You are on page 1of 12

EARTHQUAKE-INDUCED DISPLACEMENTS OF SOLID-WASTE LANDFILLS

By Jonathan D. Bray, l Member, ASCE, and Ellen M. Rathje2

ABSTRACT: The response of municipal solid-waste landfills at high levels of earthquake shaking is explored.
Results from fully nonlinear dynamic analyses indicate that the seismic response of a landfill, particularly the
seismic loading for the cover, can vary significantly due to reasonable variations of waste properties, fill heights,
site conditions, and design rock motions. However, for the base sliding case, the maximum seismic loading
depends primarily on the amplitude (peak acceleration) and frequency content (mean period) of the design rock
motion and the dynamic response characteristics of the landfill represented by its initial fundamental period.
Downloaded from ascelibrary.org by ECOLE POLYTECH/BIBLIOTHEQUE on 02/23/22. Copyright ASCE. For personal use only; all rights reserved.

The calculated permanent base displacement is largely a function of the ky/krnv. ratio, but it is also influenced
significantly by the intensity, frequency content, and duration of the design rock motion and the dynamic response
characteristics of the landfill and its foundation. Charts with normalized parameters are presented, which aid in
the development of preliminary estimates of the expected seismic loading and displacement. These analytical
results are consistent with observations from available case records.

INTRODUCTION et al. (1995), Kavazanjian et al. (1996), and Augello et al.


(1998), which include back-analyses of the instrumented Op-
Fully nonlinear dynamic analyses of municipal solid-waste
erating Industries, Inc. (OIl) landfill and geophysical mea-
landfills (MSWLFs) are performed using reasonable variations
surements of the shear wave velocity of MSW, define the pro-
in waste fill properties, height, and foundation conditions to
fession's current state of knowledge regarding the dynamic
investigate the seismic response of MSWLFs at relatively high
properties of MSW. Lastly, shaking table tests [e.g., Yegian et
levels of maximum horizontal ground accelerations (MHAs)
al. (1995)] have provided useful measurements of the dynamic
(Le., MHAs from 0.2g to 0.8g). Previous seismic response
interface shear response of geosynthetics.
studies of MSWLFs [e.g., Bray et al. (1995)] have largely been
Well-documented case records have also provided insight.
restricted to MHAs below OAg, although Environmental Pro-
An evaluation of important case records, in particular, the seis-
tection Agency (EPA) regulations specify prescriptive MHAs
mic performance of MSWLFs during the 1994 Northridge
at the 10% probability of exceedance in the 250-years level in
earthquake, is presented by Augello et al. (1995b). The ob-
excess of 0.8g in some areas of the United States. Results from
served performance of solid-waste landfills during recent
the nonlinear analyses are processed to evaluate the relative
earthquakes has been encouraging in that no global instability
importance of factors that influence seismically induced per-
has occurred. However, significant damage in the form of geo-
manent deformations within the base liners and covers of
membrane tears, cover cracking, broken gas header lines, and
solid-waste landfills. The accuracy of the commonly used de-
loss of power to gas extraction systems was experienced at
coupled approximation for calculating earthquake-induced
several landfills during the 1994 Northridge earthquake. These
displacements, which is also used in this study, is reassessed.
case records are providing important lessons, but due to the
The proposed analytical results are compared to field obser-
uncertainties involved, many questions remain unanswered.
vations, and recommendations based on insights gained from
This study examines the factors involved in assessing the po-
this study are presented.
tential seismic performance of waste fills at high levels of
BACKGROUND shaking through fully nonlinear response analyses with up-
dated waste fill property characterizations and deformable slid-
Concerns about the seismic performance of MSWLFs have ing block displacement calculations.
motivated a number of recent studies that investigate dynamic
properties of waste fill and seismic analysis procedures for BASIS OF NONLINEAR DYNAMIC ANALYSES
landfills. Anderson and Kavazanjian (1995) summarize the re-
suits of many of these studies. In short, regulatory concerns D-MOD Program
are identified and a state-of-the-practice survey is reported in The nonlinear dynamic analysis program D-MOD (Mata-
Seed and Bonaparte (1992). Commonly used seismic stability sovic 1993; Matasovic and Vucetic 1995) was used to study
procedures are reviewed, and insights from a one-dimensional the seismic response of landfills at high acceleration levels. D-
equivalent-linear parametric response study are presented in MOD is essentially DESRA-2 (Lee and Finn 1978), except
Bray et al. (1995) and Augello et al. (l995a). Singh and Sun that it incorporates a Modified Konder-Zelasko (MKZ) hyper-
(1995) and Ling and Leshchinsky (1997) study the seismic bolic stress-strain representation. With two additional param-
loading and stability, respectively, of landfill cover systems. eters, the MKZ model allows for more control over the shape
Kramer and Smith (1997) propose a modified Newmark model of the strain-dependent shear modulus reduction and damping
for calculating seismic displacements for waste fills. Kavazan- curves (Matasovic and Vucetic 1995). D-MOD uses a one-
jian et al. (1995) present MSW dynamic properties. More re- dimensional (ID) lumped mass model and solves the equations
cent studies, such as those by Idriss et al. (1995), Matasovic of motion in the time domain using the Newmark integration
'Assoc. Prof., Dept. of Civ. and Envir. Engrg., Univ. of California,
procedure, updating the soil stiffness at each time step. D-
Berkeley, CA 94720-1710. MOD and other DESRA-based nonlinear codes have been
2Asst. Prof., Dept. of Civ. Engrg., Univ. of Texas, Austin, TX 78712- shown to calculate seismic responses similar to the established
1076. equivalent-linear program SHAKE91 (Idriss and Sun 1992) at
Note. Discussion open until August I, 1998. To extend the closing low acceleration levels where nonlinearity is less pronounced
date one month, a written request must be filed with the ASCE Manager [Fig. 1; see also Finn (1988); Matasovic (1993); Kavazanjian
of Journals. The manuscript for this paper was submitted for review and and Matasovic (1995)], but D-MOD typically calculates a
possible publication on February 24, 1997. This paper is part of the lour-
nal of Geotechnical and Geoenvironmental Engineering, Vol. 124, No. lower response at high acceleration levels in agreement with
3, March, 1998. ©ASCE, ISSN 1090-0241/98/0003-0242-0253/$4.00 + recorded strong motion data at rock and soft soil sites (e.g.,
$.50 per page. Paper No. 14107. Borcherdt (1994); Seed et al. (1991)]. Back-analyses of several
242/ JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / MARCH 1998

J. Geotech. Geoenviron. Eng., 1998, 124(3): 242-253


30 m MSW, High Vs profile S%Damping

3.0
Rock Site
D-MOD
....
'::
20
'bil
'-'
SHAKE91 ./ \
2.5 Input, Pac. Dam Dnstrm .: \.
§ .
,.
,,
,,
'i ,., , ,,
40

-8
~ 2.0
,.,
.
,
,f "

.,,,
0
<: 1.5 . 60
'a
J:l
Downloaded from ascelibrary.org by ECOLE POLYTECH/BIBLIOTHEQUE on 02/23/22. Copyright ASCE. For personal use only; all rights reserved.

