Arias Intensity Assessment of Liquefaction Test Sites on the East Side of San Francisco Bay Affected by the Loma Prieta, California, Earthquake of 17 October 1989

ROBERT E. KAYEN¹ and JAMES K. MITCHELL²

¹Research Civil Engineer, US. Geological Survey, 345 Middlefield Road, Menlo Park, CA 94025 U.S.A. rkayen@usgs.gov
²University Distinguished Professor of Civil Engineering, Virginia Tech, 109B Patton Hall, Blacksburg, VA 24061-0105 U.S.A. e-mail: jkm@vk.edu

Reprinted from:
Natural Hazards 16: 243-265, 1997.
© 1997 Kluwer Academic Publishers. Printed in the Netherlands.

CONTENTS:

Abstract
Introduction
Methods
Field Investigations
Discussion
Conclusions
References

Abstract. Uncompacted artificial-fill deposits on the east side of San Francisco Bay suffered severe levels of soil liquefaction during the Loma Prieta earthquake of 17 October 1989. Damaged areas included maritime-port facilities, office buildings, and shoreline transportation arteries, ranging from 65 to 85 krn from the north end of the Loma Prieta rupture zone. Typical of all these sites, which represent occurrences of liquefaction-induced damage farthest from the rupture zone, are low cone penetration test and Standard Penetration Test resistances in zones of cohesionless silty and sandy hydraulic fill, and underlying soft cohesive Holocene and Pleistocene sediment that strongly amplified ground motions. Postearthquake investigations at five study sites using standard penetration tests and cone penetration tests provide a basis for evaluation of the Arias intensity-based methodology for assessment of liquefaction susceptibility.

Key words: acceleration, arias intensity, earthquake, geotechnical, liquefaction, field test, penetration resistance, seismometer.

1. Introduction

The 1989 Loma Prieta earthquake (M_s = 7. 1) occurred when a segment of the San Andreas fault northeast of Santa Cruz, Calif. ruptured over a length of approximately 45 km (Borcherdt, 1994; Spudich, 1996). The hypocenter which was located at lat. 37.037º N. and long. 121.883º W. at 18 km depth, and during the next 8 to 10 s ruptured bilaterally for approximately 20 km northward and 20 km southward. The rupture propagated toward the Earth's surface, up to a depth of approximately 5 to 7 km.

Peak horizontal accelerations on sites underlain by rock and stiff alluvium in the east bay generally ranged from 0.08 to 0.12 g, but amplification due to the presence of soft and deep cohesive soil deposits underlying artificial fills produced peak accelerations from 0.11 to 0.29 g at strong-motion recording estimated from strong motion records obtained near the east bay shoreline. On the basis of available strong motion data (Shakal et al., 1989; Kayen et al., 1992), the peak horizontal accelerations on bayshore fills in the vicinity of Oakland International Airport, and Bay Farm Island were apparently about 0.27 g, and at the San Franc isco-Oakland Bay Bridge toll plaza and Port of Oakland's container terminal at Seventh Street were probably about 0.28 to 0.29 g (Figure 1). Arias intensities, integrated from the strong motion records ranged from 0.8 m/s at Oakland International Airport, Bay Farm Island, and Alameda Naval Air Station to 1.71 m/s at the Port of Oakland's container terminal at Seventh Street and the Bay Bridge toll plaza (Kayen and Michell, 1997).

Figure 1: San Francisco east-bay shoreline, showing locations of sites where liquefaction-induced damage occurred from the 1989 Loma Prieta earthquake and locations of study sites where soil investigations were performed. Click on figure for larger image (84K).

At the Port of Richmond, the most northerly study site, no nearby strong-motion recordings were obtained for sites with similar soil conditions (Figure 1). Approximately 2 km to the north, however, a peak horizontal acceleration of 0. 13 g on the south component was recorded at Richmond's City Hall. We used this record, the CSMIP strong motion recording Station No. 58505, to model the site response at the Port of Richmond study site where soft, deep cohesive-soil deposits underly artificial fill, by means of one-dimensional equivalent-linear seismic site-response analyses, using the computer program SHAKE90 (a modified version of the computer program SHAKE; Schnabel et al., 1972). These results indicate that peak horizontal accelerations at the Port of Richmond were probably about 0.18 g. Synthetic seismograms from the SHAKE90 analysis were integrated and we estimate that the Arias intensity at the Port of Richmond site to be approximately 0.6-0.7 m/s.

This paper describes soil liquefaction, ground deformations and associated liquefaction-damages in artificial fill deposits on the east-bay shoreline from Oakland International Airport to the Port of Richmond, and presents the results of our post earthquake studies at five study sites near the east bay shoreline: (1) Oakland International Airport (Figures 1 and 2), (2) Bay Farm Island and Harbor Bay Plaza (Figures 1 and 2), (3) the Port of Oakland's Marine Container terminal at Seventh Street (Figures 1 and 3), (4) the San Francisco-Oakland Bay Bridge Toll Plaza (Figures 1 and 3), and (5) the Port of Richmond (Figures 1 and 4). The collapse of the Interstate Highway 880 Cypress Street Viaduct occurred near sites 3 and 4, along a stretch of viaduct that is founded on Holocene Bay Mud. Liquefaction was not observed at this site and so we do not discuss damage to the Cypress Freeway here. Detailed discussion of the geotechnical aspects of the Cypress Freeway collapse can be found in Seed et al. (1990).

