FROM ALEXANDRIA TO NEW YORK AND ON TO DUBAI, RULERS of rich cities have sustained an irrepressible urge to impress their rivals and intimidate their subjects by building bigger and taller.
One thousand years ago, churches across western Europe inherited Roman aqueduct construction techniques, arch upon arch, in the Romanesque style. When the Normans brought this design across the Channel to England, after 1066, and set about applying architectural “shock and awe” to cow the troublesome and hostile inhabitants, they built massive castles for the Norman lords and magnificent cathedrals to impose the power of their appointed bishops; these skyscrapers were two or three times higher than the Anglo-Saxon churches they replaced.1 Within a few years of the conquest, with the invaders’ colossal ambition to raise pinnacles of stone hundreds of feet in the air, thousands of English peasants had been recruited for cathedral construction as quarrymen, carters, masons, and scaffolders to work alongside stonemasons brought over from France. This was a veritable “space race” to reach for the skies.
The most architecturally challenging location for a tower was at the center of the cross-shaped plan, where the apse and choir meet the transept to create the crowning central summit of an English cathedral. To span the space, all the weight of the tower must be transferred to the four massive corner columns.
For the first 200 years, the success rate of these late-eleventh and early-twelfth-century attempts to overcome gravity was comparable to the success rate of the first US space rocket launches between 1957 and 1959—except that the effect of gravity is speeded up in the few-second trajectory of a rocket as compared with the century or two it took a 1,000-ton cathedral tower to crack through its thick stone pillars and plunge to the chancel floor.2 In England twelve colossal cathedral building projects were under way within thirty years of the Norman landings.3 Another four were initiated through the first decades of the twelfth century.4
The “half-life” of an English Norman tower proved to be somewhere between 100 and 200 years. Of the first crop of late-eleventh-and early-twelfth-century buildings, the central Norman tower at Winchester collapsed in 1202; the southwest tower at Chichester collapsed in 1210; the newly built central tower at St. Davids (in Wales) collapsed in 1220; Lincoln’s central tower collapsed in 1237; and the great Norman tower at Ely lasted until 1322. At the largest cathedrals, in fact, Norman towers survived only at Norwich and Durham (although both were significantly reconstructed in the fifteenth century).
(In November 1860, almost 500 years after it was built, jagged cracks began to appear in the piers surrounding the Chichester cathedral crossing.5 After the choir screen and altar were removed, great tree trunks were brought in to create a wooden framework to carry the colossal load. On Thursday, February 21, 1861, at 3:30 a.m., the final death throes began as one of the upper piers gave way in a series of terrifying explosions, projecting shards of shrapnel limestone across the nave before whole stones began bursting out of the columns. All the workers had withdrawn when, ten hours later, the spire and central tower collapsed into the apse.)
Faced with the early-thirteenth-century stories of tumbling Norman towers, transmitted along the “Masonic grapevine,” cathedral builders began to check some of their wilder ambitions.6 Without even knowing the cause, the news of crashes in England discouraged further high flying.
In France in the late twelfth and thirteenth centuries, cathedral builders evolved new styles of lightweight architecture with pointed arches and windows (called “Frankish” at the time and only later disparaged as “Gothic”).7 Lightweight structures were fine for handling vertical loads, as there was simply less weight to be carried. However, the mass of the thick Norman stone columns and walls served to prevent a structure from becoming pushed outward, whether by the load of a pitched roof or from storm winds.8
At the Church of Notre Dame Paris, during the thirteenth century, the lightweight walls started to suffer stress fractures and bulge out of line, inspiring the agile improvisation of flying buttresses, spider-leg props made to look as if they were part of the original design.9
As some of the largest of all structures, cathedrals were the most susceptible to the long-period ground motions of great earthquakes.10 In Lisbon the naves of almost all the principal churches and chapels of the city collapsed on November 1, 1755. Spanish attempts to build European-style cathedrals in their colonies, as at Lima, Arequipa, Santiago, Concepcion, and Manila, were all demolished by earthquakes.
