LIQUEFACTION-INDUCED DAMAGE TO

STRUCTURES DURING THE 2011 GREAT EAST JAPAN EARTHQUAKE

Susumu YASUDA1, Ikuo TOWHATA2, Ichiro ISHII3, Shigeru SATO4 and Taro UCHIMURA2

1Member of JSCE, Professor, Dept. of Civil and Environmental Eng., Tokyo Denki University

(Hatoyama, Hiki-gun, Saitama 350-0394, Japan) E-mail: yasuda@g.dendai.ac.jp

2Member of JSCE, Dept. of Civil Eng., University of Tokyo (7-3-1, Hongo, Bunkyo-ku, Tokyo 113-8656, Japan) E-mail: towhata@geot.t.u-tokyo.ac.jp, UCHIMURA@civil.t.u-tokyo.ac.jp

3Member of JSCE, Deputy Mayor, Urayasu City (1-1-1, Nekozane, Urayasu, Chiba 279-8501, Japan) E-mail: ishii.ichirou@city.urayasu.lg.jp

4Member of JSCE, Consultant Engineer, Pacific Consultants Co., Ltd. (6-8-1, Nishi-shinjyuku, Shinjyuku, Tokyo 163-6018, Japan)

E-mail: shigeru.satou@tk.pacific.co.jp

The 2011 Great East Japan Earthquake caused liquefaction in many places in the Tohoku and Kanto regions. Because of some areas’ geomorphologic condition, liquefaction occurred in artificially reclaimed lands along Tokyo Bay and the Pacific Ocean, filled lands on former ponds and marshes, and sites of ex-cavation to get iron sands or gravels which had been filled. The most serious damage occurred in Urayasu City, near Tokyo. About 85% of the Urayasu City area liquefied and wooden houses, roads and buried pipes were damaged. However, buildings and bridges supported by piles were not damaged even though the ground around them liquefied. In the Tohoku and Kanto regions, river dikes were damaged at about 2,100 sites, mainly because of the liquefaction of the foundation ground and/or embankments. Sewage pipes and manholes sustained two types of damage: i) lifting due to the liquefaction of replaced fill soils, and ii) the shear failure of manholes and disconnection of pipe joints due to the liquefaction of filled soils and the surrounding ground. Oil tank yards, harbor structures and a tailings dam were also damaged. Key Words : liquefaction, sandy soil, road, river dike, buried pipe

1. INTRODUCTION

Soil liquefaction during earthquakes causes severe damage to many structures. Buildings, tanks, and embankments settle because the soil beneath them loses strength. Underground tanks and buried pipes are lifted because the ground becomes like a liquid and floats owing to buoyancy. Quay walls move towards the sea due to increased earth pressure on them. Piles bend by the horizontal seismic inertial force of superstructures because their bearing capac-ity is reduced. Moreover, if the ground faces a wa-terfront or has a gentle slope, liquefaction tends to cause it to flow, often inducing extreme damage, such as the collapse of pile foundations.

During the 2011 Great East Japan Earthquake, soil liquefaction occurred in the Tohoku region of northeastern Japan and in the Kanto region sur-

rounding Tokyo because the earthquake was huge, with a magnitude of Mw=9.0. Liquefaction occurred in a wide area of reclaimed land along Tokyo Bay, though the epicentral distance was very large, about 380 to 400 km. Large amounts of boiled sand, large settlements and a kind of sloshing of liquefied grounds were observed in the Tokyo Bay area. Many houses, roads, lifeline facilities, and river dikes were severely damaged by soil liquefaction1), 2). The most seriously damaged city was Urayasu City, where about 85% of the city area liquefied. Some port and harbor facilities, tank yards, electric power stations and tailings dams were also damaged in the Tohoku and Kanto regions. However, many new bridges, buildings, tanks, port and harbor facilities, etc. es-caped damage because these structures had been designed to withstand the effect of liquefaction. Old structures which had not been originally designed to

Journal of JSCE, Vol. 1, 181-193, 2013Special Topic - 2011 Great East Japan Earthquake (Invited Paper)

181

withstand the effect of liquefaction but had been repaired to prevent liquefaction-induced damage were not damaged either. In this paper, liquefac-tion-induced damage to structures during the 2011 Great East Japan Earthquake is introduced by fo-cusing on the damage in Urayasu City. 2. BRIEF HISTORY OF THE DESIGN

METHODS FOR LIQUEFACTION IN JAPAN

In 1964, the Niigata and Alaska earthquakes in-

flicted huge damage on buildings, bridges and other structures by liquefying loose sandy soils. After these earthquakes, studies on liquefaction began to be widely carried out using laboratory cyclic shear tests, shaking table tests, and site investigations and anal-yses. Based on these studies, methods for the pre-diction of liquefiable layers, the design of structures in liquefied grounds and measures to prevent lique-faction were developed. In Japan, since 1970, design methods for liquefaction have been introduced in the seismic design codes for port and harbor facilities, highway bridges, railway structures, and buildings, as shown in Table 1. Later design methods were introduced for almost all important facilities, such as oil tanks, LNG tanks, high-pressure tanks, water pipes, sewage pipes and gas pipes. However, for water and sewage facilities, these design methods were applied only to major pipes in the early stage, and then gradually applied to normal pipes.

Countermeasures against liquefaction have also been developed. During the 1964 Niigata earthquake, many oil tanks settled due to liquefaction. However, some tanks were not damaged because their founda-tion ground had been compacted by vibro-flotation3).

This was the first time that the effectiveness of ground compaction against liquefaction was recog-nized. After this, many other kinds of remediation methods were developed in Japan, including sand compaction piles and gravel drains. These remedia-tion methods were summarized in a book by the Japanese Geotechnical Society (JGS) in 1993 in Japanese, which was translated into English in 19984).

