Researchers ignited a series of fires within a seismically tested structure to examine how earthquake-induced damage influences a building’s fire performance. UCSD/WPI
New research shows that earthquake damage can significantly undermine a building’s fire performance.
April 16, 2013—Although some of the largest fires in history followed earthquakes, little is known about how buildings respond to fire in the wake of a seismic event. But newly released research findings indicate that earthquake-induced damage can have a significant effect on a building’s fire performance in that it can compromise escape and rescue routes and make it easier for fire and smoke to spread through a structure.
In 2012 the University of California at San Diego (UCSD) conducted groundbreaking shake table tests on a five-story concrete building in which the top two floors were designed and equipped to resemble a hospital. The primary goal of the motion testing was to provide a better understanding of how the contents of a building that is important to the life of a community respond to seismic forces. (See “Building and Contents Tested on Shake Table,” Civil Engineering online.) But given the unusual opportunity to test a full-scale model, UCSD didn’t stop there.
The university teamed up with Worcester Polytechnic Institute (WPI) to also examine the building’s fire performance after the seismic motion tests. Such systematic studies had never been conducted before, even though it is well known that postearthquake fires can exact a frightening toll, as seen in the 1906 San Francisco earthquake and the 1923 Great Kanto and 1995 Kōbe earthquakes, in Japan. “Postearthquake fire is a really challenging problem, and it is a real issue for society,” said Tara Hutchinson, Ph.D., P.E., M.ASCE, a professor of structural engineering at UCSD’s Jacobs School of Engineering and the principal investigator in the seismic phase of the study, in written responses to questions posed by Civil Engineering online. “We had a tremendously unique opportunity to utilize the seismically tested building specimen to study this issue, [and] for us it was important to capitalize on this opportunity.”
WPI researchers conducted six fire tests in four different compartments within the building’s third floor, which had been designed to resemble a typical office with fire-rated gypsum walls, swinging doors, and a fire-rated ceiling system. They also tested an area behind the building’s elevator shaft on that floor. (See a UCSD video on the tests.)
“Those four different rooms allowed us to look at compartmentation, the performance of fire protection systems, fire-stop systems, how the exterior facade performed relative to allowing the escape of fire and smoke, and then how fire and smoke could be transmitted through interior shafts like the elevator,” says Brian Meacham, Ph.D., P.E., an associate professor of fire protection at WPI and the principal investigator in the postearthquake fire study. Researchers used liquid heptane to ignite “pan fires” within the building. “The fire tests were limited in fire size, duration, and location so as to provide large enough fires to assess building system performance without risking the potential for collapse by further damaging the structure,” Meacham says.
Released on March 13, the fire test results show that earthquake-induced damage can significantly undermine a building’s fire performance. The stairways, elevators, and doorways were so damaged by the largest motion test that it would have been difficult for occupants to escape or for rescue personnel to respond to the fire. What is more, the building’s interior and exterior walls separated in some places, leaving gaps through which fire and smoke could easily have spread from one floor to the next. “We were able to clearly show that the smoke would spread up the elevator shaft and distribute on the higher floors,” Meacham explains. Another finding was that “the particular facade system that was on the lower three floors separated by about three inches from the floor decks, which meant that you would have had the potential for unrestricted fire spread from floor to floor had we allowed the fire to go into full involvement.”
While the tests provided unprecedented results, they were not comprehensive. The building did not have windows, so researchers were unable to test how windows would have affected the building’s performance under the seismic forces and fire conditions. And although the sprinkler system withstood the motion tests, it was not tested during the fire simulation. “In a real event, I would expect an operating sprinkler system to be a significant benefit,” Meacham says. Furthermore, the structural connections that failed as a result of the earthquake motions were not tested. “Should such failures occur in a real event followed by a fire, the fire could further weaken the connections and bring the structure to failure even if it withstands the initial earthquake load,” Meacham says. “Knowing that buildings could be in that condition under these large events is of concern.”
WPI researchers hope to continue studying postearthquake fire performance to test different types of building structures. In fact, they have teamed up again with UCSD, this time submitting a proposal to the U.S. Fire Administration for funding to explore postearthquake fire performance in the types of lightweight construction systems that are often used for low-rise buildings. “Our [initial] findings indicate that under the largest, reasonably expected earthquakes, we could see damage to structural and nonstructural systems [that] can negatively impact the fire and life safety performance of buildings,” Meacham says. “While we are looking to try and quantify the magnitude of the potential impacts and suitable responses for building design practice and regulation, we have to recognize that this was one test series and one set of data. Much more data on similar and different construction materials and building configurations [are] needed to better characterize and address issues of concern.”