Proceedings of The Thirteenth (2003) International Offshore and Polar Engineering Conference

Honolulu, Hawaii, USA, May 25 –30, 2003

Copyright © 2003 by The International Society of Offshore and Polar Engineers
ISBN 1 –880653 -60 –5 (Set); ISSN 1098 –6189 (Set)

Damage to Gravity Type Quay Wall Associated with Earthquake Induced

Liquefaction and Its Seismic Design
Takahiko Sasajima
Hokkaido Regional Development Bureau., Hakodate, Japan
Masahiro Kitahara
Civil Engineering Research Institute of Hokkaido, Sapporo, Japan
Hiroaki Ueda
Hokkaido Regional Development Bureau., Sapporo, Japan
Natsuhiko Otsuka
North Japan Port Consultants Co., Ltd., Sapporo, Japan
Kinya Miura
Toyohashi University of Technology, Toyohashi, Japan

ABSTRACT logistical function of ports and harbors. Furthermore, a lot of

This paper aims to clarify the characteristics on the stability
of gravity type quay wall during earthquake by investigating
the damages caused by earthquakes. Close examinations on the
damages and stability factor of quay walls were carried out for
the recent two big earthquakes in Hokkaido Island, i.e.,1993
Kushiro-oki and 1994 Hokkaido Toho-oki earthquakes. As a
result, it is indicated that the occurrence and the degrees of the
damages are strongly related to the occurrence of the
liquefaction in the backfill ground. The damages were large at
the quay walls where liquefaction occurred in the backfill
ground. On the other hand, the quay walls whose backfill
ground had not been liquefied, could have survived the severe
shakes which exceeded their stability limit in earthquakes.
Here the stability limit had computed by seismic coefficient
method in which Mononobe-Okabe’s formula was applied to
seismic earth pressure acting on the quay walls.
From the results mentioned above, the interaction between
caisson and backfill ground might improve the availability of
seismic coefficient method for gravity type quay wall. And it is
considered that there is a possibility that the seismic force
acting on gravity type quay walls whose backfill ground had
been improved against liquefaction could be lessened in the
seismic coefficient method.
costs were needed for post-disaster construction and damage
repair. Therefore, it is very important issue to enhance seismic
resistance of quay walls in Japan.
Close examination on the damages to gravity type quay
walls caused by recent large earthquakes indicated that the
damages are strongly dependent on the occurrence of the
liquefaction of backfill. On the other hand, the quay walls with
remedial treatment of the backfill against liquefaction by sand
compaction pile method and gravel drain method survived
these earthquakes with slight damage (Japanese Geotechnical
Society, 1994; Hokkaido Development Bureau, 1996). In the
structural design of these gravity type quay walls against
earthquakes, the seismic coefficient method is usually applied.
However, this design method does not take account of the
liquefaction of backfill ground. Moreover, it is not clear about
the dynamic behavior of caisson, backfill ground and the
seismic earth pressure in the case of non-liquefaction. In this
study, close examination about the safety factors of gravity
type quay walls that had been stricken by strong earthquake
motion during two big earthquakes that occurred in Hokkaido
Island, “1993 Kushiro-oki” and “1994 Hokkaido Toho-oki”
earthquakes. And the characteristics on the seismic earth
pressure were discussed based on the extent of damage and
seismic forces estimated by the seismic coefficient method.

Keywords: Earthquake, Quay wall, Seismic design, earth SEISMIC COEFFICIENT METHOD AND SEISMIC
pressure, Liquefaction EARTH PRESSURE

INTRODUCTION It shall be standard to use seismic coefficient method for

During the recent big earthquakes in Japan, significant
damages were caused in the port and harbor facilities,
especially in reclaimed lands. For gravity type quay walls, their
function was lost because of subsidence and displacement of
caisson toward sea, and also foundations of structures and
underground structures were damaged due to the lateral flow
deformation of the backfill ground, which was triggered by the
displacement of quay walls. And this functional depression
increased the damage to the regional economy as well as
determining the seismic forces for gravity type quay walls. In
this method, seismic force (inertial force) to caisson is given
by multiplying the seismic coefficient by mass of the caisson.
And in the stability calculation, static equilibration between the
resultant horizontal force acting on the wall, resultant vertical
force and friction force between the bottom of the wall and the
foundation are considered using safety factor of sliding,
overturning and bearing capacity of foundation. The safety
factor against sliding of a gravity type quay wall can be
calculated by equation (1). The value of safety factor F s shall