Figure 2: The northern part of Oakland International Airport and Bay Farm Island, showing locations of boreholes (dots). Click on figure for larger image (64K).

Figure 3: Emeryville-Oakland shoreline, showing locations of boreholes (dots) near San Francisco-Oakland Bay Bridge Toll Plaza and Port of Oakland Marine Container Terminal at Seventh Street. Click on figure for larger image (80K).

Figure 4: Port of Richmond and Inner Harbor Basin, showing locations of boreholes (dots). Click on figure for larger image (76K).

Back to Top

2. Methods

Standard penetration tests (SPT) were performed following the guidelines of the American Society for Testing and Materials (1984). The electronic cone-penetration-tests (CPT) followed the procedures of the American Society for Testing and Materials (1986). To remove the influence of effective overburden stress on the field penetration resistance measurements so that we may evaluate the liquefaction potential of soil in a manner that is independent of soil depth, we normalized the field measurements to values at a corresponding reference stress. For example, N values, taken during SPT tests, can be normalized to a reference effective overburden stress of 1 atm ( _vo = 0.096 MPa) by correcting for overburden effects on penetration resistance, using the relation of Seed and Idriss (1971)

We found that the overburden-stress-correction factor, C_n , can be resonably expressed as

where N₁, is the overburden- stress-corrected standard penetration resistance, _vo is the effective overburden-stress, and _ref is a reference stress of 1 atm. Likewise, to evaluate q_c values measured during CPT tests we adjust the measured q_c, values as follows:

We found that C_q can be reasonably expressed as

Liquefiable layers at the Toll Plaza, Port of Oakland, Bay Farm Island, and Oakland International Airport study sites can be characterized as clean sand with a mean grain diameter (D₅₀) of at least 0.25 mm and a fine-particle content of less than 5%. The liquefiable layer at the Port of Richmond study site has a D₅₀ value of approximately 0.06 to 0.07 mm and a fine-particle content of >35%. Silty sands, like those at the Port of Richmond site, with the same liquefaction resistance to that of a corresponding reference 'clean' sand are typically observed to have distinctly lower measured penetration resistance values. Seed et al. (1984) presented boundary curves that effectively convert penetration resistance values of silty sands to that for clean sands with no fine particles, and defined a SPT-fines content correction factor (Seed and de Alba, 1986):

(N1, fc )clean sand equivalent = (N1) silty sand measured + N1.

(5)

The fines-correction factor N₁ is a function of the fines content, FC, in percent. Following the guidelines presented in the consensus statement from the NCEER Workshop on Evaluation of Liquefaction Resistance, 4-5 January 1996, Salt Lake City, Utah, we used the following equations for correcting SPT-values of low-plasticity silty sands to equivalent values for clean sand

We applied these correction factors to the Port of Richmond site SPT data. Detailed description of the borehole logs from all of the liquefaction test sites, including those discussed herein, can be found in Kayen et al. (1992, 1997), Kayen (1993) and Mitchell et al. (1994).

Back to Top

3. Field Investigations

Oakland International Airport: Soil liquefaction caused considerable damage to the main jet runway (11-29) at Oakland International Airport (Figure 2). Additional evidence of liquefaction, including sand boils, settlement and lateral spreading, occurred over wide areas of airport fill to the north, south and east of the damaged runway section. Much of the runway and inboard taxiway area is built on loose, sandy fill underlain at shallow depths by soft bay mud. The perimeters of the airport fill have dikes to prevent inundation during unusually high tides and storms.

Extensive soil liquefaction occurred in the western section of the airport fill, damaging the northwestern 900 m of the 3000 m-long main runway. The adjacent taxiway pavement was also heavily damaged. Most of the runway damage was repaired within 4 weeks, and as a result, the airport was able to resume essentially full operations with a shortened operational runway 2700 m long on 20 November 20 1989.

In addition to liquefaction, settlement, and lateral spreading of the main runway fill, settlement and lateral spreading also occurred in several places in the surrounding perimeter dikes at the west end of the runway fill. The maximum observed levee settlement of the perimeter dikes was approximately 0.5 to 0.7 m, and lateral deformations were similar in magnitude. Liquefaction was also observed near the main terminal buildings.

The logs of three boreholes drilled at the Airport show little or no surficial crust between 0 and 2 m depth, where the water table was penetrated (log ACPT7 is shown in Figure 5). From 2 to 4.5-5.8 m depth, is an extremely loose deposit of fine-sand hydraulic fill that almost certainly was responsible for the observed liquefaction and lateral spreading. q_c, values within this deposit range from 2 to 14 MPa, and N values from 1 and 4 blows/ft. The material in this layer correlates with sand boil material observed at the surface.