Recently, almost all new structures have been de-signed to prevent damage due to liquefaction. However, many structures remain which were con-structed before their seismic design codes considered liquefaction, and many small structures are still de-signed without consideration of the effect of lique-faction. Fig.1 summarizes the earthquakes that caused liquefaction-induced damage to structures from 1964 to 2011, when the Great East Japan Earthquake occurred. As shown in Fig.1, liquefac-tion occurred during 25 earthquakes in those 47 years in Japan.

During the 2011 Great East Japan Earthquake, many wooden houses were damaged due to lique-faction, though no one died as a result. According to the Ministry of Land, Infrastructure, Transport and Tourism, about 27,000 houses were damaged due to liquefaction. This large number of houses suffered damage because no design method for liquefaction has been introduced for wooden houses, since liq-uefaction-induced settlement of houses does not cause serious problems or threats to inhabitants, such as loss of life. Many of the buried structures and river dikes that were damaged had been constructed before liquefaction was considered in their seismic design codes.

On the other hand, many structures, such as ele-vated bridges, buildings, high-pressure gas pipelines,

Table 1 Years when consideration of liquefaction was in-troduced in seismic design codes in Japan.

Design codes and standards Year

Design standard for port and harbour facilities 1970

Seismic design manual for highway bridges 1972

Design criteria of building foundation structures and commentaries 1974

Design standard for railway structures, foundation and retaining wall 1974

Notification specifying particulars of technical standards concerning control of

hazardous materials

1974

Guideline for remedial measures of water works facilities against earthquakes 1979

Specification of construction of tailing dams and commentary 1980

Recommendation practice for LNG in-ground storage 1981

Guidelines for remediation measures of sewage works facilities against

earthquakes

1981

The Ministerial Notification for Aseismic Design (MITI Notification

No.515,1981)

1983

Seismic design standard of land reclamation for Agriculture (Draft) 1984

Seismic design manual for housing sites 1984

Design manual for common utility ducts 1986

Highway earthquake series, manual for soft ground remediation 1986

Technical guidelines for aseismic design of nuclear power plants 1987

Design method of high-pressure gas pipeline for liquefaction 2001

1994 Hokkaido‐toho‐oki Eq. 

1993 Hokkaido‐nansei‐oki Eq.

1968 Tokachi‐oki Eq. 

2003 Tokachi‐oki Eq.

1982 Urakawa‐oki Eq. 

1978 Miyagiken‐oki Eq. 

1978 Izuohshima‐kinkai Eq. 

1973 Nemurohanto‐oki Eq. 

1964 Niigata Eq

1983 Nihonkai‐chubu Eq。

2008 Iwate‐Miyagi‐nairiku Eq. 

1987 Chibaken‐toho‐oki Eq. 

1993 Kushiro‐oki Eq. 

1993 Noto‐hanto Eq. 

1994 Sanriki‐haruka‐oki Eq. 

2007 Noto‐hanto Eq. 

2007 Niigataken‐chuestu‐oki Eq. 

2004 Niigataken‐chuetsu Eq. 

2000 Tottoriken‐seibu Eq. 

2005 Fukuokaken‐seiho‐oki Eq. 

1995 Hyogoken‐nambu (Kobe) Eq.

2009 Surugawan‐okiNo Eq. 

2001 Geiyo Eq.

2003 Miyagiken‐hokubu no Eq. 

2011 Great East Japan Eq. 

Fig.1 Sites of soil liquefaction in Japan caused by earthquakes from 1964 to 2011.

182

and oil tanks, were not damaged, even though severe soil liquefaction occurred, because these structures had been designed to resist liquefaction damage by incorporating pile supports or improving the foun-dation soil. 3. GEOMORPHOLOGICAL FEATURES

OF LIQUEFIED SITES

Liquefied sites in the Tohoku and Kanto regions were investigated by many members of the Japan Society for Civil Engineering (JSCE), the JGS, and others. As the liquefaction-induced damage to houses, river dikes, roads, and lifeline facilities was serious, the Kanto Regional Development Bureau of the Ministry of Land, Infrastructure, Transport and Tourism conducted joint research with JGS to iden-tify liquefied sites. The results of their joint research were published on the government ministry’s web-site5). Fig.2 is a map of the liquefied sites in the Kanto region. Because of their geomorphologic condition, the following sites liquefied: a) Artificially reclaimed lands along Tokyo Bay and the Pacific Ocean b) Filled lands on former ponds and marshes c) Sites excavated to get iron sands or gravels and then refilled.

In addition, embankments and foundations of river dikes and refilled soils for sewage pipes liquefied.

Liquefaction occurred along the Pacific coast of the Tohoku region, but the liquefied sites were ob-scured by a huge tsunami that hit the coastal area

about 30 minutes after the earthquake. According to several photos and videos taken after the earthquake and before the tsunami, liquefaction may have oc-curred in natural ground, as well as in reclaimed land. In higher ground unaffected by the tsunami, filled soils at many housing lots and a tailings dam lique-fied. 4. EFFECT OF LIQUEFACTION ON

STRUCTURES IN URAYASU CITY (1) Liquefied area in Tokyo Bay area

Fig.3 is a map of the liquefied zones in the Tokyo Bay area5). The whole area of the northern part of Tokyo Bay liquefied, but in the eastern and western parts of Tokyo Bay, liquefaction was observed only in spots. In the northern part of Tokyo Bay, the ground surface was covered with boiled sands all around the reclaimed lands in Shinkiba in Tokyo, Urayasu City, Ichikawa City, Narashino City and western Chiba City. In contrast, boiled sands were observed only here and there in the reclaimed lands in Odaiba, Shinonome, Tatsumi, Toyosu and Seishin in Tokyo and in eastern Chiba City. The total lique-fied area from Odaiba to Chiba City reached about 41 km2. Many houses, roads, and lifeline facilities were severely damaged in the liquefied zones. The most serious damage was in Urayasu City, where about 85% of the city liquefied.