0
~
l:l.
1.0
en 80
0.5

100 L...L............L-1....L....L............J....I...L..JL....L..L...L............L-1....L....L............J....I...L..J.......
0.0 o 100 200 300 400 500 600
0.01 0.1 1 10 Vs (m/sec)
Period (sec)
FIG. 2. Shear Wave Velocity Profiles for MSW
FIG. 1. D-MOD and SHAKE91 Comparison at 'TWo Accelera-
tion Levels
reduction and damping curves for a clay with a plasticity index
well-documented case histories, such as Treasure Island during of 30. For these analyses, the maximum shear strain in the
the Lorna Prieta earthquake and the Wildlife Liquefaction site, waste fill typically ranged from 0.2% to 2%, and the maximum
have validated the use of D-MOD for ground response analysis shear stress never exceeded the dynamic strength of waste fill
(Matasovic 1993). [i.e., an effective friction angle of 35°, Augello et al. (1995b)].
The use of a 1D model to represent the seismic response of Three site profiles represented disparate landfill foundation
an earth/waste fill has been discussed in Vrymoed and Cal- conditions: rock, shallow sand, and deep soft clay. The rock
zascia (1978), Elton et al. (1991), and Bray et al. (1996), and site had 3 m of weathered rock (V. = 760 mls) overlying hard
it has been found that dynamic shear stresses near the base of bedrock. The sand site had 30 m of medium dense sand
a two-dimensional (2D) earth/waste fill can be approximated overlying bedrock. The shear wave velocity profile in the sand
reasonably well with 1D analysis. Capturing the cover re- varied with overburden pressure and ranged from 135-300 mI
sponse is more tenuous, but as the primary focus of the present s for low waste fill heights (10-20 m) to 350-400 mls for
study is to examine base sliding and key factors affecting seis- high waste fill heights (60-90 m). The deep, soft clay site
mic response (other than 2D geometry effects for cover sys- contained 21 m of soft clay, which was overlain by 3 m of
tems), the use of 1D analytical procedures is judged to be stiff clay and underlain by 67 m of stiff clay over bedrock.
appropriate. The shear wave velocity of the soft clay was varied with over-
burden pressure with V. = 100-200 mls for low waste fill
Cases Analyzed heights and V. = 150-230 mls for high waste fill heights. The
stiff clay was varied in the range of 250-425 mls. Modulus
The nonlinear analyses performed in this study encom- reduction and damping curves proposed by Seed et al. (1984)
passed a large number of landfill configurations to allow eval- for sand and Vucetic and Dobry (1991) for clay were used,
uation of the relative importance of key parameters on the with reasonable unit weight profiles.
seismic response of a MSWLF. The landfill configurations in- Four dissimilar baseline rock motions were selected to study
cluded waste heights (H) of 10m, 20 m, 30 m, 45 m, 60 m, the effects of the input earthquake motion characteristics. Two
and 90 m, and three shear wave velocity (V.) profiles (Fig. 2), of these records are from the western United States, one is
resulting in initial 1D fundamental periods (T.-waste = 4H/ from eastern Canada, and one is a synthetic record, developed
V., where H = height of waste fill and V. = average initial by Abrahamson (personal communication, Dec. 1, 1995) for
shear wave velocity of the waste fill) ranging from 0.17 s to the analysis of the west span of the Oakland-San Francisco
2.74 s. Recent shear wave velocity measurements at six Bay Bridge. Significant duration (Trifunac and Brady 1975)
MSWLFs in southern California indicate that the mean shear ranged from 4.4 to 25 s for these records from Moment Mag-
wave velocity of municipal solid waste is generally stiffer than nitude 5.8, 6.7, 6.9, and 8.0 earthquakes. Fig. 4 shows the
previously thought and lies within the Kavazanjian et al. normalized acceleration response spectra for the four baseline
(1996) band shown in Fig. 2. Based on these data, as well as motions and presents key characteristics of these ground mo-
data from other landfills, the lower, medium (best), and high tions. The frequency content of these motions is quite differ-
V. profiles shown in Fig. 2 were used as reasonable variations. ent, and they cover a reasonably wide range of possible input
The unit weight of MSW was selected to vary from 6.3 kN/ motions. Mean period (Tm ) is used, as well as predominant
m 3 (40 pcf) at the surface to 11.9 kN/m 3 (75 pcf) at a depth period (Tp ). Tm is defined as (Rathje et al. 1998)
of 45 m. Below this depth, the unit weight remained constant.
These values are consistent with those recommended by Ka-
vazanjian et al. (1995). Fig. 3 shows the strain-dependent shear
modulus reduction and damping curves used for MSW, as rec-
~ C;'(i)
Tm = for 0.25 Hz ::5: fi ::5: 20 Hz (1)
ommended by Kavazanjian et al. (1995). More recent studies
[e.g., Idriss et al. (1995); Matasovic et al. (1995); Augello et
2: cf I
al. (1998)] have indicated that MSW may respond more elas-
ticalIy than initially thought. Therefore, analyses were also where Cj = Fourier amplitudes of entire accelerogram; and fi
performed with the Vucetic and Dobry (1991) shear modulus = discrete Fourier transform frequencies between 0.25 and 20
JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / MARCH 1998/243

J. Geotech. Geoenviron. Eng., 1998, 124(3): 242-253


1.0 30
0.9
0.8 25
0.7 20 -..
/ ~
~ 0.6 '-"

.r!
~
0.5 15
I:l.
0.4
10 0
i
0.3
Downloaded from ascelibrary.org by ECOLE POLYTECH/BIBLIOTHEQUE on 02/23/22. Copyright ASCE. For personal use only; all rights reserved.

0.2
5
0.1
0.0 0
0.0001 0.001 0.01 0.1 1 10
Shear Strain (%)
FIG. 3. Modulus Reduction and Damping Curves for MSW

4.0 NONLINEAR RESPONSE CHARACTERISTICS OF


StF......ol 5% damping
MSWLFs
Magnitude 8 (Bay Bridge)
3.5
Pacoima Dam Downstream The importance of fill height, waste stiffness, and founda-
d " tion conditions for the Pacoima Dam Downstream ground mo-
0 Yerba Buena Island "
'+:3 3.0 tion with its MHA scaled to 0.6g is illustrated in Fig. 5. The
E!
~ relative importance of these factors is similar for the other
] 2.5 ground motions used in this study. Examining Fig. 5, the ac-
celeration response spectrum of the input motion is shown as

I
tI)
2.0
a bold line, and the remaining lines show acceleration response
spectra computed at the top of the MSWLF. First [Fig. 5(a)],
comparing the 10 m and 45 m high MSWLFs overlying rock,
] 1.5 both with the medium V, profile, the response of the 10 m
waste fill is higher at periods less than about 1 s. Yet at longer
-.a
periods, the deeper waste fill, with its higher initial fundamen-
~
Z
1.0 tal period, has a slightly higher response. If the waste stiffness
of the 45 m waste fill is increased to the high V, profile, the
0.5 waste fill's response at periods of less than 2 s is significantly
higher, with the MHA increasing from 0.31g to 0.58g,
whereas, if the waste stiffness is decreased to the low V, pro-
0.0 file, the waste fill's response is substantially reduced at periods
0.01 0.1 10
Period (sec) of less than 2 s, with the MHA decreasing to O.09g. However,
the response at periods of greater than 2 s is higher for this
case than those calculated for the previous cases.
StatloD StFemol Pacoima YerbaBucna MagDitude8 The response of the 45 m high waste fill of medium stiffness
Dam Island Bay Bridge is largely unaffected by changing from a rock foundation to a
DoWllSll'e3Jll
Eartbquake Sal!Uenav Northridae LomaPrieta Svnthetic
sand foundation [Fig. 5(b)]. The importance of site conditions
Mw 5.8 6.7 6.9 8.0 at high periods is more apparent for the deep, soft clay site,
Distance CIuD) 114 9.3 77 15 which enhances long period motions that are amplified further
TDII« 0.16 0.4 0.64 0.25
1.0 0.58
by the waste fill. This result is important, because earthquake-
Tmll« 0.13 0.47
MHA(, 0.12 0.42 0.067 0.59 induced displacements are strongly influenced by long period
Arias IllteDlitv (emil) 9.9 94 4.3 1013 motions; hence, site conditions may be important in this re-
Slmlf1eut Duratloa (aee) 21.5 4.4 8.3 25
spect. Conversely, for this relatively high level of shaking, the
FIG. 4. Baseline Input Rock Motion Characteristics low period motions are reduced, with the MHA at the top of
the waste fill decreasing from O.3g to 0.2g when the waste fill
is underlain by soft clay. However, the MHA at the top of a
Hz. The Fourier amplitude is defined as the square root of the landfill situated atop a soft clay foundation would typically be
sum of the squares of the real and imaginary parts of the Fou- amplified at lower levels of shaking.
rier coefficient. Mean period can be estimated based on mag- These results are consistent with numerous other seismic
nitude, distance, and site conditions (Rathje et al. 1998). The site response studies, and they are important, as they demon-
four baseline rock motions were scaled to MHAs ofO.2g, O.4g, strate the significant variation in waste fill response due to
0.6g, and 0.8g to investigate the sensitivity of the results to reasonable variations of waste fill dynamic properties, fill
shaking intensity. Additionally, 19 other high intensity re- heights, and site conditions. These results emphasize the im-
corded and synthetic rock motions were used to assess the portance of waste material, fill geometry, and site characteri-
generality of the results. zation on a project-specific basis. As the seismic stability of
244/ JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAl ENGINEERING / MARCH 1998