IN THE MIDDLE OF THE NINETEENTH CENTURY, THE CONSTRUCTION of tall buildings underwent a revolution from the application of physics. In 1866, William Rankine, engineer and physicist, articulated the concept of the “safety factor”: “the ratio of the ultimate strength of the material to the maximum stress permissible” from the loads acting on a structure.11 Engineers should work to a safety factor high enough to ensure that engineered structures, such as bridges or tall buildings, stayed standing in all anticipated physical conditions. The weight of the structure provided the first set of loads; then there were the “live loads,” such as from trains passing over a bridge or the people crowded into a building; last were the infrequent extraordinary loads from strong winds or earthquakes.12
Instead of using traditional “assembled” building materials, such as brick and rubble walls or wooden beams, engineers switched to manufactured elements like iron and steel girders, whose strength and ability to withstand compression and tension or to resist sideways shear could be measured and predicted. Based on the physics, structural engineers could design tall structures that would remain standing indefinitely in effect. Under normal conditions, skyscrapers simply do not fall down.13
To resist storm winds, a load of 30 pounds per exposed square foot of the walls was assumed for the first tall buildings in New York and Chicago, with 20 pounds per square foot assumed for the calmer climate of San Francisco. In rebuilding San Francisco after 1906, engineers had no idea how to assess earthquake loads on a tall structure. Instead, they assumed that an earthquake raised the wind load to 30 pounds per square foot. Concerned with the extra cost, however, architects soon argued the assumed load back down to only half this level.
These assumptions were based on a misunderstanding of the basic difference between earthquake and wind forces. Wind attempts to push a building over, and the mass of the building and its foundations must be built to prevent that. An earthquake arrives through the foundations. When the mass of the building resists the horizontal movement through inertia, the building is left behind. How far and how fast it is left behind will determine whether it stays standing.
THE IDEA FOR STRENGTHENING CONCRETE WITH IRON AND STEEL rods had no single inventor but was the product of many practical artisans and engineers in France, Belgium, Britain, and America, most of whom had little interest in documenting their experiments.14 Marc Isambard Brunel in 1825 had the original idea to interlace iron reinforcing rods through brickwork on the original Thames Tunnel.15 New stronger Portland cement, fired at higher temperatures, was patented in England around 1850; rich in calcium silicates, it reacts with water to crystallize around the sand or gravel and create a strong artificial stone.
Take a cement paving stone and bang one edge down on something hard and solid and it will break like porcelain. Do the same with a bar of steel and you will hurt your arm. Unlike concrete, steel does not simply break when stretched or knocked. Yet steel is expensive, and it rusts when exposed to the air. So the idea emerged to unite steel’s strength when pulled or knocked with the enclosing rigid properties of the concrete—yang and yin—to make a strong and resistant composite medium. Concrete and steel are also strangely compatible in how much they expand or contract with temperature.
By the start of the 1890s, this experimental material, known by its French name “ferroconcrete,” had become mainstream, and the first rules for how steel and concrete should be united had been published. The composite medium seemed to have infinite potential.
Rather than devise more elaborate reinforcing to salvage brick, the solution finally emerged, like a Zen koan: keep the reinforcing but discard the brick. Instead of elaborating ways to interlace brick with steel bars, the solution was to pour a liquid into formwork shaped to create a wall, roof, or pillar and let it set as an artificial rock around the steel. A wall of concrete could be poured much faster than a wall of brick could be laid, and a reinforced-concrete building frame was far cheaper than one manufactured from steel alone.
THE SEVENTEENTH-CENTURY CAMPAIGN TO SWITCH CONSTRUCTION from wood and thatch to brick and tile marked the first round in a campaign to make the fire-resistant city. Yet, while brick walls provided fire-resistant shielding, inside buildings contained the same wooden beams, floors, rafters, and window frames as in their medieval ancestors.
Truly “fire-resistive” construction required that buildings be made with only incombustible materials, so that floors, roofs, and walls would all be firebreaks. Once it was learned how to build taller iron-and-steel-frame buildings, such structures were not allowed in the central business districts of the principal US cities until, for the safety of the occupants, they could be made fire-resistive. Initially, floors were constructed out of hollow brick arches between iron or steel beams. In the 1870s, a British architect, Richard Dawnay, laid a wide concrete floor, reinforced with iron bars and small joists, over a span of 20 feet (6 meters) and went on to install this novel type of fire-resistant flooring in more than 3,000 buildings.16
Two great city fires, in Baltimore in 1905 and San Francisco in 1906, provided the final clues in the mission to create the fireproof city. In both cities, flames from adjacent properties had gotten inside buildings believed to be fireproof through heat-shattered windows. But by 1910, once builders had removed the flammable neighbors and installed powerful sprinkler systems, supplied by rooftop tanks, the conflagrations that had consumed US city centers for more than 100 years had gone away. It was like the successful eradication of a lethal infectious disease.
