Professional Documents
Culture Documents
Earthquake-Resistant Design of Earth Dams
Earthquake-Resistant Design of Earth Dams
Scholars' Mine
International Conferences on Recent Advances 1981 - First International Conference on Recent
in Geotechnical Earthquake Engineering and Advances in Geotechnical Earthquake
Soil Dynamics Engineering & Soil Dynamics
Recommended Citation
Seed, H. Bolton, "Earthquake-Resistant Design of Earth Dams" (1981). International Conferences on
Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics. 23.
https://scholarsmine.mst.edu/icrageesd/01icrageesd/session07/23
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.
This Article - Conference proceedings is brought to you for free and open access by Scholars' Mine. It has been
accepted for inclusion in International Conferences on Recent Advances in Geotechnical Earthquake Engineering
and Soil Dynamics by an authorized administrator of Scholars' Mine. This work is protected by U. S. Copyright Law.
Unauthorized use including reproduction for redistribution requires the permission of the copyright holder. For more
information, please contact scholarsmine@mst.edu.
Earthquake-Resistant Design of Earth Dams
H. Bolton Seed, Professor of Civil Engineering
University of California, Berkeley CA
SYNOPSIS Lessons gained from observations of the field performance of earth darns during earthquakes
are reviewed and used to illustrate the primary problems of concern. Defensive design measures which
may be taken to mitigate the various hazards are reviewed and illustrated. Analytical approaches for
evaluating seismic stability and the deformations of earth darns during earthquakes are discussed,
together with recent developments which facilitate their implementation in special cases; situations
which require careful consideration of special effects such as the three-dimensionality of the darn-
valley system and pore pressure re-distribution following an earthquake are discussed and illustrated.
1157
1158
Fig . 1. View of Lower San Fernando Dam after Upstream Slope Slide in
Earthquake of Feb . 9, 1971
five miles of the fault) on which a magnitude structures the sand was apparently not saturated.
8-1/4 earthquake occurred, so there can be little It is reasonable to conclude therefore that dams
doubt that all of these dams were subjected to built of clayey materials seem to exhibit high
strong shaking for a prolonged period of time resistance against slope failures during earth-
(over 1 min). Based on recent correlations of quakes--but the field data provide no informa-
ground motions with distance in California tion on the possible behavior of saturated
earthquakes, it seems reasonably sure that all sands .
of these dams were subjected to ground motions
having peak accelerations greater than 0.2Sg and This important conclusion is reinforced by the
those within five miles of the fault to motions study by Akiba and Semba (1941) of the perfor-
with peak accelerations greater than about 0.6g. mance of dams in the 1939 Ojika earthquake in
Yet significantly none of these old dams suf- Japan. As a result of this earthquake 12 cases
fered any significant damage and there was cer- of complete dam failures occurred,together with
tainly no evidence of slope instability. It is about 40 cases of reported slope failures. The
not possible to attribute this to the use of main conclusions of this study were:
flat slopes for the dams (slopes varied typi-
cally from 1 on 2 to 1 on 3) or to the high (a) There were very few cases of dam failures
quality of construction (most of the dams were during the earthquake shaking, most of the
not compacted with rollers but by moving live- failures occurring either a few hours or
stock or by teams and wagons) . However, a up to 24 hours after the earthquake.
significant characteristic of all of the dams is
that they were constructed of clayey soils on (b) The majority of the damaged and failed
rock or clayey soil foundations. Only two of embankments consisted of sandy soils and
the dams were built largely of sand and for these no complete failures occurred in embank-
1159
It is clear that these observations tend to con- (c) Dams constructed of clay soils on clay or
firm those concerning dams built of clay soils rock foundations have withstood extremely
in the 1906 San Francisco earthquake. In strong shaking ranging from 0.35 to 0.8g
addition,the Ojika earthquake experience pro- from a magnitude 8-1/4 earthquake with no
vides clear evidence of the vastly superior sta- apparent damage.
bility of embankments constructed of clay soils
under strong seismic loading conditions over (d) Two rockfill dams have withstood moderately
those constructed of saturated sands--a fact strong shaking with no significant damage
fully in accord with Terzaghi's insightful con- and if the rockfill is kept dry by means of
siderations on this question. Moreover, the a concrete facing, such dams should be able
performance record clearly suggests that the to withstand extremely strong shaking with
critical period for an embankment dam subjected only small deformations.
