[Please Help] Post installation test on MV cable

[JuanBC]

Hello,

I'm not very familiar with tests that are done on cables for which I would like to describe the following situation to get your opinion

- The company for which I work contract another company to carry out post-installation tests on 15 kV-cables that will work in a 3x13200V network

- The contracted company performs the following test on the cables: They apply 10kV 50Hz between each conductor and the screen, with the following cycle: 30 sec - wait - 1 min - wait- 1 min 30 sec - end

They measure 20mA of leakage current in each case

Conclusion?

Thanks!

 

[waross]

Testing a 15 kV cable with 10 kV?
Short answer:
Get another contractor.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[7anoter4]

Although I was cable designer and I prepared technology proceeding and for cable manufacturing lines -extrusion, copper drawing , conductor stranding, and other- I was not a cable testing engineer.
In my opinion there are IEC 60502-2 and IEC 60006-3 relevant standards [or IEEE 400/2012 cptr. 6.1] for withstand voltage test requirements.
However, since the measured current was 20 mA means that the cable withstood the test and this what the standards required.
According to IEC 60006-3 5.6 Withstand voltage test procedure:
"The requirements of the test are satisfied if there is no disruptive discharge."
The test voltage which has to be applied 75% of rated [13.2*0.75=9.9kV] at first and from here up to U=13.2 kV the rate of rise is about 2 % of U per second when the applied voltage is above 75 % of U.
"The test duration shall be specified by the relevant technical committee, taking into consideration that the time to reach the steady-state voltage distribution depends on the resistances and capacitances of the test object components. When not otherwise specified by the relevant technical committee, the duration of a withstand test shall be 60 s."
IEC 60502-2 cptr.20 Electrical tests after installation/20.2.1 AC testing states:
"By agreement between the purchaser and the contractor, an a.c. voltage test at power frequency in accordance with item a) or b) below may be used:
a) test for 5 min with the phase-to-phase voltage of the system applied between the conductor and the metallic screen/sheath.

 

[JuanBC]

Bill: I agree with you, but it is not my call. I was simply consulted by a precomissioning member about whether "20mA" was a expected value or not (and I had no idea until I made some calculations with the capacitance of the cable and it's lenght)

7anoter4: I have read both standards (IEC) but I would like to have the opinion of someone with experience in the field.

IEC does not take position when it says "The test duration shall be specified by the relevant technical committee" or "By agreement between the purchaser and the contractor, an a.c. voltage test at power frequency in accordance with item a) or b) below may be used"

JBC
.......
"The more I read, the more I acquire, the more certain I am that I know nothing"

 

[ausphil]

JuanBC,

Unfortunately you can't make a call as to whether it has passed a test or not, as you don't have a known test specification that you can make a pass/fail decision on.

7anoter4 got part way there in specifying IEC 60502.

However the test that your contractors have performed does not comply with this standard. The voltage level is 10 kV, not 13.2 kV. Also the time of application of the test voltages does not meet the 5 minute duration, so you can't say that it passes or fails with respect to this standard.

You may have a local specific national standard or the electricity company may have an inhouse standard that specifies that particular test regime, and if they do, it should have a pass/fail criteria associated with it. In reality though, you are not really overstressing the cable from its normal operating voltage (about 30% at 10 kV).

To your initial question of "conclusion?", my answer is that you have a cable with a capacitance of 6.4nF. Assuming it is XLPE, the resistive current will be almost negligible, so the capacitive current will be essentially 100% of your test current. So this test current, at 50 Hz seems to infer that it is a very short cable (say the capacitance for this cable is in the range of 0.3-0.6 uF/km, this would lead to a cable length of 10-20m). If this sounds right, then the current measurement is probably pretty good. But it is not giving you any comfort about its ability to withstand a high voltage test to a known standard.

ausphil

 

[JuanBC]

Hi ausphil,

Unfortunately you can't make a call as to whether it has passed a test or not, as you don't have a known test specification that you can make a pass/fail decision on.

7anoter4 got part way there in specifying IEC 60502.

