FLAME SPREAD–THE SECRET KILLER
BY FRANCIS L. BRANNIGAN
A complete lack of knowledge of flame spread/fire growth led to a “training exercise” in which firefighters without a hoseline were sent into a converted school bus with fire in a foam plastic couch. There were serious injuries. The state of New Jersey recently paid out almost $2 million to the victims.
There seems to be an element of resistance in the suppression forces to learning about flame spread/ fire growth. In the Happyland Social Club fire in New York City, 87 people died before a hoseline could be stretched. In the Dublin Disco Fire–50 fatalities were related to flame spread. The Irish government had an exact replica of about 25 percent of the disco rebuilt and burned to answer the question, Why did the fire spread so fast? You and your family will probably spend time in similar places. Do your children know what to do in a similar situation? Do you evaluate flame spread potential when preplanning a building? You should.
BULLETIN: POTENTIALLY DEADLY WOODEN
The following information is from Chief Robert Bruce Graham, Lake Bluff, Illinois. The chipboard webs of wooden I-beams disintegrated reportedly under only moderate heat conditions.
In an unfurnished townhouse, a 30-gallon plastic garbage can with cleanup contents was located on a wooden floor supported by wood I-beams with chipboard webs. When cleanup workers returned from lunch, they noted smoke coming from the structure. The garbage can had ignited, possibly from a cigarette. The first floor was smoldering; only smoke was in the basement. The floor was “spongy.” The order to evacuate was given. Within a minute, the floor collapsed. Slight burning was noted on the beams, but the beam web had disintegrated into wood chips.
A new structure had “smoke showing.” The floor was weak. Firefighters evacuated. The floor fell into the basement. A torpedo-type gas heater had been used on the first floor to dry drywall taping.
About 10 years ago, some trusses with glued finger joints failed in Oregon.
Each year the National Fire Protection Association (NFPA) issues a statistical report of firefighter fatalities, with some case studies. The following two items are drawn from “1996 Firefighter Fatalities” in NFPA Journal, July/August 1997.
Illinois. The fire was in a vacant boarded-up building. An offensive attack was abandoned when a portion of the ceiling fell as firefighters were making a three-foot advance into the building. A mutual-aid firefighter was killed when he was hit by a falling wall, pushed by a roof collapse, while attempting to remove a plywood covering from a ground-floor window to facilitate a master stream attack.
Old, poorly maintained buildings get more hazardous every day. We must carefully weigh the risk of death or injury against the goal to be accomplished. Just “putting out the fire” is not enough.
Nebraska. Firefighters searching a one-story store for fire noticed a glow through an opening in a heavy plaster ceiling. As they opened the ceiling, a large section collapsed, trapping the captain. The other two firefighters got out. The captain was found a few feet from the main door; his face piece was in place, but he had run out of air. The roof was described as a wood frame assembly with a slight peak running from front to back. A truss?
The fire department plans to establish a red-flag system to identify dangerous buildings. A city ordinance requiring that placards be placed on buildings with truss roofs may be proposed (such an ordinance was passed in the state of New Jersey; see Ol` Professor, March 1997). [Commissioner Martin Pierce informed me that two years ago, the Boston (MA) Fire Department surveyed all buildings for truss roofs.] I must reiterate: You can have raging fire over your head and not know it. Any clear span may have been accomplished by a truss. If in doubt, determine the roof construction before you go on it or under it.
Years ago, when I was an AEC safety and fire protection engineer, Columbia University researchers had erected a dangerously unstable 75-pound solid concrete block shielding wall around a vertical Van De Graff generator to protect the staff from neutrons. I went up to see the redoubtable Dr. John Dunning, who typically had a great respect for radiation hazards and little regard for mundane hazards such as falling bricks–and none for “federal bureaucrats.” I charged into his office like Paul Revere, with the bulletin: “There is a terrible neutron hazard in the basement!” He was all attention. “Seventy-five pounds of neutrons and protons are going to smash somebody`s skull–and don`t forget, gravity-accelerated nuclear particles are the very worst.” The wall came down that day. Perhaps we could emulate the Dunning incident by considering a truss-fire situation as a haz-mat incident, where good accepted practice is to determine what you are getting into before you act.
