Posts Tagged ‘smoke explosion’

There has been an increasing awareness that smoke is fuel and that hot smoke overhead results in thermal insult (due to radiant heat transfer) and potential for ignition. However, the hazard presented by smoke as gas phase fuel can extend a considerable distance from the current area of fire involvement.

Reading the Fire

Print a copy of the B-SAHF Worksheet. Use the worksheet to document observed fire behavior indicators as you watch the first six minutes of the following video of an apartment fire that occurred on May 10, 2013 at the corner of Park Creek Lane and Hill Park Court in Churchville, NY. In particular, focus on fire behavior indicators that may point to changes in conditions. Don’t focus too much on the flame indicators presenting from the area involved, but pay particular attention to Building, Smoke, and Air Track indicators.

The following satellite photo and view of the Alpha/Delta Corner prior to the fire are provided to help orient you to the incident location. You can also go to Google Maps Street View and do a walk around on Sides Alpha (Hill Park Court) and Delta (Park Creek Lane) to view all four sides of the building.

The following time sequence from the video of this incident illustrates the conditions immediately prior to and during the explosion. The extremely rapid increase in heat release rate during the explosion was not sustained (a transient event) as evidenced by conditions illustrated at 06:25.

Building Factors

This building is of Type V construction with a wood truss roof system. In a large apartment building such as this, the trussloft is typically subdivided with draft stops comprised of gypsum board applied to one (or both) sides of a truss to stop rapid spread of fire within the trussloft. Draft stops should be thought of as speed bumps rather than a barrier (such as a firewall that extends through the roofline). While draft stops slow fire and smoke spread, they do not stop it completely and it is common for smoke to spread beyond the fire area despite the presence of draft stops.

The small dimension framing materials used in truss construction have a high surface to mass ratio, increasing the speed with which they can be heated and increasing pyrolysis products in the smoke when heated under ventilation limited conditions.

Note: The possible location of the draft stops is speculative as specific information regarding the construction of this building was not available at the time of this post. However, draft stops may be provided between the trussloft between units or based on the size of the trussloft without regard to the location of walls between units. Preplan inspections provide an opportunity to examine building factors that may be critical during an incident!

Smoke and Air Track Indicators

An important air track indicator in this incident was the strong wind blowing from the Alpha/Bravo Corner towards the Charlie/Delta Corner. The wind may have had some influence on ventilation in the trussloft above Exposures Bravo and Bravo 2, and definitively influenced other Smoke and Air Track indicators.

From the start of the video light colored smoke is visible at the peak of the roof above Exposure Bravo and Bravo 2, indicating that smoke had infiltrated areas of the trussloft that had not yet become involved in fire. Smoke that is light in color may be comprised of pyrolysis products and air and may be to lean or too rich to burn or it may be explosive See the video Smoke on the Firegear website for a good discussion of the characteristics of smoke (note that this video is currently undergoing validation).

The volume and color (smoke indicators), velocity and direction (air track indicators) above exposure Bravo 2 vary considerably from the start of the video until shortly before the explosion that occurred at 06:12 in the video. At 02:52 a firefighter entered Exposure Bravo 2 and a short time later at 03:47 a hoseline (dry) was stretched into this exposure and charged. It is unknown from watching the video if the firefighters on this line advanced to Floor 2 or if they took any action to change the ventilation profile (other than opening the door on Floor 1, Side Alpha). The exited after the explosion, but without haste, so it is likely that they were not on Floor 2 at the time of the explosion.

Smoke Explosion

Smoke explosion is described in a number of fire dynamics texts including Enclosure Fire Dynamics (Karlsson and Quintiere) and An Introduction to Fire Dynamics (Drysdale). However, Enclosure Fires by Swedish Fire Protection Engineer Lars-Göran Bengtsson (2001) provides the most detailed explanation of this phenomenon. Paraphrasing this explanation:

A smoke or fire gas explosion occurs when unburned pyrolysis products and flammable products of combustion accumulate and mix with air, forming a flammable mixture and introduction of a source of ignition results in a violent explosion of the pre-mixed fuel gases and air. This phenomenon generally occurs remote from the fire (as in an attached exposure) or after fire control.

In some cases, the fire serves as a source of ignition as it extends into the void or compartment containing the flammable mixture of smoke (fuel) and air.

Conditions Required for a Smoke Explosion

The risk of a smoke explosion is greatest in compartments or void spaces adjacent to, but not yet involved in fire. Infiltration of smoke through void spaces or other conduits can result in a well-mixed volume of smoke (fuel) and air. Smoke explosion creates a significant overpressure as the fuel and air are premixed and ignition results in a very large energy release. Several factors influence the violence of this type of explosion:

• The degree of confinement (more confinement results in increased overpressure)
• Mass of premixed fuel and air within the flammable range (more premixed fuel results in a larger energy release)
• How close the mixture is to a stoichiometric concentration (the closer to an ideal mixture the faster the deflagration)

Potential Smoke Explosion Indicators

It is very difficult to predict a smoke explosion. However, the following indicators point to the potential for this phenomenon to occur:

• Ventilation controlled fire (inefficient combustion producing substantial amounts of unburned pyrolysis products and flammable products of incomplete combustion)
• Relatively cool (generally less than 600^o C or 1112^o F) smoke
• Presence of void spaces, particularly if they are interconnected
• Combustible structural elements
• Infiltration of significant amounts of smoke into uninvolved compartments in the fire building or into exposures

Preventing a Smoke Explosion

As it is difficult to predict a smoke explosion, there are challenges to preventing their occurrence as well. However, general strategies would include 1) preventing smoke from accumulating in uninvolved spaces or 2) removing smoke that has accumulated remote from the fire (e.g., in attached exposures), or 3) a combination of the first two approaches.

Tactics to implement these strategies may include:

• Pressurizing uninvolved spaces with a blower to prevent infiltration of smoke. This involves use of a blower for anti-ventilation by applying pressure without creating an exhaust, similar to what is done to pressurize a highrise stairwell. It is essential to check for extension prior to implementing this tactic!.
• Horizontal ventilation of attached exposures to remove smoke, checking for extension, and then pressurization with a blower to prevent continued infiltration of smoke. If fire extension is found, pressurization without an exhaust opening must not be implemented!

Additional Resources

The following previous posts on the CFBT-US Blog may also be of interest in exploring the smoke explosion phenomena.

• Smoke Explosion or Backdraft
• Fire Gas Ignitions
• Language and Understanding: Extreme Fire Behavior
• Extreme Fire Behavior: An Organizational Scheme (Ontology)
• Gas Explosions
• Sudden Blast
• Explosions During Structural Firefighting

References

Bengtsson, L. (2001). Enclosure Fires. Retrieved May 12, 2013 from https://www.msb.se/RibData/Filer/pdf/20782.pdf .

Updated with Additional Video

On March 10, 2013 five Harrison, New Jersey firefighters were injured in an explosion while working at a commercial fire at 600-602 Frank E. Rodgers Boulevard. The fire originated in a two-story commercial building at the corner of Frank E. Rodgers Boulevard North and Davis Street and extended into Exposures Charlie and Delta, two-story residential buildings.

Figure 1. Alpha/Bravo Corner and Exposure Charlie

Image from Google Maps, click on the link to walk around using Street View.

Reading the Fire

Before watching the video (or watching it again if you have already seen it), download and print the B-SAHF Worksheet. Using the pre-fire photo (figure 1) and observations during the video, identify key B-SHAF indicators that may have pointed to potential for extreme fire behavior in this incident.

Important! Keep in mind that there is a significant difference between focusing on the B-SAHF indicators in this context and observing them on the fireground. Here you know that an explosion will occur, so we have primed the pump so you can focus (and are not distracted by other activity).