The epicentral distance of the liquefied area in Tokyo Bay is very long, around 400 km. However, the distance from the boundary of the rupture plane is

Fig.2 Sites of soil liquefaction in the Kanto Region caused by the Great East Japan Earthquake of 20115).

2km

D’

D

Funabashi

Urayasu

Seishin

Shinkiba

Odaiba

Tatsumi

Shinonome

Takahama

Takasu

Liquefied zone

Fig.3 Liquefied areas from Odaiba in Tokyo to Chiba5).

Tokyo

Kanto Region

4

Liquefied site

20 0 20(km)

Kanagawa

Chiba

Saitama

Ibaraki

TochigiGumma

183

only 110 km because the rupture plane is very wide. (2) History of reclamation and soil in Urayasu

City The reclamation of coastal areas in Tokyo Bay

started in the seventeenth century and accelerated with an increase in population. The history of rec-lamation work in Urayasu City is summarized in Fig.46). Urayasu City is divided into three towns, Motomachi, Nakamachi and Shinmachi, which mean old, middle and new towns, respectively. The ground of Motomachi formed naturally at an estuary of Edo River. On the other hand, zones A, B and C in Nakamachi were reclaimed from 1965 and zones D, E and F in Shinmachi were reclaimed from 1972. The area of each zone is 1.0 to 3.5 km2, and the total area of Nakamachi and Shinmachi is 14.36 km2, which is about 85% of the area of Urayasu City.

Four months after the earthquake, a technical committee chaired by Prof. K. Ishihara was orga-nized by the Urayasu City Government to study the mechanism of liquefaction-induced damage and restoration works. The committee conducted detailed soil investigations, laboratory tests and analyses, and reported valuable results6). The following is a brief summary of the report.

Fig.5 shows a soil cross-section along the line Urayasu D-D’ in Figs.3 and 4 through Motomachi, Nakamachi and Shinmachi. In the reclaimed zone, a filled layer of mainly hill sand (B) and a filled layer of dredged sandy soil (F) with low SPT N-values averaging 5.46 are deposited with a thickness of several meters. Though layer F had to be liquefied, the fines content of layer F is very large, with an average value of about 44%, and very scattered from FC=0% to 100%, as shown in Fig.6. An alluvial sand layer (AS) with an average SPT N-value of 12.9 un-derlies the filled layer. The average fines content of layer AS is 30.9%. A very soft alluvial clay layer (Ac) is deposited under the AS layer with a thickness of 10 to 40 m, increasing in thickness as it approaches the sea. A diluvial (Pleistocene) dense sandy layer (Ds) with an SPT N-value of more than 50 underlies the alluvial clay layer. The water table is shallow, aver-aging 1 to 2.5 m below ground level (G.L.).

Urayasu City contains light structures, such as wooden houses, and heavy structures, such as high-rise buildings and elevated bridges. These heavy structures are supported by long piles which mainly penetrate to the depth of the diluvial dense layer.

(3) Zones of structural damage

Fig.7 shows the areas of Urayasu City where liq-uefaction damaged sewage facilities, roads and

houses. Liquefaction occurred only in Nakamachi and Shinmachi, as Motomachi is built on natural land. In Motomachi an alluvial sand (AS) layer is deposited just under the water table, as shown in Fig.5. If the AS layer had liquefied, sand boils should have appeared. However, no sand boils, nor damage to houses and lifeline facilities, occurred. Therefore, it can be estimated that the AS layer basically was not liquefied by the 2011 Great East Japan Earthquake, though some loose parts of the layer might have

Fig.5 Detailed soil cross-section of Urayasu City (Slightly modified from the report by Urayasu City6)).

F B

As

Ac

AsAs

Na

Altitude(m

)

Altitude(m

)

‐25

‐50

‐75

‐25

‐50

‐75

D D’

1km

Ds

Fig.4 History, location and extent of land reclamation in Urayasu6).

Fig.6 Frequency distributions of fines content FC for dredged soils6).

0

50

100

150

200

250

300

350

0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100

デー

タ数

FC (%)

μ‐σ μ+σμ

平均値μ=43.8 標準偏差σ=30.4

サンプル数N=410

Num

ber

of s

ampl

es

Fines content, FC(%)

Average FC=43.8%

184

liquefied and some parts of the dredged sandy soil under the water table might have liquefied in Nakamachi and Shinmachi.

Sewage pipes, manholes, roads and houses were damaged in the liquefied areas, but the severity of damage to roads and houses varied from place to place.

(4) Accelerations during main shock and after-

shock The duration of shaking in the 2011 Great East

Japan Earthquake was extremely long, and the main shock was soon followed by a big aftershock. Surface acceleration during the main shock was not high, around 160 Gal. Fig.8 shows the accelerographs and velocities recorded during the main shock and the aftershock by K-NET7) in Urayasu on ground where liquefaction did not occur.