Firefighters should not be on or under a truss on fire. Opening a ceiling admits air and thus intensifies the fire. Opening a ceiling under a burning truss roof is like playing Russian roulette. The roof system is the minimum required to defeat gravity, and it is being weakened minute by minute by the fire. Gravity always acts instantly and, particularly for trusses, without warning.
The NFPA article also includes a “Special 10-Year Analysis of Firefighter Deaths While Responding to or Returning from Alarms” (page 54), which is well worth study and discussion.
The International Society of Fire Service Instructors (ISFSI) publishes Instructograms by a long-time friend, Assistant Chief Roger McGarry, Montgomery County, Maryland. The July/August issue tells of some terrible disasters (I just can`t call them “accidents”) in training activities.
A converted chicken coop was the drill site. Heat continued to build up during a breathing apparatus drill. Fire flashed over tarpaper (known to have a high flame spread); there was no backup line. Death and injury resulted.
A rookie volunteer less than a month in the department was sent into a live fire drill without proper supervision. The structure flashed over. The rookie panicked, stood up, took off his gloves, and lost the use of his hands.
An instructor was demonstrating rappelling off a tower. He wrapped the rope the wrong way around the belt clip and fell to his death. An instructor who used an undersized rope fell 60 feet and was out of work for 18 months.
The lower part of a wooden tower was being used for SCBA training and the top level for removal of persons from the roof by an aerial. Some freelancer thought there was not enough smoke, so he added half a tire and a flammable liquid. The fire accelerated and ignited the tower. Because of competent instructors, the fire was extinguished and all were removed from the roof. Two suffered second-degree burns.
The bulletin contains a wealth of practical suggestions that can help to prevent a heartbreaking disaster.
NOW WE KNOW WHO DID IT
Georgia Tech`s Engineering Hall of Fame recently enshrined J. Calvin Jureit, who in 1955 invented the gang nail (or gusset plate) truss. A Miami Herald article notes that Jureit`s house in Gables Estates has a huge 40- 2 32-foot living room with a 14-foot-high ceiling. (Local firefighters, take note: Some sort of truss?) I wonder if Jureit is aware of the loss of life that has occurred in fires involving lightweight wood trusses and the so far unsuccessful work done to attempt to improve their fire characteristics.
THE RESEARCH HAS BEEN DONE
As reported in BCFS3 (p. 551), the Department of Agriculture`s Forest Products Laboratory in Madison, Wisconsin, had received a grant to address “critical barriers to advances in structural wood engineering associated with concerns for the fire safety of new structural members such as trusses.” The project summary noted that “The concern for the safety of firefighters has been most visibly expressed by F.L. Brannigan in Building Construction for the Fire Service.”
Our first consideration in any rating, which is done using E119 fire resistance test (the standard test), is to be aware of the deficiencies of the test from our point of view:
Loads: The test structure is loaded with a 30-40 pound static load–not nearly representative of a moving crew of firefighters. The test does not envision any of the following:
fire burning down through the floor,
penetrations of the gypsum sheath,
fire starting in the void as in the case of a defective electrical fixture,
the accumulation of CO in the truss void, and
the spread of fire through the truss void.
In short, the fact that a gypsum board wood assembly passes the ASTM E119 test and gets a one-hour fire resistance rating has no relationship at a real-fire situation. In this research, the units tested were without a gypsum membrane ceiling and were thus directly exposed to the fire.