755
be set appropriately in accordance with the characteristics of k: seismic coefficient
the structure (OCDI, 2002a). γw : unit weight of water
fW H : height of structure below the still water level (m)
Fs ≤ (1) Y : depth from the still water level (m)
P
The earth pressure of sandy soil acting on a gravity type quay
where,
wall during an earthquake can be calculated by the theories
W: resultant vertical force acting on the wall (kN/m)
proposed by Mononobe (1933) and Okabe (1924), as shown in
P : resultant horizontal force acting on the wall (kN/m)
equation (4) and (5).
f : coefficient of friction between the bottom of the wall
Pai = K ai [Σγ i hi + w]
body and the foundation
F s : safety factor (4)
Generally the resultant horizontal force should include the cos (φ i − θ )
2
following (Figure-1): K ai = 2
(5)
(a) Horizontal component of the earth pressure with a  sin(φ i + δ ) sin(φ i − θ ) 
surcharge applied. cos θ cos(δ + θ ) 1 + 
(b) Residual water pressure.  cos(δ + θ ) 
(c) Seismic force acting on the mass of wall body and where
dynamic water pressure during earthquake. θ: composite seismic angle:
θ=tan -1 k (above water level)
RWL θ=tan -1 k’ (below water level)
k: seismic coefficient
(c) (b k’: apparent seismic coefficient applied for the area
(a)
Dynamic below the residual water level
Water However, Mononobe (1924) indicated that since the seismic
pressure
coefficient is calculated from maximum ground acceleration
and this formula was based on the combination of static force,
there could be a possibility of over estimation for the resultant
Figure-1 Horizontal Force Acts on the Gravity Type Quay Wall seismic earth pressure. As a result of close examination for the
damage to gravity type quay walls, Noda (1975) indicated that
In the examination of the bearing capacity of foundation for it might provide over estimation when calculating the design
gravity type quay walls, the circular arc analysis method for seismic coefficient by dividing maximum ground acceleration
eccentric loads can be applied (OCDI, 2002b). by apparent gravity when the maximum ground acceleration is
bigger than 2.0m/sec 2 . Kazama (1993) indicated that the
1 (cb + W ' tan φ ) secα
Fs = Σ (2)
dynamic component of earth pressure during earthquake had
1 1 + (tan α tan φ ) / Fs the tendency to work in the opposite direction to the inertia
ΣW sin α +  ΣHa force of the caisson based on the shaking table test with model
 R quay wall backfilled by dry sand. Miura (1997,1999) indicated
where that if liquefaction does not occur in the backfill, the
F s : safety factor against circular failure according to the fluctuating component of the earth pressure works in opposite
simplified Bishop method to the inertia force, whereas the fluctuating component of the
W: total weight of slice element per unit length (kN/m) earth pressure and inertia force are same in the direction with
α: angle at which the bottom of a slice element intersects no phase angle difference when liquefaction occurs. And the
the horizontal plane dynamic earth pressure with fluctuation from liquefied backfill
R: radius of circular arc slip circle (m) can be evaluated from the formula for the dynamic water
H: horizontal external force acting on soil inside the pressure on dams derived by Westergaard (1933). Thus it is
circular slip surface (kN/m) pointed out that the dynamic interaction between caisson and
a: arm length from the center of circular slip surface to the backfill should be considered on the seismic design of a gravity
acting position of horizontal external force H(m) type quay wall.
c: undrained shear strength, or apparent cohesion under
drained condition in sandy ground (kN/m 2 )
b: width of a slice element DAMAGE TO GRAVITY TYPE QUAY WALLS
W’: effective weight of a slice element per unit length
(kN/m) In 1993 and 1994, strong earthquakes hit Northeastern Coast
φ : internal friction angle under drained condition for sandy of Hokkaido Is.; 1993 Kushiro-oki Earthquake and 1994
ground Hokkaido Toho-oki Earthquake. The maximum ground
The dynamic water pressure during an earthquake can be acceleration of each earthquakes were over 2.0(m/s 2 ) and a lot
calculated by Westergaard’s formula as shown in equation (3). of damages to port and harbor facilities in this region as well
as roads, river banks and lifelines were caused. Figure-1 and
7 Figure-2 show the epicenter and maximum ground surface
Pdw = ± kγ w Hy (3) acceleration of 1993 Kushiro-oki Earthquake and 1994
8 Hokkaido Toho-oki Earthquake respectively. Table 1 shows the
summary of each earthquake.
where, According to the strong motion of the ground, damages to
P dw : dynamic water pressure (kN/m2) the gravity type quay walls had occurred in the Port of Kushiro,