IN SAN FRANCISCO BEFORE 1906, THE BRICKLAYERS UNION HAD seen the introduction of reinforced concrete as a direct threat to its members. Large quantities of concrete could be poured without having to employ the trowel skills of a Union “brickie.” San Francisco was a union town, and the bricklayers used their influence to get reinforced-concrete construction banned.17
After the 1906 earthquake, the union’s attitude contributed to the enthusiasm now lavished on the reinforced-concrete buildings that had been built in other towns around the San Francisco Bay. In Alameda, “a two story reinforced concrete building stands intact, while a brick building only 50 feet away is badly wrecked.”18 John B. Leonard wrote that “an inquiry among architects and engineers, together with my own observations, has failed to reveal any instance of failure on the part of reinforced concrete” in the whole Bay Area.19
The ban on reinforced-concrete (RC) construction in San Francisco was lifted ten weeks after the earthquake.20 Within a year, there were more than seventy RC office buildings and warehouses in the city. Apart from Edward J. Brandon of the Bricklayers Union, who urged the supervisors to reject the new building code, the only opposition came from the manufacturers of the more expensive steel frames. They had a point: none of the new iron-and-steel-frame high-rise building designs were significantly damaged by the earthquake, and while several were burned out, many were later rehabilitated.
Perhaps the most neutral and informed opinion came (once again) from the young associate professor of engineering at the University of California. Charles Derleth Jr. warned of the dangers of assuming that RC construction provided a panacea; weaknesses in the way vertical and horizontal beams were interlinked in RC structures, he cautioned, could make them prone to collapse under earthquake loads.21 His note of caution was not lost on the Japanese engineer Toshikata Sano, who returned to Japan from what he had seen in San Francisco a strong champion of properly designed RC construction. In 1915 he proclaimed that ferroconcrete was the solution to making buildings in Japan both fire- and earthquake-proof—and the answer to brick’s inadequacies.22 “In earthquake country like Japan,” Sano wrote, “construction took precedent over design” while aesthetic concerns were “only women’s things.” His proposals proved enormously influential and led to the proliferation across Tokyo of plain, two- to five-story reinforced-concrete buildings.
Sano proposed that the earthquake forces on a structure could be represented as a horizontal acceleration.23 This notion gave engineers a parameter that they could plug into their physics-based models of building design. In 1920 the architect Naitō Tachū calculated the earthquake loads in the design of all his new buildings and then had the good fortune to see his engineered buildings survive the 1923 earthquake.24
A 1920 law in Tokyo had limited buildings to 100 feet (30 meters) high. Over the next two decades, concrete cityscapes rolled out across Japan. Even up to the 1960s, there were very few buildings in Japan that reached above the sixth or eighth story.
BETWEEN 1910 AND 1920, THE MOST ADVENTUROUS ARCHITECTS adopted ferroconcrete to define a new aesthetic—a style that made a clean break with the traditions and ornamentation of nineteenth-century architecture.
In 1926 the Swiss French architect Le Corbusier, self-proclaimed leader of the modernist movement, set out his manifesto: “the five points for a new architecture.”25
1. An open first floor in which the weight of the building is carried by a small number of pilotis, thin columns that operate like timber piles supporting a platform in the sea
2. The free plan, in which there is no restriction on the positions of interior walls, as they do not need to be load-bearing
3. The free facade, an entirely open front
4. Strip windows—large sheets of glass in the curtain walls
5. Roof terraces on the flat concrete roofs
These innovations were all facilitated by reinforced concrete, a medium that enabled exhilarating architectural freedom. No longer did a room have to have four walls, or doors and windows that occupied only one-quarter of the wall space, or walls directly above the walls on the floor below. The Villa Savoye in Poissy, constructed between 1928 and 1931, combined all five points of Le Corbusier’s new “international style”: a raised, white, single-story rectangular structure that had long strip windows and “floated” high off the ground on thin concrete pilotis.26
The new architecture would redefine how people lived. In 1922, Le Corbusier introduced his vision of the “Contemporary City”: sixty-story skyscrapers, arranged on a grid, housing 3 million people, and surrounded by open spaces.27 In 1925 he proposed that eighteen of these “Croix de Lorraine” cruciform-plan buildings replace the cluttered center of Paris north of the River Seine. In 1933 he launched his “Obus” proposal, in which the corniche at Algiers would be graced with a 9-mile-long (15-kilometer-long), fourteen-floor apartment building, housing 180,000 people, and capped with a rooftop highway.28
There was one small problem with Le Corbusier’s new aesthetic: it was deadly in earthquakes. It took forty years for the flaw to become exposed. The only mass housing complex he did actually complete was in Marseilles, immediately after the Second World War, but by then a whole generation of his architect disciples had gone out to design and build the massive, concrete, high-rise public housing projects that came to characterize cities of the world in the 1950s and 1960s.
The role of the open floor in building collapse was first spotted in the aftermath of the Magnitude 6.5 earthquake close to Caracas, Venezuela, in 1967.29 Four multistory apartment buildings collapsed and 240 people died. The Palace Corvin had an H-shaped plan, with two identical rectangular buildings (the verticals of the H) joined in the center by a “circulation block” of staircases. One of the two buildings had an open ground floor of columns to create space for parking, while the other building had walled apartments all the way through the ground floor. The building with the open ground-floor parking space collapsed. The other building was completely undamaged. The lesson could not have been clearer: the open floor had created a fatal weakness.
