Select Liquefaction Case Histories From The 2010-2011 Canterbury Earthquake Sequence
Select Liquefaction Case Histories From The 2010-2011 Canterbury Earthquake Sequence
Select Liquefaction Case Histories From The 2010-2011 Canterbury Earthquake Sequence
INTRODUCTION
The objective of this study is to present high-quality liquefaction case histories from the
20102011 Canterbury earthquake sequence and to use these case histories to evaluate three
commonly used, deterministic, cone penetration test (CPT) based simplified liquefaction
evaluation procedures. The 20102011 Canterbury earthquake sequence began with the
4 September 2010 M w 7.1 Darfield earthquake and included up to ten events that induced
liquefaction (Quigley et al. 2013). However, most notably, widespread liquefaction
was induced by the M w 7.1, 4 September 2010 Darfield and the M w 6.2, 22 February
2011 Christchurch earthquakes. The ground motions from these events were recorded across
Christchurch and its environs by a dense network of strong motion stations (e.g., Cousins and
McVerry 2010, Bradley and Cubrinovski 2011, Bradley 2012a). Also, due to the
severity and spatial extent of liquefaction resulting from the 2010 Darfield earthquake,
a)
Department of Civil and Environmental Engineering, Virginia Tech, 200 Patton Hall, Blacksburg, VA 24061
b)
Department of Civil and Natural Resources Engineering, University of Canterbury, Private Bag 4800,
Christchurch 8140, New Zealand
c)
Department of Civil, Architectural, and Environmental, University of Texas at Austin, 301 E. Dean Keeton St.,
Austin, TX 78712-1056
d)
Department of Civil and Environmental Engineering, University of Auckland, Private Bag 92019, Auckland
1142, New Zealand
131
Earthquake Spectra, Volume 30, No. 1, pages 131153, February 2014; 2014, Earthquake Engineering Research Institute
132 GREEN ET AL.
the New Zealand Earthquake Commission (EQC) funded an extensive subsurface character-
ization program for Christchurch, with over 10,000 CPT soundings performed to date.
The combination of well-documented liquefaction response during multiple events, den-
sely recorded ground motions for the events, and detailed subsurface characterization pro-
vides an unprecedented opportunity to add numerous quality case histories to the liquefaction
database. However, as discussed in Idriss and Boulanger (2012), the position of the cyclic
resistance ratio (CRR) curve, which is central to simplified liquefaction evaluation proce-
dures, is often controlled by relatively few case histories. Toward this end, the authors
selected 25 sites to analyze in detail, many of which had minor surficial liquefaction man-
ifestations resulting from the Darfield or Christchurch earthquake. The sites were evaluated
during both these events, resulting in 50 high-quality case histories. The sites selected for
detailed evaluation were located relatively close to strong ground motion stations and were
characterized by both CPT soundings and surface wave testing.
The authors use the case histories to evaluate existing deterministic, CPT-based simpli-
fied liquefaction evaluation procedures, namely the procedures proposed by Robertson and
Wride (1998; hereinafter, R&W98), Moss et al. (2006; hereinafter, MEA06), and Idriss and
Boulanger (2008; hereinafter, I&B08). These procedures were assessed by comparing pre-
dicted and observed liquefaction responses at the sites, where the liquefaction responses were
determined using one or more of these three methods: (1) post-earthquake site visits by the
authors; (2) examination of high resolution aerial and satellite imagery; and/or (3) interviews
with residents who lived near the case history sites. An error index is proposed to quantify the
predictive capabilities of the three CPT-based procedures for the 50 selected case histories.
Background information on geology of the Canterbury Plains and on the Darfield and
Christchurch earthquakes is presented first, with emphasis on information relevant to lique-
faction. Next, the ground motions recorded during the Darfield and Christchurch earthquakes
are discussed in relation to how the peak ground accelerations (PGA) at the case history sites
were estimated, followed by a discussion of how the liquefaction case histories were inter-
preted with a subsequent detailed discussion of five of the 25 case history sites. The results
from the case histories are then used to evaluate the R&W98, MEA06, and I&B08 proce-
dures. An extensive Electronic Supplement is provided that details all the case histories and
the authors interpretations. This will allow subsequent researchers to perform their own
interpretations, if they desire to do so.
Most of Christchurch was once low-lying floodplains and swamps behind a series
of barrier dunes (composed of fine-grained beach/dune sand), estuaries, and lagoons
(underlain by fine-grained deposits) of Pegasus Bay. The Waimakariri River regularly
flooded Christchurch before levee construction and river realignment, shortly after the
city was established in 1850. The original city center was constructed on slightly higher
ground compared to areas to the north and east. Of particular relevance to liquefaction
susceptibility in Christchurch and its environs are the locations of abandoned paleo-channels
of the Waimakariri, Heathcote, and Avon Rivers, and former swamps. These areas are under-
lain by, and filled with, young loose sandy sediments, with shallow groundwater levels (from
1 m to 5 m below ground surface), which are highly susceptible to liquefaction (e.g.,
Wotherspoon et al. 2012).
