Posts Tagged ‘practical fire dynamics’

At a formal dinner on 23 January 2010, Chief Ed Hartin was recognized as an honorary member of Company 1 “Germania” of the Valdivia, Chile Fire Department. In addition, he was awarded a commendation for supporting the ongoing professional development of the members of Company 1 “Germania” of the Valdivia, Chile Fire Department and encouraging them in their efforts to share their knowledge with Chile’s fire service.

Commendation for Support of Company 1 “Germania”

Left to Right: Teniente Juan Esteban Kunstmann, Chief Ed Hartin, Capitán Francisco Silva V.

On 24-27 January 2007, the Company 1 “Germania” of the Valdivia, Chile Fire Department hosted the first international fire service congress to be held in South America. Participants included over 150 firefighters and officers from Chile, Peru, Argentina, and the United States. The congress provided an opportunity to participate in both classroom and hands-on workshops on a wide range of fire service topics including fire behavior, ventilation, search, rapid intervention, technical rescue, and extrication. While topical areas were diverse, the congress had a substantive emphasis on compartment fire behavior with lectures presented by CFBT-US Chief Instructor Ed Hartin and Geraldo Crespo of Contraincendio in Buenos Aires, Argentina and practical training sessions conducted by Ed Hartin and Juan Esteban Kunstmann of the Valdivia Company 1 “Germania”.

Lecture Presentation

Lecture presentations by CFBT-US Chief Instructor Ed Hartin included (click on the links for a copies of the presentations):

• Fire Development and Flashover: Foundational Knowledge
• Extreme Fire Behavior: Understanding the Hazard
• 3D Firefighting: Controlling the Fire Environment
• Reading the Fire: B-SAHF
• Modern Buildings: Fires in the Built Environment
• Large Area Fires: Tactical Implications
• Fire and Smoke in Void Spaces: Hidden Hazards
• Live Fire Training as Simulation: The Role of Fidelity

CFBT practical skills sessions were held at the Valdivia Fire Department’s training center and focused on developing basic skill in nozzle technique and understanding fire development in a compartment.

This is My Nozzle! There are many like it, but this one is mine…

Center: Ed Hartin

Practicing Nozzle Techniques

Right: Teniente Juan Esteban Kunstmann

International Collaboration

Left to Right: Battalion Chief Danny Sheridan, FDNY and Capitán Giancarlo Passalacqua Cognoro, Lima, Pe?u Cuerpo General de Bomberos Voluntarios

Congratulations to the members of Company 1 “Germania” for their success with the first Congreso Internacional Fuego y Rescate! I look forward to working with these outstanding fire service professionals in their ongoing efforts to learn and share knowledge with the fire service throughout Chile, Latin America, and the World.

Ed Hartin, MS, EFO, MIFireE, CFO

As we start the New Year it is a good time to reaffirm our commitment to mastering our craft. Developing and maintaining proficiency in reading the Fire using the B-SAHF (Building, Smoke, Air Track, Heat, and Flame) organizing scheme for fire behavior indicators, requires practice. This post provides an opportunity to exercise your skills using a video segment shot during a residential fire.

Residential Fire

Early on the morning of December 23, 2009, the Cheektowaga Police department was dispatched to 305 Highland Drive in Cheektowaga to investigate a 911 call for an unknown type problem. The female caller was screaming, but the dispatcher was unable to determine the nature of the emergency. The first arriving police unit discovered a residential fire with persons trapped, and requested fire response. The police officers rescued a male victim from just inside the door, but fire and smoke conditions prevented them from assisting the other occupants.

The Hy-View Volunteer Fire Company responded with a first alarm assignment and observed flames showing on Side C.

Download and the B-SAHF Worksheet.

Watch the first 1 minute 10 seconds (1:10) of the video. This segment was shot from Side B at the B/C Corner.  First, describe what you observe in terms of the Building, Smoke, Air Track, Heat, and Flame Indicators; then answer the following five standard questions?

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? If presented with persons reported (as the first arriving companies were) how would you assess potential for victim survival?
5. How would you expect the fire to develop over the next two to three minutes

Now watch the remainder of the video clip and answer the following questions:

1. Did fire conditions progress as you anticipated?
2. A voice heard in the video states that this was a backdraft. Do you agree? Why or why not?

Hy-View Volunteer Fire Company personnel recovered two female civilian victims from the residents. However, all three victims died as a result of smoke inhalation.

Ed Hartin, MS, EFO, MIFIreE, CFO

The ability to read the fire and predict likely fire behavior is a critical skill for both firefighters and fire officers. Previous posts have examined how to use the B-SAHF scheme to recognize critical fire behavior indicators and identify the stage of fire development, burning regime, and potential for extreme fire behavior such as flashover or backdraft. However, there is something missing!

Experience is critical to adapting standard procedures and practices to a complex and dynamic operational environment. However, learning about fire behavior and changes in fire conditions based on fireground observations are a bit like a black box test. Black box testing is a technique for testing computer software in which the internal workings of the item being tested are not known by the tester. This is not entirely true in the case of fire behavior, but there is much that we don’t know when assessing conditions on the fireground. How long has the fire been burning? What are the specific characteristics of the fuel? What sort of internal compartmentation is present? What exactly is the ventilation profile? Some of these factors can be determined during fire investigation and it is also possible to determine (with some degree of uncertainty) what influence these factors had on the outcome of the incident. Did you ever wonder how fire behavior would have changed if you had used different tactics? Unfortunately, in real life there are no “do overs”!

UL Tactical Ventilation Research Project

One of the people who has asked himself the question of what would have changed if different tactics were used is Underwriters Laboratories Fire Protection Engineer Steve Kerber.

Underwriters Laboratories (UL) has received a Firefighter Safety Research and Development Grant from the Department of Homeland Security (DHS). This research project will investigate and analyze the impact of natural horizontal ventilation on fire development and conditions in legacy (older, more highly compartmented) and contemporary (multi-level, open floor plan) residential structures.

Preliminary work has included review of literature related to horizontal ventilation and incidents in which ventilation had a significant influence on firefighter injuries and fatalities. In addition, UL has done preliminary work on the performance of various structural components such as single and multi-pane windows as preliminary input for design of full scale residential fire experiments.

In mid-December 2009, Steve Kerber met with the project advisory panel comprised of Captain Charles Bailey, Montgomery County (MD) Fire Department; Lieutenant John Ceriello New York City Fire Department, Firefighter James Dalton and Director of Training Richard Edgeworth, Chicago Fire Department, Chief Ed Hartin, Central Whidbey Island (WA) Fire & Rescue, Chief Otto Huber Loveland-Symmes (OH) Fire Department, and Chief Mark Nolan, Northbrook (IL) Fire Department. In addition, the advisory panel includes Fire Protection Engineers Dan Madrzykowski from the National Institute of Standards and Technology (NIST) and Dr. Stefan Svensson, a research and development engineer from the Swedish Civil Contingencies Agency.

Figure 1. Defining Experiment Parameters for the Contemporary Structure

The main task presented to the advisory panel at the first meeting was to aid in defining the parameters for the experiment; including fire location, changes in ventilation profile, timing of these changes, and instrumentation to measure effects on fire development and conditions.

