allowable leakage current during hipot testing of cable 1

[elecsun]

WE are carrying out installationg and commissioning test for medium voltage(36kV) grade Cu/XLPE/SC/Cut/PVC/SWA/PVC cable thru DC Hipot test kits. Cable lengthe are varying as per site condtion from 100Mtrs. to 2500 Mtrs.I would like to know the allowable leakage currents(milliamps) during this test.In which IEC/IEEE stanadard this has been detailed >Whats the theoratical formulae to calculate leakage current for such HIPOT test?Experts guidence please .....

 

[alehman]

I know of no way to calculated expected leakage current. You might want to search this forum a bit. This subject has been discussed extensively. Here's a thread to get you started.

thread238-80814

 

[electricpete]

In most cases a hi-pot is not judged by the leakage current but by the fact that the test set doesn't trip off. If the equipment survives, you passed.

One exception is the "step voltage" test often used on rotating machinery. There you increase your voltage in increments, allow approx 1 minute stabilization and log the current. Then the equipment condition is judged by the shape of the curve. Linear is good... usually means you are some distance away from breakdown voltage. A sharp increase in the rate of rise of current (vs voltage) may indicate you are approaching breakdown voltage.

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

 

[mc5w]

You should be able to get a data sheet from the cable manufacturer as to how much leakage current the cable will tolerate. However, DC hipot testing has been discredited for any medium voltage testing other than a go/ no go test for a brand new installation.

A more fancy test energizes the line with power frequency voltage and then looks for the characteristic noise from partial discharges. You should do this first at a lower voltage level as energizing the cable could cause a ground afult at a defect. If this defect exists then energizing the cable first at say 120 volts via a resistor and then at higher voltages via a resistor will be less spectacular.

I have seen a sustained arcing fault in open air at a 2,400Y4,160 volt substation. The cause was a cracked porcelain insulator.

 

[rbulsara]

It is usually few microamps, 10-15 or less, trying to recall from top of my head, for some 15 kV cable tests results I have seen. NETA-Acceptance Test specificaitons must have some guidelines, I will check that.

The key is to see the plot of the leakage current vs. time. There will be peak in first few seconds, than it will taper off and go to a steady state value after a minute or so, this will be sign of good cable. If the current keep rising,it is indicative of a bad insulation.

 

[elecsun]

Thanks Mr.Bulsara ..i am looking specifiic values alloawable ..if u find something in NETA please send me the copy ....

 

[dadfap]

My understanding of the leakage current is that if it stabilises or reduces during the rest period when the ultimate test voltage is reached the cable is good.
If the cable leakage current continues to increase then the cable is problematic.

 

[rbulsara]

If you are involved in testing of such important systems, buy the NETA specifications from

http://www.netaworld.org

If not, probably it is not that important to you or your organization, I would guess.

 

[jghrist]

Also see IEEE Guide for Making High-Direct-Voltage Tests on Power Cable Systems in the Field, IEEE Std 400.

There are no specific values of leakage current allowable.

 

[electricpete]

I agree with jghrist and dadfap and electricpete ;-)

- As jghrist said, IEEE400-2001 does not give an acceptance criteira in current.
- NETA MTS-1-2001 (Maintenance Testing Spec) section 7.3.3 describes hi-pot and does not give an acceptance criteria in current.
I would be really really surprised to see anything different in NETA Acceptance Testing Spec.




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

 

[rbulsara]

As suspected no accpetable leakage current values are indicated in NETA-ATS. All you do is plot the leakre current vs. time and see that it tapers off.

NETA-ATS suggests setting DC test voltage to zero at the end then measure the residual voltage for each condutor at regular intervals. This residual voltage shall be more or less equal for each phase condutor or parallel conductor.

As for the formula: I=V/R for DC . so you can figure that if you know your insulation R and test voltage. But this all chages with temperature, voltage applied etc, so use your judgement. Best thing is to compare values with that of a known good cable.

Ask the testing company for some results of previous successful tests.

 

[rbulsara]

and oh yes, there is no beter resource than the engineerig dept of the cable mfr.!!

 

[alehman]

DC hipot is basically a pass/fail test.
Steady or decreasing leakage = pass
Increasing leakage = fail

If there are large current differences among phases of the same run and same type of cable and you have ruled out external influences (differences in humidity, method of space charge containment, and other leakage paths), you may have a problem.

