National Academies Press: OpenBook

Practical Lessons from the Loma Prieta Earthquake (1994)

Chapter: 2. The Geotechnical Aspects

« Previous: 1. Legacy of the Loma Prieta Earthquake: Challenges to Other Communities
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

2
The Geotechnical Aspects

G. Wayne Clough, James R. Martin, II, and Jean Lou Chameau

INTRODUCTION

Evidence obtained immediately following the Loma Prieta earthquake and in subsequent studies indicated a strong geotechnical influence on the observed behavior and damages. Much of the response could be termed "expected," but research following the earthquake has led to a refined understanding of previously defined problems and development of new areas of focus. For example, the earthquake allowed (1) a first-time "test" of sites that had been improved to resist liquefaction; (2) evaluation of soil density changes by comparing pre- and post-earthquake site test results; (3) direct measurement of effects of site amplification; (4) at least limited documentation of liquefaction-induced settlements and lateral movements; and (5) indirect measurement of the response of major landfills, underground structures, and reinforced earth retaining systems. Thus, considerable useful experience can be derived from the Loma Prieta earthquake for geotechnical engineering.

Although much has been learned from the earthquake, and more knowledge is to come, extrapolation of the information for the geotechnical community has to be tempered by the knowledge that special conditions ameliorated the damages. For example, even though up to 4,000 landslides occurred (Keefer, in press), such events were moderated by the effects of a four-year drought in Northern California. Other factors that should be considered in attempting to extrapolate lessons from the earthquake include the moderate size of the earthquake, the distance of the epicenter from large population centers and soils susceptible to liquefaction, the unique nature of the fault break and the relatively short duration

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

of strong shaking, and the presence of low reservoir levels behind earth dams and embankments. These conditions make it essential that care is taken in using the lessons learned from the earthquake. If the duration of the event had been longer, water tables and reservoir levels higher, or the epicenter closer to San Francisco damages could have been greater, and what appeared to be successful performance could have translated into unsuccessful behavior.

The timing of this conference is well-matched to the discovery phase of the research on the Loma Prieta earthquake. It is notable that in the literature search for this paper many of the early, sometimes seemingly sensational, documents now seem dated. Their value for the future will not lie in the profundity of the insights developed but in the raw observations made of patterns of behavior. Clearly, the more recent publications, which reflect careful studies conducted in the intervening years since the event itself, are beginning to decipher properly the true causes of behavior. Also, with the publication of more research results, patterns are emerging that were not obvious before. This paper should be viewed as a summary of findings to date. Further useful results will undoubtedly be forthcoming.

OVERVIEW

It is useful to review some aspects of the Loma Prieta earthquake that are important to the geotechnical response associated with it. The Ms = 7.1 event (Mw = 6.9) was a moderate earthquake, with an epicenter located in the Santa Cruz Mountains, about 11 miles (18 km) from Santa Cruz and 60 miles (97 km) from the San Francisco Bay area. The causative fault rupture was bilateral, with a medial location of the epicenter. As a result, the strong shaking lasted only 8 to 15 seconds, shorter by as much as a factor of two relative to durations normally associated with an event of this magnitude. For the subsurface materials, this translates to a smaller number of cycles of loading than would have occurred otherwise.

The map in Figure 2-1 shows the position of the epicenter, major population centers, and locations of liquefaction-induced damage and landslides. As expected, there is a concentration of damages and landslides near the epicentral area, which reflects the high level of accelerations and steep terrain in this vicinity. South of the epicenter, in the vicinity of Santa Cruz, Watsonville, and Moss Landing, certain land masses were particularly susceptible to ground movement due to liquefaction and landsliding. Damages were also concentrated to the north of the epicenter in the San Francisco Bay area, where the type of earthquake motions and the soil conditions combined to produce site amplification and liquefaction in a heavily populated area. These effects are explored in more detail subsequently.

Table 2-1 lists peak horizontal accelerations and durations of strong shaking for a number of locations near the epicenter and in the San Francisco Bay area.

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

FIGURE 2-1 Regional map of earthquake damage due to liquefaction and landsliding (after Seed et al., 1991 ).

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

TABLE 2-1 Peak Horizontal Accelerations at Selected Sites in the Loma Prieta Earthquake

Location

Epicentral Distance (miles)*

Peak Ground Accelerations (gs)

Ground Condition

Epicenter

0

0.65

Rock

Santa Cruz

11

0.47

Rock

Watsonville

12

0.40

Rock

San Jose

14

0.25

Stiff Soil

San Francisco Airport

52

0.33

Fill/Soft Soil

Ricon Hill, San Francisco

63

0.10

Rock

Yerba Buena Island

64

0.07

Rock

Treasure Island

64

0.16

Fill/Soft Soil

Emeryville

65

0.24

Fill/Soft Soil

* (1 mile = 1.6 km).

The highest recorded accelerations were 0.6 g near the epicenter, and attenuation patterns of accelerations with distance from the epicentral region followed largely expected trends with some exceptions. Some 40 to 60 miles (64 to 97 km) from the epicentral region in the San Francisco Bay area, the peak accelerations varied from 0.05 g to 0.33 g, with the higher values associated with soft soil sites and the lower values recorded in rock and hard soil sites.

LIQUEFACTION

Occurrence And Recurrence

Liquefaction during the Loma Prieta earthquake was common near the shoreline of the San Francisco Bay and adjacent to rivers and bodies of water near the Pacific Ocean west of the epicentral region (Figure 2-1). There were few surprises as to the locations of liquefaction, since most of the areas where it was evidenced fit expected criteria for liquefaction. In a number of cases, liquefaction was accurately predicted prior to the earthquake (Clough and Chameau, 1983; Dupre and Tinsley, 1980). Recurrence of liquefaction in the same locations as in the 1906 San Francisco earthquake (Ms = 8.3) was not uncommon (Seed et al., 1991; O'Rourke et al., 1991). However, where damage patterns due to liquefaction in the earthquake mimicked those from the 1906 earthquake, the severity of liquefaction and damage from the earthquake was less than that associated with the 1906 event.

Well-documented liquefaction failures were associated with uncompacted, saturated, sandy fills in the central San Francisco Bay region (EERI, 1990). Figure 2-2 indicates where sandy fills are present on the eastern side of San

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

FIGURE 2-2 Locations of waterfront fills in San Francisco and test sites TH and YBC. Liquefaction phenomena observed in Loma Prieta earthquake is noted (modified from Seed et al., 1991).

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

Francisco. The majority of these fills were placed in the late 1800s or the early to mid-1900s and consist of sands that were dumped or dredged into place and allowed to settle in suspension (Dow, 1973). Fills placed after 1950 tended to have been subjected to some form of compaction. Table 2-2 provides a description of many of the major fills and their placement processes. It is important to note that the fills that were dumped into the bay consisted of a wide range of materials, including rubble from construction and demolition. Another commonly dumped fill material was dune sand, a soil that was abundant in the early days of filling of the waterfront areas. Dune sand has a uniform, medium gradation and is largely free of fines. With time, sands were also dredged from sediments in San Francisco Bay. These materials typically contained fines, in contrast to the clean dune sands, and they attained lower fill densities than the dumped clean sands. During the Loma Prieta earthquake, differences in responses of fills created by different placement techniques was exhibited in areas like the Marina District (Bonilla, 1992; O'Rourke et al., 1990, 1991). The dredged fills exhibited a tendency to liquefy more readily than the dumped fills.

There were major fills around the bay that behaved well in the earthquake. In almost all cases, specific measures had been taken to compact the fills while they were being placed, or after placement. Those densified after placement are described in a subsequent section of this paper. The fills at the San Francisco Airport, and those at Foster City and Redwood Shores, were at least partially compacted during placement and exhibit medium to dense densities (EERI, 1990). In some cases, the fills also contain shells and are partially cemented. No significant liquefaction was found in these fills.

The Loma Prieta earthquake provided the first opportunity to assess the accuracy of regional liquefaction susceptibility maps (Tinsley and Dupre, in press). The susceptibility map in Figure 2-3 was developed by Dupre and Tinsley (1980) for the Monterey Bay region using the procedures of Youd and Perkins (1987). Liquefaction susceptibilities were based on the occurrence of a large event like the 1906 earthquake. In the Loma Prieta earthquake, the Dupre and Tinsley mapping accurately defined locations of major occurrences of liquefaction and lateral spreading. At the same time, broad regions within areas mapped as susceptible to liquefaction showed no response. This can be explained in terms of (1) the small size of the earthquake relative to the 1906 event, (2) lower water tables than those expected, and (3) local differences between grain sizes of deposits identified as liquefiable. The latter item was important in that flood plain deposits that were clay-rich did not fail, whereas areas of sand-rich tidal flat and abandoned channel deposits did fail. While broad mapping of liquefaction susceptible soils inherently has difficulty in capturing details, such as fines content or water table fluctuations, it is a valuable guide in identifying areas of potential problems.