UL Large Fire Research Facility

The ventilation experiments will be conducted at the UL Large Fire Research Facility in Northbrook, IL. From the exterior, this facility simply looks like a large industrial building (see Figure 2). However, the interior of the structure includes a unique facility for fire research.

Figure 2. UL Large Fire Research Facility

One of the facilities inside this building is a 100’ x 120’ (30.48 m x 36.58 m) with a ceiling height that is adjustable up to 50’ (15.24 m) (see Figure 3). All of the smoke resulting from tests in this facility is exhausted through a system designed to oxidize unburned fuel and scrub hazardous products from the effluent prior to discharge to the atmosphere. Tests are monitored from a control room that overlooks the large burn room.

Figure 3. Large Burn Room

Over the next month, the two residential structures to be used for the ventilation experiments will be constructed inside the large burn room at the UL Large Fire Test Facility. After construction is complete, a series of 16 full scale fire experiments is planned to evaluate a range of different horizontal ventilation scenarios.

Research with the Fire Service

Steve Kerber has often stated that it is essential that scientists and engineers conduct research with, not for, the fire service. Engagement between researchers and firefighters on the street is essential in advancement of our profession. With this ventilation research project, Underwriters Laboratories is actively engaged in this process.

The outcome of this project will not simply be an academic paper (but there might be one or more of those as well). As part of the DHS grant, UL will be developing an on-line course to present the results of the experiments and their practical application on the fireground.

Happy Holidays,

Ed Hartin, MS, EFO, MIFireE, CFO

The first post in this series, Effective and Efficient Fire Streams, discussed theoretical cooling capacity, fire stream efficiency, and flow rate. This post extends the discussion, by examining how nozzle design characteristics and methods of use influence efficiency.

Think!

I do not ask that anyone believe anything that I say (or write in this blog) simply because I said so. Firefighters� opinions about nozzles are as strong as their opinions about what color fire apparatus should be painted and what type of helmet should be used to protect our heads. Each kind and type of nozzle discussed in this post is being used by firefighters all over the world to extinguish fires in buildings. This does not mean that they are all equally effective, or appropriate in all circumstances. I challenge you to think about the physics of fire control and examine your assumptions about nozzles, fire stream characteristics, and how to develop effective and efficient fire streams.

Nozzle Classification

There are several different ways to approach classification of nozzles used in structural firefighting. One simple approach is to consider the pattern or patterns in which water can be applied:

Solid Stream/Smooth Bore: This type of nozzle provides a single pattern consisting of a jet of water that maintains coherence throughout its effective reach (breaking up into extremely large droplets beyond that point)

Combination: This type of nozzle can produce a variety of patterns from a straight stream to a fog cone. Both the straight stream and fog cone are comprised of small droplets of varying diameters. Droplet diameter and consistency of droplet size is dependent on nozzle design and operating pressure (higher pressure results in smaller droplets).

Special Purpose Nozzles: In addition to solid stream and combination nozzles, there are a variety of other specialized nozzles such as piercing nozzles (fog nails), cellar nozzles (of various types), and ultra high pressure solid stream nozzles that can be used for cutting through a variety of materials as well as to produce a fog pattern with extremely small droplets. Specialized nozzles and in particular high pressure and ultra-high pressure systems will be examined in detail in a subsequent post in this series.

Nozzle Characteristics

Beyond simple classification of structural firefighting nozzles as solid stream, combination, or special purpose, nozzles may be further classified based on a number of other characteristics such as the flow rate(s) or flow range and designed operating pressure.

Single Flow: Some nozzles are designed to provide a specific, fixed flow rate at their designed operating pressure. This includes solid stream nozzles with a single sized tip and fixed flow rate combination nozzles. While these nozzles are considered to provide a single flow rate, this is not exactly true. The nozzle orifice is of fixed size, providing a given flow rate at a specified nozzle pressure. As discussed in Under Pressure, flow rate from an opening is based on the area of the opening and the velocity of the water being discharged. Increased or decreased nozzle pressure influences flow rate. For example, increasing the nozzle pressure on a solid stream nozzle from 50 psi (345 kPa) to 80 psi (352 kPa) increases flow rate by approximately 22%.

Variable Flow: Nozzles may also be designed to allow orifice size to be changed, providing variable flow rates at a given nozzle pressure. With solid stream nozzles, this is accomplished by changing the tip size. With some combination nozzles, flow and pattern vary together (e.g., the fog pattern has a lower flow rate than the straight stream setting). However, most modern combination nozzles used for structural firefighting allow adjustment of the spray pattern while maintaining flow rate. Variable flow combination nozzles may be manually adjustable with several different flow rate settings at a specified nozzle pressure.

Automatic Nozzles: Another type of nozzle that allows variation of flow rate is the automatic nozzle. With this design the nozzle adjusts flow rate by varying orifice size automatically to maintain a relatively constant nozzle pressure. With automatic nozzles the flow range specifies the lowest and highest flow rate at the designed nozzle pressure. Some automatic nozzles allow adjustment of the nozzle pressure setting to allow operation at two different nozzle pressures such as 100 psi (690 kPa) and 50 psi (345 kPa).

Nozzle Pressure: At one time the question of nozzle pressure was fairly simple, combination nozzles generally were designed to operate at 100 psi (690 kPa) nozzle pressure. However, today it is not that simple. For a variety of reasons ranging from limited pressure available from high-rise standpipe systems to the desire for lower nozzle reaction force, nozzle manufacturers are producing combination nozzles with varied designed operating pressures (commonly 50 psi (345 kPa), 75 psi (517 kPa), and 100 psi (690 kPa).

Nozzle Performance

Floyd Nelson (1989) captured the essence of nozzle performance in the following statement: �In principle, firefighting is very simple. All one needs to do is put the right amount of water in the right place and the fire is controlled� (p. 102).

Nozzles used for gas cooling must produce small droplets, be capable of varying the angle of the fog cone to have sufficient reach to cover varied sizes of compartments. Droplets with a diameter of 0.3 mm are small enough to vaporize readily in the hot gas layer, but also have sufficient mass to travel a reasonable distance (Herterich, 1960). Droplets larger than 1 mm are likely to penetrate some distance through hot gases and flames without completely vaporizing (S�rdqvist, 2001). In reality, while we know a fair bit about droplet size and performance. We don�t know much at all about the droplet sizes produced by the nozzles we are using.

What we do know is that lower pressure nozzles develop larger droplets than higher pressure nozzles of the same general design. Specific design characteristics such as the angle that the water must take as it exits the orifice and forms the fog cone also impact on droplet size. This can be illustrated using a nozzle such as the Akron Turbojet. When set on 30 gal/min (115 l/m) or 60 gal/min and operated at a nozzle pressure of 100 psi, droplet size is extremely small, providing excellent gas cooling performance. However, when flow rate is increased to 95 gal/min (360 l/min) or 125 gal/min (473 l/min) droplet size increases dramatically. While still effective for gas cooling, water application at these flow rates is less efficient.