Figure 5: Log of borehole ACPT7 in fill deposits at Oakland International Airport test site (see Figure 2 for location). Click on figure for larger image (60K).

Bay Farm Island and Harbor Bay Plaza: At Bay Farm Island, immediately north of Oakland International Airport, considerable liquefaction occurred at the northwest corner and at points along the west edge of the fill (Figures 1 and 2). Most of the western part of Bay Farm Island consists of sandy hydraulic fill, underlain by bay mud and deeper, stiffer alluvium and estuarine sediment. Soil liquefaction on the island was largely confined to undeveloped, artificially filled land. Numerous sand boils, many relatively large (with diameters more than 3 m) occurred in this area. Fill in a perimeter dike that bounds the western side of Bay Farm Island was densified by dynamic compaction, which appears to have successfully prevented soil liquefaction of the dike during the earthquake.

The logs of three boreholes drilled at the Bay Farm Island study sites suggest a complex stratigraphy of relatively thin (less than 1 m thick) interbeds of sand and finer material. A total of 20 CPT logs were also taken at one of the liquefaction test sites to observe any changes in strength over time; a representative log (HBI-CPT 1) is presented in Figure 6. The logs in the unimproved fill indicate a noncohesive layer of sand and silty sand at 2 to 3 m, with q_c values of typically 2 to 16 MPa and N values of 8 to 18 blows/ft. The water table is within this layer at between 1.5 and 2 m depth. Below this layer, to approximately 6.5 m depth, is a set of finer interlayered deposits exhibiting decreasing q_c values, and elevated B_q and FR values. Below 6.5 m depth, silty sand q_c defines the rest of the logged soil column. Liquefaction probably occurred in the fine sand and silty sand at 2 to 3 m depth. on the basis of correlation of sand boil materials with recovered samples.

Figure 6: Log of borehole HBI-CPT1 in fill deposits at Bay Farm island test site (see Figure 2 for location). Click on figure for larger image (60K).

CPT and SPT logs were taken along the improved perimeter dike, which did not liquefy during the earthquake. At the elevated dike structure, the water table is somewhat deeper in the soil profile at 2.5 m depth. However, within the sandy deposits at 2.5 to 7 m depth, q_c values are noticeably higher, ranging from 10 to 3 5 MPa, and N values of 29 to 58 blows/ft.

The Port of Oakland: Immediately south of the San Francisco-Oakland Bay Bridge, soil liquefaction caused considerable damage to marine-container facilities at several places in the Port of Oakland, adjacent to the Oakland Outer Harbor, as shown in Figures 1 and 3. Much of the extreme western part of the area near the Oakland shoreline, south of the mole, is filled land underlain by a relatively thin layer of soft, normally consolidated Holocene marine clay (bay mud). These shallow surficial units are, in turn, underlain by upwards of 180 m of stiffer Pleistocene glacial and interglacial deposits.

Most of the surface fill at the Port of Oakland was hydraulically placed to sea level, above which a combination of hydraulic and dumped fill was placed. After placement of the unimproved hydraulic fill, a relatively thick asphalt-cement pavement was laid in many of the terminal areas to support heavy vehicles and shipping containers. All of the Port terminals have pile-supported concrete wharves at the loading-dock edge of the terminal fill. The most severe liquefaction-induced damage to Port facilities occurred at the Marine Container Terminal at Seventh Street. Liquefaction of the hydraulic fill resulted in settlement, lateral spreading, and cracking of the pavement over large areas of the terminal. Maximum settlements of the paved container yards inboard of the wharves were about 0.3 m. Several cranes which operate along the edges of the fill, traverse laterally along the wharves on heavy rails. The outboard rail, which was pile supported on the concrete wharf, did not settle appreciably. The inboard crane rail, however, which was supported on fill throughout the entire terminal, underwent differential settlement, damaging several of the loading cranes and the dock facility was inoperable after the earthquake.

The logs of six boreholes drilled at the Port of Oakland study site (Figure 2) near the Marine Container Terminal at Seventh Street were used to assess the liquefaction susceptibility of the Port. We note that evidence of liquefaction of the subroad base material appears to have extended across essentially the entire western part of the Marine Container Terminal, as evidenced by settlement, sand boils, and tension cracking of the asphalt. The logs of boreholes P007-2 and P007-3 are shown in Figures 7 and 8. These sites typically have a resistant surface layer in the upper 3 to 4 m with q_c values typically 25 to 35 MPa and N values 25 to 36 blows/ft. The water table within this layer averages approximately 2 to 2.3 m depth and fluctuates with tidal action. Below this surface layer, the fill consists of looser deposits of fine marine sand with q_c values of 8 to 15 MPa, and N values of 10 and 25 blows/ft. Liquefaction appears to have occurred in the materials with in the depth range 4-8 m, on the basis of correlations of SPT samples with surface sand-boil material recovered.

