Reexamining Premises for High-Rise Design

Reexamining Premises for High-Rise Design

By Michael A. Gips

Engineers, building designers, and other experts have pored over the wreckage of the World Trade Center. Can they provide clues for making high rises safer?

When the World Trade Center buildings succumbed last September after direct hits from two terrorist-driven planes, among the many ramifications was the need for government and business experts to reexamine the premises on which high rises have for decades been built. Since that day, experts have sifted through debris, studied the site, viewed tapes, and analyzed a host of other sources of information in an attempt to answer one question: Could any building have withstood such an attack, and if so, how should future building designs be changed to achieve that objective?

W. Gene Corley, the Chicago-area engineer who headed the team conducting the study, says that so far the answer has proven elusive. There is as yet no blueprint for a better building. Still, much has been learned. This article looks at those lessons and where building design is likely to go.

The Evidence

The events of that day are well known, but it's important to understand the details of what happened from a structural perspective so that the analysis can be put into context.

At 8:46 a.m. on September 11, 2001, hijackers flew a Boeing 767 jet loaded with fuel for a cross-country journey into the north tower of the World Trade Center (One WTC) between the 94th and 98th floors. The plane slammed into the building at an estimated 470 miles per hour.

Seventeen minutes later, a second 767 plowed into the south tower (Two WTC) between the 78th and 84th floors. Its speed at impact was estimated to be 550 mph. Both buildings burst into flames--with furniture, drapes, paper, and other combustibles in the buildings feeding the fire.

At 10 a.m., after surviving the impact for almost an hour, Two WTC collapsed. One WTC crumbled approximately 30 minutes later, about one hour and 43 minutes after impact. Because the buildings were able to resist immediate collapse, about 55,000 complex occupants successfully evacuated. Nearly 3,000 others, including responding police and firefighters, could not.

Pelted by burning debris, the five other buildings in the WTC complex also collapsed, at least partially; for example, Seven WTC, a 47-story building, burned for seven hours before crumbling, and Three WTC, the 22-story Marriott World Trade Center Hotel, was destroyed. Hundreds of nearby buildings and water mains were also damaged.


Much of what is known about the collapse comes from an analysis of the wreckage and related evidence by the WTC Building Performance Assessment Team (BPAT), a group of experts--including structural and fire engineers, blast-effects specialists, building designers, and investigators--that was assembled by the Structural Engineering Institute of the American Society of Civil Engineers (SEI/ASCE) under the auspices of the Federal Emergency Management Agency (FEMA). The team's investigation yielded an extensive report, the World Trade Center Building Performance Study: Data Collection, Preliminary Observations, and Recommendations. (The report will be referred to as the FEMA/ASCE report.)

The study team based its conclusions on an analysis of debris, video and photographic materials, first-hand accounts, and other evidence--much of it provided by organizations performing recovery efforts in the immediate aftermath of the attacks, such as the Structural Engineers Association of New York, the U.S. Army Corps of Engineers, and other private and government consultants.

The report discusses the design and construction of the seven World Trade Center buildings as well as other affected buildings in the area. Its primary focus is to determine the causes of the collapse and any lessons that can be learned.

The report notes that while the Twin Towers were designed to resist a jet impact, that jet was the much smaller 707, and engineers assumed that any impact would take place at landing speeds, not 500 mph. The team found that certain design features were critical in keeping the buildings standing for as long as they did. For example, the study notes that following the impact, floor loads in One WTC originally supported by destroyed exterior columns were transferred to "other load paths," such as adjacent perimeter columns.

Three WTC, the Marriott Hotel, also bore up remarkably well, according to the report, despite being pelted by debris from One and Two WTC. The report states: "It is noteworthy that the building resisted both vertical and horizontal progressive collapse when subjected to debris from Two WTC. The overloaded portions were able to break away from the rest of the structure without pulling it down, and the remaining structural system was able to remain stable and support the debris load." (The building did later collapse when Two WTC came down on top of it. See page 59 for more about the Marriott on September 11 and lodging lessons from that event.)

Other features of the Twin Towers support structure--such as trusses, stairwells, and fireproofing--were identified as playing a role in how and when the buildings fell, preventing occupants above the crash point from escaping. But that doesn't mean they were faulty or should be changed in future construction. "These features should not be regarded either as design deficiencies or as features that should be prohibited in future building codes," states the report. Instead, those aspects of the building should be evaluated to see whether they performed satisfactorily.