There is no standardized test used for nozzles that determines the range of droplet sizes produced under different flow rates, nozzle pressures, cone angles, etc. However, there is light at the end of the tunnel. The technology exists to answer this interesting (and I believe important) question. Figures 1 and 2 illustrate a system comprised of lasers and a high speed camera that is used to determine droplet size from sprinkler heads. This system could also be used to assess droplet size developed by handline nozzles (if funding was available).

Figure 1. UL Sprinkler Droplet Size Test Facility

Note: Underwriters Laboratories, Northbrook, IL

Figure 2. Laser and Camera Used for Measuring Droplet Size

Note: Underwriters Laboratories, Northbrook, IL

One factor that complicates things when considering droplet size and nozzle performance is that the nozzle is only one part of the equation. The nozzle operator has a significant influence on performance. For example, in a short pulse, if the nozzle is opened quickly, more of the droplets are formed with the nozzle operating at full nozzle pressure than if it is opened slowly (providing a lower pressure at the start of the pulse). The same is true if the nozzle is closed slowly rather than quickly. This is less significant with long pulses as the opening and closing phase of the pulse comprises a small percentage of the total operating time.

In direct and indirect attack, water must pass through the hot gas layer and reach burning fuel (direct attack) and/or hot surfaces (indirect attack) before significantly evaporating. If distances are not great or the temperature of the hot gas layer is not extremely high, a straight stream or narrow fog cone comprised of small droplets may be effective in accomplishing this task. When gas cooling precedes direct attack, this is often the case. However, if the distance between the nozzle and intended target is large and/or the temperature of the hot gases is high, larger droplets (or a solid stream) may be much more effective.

Selection

Many factors can (and should) be considered when selecting a system to develop effective and efficient fire streams. I used the word system intentionally as a nozzle is useless without hose, a pump, source of water, and most importantly knowledgeable and skilled firefighters to operate it.

The ideal system to develop effective and efficient fire streams for offensive firefighting would have the following capabilities (but not necessarily at the same time).

• Can produce small droplets for gas cooling
• Can produce larger droplets to penetrate hot gases and reach burning fuel or hot surfaces
• Adjustable fog cone angle (allows effective reach in varied size compartments)
• Adequate reach and stream cohesion when adjusted to straight stream
• Ability to vary flow rate depending on fire conditions and tactical application
• Light weight and high level of maneuverability
• Ease of operation to simplify training requirements

At present, it is unlikely that any single system meets all of these requirements (but that is open to debate). Future posts will examine a variety of systems including those that use low, medium, high, and ultra-high nozzle pressure as well as a range of flow rates.

Ed Hartin, MS, EFO, MIFireE, CFO

References

Nelson, F. (1991). Qualitative fire behavior. Ashland, MA: International Society of Fire Service Instructors.

Herterich, O. (1960). Wasser als loschmittel [in German]. Heidelberg, Germany: Alfred Huthig

S�rdqvist, S. (2001). Water and other extinguishing agents. Karlstad, Sweden: R�ddnings Verket.

It is often stated and commonly believed that it takes gpm to overcome Btu. While I suspect that firefighters understand the underlying intent of this statement, it is actually incorrect as it is comparing apples and oranges. Flow rate is expressed in terms of volume and time (gal/m or l/m). However, Btu (or Joules) is a measure of quantity (more like volume than flow rate).

You can say that it takes gallons (or liters) to overcome Btu (or Joules), But the rate at which energy is absorbed by a fire stream must overcome heat release rate (energy released/unit of time). This concept points to the need for a higher flow rate when the heat release rate from a fire is larger. This leads to another common fire service saying: “Big Fire, Big Water”. While this is not completely incorrect, it is a bit misleading as it does not account for the efficiency of the fire stream in absorbing energy. Not all of the water that leaves the nozzle absorbs the same amount of energy.

Theoretical Cooling Capacity

Water is an excellent extinguishing agent because it has a high specific heat (energy required to raise its temperature) and high latent heat of vaporization (energy required to change it from water to steam). As illustrated in Figure 1, conversion of water to steam is most significant as it absorbs 7.5 times more energy than heating water from 20^o C (68^o F) to its boiling point.

Figure 1. Theoretical Cooling Capacity

However, this only tells us the theoretical cooling capacity of a single kilogram of water at 20^o C (68^o F) if it is raised to 100^o C (212^o F) and completely vaporized. Examining theoretical cooling capacity in terms of flow rate requires a bit more work.

Flow is defined in terms of gallons per minute (gal/m) or liters per minute (l/m) and theoretical cooling capacity of water was defined in terms of energy absorbed per second per unit mass (MJ/kg) we need to work through conversion to common units of measure.

While SI units are simpler to work with, I have worked cooling capacity out in both liters per minute (LPM) and gallons per minute (GPM). However, in that specific heat and latent heat of vaporization are applied to mass rather than volume and Watts are joules per second, it is first necessary to covert flow rate into kg/s

Figure 2. Flow Rate and Theoretical Cooling Capacity

This example assumes instantaneous heat transfer and 100% efficiency in conversion of water to the gas phase. Neither of which is possible in the real world!

Factors influencing effectiveness and efficiency of heat transfer (Svennson, 2002) include:

• Diameter (in the gas layer and on surfaces)
• Temperature (in the gas layer and on surfaces)
• Velocity (in the gas layer)
• Film formation (on surfaces)
• Temperature of the gas layer
• Surface temperature

Fire Stream Efficiency

The firefighter’s power is not simply related to flow rate, but flow rate effectively applied to transfer heat from hot gases and surfaces by changing its phase from liquid (water) to gas (steam). Extinguishing a compartment fire generally involves converting a sufficient flow (gal/m or l/m) of water to steam. So while the “steam” itself does not generally extinguish the fire, the energy absorbed in turning the water to steam has the greatest impact on fire extinguishment.

Experimental data (Hadjisophocleous & Richardson, 2005; Särdqvist, S., 1996) has shown that the cooling efficiency of water in both laboratory experiments and actual firefighting operations ranges from 0.2 to 0.6. Särdqvist (1996) suggests that an efficiency factor of 0.2 be used for interior fog nozzles. Based on my personal observations (but no experimental data), I think that Särdqvist’s efficiency factor of 0.2 might be just a bit on the low side. Barnett (as cited in Grimwood,2005) suggests that an efficiency factor of 0.5 be used for solid or straight stream application and 0.75 for fog application. The following table takes a slightly more conservative approach, using 0.6 as an average efficiency factor.

Figure 3. Flow Rate and Adjusted Cooling Capacity

Figure 3 is provided to illustrate the impact of efficiency (or lack thereof) on fire stream cooling capability. The key point is that actual cooling capability is considerably less than the theoretical cooling capacity. Another complication is that in addition to nozzle performance characteristics, nozzle efficiency is also dependent on the skill of the nozzle operator, the manner in which water is applied (straight stream, narrow fog pattern, wide fog pattern), the configuration of the space, and fire conditions. Unfortunately, there is no standardized test with specified conditions that permits comparison of different nozzles and/or methods.