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

TABLE 2-2 Behavior of Fill Soils in Central San Francisco Bay Region in the 1989 Loma Prieta Earthquake

Sites in City of San Francisco

Fill Type

Soil Type

Fill Density

Liquefaction Damage During Loma Prieta Earthquake*

Embarcadero

Unimproved End-Dumped

Fine Sand

Loose to Medium

Moderate to Minor

Hunter's Point Cofferdam

Unimproved Hydraulic

Fine, Silty Sand

V. Loose to Loose

Severe

Manna District

Unimproved Hydraulic,

Fine, Silty Sand

V. Loose to Loose

Severe

 

Unimproved End-Dumped,

Fine Sand

Loose to Medium

Moderate

 

Natural Ground

Find Sand

Medium to Dense

None

Mission District

Unimproved End-Dumped

Fine Sand, Rubble

Loose to Medium

Moderate to Minor

Pier 80, 84

Improved Dumped

Fine Sand

Medium to Dense

None

Pier 45

Unimproved End-Dumped

Fine Sand

V. Loose to Loose

Severe

Sites Outside

 

 

 

 

San Francisco

 

 

 

 

Alameda Island

Unimproved Hydraulic

Fine, Silty Sand

V. Loose to Loose

Severe

Bay Farm Island

Unimproved Hydraulic

Fine, Salty Sand

V. Loose to Loose

Severe

Sites outside City of San Francisco

Fill Type

Soil Type

Fill Density (V-Very)

Liquefaction Damage During Loma Prieta Earthquake*

Foster City

Roller-Compacted

Fine Sand, Some Cementation

Medium to Dense

None

Oakland Airport

Unimproved Hydraulic

Fine, Silty Sand

V. Loose to Loose

Severe

San Francisco Airport

Roller - Compacted

Fine Sand

Medium to Dense

None

Seventh Street Terminal

Unimproved Hydraulic

Fine, Silty Sand

V. Loose to Loose

Severe

Treasure Island

Unimproved Hydraulic,

Fine, Silty Sand

V. Loose to Loose

Severe

 

Improved Hydraulic

Fine, Silty Sand

Medium to Dense

None

*NOTES: Minor = Slight lateral spreading and/or settlements, little surficial evidence.

Moderate = Minor lateral spreading and/or limited settlements, sand boils, etc.

Severe = Large lateral deformation and/or settlements, large & numerous sand boils, etc.

V. = Very

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

FIGURE 2-3 (a) Site map showing areas of Monterey Bay Region for which liquefaction susceptibility maps were developed; 

(b) Liquefaction susceptibility map developed for area indicated by inset in (a) (adapted from Tinsley and Dupre, 1992).

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

Settlements

Differential settlements due to liquefaction were widespread during the Loma Prieta earthquake. In level ground areas, the settlements were generally caused by a loss of fill volume during sand boiling and consolidation of the fills following pore pressure build-up. However, in instances where liquefaction occurred in the presence of a slope (even a very mild slope), downslope lateral movements in the soils caused settlements in the upper reaches of the soil mass that moved (see next section). Magnitudes of the settlements were essentially a function of the thickness of the liquefiable soils and the liquefaction potential of the soils. Liquefaction-induced settlements were believed to have caused failures in the San Francisco Municipal Water Supply System (Scawthorn et al., 1991); structural damages in the Marina District (Mahin, 1991), the South of Market area (Seed et al., 1991), Fisherman's Wharf and the Embarcadero (Chameau et al., 1991), and Treasure Island and the Oakland Port (EERI, 1990; Egan and Wang, 1991; Seed et al., 1991); and damages to highways and runways, for example, the Oakland Airport (EERI, 1990). It should be noted that even at sites where seawall and containment dikes were present, large settlements still occurred in hydraulic fills behind these support systems.

Accurate measured values of settlement due to liquefaction could not be derived from the information available to investigators. However, reasonable estimates could be made in some cases, and O'Rourke et al. (1991) were able to test existing methods of prediction of settlement caused by liquefaction. The procedures use results from Standard Penetration Tests (SPT) or Cone Penetration Tests (CPT) as input parameters. It was concluded that the methods worked well for clean sands but were not accurate in sands with fines unless corrections were applied to account for the effects of fines on the blow counts or cone penetration resistances.

Lateral Spreading

Lateral spreading was observed in most areas where significant liquefaction occurred during the Loma Prieta earthquake. In San Francisco, lateral movements were primarily associated with the sandy fills along the waterfront (Figure 2-2). These fills typically slope from 0.5 percent to 2 percent downhill toward the bay and are restrained by seawalls that run along the perimeter of the waterfront. Indications of lateral spreading in these fills during the earthquake were relatively minor. Much of the prominent pavement buckling in the central Marina District was attributed to oscillatory movements, not spreading. Except for an area near the marina where about 2 ft (0.6 m) of lateral movement occurred at St. Francis Spit (Taylor et al., 1992), permanent downslope (toward the bay) displacements of the fills were typically less than several inches. It must be kept in mind, however, that these movements were undoubtedly partially controlled by

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

the presence of seawalls. Interestingly, Mitchell et al. (1991) have presented the thesis that the presence of large box culverts—typically 25 ft (7.6 m) wide and 30 ft (9.2 m) deep—buried along the perimeter of Marina Green helped to control the lateral spreading of the fills that liquefied in this area. This seems reasonable, and the stabilizing effect could have been amplified by the densification of the fills around the culverts that occurred as the sheet piles for the excavation support were vibrated into place (Clough and Chameau, 1980).

The relative lack of ground movements in the fills in the Old Mission Bay area was surprising, considering that lateral movements of up to 8 ft (2.4 m) occurred in this region during the 1906 earthquake (Youd and Hoose, 1978). The lack of significant lateral movements is partly attributed to the lesser magnitude of the Loma Prieta earthquake relative to the 1906 event, but other factors may have been at work. Possibilities include densification as a result of the 1906 earthquake, or that additional filling has occurred along the waterfront in this area since the 1906 earthquake. It is known that much of the rubble from buildings damaged during the 1906 event was pushed into the bay near the mouth of the old Mission Creek channel (Dow, 1973), and this could provide some buttressing support to the fills located farther inland.

Lateral spreading in the central San Francisco Bay region also occurred at Treasure Island, the Oakland Port, and other areas along the eastern bay shore. Similar to the occurrences in San Francisco, movements at these sites were also restrained by seawalls and containment dikes. Because the influence of the containment structures upon the observed movements is difficult to quantify, the actual deformation behavior of the soils is somewhat moot, and the movement data from these sites would be of limited use in the development of methods to predict lateral movements in liquefied soils.

Useful data on lateral spreading was obtained from the Monterey Bay area, where approximately 50 lateral spread sites were investigated by Tinsley and Dupre (1992). Lateral spreading throughout this region was strongly related to geologic facies. Approximately 95 percent of the spreads occurred in late Holocene fluvial point-bar deposits, fluvial channel deposits, and estuarine deposits. Beach and alluvial fan deposits rarely liquefied (see Figure 2-4). Geotechnical data from the field sites in the Monterey Bay area have yet to be sufficiently analyzed to develop relationships between the magnitude of the movements and the factors that controlled the movements. However, preliminary analyses suggest that the movements were not merely a function of simplified parameters such as slope angle, free-face height, etc. Although it was clear at all sites that the largest movements occurred near the free faces of laterally displaced slopes, there was no consistency in the ''safe setback'' distance from the free faces. At each site, the overall size of the soil mass that moved laterally was apparently controlled by the extent of the geologic unit in which the spread developed. Consistency for this selective behavior was confirmed by SPT and CPT measurements, which indicated soil conditions in the young fluvial and estuarine

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

FIGURE 2-4 Histogram showing distribution of lateral-spread ground failures according to sedimentary facies for the Loma Prieta earthquake (after Tinsley and Dupre, 1992).

sediments to be the most favorable for liquefaction relative to older geologic units within the region.

Tinsley (1993) compared measured lateral movements in sandy soils in the Monterey Bay region with movement predictions obtained using the Liquefaction Severity Index (LSI) method of Youd and Perkins (1987). The results indicated the upper-bound estimates for lateral movements from the LSI method were too low. The LSI method predicts lateral movements based on earthquake magnitude and distance to the vertical projection of the fault rupture; it does not consider soil parameters. A more rigorous technique, developed by Bartlett and Youd (in press), as well as other prediction methods, have yet to be evaluated with the new field data.

Fill Densification As A Result Of Earthquake Shaking

Following the Loma Prieta earthquake, SPT and CPT were conducted at several sites along San Francisco's waterfront, where similar tests had been performed in the late 1970s, which provided an opportunity to measure changes in density that were due to earthquake shaking (Clough and Chameau, 1983; Chameau et al., 1991). Two of the principal study areas, known as the TH and YBC sites, were located along the Embarcadero north of the Bay Bridge and Market

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

FIGURE 2-5 Average CPT tip resistances measured at the YBC and TH sites  before and after the Loma Prieta earthquake (after Chameau et al., 1991).

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

Street (Figure 2-2). The soil conditions consist of dune sand fills about 30 ft (9.2 m) thick, which are underlain by soft recent bay mud; the ground water table is at a depth of 10 ft (3 m). Using conventional methods for liquefaction prediction, liquefaction would have been expected in the loosest portions of the fills at the YBC site but not in any of the fills at the TH site.

In Figure 2-5, the relative density profiles estimated for the pre-earthquake and post-earthquake conditions for the YBC and TH sites are plotted versus depth. Relative densities in the loosest sections of the fills at the YBC site were estimated to have increased from 45 percent to 60 percent due to the earthquake. Observations at the YBC site after the earthquake indicated minor sand boiling and lateral spreading and settlements. In contrast, no surficial evidence was found for liquefaction at the TH site, and no significant density changes were observed (average relative densities were estimated at 60 percent both before and after the earthquake). The implications from these findings are that moderate shaking and marginal liquefaction can be effective in densifying soils that are in a loose to very loose condition. Chameau et al. (1991) provide data for other sites to support the behavior observed along the Embarcadero.