Backdraft or Smoke Explosion

While smoke explosion and backdraft are often confused, there are fairly straightforward differences between these two extreme fire behavior phenomena. A smoke explosion involves ignition of pre-mixed fuel (smoke) and air that is within its flammable range and does not require mixing with air (increased ventilation) for ignition and deflagration. A backdraft on the other hand, requires a higher concentration of fuel that requires mixing with air (increased ventilation) in order for it to ignite and deflagration to occur. While the explanation is simple, it may be considerably more difficult to differentiate these two phenomena on the fireground as both involve explosive combustion.

1. Did you observe any indicators of potential backdraft prior to the explosion?
2. Do you think that this was a backdraft?
3. What leads you to the conclusion that this was or was not a backdraft?
4. If you do not think this was a backdraft, what might have been the cause of the explosion?

For more information in Backdraft, Smoke Explosion, and other explosive phenomena on the fireground, see:

• Smoke Explosion and Backdraft
• Explosions During Structural Firefighting
• Lima, Peru: Backdraft
• Chicago Extreme Fire Behavior: Analysis of Fire Behavior Indicators
• Sudden Blast
• Gas Explosions
• Extreme Fire Behavior: An Organizing Scheme
• Backdraft at 62 Watts Street
• Fire Gas Ignitions

Back at it!

I would like to say thanks to all of you who have sent e-mail or contacted me on Facebook inquiring about the status of the CFBT-US blog. The last several years have been extremely busy at Central Whidbey Island Fire & Rescue and my focus has been almost exclusively on the fire district. However, I am renewing my commitment to developing knowledge of practical fire dynamics throughout the fire service and will endeavor to return to posting on a regular basis. In addition, I am working on a series of short (10-minute) drills on fire dynamics that will be cross posted on the CFBT Blog and the Fire Training Toolbox.

Ed Hartin, MS, EFO, MIFIreE, CFO

Video of several incidents involving explosions during structural firefighting operations have been posted to YouTube in the last several weeks. Two of these videos, one from New Chicago, IN and the other from Olathe, KS involve residential fires. The other is of a commercial fire in Wichita, KS.

When a video shows some sort of spectacular fire behavior there is generally a great deal of speculation amongst the viewers about what happened. Was it a smoke (fire gas) explosion, backdraft, flashover, or did something else happen? Such speculation is useful if placed in the framework of the conditions required for these phenomena to occur and the Building, Smoke, Air Track, Heat, and Flame (B-SAHF) indicators that provide cues of to current fire conditions and potential fire behavior.

Occasionally, what happened is fairly obvious such as flashover resulting from increased ventilation under ventilation controlled conditions. However, the phenomena and its causal factors are often much more of a puzzle.

Download and print three copies of the B-SAHF Worksheet.

Residential Fire-Olathe, KS

Limited information was posted along with this pre-arrival video of a residential fire in Olathe, KS. The home was unoccupied when the fire occurred.

Watch the thirty seconds (0:30) of the video. First, describe what you observe in terms of the Building, Smoke, Air Track, Heat, and Flame Indicators; then answer the following five standard questions (based only on what you observe during the first thirty seconds of the video)?

1. What additional information would you like to have? How could you obtain it?
2. What stage(s) of development is the fire likely to be in (incipient, growth, fully developed, or decay)?
3. What burning regime is the fire in (fuel controlled or ventilation controlled)?
4. What conditions would you expect to find inside this building?
5. How would you expect the fire to develop over the next two to three minutes

Watch remainder of the video and consider the following questions:

1. Did fire conditions progress as you anticipated?
2. What changes in the B-SAHF indicators did you observe?
3. What may have caused the explosion (consider all of the possibilities)?
4. Were there any indications that may have given warning of this change in conditions?

Residential Fire-New Chicago, IN

Companies from New Chicago and Hobart were dispatched to a reported house fire at 402 Madison in New Chicago, IN on February 17, 2012.

Watch the thirty seconds (0:30) of the video. First, describe what you observe in terms of the Building, Smoke, Air Track, Heat, and Flame Indicators; then answer the following five standard questions (based only on what you observe during the first thirty seconds of the video)?

1. What additional information would you like to have? How could you obtain it?
2. What stage(s) of development is the fire likely to be in (incipient, growth, fully developed, or decay)?
3. What burning regime is the fire in (fuel controlled or ventilation controlled)?
4. What conditions would you expect to find inside this building?
5. How would you expect the fire to develop over the next two to three minutes

Watch remainder of the video and consider the following questions:

1. Did fire conditions progress as you anticipated?
2. What changes in the B-SAHF indicators did you observe?
3. What may have caused the explosion (consider all of the possibilities)?
4. Were there any indications that may have given warning of this change in conditions?

Commercial Fire-Wichita, KS

Wichita Fire Department on scene of a working building fire in large, non-combustible commercial building. Extreme heat and fire conditions cause an unknown cylinder to explode.

Keep in mind that gas cylinders and other closed containers can result in explosions during structural firefighting operations. Unlike backdraft and smoke explosion, the only clue may be building factors related to occupancy (and this may not be a good indicator when operating at a residential fire).

Wichita Fire Department on scene of a working building fire in a large metal structure. Extreme heat and fire conditions cause an unknown cylinder to explode. If you listen close, you can hear it vent before it goes off. Concussion actually cuts out my audio for just a couple seconds. No one was injured.

Video by Sean Black Photography http://seanblackphotography.smugmug.com/

Firefighter Safety

Potential for explosions related to extreme fire behavior such as backdraft and smoke explosion may be recognized based on assessment and understanding the B-SAHF (Building, Smoke, Air Track, Heat, and Flame) indicators. Other types of explosions such as those resulting from failure of closed containers (e.g., containing liquids or gases) may be a bit more difficult as this potential is likely to be present in most types of occupancies. However, commercial and industrial occupancies present greater risks.

Recognizing that even with sound experienced judgment, there may be undetected hazards on the fireground. Managing the risk requires developing a solid knowledge base and skills and operating within sound rules of engagement such as the IAFC Rules of Engagement for Structural Firefighting. However, considering the hazards presented by rapid fire progression and potential for changes in conditions following explosive events, I would add the following:

• Base your strategies and tactics on current and anticipated fire behavior and structural stability.
• Ensure that members correctly wear complete structural firefighting clothing and SCBA when working in the hazard zone and practice good air management. Buddy check before entry!
• Crews operating on the interior should have a hoseline or be directly supported by a crew with a hoseline. If conditions deteriorate, a hoseline allows self-protection and provides a defined egress path.
• Have well practiced battle drills for tactical withdrawal and abandoning the building (depending on conditions). See Battle Drill, Battle Drill Part 2, and Battle Drill Part 3.

Next…

My next post will address the impact of a closed door on tenability during a residential fire as the ninth tactical implication identified in the UL study on the Impact of Ventilation on Fire Behavior in Legacy and Contemporary Residential Construction.

Subsequent posts will come back to the Olathe, KS and New Chicago, IN residential fires to examine potential impacts on fire behavior and explosions that resulted during these incidents.

Ed Hartin, MS, EFO, MIFIreE, CFO

My last post, Hazards Above, provided a brief overview of three incidents involving extreme fire behavior in the attic or truss loft void spaces of wood frame dwellings. This post will examine the similarities and differences between these lessons and identify several important considerations when dealing with fires occurring in or extending to void spaces. At the conclusion of Hazards Above, I posed five questions:

1. What is similar about these incidents and what is different?
2. Based on the limited information currently available, what phenomena do you think occurred in each of the cases? What leads you to this conclusion?
3. What indicators might have pointed to the potential for extreme fire behavior in each of these incidents?
4. How might building construction have influenced fire dynamics and potential for extreme fire behavior in these incidents?
5. What hazards are presented by fires in attics/truss lofts and what tactics may be safe and effective to mitigate those hazards?

Similarities and Differences

The most obvious similarities between these incidents was that the buildings were of wood frame construction, the fire involved or extended to an attic or truss loft void space, and that some type of extreme fire behavior occurred. In two of the incidents firefighters were seriously injured, while in the other firefighters escaped unharmed.