J. Geotech. Geoenviron. Eng., 1998, 124(3): 242-253


3.0 As developed by Seed and Martin (1966) for earth em-
- Input. Pee. Dam Down_ 0.6 1 bankments, the horizontal equivalent acceleration (HEA) rep-
10 m MSW. Me<! Va, Rock Site. 1.-0.25 I resents the shear stress at the base of a potential sliding mass,
2.5 45m MSW, Low VI. Rock Site, 11-1.11 and thus is an appropriate parameter to represent seismic load-
45m MSW, Me<! VI, Rock Site, 11-0.15 • ing. The equivalent acceleration captures the cumulative effect
2.0 of the nonuniform acceleration profile in the potential sliding
45 m MSW. Hilh VI, Rock Site. 1.-0.58 I
mass, and hence, it is used herein to represent the seismic
~%Damping
loading for a potential sliding mass. Its maximum value, the
1.5 MHEA, which acts for only an instant during the earthquake,
can be expressed as MHEA = (Th.mu/rrv)g, where MHEA =
maximum horizontal equivalent acceleration acting on column
1.0 of fill above depth Z; Th,mu = maximum horizontal shear stress
Downloaded from ascelibrary.org by ECOLE POLYTECH/BIBLIOTHEQUE on 02/23/22. Copyright ASCE. For personal use only; all rights reserved.

at depth z; and rr v = total vertical stress at depth z (Bray et al.


0.5 1995).

0.0 ~i.:.f.l~a::::1::L:u.u..u.L.....J::i:~lUJ Base Stability


0.01 0.1 1 10 At lower levels of acceleration, where the landfill response
(a) is fairly linear, Bray et al. (1995) found that MHEA for base
Period (sec)
sliding increased proportionally with the peak acceleration
3.0 (MHA) of the earthquake rock motion. Numerous studies [e.g.,

2.5
-
r--T""T''TTTmr---r''T"Tnmrr--T'''T'TTTnn
Input, PIc. Dam DownsIJeIm 0.61
45m MSW. Me<! Va, Rock Site
Borcherdt (1994); Seed et al. (1991)] have shown that the peak
ground acceleration at the top of a soil site is amplified at low
rock input peak accelerations and attenuated at high rock input
45m MSW. Me<! Va, Snl Site
peak accelerations. As an example, the semiempirical/semian-
45m MSW. Me<! Va, Clll)' Site alytical Seed et al. (1991) MHA-site versus MHA-rock rela-
2.0 ~%Damping tionship for a deep cohesionless or stiff cohesive soil site is
shown in Fig. 6(a). To account for the nonlinearity expressed
1.5 in this type of relationship, the nonlinear response factor
[(NRF) = MHA-siteIMHA-rock] was developed. This factor
recognizes that the MHEA for a potential sliding mass will
1.0 not be directly proportional to the input MHA-rock across a
wide range of shaking intensities. The nonlinear response fac-
tor shown in Fig. 6(b) is based on the Seed et al. (1991) re-
0.5 lationship shown in Fig. 6(a), because the range in dynamic

0.0 1-...&..J:.Lu.u.IL-........L.J...LU.W-......L....&.;;l-.lWl 1.0


0.01 0.1 1 10
(b) 0.8
Period (sec)
FIG. 5. Comparison of Responses from Different Landfill Con-
~
0.6
figurations: (a> Effect of Waste Fill Height and Stlffne..; (b> Ef- .~
fect of Site Conditions

an MSWLF depends on its dynamic response and strength


characteristics, as well as the design rock motions, developing
i 0.4

0.2
good waste fill property characterizations remains a top pri-
ority, Moreover, a simplified design procedure that does not 0.0
capture the importance of the waste fill dynamic stiffness and 0.0 0.2 0.4 0.6 0.8 1.0
height (Le., its fundamental period) is judged to be inappro- (a) MHA, rock (g)
priate.
2.0 I
SEISMIC LOADING
Equivalent Acceleration
The pseudostatic slope stability method (Seed and Martin
J 1.5 1-,
, -
1966) is often used in practice because of its relative simplic-
ity, its general consistency with case histories for nondegrad-
ing materials (Seed 1979), and its citation in governing state
and federal regulations and guidelines [e.g., "Solid Waste"
(1993)). With this method, the seismic loading is represented
h
j
1.0

0.5 ~
--- ---
-
by a seismic coefficient that acts as a horizontal destabilizing
~
force on the landfill. The seismic coefficient, in conjunction 0.0 I I
with compatible factor of safety (FS) and dynamic strengths, 0.0 0.2 0.4 0.6 0.8 1.0
is crucial if the results of a pseudostatic analysis are to be MHA, rock (g)
meaningful. A simplified seismically induced permanent de- (b)
formation analysis [e.g., Makdisi and Seed (1978)) also re- FIG. 6. Diagram of: (a> Site Amplification Factor (Seed et al.
quires an estimate of the seismic loading. 1991); (b) Nonlinear Response Factor

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / MARCH 1998/245

J. Geotech. Geoenviron. Eng., 1998, 124(3): 242-253


2.0

1.8 •<> KlMmlnjillne( aI. (l99S) MSWclyn8mic properli<s. MHA~.4g


KlMmlnjilll c( aI. (l99S) MSWclyn8mic properties, MHA>O.4g
1.6
..
x
x
X
0
Vucetic am ~(1991) ClayPl.-3O d)namic JI'OIlSIies
Results from othor inpt tmliCIII
1.4

1.2
. ~ x RllCIc site beot fit am +(- one staOO8rd error_
Base1ine. tmliCIII:
I'lIc:obm DBrn IIcMnstre8m 0.2, 0.4, 0.6, 0.8 g
1.0 St. F~ 0.2, 0.4, 0.6, 0.8 g
Yalle Buom Islam 0.2, 0.4
~8 0.2, 0.4, 0.6, O. g
f
0.8
MSW . : 10, 20, 30, 45, 60, 90 m
~n: Low, Medillll, ani Higb IIlII&"S
0.6

0.4
Downloaded from ascelibrary.org by ECOLE POLYTECH/BIBLIOTHEQUE on 02/23/22. Copyright ASCE. For personal use only; all rights reserved.

0.2

0.0
(a) 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0
Ts-waste/fm-eq
1.2
RllCIc site best fit (R'=O.86)
1.0 0 SarJj site beot fit ~.80)

~
0 SarJj site daIa

0.8 ==-~~.79)am+{-one
Oay site dala
*
Baseline InpJI tmlions:

i
0.6 Pacolml DBrn Dowmtretm 0.2, 0.4, 0.6, 0.8 g
St. F~ 0.2, 0.4, 0.6, 0.8 g
Yalle Buena Island 0.2, 0.4
~ 80.2, 0.4, 0.6, O. g
f
0.4 MSW .gIIt: 10,20, 30,45,60, 90m
Vs profile: Low, Medium, am Higb rt\lllOS

