Special Lectures

Ishihara Lecture
 Earthquake Geotechnical Engineering for Protection and Development of
Environment and Constructions – Silvestri & Moraci (Eds)
© 2019 Associazione Geotecnica Italiana, Rome, Italy, ISBN 978-0-367-14328-2

Summarizing geotechnical activities after the 2011 Tohoku

earthquake of Japan

I. Towhata
Kanto Gakuin University, Yokohama, Japan

ABSTRACT: After the 2011 Tohoku earthquake in Japan, I as the Vice President and then
as the President of the Japanese Geotechnical Society took initiatives in damage reconnais-
sance, developing new safety criterion for liquefaction-prone residential areas, quantitative
evaluation of ageing effects on liquefaction resistance of sand, investigation on damage mech-
anism of river levees, recovery from the tsunami incidence of Fukushima No.1 Nuclear Power
Plant and ground improvement in existing residential area. Some of them were conducted
urgently within several months after the earthquake, while others encountered many difficul-
ties from the viewpoint of public agreement. This paper summarizes the post-earthquake
activities of geotechnical engineering in restoration of the community.

1 INTRODUCTION

The 2011 Tohoku earthquake of Japan registered the seismic magnitude of Mw=9.0. This
event forced people to change their idea on safety during natural disasters. Before this earth-
quake, people demanded engineers to achieve zero risk during natural disasters and engineers
believed that it is possible. After this earthquake, in contrast, people understood that zero risk
is impossible under such a rare event. Nevertheless, it appears that they are not yet prepared
reasonably for possible extreme events. They are either unnecessarily scared by the worst but
rare event or still overlook the risk of natural disasters. This paper addresses the experience of
the author who had opportunities after the 2011 earthquake to be involved in a variety of
post-disaster activities that were related with geotechnical engineering.

2 SEISMOLOGICAL ASPECTS OF THE 2011 TOHOKU EARTHQUAKE

The Pacific Coast of Eastern Japan (Figure 1) is a seismically active region. Because of the subduc-
tion of the Pacific Ocean Plate, many gigantic earthquakes occurred there in the past and some of
them triggered disastrous tsunamis. Noteworthy is that those earthquakes and tsunamis in the past
occurred in the northern part of Tohoku coast (Figure 1) and that much was not known about the
seismic vulnerability in the southern part (HERP, 2009). This lack of knowledge became important
later in the context of safety of nuclear power plants in the southern coast. Figure 2 shows the
historical earthquake disasters in the Pacific Coast of Tohoku. They occurred in the Sendai City
area and in the northern part of the Pacific coast. Obviously many gigantic earthquakes occurred
in this region and tsunami disasters were repeated. The problem is the lack of earthquake and tsu-
nami information from 9th to 16th Century. Because the Tohoku Region was far from the capital
(Kyoto), written record was not preserved of natural disasters. Even the official record on the AD
869 earthquake had not been fully trusted because of its unclear description.
Based on the repetition of gigantic earthquakes in the past, the next big one had been expected
for decades. Figure 3 illustrates the occurrence of many earthquakes in the Pacific Coast area of
Tohoku and its surroundings before and after the 2011 main shock. When the one on March 9
(two days before the catastrophic one) registered M=7.3 off Sendai, it was thought that the
expected one occurred. It was unfortunately wrong and most probably this event was a foreshock.

3
 Figure 2. History of earthquakes in Pacific Coast
Region of Tohoku, Japan (Northern Tohoku)
Figure 1. Tohoku Region and Sendai. (Most data after Hatori, 1975).

Figure 4. K-Net acceleration record of Tohoku

Figure 3. Aftershocks in Tohoku Region (data Meteorological Agency;earthquake at Ishinomaki
by Japan. maximum seismic intensity>3). (NS component).