Figure 7: Log of borehole POO7-2 in fill deposits at Port of Oakland (see Figure 3 for location). Click on figure for larger image (60K).

Figure 8: Log of borehole POO7-3 in fill deposits at Port of Oakland (see Figure 3 for location). Click on figure for larger image (64K).

San Francisco-Oakland Bay Bridge Toll Plaza: The San Francisco-Oakland Bay Bridge mole (peninsula-approach fill), immediately south of Emeryville, was extensively damaged by soil liquefaction (Figures 1 and 3). Appreciable settlement (max. 40 cm) occurred over most of the mole. Lateral spreading also was significant along most of the fill, causing numerous longitudinal fissures in the pavement parallel to the fill edges. Settlements of the mole and the elevated structures of Interstate Highway 580 and West Grand Avenue were severe, resulting, at each site, in pavement collapse and open fissures. as much as 0.3 m wide, at the soil-structure interface. The approach mole also settled below the bridge- and ramp-structure road level by as much a 0.5 m. The fissures and uneven pavement surface interrupted ground transportation in the interchange areas after the earthquake.

The logs of five boreholes drilled at the San Francisco-Oakland Bay Bridge Toll Plaza study site indicate a complex stratigraphy. Five sets of logs were taken at locations along the median strip and open space between the east bound and west bound lanes of Interstate Highway 80. The log of borehole SFOBB-1 was taken at the easternmost end of the Toll Plaza parking lot. The log is marked by a moderately dense surface crust between 0 and 2.8 m depth, followed by an apparently interbedded sequence of cohesionless silty sand and cohesive clay to 4.5 m depth, as noted by the local variations in FR and q_c values. The water table during sampling was in the upper part of the section at approximately 2 m depth. Between 4.5 and 7.5 m lies cohesionless material with low q_c values typically 5 MPa and N values of typically 5 to 17 blows/ft with a maximum N value of 28 blows/ft at 5.2 m, and a minimum N value of 0 blows/ft (drill rod and hammer sank under their own weight) at 6.6 m depth. This layer appears to have been responsible for the observed liquefaction, on the basis of both inplace testing and correlation with surface sand-boil material.

Figure 9: Log of borehole SFOBB-1 on east shoreline of San Francisco-Oakland Bay Bridge mole (see Figure 3 for location). Click on figure for larger image (68K).

The logs taken of two boreholes SFOBB-3 and SFOBB-4 (not shown), drilled in the central part of the median strip directly south of the Toll Plaza and maintenance buildings, show similar soil profiles. The upper 4 m sections of these two logs suggest moderately loose sand (q_c = 5-15 MPa) interbedded with finer materials. Below, there is a somewhat uniform, more sand rich deposit to 8-m, with q_c, values of typically 4 to 15 MPa. This layer appears to be the principal liquefied layer. No SPT measurements were made at boreholes SFOBB-3 or SFOBB-4. The logs of two boreholes, SFOBB-2 and SFOBB-5 (Figure 10), in a wooded area at the west end of the median strip show gravelly sand to 6 m depth beneath a dense crust. From 6 to 8.5 m depth is a layer of silty and sandy soil, q_c values of 5 to 10 MPa and N values of 7 to 19 blows/ft, within which liquefaction apparently occurred.

Figure 10: Log of borehole SFOBB-5 on east shoreline of San Francisco-Oakland Bay Bridge mole (see Figures 3 for location). Click on figure for larger image (60K).

Port of Richmond: Soil liquefaction occurred at a study site in the western part of Richmond Inner Harbor (Figures 1 and 4). The area that liquefied is a zone approximately 75 m wide and 300 m long at the foot of Harbor Way Road (10th Street). Approximately 85 km north of the Loma Prieta rupture zone, this site represents the most distant point from the zone of energy release to undergo soil liquefaction sufficient to damage structures. Much of the land at the edge of the Richmond Inner Harbor had been created by placement of uncompacted sandy hydraulic fill. The liquefaction zone overlies deposits of soft bay mud, which are underlain, in turn, by deeper deposits of stiffer, overconsolidated noncohesive and cohesive soils that naturally fill a deep fluvial channel at the west end of the harbor. Four large and a dozen smaller sand boils vented fine sand and silty sand from the underlying fill. In addition, minor settlements of approximately 2 to 8 cm and fissures of similar magnitude occurred at the edge of the harbor.

Figure 11: Log of borehole POR-2 in fill deposits at Port of Richmond test site (see Figure 4 for locations). Click on figure for larger image (56K).