The four stories of the 1960s Olive View Hospital in Sylmar, on the northeast edge of Los Angeles, were supported by open piloti columns on the lower floors. In the 1971 San Fernando earthquake, the columns were so badly damaged that the whole building had to be demolished.30
Le Corbusier’s first three points all referenced the open floor, straddled by his favorite pilotis. The idea of the open floor caught on, not simply for its aesthetics but because of its practicality. Elevating the building created space: both public space for pedestrians and private space for car parking, shops with their large glass facades, and reception areas in hospitals, restaurants, and hotel ballrooms. The open ground floor, made possible by reinforced concrete, became a standard functional component of urban design.
When a building with an open ground floor swings backward and forward, however, all the motion concentrates on the open floor, stressing the concrete pillars or the remaining walls, potentially all the way to the breaking point. If the pillars break, the whole building can collapse. The open story is “soft”: lacking in rigidity, it is the weakest link. Engineers named this behavior the “soft story failure.”
However, it took the earthquake damage in Mexico City in 1985 before the scale of the problem was recognized.31 The city center buildings had not been designed in anticipation of strong and slow earthquake shaking. A large, long-lasting, distant earthquake off the shore of central Mexico caused the former lake basin at the heart of Mexico City to oscillate like a drum, about once every two seconds—the same resonant frequency as ten- to twenty-story buildings. As a result, 400 multistory buildings partly or completely collapsed. At the Tlatelolco apartment complex, the fifteen-story Nuevo León building pulled out of its foundations on one side and toppled to the ground like an overturned tree. At the steel-frame Conjunto Pino Suárez building, a twenty-one-story tower suffered column buckling and took out the adjacent fourteen-story tower in its fall. The most common cause of collapse was found to be inadequate linkages between columns and beams and between columns and slabs. Many buildings were overloaded relative to the strength of their supports.
The evidence for soft story failure was now so pressing that in 1987 and 1988 it was included in the next generation of Mexican and US building codes, to be applied for new construction.32 These codes stressed the need to check the “degree of vertical structural irregularity.” Where shear walls form the main resistant elements of the upper floors of a building, they need to be continuous all the way down to the ground level; “to interrupt this load path is a fundamental error.” In the Bay Area (“Loma Prieta”) earthquake of 1989 in California, half of the homes that became uninhabitable were the victims of soft story damage.33
It was fortunate that, because he worked principally in Europe, the buildings that Le Corbusier had himself designed rarely encountered strong shaking. This was not the case, however, for the National Museum of Western Modern Art, established in Tokyo in 1959; based on one of Le Corbusier’s designs, it included his familiar piloti structure of open columnar supports on the ground floor. In the early 1990s, a leading Japanese professor of earthquake engineering commented: “My apologies to Le Corbusier but the evaluation of seismic capacity of this building was terrible. [We] had to conclude that this building survived thus far only because no big earthquake had hit the building. If left untreated visitors may be in danger.”34 Unable to transform the piloti structure without ruining the design, and “because everyone admitted that this architecture was successfully meeting various emotional requirements that are deeply associated with our own time,” in 1995 the Japanese added an expensive “base isolation system” beneath the building’s foundations so as to prevent the vibrations of the earth from demolishing the structure.35
Meanwhile, as the epitome of twentieth-century modernism, buildings with open ground floors have continued to proliferate. In every city from Ashgabat, Turkmenistan, to Lima, Peru, there are now multistory offices above storefronts with thin load-bearing pillars or twenty-story tower blocks perched on stilts. In San Juan, Puerto Rico, there are fifteen-story office buildings on top of five stories of open car parking. There seems to be no place in the language of architects and town planners commissioning more and more open ground floors in their designs for the vocabulary of earthquake engineers seeking to outlaw “vertical structural irregularity.”
BY THE 1980S, BETTER RECORDINGS OF ACTUAL EARTHQUAKE VIBRATIONS, allied with computer modeling, had transformed the understanding of how earthquakes damage buildings. A building has a “resonant period”—the pulse at which it will naturally oscillate (like rocking a dead tree). If the earthquake is big enough and the vibrations last for tens of seconds, the vibrations may tune in to the building’s natural resonant period, and then the shaking can become amplified with each vibration.36 Moreover, the clashing resonant vibrations of adjacent short and tall buildings can smash up against one another. Close to a breaking fault, the ground can move by several meters and at speeds sufficient to threaten the elastic response of a tall building.