Samples of liquefaction ejecta were collected from several sites around Christchurch and
Kaiapoi. The grain characteristics of these samples were analyzed using a scanning electron
microscope (SEM), energy-dispersive X-ray spectrometry (EDS), X-ray diffraction (XRD),
and a laser particle size analyzer. Although the characteristics of the collected samples varied,
most can be described as silty fine sand having subrounded particle shapes. The EDS and
XRD analyses showed that the ejecta is predominantly quartz and feldspar, which is con-
sistent with the hypothesis that the material was derived from Torlesse Greywacke sandstone
in the Southern Alps. The grain size distributions obtained from the laser particle size ana-
lyses indicate that the samples classify as SP, SM, and SP-SM, per ASTM D-2487
(ASTM, 2011).
GROUND MOTIONS
The 20102011 Canterbury earthquake sequence started at 4:35 am on 4 September 2010
NZ Standard Time (16:35 3 September 2010 UTC), when a previously unmapped fault west
of Christchurch ruptured, producing the M w 7.1 Darfield earthquake (Bradley et al. 2014).
Although the earthquake caused major damage to the built environment and induced wide-
spread liquefaction, there were no fatalities or major injuries. The Canterbury earthquake
sequence included twelve other events having M w 5.0 with epicentral locations within
20 km of Christchurch (GeoNet, 2012), and up to ten of these larger events are known
to have induced liquefaction (Quigley et al. 2013). However, the M w 6.2, 22 February
2011 Christchurch earthquake was the most damaging event, due to the close proximity
of its rupture plane to Christchurch, resulting in 185 fatalities and causing widespread lique-
faction (e.g., Cubrinovski et al. 2011, Cubrinovski et al. 2012, Green et al. 2011, Maurer et al.
2013, Orense et al. 2011, Robinson et al. 2013).
The motions from both the Darfield and Christchurch earthquakes were recorded by a
dense network of strong ground motion stations (e.g., Cousins and McVerry 2010, Bradley
and Cubrinovski 2011, Bradley 2012a and 2012b, Bradley et al. 2014). To evaluate the factor
of safety against liquefaction per the simplified procedures used herein, the amplitude of
cyclic loading is proportional to the PGA at the ground surface and the duration is related
to the earthquake magnitude. Using the accelerograms recorded at the strong motion stations
(GeoNet 2012), the conditional PGA distributions at the case history sites were computed
using the procedure briefly outlined in Part A of the Electronic Supplement and discussed in
detail by Bradley (2013a). This approach is similar to that used by Green et al. (2011) and
134 GREEN ET AL.
Figure 1. Contours of computed conditional PGAs (g): (a) Darfield earthquake and
(b) Christchurch earthquake.
Maurer et al. (2014). Contour maps of the computed conditional PGAs for the Darfield and
Christchurch earthquakes are shown in Figure 1.
Figure 2. (a) Areas that liquefied during the Mw 7.1 Darfield earthquake (bound by dark lines)
and areas that liquefied during the Mw 6.2 Christchurch earthquake (white shaded areas). (b) Loca-
tions of case history sites (numbered, dark dots) and strong motion seismograph stations (labeled,
white dots).
immediately after the earthquakes. Finally, high resolution aerial photographs and satellite
imagery taken within a few days after the earthquakes were used to further determine the
liquefaction response at the selected sites.
was analyzed using the frequency domain beamformer method (Zywicki 1999). Once the
surface wave dispersion trends from each method were obtained, a composite dispersion
curve was generated by combining the dispersion data from SASW and MASW. The
mean and standard deviation of the dispersion data was calculated based on Rayleigh
wave phase velocity (V R ) and wavelength according to Cox and Wood (2011). The
shear wave velocity profile was then determined by fitting a 3D theoretical solution (i.e.,
effective mode inversion) to the mean experimental dispersion curve using the software
WinSASW. The water tables used in the forward modeling were determined using either
piezo-CPT sounding data (i.e., CPTu data) and/or P-wave refraction results. The shear
wave velocity profiles obtained from forward modeling of each site were limited to the max-
imum experimental wavelength divided by two (i.e., max 2). For additional information
regarding the surface wave testing conducted as part of this study refer to Wood et al. (2011).
As discussed below, the shear wave velocity profiles were used to help interpret the soil
profile at each site and to compute the average shear wave velocity over the top 12 m (V S12 ),
which was used in conjunction with the MEA06 simplified liquefaction evaluation proce-
dure. The shear wave velocity profiles were also used to evaluate liquefaction susceptibility
using various simplified V S liquefaction triggering relationships. However, these evaluations
are still ongoing and are not presented in this paper.
pressures but not necessarily liquefying, or from a relatively dense, thick, shallow
layer marginally liquefying. The thickness of the loose stratum would have to
increase the deeper the stratum is below the surface, while the density of the
medium-dense stratum would have to decrease the deeper the stratum is.
