 |
 | Home | Contact Us | FAQ | Search & Site Map | Link to Us |
 | Sign In | Join | Other 45 Sites in Network |
| HomeArticlesDiving DestinationsLearning Scuba DivingMarine LifeMiscellaneousDiscussion GroupsGeneralScuba EquipmentScuba LocationsAustralian ScubaUK ScubaDirectoryScuba Clubs |
Scuba Forum / General / February 2006Aluminum cylinder explosions in fire.
We all watched in horror as the aluminum cylinders exploded onboard the evacuation bus during the Katrina disaster. An I wondered then "Do we need to know more about this failure mode?". Now it has hit very close to me personally. Both my former boss and his next door neighbor were on supplemental oxygen. They both had several of the aluminum oxygen cylinders in their homes. A fire broke out in his neighbors home and one by one every single cylinder exploded!! Not one cylinder was protected by the burst disk! Every cylinder blew apart. The entire front wall of my freinds house collapsed from the repeated shock waves of the blasts. It seems the annealing of the aluminum weakens in the heat of the fire and the cylinder bursts long before the burst disk ever blows! Please respect this fact about aluminum cylinders in a fire, this would include your scuba cylinder in a house or car fire! The cylinders that blew my former boss's house apart were only the small size of perhaps 20 cubic feet each. I still dive with aluminum tanks on occasion, and I don't fear them if they are well maintained and properly inspected, but alas that doesn't mean a damn thing in the event of fire. Manufactures, please wake the hell up and tell us of this imposing danger or fix your worthless burst disk malfunction!! If I ever see an effort to spearhead a major lawsuit against you, I will surely be on the bandwagon this time. Your cylinders WILL NOT survive a fire without exploding!! I value my life and my safe escape time!! Anyone here have their own horror story to add? Incidently, the neighbor lady died one day later and I will be attending the funeral tomorrow. > We all watched in horror as the aluminum cylinders exploded onboard > the evacuation bus during the Katrina disaster. An I wondered then [quoted text clipped - 5 lines] > > Not one cylinder was protected by the burst disk! That's not DIR. No aluminum cylinder is protected by burst disks, burst disks are in the valves. Burst disks are designed to rupture and approximately 150% of working pressure. Aluminum is GREATLY weakened by high temperatures, if the cylinder is in a fire the metal may weaken before the burst disk goes. Some of the older tank valves had plastic dip tubes. these would melt in a fire and prevent the burst disk from blowing. That being said have a look at propane cylinders or O2 cylinders no burst disks anywhere. Don't even think about spray cans. Timothy Ewing > We all watched in horror as the aluminum cylinders exploded onboard > the evacuation bus during the Katrina disaster. An I wondered then [quoted text clipped - 29 lines] > lady died one day later and I will be attending the funeral tomorrow. > > We all watched in horror as the aluminum cylinders exploded onboard > the evacuation bus during the Katrina disaster. No, it was the Rita evacuation... > A fire broke out in his neighbors home and one by > one every single cylinder exploded!! Aluminum isn't as resistant to heat as steel... Not sure what would have happened with steel tanks, but they would have probably been able to cook a bit longer than the steel ones... The main thing though is not that they were pressurized tanks, but rather the fuel source (i.e. O2)... If they had been pressurized air, it would have been different... > Manufactures, please wake the hell up and tell us of this imposing danger > or fix your worthless burst disk malfunction!! If I ever see an effort > to spearhead a major lawsuit against you, I will surely be on the > bandwagon this time. Your cylinders WILL NOT survive a fire without > exploding!! I value my life and my safe escape time!! I don't think they ever claimed it would... Hell, they even say that you shouldn't paint the tanks with anything that requires the paint to be baked on it... That's a hell of a lot less temperature than would occur in a fire... I think that the person in question with the O2 tanks would be in a different situation if the tanks had been air instead of O2... The O2 added fuel to the fire, causing more things to burn rapidly, causing more gas to be produced than the amount that was in the actual tanks... > > A fire broke out in his neighbors home and one by > > one every single cylinder exploded!! > > Aluminum isn't as resistant to heat as steel... Not sure what would have > happened with steel tanks, but they would have probably been able to cook a > bit longer... The basic mechanism is identical: a) the fire causes heating, which as per PV=nRT, will cause the tank's internal pressure to rise. b) the internal pressure increases until something fails - - either the burst disk (if present), or the structure of the tank. c) the same heating may also cause mechanical weakening of the pressure vessel, which lowers the internal pressure required for it to fail. This is dependent on several factors, but the major ones of interest here are its material (susceptibility) and its design (magnitude of design safety margin). As a general expectation, I'd expect an AL tank to