756
Port of Nemuro and Port of Kiritappu. The damages were as Table-2 Damage to gravity type quay walls among earthquake
follows: horizontal settlement toward the sea, forward Earthquake Port Nu δx δy 1 δy 2
inclination of caisson, vertical deformation of caisson,
deformation of backfill and wreck of apron pavement. And in 1993 Kushiro 19 0.0 to 0.05 to 0.1 to
many of these damaged facilities, traces of sand boiling were Kushiro-oki 1.9 0.5 0.6 or
discovered in the backfill ground. Table-2 shows the summary Earthquake more
of damage to the gravity type quay walls. And Figureure-3 and Kiritappu 8 0.1 to 0.05 to 0.1 to
Figureure-4 show the location of damaged facility in the 0.45 0.2 0.35
Kushiro Port and the typical view of damage to a gravity type 1994 Nemuro 23 0.05 to 0.05 to 0.0 to
quay wall respectively. Hokkaido (Hanasaki) 1.65 0.47 0.6
Toho-oki Nemuro 12 0.0 to 0.05 to 0.1 to
Earthquake 0.35 0.3 0.3
Table-1 Details of Earthquakes
Kiritappu 13 0.16 to 0.06 to 0.0 to
Details 0.90 0.39 0.5
1993 Date : Jan.15 1993 Kushiro 5 0.07 to 0.4
Kushiro-oki Epicenter : N42-53,E144-22, Depth 103km 0.32
JMA Magnitude : 7.8 Here, Nu :Number of damaged facility
1994 Date : Oct.4 1994 δx :Horizontal deformation
Hikkaido Epicenter : N43-23,E147-43, Depth 23km δy 1 :Vertical deformation
Toho-oki JMA Magnitude : 8.1 δy 2 :Vertical deformation of backfill
Unit : m

Hanasaki-F
110(N20E) Pacific Ocean
75 (E20S)
43 (UD)

Epicenter
Kushiro-G Kushiro-GB The Port of Kushiro
314(NS) 155(NS)
246 (EW) 185 (EW)
91 (UD) 50 (UD)
(GL-77.0m)
Tokachi-M
209(NS)
260 (EW)
96 (UD) Damaged Facility
Figure-2 Observed Maximum Accelerations of Kushiro-oki
Earthquake (cm/s 2 )
Figure-4 Plane view of damaged facilities at Kushiro Port in
1993 Kishiro-oki Earthquake
Hanasaki-F
227(N20E)
277 (E20S)
142 (UD)

Muroran-G
61 (NS)
67 (EW) before the earthquake
27 (UD) Epicenter after the earthquake
Kushiro-GB Kushiro-GB
163 (NS) 72(NS)
212 (EW) 76 (EW)
68 (UD) 34 (UD)
(GL-77.0m)
Tokachi-M
97(NS)
129 (EW)
50 (UD)

Figure-3 Observed Maximum Accelerations of Hokkaido

Toho-oki Earthquake (cm/s 2 )

Figure-5 Typical case of damaged gravity type quay wall

(the Kushiro Port)

757
EXAMINATION ON THE SAFETY FACTOR AND 0.30
DAMAGE TO GRAVITY TYPE QUAY WALLS Damaged Facilities

Limit Seismic Coefficient of

Limit Seismic Coefficient and Damage to Quay Walls 0.25
The seismic coefficient of each earthquake was defined as

Quay Wall
the operated seismic coefficient “kh op ” for the facilities in
Table-2 individually. The operated seismic coefficient was 0.20
calculated from the maximum acceleration in the normal
direction of the facility that was calculated from the observed
maximum ground surface acceleration of NS and EW 0.15
component. On the other hand, the limit seismic coefficient for
sliding of caisson and bearing capacity of foundation at a
condition of which safety factor equal to 1.0 were calculated. 0.10
In the safety factor calculation, surcharge load was neglected 0.10 0.15 0.20 0.25 0.30 0.35
according to the actual condition of the facility; and the
observed sea water level was applied. And the limit seismic Operated seismic Coefficient(α max/g)
coefficient “kh lim ” was defined by choosing smaller one of
these two limit seismic coefficients. ;liquefied, 10 to 30cm
;liquefied, over 30cm
Figure-6 compares the operated seismic coefficient and limit
;non-liquefied,10 to 30cm
seismic coefficient of each gravity type quay wall of which the ;non-liquefied, over 30cm
horizontal deformation to the seaside was more than 10cm. In ;unconfirmed, 10 to 30cm
this Figure, painted plots show the facilities of which the marks ;unconfirmed, over 30cm
of liquefaction were found on the backfill ground. Figure-7
compares the kh op and kh lim of each gravity type quay wall of Figure-6 Limit Seismic Coefficient and Damage of Quaywalls
which the horizontal deformation to the seaside was less than During Earthquake(Non-Liquefaction)
10cm. In this group of facilities, there was no liquefied mark in
the backfill. The white painted plots show the facilities with 0.3 no damage
the remedial treatment of the backfill against liquefaction. As