Improved Ground

The Loma Prieta earthquake provided an opportunity to test the effectiveness of modern ground improvement techniques (e.g., vibraflotation, compaction piles, etc.) used to compact sandy fills and reduce their liquefaction susceptibility. There were no indications of liquefaction at sites where soils were improved using these procedures, although significant liquefaction damages occurred in adjacent areas of unimproved ground (Mitchell and Wentz, 1991). In spite of this positive behavior, it should be noted that the earthquake did not represent the design earthquake for most of the improved sites. Further, the ground motions at the improved sites were not significantly different from those at unimproved sites. Thus, while ground improvement may serve to inhibit liquefaction, it does not limit shaking of structures founded in the area.

SITE AMPLIFICATION AND RESPONSE SPECTRA

Site amplification is a term used to define the occurrence of ground motions at the surface of sites that are larger than those that would occur if the site were composed of bedrock. In the Loma Prieta earthquake, amplification occurred in the San Francisco Bay area at sites underlain by weak rock, stiff soil and soft soil (Borcherdt and Glassmoyer, 1992). The most dramatic effects were observed for sites with soft soil, and these are predominately found in the Bay Area around the fringes of San Francisco Bay. The soft soils were created by deposition over the past 10,000 years as San Francisco Bay was submerged under rising sea levels caused by the melting of the glaciers. These soils are termed locally as

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

"recent bay mud," and they are generally near normally consolidated silty clays, although they can be either highly plastic clays or sands. The occurrence of soil amplification in the soft soils was not a surprise, since the technology available was able to predict this phenomenon. However, the experience of the earthquake provided insight into the degree of accuracy of predictive tools and important details about soil amplification.

The existence of site amplification can be seen in Figure 2-6, where ratios of peak transverse site accelerations to peak transverse accelerations for a nearby "bedrock" site are shown as a function of site conditions. All of the sites are located in the Bay Area some 40 to 60 miles (64 to 97 km) from the epicentral region. The harder rock sites show ratios of about one (as expected, no amplification), but the weak rock and soil sites have ratios above one, with average ratios of 1.2 for weak rock, 1.5 for stiff soil sites, and 2.5 for soft soil sites. The largest ratio for a soft soil site was 3.7 at the San Francisco Airport. In absolute terms, measured accelerations on rock sites were 0.1 g or less in the San Francisco/Oakland area. At instrumented sites that were underlain by soft bay mud, locations like Foster City, Treasure Island, or in the shoreline area of the East Bay, peak accelerations were 0.2 to 0.3 g. These accelerations were also rich in long period motions and, thus, possessed enhanced damage potential for weak structures (e.g., unreinforced masonry) and structures with long natural periods.

FIGURE 2-6 Acceleration amplification as a function of site conditions from Bay Area sites  (adapted from Borcherdt and Glassmoyer, 1992).

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

Idriss (1990) was the first to report analyses of site motions that were conducted using conventional one-dimensional simulations based on piece-wise linear-elastic models of the ground response. He noted that this tool was successful in reasonably predicting response spectra of instrumental recordings at a number of soft soil sites if allowances were made for stiffer soil behavior and reduced damping in the soil model over earlier, accepted values. This conclusion was supported by Seed et al. (1991) and Dickenson et al. (1991), who performed analyses of a number of sites where observed and predicted response spectra were compared. Bardet et al. (1992) investigated the response of the soils in the Marina District of San Francisco using one-dimensional, two-dimensional, and three-dimensional analysis tools. The Marina District region is of interest in that the soils fill a basin with a three-dimensional shape where micro-tremor instrument recordings suggested an effect of this geometry (Boatwright et al., 1992). Conclusions from the Bardet et al. (1992) analyses were (1) one-dimensional analysis tools reasonably predict the response of the basin in the central regions, and (2) accounting for bedrock profile changes and wave propagation effects in two- and three-dimensional models can generate higher accelerations than are predicted in one-dimensional models. The results of all of the site motion analyses lead to several interpretations. First, it seems that for day-to-day design, the simplified one-dimensional models remain the most reasonable tool for site-specific analysis of ground motions, given the inherent unknowns about input motions and subsurface characteristics. However, for purposes of research studies of observed behavior, the more sophisticated analyses procedures would appear to warrant serious consideration.

While the site amplification effects in soft soils in the Loma Prieta earthquake could have been anticipated, the one-dimensional analyses showed that the largest amplified accelerations were not generated at the fundamental frequencies of the sites (Seed et al., 1991). Figure 2-7 shows response spectra determined from measured ground motions at Treasure Island (soft soil site) and the adjacent Yerba Buena Island (rock site). Both of the spectra exhibit peaks at periods of about 0.6 s (seconds) with motions for Treasure Island amplified about three times those of Yerba Buena Island. Seed et al. (1991) noted the amplification effect was generated by the focused energy in the input rock motion at a period of 0.6 s. With the predominant natural frequency of the Treasure Island site at approximately approximately 1.3 s, the peak spectral response that occurred at a period of 0.6 s was associated with approximately the period of the second mode of the site, not its first mode. At periods at or near 1.3 s, the response spectrum shows magnifications of four to five, but because the energy of the rock motions at this period was lower than that at 0.6 s, the magnitudes of accelerations at this period were not as large as those at a period of 0.6 s.

Figure 2-7 also presents the design spectrum proposed for the S4 soil profile for purposes of comparison to the measured spectra. The S4 condition was developed to provide design guidance for cases with deep soft soils. It is seen

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

FIGURE 2-7 Measured response spectra at Yerba Buena Island (rock site) and Treasure Island (soft site) from the Loma Prieta earthquake with recommended S4 design code spectrum (adapted from Dickinson et al., 1991).

that the measured response spectrum for the Treasure Island site significantly exceeds the design spectrum at periods where the maximum amplification occurred. Dickenson et al. (1991) show this effect applies for a range of spectra from soft soil sites in the Bay Area. This puts at issue the degree to which even the latest attempt to capture a conservative design approach for deep soft soil sites is adequate. As per the recommendations in such cases, a site-specific ground motion analysis is advisable.

SLOPES, FILLS, AND EMBANKMENTS

Landslides, downslope movements, and cracks developed in natural slopes, road fills, embankment dams, and landfills during the Loma Prieta earthquake (Figure 2-2). One fatality was attributed to a rockfall along the Pacific Coast, and damages were induced in over two hundred residences and fifteen earth dams. In addition, a major highway was blocked in two locations by landslide masses for extended periods of time.

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

Natural Slopes

A summary of the findings concerning the response of natural slopes in the Loma Prieta earthquake is to be provided in a forthcoming U.S. Geological Survey Professional Paper edited by Keefer (in press). Excerpts from this document provided considerable useful information for the authors. Keefer notes as many as 4,000 landslides may have occurred in the earthquake, with the most common types of landslides being small rockfalls, rock slides, and disrupted soil slides. Deeper seated and more-coherent slumps and block slides were less common and more likely to be located near the epicentral region around Summit Ridge. The large number of landslides in the earthquake is not unusual, since it is estimated that about 10,000 landslides occurred in the 1906 San Francisco earthquake. The four-year drought at the time of the earthquake is thought to have prevented more-extensive movement and landsliding.

The natural slope landslides were concentrated in formations that had been previously identified as susceptible to sliding. Keefer reports that 85 percent of the landslides in the earthquake were located southwest of the fault rupture, mainly in the southern Santa Cruz Mountains, in the poorly indurated sedimentary rock formations. They were primarily composed of sandstones, siltstones, mudstones, and shales. Many of the landslides were in man-made cuts for highways, and two of these blocked different portions of California Highway 17 for several weeks. The highway cuts were materials that were highly fractured, and which were likely in a state of incipient failure prior to the earthquake.

Coherent slides were less common than the rockfall variety, but these were disruptive, since many were often located so as to impact residential areas or roads (Seed et al., 1991; Spittler and Harp, 1990). As many as 30 percent of the coherent slides occurred in road cuts, fills, or embankments, and most of these were small (Keefer, in press). The largest of the slumps occurred in the Summit Road area of the southern Santa Cruz Mountains. It involved up to 27 × 106 yd3 of soil and rock with landslide depths of up to 300 ft (92 m). Trenches dug for purposes of identifying the slide planes for the Summit Road landslides showed that up to three displacement events in addition to that associated with the Loma Prieta earthquake had occurred over a period of 2,000 to 3,200 years (Nolan and Weber, in press). Cole (1991) attempted to use analytical procedures for slope and movement analysis to predict response of the coherent slides. Material parameters were determined from laboratory tests on small samples from the landslide areas. The calculations predicted movements that were smaller than those that actually occurred. It was recommended that more realistic predictions could be achieved using material parameters determined from back-analysis of previous slope failures. The effects of geometry and wave propagation should also be considered, since it was noted that block slides occurred in a number of locations at the end of prominent ridges.

In addition to the landslides, many ground cracks were opened in the Sum-

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

mit Road area. Some of the cracks were linked to adjacent landsliding, while others were indicative of incipient landsliding. Still other crack sets and fissures were found near the ends of prominent ridges. This may have been caused by wave reflection that would tend to occur at the face of a steep slope and by the known tendency for slopes to intensify motions near the face of the slope (Sitar and Clough, 1983).

A significant number of landslides and rockfalls were observed in the steep bluffs along the Pacific Coast (Griggs and Plant, 1993). While the coastal materials are often referred to as sedimentary rocks, they commonly are either highly fractured soft rock or cemented soils. Sitar and Clough (1983) showed the response of coastal bluffs varied depending on the tendency toward matching of the natural frequency of the slope and the characteristic period of the earthquake. For steep bluffs in cemented soils, the natural period for bluff heights of 25 m would fall in the range of 0.5 to 1, values that could coincide with those from the earthquake at distance from the epicenter.

