Acid Fumes from Corona or Partial Discharge?
Acid Fumes from Corona or Partial Discharge?
(OP)
Can corona or partial discharges generate acid fumes?
If it is in an enclosed space can enough acid fumes build up to be noticeable?
After an inspection of a 21 KV generator terminal enclosure, a client sent this comment.
"The generator termination box door was opened, a strong smell of ozone was present and our eyes immediately had a burning sensation. Ozone and nitric acid are indicators of arcing in generator windings."
I am familiar with the ozone smell and have experienced that many times. The nitric acid is a new one to me.
Has anyone else heard of or witnessed acid fumes from electric discharges?
Visual inspection of the end windings did find discolorations and what appear to be pinholes, possibly due to partial discharge. The vendor is repairing and doing a root cause analysis. Generator vendor's engineers said the burning sensation was probably due to paint or insulating compound volatiles baking off this relatively new unit. (<1500 hours on line.)
If it is in an enclosed space can enough acid fumes build up to be noticeable?
After an inspection of a 21 KV generator terminal enclosure, a client sent this comment.
"The generator termination box door was opened, a strong smell of ozone was present and our eyes immediately had a burning sensation. Ozone and nitric acid are indicators of arcing in generator windings."
I am familiar with the ozone smell and have experienced that many times. The nitric acid is a new one to me.
Has anyone else heard of or witnessed acid fumes from electric discharges?
Visual inspection of the end windings did find discolorations and what appear to be pinholes, possibly due to partial discharge. The vendor is repairing and doing a root cause analysis. Generator vendor's engineers said the burning sensation was probably due to paint or insulating compound volatiles baking off this relatively new unit. (<1500 hours on line.)






RE: Acid Fumes from Corona or Partial Discharge?
High-enegy electrons strip O2 and N2 and H2 into free atoms. There is a law saying that if a stable compound can be formed, it will.
Ozone is O3, and that can be formed from oxygen atoms - easily. Nitric acid is HNO3, which can be formed from hydrogen, nitrogen and oxygen atoms. So, I think that is a possibility. In "our" accelerator it formed in quite large quantities and could be found in puddles on the plastic floor. It took out the ventilation equipment completely.
If ozone can be formed in a partial discharge environment, I think that nitric acid can, too.
Gunnar Englund
www.gke.org
--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...
RE: Acid Fumes from Corona or Partial Discharge?
Have a look at Photo 1 and Photo 2 for a couple of images of some really bad PD activity caused by poor installation. The lab determined that the corrosion was the result of a mixture of nitric and sulphuric acids. Images are about 750kB each for anyone on a slow connection.
----------------------------------
If we learn from our mistakes I'm getting a great education!
RE: Acid Fumes from Corona or Partial Discharge?
This website also explains the phenoomenon.
http:
RE: Acid Fumes from Corona or Partial Discharge?
I am sure that your lab was correct in determining that acids caused the corrosion seen in the photos, it would be foolish for me to assume otherwise. Also, I am sure that I am right in assuming that there was no source of water leakage from above the junction box since you would have seen that as an obvious cause of the corrosion. But, looking at the pictures, it appears as if there are flow marks from some liquid causing or contributing to the corrosion you are talking about.
I have seen this type of corrosion before where the outside of the junction box was corrosion free but the inside was corroded with signs of condensed liquid flow. In these cases I have concluded that there must be some source of moisture other than a leak from above. A clue that this is the case is if only one junction box is corroded. Your pictures appear to show the HV (2.4-15kV) junction box. Was there similar corrosion in the LV side?
One possibility that I have considered in this situation is outside air flow through the conduits carrying cool, moist air that condenses in the warm transformer junction box. This flow could be caused by temperature or pressure differences between the inside of the junction box and the outside. This is especially a possibilty when the transformer is located inside of a building and the conduit is running to an outside enclosure. Of course, the flow could go the reverse direction into the junction box through the cover and out through the conduit but I would think this is a less likely scenario. In either case, I do see that the mounting flange for the enclosure cover is also corroded. Was there a gasket in place?
Of course, once the high humidity exists, I agree that the resulting corona discharge may develop acids that would accelerate the oxidation effect and produce the type of gummy, sticky looking rust that your pictures show.
RE: Acid Fumes from Corona or Partial Discharge?
I cannot imagine that cool moist air condenses in a warm junction box. As you recall from physics class, cool air contains less water than warm air does and it is usually so, that water vapour in warm air condenses when it meets a wall with lower temperature. Not the opposite.
Gunnar Englund
www.gke.org
--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...
RE: Acid Fumes from Corona or Partial Discharge?
RE: Acid Fumes from Corona or Partial Discharge?
When sparks occur from partial discharge or corona, chemical reactions take place. One of these reactions is the formation of Ozone. Although ozone generally gets the credit, its is not the primary cause of the smell. The primary contributor is nitric oxide, which is a very sharp smelling gas. If high humidity is present, a fine film of nitric acid can be created on nearby surfaces. The nitric acid corrodes metal, creating a greenish (verdigris) discoloration of copper or may rust steel after some time.
RE: Acid Fumes from Corona or Partial Discharge?