However, the concept of efficiency gives rise to an interesting question. Does a nozzle flowing 100 gpm with an efficiency factor of 0.6 have the same extinguishing capability as 200 gpm nozzle with an efficiency factor of 0.3. This is simple math! The cooling capacity would be identical. While the practical application is more complex (as we do not really know the efficiency factors for the two nozzles and manner in which they are being used), this is worth thinking about.

Flow Rate or Heat Absorption Capacity

CFBT-US Senior Instructor Trainer Matt Leech (LT Tualatin Valley Fire & Rescue) proposed (half in jest) that nozzles should be labeled with their potential cooling capacity rather than flow rate. While this idea did not get significant traction, it is important for firefighters to recognize that flow rate and fire stream characteristics have a significant impact on potential cooling capacity.

Fire Stream Effectiveness

Safe, effective and efficient fire control requires:

• Water application to control the fire environment as well as direct attack on the fire
• Appropriate flow rate for the tactical application (cooling hot, but unignited gases may be accomplished at a lower flow rate than direct attack on the fire)
• Direct attack to exceed the critical flow rate based on the fire’s heat release rate
• Sufficient reserve (flow rate) be available to control potential increases in heat release rate that may result from changes in ventilation
• Water application in a form appropriate to cool its intended target (i.e., small droplets to cool hot gases or to cover hot surfaces with a thin film of water)
• Water to reach its intended target (fog stream to place water into the hot gas layer and a straight or solid stream to pass through hot gases and flames and reach hot surfaces)
• Control of the fire without excessive use of water

Accomplishing this requires different stream characteristics at different times. The characteristics that are optimal for gas cooling are likely quite different than for cooling hot surfaces, particularly when dealing with fully developed fire conditions in a large compartment.

Priorities

As regular readers have likely noted my posting schedule has been a bit off of late. My responsibilities as the new Fire Chief with Central Whidbey Island Fire & Rescue preclude the necessary research and writing necessary to constantly post twice weekly. I will be scaling back to a single post on Thursday for the next few months while I get a handle on my new job and get my family moved to Whidbey Island.

Ed Hartin, MS, EFO, MIFIreE, CFO

References

Grimwood, P. (2005) Firefighting Flow Rate: Barnett (NZ) – Grimwood (UK) Formulae. Retrieved January 26, 2008 from http://www.fire-flows.com/FLOW-RATE%20202004.pdf

Hadjisophocleous, G.V. & Richardson, J.K. (2005). Water flow demands for firefighting. Fire Technology 41, p. 173-191.

Särdqvist, S. (1996) An Engineering Approach To Fire-Fighting Tactics Sweden, Lund University, Department of Fire Safety Engineering

Svennson, S. (2002). The operational problem of fire control (Report LUTVDG/TVBB-1025-SE). Sweden, Lund University, Department of Fire Safety Engineering.

This morning I begin the process of relocation to beautiful Whidbey Island, Washington. Later this week I begin my new job as Fire Chief with Central Whidbey Island Fire & Rescue.

Coupeville Harbor at Sunrise

As I have been packing and preparing for my move all weekend, I have not had time to develop an in-depth Monday morning post. However, I did run across an interesting video clip a few weeks ago that ties in well with our ongoing effort to develop skill in reading the fire.

R-Fire

On March 27, 2009 the Emerson and Red Oak Fire Departments were dispatched for a residential fire at 901 Lowell Ave in Emerson, IA. First arriving companies observed a fire on Floor 1 and smoke throughout the structure.

The following video clip appears to have been shot early in incident operations as positive pressure ventilation is being implemented.

Download the B-SAHF Worksheet to use as a reminder while watching the video clip.

As you view the video clip, what do the fire behavior indicators (particularly smoke and air track) tell you about the stage of fire development, burning regime, and effectiveness of tactical operations?

Questions

In addition to your general observations of B-SAHF indicators while you watched the video the first time, consider the following questions:

• Watch the video clip again and examine conditions at the inlet between 0:21 and 0:25. What does the presence of smoke (and particularly dark smoke) pushing from the inlet indicate?
• Continue the video and examine conditions between 1:02 and 1:10. What do you observe at this point? What do smoke and air track conditions indicate now?
• Continue on and examine conditions closely between 2:28 and 2:48. What does the variation in smoke and air track indicators at different points on the exterior of the structure tell you?

Back on Schedule!

I plan to be back on track with the next post in the series examining nozzle techniques on Thursday, November 12. I have been having an e-mail conversation BC Mike Walker of the Oklahoma City Fire Department regarding nozzle techniques and flashover. Mike is working on a research project regarding “right for reach and left for life”. Interestingly, when I received Mike’s first e-mail, I was in the process of outlining an upcoming blog post on the concept of “Battle Drills” to deal with or escape from conditions such as those resulting from unplanned changes in ventilation (window failure, wind, etc.).

Ed Hartin, MS, EFO, MIFireE, CFO

Open Enrollment CFBT Level I & Instructor Courses

CFBT-US, LLC and the Northwest Association of Fire Trainers (NAFT) will be offering CFBT Level I and Instructor Courses at the Clackamas County (OR) Fire District I CFBT facility.

CFBT Level I
7-9 November 2009
Course Fee: $335

CFBT Instructor
9-13 November 2009
Course Fee: $915

Instructor course participants receive a copy of 3D Firefighting: Training, Techniques, & Tactics and an extensive 2-DVD library of CFBT resources including the CFBT Level I curriculum. For information on these courses download a NAFT CFBT Brochure and the CFBT Level I and CFBT Instructor Course Information Sheets.

CFBT Workshop in Sandö, Sweden

From 12-16 October 2009, I will be participating in a CFBT workshop in Sandö, Sweden along with a small group of instructors from around the world. We will be studying the compartment fire behavior curriculum at the Swedish Civil Contingencies Agency (Myndigheten för samhällsskydd och beredskap (MSB)) College in Sandö.

Figure 1. Fire Behavior Training in Sandö

In January of 2009 MSB replaced the Swedish Rescue Services Agency, the Swedish Emergency Management Agency, and the Swedish National Board of Psychological Defense. The MSB maintains two fire service colleges, one in Sandö (see Figure 2) and the other in Revinge.

Figure 2. MSB College in Sandö

The International Conference of Fire and Rescue, Valdivia – Chile 2010 CIFR

My brothers with Company 1 “Germania” of the Valdivia, Chile Fire Department have taken on a tremendous task with delivery of the first International Conference of Fire & Rescue in Valdivia. The conference will be held 23-27 January 2010.

Conference presenters include a diverse cadre of instructors from around the world. I will be presenting a series of seminars on fire behavior as well as a hands-on CFBT workshop. Presentations will be simultaneously translated into English and Spanish (as applicable). Have a look at the Conference Web Site for more information on this tremendous learning opportunity.

NIOSH Death in the Line of Duty F2007-02

On November 23, 2006, Firefighter Steven Solomon, a 33-year-old career fire fighter was seriously injured during a ventilation induced flashover or related fire behavior event in an abandoned single story duplex in Atlanta, GA; he died as a result of these injuries 6 days later.