The investigators found excellent results with an all-wood truss (no gusset plates) when they enclosed all four sides with 13- or 16-mm fire rated gypsum board, with edges taped (and presumably nails set). I photographed such trusses (I don`t know if all wood or gusset plate) in Alexandria, Virginia, more than 25 years ago. Unfortunately, that was before I had computerized my slide inventory.
The addition of small nails or screws was not effective. Trusses composed of all wood, without metal plates, lasted somewhat longer than metal-plate-connected trusses. The older truss floors were built that way before the metal plates were developed. When metal plates were covered with gypsum board, the trusses lasted about as long as the all-wood trusses.
How does the following, paraphrased from the report, sound as a warning to firefighters about a forthcoming collapse?
Firefighters are concerned because wood truss systems fail without warning in fires. It is possible to design a more gradual failure mechanism into the truss by protecting some selected plates and leaving others bare so the sagging will be a warning.
Personally, I think installing in the void strings of Chinese firecrackers that would ignite and sound off would be more effective! Again, all this work was done in the never-never land of unrealistic loading of the trusses.
Know your buildings. Trusses collapse without warning. Examine trusses for possible fire. Do not be on or under a truss on fire.
Truss voids throughout the building are a typical feature of multistory combustible structures built in recent years. The voids are interconnected by utility openings, firestopping not withstanding. A fire that starts in or gets into the void through one of the many penetrations in the gypsum “sheath” can spread through the whole building like a cancer.
A very interesting, serious fire occurred in Austin, Texas, as described in the following letter from Chief Robin Paulsgrove.
“Firefighters responded to a structure fire at the Centennial Condominiums (501 West 26th Street) at approximately 5 p.m. on Friday, December 13, 1996. They found a dumpster in the underground parking garage on fire with flames extending into a trash chute. The chute ran from above the dumpster to the attic. Austin Fire Department personnel extinguished the fire with no resulting structural damage. Firefighters remained on the scene for nearly two hours. They used a thermal imaging camera to check for extension in the trash chute in the attic. In addition to searching the trash chute, firefighters searched the adjoining storage space on each floor.
“Investigators believe that a stray ember from the dumpster fire ended up in the chute between the first and second floors and smoldered undetected for hours before breaking into fire around midnight. The investigation determined that the fire originated in an area that would have been undetectable by firefighters responding to the previous dumpster fire. It seems as if the chute construction does not match the plans.
“The blueprints provided the builder with two chute designs. One option was a chute constructed of concrete masonry blocks faced with a brick veneer. The second option described in the plans required a wood frame faced with SheetrockTM on the interior, sheathing on the outside and a brick veneer over the sheathing. The builder used the second option.
“The actual structure does not appear to have had the SheetrockTM on the interior of the shaft, nor was there any fire stopping at the top or bottom of the shaft. The shaft had a sprinkler head on each floor and at the termination point above the dumpster. The sprinkler head did not activate. Fire department investigators found the valve to the sprinkler head at the termination point above the dumpster closed.
“The builder used parallel trusses for the structural component of the floor-ceiling assembly. Draft stopping was in place but proved ineffective. Fire prevention staff was able to find evidence of a two-hour fire partition wall. We do not know if the draft stopping called for in the attic was in place. The design of the truss layout would have prohibited the draft stopping from effectively working. The fire could easily go around the draft stops.”
Note that had a 13R sprinkler system been installed in this building, it likely would have proved totally ineffective. At the time the areas to be sprinklered were selected by the code committee, the hazard of these extended interconnected voids was not appreciated.
An article in Fire Marshals of North America by Dick Hughey, PE, manager of engineering and standards for the Insurance Services Office (ISO) (NFPA, September 1996), makes it clear that the 13D and 13R sprinkler systems cannot be relied on for property protection. They are not only used for residences but for “motels, group houses, dormitories, halfway houses, fraternity and sorority houses, rooming houses, small hotels, care houses, and many other buildings with habitational occupancies.” He points out that ISO is now seeing property losses in structures with substantial areas (up to 67 percent) not covered by automatic sprinklers.