Given the limited information available from news reports and photos taken after the occurrence of the extreme fire behavior events, it is not possible to definitively identify what types of phenomena were involved in these three incidents. However, it is interesting to speculate and consider what conditions and phenomena could have been involved. It might be useful to examine each of these incidents individually and then to return to examine fire behavior indicators, construction, and hazards presented by these types of incidents.

Minneapolis, MN

In the Minneapolis incident the fire occurred in an older home with legacy construction and relatively small void spaces behind the knee walls and above the ceiling on Floor 3. The triggering event for the occurrence of extreme fire behavior is reported to be opening one of the knee walls on Floor 3. As illustrated in Figure 1, the fire appeared to transition quickly to a growth stage fire (evidenced by the dark smoke and bi-directional air track from the windows on Floor 3 Side A. However blast effects on the structure are not visible in the photo and were not reported.

Figure 1. Minneapolis MN Incident: Conditions on Side A

Note: Photo by Steve Skar

Potential Influencing Factors: While detail on this specific incident is limited, it is likely that the fire burning behind the knee wall was ventilation controlled and increased ventilation resulting from opening the void space resulted in an increase in heat release rate (HRR). Potential exists for any compartment fire that progresses beyond the incipient stage to become ventilation controlled. This is particularly true when the fire is burning in a void space.

Extreme Fire Behavior: While statements by the fire department indicate that opening the knee wall resulted in occurrence of flashover, this is only one possibility. As discussed in The Hazard of Ventilation Controlled Fires and Fuel and Ventilation, increasing ventilation to a ventilation controlled fire will result in increased HRR. Increased HRR can result in a backdraft (if sufficient concentration of gas phase fuel is present), a vent induced flashover, or simply fire gas ignition (such as rollover or a flash fire) without transition to a fully developed fire.

Harrisonburg, VA

The Harrisonburg incident involved extreme fire behavior in Exposure D (not the original fire unit). The extreme fire behavior occurred after members had opened the ceiling to check for extension. However, this may or may not have been the precipitating event. As illustrated in Figure 2, as members prepare to exit from the windows on Floor 3 , Side C, flames are visible on the exterior at the gable, but it appears that combustion is limited to the vinyl siding and soffit covering. There are no indicators of a significant fire in Exposure D at the time that the photo was taken. However, it is important to remember that this is a snapshot of conditions at one point in time from a single perspective.

Figure 2. Harrisonburg, VA Incident: Conditions on Side C

Note: Photo by Allen Litten

Potential Influencing Factors: The truss loft was likely divided between units by a 1 hour fire separation (generally constructed of gypsum board over the wood trusses). While providing a limited barrier to fire and smoke spread, it does not generally provide a complete barrier and smoke infiltration is likely. Sufficient smoke accumulation remote from the original fire location can present risk of a smoke explosion (see NIOSH Report 98-03 regarding a smoke explosion in Durango, Colorado restaurant). Alternately, fire extension into the truss loft above an exposure unit can result in ventilation controlled fire conditions, resulting in increased HRR if the void is opened (from above or below).

Extreme Fire Behavior: Smoke, air track, and flame indicators on Side C indicate that the fire in the truss loft may not have continued to develop past the initial ignition of accumulated smoke (fuel). It is possible that smoke accumulated in the truss loft above Exposure B and was ignited by subsequent extension from the fire unit. Depending on the fuel (smoke)/air mixture when flames extended into the space above Exposure B ignition could have resulted in a smoke explosion or a less violent fire gas ignition such as a flash fire.

Sandwich, MA

In the Sandwich incident, the extreme fire behavior occurred shortly after the hose team applied water to the soffit. However, this may or may not have been the precipitating event. As illustrated in Figure 3, the fire transitioned to a fully developed fire (likely due to the delay in suppression as the injured members were cared for). Blast effects on the structure are obvious.

Figure 3: Sandwich, MA: Conditions on Sides C and D

Note: Photos by Britt Crosby (http://www.capecodfd.com)

Potential Influencing Factors: The roof support system in this home appears to have been constructed of larger dimensional lumber (rather than lightweight truss construction). In addition, it is likely that the attic void spaces involved in this incident were large and complex (given the size of the dwelling and complex roof line). It appears that at least part of the home had a cathedral ceiling. Fire burning in the wood framing around the metal chimney would have allowed smoke (fuel) and hot gases to collect in the attic void in advance of fire extension.

Extreme Fire Behavior: The violence of the explosion (see blast damage to the roof on Side D in Figure 3) points to the potential for ignition of pre-mixed fuel (smoke) and air, resulting in a smoke explosion. However, it is also possible that failure of an interior ceiling (due to water or steam production from water applied through the soffit) could have increased ventilation to a ventilation controlled fire burning in the attic, resulting in a backdraft).

Fire Behavior Indicators

The information provided in news reports points to limited indication of potential for extreme fire behavior. One important question for each of us is how we can recognize this potential, even when indicators are subtle or even absent.

Important! A growth stage fire can present significant smoke and air track indicators, with increasing thickness (optical density), darkening color, and increasing velocity of smoke discharge. However, as discussed in The Hazard of Ventilation Controlled Fires, when the fire becomes ventilation controlled, indicators can diminish to the point where the fire appears to be in the incipient stage. This change in smoke and air track indicators was consistently observed during the full-scale fire tests of the influence of ventilation on fires in single-family homes conducted by UL earlier this year.

Even with an opening into another compartment or to the exterior of the building, a compartment fire can become ventilation controlled. Consider building factors including potential for fire and smoke extension into void spaces in assessing fire conditions and potential for extreme fire behavior. A ventilation controlled fire or flammable mixture of smoke and air may be present in a void space with limited indication from the exterior or even when working inside the structure.

Building Construction

Each of these incidents occurred in a wood frame structure. However, the construction in each case was somewhat different.

In Minneapolis, the house was likely balloon frame construction with full dimension lumber. As with many other structures with a “half-story”, the space under the pitched roof is framed out with knee walls to provide finished space. This design is not unique to legacy construction and may also be found with room-in-attic trusses. The void space behind the knee wall provides a significant avenue for fire spread. When involved in fire, opening this void space can quickly change fire conditions on the top floor as air reaches the (likely ventilation controlled) fire.

The incident in Harrisonburg involved a fire in a townhouse with the extreme fire behavior phenomena occurring in an exposure. While not reported, it is extremely likely that the roof support system was comprised of lightweight wood trusses. In addition, there was a reverse gable (possibly on Sides A and C) that provided an additional void. As previously indicated, the truss loft between dwelling units is typically separated by a one-hour rated draft stop. Unlike a fire wall, draft stops do not penetrate the roof and may be compromised by penetrations (after final, pre-occupancy inspection). Installed to code, draft stops slow fire spread, but may not fully stop the spread of smoke (fuel) into the truss lofts above exposures.

Firefighters in Sandwich were faced with a fire in an extremely large, wood frame dwelling. While the roof appeared to be supported by large dimensional lumber, it is likely that there were large void spaces as a result of the complex roofline. In addition, the framed out space around the metal chimney provided an avenue for fire and smoke spread from the lower level of the home to the attic void space.

Hazards and Tactics

Forewarned is forearmed! Awareness of the potential for rapid fire development when opening void spaces is critical. Given this threat, do not open the void unless you have a hoseline in hand (not just nearby).

Indirect attack can be an effective tactic for fires in void spaces. This can be accomplished by making a limited opening and applying water from a combination nozzle or using a piercing nozzle (which further limits introduction of air into the void).

If there are hot gases overhead, cool them before pulling the ceiling or opening walls when fire may be in void spaces. Pulses of water fog not only cool the hot gases, but also act as thermal ballast; reducing the potential for ignition should flames extend from the void when it is opened.