NOSH Report F2007-02 provides an excellent description of fire behavior indicators observed prior to the occurrence of extreme fire behavior and correctly identifies that increased ventilation without coordinated fire attack resulted in worsening fire conditions.

Several conclusions in the report were based on computational fluid dynamics (CFD) modeling using the National Institute of Standards and Technology (NIST) Fire Dynamics Simulator software. As discussed in a previous post computer modeling is an excellent tool, but it is important to understand both its capabilities and limitations (see Townhouse Fire-Washington, DC: Computer Modeling)

It is crucial to bear in mind that fire models do not provide a reconstruction of the reality of an event. They are simplified representation of reality that will always suffer from a certain lack of accuracy and precision. Under the condition that the user is fully aware of this status and has an extensive knowledge of the principles of the models, their functioning, their limitations and the significance attributed to their results, fire modeling becomes a very powerful tool (Dele´mont & Martin, J., 2007, p. 134).

Review NIOSH Report F2007-02 and see if you agree or disagree with the conclusions regarding the type of extreme fire behavior phenomena involved in this incident.

Ed Hartin, MS, EFO, MIFireE, CFO

Chicago Dollar Store Fire

On the morning of October 1, 2009 the Chicago Fire Department (CFD) responded to a fire in the Super Dollar and Up store at 3952 West Cermak Road. CFD Senior Fire Alarm Operator and Fire Photographer Steve Redick captured early incident operations on video.

The first segment of the video was shot in the alley on Side C from the B/C Corner. The next several minutes of video are shot from positions on Side A as indicated in Figure 1.

Figure 1. Plot Plan and Approximate Video Camera Locations

Download the B-SAHF Worksheet.

Watch the first 60 seconds of Video Segment 1. Consider the information provided in this segment of the video clip. First, describe what you observe in terms of the Building, Smoke, Air Track, Heat, and Flame Indicators and then answer the following five standard questions?

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?

After completing the B-SAHF worksheet and answering the five standard questions, watch the next minute and twenty seconds of the video.

1. Did you anticipate this change?
2. What factors may have influenced this change in conditions?

Visit Steve Redick’s Web Site for additional video and excellent photos of this incident.

Memphis Dollar Store LODD

The rapidly changing conditions in the Chicago incident reminded me of the fire in Memphis, Tennessee that took the lives of Lieutenant Trent Kirk and Private Charles Zachary. Similar to the fire in Chicago, this incident involved a fire in a one-story, non-combustible building containing multiple commercial occupancies. As companies arrived they observed a small volume of smoke from the roof and little smoke inside the building. Approximately nine minutes after arrival conditions worsened with a large volume of smoke pushing from the doorway on Side A. Crews became disoriented as a result of rapid fire progression, and Lieutenant Kirk and Private Zachary were trapped.

For additional information on this incident see NIOSH Death in the Line of Duty Report F2003-18 and Memphis Fire Department Director’s Review Board Family Dollar Store Fire report.

Dollar Stores as a Target Hazard

Dollar stores and similar types of commercial occupancies should be considered as a target hazard that presents a significant threat to firefighters. These types of stores are generally in an enclosed building (good access from the front, but not generally from the other sides of the building) with high ceilings and a cockloft or other ceiling void space. In addition, this type of store contains a large fuel load comprised predominantly of synthetic fuel with a high heat of combustion (think high energy) and potential for extremely rapid fire development.

Fires in this type of occupancy are not uncommon! A quick search uncovered 15 similar incidents across the United States in the last three years (and 11 in 2009). There were likely more (as the scope of this search looked for fires in “dollar stores” and stopped after the first several hundred hits with the Google search engine).

• Broadview, IL (June 9, 2009)
• Flint, MI (August 24, 2009)
• Lubbock, TX (September 15, 2009)
• Terre Haute, IN (June 29, 2009)
• New York, NY (June 9, 2009)
• Midlothian, IL (February 6, 2008)
• Highland Park, MI (October 7, 2007)
• Denver, CO (June 29, 2009)
• Sanford, FL (March 23, 2009)
• Chattanooga, TN (April 14, 2009)
• Conklin, NY (August 27, 2009)
• Muncie, IN (September 16, 2009)
• Lake Worth, TX (November 25, 2006)
• Omaha, NE (April 8, 2008)
• Bells Corner, PA (June 3, 2009)

Building Factors and Fire Behavior

Building factors include the construction, configuration, and contents of a structure. These factors are critical fire behavior indicators that must be assessed during pre-planning and in the course of size-up and incident operations. Consider how building size (particularly volume, ceiling height, and presence of ceiling, attic, or cockloft void spaces) impacts on both fire behavior and how the other B-SAHF indicators present.

Reporting on the Dollar Store fire in Chattanooga, TN in April 2009, a Chattanooga Fire Department spokesperson said:

At first, it appeared that the firefighters would be able to get the fire under control fairly quickly, but the fire got into the attic and was difficult to locate in the thick, black smoke… The firefighters made an interior attack and tried to use thermal imaging cameras to locate the fire. However, other firefighters noticed that the roof was beginning to sag, so the order was given to evacuate the building for the safety of the firefighters.

It is essential to recognize potential for worsening conditions and extreme fire behavior. This is particularly important when faced with an incident outside the norm of fires in residential structures such as one and two-family dwellings and apartments.

Posts from Sandö, Sweden

Next week I will be posting from Sandö, Sweden as 12-16 October I will be participating in a Compartment Fire Behavior Training Workshop at the Swedish Civil Contingencies Agency College. Along with representitives from Australia, Canada, Germany, and Spain, I will be studying contemporary approaches to fire behavior training as well as the evolution of Swedish fire behavior training since the 1980s. This workshop provides a tremendous opportunity to learn along with Mats Rosander, Nils Bergström, and Marcos Dominguez, poneers in the evolution of fire behavior training in Sweden and around the world.

Ed Hartin, MS, EFO, MIFIreE, CFO

This post continues study of an incident in a townhouse style apartment building in Washington, DC with examination of the extreme fire behavior that took the lives of Firefighters Anthony Phillips and Louis Mathews. As discussed in Townhouse Fire: Washington, DC-Computer Modeling Part I, this was one of the first cases where the NIST Fire Dynamics Simulator (FDS) software was used in forensic fire scene reconstruction (Madrzykowski and Vettori, 2000).

Quick Review

As discussed in prior posts, crews working on Floor 1 to locate the fire and secure the door to the stairwell were trapped and burned as a result of rapid progression of a fire in the basement up the open interior stairway after an exterior sliding glass door was opened to provide access to the basement. For detailed examination of incident operations and fire behavior, see:

• Fire Behavior Case Study Townhouse Fire: Washington, DC
• Townhouse Fire: Washington DC-What Happened
• Townhouse Fire: Washington DC-Extreme Fire Behavior
• Townhouse Fire: Washington DC-Computer Modeling

Figure 1. Conditions at Approximately 00:28

Note: From Report from the Reconstruction Committee: Fire at 3146 Cherry Road NE, Washington DC, May 30, 1999, p. 29 & 32. District of Columbia Fire & EMS, 2000.