4. The critical layer associated with lateral spreading can be relatively thin, with its
thickness and density increasing and decreasing, respectively, the larger the amount
of the surface ejecta and the wider the lateral spread cracks, respectively. However,
the critical layer should be selected by also considering the slope of the ground
surface, the height of the free face, and the lateral continuity of the liquefiable stra-
tum relative to the lateral spread area.
As described above, the liquefaction responses for the case histories are categorized as
No Liquefaction, Minor Liquefaction, Moderate Liquefaction, Severe Liquefaction,
and Lateral Spreading. As the naming scheme implies, No Liquefaction includes all the
cases where no liquefaction surface manifestations were observed, Minor Liquefaction
includes all the cases where minor surficial liquefaction manifestations were observed,
Moderate Liquefaction and Severe Liquefaction include all cases where the observed
liquefaction surface manifestations were more severe than Minor Liquefaction, and
Lateral Spreading includes cases where liquefaction was manifested at the ground surface
in the form of lateral spreading cracks, etc. Table ESB-1 in Part B of the Electronic Supple-
ment provides quantitative metrics for the severity categorization used herein. However,
because the severity of surficial liquefaction manifestations is a continuum ranging from
none to very severe, any sort of discrete categorization of Minor, Moderate, and Severe
is inherently subjective, regardless of the best efforts to quantify liquefaction severity. To
help reduce ambiguity in how the authors classified the case histories presented herein,
Part B of the Electronic Supplement also gives examples of high resolution aerial images
of the different severity manifestation categories.
Implementing the above rules to select the critical layers at each case history site requires
considerable judgment, and several of the 60 sites that were initially selected for detailed
analysis were removed from further consideration because of the significant ambiguity in
constraining the depth/thickness of the critical layers. Additionally, in several of the remain-
ing sites, both preferred and alternative critical layers were selected. Finally, it is debatable
whether or not lateral spreading case histories should be included in the liquefaction trigger-
ing database. These case histories are often more difficult to interpret than level ground lique-
faction cases, and require extensive in situ test data to properly interpret (more extensive than
is often performed). Accordingly, it is the opinion of the authors that if lateral spread case
histories are included in the liquefaction triggering database, they should generally be given
less weight when using them to develop or to evaluate CRR curves. Of the 25 sites included
in this study, only two were lateral spread cases, one of which is presented in detail below,
and both are briefly discussed in relation to the observed surface manifestations versus their
predicted response by R&W98, MEA06, and I&B08.
Before applying the above general rules for selecting the critical layers, the soil type and
soil density needed to be estimated from the available in situ test data. Toward this end, CPT
logs were used to compute the soil behavior type index, I c , for each site as a function of
depth. Table 1 lists values of I c and the corresponding inferred soil type (Robertson and
138 GREEN ET AL.
Wride, 1998). Following Youd et al. (2001), in the absence of site specific sampling and
testing, soils having an I c > 2.4 were considered to be non-liquefiable. The relative densities
(Dr ) were then estimated for soils having I c < 2.4 using the relationship derived from
Robertson and Cabal (2012):
1 qc pa
Dr % (1)
c 0 v 0.5
EQ-TARGET;temp:intralink-;e001;41;452
pa
Once the I c and Dr profiles were computed for all the sites, the critical layers for each site
were selected using the general rules listed above, with the estimated Dr values used in a
relative sense for a given profile as opposed to being used in an absolute sense. That is for a
given profile, the computed Dr values are used to determine density of one stratum relative to
another as opposed to determining the relative density of the strata in absolute terms.
As may be noted, the above rules for selecting critical layers do not explicitly take into
account the variation of the induced cyclic stress as a function of depth, although this is
somewhat taken into account through judgment in selecting the critical layer. However,
to assess the influence of the depth dependency of the induced seismic demand, the factor
of safety against liquefaction (FSliq ) was computed using the various simplified liquefaction
evaluation procedures to determine if there were any credible alternative critical layers for the
profiles. If there were, judgment was be used in revising the selected critical layers. However,
this is a secondary step, and the authors view it as being different from using a given sim-
plified liquefaction evaluation procedure to select the critical layers and then turning around
and using these case histories to validate the predictive capabilities of the same simplified
liquefaction evaluation procedure (or to invalidate the predictive capabilities of different sim-
plified liquefaction evaluation procedures).
