Earthquake engineering is the collective effort of earth scientists, geotechnical engineers, structural engineers, and public policymakers to provide a built environment that is safe in the event of an earthquake. A significant part of this effort and the focus here is related to structural engineering, which involves the design and construction of structures and the anchorage of nonstructural building contents. Additionally structural evaluations and targeted retrofit of existing structures can be utilized to mitigate the risk of human and economic loss from an expected maximum probabilistic earthquake at a given site due to building collapse, loss of building contents, or economic downtime. Earthquake engineering thus constitutes a case study in specific relations between science, technology, and ethics.
Historical and Technical Background
Interest in constructing buildings to provide greater resistance to earthquakes arose in association with the scientific and professional development of engineering, especially from the late 1800s and early 1900s, in response to large earthquake damages that occurred in Japan, Italy, and California. For instance, the earthquake near San Francisco, in April 1906 (magnitude M = 7.8 on the Richter scale, 3,000 fatalities) destroyed structures in an area 350 miles long by 70 miles wide, and was the most expensive natural disaster in U.S. history until hurricane Andrew in 1992, with $500 million in damages (equivalent to $10 billion in 2004 dollars).
In order to defend investments and continue growth, initial press reports from San Francisco minimized the quake itself and focused instead on the fires started by downed electrical wires, cracked gas lines, and broken stoves (Geschwind 2001). Yet shaken by this and related events, California has become one of the most progressive states in the public reduction of earthquake risk through engineering design. More recent major losses in August 1999 in Izmit, Turkey (M = 7.6, 17,000 fatalities); January 2001 in Gujarat, India (M = 7.7, 20,000 fatalities); and December 2003 in Bam, Iran (M = 6.6, 43,000 fatalities) have promoted recognition of the need to deal systematically with earthquakes in the regions affected.
Despite the length of time since public attention was first drawn to earthquake risks, earthquake engineering remains a young science because of the relative infrequency of large quakes and the tremendous number of variables involved. Since the 1960s, earthquake-engineering development has made important progress by moving to incorporate knowledge from the pure geosciences with structural engineering, moving even toward multidisciplinary efforts to include sociology, economics, lifeline systems, and public policy (Bozorgnia et al. 2004). The scientific study of earthquakes or seismology is also relevant (see Bolt 1993).
Complete or partial structural collapse is the major cause of fatalities from earthquakes worldwide; earthquakes themselves seldom kill people, collapsing buildings do. Earthquake energy causes structures not sufficiently designed to resist earthquakes to move laterally. At this point, a building may lose its load carrying capacity and collapse under its own weight. Portions of buildings (such as roof parapet walls) or the interior nonstructural contents (refrigerators, bookshelves, and so on) can topple onto inhabitants inside or outside the building. Directly adjacent buildings can pound into each other, serving sometimes to stabilize each other when neighboring structures are on both sides (termed bookends) or to cause additional damage if a neighboring structure is on one side only or the floors do not align. Buildings on corners of city blocks are known to perform poorly, being pushed into the street due to one-sided pounding. Tsunamis, or tidal waves triggered by seafloor seismic movements, are another source of damage. Fires can be initiated from broken gas or electrical lines. Water saturated soil can lose its strength during dynamic shaking, and landslides or soil liquefaction may cause buildings to slide, be buried, or sink as if into quicksand.
To affect a structure, earthquake energy must first transmit through the bedrock from the epicenter, or the fault rupture location, through the soil above the bedrock (if any), through the foundation system, and then up through the building itself. All of these elements between the epicenter and building structure affect the level of lateral force (termed base shear) used for structural design. Frequency of ground motion can vary with distance from the epicenter, directivity, type of fault rupture, and magnitude. In the United States, the U.S. Geological Survey (USGS) maintains probabilistic earthquake demand topography maps based on statistical analyses of seismi-city, referenced by building codes and design standards andusedbystructuralengineersfordesign.