Limit Seismic Coefficient of

can be seen in these two figures, it is very clear that damages
are strongly dependent on the occurrence of liquefaction of the 0.25
backfill. And in Figure-7, there were many facilities whose
Quay Wall
limit seismic coefficient were smaller than the operated ones
which indicates the expectation of damage to occur. Table-3 0.2
compares the horizontal deformation of gravity type quay walls
with the remedial treatment of the backfill against liquefaction
and without it in the Kushiro Port caused by 1993 Kushiro-oki 0.15
Earthquake. Each facility with the same depth had the same
structure and located the same direction. It is clear that the
damages of which treated against liquefaction facilities were 0.1
obviously small. 0.1 0.2 0.3
Operated Seismic Coefficient(α m ax/g)
Seismic Earth Pressure
If the liquefaction would not occur in the backfill ground, Horizontal Displacement of the caisson
forces acting on the caisson during earthquake should consist treated ; 0cm
of the residual water pressure, seismic force acting on the mass treated ; 0 to 10cm
un-treated ; 0cm
of wall body (inertia force), dynamic water pressure during
un-treated ; 0 to 10cm
earthquake and seismic earth pressure. In general, the inertia
force would be the main force among these forces when strong
acceleration would act to a quay wall. And it is experimentally Figure-7 Limit Seismic Coefficient and Damage of Quay Walls
observed that the dynamic water pressure tends to act on the During Earthquake (Non-Liquefaction)
quay wall in the same direction as the inertia force (Miura,
1997). And the earth pressure would not always take maximum Table-3 Effectiveness of the Remedial Treatment of Backfill
value statically because it is considered to be fluctuating Against Liquefaction
dynamically during earthquake. Therefore, the displacement of Facility Maximum Average
the caisson might occur when the maximum inertia force and deformation deformation
dynamic water pressure according to the maximum acceleration -7.5m Non-treated 25cm 13cm
would arise, and the seismic earth pressure would act to the Treated 15cm 5cm
caisson additionally. -10m Non-treated 5cm 3cm
Treated 0cm 0cm
-12m Non-treated 26cm 23cm
Treated 6cm 1.6cm

758
 In order to confirm the earth pressure when the caisson
Seismic Active Earth Pressure
yields to move during the earthquake, the resultant horizontal
at kh op by Eq.(4)
force at the limit seismic coefficient are compared to the sum Un-Stable
of the inertia force, dynamic water pressure and residual water Active Earth Pressure
pressure at the operated seismic coefficient (Figure-8). The on a steady basis
difference between these two resultant force “H’ dif ” can be
thought to corresponds to the resultant effective horizontal
earth pressure when the maximum inertia force acts to the
H diff
caisson of which backfill did not liquefy. In Figure-9, this
differential factor H’ dif was normalized by the resultant
effective active earth pressure on a steady basis and plotted in Stable the Inertia Force
proportion to the extent of damage. In Figure-9, H’ dif was + Dynamic Water Pressure
normalized by the resultant effective active earth pressure + Residual Water Pressure
during earthquake calculated by equation (4) using the operated at kh op
seismic coefficient kh op . These Figures indicate that no damage
can be seen when H’ dif is bigger than resultant active earth
pressure, and bigger than half the value of effective seismic Resultant Horizontal Force at the
earth pressure calculated by equation (4). Limit Seismic Coefficient
According to this, it can be considered that the maximum as the Stability Limit of Caisson
resultant earth pressure would not occur at the same time as the
maximum inertia force when the backfill does not liquefy. And
the damage to the gravity type quay wall would be occurred Figure-8 Schematic of Examining the Earth Pressure During
when the maximum resultant horizontal force would exceed the Earthquake in Figure-9 and 10
sum of maximum inertia force, dynamic water pressure,
residual water pressure and 50% of seismic active earth Displacement
pressure. 3 :no damage
:less than 10cm