Some inhabitants of Urayasu City observed the boiling of muddy water immediately after the main shock, while others saw spouts of muddy water five to nine minutes after the main shock8). Surprisingly, some inhabitants testified that boiling did not occur during the main shock but occurred during the af-tershock. Some excess pore-water pressure may have built up during the main shock, causing complete liquefaction during the aftershock. It is impossible to evaluate the liquefaction time, but it must have dif-fered by place because of the long duration of the main shock and the subsequent large aftershock. Shaking continued for a long time after the occur-rence of liquefaction. (5) Damage to flat roads

As described in Table 2, there are four grades of planar roads in Urayasu City: national roads, pre-

応急危険度調査対象

特に建物被害の多い所

道路の被害中エリア

道路の被害小エリア

護岸の被害が大きい所

明海二丁目

明海一丁目

明海四丁目

明海三丁目

明海五丁目

明海六丁目

明海七丁目

日の出八丁目日の出七丁目

日の出五丁目

日の出六丁目

日の出三丁目

日の出四丁目

日の出二丁目

日の出一丁目

高洲九丁目高洲八丁目

高洲七丁目

高洲六丁目

高洲五丁目

高洲三丁目

高洲二丁目

高洲一丁目

当代島三丁目

猫実三丁目

鉄鋼通り一丁目

鉄鋼通り二丁目

千鳥

高洲四丁目

弁天三丁目

鉄鋼通り三丁目

舞浜三丁目

舞浜

入船四丁目

美浜二丁目

今川三丁目

今川四丁目

弁天二丁目

富岡三丁目

富岡四丁目

東野二丁目

東野一丁目

富岡一丁目

富岡二丁目

美浜一丁目

入船六丁目

入船一丁目

入船二丁目入船三丁目

今川一丁目

今川二丁目

舞浜二丁目

東野三丁目

富士見五丁目

富士見四丁目

富士見一丁目

富士見三丁目

弁天四丁目

弁天一丁目

入船五丁目

美浜五丁目

美浜三丁目

海楽一丁目堀江一丁目

猫実一丁目

猫実二丁目

北栄三丁目

北栄四丁目

海楽二丁目

美浜四丁目

富士見二丁目

堀江五丁目

堀江六丁目

堀江二丁目

堀江三丁目

猫実四丁目

猫実五丁目

堀江四丁目

北栄一丁目

当代島二丁目

北栄二丁目

当代島一丁目

港

ンイラト

ーゾリ

ーニズィデ

ＪＲ　京葉線

ＪＲ　京葉線

ＪＲ　京葉線

営団地下鉄東西線

営団地下鉄東西線

営団地下鉄東西線

液状化及び下水道施設破損エリア

道路の被害大エリアZones where severe damage to houses occurred

1km

Areas surrounded by linesBlack: Liquefied and sewage facilities

were damagedRed: Roads were severely damaged Blue: Roads were damagedYellow: Roads were slightly damaged

Fig.7 Zones of Urayasu City where roads, sewage pipes and houses were severely damaged6).

Fig.8 Acceleration and velocity waves measured by Urayasu K-NET during the main shock and the biggest aftershock7).

-200 -100

0

100

200

0 50 100 150 200 250 300

Time (sec)

CHB008EW(14:46) EW

Acc

eler

atio

n (G

al)

-200

-100

0

100

200

0 50 100 150 200 250 300Time (sec)

Acc

eler

atio

n (G

al)

CHB008EW(15:15) EW

-40

0

-20

20

40

0 50 100 150 200 250 300Time (sec)

CHB008EW(14:46) EW

Ve

loci

ty (

Kin

e)

-40

-20

20

0

40

0 50 100 150 200 250 300Time (sec)

Vel

ocity

(K

ine)

CHB008EW(15:15) EW

(1) Main shock (May 11,14:46) (2) Afteshock (May 11, 15:15)

185

fectural roads, main city roads and normal city roads. The total length of damaged roads amounted to about 80 km. These roads were designed without consid-ering liquefaction because no seismic design codes which consider liquefaction have been developed for flat roads. Liquefaction caused several phenomena to interrupt traffic: i) much muddy water boiled onto roads (see Fig.9); ii) road asphalt heaved, was thrusted or was deformed

into waves at many sites (see Fig.10); iii) many small road cave-ins occurred several

months after the earthquake. (6) Damage to river and sea walls

Three sides of Urayasu City face Tokyo Bay and

are protected by sea walls. Two rivers, the Sakai River and the Miake River, transverse Urayasu City from northwest to southeast and are protected by river walls. These sea and river walls were not de-signed to withstand liquefaction damage because reclamation work for the walls started only a few years after the 1964 Niigata Earthquake, when liq-uefaction was first recognized. Though many sea and river walls are found in the liquefied area, only a few walls were damaged, as shown in Fig.4. These sea and river walls slightly expanded and settled. Fig.11 shows the damaged concrete river wall at Sakai river in Urayasu City, and Fig.12 shows a cross-section of this wall.

During the 1995 Hyogoken-nambu Earthquake, many caisson quay walls and sea walls moved and tilted toward the sea in the liquefied area, resulting in large lateral flows of the ground behind the walls. The failure of the quay wall is a complex dynamic phenomenon. To explain from the point of view of the design, the failure process can be summarized as follows: i) a strong inertial force due to very high acceleration, ii) a decrease in shear strength due to an increase in excess pore-water pressure in the replaced sand beneath the caisson walls, and iii) an increase in earth pressure due to a rise in excess pore-water pressure in the ground behind the wall. In contrast, only a few sea and river walls moved and tilted in Urayasu and other cities along Tokyo Bay during the 2011 Great East Japan Earthquake. The reasons for the difference in damage to these walls caused by the two earthquakes are uncertain, but they may be the different wall types and strengths, different soil conditions, different slopes of the ground behind the walls, the depth of liquefied soil and different ac-celerations. (7) Damage to lifeline structures

Lifeline facilities for water, sewage, gas, electric power and telegraphs were severely damaged in the liquefied area. Though seismic design codes for these

Table 2 Extent of road damage in Urayasu City6).

Fig.9 Cars stuck in boiled muddy water in Urayasu City6).

Fig.10 Thrusted road pavement in Urayasu City6).

Length (km)

Damage length (km)

Damage rate (%)

National road 4.5 ‐ ‐

Prefectural road 7 1.9 27

City main road 28 12 43

City normal road 195 66 34

Fig.11 Moved and settled wall along Sakai River in Urayasu City6).