rupture at a lower pressure due to heating than a Steel tank because many AL alloys are known to have reduced material strength properties at "relatively" low (ie, below 350F) temperatures. > The main thing though is not that they were pressurized tanks, but > rather the fuel source (i.e. O2)... If they had been pressurized air, > it would have been different... Ummmm...not exactly (or more accurately, not for the initial mode). Consider the simpler case of a single tank: so long as the tank isn't leaking, how could the fire have knowledge or access to the tank's contents for it to use (ie, burn more intensely)? Answer: It doesn't. The fire pragmatically has no way of using whats inside the tank until the tank ruptures. Still considering the simple single tank scenario, there is the possibillity of a leak prior to rupture. Said leak would be venting either via burst disk or burned off supply hose, and this could provide a "blow torch" effect which would accelerate the fire and could cause a rapid enough rise in temperature-caused pressure that the tank could rupture despite the presence of pressure venting (ie, the "leak"). Here, since the internal gas is being actively released into the fire, the fire will be able to "know" what it is. For multi-tank situations, tank #5 will likely have had a hotter fire due to the contents of venting or ruptured tanks #4, #3, etc. To this end, if you have a full tank sitting with a bunch of empties, the burst disk of the full one will pop, and there's your fire acceleration source to more rapidly heat up the empties to get them to the temperature where they have metallurgical failure ruptures prior to reaching the pressure where their burst disks will fail. > > Manufactures, please wake the hell up and tell us of this imposing danger > > or fix your worthless burst disk malfunction!! If I ever see an effort [quoted text clipped - 3 lines] > > I don't think they ever claimed it would... The manufacturers are probably shielded from liability by the DOT CFR regulations that require a certain type of protection system, with an "industry standard" testing method. Furthermore, when you look at the system and this failure mechanism, the reality is that tanks are significantly more likely to suffer such a catastrophically blow when the consumer has left them laying around half-empty...and who's responsible for that? Similarly, this all also suggests that full tanks (which will pop their burst disks) shouldn't be stored with empties (pendant note: "unless their contents act as a fire suppressant") and this is also the responsibility of the consumer. -hh > We all watched in horror as the aluminum cylinders exploded onboard > the evacuation bus during the Katrina disaster. An I wondered then > "Do we need to know more about this failure mode?". > It seems the annealing of the aluminum weakens in the heat of the > fire and the cylinder bursts long before the burst disk ever blows! What more do you need to know? Do you think we need to imprint each cylinder with the words "do not dispose of in a fire?" > Manufactures, please wake the hell up and tell us of this imposing danger > or fix your worthless burst disk malfunction!! If I ever see an effort > to spearhead a major lawsuit against you, I will surely be on the > bandwagon this time. Your cylinders WILL NOT survive a fire without > exploding!! I value my life and my safe escape time!! Do you buy your coffee at McDonalds? Do you think we need to imprint each cylinder with the words "do not dispose of in a fire?" Here's a clue. By the time a fire is hot enough to weaken an aluminum tank, it's beyond hot enough to burn a human body. Try to be out before things get to that point. Since you need to be told it's not a good idea to burn full tanks, here's a few more. Black powder is explosive. conventional powder, if contained, acts pretty much the same and, even when not contained, burns very fast and hot. Gas burns fast and hot gas fumes are explosive. Pressurized cans of all kinds can explode in a fire and are quite impressive if filled with something flamable. Paints, bug sprays, LP gas, hair spray and some silly string cans are filled with something flamable. Lee >> We all watched in horror as the aluminum cylinders exploded onboard >> the evacuation bus during the Katrina disaster. An I wondered then [quoted text clipped - 28 lines] > >Lee Yes, you are correct, and I understand that common sense plays a key factor here, and that the Darwin awards show that many have no common sense, or perhaps any sense. I guess my rant would really suggest that I would like to see burst disks for protection of aluminum cylinders made of a very low melting point material, something that would let go long before the cylinder did. For those that suggested the bursting tanks added FUEL to the fire, in fact they added an OXIDIZER to the fire. By the way, propane tanks bigger than 14 ounces have relief valves. That's what that funny looking thing like an inside out tire valve is. It's not the filler, it's an exit point. The steel propane tanks have a decent fire survival rate, the aluminums do not. Just look at the recall of aluminum propane tanks by the maker of hot air balloons called Aerostar. I won't put my AL50 pony in a fire on purpose, It would just be nice to know it won't blow a wall out of my house or kill firefighters in the event it