Engineering judgment, based on experience and observation of damage during past earthquakes, is relied on heavily in approximating earthquake demand, structural analysis, and overall structural design. Geotechnical engineers determine site soil conditions and site-specific seismic hazard. Structural engineers model the structural mass and stiffness, or how much a building moves when pushed laterally, based on the earthquake-resisting structural system used in design. Dynamic force and displacement limits are assumed based on structural detailing of connections and experimental testing results. Though material standards are used to set minimum criteria for material properties, there still exists some variability in material strength and ductility, requiring designing for a range of properties. Due to these many variables, two identical structures at different locations may require quite different earthquake-resisting systems.
After an earthquake, it is often difficult to know immediately if a building is severely damaged. The structure is typically covered by finishes, suspended ceilings, and fireproofing that need to be removed for visual investigation of connections, cracking, and other damage. In the United States, structural engineers may travel thousands of miles to aid in the initial building tagging and reconnaissance efforts, to quickly assign a red (no entry, evacuate), yellow (limited entry), or green (functional) placard at the entrance points. Developments in instrumentation have allowed for real-time building motions to be streamed over the Internet, which facilitates accuracy in initial tagging, but visual observation remains the primary basis for evaluation. In the case of a large office building, red tagging means the loss of weeks or months of revenue. In the case of hospitals and emergency response centers, a decision to evacuate means disruption of critical care during an emergency situation, increasing the death rate. For such reasons, engineers have an ethical responsibility to be extremely careful about a recommendation to evacuate a damaged building.
As architecture, construction materials, technology, and economics of construction evolve, seismic engineering evolves as well. Assumptions made during design are put to the test in future earthquakes, both validating previous thinking and exposing flaws. After the January 1994 earthquake in Northridge, California (M = 6.7, 60 fatalities, $40 billion in damage), it was found that many steel beam-to-column connections in relatively new structures had fractured at yield stress in buildings across the city, much different than the ductile behavior assumed in design. The structural engineering community initiated a six-year research project funded primarily by the U.S. Federal Emergency Management Agency (FEMA) to determine the cause of the poor performance, devise repair schemes, determine new design procedures that would produce desired ductile behavior, and modify building codes to avoid similar failures in future earthquakes (SAC 2000).
Building Codes, Economics of Construction, and Seismic Loss
In general the purpose of building codes is to protect public safety. But building codes and design standards, like the structures and societies in which they exist, are not permanent static entities, but dynamic and evolving to meet the demands and knowledge of changing times.
To minimize construction costs, building codes function as minimum requirements to permit damage from an earthquake but prevent collapse of the main structure, structural attachments, or contents. New buildings are expected to be repairable after a major earthquake, but some may be too costly to repair. Engineers have a responsibility to inform clients that building codes are not intended to preserve a structure, but do provide opportunities to increase the structural capacity or add special elements such as supplemental energy dissipation devices (viscous dampers and friction dampers, among others) or base isolation to reduce damage permissible by design codes.
Building owners arethusabletoincreaseabuilding's earthquake performance level if they are willing to pay the additional construction and design costs. Generally, however, it is difficult to sell higher performance engineering and construction costs to owners in the United States. In Japan and New Zealand, by contrast, higher performance structural elements are more frequently used. Building codes increase earthquake demand for critical structures, such as hospitals, schools, and communications hubs, with the intent that less damage occur during a major earthquake allowing the structure to remain operational afterward. In capitalist societies, history has shown that either economic incentives (tax breaks) or the threat of a facility being closed are often required to make building owners decide to retrofit. Both tactics are used in California (Geschwind 2001).