 

[benlanz]

You are fooling yourself if you think that a DC test is an effective commissioning test for extruded cable. Extruded materials fail by a mechanism associated paritial discharge (PD). Defects in extruded cable look like an open circuit to DC. A DC test detect low impedance defects that fail by conduction, such as in paper insulated cables!!! Measuring conduction is a waste of your time and money and will not find defects which will fail later.

Forget about NETA. They are still struggling to adopt IEEE standards from 2001! The definition of a shielded power cable acceptance test according to IEEE 400, the IEEE Standard Guide for Field Testing and Evaluation of the Insulation of Shielded Power Cable Systems, is the following:

"A field test made after cable system installation, including terminations (see IEEE 48) and joints (see IEEE 404), but before the cable system is placed in normal service. The test is intended to further detect installation damage and to show any gross defects or errors in installation of other system components."

The DC HIPOT withstand test no longer fulfills IEEE 400’s definition of an acceptance test. IEEE 400, section 4.2, states:

"Furthermore, from the work of Bach [Bach, R., et al, “Voltage Tests to Assess Medium Voltage Cable Systems,” Elektrizitaetswirtschaft, Jg. 92, H. 17/18. pp 1076-1080, 1993.], we know that even massive insulation defects in extruded dielectric insulation cannot be detected with dc at the recommended voltage levels."

If the purpose of an acceptance test, according to the definition above, is to show any gross (massive) defects or errors in installation of other system components, then, according to IEEE, the DC HIPOT not suitable for extruded cable installations.

If you want an effective HIPOT on extruded cable you have to use an AC HIPOT such a VLF HIPOT according to IEEE 400.2. However, no HIPOT will not be able to predict future peformance. If you want a predictive test you must use the same test as the cable and accessory manufacturers use, which is a off-line PD test. If you want to use the manufacturers specifications for accpetable PD levels you must use a 60Hz, Offline PD test. IEEE 400 section 7.4 states:

"If the cable system can be tested in the field to show that its partial discharge level is comparable with that obtained in the factory tests on the cable and accessories, it is the most convincing evidence that the cable system is in excellent condition."

According to IEEE 400 “the most convincing evidence that the cable system is in excellent condition” is a test which can best replicate the factory test. A factory test is a 60Hz, off-line PD test.


Cheers

-Ben





Benjamin Lanz
Sr. Application Engineer
IMCORP

 

[electricpete]

And needless to say, Ben will make money if everyone believes their dc hi-pot tests sets have suddenly become useless. I'm not arguing the technical aspects, just pointing out the conflict of interest in the opinion expressed.

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

 

[rbulsara]

Ben:

Reputation of a manufacturer also counts for something. So does experience over years. Acceptance tests are just that, acceptance tests not factory tests. Mfr's are obligted and they do all the necessary tests to make sure their products meet the published standards they claim to be meeting.

Acceptance test in general to verify primarily that the products integrity is not grossly compromised during shippng, handling or installtion or its not a complete 'lemon' as it may turn out time to time. It would be foolish to expect acceptance tests to replicate factory tests.

 

[benlanz]

Thanks for the word of commendation. The fact that DC will not find issues with extruded products which have been grossly compromised during shipment, handling, or installation is exactly the point IEE400 is trying to make. When I by a product I expect it work as specified. The DC test does not give any indication that the product is not working to specification.

For those who are skeptics:
Go ahead, do the experiment yourself. Cut half way through your extruded cable's insulation and apply a standard DC acceptance test. The cable will pass without a problem. A VLF HIPOT may fail and 'find' this defect, and a through non destructive PD test with 5 pC sensitivity definitely will locate the defect without failing the cable.

A Short History of Cable Testing:
I understand your reluctance to accept that IEEE is no longer recommending a DC test as an acceptance test for extruded cable. I was in your same position at one time, until someone explained the problem to me. To understand where the great confusion about cable testing in our industry has come from, we must review the history of cable testing.