Smokeview

Smokeview is a visualization program used to provide a graphical display of a FDS model simulation in the form of an animation or snapshot. Snapshots illustrate conditions in a specific plane or slice within the building. Three vertical slices are important to understanding the fire dynamics involved in the Cherry Road incident: 1) midline of the door on Floor 1, Side A, 2) midline of the Basement Door, Side C, and midline of the Basement Stairwell (see Figure 2). Imagine that the building is cut open along the slice and that you can observe the temperature, oxygen concentration, or velocity of gas movement within that plane.

Figure 2. Perspective View of 3146 Cherry Road and Location of Slices

Note: From Simulation of the Dynamics of the Fire at 3146 Cherry Road NE Washington D.C., May 30, 1999, NISTR 6510 (p. 15) by Dan Madrzykowski and Robert Vettori, 2000, Gaithersburg, MD: National Institute for Standards and Technology.

In addition to having an influence on heat release rate, the location and configuration of exhaust and inlet openings determines air track (movement of smoke and air) and the path of fire spread. In this incident, the patio door providing access to the basement at the rear acted as an inlet, providing additional air to the fire. The front door and windows on the first floor opened for ventilation served as exhaust openings and provided a path for fire travel when the conditions in the basement rapidly transitioned to a fully developed fire.

Figures 3-10 illustrate conditions at 200 seconds into the simulation, which relates to approximately 00:27 during the incident, the time at which the fire in the basement transitioned to a fully developed stage and rapidly extended up the basement stairway to Floor 1. Data is presented as a snapshot within a specific slice. Temperature and velocity data are provide for each slice (S1, S2, & S3 as illustrated in Figure 2).

Figure 3. Temperature Along Centerline of Basement Door Side C (S1) at 200 s

Note: From Simulation of the Dynamics of the Fire at 3146 Cherry Road NE Washington D.C., May 30, 1999, NISTR 6510 (p. 17) by Dan Madrzykowski and Robert Vettori, 2000, Gaithersburg, MD: National Institute for Standards and Technology.

Figure 4. Vector Representation of Velocity Along Centerline of Basement Door Side C (S1) at 200 s

Note: From Simulation of the Dynamics of the Fire at 3146 Cherry Road NE Washington D.C., May 30, 1999, NISTR 6510 (p. 18) by Dan Madrzykowski and Robert Vettori, 2000, Gaithersburg, MD: National Institute for Standards and Technology.

Figure 5. Oxygen Concentration Along Centerline of Basement Door Side C (S1) at 200 s

Note: From Simulation of the Dynamics of the Fire at 3146 Cherry Road NE Washington D.C., May 30, 1999, NISTR 6510 (p. 23) by Dan Madrzykowski and Robert Vettori, 2000, Gaithersburg, MD: National Institute for Standards and Technology.

Figure 6. Temperature Slice Along Centerline of Basement Stairwell (S2) at 200 s

Note: From Simulation of the Dynamics of the Fire at 3146 Cherry Road NE Washington D.C., May 30, 1999, NISTR 6510 (p. 21) by Dan Madrzykowski and Robert Vettori, 2000, Gaithersburg, MD: National Institute for Standards and Technology.

Figure 7. Vector Representation of Velocity Along Centerline of Basement Stairwell (S2) at 200 s

Note: From Simulation of the Dynamics of the Fire at 3146 Cherry Road NE Washington D.C., May 30, 1999, NISTR 6510 (p. 22) by Dan Madrzykowski and Robert Vettori, 2000, Gaithersburg, MD: National Institute for Standards and Technology.

Figure 8. Oxygen Concentration Along Centerline of Basement Stairwell (S2) at 200 s

Note: From Simulation of the Dynamics of the Fire at 3146 Cherry Road NE Washington D.C., May 30, 1999, NISTR 6510 (p. 24) by Dan Madrzykowski and Robert Vettori, 2000, Gaithersburg, MD: National Institute for Standards and Technology.

Figure 9. Temperature Slice Along Centerline of Floor 1 Door Side A (S3) at 200 s

Note: From Simulation of the Dynamics of the Fire at 3146 Cherry Road NE Washington D.C., May 30, 1999, NISTR 6510 (p. 19) by Dan Madrzykowski and Robert Vettori, 2000, Gaithersburg, MD: National Institute for Standards and Technology.

Figure 10. Vector Representation of Velocity Along Centerline of Floor 1 Door Side A (S3) at 200 s

Note: From Simulation of the Dynamics of the Fire at 3146 Cherry Road NE Washington D.C., May 30, 1999, NISTR 6510 (p. 20) by Dan Madrzykowski and Robert Vettori, 2000, Gaithersburg, MD: National Institute for Standards and Technology.

Figure 11. Perspective Cutaway, Flow/Temperature, Velocity, and O₂ Concentration

Figure 12. Thermal Exposure Limits in the Firefighting Environment

Note: Adapted from Measurements of the firefighting environment. Central Fire Brigades Advisory Council Research Report 61/1994 by J.A. Foster & G.V. Roberts, 1995. London: Department for Communities and Local Government and Thermal Environment for Electronic Equipment Used by First Responders by M.K. Donnelly, W.D. Davis, J.R. Lawson, & M.J. Selepak, 2006, Gaithersburg, MD: National Institute of Standards and Technology.

Compartment Fire Thermal Hazards

The temperature of the atmosphere (i.e., smoke and air) is a significant concern in the fire environment, and firefighters often wonder or speculate about how hot it was in a particular fire situation. However, gas temperature in the fire environment is a bit more complex than it might appear on the surface and is only part of the thermal hazard presented by compartment fire.

Tissue temperature and depth of penetration determine the severity of a thermal burn. Temperature and penetration are dependent on the amount of energy absorbed and the duration of the thermal insult as well as the properties of human tissue. In a compartment fire, firefighters absorb energy from any substance that has a temperature above 37^o C (98.6^o F), including hot compartment linings, contents, the hot gas layer, and flames. The dominant mechanisms of heat transfer involved in this process are convection and radiation (although conduction through personal protective equipment is also a factor to be considered).

The total thermal energy received is described in joules per unit area. However, the speed or rate of energy is transferred may be more important when assessing thermal hazard. Heat (thermal) flux is used to define the rate of heat transfer and is expressed in kW/m² (Btu/hr/ft²).

One way to understand the interrelated influence of radiant and convective heat transfer is to consider the following scenario. Imagine that you are standing outside in the shade on a hot, sunny day when the temperature is 38^o C (100^o F). As the ambient temperature is higher than that of your body, energy will be transferred to you from the air. If you move out of the shade, your body will receive additional energy as a result of radiant heat transfer from the sun.

Convective heat transfer is influenced by gas temperature and velocity. When hot gases are not moving or the flow of gases across a surface (such as your body or personal protective equipment) is slow, energy is transferred from the gases to the surface (lowering the temperature of the gases, while raising surface temperature). These lower temperature gases act as an insulating layer, slowing heat transfer from higher temperature gases further away from the surface. When velocity increases, cooler gases (which have already transferred energy to the surface) move away and are replaced by higher temperature gases. When velocity increases sufficiently to result in turbulent flow, hot gases remain in contact with the surface on a relatively constant basis, increasing convective heat flux.