Finally, while most of the landsliding occurred near the epicentral area, Seed et al. (1991) report a landslide that occurred in San Francisco in sands that damaged 30 homes, and one north of San Francisco on the west coast of the Marin Peninsula. These slides were located more than 60 miles (97 km) from the epicenter.

Engineered Embankments And Dams

Harder (1991) and Seed et al. (1991) reported on the behavior of earth dams that were impacted by the Loma Prieta earthquake, and much of the material that follows is drawn from these sources. There were 111 earth dams located within 120 miles (75 km) of the fault rupture of the earthquake (Figure 2-8). About half of the dams were built prior to 1950, and 21 of these were constructed before the 1906 earthquake. The height of the dams ranged from under 10 ft (3 m) to over 300 ft (92 m). Most were constructed of relatively homogeneous clayey soils, even those built before 1906. Five were constructed by hydraulic filling.

Of the dams that were built prior to 1906, none were damaged, a behavior consistent with experiences described for the 1906 San Francisco event (Seed et al., 1978). No significant damages were observed in any of the hydraulically filled dams, although Mill Creek Dam and Hawkins Dam experienced some minor cracking. Notably, the hydraulically filled dams were either located at a considerable distance from the epicenter or had dry reservoirs at the time of the earthquake.

Table 2-3 gives the key characteristics of ten dams that experienced noticeable ground movement or cracking in the earthquake. The majority of the dams exhibited cracking near the crest or in the materials adjacent to the abutment. Eight of the dams were earth/rockfill embankments and two were hydraulically filled. Not surprisingly, the closer the dam was to the epicenter, the more likely there was damage. Austrian Dam, which was only one mile (1.6 km) from the

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

FIGURE 2-8 Location map for earth and rockfill dams and major  landfills (adapted from Harder, 1991 ).

epicenter, sustained horizontal accelerations over 0.5 g, and it was the most severely damaged dam. Austrian Dam is 185 ft (56.7 m) high with 2.5:1 to 3:1 slopes, and it is composed of a compacted clayey, sandy gravel fill. The dam settled as much as 2.5 ft and displaced up to 0.5 ft upstream. The concrete spillway was extensively damaged, and cracking formed in the embankment longitudinally near the crest and transverse to the dam near the abutments. Water levels in five open standpipe piezometers set into the embankment materials

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

TABLE 2-3 Earth and Rockfill Dams Cracked in the Loma Prieta Earthquake (After Harder, 1991)

Dam

Type*

Date Completed

Height (ft)

Epicentral Distance (miles)**

Peak Ground Acceleration (g)

Comments

Austrian

Earth

1950

185

1

0.6

Moderate settlement, transverse and longitudinal cracking

Lexington

Earth

1953

205

2

0.5

Moderate transverse and longitudinal cracking, settlement

Guadeloupe

Earth

1935

142

6

0.4

Moderate cracking

Newell

Earth

1960

182

6

0.4

l-9-inch longitudinal cracks in U/S slope

Chesbro

Earth

1955

95

8

0.4

Moderate longitudinal cracking minor transverse cracking

Soda Lake

Earth

1978

35

10

0.3

Local slumping

Mill Creek

Hyd.

1889

76

12

0.3

Minor longitudinal cracking

Vessey

Earth

1945

20

13

0.3

Minor longitudinal cracking

Anderson

Earth

1931

72

21

0.2

Minor longitudinal cracking

Hawkins

Hyd.

1931

72

21

0.2

Minor longitudinal cracking

* Earth = earth/rockfill embarkment; Hyd. = hydrolically filled.

** (1 mile = 1.6 km)

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

rose after the earthquake, one by as much as 54 ft (16.5 m). It is not clear if these are representative of pore pressure increases in the embankment or other causes, but it was determined that several of the piezometer casings were bent in a manner indicative of spreading of the embankment near the lower portions of the fill. The damages to Austrian Dam are attributed by Harder to settlement and spreading of the fill. Had it not been for a low reservoir, Austrian Dam could have suffered greater damages.

Damages to most of the dams listed in Table 2-3 were caused by shaking and resultant settlement and lateral spreading of the embankment. However, the problems at Soda Creek Dam, and possibly Mill Creek Dam, were attributed to liquefaction. Soda Creek Dam retains tailings and was constructed with some saturated tailings left in the embankment section. It is believed that these materials liquefied during the earthquake, causing the embankment movements. Before the earthquake, it was thought the hydraulic fills within Mill Creek Dam were susceptible to liquefaction, since they were known to have low blow counts (the exact level of blow count could not be identified, since there was some controversy over the way in which the drill holes were supported). After the earthquake, small cracks were observed in the crest area of Mill Creek Dam, but no major damage occurred. The lack of damage was somewhat of a surprise, given the potential believed to exist for liquefaction.

Direct measurements of ground motions were made at eight of the dam sites, providing a resource for future studies. Harder reports that crest amplification effects appeared strongest in the cases where the base motions were the smallest. The reduced amplification in cases with higher accelerations was thought to be caused by increased damping or soil yielding in the embankment's soils that would occur if strains were relatively high. It could also relate to differences that exist between the natural period of the embankment and the characteristic period of the earthquake (higher motions are usually associated with higher frequencies). Two analytical studies of the measured embankment response concluded that finite element analyses could reasonably model the observed behavior (Makdisi et al., 1991; Sayed et al., 1991).

Sanitary And Hazardous Waste Landfills

A large number of major landfills and hazardous waste sites were shaken by the Loma Prieta earthquake. Table 2-4 presents a compilation of information on fifteen of these sites as derived from Sharma and Goyal (1991), Johnson et al. (1991), and Buranek and Prasad (1991) (see Figure 2-8 for locations). The landfills came in varied shapes, with some placed in canyons and others formed in mounds. The heights ranged up to 250 ft (76.7 m), with side slopes from 2:1 to 4:1. The Ben Lomond and Santa Cruz landfills were the closest to the epicenter and were subjected to horizontal accelerations of around 0.4 g. No major damages were observed at any of the landfills, although cracking was found at

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

TABLE 2-4 Major Landfills Affected by the Loma Prieta Earthquake (Adopted from Johnson et al., 1991, and Buranek and Prasad, 1991)

Landfill

Location

Type

Landfill Slopes

Effects of Quake

Peak Ground Accelerations (g)

Kirby Canyon Landfill

San Jose

Canyon fill

200 to 250 ft high, 2:1

No damage to slopes

0.3

Buena Vista Landfill

Watsonville

Modified gravel pit

up to 100 ft high, 3:1

Minor cracks or failures on trash slopes

0.4

Ben Lomond Landfill

Ben Lomond

Side hill fill

up to 150 ft high, 3:1

Minor cracking at contact with natural and on slope benches

0.5

Durham Road

Fremont

Mound

up to 90 ft high, 3:1

No damage

0.1

Newby Island

San Jose

Mound

100 ft, 3:1

No damage

0.1

John Smith

Hollister

Canyon fill

Gentle Slope

No damage

0.2

Ox Mountain

Half Moon Bay

Canyon fill

200 ft, 3:1

No damage to slopes; minor settlement

0.1

Dipauli/Vasco Road

Livermore

Canyon fill

150 ft, 3:1 to 4:1

No damage

0.05

Palo Alto City

Palo Alto

Cut and fill

60 ft, 3:1 to 4:1

Minor settlement at cut/fill contact

0.3

Santa Cruz City

Santa Cruz

Canyon fill

150 ft, 2:1

No failures of slopes; minor cracking at contact between fill and natural

0.5

Guadeloupe Landfill

San Jose

Canyon fill

up to 100 ft, 2.5:1

Minor cracking and downslope movement

0.4

Pacheco Pass Landfill

Pacheco Pass

Rock Site

125 ft, 3.6:1

Minor cracking

0.2

Marina Landfill

Marina

Area fill

up to 90 ft, 3:1

No damage

0.1

Zanker Road Landfill

Freemont

Area fill

up to 75 ft, 3.2:1

No damage

0.2

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

the contact between the landfill materials and the natural ground at sites closest to the epicenter. In computer simulations, Sharma and Goyal found that for landfills up to 50 ft (15.3 m) the base accelerations were increased at the top of the fill, but for landfill heights above this, the landfill acts as a ''damper'' and leads to accelerations less than those input at the base. Johnson et al. (1991) observed that the low density of the landfill material causes it to act as an energy absorber at the boundaries between the landfill and the natural ground. They also suggested that the landfill would offer considerable damping and that random materials within the landfill, such as boards and metal parts, would act as reinforcement for the landfill slopes.

WATERFRONT CONTAINMENT STRUCTURES, PIERS, AND LATERAL RETAINING STRUCTURES

The Loma Prieta earthquake provided the opportunity to evaluate the performance of waterfront containment structures including seawalls, rockfill dikes, levees, cellular cofferdams, piers and wharves, and lateral retaining structures. Rockfill seawalls constructed in the 1800s and early 1900s along the perimeter of San Francisco's waterfront and founded on bay mud generally performed well (notably the fills in these areas apparently were only marginally liquefied). At many locations where fills were located behind the walls, cracks were observed in the streets running parallel to the seawalls, suggesting minor lateral spreading toward the bay. Seawalls along the waterfront of the Marina District are supported on piles, and these exhibited little or no movement. In an adjacent area at the St. Francis Yacht Club, cobblestone seawalls that are founded on spread footings on sand fill suffered considerable damage, with movements of up to 1 ft (0.3 m) toward the bay (Taylor et al, 1992).

