Can cold water cause a BLEVE?

[LawFires]

We were discussing BLEVEs at the station yesterday, and the SOP for a potential BLEVE involves, among other things, trying to cool down a container that is a potential BLEVE before it Explodes.

I was thinking about this, and I know from my welding experience that if you have metal that is hot and under tension, and you quench it suddenly with cold water, you can fatigue the metal and/or make it very brittle, causing it to lose its tensile strength very rapidly.

I imagine this is less of an issue on a large potential BLEVE container, as the cold water wouldn't have enough 'chill' to cool it rapidly enough to matter, but on a small potential BLEVE (e.g. small LP tank that's been heated up by a fire), is there the potential that a shot of cold water from a hose could fatigue the metal and actually cause the BLEVE?

My suspicion is that the sage answer here is: maybe, but it's the lesser of the available evils...But should the SOP involve a more gradual application of water to the container, rather than a sudden direct blast of cold water?

[fireman4949]

My thought is no.

I believe that a container (and the liquid inside it) that is reaching temps that are approaching those high enough cause a BLEVE will have a slow enough heat dissapation rate that the amount of cold water applied via a fire stream will be insufficient to cause the metal to fracture. The rate of cooling of the liquid inside the container, as well as the rate of cooling of the container itself from the applied stream should be slow enough to allow the metal to react without losing its integrity.

Perhaps if the container was suddenly, totally immersed in ice cold water there may be a greater risk of failure, but I don't think a fire stream will have the potential to cause it.




Kevin

[GeorgeWendtCFI]

My thoughts are along the same lines as Kevin. If it is a large container, I am not sure that the rate of effective water application will result in a rapid dissipation of heat. Remember, you are not submerging this container. Only a fraction of the water applied to the tank will actually be cooling (absorbing the heat) it.

I understand completely what you are saying, and your analysis is sound, but I think a fire/BLEVE situation is different.

[don120]

Bleves are caused by the expanding LP in the tank...this increase of internal
pressure is what causes the tank to fail.

LP is liquid in the tank when under pressure with some vapour.
LP boils at -40
LP when heated to ambiant temps expands to 270 times it's liquid volume.
It is this pressure increase that results in a bleve.

The tank could not get that hot , where the cold water would
damage the tank and cause a bleve,before the expanding LP would
result in internal pressures great enough to bleve.

So if you have fire impinging on a tank keep water on the tank ....
large tanks will could require an unmanned stream set up.

Hope this helps

Don

[mcaldwell]

I am on side with Kevin.

And Don, don't forget that significant flame impingement on the metal above the vapour space of a tank can cause the metal to fatigue and fail long before the liquid on the bottom reaches a critical temp and pressure. The convective cooling of the metal by the contents is dramatically less on the top half of the tank (above the liquid line), than it is on the bottom half of the tank, so the relief valve can be functioning within acceptable volume, and you may still see a rupture.

But ultimately, IMO the best reason for cooling the tank is this; Failing to cool the tank WILL result in a BLEVE.

[LawFires]

Based on that analysis, then I'm guessing the best area to direct the stream of water would be (in order of priority): 1) the source of the heat; and then, 2) the upper portion of the tank (perhaps top 1/2)?

[mcaldwell]

Based on that analysis, then I'm guessing the best area to direct the stream of water would be (in order of priority): 1) the source of the heat; and then, 2) the upper portion of the tank (perhaps top 1/2)?

Yes, when there is flame impingement on the upper section of the tank, or above the vapour line (which is not always identifiable).

When the flame is affecting the bottom of the tank, you will see the excessive pressure buildup Don mentioned, resulting in either a discharge of the fuel vapour through the safety valve, or eventually a rupture of the tank due to excessive pressure.

When heated from below, the fuel in the bottom of the tank needs to be cooled to keep the evaporation rate down. Most BLEVE's occur once most of the fuel has vented off, and there is not enough liquid left to internally cool the cylinder walls fast enough. If you cool the bottom of the tank in that scenario, you can help reduce the rate at which the liquid vapourizes and blows out the relief valve. This may buy you more time in which to fight the source fire. Either way, as the relief valve discharge volume and noise increases, so should your distance (and rapidly ).

[don120]

I will respectfully disagree with the posts above that priority should be
given to the source fire....