It is cheaper by far to allow for seismic forces during initial design than to incur damage or to retrofit later. Considering seismic forces initially may increase construction costs by 2 to 5 percent. Retrofit costs are typically on the order of 20 to 50 percent of original construction costs, excluding design fees and business interruption costs (Conrad 2004). Though seemingly inexpensive in comparison with the potential loss of the entire structure, there is major resistance to a 5 percent increase in construction cost from building owners, developers, and engineers not familiar with seismic design, especially in areas where the earthquake return period is longer than 100 years, when building codes (as in the United States) assume the typical building life to be fifty years. The area along the Mississippi River between St. Louis, Missouri, and Memphis, Tennessee, experienced three magnitude 7.8 to 8.1 earthquakes in 1811 and 1812, which reportedly moved furniture in James Madison's white house and rang church bells in Boston, yet many in the local communities maintain that designing for earthquakes is too costly. Money not spent on seismic retrofit for public facilities could theoretically be spent on the salaries of police and teachers, better hospital care, highway upgrades, and social programs. However probabilistic risk analysis demonstrates that ignoring earthquakes in design is often much more costly in the long run than short-term benefits of construction savings or budget reallocations.
In addition to loss of life, earthquake damages can significantly affect the local and world economies. The January 1995 earthquake in Kobe, Japan (M 6.9, 5,502 fatalities) caused more than $120 billion in economic loss. It is estimated that a similar earthquake in a major metropolitan area in the United States could result in a comparable loss (House Committee on Science, Subcommittee on Research, 2003). In the United States, earthquakes pose significant risk to 75 million Americans in 39 states. Averaging single event losses over the time between events, total annualized damages in the United States have been estimated at approximately $4.4 billion (House Committee on Science, Subcommittee on Research, 2003). When industrial transportation and utility losses are considered,
|*Fatalities estimated as high as 655,000.|
|SOURCE: U.S. Geological Survey, Earthquakes Hazards Program. Available from http://earthquake.usgs.gov; National Geophysical Data Center. Available from http://www.ngdc.noaa.gov/seg/.|
|January 23, 1556||Shensi, China||34.5||109.7||830,000||–||~8|
|November 1, 1755||10:16||Lisbon, Portugal||36.0||−11.0||70,000||–||~8.7|
|December 16, 1811||08:00||New Madrid, MO, USA||36.6||−89.6||–||~8.1|
|January 23, 1812||15:00||New Madrid, MO, USA||36.6||−89.6||12||~7.8|
|February 7, 1812||09:45||New Madrid, MO, USA||36.6||−89.6||12||~8|
|August 31, 1886||02:51||Charleston, SC, USA||32.9||−80.0||60||–||~7.3|
|June 15, 1896||19:32||Sanriku, Japan||39.5||144.0||–||~8.5|
|June 12, 1897||11:06||Assam, India||26.0||91.0||1,500||–||~8.3|
|April 18, 1906||13:12||San Francisco, CA, USA (San Andreas fault from Cape Mendocino to San Juan Bautista)||3,000||11||7.8|
|August 17, 1906||00:40||Valparaiso, Chile||−33.0||−72.0||20,000||11||8.2|
|December 16, 1920||12:05||Ningxia-Kansu, China||36.60||105.32||200,000||–||8.6|
|September 1, 1923||02:58||Kanto, Japan||35.40||139.08||143,000||–||7.9|
|May 22, 1927||22:32||Tsinghai, China||37.39||102.31||200,000||–||7.9|
|March 2, 1933||17:31||Sanriku, Japan||39.22||144.62||2,990||–||8.4|
|March 11, 1933||01:54||Long Beach, CA, USA||33.6||−118.0||115||–||6.4|
|December 26, 1939||23:57||Erzincan, Turkey||39.77||39.53||32,700||11||7.8|
|May 22, 1960||19:11||Chile||−38.24||−73.05||5,700||11||9.5|
|March 28, 1964||03:36||Prince William Sound, AK, USA||61.02||−147.65||125||–||9.2|
|February 9, 1971||14:00||San Fernando, CA, USA||34.40||−118.39||65||11||6.7|
|July 27, 1976||19:42||Tangshan, China||39.61||117.89||255,000*||10||7.5|
|September 19, 1985||13:17||Michoacan, Mexico||18.44||−102.36||9,500||9||8.0|
|October 18, 1989||00:04||Loma Prieta, CA, USA||37.14||−121.76||63||9||6.9|
|January 17, 1994||12:30||Northridge, CA, USA||34.18||−118.56||60||9||6.7|
|January 16, 1995||20:46||Kobe, Japan||34.57||135.03||5,502||11||6.9|
|August 17, 1999||00:01||Izmit, Turkey||40.77||30.00||17,118||–||7.6|
|January 26, 2001||03:16||Gujarat, India||23.39||70.23||20,085||–||7.7|
|December 26, 2003||01:56||Bam, Iran||29.00||58.34||26,200||9||6.6|
|December 26, 2004||00:58||offshore Sumatra, Indonesia||3.31||95.85||225,000 (est.)||–||9.0|
the estimated annual damage approaches $10 billion (Bonneville 2004). The September 11, 2001, terrorist attacks in the United States caused approximately 3,000 deaths and $100 billion in losses, roughly the same proportions as a major earthquake. Just as insurance, travel, and security measures have been increased throughout the world in response to the attacks of September 11, 2001, preparing for the next major earthquake would lessen the worldwideeconomiceffectsoffutureevents.