The earthquake of Mw=9 in 2011 was triggered by tectonic rupture off the entire Pacific Coast
of the Tohoku Region. The size of rupture was 500 km in NS direction and 200 km in EW direc-
tion (JMA, 2011), thus generating tsunami wave over the entire coast of Tohoku. Figure 4 illus-
trates the time history of acceleration that was recorded by K-Net Ishinomaki Station which is
near Sendai. This shaking is characterized by its long duration of the strong phase, consisting of
many number of cycles, and two particularly strong phases of shaking at around 50 and 100
seconds, respectively. The latter feature was produced by two separate ruptures in the earth
crust. The Tohoku earthquake was generated by consecutive rupture of many crustal pieces.
Aftershocks continued for months. Figure 3 suggests that the number of events came back
to the original (background) level about 200 days after the main shock. During the aftershock
period, there was a rumor that another catastrophic quake was coming to Tokyo Region. In
such a situation, it was difficult to initiate a reconstruction plan of the community except
emergency activities.

3 GEOTECHNICAL DAMAGE

Because many reports have been published on damage, this chapter introduces only typical
ones that facilitate the readers’ understanding of the following chapters.

3.1 Tsunami
It is obvious that tsunami on March 11th was the major cause of catastrophic damage. Its height
exceeded the previous expectation and overtopped many sea walls. The tsunami came several
times (Figure 5). It appears that the first tsunami destroyed sea walls at many places and there was
no safety measure during the following tsunami attacks. Note that the sea level did not retreat
prior to the first tsunami, contradictory to what some people believe. The sea wall in Figure 6 was
damaged by in-flow of water that scoured the toe on its land side. After loss of lateral support in
its foundation, the sea wall was easily transported landwards by later in-coming tsunami. This
appears different from the experience in Banda Aceh where harbor structures moved seawards.

4
Figure 5. Sea level change in Onahama Harbor
during the tsunami attack on March 11, 2011
(after Japanese Meteorological Agency). Figure 6. Damaged seawall near the Abukuma
River mouth.

Figure 7. Destroyed community behind seawall

(Tohni-Hongo). Figure 8. Rikuzen-Takata after tsunami washing.

Figure 9. Shrine atop a slope in Onagawa Figure 10. Shrine in Ohzuchi at very low eleva-
where where tsunami stopped in its front. tion tsunami stopped in its front.

Figure 7 shows the situation in Tohni-Hongo where the local community had been
relocated to higher places after the previous tsunami in 1933. Thereafter, a sea wall was con-
structed; its height was higher than the tsunami in 1933. As a consequence, people felt safe
and many of them came back to the coastal area. This was a reasonable decision because most
people lived on fishing. In 2011, the tsunami was higher than the wall and the community
behind the wall was totally destroyed. This is an unfortunate evidence to show that empiricism
(height of previous tsunami) is not fully reliable under rare catastrophic natural disasters.
Overtopping occurred in many other towns as well (Figure 8).
The risk of catastrophic but rare disaster may be more reasonably assessed by non-engineer-
ing studies. Along the Pacific Coast of Tohoku region, there are many shrines (temples of Jap-
anese gods). It is likely that those shrines were destroyed by past tsunamis and that local
people reconstructed them at safer (higher) places. Consequently, the tsunami run-up in 2011
stopped in front of the shrine at many places (Uda et al., 2012; Matsuura, 2013). Figure 9 is

5
one of the examples for this. Since the tsunami height varies with the submarine topography
(bathymetry), it was very low at some places. Figure 10 is an example of such a place where
the shrine at low elevation survived the disaster. These facts imply the importance of historical
knowledge that may not appear scientific from the modern technological viewpoints. Note,
however, that Yamazaki (2013) together with Endo and Mazereeuw (2015) attributes the
shrine location to other historical/political and cultural reasons.