Lastly, react immediately and appropriately when faced with worsening fire conditions. Review my previous posts on Battle Drill (Part 1, Part 2, and Part 3). An immediate tactical withdrawal under the protection of a hoseline is generally safer than emergency window egress (particularly when ladders have not yet been placed to the window).

Ed Hartin, MS, EFO, MIFireE, CFO

Finally! It has been quite some time since my last post, but the CFBT-US web site and blog have been attacked twice by hackers WordPress and ISP upgrade issues have been a major challenge and it has taken some time to get things back to normal.

A Big Improvement, But More Work is Needed

The Fire Service in the United States saw a considerable reduction in firefighter line-of-duty deaths in 2009. However, our efforts to improve firefighter safety must persist. Recent events reinforce the need to ensure understanding of practical fire dynamics and have the ability to apply this understanding on the fireground.

Three recent incidents involving extreme fire behavior present an opportunity to examine and reflect on the hazards presented by fires and accumulation of excess pyrolizate and unburned products of combustion in attics and other void spaces.

Minneapolis, MN Residential Fire

At 1130 hours on Saturday, July 3, 2010 Minneapolis firefighters responded to a residential fire at 1082 17th Avenue SE. First arriving companies observed light smoke and flames showing from a two and one-half story wood-frame home. A crew opening up the kneewall on the A/D corner of Floor 3 was trapped on the third floor by rapid fire progress.

Note: Photo by Steve Skar

A department spokesperson indicated that as they opened up the walls “it flashed over on them”. News reports indicated that the blast threw Firefighter Jacob LaFerriere, across the room and that he was able to locate a window, where he exited and dropped to the porch roof, one floor below. Capt. Dennis Mack was able to retreat into the stairwell where he was assisted to the exterior by other crews operating on the fireground (Mathews, 2010; Radomski & Theisen, 2010).

News reports also reported that a witness stated that the “flashover was quite loud and within seconds heavy fire was venting from the attic area” (Mathews, 2010). A later statements by department spokespersons indicated introduction of oxygen when the wall was opened resulted in the flashover (Porter, 2010) and that a burst of flames blew out the south side of the roof (Radomski & Theisen, 2010).

Firefighter Jacob LaFerriere suffered third degree burns on his arms and upper body. Capt. Dennis Mack suffered second degree burns (Radomski & Theisen, 2010) and are as of Sunday, July 4 were in satisfactory condition in the Hennepin County Medical Center Burn Unit.

Harrisonburg, VA Townhouse Fire

On June 24, 2010 Harrisonburg, Virginia firefighters responded to an apartment fire off Chestnut Ridge Drive. First arriving companies encountered a fire in a townhouse style, wood frame apartment. Investigating possible extension into Exposure Bravo, Firefighters Chad Smith and Bradly Clark observed smoke and then flames in the attic. They called for a hoseline, but when the pulled the ceiling, conditions worsened as the room ignited. Both firefighters escaped through a second floor window (head first, onto ladders placed by exterior crews). Four other firefighters were inside Exposure B when the extreme fire behavior occurred. Two received second degree burns, one was treated for heat exhaustion, and the fourth was uninjured (Firehouse.com News, 2010; WHSV, 2020). Department spokespersons indicated that a backdraft occurred when fire gases built up in the attic.

Note: Photo by Allen Litten

Sandwich MA Residential Fire

At around noon on Memorial Day, Sandwich, Massachusetts firefighters responded to a residential fire at 15 Open Trail Road. On arrival they found a 5,000 ft² (464 m²) wood frame single-family dwelling with a fire on Side C (exterior) with extension into the home. Firefighters Daniel Keane and Lee Burrill stretched a handline through the door on Side A, knocking down the fire and extending the line out onto a deck on Side C. Fire was extending through a void containing a metal chimney flue on the exterior of the building. The crew on the hoseline was making good progress until they hit the soffit with a straight stream and an explosion occurred. The force of the blast knocked the crew over the deck railing and caused significant structural damage. Firefighter Keane suffered fractures of his neck and back while Firefighter Burrill experienced a severely fractured ankle (Fraser, 2010; D LeBlanc personal communication June 2010).

Note: Photos by Britt Crosby (http://www.capecodfd.com/)

Questions

One of these fires occurred in an older home of legacy construction, the other two occurred in relatively new buildings. One was a large contemporary home, likely with an open floor plan and large attic/trussloft voids. The other two occurred in buildings with smaller void spaces in the attic/trussloft.

1. What is similar about these incidents and what is different?
2. Based on the limited information currently available, what phenomena do you think occurred in each of the cases? What leads you to this conclusion?
3. What indicators might have pointed to the potential for extreme fire behavior in each of these incidents?
4. How might building construction have influenced fire dynamics and potential for extreme fire behavior in these incidents?
5. What hazards are presented by fires in attics/trusslofts and what tactics may be safe and effective to mitigate those hazards?

Late Breaking Information

Two firefighters and an officer from the Wharton Fire Department were trapped by rapid fire progress in a commercial fire at the Maxim Production Company in Boling, TX on July 3, 2010. The crew had advanced a hoseline into the 35,000 ft² (3252 m²) egg processing plant to cut off fire extension when they encountered rapidly worsening fire conditions. The two firefighters were able to escape, but Captain Thomas Araguz III was trapped and killed (Statter, D., 2010). More information will be provided on this incident as it becomes available.

References

Mathews, P. (2010). Two Minn. ffs burned in flashover. Retrieved July 4, 2010 from http://www.firehouse.com/news/top-headlines/two-minneapolis-firefighters-burned-flashover

Radomski, L & Theisen, S. (2010). Firefighters hospitalized after flashover identified. Retrieved July 4, 2010 from http://kstp.com/news/stories/S1637495.shtml?cat=1

Porter, K. (2010). 2 firefighters burned in Mpls. fire ID’d. Retrieved July 5, 2010 from http://www.kare11.com/news/news_article.aspx?storyid=856556&catid=396

WHSV. (2010) Harrisonburg firefighters talk about their close call. Retrieved July 5, 2010 from http://www.whsv.com/home/headlines/97127924.html

Firehouse.com News. (2010). Harrisonburg, Va. firefighters forced to bail out. Retrieved July 5, 2010 from http://www.firehouse.com/showcase/photostory/harrisburg-va-firefighters-have-bail-out

Fraser, D. (2010). Mass. firefighters thrown more than 30 Ft. by blast. Retrieved July 5, 2010 from http://www.firehouse.com/news/top-headlines/blast-throws-mass-firefighters-more-30-feet

Statter, D. (2010). Update: Captain Thomas Araguz III killed during 4-alarm fire at egg plant in Boling, Texas. http://statter911.com/2010/07/04/firefighter-killed-during-4-alarm-fire-at-egg-plant-details-from-wharton-county-texas/

Quick Review

The previous post in this series presented a video clip of an incident on the afternoon of February 18, 2010 that injured four Chicago firefighters during operations at a residential fire at 4855 S. Paulina Street.

First arriving companies discovered a fire in the basement of a 1-1/2 story, wood frame, single family dwelling and initiated fire attack and horizontal ventilation of the floors above the fire. Based on news accounts, the company assigned to fire attack was in the stairwell and another firefighter was performing horizontal ventilation of the floors above the fire on Side C when a backdraft or smoke explosion occurred. Two firefighters on the interior, on at the doorway and the firefighter on the ladder on Side C were injured and were transported to local hospitals for burns and possible airway injuries.

In analyzing the video clip shot from inside a nearby building, we have several advantages over the firefighters involved in this incident.

Time: We are not under pressure to make a decision or take action.

Reduced Cognitive Workload: Unlike the firefighters who needed to not only read the fire, but also to attend to their assigned tactics and tasks, our only focus is analysis of the fire behavior indicators to determine what (if any) clues to the potential for extreme fire behavior may have been present.

Repetition: Real life does not have time outs or instant replay. However, our analysis of the video can take advantage of our ability to pause, and replay key segments, or the entire clip as necessary.