Traditionally, cable acceptance tests have been carried out by applying a direct current (DC) voltage to a cable at a specific voltage level and for a prescribed duration. The DC high potential withstand test, or HIPOT, was, and still is, a good choice of high potential withstand test for paper insulated lead covered (PILC) cables. The DC HIPOT is a good choice for PILC cables because PILC cables commonly fail by conduction which can be measured by a test in the form of power loss or ‘leakage’ current. In the 1960s the electric utility industry decided to embrace a new technology and adopt cables with extruded insulation. The extruded cable systems promised to be more economical, simpler to install, and have a relatively long service life (compared to PILC cables). What the electric utilities did not know is that extruded materials, such as cross linked polyethylene (XPLE) and ethylene propylene rubber (EPR), do not typically fail by conduction. They fail in the presence of an alternating current (AC) field through a mechanism associated with PD. This failure mechanism causes the insulation to be eroded over time until a fault channel bridges its entire thickness. Some extruded materials, such as XLPE, cannot tolerate continuous PD for very long. Other extruded materials, such as some formulations of EPR, can last years before failure, even when under continuous PD conditions. In the early days of extruded cable and accessory development, manufacturers understood the problem of PD and quickly switched the quality control test from traditional dielectric loss measurements, such as power factor and tangent delta, to a calibrated 60Hz PD test. A calibrated 60Hz PD test was not available outside the factory shielded test laboratories until the mid 1990s due to high radio frequency (RF) noise in the environment where cables are installed. Today, calibrated 60Hz PD diagnostic tests comparable to the factory tests are readily available for field application.

I am just presenting the facts, the choice is yours. You can find a standard, such as NETA, that will cover you if anyone questions your DC acceptance test but, if you actually want to prove that your cable system is free from gross defects you are going to have to do a VLF HIPOT (per IEEE 400.2), or, better yet, a off-line PD test (per IEEE 400.3)

I truly hope this information is helpful. I welcome your questions and comments on the technical merit of the course of action that I am recommending.

-Cheers



Benjamin Lanz
Sr. Application Engineer
IMCORP

 

[DanDel]

elecsun, DC Hipot testing is still the standard for Acceptance testing of new MV cable. PD testing is a very effective (and expensive) test typically used for Maintenance testing of cable in service, favored because it does not require a shutdown as hipot testing does.

Maintenance testing by DC hipot is not usually recommended anymore, at least not anywhere above operating voltage.

As far as your initial question, I have tested new cable by DC hipot and have typical results of <1.0 microamp for cables up to 300-400' long, with proper corona suppression techniques at each end. I have also seen results of up to 100 microamps for longer cables, high humidity, less successful corona suppression, etc. The key is, as already expressed here, that the leakage increases linearly with the step voltage increases, peaks at the maximum test voltage, and decreases (while holding that voltage) over the first few minutes. It should then reach a steady-state value of leakage.

 

[benlanz]

elecsun,

STANDARD -ICEA
If you what a standard to reference you can you use the cable manufacturers standard ICEA S94-649-2000 appendix E5 which describes DC voltage levels and times for acceptance tests. This document is in direct conflict with IEEE 400 and is soon to be revised. This document is where NETA gets their specification. Of course it is in the cable manufacturers’ interest that you do not find manufacturing defects so they are a little reluctant to adopt IEEE 400. ICEA recommends 100kV DC for 15minutes for 35kV class systems with 8.76mm of insulation.

DanDel,

I agree that there are standards that have not been updated to the state-of-the-art, and if you want to do a DC acceptance test, linearity is the key. The installer will be covered by these old standards if the cable system passes the DC test.

However, don't assume the extruded cable is defect free if it passes. The only thing you have proven with a DC test is that there are no conduction paths (i.e. grounds left on or that the cable is already failed). This is useful information (I recommend a simple megger test because they're inexpensive, and simple to use), but does not prove that the cable is free from 'gross' or 'massive' defects (per IEEE 400).

ONLINE PD Test
I disagree with your statement that PD is favored because it can be done on-line. On-line PD testing is a preliminary test at best. If you find something wrong with an on-line PD test, you have probably saved yourself a failure. However, if you don't find anything you can say nothing of the future reliability of the cable. Only off-line PD tests (like the cable manufacturers test) have been proven to reliably predict future performance. Off-line PD is favored because it is the only non-destructive test which can locate defects in new extruded cable system and predict future performance reliably.

If you don't want the expensive of the PD test and you realize that the DC test is not a reliable HIPOT I highly recommend a VLF HIPOT. The equipment is relatively inexpensive and is a much more effective HIPOT.

The key to using any kind of test is to know its limitations and the cost benefit economics.

I agree that PD tests are expensive initially, but the test pays off in the future because you will have found around 95% of the defects instead approximately 1% that a DC test can find. What is the cost benefit if you have around 94% fewer failures?

Regards,

-Ben


Benjamin Lanz
Sr. Application Engineer
IMCORP