SELECT LIQUEFACTION CASE HISTORIES 139
Site 1: CPT-SHY-09
Site 1: CPT-SHY-09 is located on Hercules Street, just east of the intersection of Hercules
and Hope Streets in the northeastern suburb of Richmond. The site is approximately 0.27 km
from the SHLC strong motion seismograph station, and the estimated geometric means of the
horizontal PGAs at the site during the Darfield and Christchurch earthquakes are 0.187 g and
0.347 g, respectively. There were no observed surface manifestations of liquefaction follow-
ing the Darfield earthquake, but there was evidence of Moderate Liquefaction at the site
following the Christchurch earthquake.
The measured CPT qc , friction ratio (Rf f s qc ), I c , and Dr as functions of depth for
SHY-09 are plotted in Figure 3. Superimposed on these plots is the selected critical layer for
both the Darfield and Christchurch earthquakes. The critical layer is relatively shallow and is
140 GREEN ET AL.
Figure 3. Measured CPT tip resistance (qc ), friction ratio (Rf ), soil behavior type index (I c ), and
estimated relative densities (Dr ) as functions of depth for Site 1: CPT-SHY-09. Superimposed on
these plots is the selected critical layer for both the Darfield and Christchurch earthquakes.
approximately 2 m thick. Using the I c soil type correlation presented in Table 1 and
Equation 1, the inferred soil type for the layer is loose clean sand to silty sand. Following
the general rules outlined above for selecting critical layers, the authors selected the layer
shown in Figure 3 because its depth-thickness-density combination is believed to be con-
sistent with the surficial liquefaction manifestations observed during the both the Darfield
and Christchurch earthquakes. That is, in the authors judgment, the depth-thickness-density
combination of the critical layer is such that surface manifestations would have resulted if this
layer liquefied during the Darfield earthquake; similarly, if this layer liquefied during the
Christchurch earthquake, its depth-thickness-density combination is sufficient to have caused
moderate surface liquefaction manifestations observed at the site.
Site 4: CPT-CBD-21
Site 4: CPT-CBD-21 is located near the northwest corner of the intersection of Durham
and Salisbury Streets in Christchurchs CBD. The site is approximately 0.3 km from the
REHS strong motion seismograph station, and the estimated geometric means of the hori-
zontal PGAs at the site during the Darfield and Christchurch earthquakes are 0.219 g and
0.460 g, respectively. There were no observed surface manifestations of liquefaction follow-
ing the Darfield earthquake, but there were minor surficial liquefaction manifestations fol-
lowing the Christchurch earthquake. Because of the limited consequences of minor
liquefaction to the built and natural environment, other investigators in previous studies
sometimes have classified sites that had minor surficial liquefaction manifestations as
no liquefaction. However, due to the potential for minor liquefaction case histories to
help constrain the position of the CRR curve, the authors prefer to categorize these sites
separately with the designation of Minor Liquefaction. This is because excess pore
SELECT LIQUEFACTION CASE HISTORIES 141
water pressures were certainly elevated at these sites as a result of the earthquake shaking, but
the thickness-depth-severity combination of the critical layer was such that the surface man-
ifestations were minor. To illustrate Minor Liquefaction as defined in this study, a high reso-
lution aerial photograph of CBD-21 taken two days after the Christchurch earthquake is
shown in Figure 4. At this site, a small amount of liquefaction ejecta surfaced along the
edge of road (a few cm thick at most). There was no visible distress to the adjacent structure,
to the pavement, or buried utilities (if present) due to liquefaction at this site (i.e., after the
limited amount of ejecta was cleaned up, there was no evidence of liquefaction at the site;
note that authors returned to this site periodically from the time of the earthquakes to the time
of the writing of this paper).
The measured CPT tip resistance (qc ), friction ratio (Rf ), soil behavior type index (I c ), and
estimated relative densities (Dr ) as functions of depth for CBD-21 are plotted in Figure 5.
Superimposed on these plots is the selected critical layer for both the Darfield and
Christchurch earthquakes. The critical layer is 4.5 m deep and is approximately 2 m
thick. Using the I c soil type correlation presented in Table 1 and Equation 1, the inferred
soil type for the layer is dense clean sand to silty sand. Per the general rules outlined above
for selecting critical layers, the authors selected the layer shown in Figure 5 because its
depth-thickness-density combination is believed to be consistent with the liquefaction surface
manifestation (or lack thereof) observed during the both the Darfield and Christchurch
earthquakes.
Figure 4. High resolution aerial photograph of Site 4: CPT-CBD-21 taken two days after the
Christchurch earthquake illustrating minor liquefaction as defined in this study.
142 GREEN ET AL.
Figure 5. Measured CPT tip resistance (qc ), friction ratio (Rf ), soil behavior type index (I c ),
and estimated relative densities (Dr ) as functions of depth for Site 4: CPT-CBD-21.
Superimposed on these plots is the selected critical layer for both the Darfield and Christchurch
earthquakes.