0.2

0.0
(b) 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0
Ts-waste!fm-eq
FIG. 7. Normalized Maximum Horizontal Equivalent Acceleration versus Normalized Fundamental Period of Waste Fill: <a> Rock Site;
<b) All Sites

stiffnesses of this site category is close to that used for waste and height), a number of other trends are important. For in-
fill. stance, to a lesser degree, the modulus reduction and damping
The nonlinear response factor captured the nonlinear vari- curves used to represent the waste fill's response at larger
ation in seismic loading across a range of ground motion in- strains are important, with the slower reducing Vucetic and
tensities. In Fig. 7, the MHEA at the base of the landfill nor- Dobry (1991) PI = 30 cohesive soil curves giving uniformly
malized with respect to the MHA and corresponding NRF of higher responses at comparable T,ITmvalues than analyses per-
the input rock motion is plotted against the initial fundamental formed using the Kavazanjian et al. (1995) waste fill curves.
period of the waste fill (T,-waste) normalized by the mean Site condition effects are displayed in Fig. 7(b). Regression
period of the input rock motion (Tm-eq). Introduction of the curves for the rock, sand, and clay site results for the various
nonlinear response factor and replacement of the predominant landfill configurations and input rock motions described pre-
period by the mean period in this normalization reduced the viously are shown in this figure. In terms of MHEA, the
variation of the data about the median relationship (i.e., R 2 MSWLF responses at rock and sand sites are comparable, but
increased from 0.74 to 0.86). For the rock site cases shown in the response at clay sites is lower. At significant levels of shak-
Fig. 7(a), the data (which include 14 rock motions ranging in ing, nonlinearity within the deep, soft clay reduces the inten-
intensity from 0.2g to 0.8g, three V, profiles, two pairs of mod- sity of the seismic loading. Due to the long period motion
ulus reduction and damping curves, and six waste fill heights, amplification at deep, soft clay sites [Fig. 5(b)], these sites may
all together 324 analyses) follow a well-defined trend, except not necessarily be less critical in terms of earthquake-induced
near the resonance condition (T,-wasteITm-eq < 1). Note that displacements, and this will be discussed later in this paper.
due to modulus reduction, resonance does not occur at T,ITm Fig. 7 is not meant to replace site-specific seismic response
= 1, because T, is defined as the initial (small strain) funda- analyses; however, it does provide useful insight on the im-
mental period of the waste fill system. A degraded T" as was portance of the waste fill's dynamic characteristics and the
used by Makdisi and Seed (1978), is difficult to define in a input rock motion's intensity and frequency content on the
fully nonlinear analysis, and attempts to estimate it with an calculated MHEA. As this graph has been prepared with nor-
equivalent-linear approximation did not reduce the scatter. As malization parameters that may be estimated based on avail-
a check, additional analyses were performed with three landfill able information for many projects, Fig. 7 may be used as a
configurations using an additional 19 recorded and synthetic guide in the selection of an appropriate seismic coefficient for
high intensity rock motions, and the consistency of the results simplified pseudostatic and deformation analyses. It should be
indicates that Fig. 7(a) is applicable for motions other than remembered, however, that duration of strong shaking is an
those used in this sensitivity study. important earthquake parameter that is not captured by
Although Fig. 7(a) highlights the importance of the funda- MHEA.
mental period of the waste fill (i.e., its shear wave velocity For comparison with the normalized graph presented by
246/ JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / MARCH 1998

J. Geotech. Geoenviron. Eng., 1998, 124(3): 242-253


Bray et al. (1995) based on ID equivalent-linear analyses, the Cover Stability
response of a 30 m high waste fill with average waste prop-
erties (Ts-waste = 0.63 s) to an earthquake record with Previous analytical results [e.g., Bray et al. (1995); Kava-
Tm-eq = 0.6 (Tp-eq = 0.4) at two intensities is examined. For zanjian and Matasovic (1995); Singh and Sun (1995)] and case
this case with MHA-rock = 0.2g, the D-MOD results shown records from the instrumented 011 landfill [e.g., Idriss et al.
in Fig. 7 indicate MHEA = 0.12g-0.17g, which is in good (1995)] have shown that ground motions can be amplified as
agreement with the Bray et al. (1995) estimate of O.l1g- they propagate up through a waste column. Given the poten-
0.17g. However, at MHA-rock = 0.8g, the fully nonlinear tially low strengths of geosynthetic interfaces in cover systems
D-MOD analyses provide MHEA estimates between 0.32g- due partly to the relatively low normal stresses at these shallow
0.45g, whereas extrapolation of the Bray et al. (1995) results, depths, the seismic stability of cover systems is an important
which were developed from less intense ground motions, design concern. Although 2D analyses may be warranted for
would suggest relatively high levels of MHEA = 0.44g-0.68g. analyzing landfill cover systems in many cases, the results
Downloaded from ascelibrary.org by ECOLE POLYTECH/BIBLIOTHEQUE on 02/23/22. Copyright ASCE. For personal use only; all rights reserved.

Thus, Fig. 7 appears to provide more reasonable estimates from these ID analyses will be discussed, as a number of
of MHEA for base sliding over a wide range of intensity lev- insights are relevant.
els. Partial results from the D-MOD analyses at rock sites are

0.0
(a) 0.0 0.2 0.4 0.6 0.8 1.0
MHA, rock (g)

2.2
()
X MediwnVs
2.0 ()
o HighVs
1.8 Best fit and +/- one stmIard
deviation curves

f 1.6
1.4
1.2
1.0
o
0.8
0.6 x

0.4 •
0.2
0.0
(b) 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0
Ts-waste/Tm-eq
FIG. 8. Diagram of: (a) MHA, Top versus MHA, Rock Calculated by D-MOD and Observed at 01/ Landfill; (b) Normal/zed Maximum Ac-
celeration versus Normalized Fundamental Period of Waste Fill

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / MARCH 1998/247

J. Geotech. Geoenviron. Eng., 1998, 124(3): 242-253


shown in Fig. 8(a), in which the peak acceleration at the top zation scheme used to develop Fig. 7 was applied to these
of the waste fill (MHA, top) is plotted against the peak rock data, and the resulting plot is shown in Fig. 8(b). Note that
acceleration (MHA, rock). Notice that the trend of the D-MOD reasonable variations in just the waste's V. profile produce sig-
results is consistent with observations at the 011 landfill as nificant differences.
well as the Seed et al. (1991) relationship for stiff soil sites. This 1D response study emphasizes the importance of per-
These results are also consistent with other previously refer- forming project-specific studies to evaluate cover stability.
enced landfill studies, with the data being enveloped by a sim- Simplified stability procedures that require an estimate of the
ilar relationship developed by Harder (1991) for earth dams. peak acceleration at the top of the waste fill should also be
However, the scatter in the data is significant, which should viewed with caution. These findings highlight the relative ro-
be a source of concern for engineers responsible for evaluating bustness of using Fig. 7 directly to estimate the seismic load-
the seismic performance of cover systems. The scatter in Fig. ing (MHEA) for base sliding as opposed to using a procedure
8(a) is due to variations in the fill height, waste stiffness, and that requires the peak acceleration at the top of the waste fill
Downloaded from ascelibrary.org by ECOLE POLYTECH/BIBLIOTHEQUE on 02/23/22. Copyright ASCE. For personal use only; all rights reserved.

input rock motions. Attempts to reduce this scatter through to be estimated first in order to evaluate MHEA within the
normalizations were only somewhat successful. The normali- landfill [e.g., a Makdisi and Seed (1978) approach].

K)""O.
Dec:oupled
60
•••••• Coupled
- - RigidBloclc:
40
Input MHA~.4 g
Damping 15%
20 t--------"-'~~-1

1 2 3 4 123 4
Tslfm Tslfm
(a)
1000 r-r-...-.,......,.....,.-..,.....,."""'T"""T"""W
<>

100 100
• <>
• <>
. ~

10 10 •
1 1 •
Tsffm-1.0 • TIff.......O
0.1 • Coupled
<> 0.1 • Coupled •
o Decouplcd o Decoupled
0.01 0.01 ..........................-"-............-.&-............
o 0.2 0.4 0.6 0.8 1 o 0.2 0.4 0.6 0.8 1
Ky/Kmax Ky/Kmax
(b)
100 .........-r-...,......,..-..--r-...,......,..-..--.
• TslTm= 1.0 • • TslTm=4.0
• •




•t •·
•• ••
• ••

• Damping 15%
-50 L..-L...........L-.........- - ' -......L.......L.--1 -50 .....
0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0
KylKrnax KylKrnax

FIG. 9. Comparison of Displacements Calculated Using Decoupled and Coupled Deformable Sliding Block Analyses with 15% Ma-
terial Damping: (a) Fundamental Period; and (b) k,llcm.. Effects for Synthetic Magnitude 8 Input Motion (MHA = 0.4g); and (c) Differ-
ences for Several Input Motions

