“Good Enough” Bridges Aren’t Good Enough

Bridges are built to specific standards of weight and length of use, but these specifics and the laws surrounding them need to be reconsidered as times change and traffic increases.


Present-day bridge design rarely considers the increasing number of cars and heavy trucks on the road, causing bridges to deteriorate and be rebuilt more frequently, and throwing off budgeting and advanced planning both locally and nationally.

Photo by Fotolia/dpmartin123

Content Tools

In America, our economy relies on roads, highways, and bridges. Yet the costs of construction and maintenance for these avenues climb while resources and funding drop off. In Henry Petroski’s book The Road Taken, the complex history and growth of the transportation industry is examined, as well as its flaws, its failures, and its future. From soaring structural feats that will engage engineers and economists, to the upkeep of residential roads sure to interest any homeowner, Petroski has a call to action when it comes to rebuilding our national prosperity in a way too often overlooked: infrastructure.

Bridge failures follow a cyclic pattern, occurring as they do after a prolonged period of unremarkable service. As I have written elsewhere, this period of perceived quality known as success lulls engineers into thinking they have solved all the relevant problems with bridge design and construction, and so they relax their guard, their standards, and their vigilance, meaning that they design less conservatively, build with less attention to detail, inspect more casually, and generally lower the quality of bridges being built. In the meantime, they also tend to push the limits of their technology, making longer spans and more daring new structures. The duration of success is typically remarkably close to three decades — about the extent of a professional generation — and by then complacency has become so common that any warning signs of impending disaster there might be are neither noticed nor heeded. When a major bridge failure does occur — like the destruction of the Tacoma Narrows Bridge in the wind in 1940 or the spontaneous collapse of the Minneapolis Interstate 35W bridge in 2007 — it is a clear wakeup call that complacency had indeed set in. Subsequent to the failure, engineers —  with the public and politicians looking over their shoulders — gain a new sense of urgency to design and oversee construction and maintenance with less carelessness and bravado and more quality and care. And so the cycle repeats.

Even given such a cycle, it is a fact that bridge failures are simply not very common. One major failure every thirty years is certainly no indictment that engineers and builders are without some considerable understanding of the workings of the daring structures that carry automobile, truck, and rail traffic across wide rivers and valleys. But the accusation of survivor bias with regard to praising the quality of older homes might make us wonder if such a phenomenon is also at play with regard to bridges. Were older bridges made better than newer ones? Any suggestion that they were is certainly counterintuitive. Nevertheless, we do celebrate ancient Roman aqueducts still standing in France and Spain; the countless centuries-old stone bridges still in service throughout Italy and the rest of the world; the first iron bridge, which has been standing since 1779 at Coalbrookdale, England; and, of course, the Brooklyn Bridge, which since 1883 has been faithfully serving New York City commuters.

At the same time, we see highway bridges that have served our interstates for mere decades being torn down after being replaced with newer, wider models. The multi-span Tappan Zee Bridge, when barely a half century old, was being replaced with a grand new crossing because the old steel cantilever was not considered up to current structural or functional standards or worth maintaining. And as we have seen, the east span of the San Francisco–Oakland Bay Bridge, which early in its sixth decade was damaged by an earthquake, was entirely replaced at a cost in excess of $6 billion. Did they or did they not really build them better in the old days?

The Eads Bridge across the Mississippi River at St. Louis is an iconic structure that in 2014 reached its 140th anniversary. The bridge had not been without trouble: its construction claimed the lives of fifteen men who worked in the pneumatic caissons driving the river piers down to bedrock; the bridge’s daring arches made of the new material steel were extremely difficult to complete; the bridge company went bankrupt within a year of the span being opened to traffic; part of the bridge’s east approach masonry arcade was destroyed by a tornado in 1896; railroads stopped using the bridge in 1974; and in 1991 its upper roadway was closed to motor vehicle traffic for major rebuilding. The structure was reopened to automobiles in 2003, a decade after its lower deck began to be used by Metrolink, the light-rail system that serves the St. Louis metropolitan area. Yet in spite of its vicissitudes, the Eads Bridge is a symbol of American innovation and resolve, and a clear success in its innovative use of the steel arch. It holds its own, even if as a low-profile architectural and engineering landmark in the shadow of the boldly soaring Gateway Arch.