I can think of 2 incidents involving large and medium size propane tanks,where
those tactics were used and the results were six firefighters killed and over
a dozen injured. In both cases bleve occurred within 10 minutes of arrival
of the FD in both cases the tanks were venting and the attack lines were
directed at the source fire.The large tank incident was in the U.S. and was investigated by NIOSH it occurred in 1998...the medium size tank incident
was in Warwick Quebec Canada,June 27,1993.

In ALL cases where a propane tank is involved priority is the tank.

Stay safe
Don

Just found the US incident: April 9,1998 in Albert City Iowa.

[fireman4949]

I will respectfully disagree with the posts above that priority should be
given to the source fire....

I can think of 2 incidents involving large and medium size propane tanks,where
those tactics were used and the results were six firefighters killed and over
a dozen injured. In both cases bleve occurred within 10 minutes of arrival
of the FD in both cases the tanks were venting and the attack lines were
directed at the source fire.The large tank incident was in the U.S. and was investigated by NIOSH it occurred in 1998...the medium size tank incident
was in Warwick Quebec Canada,June 27,1993.

In ALL cases where a propane tank is involved priority is the tank.

Stay safe
Don

Just found the US incident: April 9,1998 in Albert City Iowa.

Absolutely!

[mcaldwell]

I will respectfully disagree with the posts above that priority should be
given to the source fire....

In ALL cases where a propane tank is involved priority is the tank.

I think you were misreading my post Don (or I wasn't clear in my description). I was talking specifically about cooling the tank in all scenarios.

The difference comes in where the source flame is impinging on the tank, and whether or not to cool the top or bottom of the tank. You cool the hottest spot of the tank first, as that is usually going to be the failure point, or at least the secondary cause of the failure.

[MEDIC0372]

If you are not cooling it, it is getting hotter...so cool it ASAP!

[Catch22]

From the training I've had, it's better to cool the tank with a solid stream (unmanned master stream) below the vapor line. This will #1 cool the liquid that's boiling, #2 prevent the rapidly cooling vapors from causing a vacuum and collapsing the tank.

If you're one of those guys that has to "see it to believe it" like I am, make a little hanger that'll fit the top of a coke can. Fill the can half full with water and cap it with a latex glove. Hang the can on something and put a torch at the bottom. The glove expands as water turns to gas so you can see how it works. Then take a spray bottle and spray the can using different areas and different "streams." I will mention, you might want a few cans, when you put water on the vapor space areas you'll collapse them in a hurry.

I have seen this happen with some fuel tanks that were on stands next to a structure fire. Everything was fine until the mutual aid dept. decided to cool the tanks. Vents closed, tanks collapsed as they sprayed the tops of the tanks.

[BirkenVogt]

Those would have to be liquid fuel tanks. A propane tank would have to be cooled below -40° to attain a negative pressure. There is no danger of collapsing one of them.

The reason for applying water on the vapor portion of the tank is that the remaining liquid in the tank usually does a pretty good job of cooling that part by its own boiling. The vapor portion has no such benefit and approaches the temperature of the flame which is what causes it to fail so cooling that portion helps the metal retain its strength.

Birken

[Standpipe]

Brannigan was asked a couple of years ago when the NFA's STICO class first came out about cooling steel in structures (the text recommends not to do that). Granted, the steel being discussed is not under pressure from a boiling liquid, but the response was general to the properties of steel. He quoted the 17th Edition of the Fire Protection Handbook, pg. 6-136 and page 259 of Building Construction for the Fire Service, 3rd edition. He said, "Steel being heated by fire should be cooled by hose streams if a safe operating position is available. If the steel is elongating, it will retract to its original length when cooled. If it has started to fail, it will be frozen in the shape it had reached. Contrary to a once prevalant myth, the water does not cause the distortion of steel." He went on to explain where that information came from. He said, "My authority way back was the American Iron and Steel Institute. IFSTA had the error in the first edition of the "Science for the Fire Service". When I sent them my letter from the Iron and Steel Institute they contacted an engineer who wrote the material on page 259 of bcfs3 and issued it as an errata in newly purchased copies of Science, etc. but did not publicize it to past purchasers."