248/ JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / MARCH 1998

J. Geotech. Geoenviron. Eng., 1998, 124(3): 242-253


SEISMICALLY INDUCED PERMANENT the average overestimation from the decoupled approach at
DISPLACEMENTS 15% material damping was on the order of 20%. They con-
cluded that "the errors arising from the decoupling assumption
Sliding Block Analysis are insignificant compared to other uncertainties involved in
the use of the sliding block analogy." Gazetas and Uddin
The MHEA is useful for representing the maximum desta- (1994) performed coupled and decoupled 20 finite element
bilizing force induced by the earthquake shaking, but it does analyses of earth dams and similarly found that "the two-step
not address directly the issue of earthquake-induced permanent approximation leads, in general, to very good results, although
displacements. In areas of high seismicity, it may be overly it often provides conservative estimates of sliding deforma-
conservative to design MSWLFs to resist the maximum desta- tions." Kramer and Smith (1997) agreed that the decoupled
bilizing force without allowing some seismically induced per- approach was somewhat conservative for stiff and/or shallow
manent displacement. Thus, the Newmark (1965) sliding rock sliding masses, but that "it may produce unconservative esti-
Downloaded from ascelibrary.org by ECOLE POLYTECH/BIBLIOTHEQUE on 02/23/22. Copyright ASCE. For personal use only; all rights reserved.

approach, which was originally developed for analyzing earth mates for failure masses that are soft and/or deep." Chopra
embankments, may be applied to MSWLFs to estimate earth- and Zhang (1991) reexamined the decoupling assumption in
quake-induced permanent displacements. The assumptions and the context of earthquake-induced base sliding of concrete
limitations involved in applying a Newmark deformation anal- gravity dams. These analyses computed the linear elastic re-
ysis to MSWLFs have been discussed in Bray et al. (1995). sponse of a dam using its fundamental mode of vibration based
Due to the simplicity of the analysis and the uncertainties in- on distributed mass, and included reservoir induced forces and
volved in characterizing geomaterials and earthquake ground a rigid perfectly plastic stick-slip foundation interface. They
motions, the calculated earthquake-induced displacement found that the decoupled approximation could provide an or-
should be viewed as merely an index of earthquake perfor- der-of-magnitude estimate of the sliding displacement calcu-
mance. The Franklin and Chang (1977) calculational proce- lated through a coupled procedure, which was conservative for
dure implemented in the program TNMN (Pyke and Beikae large displacements but not necessarily for small displace-
1991) is used to estimate permanent displacements based on ments. Because commonly accepted tolerable earthquake-in-
the computed horizontal equivalent acceleration-time history duced base displacements for MSWLFs (e.g., 2-30 em) are
at the depth of sliding. Use of equivalent acceleration allows considerably less than those values often considered accepta-
for the seismic response of the deformable potential sliding ble for earth or concrete gravity dams (e.g., 1 m or more), this
mass to be represented in the Newmark rigid sliding block is potentially an important issue.
procedure (Makdisi and Seed 1978). However, a potentially The writers used the procedures developed by Chopra and
significant limitation of this simplified procedure is that the Zhang (1991), except dynamic properties appropriate for
computation of the seismic response of the potential sliding MSWLFs and a mode shape satisfying level ground boundary
mass is decoupled from the subsequent double-integration of conditions were used without the reservoir to reexamine this
the computed equivalent acceleration-time history. issue. Representative results are shown in Fig. 9, where ley is
A number of investigators have addressed this concern. Us- the yield acceleration and !emu = MHEAlg. In Fig. 9(a), de-
ing a single lumped mass, stick-slip model, Lin and Whitman coupled displacements are conservative, except for large fun-
(1983) found that the decoupled approximation provided rea- damental periods (Ts ). Note that a rigid sliding block analysis
sonably conservative estimates of the seismically induced per- (i.e., direct integration of the input acceleration-time history)
manent displacements calculated by a coupled analysis. For can be significantly unconservative, and it should not be used.
example, for simulated earthquake motions in which the yield Examining Fig. 9(b), the decoupled approximation does pro-
acceleration of a sliding block representing a deep base failure vide a reasonable estimate of the coupled earthquake-induced
was about half of the expected maximum block acceleration, displacements (i.e., within a factor of three). However, a de-

1000 Rock Site Results


All landfill configurations

6 Sl Fcrrcol 0.2 g

100 ... StFe.,..,1 0.4 g


t:. 51 Fcrrcol 0.6 g
A 51 Ferrool 0.8 g
-.
~ 10 •<> P1Ic. Dam Dnstnn 0.2 g

--
~
- 0
Pac. Dam Dnlllrm 0.4 g
Pac. Dam Dnstnn 0.6 g

e <>
X
Pac. Dam Dnstnn 0.8 g
YBIO.2g

-i
YB10.4g

~ *
0 MlIlP'ilude 8 0.2 g
.~ 0.1
0 0
181 MlIlP'ilude 8 0.4 g
0 0 MlIlP'ilude 8 0.6 g
&
6
EEl MlIlP'ilude 8 0.8 g
• 100prob.ofex_
0.01
I~
....
16% prob. of excoedonoe
31% prob. ofexcoedonoe
_0 '0% prob. of exceedInoe
0.001 0 Remits from other Input
motion.
0.0 0.2 0.4 0.6 0.8 1.0
kylkmax
FIG. 10. Base Liner Sliding Block Displacements

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / MARCH 1998/249

J. Geotech. Geoenviron. Eng., 1998, 124(3): 242-253


coupled analysis may predict smaller displacements than a scaled to MHAs of 0.2, 0.4, 0.6, and 0.8g, the upper bound,
coupled analysis for systems with larger values of kylkmax . In median, and lower bound calculated seismically induced per-
these cases, the displacements calculated from both analyses manent displacements are shown for all landfill configurations
are generally small (i.e., less than a few centimeters). Differ- sited on rock, sand, and deep soft clay. For these input mo-
ences between decoupled and coupled displacements for sev- tions, the median and upper bound displacements are signifi-
eral input motions are shown in Fig. 9(c). For cases applicable cantly higher for cases where the MSWLF is situated atop a
to landfills where only minor earthquake-induced base dis- deep soft clay foundation. Even the sand site produces signif-
placements are generally tolerable (Le., ky/kmax > 0.5), the de- icantly larger displacements than the rock site at low ky/kmax
coupled approximation is reasonable, so it will be used in this values. Hence, site conditions are important in evaluating seis-
study to evaluate the factors influencing earthquake-induced mically induced permanent base displacements. Note, how-
displacements of MSWLFs. At lower kylkmax ratios (especially ever, that the larger displacements calculated at soft soil sites
at higher T,/Tm ratios) where the calculated displacements are for a specified ky/kmax ratio are offset by the results shown in
Downloaded from ascelibrary.org by ECOLE POLYTECH/BIBLIOTHEQUE on 02/23/22. Copyright ASCE. For personal use only; all rights reserved.

large, the decoupled approximation is less reliable, and this Fig. 7(b), indicating that deep soft clay sites produce lower
may be important in evaluating earth dams. kmax values for identical rock motions and landfill configura-
tions. Thus, for identical landfill configurations and rock mo-
Results tions, the base sliding displacement calculated at soft sites is
comparable to that calculated at stiff sites (i.e., only slightly
Calculated seismically induced permanent displacements higher when ky < 0.1, but slightly lower at higher ky values).
(U) for the base sliding case for all landfill configurations [see
Calculated earthquake-induced cover displacements for the
Fig. 7(a)] sited on rock undergoing the 14 input rock motions
listed are shown in Fig. 10. In this figure, U is plotted versus 1000
selected ky/k max ratios of 0.2, 0.4, 0.6, and 0.8. Additionally, t::. Mw 6.25 - SId Emr 0.35
three landfills were analyzed with another 19 input rock mo-
tions, so that 309 data points are plotted at each kylkmax ratio.
There is considerable scatter, both with respect to results from
j 100 .... ~ .. ~x
X
<>
Mw 7.0· SUI. Emr 0.33
Mw 8.0 - SUI. Emr 0.36

different input motions and results from different landfill con-


figurations undergoing the same input motion. Much of this
!'[ 10
scatter is expected. For example, the longer duration Magni- ~~
:= .,r
tude 8 earthquake produces large calculated displacements, tnO\
1
~~
and the high frequency St. Ferreol motion produces relatively
small displacements. Moreover, at the same intensity level, the ••
Yerba Buena Island record generally produces larger displace-
ments than the Pacoima Dam Downstream record because of
its significant long period motion, which better matches the
!j 0.1
x