3.2 Coseismic subsidence due to tectonic action

It is well known that gigantic earthquakes in the subduction zone are triggered by
rebound of continental tectonic plates after rupture in its interface with the ocean plates.
The rebound results in the uplift of the sea bed and triggers tsunami. At the same time,
the inland part of the continental plate subsides because the rebound-induced elongation
in the horizontal direction is associated with vertical contraction; recall Poisson ratio.
Figure 11 is an example of such coseismic subsidence in Banseki-Ura near Sendai where
the ground subsidence exceeded 1m. Similar subsidence occurred during the 1946 Nankai
earthquake of Mw=8.4 in Japan (Kawasumi and Sato, 1949), the 1960 Chile earthquake
of Mw=9.5 (Housner, 1963) and the 1964 Alaska earthquake of Mw=9.2 (Plafker, 1969).
Figure 12 exhibits the post-earthquake situation of Onagawa Nuclear Power Plant that
was situated at 14.8 m above the sea level. This height was sufficient to survive the tsu-
nami inundation but the tsunami water came into the power plant through an under-
ground conduit into one of the excavated basement spaces. Fortunately, the water was
stopped by a water-tight door at the entrance of the reactor building. Noteworthy is that
the power plant site had subsided one meter by the tectonic mechanism, without which
the plant would have been safer.
The tectonic subsidence together with the loss of sea wall due to tsunami caused fear of
safety against high sea wave during the typhoon season of 2011. It was fortunate that no
typhoon came to the Tohoku region in that year. Figure 13 shows the time history of recovery
of ground level after the earthquake. More profound subsidence occurred in the central part
of Tohoku region (Miyako, Taro, Miyagi Yamoto, Onagawa and then Iwaki), while less sub-
sidence occurred at far places. It is interesting that the rate of recovery is very slow in contrast
to the experience after the 1946 Nankai earthquake after which the ground level recovered
substantially within half year (Sect. 16.10 in Towhata, 2008).

3.3 Subsoil liquefaction

Japanese Geotechnical Society and the Kanto Regional Development Bureau of the Ministry
of Land, Infrastructure, Transportation and Tourism jointly investigated the location of lique-
faction in the Kanto Region and the result is shown in Figure 14 in which red dots account
for the sites of liquefaction. It is evident that liquefaction occurred in two kinds of geomorph-
ology which are young man-made islands along the coast of Tokyo Bay and abandoned river
channels and former lakes along the Tone River and other big rivers.
There are many man-made islands in the Tokyo Bay. They were constructed by dumping
either hill sand or dredged sea-bed sand mostly in the second half of the 20th Century.
Figure 14 indicates that liquefaction was significant in those islands between Tokyo and
Chiba while not much happened in the eastern coast of the Bay (south of Chiba) and between
Tokyo and Kanagawa on the western side. One reason for the “good” performance in the
eastern coast is the lack of information. Those “good” islands are industrial areas and factor-
ies therein do not release “negative” information nowadays. The other reason is that industries
had improved soil conditions in their factories after the 1987 Chiba-ken-Toho-oki earthquake
when liquefaction was significant in industrial islands (Kotoda and Wakamatsu, 1988). Con-
sequently, liquefaction in 2011 was concentrated in young residential islands. In contrast,
liquefaction hardly occurred in aged subsoil regions except young backfill soil of buried
pipelines.

6
Figure 11. Tectonic subsidence in Banseki-Ura Figure 12. Onagawa Nuclear Power Plant that
near Sendai. marginally survived the tsunami-induced disaster.

Figure 13. Slow recovery of ground elevation after tectonic subsidence in 2011 (after Earth Observation
Network, Geospatial Information Authority of Japan).

Figure 14. Sites of liquefaction in Kanto Region (Towhata et al., 2014).

7
Figure 15. Floating of underground parking
Figure 16. Floating of sewage manhole (Itako,
structure (Yokohama).
Ibaragi).

Figure 17. Floating of sewage pipeline in Figure 18. Tilting of house caused by subsoil liquefac-
Itako. tion (Kashima, Ibaragi).

Figure 19. Seismic damage of residential land in the suburban hill area of Sendai.

Figure 15 illustrates floating of an underground car parking. Figure 16 shows floating of a

manhole and Figure 17 indicates floating of a sewage pipeline. Thus, the vulnerability of
sewage lifeline network became evident at many places. Those buried structures that were
lighter than unit weight of liquefied sand floated and stopped functions. Tilting of house was
repeated as well due to softening of liquefaction-prone subsoil in the foundation. The situation
in Figure 18 is the worst and other affected houses exhibited much smaller tilting without
structural problem. However, it is nowadays known that tilting of as small as 0.6% causes

8
Figure 20. Crest subsidence and lateral spreading Figure 21. Relation between levee height and
of Abukuma-River levee (Sasaki et al., 2012). subsidence (Sasaki et al., 2012).