At sites where the restrained fills were strongly liquefied, rockfill dikes were not as effective in restraining movement of the fills as those along the San Francisco Waterfront. At Treasure Island, widespread liquefaction in hydraulic fills caused lateral movements on the order of 1 ft (0.3 m) of many of the levees surrounding the island. Similar behavior was observed at rockfill containment dikes along the East Bay shoreline in Oakland and Alameda.

Areas of the waterfront at the Hunter's Point Naval Yard were constructed on cellular cofferdams filled in the 1940s by dumped sands. Large settlements were observed in these structures (Chameau et al., 1991), and distress was evident at the interlocks of the cells. Complete failure occurred in one of the cells that formed the end of the cofferdam. This was caused in part by liquefaction of the fill and deterioration of the cell due to corrosion.

Piers 80 and 94, located along the eastern waterfront of San Francisco, behaved well in the earthquake. These piers are active container ports and were formed first by below-water dredging a wedge-shaped excavation in the underlying bay muds and subsequently by filling the excavated area by barge-dumping

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

sands. The pier structures were supported on piles driven into the sands. The pile driving was expected to densify the sand fills. Tests performed after the earthquake showed the fills to average 60 percent relative density, suggesting the pile driving had densified the fills. Notably, this level of density may not be adequate to prevent damage in an event larger than the Loma Prieta earthquake.

At Seventh Street Terminal in Oakland, lateral movement and failure of a reinforced concrete wharf occurred. The wharf was supported by several vertical piles along the inboard portion of the wharf and by two rows of battered piles along the waterfront edge. The battered piles were designed to restrain lateral movements of the wharf, but they suffered severe damage at their tops during the earthquake, causing significant damages to the wharf. Further insight into the failure of this system is given in the next section, "Foundations."

A variety of lateral retaining structures were subjected to the effects of the earthquake. These included temporary-braced and tied-back walls, crib walls, soil-nailed and reinforced earth walls, and conventional retaining walls. Relatively little attention was given to these systems, since they behaved well in the earthquake. There were no reported failures, although some of the crib walls that supported fills along highway cuts were distressed slightly as a result of settlement of the fill (EERI, 1990).

FOUNDATIONS

The Loma Prieta earthquake provided the opportunity to observe the performance of foundation systems under earthquake loading and delineate differences in their behavior, as well as to test newer generation foundation systems such as base-isolated supports. Although some trends were identified and lessons learned, the fact that the Loma Prieta earthquake was of moderate severity meant that, in many cases, performances of foundation systems did not indicate limit behavior.

There were many cases throughout the bay region where damages to buildings and other constructed facilities varied by foundation type. The most prominent contrast was the behavior of structures founded on deep foundations compared with that of those on shallow footings. For instance, in the Marina and Mission districts of San Francisco, where liquefaction was prominent, structures founded on shallow footings were damaged, while those supported on piles embedded in non-liquefiable materials below the liquefied soils performed satisfactorily. A selective damage pattern that varied with foundation type was also observed in other liquefaction areas south of San Francisco, such as the Moss Landing area. Observations made during the earthquake also indicated that damages appear to be more likely if more than one type of foundation is used to support the same structure. For example, a building at the St. Francis Yacht Club in the Marina District was severely damaged due to large differential settlements between a newer pile-supported part of the structure and an older spread-

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

footing supported portion of the structure. This was also demonstrated at Pier 45 where differential settlements of up to 1 ft (0.3 m) between the floors and walls of two large storage buildings forced closure of the structures. The floors of the building were founded directly on liquefiable sandy fill, but the walls and roofs of the sheds were supported by piles embedded into the bay mud underlying the liquefiable soils.

The earthquake also demonstrated the vulnerability of bridges founded on liquefiable alluvial soils to damage from lateral movements. Failures of this type were primarily located in the Monterey Bay area, near the epicenter. One of the most significant failures involved a seven-span bridge over the Pajaro River that had to be closed after the central pier of the bridge displaced laterally. The foundation soils consist of saturated, fine, and silty sands that liquefied during the earthquake and moved laterally, downslope toward the river. Evidence of liquefaction at the site was in the form of sand boils around the bridge piers, sand ejected from cracks running parallel to the stream, and minor slumping of the ground. Similar behavior was observed at the Neponset Bridge, a railroad bridge spanning the Salinas River. During the earthquake, liquefaction and subsequent lateral movement of liquefied soils toward the river bank caused several inches of lateral displacement of one of the bridge piers and a sharp curvature in the rail track. Interestingly, this bridge, which was built around 1903, suffered almost identical damages during the 1906 earthquake.

The failure of a bridge across Struve Slough on Highway 1 near Watsonville demonstrated that piles embedded in soft subsoils can lead to significant damages during earthquake shaking. The bridge was supported on concrete piles-columns that were embedded into very soft clay and peat underlying the site. During the earthquake, the piles settled and moved laterally, causing plastic deformation and shear failure of the top of the piles (EERI, 1990). There was also evidence of lateral movement of the underlying subsoils. Following the earthquake, gaps of 11.8 to 17.7 inches (30 to 45 cm) were measured between the piles and surrounding soils. The relative lateral displacement between the top of the columns and the base of the columns was estimated at about 30 cm. Although it was initially suspected that the foundation movements were related to liquefaction, there was little hard evidence of liquefaction at the site (no sand boils, etc.). Thus, it is uncertain what role, if any, liquefaction played in producing lateral movements at this site.

The earthquake provided insight into the ability of battered pile foundation systems to resist lateral loadings. As discussed earlier, failure of a concrete wharf occurred at the Seventh Street Terminal in Oakland due to the failure of two rows of battered piles installed along the outboard edge of the wharf to resist lateral loads. The piles were anchored into a dense sand layer, which underlies a rockfill containment dike and a layer of uncompacted hydraulic sand fill. Failure of the battered piles occurred when the containment dike and liquefied hydraulic fill material moved laterally toward the bay during the earthquake (EERI,

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

1990). Notably, none of the vertical piles at the wharf failed, nor did failure occur at adjacent wharves that were constructed with only vertical piles.

The earthquake provided the opportunity to test the performance of a base-isolated structure, the Sierra Point Bridge. The bridge, which is located on Highway 101 in south San Francisco, is 616 ft (188 m) long and supported on 30-inch (0.8-m) diameter concrete columns. In 1985, the bridge was retrofitted with seismic isolators that consisted of neoprene bearing pads that contained a central lead core. The isolators were placed at the top of the columns beneath the superstructure in an effort to reduce the seismic motions that would be induced in the structure during earthquake shaking. It was estimated that the increased period of the structure coupled with the hysteretic damping provided by the isolators would reduce the required seismic shear force by a factor of six (Mitchell et al., 1991). The bridge was designed for a peak acceleration of about 0.6 g at the ground surface. During the earthquake, a peak acceleration of about 0.09 g occurred at the base of the structure, an acceleration of 0.42 g was measured at the top of the columns below the isolator, and a reduced acceleration of 0.33 g occurred in the superstructure above the isolator. No structural damage was observed following the earthquake. Although the isolation system did reduce motions between the tops of the columns and the superstructure, it appears that the bridge did not behave entirely in the manner intended for a base-isolated structure. This is attributed to the fact that the abutments were not modified with a proper seismic gap to allow free movement of the structure on the isolator pads and the fact that the neoprene bearings were very stiff at the relatively low levels of strain induced by the earthquake.

One interesting phenomenon observed in the earthquake involves the relative motions of the supports of some extended structures (Earthquake Engineering Research Center, 1992). It suggests that even where soil conditions are similar, earthquake ground motions can vary between the supports of long-span structures, such as multiply supported bridges. The differential support motions presumably induce forces that are larger than those induced during uniform support motions. This effect is suspected to have been the primary cause of several bridge failures during the Loma Prieta earthquake, although additional research is needed to fully understand this phenomenon.

TUNNELS AND UNDERGROUND STRUCTURES

Underground infrastructure is reasonably well developed in the Bay Area. Most of it is related to lifelines, a topic that is the subject of another paper at this conference. However, it is appropriate in this work to consider the major underground structures such as those associated with the Bay Area Rapid Transit (BART) and the Clean Water Project (CWP) in San Francisco. The BART system is reasonably well known, and it travels underground in portions of the East Bay and San Francisco. BART passes beneath the bay in a submerged tube

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

tunnel system that was placed in a dredged cut and subsequently filled to provide a cover over the tubes. The CWP is less well known. It consists of a massive network of near-surface culverts and tunnels that are linked to several pumping stations. The purpose of the CWP is to allow storage of storm runoff followed by pumping of the runoff to treatment facilities as water flows subside. Figure 2-9 provides a plan view of the location of the BART and CWP systems.

The designers of BART were aware of the need to accommodate earthquake effects in the underground structures (Kuesel, 1968; Douglas and Warshaw, 1971). The BART tubes pass through different types of materials, ranging from rock in some locations to soft bay mud under the bay and as the tubes enter land on the San Francisco side. Considerations were given to special earth loadings and earthquake wave-induced structural deformations for the underground sta-

FIGURE 2-9 Locations of BART tunnels and large CWP culverts.