In a few cases, building designers have already been modifying practices to improve safety, for example with regard to stairwells and fireproofing. But Corley, the Chicago-area engineer who headed the team conducting the analysis, says that most of the report's recommendations involve future research, not suggestions for specific design changes, because solid answers were not forthcoming and because the team was not convinced that the future threat of similar attacks was sufficient to justify changes to the building code.

Following are the report's findings on specific building issues as well as a discussion of some steps already being taken by industry groups.

Support structures.

When the planes struck on September 11, they took out between 31 and 36 of the 59 columns that ran across the face of One WTC and between 27 and 32 columns of the 59 girding Two WTC. Remarkably, neither building collapsed immediately, which experts attribute to the sheer size of the buildings and to the exterior columns' being placed very close together. But ultimately the buildings did suffer a progressive collapse--the domino effect that occurs when one or more support structures fail, causing one floor to fall into the next.

According to the FEMA/ASCE report, what eventually took down the towers was the fire generated from the jet fuel that ignited much of the buildings' contents. With the emergency water supply having been disrupted by the impacts, the sprinkler system had no chance to quell the fire.

Even if the water supply had been left intact, the sprinkler system probably would have been ineffective, the report says, in part because the initial spread of flames would have overwhelmed the system. In fact, the water supply was left intact in the nine-story building, Five WTC, but it does not appear to have functioned in light of the lack of water damage there.

The raging fire in the towers weakened the already damaged vertical support members, leading to collapse. Corley, who chaired the FEMA/ ASCE study and who spoke at a National Institute of Standards and Technology (NIST) hearing on the issue, has said the industry needs to learn more about how loads are distributed when large portions of a building are removed.

At the same hearing, Brian Meacham, an engineer with the global firm Ove Arup & Partners, called for investigation of "the effect of load transfer from critical zones to elements under duress and the composite action of the core slabs and connections to the frame...."

Codes in other countries do address progressive collapse, and they could be good starting points, experts say. For example, according to Jon Magnusson, a member of the study team and a partner at the Seattle-based construction engineering firm Skilling Ward Magnusson Barkshire, Inc., some European codes specify that if a column falls, floors must have enough ties to transfer weight loads, preventing collapse.

The lack of codes isn't holding back some developers. For example, Ysrael Seinuk, a construction engineer who has worked on high rises in the United States, Britain, and Israel, says that clients are asking for collapse-resistance above and beyond code.

He notes that one potential vulnerability is columns that are close to the street or otherwise exposed. In a blast, columns can be stressed to their breaking point. When deemed appropriate for the risk, Seinuk has been protecting those columns with "sacrificial" material, such as steel or concrete shells. Similarly, he is designing loading docks, a likely site for a car or truck bomb, so that they would direct an explosion away from the building.


Fireproofing involves insulating structural elements so as to retard heat transfer from a fire to the protected structure. It can take the form of a spray-on material, thin coatings, or insulating boards or blankets (often used in walls). According to a paper by International Protective Coatings, the main classes of fireproofing material are cementitious products (such as concrete coatings), intumescent materials (thin films such as vinyl), fibrous materials (boards and blankets of mineral wool and ceramic fiber), and composite materials (which combine elements of the other three classes).

Fireproofing plays a critical role in preventing collapse. In One WTC, structural elements were originally protected up to the 39th floor with a spray-applied product containing asbestos. For environmental reasons, the asbestos-containing material had been subsequently either encapsulated or replaced by a safer fireproofing substance (though some experts say that the asbestos-containing material is more effective).

On all other floors of One WTC and throughout Two WTC, a spray-applied asbestos-free mineral fiber was used. The spray was applied to each element of the steel floor trusses.

The average thickness of the fireproofing on the trusses was 3/4 inch. In the mid-1990s, a decision was made to increase the thickness to 1 1/2 inches. The upgrade was made to individual floors as they became vacant. By September 11, a total of 31 stories in the two towers had been upgraded, including every floor in the impact zone of One WTC, but only in the lowest impacted floor in Two WTC.

Spandrels (areas between the head of a window on one level and the sill of a window immediately above) and girders also received spray fireproofing. Thicker steel column sections received less fireproofing relative to thinner sections. Gypsum board was used for fireproofing in elevator shafts and stairwells.

