Explosion in X-Proof Starter 1

[acwalsh]

I was called to investigate an explosion of a 30 year old Siemens combination starter. The starter is an explosion proof design with a threaded top and bottom cover; the starter is used to control a capstan on our jetty. Following the explosion, one half of the top cover was found approximately 40 feet away, the other half presumed to have gone overboard. The part that was found showed that a crack, approximately 4" long, had existed in the cover for some time. There was no visible signs of an explosion, no soot, no charring and no broken parts (except for the cover). There was no telltale odour of the kind typically associated with electrical arcing or fire. Phase to phase and phase to ground megger readings of both L and T leads were good. A replacement cover was found and the capstan was placed back in service. Can anybody offer a plausible explanation as to what happened?

 

[jraef]

Explosive gas of some sort got in. Happens all the time.

This may be news to you, but "Explosion Proof" does not mean what you probably think it does! All it means is that IF an explosion happens inside of the enclosure, the hot gasses released will be sufficiently cooled off so as to not ingnite any surrounding gasses in the atmosphere (of the type the enclosure is listed for). So if your entire facility did not go up in a mushroom cloud, the enclosure actually performed admirably! It gave it's life in the call of duty, saving all those around it.

"Our virtues and our failings are inseparable, like force and matter. When they separate, man is no more." Nikola Tesla
Make the best use of Eng-Tips.com Read the Site Policies at faq731-376
Member, [blue]P³[/blue]

 

[itsmoked]

Further "explosion proof" because of the cooling paths, mentioned by jraef, almost guarantees that the enclosure is NOT air tight. So gasses could easily get in.

Next question would be, "could an explosive gas been present at the time, or recently?"

 

[jraef]

Generally outdoor rated Ex-Proof enclosures are lightly sealed (i.e. a large O ring in a groove or a grease on the mating surfaces) so that water and therefore gasses can't enter easilly, but you are right, not all are and it is not specifically intended to keep everything out. That old crack could have been a point of entry if the gasses/vapors were local. If so, you were lucky the explosion cooled enough going through that to avoid igniting anything else.

Another thing that happens is that conduit seals eventually fail letting gasses enter from some other area, and/or exposed conduit heats during the day, then cools at night, creating a vacuum that draws local gasses into the box. Most facilities have a scheduled maintenance item to check conduit seals periodically or replace them every so often just for that reason. The fact that the covers blew off, and you had a potentially dangerous crack yet neither circumstance caused local ignition would point to gasses entering the enclosure from somewhere else. I would follow the conduits and see where that may have been. A fuel spill perhaps?

"Our virtues and our failings are inseparable, like force and matter. When they separate, man is no more." Nikola Tesla
Read the Eng-Tips Site Policies at faq731-376
Member, [blue]P³[/blue]

 

[itsmoked]

THAT is probably a much better source of gasses then the starters local environment.

 

[GTstartup]

Since the enclosure blew itself apart it did not "perform admirably"

From NFPA 70

Explosionproof apparatus: "Apparatus enclosed in a case that is capable of withstanding an explosion of a specified gas or vapor that may occur within it and of preventing the ignition of a specified gas or vapor surrounding the enclosure by sparks, flashes, or explosion of the gas or vapor within and that operates at such an external temperature that a surrounding flammable atmosphere will not be ignited thereby." (See NFPA 70)

I don't know for sure what caused it but have to agree that somehow combustible gases built up and ignited when the contactor operated. I will say however, that I would replace the whole enclosure as it clearly failed and the next time might take somebody's head off.

 

[itsmoked]

GTstartup I would agree totally with the "replace it" endorsement.

Now. Are not EXP enclosures rated for a specific volume of enclosed space? Hence an empty one would maybe fail/explode whereas one with say 3/4ths of it's volume filled with solids/equipment wouldn't? Or am I remembering wrong(happened once b4). I could see where that might make testing them a pain but I can also see where having a rule like this could save huge amounts of weight and material.

