One small defining moment of pride greeted Mahendra Singh when he returned to India to evaluate the damage from the magnitude-7.7 earthquake that struck the country on Jan. 26. About 80 miles southeast of the city of Bhuj was a railroad bridge that the Virginia Tech engineering professor had helped design and build back in 1964, when he was fresh out of the University of Roorkee. And after Jan. 26, the bridge was still standing. Many other structures were not.
[At right: Researchers from the U.S. Mid-America Earthquake Center investigate a damaged building in Ghandidham, a city 100 kilometers southeast of the Jan. 26 quake’s epicenter. Image courtesy of the Mid-America Earthquake Center.]
When the earthquake hit the state of Gujarat, it turned the cities of Bhuj, Anjar, Bhachau, Gandhidham, Kukuma, Ratnar, Lodai and Kottar along with nearby villages to ruins in less than two minutes. In Ahmedabad, 240 kilometers from the epicenter, upper- and middle-class apartment buildings pancaked on their lower levels.
The quake left more than 20,000 people confirmed dead, with thousands more believed buried under the rubble.
Devastation resulted not from the size of the quake alone, but more so from inadequate construction and poor adherence to advised, but not mandatory, building codes, engineers say.
When the Gujarat earthquake hit, it ruptured a fault deep within the crust that scientists are still looking to identify. It broke the crust between 17 and 40 kilometers depth — “that’s 17 kilometers of crust that didn’t break, or at least 10. We don’t know how shallow it actually broke up to,” says William Lettis, an engineering geologist in Walnut Creek, Calif. “There is no evidence of any primary surface fault rupture or concentrated folding or warping.” The quake hit on what’s known as a blind fault and may give insight to geologists studying the New Madrid earthquakes that struck the Midwestern United States in 1811 and 1812.
“There are a great many lessons to be learned from this earthquake. We just don’t know which classroom we are in yet,” Lettis says. “The lessons to be learned are dependent on whether we are in an intraplate tectonic setting similar to New Madrid or to an active fold and thrust belt transitional to the Himalayan plate margin about 400 kilometers to the north.”
In 1819 an earthquake on a fault near the Pakistan border felled buildings in India, killing close to 2,000 people. It was this quake that eventually classified the state of Gujarat as a level 5 seismic zone, the highest level for building codes in India.
“Earthquake-resistant measures were developed as early as 1962 in the form of standard design criteria and in 1967 as an earthquake-resistant building construction code of practice,” wrote Anand S. Arya in a presentation to the 12th World Conference on Earthquake Engineering in New Zealand last year. “Since these are not mandatory by law, the construction in the informal sector proceeds without considering the safety provisions.” Lack of regulation either for planning or settlements or in the building bylaws allows buildings to be added every year without incorporating earthquake-resistant measures, Arya added.
Many of the structures in the hard-hit cities were unprepared for a quake. Many apartment complexes built in the last 20 years have no walls on the ground floor, for example. Instead, columns support the upper stories, providing a ground-floor garage to protect cars from the sun. “But really the garage is causing the problem,” says Eduardo Miranda of Stanford University’s Civil and Environmental Engineering School. When the earthquake struck, the columns buckled and snapped under the weight of the high-rise apartment structures. “This first floor disappears and the whole building comes down after that — a person with no feet can’t stand,” he says.
Indeed, Miranda says, much of the damage in Ahmadabad could be linked to column-supported buildings. Only one strong motion record could be obtained from the city, and the change in ground velocity that it measured was slightly higher than what would have been expected for a magnitude-7.7 earthquake in a city that is 250 kilometers from the epicenter. Calculations estimated that such a quake should have pushed the buildings sideways with a force equal to 6 to 7 percent of their weight, roughly what many were designed to withstand. But the one record indicated that the lateral acceleration was 10 percent of the acceleration of gravity (or g).
“Even though the peak ground acceleration was slightly higher than expected, 10 percent g does not explain the amount of damage observed in the city. The damage that took place correlated with poor construction practice,” Miranda says. “Ten percent of the acceleration of gravity is relatively small compared to something that would produce a lot of damage. The 1989 Loma Prieta earthquake was 60 percent g. The Northridge 1994 earthquake record was more than 100 percent g — that means for a fraction of a second, it’s like putting the building sideways, feeling its whole weight of the structure pushed laterally.”
Rakesh Goel of California Polytechnic State University in San Luis Obispo agrees. Most of the buildings in Ahmadabad are constructed to support only 4 to 6 percent of their weight pushed sideways, Goel wrote in a report for the Earthquake Engineering Research Institute in Oakland, Calif. “However, the extent of damage observed was significantly more than expected,” Goel adds. Much of the damage can be ascribed to the typical open-framed ground floor construction, but low quality material could also be a key factor. “In many instances, only one out of several similarly constructed buildings in the same apartment complex collapsed. … The concrete disintegrated within the reinforcement cage; when touched, the concrete appeared to have very little cement content.” Goel noted that preparing concrete with local groundwater, which has a high salinity, may have affected the concrete quality.
Mud and stone are commonly used in villages where buildings are made with material that is easily available. In Ratnar, a village between Anjar and Bhuj, the collapse of the buildings reportedly killed 300 people. Most of the houses are kutcha-pucca type houses, made of brick and mud mortar, or houses made simply of mud walls. The extensive shaking during the earthquake first cracked and then crumbled the homes. The initial P motion lasted 15 seconds while the main S waves were recorded as lasting 85 seconds followed by several minutes of lower level shaking. In Seattle, the Feb. 28 Nisqually earthquake lasted 45 seconds.
When Mahendra Singh visited India after the Jan. 26 earthquake, “I was very sad to see the damage in Ahmadabad and these areas; sad to see the whole thing,” he says. He traveled to the small village of Surajbari near the bank of the Arabian Sea on the Gulf of Kutch, about 35 miles from the epicenter, and found the old highway bridge that led across the tidal creeks severely damaged.
Nearby, the railroad he had worked on in the 1960s, as a beginning professional engineer for Indian Railways, came out of the earthquake with only its tracks misaligned. And the newest highway bridge crossing the sea’s ingress went practically unscathed.
When the railroad company constructed its bridge “we were concerned about earthquakes,” Singh says. In 1956, an earthquake disrupted railway services after hitting the city of Anjar, about 30 miles south of Bhuj. Indian Railways designed the bridge near Surajbari so that in case of shaking it could withstand 10 percent of its weight pushed sideways. The railroad company built the bridge using 62 spans of 60-foot steel girders and imported salt-free water to mix the concrete. Although the highway bridge was also built in the 1960s, it was made of reinforced concrete box-girders, making it heavier and more vulnerable to earthquakes than the railroad bridge.
After the Jan. 26 quake, the railway crew could run the trains again after only a few hours spent realigning the girders and fixing other minor damages. In contrast, all the bearings under the old highway’s box girders were smashed and one of the piers had tilted, forcing traffic to a single lane unsuitable for heavy vehicles.
Near Surajbari, about 200 yards east of the old highway bridge, a new highway bridge is nearly completed. “The new highway bridge did not have much problems in this quake,” Singh says. It uses three pre-stressed concrete girders under the deck.
Cranes hoisted the 32.5-meter-long girders, which were cast on site in a shop. Only one span in the middle remained to be completed at the time of the quake. The new bridge construction clearly indicates a better design — for highway bridges at least, Singh says.
Geological Survey of India
Mid-America Earthquake Center
IRIS Consortium site on India quake
USGS site on India quake
USGS National Earthquake Information Center