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

tions and tunnels. However, none of these resulted in any significant design changes, since allowances were also made for larger tolerances for structural stresses during earthquake loadings. Perhaps the most unique feature added to the BART system for earthquake protection was the special "seismic joint" used where the tubes rise from the bay crossing and tie into the land-based ventilation structures. Figure 2-10 shows a section through the seismic joint that allowed for up to 6 inches (15 cm) of relative longitudinal, shearing, and torsional movements between the tunnels and the ventilation structures. Teflon coatings were used on the contiguous steel surfaces to prevent binding. On the San Francisco side, the tubes lie in soft bay mud as they reach the ventilation structure, but on the Oakland side, the tubes pass from bay mud into natural sand at the ventilation structure.

The CWP culverts are large, typically 25 ft (7.7 m) wide and 30 ft (9.2 m) deep, and they lie at shallow depths (average depth of 5 to 10 ft [1.97 to 3.94 m]). The alignment of the culvert takes it along the Marina Green in the Marina District and along the waterfront from the Financial District to the Mission District. This places the culvert fully in fills that are known to contain loose sands and to be subject to liquefaction. The culverts were not pile founded, and they are subject to movement if large portions of the surrounding soils were to liquefy. Consideration of the likelihood of liquefaction led to design of the culverts for the fills acting in a full fluid condition.

FIGURE 2-10 Schematic of seismic joint to connect BART tunnels to ventilation structures on San Francisco and Oakland sides of the San Francisco Bay (adapted from Douglas and Warshaw. 1971).

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

During the Loma Prieta earthquake, the BART and CWP were subjected to motions amplified by the bay mud, and sections of the fills in the Marina. Financial, and Mission districts at least partially liquefied around the CWP culvert. No damages were observed in either the BART or CWP systems after the earthquake, although there was evidence for small permanent movements in a few locations. Inspections of the seismic joints for BART showed no movement on the San Francisco side, but about 0.75 inches (1.9 cm) on the Oakland side (Chiu, 1993). The reasons for the Oakland side showing movement but not the San Francisco side are not obvious. Movements in the CWP culverts were observed at the joints between culvert boxes, and these could have been induced by the earthquake. On the other hand, it is also argued that these may have existed prior to the earthquake and were caused by general settlement. The lack of damage to the CWP culverts passing along the Embarcadero is to be contrasted with the severe damage that occurred to the Embarcadero Freeway structure that was above ground and founded on very deep piles. This illustrates the positive effects found in structures that are able to move with the ground. The CWP culverts as structures also possess a bending resistance that helps to resist local movements. It is not clear whether the CWP culverts would perform as well if extensive liquefaction were to occur under an event larger than the Loma Prieta earthquake.

LESSONS LEARNED FROM THE GEOTECHNICAL ASPECTS OF THE LOMA PRIETA EARTHQUAKE

There are many lessons that can be derived, and are being derived, from the experiences of the Loma Prieta earthquake. More is yet to be gained as further studies are conducted using the information that is still being obtained from investigations of the site conditions at locations where field behavior was documented. Care is required in extrapolation of the lessons learned, so that the factors that served to limit damages in the earthquake are properly included in the studies. This is likely not to be a problem with experienced earthquake engineers. However, there is a real danger that planners or managers who are unfamiliar with the technical details will underestimate the potential for damages in a larger event than the Loma Prieta earthquake or in different conditions with a wetter climatic condition. Consistent messages need to be delivered from studies of the effects of the earthquake. The key lessons learned from this event deal with liquefaction; site amplification; slopes, fills, and embankments; waterfront containment structures, piers, and retaining structures; foundations; underground structures; and areas for future concern.

Liquefaction

  • Liquefaction susceptible conditions were accurately predicted using existing SPT- and CPT-based technology.

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
  • Compacted or improved sand fills were made resistant to liquefaction in the Loma Prieta earthquake, but such improvements did little to reduce ground motions. The earthquake did not provide full design loading for the improved sites.

  • There is a need for improved methods for predicting vertical and lateral movements caused by liquefaction, with allowances for the effects of fines on sand behavior.

  • Broad mapping of liquefaction susceptibility identified areas of potential concern, but details of the subsoil conditions controlled the actual occurrence of liquefaction.

  • Liquefaction, in a number of cases in the earthquake, represented a recurrence of liquefaction that occurred in the 1906 San Francisco event.

  • Densification of sands after liquefaction occurred in a number of loose sandy fills. No densification was observed if the sands were of a medium density.

Site Amplification

  • Although site amplification was most prominent for soft soils, it also occurred in stiff soils and weak rocks under specialized conditions.

  • Site effects not only amplified accelerations but also changed the frequency content of the motions.

  • Primary amplification can occur by matching of the characteristic rock motion with the second fundamental period of the site.

  • Recent design spectra proposed for use with deep soft soil sites did not capture the maximum response periods in areas where site amplification occurred.

  • One-dimensional site-response analysis tools were effective in explaining most ground motion issues, but more sophisticated methods were needed in cases where the underlying bedrock surface was non-uniform.

Slopes, Fills, And Embankments

  • Slope failures left major transportation arteries closed for extended periods of time.

  • Prominent topographical features (bluffs, promontories, etc.) apparently amplified accelerations and in some cases were subject to rapid failure because of resonance effects, wave reflection, and lack of confinement.

  • Coherent, deep-seated sliding occurred only near the seismic energy source zone.

  • Predictions of the amount of movement of coherent landslides under the earthquake loading based on laboratory-derived material parameters were not successful; back-analysis of existing slides appeared to offer a better alternative to obtain material parameters.

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
  • Not all landslides induced by pre-Loma Prieta earthquakes were re-activated by the Loma Prieta earthquake, but incipient landslides were activated by the event.

  • Well-designed earth dams sustained large levels of shaking, but localized damages could have led to problems if reservoir levels had been high.

  • Valuable measurements were made of the acceleration response of earth dams that can be used in future studies.

  • Sanitary landfills performed well in the earthquake, apparently in part due to their high damping capacity.

Waterfront Containment Structures, Piers, And Retaining Structures

  • Seawalls limited ground movements in the presence of moderate liquefaction but were less effective where liquefaction was extensive.

  • Liquefaction of fills contained within, or underlying, pier facilities was the primary source of damages to retaining systems and piers.

  • Conventional retaining structures and more-recently developed systems, such as soil nailed walls, performed well, even where subjected to large accelerations.

Foundations

  • In areas of prominent liquefaction, structures founded on shallow footings were damaged, while those supported on piles embedded in non-liquefiable materials performed satisfactorily. Notably, lateral movements of the soil surrounding the piles were typically less than one foot.

  • Several bridges founded on alluvial soils in the Monterey Bay area suffered damage due to lateral movements of liquefied foundation soils.

  • Structures that were constructed with batter piles to provide lateral restraint suffered heavier damage than adjacent structures with vertical piles.

  • Seismic motions varied between the foundation supports of some long-span structures.

  • Differential support motions presumably induce forces larger than uniform support motions and apparently caused substantial damage to several bridges.

Underground Structures

  • Underground structures other than pipelines were not damaged by the earthquake, because they largely moved with the ground or provided bending resistance to potential lateral movements.

  • The positive performance of the CWP culverts and the BART tunnels maintained the viability of essential portions of the infrastructure in the Bay Area. The lack of damage for the CWP culverts and the BART tunnels at the

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

San Francisco Waterfront was in sharp contrast to the damages to the aboveground Embarcadero elevated freeway system.

  • The CWP culverts performed well even in the presence of liquefied fills, because they were designed for this loading condition. This has to be tempered by the fact that the Loma Prieta earthquake did not produce the large-scale liquefaction that might occur in a 1906-type event.

Areas For Future Concern

Areas for future concern include:

  • the existence of liquefiable sandy fills in the waterfront areas of San Francisco Bay;

  • lack of consideration of site amplification effects in design of many older structures;

  • potential problems that continue to exist with landslides along the Pacific Coast and in the Coastal Mountain Range;

  • the need for better understanding of how to incorporate site amplification effects into design spectra; and

  • the need for re-assessment of what was perceived to be satisfactory performance in the Loma Prieta earthquake relative to what might occur in a larger event.

ACKNOWLEDGEMENTS

The writers would like to gratefully acknowledge the assistance received in preparation of this paper. Information and advice was provided by T. L. Youd, D. Koutsoftas, M. M. Chiu. T. D. O'Rourke, M. S. Power, S. E. Dickenson, J. C. Tinsley, and D. K. Keefer. Assistance was also received from the Earthquake Engineering Research Institute, the Earthquake Engineering Research Center of U.C. Berkeley, and the National Center for Earthquake Engineering of the State University of New York at Buffalo. To all those who helped, we owe a significant vote of thanks.

REFERENCES

Bardet, J.P., M. Kapuskar, G.R. Martin, and J. Proubet. 1992. Site Response Analysis, The Loma Prieta California Earthquake of October 17, 1989—Marina District. U.S. Geological Survey Professional Paper, 155 1-F. Washington, D.C.: United States Government Printing Office.

Bartlett, S.F., and T.L. Youd. In press. Empirical Prediction of Lateral Spread Displacement. Proceedings, 4th U.S.-Japan Workshop on Earthquake Resistant Design of Lifeline Facilities and Countermeasures for Soil Liquefaction," Honolulu, Hawaii, May, 1992.

Boatwright, J., L.C. Seekins, T.E. Furnal, H. Liu, and C.S. Mueller. 1992. Ground-Motion Amplification, The Loma Prieta, California, Earthquake of October 17, 1989—Marina District. U.S. Geological Survey Professional Paper, 1551-F. Washington, D.C.: United States Government Printing Office.