Seismic Risk Analysis and Societal Response
Since 1990 financial risk management analysis has been increasingly utilized by various levels of government, large corporations, and universities to understand and work toward reducing the financial impacts of major earthquakes. For example, a small one-story structure storing landscaping equipmentmaynotbeasimportanttoaclientasaone-story structure that houses emergency generators and an essential communications antenna. If the one-story structure is a collapse hazard, the owner may decide to strengthen the structure or move essential components to reduce risk.
Risk analyses use various loss estimating measures. The most common is the probable maximum loss (PML) due to a major earthquake, presented as a percentage of the building value. A 50-percent PML anticipates that half of the building will be damaged beyond repair in a major earthquake. Risk assessments need to be periodically updated to show progress and to reevaluate a client's portfolio with the ever-improving tools available to structural engineers produced through new research, analysis software, code developments, and observed damage. Values of PML studies need to be defined and investigated carefully as each methodology or computer program assumes slightly different parameters (Dong 2000).
Three requirements must be satisfied for a successful earthquake resistant design protocol. First, there must be practical structural design standards that reflect current observations and research, standards that are used by engineers and legally enforced as minimum requirements. Second, there must be thorough structural engineering performed by qualified and licensed engineers that leads to clear and explicit construction drawings. Third, construction must be monitored by qualified inspectors or by the designing engineers to ensure that the intended materials are used and construction proceeds as shown in the drawings and specifications. In case of unforeseen construction difficulties, the structural engineer must be involved in a solution that meets the intent of the design without compromising the structure, but also is as economic as possible.
If one or two of these three requirements are satisfied, the protocol is not successful. For example, after the 1999 earthquake in Izmit, Turkey, reports focused on shoddy construction and unenforceable building codes. Building codes are quite good in Turkey, closely following the standards published in the United States. However, for cultural reasons the building codes are frequently not enforced when a design is reviewed, and the contractor is held neither to building to the design standard nor to having an engineered design (EERI 1999).
Due to the Izmit earthquake, efforts to mitigate current and future earthquake risk in Europe are underway in Turkey, as well as Greece, Portugal, Italy, and the rest of the European Union (Spence 2003). All countries with moderate or high earthquake risk have their own cultural, financial, and political barriers toward earthquake risk mitigation. However, as has been demonstrated in the United States, Japan, and elsewhere when the three requirements of practical codes, sound structural design, and construction monitoring work together, earthquake risk is decreased as new buildings replace older ones.
It is extremely difficult for developing countries to mitigate seismic risk. Priorities are on more immediate needs such as food, clean water, and disease prevention and on the effects of poverty and war. Construction uses available materials and follows traditional methods without structural calculations. While economic losses in developing countries may not be as high as in the United States, loss of life is much more severe, potentially approaching the proportions of the July 1976 earthquake in Tangshan, China (M 7.5), where between 250,000 and 655,000 people were killed and more injured when nearly the entire city was razed.