The Eads Bridge and the associated surface and tunnel infrastructure that integrated the river crossing into the city’s existing street and railway network were clearly built to last, and they carried a correspondingly serious $10 million price tag. There were other river crossings proposed for St. Louis in the mid-nineteenth century, and had one of them been built instead of the Eads, it is not likely that it would still be standing in its original form the way the Eads essentially is. The most likely alternative was one promoted by Lucius Boomer, a bridge financier, and the engineer Simeon S. Post, whose eponymous truss was to be the main structural element used in their bridge.

Although the railroads clearly needed bridges to carry their tracks across rivers and valleys, they did not necessarily have the finances to do so at will. According to historian of technology John K. Brown, the iron truss bridges of the mid-nineteenth century were generally made only “good enough,” the bridge-building companies giving the railroads just what they wanted — and very little more. But the railroads were investing in more than rails and bridges; they were also developing heavier and heavier rolling stock. This meant that in as short a period as a decade or so bridges that had once been good enough no longer were adequate to carry the increased loads. Older bridges had to be strengthened or torn down and replaced before they collapsed under a train. Thus, the fate of a bridge built in the 1860s might be to last only into the 1880s, and the replacement only into the early twentieth century. The Eads Bridge, with its use of the stronger material steel and with its robust design, was an exception. Were it not for James Buchanan Eads’s seeking and securing investment capital outside the railroad business, his bridge, which came at a price at least twice what a good enough one would have carried, might never have been completed.

Given the way bridges were financed, designed, and built in that earlier era, there obviously can be a clear danger of employing survivor bias in trying to assess the quality of historic bridge-building technology only through its survivors. But in the case of railroad bridges, at least, no one wished to build structures that were not good enough to serve at the time. After all, no railroad would have wished to see a bridge collapse under a speeding train. Lives likely would be lost, significant cleanup and repair costs would be incurred; and since a bridge collapse would be big news, the publicity would not be good for business. A railroad may have wanted its bridges only to be good enough, but that goal could still demand a quality product, relatively speaking. Looked at from this point of view, concern over survivor bias does not apply. A bridge torn down or replaced by design because it had reached the end of its intended useful life should not be judged against examples like the Eads, which might have been considered a contemporary extravagance. And a bridge built within its budget should not necessarily be considered a substandard structure in its own context.

Today, highway bridges can face ever increasing loadings analogous to those that the railroad bridges of the nineteenth century and road bridges of the early twentieth did. How heavy a truck is allowed to use a highway or a bridge is decided not by the engineer designing the structure but by legislators yielding to pressure from lobbyists, such as those representing the trucking industry. When a bridge is on the drawing board, engineers size its beams and girders to carry a load imposed by trucks permitted by then-current laws, plus more via a factor of safety. Those laws might limit vehicles using a certain class of highway to a gross weight of, say, 80,000 pounds, with that weight typically distributed over five axles. The road and its bridges will have been designed to these specifications, but future legislators may increase the limit to, say, 110,000-pound trucks, with no change in road or bridge structures. Michigan has allowed trucks weighing up to 164,000 pounds, typically distributed over eleven axles, to use its highways without special permits. Trucks with gross weights as high as 1.5 million pounds have been able to pay nominal fees to obtain special permits to travel over Michigan roads. Factors of safety employed by the designers clearly do enable heavier loads to be carried, but if that were done routinely, highways and bridges would age prematurely. In 2014, proposed legislation to cut the weight limit in half in order to reduce damage to roads was rejected decisively by the Michigan state senate.

Because of reserve strength built into them, today’s roads and bridges are able to support greater weights than they are nominally designed for, but they will do so at a cost. Allowing heavier trucks to use a road and its bridges will cause them to deteriorate at a faster rate than anticipated by engineers, and so whatever long-range planning a state highway department may have done will have to be revised, for the road surface will develop potholes faster and the bridge structure cracks earlier than predicted. The upshot of this will be, of course, that repaving of the roadway and strengthening or replacement of the bridge will have to be done earlier than planned. This wreaks havoc with long-range budgeting, and all the advances in careful, thoughtful, and rational transportation planning are essentially nullified. Dealing with infrastructure is thrown back to earlier times, when repair, maintenance, and replacement were done reactively rather than proactively. Such predicaments can only be done away with if the different branches of government work in a cooperative rather than an autonomous way.

Reprinted with permission from The Road Taken: The History and Future of America's Infrastructure by Henry Petroski, published by Bloomsbury USA, an imprint of Bloomsbury Publishing, 2016.