Site 5: CPT-FND-01
Site 5: CPT-FND-01 is located northwest of the intersection of Fendalton Rd, Deans Ave,
and Harper Ave in the suburb of Fendalton and is the western most site analyzed in this study.
The site is approximately 0.9 km from the CBGS strong motion seismograph station, and the
estimated geometric means of the horizontal PGAs at the site during the Darfield and
Christchurch earthquakes are 0.199 g and 0.382 g, respectively. The site experienced Lateral
Spreading during the Darfield earthquake and Severe Lateral Spreading during the
Christchurch earthquake.
The measured CPT tip resistance (qc ), friction ratio (Rf ), soil behavior type index (I c ),
and estimated relative densities (Dr ) as functions of depth for CBD-21 are plotted in
Figure 6. As inferred from these plots, the entire profile consists of relatively thin, alter-
nating layers of coarse- and fine-grained soils, where the former are assumed to be liquefi-
able and the latter are assumed not to be. The observed surface manifestations of
liquefaction likely resulted from several of the thin, coarse-grained layers liquefying, rather
than just from a single layer liquefying. However, in lieu of averaging the properties of
multiple layers having various inferred soil types, the authors opted to select one of the
thin coarse-grained layers as being representative of all the thin, coarse-grained layers
that likely liquefied during both the Darfield and Christchurch earthquakes. This layer
is superimposed on the plots in Figure 6. Because the selected representative critical
layer is relatively thin (0.3 m thick), thin layer corrections were used in computing
the normalized CPT tip resistances in accordance with the respective procedures outlined
in R&W98, MEA06, and I&B08.
SELECT LIQUEFACTION CASE HISTORIES 143
Figure 6. Measured CPT tip resistance (qc ), friction ratio (Rf ), soil behavior type index (I c ),
and estimated relative densities (Dr ) as functions of depth for Site 5: CPT-FND-01. Superimposed
on these plots is the selected representative critical layer for both the Darfield and Christchurch
earthquakes.
Figure 7. Measured CPT tip resistance (qc ), friction ratio (Rf ), soil behavior type index (I c ), and
estimated relative densities (Dr ) as functions of depth for Site 11: CPT-KAN-26. Superimposed
on these plots are the selected critical layers for the Darfield and Christchurch earthquakes.
development of iron oxide byproducts that result from exposure to oxygen through a fluc-
tuating groundwater table near the ground surface or from being a fill material (i.e., a shallow
layer). On the contrary, the moderate liquefaction surface manifestations that formed during
the Darfield earthquake included both brown and blue-gray sand ejecta, which for the latter
ejecta indicate the sand was in a reducing atmosphere, sealed from oxygen under the ground-
water table, which is only possible for a deeper deposit (i.e., a deeper layer).
Figure 8. Measured CPT tip resistance (qc ), friction ratio (Rf ), soil behavior type index (I c ), and
estimated relative densities (Dr ) as functions of depth for Site 19: CPT-NBT-03. Superimposed
on these plots is the selected critical layer for both the Darfield and Christchurch earthquakes.
for the layer is clean sand to silty sand with thin strata of medium-dense silty sand to sandy
silt. Although the properties of the selected critical layer (e.g., inferred relative density and
soil type) vary as a function of depth, the authors believe that the depth-thickness-density
combination of the selected layer is consistent with the liquefaction surface manifestation
observed during both earthquakes.
Figure 9. Case history data plotted together with CRRM7.5 curves: (a) R&W98; (b) MEA06;
(c) I&B08. Note similar plots are contained in Part D of the Electronic Supplement, wherein
each of the case histories is identified by number in the plots which corresponds to those tabulated
in Table ESC-1.
the critical layers were estimated using two different I c -FC correlations, a generic correlation
proposed by Robertson and Wride (1998) and a Christchurch-soil-specific correlation devel-
oped by Robinson et al. (2013). It should be noted that Idriss and Boulanger (2008) do not
recommend the use of generic I c -FC correlations to estimate FC, but rather recommend the
development and use of project-specific I c -FC correlations such as that developed by
Robinson et al. (2013). The two correlations used in this study are shown in Figure 10. How-
ever, so as not to misclassify very loose clean sands as denser sands containing fines, an
SELECT LIQUEFACTION CASE HISTORIES 147
additional criterion was applied in estimating FC: FC 5% if 1.64 < I c < 2.36 and
F < 0.5%, where F is the normalized friction ratio and is given by (Robertson and
Wride, 1998):
fs
F and v total vertical stress (2)
qc v
EQ-TARGET;temp:intralink-;e002;62;350
The authors computed normalized CPT tip resistances per I&B08 for the case histories
using the FC estimated using both I c -FC correlations. The CSRM7.5 case history data for
I&B08 are plotted in Figure 9c as a function of the normalized CPT tip resistances computed
using both sets of estimated FC values. In this figure, the triangular and circular symbols
correspond to the normalized CPT tip resistances computed using FC values estimated
using the Christchurch-soil-specific I c -FC correlation. The end of the tails extending
from the triangular and circular symbols correspond to the normalized CPT tip resistances
computed using FC values estimated using the generic I c -FC correlation.