Analysis of the debris from the September 11 attacks revealed that the impact of the planes knocked off much of the fireproofing, which the team concluded probably hastened the structural failure of the towers. Corley says that, therefore, one lesson from the WTC incident is that target buildings need to have fireproofing that stays in place during an impact. He also says that target buildings, those existing as well as those yet to be built, should increase the level of fireproofing in parts of the structure with a higher fire load--those with highly flammable materials.

For example, a typical office space that has been converted to a law library--which has more combustibles--may well need extra fireproofing. When a fire load increases, Corley says, "you check your fireproofing and [you] might have to open the walls and put more fireproofing in there."

Corley surmises that in Seven WTC, thousands of gallons of diesel fuel for emergency generators were present on the fifth floor. The blaze was likely enhanced in intensity and duration by that fuel, the hypothesis goes, which helped the fire to eat away at the fireproofing of transfer trusses between floors five and seven, eventually crippling those trusses and causing the building to implode. These structural members might have held up a bit longer with additional fireproofing.

Part of the problem, says Corley, is that in protected steel structures, the main part of the building is fire tested, but connections (joints uniting two or more distinct building elements) are not. What's more, the same thickness of fireproofing is often placed on connections as on other structural members, although connections are often made of different material. "We don't know if that difference is significant or not," he says.

With regard to how the WTC withstood the fire, some experts say that fireproofing was too thin as well. Speaking at the NIST hearing, Professor James Quintiere of the Department of Fire Protection Engineering at the University of Maryland noted that the collapse times of the Twin Towers were proportional to the thickness of the fireproofing.

Specifically, One WTC, which had 1 1/2 inch fireproofing, fell in 104 minutes, while Two WTC, which had 3/4 inch fireproofing, crumbled in only 56 minutes. "The insulation used [was] too little, according to our calculations," he said.

Robert Solomon, assistant vice president for building and life-safety codes for the National Fire Protection Association (NFPA), explains that fire codes specify how long fireproofing must withstand heat; they do not address thickness issues because the same thickness of different materials might produce different results.

But Solomon questions whether the thickness of fireproofing is related to how fast the buildings collapsed. He notes that Two WTC, which fell first, was struck at a lower point than One WTC, and thus the damaged Two WTC had more weight to support. The relative times to collapse "probably had more to do with the additional weight that [Two WTC] was trying to support," he says.

At the same hearing, Roger Morse, an architect who investigated the WTC's fireproofing from the early 1990s to June 2001, said that the towers suffered from the same sorts of deficiencies as many other high-rise office buildings in the United States and Europe. He noted that fireproofing on long-span joists was often "extremely thin" (less than the 3/4 inch specified in the FEMA/ASCE report) and that some structural elements were never fireproofed in places because ductwork prevented ready access. Moreover, he observed that fireproofing on the columns had been coming off because it had been applied over the rust that had built up on the columns, and the rust was flaking from the steel. (Building codes don't get into this degree of specificity, but such practices go against the manufacturer's recommendations.) Finally, he observed that inspectors assess fireproofing before construction has finished, and that fireproofing is often damaged prior to occupancy.

While other design elements can be difficult if not impossible to retrofit, Corley says that improving or adding fireproofing is feasible. As mentioned previously, One WTC had a program to replace its asbestos fireproofing with a safer alternative. As someone moved out of an office, fireproofing would be replaced. "It's not that difficult to do, but it's fairly costly," he says.


One of the problems noted at the Twin Towers was the proximity of the emergency stairwells to each other, which meant that a number of potential escape routes were cut off simultaneously by the plane's impact. "We think it's desirable to spread stairwells out, so if there is an impact...there's a better chance that the impact won't take out all the stairwells, as it did in building one," says Corley.

The analysis also suggests that in buildings that are potential targets, stairwells should be enclosed using a more impact-resistant system. Plasterboard can be used in combination with metal studs (as it was in the WTC), Corley adds, but it must be much more firmly attached. In the WTC, it was attached according to code, which required just nominal adhesion.

But Vincent Dunn, a fire-safety consultant and ex-deputy chief of the New York City Fire Department, warns that some materials that are more impact resistant, such as gypsum board or plasterboard, are less likely to hold up under streams of water. He, therefore, cautions against substituting those materials for concrete blocks enclosing stairways. "Our master streams and powerful hose streams collapse the plasterboard. [They don't] do that with concrete," he says.

Another problem that the experts noted with the evacuation was that the narrow stairways became congested, slowing egress. (Each tower had three emergency stairwells. Two were 44 inches wide, and one was 56 inches wide.)