to earthquake shaking is not only the period of
shaking itself, but also a period of hours (e) Dams which have suffered complete failure
following an earthquake, possibly because piping or slope failures as a result of earthquake
may occur through cracks induced by the earth- shaking seem to have been constructed pri-
quake motions or because slope failures may marily with saturated sand shells or on
result from pore pressure redistribution. saturated sand foundations and these types
of dams require careful attention to ensure
More recently, important earthquakes in which their seismic safety.
dam performance was observed include: the 1968
Tokachi-Oki earthquake in which a large number (f) Since there is ample field evidence that
of slope failures occurred at shaking levels of well-built dams can withstand moderate
the order of 0.2g in dams constructed of loose shaking with peak accelerations up to at
volcanic sand (Moriya, 1974) and the 1971 least 0.2g with no harmful effects, we
San Fernando earthquake in which slope failures should not waste our time and money
occurred in the Upper and Lower San Fernando analyzing this type of problem--rather we
Dams, both having sand shells (Seed et al., l975b), should concentrate our efforts on those
while generally excellent performance was dams likely to present problems either
observed in 25 rolled earth fill dams at shaking because of strong shaking involving
levels between 0.2 and 0.4g (Seed et al., 1978). accelerations well in excess of 0.2g or
The latter event, together with the performance because they incorporate large bodies of
of hydraulic fill dams in Russia (Ambraseys, cohesionless materials (usually sands)
1960), also showed that even dams constructed which, if saturated, may lose most of their
with hydraulically deposited sand shells can strength during earthquake shaking and
withstand levels of shaking up to about 0.2g thereby lead to undesirable movements.
from magnitude 6-1/2 earthquakes without detri-
mental effects. (g) For dams constructed of saturated cohesion-
less soils and subjected to strong shaking,
The slide movements in the San Fernando Dams a primary cause of damage or failure is the
were of special importance since in both cases, build-up of pore water pressures in the
field observations showed that the shaking embankment and the possible loss of
induced by the earthquake caused a dramatic in- strength which may accrue as a result of
crease in pore-water pressure in the shells of these pore pressures. Methods of stability
the dams and, in the case of the Lower analysis which do not take these pore pres-
San Fernando Dam, a condition of liquefaction sure increases and associated loss of
which led to a major slide resembling a flow strength into account are not likely to
slide, in the upstream shell. In both cases the provide a reliable basis for evaluating
slide movements were apparently associated with field performance.
a loss of strength associated with these pore
water pressure increases.
DEFENSIVE DESIGN MEASURES
The general conclusions (Seed et al., 1978) which
seem to follow from a close study of embankment It may be noted that most of the potential prob-
dam performance during earthquakes are as follows lems which may develop as a result of earth-
quake action do not require analytical treatment
(a) Hydraulic fill dams have been found to be but simply the application of commonsense defen-
vulnerable to failures under unfavorable sive measures to prevent deleterious effects.
conditions and one of the particularly un- Thus to prevent a dam being disrupted by a fault
favorable conditions would be expected to movement in the foundation may simply require
be the shaking produced by strong earth- the identification of potentially active faults
quake,s. However, many hydraulic fill dams and the selection of a site where such faults
have performed well for many years and when do not exist. Similarly the potential for
they are built with reasonable slopes on settlement, slumping,or tectonic movements, all
good foundations they can apparently sur- of which could lead to loss of freeboard, can be
vive moderately strong shaking--with ameliorated by the provision of additional
1160
freeboard so that the loss of some portion would from active faults. The occurrence of the
not have serious consequences. In short, many Oroville earthquake in 1975 (State of California
of the potentially harmful effects of earth- Department of Water Resources, 1977) , some 60
quakes on earth and rockfill dams can be elimi- miles to the north, led to a re-evaluation of
nated by adopting defensive measures which this situation and the determination that poten-
render the effects non-harmful. A list of such tially active faults exist very close to or
defensive measures would include the following possibly even across the proposed dam-site.
(Seed, 1979): Clearly this has led to a re-evaluation of the
desirability of constructing a thin concrete
(a) Allow ample freeboard to allow for settle- arch dam at this location. Such occurrences
ment, slumping or fault movements. point up the need for prudence in evaluating
the potential seismicity of any dam site.
(b) Use wide transition zones of material not
vulnerable to cracking. Defensive measures, especially the use of wide
filters and transition zones, provide a major
(c) Use chimney drains near the central portion contribution to earthquake-resistant design and
of the embankment. should be the first consideration by the pru-
dent engineer in arriving at a solution to
(d) Provide ample drainage zones to allow for problems posed by the possibility of earthquake
possible flow of water through cracks. effects.