Figure 22. Reason for the maximum crest subsidence not exceeding 75% of the levee height.

Figure 23. Consequence of liquefaction in levee body (Naruse River, Miyagi Prefecture).

such health problems as headache and dizziness to residents (Kitahara and Uno, 1965; Keino
et al., 2011; Kohiyama and Keino, 2012; Kohiyama et al., 2012; Yasuda, 2004).

3.4 Collapse of residential land in hilly area

Sendai area developed its suburban residential land in hilly areas by cutting hills and filling val-
leys, thus producing level ground. This is in clear contrast with the situation in Kanto Region
(Figure 14) where man-made islands were used for new residences. After completion, ground
water in hills infiltrated into new fills and increased the weight of soil. Moreover, compaction of
fill was insufficient to resist the seismic load and that the shear resistance in the interface between
fill and the original valley surface was, most probably, poor. Accordingly, the fill parts collapsed
during the earthquake (Figure 19a) and the value of the real estate was lost. Conversely, the cut

9
part in the vicinity was intact (Figure 19b). The problem is that the ground surface is flat after
construction and people cannot capture the original condition (fill or cut) of the land when they
purchase it.

3.5 River levee

Significant subsidence occurred in many sections of river levee in both Kanto Plain around
Tokyo and Sendai Plain (Figure 1). Figure 20 shows the induced subsidence and lateral spreading.
According to the damage summary, the extent of subsidence (S) did not exceed 75% of the ori-
ginal levee height (H) (Figure 21), which is consistent with the previous summary of seismic
damage of levees since the 1891 Nobi earthquake of M=8.0. The mechanical reason for the upper
bound of 75% subsidence is schematically illustrated in Figure 22; compare the unit weight of
water-saturated liquefied subsoil (typically 20 kN/m3) and that of levee with less degree of satur-
ation (typically 15 kN/m3). Because of the support provided by the bearing capacity of subsoil,



Buoyancy 20 kN=m3
SðmÞ5Gravity 15 kN=m3
HðmÞ ð1Þ

The weight of the levee and the buoyancy in its submerged part reach equilibrium at the
subsidence (S) equal to 75% of the original levee height (H). 75% subsidence does occur in the
worst case in which liquefaction occurs from the ground surface and lasts for a sufficiently
long time to allow for the maximum possible extent of ground deformation.
A peculiar mode of levee deformation is shown in Figure 23. At this site, the levee crest subsided
significantly, while two shoulders of the slope maintained their original height. Moreover, the local
soil condition at the surface in this area is clayey and there is no evidence of liquefaction on the
ground surface (Figure 23b). This phenomenon is attributed to liquefaction inside the levee body.

4 ASSESSMENT OF LIQUEFACTION RESISTANCE OF SANDY SUBSOIL

4.1 Ageing of liquefaction resistance of sand

The Tohoku earthquake caused liquefaction so vastly that concern on the validity of existing
liquefaction code arose. Ishihara and Sasaki (2012) examined the Highway Bridge Design
Code by comparing calculated risk of liquefaction and reality. It was indicated by them that
the existing code is reasonable with some safety margin. The margin means that the code
slightly underestimates the liquefaction resistance of aged soil. The author believes that this
conservatism is important in design of important infrastructures. Noteworthy is that liquefac-
tion risk of private residential land attracted concern at the same time (Figure 18) and people

Figure 24. Distribution of liquefaction only in Figure 25. Two similar bore hole profiles at sites
recent man-made island, Urayasu, Chiba. with and without liquefaction.