In truss floor buildings, the truss floor void (or trussloft, coined simultaneously by Tom Brennan and yours truly) and trussed attic represent huge concealed accumulations of kindling-size wood, with adequate available air supply. A standard NFPA 13 sprinkler system would control fire in these voids. This coverage was eliminated in the tradeoff to get life safety sprinklers into residential buildings.
Let us ensure that part of the tradeoff is not firefighter injuries and death. When the occupants have been removed, the incident manager must decide with regard to the justification for risking lives to save property when proven fire protection systems were omitted to save money.
MORE ON HUGE WAREHOUSES
Recently, the Washington Post published a story on huge warehouses in the Washington area, among which those below were listed. Most metropolitan areas could have a similar story.
Winchester, VA: Kohl`s 350,000 sq. ft., 600 jobs;
Aberdeen, MD: Saks Fifth Avenue, 500,000 sq. ft., 700 jobs;
Joppa, MD: Hecht`s, 800,000 sq. ft., 230 jobs;
Frederick, MD: Toys R Us, 673,000 sq. ft., 260 jobs;
Hagerstown, MD: Staples, 850,000 sq. ft., 700 jobs;
Bowie, MD: Nordstrom, 410,000 sq. ft., 185 jobs;
Baltimore, MD: Warner, 400,000 sq. ft., 100 jobs.
A huge warehouse is a major employer and taxpayer. Its protection from loss should be a serious concern to the political authorities. Sometimes they seem more interested in protecting the management from “harassment” by the fire department. I recommend concerned fire departments do the following:
Form a focus group on the local warehouse.
Obtain a copy of the report from NFPA.
Study Chapter 14, “Rack Storage” in BCFS3.
Review Chapter 13, “Sprinklers” in BCFS3.
Organize heavy mutual aid.
Share the fire department`s concerns with the management of the warehouse. See the approach suggested on page 6 of BCFS3.
Review the article about the $280 million disaster in a sophisticated sprinklered warehouse in the October 1997, Ol` Professor.
No matter how many fires you have been to, you are a rookie at huge warehouse fires. Keep in mind that the building is supposed to be able to take care of itself. Firefighter safety is the first consideration.
A RECOMMENDED ARTICLE
Maybe you were on vacation and missed it. Cut out, copy, and send to every officer Chief Robert Bingham`s “Preparation and Training Key To Emergency Operations” (Fire Engineering, August 1997). It is like what New Yorkers called a “blivet”–five pounds of superb experience and thinking in a one-pound bag.
When I first started lecturing about building construction, I was given a slide of an aluminum beam. My reaction? “Who would be so stupid as to use a material that melts so readily as a structural element?” Well, the Ol` Professor was wrong. A letter from B.C. Robert Meirop of the Bolingbroke (IL) Fire Department tells the story (see photo on page 105). Some wood-frame townhouses built in 1972 have aluminum beams. Arriving firefighters found a fully involved first floor. A defensive attack was initiated. Crews did not enter until the extent of structural damage was assessed. The arrow indicates a damaged aluminum beam. n
© Francis L. Brannigan 1997.
All new townhouse construction featuring lightweight wood roof trusses, floor trusses, and metal studs. In some cases, the metal studs are nonfirestopped, giving a channel for smoke and heat to spread unchecked throughout the building if the fire breaks through the drywall skin. (Photo by author.)
Room of origin. Heavy destruction of the floor joist. Note the damaged aluminum beam, marked by an arrow.
FRANCIS L. (FRANK) BRANNIGAN, recipient of Fire Engineering`s First Lifetime Achievement Award, has devoted more that half of his 56-year career to the safety of firefighters in building fires. He is well known for his lectures and videotapes and as the author of Building Construction for the Fire Service, Third Edition, published by the National Fire Protection Association. Brannigan is an editorial advisory board member of Fire Engineering. He may be reached at (301) 855-1982.