If the earthquake was generated by vertical movement along a fault, it will radiate vertical accelerations. As the ground lurches upward, the building is launched as if on a trampoline, only to come crashing down again, so that the whole weight of the overlying structure can be doubled on the floors below. In addition, all the individual components of the building frame will be moving, concentrating stress at the joints. A rectangle of beams will skew one way and then the other. If a floor beam is simply resting, the ledge may move enough to let it fall. Stresses will be raised where load becomes concentrated on narrow columns.
Engineers learned that there are two ways of making frame buildings earthquake-resistant. In the first, the steel or heavily reinforced concrete “moment-resisting” frame of the building is so strong, and the beams and columns so rigidly connected, that it holds together through the motions.37 If flexed too far, the frame components are expected to buckle rather than break. The floors and walls are tied to the frame, but internal panel walls may become smashed as the frame flexes and distorts, so the panels must not be made so strong that they crack the adjacent beam. An alternative way to resist sideways forces is to build strong shearwalls: out of vertical plywood sheets in a wood frame house, or from masonry wall panels compressed within cast-in-situ reinforced concrete beams, thus creating a honeycomb of walls and beams that resists distortion, as all the components are tightly locked together.
Overall, since the lessons acquired in Mexico City in 1985, the most modern high-rise buildings have proved robust in strong shaking. No tall buildings collapsed in the 1994 Northridge, California, earthquake (although months later it was discovered that one in twenty of the rigid welded joints of steel frame buildings had broken).38 In the 2010 Chile earthquake, only one high-rise building collapsed, after swaying so strongly that it tore off its cavernous foundations.39
Yet no one really knows how high-rise buildings will perform when tested in great earthquakes, especially within a few miles of a big fault rupture. The number of tall buildings continues to rise, and hundreds of them are located in coastal cities exposed to the largest earthquakes. There are more than 1,100 in Tokyo, 650 in Vancouver, 600 in Taiwan, and almost 300 each in Lima and Panama City (where the towers are particularly slender).40
THE KEY TO GETTING CONCRETE AND STEEL TO UNITE THEIR YIN-AND-YANG properties is “detailing.”41 Detailing defines precisely what steel reinforcing rods must be used and how they are to be arranged—the inner metal skeleton supporting the rigid concrete matrix. The twisted steel rods need to be of the right width—not too fat and not too thin—to strengthen the concrete beam or slab. They cannot be placed too close to one another or the beam could split. Bars near the edge of a beam could lead to the concrete “spalling”—breaking off in flakes. The vertical steel bars inside a column need to be bound to one another with confining stirrups so that one bar does not get pulled out of place when the beam is stressed. Where the beam is longer than the steel, the bars must be carefully overlapped and spliced. The detailing becomes most critical at junctions, where a column meets a horizontal concrete beam, or a column supports a slab, or a balcony overhangs from a wall.42
Today the configuration of steel rebar is specified by CAD software, which models the stresses and strains to which the concrete structure will be subject, both under the live loads in the building and under earthquake loads. The contractor then has to create the skeleton of steel bars according to the engineer’s precise specification. Bending and cutting thick steel rods into the right set of lengths and angles requires a large and expensive machine. The steelwork is laid in position by the contractor, and the engineer returns to check that it has been done exactly as specified. The engineer tests whether the fresh concrete mix is the right composition and ensures that it has not started to set; then the mix is poured into the temporary shuttering, vibrated down to eliminate air pockets, and allowed to slowly set.
You get the picture! This is complicated stuff, more difficult than the most elaborate Escoffier recipe, and with far more serious consequences than a disconsolate diner if the recipe (the detailing) goes wrong.
The engineers have to check each other’s work and keep detailed records. Structural engineers have spent four or five years in college to acquire strong backgrounds in mathematics and physics and experience of construction projects. Indeed, it was the challenges of working with reinforced concrete and steel that created the profession of structural engineer.
ASK ANYONE: WHAT IS THE DEADLIEST WHITE POWDER SOLD ON the street, first synthesized in the nineteenth century? Heroin? Cocaine? they may suggest. Crystal meth?
The answer? Portland cement.
Worldwide hundreds of thousands have already died and more than a million more will inevitably follow over the next forty years. Yet the majority of people entombed in their houses never bought or touched the powder, but were simply innocent secondary victims.
Since the middle of the twentieth century, the mountain roads and desert tracks of Turkey, Iran, and northern Pakistan; the jungle trails of Venezuela, Guatemala, and Ecuador; the barren steppes of Tajikistan and Azerbaijan—all have echoed to the roar of overloaded lorries, straining uphill, stacked with bags of cement. There are now few villages in Asia or Central America that have not seen a smoke-spewing, open-backed, dust-covered truck with broken shock absorbers, dumping heavy paper sacks of cement to be sold in the local builder’s yard.