186

facilities started to consider the effect of liquefaction from about 1980, as shown in Table 1, many lifeline facilities in Urayasu City were constructed before that date. So, many facilities were damaged due to liquefaction. Detailed assessments of damage to buried pipes are not available because not all dam-aged pipes have been excavated and permanently restored.

Leakage occurred at 608 sites along water pipes, and water service for 33,000 households was stopped. Water pipes were broken at 122 of the leakage sites, and joints were pulled out or sheared at

366 sites. Sewage pipes and manholes for waste water and

rain water were severely damaged in a wide area. Urayasu City was served by 20.7 km of main sewage pipes and 191.5 km of branch sewage pipes. Of these pipes in Nakamachi and Shinmachi, 20% and 14%, respectively, were damaged, whereas no pipes were damaged in Motomachi, as summarized in Table 3. Fig.13 shows the sites of damaged waste water pipes and manholes. Sewage pipes were deformed, cracked, broken and meandered, and joints were sheared or disconnected, as shown in Fig.14 (1).

0.0

1.0

3.0

2.0

4.0

5.0

-1.0

-2.0

A.P　m

H.W.L　A.P+2.0m

Wood Pile

Concrete sheet pile

Blast furnace slag

Blast furnace slag

Concrete

Fig.12 Cross-section of the moved and settled wall along Sakai River in Urayasu City6).

Table 3 Extent of damage to public waste water pipes and manholes in Urayasu City6).

Motomachi Nakamachi Shinmachi

Pipe Length (km) 84.2 94.4 33.6

Damaged length (km)

0 19.2 4.6

Damage rate (%) 0 20 14

Manhole Number 2,348 2,799 1,069

Damaged number 0 328 147

Damaged rate (%) 0 12 14

Fig.13 Sites of damaged waste water pipes and manholes in Urayasu City6).

管路被害（要補修箇所）

マンホール浮上

マンホール沈下

マンホール破損・ズレ

マンホール蓋のみ

Damaged pipeUplift of manholeSettle of manholeShear of manholeDamage to lid of manhole

TokyoDisneyland

Maihama

Chidori

Minato

TakasuTekkodori

Benten

Imagawa

Tomioka

HigashinoFujimi

Horie Nekozane

Kairaku

Mihama

Irifune

Hinode

Akemi

Damaged pipeUplift of manholeSettling of manholeShearing of manholeDamage to lid of manhole

187

Muddy water seeped into the damaged pipes for 127.2 km and closed pipes completely for about 60 km.

Many sewage manholes were cracked and sheared in a horizontal direction and filled with muddy water, as shown in Fig.14 (2), while a few manholes were lifted or slightly settled, as shown in Fig.14 (3). Muddy water filled almost all manholes in Nakamachi and Shinmachi. It took 34 days to restore these damaged pipes and manholes provisionally.

Shear damage to manholes had not occurred dur-ing past earthquakes in Japan. Many sewage man-holes were lifted but not sheared during the 1993 Kushiro-oki earthquake, the 2003 Tokachi-oki earthquake and the 2004 Niigataken-chuetsu earth-quake9). During the construction of the manholes and

pipes, the ground was excavated to a width of about 2 m. After placing the manholes and pipes, the exca-vated area was filled with sand. During past earth-quakes, only the fill soils were liquefied, but during the 2011 Great East Japan Earthquake, both the fill soils and the surrounding soils were liquefied in Urayasu. Therefore, the horizontal displacement of the liquefied soil around the pipes in Urayasu caused by the 2011 earthquake must have been larger than the horizontal displacement caused by previous earthquakes. Moreover, as the main shock of the 2011 earthquake was very long and was followed by a big aftershock, shaking must have continued for a long time after the occurrence of liquefaction. The large horizontal displacement of liquefied ground had to have caused the disconnection of the pipe joints and the shear failure of the manholes, allowing the influx of muddy water into the pipes and man-holes.

In gas facilities, gas holders, governor stations and high-pressure pipes were not damaged and mid-dle-pressure pipes were only slightly damaged. On the other hand, some low-pressure gas pipes were broken and muddy water filled the pipes to a length of 11 km, resulting in the shut-down of gas service for 8,631 homes.

Buried cables for electric power distribution were damaged at six sites. Manholes for telegraph lines were sheared similar to the sewage manholes. Moreover, electric and telegraph poles seriously settled and tilted here and there.

A steel water tank for emergency use buried to a depth of G.L.-2.4 m was lifted, as shown in Fig.15, though the tank did not break. The tank is 3.0 m in diameter, 15.1 m in length, 100 m3 in capacity and fixed with belts to prevent uplift; the belts must have failed. (8) Damage to houses

According to the Ministry of Land, Infrastructure, Transport and Tourism (MLIT), about 27,000 wooden houses in Japan were damaged due to liq-uefaction caused by the 2011 Great East Japan Earthquake. Strip footing foundations or mat foun-dations are used for houses in Japan. In the design of wooden houses, liquefaction has not been consid-ered, whereas the design code for other buildings has considered liquefaction since 1974, as shown in Ta-ble 1. This is the main reason such a large number of houses were damaged. In contrast, some houses supported by steel piles or cement-mixed soil piles penetrating to the depth of the non-liquefied layer did not settle.

In Urayasu City, severely damaged houses were lo-cated in the zones shown in Fig.7. Many houses settled and tilted, though they suffered no damage to walls and

Fig.15 Uplifted water tank for emergency use in Urayasu City6).

(1) Meandered pipe with disconnected joints

(3) Uplifted manhole

(2) Sheared

Fig.14 Schematic diagram of damage to sewage pipes and manholes in Urayasu City6).