Radiant heat transfer is influenced by proximity and temperature of the radiating body. Radiation increases by a factor of four when distance to the hot material is reduced by half. In addition, radiation increases exponentially (as a function of the fourth power) as absolute temperature increases.

Thermal hazard may be classified based on hot gas temperature and radiant heat flux (Foster & Roberts, 1995; Donnelly, Davis, Lawson, & Selpak, 2006) with temperatures above 260^o C (500^o F) and/or radiant heat flux of 10 kW/m² (3172 Btu/hr/ft²) being immediately life threatening to a firefighter wearing a structural firefighting ensemble (including breathing apparatus). National Institute of Standards and Technology (NIST) experiments in a single compartment show post flashover gas temperatures in excess of 1000^o C (1832^o F) and heat flux at the floor may exceed 170 kW/m² (Donnelly, Davis, Lawson, & Selpak, 2006). Post flashover conditions in larger buildings with more substantial fuel load may be more severe!

Figure 11 integrates temperature, velocity, and oxygen concentration data from the simulation (Figures 3-10). Detail and accuracy is sacrificed to some extent in order to provide a (somewhat) simpler view of conditions at 200 seconds into the simulation (approximately 00:27 incident time). Note that as in individual slices, data is presented as a range due to uncertainty in the computer model.

Alternative Model

In addition to modeling fire dynamics based on incident conditions and tactical operations as they occurred, NIST also modeled the incident with a slightly different ventilation profile.

The basic input for the alternate simulation was the same as the simulation of actual incident conditions. Ventilation openings and timing was the same, with one exception; the sliding glass door on Floor 1, Side C was opened at 120 s into the simulation. Conditions in the basement during the alternative simulation were similar to the first. However, on Floor 1, the increase in ventilation provided by the sliding glass door on Side C resulted in a shallower hot gas layer and cooler conditions at floor level. A side-by-side comparison of the temperature gradients in these two simulations is provided in Figure 13.

Figure 13. Comparison of Temperature Gradients Along Centerline of Basement Stairwell (S2) at 200 s

Note: Adapted from Simulation of the Dynamics of the Fire at 3146 Cherry Road NE Washington D.C., May 30, 1999, NISTR 6510 (p. 21 & 27) by Dan Madrzykowski and Robert Vettori, 2000, Gaithersburg, MD: National Institute for Standards and Technology.

The NIST Report (Madrzykowski & Vettori, 2000) identified that the significant difference between these two simulations is in the region close to the floor. In the alternative simulation (Floor 1, Side C Sliding Glass Door Open) between the doorway to the basement and the sofa, the temperatures from approximately 0.6 m (2 ft) above the floor, to floor level are in the range of 20 °C to 100 °C (68°F to 212 °F), providing at least an 80 °C (176 °F) temperature reduction.

While this is a considerable reduction in gas temperature, it is essential to also consider radiant heat flux from the hot gas layer. Given the temperature of the hot gases from the ceiling level to a depth of approximately 3′ (0.9 m), the heat flux at the floor would likely have been in the range of 15-20 kW/m² (or greater).

Questions

1. Temperatures vary widely at a given elevation above the floor. Consider the slices illustrated in Figures 3, 6, and 9, and identify factors that may have influenced these major differences in temperature.
2. How might the variations in temperature illustrated in Figures 3, 6, and9 and location of Firefighters Phillips (basement doorway), Mathews (living room, C/D corner), and Morgan (between Phillips & Mathews) have influenced their injuries?
3. Examine the velocity of gas movement illustrated in Figures 4, 7, and 10 and integrated illustration conditions in Figure 11. How does this correlate to the photos in Figure 1 illustrating incident conditions at approximately 00:28?
4. Explain how the size and configuration of ventilation openings resulted in a bi-directional air track at the basement door on Side C.
5. How did the velocity of hot gases in the stairwell and living room influence the thermal insult to Firefighters Phillips, Mathews, and Morgan? What factors caused the high velocity flow of gases from the basement stairwell doorway into the living room?
6. Rescue 1B noted that the floor in the living room was soft while conducting primary search at approximately 00:30. Why didn’t the parallel chord trusses in the basement fail sooner? Is there a potential relationship between fire behavior and performance of the engineered floor support system in this incident?
7. How might stability of the engineered floor support system have differed if the sliding glass door in the basement had failed prior to the fire departments arrival? Why?
8. How might the double pane glazing on the windows and sliding glass doors have influenced fire development in the basement? How might fire development differed if these building openings had been fitted with single pane glazing?
9. What was the likely influence of turbulence in the flow of hot gases and cooler air on combustion in the basement? What factors influenced this turbulence (examine Figures 4, 7, and 10) illustrating velocity of flow and floor plan illustrated in conjunction with the second question)?
10. How did conditions in the area in which Firefighters Phillips, Mathews, and Morgan were located correlate to the thermal exposure limits defined in Figure 12? How did this change in the alternate scenario? Remember to consider both temperature and heat flux.

Extended Learning Activity

The Cherry Road case study provides an excellent opportunity to develop an understanding of the influence of building factors, burning regime, ventilation, and tactical operations on fire behavior. These lessons can be extended by comparing and contrasting this case with other cases such as the 1999 residential fire in Keokuk, Iowa that took the lives Assistant Chief Dave McNally, Firefighter Jason Bitting, and Firefighter Nathan Tuck along with three young children. For information on this incident see NIOSH Death in the Line of Duty Report F2000-4, NIST report Simulation of the Dynamics of a Fire in a Two Story Duplex, NIST IR 6923.and video animation of Smokeview output from modeling of this incident

Ed Hartin, MS, EFO, MIFireE, CFO

References

District of Columbia (DC) Fire & EMS. (2000). Report from the reconstruction committee: Fire at 3146 Cherry Road NE, Washington DC, May 30, 1999. Washington, DC: Author.

Madrzykowski, D. & Vettori, R. (2000). Simulation of the Dynamics of the Fire at 3146 Cherry Road NE Washington D.C., May 30, 1999, NISTR 6510. August 31, 2009 from http://fire.nist.gov/CDPUBS/NISTIR_6510/6510c.pdf

National Institute for Occupational Safety and Health (NIOSH). (1999). Death in the line of duty, Report 99-21. Retrieved August 31, 2009 from http://www.cdc.gov/niosh/fire/reports/face9921.html

This post continues study of an incident in a townhouse style apartment building in Washington, DC with examination of the extreme fire behavior that took the lives of Firefighters Anthony Phillips and Louis Mathews.

A Quick Review

Prior posts in this series, Fire Behavior Case Study of a Townhouse Fire: Washington, DC, Townhouse Fire: Washington, DC-What Happened,and Townhouse Fire: Washington, DC-Extreme Fire Behavior examined the building and initial tactical operations at this incident. The fire occurred in the basement of a two-story, middle of building, townhouse apartment with a daylight basement. This configuration provided at grade entrances to Floor 1 on Side A and the Basement on Side C.