 

[acwalsh]

Thanks to you all for your interest and feedback. Some further developments. It has since been discovered that preventive maintenance was performed on the starter only hours before the explosion. The maintenance involved the use of an alcohol based contact cleaner.................. It is very probably that alcohol vapourized inside the sealed enclosure and exploded either when the contactor pulled in or released. Alcohol burns very cleanly; this may explain the lack of evidence afterwards. As to the failure of the "explosion proof" starter to contain the explosion, the integrity of the top had already been compromised. We are taking the following steps:
1. Check all tops for wear or cracking and replace any that are suspect.
2. Try to find alternate, non-volatile, contact cleaner. Failing that, use cleaner sparingly and leave both top and bottom covers off starter for a minimum 30 minutes following application to allow alcohol to flash off.

Thanks,
Andy Walsh

 

[rbalex]

Neither explosionproof enclosures nor seals are designed to be gas-vapor tight. See 2005 NEC Section 501.15(C) Fine Print Note #1 or [501.5(C) FPN No1 from earlier editions]; i.e., ingress of gas or vapor is not a failure.

The starter enclosure is most likely listed for Class I, Group C&D and the alcohol-based cleaner alone should not have been a problem. Depending on its age, if you look inside the enclosure, you will likely find it is listed for use on systems with available faults no greater than 10kAic - no matter what the interrupting duty of the starter is.

Arc-pressure forms even for simple load interruptions and, for typical XP enclosures, low available displacement volume caused by a “full” enclosure exacerbates the effect; however, basic explosion proof enclosures are tested both hydrostatically and by direct application at full volume. A few specialized enclosures, such as formed plastic or ceramic-type push-button contact blocks, depend on limited available volumes to limit total available ignition energies.

The most likely problem, as you pointed out, was the already compromised cover; it may well have blown off at any time from arc-plasma whether there was gas inside or not.

 

[jraef]

I think your explaination is simple and very plausible, assuming of course that what you meant was that the cover was removed, the contact cleaner was sprayed on the contactor and other components inside, and then the cover was immediately replaced before the vapors had a chance to evaporate. If trapped inside the enclosure when the contactor energized, the spark would have ignited those residual vapors for sure. Good catch.

I also think that if you read the can of contact cleaner, it will tell you specifically to avoid energizing any equipment where it has been used for a specific amount of time. 30 minutes may be a little overkill, but better than too short of a wait.

By the way, I doubt you will find a contact cleaner that is non-flamable. But a better question is, why do you feel it is necessary to clean contacts in a sealed enclosure like that?

"Our virtues and our failings are inseparable, like force and matter. When they separate, man is no more." Nikola Tesla
Read the Eng-Tips Site Policies at faq731-376
Member, [blue]P³[/blue]

 

[dpc]

I think even if the contact cleaner was the immediate cause, the cover should not have been blown off an explosion-proof box.

The explosion-proof boxes generally have machined surfaces that force a long enough exhaust path for gas produced in an internal explosion that the exhaust gas is below the ignition temperature of whatever hazard group the equipment is rated for. It is difficult to find enclosures that are both explosion-proof and completely weather-proof or at least it was the last time I tried.

Since the enclosure is old, is it possible that someone at some time decided to "improve" the enclosure by adding a gasket or liquid sealant that was not part of the original design? This might have allowed sufficient pressure build up during an internal explosion to damage the enclosure.

I agree with earlier comments that the entire enclosure should be replaced.

 

[rbalex]

He already mentioned it was a threaded cover design. If it was a cast-cover (very likely) a hairline fracture is all it would take to fail over time. Bluntly, if this is a Division 2 application, in absence of any other evidence of enclosure damage, replacing the cover (as they did) is about all they needed to do.

 

[jraef]

Agreed. In those spin top starter boxes, the cover is usually a can with female threads screwing onto a cast body with male threads. The flame path is the number of threads, not a machined surface. A crack in the cover could easilly have allowed it to spread and pop off with even minimal internal force.

"Our virtues and our failings are inseparable, like force and matter. When they separate, man is no more." Nikola Tesla
Read the Eng-Tips Site Policies at faq731-376
Member, [blue]P³[/blue]