WATER TREEING IN XLPE CABLE 4

[YUSNIDA83]

Any body can help me about water treeing? is there any product or analysis to spot the pin point of water treeing?

 

[TurbineGen]

Sure there is... sort of. A thumper will tell you where the cable treeing caused it to fail. Other than that, I know of no other way to detect insulation damage other than to remove the cable and inspect the entire thing.

------------------------------------------------------------------------
If it is broken, fix it. If it isn't broken, I'll soon fix that.

 

[cuky2000]

The IEEE Std 400 "Guide for Field Testing Shielded Power Cables" could be a good starting point. Type 2 field diagnostic test provide guide not only to assess the overall condition of the cable but also detect and locate defect in a cable system. Also ANSI / ICEA standard S-649-2004 and AEIC CS8-07, CENELEC could be a reference to check for production test.

Insulation Condition Water treeing assessments testing.
• Hz dissipation factor (Tang Delta)
• Hz Partial Discharge (PD)
<sub>NOTE: Testing that combine the two tests above provide good results

Other test to investigate:
• Isothermal relaxation current
• Return voltage (RV) method
• Oscillating wave partial discharge (PD) testing
• Dielectric spectroscopy
• AC breakdown strength.</sub>

 

[VLFit]

The best way to test for water treeing in xlpe cable is to perform a tan delta test. This test will tell you how deteriorated the insulation is, as the worse the cable, the less the phase shift of 90 degrees between the voltage and capacitive current. similar to a power factor test. this test details the overall degradation from end to end. If resources are available, you may want to also then perform a partial discharge test if you are in a position to replace the cable or repair depending on what is found. There are really so many vaiables that go into selecting the right diagnostic routine. there is no short answer. some methods are great for the situation while others a waste of time and money. You have to talk to someone more in the know to make the determination. for an impartial engineering analysyis, go to

http://www.neetrac.gatech.edu

 

[YUSNIDA83]

Thankyou guys..

 

[YUSNIDA83]

To Mr cuky2000,

Peace upon to you, Tan delta is the formula for whole cable detection, can I refer any journal or link to study about it?

 

[benlanz]

YUSNIDA83,

As you my know, water treeing in power cable is a topic that has been discussed since the 1970's. Currently there is no commercially available way to pinpoint water treeing in a cable system.

We have a great deal of experience performing Tangent Delta (TD) tests in the field with our clients. While it is true that a correctly applied TD test can detect losses caused by water treeing and can be a useful tool, it also measures losses coming from all other cable system components. TD cannot locate losses but, it simply provides a value for the entire cable system under test. Most TD sets for field use do not use a guard circuit to bypass leakages on the surface of terminations. Losses coming from joints and terminations are sometimes 10, 100, or 1000 times larger than that which the cable component manufacturers recommend. Yet, "leakage" on the surface of a termination is not typically an issue of concern. Also there is no commercially available method to calibrate the TD test in the field to assure that the results are accurate. I know of a case study where someone replaced approximately 36 out of 48 terminations (78%), and found that the TD of the cable system improved by more than a factor of 10. When asked, how they knew which termination to replace, they stated that the always replace both because they couldn't tell. This of course was good news for the testing company because they were also the repair contractor. In the vast majority of field test studies that I have reviewed, TD values on aged solid dielectric cable systems are much higher than manufacturers' recommendations for new systems and (on a standard basis) would be recommended for replacement. Thus, a TD test on aged cable is many times a self fulfilling prophesy. Some of our clients have called it the “calendar” test as aged cables typically have higher losses than newer cables.

On the other hand, we have found that an offline 50/60Hz PD test, is very effective at finding electrical trees which are associated with water trees. After all, the only water trees that are likely to go to failure are those which initiate an electrical tree (small tree like fault channel). Since electrical trees typically only grow during voltage transient situations and are delayed in growth due to the water tree, they typically take years to go to failure and are observable and can be pinpointed by a standardized offline 50/60Hz PD test which is calibrated in the field and demonstrates better than 5pC sensitivity (cable component manufacturers’ requirement). While this approach can not detect losses, unless they are associated with PD, most of our clients prefer the PD test as it is much more precise, it is comparable with standardized factory PD tests and the results and can be calibrated and verified in the field.

I would like to caution you to use consensus documents as a reference. (IEEE, IEC) Many of the available documents are not consensus documents and are written by the organization performing much of the testing (a biased position). In addition, we typically only recommend studies in which there are control and experimental data so that actually performance of tests and cable systems can be compared.

Happy New Year!

Cheers!




Benjamin Lanz
Past Chair of IEEE 400
Sr. Application Engineer
IMCORP- Power Cable Reliability Consultants

 

[cuky2000]

Tan delta is more than a formula for cable detection. The principle of the Tan(?) test is based in measuring the phase shift between the voltage and the current. Cable behave essentially as a capacitor considering ? =90° for an ideal or perfect capacitor. Water tree, impurities and dielectric losses will decrease ? < 90°. The interpretation of the test data is used to correlate the condition assessment of the insulation.

For instance IEEE std 400 consider fairly new cables are still in good condition, if :
a) tan([&delta;]) @2Vo < 0.012 and
b) tan([&delta;])@2Vo - tan([&delta;])@Vo< 0.006

Search in the internet for "dissipation factor" "tangent delta" "dielectric loss angle" and/or check in the following organizations:

IEEE,IEC
Doble Test Eng
NETA
Hafely

 

[YUSNIDA83]

thank you very much guys, i will try..

 

[YUSNIDA83]

To Benlaz,

thank you sir .. I really want to make a study of water trees, while my country is new again in this field, now I'm collecting information about the water treeing and try to find what is an appropriate for my research topic on the my master degree proposal.. anyway thank you willing to help .. I really appreciate it.

Happy New Year too!

 

[cuky2000]

To follow up this post I would like to share with the results of an acceptance test for three 345 kV UG XLPE feeders connecting the generating plant GSUs with a GIS switches collector bus.

The PD acceptance show two possible weak points both in phase C as follow:
a) 94% Confidence level of approximately at 211 ft from the termination. Testing company recommend repairing the cable after verifying the location of the PD using electronic location matching.
b) 70% confidence level for a problem identified in the termination. Defer action was recommended because low sensitivity.

For the case a, our team fill that repairing the cable is not the best practical option. For case b, we are recommending revisit the cable termination and inspect closely for any defect.

Any suggestions/comment?

_{For illustration, see preliminary test result and physical arrangement of the HV UG cabling system.}