Engine 26, the first arriving unit reported heavy smoke showing from Side A and observed a bi-directional air track at the open front door. Engines 26 and 10 operating from Side A deployed hoselines into the first floor to locate the fire. Engine 17, the second due engine, was stretching a hoseline to Side C, but had insufficient hose and needed to extend their line. Truck 4, the second due truck, operating from Side C opened a sliding glass door to the basement to conduct search and access the upper floors (prior to Engine 17’s line being in position). When the door on Side C was opened, Truck 4 observed a strong inward air track. As Engine 17 reached Side C (shortly after Rescue 1 and a member of Truck 4 entered the basement) and asked for their line to be charged. Engine 17 advised Command that the fire was small.

Conditions changed quickly after the door on Side C was opened, as conditions in the basement rapidly transitioned to a fully developed fire with hot gases and flames extending up the interior stairway trapping Firefighters Phillips, Mathews, and Morgan. Confusion about building configuration (particularly the number of floors and location of entry points on Side A and C) delayed fire attack due to concern for opposing hoselines.

Modeling of the Cherry Road Incident

National Institute for Standards and Technology (NIST) performed a computer model of fire dynamics in the fire at 3146 Cherry Road (Madrzykowski and Vettori, 2000) using the NIST Fire Dynamics Simulator (FDS) software. This is one of the first cases where FDS was used in forensic fire scene reconstruction.

Fire Modeling

Fire modeling is a useful tool in research, engineering, fire investigation, and learning about fire dynamics. However, effective use of this tool and the information it provides requires understanding of its capabilities and limitations.

Models, such as the National Institute of Standards and Technology (NIST) Fire Dynamics Simulator (FDS) relay on computational fluid dynamics (CFD). CFD models define the fire environment by dividing it into small, rectangular cells. The model simultaneously solves mathematical equations for combustion, heat transfer, and mass transport within and between cells. When used with a graphical interface such as NIST Smokeview, output can be displayed in a three-dimensional (3D) visual format.

Models must be validated to determine how closely they match reality. In large part this requires comparison of model output to full scale fire tests under controlled conditions. When used for forensic fire scene reconstruction, it may not be feasible to recreate the fire to test the model. In these situations, model output is compared to physical evidence and interview data to determine how closely key aspects of model output matched events as they occurred. If model output reasonably matches events as they occurred, it is likely to be useful in understanding the fire dynamics involved in the incident.

It is crucial to bear in mind that fire models do not provide a reconstruction of the reality of an event. They are simplified representation of reality that will always suffer from a certain lack of accuracy and precision. Under the condition that the user is fully aware of this status and has an extensive knowledge of the principles of the models, their functioning, their limitations and the significance attributed to their results, fire modeling becomes a very powerful tool (Dele´mont & Martin, J., 2007, p. 134).

FDS output included data on heat release rate, temperature, oxygen concentration, and velocity of gas (smoke and air) movement within the townhouse. As indicated above, model output is an approximation of actual incident conditions.

In large scale fire tests (McGrattan, Hamins, & Stroup, 1998, as cited in Madrzykowski and Vettori, 2000), FDS temperature predictions were found to be within 15% of the measured temperatures and FDS heat release rates were predicted to within 20% of the measured values. For relatively simple fire driven flows such as buoyant plumes and flows through doorways, FDS predictions are within experimental uncertancies (McGrattan, Baum, & Rehm, 1998, as cited in Madrzykowski and Vettori, 2000).

Results presented in the NIST report on the fire at 3146 Cherry Road were presented as ranges to account for potential variation between model output and actual incident conditions.

Heat release rate is dependent on the characteristics and configuration of the fuel packages involved and available oxygen. In a compartment fire, available oxygen is dependent on the ventilation profile (i.e., size and location of compartment openings). The ventilation profile can change over time due to the effects of the fire (e.g., failure of window glazing) as well as human action (i.e., doors left open by exiting occupants, tactical ventilation, and tactical anti-ventilation)

In this incident there were a number of changes to the ventilation profile. Most significant of which were, 1) the occupant opened the second floor windows on Side C (see Figure 3), 2) the occupant left the front door open as they exited (see Figures 1 &2 ), 3) tactical ventilation of the first floor window on Side A, and opening of the sliding glass door in the basement on Side C (see Figures 1-3). In addition, the open door in the basement stairwell and open stairwell between the Floors 1 and 2 also influenced the ventilation profile (see Figure 1).

Figure 1. Cross Section of 3146 Cherry Road NE

Figure 2. Side A 3146 Cherry Road NE

Figure 3. Side C 3146 Cherry Road NE

Figure 4 illustrates the timing of changes to the ventilation profile and resulting influence on heat release rate in modeling this incident. A small fire with a specific heat release rate (HRR) was used to start fire growth in the FDS simulation. In the actual incident it may have taken hours for the fire to develop flaming combustion and progression into the growth stage. Direct comparison between the simulation and incident conditions began at 100 seconds into the simulation which corresponds to approximately 00:25 during the incident.

Figure 4. FDS Heat Release Rate Curve

Note: Adapted from Simulation of the Dynamics of the Fire at 3146 Cherry Road NE Washington D.C., May 30, 1999, NISTR 6510 (p. 14) by Dan Madrzykowski and Robert Vettori, 2000, Gaithersburg, MD: National Institute for Standards and Technology.

Questions

The following questions are based on heat release rate data from the FDS model presented in Figure 4.

1. What was the relationship between changes in ventilation profile and heat release rate?
2. What would explain the rapid increase in heat release rate after the right side of the basement sliding glass door is opened?
3. Why might the heat release rate have dropped slightly prior to opening of the left side of the basement sliding glass door?
4. Why did the heat release rate again increase rapidly to in excess of 10 MW after the left side of the basement sliding glass door was opened?
5. How does data from the FDS model correlate to the narrative description of events presented in prior posts about this incident (Fire Behavior Case Study of a Townhouse Fire: Washington, DC, Townhouse Fire: Washington, DC-What Happened,and Townhouse Fire: Washington, DC-Extreme Fire Behavior)?

More to Follow

In addition to heat release rate data the computer modeling of this incident provided data on temperature, oxygen concentration, and gas velocity. Visual presentation of this data provides a more detailed look at potential conditions inside the townhouse during the fire. The next post in this series will present and examine graphic output from Smokeview to aid in understanding the fire dynamics and thermal environment encountered during this incident.

Ed Hartin, MS, EFO, MIFireE, CFO

References

District of Columbia (DC) Fire & EMS. (2000). Report from the reconstruction committee: Fire at 3146 Cherry Road NE, Washington DC, May 30, 1999. Washington, DC: Author.

Madrzykowski, D. & Vettori, R. (2000). Simulation of the Dynamics of the Fire at 3146 Cherry Road NE Washington D.C., May 30, 1999, NISTR 6510. August 31, 2009 from http://fire.nist.gov/CDPUBS/NISTIR_6510/6510c.pdf

National Institute for Occupational Safety and Health (NIOSH). (1999). Death in the line of duty, Report 99-21. Retrieved August 31, 2009 from http://www.cdc.gov/niosh/fire/reports/face9921.html