Three boreholes (POR-2 through POR-4) drilled in the open space north of the Harbor seawall (Figures 1 and 4). SPT's were performed in shallow boreholes adjacent to each CPT. The logs of the boreholes drilled at the Port of Richmond study site (Figure 2) are all similar and are characterized in the upper part of the section by an oxidized, tan-brown crusted silty sand to 0.8 m depth; oxidized tan-brown silty sand to 1.8 m depth; and oxidized, tan-brown sandy clay to reduced sandy clay to 4.1 m depth. The water table was approximately at a depth of 2.5 m during sampling. This section has low q_c values of approximately 4 to 5 MPa, FR values greater than 3% and pore pressure generation, and N values ranging from 1 to 6 blows/ft. Below this section, at 4.1 to 7.8 m depth, is a hydraulic-fill layer of olive gray fine-silty sand containing shell fragments. This layer almost certainly was responsible for the observed liquefaction at the surface, and surface sand-boil material is identical to SPT samples of this layer. This layer has extremely low q_c values 1 to 3 MPa, average FR values of 0.3 to 1.0%, and minor pore pressure generation during cone penetration, and N values of 2 to 11 blows/ft, but typically, 2 to 5 blows/ft. Below this layer is a thin deposit of soft bay mud to 9.5 m depth, underlain by dense sand deposits.

Back to Top

4. Discussion

We compared both the observed liquefaction behavior and the liquefaction potential of the study sites, as measured by q_c and N, with the measured earthquake shaking intensity recorded at nearby seismometer sites. Two approaches were used to assess the intensity of seismic shaking: the first approach used the conventional cyclic stress method (CSR) of Seed and others (1984); and the second approach utilizes a new method based on the Arias intensity of ground motion recorded near the liquefaction study sites (Kayen and Mitchell, 1997). The first method utilizes a simplified uniform cyclic stress ratio to represent the complex and irregular earthquake-induced stress-time history. This equivalent series of cyclic loads of uniform amplitude is expressed as follows (Seed and Idriss, 1982)

where a_max is the peak ground acceleration, g is the gravitational acceleration (9.81 m/s²), _vo is the total overburden stress, _vo is the effective overburden stress, and r_d is a depth-reduction factor, which can be estimated in the upper 10 m of the soil column by the following relation:

where z is the depth in meters (Kayen et al., 1992).

The second approach utilizes a quantitative measure of earthquake shaking intensity, termed Arias intensity (Arias, 1970). It is the sum of the energy absorbed by an evenly-spaced population of idealized undamped simple oscillators in response to the earthquake motion. For the two horizontal components of motion, the Arias intensity I_h is calculated as follows:

where I_xx is the Arias intensity measured in the x direction; I_yy is the Arias intensity in the y direction; g is the acceleration due to gravity; t₀ is the duration of earthquake shaking; and a_x and a_y are the transient accelerations in the x and y directions. The Arias intensity measure has the dimensional units of velocity.

As in the case of a_max, we find that I_h typically diminishes with depth in the soil column. For liquefaction assessment we need to know the I_h profile within the soil column. We estimated the attenuation of I_h with depth through a parametric study of synthetic accelerograms propagated through shear modulus representations of soil columns at our east bay study sites. Synthetic seismograms were generated for the surface and depth nodes of soil column input files with the ground-response computer-program SHAKE90, a modified version of the program presented by Schnabel et al. (1972). Strong motion records from eight earthquakes were propagated through a soil column representing loose sandy fill (V_s = 150 m/s.) overlying a cohesive soil with shear wave velocity profiles for San Francisco bayshore sites (Sun et al., 1989; Golesorkhi, 1989; Kayen, 1993). Each profile was underlain by an elastic half-space with a shear wave velocity of 2500 m/s, representative of the Mesozoic Franciscan Formation underlying the east bay sites. Output synthetic acceleration-time histories for layer nodes at depth in the model soil profiles were integrated to calculate I_h and then normalized by the value of I_h at the surface. This normalization process allows us to evaluate the depth-dependency of I_h over a broad range of earthquake and site conditions by collapsing the profiles to a common reference value (unity) at the ground surface. We define the Arias intensity depth-of-burial reduction parameter, r_b, as the ratio of the buried-to- surface Arias intensity.

The r_b parameter is analagous to r_d and can be calculated from either one- or two-component horizontal Arias Intensity. The mean and ±1 responses calculated from the suite of SHAKE90 runs are plotted in Figure 12. The mean attenuation of r_b falls somewhat linearly from 1.0 to a value of 0.58 at 6 m depth, then decreases to 0.46 at 10 m. The value of r_b remains essentially constant at depths below 10 meters (30.5 ft). We calculated the Arias intensity in the soil column, I_hb, as the product of I_h measured from recorded seismograms in the east bay, and the reduction parameter r_b estimated from the SHAKE90 study.

Figure 12: Arias intensity depth reduction profiles modeled using the ground response program SHAKE are presented in Figure 12a. A statistical synthesis of the SHAKE-study is presented in Figure 12b. Click on figure for larger image (40K).

Figure 13: (a) Ihb versus (N1)60 for sites in the United States and Japan with fines content correction to equivalent 'clean sand'.; (b) Ihb versus qc1 for sites listed in Table 1. Boundary curves for D50 > 0.25 mm and D50 < 0. 15 mm are based on qc - N relation of Robertson and Campanella (1985). Click on figure for larger image (108K).