As it turns out, a code change relating to the width of stairwells is already in the works at the NFPA, though not because of 9-11, according to the NFPA's Solomon. Stairwells that must accommodate more than 2,000 occupants will soon be required to have 48 inches of width, up from the current 44. That change recognizes that building occupants tend to be "reluctant" to comply with phased or partial evacuation, he says. NFPA codes and standards are voluntary, however, until a jurisdiction adopts them.

Some jurisdictions may not act; others may go further than the NFPA code. For example, the Fire Safety Directors of Greater New York advocates an even larger passageway. The group is calling for stair width to increase to 55 inches, according to vice chairman Jack Murphy.

Caveat Builder

Despite the nonprescriptive nature of the FEMA/ASCE report, changes in the design and construction of future buildings are inevitable, many experts say. But are they advisable?

Jonathan Barnett, a member of the study team and a professor of fire protection engineering at the Center for Fire Studies, Worcester Polytechnic Institute, points out that no collapse due to fire had ever occurred before at a fire-protected high rise. Also, he says, building codes have changed considerably since the World Trade Center was built in the sixties and early seventies, and those changes may have already addressed any code deficiencies that are found. "I'm not convinced based on [the] limited data I have that there is a need for any building code changes," he says

Similarly, Robert C. Wible, executive director of the National Conference of States on Building Codes and Standards (NCSBCS), worries that changes made in response to 9-11 may just be a knee-jerk reaction. "There are just some things that a building will not stand up to," he says. Even if code changes do emerge from this, he says, jurisdictions won't necessarily be on the same page.

Wible points out that the International Code Council and the NFPA are essentially involved in code wars, with each group vying to get its rules adopted by the 44,000 U.S. jurisdictions with building codes. Wible urges the two groups to set aside rivalries and join forces for a common good: "It makes more sense for the country to operate off of a set of coordinated codes," he says.

Magnusson, of the study team, strongly questions the need for change, calling design issues raised by 9-11 a "red herring."

Building codes don't exist in a vacuum, he says; they address specific hazards, and performance objectives are designed for them. He says that high-rise fire codes are generally adequate because performance objectives are carried out in multiple ways: fireproofing, sprinklers, emergency lighting, and so on.

"There is no performance objective for a 767 flying into a building," Magnusson says. If planes were recognized as a hazard, he continues, then building codes would have to address the most potentially destructive aircraft (if not, terrorists would just use aircraft that exceeded building code requirements, he reasons). For example, a 747-400 at takeoff could carry more than five times the fuel that each WTC tower plane contained, would weigh three times as much, and would have a much larger wingspan. An Airbus (A380 Jumbo) in the design stage is even bigger.

In addition, the 767s that the terrorists flew into the WTC blasted holes about 140 feet across in each building, which were about 209 feet across. "The fact that the towers were able to swallow up a 767 and remain standing gave a false expectation to the public [and] the media," Magnusson says.

Many tall buildings are closer to 140 feet across, he observes. "You don't need to be an engineer to know what would happen if you put a 140-foot hole in a 140-foot building." In sum, it's impossible to supply the capacity to meet performance standards for an airplane attack, he says. That threat, he advises, is better addressed in the aviation sector.

Digging Deeper

To better assess the need for changes, experts agree, more study is needed. Many of the subjects suggested for future research in the FEMA/ASCE report are being picked up by NIST, which hopes to get funding for a long-term study of the causes of the WTC collapse that could lead to improvements in the way buildings are designed, constructed, maintained, and used. NIST Director Arden Bement predicts it will lead to revisions to fire and building codes, standards, and practices.

If the project is funded, NIST, under the aegis of a federal advisory committee, will conduct a two-year building and fire-safety investigation into contributing factors to the collapse of One, Two, and Seven WTC.

For example, it will use steel recovered from the WTC site to examine the mechanical and metallurgical behavior of various grades of structural steel. Studying each building's construction, materials, and technical conditions to determine what led to its collapse "is expected to benefit buildings of all designs," according to the proposal.

Another aspect of the study would examine issues relating to fire safety, the prevention of progressive collapse, and equipment standards for first responders. Current and historical building codes from various cities will be compared as well. This task will require reviewing design calculations, building construction, passive fire-protection features, and other data.

