Breakdown of Surge Arrester

<...Is it possible that during the line-to-ground- fault the energy created by the fault surpassed the rating of the arrester? if not, what is the possible cause of arrester breakdwon?...>

It is certainly possible. Among the other possibilities are:

• The equipment was incorrectly selected for the application.
• The equipment was incorrectly installed.
• The equipment was incorrectly maintained.
• Something else was there that should not have been there.
• Etc.

However, as the forum cannot see the installation, the exact cause cannot be determined from here.

Thank you for the reply.

The arrester is good operating condition based on our electrical test. Actually it was nely replaced. Just operating for about 8 months. We had the same incident last year. Also open loop of the feeder line then the same phase arrester was damaged. So, we replaced it, then the incident happen again.

The arrester has mcov of 48kV and duty cycle rating of 60kV. The system voltage is 69kV.

The arrester was based on our transformer which is wye grounded on 69kV side. However, the customer's transformer is delta. The customer's transformer is installed after the metering point. And the arrester is at the meetering point. Will we specify the arrester based on the delta customer or our system?

<...Will we specify the arrester based on the delta customer or our system?...>

The question is unanswerable, as <...we...our...> is undefined and as a consequence there is no way for the forum to obtain future information from <...we...> with which to answer it.

Hi, thank you for your reply.

Actually, the delta transformer (69/13.8kV) is no longer ours. It is the transformer of the customer.

Our transformer (230/69kV) is the source which is Wye on the 69kV side. about 20 kms away form the transformer is the customer. But, before the delta transformer of the customer is our metering LA, PT, and CT. And, our metering LA is rated based on the primary source which is wye. Thus, its duty cycle rms is 60kV and MCOV of 48kV.

Do we need to refer it is the primary of the customer? If we will increase it to line-to-line voltage rating, we will also replace the metering CT and PT installed.

For other trippings of the line, the LAs served its purpose well. Only with this kind of condition. Whenever the line encounters open-loop (meaning the line has a cut/open jumper loop), the LA fails.

I am not experienced with a utility single phasing an industrial on a regular basis, at least enough to build up a history. Our delta primary transformers always used line rated arresters, regardless of how (electrically) close the utility neutral was.

Normal switching surges on longer lines (160kM) have produced typical line to ground potentials 1.7 to 2.15 times the voltage. Your arrester needs to be able to withstand that sort of activity, but something in your circuit is extending the duration of the surge. Line dropping typically results in surges 2 to 2.2 time the voltage on line to ground equipment, even a higher risk. I think you have enough experience that the statistical nature of switching surge overvoltages should have allowed a few operations, recovering from single phasing, that did not pop the arresters.

My guess, and only a guess, is that there is enough distributed capacitance in the transmission line phase lines /reactance in the neutral return that the local ground at the arrestor shifts to become line to line potential, really the only way you can systematically flash over the arrester.

I assume your standards created by experience tell you to switch out the metering equipment, however in this case, the SWC of the equipment might tolerate the transient that you seem to get when re-energizing the open line. If your potential xfrmrs are connected L-N, it would be interesting to see recordings of these values when one of these events occurs.

When you go single phase, do you disconnect the customer, and close in all 3 phases at once, or do you switch the open phase while customer connected? Research actually indicates that single phase switching lowers the transients, not what I would have expected.

Thank you very much for those valuable information.

During the open loop, the 69kv line was left energize for the sake of customers before the open jumper loop to have continuous power supply. However, for customers after the open jumper loop, was disconnected from the 69kv line via recloser of the utility at 13.8kv side.

So technically speaking, the load was out all at once. Then there was a shutdown of whole line to correct the open jumper loop. For the energization, the 69kv line was restored three phase simaultaneously. That is where the LA failed.

Reenergize sounds as normal a procedure as you can get, you should have similar history on other sites, but you do not.

Maybe the arrester fails before you reclose on to it?

Is it possible that the arrester fails during the loop failure (opening the circuit), and you don't detect it until reclosing? As I found in some literature, opening a line can lead to a higher than normal surge voltage, usually a big concern for the circuit breaker doing the interruption, recovery voltage, not surge arresters.

Perhaps with transients & restrikes with a loop flailing about, you may get repeated surges that exceed the arrester rating, if it has failed upon initial fault.

If it does fail upon re-energization, then I think you have something unusual with your earth return equipment, perhaps right at your customer's site, or in the ~12mile return path, big distance for my experience, but a short hop for utilities.

In any event, I suspect your only economical solution is to increase the L-N rating, and check the BIL of the customer's transformer, I suspect it will be OK. If not, you may need to get bigger capacity arresters, or install multiple paths to carry the surge. Maybe air gap & MOV arrester types can be coordinated to help each other?

One reference I ran across shows the gap acts as a current limiter, reducing the surge current the MOV experiences.

As several others here have already pointed out, there are far too many undefined pronouns and conditions for anyone to make a definitive conclusion on the root cause of this failure. A thorough forensic analysis of the failure would require "hands on" examination of the failed parts and probably careful dissection of these parts to identify the first failure point. Then analysis of plausible causes of that failure mechanism can be debated. Several of the plausible causes have been already identified but a few more come to my mind:

• Counterfeit parts
• Improper testing
• Actual surge caused by connection sequence, fly back loading or simply a coincident event (lightning) on the grid

Identifying how and why a failure happens can be one of the most valuable tasks an engineering team can provide. It cannot be done by forum discussion. Knowledgeable people must inspect the aftermath. Often a lot of analysis discussion and reexamination of parts to verify or refute a scenario must also be performed. This is not a trivial task or expense. It may not be worth the expense but it could also be invaluable in solving a reoccurring problem.

The arrester could have been damaged or compromised in several different ways...It should have been inspected before energization....careful inspection now may yield clues, in any case if you are acquiring an arrester from the same supplier, I would keep it under regular inspections for cracks or other signs of premature failure in the materials...You may need somebody who is an expert in this area you can call on for inspections....

Familiarize yourself with the most prevalent causes of failure listed in the link below....and the term 'moisture ingress'...

http://www.ijoee.org/uploadfile/2013/1006/20131006033843423.pdf