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

Bonilla, M.G. 1992. Geological and Historical Factors Affecting Earthquake Damage. The Loma Prieta. California, Earthquake of October 17, 1989—Marina District. U.S. Geological Survey Professional Paper, 1551-F. Washington. D.C.: United States Government Printing Office.

Borcherdt, R.D., and G. Glassmoyer. 1992. On the Characteristics of Local Geology and Their Influence on Ground Motions Generated by the Loma Prieta Earthquake in the San Francisco Bay Region, California. Bulletin of the Seismological Society of America. 82(2):603-641.

Buranek. D., and S. Prasad. 1991. Sanitary Landfill Performance During The Loma Prieta Earthquake. Pp. 1655-1660, in Proceedings, Second International Conference on Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics, Rolla, Missouri, Vol. II.

Chameau, J.L., G.W. Clough, F. Reyna, and J.D. Frost. 1991. Liquefaction Response of San Francisco Bayshore Fills. Bulletin of the Seismological Society of America, 81 (5):2998-2018.

Chiu, M.M. 1993. Personal communication.

Clough, G.W., and J.L. Chameau. 1980. Measured Effects of Vibratory Sheet Pile Driving, Journal of the Geotechnical Engineering Division, ASCE, 106(GT10):108-1100.

Clough, G.W., and J.L. Chameau. 1983. Seismic Response of San Francisco Waterfront Fills. ASCE Journal of the Geotechnical Engineering Division, ASCE, 109( ):491-506.

Cole, W.F. 1991. Landslides Triggered by the Loma Prieta Earthquake, Implication for Zonation. Pp. 653-660 in Proceedings. Fourth International Conference on Seismic Zonation, Stanford, California, August, 1991.


Dickenson, S.E., R.B. Seed, J. Lysmer, and C.M. Mok. 1991. Response of Soft Soils During the 1989 Loma Prieta Earthquake and Implications for Seismic Design Criteria. Proceedings, Pacific Conference on Earthquake Engineering, Auckland, New Zealand, November, 1991.

Douglas, W.S., and R. Warshaw. 1971. Design of Seismic Joint for San Francisco Bay Tunnel. ASCE, Journal of the Structural Engineering Division, 7:1129-1141.

Dow, G.R. 1973. Bay Fill in San Francisco: A History of Change. San Francisco State University, M.A., Thesis.

Dupre, W.R., and J.H. Tinsley. 1980. Maps Showing Geology and Liquefaction Potential of Northern Monterey and Southern Santa Cruz Counties, California. U.S. Geological Survey, Miscellaneous Field Studies Map, MF-1199.


EERI. 1990. Loma Prieta Earthquake Reconnaissance Report, Earthquake Spectra, Supplement Vol. 6.

Earthquake Engineering Research Center. 1992. News. Vol. 13. No. 2. June.

Egan, J.A., and Z.L. Wang. 1991. Liquefaction-Related Ground Deformation and Effects on Facilities at Treasure Island, San Francisco, During the 17 October 1989 Loma Prieta Earthquake. Pp. 57-76. Proceedings, 3rd Japan-U.S. Workshop on Earthquake Resistant Design of Lifeline Facilities and Countermeasures for Soil Liquefaction, Buffalo, NY, February, 1991.


Harder, L.F. 1991. Performance of Earth Dams During the Loma Prieta Earthquake. Pp. 1613-1629 in Proceedings, 2nd International Conference on Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics, Rolla, Missouri, Vol. II, March, 1991.


Idriss, I.M. 1990. Response of Soft Soil Sites During Earthquakes. Pp. 273-290 in Proceedings, H.B. Seed Memorial Symposium, Vol. 2.


Johnson, M.E., M. Lew, J. Lundy, and M.E. Ray. 1991. Investigation of Sanitary Landfill Slope Performance During Strong Motion From the Loma Prieta Earthquake of October 17, 1989. Pp. 1701-1708 in Proceedings, 2nd International Conference on Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics, Rolla, Missouri, Vol. II.


Keefer, D.K., ed. In press. The Loma Prieta, California, Earthquake of October 17, 1989—Landslides. NEHRP Report to Congress, U.S. Geological Survey, 1993.

Kuesel, T.R. 1968. Structural Design of the Bay Area Rapid Transit System. Civil Engineering, ASCE, April, pp. 1-6.


Mahin, S.A. 1991. The Loma Prieta Earthquake: Implications of Structural Damage LP03. Pp. 1587-1600 in Proceedings, 2nd International Conference on Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics, Vol. II, March, 11-15, Rolla, Missouri.

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

Makdisi, F.I., C.Y. Chang, Z.I. Wang, and C.M. Mok. 1991. Analysis of the Recorded Response of Lexington Dam During Various Levels of Ground Shaking. Seminar on Seismological and Engineering Implications of Recent Strong-Motion Data, Sacramento, Calif., May 30, pp. 10-1 to 10-10.

Mitchell, D., R. Tinawi, and R.G. Sexsmith. 1991. Performance of Bridges in the 1989 Loma Prieta Earthquake—Lessons for Canadian Designers. Canadian Journal of the Geotechnical Engineering Division, No. 18, pp. 711-734, January.

Mitchell, J.K., and F.J. Wentz. 1991. Performance of Improved Ground During the Loma Prieta Earthquake. Earthquake Engineering Research Center Report 91/12, University of California at Berkeley, October.

Mitchell, J.K., T. Masood, R.E. Kayen, and R.B. Seed. 1990. Soil Conditions and Earthquake Hazard Mitigation in the Marina District of San Francisco. Earthquake Engineering Research Center Report 90/08, University of California, Berkeley.

Nolan, J.M., and G.E. Weber. In press. Trenching Studies in the Summit Ridge Area. In Keefer, D.K., ed., The Loma Prieta, California Earthquake of October 17, 1989: Landslides and Stream Channel Change . U.S. Geological Survey.


O'Rourke, T.D., T.E. Gowdy, H.E. Stewart, and J.W. Pease. 1990. Lifeline Performance and Ground Deformations in the Marina During the 1989 Loma Prieta Earthquake. Pp. 129-146 in Proceedings, 3rd U.S.-Japan Workshop on Earthquake Resistant Design of Lifeline Facilities and Countermeasures for Soil Liquefaction, San Francisco.

O'Rourke, T.D., J.W. Pease, and H.E. Stewart. 1992. Lifeline Performance and Ground Deformation During the Earthquake. The Loma Prieta California, Earthquake of October 17, 1989— Marina District. U.S. Geological Survey Professional Paper 1551-F. Washington, D.C.: United States Government Printing Office.


Plant, N., and G.B. Griggs. 1990. Coastal Landslides Caused by the October 17, 1989 Earthquake, Santa Cruz County, California. California Geology, 43(4):75-84.


Sayed, H.S., A.M. Abdel-Ghaflar, and S.F. Masri. 1991. Parametric System Identification and Seismic Performance Evaluation of Earth Dams During the October 17, 1989, Loma Prieta Earthquake. Report No. CRECE-91-03, University of Southern California, Department of Civil Engineering, July.

Scawthorn. C.R.. T.D. O'Rourke, M.M. Khater, and F. Blackburn. 1991. Loma Prieta Earthquake and The San Francisco AWSS: Analysis and Observed Performance. Pp. 527-540 in Proceedings, 3rd Japan-U.S. Workshop on Earthquake Resistant Design of Lifeline Facilities and Countermeasures for Soil Liquefaction, Buffalo, NY, February.

Seed, H.B., F.I. Makdisi. and P. DeAlba. 1978. Performance of Earth Dams During Earthquakes. Journal of the Geotechnical Engineering Division, ASCE, GT-7, July.

Seed, R.B., S.E. Dickenson, and I.M. Idriss. 1991. Principal Geotechnical Aspects of the 1989 Loma Prieta Earthquake. Soils and Foundations, Japanese Society of Soil Mechanics and Foundation Engineering, 31 (1):1-26.

Sharma, H.D., and H.K. Goyal. 1991. Performance of a Hazardous Waste and Sanitary Landfill Subjected to Loma Prieta Earthquake. Pp. 1717-1725 in Proceedings, 2nd International Conference on Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics, Rolla, Missouri, Vol. II.

Sitar, N., and G.W. Clough. 1983. Seismic Response of Steep Slopes in Cemented Soils. Journal of the Geotechnical Engineering Division, ASCE, 109(2):210-227.

Spittler. T.E., and E.L. Harp, compilers. 1990. Preliminary Map of Landslide Features and Coseismic Fissures in the Summit Road Area of the Santa Cruz Mountains Triggered by the Loma Prieta Earthquake of October 17, 1989. U.S. Geological Survey Open-File Report 90-688, Scale 1:4,800.


Taylor, H.T., J.T. Cameron, S. Vahdani, and H. Yap. 1992. Behavior of the Seawalls and Shoreline During the Earthquake. The Loma Prieta California Earthquake of October 17, 1989—Manna District, U.S. Geological Survey Professional Paper, 1551-F, Washington, D.C.: United States Government Printing Office.

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

Tinsley, J.C. 1993. Personal communication.

Tinsley, J.C., and W.R. Dupre. In press. Liquefaction Hazard Mapping, Depositional Facies, and Lateral Spreading Ground Failure in the Monterey Bay Area, Central California. Proceedings. 4th Japan-U.S. Conference on Earthquake Resistant Design of Lifeline Facilities and Countermeasures for Soil Liquefaction, Honolulu, Hawaii, May, 1992.

