Transformer Pressure Relief Devices 2

[jburn]

I am having a discussion with out Mechanical Integrity engineer about the need to pressure test the pressure relief devices on our oil filled transformers. He wants to know if we can raise the pressure on a transformer, to measure the pressure that the pressure relief device opens. I have never heard of anyone routinely testing these devices on transformers, and I question the need to do so. He is looking at it from the standpoint that we have a tank, with a pressure relief device, and a means to generate pressure. (Transformer fault) He wants to add these devices to the testing program with all of the other pressure relief devices in the plant. Is anyone doing this? Also, does anyone routinely test their sudden pressure switch on their transformers? Not sure how you would do it. I have not told him about the SP switches yet. Would like some discussion on this topic from the group. Thanks.

 

[NormGA]

During my 30 years or so as a substation test engineer I never saw or heard of anyone even wanting to test a pressure relief device. A couple of problems I think you might have, one is the very real likelihood that the device will not reseat itself and you will have a gas leak and have to replace it. The other is contamination of the gas blanket with the test gas, unless you are using clean dry nitrogen. Even then I would be concerned about causing some of the nitrogen to be dissolved in the gas only to become bubbles some time later, possibly causing dielectric failure. I doubt that would happen because of the short time involved, but I wouldn't want to be responsible for it if it did. I suggest you talk to the manufacturer of the transformer or the pressure relief device before embarking on such a test. If the pressure relief device has an alarm switch it can and should be tested. It should make the MIE happy to test the pressure-vacuum bleeder, which you didn't mention. They are easy to test either on or off the transformer.

The sudden pressure switch (I assume you are talking about the sudden pressure relay) is easy to test. The relay instruction book has a test procedure for it and it should be tested on the same schedule as the protective relays if used for relaying. I don't remember the numbers but it involves using something like a blood pressure cuff bulb and valve and applying a few mm of pressure to the relay bleed port for a few seconds then suddenly releasing it. You then do the same with reduced pressure. It's basically a go-no-go test. Just be sure you disable all tripping before doing this unless you really mean to trip something. They have been known to trip by just closing and opening the isolation valve.

 

[jbartos]

Suggestion: Theoretically, you could heat the transformer externally thus increasing internal pressure. This is very risky. Supposing that the pressure switch malfunctions or there is some material defect in the transformer enclosure, it could end up in the substantial damage. Therefore, it is not worth it. It is better to wait for a routine transformer maintenance including this pressure relief device service.

 

[electricpete]

I think that periodic testing of sudden pressure devices is fairly common (most designs can be tested relatively easily).

I remember reading a question among Doble Maintenance Engineers on the exact subject of testing transformer relief devices. There was a small percentage that periodically tested them. But in those cases they removed the relief and took it to the shop for testing (requires a homemade test fixture to be able to seal the relief to apply test pressure). Also the ability to hold pressure after reseated should be tested and a spare relief should be on hand in case of failure.

As jbartos mentioned, it would be foolhardy to raise the entire tank pressure to the relief setpoint. I have seen a radiator rupture and our guestimate was that actual pressure was approx 7 psig (relief setpoint 10 psig). That's not the design, but it happens in the real world, and depending on location of the leak, you can have a real mess on your hands.

 

[Newton1Law]

I have about the same experience as NormGA in that Ihave been a System Maintenance Engineer for 26 years working on and with substation equipment. If your transformer has sudden pressure relays (SPR) on it they can and should be tested. The overpressure relief device you are speaking of is designed to operate for a sustained low energy internal fault that is two slow for the sudden pressure type relay. It may however operate in conjunction with a SPR but only serves to point out that you have grave problems with your transformer. The indicator switch attached to the overpressure or pressure relief device (PRD) is usually an over-center type toggle switch that will remain on after the PRD has operated and reset. Most PRD are designed to operated at about 10 PSI, lifting 6 to 10 inch diameter plate off the transformer and resealing itself due to the spring pressure that holds the plate in place. Motion of this plate may also set up a yellow alarm flag ot post depending on which type you have.

Thes devices are very reliable and I have never seen one fail to operate when called upon. The manufacturer's do not suggest testing these units on the transformer. This is because they are such a simple device and the probably of failure is extremely small. The toggle switch should be checked routinely and may be operated without setting of the PRD. Remove the transformer from service and flip the toggle switch on and off looking for freedom and ease of motion.

Last, trying to over pressure the transformer tank with nitrogen would require a lot of gas delivery in a moderate to high volume. Not an easy task. Second, the transformer normally operates between +/- 5psi. Trying to operate the 10 PSI PRD may cause the transformer to develop oil leaks it did not have prior to the test. (This assumes we are talking about a power transformer 5MVA or larger).

My advice would be not to perform the test as it is a waste of time and may cause further problems.

 

[electricpete]

Good comments from all. I don't have the multiple decades in that area that they do, but perhaps a total of 3 years working on power transformer testing/maintenance full time scattered thru my career.

Newton mentioned that he has never seen a relief device fail to operate. To qualify the significance of that observation, it would be interesting to see how many times he has seen one operate sucessfully. I saw about 10 transformer trips that damaged transformer to point requiring repair. In none of those cases did relief operate (presumably pressure did not get that high). I did see one occasion when generator stepup transformer relief lifted due to air bubble buildup within the transformer due to leak at pump suction under vacuum.

In deciding whether or not to do the test, it's worth mentioning the potential consequences of failure of a relief. I disagree with the statement that the relief "only serves to point out that you have grave problems with your transformer." The relief is in fact the last line of defense against tank rupture. Tank rupture can release all of the oil in the transformer, and also can be a bear to repair. I don't know the full design but relief might also play a role in preventing doomsday scenario where transformer erupts into ball of flame.

And if you have a device that has been sitting for 20 years and never tested, how much confidence do you really have that it'll work when called upon for the first time.

As I mentioned the "norm" is to not test the device. But I just wanted to play devils advocate and discuss the potential risks.

 

[electricpete]

IEEE 62-95 Part 1 is IEEE Guide for Field Testing of Transformers. It has a fairly extensive list of tests/checks commonly performed. There is no mention whatsoever of relief valve testing.

PS - Newtonlaw - I reread my last message and realize it was pretty argumentative. Sorry about that... I was just trying to contrast your view with the opposite view for purposes of discussion.

 

[jburn]

Thanks for all of the good discussion on this topic. So far I have not heard of anyone testing the pressure relief, nor have I found any reference on transformer testing and inspection, that would recommend doing so. Everything just talks about a visual inspection. At this point I think I will recommend that we change the device at ten year inspection if the unit looks rusted or corroded. I have two transformers we plan to do some major work on this fall, and I plan on replacing the units. They are over 25 years old. I'll let our mechanical engineers test the old device if they wish. Thanks.

 

[jbartos]

Suggestion: Another way to go about this would be to check with the manufacturer of the pressure relief devices at the transformer to find out for how long their product should work without checking, testing and replacement. This could eliminate an unnecessary test or initiate tests/replacements that might have been overlooked.