged cable, a VLF or Tan Delta test will confirm this.