Discussion Central

A BLEVE is an explosion as the consequence of the catastrophic rupture of a pressure vessel containing a liquefied gas. The catastrophic rupture of the vessel in a BLEVE is caused by the sudden isentropic drop of pressure from an initial state of equilibrium to a state, normally at atmospheric ambient conditions, that generates enough superheat to initiate a kind of boiling throughout the bulk liquid called homogenous nucleation.

The temperature at atmospheric pressure above which homogenous nucleation takes place is called the superheat limit temperature. If the temperature in the vessel is above this value, then homogeneous nucleation, a BLEVE, will occur when the contents are isentropically reduced to atmospheric conditions. 

(NOTE: this is the primary theory behind BLEVEs, but there is a lot of research into a phenomenon called a Cold BLEVE, with is not dependent upon the concept of a superheat limit temperature.) 

In homogenous nucleation, vaporization occurs simultaneously and in microseconds throughout the entire bulk liquid mass. This causes an immediate and explosive failure of the vessel. The contents exit at sonic velocity. 

If the gas is flammable and an ignition source is available, a fireball will ensue, with the flame front chasing the sonic fluid front as it expands from the vessel. 

The stereotypical BLEVE occurs when a flame impinges on the vapor space of an LPG tank (say, the outlet of a relief valve ignites…) creating a weakness and a temperature rise above the superheat limit temperature. The metal weakens and fails, leading to enough of a pressure drop to initiate homogeneous nucleation, and the BLEVE occurs.

An external fire is not necessary. A CSB investigation of the BLEVE of a reboiler at the Williams Geismer facility in Louisiana found the initial crack and pressure release was due to thermal expansion.

=-=-=-=

Chemical Safety Board Website. Williams Olefins Plant Explosion and Fire. https://www.csb.gov/williams-olefins-plant-explosion-and-fire-/

Birk. AM, et al., Hot and Cold BLEVEs: Observation and Discussion of Two Different Kinds of BLEVEs, AIChE Symposium Series, vol. 8, 1993 .

Mengmeng, Xie, Thermodynamic and Gas Dynamic Aspects of a BLEVE, Delft University of Technology, The Netherlands, October 1, 2007.

Mannan, S, Lees' Loss Prevention in the Process Industries, Section 17.29 Boiling Liquid Expanding Vapor Explosions, 4th ed., 2012.

Center for Chemical Process Safety (CCPS). Guidelines for Vapor Cloud Explosion, Pressure Vessel Burst, BLEVE and Flash Fire Hazards, 2nd ed.; John Wiley & Sons, Inc.: Hoboken, New Jersey, 2010.

------------------------------
Steve Cutchen
Investigator, retired
US Chemical Safety Board
Houston TX
------------------------------

Original Message:
Sent: 09-04-2020 10:36
From: Marlon Guerra Mutis
Subject: Open discussion: BLEVE vs. BOILOVER

Hi, 

Some experienced / technical thoughts / references about differences between BLEVE and BOILOVER, and conditions to define the classification of such events are welcomed.

Regards;

--

Marlon H. Guerra Mutis

Here's a second explanation of a BLEVE, which may or may not be easier to understand... Basic references would be the same.
=-=-=-=

The issue is that normal vaporization of the contents through a crack, even one increasing in size, would not likely lead to the BLEVE signature of flattening the shell. What makes a BLEVE so destructive is how fast the conversion from liquid to vapor occurs. Normal vaporization requires heat input to generate the heat of vaporization, because as the liquid flashes, it cools. 

At Channelview, we tried flaring a tank of methylamine to the flare. It went OK for about 15 minutes, then the tank got so cold that the MA quit flashing. And the outside of the tank iced up from humidity, insulating the tank and stopping even modest heat transfer from ambient conditions. We had to put steam hoses on the outside to finish deinventorying the tank.

So here's a description of a vessel failure. A crack occurs and material escapes the vessel. In the classic case, this crack is from weakened metal due to fire. In the Williams case, it was thermal expansion. This drops the pressure inside the vessel, and will cause one of two events. If the level of superheat at the new pressure is not too large, the liquid will continue to expand through the crack as a jet. If it is flammable, a jet fire could occur, especially if fire was the reason for the initial weakening. There's your ignition source. Vaporization is along the lines of normal boiling, and if it is not too forceful, the crack, even if it grows, remains fairly stable. The liquid autorefrigerates, actually reducing the internal pressure and slowing the jet flow.