(e) Use wide core zones of plastic materials At the same time it is necessary to recognize
not vulnerable to cracking. that all reasonable steps should be taken to
ensure that sliding such as that which occurred
(f) Use a well-graded filter zone upstream of at the Lower San Fernando Dam, the Sheffield
the core to serve as a crack-stopper. Dam or a number of dams in Japan does not in-
validate the beneficial effects of defensive
(g) Provide crest details which will prevent measures; and many members of the public,
erosion in the event of overtopping. regulatory agencies and even leading dam
engineers will readily acknowledge that rela-
(h) Flare the embankment core at abutment con- tively sophisticated analyses are warranted in
tacts. many cases to provide guidance on the possibi-
lity of slide movements developing during
(i) Locate the core to minimize the degree of earthquakes and their possible extent. Accord-
saturation of materials. ingly it is of interest to review the approaches
taken to check on this possibility and evaluate
(j) Stabilize slopes around the reservoir rim their significance in the design procedure--if
to prevent slides into the reservoir. for no other reason than to provide some in-
sight into their effects in producing safer
(k) Provide special details if danger of fault designs. Early attempts in this direction are
movement in foundation. discussed in the following pages.
to deformation as a result of earthquake or simu- The Newmark (1965) type of analysis leading to
lated earthquake loading and thus pseudo-static evaluatio ns of slope displacements by double
analyses will generally provide an acceptable i ntegration of that portion of the induced
method of ensuring adequate performance for acceleration response of a potential slide mass
embankments constructed of these types of soil . e xceeding the computed yield acceleration for
In cases of doubt , however , a careful laboratory the mass (see Fig. 4), represented a major step
study will invariably provide the information forward in both the philosophy of evaluat i ng
from which an appropriate engineering decision embankment performance duri ng earthquakes and in
concerning the applicability of the method can suggesting the means for it s implementation. It
be made . It should also be noted that even some is an interesting by- product that the use of this
soils which might be vulnerable to the develop- improved method based on the prediction of
ment of large pore pressures and some strength embankment displacements should ultimately lead
loss under conditions of strong shaking may show to the conclusion that for some types of soils
little evidence of these effects under less in- and conditions (i . e. , those which do not build
tense shaking , in which case the principles dis- up large pore pressures or cause significant
cussed a b ove would still be applicable. strength loss due to earthquake shaking and
1163
safety against failure of the embankment indicates that for these conditions a slope
either during or following the earthquake. failure would develop.
(h) If the embankment is found to be safe An example of a seismic stability analysis where
against failure, use the strains induced by failure did not occur is illustrated by the com-
the combined effects of static and dynamic puted response of the Upper San Fernando Dam in
loads to assess the overall deformations of the same earthquake. Again extensive zones of
the embankment. high pore-water pressure were developed within
the embankment as a result of the earthquake
(i) Be sure to incorporate the requisite amount shaking but they were not sufficiently extensive
of judgment in each of steps (a) to (h) as to cause a failure, as illustrated in Figs. 7
well as in the final assessment of probable and 8. In this case the embankment suffered
performance, being guided by a thorough significant deformations, the crest moving down-
knowledge of typical soil characteristics, stream about 5 ft. Deformation analyses based
the essential details of finite element on the method described above predicted a move-
analysis procedures, and a detailed know- ment of about 3.8 ft.
ledge of the past performance of embank-
ments in other earthquakes. As with all analytical procedures used in geo-
technical engineering, the method should only be
This procedure may seem rather long and cumber- used if it is found to work--that is, if it
some but it also seems to incorporate the provides reasonable evaluations of behavior for
essential steps in evaluating such a complex cases where the behavior has been or can be
problem as the response of earth dams to earth- observed. In all, the general procedure des-
quake effects. cribed above has been used to study the per-
formance of eight dams whose performance during
It lends itself naturally, however, to somewhat earthquakes is known. Two of these had major
simplified versions of the method, which have slides, one underwent large deformations, one
often been used for reason of time and economy underwent small deformations and four had no
(e.g., Finn, 1967; Klohn et al., 1978; Lee and discernible damage. The behavior predicted by
Walters, 1972; Lee, 1978; Leps et al., l978a and the analysis was similar to that observed in
b; Vrymoed and Galzacia, 1978; etc.). The ulti- the field in each case, and while it is true
mate simplification is, of course, the total that each of these cases was in fact studied
elimination of all analysis procedures and a after the event involved, it seems that the
simple evaluation, based on a knowledge of the procedure has the c~pability of giving con-
materials comprising the dam and the judgment siderable insight into the possible behavior of
resulting from conducting many previous analyses embankments subjected to earthquake effects.