188

windows, as shown in Fig.16. In the greatly tilted houses, inhabitants felt giddy, sick and nauseated, and found it difficult to live in their houses after the earth-quake. In May 2011, the Japanese Cabinet announced a new standard for the evaluation of damage to houses based on two factors, settlement and inclination. A new class of “large-scale half collapsed house” was also introduced, and houses tilted at angles of more than 1/20, of 1/20 to 1/60, and of 1/60 to 1/100 were judged to be totally collapsed, large-scale half collapsed and half collapsed houses, respectively, under the new standard. The number of damaged houses in Urayasu City by the new standard in listed in Table 4. The number of totally collapsed, large-scale half collapsed, half collapsed partially damaged and undamaged houses in Urayasu City was 10, 1,509, 2,102, 4,848 and 963, respectively, as of Sept. 3, 2011. According to a previous study on the non-uniform settlement of hous-es10), several factors affect such settlement. Among them, the effect of adjacent houses was dominant in the Abehikona housing lot during the 2000 Tottori-ken-seibu Earthquake10). A similar tendency was ob-served in the Tokyo Bay area. If two houses are close to

each other, they tilt toward each other, and if four houses are close together, they tilt toward their center point8), 11) as shown in Fig.16.

Even in liquefied areas, several zones did not liq-uefy or suffer damage to houses because the ground in these zones had been improved. One example is a housing lot in Irifune where sand compaction piles (SCPs) were installed in the foundations of mainly three-story houses to increase the bearing capacity and to prevent liquefaction, and gravel drains (GDs) were installed in the foundations of mainly two-story houses to mitigate liquefaction8). Each type of building had a spread foundation (continuous foun-dation) supported by nodular piles 8 m long. Fol-lowing the earthquake, sand boils were observed in unimproved areas around the housing lot. However, no sand boils were seen within the improved area and no damage occurred to buildings constructed on ground improved by either the SCP method or the GD method. (9) Damage to heavy structures supported by

piles Many road and railway bridges supported by piles

were not damaged, though liquefaction occurred around the piles. Fig.17 shows an elevated bridge on the Keiyo Railway Line which is supported by pre-stressed concrete (PC) and steel and concrete com-posite (SC) piles, as illustrated in Fig.18. In the design of the pile foundation, liquefaction was as-sumed to be probable and the lateral bearing capacity of the liquefiable layer, from the water level to G.L. -10 m, was assumed to be zero12). Thus, SC piles, which are stronger than PC piles, were installed to a depth of -10 m.

There are many buildings supported by piles in the liquefied area. These buildings were not damaged, but the inhabitants of some apartment buildings were forced to live in inconvenient conditions for weeks because of the shutdown of lifeline facilities and because of ground settlement at the entrances of their apartment houses. At the fire station shown in Fig.19, ladder trucks could not enter the fire station due to the settlement of the road in front of the station.

During the 1995 Hyogoken-nambu Earthquake, many piles for road bridges and apartment buildings were damaged due to liquefaction in Kobe City. In contrast, no damage occurred to road bridges and apartment buildings with similar pile foundations in Urayasu and other cities in the Tokyo Bay area. This difference in damage may be due to the larger am-plitude of shaking in Kobe, where the recorded ac-celerations were about three times the accelerations recorded in the Tokyo Bay area.

Fig.16 Settled and tilted houses in Urayasu City.

Table 4 Number of damaged houses in Urayasu City under old and new standards (Counted on Sept. 3, 2011).

Grade of damage Number of houses

Old standard New standard

Totally collapsed 8 10

Large-scale half

collapsed 0 1,509

half collapsed 33 2,102

Partially damaged 7,930 4,848

No damage 1,028 963

Total 8,999 9,432

189

5. LIQUEFACTION-INDUCED DAMAGE TO SEVERAL STRUCTURES IN TOHOKU AND KANTO REGIONS

(1) Damage to river dikes

River dikes settled and slid seriously at many sites in the Tohoku and Kanto regions. According to the

Tohoku and Kanto Regional Development Bureaus of the MLIT, there were 1,195 sites of dike damage in the Tohoku region and 939 sites in the Kanto region. Most river dikes in these regions were designed and constructed without considering liquefaction. In the Tohoku region, a few dikes which had been damaged during earthquakes in 1978 and 2003 and restored to prevent liquefaction-induced damage suffered no damage from the 2011 Great East Japan Earthquake. After this earthquake, technical committees were organized by the two regional development bureaus to ascertain the mechanism of failure and select ap-propriate restoration methods.

In the Kanto region, dikes along the Tone, Kokai, Kinu, Hinuma, Naka and other rivers were damaged, as shown in Fig.2013). The types of damage included settlement, as shown in Fig.21, slides, longitudinal cracks and transverse cracks. Serious damage oc-curred at 55 sites. The mechanism of the damage at the seriously damaged sites was studied based on soil investigation. It was concluded that 51 sites suffered damage triggered by liquefaction. As shown in

Keiyo Railway Line

Fig.17 Undamaged elevated bridge on the Keiyo Railway Line at Shin-Urayasu Station.

Fig.19 Road subsidence in front of a fire station.

Fig.18 Diagram of pile foundation of Keiyo Railway Line12).

19.8m

Upper pile

Lower pile

Piles (φ=600mm)Upper pile : SC pileLower pile : PC pile

0 50

10

20

30

40

50

60

SPT N

0 30 60km

Ibaraki

Tochigi

Gunma

Saitama

Tokyo

Kanagawa

Chiba

Mt.Fuji

Kuji River

Naka River

Tone River

Lake Kasumigaura

Kokai River

Kinu River

Watarase River

Ara River

Tama River

Tone River

N

: Damage zone

Fig.20 Sites of river dike damage in the Kanto region(modified from MLIT13)).

Fig.21 Settled river dike along the right bank of Tone River.