was in a house fire or car fire. That was kind of my point. <snip> > I guess my rant would really suggest that I would like to see burst > disks for protection of aluminum cylinders made of a very low melting > point material, something that would let go long before the cylinder did. > <snip> Early scuba cylinders had just such a device. It was a straight through burst plug that was filled with lead. It was supposed to go at a preset melt point. I only remember a few of them breaking down or corroding to a point that the lead would become dislodged and shoot out like a bullet! Many dive shops refused to fill cylinders that had them for fear of being shot. There was a lot of confusion as to the use and some nimnals would back them up with the conventional copper disc ... virtually eliminating the chance of them ever relieving! In the early 80s, King Neptune (remember them you old timer diver shops?) pumped a cylinder valve up to over 10,000 psi and couldn't get one to fail. In fact, the valve casting failed first! Those burst plugs came from the similar oxygen cylinder post valves in use at the time. There was an outcry then to BAN all lead filled plugs. And they were! Seems that the manufactures just have to make us a valve that has both an over pressure plug and a high temperature plug. Of course, we could also follow the guidelines of never storing a half full cylinder at home. Leave it either full so any heat at all will get it to burst pressure or leave it almost empty so that it won't be an explosive hazard in a fire. Just my $.02  SignatureRay Contreras =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Webmonkey for: http://www.ossystems.com http://www.bobs-garage.com http://www.clanwolverine.org http://www.rayzplace.com > Of course, we could also follow the guidelines of never storing a half > full cylinder at home. Leave it either full so any heat at all will get > it to burst pressure or leave it almost empty so that it won't be an > explosive hazard in a fire. Store the cylinder underwater so that the most that would happen is that it get to 212F? > For those that suggested the bursting tanks added FUEL to the fire, > in fact they added an OXIDIZER to the fire. Yeah, I know that technically O2 is an oxidizer, not a fuel, but I was not using it as a purely technical term... I was using it as basically meaning something that promotes burning... I think that the O2 tanks exploding would cause the fire to burn hotter and thus more gas be generated from the burning in a shorter period of time than if it has been air in the tanks... One might wonder if just having air in the tanks would have been something like the blowing out of a candle... I'm not sure, but I suspect that if you're in a room with tanks that have been heated to the point of metal fatigue, the failure of the tanks should not be your most immediate concern... >>> We all watched in horror as the aluminum cylinders exploded onboard >>> the evacuation bus during the Katrina disaster. An I wondered then [quoted text clipped - 48 lines] >to know it won't blow a wall out of my house or kill firefighters in >the event it was in a house fire or car fire. That was kind of my point. Do you have a lawn mower? You probbly have a 1 gallon gas can. There's a lot of umph in that empty can of gas. Same thing with that hungry SUV in the garage that you're going to fillup in the morning. Signaturedillon Could have been is in the past Could be is in the future There is only the now > I guess my rant would really suggest that I would like to see burst > disks for protection of aluminum cylinders made of a very low melting > point material, something that would let go long before the cylinder did. Even if made of candle wax, burst disks would be useless in a fire, where the fire is raising the pressure inside the tank faster than the burst disk can relieve the pressure. > >> We all watched in horror as the aluminum cylinders exploded onboard > >> the evacuation bus during the Katrina disaster. An I wondered then [quoted text clipped - 36 lines] > disks for protection of aluminum cylinders made of a very low melting > point material, something that would let go long before the cylinder did. If you keep your cylinder full or empty it would. Problem solved. > For those that suggested the bursting tanks added FUEL to the fire, > in fact they added an OXIDIZER to the fire. By the way, propane tanks [quoted text clipped - 7 lines] > to know it won't blow a wall out of my house or kill firefighters in > the event it was in a house fire or car fire. That was kind of my point. <snip> > Both my former boss and his next door > neighbor were on supplemental oxygen. They both had several of the [quoted text clipped - 7 lines] > It seems the annealing of the aluminum weakens in the heat of the > fire and the cylinder bursts long before the burst disk ever blows! <snip> How much pressure was in the cylinders? If they were full I would think the burst discs would usually blow, but if they were partly full (say, 1000 PSI), I think there is a good chance they will heat enough to weaken in a fire before the pressure causes the burst disk to blow. Burst discs are supposed to burst at or below the hydro test pressure for the tank. Hydro test pressure is 5/3 of the rated pressure. For a 3000 PSI tank, that's 5000 PSI. Say the tank is filled to 3000 PSI at 