In our study, we associate CSR and I_hb profiles with the liquefaction field performance data (N₁)₆₀, and q_c1 from our study sites (Figure 13). Detailed discussion of the Arias intensity method for liquefaction potential assessment is presented in Kayen (1993), Kayen and Mitchell (1997), and Kayen et al. (1997). CSR and I_hb values used in the analyses of critical soil layers at the east bay study sites (Figure 1) are summarized in Table 1.

The SPT boundary curve for the assessment of soil liquefaction potential based on Arias intensity is presented in detail in Kayen and Mitchell (1997) and is presented here for soils with low-plasticity fines content of less than 5 % (Figure 13a). An equivalent cyclic stress ratio based boundary for assessment of liquefaction potential was presented by Seed and de Alba (1986). The insensitivity of CSR to earthquake magnitude and duration requires that magnitude correction factors be applied to CSR-values to scale for the severity of the earthquake shaking. This results in the suite of magnitude dependent boundary curves. Recent work questions the CSR boundary curves presented in Seed and de Alba (1986) as derived from the scaling factors of Seed and Idriss (1982) (Ambraseys, 1988; Arango, 1996). Arias intensity incorporates magnitude, frequency and earthquake shaking duration elements of strong motion and therefore needs no magnitude correction factors.

Table 1: Representative values of parameters used in liquefaction-susceptibility analyses for critical soil layers at east bay study sites (see Figures 1, 2, 3, and 4, for locations of boreholes. PGA, peak ground acceleration; qc1, normalized cone-penetration resistance; (N1)60, 60%-hammer-efficiency normalized standard penetration resistance; CSR, cyclic stress Ratio; Ihb, Arias intensity; fc fines content; (N1)60, fc fines content corrected SPT-resistance). Click on table for larger image (76K).

The I_hb at the liquefaction boundary in Figure 13A represents the minimum threshold seismic intensity (I_hb,₁) required to induce liquefaction in soil of a given (N₁)₆₀ or q_c1 value. A similar threshold in terms of CSR can be determined based on the boundaries presented by Seed and de Alba (1986). CPT-based boundary curves for CSR have been proposed by Robertson and Campanella, 1985; Seed and de Alba, 1986; Shibata and Teparaksa (1988) and Mitchell and Tseng (1990). The proposed boundaries of Robertson and Campanella (1985) and Seed and De Alba (1986) are based on q_c/N ratios rather than direct field investigations of liquefaction test sites. The method of Shiibata and Teparaksa (1988) is based on a limited data set of direct measurements at liquefaction sites for earthquakes in Japan, China, and the United States. Mitchell and Tseng (1990) presented boundary curves developed from q_c - D_r, (relative density)-CSR relationship at a given effective confining stress which are based on cavity expansion theory and cone chamber test results.

We compared CSR₁ and I_hb ,₁ with the corresponding seismic-shaking intensity function induced by the earthquake (CSR_eq and I_hb , _eq) using Equations (7) through (10) to determine profiles of factor of safety against liquefaction. The Arias intensity based factor of safety FI_hb against liquefaction occurrence is defined as the ratio of the Arias intensity required to cause liquefaction to the Arias intensity imparted by the earthquake.

This Arias intensity definition of factor of safety is analogous to that defined by Seed and his colleagues based on cyclic stress (F_CSR).

We converted the N profiles from Figures 5-11 into factor of safety profiles for both FI_hb and F_CSR in Figure 14, based in the calculated CSR and I_hb for the study sites. At Bay Farm Island both methods accurately predict no-liquefaction at the improved-dike and liquefaction at between 3.5 and 4.5 m at HBI-CPT1, and at Oakland International Airport both methods predict liquefaction in a zone between approximately 2 and 5 m. The profiles at the Port of Oakland for POO7-2 are in agreement and identify the zone of liquefaction between 5 and 8 m, and POO7-3 indicates thins seams of loose potentially liquefiable material at 6.5 and 7.5 m, but otherwise a profile of non-liquefaction susceptibility. Profiles of F_CSR and FI_hb for the San Francisco-Oakland Bay Bridge site correctly indicate that liquefaction occurred principally at a depth of between approximately 5.5 and 7.5 m at SFOBB-1 and between 6.0 and 9.0 m at SFOBB-5, based on field observations. At the Port of Richmond, the profiles accurately identify the zone of liquefaction between 4 and 5 m in the profile POR-2. At POR-Hall St., both profiles accurately indicate no liquefaction.

Figure 14: Comparison of the factor of safely against liquefaction (FS 1.0 = liquefaction; FS > 1.0 = no liquefaction) for the study sites on the east side of San Francisco Bay using the Arias intensity and the Cyclic Stress methods. Click on figure for larger image (148K).

The CPT data used to construct the I_hb - q_c1 association in Figure 13b are derived almost exclusively from the Loma Prieta dataset. Thus, FI_hb profiles when presented in a manner similar to Figure 14, are entirely consistent and in agreement with the results presented in Figure 13b where soil layers are summarized as individual I_hb / q_c1 points. The new Arias intensity boundary for liquefaction assessment, presented here, is therefore based on a limited data set from the Loma Prieta earthquake. This boundary will enhanced by new liquefaction field data that become available.