Before 1960, cement was available in developing countries only for big government-financed infrastructure projects like bridges and dams. In 1947, there were four cement factories in Pakistan, producing half a million tons each year. By 2010, production was 44 million tons, a nearly hundredfold increase in sixty years. In many countries there is now a cement glut, which keeps prices low. In place of patronizing a local lime-kiln for a key building material, or digging out a supply of mud, villagers now purchase cement off some distant corporation. Yet cement has also brought modernity and globalization. Small-town builders can create five-story apartment buildings and the large shop window openings they have only previously seen in the movies.
There is only one problem with the wild architectural dreams triggered by the white powder. No one in these provincial towns knows how to build safely with cement. The same artisan builders who mortared stones or bricks to make a wall have now accommodated themselves to the new material. Steel reinforcing is a particular mystery. Steel is expensive, and most local builders have no chance of finding the equipment to bend steel rods for detailing at connections.
Throughout the twentieth century, a widening schism opened between rich and poor countries around reinforced concrete. In rich countries, concrete construction designed using the latest finite element software is supervised by qualified engineers. In developing countries there are no engineers. Anyone bright and motivated enough to become an engineer goes to college and emigrates to America or Dubai or Germany. Engineers expect to be well paid.
If reinforced concrete were no more than a formula, there would be some chance that the recipe could be followed anywhere in the world. For laying a concrete path, it does not really matter if the aggregate is laced with sea salt or the proportion of cement is cut by half. But for an eight-story building, what is happening inside each beam and joint matters. It is impossible to understand what “detailing” accomplishes unless the designer has been trained as a structural engineer and understands about materials and forces and failure tolerance. Once the concrete is poured, the detailing is completely invisible—unless and until the building collapses. (I have seen drink cans, cigarette packets, and purely cosmetic offcuts of rebar with no concrete attached.)
After the modest February 29, 1960, earthquake at Agadir, Morocco, in which 10,000 died, J. Despeyroux, the French government chief engineer, visited the city and in his report triaged the buildings into three classes: (1) “the old or poor masonry houses”; (2) “the smaller city-like buildings, erected without any technical care”; and (3) “the Modern European Buildings.” “The first two classes,” he wrote, “are of no interest for our purposes.”43 In a 230-page report from the US Earthquake Engineering Research Institute on the 1972 Managua earthquake, only two were devoted to the “ordinary buildings” that were the source of almost all the 11,000 fatalities.44 From the perspective of a structural engineer writing for other engineers, there is nothing to learn from these dangerous informal buildings. Only the “Modern European Buildings,” Despeyroux proclaimed, “held any useful lessons that can be taken and applied by engineers in future buildings.” Among the thousands of engineers who visit earthquake-damaged cities, almost all ask this same question: what engineering lessons can be learned? It would take orthogonal thinking to ask instead: what action(s) would most improve the construction of ordinary buildings in this country?
The lessons on building failures collected on earthquake field missions in Pakistan or Haiti today are pretty much identical to those gathered fifty years ago (so much so that structural engineers get bored witnessing them). It all comes down to bad design, bad execution, bad reinforcing, and bad concrete. The schism between the rich world and the developing world over reinforced-concrete construction has been known for seventy years. After the 1944 earthquake demolished the adobe city of San Juan in Argentina, a local judge demanded that all future construction be entrusted to engineers, “not builders with no technical knowledge or homeowners themselves.”45 To allow an untrained builder to create a building out of concrete without the attentions of a structural engineer is like giving a child a loaded automatic weapon.
You can get some idea of the impact of the great wave of concrete construction in lower- and middle-income countries by surveying what has happened in earthquakes. In the Magnitude 7.3 earthquake that hit Adapazari, Turkey, in 1967, eighty-six people died. The majority of residential buildings were of wood and stone construction. They took some damage but generally stayed standing. In a similar-sized (Magnitude 7.4) earthquake that broke thirty years later, on August 17, 1999, to the west of the 1967 shock, at Kocaeli 40 miles (60 kilometers) south of Istanbul, 115,000 buildings collapsed and 17,118 died. The fallen buildings included more than 3,000 total “pancake” collapses of unlicensed five- to seven-story concrete apartment blocks, built within the previous thirty years.46 (Above that height an elevator would be required, inviting the attentions of an engineer.) In the 1999 Chi-Chi earthquake in Taiwan, one-third of the more than 2,000 people killed were in reinforced-concrete structures as much as ten to fifteen stories high. Only buildings above 165 feet (50 meters) in height were required to have their structure reviewed by an engineer.47
I HAD A MORNING MEETING IN A BALTIMORE HOTEL TO TALK TO insurers about storm risks. The plan was to leave at lunchtime to take a train to catch an afternoon flight from Newark. (I still have my unused United Airways Newark–to–San Francisco ticket from September 11, 2001.)