I
fuJI Dala Set Rqp'essi0ll
long period response characteristics of most waste fills. For a
0.01 16% IlId 84% prob. ofCllCl:IlOdance x
given input motion, significantly larger displacements are cal-
S% '"'" 95% prob. of c:x=lonce
culated for landfill configurations with stiffer response char- log(Uikrlllx°DS-9S) - 1.87·3.477"kylknolx, SUI. Emr 0.35
acteristics that more closely match the short period motions 0.001
contained in most rock records. As the landfill's fundamental 0.0 0.2 0.4 0.6 0.8 1.0
period increases, due to increasing height or decreasing shear kylkmax
wave velocity, the calculated displacements decrease. This
finding is consistent with the results presented in Fig. 7(a) and FIG. 11. Normalized Base Liner Displacements
results presented in Augello et al. (1995a). In fact, at a spec-
1000 ~""-r---r--r---r--'r-.....,.-...----r--:r
ified kylkmax ratio, the calculated displacement is roughly pro-
portional to MHEA, indicating that those factors that have
been shown to affect MHEA also affect U. However, there is
considerably more scatter in the calculated displacements, es-
pecially at higher MHEA values. 100
Several attempts were made to normalize the calculated dis-
placement data presented in Fig. 10. These attempts were of
limited success. The best normalization for the cases analyzed
is shown in Fig. 11. In this figure, the calculated seismically 10
induced permanent displacement is normalized by kmax
(MHENg) and significant duration (D S - 9S %) of the input mo-
tion. MHEA has been shown to capture the important effects
of earthquake intensity and frequency content (e.g., Fig. 7),
and significant duration captures another key ground motion
characteristic. The normalized displacement decreases with in-
creasing kylkmaxo and shows considerably less inter- and in- -+- PIlC. Dam Dnstrm, rock site

traearthquake scatter. Hence, an order-of-magnitude estimate ~ PIlC. Dam Dnstrm, sand site
of earthquake-induced displacement can be made given an es- 0.1 ~ PIlC. Dam Dnstrm. clay site
timate of the intensity (MHA), frequency content (Tm ), and ....... M"llIIitudc 8. rock site
duration (D S - 9S %) of the design rock motion, and the dynamic ---w- M"llIIitudc 8, sand site
response characteristics (T,) and strength (ky ) of the landfill. -e- M"llIIitude 8, clay site
With this information, the seismic coefficient kmax (which is 0.01 L.--........._..L---L_~.......lL...... ......._""----'-_.....I----I
MHENg) can be estimated using Fig. 7(a), and the seismically 0.0 0.2 0.4 0.6 0.8 1.0
induced displacement (U) can then be estimated using Fig. 11. kylkmax
Site effects are apparent in Fig. 12. For both the Pacoima
Dam Downstream and Synthetic Magnitude 8 rock motions FIG. 12. Site Condition Effects on Base Liner Displacements

250 I JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING I MARCH 1998

J. Geotech. Geoenviron. Eng., 1998, 124(3): 242-253


1000 Rock Site Results
0 All landfill oonfiguratiom
I:::. St Forreol 0.2g
100
• St Ferreol 0.4g

B
-..-.-
6.
JJ..
St Forreol 0.61
St Femol 0.8g
::> 10 • Pac. DIm D1slrm 0.2g

I!
¢ Pac. Dam D1slrm 0.4g

0 Pac. Dam Dnslnn 0.61


~ Pac. DlmD1slrmO.8g
1 X YBIO.2g
Downloaded from ascelibrary.org by ECOLE POLYTECH/BIBLIOTHEQUE on 02/23/22. Copyright ASCE. For personal use only; all rights reserved.

YBlO.4g

.." *0 Magnitude 8 0.2g

~ 0.1 181
0
Maanitude 8 0.4g
MIgllitude 8 0.61
~ EE Magnitude 8 0.8g

8 0.01 ..
10',1, prob. of exceedmcc
16% prob. of exceedcnce
~O% prob. of exceedcnce
A
"a<> 16% P"'!!. ofexceedmcc
fll' bise deflI'maticns
0.001 0 ~r:sfrcm other inpIl
0.0 0.2 0.4 0.6 0.8 1.0
ky/kmax
FIG. 13. Cover Liner Sliding Block Displacements

rock site cases discussed previously are shown in Fig. 13. ever, the engineer must decide whether or not 15 cm of per-
Again, the data exhibit considerable scatter, emphasizing the manent displacement along the landfilllliner interface at the
need for project-specific earthquake-induced deformation anal- site is tolerable. Moreover, site effects may be significant at a
ysis of cover systems. In general, landfills with the low V. kylkmax ratio of 0.6 if the landfill foundation has a tendency to
profile had larger calculated cover displacements than those amplify long period motions, as is the case for a deep soft
with the high V. profile, due to the long period motion ampli- clay site (see Fig. 12). For these special soil conditions, pro-
fication in the softer waste fill. The calculated cover displace- ject-specific analyses should be performed or the seismic co-
ments are significantly larger than those calculated for base efficient should be increased accordingly. For instance, Fig. 12
sliding (note the 16% probability of exceedance lines for both suggests that the seismic coefficient for landfill base sliding at
cases shown in Fig. 13). Combining this observation with the a deep, soft clay site should be 80% of kmax to ensure that the
unknown reliability of calculating cover displacements using calculated seismically induced permanent displacements are
ID analyses, the seismic stability of MSWLF cover systems just as likely to be less than 15 cm.
requires additional study. Although not shown (because the
trend is similar to that shown previously for the base sliding COMPARISON OF ANALYTICAL RESULTS TO
case in Fig. 12), deep soft clay sites produced larger cover OBSERVED PERFORMANCE
displacements than those calculated at rock sites at similar
kylkmax ratios. The results of this analytical study (e.g., Figs. 7 and 11)
were used to estimate the expected range of earthquake-in-
Calibrated Pseudostatic Slope Stability Analysis duced permanent displacements for a number of California
landfills that were strongly shaken by the 1989 Loma Prieta
A conservative, preliminary pseudostatic slope stability and 1994 Northridge earthquakes. Landfill configurations and
analysis may be performed with a seismic coefficient cali- seismic performances are described in Buranek and Prasad
brated with the preceding seismically induced permanent de- (1991) and Augello et al. (l995b). Calculations and observa-
formation results. For the sake of simplicity, one could use the tions are summarized in Table 1. The yield acceleration rec-
MHEA/g directly as the seismic coefficient with FS =:: 1 and ommended by Buranek and Prasad (1991) was used for the
conservative strength values, because MHEA represents the Loma Prieta earthquake case records, and ky was calculated
maximum destabilizing force during the earthquake, which with the 2D limit equilibrium slope stability analysis program
aCts for only an instant. Conservative strength values and a UTEXAS3 (Wright 1990) for the Northridge earthquake cases
calculated FS of at least one means that the resisting force is using the strength values recommended by Augello et al.
always equal to or higher than this maximum destabilizing (1995b). The mean and plus one standard error kmax values
force, and hence no seismically induced deformations would were estimated using Fig. 7, based on the estimates of T"
be calculated for this case. For cases where some level of MHArocko and Tm shown. The calculated displacements, U,
deformation is acceptable, Fig. 10 indicates that for a wide were then estimated using the median and 16% probability of
range of landfill heights, waste fill properties, and input rock exceedance lines shown in Fig. 11, with significant duration
motions, the calculated base sliding permanent displacements estimated using the Abrahamson and Silva (1996) relationship.
for the rock site case are less than about 15 cm (with about a Median and 84% values were used to develop reasonably con-
10% probability of exceedance) when the yield acceleration is servative displacement estimates. As shown in Table 1, the
greater than 60% of the kmax • Hence, 60% of the MHEA/g may earthquake-induced displacements calculated using the results
be used as an estimate of the seismic coefficient to be used in from the present study are consistent with the magnitude of
pseudostatic analyses in combination with conservative dy- permanent displacements observed at these landfills after the
namic strength properties and a factor of safety of one. How- earthquakes. Except for two cases, no cracking was observed
JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / MARCH 1998/251