190

Fig.22, there were three patterns of liquefied soil: 1) the foundation ground liquefied, 2) the lower part of the embankment liquefied, and 3) both foundation ground and embankment liquefied. In the latter two types, the water tables inside the embankments were higher than the water tables in the surrounding-ground. In the restoration work, several appropriate methods were selected in accordance with the type and severity of damage to prevent damage during future earthquakes, as shown in Fig.23.

(2) Damage to sewage pipes and manholes

According to the MLIT, sewage pipes and man-holes were damaged in 132 cities, towns and villages. Of 65,001 km of sewage pipes in the cities, towns and

villages, 642 km were damaged14). As shown in Fig.24, 65.9% of the damage to pipes was due to the liquefaction of filled soils and 24.5% was due to the liquefaction of both filled soils and the surrounding ground. Of the damage to manholes, 41.7% was due to the liquefaction of filled soils and 26.7% was due to the liquefaction of filled soils and the surrounding ground. The damage due to the liquefaction of both filled soils and surrounding ground mainly occurred along Tokyo Bay, as shown in Fig.25, because liq-uefaction occurred in a wide area of reclaimed land, as mentioned before.

Most of the damaged pipes and manholes were constructed without consideration of liquefaction. A few pipes and manholes that had been restored with

(2) Liquefaction of embankment

Embankment

Liquefaction

Clay

(3) Liquefaction of ground and embankment

Embankment

Liquefaction

Liquefaction Sand

Clay

(1) Liquefaction of ground

Embankment

Liquefaction Sand

Clay

Embankment

Liquefaction Sand

Clay

(4) Liquefaction of ground and embankment

Fig.22 Patterns of soil liquefaction along river dikes13).

Fig.23 Diagram of methods to restore damaged river dikes in the Kanto region (modified from MLIT13)).

Liquefied part Restoration method Schematic diagram

<Before earthquake>

Drainage

Densification by sand compaction

pile

Deep mixing

<Before earthquake>

Sheet pile

Deep mixing

Embankment+

ground

Drainage+

Sheet pile

Lower part ofembankment

Foundationground

River level

Ground water levelLiquefied layer

Non-liquefied layer

Deep mixing method

River level

Ground water level Liquefied layer

Non-liquefied layer

Sheet pile +Drain + Waste channel

Deep mixing method

Sheet pile +Drain + Waste channel

Drain + Waste channel

S.C.P + Waste channel

Seismic force

Liquefaction of filled soil

Liquefaction of surrounding ground and filled soilTsunami

Slide or settlement of fill in housing lotsOthers

Fig.24 Causes of damage to sewage pipes in the 132 damaged cities, towns and villages14).

Uplift of sewage pipes and manholes due to the liquefaction of filled soils

Remarkable damage to sewage pipes and manholes due to the liquefaction of surrounding grounds and filled soils

Fig.25 Areas where two patterns of damage to sewage pipes and manholes dominated14).

Fig.26 Diagrams of successful measures for preventing the uplift of existing manholes14).

Non liquefiable Non liquefiable

(1) Dissipation of excess pore water pressure

(2) Balance by coun-ter weight

191

countermeasures after the 2008 Iwate-Miyagi- Nairiku Earthquake were not lifted during the 2011 Great East Japan Earthquake. These pipes and manholes had been restored by one of two measures proposed by the Technical Committee on the Sewer Earthquake Countermeasures in 2005: i) fill ditches with gravel instead of sand, and ii) mix the filled sand with cement.

Two other recently developed methods to prevent the uplift of existing manholes, illustrated in Fig.26, had been applied in some parts of the liquefied areas and successfully served their purpose. (3) Damage to plants

There are many electric power plants, oil tank yards, factories and warehouses in the liquefied areas along the coasts of Tokyo Bay and the Pacific Ocean. Fig.27 shows a tilted oil tank yard fence15). The ex-tent of damage to plants is still unknown because most of the plants belong to private companies and site surveys are difficult to undertake. However, serious damage, such as the burning of oil tanks, did not occur in the liquefied areas. Most likely, the main reason for this lack of serious damage is that the important facilities, such as oil tanks and electric power plants, were designed based on seismic design codes that considered liquefaction and required the use of appropriate pile foundations or soil im-provement. Many old oil tanks constructed before liquefaction was considered in seismic design codes had been strengthened against liquefaction by special measures, such as surrounding the tanks with sheet piles to prevent liquefaction-induced settlement. (4) Damage to port and harbor structures

There are many ports and harbors along the coasts of the Pacific Ocean and Tokyo Bay. No serious damage was inflicted on ports and harbors in the liquefied areas along Tokyo Bay. Along the Pacific

coast, it is not easy to judge whether damage oc-curred due to liquefaction or not because a huge tsunami washed away all traces of liquefaction. Sugano reports that obvious liquefaction-induced damage occurred at two ports, Souma Port and Onahama Port, where the corners of quay walls moved toward the sea16). At Onahama Port, an apron behind a quay wall settled below the rails for cargo loaders.

At Sendai Airport, an underpass road and an un-derground conduit had been constructed beneath the tarmac at four points. For their construction, the ground had been excavated. After their construction, the excavated area was filled with sand. Subse-quently, special techniques to prevent liquefaction, i.e., compaction grouting, and high-pressure injec-tion, were applied to the sand fill at three of these points. No liquefaction accompanying the Great East Japan Earthquake occurred at these three points. However, at the fourth point, which had not been improved, the tarmac settled17).

(5) Damage to a tailings dam

Three abandoned tailings dams failed in the Tohoku and Kanto regions. The Kayakari tailings dam built by Ohya Mining, located near Kesen-numa City in Miyagi Prefecture, failed due to the liquefac-tion of gold slime, as shown in Fig.28. A cross-section of the failed slope is shown in Fig.2918). Slime from the failed slope flowed down a small river and destroyed a house.

Fig.28 Liquefied and failed slope of Kayakari tailings dam.