Perspective: Since the field of view in the video clip is limited by the window and the fidelity of the recording is less than that seen in real life, it presents a considerably different field of view than that of the firefighters observed in operation and does not allow observation of fire behavior indicators and tactical operations on Sides A, B, and D.

Initial Size-Up

What B-SAHF indicators could be observed on Side C up to the point where firefighters began to force entry and ventilate the basement (approximately 02:05)?

Figure 1. Conditions at 01:57 Minutes Elapsed Time in the Video Clip

Building: The structure is a 1-1/2 story, wood frame, dwelling with a daylight basement. The apparent age of the structure makes balloon frame construction likely, and the half story on the second floor is likely to have knee walls, resulting in significant void spaces on either side and a smaller void space above the ceiling on Floor 2. One window to the left of the door on Side C appears to be covered with plywood (or similar material). Given the location of the door (and door on Side A illustrated in the previous post in this series), it is likely that the stairway to the basement is just inside the door in Side C and a stairway to Floor 2 is just inside the door on Side A.

Smoke: A moderate volume of dark gray smoke is visible from the Basement windows and windows and door on Floor 1 as well as a larger volume from above the roofline on Side B. While dark, smoke on Side C does not appear to be thick (optically dense), possibly due to limited volume and concentration while smoke above the roofline on Side B appears to be thicker. However smoke on Side C thickens as time progresses, particularly in the area of the door on Floor 1. The buoyancy of smoke is somewhat variable with low buoyancy on Side C and greater buoyancy on Side B. However, smoke from the area of the door on Floor 1 Side C intermittently has increased buoyancy.

Air Track: Smoke on Side C appears to have a faintly pulsing air track with low velocity which is masked to some extent by the effects of the wind (swirling smoke due to changes in low level wind conditions). Smoke rising above the roofline on Side B appears to be moving with slightly greater velocity (likely due to buoyancy).

Heat: The only significant heat indicators are limited velocity of smoke discharge and variations in buoyancy of smoke visible from Sides B and C. Low velocity smoke discharge and low buoyancy of the smoke on Side C points to relatively low temperatures inside the building. The greater buoyancy and velocity of smoke observed above the roofline on Side B indicates a higher temperature in the area from where this smoke is discharging (likely a basement window on Side B).

Flame: No flames are visible.

Initial Fire Behavior Prediction

Based on assessment of conditions to this point, what stage(s) of development and burning regime(s) is the fire likely to be in?

Dark smoke with a pulsing air track points to a ventilation controlled, decay stage fire.

What conditions would you expect to find inside the building?

Floors 1 and 2 are likely to be fully smoke logged (ceiling to floor) with fairly low temperature. The basement is likely to have a higher temperature, but is also likely to be fully smoke logged with limited flaming combustion.

How would you expect the fire to develop over the next few minutes?

As ventilation is increased (tactical ventilation and entry for fire control), the fire in the basement will likely remain ventilation controlled, but will return to the growth stage as the heat release rate increases. Smoke thickness and level (to floor level) along with a pulsing air track points to potential for some type of ventilation induced extreme fire behavior such as ventilation induced flashover (most likely) or backdraft (less likely). Another possibility, would be a smoke explosion; ignition of premixed gas phase fuel (smoke) and air that is within its flammable range (less likely than some type of ventilation induced extreme fire behavior)

Ongoing Assessment

What indicators could be observed while the firefighter was forcing entry and ventilating the daylight basement on Side C (02:05-02:49)?

There are few changes to the fire behavior indicators during this segment of the video. Building, Heat, and Flame indicators are essentially unchanged. Smoke above the roofline appears to lighten (at least briefly) and smoke on Side C continues to show limited buoyancy with a slightly pulsing air track at the first floor doorway.

What B-SAHF indicators can be observed at the door on Side C prior to forced entry (02:49-03:13)?

Figure 2. Conditions at 03:06 Minutes Elapsed Time in the Video Clip

Figure 3. Conditions at 03:08 Minutes Elapsed Time in the Video Clip

Building, Smoke, Heat and Flame indicators remain the same, but several more pulsations (03:05-03:13) providing a continuing, and more significant indication of ventilation controlled, decay stage fire conditions.

What indicators can be observed at the door while the firefighter attempts to remove the covering over the window adjacent to the door on Floor 1 (03:13-13:44)?

No significant change in Building, Heat, or Flame Indicators. However, smoke from the doorway has darkened considerably and there is a pronounced pulsation as the firefighter on the ladder climbs to Floor 2 (03:26). It is important to note that some of the smoke movement observed in the video clip is fire induced, but that exterior movement is also significantly influenced by wind.

What B-SHAF indicators do you observe at the window on Floor 2 prior to breaking the glass (03:44)?

Figure 4. Conditions at 03:43 Minutes Elapsed Time in the Video Clip

The window on Floor 2 is intact and appears to be tight as there is no smoke visible on the exterior. It is difficult to tell due to the angle from which the video was shot (and reflection from daylight), but it would be likely that the firefighter on the ladder could observe condensed pyrolizate on the window and smoke logging on Floor 2. It is interesting to note limited smoke discharge from the top of the door and window on Floor 1 in the brief period immediately prior to breaking the window on Floor 2.

What indicators are observed at the window on Floor 2 immediately after breaking the glass (03:44-03:55)?

Figure 5. Conditions at 03:52 Minutes Elapsed Time in the Video Clip

No significant changes in Building, Heat, or Flame indicators. Dark gray smoke with no buoyancy issues from the window on Floor 2 with low to moderate velocity immediately after the window is broken.

What B-SAHF indicators were present after the ventilation of the window on Floor 2 Side C was completed and 04:08 in the video clip (03:44-04:08)?

Buoyancy and velocity both increase and a slight pulsing air track develops within approximately 10 seconds. In addition, the air track at the door on Floor 1 shifts from predominantly outward with slight pulsations to predominantly inward, but with continued pulsation (possibly due to the limited size of the window opening on Floor 2, Side C.

Anticipating Potential Fire Behavior

Unlike the firefighters in Chicago who were operating at this incident, we can hit the pause button and consider the indicators observed to this point. Think about what fire behavior indicators are present (and also consider those that are not!).

Initial observations indicated a ventilation controlled decay stage fire and predicted fire behavior is an increase in heat release rate with potential for some type of extreme fire behavior. Possibilities include ventilation induced flashover (most likely) or backdraft (less likely), or smoke explosion (less likely than some type of ventilation induced extreme fire behavior).

Take a minute to review the indicators of ventilation controlled, decay stage fires as illustrated in Table 1.

Table 1. Key Fire Behavior Indicators-Ventilation Controlled, Decay Stage Fires

Which of these indicators were present on Side C of 4855 S. Paulina Street?

Building: The building appeared to be unremarkable, a typical single family dwelling. However, most residential structures have more than enough of a fuel load to develop the conditions necessary for a variety of extreme fire behavior phenomena.

Smoke: The dark smoke with increasing thickness (optical density) is a reasonably good indicator of ventilation controlled conditions (particularly when combined with air track indicators). Lack of buoyancy indicated fairly low temperature smoke, which could be an indicator of incipient or decay stage conditions or simply distance from the origin of the fire. However, combined with smoke color, thickness, and air track indicators, this lack of buoyancy at all levels on Side C is likely an indicator of dropping temperature under decay stage conditions. This conclusion is reinforced by the increase in buoyancy after ventilation of the window on Floor 2 (increased ventilation precipitated increased heat release rate and increasing temperature).

Air Track: Pulsing air track, while at times quite subtle and masked by swirling smoke as a result of wind, is one of the strongest indications of ventilation controlled decay stage conditions. While often associated with backdraft, this indicator may also be present prior to development of a sufficient concentration of gas phase fuel (smoke) to result in a backdraft.

Heat: Velocity of smoke discharge (air track) and buoyancy (smoke) are the only two heat indicators visible in this video clip. As discussed in conjunction with smoke indicators, low velocity and initial lack of buoyancy which increases after ventilation is indicative of ventilation controlled, decay stage conditions.