J. Geotech. Geoenviron. Eng., 1998, 124(3): 242-253


TABLE 1. Back Analysis of Landfill Performance during Recent Earthquakes
MHA, rock
Idriss (1991) Tm D5 - 1l5% T. kmax Calculated U
Landfill EO (g) (s) (s) (s) ky (mean/+1a) (em) Observed damage
(1 ) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Buena Vista LP 0.40 0.6 13 0.64 0.26" 0.17/0.23 None No cracking"
Guadalupe LP 0.45 0.6 13 0.64 0.20" 0.24/0.32 0.3-4 Minor downslope movement"
Pacheco Pass LP 0.30 0.6 14 0.76 0.30" 0.16/0.21 None No cracking"
Marina LP 0.21 0.6 15 0.59 0.26" 0.1210.16 None No cracking"
Zanker Road LP 0.19 0.6 15 0.56 0.14" 0.07/0.10 None No cracking"
Lopez Canyon C-A NR 0.42 0.5 11 0.64 0.27 0.18/0.25 None No cracking"
Lopez Canyon CoB NR 0.42 0.5 11 0.45 0.35 0.24/0.33 None No cracking"
Chiquita Canyon C NR 0.33 0.6 11 0.64 0.09 0.18/0.24 3-20 24 cm tear in high-density poly-
Downloaded from ascelibrary.org by ECOLE POLYTECH/BIBLIOTHEQUE on 02/23/22. Copyright ASCE. For personal use only; all rights reserved.

ethylene (HDPE)"
Chiquita Canyon D NR 0.33 0.6 11 0.64 0.10 0.18/0.24 2-15 30 cm tear in HDPE"
OIl Section HH NR 0.11 0.6 16 0.47 0.14 0.10/0.13 None 5-15 cm cover cracks"
OIl Section HHc - - - - 0.81 0.08 0.13/0.16 2-25 5-15 cm cover cracks"
Sunshine Canyon NR 0.46 0.5 11 0.77 0.31 0.17/0.23 None 2-30 cm cover cracks"
Sunshine CanyonC - - - - 1.3 0.25 0.30/0.37 0.2-5 2-30 cm cover cracks"
"Observations of damage and k,; values from Buranek and Prasad (1991).
'Observations of damage from Augello et ai. (1995b).
CAnalyses performed for shallow sliding along cover system using Fig. 13.

when the calculated displacement was zero, and some distress rock motion, and by the foundation soils. The waste fill's dy-
was observed for cases when the calculated displacement was namic response characteristics are also important, and many
nonzero. As the calculated earthquake-induced displacement of these factors are captured by km... Normalizing the calcu-
should be viewed as merely an index of earthquake perfor- lated seismic displacement by kmax and significant duration re-
mance, the results presented in this paper are judged to provide duced the scatter in the results considerably. For the base slid-
an index of expected seismic performance that correlates rea- ing case at a rock or shallow, stiff soil site, if kylkmax is greater
sonably well with the actual seismic performance of these than 0.6 (or in using a seismic coefficient equal to 0.6 times
landfills. Moreover, these results are consistent with results MHEAlg with conservative strength values, the pseudostatic
obtained by double integration of the equivalent acceleration factor of safety exceeds one), the calculated permanent dis-
time history computed from a project-specific dynamic anal- placements are likely to be less than 15 cm (Fig. 10). At deep,
ysis. A simplified seismic design procedure for solid-waste soft sites, the seismic coefficient would need to be increased
landfills, which is based on the results of this study, is de- to 0.8 times the MHEAlg to ensure the same likeliness that a
scribed in Bray et al. (1998). displacement of 15 cm was not exceeded (Fig. 12), but re-
member that the MHEA calculated at a deep, soft clay site is
CONCLUSIONS generally lower than that at a rock site. These analytical results
In areas of high seismicity, where EPA regulations require are consistent with observations from available case records,
the use of relatively high intensity earthquake motions, fully and may be used for a simplified design procedure (e.g., Bray
nonlinear seismic response analyses are appropriate. Results et al. 1998).
from the nonlinear D-MOD analyses performed in this study
indicate that the dynamic response of an MSWLF can vary ACKNOWLEDGMENTS
significantly due to reasonable variations of dynamic waste Financial support was provided by the David and Lucile Packard
properties, fill heights, site conditions, and input rock motions. Foundation and National Science Foundation under grant BCS-9157083;
However, for the case of base sliding along the landfilVfoun- this support is gratefully acknowledged. Dr. Matasovic of GeoSyntec
dation interface, the maximum seismic loading represented by Consultants shared his D-MOD program and expertise with us. Reviews
by A. J. Augello and Dr. Seed of the University of California, Berkeley,
MHEA is shown to follow a well-defined trend. MHEA de- were particularly helpful.
pends largely on the mean period (Tm ) and MHA of the input
rock motion adjusted by the NRF and the initial fundamental APPENDIX. REFERENCES
period of the waste fill (T, =4HIV,). MHEA may be estimated
using the normalization presented in Fig. 7. Abrahamson, N. A., and Silva, W. J. (1996). "Empirical ground motion
The results of this study indicate that the MHA (or MHEA) models." Rep. Prepared for Brookhaven Nat. Lab.
Anderson, D. G., and Kavazanjian, E. Jr. (1995). "Performance of land-
calculated at the top of a landfill can vary erratically for rel- fills under seismic loading." 3rd Int. Con! on Recent Adv. in Geotech.
atively modest variations in landfill configurations and input Engrg. and Soil Dyn., Univ. of Mo. at Rolla, St. Louis, Mo., III, 1557-
motions. Hence, simplified procedures for estimating the seis- 1587.
mic loading for cover systems are judged to be lacking, as Augello, A. J., Bray, J. D., Leonards, G. A., Repetto, P. C., and Byrne,
significant conservatism would need to be employed with R. J. (l995a). "Response of landfills to seismic loading." Proc.,
these procedures. Moreover, procedures that require an esti- Geoenvironment 2000, ASCE Geotech. Spec. Publ. No. 46, ASCE, Res-
ton, Va., 1051-1065.
mate of the MHA at the top of a landfill to estimate the seismic Augello, A. J., Matasovic, N., Bray, J. D., Kavazanjian, E. Jr., and Seed,
loading within and at the base of the landfill should be used R. B. (l995b). "Evaluation of solid waste landfill performance during
with caution, due to the relatively large uncertainty in esti- the Northridge earthquake." Earthquake design and peiformance of
mating MHA at the top of a landfill. solid waste landfills. ASCE Geotech. Spec. Publ. No. 54, M. K. Yegian
A reasonable estimate of the seismically induced permanent and W. D. L. Finn, eds., ASCE, Reston, Va., 17-50.
displacement can be obtained using the decoupled approxi- Augello, A. J., Bray, J. D., Abrahamson, N. A., and Seed, R. B. (1998).
"Dynamic properties of solid waste based on back-analysis of 011 land-
mation delineated in Makdisi and Seed (1978). The magnitude fill." J. Geotech. and Geoenvir. Engrg., ASCE, 124(3),211-222.
of earthquake-induced permanent displacements is intimately Borcherdt, R. D. (1994). "Estimates of site-dependent response spectra
linked to the ratio of kvlkm... but it is also influenced signifi- for design (methodology and justification)." Earthquake Spectra,
cantly by the intensity, frequency content, and duration of the 10(4), 617-653.