and observing the performance of existing dams. For this reason, presumably, it has been adopted
However, it should be noted that each of the in studies of many dams throughout the world as
steps is an essential element of the procedure a guide to final assessment of their probable
and if one of them is performed incorrectly, the performance during earthquakes (e.g., Seed etal.,
results of the analysis may be grossly mislead- 1969, 1973, 1975b; Gordon et al., 1974; Kramer
ing. In such cases, where the job cannot be et al., 1975; Marcuson et al., 1977; Makdisi
done properly, it may be better not to do it at et al., 1978; Sadigh et al., 1978; State of
all rather than to be misled by the erroneous California Department of Water Resources, 1979;
results which may ensue. It is for this reason etc.).
that judgment is necessary at each step in the
development. If the permeability of the shell material for a
dam becomes sufficiently high, say of the order
In the most modern versions of the method, the of the 1 cm/s, then it may be impossible for an
assessment of pore water pressures during and earthquake to cause any build-up of pore pres-
following the earthquake shaking may involve sures in the embankment since the pore pressures
studies of simultaneous pore pressure generation can dissipate by drainage as rapidly as the
and dissipation using appropriate computer earthquake can generate them by shaking. A good
programs (Booker et al., 1976; Finn et al., 1978) example of such a situation is the upstream shell
and the evaluation of the final configuration of of Dartmouth Dam (Seed, 1974). This 650 ft high
the structure using a strain-harmonizing tech- rock fill structure (see cross-section in Fig. 9)
nique, again involving finite element procedures has highly pervious shells having a 10% size of
(Lee et al., 1974; Serff et al., 1976) . about 2.5 em and a permeability coefficient
possibly of the order of 100 cm/s. In such a
The particular procedure used in any given case case analysis shows that the pore pressure build-
should depend on the complexity of the case being up in 10 s of earthquake shaking would be a neg-
considered, the margin of safety provided for the ligible proportion of the initial effective
level of earthquake shaking likely to develop, overburden pressure (Seed, 1979) even for an
and the judgment and experience of the engineer earthquake which might hypothetically be con-
responsible for the study. sidered to produce shaking of sufficient inten-
sity to generate a pore pressure ratio of 100%
An interesting example of the use of this method throughout the shell if it were undrained.
to analyze a slope failure is provided by the Clearly in a case such as this, the upstream
analysis of the Lower San Fernando Dam (Seed, shell can for practical purposes be considered
1979). The computed response of this dam to the to be fully drained during any reasonable period
earthquake' ground motions is shown in Fig. 5 with of earthquake shaking and stability can be
the dark area indicating the zones where the evaluated on this basis.
residual pore water pressure at the conclusion
of the earthquake was equal to 100%. A stability A similar conclusion has been drawn in the eval-
analysis of this section, shown in Fig. 6, clearly uation of the seismic stability of the Watauga
1165
06,-------------------~-------------------.-------------------.
Ttl'l'le- sec.one11
1160
--+--
06,--------------------.---------------------.--------------------~
- zonu of failure dut to ilq~efoctton indtcoted by analysts
ofttr 10 5 seconds of shok1ng
~ !f,':',;.'~'~~~~~d~u~,'~~!~~~!octton Indicated by analysiS
~ !ft:'t~'.!~~~~: ~f~k:~~uefaction tndicoted by anolys••
o•r---------------------~-----------r------~----~------------------~
060~--------------------~------------------L_~I~0----------------------115
(otter R F Stolt)
Fig. 5. Analysis of Response of Lower Dam During San Fernando Earthquake to Base Motions Determined
from Seismoscope Record
......