Flame: Lack of visible flame is often associated with ventilation controlled decay and backdraft conditions. However, there are a number of incidents in which flames were visible prior to occurrence of a backdraft (in another compartment within the structure). Lack of flames must be considered in conjunction with the rest of the fire behavior indicators. In this incident, lack of visible flames may be related to the stage of fire development, but more likely is a result of the location of the fire, as there is no indication that flames were present on Side C prior to the start of the video clip.

What Happened?

Firefighters had entered the building for fire attack while as illustrated in the video clip, others were ventilating windows on Side C. It is difficult to determine from the video if a window or door at the basement level on Side C was opened, but efforts were made to do so. A window on Floor 2 had been opened and firefighters were in the process of removing the covering (plywood) from a window immediately adjacent to the door on Floor 1. At 04:12, an explosion occurred, injuring two firefighters on the interior as well as the two firefighters engaged in ventilation operations on Side C.

Starting at approximately 03:59, velocity of smoke discharge from the window on Floor 2 Side C increases dramatically. At 04:08 discharge of smoke begins to form a spherical pattern as discharged from the window. This pattern becomes more pronounced as the sphere of smoke is pushed away from the window by increasing velocity of smoke discharge at 04:12, immediately prior to the explosion. Velocity of smoke discharge at the door increases between 03:59 and -4:12 as well, but as the opening is larger, this change is less noticeable. As pressure increases rapidly during the explosion a whooshing sound can be heard. After the explosion, there was no noticeable increase in fire growth.

Figure 6. Conditions at 04:08 Minutes Elapsed Time in the Video Clip

Figure 7. Conditions at 04:09 Minutes Elapsed Time in the Video Clip

Figure 8. Conditions at 04:10 Minutes Elapsed Time in the Video Clip

Figure 9. Conditions at 04:11 Minutes Elapsed Time in the Video Clip

Figure 10. Conditions at 04:12 Minutes Elapsed Time in the Video Clip

Figure 11. Conditions at 04:13 Minutes Elapsed Time in the Video Clip

Based on observation of fire behavior indicators visible in the video clip, we know that a transient extreme fire behavior event occurred while a crew was advancing a hoseline on the interior and ventilation operations were being conducted on Side C. What we don�t know is what firefighting operations were occurring on the other sides of the building or in the interior. In addition, we do not have substantive information from the fire investigation that occurred after the fire was extinguished.

The Ontology of Extreme Fire Behavior presented in an earlier post classifies these types of phenomena on the basis of outcome and conditions. As a transient and explosive event, this was likely a backdraft or smoke explosion. In that this occurred following entry and during ongoing ventilation operations, I am inclined to suspect that it was a backdraft.

Indicators visible on Side C provided a subtle warning of potential for some type of ventilation induced extreme fire behavior, but were likely not substantially different from conditions observed at many fires where extreme fire behavior did not occur.

As the title of the wildland firefighting course S133 states; Look Up, Look Down, Look Around! Anticipation of fire development and extreme fire behavior requires not only recognition of key indicators, but that these indicators be viewed from a holistic perspective. Firefighters and/or officers performing a single task or tactical assignment may only see part of the picture. It is essential that key indicators be communicated to allow a more complete picture of what is occurring and what may occur as incident operations progress.

Ed Hartin, MS, EFO, MIFireE, CFO

Updated March 7, 2010 with Longer Video Clip of this Incident

On the afternoon of February 18, 2010, firefighters in Chicago responded to a residential fire at 4855 S. Paulina Street. First arriving companies discovered a fire in the basement of a 1-1/2 story, wood frame, single family dwelling and initiated fire attack and horizontal ventilation of the floors above the fire.

Based on news accounts, the company assigned to fire attack was in the stairwell and another firefighter was performing horizontal ventilation of the floors above the fire on Side C when a backdraft or smoke explosion occurred. Three firefighters on the interior and the firefighter on the ladder on Side C were injured and were transported to local hospitals for burns and possible airway injuries.

Figure 1. Consider Key Fire Behavior Indicators

B-SAHF Indicators

Recognizing subtle fire behavior indicators during incident operations can be difficult and important indicators are often only visible from one location (other than where you are). What Building, Smoke, Heat, and Flame (B-SAHF) indicators would you anticipate seeing if potential backdraft conditions exist (or may develop as the incident progresses)? How would this differ from the indicators that conditions may present risk of a smoke explosion?

For more information on key fire behavior indicators related to ventilation controlled burning regime, decay stage fires, backdraft, and smoke explosion, see the following posts:

• Recent Extreme Fire Behavior
• Decay Stage Fires: Key Fire Behavior Indicators
• Real �Backdraft�
• Hazard of Ventilation Controlled Fires
• Smoke Explosion or Backdraft?
• Language & Understanding: Extreme Fire Behavior
• 15 Years Ago: Backdraft at 62 Watts Street
• 62 Watts Street: Modeling the Backdraft
• Extreme Fire Behavior: An Organizational Scheme (Ontology)
• Fires and Explosions
• Gas Explosions
• Gas Explosions Part 2
• Extreme Fire Behavior: An Organizational Scheme
• Fuel and Ventilation

Incident Video

A video of the incident at 4855 S. Paulina Street was recently posted on YouTube (a shorter version is posted on Firevideo.net). It appears that the video may have been shot through a window by an occupant of the D2 exposure. The title of this video is �Chicago Smoke Explosion�. After watching the video and answering the questions posed in this post, do you think that this was a backdraft or smoke explosion? Why?

One of the great assets of using video as a learning tool is the ability to stop the action and go back to review key information. Watch the video and stop the action as necessary to answer the following questions�

• Pause at 02:05. What B-SAHF indicators could be observed on Side C up to this point in the video clip?
• Pause at 02:49. What indicators could be observed while the firefighter was forcing entry and ventilating the daylight basement on Side C?
• Pause at 03:13. What B-SAHF indicators can be observed at the door on Side C prior to forced entry?
• Pause at 03:35. What indicators can be observed at the door after forcing the outer door (prior to ventilation of the window on Floor 2)?
• Pause at 03:44. What B-SHAF indicators do you observe at the window on Floor 2 prior to breaking the glass?
• Pause at 03:55. What indicators are observed at the window on Floor 2 immediately after breaking the glass?
• Pause at 04:08. What B-SAHF indicators were present after the ventilation of the window on Floor 2 Side C was completed and 04:08 in the video clip?

After answering the questions, watch the complete clip. Do you think that this was a backdraft or smoke explosion? If you thought that this was a backdraft: Did you see potential indicators? If so what were they? If not, why do you think that this was the case? If you think that this was a smoke explosion, what indications lead you to this conclusion? What indicators were present?

You may want to watch this video clip several times and give some thought to what factors were influencing the B-SAHF indicators (particularly smoke, air track, and heat). Were these indicators consistent with your perception of backdraft indicators? Is so, how? If not, what was different? What indicators may have been visible from other vantage points. Remember that the video provides a view from a single perspective (and one that is considerably different than the crews working at this incident).

The next post in this series will take a closer look at the video and key fire behavior indicators.

Ed Hartin, MS, EFO, MIFireE, CFO

Unanticipated smoke explosion and building collapse nearly kills three firefighters.

Portsmouth, VA Near-Miss Incident

Firefighter Eric Kirk gives a firsthand account of a near-miss incident involving a smoke explosion in the June 2009 issue of FireRescue magazine. On a December morning in 2007, firefighters in Portsmouth, Virginia responded to a fire in a church. On arrival, the building was well involved and defensive operations were initiated to protect exposures and confine the fire. Over the course of the fire, smoke extended into an attached, three-story, brick building and formed a flammable fuel/air mixture. Subsequent extension of flames from the church to the interior of the exposure resulted in ignition and explosive combustion of this fuel (smoke)/air mixture.