As may be observed in Figure 9, several of the data points have double symbols (i.e., a
circle within a circle or triangle within a triangle). These are cases where credible alternative
critical layers were identified (i.e., ambiguous cases). However, these points are plotted for
the authors preferred critical layer, not the alternative critical layer, but similar plots showing
the data plotted for the alternative critical layers are given in Part D of the Electronic
Supplement.
DISCUSSION
As may be observed from Figure 9, R&W98, MEA06, and I&B08 correctly predicted the
majority of the observed liquefaction responses for the case histories, but none of the
148 GREEN ET AL.
procedures correctly predicted all cases. In assessing which of the three procedures gives the
most accurate predictions for the data analyzed, it must be remembered that the liquefaction
responses for all of the case histories presented herein are based on surface manifestations, as
is the case for almost all previous liquefaction case history studies whether explicitly stated or
not. As a result, a No Liquefaction case may mean that no stratum in the profile liquefied,
or, that the severity-depth-thickness combination of a liquefied stratum was such that it did
not manifest on the ground surface. On the contrary, a Liquefaction case (i.e., cases where
surface manifestations are more severe than Minor) can only mean that a stratum in the
profile liquefied (otherwise no liquefaction surface manifestations would have occurred).
Finally, a Minor Liquefaction case could result from a few different scenarios. For exam-
ple, minor surface liquefaction manifestations could result from a relatively loose, thin, deep
layer severely liquefying, from a relatively loose to medium dense, thick, shallow layer hav-
ing highly elevated excess pore pressures but not necessarily liquefying, or from a relatively
dense, thick, shallow layer marginally liquefying. The latter two scenarios should yield data
that will constrain the position of the CRRM7.5 curve, while data from the first scenario may
plot well above/to the left of the CRRM7.5 curve.
To quantitatively assess which of the three CPT-based liquefaction evaluation procedures
yields the most accurate predictions for the data analyzed, the authors propose the follow-
ing error index (E I ):
Xn
EI R
i1 i
(3) EQ-TARGET;temp:intralink-;e003;41;401
0 K
8
< CSRM7.5 CRR
for No Liq cases where CSRM7.5
M7.5 < CRRM7.5
Ri K K
: for No Liq cases where CSRM7.5
CRRM7.5
EQ-TARGET;temp:intralink-;sec6;41;286
0 K
The proposed error index will equal zero if all the predictions correctly match the field
observations, but will increase in value as the number and magnitude of the mispredic-
tions increases. On an individual case basis, R equals zero for a correct prediction of a
Liquefaction/Minor Liquefaction or No Liquefaction case, but is equal to the vertical dis-
tance between the CRRM7.5 curve and the plotted point for a mispredicted Liquefaction/
Minor Liquefaction or No Liquefaction case.
The computed E I values for each CPT-based liquefaction evaluation procedure are pre-
sented in Table 2. As may be observed, the I&B08 gives the most accurate predictions of the
case histories analyzed in the this paper, with the computed E I values for I&B08 being rela-
tively independent of whether the FC of the critical layers were estimated using the generic
I c -FC correlation proposed by Robertson and Wride (1998) or the Christchurch-soil-specific
correlation developed by Robinson et al. (2013). The R&W98 gives the next most accurate
SELECT LIQUEFACTION CASE HISTORIES 149
Table 2. Error indices for the three CPT-based liquefaction evaluation procedures
EI
I&B08 I&B08
MEA06 MEA06 (Generic I c -FC (Chch I c -FC
R&W98 (V S12 dep r d ) (V S12 indep r d ) correlation) correlation)
Darfield EQ 0.264 0.379 0.411 0.120 0.120
Christchurch EQ 0.275 0.225 0.290 0.047 0.054
Total 0.539 0.604 0.701 0.167 0.174
predictions of the analyzed cases from the Darfield earthquake, while the MEA06 with
the V S12 -dependent r d yielded more accurate predictions of the analyzed cases from the
Christchurch earthquake. The least accurate predictions for the cases analyzed from both
earthquakes result from the MEA06 procedure with the V S12 -independent r d relation. How-
ever, the authors emphasize that the noted trends in the E I values are for the case histories
analyzed in this study; these trends may not hold for other scenarios that are not represented
by the case histories analyzed (e.g., deep/dense liquefiable layers).