252/ JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / MARCH 1998

J. Geotech. Geoenviron. Eng., 1998, 124(3): 242-253


Bray, J. D., Augello, A. J., Leonards, G. A., Repetto, P. C., and Byrne, Lin, J. S., and Whitman, R. V. (1983). "Decoupling approximation in the
R. J. (1995). "Seismic stability procedures for solid waste landfills." evaluation of earthquake-induced plastic slip in earth dams." Earth-
J. Geotech. Engrg., ASCE, 121(2), 139-151. quake Engrg. and Struct. Dyn., 11, 667-678.
Bray, J. D., Augello, A. J., Leonards, G. A., Repetto, P. C., and Byrne, Ling, H. I., and Leshchinsky, D. (1997). "Seismic stability and permanent
R J. (1996). "Closure to 'Seismic stability procedures for solid-waste displacement of landfill cover systems." J. Geotech. Engrg., ASCE,
landfills.''' J. Geotech. Engrg., ASCE, 122(11),952-954. 123(2), 113-122.
Bray, J. D., Rathje, E. M., Augello, A. J., and Merry, S. M. (1998). Makdisi, F. I., and Seed, H. B. (1978). "Simplified procedure for esti-
"Simplified seismic design procedure for lined solid-waste landfills." mating dam and embankment earthquake-induced deformations." J.
Geosynthesis Int. J., 4, in press. Geotech. Engrg., ASCE, 104(7), 849-867.
Buranek, D., and Prasad, S. (1991). "Sanitary landfill performance during Matasovic, N. (1993). "Seismic response of composite horizontally-lay-
the Lorna Prieta earthquake." Proc., 2nd Int. Conf. on Recent Adv. in ered soil deposits." Ph.D. dissertation, Dept. of Civ. Engrg., Univ. of
Geotech. Earthquake Engrg. and Soil Dyn., Univ. of Mo. at Rolla, St. California, Los Angeles, Calif.
Louis, Mo., 1655-1660. Matasovic, N., Kavazanjian, E. Jr., and Abourjeily, F. (1995). "Dynamic
Chopra, A. K., and Zhang, L. (1991). "Base sliding response of concrete properties of solid waste from field observations." Proc., lst Int. Conf.
Downloaded from ascelibrary.org by ECOLE POLYTECH/BIBLIOTHEQUE on 02/23/22. Copyright ASCE. For personal use only; all rights reserved.

gravity dams to earthquakes." Rep. No. UCB/EERC-9//05, Earthquake on Earthquake Geotech. Engrg, Tokyo, Japan, 1, 549-555.
Engrg. Res. Ctr., Univ. of California, Berkeley, Calif. Matasovic, N., and Vucetic, M. (1995). "Seismic response of soil deposits
Elton, D. J., Shie, C.-F., and Hadj-Hamou, T. (1991). "One- and two- composed of fully-saturated clay and sand layers." Proc.. 1st Int. Conf.
dimensional analysis of earth dams." Proc., 2nd Int. Conf. on Recent on Earthquake Geotech. Engrg., Tokyo, Japan, 1,611-616.
Adv. in Geotech. Earthquake Engrg. and Soil Dyn., 1043-1049. Newmark, N. M. (1965). "Effects of earthquakes on dams and embank-
Finn, W. D. L. (1988). "Dynamic analysis in geotechnical engineering." ments." Geotechnique, 15(2), 139-160.
Earthquake engineering and soil dynamics II. ASCE Geotech. Spec. Pyke, R, and Beikae, M. (1991). "TNMN." Taga Engineering Software
Publ. No. 20, J. L. Von Thun, ed., ASCE, Reston, Va., 523-591. Services, Lafayette, Calif.
Franklin, A. G., and Chang, F. K. (1977). "Earthquake resistance of earth Rathje, E. M., Abrahamson, N., and Bray, J. D. (1998). "Simplified con-
and rockfill dams." Misc. Paper S-71-17, U.S. Army Wtrwy. Experi- tent estimates of earthquake ground motions." J. Geotech. and Geoen-
ment Station, Vicksburg, Miss. vir. Engrg., ASCE, 124(2), 150-159.
Gazetas, G., and Uddin, N. (1994). "Permanent deformation on preex- Seed, H. B., and Martin, G. R (1966). "The seismic coefficient in earth
isting sliding surfaces in dams." J. Geotech. Engrg., ASCE, 120(11), dam design." J. Soil Mech. and Found. Div., ASCE, 92(3), 25-58.
2041-2061. Seed, H. B. (1979). "Considerations in the earthquake-resistant design of
Harder, L. S. Jr. (1991). "Performance of earth dams during the Lorna earth and rockfill dams." Geotechnique, London, England, 29(3),
Prieta earthquake." Proc., 2nd Int. Conf. on Recent Adv. in Geotech. 215-263.
Earthquake Engrg. and Soil Dyn., Univ. of Mo. at Rolla., St. Louis, Seed, H. B., Wong, R. T., Idriss, I. M., and Tokimatsu, K. (1984). "Mod-
Mo., 1673. uli and damping factors for dynamic analysis of cohesionless soils."
Idriss, I. M. (1991). "Procedures for selecting earthquake ground motions Rep. UCB/EERC-84/14, Earthquake Engrg. Res. Ctr., Univ. of Califor-
at rock sites." Rep. Prepared for Nat. Inst. of Standards and Technol., nia, Berkeley, Calif.
Univ. of California. Seed, R B., and Bonaparte, R. (1992). "Seismic analysis and design of
Idriss, I. M., Fiegel, G. L., Hudson, M. B., Mundy, P. K., and Herzig, R
lined waste fills: Current practice." Proc., ASCE Spec. Conf. on Sta-
(1995). "Seismic response of the operating industries landfill." Earth-
bility and Perf. of Slopes and Embankments, II, ASCE, Reston, Va.,
quake design and performance of solid waste landfills. ASCE Geotech.
1152-1187.
Spec. Publ. No. 54, M. K. Yegian and W. D. L. Finn, eds., ASCE,
Seed, R. B., Dickenson, S., and Mok, C. M. (1991). "Seismic response
Reston, Va., 83-118.
Idriss, I. M., and Sun, J. I. (1992). User's manual for SHAKE9I. Ctr. for analyses of soft and deep cohesive sites: A brief summary of recent
Geotech. Modeling, Dept. of Civ. and Envir. Engrg., Univ. of Califor- lessons." Proc., 1st Annu. CALTRANS Seismic Res. Seminar, CAL-
nia, Davis, Calif. TRANS.
Kavazanjian, E. Jr., and Matasovic, N. (1995). "Seismic analysis of solid Singh, S., and Sun, J. (1995). "Seismic evaluation of municipal solid
waste landfills." Proc., Geoenvironment 2000, ASCE Geotech. Spec. waste landfills." Proc., Geoenvironment 2000, ASCE Geotech. Spec.
Publ. No. 46, ASCE, Reston, Va., 1066-1080. Publ. No. 46, ASCE, Reston, Va., 1081-1096.
Kavazanjian, E. Jr., Matasovic, N., Bonaparte, R, and Schmertmann, G. Trifunac, M. D., and Brady, A. G. (1975). "A study of the duration of
R. (1995). "Evaluation of MSW properties for seismic analysis." strong earthquake ground motion." Bull., Seismological Soc. of Am.,
Proc., Geoenvironment 2000, ASCE Geotech. Spec. Publ. No. 46, 65, 581-626.
ASCE, Reston, Va., 1126-1141. "Solid waste disposal facility criteria." (1993). Technical manual
Kavazanjian, E. Jr., Matasovic, N., Stokoe, K. H., and Bray, J. D. (1996). EPA530-R-93-017, PB94-IOO-450. U.S. Envir. Protection Agency.
"In situ shear wave velocity of solid waste from surface wave mea- Vrymoed, J. L., and Calzascia, E. R (1978). "Simplified determination
surements." Proc., 2nd Int. Congr. on Environ. Geotechnics, 1,97-102. of dynamic stresses in earth dams." Proc., Earthquake Engrg. and Soil
Kramer, S. L., and Smith, M. W. (1997). "Modified Newmark model for Dyn. Conf., ASCE, Reston, Va., 991-1006.
seismic slope displacement of compliant slopes." J. Geotech. and Vucetic, M., and Dobry, R. (1991). "Effect of soil plasticity on cyclic
Geoenvir. Engrg., ASCE, 123(7),635-644. response." J. Geotech. Engrg., ASCE, 117(1), 89-107.
Lee, M. K. W., and Finn, W. D. L. (1978). "DESRA-2, dynamic effective Wright, S. G. (1990). UTEXAS3, a computer program for slope stability
stress response analysis of soil deposits with energy transmitting calculations, Dept. of Civil Engineering of Texas, Austin, Tex.
boundary including assessment of liquefaction potential." Soil Mech. Yegian, M. K., Yee, Z. Y., and Harb, J. N. (1995). "Seismic response of
Ser. No. 36, Dept. of Civ. Engrg., Univ. of British Columbia, Vancou- geosynthetic/soil systems." Proc., Geoenvironment 2000, ASCE Geo-
ver, Canada. tech. Spec. Publ. No. 46, ASCE, Reston, Va., 1113-1125.

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING / MARCH 1998/253

J. Geotech. Geoenviron. Eng., 1998, 124(3): 242-253

You might also like