Incident Photos from PilotOnline.com

Smoke Explosion

This post expands on Smoke is Fuel (Hartin, 2009), a sidebar that I wrote for FireRescue that accompanies Eric’s article examining the Portsmouth, VA smoke explosion incident.

Smoke explosions have resulted in three firefighter fatalities in the United States since 2005, two in Wyoming (see NIOSH Report F2005-13) and one last year in Los Angeles California (NIOSH report pending). In addition, there have been a number of near miss incidents including this one in Virginia and another in Durango, Colorado (see NIOSH Report F2008-02)However, many firefighters have not heard of or misunderstand this fire behavior phenomenon.

The terms backdraft and smoke explosion have typically been used to describe explosions resulting results from confined and rapid combustion of pyrolysis and unburnt products of incomplete combustion. Describing a backdraft incident at a Chatham, England Mattress Store in 1975, Croft (1980) states “this is not an entirely new phenomenon, the first formal description of what have been called ‘smoke explosions’ having been given in 1914” (p. 3).

As an explanation of many contradictory statements in reference to explosions that are reported to have occurred in burning buildings, where it is also testified that explosives were non-existent, we may cite so-called “smoke explosions.”

Distinct from, yet closely allied with explosions of inflammable dust, are explosions caused by the ignition of mixtures of air with the minute particles of unconsumed carbon and invisible gaseous matter in smoke from the imperfect combustion of organic substances…

These “smoke explosions” frequently occur in burning buildings and are commonly termed “back draughts” or “hot air explosions” (Steward, 1914).

As discussed in my earlier post, Fires and Explosions, the term Smoke Explosion was a synonym for Backdraft. In fact, if you look up the definition of smoke explosion in the National Fire Protection Association (NFPA) 921 (2008) Guide for Fire and Explosion Investigation, it says “see backdraft” (p. 921-15). However, today it identifies a different, and in many respects more dangerous extreme fire behavior phenomenon. Smoke (or Fire Gas) Explosion is described in fire dynamics textbooks such as Enclosure Fire Dynamics (Karlsson and Quintiere) and An Introduction to Fire Dynamics (Drysdale) and Enclosure Fires (Bengtsson). Of these, the text Enclosure Fires by Swedish Fire Protection Engineer Lars-G�ran Bengtsson provides the best explanation of how conditions for a smoke explosion develop. However, this phenomenon is less well known among firefighters and fire officers. In fact many well known fire service authors continue to use backdraft and smoke explosion interchangeably.

A smoke or fire gas explosion occurs when unburned pyrolysis products accumulate and mix with air, forming a flammable mixture and introduction of a source of ignition results in a violent explosion of the pre-mixed fuel gases and air. This phenomenon generally occurs remote from the fire (as in an attached exposure) or after fire control.

Conditions Required for a Smoke Explosion

The risk of a smoke explosion is greatest in compartments or void spaces adjacent to, but not yet involved in fire. Infiltration of smoke through void spaces or other conduits can result in a well mixed volume of smoke (fuel) and air within its flammable range, requiring only a source of ignition.

Smoke explosions create a significant overpressure as the fuel and air are premixed. Several factors influence the violence of this type of explosion:

• The degree of confinement (more confinement results in increased overpressure)
• Mass of premixed fuel and air in the compartment (more premixed fuel results in a larger energy release)
• How close the mixture is to a stoichiometric concentration (the closer to an ideal mixture the faster the deflagration)

For additional information on transient, explosive, fire phenomena see earlier posts: Gas Explosions and Gas Explosions Part 2.

Indicators Smoke Explosion Potential

It is very difficult to predict a smoke explosion. However, the following indicators point to the potential for this phenomenon to occur.

• Ventilation controlled fire (inefficient combustion producing substantial amounts of unburned pyrolysis products and flammable products of incomplete combustion)
• Relatively cool (generally less than 600^o C or 1112^o F) smoke
• Presence of void spaces, particularly if they are interconnected
• Combustible structural elements
• Infiltration of significant amounts of smoke into uninvolved exposures

Mitigating the Hazard

As with recognizing the potential for a smoke explosion, mitigation can also be difficult. The gases are relatively cool, so application of water into the gas layer may have limited effect. Tactical ventilation to remove the smoke is the only way to fully mitigate the hazard and establish a safe zone. However, use care not to create a source of ignition (such as the sparks created when using an abrasive blade on a rotary saw).

The best course of action is to prevent infiltration of smoke into uninvolved spaces using anti-ventilation (confinement) tactics. Anti-ventilation is the planned and systematic confinement of heat, smoke, and fire gases, and exclusion of fresh air (from the fire). In this case, anti-ventilation may involve pressurizing the uninvolved are to prevent the spread and accumulation of smoke.

Ed Hartin, MS, EFO, MIFIreE, CFO

References

Bengtsson, L. (2001). Enclosure fires. Karlstad, Sweden: R�ddnings Verket.

Croft, W. (1980) Fires involving explosions-a literature review. Fire Safety Journal, 3(1), 3-24.

Drysdale, D. (1998). An introduction to fire dynamics (2^nd ed.). New York: John Wiley & Sons.

Hartin, E. (2009, June). Smoke is fuel. FireRescue, 27(6), 54.

Karlsson, B. & Quintiere, J. (2000). Enclosure fire dynamics. Boca Raton, LA: CRC Press.

Kirk, E. (2009, June). Sudden blast: Unanticipated smoke explosion & building collapse nearly kills 3 firefighters. FireRescue, 27(6), 52-54.

National Fire Protection Association. (2008). NFPA 921 Guide for fire and explosion investigations. Quincy, MA: Author.

National Institute for Occupational Safety and Health (NIOSH). (2006) Death in the Line of Duty Report F2005-13. Retrieved June 22, 2009 from http://www.cdc.gov/niosh/fire/pdfs/face200513.pdf

National Institute for Occupational Safety and Health (NIOSH). (2009) Death in the Line of Duty Report F2008-02. Retrieved June 22, 2009 from http://www.cdc.gov/niosh/fire/pdfs/face200803.pdf

Steward, P. (1914). Dust and smoke explosions, NFPA Quarterly 7, 424-428.

My last post (Gas Explosions) examined flammability and ignition of fuel/air mixtures as related to gas explosions. Deflagration of a fuel/air mixture can result in a significant energy release, when confined, this results in a significant pressure increase.

Pressure

If a confined gas is heated, pressure will increase as indicated in Gay-Lussac’s Law.

Gay-Lussac’s Law: When the volume of a gas remains the same and temperature is increases, pressure increases in proportion to the absolute temperature of the gas.

Pressure generated in a gas explosion is dependent on the speed with which flames move through the fuel and the degree to which expanding hot gases are confined.

The speed with which flames propagate through unburned pyrolysis and flammable combustion products is subsonic (slower than the speed of sound), making this a deflagration. Flame propagation in backdraft may be several meters per second (Guigay, G., Eliasson, J., Gojkovic, D., Bengtsson,L., & Karlsson, B., 2008). The pressure generated by this type of explosion inside a compartment or building can easily break windows (changing the ventilation profile) and in many cases can be sufficient to result in structural damage.

When pre-mixed fuel and air is ignited, it pushes unburned bas ahead of the flame, producing turbulence. Flame propagation into this turbulent, pre-mixed fuel will result in an increased rate of combustion, increasing velocity and turbulence even further. This feedback loop results in acceleration of flaming combustion and high pressure from expansion of hot gases. When this reaction is confined (e.g., ventilation is limited to a single opening such as a door or window), pressure can increase to an even greater extent.

In an explosion of unburned pyrolysis and combustion products and air, the severity of the reaction will depend on the total mass and concentration of fuel, location of the ignition point, strength of the ignition source, and extent of confinement. While it is not possible to evaluate these factors under fire conditions, understanding the variables aids in understanding the processes involved. For example, as illustrated in Figures 1 and 2, ignition that occurs inside a compartment can expel a mass of unburned fuel which may subsequently ignite (note that the opening may not be to the exterior, but may simply be to another interior compartment, stairwell, etc.).