The authors note that a misprediction of a Liquefaction case as a No Liquefaction case
can have greater consequences than the misprediction of a No Liquefaction case as a Lique-
faction case. Additionally, the authors note that minor surficial liquefaction manifestations
can occur in very loose soil when excess pore pressures are elevated, even if liquefaction is
not triggered (e.g., r u slightly less than 1.0, where ru is the excess pore pressure ratio). To
examine the significance of a misprediction and to further distinguish Minor Liquefaction
cases from Liquefaction cases, the authors present modified versions of the equations for Ri
in Part E of the Electronic Supplement. Furthermore, to determine the influence of the alter-
native versus the preferred critical layers on the computed E I values, the authors computed
the E I values for the following scenarios: (1) authors preferred critical layers for all cases
(Table 2); (2) authors preferred critical layers for 18 cases and alternative critical layers for
the 7 ambiguous cases; and (3) authors preferred critical layers for 18 cases, with the 7
ambiguous cases removed. As may be observed from the resulting E I values tabulated in
Part E of the Electronic Supplement, the ranking of the three procedures remains relatively
unaffected for the cases histories analyzed in this study.
Finally, as mentioned previously, it is debatable whether or not lateral spreading case
histories should be included in the liquefaction triggering database due to the difficulty
in their interpretation and the often lack of required field test data to properly interpret
them. However, all the simplified liquefaction evaluation procedures correctly predicted
the observed liquefaction response for the case histories analyzed herein where liquefaction
manifested in the form of lateral spreading (i.e., Site 5: FND-01 and Site 23: Z2-6). Whether
the correct prediction of these case histories is the result of judicious critical layer selection by
the authors or coincidence is unknown at this time. However, further investigations are being
conducted to evaluate the prediction of liquefaction triggering at sites subject to lateral
spreading (e.g., Robinson et al. 2013).
150 GREEN ET AL.
ACKNOWLEDGMENTS
The authors gratefully acknowledge the Canterbury Geotechnical Database and the
New Zealand GeoNet project and its sponsors EQC, GNS Science, and LINZ, for providing
some of the data used in this study. Also, the authors are grateful to Dr. Josh Zupan and
Dr. Jonathan Bray who oversaw the performance of some of the CPT soundings presented
in this study (Bray et al. 2014), to Mr. Kevin Foster for performing SEM and EDS analyses
on liquefaction ejecta samples, and to Dr. Matthew Hughes for computing the grade at the
case history sites. Additionally, the authors thank Drs. Robb Moss and Ross Boulanger for
answering questions and providing information about the MEA06 and I&B08 liquefaction
evaluation procedures, respectively. The primary support for the US authors was provided by
U.S. National Science Foundation (NSF) grants CMMI-1030564, CMMI-1137977, and
CMMI-1306261. Also, L. Wotherspoons position at the University of Auckland is funded
by the Earthquake Commission (EQC). However, any opinions, findings, and conclusions or
recommendations expressed in this material are those of the authors and do not necessarily
reflect the views of the National Science Foundation or the other funding agencies.
REFERENCES
Abrahamson, N. A., and Youngs, R. R., 1992. A stable algorithm for regression analyses using
the random effects model, Bulletin of the Seismological Society of America 82, 505510.
ASTM International (ASTM), 2011. D2487-11 Standard Practice for Classification of Soils for
Engineering Purposes (Unified Soil Classification System), West Conshohocken, PA.
SELECT LIQUEFACTION CASE HISTORIES 151
Forsyth, P., Barrell, D., and Jongens, R., 2008. Geology of the Christchurch Area, Institute of
Geological and Nuclear Sciences GNS Science 1:250,000 Geological Map 16, 67 pp.
GeoNet, 2012. Quake Search, available at http://magma.geonet.org.nz/resources/quakesearch/
(last accessed 26 Nov 2012).
Goda, K., and Hong, H. P., 2008. Spatial correlation of peak ground motions and response spec-
tra, Bulletin of the Seismological Society of America 98, 354465.
Green, R. A., Allen, A., Wotherspoon, L., Cubrinovski, M., Bradley, B., Bradshaw, A., Cox, B.,
and Algie, T., 2011. Performance of levees (stopbanks) during the 4 September Mw 7.1
Darfield and 22 February 2011 Mw 6.2 Christchurch, New Zealand, earthquakes, Seismologi-
cal Research Letters 82, 939949.
Green, R. A., Obermeier, S. F., and Olson, S. M., 2005. Engineering geologic and geotechnical
analysis of paleoseismic shaking using liquefaction effects: Field examples, Engineering
Geology 76, 263293.
Green, R. A., Wood, C., Cox, B., Cubrinovski, M., Wotherspoon, L., Bradley, B., Algie, T.,
Allen, J., Bradshaw, A., and Rix, G., 2011. Use of DCP and SASW tests to evaluate liquefac-
tion potential: Predictions vs. observations during the recent New Zealand earthquakes,
Seismological Research Letters 82, 927938.
Idriss, I. M., and Boulanger, R.W., 2008. Soil liquefaction during earthquakes. Monograph
MNO-12, Earthquake Engineering Research Institute, Oakland, CA, 261 pp.