[RFDACM]

Birken hits it on the head! Cool the vapor space first as a rule. The water stream can cool the vapor space area, but probably has very little chance of touching the area below. Remember the liquid is beginning to boil at -40 F so what is the temp at vs. what your water temp? Are you heating the boiling liquid? As for the cold water affecting the tank and causing failure? I thinks its unlikely that if the heat is great enough to degrade the tank that any water will reach the target and still be "cold". It may slowly reduce the heat and eventaully arrive cold, but instantly? Doubtful, unless you're close in with a large smoothbore gun (and suicidal).

[Dalmatian190]

I'm definitely no welding expert!

But I believe the biggest concern with "quenching" improperly or otherwise heating metal unevenly isn't with a instaneous failure.

It's with the fatique caused later causing a failure as microscopic cracks expand over time...which could cause a catastrophic failure down the road.

If you have a tank that's in BLEVE danger, cool it.

It's life expectancy is being measured in days at this point, not decades. Certainly any bulk container is going to be getting off-loaded ASAP by the clean up contractors.

Makes me wonder if anyone flares off the 20# cylinders that have been involved in a fire but didn't fail???

[doughesson]

Those would have to be liquid fuel tanks. A propane tank would have to be cooled below -40° to attain a negative pressure. There is no danger of collapsing one of them.

The reason for applying water on the vapor portion of the tank is that the remaining liquid in the tank usually does a pretty good job of cooling that part by its own boiling. The vapor portion has no such benefit and approaches the temperature of the flame which is what causes it to fail so cooling that portion helps the metal retain its strength.

Birken

From the reading I've done,and do not have at hand,don't LP tanks on truck and in tank farms have to have vacuum breaker valves as well as pressure relief valves?I've worked on petroleum barges for a few years and they have them to make sure nothing over presses or vacuums up and tears up the hull.

[BirkenVogt]

My point is that liquid water (above 32°F) will always leave positive pressure in a LPG cylinder. It might collapse say a gasoline or diesel tank however. But I would rather have a collapsed tank than an exploded one.

If the quenching is going to be what causes final failure of the cylinder, it was already just about to fail anyway. That is a risk we have to assess and take when deciding to cool a cylinder to prevent BLEVE. But once you do, take agressive action and cool above the liquid line. This is no time for pussyfooting.

Birken

[AC1503]

Standpipe, you nailed it on the head. But more than a couple of years ago did Frank Brannigan teach that steel is a thermoplastic . His position was the same 25 years ago when I attended his "Beware the truss" class. It was true then and it is still true today.

It is too bad that new firefighters will not get to know him, but at least they can read what he taught.

IACOJ

[Catch22]

My point is that liquid water (above 32°F) will always leave positive pressure in a LPG cylinder. It might collapse say a gasoline or diesel tank however. But I would rather have a collapsed tank than an exploded one.

If the quenching is going to be what causes final failure of the cylinder, it was already just about to fail anyway. That is a risk we have to assess and take when deciding to cool a cylinder to prevent BLEVE. But once you do, take agressive action and cool above the liquid line. This is no time for pussyfooting.
Birken

I'm going to see if I can explain my thought/training process without boggling it up. Maybe I'm thinking or was trained wrong, but you'll have a better understanding of where I'm coming from in the thought process.

Basically, pressure is keeping the LPG in a liquid state inside the container, the amount is dependant on the outside ambient temp. If the tank's filled in the morning when it's 60 degrees, as the temp increases to 80 degrees it'll expand and vent off, or just increase the amount of pressure in the tank.

Throw some fire around the tank, increasing the temp around it, it's going to expand even more. If the tank weakens or the pressure relief devices can't keep up or fail, you get an explosion.

We arrive and start spraying our cooler water. If we spray it in the vapor space, we're cooling the vapor (by cooling the tank), causing it to compress, and creating a vacuum. If the vacuum relief valves (if equipped) can't keep up, we collapse the tank.

Now, if we cool the liquid, it cools it at a slower rate, first stabilizing, then reducing the pressure inside the container. The vacuum effect is slowed, the safety devices don't get overwhelmed, and we don't have a tank collapse.

Not saying it's right or wrong, just what I've been taught. Maybe you more scientic guys can explain it better, or show me where I'm thinking wrong whichever is the case