Stanford University

News Service



CONTACT: Kathleen O'Toole, News Service (650) 725-1939;

Team to assess earthquake vulnerability of transportation networks

A team of university researchers will study the earthquake vulnerability of the San Francisco Bay Area's transportation system through a demonstration project launched by a consortium of West Coast universities at Stanford on July 10.

Researchers from Stanford and the University of Southern California will lead the effort to look at not only how emergency workers can get where they are needed in the immediate aftermath of a quake, but how a region will recover economically when key transportation links are out of service for days, months or years. Their goal is to develop better software tools and analytical models for transportation planning and response in all earthquake-prone urban areas.

The Bay Area's system of roads and bridges, railways, airports and ports was selected for the detailed, three-year study by about 40 government, academic and private industry experts who met July 10 and 11 at Stanford's Blume Earthquake Engineering Center. The workshop was co-sponsored by the recently formed Pacific Earthquake Engineering Research Center, a consortium of eight West Coast research universities that aspire to improve earthquake hazard mitigation by integrating research in engineering, earth sciences and social sciences.

The Bay Area was chosen for the consortium's first demonstration project because "the transportation network in the Bay Area is more complex than the network in Los Angeles, and so the group decided it would be a better testbed," said Anne Kiremidjian, Stanford professor of civil engineering and director of the Blume center. "There is not as much redundancy in the Bay Area system, and the reconstruction of the Bay Bridge over the next two years also might provide valuable information on time delays and traffic re-routing."

Kiremidjian, who has been involved in developing a new system for analyzing the vulnerability of bridges, will head the project, which will include two other principal investigators ­ Professors James Moore of USC, who has studied the economic impact of the 1994 Northridge quake in Southern California, and Samuel Chiu of Stanford's Engineering-Economic Systems and Operations Research Department, who teaches transportation systems analysis.

Some workshop participants also said the Bay Area would provide a good study site because it would prompt researchers to consider political variables. Residents of the Los Angeles area that was struck by the Northridge quake developed an immediate consensus to rebuild damaged freeways, bridges and overpasses, but it has taken the Bay Area nearly 10 years to complete the rebuilding of the Cypress freeway structure, which fed traffic to and from the Bay Bridge in Oakland until a section of its top deck collapsed in the 1989 Loma Prieta earthquake. No consensus has yet been reached on how or if to replace the Embarcadero Freeway that was badly damaged in San Francisco.

A major difference from past studies is this project's focus on developing methods for analyzing the indirect economic impacts of earthquake damage to transportation systems, not just the direct costs of replacing and repairing the facilities themselves. Decision makers could use the information to aid them in setting priorities for seismic retrofitting and post-quake construction projects, Kiremidjian said. The workshop group "recommended we look in particular at the impact on Silicon Valley electronic business because it is such a major component of the economy of the Bay Area and perhaps another industry, such as the trucking industry, which interacts with many others."

Funding for the project will come from the Pacific Earthquake Engineering Research Center, which has received $2 million from the National Science Foundation and $2 million in matching funds from the state of California, including $1 million from Caltrans.

Multiple, major transportation links were damaged and closed in the aftermath of the Loma Prieta and Northridge earthquakes, but emergency personnel still were able to respond in a timely fashion, workshop participants said. Most key arteries and bridges were repaired more quickly than initially estimated, especially in Southern California.

The situation can be worse, however, as it was in 1995 in Kobe, Japan, when an earthquake hit the heart of a major metropolitan area. Kobe suffered railroad and roadway bridge failures, and subway line failures, and its harbor facilities were destroyed. The Kobe port was closed for more than a year, and probably will never recover its pre-earthquake business, Kiremidjian said.

"In the first 48 to 72 hours after the Northridge and Loma Prieta earthquakes, most people stayed away from the freeways. Only the people needed for emergency response used them and that helped the traffic tremendously. For several months following the Northridge earthquake, many employees worked from their homes," she said.

"This is in total contrast to what happened in Kobe, where as soon as the announcement of the earthquake was out, [residents'] relatives from the north part of the island descended on Kobe, creating major gridlock," she said. Kobe's recovery also was hampered by the collapse of buildings onto city streets.

A number of expert systems and software modeling tools have been developed for assessing the risk of earthquakes. Stanford researchers, for example, have contributed to HAZMUS, a national hazard analysis system funded by the Federal Emergency Management Agency that eventually will include analysis of risks from hurricanes and floods as well as earthquakes. It is intended to aid cities, states and the federal government in estimating the potential cost of disasters, and in developing contingency plans for emergency handling of casualties, injuries, shelter and recovery.

The Pacific consortium participants believe there is a need for an analysis tool that is more specifically focused on transportation ­ one that analyzes the whole transportation system as well as vulnerable individual components such as bridges and overpasses, Kiremidjian said. Transportation planners, for example, could use a software system that kept track of the width and condition of roadway shoulders so they could quickly decide where it was feasible to add an extra lane of traffic to replace a damaged link elsewhere.

Predicting how people and cargo will move around after an earthquake has damaged major arterials is even harder than predicting where damage will occur, many workshop participants said. Existing traffic modeling systems do not take into account the altered mindsets of people after a disaster, said Samer Madanat, an associate professor of civil and environmental engineering at the University of California-Berkeley. If properly used by emergency response officials, the news media can be a great help in getting people to minimize the usage of essential infrastructure in the first few hours after a disaster, he said. But once the immediate emergency has passed, media reports on traffic conditions generally don't reduce congestion and often increase it, according to research results, he said. Individuals use both their own travel experience and reports on traffic conditions to frequently change their travel routes and times in the days following a disaster, making it difficult for planners to predict the unstable traffic patterns.

Predicting which roads or bridges might fail is also difficult, and most existing models have predicted more failures than actually have occurred, participants said. This is partly because the models do not take into account the high degree of uncertainty involved in most of the input data, participants said. Some overprediction of damage is also considered prudent and at least provides a priority list for which facilities to inspect first.

"There are a lot of holes in the bridge evaluation system," Kiremidjian said after the workshop. To help plug those, she and former graduate student Nesrin Basöz, now a senior staff engineer at K2 Technologies in San Jose, developed a new bridge classification system for bridge components that has garnered much interest from other countries and states. Basöz developed a multiple origin-and-destination network analysis method for deciding how to allocate resources when multiple roadway links have failed and when there are several choices of hospitals, police and fire stations to respond to an emergency.

Basöz also analyzed the bridge damage data reports from the Northridge and Loma Prieta earthquakes and found inconsistencies in the way inspectors evaluated the damage, Kiremidjian said. "Her research suggests that better standards for assessing damage should be set to provide more robust data. Software tools to guide inspectors in the data collection would be very helpful."

The demonstration project hopes to bring such disparate elements together by applying the best methods available and developing new approaches for transportation systems.


By Kathleen O'Toole

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