Stanford University

News Service



Casualty, damage estimates of great quakes revised upward

STANFORD -- If the worst possible great earthquakes struck San Francisco, Los Angeles or Tokyo, economic losses would be substantially higher than previously estimated, ranging between $170 billion and $3.3 trillion. In the two U.S. cities, loss of life also would be greater than earlier estimates.

The upward revisions are due to the incorporation of the latest information from the 1994 Northridge earthquake in California and the 1995 Kobe, Japan, earthquake into a detailed analysis performed by Haresh C. Shah, professor of civil engineering at Stanford, and associates at Risk Management Solutions Inc. (RMS) of Menlo Park.

According to the studies:

  • A repeat of the famous 8.3 magnitude 1906 earthquake in the San Francisco Bay Area could result in 3,000 to 8,000 deaths, 8,000 to 18,000 serious injuries, and a total economic loss of $170 billion to $225 billion.
  • A magnitude 7.0 temblor on the Newport-Inglewood fault in the Los Angeles basin could result in 3,000 to 8,000 deaths, 11,000 to 20,000 hospitalizations, and a total economic loss of $175 billion to $220 billion.
  • Reoccurrence of the 7.9 magnitude Great Kanto Earthquake, which destroyed Tokyo in 1923 and killed 140,000, could result in 30,000 to 60,000 deaths, 80,000 to 100,000 serious injuries, and economic losses totaling between $2.1 trillion and $3.3 trillion.

These estimates represent worst-case scenarios. They assume that the largest earthquakes considered likely to occur in the next 30 to 50 years would strike as close as possible to the most highly developed areas in the three locations.

The new figures represent a substantial increase in estimated economic losses from those contained in preliminary estimates produced by Shah and RMS that were released in August 1994. Casualty estimates for Los Angeles and San Francisco are also higher as well. The revisions were the result of evidence that the ground can shake more violently at the center of a quake than experts had thought, and a recognition that demand for construction materials and labor after a major quake will inflate the costs of reconstruction.

"We now know a lot more than we did before Northridge and Kobe, which were the first large earthquakes centered under major urban areas in recent times," said Shah, the Obayashi Professor of Engineering. "Two years ago, we didn't know for sure how bad the ground shaking would be in the area right above the fault. We had estimated that peak accelerations might reach as much as 60 percent that of gravity. Now we know that it can be more than 100 percent. As a result, even modern structures could be at risk," Shah said.

Measurements taken at Northridge and Kobe also indicate that, in addition to the shaking being stronger, the speed with which ground moves near the earthquake's center can be greater than experts had thought. These large velocity pulses seriously damaged steel structures that seismic engineers previously had thought were safe, Shah said.

As a result, Shah and RMS were forced to increase their worst-case casualty estimates from a range of 2,000 to 6,000 deaths to 3,000 to 8,000 deaths in San Francisco and from 2,000 to 5,000 deaths to 3,000 to 8,000 deaths in Los Angeles.

Estimates of the overall economic costs of the two great quakes in the United States also were revised upward with the cost rising from between $115 billion and $225 billion to between $170 billion and $225 billion for the Bay Area and from between $125 billion and $145 billion to between $175 billion and $220 billion for Los Angeles.

According to Shah, the increased intensity of projected ground shaking was a major contributor to this increase, but there was a socioeconomic factor as well. In previous studies, they had based their estimates on the cost to repair damages on current material and labor costs.

"But Northridge and Kobe illustrated that, following such a natural disaster, the cost of material and labor rises substantially," Shah said. As a result, the study includes significant increases in the estimated costs of materials and labor during the recovery period.

A trillion-dollar earthquake

In the case of Tokyo, casualty estimates remained largely the same, but the estimated economic losses nearly doubled, jumping from the range of $800 billion to $1.2 trillion to $2.1 trillion to $3.3 trillion.

The epicenter of the worst-case earthquake for Tokyo would occur under Tokyo Bay, not the city. As a result, there was no change in the study's estimate of ground shaking.

On the other hand, the material and labor inflator was applied to the Tokyo projection. That and several other factors brought to light by the experience at Kobe were responsible for the large increase in the economic loss estimates, according to Shah. These factors include:

  • Recognition that older Japanese houses have a much higher failure rate than previously estimated because they have very heavy roofs, designed to withstand typhoon winds, that make them more likely to sustain damage in an earthquake.
  • Increased estimates of the damage that would be sustained by buildings constructed between 1970 and 1982, when the seismic features in the Japanese building code were much looser than they are today.
  • Acknowledgment that the population density in the area is so great that Japanese emergency response efforts would be overwhelmed and, as a result, largely ineffectual for several days after the event. "This was the case in Kobe, and Tokyo would be Kobe multiplied by a factor of 20," Shah said.
  • The discovery that a number of residences had not been included in the study's original database, which was obtained from Japanese insurance companies. The researchers learned that a large number of houses in Kobe and Tokyo are insured by Zankyoran, an insurance group that is an arm of the Ministry of Agriculture.
  • The discovery that the commercial losses that are likely to occur are substantially greater than originally calculated based on insurance industry estimates.

New method for estimating quake damage

The Shah/RMS estimates are based on a new and potentially more accurate methodology than that used in previous studies.

In the past, such figures were prepared by doing detailed analyses of the buildings in several small but representative neighborhoods and then statistically generalizing the results of these small-scale studies to the entire area that would be affected.

Shah and RMS, however, have used a relatively new computer-mapping technology, called a geographic information system. It allows them to combine information about individual structures, such as their age, type and assessed value, with geological conditions - such as slope and soil type - that influence the behavior of the ground during an earthquake. In this fashion, they can estimate the amount of damage that every building and structure in an area might sustain if an earthquake of a given size occurs in a given location.

"Because it works with many of the largest insurance companies, RMS has developed the most significant database of its type in the world," said Tom Hutton, company president.

According to Shah, having such a database is very important. "For any one house we can easily be off by 100 percent. For 10 houses of the same type, we can do somewhat better. And for 1,000, we can do pretty well." Despite all the effort that goes into such estimates, however, uncertainties of less than 50 percent are considered quite good, he acknowledges.

"These estimates are so big, and include so many factors, that I won't feel comfortable until our work has been independently confirmed by other groups," Shah said.

Several groups in Japan are beginning independent studies of the Tokyo case. These include the National Land Agency, Disaster Prevention Bureau; the Tokyo Disaster Prevention Congress; and Tokai Bank Limited. But Shah said he has heard of no comparable efforts in the United States.

Risk Management Solutions Inc. is a private company that evaluates earthquake risks for insurance companies. It was founded by Shah's son, Hemant. Hutton, the company president, is a Stanford graduate. Professor Shah is a member of the board of directors. The company licenses earthquake risk assessment technology from Stanford.






Download this release and its related files.

The release is provided in Adobe Acrobat format. Any images shown in the release are provided at publishing quality. Additional images also may be provided. Complete credit and caption information is included.

© Stanford University. All Rights Reserved. Stanford, CA 94305. (650) 723-2300. Terms of Use | Copyright Complaints