Slime

Rock

0

20

40(m)

0 25 50 (m)

Before failure

After failure

Fig.29 Cross-section of failed slope at Kayakari tailings dam18).

Fig.27 Tilted oil tank yard fence along the Pacific Ocean coast15).

192

The Kayakari tailings dam was constructed by the inner filling method and finished in 1966. After the Mochikoshi tailings dam failed due to liquefaction during the 1978 Izuohshima-kinkai Earthquake, liquefaction consideration was introduced in the design code for tailings dams in 1980. Then, the probability of liquefaction-induced damage at many tailings dams was estimated by conducting soil in-vestigations and tests. The Kayakari tailings dam was one of those examined, and the examiners concluded that there was little or no probability of such damage. However, the seismic intensity that hit the Kayakari tailings dam during the Great East Japan Earthquake was fairly higher than the examined intensity. 6. CONCLUSIONS

Liquefaction-induced damage to structures during the 2011 Great East Japan Earthquake has been de-scribed based on site investigations conducted by the authors and others, and the following conclusions are derived: 1) Because of their geomorphologic condition, arti-ficially reclaimed lands along Tokyo Bay and the Pacific Ocean, filled lands on former ponds and marshes, and sites of excavation to get iron sands or gravels that had been refilled, liquefied. 2) In Urayasu City, where about 85% of the city area liquefied, wooden houses, flat roads and buried pipes were seriously damaged due to liquefaction, whereas buildings and bridges supported by piles were not damaged even though the ground surrounding them liquefied. 3) River dikes were damaged at about 2,100 sites, mainly because of the liquefaction of the foundation ground and/or embankments. 4) Sewage pipes and manholes sustained two types of damage: i) lifting due to the liquefaction of fill soils, and ii) the shear failure of manholes and pull-out of pipe joints due to the liquefaction of fill soils and the surrounding ground. ACKNOWLEDGMENT: Some of the results cited in this paper are quoted from reports by Technical Committees organized by the Ministry of Land, In-frastructure, Transport and Tourism, Ministry of Economy, Trade and Industry, and Urayasu City Government. The authors would like to express their sincere appreciation to them. REFERENCES 1) Yasuda, S. and Ishikawa, K. : Several features of liquefac-

tion-induced damage to houses and buried lifelines during the 2011 Great East Japan Earthquake, Proc. of the Inter-national Symposium on Engineering Lessons Learned from the 2011 Great East Japan Earthquake, pp.825-836, 2012.

2) Towhata, I., Kiku, H. and Taguchi, Y. : Technical and societal problems to be solved in geotechnical issues, Proc. of the International Symposium on Engineering Lessons Learned from the 2011 Great East Japan Earthquake, pp.837-848, 2012.

3) Watanabe, T. : Damage to oil tank refinery plants and a building on compacted ground by the Niigata earthquake and their restoration, Soils and Foundations, Vol.6, No.2, pp.86-99, 1966.

4) The Japanese Geotechnical Society : Remedial Measures Against Soil Liquefaction, Balkema, 1998.

5) http://www.ktr.mlit.go.jp/ktr_content/content/000043569.pdf. (in Japanese)

6) Technical Committee on Measures Against Liquefaction in Urayasu City : Report on measures against liquefaction in Urayasu City, 2012. (in Japanese) http://www.city.urayasu.chiba.jp/secure/26052/lasthoukoku01.pdf

7) National Research Institute for Earth Science and Disaster Prevention (NIED). “K-NET WWW service”, Japan (http://www.k-net.bosai.go.jp/).

8) Yasuda, S., Harada, K., Ishikawa, K. and Kanemaru, Y. : Characteristics of the liquefaction in Tokyo bay area by the 2011 Great East Japan earthquake, Soils and Foundations, 2012 (in printing).

9) Yasuda, S. and Kiku, H.: Uplift of sewage manholes and pipes during the 2004 Niigataken-chuetsu earthquake, Soils and Foundations, Vol. 46, No. 6, pp. 885-894, 2006.

10) Yasuda, S. and Ariyama, Y.: Study on the mechanism of the liquefaction-induced differential settlement of timber houses occurred during the 2000 Tottoriken-seibu earth-quake, Proc. of 14th World Conference on Earthquake Engineering, Beijing, Paper No.S26-021, 2008.

11) Tokimatsu, K. and Katsumata, K.: Liquefaction-induced Damage to Buildings in Urayasu City During the 2011 Tohoku Pacific Earthquake, Proceedings of the Interna-tional Symposium on Engineering Lessons Learned from the 2011 Great East Japan Earthquake, pp.665-674, 2012.

12) Wada, A., Suzuki, H., Fujiwara, T. and Takasaki, H. : Effect of liquefaction on railway facilities of Keiyo Line, The Foundation Engineering & Equipment, Vol.40, No.4, pp.71-73, 2012. (in Japanese)

13) http://www.ktr.mlit.go.jp/ktr_content/content/000045202.pdf. (in Japanese)

14) http://www.mlit.go.jp/common/000211317.pdf. (in Japa-nese)

15) Yasuda, S., Fujita, S. and Minagawa, K. : Liquefaction in Kanto Regions in the Great East Japan Earthquake, Me-chanical Engineering Congress, 2012, S101012, 2012. (in Japanese)

16) Sugano, T. : Damage and restoration of port and airport facilities, The Foundation Engineering & Equipment, Vol.40, No.4, pp.39-42, 2012. (in Japanese)

17) Mizukami, J. : Damage of major facilities of Sendai Airport, The Foundation Engineering & Equipment, Vol.40, No.4, pp.71-73, 2012. (in Japanese)

18) http://www.meti.go.jp/policy/safety_security/industrial_safety/oshirase/2012/06/240622-1-2.pdf. (in Japanese)

(Received, October 12, 2012)

193