In a BLEVE, the initial crack results in a pressure inside the vessel that generates significant superheat. If the superheat is above a critical level, an entirely different type of vaporization occurs, a homogeneous boiling that occurs explosively. It is extremely rapid, on the order of milliseconds, and the material cannot exit the crack fast enough to prevent a massive pressure rise inside the vessel. The vessel fails catastrophically; effectively instantaneously. This is why it ends up flattened. A BLEVE can occur even with a non-flammable liquid, but the common case is the fire-induced crack on an LPG tank, followed by the BLEVE, followed by the fire igniting the resulting vapor cloud (expanding at shock wave speed). The flame front does not even keep up with the vapor expansion, and in slow motion, you can see liquid leaving the vessel as flames climb up through the vapor cloud, catching up as momentum of the cloud falls.

------------------------------
Steve Cutchen
Investigator, retired
US Chemical Safety Board
Houston TX
------------------------------

Original Message:
Sent: 09-04-2020 10:36
From: Marlon Guerra Mutis
Subject: Open discussion: BLEVE vs. BOILOVER

Hi, 

Some experienced / technical thoughts / references about differences between BLEVE and BOILOVER, and conditions to define the classification of such events are welcomed.

Regards;

--

Marlon H. Guerra Mutis

Very good info Steve!

Side Story time.....I am no expert in a BLEVE, but it brings me back to my university days.​  We had a professor from MIT that taught us advanced thermodynamics (graduate level course) at a Canadian University.  A bunch of us undergrads took it because we really liked the professor.  He did a lot of work around fire propagation especially on US Navy ships etc.  He also was a BLEVE expert as we found out later in the course.  He did a lot of research using high speed filming of purposely induced BLEVEs.  He would buy empty 20lb propane BBQ tanks and install thermocouples in them.  He would get them filled with propane (which of course was illegal since you modified the tank).  He would take them to a gravel pit where he had a steel cage built.   They were placed on a fire to "cook for awhile."  He had a friend in the RCMP (Royal Canadian Mounted Police) that was a trained snipper.  The snipper would shoot a high powered riffle to induce a BLEVE event.  Now this was like 40-45 years ago and the high speed cameras were the old "film type."  We spent a whole week watching and analyzing these high speed videos in class.  It was really awesome to take the thermodynamics that he was teaching us and applying it to what we were watching.  We did the pressure wave calculation to understand the amount of energy that the pressure wall was capable of.  I remember one video he showed us from a news crew that watched a real railcar BLEVE occur.  The camera man was like a mile or two away or something like that.  After we watched it, we were all talking about what we just seen....then the professor told us that the cameraman was killed in the pressure wave.  His point was that it isn't the fire ball that's the killer....it was the pressure wave that hits you like a brick wall.  Just like you see in those old nuclear videos.  And just like we have recently seen in the Beirut event.  Thanks for bringing me back to my old university days.

Have a safe day,

------------------------------
John Howell PE
Senior Process Engineer
Stantec
Fredericton NB
Canada
------------------------------

Original Message:
Sent: 09-05-2020 00:34
From: Steve Cutchen
Subject: Open discussion: BLEVE vs. BOILOVER

Here's a second explanation of a BLEVE, which may or may not be easier to understand... Basic references would be the same.
=-=-=-=

The issue is that normal vaporization of the contents through a crack, even one increasing in size, would not likely lead to the BLEVE signature of flattening the shell. What makes a BLEVE so destructive is how fast the conversion from liquid to vapor occurs. Normal vaporization requires heat input to generate the heat of vaporization, because as the liquid flashes, it cools. 

At Channelview, we tried flaring a tank of methylamine to the flare. It went OK for about 15 minutes, then the tank got so cold that the MA quit flashing. And the outside of the tank iced up from humidity, insulating the tank and stopping even modest heat transfer from ambient conditions. We had to put steam hoses on the outside to finish deinventorying the tank.

So here's a description of a vessel failure. A crack occurs and material escapes the vessel. In the classic case, this crack is from weakened metal due to fire. In the Williams case, it was thermal expansion. This drops the pressure inside the vessel, and will cause one of two events. If the level of superheat at the new pressure is not too large, the liquid will continue to expand through the crack as a jet. If it is flammable, a jet fire could occur, especially if fire was the reason for the initial weakening. There's your ignition source. Vaporization is along the lines of normal boiling, and if it is not too forceful, the crack, even if it grows, remains fairly stable. The liquid autorefrigerates, actually reducing the internal pressure and slowing the jet flow.

In a BLEVE, the initial crack results in a pressure inside the vessel that generates significant superheat. If the superheat is above a critical level, an entirely different type of vaporization occurs, a homogeneous boiling that occurs explosively. It is extremely rapid, on the order of milliseconds, and the material cannot exit the crack fast enough to prevent a massive pressure rise inside the vessel. The vessel fails catastrophically; effectively instantaneously. This is why it ends up flattened. A BLEVE can occur even with a non-flammable liquid, but the common case is the fire-induced crack on an LPG tank, followed by the BLEVE, followed by the fire igniting the resulting vapor cloud (expanding at shock wave speed). The flame front does not even keep up with the vapor expansion, and in slow motion, you can see liquid leaving the vessel as flames climb up through the vapor cloud, catching up as momentum of the cloud falls.