About 9:00 a.m., I was setting up a presentation when someone beckoned us to come and watch the TV in the hotel bar. Transfixed by what was unfolding, I recall the extraordinary moment close to 10:00 a.m. when, in the background, the South Tower of the World Trade Center began its terrible descent. The reporter, speaking to the camera, had yet to notice. The avalanche–dust-storm–squall-front surging through the canyon streets looked like the collapse of a volcanic plug.
The next day I took a train to Manhattan and slipped through the police barriers to check out how far the collateral damage had spread. The thick drifts of dust I tramped through had carpeted the streets and looked just like the aftermath of an eruption.
The collapse of the World Trade Center towers was shattering and extraordinary. People in northern Europe, the United States, and Japan go through a lifetime never witnessing a building collapse. Yet, if you are one of the 12 million inhabitants of Cairo, the collapse of a building would seem so everyday an occurrence as to be hardly worth retelling. Like car accidents, building collapses in many locales gain local headlines only when more than a handful of people have been killed.
The collapse of an eleven-story building in Alexandria in June 2012 inspired the first-ever survey of spontaneous demolition in Egypt. Over the following twelve months, almost 400 buildings collapsed in that country, killing a total of 192 people.48 These collapses are what structural engineers call “intensity zero damage”—“intensity zero” because there was no earthquake. The toll only increases when some ground shaking is added to the mix. On October 12, 1992, a modest (Magnitude 5.9) earthquake 21 miles (35 kilometers) south of Cairo resulted in the collapse of 200 buildings in the city, including one fourteen-story building in which seventy-two died.49 Many of these buildings would have collapsed anyway in time, but they would have done so with more crackings and grumblings in the fabric of the structure—these symptoms, which typically extend over days or weeks and indicate that a collapse is under way, serve as warnings for the inhabitants to move out ahead of the inevitable. The 1992 earthquake “harvested” these buildings early, like forest treefall in a winter storm. Almost 4,000 buildings took some structural damage; before embarking on their inevitable downward journey, they were classed as “severely damaged.” Three thousand families had to be rehoused.
Why are buildings in Cairo so prone to fall? For every twenty-five buildings in Egyptian cities, only two are officially registered, and only one does not violate the city’s building code in one way or another. A building may start off legal but then drop out of the system. The fourteen-story building that collapsed in the earthquake was a licensed eight-story building to which the owner had informally added six additional stories, almost doubling his rental income.
In 1999, Salah Hassaballah, the former minister of housing, claimed that three-quarters of the buildings across Egypt might collapse in the next twenty years as a result of “insufficient maintenance.”50 Ten years after the 1992 earthquake, the Ministry of Housing identified 180,000 buildings in Cairo and 2 million nationwide “on the brink of collapse.”
Cairo is not alone. Many other crowded cities in the developing world suffer from “falling building sickness.” In Nigeria, the city engineer’s office can only keep tally of the collapses: one every fifty-two days between January 2000 and June 2007.51 Spontaneous demolition does not discriminate: apartment buildings, houses, hotels, a mosque, and several schools have all fallen down. In one 1999 school collapse at Port Harcourt, fifty students were killed as the owner was adding extra floors during the school day.52 But this is not just an African problem.53 In any country without a well-enforced building code, even new buildings collapse.
Four thousand years ago, Babylon was the original crowded thriving city: space was tight within the protective city wall, and the city was beset by falling buildings. Around 1700 BC, King Hammurabi set out the responsibilities of a builder for his construction defects: “If a builder builds a house for someone, and does not construct it properly, and the house which he built falls in and kills its owner, then that builder shall be put to death.”54 Building collapse had become a hot-ticket political problem. Hammurabi’s edicts are the oldest surviving “building regulations”—acknowledgment that a building is a loaded weapon for which the builder maintains the liability.
Building collapse was a problem for all ancient towns and cities. In a reflection on whether those killed accidentally are innocent, Jesus asks: “Or those eighteen, upon whom the tower in Siloam fell, and slew them, think ye that they were sinners above all men that dwelt in Jerusalem?” (Luke 13:4). The Siloam tower tragedy was evidently common knowledge to his audience. Building collapse was the most salient example of mass accidental death, just as an air crash might be to us today. Around 500 years earlier, Simonides of Ceos is said to have invented the “Art of Memory” when challenged to identify all the other diners who were crushed by a collapsed building after he alone had stepped outside of the banquet hall.