Related aspects of the investigation will delve into structural performance and aircraft-impact damage prediction, forensic analysis of structural steel, prediction of thermal and tenability environment, and structural fire response and collapse.

Several areas of inquiry even go beyond the FEMA/ASCE report. For example, NIST plans to investigate the performance of design, construction, and approval processes used to ensure safety when innovative structural systems are used or where code variances are required.

All along the way of this multiyear effort, there will be opportunity for public input. The first of such opportunities was an all-day hearing held June 24, at which experts from the relevant fields, such as engineering and architectural design, weighed in on NIST's proposal.

Most participants commended the effort, but some experts worried that NIST was taking on too much. The University of Maryland's Quintiere contended at the hearing that "the plan is a dilution of what I believe should be the primary objective--the collapse."

As of this writing, the study is not a fait accompli. According to NIST spokesman Michael Newman, while FEMA has earmarked $16 million of a supplemental budget request to Congress to cover two years of the investigation, that supplemental funding might not be approved by President Bush.

Additional federal funding would have to be found for the parts of the investigation that extend beyond two years. At this point, Newman says, NIST is "gearing up," getting the WTC steel ready for examination, establishing guidelines and procedures for the analysis, and assembling its investigation team.

Assuming that the project proceeds as planned, four interim reports will document the findings. The first, expected to be released six months after the investigation begins, will report on the design, operation, maintenance, and performance of emergency access and evacuation systems in the WTC towers. The second report, due three months later, will document codes related to high-rise construction and structural fire safety.

The third report, due to be issued at the same time as the second, will cover regulations and practices related to the passive fire-protection systems. Finally, one month later, a report is expected to document maintenance and modifications made to the WTC that may have affected fire-protection systems.

It took months for the dust to settle at ground zero. It will likely take much longer than that for the structural and design lessons of the disaster to make their way into building codes and common practice. But in the end, it is hoped that high-rise designs and construction methods that keep building occupants safer will be one positive legacy of last year's terrorist attacks.

Michael A. Gips is senior editor at Security Management.

A Separate Code?

One oft-mentioned issue is whether a separate, more rigorous set of codes and standards should apply to extremely tall buildings. In general, in the United States a high rise is considered to be seven stories or greater. Codes and standards apply equally to a seven-story Chicago apartment building and the 1,454-foot, 110-story Sears Tower.

Robert Solomon of the National Fire Protection Association (NFPA) is one of the experts who recommends that architects and engineers consider categorizing high rises by height and perhaps tailor requirements accordingly. For example, 7- to 40-story buildings might be labeled "tall buildings," buildings with 41 to 90 floors could be deemed "high rises," and those 91 floors and more could be referred to as "supertall" structures.

There is precedent for code provisions to vary in high rises by building height. Recently, the NFPA adopted a change requiring that all buildings taller than 420 feet (roughly 40 stories) increase their fireproofing for structural members. Over the past 25 or so years, says Solomon, buildings over 40 stories required "three-hour construction," meaning that structural members would have to be protected by fireproofing that lasted at least three hours. The change will bump that standard up to four hours.

Jon Magnusson, a partner in a Seattle-based engineering practice, sees problems in differentiating codes by building height. He says that terrorists will simply switch to targets with less rigorous codes (such as short or older buildings) that still carry symbolic weight, such as corporate campuses.

Shy of the Sky

Experts stress that design improvements need not necessarily be made in all new construction. As in every other security discipline, threat, risk, and vulnerability analyses should be conducted to determine what design and structural enhancements are prudent. For example, engineer Ysrael Seinuk says that he has developers "look at all possible threats and decide which are the main ones they want to address."

Mary Lynn Garcia, a senior member of the technical staff at Sandia National Laboratories, expresses a similar sentiment. "Let's identify buildings that look like the best targets," she says. For example, a skyscraper in San Francisco is a much more likely target than a bank in Albuquerque. Sandia is trying to get relevant parties to use a consistent methodology to identify these targets, she says.

Given that tall buildings can be targets and might require more stringent protections or more target hardening of the structure, some companies are simply deciding they no longer need to reach for the sky. "You don't want to become a symbol of the city," says Adrian Smith, a design partner in the Chicago office of Skidmore, Owings & Merill. "You become a potential target." For example, partly for this reason (as well as economic reasons), he says, planners downsized a proposed 2,000-foot-tall tower in Chicago to make it shorter than buildings such as the Sears Tower and the Hancock Building.

eXTReMe Tracker