Back to Top

5. Conclusions

Extensive soil liquefaction occurred during the 1989 Loma Prieta earthquake in uncompacted artificial-fill deposits of the east bay from Oakland International Airport to the Port of Richmond, from 65 to 85 km from the north end of the rupture zone. We present the results of studies at five sites near the east bay shoreline: at the Port of Richmond, San Francisco-Oakland Bay Bridge Toll Plaza, Port of Oakland's Seventh St. Marine Container Facility, Bay Farm Island and Harbor Bay Plaza, and Oakland International Airport. Typical of all these sites are extremely-low to low N and q_c , values in layers of cohesionless hydraulic fill that overlie deep, primarily cohesive soil deposits. Two factors, low penetration resistance and amplification of seismic shaking by the underlying soils, combined to give the sites a relatively high liquefaction susceptibility.

A direct comparison was made of the cyclic stress and Arias intensity-based methods for assessing the potential of soil liquefaction using the SPT field test. We found that when using SPT data, the two methodologies are largely in agreement in predicting the occurrence or nonoccurrence of liquefaction in soils layers at our study sites during the earthquake. Field data from the Loma Prieta study sites allow us to construct a new Arias intensity-cone penetration resistance boundary for liquefaction assessment.

For liquefaction assessment, the advantages in using Arias intensity over peak ground acceleration (a_max), used in the cyclic-stress approach, are that (1) Arias intensity is derived from the acceleration records of both horizontal components of motion over the entire duration of motion, whereas a_max utilizes a single arbitrarily selected value; (2) Arias intensity incorporates the intensity of motions over the full range of recorded frequency, whereas, a_max is often associated with high-frequency motion; and (3) the breakdown of soil structure that results in liquefaction is fundamentally more dependent upon input-energy than on a single level of acceleration (Liang, and others., 1996).

References

Ambraseys, N. N.: 1988, Engineering seismology, Earthquake Eng. Struct. Dynam. 17(l), 1-105.

American Society for Testing and Materials: 1984, Standard method for penetration test and split barrel sampling of soils, Annual Book of ASTM Standards, Section 4, Volume 04.08, D 1586-84. American Society for Testing and Materials: 1986, Standard method for quasi-static, cone and friction cone penetration test of soil, Annual Book of ASTM Standards, Section 4, Volume 04.08, D3441-86. American Society for Testing and Materials: 1986, Standard test method for stress wave energy measurement for dynamic penetrometer testing systems: Annual Book of ASTM Standards, Section 4, Volume 04.08, D4633-86, pp. 943-946.

Arango, I.: 1996, Magnitude scaling factors for soil liquefaction evaluations, ASCE J Geotech. Eng.122(11), 929-936.

Borchedt, R D. (ed.): 1994, The Loma Prieta, California, Earthquake of October 17, 1989-Strong Ground Motion, U.S. Geological Survey Professional Paper 1551-A.

Campanella, R. G. and Robertson, P. K.: 1982, State-of-the-art in in-situ testing of soils; developments since 1978, Engineering Foundation Conference on Updating Subsurface Sampling of Soils and Rocks and Their In-Situ Testing, Santa Barbara, Calif, Proceedings, p. 23.

Farrar, J. A.: 1991, Field Energy Measurements of Standard Penetration Testing, M.S. Thesis, Denver, University of Colorado, Denver, 400 pp.

Hardin, B. 0. and Drnevich, V P.: 1972, Shear modulus and damping of soils: Measurement and parameter effects, Journal of Soil Mechanics and Foundation Division ASCE 98(SM6), Proc. Paper 8977.

Kayen, R. E.: 1993, Accelerogram-Energy Approach for Prediction of Earthquake-Induced Ground Liquefaction, PhD Dissertation, University of California, Berkeley, 289 pp.

Kayen, R. E., Mitchell, J. K., Lodge, A., Seed, R. B., Nishio, S., and Coutinho, R.: 1992, Evaluation of SPT-, CPT-, and shear wave-based methods for liquefaction potential assessment using Loma Prieta data: Proceedings of the Fourth Japan - U.S.Workshop on Earthquake Resistant Design of Lifeline Facilities and Countermeasures for Soil Liquefaction, in M. Hamada and T. D. O'Rourke (eds.), Technical Report NCEER-92-0019, Vol. 1, pp. 177-204.

Kayen, R. E., Mitchell, J. K., and Holzer, T. L.: 1994, Ground motion charateristics and their relation to soil liquefaction at the Wildlife Liquefaction Array, Imperial Valley, California: Proceedings of the Fifth U.S. - Japan Workshop on Earthquake Resistant Design of Lifeline Facilities and Countermeasures for Soil Liquefaction, in T. D. O'Rourke and M. Hamada (eds.), Technical Report NCEER-94-0026, pp. 267-283.