Population expansion, and hence the rate of construction using traditional (seismically unsafe) methods, is at a much higher rate in countries such as India or Nepal than in the United States, exponentially increasing the earthquake risk in these countries. It is estimated that the risk of fatalities in developing countries compared to industrialized countries is 10 to 100 times greater—and increasing. This trend is the largest ethical and functional difficulty worldwide with regard to earthquake risk. In addition to moral obligations to reduce earthquake risk in developing countries, there are financial reasons as well. Due to economic globalization, a major disaster in a developing country has direct immediate and long-term financial impact on the world economy.
JAMES P. HORNE
DAVID R. BONNEVILLE
Bolt, Bruce A. (1993). Earthquakes, newly revised and expanded. W. H. New York: Freeman. Contains information about seismology, structural engineering, and public response to historical earthquakes.
Bonneville, David. (2004). "Securing Society Against Earthquake Losses." Structural Engineers Association of Northern California Newsletter 59(2004): 1–3.
Bozorgnia, Yousef, and Vitelmo V. Bertero, eds. (2004). Earthquake Engineering: From Engineering Seismology to Performance-Based Engineering. Boca Raton, FL: CRC Press. Describes the current knowledge of practicing engineers' and researchers' techniques in structural engineering for earthquake resistance.
Earthquake Engineering Research Institute (EERI). (2000). Financial Management of Earthquake Risk/EERI Endowment Fund White Paper. Oakland, CA: EERI. An excellent description of the key aspects of seismic risk analysis techniques after a decade of development and private practice.
Geschwind, Carl-Henry. (2001). California Earthquakes: Science, Risk, and the Politics of Hazard Mitigation. Baltimore, MD: Johns Hopkins University Press. An extremely well developed text from start to finish of seismic legislation and public policy related to curbing seismic loss in California.
SAC Joint Venture. (2000). Federal Emergency Management Agency (FEMA) Publication 354. The SAC joint venture is a very thorough research project that employed a great deal of coordination and cooperation between academia and practicing engineers, in an effort to better understand the poor behavior that was observed after the 1994 Northridge earthquake, which was exactly the opposite of the intention of the building codes which produced the results. Engineers and researchers took it upon themselves to publicize the problem, obtain funding, and publish these documents to avoid the poor behavior in future buildings, as well as publish ways to retrofit existing structures with similar characteristics.
Spence, Robin. (2003). "Earthquake Risk Mitigation in Europe: Progress towards Upgrading the Existing Building Stock." In Proceedings of the Fifth National Conference on Earthquake Engineering. Istanbul, Turkey: Fifth National Conference on Earthquake Engineering. A discussion on the European view of earthquake engineering, the cultural resistance, and potential for future progress.
Applied Technology Council. Available from www.atcouncil.org. Organization web site.
Conrad, Katherine. (2004). "Seismic Retrofit May Limit Bomb Damage." East Bay Business Times. Available from http://eastbay.bizjournals.com/eastbay/stories/2004/04/05/focus1.html
Consortium of Universities for Research on Earthquake Engineering. Available from www.curee.org. Organization web site.
Earthquake Engineering Research Institute (EERI). "Learning From Earthquakes: The Izmit (Kocaeli), Turkey Earthquake of August 17, 1999." Available from: http://www.eeri.org/lfe/pdf/turkey_kocaeli_eeri_preliminary_report.pdf.
House Committee on Science, Subcommittee on Research. The National Earthquake Hazards Reduction Program: Past, Present and Future. 108th Cong, 1st sess. May 8, 2003, 108–114. Available from http://commdocs.house.gov/committees/science/hsy86870.000/hsy86870_0.HTM. An interesting interpretation of where and how funds in the United States are appropriated, and the information this is based on.
Structural Engineers Association of California. Available from www.seaoc.org. Organization web site.