Hdiff / Resultant active earth

CONCLUSION :10cm to 30cm
2 :over 30cm
The characteristics on the stability of gravity type quay wall pressure
during earthquake were investigated. The following 1
conclusions were obtained from close examinations on the
damages and stability factor of quay walls for the recent two
big earthquakes in Hokkaido Island. 0

(1) The damages to the gravity type quay walls were strongly
dependent on the occurrence of liquefaction of the backfill. -1
And a lot of quay walls whose backfill ground had not 0.15 0.2 0.25 0.3 0.35
been liquefied, could have survived the severe shakes
which exceeded their stability limit in earthquakes. Ο perated seismic coefficient (α max / g)
(2) In order to confirm the earth pressure when the caisson
yields to move during the earthquake, the resultant Figure-9 Comparison Between the Active Earth Pressure on a
horizontal force at the limit seismic coefficient are Steady Basis and H’ dif
compared to the sum of the inertia force, dynamic water
pressure and residual water pressure at the operated
seismic coefficient. As a result, no damage can be seen Displacement
Hdiff /Resultant active seismic earth

1.5 :no damage

when the difference between these two factors as ”H’ dif “ is
:less than 10cm
bigger than the resultant effective earth pressure on a 1 :10cm to 30cm
steady basis, and bigger than half the value of effective :over 30cm
seismic earth pressure calculated by Mononobe/Okabe’s
0.5
pressure

formula.
(3) It can be considered that the maximum resultant earth
pressure would not occur at the same time as the maximum 0
inertia force when the backfill ground did not liquefy. And
the damage to the gravity type quay wall would be -0.5
occurred when the maximum resultant horizontal force
would exceed the sum of maximum inertia force, dynamic -1
water pressure, residual water pressure and 50% of seismic 0.15 0.2 0.25 0.3 0.35
active earth pressure.
Ο perated seismic coefficient (α m ax / g)

Figure-10 Comparison Between the Seismic Active Earth

Pressure and H’ dif

759
 pp.512-519
REFERENCES Mononobe, N. (1933), “Seismic Design on Civil Engineering”,
Tokiwa Syobou, (in Japanese)
Hokkaido Development Bureau (1996). “Report on the Mononobe, N. (1924) “The Effects of Vertical Motion During
Improvement of the Seismic Resistance of Port Facilities.”, Earthquake”, Journal of Japan Society of Civil Engineers,
Cold Region Port and Harbor Research Center. (in Japanese) Vol.10, No.5, pp.1063-1094, (in Japanese)
Japanese Geotechnical Society (1994). “Report of the Damages Noda, S., Uwabe, T., Chiba, T. (1975) ”Relation between
during 1993 Kushiro-oki and Notohanto-oki Earthquakes.” seismic coefficient and ground acceleration for gravity
pp.175-182. (in Japanese) quaywall”, Rept. of PHRI, Vol. 14, No. 4, (in Japanese).
Kazama, M. (1993), “Seismic Stability of Gravity-type Harbor Okabe, S. (1924) “General Theory on Earth Pressure and
Structures Subjected to Earth Pressure”, TECHNICAL NOTE Seismic Stability of Retaining Wall and Dam”, Journal of
OF THE PORT AND HARBOUR RESEARCH INSTITUTE, Japan Society of Civil Engineers, Vol.10, No.6, pp.1277-1323
MINISTRY OF TRANSPORT JAPAN, No.752, (in Japanese) The Overseas Coastal Area Development Institute of Japan
Miura, K., Kohama, E., Kurita, S., Ohtsuka, N., Yoshida, N. (2002a), “TECHNICAL STANDARDS AND
(1997) “Behavior of Gravity Type Quay Wall during COMMENTARIES FOR PORT AND HARBOUR
Earthquake Observed in Model Shaking Table Test”, The FACILITIES IN JAPAN” pp.408-411
Proceedings of The 7th International OFFSHORE AND The Overseas Coastal Area Development Institute of Japan
POLAR ENGINEERING CONFERENCE, pp.683-688 (2002b), “TECHNICAL STANDARDS AND
Miura, K., Kohama, E., Inoue, K., Ohtsuka, N., Sasajima, T., COMMENTARIES FOR PORT AND HARBOUR
Yoshida, N. (1999), “Instability of Gravity Type Quay Wall FACILITIES IN JAPAN”, pp.277
during Earthquake with Regard to Dynamic Interaction with Westergaard, H. M. (1933). “Water Pressures on Dams during
Backfill Ground”, The Proceedings of The 9th International Earthquakes”, Trans. ASCE, Vol.98, pp.418-432.
OFFSHORE AND POLAR ENGINEERING CONFERENCE,

760