Kayen, R. E. and Mitchell, J. K.: 1997, Assessment of liquefaction potential during earthquakes by Arias intensity, Am. Soc. Civil Eng.: J Geotech. Geoenviron. Eng. 123(12), 1162-1174.

Kayen, R. E., Mitchell, J. K., Seed, R. B., and Nishio, S.: 1998, Soil Liquefaction along the eastern shoreline of San Francisco Bay during the Loma Prieta Earthquake, 17 October 1989, in T. L. Holzer, (eds.), The Loma Prieta, California, Earthquake of October 17, 1989, Professional Paper II - Earthquake Effects: U.S. Geological Survey Professional Paper 1551-B, NEHRP Report to Congress.

Mitchell, J. K. and Tseng, D.-J.: 1990, Assessment of liquefaction potential by cone penetration resistance, in J. M. Duncan (ed.), H. Bolton Seed Memorial Symposium, Vol. 2, Berkeley, Calif, BiTech, pp. 335-350.

Mitchell, J.K., Lodge, A. L., Coutinho, R. Q., Kayen, R. E., Seed, R. B., Nishio, S, and Stokoe, K. H. II: 1994, In situ test results from four Loma Prieta earthquake liquefaction sites: SPT, CPT, DMT and Shear Wave Velocity: Berkeley, University of California Earthquake Engineering Research Center Report UCB/EERC-94/04, 179 pp.

Robertson, P.K., 1986. In situ testing and its application to foundation engineering: Can. Geotech. J 23(4), 573-594.

Robertson, P. K.: 1990, Soil classification using the cone penetration test, Can. Geotech. J 27(l), 151-158.

Robertson, P. K. and Campanella, R. G.: 1985, Liquefaction of sands using the CPT, Am. Soc. Civil Eng. Proc., Geotech. Div. J 111 (GT3), 3 84-403.

Robertson, P. K., Campanella, R. G., Gillespie, Don, and Rice, A.: 1986, Seismic CPT to measure in situ shear wave velocity, Journal of Geotechnical Engineering 112(8), 781-803.

Schnabel, P. B., Lysmer, John, and Seed, H. B.: 1972, SHAKE - a computer program for earthquake response analysis of horizontally layered sites, Report UCB/EERC-72/12, Earthquake Engineering Research Center, University of California, Berkeley.

Seed, H. B. and de Alba, Pedro: 1986, Use of SPT and CPT tests for evaluating the liquefaction resistance of sands, in S. P. Clemence (ed.), Use of In-Situ Tests in Geotechnical Engineering, American Society of Civil Engineers, New York, pp. 281-302.

Seed, H. B. and Idriss, I. M.: 1971, Simplified procedure for evaluating soil liquefaction potential, Am. Soc. Civil Eng. Proc., Soil Mech. Found Div. J 97(SM9), 1249-1273.

Seed, H. B. and Idriss, I. M.: 1982, Ground Motions and Soil Liquefaction during Earthquakes, Earthquake Engineering Research Institute, University of California, Berkeley, 139 pp.

Seed, H. B., Idriss, I. M., and Arango, Ignacio: 1983, Evaluation of liquefaction potential using field performance data, J Geotech. Eng. ASCE 109(3), 458-482.

Seed, H. B., Tokimatsu, Kohji, Harder, L. H., and Chung, R. M.: 1984, The influence of SPT procedures in soil liquefaction evaluations, Report UCB/EERC-90/05, Earthquake Engineering Research Center, University of California, Berkeley, 50 pp.

Seed, R. B., Dickenson, S. E, Reimer, M. F., Bray, J. D., Sitar, Nicholas, Mitchell, J. K., Idriss, I. M., Kayen, R. E., Kropp, Alan, Harder, L. F., Jr., and Power, M. S., 1990, Preliminary report on the principal geotechnical aspects of the October 17, 1989 Loma Prieta earthquake, Report UCB/EERC-90/05, Earthquake Engineering Research Center, University of California, Berkeley, 137 pp.

Shakal, A. F., Huang, M. J., Reichle, Michael, Ventura, C. E., Cao, Tianging, Sherboume, R. W., Savage, M. K., Darragh, R. B., and Petersen, C. P.: 1989, CSMIP strong-motion records from the Santa Cruz Mountains (Loma Prieta), California earthquake of 17 October 1989, Report OSMS 89-06, Office of Strong Motion Studies, California Division of Mines and Geology, 196 pp.

Shibata, Toru and Teparaksa, Wanchai: 1988, Evaluation of liquefaction potentials of soils using cone penetration tests, Soils and Foundations, 28(2), 49-60.

Spudich, P. (ed.): 1996, The Loma Prieta, California, Earthquake of October 17, 1989 - Main-shock characteristics, U.S. Geological Survey Professional Paper 1550-A, pp. 297.

Wroth, C. P.: 1984, Interpretation of in situ soil test, 24th Rankine Lecture, Geotechnique 34,449-489.

Back to Top