Without Redundancy

In the wake of the Minneapolis tragedy, our columnist asks: As the interstate highway system ages, are other failures inevitable?

Soon after the I-35W bridge collapsed in Minneapolis this summer, news reports began offering tentative explanations. We kept hearing the phrase “nonredundant structure”—meaning that if one member of the bridge’s steel truss failed, the whole thing would give way. Newscasters repeated nonredundant as if it were a special quality, a quirk of this one ill-fated bridge, but the truth is that at this point in human history the man-made world is a nonredundant place. We’re surrounded by buildings and infrastructure that would instantly give way to the forces of gravity and nature were it not for our cleverness, diligence, and luck.

“In the past, with a lack of detailed knowledge of analysis and even materials, structures were built heavily and therefore were pretty safe and could take a tremendous amount of abuse,” observes architectural engineer Matthys Levy, coauthor with Mario Salvadori of the classic Why Buildings Fall Down. According to Levy, redundancy (or the lack thereof) is actually a marker dividing the modern age from everything that came before it. “As we’ve progressed,” he continues, “and certainly as both our materials have become lighter and stronger and our methods of analysis have become more accurate and faster, we’ve come to the point now where we can analyze and therefore optimize anything. There’s no question that lighter and lighter structures have been built.”

When the I-35W bridge collapsed, I started thinking about the history and implications of nonredundancy. I recall a pair of articles I wrote in the 1990s about the causes of structural failure in modern New York buildings. One was prompted by the collapse of a brick facade attached to a 39-story postwar building in Midtown Manhattan. The brick wall was not designed to be load-bearing. It was just a veneer connected to the frame of the building with thin ties of corrugated metal spaced, in theory, at regular intervals. Except the building’s contractors had left out two-thirds of the specified ties, something the new owner soon discovered when he decided to punch windows in the wall and bricks began showering Madison Avenue.

Around the same time, I reported on the Whitney Museum’s prudent decision to reinstall the 1,500 600-pound granite slabs that make up its facade. An inspection had revealed that the hardware used to hang the massive stone slabs was not to specification. The project’s restoration architect, Diane Kaese—then with engineering firm Wiss, Janney, Elstner Associates—told me what she discovered behind the walls: “When we took off the skin, we found galvanized steel, brass, stainless steel, and regular carbon steel. We found painted steel, unpainted steel. You name it, we found it. It was crazy. It was wacko.”

In the course of my research I spent a lot of time talking with forensic engineers and discovered that they all had horror stories about shoddy building practices. They suggested that buildings erected during a boom, like that of the 1960s, are less likely to be built to specification because contractors have an incentive to work faster and cut corners, and busy inspectors are less likely to catch problems. Theodore Prudon, an architect who taught a course at Columbia University called “Building Pathology,” also emphasizes the “disjunction between aesthetic requirements and technology” that occurred in the 1950s and 1960s. “You want a certain look,” he says of that moment, “and you don’t know how to build it yet.” But perhaps there was an even deeper issue: Could all this dysfunction be a direct product of the postwar mainstreaming of Modernism? Maybe it’s what happens when the rules change—when you tell craftsmen that craft no longer matters, when a society shrugs off the idea of permanence.

The I-35W bridge was completed in 1967 as part of the interstate highway system. Most of it, more than 46,000 miles of roadway, was built between 1956 and the early 1980s. Of the more than 45,000 bridges that were constructed during that time, 24,314 went up in the 1960s. Not all of the bridges were truss structures, and not all of them were as long as the one in Minneapolis, which spanned the Mississippi River. But all were surely products of a confident, exuberant period in our history. The engineers who built the interstate were the contemporaries of the NASA engineers, but instead of sending a man to the moon, they were sending streams of traffic into America’s cities and countryside.

The I-35W bridge is not the first interstate span to fall in a river. In 1983 a section of the Mianus River Bridge, a cantilevered overpass built in the 1950s that was part of heavily traveled I-95 in Connecticut, separated and dropped, taking two cars and two semitractors into the water with it. Accord­ing to Why Buildings Fall Down, the postmortem revealed that at least one crucial bit of hardware, the metal caps intended to prevent the hangers that held the roadway from slipping off the steel pins on which they were suspended, were “flimsy”—half the width called for in the 1957 design regulations. Worse, the hanger assembly was rusted because, a decade before the collapse, the roadway was repaved and the drainage system paved over, allowing water to get deep into the joints. The structure’s design may have been good on paper, but it was undone by heedlessness and neglect.

It’s hardly surprising that the Mianus investigation turned up substandard hardware. Dr. Fabian Hadipriono Tan, a professor of civil engineering at Ohio State University and coauthor of the 2003 paper “Analysis of Recent Bridge Failures in the United States,” says those structural connections “are the most vulnerable places.” Indeed, the New York Times reported that there might be a problem with the I-35W bridge’s gusset plates, which are used to reinforce the connections between the steel members. Widely used through­out the interstate system, gusset plates were attached in the 1960s with rivets. Later, in the 1970s, designers switched to bolts, which are stronger.

Since the I-35W collapse, there have been endless editorials about our neglected infrastructure and how the trillions of dollars we’ve blown on the Iraq war could have been spent on maintaining our roads and bridges. I can’t argue with any of that. But I wonder whether we’re now on the edge of a situation that is going to get worse as our most traveled highways, roads originally designed to handle traffic loads 20 years into the future, turn 40 or 50. It seems clear that as the interstate highways and the 55,315 bridges that are part of the system age, failure becomes increasingly likely. Of course, we can, as Hadipriono suggests, scrutinize these bridges with “a new attitude.” He stresses the need for highly trained inspectors equipped with “strain gauges, lasers, and optical devices.” Better still, now that we’ve finally figured out that technology and craft are not mutually exclusive, it might be a good time to start aggressively replacing what was built during the interstate years. Perhaps we could even take this opportunity to incorporate, as the mayor of Minneapolis has suggested, mass transit. A more enlightened federal government might even regard the decay of the interstates as a blessing in disguise, a chance to rethink long-distance ground transportation.

As for the lack of redundancy, maybe it’s not a structural situation but a philosophical one. May­be the problem that caused the I-35W bridge to fall isn’t just a bad gusset plate, or the effect of 40 years of road salt on steel trusses, but the result of a grand infrastructural undertaking launched at the exact moment when we stopped building things to last.

Find out more facts about this subject on the Reference Page: October 2007

Recent Programs