In the cities of ancient Greece and Rome, the most important civic and religious buildings—the temples and triumphal arches, the Colosseum and the Pantheon—were “overengineered”: they were made more robust than was strictly necessary by builders who would have done anything to ensure that their reputation did not collapse with the structure. As a result, many were built so strongly that they remain standing today. Yet these massive public buildings were not typical of the residences of the majority of people in Rome: crowded six- to nine-story insula tenement buildings that were prone to collapse, just like the buildings of ancient Babylon or twenty-first-century Cairo.55
Beyond threatening the life of the builder, a building code requires the rule of law and only works if there are penalties when the code is ignored. Officials have to scrutinize all building plans and visit construction sites to check for compliance. Inspectors need to be vigilant and to have not only impeccable ethics but sufficient salary, so that they cannot readily be bribed. Buildings have to be torn down if constructed outside the code. All of these conditions sound possible in a rich country.
Many poorer countries have acquired building codes, but in most of them these codes are not applied. This is how it works in the informal economy: someone constructs a building without official authorization or planning approval. The builder meets an inspector in a café. Coffee is drunk and cigarettes are smoked, some money changes hands, and now there is a certificate to show that the building is “authorized” and complies with the regulations.56
If you do not have the capacity to borrow a building code, you are inevitably going to have problems with buildings and disasters. Corruption and earthquake casualties are both manifestations of weak governance, and in a corrupt country, construction is the most corrupt sector.57 In 2010 Haiti, so poor and ungoverned, there was no building code, and Haiti is number 175 out of 178 on Transparency International’s list of the most corrupt countries.
Spontaneous collapse reveals how the building stock will perform in a big earthquake. On November 7, 2008, a school collapsed in Pétionville near Port au Prince, leaving eighty-four dead.58 Four days later, with the media still focused on the Pétionville story, it was reported that at least nine people were injured when a private school collapsed in the capital. Fortunately, the schoolchildren were outside at break.
In the shantytown of Cité Soleil, home to more than a quarter of a million of the poorest of the poor in Haiti, the houses have cement walls with corrugated metal or polythene roofs—scavenged materials that are too insubstantial to cause much injury in an earthquake. With a little more money, people could pay to escape damage from the frequent hurricanes by living in blockwork buildings with concrete roofs. In the absence of any controls, the poorest lived at the bottom of the steep slopes of the ravines, where the sewage flowed when it rained, while those at the top of the slope even had balconies for the view. When the slab roof of one house was liberated from a collapsing wall in the 2010 earthquake, it slid into the next house, and then so on down the slope, all of them collapsing like houses of cards. The casualties completely overwhelmed the municipal authorities.59
Around 30 percent of the 400,000 concrete buildings were repairable, and aid agencies brought in structural engineers to supervise repairs on 10,000 of them.60 In the weeks following, desperate to escape from the temporary encampments, the majority of owners returned to patch up the cracks gashed through their concrete roofs and walls, wounds that will be exploited the next time the city is strongly shaken.
A plate boundary fault cuts through the hills on the edge of the city. After the 2010 earthquake, Port au Prince should have been relocated, but there was not enough government even to propose the option. Like New Orleans after Katrina, both emotionally and politically, the city is too big and has too much history and pride.
TENS OF MILLIONS OF FRAGILE CONCRETE BUILDINGS HAVE BEEN constructed in earthquake zones. Multistory buildings, poised on weak and unconnected supports, with all their weight suspended in the air, are weapons of mass destruction, raised blunt guillotine blades, and all it will take is the wall props moving and the concrete floors and roof beams slipping for the deadly weight of these buildings to be sent crashing down, potentially killing everyone inside. Hundreds of thousands of buildings will collapse in earthquakes over the next twenty years. Many of the people these buildings are going to kill are already living in them.
We can attribute responsibility to the slogan: “It’s not the earthquake that kills you, it’s the builders.”
HOW WOULD WE NOW REWRITE “THE STORY OF THE THREE LITTLE Pigs and the Big Bad Wolf” to reflect earthquake construction in poor countries?
The first little pig meets a man with a load of mud bricks and rubble stones and constructs a heavy-walled traditional house. The Quake-Wolf comes along and shakes and quakes the house, and it falls down and injures the little pig.
The second little pig meets a man with a load of cement and, without reinforcement or an engineer, creates a set of concrete columns and a heavy concrete lid roof. Then, with little effort, the Quake-Wolf shakes the building to destruction, killing the little pig.
The third little pig meets a man with a load of wooden beams and branches, out of which he constructs a boat builder’s frame of a half-timbered house, filled with rigid triangles. The Quake-Wolf shakes and quakes the house, but while some of the plaster panels fall out, the building does not fall down.