Idriss, I. M., and Boulanger, R. W., 2012. Examination of SPT-based liquefaction triggering
correlations, Earthquake Spectra 28, 9891018.
Johnson, R. A., and Wichern, D. W., 2007. Applied Multivariate Statistical Analysis, Prentice
Hall, Upper Saddle River, NJ
Maurer, B. W., Green, R. A., Cubrinovski, M., and Bradley, B. A., 2014. Evaluation of
Liquefaction Potential Index (LPI) for Assessing Liquefaction Hazard: Case Study
Christchurch, New Zealand, Journal of Geotechnical and Geoenvironmental Engineering,
ASCE, in press.
Moss, R. E. S, Seed, R. B., Kayen, R. E., Stewart, J. P., Der Kiureghian, A., and Cetin, K. O.,
2006. CPT-based probabilistic and deterministic assessment of in situ seismic soil liquefac-
tion potential, Journal of Geotechnical and Geoenvironmental Engineering, ASCE, 132,
10321051.
Olson, S. M., Green, R. A., and Obermeier, S. F., 2005. Engineering geologic and geotechnical
analysis of paleoseismic shaking using liquefaction effects: A major updating, Engineering
Geology 76, 235261.
Orense, R. P., Kiyota, T., Yamada, S., Cubrinovski, M., Hosono, Y., Okamura, M., and Yasuda,
S., 2011. Comparison of liquefaction features observed during the 2010 and 2011 Canterbury
earthquakes, Seismological Research Letters 82, 905918.
Pinheiro, J., Bates, D. M., DebRoy, S.Sarkar, D., and the R Core Team, 2008. nlme: Linear
and Nonlinear Mixed Effects Models, R package version 3.1, 89 pp.
Quigley, M. C., Bastin, S., and Bradley, B. A., 2013. Recurrent liquefaction in Christchurch,
New Zealand, during the Canterbury earthquake sequence, Geology 41, 419422.
Robertson, P. K., 2009. Performance based earthquake design using the CPT, in Performance
Based Design in Earthquake Geotechnical Engineering: From Case History to Practice
(T. Kokusho, Y. Tsukamoto, and M. Yoshimine, eds.), Taylor & Francis Group, London,
320.
SELECT LIQUEFACTION CASE HISTORIES 153
Robertson, P. K., and Cabal, K. L., 2010. Estimating soil unit weight from CPT, 2nd Interna-
tional Symposium on Cone Penetration Testing, Paper #2-40, May 2010, Huntington
Beach, CA.
Robertson, P. K., and Cabal, K. L., 2012. Guide to Cone Penetration Testing for Geotechnical
Engineering, 5th edition, Gregg Drilling & Testing, 131 pp.
Robertson, P. K., and Wride, C. E., 1998. Evaluating cyclic liquefaction potential using the cone
penetration test, Canadian Geotechnical Journal 35, 442459.
Robinson, K., Cubrinovski, M., and Bradley, B. A., 2013. Sensitivity of predicted liquefaction-
induced lateral displacements from the 2010 Darfield and 2011 Christchurch earthquakes, in
Proc. 2013 Conference of the New Zealand Society for Earthquake Engineering (NZSEE
2013), 2628 April, Wellington, New Zealand.
Whitman, R. V., 1971. Resistance of soil to liquefaction and settlement, Soils and Foundations
11, 5968.
Wood, C. M., Cox, B. R., Wotherspoon, L. M., and Green, R. A., 2011. Dynamic site character-
ization of Christchurch strong motion stations, Bulletin of the New Zealand Society for Earth-
quake Engineering 44, 195204.
Wotherspoon, L. M., Pender, M. J., and Orense, R. P., 2012. Relationship between observed
liquefaction at Kaiapoi following the 2010 Darfield earthquake and former channels of the
Waimakariri River, Engineering Geology 125, 4555.
Youd, T. L., Idriss, I. M., Andrus, R. D., Arango, I., Castro, G., Christian, J. T., Dobry, R.,
Finn, W. D. L., Harder, L. F., Hynes, M. E., Ishihara, K., Koester, J. P., Liao, S. S. C.,
Marcuson, III, W. F., Martin, G. R., Mitchell, J. K., Moriwaki, Y., Power, M. S., Robertson,
P. K., Seed, R. B., and Stokoe, II, K. H., 2001. Liquefaction resistance of soils: Summary
report from the 1996 NCEER and 1998 NCEER/NSF workshops on evaluation of liquefac-
tion resistance of Soils, Journal of Geotechnical and Geoenvironmental Engineering 127,
297313.
Zywicki, D.J., 1999. Advanced signal processing methods applied to engineering analysis of
seismic surface waves, Ph.D. Thesis, Georgia Institute of Technology, Atlanta, GA, 357 pp.
(Received 5 March 2013; accepted 27 August 2013)