I
c:
0
!!!.
l.tJ
100
1/X\V?A<Vh<'V>.X' V4 )>X<WX<V>X'
Fig. 6 • Stability of Lower San Fernando Dam a Short Time after Earthquake Motions Stop
1166
1120
110
1000 or>-.---------.------'----~
-r- #JX41NJEJIW fiNIJqoJIG'>zq;;c
-j
0~0-- -------------------
I
07!'>0- -- -----------------t---- r,,..._Meoncn
,'o ,,
Tune· Ho!.!ory of AcceterOioOn n Bow Rock
Fig. 7. Analysis of Response of Upper San Fernando Dam During San Fernando Earthquake
r-
' -A••S ol Darn
05r---------r---------.---------,---------. 0.5r----------r----------.----------r----------,
£ Limtts set by Chinese Code (1974)
i e L tquefoction I
~ I
: I
0 No liquefoctton I
~ I
I '<:/'1
:S,I
I
I
'<:f'l
"'" 1 '{::'I
I I
n.tf
0
.Q
g. 04
-'
I
I
I
I
ql
<nj <Df <D/ ; :;:
Ql
... ...1
It/ II/
,._II -<:
../!II
I
Q:)l
.:;s:_l
e 0
ee e e 0 0
e .-_ ..~.___ N~tgato
0
e e 0 0 0
e ?
.., eo eo o
- !E
0
0
a:: -t 0
oo
0
0
0
0
.."'"'
~
_J
Lake Amotttlan u _J 0.1
~
u
,..
u
u
0 0
0 20 30 40 0 10 20 30 40
Modified Penetration Resistance, N1 - blows/It Madtfted Penetration Resistance, N -blows/ft.
1
Fig. 10. Correlation Between Field Liquefaction Fig. 11. Design Curves for Evaluating Field
·Behavior of Sands for Level Ground Liquefaction Resistance of Sands Under
Conditions and Penetration Resistance Level Ground from Standard Penetration
(Earthquakes with Magnitude ~ 7-1/2) Test Data
1168
TABLE 3. Performance During Earthquakes of Five Darns with Sandy Shells and Central Clay Cores
(see material characteristics in Fig. 17)
Darn (K )
2 max
* Field Performance
..~ ~
(. .6 1.0
'•i
_t:
Ill
Q)
t..:> L
31::J .5 de ~:II
-~ Q)
::J
.8
a C(: "'t'fc
0( =0.3
-J:
.9
'tl .4
c1,c
o<:Q2
J131
~£
.6
.~ o;§
::J
0' o< =0.1 .,o
.,c
::Ju .4
&3 iii+'
>~
~ 0( :0. GIL
>II
....L.
Ill .2
:;:.;t .2
ruo
L.
~ &iii
s::.
Ill
.o
.1 0. 1. 2. 3. 4. 5. 6. 7. 8.
.!:!
Effective Overburden Pressure(o;,' )- Kg/cm2
~
u .0
0. 1. 2. 3. 4. 5. 6. 7. B.
Initial Effective Vertical Stress <tc- Kg/cm2 Fig. 15. Typical Reduction in Cyclic Stress
Ratio Causing Liquefaction with
Increase in Initial Confining Pressure
Fig. 14. Typical Chart for Evaluating Effect of 3. From the value of N1 thus determined, read
Initial Cyclic Stresses on Cyclic off the corresponding value of the cyclic
Loading Resistance of Sands stress ratio required to cause a pore
1170
pressure ratio of 100% for cases where a= 0 2. Use of Comparative Procedures to Evaluate
from charts such as that shown in Fig. 11. Seismic Stability
4. Determine the pre-earthquake static stresses With the increasing number of dams for which (1)
Th and a ' at the point under consideration. properties and performance under e~rthquakes of 1
He~ce det~rmine the value of a = Thv;ov' for known intensities have been establ1shed and (2) ,
the point. properties and evaluations of seismic stability \
have been made, it is becoming possible to
5. From Fig. 14, determine the value of Ka cor- evaluate the probable performance of other dams
responding to the known value of a and hence by simple comparison with available performance
determine the value of or evaluation data for embankments constructed
under similar conditions and with similar con-
figurations. Thus no detailed analysis is
required--simply a good file of past performance
and analytical evaluation data.
263.
LEGEND
e C'I'CUC SHEAR STRENGTH
AT SPf.CifiC LOCA.TION Of
EACH DRILL HOLE
-e.o. 60 120. 180 240. lOO. 360. 4>0.
DI5TA.NCE FROM "Xi5. FEET
.. eDH-9
.OH-7
eDH-12
OH-<l
• I
.~
'"••4.14 K5F
DH-10 QrHO
. -·
•
• • ."*"'
DH-:H
DH-33
•
Fig. 16. Analysis of Stability of Foundation Sands at Camanche Dam Using Standard Penetration Test
Data
Fig. 17. Cyclic Loading Resistance of Shell Fig. 18. Cyclic Loading Characteristics of Shell
'Materials for Selected Dams (see Materials for Several Dams Shown by
Table 3) Analyses to be Capable of Withstanding
Strong Earthquake Shaking
1172
Leps, T. M., A. G. Strassburger and R. L. Meehan Seed, H. B., K. L. Lee, I. M. Idriss and F. I.