Youd, T.L., and S.W. Hoose. 1978. Historic Ground Failures in Northern California Triggered by Earthquakes. U.S. Geological Survey Professional Paper 993.

Youd, T.L., and D.M. Perkins. 1987. Mapping of Liquefaction Severity Index. Journal of the Geotechnical Engineering Division, ASCE, 113:1374-1392.

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

DISCUSSANTS' COMMENTS: GEOTECHNICAL ISSUES

William Cotton, William Cotton & Associates

As a geologist, these are the three things I would like to expand upon from Dr. Clough's paper: (1) seismic zonation maps, (2) landslide reactivation, and (3) risk communication. Concerning seismic zonation maps, the ground did behave as expected throughout the Bay Area and throughout the epicentral region. There were maps in place before the earthquake that predicted ground behavior. The predictive kinds of seismic hazards, for which I have maps, include liquefaction-induced ground failures, landslide reactivation, and ground rupture in the Santa Cruz mountains and along the San Andreas fault. There is work being done by Roger Borcherdt at the U.S. Geological Survey and others about trying to predict the response of the ground to certain types of earthquake excitation. Amplification capability maps will come into greater use in the future. From these maps, the state of California has a new law in effect called the Seismic Mapping Hazard Act, which will take the three other hazards—shaking, liquefaction, and landslides—and try to produce maps based upon seismic response. Good ground behavior prediction maps were available for the Marina District, the San Francisco Bay margins, the city of Santa Cruz, and the Watsonville and the Summit Road region of the Santa Cruz mountains.

Concerning landslides, a large number of old deep-seated landslides in the epicentral region were reactivated during the earthquake. Most of these ''coherent'' landslides occurred in the Santa Cruz mountains. These slope failures provided a unique opportunity to evaluate the seismic slope stability and to test the dynamic slope stability methods currently being used in geotechnical practice. In addition, seismic displacement calculations were found to be very sensitive to selected yield coefficients, shear strength values, and acceleration-time history of the ground motion.

The most challenging risk, as I see it, is the transfer of geohazard information to individuals or groups that are charged with mitigating. Engineers and scientists have a poor record of packaging their research results and transferring their knowledge to the public. Thank you.

Maurice S. Power, Geomatrix Consultants

I would like to elaborate on two locations of liquefaction during the Loma Prieta earthquake. These were the Port of Oakland's 7th Street Marine Terminal and Treasure Island. Both are areas where hydraulic fill was placed through bay waters to create land.

During the earthquake the hydraulically placed sand fill liquefied at the marine terminal. As a result, the yard area settled by several inches, as did the rear crane rail, resulting in discontinuation of the crane operation. Lateral spread-

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

ing movements caused extensive damage to the rear batter piles. The Seed-Idriss correlation for liquefaction potential predicted the occurrence of liquefaction. The Port of Oakland has reconstructed the facility replacing the batter piles with octagonal vertical piles. Of particular interest from the geotechnical viewpoint is that the ground has been improved. Post-earthquake ground improvement (vibro-replacement) was used to densify the sand to resist future lateral spreading. Comparison of pre- and post-improvement blow counts in the sand indicates that the sand was effectively densified at the Port of Oakland site using the vibro-replacement technique.

At the Treasure Island site, there was a general subsidence of about 4-6 inches (10.16-15.24 cm) during the earthquake, as well as spreading around the island perimeter. Again, the Seed-Idriss correlation agreed with the observations of liquefaction. Building sites on Treasure Island where pre-earthquake ground improvement had been implemented performed satisfactorily. There was evidence from pre- and post-earthquake survey measurements of an absence of lateral spreading movements at one location on the island perimeter where vibroflotation had been performed, whereas adjacent unimproved areas experienced spreading.

Loma Prieta provided a wealth of recorded ground motion data. Improved attenuation relationships for estimating rock motions have been developed using these data. Correlations for assessing site response effects on ground motions have been developed. Knowledge of dynamic soil properties has been improved, and techniques for measuring dynamic properties have been evaluated. Analytical procedures for assessing site response have been found to give reasonable estimations of ground motions for the levels of excitation of the earthquake. Thank you.

Thomas Hanks, U.S. Geological Survey

The phrase "lessons learned" with respect to earthquakes first came into my consciousness 22 years ago, at the time of the 1971 San Fernando earthquake, when I was a graduate student at an epicentral distance of 62 miles (39 km). Back then, the National Academy of Sciences, which is sponsoring this meeting as well, put out a report entitled something like "Lessons Learned from the 1971 San Fernando Earthquake." We have been learning lessons from earthquakes ever since (and long before, of course), and this is appropriate for a scientific and engineering discipline that relies so heavily on observations, empiricism, and experience. And it is also true that a major earthquake in or near a major metropolitan area will be a learning experience for millions of people, most of whom don't know much about earthquakes.

Nevertheless, there is a certain déjà vu about many of the lessons of the Loma Prieta earthquake that are before us at this symposium. The effects of earthquake strong ground motion, for example, on unreinforced masonry, soft

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

first stories, decayed timbers, bad foundations, hydraulic fill, and young bay mud hardly qualify as news here in San Francisco where these "lessons" had all been learned in 1906, if not before. These things keep happening, though, so we keep talking about them, but are we really making any progress in keeping these things from happening? And if not, why not?

With the latter question in mind, I learned a few things as a result of my own Loma Prieta experience. The first is that the American public, even in very affluent, very well-educated neighborhoods in the heart of earthquake country, is surprisingly uninformed about even the basics of earthquake occurrence, hazards, and risk. The second is that the practice of earthquake hazards reduction is very different from the theory of earthquake hazards reduction. In reality, earthquake hazards reduction is an intensely local happening involving large sums of money that hardly anyone wants to spend unless one absolutely has to, regardless of whether the source of funds is a government agency at the federal, state, or local level, or a neighborhood organization, or a private citizen. Third, an ounce of prevention in this business, like so many others, is worth a pound of cure, a largely unappreciated nicety, because it means spending money when you "don't have to."

I believe we should meet Tom Tobin's challenge: for all of us here and all that we represent, to take advocacy stands and more active roles to inform our citizen colleagues about what we know and to encourage our governing bodies to provide stronger incentives to practice earthquake hazards reduction in advance, so that potential hazards do not become real ones.

C. Thomas Statton, Woodward-Clyde Consultants

The Loma Prieta earthquake had some interesting timing aspects for my career in that we were developing some seismic design provisions for the New York City building code; the earthquake gave greater impetus to the team. But selling the idea of potential earthquake damage to a community that is essentially built, a community that has predominately poor ground on what land is left is not so easy. Our ability as group of scientists and engineers to convince the body politic that earthquakes occur and that earthquake engineering allows one to prevent a disaster is more difficult than perhaps on the West coast.

How does one translate the lessons learned from Loma Prieta to a different environment? One must transfer the context of the lesson as well as the craft itself. The context in the eastern United States and the western United States is quite different. For example, in the west, the sources of earthquakes are understood, but this is not so in the eastern United States. In the east, we deal with large segments of ground, and we call them seismic source zones. In the west, we deal with linear segments of ground, and we call them faults that produce earthquakes.

Another issue that needs to be addressed is that the seismic building code

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×

provisions are primarily derived from the California experience. So the rate of seismicity is important as we translate the lessons to the eastern United States context. When we look at earthquake design in the west for specific structures we find that the design earthquakes represent 80-90 percent of the maximum expected values coming from the maximum expected earthquakes. This is not true in the eastern United States, where the design values that are currently being looked at may represent 50 percent of the ground motions of the maximum expected event. So we may be designing for the same probability of occurrence of ground motion but nowhere near the same probability of being able to withstand the maximum event. In the eastern United States, the larger events that occur so infrequently may collapse structures, whereas in the western United States, the individual building performance will be significantly better.

In the east, the forward-looking view of lessons learned must be tempered by the fact that the east is largely built and built with buildings that have had no seismic attention paid to them, so the issue is how to apply retrofit lessons. Given the design event, we need to find a way to look backwards to a community that's predominately built. Thank you.

Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
This page in the original is blank.
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 29
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 30
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 31
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 32
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 33
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 34
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 35
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 36
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 37
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 38
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 39
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 40
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 41
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 42
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 43
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 44
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 45
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 46
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 47
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 48
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 49
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 50
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 51
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 52
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 53
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 54
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 55
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 56
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 57
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 58
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 59
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 60
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 61
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 62
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 63
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 64
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 65
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 66
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 67
Suggested Citation:"2. The Geotechnical Aspects." National Research Council. 1994. Practical Lessons from the Loma Prieta Earthquake. Washington, DC: The National Academies Press. doi: 10.17226/2269.
×
Page 68
Next: 3. Buildings »
Practical Lessons from the Loma Prieta Earthquake Get This Book
×
Buy Paperback | $50.00
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

The Loma Prieta earthquake struck the San Francisco area on October 17, 1989, causing 63 deaths and $10 billion worth of damage. This book reviews existing research on the Loma Prieta quake and draws from it practical lessons that could be applied to other earthquake-prone areas of the country. The volume contains seven keynote papers presented at a symposium on the earthquake and includes an overview written by the committee offering recommendations to improve seismic safety and earthquake awareness in parts of the country susceptible to earthquakes.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    Switch between the Original Pages, where you can read the report as it appeared in print, and Text Pages for the web version, where you can highlight and search the text.

    « Back Next »
  6. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  7. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  8. ×

    View our suggested citation for this chapter.

    « Back Next »
  9. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!