------------------------------
Steve Cutchen
Investigator, retired
US Chemical Safety Board
Houston TX

Original Message:
Sent: 09-04-2020 10:36
From: Marlon Guerra Mutis
Subject: Open discussion: BLEVE vs. BOILOVER

Hi, 

Some experienced / technical thoughts / references about differences between BLEVE and BOILOVER, and conditions to define the classification of such events are welcomed.

Regards;

--

Marlon H. Guerra Mutis

Steve Cutchen has some good insights on BLEVEs.

A boilover, in contrast, typically occurs when heat is deliberately added to a boiling system at liquid-vapor equilibrium, and liquid foaming material is entrained in a channel intended for vapor only.  A typical example would be in a distillation column reboiler, where foam is forced into fractionation plates, potentially plugging up bubble caps, or even expelling non-volatile constituents into the condenser, hence contaminating the distillation product.  Another example would be in a brewery, where excess foam can eject malt sugar and hops out of the boiling kettle, thus making a huge mess on the brewery floor, and also throwing off the bitterness and sugar content of the beer.  Either of these events can have serious safety consequences.

A boilover can be prevented in real time by monitoring the level of foam in a system and cutting back on the heat if it gets too high.  Depending on the chemistry of the boiling system, it may be possible to add some anti-foaming agents to the system.

It would also be possible to have a hybrid BLEVE-boilover event, where heat input is too high relative to the rate at which vapor can escape.  Pressure can build until a structural failure occurs.  This would typically occur in a system that was never intended to contain pressure.

------------------------------
Carl Townsend
Engineering Fellow
Raytheon
Los Angeles CA
------------------------------

Original Message:
Sent: 09-04-2020 10:36
From: Marlon Guerra Mutis
Subject: Open discussion: BLEVE vs. BOILOVER

Hi, 

Some experienced / technical thoughts / references about differences between BLEVE and BOILOVER, and conditions to define the classification of such events are welcomed.

Regards;

--

Marlon H. Guerra Mutis

Marlon,

 

I believe when in context with BLEVE's, a "Boil-Over" is going to be more in line with the NFPA 30 definition.  I would recommend the NFPA 30 Handbook for an initial reference on the topic. 

 

NFPA 30 3.3.6 defines boil-over as "An event in the burning of certain oils in an open-top tank when, after a long period of quiescent burning, there is a sudden increase in fire intensity associated with expulsion of burning oil from the tank."

 

The appendix, A.3.3.6, has further clarification.  For certain mixtures of low boiling and high boiling fluids that have some water present a hazardous phenomenon can occur, During a fire on an open top tank, hot residue accumulating on the top liquid surface of a tank can become dense, sink, and cause rapid boiling of water underneath.  The hot residue mixing with the water causes rapid or nearly explosive vaporization of water that causes the contents to boil over.  

 

The NFPA 30 Handbook (a good reference) has additional explanation that "…the expulsion of as much as half of the tank contents, a sudden spread of burning oil over a wide area, and an overhead fireball…This eruption of a burning oil is a boil-over and is the reason for the stricter tank separation distances."

------------------------------
Robert Cowden PE
Process Dept. Manager
QualEx
Mount Vernon IN
------------------------------

Original Message:
Sent: 09-04-2020 10:36
From: Marlon Guerra Mutis
Subject: Open discussion: BLEVE vs. BOILOVER

Hi, 

Some experienced / technical thoughts / references about differences between BLEVE and BOILOVER, and conditions to define the classification of such events are welcomed.

Regards;

--

Marlon H. Guerra Mutis

Just an FYI, the recently released 2020 Edition of the DOT's Emergency Response Guide (ERG) has an updated BLEVE section and now includes an explanation of how a BLEVE is different than a HIT (Heat Induced Tear).  Much welcomed, as these days seems everyone wants to claim every major explosion is a BLEVE.  Heck, some news outlets and "experts" have declared the Beruit AN detonation as a BLEVE, much as they did after the West, TX AN accident!  The updated ERG makes no reference to "Boil Overs" as these do not normally occur in transportation containers (albeit it could under the right conditions with a flammable liquid container).

https://www.phmsa.dot.gov/sites/phmsa.dot.gov/files/2020-08/ERG2020-WEB.pdf

------------------------------
Bryan Haywood
Founder & CEO
SAFTENG
Milford OH
------------------------------

Original Message:
Sent: 09-04-2020 10:36
From: Marlon Guerra Mutis
Subject: Open discussion: BLEVE vs. BOILOVER

Hi, 

Some experienced / technical thoughts / references about differences between BLEVE and BOILOVER, and conditions to define the classification of such events are welcomed.

Regards;

--

Marlon H. Guerra Mutis