(1978b), "Seismic Stability of Hydraulic Fill Makdisi (1973), "Analysis of Slides in the
Dams--Part 2," Water Power and Dam Construc- San Fernando Dams during the Earthquake of
tion, October, pp. 43-58. February 9, 1971," Report No. EERC 73-2,
Earthquake Engineering Research Center,
Makdisi, F. I., H. B. Seed and I. M. Idriss
(1978), "Analysis of Chabot Dam During the Seed, H. B., K. L. Lee, I. M. Idriss and F. I.
1906 Earthquake," Proc., ASCE Geotech. Engng. Makdisi (1975c), "The Slides in the
Div. Specialty Conf. on Earthquake Engineer- San Fernando Dams during the Earthquake of
ing and Soil Dynamics, Pasadena, pp. 569-587. February 9, 1971," J. Geotech. Engng. Div.,
ASCE, Vol. 101, No. GT7, pp. 651-668.
Marcuson, W. F. III, R. F. Ballard, Jr. and
R. H. Ledbetter (1979), "Liquefaction Failure Seed, H. B., F. I. Makdisi and P. DeAlba (1978),
of Tailings Dams resulting from the Near Izu "Performance of Earth Dams during Earth-
Oshima Earthquake, 14 and 15 January, 1978," quakes," J. Geotech. Engng. Div., Vol. 104,
Proc., 6th Panamerican Conf. on Soil Mech- No. GT7.
anics and Foundation Engineering, Lima, Peru.
Serff, N., H. B. Seed, F. I. Makdisi and C. K.
Marcuson, W. F. III, E. L. Krinitzky and E. R. Chang (1976), "Earthquake Induced Deforma-
Kovanic (1977), "Earthquake Analysis of tions of Earth Dams," Report No. EERC 76-4,
Fort Peck Dam, Montana," Proc., 9th Int. Conf. Earthquake Engineering Research Center,
on Soil Mech. and Found. Engrg., Tokyo, University of California, Berkeley.
Japan.
Sherard, J. L. (1967), "Earthquake Consideration
Moriya, M. (1974), "Damages Caused to Small in Earth Dam Design," J. Soil Mech. Div.,
Earthfill Dams for Irrigation in Armori ASCE, Vol. 93, No. SM4, pp. 377-401.
Prefecture by Tokachi Offshore Earthquake,"
Report prepared for the Committee on Lessons State of California Department of Water Resources
from Incidents of Dam Failures, International (1979), "The August 1, 1975 Oroville Earth-
Commission on Large Dams, 1974. quake Investigations," Dept. of Water
Resources Bulletin No. 203-278.
Newmark, N. M. (1965), "Effects of Earthquakes
on Dams and Embankments," Geotechnique, Terzaghi, K. (1950), "Mechanisms of Landslides,"
Vol. 15, No. 2, pp. 139-160. The Geological Survey of America, Engineer-
ing Geology (Berkey) Volume.
Sadigh, K., I. M. Idriss and R. R. Youngs (1978),
"Drainage Effects on Seismic Stability of Vrymoed, J. L. and E. R. Galzascia (1978),
Rockfill Dams," Proc. ASCE Geotech. Engng. "Simplified Determination of Dynamic Stresses
Div. Specialty Conf. on Earthquake Engineer- in Earth Dams," Proc. ASCE Geotech. Engng.
ing and Soil Dynamics, Pasadena, pp. 802-818. Div. Specialty Conf. on Earthquake Engineer-
ing and Soil Dynamics, Pasadena, pp. 991-
Seed, H. B. (1966), "A Method for Earthquake 1006.
Resistant Design of Earth Dams," J. Soil
Mech. Fdn. Div., ASCE, Vol. 92, No. SMl, Wahler and Associates (1981) "Seismic Reevalua-
pp. 13-41. tion of Camanche Reservoir Main Dam," Report
prepared for the East Bay Municipal Utility
Seed, H. B. (1979), "Considerations in the District, Oakland, CA, January.
Earthquake-Resistant Design of Earth and Rock-
fill Dams, Geotechnique, Vol. XXIX, No. 3,
Sept.