Figure 1. Influence of Ignition Location

Expulsion of unburned gas phase fuel from a compartment in a backdraft results in a characteristic spherical mass of fuel (as illustrated in Figure 2) which subsequently ignites, resulting in a fireball.

Figure 2. Expulsion of Gas Phase Fuel in a Backdraft

How might ignition location have influenced the nature and duration of flaming combustion in the stairwell in the Watts Street incident discussed in 15 Years Ago: Backdraft at 62 Watts Street and 62 Watts Street: Modeling the Backdraft?

Thermal and Structural Effects of Explosions

Explosions and the resulting pressure increase occur extremely rapidly, this makes the force that is applied to structures a dynamic load. How a structure responds to this type of dynamic load depends on the magnitude of the load, design, and condition of the structure before the load was applied. In addition to the pressure generated by an explosion, movement of gas at high velocity also adds to the dynamic load imposed on the structure. Figure 4 illustrates the rapid changes in pressure resulting from an explosion in a compartment.

Figure 4. Explosion Time-Pressure Curve

Note: Adapted from the Gas Explosion Handbook (GexCon, 2006).

Backdraft and smoke explosion can generate considerably more pressure and flow than is necessary to cause structural damage. Even if the pressure from an explosion is limited, it will generally be sufficient to cause failure of window glazing or damage to other building openings, resulting in a significant change in ventilation profile. When the fire is ventilation controlled, this will lead to increased heat release rate and potential for rapid transition to a fully developed fire.

Ed Hartin, MS, EFO, MIFireE, CFO

References

GexCon. (2006) Gas explosion handbook. Retreived March 20, 2009 from http://www.gexcon.com/index.php?src=handbook/GEXHBchap4.htm

Guigay, G., Eliasson, J., Gojkovic, D., Bengtsson,L., & Karlsson, B. (2008) The Use of CFD Calculations to Evaluate Fire-Fighting Tactics in a Possible Backdraft Situation. Fire Technology

Extreme fire behavior can be categorized as a step event which results in a sustained increase in heat release rate or a transient event that results in a brief increase in heat release rate. Transient events involve combustion of unburned combustion and pyrolysis products. The speed of this combustion process can vary widely depending on the concentration of fuel and oxygen, extent of mixing, confinement and a number of other factors. Transient events may simply involve rapid combustion (e.g., flash fire) or they may be explosive (e.g., backdraft, smoke explosion), resulting in a significant pressure increase within the compartment or building.

My next few posts will provide brief overview of gas explosions in general to provide a foundation for understanding explosive extreme fire behavior phenomena such as backdraft and smoke explosion.

Introduction

The Gas Explosion Handbook (GexCon, 2006) defines a gas explosion as a process where combustion of premixed gas phase fuel and an oxidizer (e.g., fuel and air) causes a rapid increase in pressure. The fuel in a gas explosion may result from release of a flammable gas normally used as a fuel or in industrial processes (e.g., methane, cyclohexane) or from accumulation of unburned pyrolysis and combustion products in a compartment fire.

An explosion involving unburned pyrolysis and combustion products in a compartment fire may occur in one of two ways: 1) air is mixed with a rich fuel/air mixture and subsequently undergoes auto or piloted ignition (backdraft), or 2) a pre-mixed, flammable, fuel/air mixture undergoes piloted ignition (smoke explosion). Exploring the basic processes involved in a gas explosion will lay a foundation for understanding these two important extreme fire behavior phenomena.

Flammable Fuel/Air Mixtures in Compartment Fires

Compartment fires generally involve combustion of natural and synthetic organic (carbon containing) materials such as wood, paper, and plastics. In order for flaming combustion to occur, fuel must be transformed into the gas phase through vaporization or pyrolysis. Incomplete combustion of organic fuels results in production of carbon monoxide, soot, and a wide range of other products of combustion (many of which are flammable). Smoke is comprised of not only the products of incomplete combustion, but also unburned pyrolysis products. As illustrated in Figure 1, Smoke is Fuel!

Figure 1. Smoke is Fuel

Gas phase fuel in smoke may ignite and burn in the plume or ceiling jet or it may burn as it exits through a ventilation opening. However, unburned gas phase fuel may also accumulate inside the compartment or building, mixing with air to form a potentially flammable mixture. In ventilation controlled fires, concentration of gas phase fuel increases and may become too rich to burn without introduction of additional air. In addition, flammable products of combustion and pyrolysis products may infiltrate into uninvolved compartments or attached exposures and mix with air to form a flammable atmosphere.

Review of Flammability

Combustion requires fuel and oxygen in the proper concentration. Under normal conditions, air contains approximately 21% oxygen and 79% nitrogen and trace amounts of other gases. The nitrogen, other gases, and water vapor are passive agents as they are not chemically part of the combustion reaction (but as energy is required to raise the temperature of passive agents, they do influence combustion).

Figure 1. Methane Flammability Diagram

At first glance, the flammability diagram in Figure 2 appears to be extremely complex. However, it simply represents the relationship between fuel, oxygen, and passive agents. In this triangular diagram, the total of the concentration of fuel, oxygen, and passive agents equals 100%. In the case of Figure 2, the triangular diagram shows all possible mixtures of methane, oxygen, and nitrogen passive agents (predominantly nitrogen in the air). The blue (air) line indicates oxygen concentration from normal 21% (by volume) to 0%. The red (stoichiometric) line indicates the ideal mixture of oxygen and fuel for complete combustion. The gray shaded area indicates the mixtures of methane, oxygen, and passive agents that will be flammable.

The area of this diagram that is of greatest interest in most compartment fires is the region to the right of the Air Line (flammable limits under normal conditions and the minimum oxygen concentration that will allow combustion). This is because most compartment fires are dependent on ambient air as a source of oxygen. If the concentration of fuel increases, it must be offset by a corresponding reduction in oxygen and passive agents and may make the mixture too rich to burn. However, if fuel escapes and is replaced with air, the concentration of the mixing gases may reenter the flammable range.

The flammability diagram applies to pre-mixed fuel and air (oxygen and passive agents). In a compartment fire, the concentration and mixing of fuel and air varies considerably due to differences in temperature and resulting density of smoke and air. Consequently, there may be pockets of fuel and air that are within the flammable envelope, while other areas may be too rich or lean to burn.

Ignition of Fuel/Air Mixtures

If smoke is flammable, why doesn’t it always ignite and burn? Ignition is dependent on having sufficient fuel and oxygen as well as an adequate ignition source. Ignition of a mixture of pre-mixed air and fuel requires an ignition source with sufficient strength. The minimum amount of energy required to initiate combustion is the minimum ignition energy. Factors that affect the minimum ignition energy include:

• Type of fuel
• Mixture of fuel and air
• Temperature
• Total energy supplied
• Rate at which energy is supplied (energy per unit time)
• Area over which energy is delivered

The minimum ignition energy for a given fuel generally corresponds to the stoichiometric (ideal) mixture of fuel and air. As concentration increases or decreases within the flammable range, ignition energy increases (i.e., ignition energy at the Lower and Upper Flammable Limits will be higher than for the stoichiometric concentration).

The concentration and specific gas species of flammable combustion and pyrolysis products is complex and will influence the energy required for ignition. The concept of ignition energy and the influence of concentration of fuel in air is important in understanding why a flammable mixture of combustion and pyrolysis products may not be ignited by surface combustion, but may be ignited by the higher energy provided by flames.

More to Follow

My next post will continue with a look at other factors that influence explosive ombustion in compartment fires.

Ed Hartin, MS, EFO, MIFireE, CFO

References

GexCon. (2006) Gas explosion handbook. Retreived March 20, 2009 from http://www.gexcon.com/index.php?src=handbook/GEXHBchap4.htm