70F. That's 530 degrees absolute. To get to 5000 PSI, the temperature has to go up to 883 absolute (530 * 5/3), or 423F. This is far below the melting point of 6061 aluminum, which starts around 1100F. So if the tank were full, I would expect the disc to blow. (However, I see that Luxfer says tanks that have been heated to 350F or more must be condemned.) But what if the tank has 1000 PSI in it? To triple the pressure, we have to triple the temperature, going from 530 degrees absolute to 1590 absolute, which is 1130F. 1130 is more than 1100. I think it's safe to conclude that the tank will explode well before the disc would blow. Somewhere, someone has mapped aluminum's strength as a function of temperature, which would let me form some opinions about the minimum and maximum safe storage pressures, but I couldn't find it easily on the web. I can see one other thing to consider. If the tank is heated very rapidly it's possible that the tank could weaken before enough heat was transferred to the gas to cause the pressure to rise enough to cause the disc to blow. This all leads to what our LDS says about aluminum cylinders: Store 'em full or store 'em nearly empty, but not in between. (Have I been religious about this? Well, no, but now that you've made me think about it....) Anyone see a flaw in this reasoning? Anyone know at what pressure the burst discs will actually blow? John Eells A coupe of years ago, one night, a dive boat in our area caught fire and burned to the waterline. Dive gear had been loaded onto the boat for a dive planned to take place the next morning. Included with the gear were several tanks, fully charged, both aluminum and steel. All of the aluminum tanks reptured. All of the steel tanks were intact but with blown overpressure disks. Now, this was far from a scientific experiment but neither was it conjecture - just a real world example. Safe diving, Cpt. Dale > A coupe of years ago, one night, a dive boat in our area caught fire and > burned to the waterline. Dive gear had been loaded onto the boat for a dive [quoted text clipped - 3 lines] > disks. Now, this was far from a scientific experiment but neither was it > conjecture - just a real world example. This may have to do with the material used for overpressure disks. I guess that ideally they should be made in a material that weakens at least as fast as that of the tanks themselves under the influence of temperature, so that they serve their purpose regardless of temperature. However If the overpressure disks are the same for both Al and steel tanks, since Al's mechanical properties are more affected by heating, then it is logical that beyond a certain point the burst disks will no longer protect the Al cylinders while they continue to protect the stell one. Cheers; Froggy in message news:FPLCf.4$5z3.3@fe10.lga... > <snip> >> Both my former boss and his next door [quoted text clipped - 27 lines] > disc to blow. (However, I see that Luxfer says tanks that have been > heated to 350F or more must be condemned.) Nice try, but not relevant. The tanks don't melt, they weaken and, when they do, the pressure ruptures them. Lee > > Burst discs are supposed to burst at or below the hydro test pressure for > > the tank. Hydro test pressure is 5/3 of the rated pressure. For a 3000 [quoted text clipped - 9 lines] > Nice try, but not relevant. The tanks don't melt, they weaken and, when > they do, the pressure ruptures them. True, that's the failure mechanism, but the basic analysis methodology is sound: it shows that the tanks would have to fail (if by no other means than by melting) before the rupture disk's pressure rating is reached. FWIW, its been awhile since I've looked up various AL alloy grades for their resistance to temperature, but I do recall that there's at least one variant that starts to have its material properties degrade at below 200F. There's also at least one AL tank manufacturer whose documentation calls for an upper temperature limit of 130F. -hh > If they were full I would think the burst discs would usually blow, but > if they were partly full (say, 1000 PSI), I think there is a good chance > they will heat enough to weaken in a fire before the pressure causes the > burst disk to blow. yup. Even more so if empty (ambient). >> If they were full I would think the burst discs would usually blow, but >> if they were partly full (say, 1000 PSI), I think there is a good chance >> they will heat enough to weaken in a fire before the pressure causes the >> burst disk to blow. > >yup. Even more so if empty (ambient). There was a great demo of this in CoCo View. Two tanks, one that looked a little slagged and the other with a huge blown out side. The slagged one had been full and popped the burst disc. The blown out one had 500-750 psi and failed dramatically. Only a thin piece of Al would have prevented that one, except that it probably wouldn't hold the 3000 psi. Can't have it both ways. As I said before, that gas can in the garage is a bigger threat. As is the gas water heater. Signaturedillon Could have been is in the past Could be is in the future There is only the now |
Reply to this Message |
 Sign In
 Join
 My Latest Posts My Monitored Threads My Blog My Photo Gallery My Profile My Homepage Start New Thread Enable EMail Alerts Rate this Thread
|
©2008 Advenet LLC Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.