Transformer aging v/s high current faults

[mmoisan]

Hello everyone.

Currently, we are feeding a distribution network (13.8 kV) through a 33 MVA 110/13.8 kV transformer. Due to the high magnitud of the line short circuit current in the 13.8 kV network (about 8kA), we are looking for a relationship between transformer aging and -for example- the number of high current faults downstream. This could help to determine if autoreclosing is suitable in our distribution network.

I d appreciate your recommendations, methodology or suggestions of scientific publications on this matter.

Thank you in advance!

 

[magoo2]

There's an old paper by GE and VEPCO about through faults and the effect on substation transformer life. They were using 3-shot reclosing and wound up going to 2-shot reclosing. Their 34.5 kV fault currents were closer to 20 kA than 8 kA, but the situation is similar.

You also might want to look at installing neutral reactors to knock down the fault current for single-phase to ground faults since this is a lot more common than others.

13.8 kV is about 8 kV ph-n. With 1 ohm of system impedance, you get 8 kA of fault current. Put in a 1 ohm neutral reactor and you cut the fault current in half. You can look at lower ohmic ratings too.

You just want to be sure you don't mess up the amount of neutral shift during single line to ground faults. Otherwise you'll fail some arresters.

A lot of utilities use neutral reactors, so it's not a big deal.

 

[dpc]

The ANSI damage curves have options for frequent and non-frequent through-faults. I believe the definition of frequent is 10 or more through-faults during the life of the transformer.

I'm not aware of any quantitative data on the impact of reclosing on transformer life. The impact of through-faults is probably cumulative to some extent, but there are so many variables, I'm not sure how meaningful a statistical analysis would be. Intuitively, I'd guess that a lot of reclosing into faults will shorten the life of a transformer.

I think a decision on reclosing should be based on the impact on system availability and customer satisfaction and not on maximizing transformer life. You might also want to look into the reclosing sequences being used.

David Castor

http://www.cvoes.com

 

[mmoisan]

Hello magoo2 and dpc and thanks for your answers. I just got a copy of the Vepco+GE paper, it seems there are some good ideas there inside.
I have been looking to the IEEE Std C57.109-1993 guide for Transformer Through-Falt-Current Duration and our protection scheme provides good triping times respecting the recommended curves. Anyway in the IEEE Standard I find no word on the cummulative effect in the transformer life of high current faults. We are just worried about that high current short circuit.
As we are fulfilling the clearance fault time recommended in the IEEE standard, the effect of reclosing should be analized in function of customer satisfaction and not in term of transformer life time, as dpc pointed out.

 

[prc]

mmoisan, what is the title/author of the GE paper and where we can get it ?

 

[dpc]

The connection between the number of through-faults and transformer life would seem to be implicit in the Frequent and Infrequent Damage Curves defined by ANSI C57. It's just trying to quantify it that is difficult.



David Castor

http://www.cvoes.com

 

[zer0sequence]

If you decide to reclose, you might look into one of the modern transformer protective relays will monitor through faults and give you lots of useful information on transformer life, etc. Unfortunately you usually have to get the expensive model to add these features though...

 

[mmoisan]

Hello people.
@prc: The article's details are:
L. Johnston, N. Tweed, D. Ward and J. Burke. An Analysis of VEPCO's 34.5 kV Distribution Feeder Faults as Related to Through Fault Failures of Substation Transformers. IEEE Transactions on Power Aparatus and Systems. Vol. PAS-97, No. 5, Sept/Oct 1978.
It is available online in the IEEE website. Let me know if you are able to download it.

@David Castor: Thanks for the link!

@zerosequence: That is a good idea. We will research how to take advantage of our relays (GE UR-T60). Thanks.

 

[DTR2011]

you can also perform Frequency Response Analysis (FRA)tests to look for mechanical deformation. If there are sister units in your fleet, comparisons are much easier. You can also perform phase to phase analysis if you have nothing else to compare to.

There are many manufacturers of test equipment worldwide.

 

[prc]

Thank you mmoisan for the paper details.

My view on the subject is, it is impossible to find a correlation between the number of faults Vs transformer life. The severity of fault current depends on the peak value of fault current. Forces developed in winding to destabilse them varies as the square of the peak value of current. The assymetry and thereby peak value depends on the point in the sinusoidal voltage wave at which the fault starts or the point where reclosing happens.This will vary and the effect of each fault current is different.It also depends on the system impedance.So transformer engineers believe that most severe fault current will never flow in many of the transformers in their whole life.

Cumulative effect is also difficult to estimate.Definitely each fault will create shock inside the winding, eating away the compressive strength of press board insulation and loosening out of clamping force on windings. But it is well known that good quality specially designed transformers can stand frequent faults during their life.eg rail side traction supply transformers or furnace transformers where number of faults every day may be several tens.

The effect of fault current on transformer winding is two fold-mechanical:dynamic forces with maximum at first cycle and diminishing fast with current decay and thermal as the heat is stored in copper during the fault duration.Thermal effects can be ignored if reclosing is done after say 5 minutes,by that time heat will be partially dissipated from copper.