Firefighting has always been inherently risky. This is especially true once fire penetrates a building’s concealed spaces and begins to chew away at structural members and the structure begins its collapse sequence. What is unknown to the incident commander is when a collapse will occur.
Recent scientific testing by the National Institute of Standards and Technology (NIST) on the effects of fire on lightweight truss systems should cause the fire service to rethink its approach to firefighting with these types of systems. The tests reveal very early collapse potential.
Lightweight truss systems typically are constructed of two-by-four-inch wood materials. The system is engineered to hold up a substantial deadweight load for its design and construction. They are most often used to construct attic systems, but may also be found in floor systems.
Lightweight truss systems are mass-produced off site and brought to a building location. These systems use metal plates, often called “gusset plates,” to hold the wood components together and replaces the old method of “nailing.” These plates have dozens of small teeth (typically no more than three-eighths of an inch in depth) that are pneumatically pressed into the wood during the assembly process (see photo 5). It is these plates that have been found to contribute to the early collapse potential. As fire penetrates a space containing a lightweight truss system, the metal gusset plates rapidly absorb heat, even prior to direct flame impingement. As the metal heats, wood charring quickly occurs around the teeth of the plate. Because of the superficial penetration of the teeth, the charring rapidly causes the loss of hold. Eventually, the plate will literally fall off. This leaves the truss system held in place purely by friction.
Once any single component of the truss fails, the entire truss system is severely compromised. Any weight or movement across the roof surface or ventilation activities can cause an immediate collapse of the roof. The same will occur if the system is disturbed from below – such as pulling ceiling, or operating hose streams into the attic space or floor assembly. It should also be noted, when lightweight construction is involved, history indicates large portions of the roof will collapse.
Over the past two decades, the National Fire Protection Association (NFPA) has cited numerous cases of lightweight construction contributing to firefighter fatalities. In one case, in 1988, two Orange County, FL, firefighters died while fighting an attic fire in a strip mall. Firefighters were operating from below and collapse occurred before the second-due company arrived on scene!
In another case, the Phoenix, AZ, Fire Department experienced a “near miss” in July 1989 when three of four members of a ladder crew fell through the roof at a residential fire. One member disappeared completely into a well-involved attic space for several seconds before he crawled back to the opening and was dragged out by fellow firefighters. All survived with minor injuries only because they were wearing self-contained breathing apparatus (SCBA), with their facepieces on their faces, and were fully encapsulated in protective clothing.
The investigation revealed a lightweight truss system. The gusset plates had fallen off many of the truss components. By contrast, of the surviving two-by-four-inch truss members, the worst char depth had penetrated less than 20% of the thickness of the truss members. Had the roof system used a conventional “nailing” method, with deep penetrating nails, the roof system most likely would not have collapsed in such a short time (see photo 6).
The Phoenix Fire Department partnered with NIST to conduct live-fire testing on lightweight truss systems. Four identical buildings were constructed and each contained identical interior contents. These contents included a bed, chairs, dresser and tables. All roof systems were identical lightweight truss systems. Walls and ceilings in the interior living areas were all dry-walled. Each roof had identical air conditioning units mounted on the roof in the same location. The roof surfaces, however, were different. Two had plywood and two had oriented strand board (OSB), also known as “chip board.” Clay tile was placed on one of each of the two different surfaces. Asphalt shingles were similarly placed on the other two surfaces.
Mounted on each roof representing firefighters were two mannequins, each wearing protective clothing, a helmet and SCBA – total weight of each was approximately 240 pounds. Each mannequin was tethered to an aerial apparatus to allow retrieval once collapse occurred (see photo 1).
Each fire was ignited electronically at the same location in each building and allowed to free burn until collapse occurred. In all four cases, roof collapse occurred on an average of 17 minutes following ignition. For most urban fire departments, with a typical response time for ladder companies, this means that rooftop ventilation operations may still be underway at 17 minutes. In other words, collapse risk will be present before ventilation is typically complete (see photos 2 and 3).
The most shocking result of the NIST test was what happened once flame penetrated the attic space. Collapse occurred on an average of only eight minutes following penetration! This eight-minute factor should be most troubling for the urban fire service. In those situations where a fire starts in the attic space, firefighters should expect collapse to occur as they arrive on scene, or very early into the initial attach (see photo 4).
The Phoenix Fire Department forwarded data and videotapes on the NIST test to Frank Brannigan, a Firehouse® contributing editor and one of the most respected authorities on building construction in the American fire service, for his review. He refers to the 17-minute and eight-minute collapse time factors as “dangerously and deceptively high.” He said he believes that collapse times can be expected to occur much earlier. He pointed out that the test mannequins (firefighters) were a static load – no movement of firefighters on the roof occurred, nor was there any cutting activity that would typically be found in most vent operations. With a fire in the attic space, and crews working on the roof, collapse may occur earlier than the test results indicate!
For the fire service, the data gained from these four scientific experiments has to change our tactics. We cannot allow members on the roof to conduct ventilation operations where these systems are used and fire has substantially penetrated these spaces. Nor can we allow members underneath such an attic system – particularly where large open spaces exist under the ceiling. To do so is to invite a catastrophic event and probable fatal injuries to firefighters.
Gary Morris will present “Future Challenges in Firefighter Survival” at Firehouse Expo 2005, July 26-31 in Baltimore.Gary Morris is the fire chief for the Rural-Metro Fire Department, Maricopa and Pinal County Operations, in Arizona. He previously was the fire chief of the Seattle Fire Department and retired as an assistant chief with the Phoenix Fire Department after a 30-year career. Morris holds a master’s degree in organizational management from the University of Phoenix.