The gasklet was certainly in good shape, although I'm very familiar with the 'breathing' phenomenon described by you and Skogs. Definitely a possibility, especially as the box design left the gasket edge exposed to the elements and thus prone to having water accumulating on them. In my opinion this is a likely source of the water, although it's clearly been in a vapour state to have reached the upper corners of the box, and then presumably condensed and run back down as a liquid.
I'm not sure if it the greenish colour is visible in the photo, but there was a fair amount of green liquid present on the earth bar.
----------------------------------
If we learn from our mistakes I'm getting a great education!
RE: Acid Fumes from Corona or Partial Discharge?
Steve
RE: Acid Fumes from Corona or Partial Discharge?
Gotta admit, this is news to me.
Reminds me of scotty's sigline. :)
Muthu
www.edison.co.in
RE: Acid Fumes from Corona or Partial Discharge?
The mess in my photos was the air insulated terminal box of a small transformer. PD doesn't happen in a hydrogen atmosphere, or at least it doesn't at typical generator voltages up to about 30kV or so. Quite what happens in hydrogen at very high voltages is something I don't know.
----------------------------------
If we learn from our mistakes I'm getting a great education!
RE: Acid Fumes from Corona or Partial Discharge?
The high voltage termination enclosure is in a cooling air plenum cast into the generator foundation. That's why there was so much "ozone" smell and burning eyes when the termination compartment was opened. We have found many more locations where PD activity has occurred on the end windings and are concerned about the longevity of the insulation system.
RE: Acid Fumes from Corona or Partial Discharge?
Three of the 20 sections very suddenly developed severe corona discharge that was audible over 20 feet away. Investigation prior to shutting down revealed the Glastic insulators involved with these sections were emitting the incredible audible corona noise. Infrared indicated a 1" ring around each of the insulators that was 20F above ambient.
Further investigation revealed that all the fiberglass components (Glastic insulators, bus supports,etc.) associated with these three sections were severely damaged due to tracking. Both line side and load side compartments were covered in a light rust inside and out. Most of the tracking damage was due to the layer of rust. There are no conduits or methods for moisture to get into these sections. The Glastic insulators were covered with an oily film.
The Glastic insulators associated with those three sections were replaced with a cycloaliphatic insulators with greater creep distance. All other insulated components were replaced and the gear cleaned and painted.
That was 8 years ago. We go through the gear each year (line and load compartments) and have found nothing wrong since then. No problems have developed with the remaining original Glastic components.
The insulators showed no evidence of tracking or discoloration, only the oily film. We performed a number of in house electrical tests on the damaged insulators at the time and determined they may have just been defective.
Now I am not so sure. I still have a couple around my shop. They are still oily. I may just have to send one out for chemical analysis.
RE: Acid Fumes from Corona or Partial Discharge?
We use Lexan to replace these failed barriers all the time with fantastic success. I get 1 or 2 such failures every momth.
What is neat is there are some new (To the US, the UK guys have been using them for awhile) handheld PD monitors that monitor Airborne Acustics and TEV via capicitive coupling, You can do a survey of a MV lineup in a few monutes and determine if any PD activity exists or not and if it does what the extent of it is to some degree. Of course if the detector finds PD activity further investigation is a good idea to pinpoint the location and severity.
There is some good info on PD detection here:
http://www.nooutageelectricaltesting.com/
RE: Acid Fumes from Corona or Partial Discharge?
The case you describe is interesting. Three of twenty units had failure of Glastic (epoxy-cast) insulators after 12 years of service, 1992-2004 (present minus 5 years). You replaced those insulators with a different type in those three units. Now, another five years has gone by and there are no additional failures on the new insulators or on the remaining 17 original insulators.
I wonder what makes the three units that failed prematurely different than those that remain in service. If there are no conduits then you would think that there would be no outside chemical contamination in those three that would not also affect the rest (no 'breathing' as ScottyUK describes it) .
It follows that I also wonder if the reason that those three units have not failed since is because of the different insulator material or if it is because the cause of the original failure was a one time event that affected only those three units (ie. lightning or some other surge).
Do you have any ideas about this? I am just curious for the sake of experience.
RE: Acid Fumes from Corona or Partial Discharge?
We monitored the equipment for a number of years on a regular basis. Ultrasonic inspections were performed every few months on the entire lineup for almost a year. Facility people walked the switchgear room each day for a year and still do each week specifically looking for corona. Our concerns were based on the fact that original problem had come up extremely quickly. In addition, the problem was not a typical insulation failure (at least to me). Standing over 20 feet away the audible noise was incredible. Loud enough I honestly thought we were going to loose the gear any moment.
The switchgear mfg. uses 5kV indoor rated (non fluted) Glastic insulators to support fluidized bed epoxy insulated bus on 15kV equipment. We had the opportunity to replace these with better material with a greater creep distance. We purchased enough to replace all insulators (over 100) associated with the line up and may actually do so over the next few years.
It has been my experience that the Westinghouse/Eaton VCP style switchgear is quite robust if installed correctly in a reasonable environment and maintained. The switchgear was (and still is) in incredibly great shape. The thing that stunned us was the contrast between these deteriorated cells and the perfect cells right next to each them. It was like looking at a section that was left outside along the road for 5 years and a brand new section about to leave the factory.
We believe there were a few defective standoffs that apparently gave off some type of gas that deteriorated the components in those compartments quickly. Reading this thread may help us look further into it and come up with a better understanding of what happened.