I've read IEC 60502-2, IEC 60060-3 and IEEE 400. I still do not know what position to take. They are quite generic with regards to electric post-installation tests.

e.g. IEC 60502-2 says:

20.3.1  AC testing 
By  agreement  between  the  purchaser  and  the  contractor,  an  a.c.  voltage  test  in  accordance with IEC 60060-3 and in accordance with item a), b) or c) as below may be used:

a)  test for 15 min with the phase-to-phase voltage U, at a frequency between 20 Hz to 300 Hz 
shall be  applied between the conductor and the metal screen/sheath; 
b)  test for 24 h with the normal rated voltage U0 of the system; 
c)  test for 15 min with the RMS rated voltage value of 3 U0 at a frequency of 0,1 Hz applied 
between the conductor and the metal screen/sheath.

What happens if I CAN NOT reach an agreement with the cable manufacturer? What should I do?

In reality though, you are not really overstressing the cable from its normal operating voltage (about 30% at 10 kV).

How much will be "enough"? IEC and IEEE fail to establish a value

To your initial question of "conclusion?", my answer is that you have a cable with a capacitance of 6.4nF. Assuming it is XLPE, the resistive current will be almost negligible, so the capacitive current will be essentially 100% of your test current. So this test current, at 50 Hz seems to infer that it is a very short cable (say the capacitance for this cable is in the range of 0.3-0.6 uF/km, this would lead to a cable length of 10-20m). If this sounds right, then the current measurement is probably pretty good.

I made same calculation after reading and thinking a bit, and yes, I came to the same conclusion. P.S. The real lenght of the cable is arround 20m.

But it is not giving you any comfort about its ability to withstand a high voltage test to a known standard.

Yeah, i think the same :/

P.S.2: Now they have to test another cable (3x70 7.6/13.2 (14.5kV) [Uo/U (Um)] - 3200m XLPE/CU cable)

They told us that since they only own an AC equipment and they can't find any CC equipment in my country (really weird) they are studying how to test this very long cable.

Any suggestions? I thought about "VLF test" or "Resonance AC equipment" but I do not know if they are suitable for a post-installation test.

Regards,
JBC

 

[ausphil]

A couple of points:
1 - IEC does specifiy the test voltage and time. It gives you the 3 options as above.
2 - the agreement is between the purchaser and contractor, the manufacturer may have nothing to do with the tests after installation. The wording under section 20.1 states that the test on the insulation may be replaced by quality assurance procedures during installation of the accessories. This said, if this agreement is not made, then you have the right to say that you want a test in accordance with one of the 3 options above. It is usually best to get this in the contract up front, rather than arguing it once the cable is in the ground and jointed.
3 - vlf or resonant test - either are satisfactory, although I would suggest that at this voltage, vlf is the better option.

ausphil

 

[VLFit]

Hello JuanBC,

For your application with only 20m of cable, you could use a power frequency AC hipot to perform the test. Assuming .4 uF/km, at 50 Hz with a test voltage of 13.2 kV applied from phase to ground, the charging current needed from the hipot is 83 mAac. A 15 kVac hipot rated for 83 mA is only 1.3 kVA. That is a readily available size. Remember though, that is only for a 20 meter cable length. For longer cables and/or higher voltages, the power requirement will grow rapidly. You will most likely need to use a series resonant system or a VLF 0.10 Hz AC hipot. The VLF is the better choice since it is so much smaller, lighter, consumes far less power to perform the test, and is very economical. You can use the VLF to perform a VLF Withstand voltage test, and if you would like, a VLF Tan Delta and VLF PD test. However, for verifying the integrity of a new installation, most would just perform a VLF withstand test. The IEC standard will say 3Uo for 15 minutes. The IEEE 400.2 standard will specify 30 kV peak @ 0.1 Hz for 30 - 60 minutes. A basic electro-mechanical 30 kV VLF rated for 0.4 uF @ 0.10 Hz sells for approximately $10,000.00. A solid state version for automated, wireless, data logging control, etc. sells for about $15,000.00, both are less than 80lbs. Both can test over 1000m of 15kV cable.

I hope this helps.
VLFit

 

[JuanBC]

Thank you ausphil & VLFit,

Finally the person in charge of the subcontract decided to hire a company with a CC equipment despite the fact that I recommended VLF

Anyway, I learned a lot with your help.

Regards,

JBC
.......
"The more I read, the more I acquire, the more certain I am that I know nothing"