10
cannot afford conservative ground improvement. In this regard, more precise assessment with
less safety margin was desired.
The author paid attention to the age effect on soil. Because of this effect, liquefaction
occurred in young man-made islands only (Towhata et al., 2016a). Figure 24 shows the distri-
bution of liquefaction in Urayasu City to the east of Tokyo where onset of liquefaction was
limited to very recent man-made islands. Figure 25 compares two similar bore hole logs; one
site was located in a man-made island and suffered from liquefaction, while the other in a nat-
ural and aged alluvium did not. It was supposed therefore that liquefaction resistance should
be assessed by considering the age of soil that is not accounted for in bore hole data.
Tokyo and its surrounding area is an ideal place for study of ageing because locations of
liquefaction during the 2011 earthquake are well recorded, many borehole data is available
within the area, earthquake motion records are available at many K-Net sites, and the history
of land reclamation since the end of 16th Century is known. By using this information
together with the formulae of the Highway Bridge Design Code and Recommendations for
Design of Building Foundations by AIJ, the factor of safety against liquefaction, FL, was cal-
culated and compared against
the age of subsoil. The aim was to determine the border factor of safety between liquefac-
tion and no liquefaction which is theoretically equal to 1.0 but may vary with soil age.
The difficulties of the ageing study were as what follows;
1. It is significantly uncertain which soil layer liquefied. Sand ejecta and other surface mani-
festation of liquefaction do not specify the depth of liquefaction.
2. It is possible that liquefaction occurred at great depth and did not produce surface
manifestation.
3. Because the majority of K-NET earthquake monitoring stations are situated on unliquefied
sites, no actual ground motion is known at liquefied sites.
4. Sites of K-NET stations and those of analyzed borehole data are not necessarily identical.
The FL values have to be calculated by using the nearest available K-NET motion data.
5. Alluvial deposits are generally assumed to be older than reclaimed manmade islands. How-
ever, in the extreme case, soil deposited in the very recent past.
6. Reported liquefaction may be the event in recent backfill of lifelines within older natural
deposits.
These difficulties were solved as described in detail in Towhata et al. (2016a) and this paper
introduces the most important idea to solve the above-mentioned uncertainty (1). At sites where
surface manifestation of liquefaction was reported (Figure 26a), the factor of safety was calculated
in the vertical direction and the minimum factor of safety, FLmin, was determined. The border
value is greater than FLmin because, otherwise, the factor of safety would be greater than the
border value throughout the depth and no liquefaction would have occurred. This is obviously
against the liquefaction report. Thus, FLmin is the lower bound. At sites without report of lique-
faction (Figure 26b), on the contrary, the maximum value in the FL profile is taken and considered

Figure 26. Idea of lower and upper bound of border factor of safety.

11
Figure 27. Temporal variation of border factor of safety between liquefaction and no liquefaction.

Figure 28. Increase of liquefaction resistance of soil with age under Japanese geological environment.

to be the upper bound of the border value; FLmax. This value is the upper bound because, other-
wise, FL value would be less than the border value throughout the profile and liquefaction should
have occurred extensively, which is obviously contradictory against what happened in reality.
The obtained FLmin and FLmax were plotted against the soil age (Figure 27). Note that several
more corrections were made in the FL calculation. For details, refer to Towhata et al. (2016a). In
Figure 27, the upper bound at sites without liquefaction is plotted by open symbols and the
border between liquefaction and no liquefaction is below them as illustrated by downward
arrows. In contrast, the lower bound at liquefied sites is plotted by solid symbols and the border
is above them, as illustrated by upward arrows. The most likely border is between upward and
downward arrows and is judged to be somewhere in the shadowed range. It is important that this
shadowed range goes down with increasing soil age, suggesting that less liquefaction happens in
aged soil in spite of the factor of safety less than 1.0. This means that actual liquefaction resistance
is greater in aged soil than what is employed in the current design codes.

12
Figure 29. Difference in earthquake activities in Japan and USA (comparison of number of earth-
quakes per million km2) (Towhata, 2018).

Because the history of Tokyo started at the end of the 16th Century, older data was not avail-
able. This shortcoming was supplemented by studying Tsukiji area in Amagasaki City near
Osaka and Kyo-ga-shima of Kobe. The former site was constructed at the beginning of the 17th
Century and liquefied totally during the 1995 Kobe earthquake. The latter island was con-
structed at the end of the 12th Century and did not liquefy during the 1995 Kobe earthquake.
For both, borehole data and earthquake motion records at nearby sites were available. Accord-
ingly, the ageing effect in liquefaction phenomenon of sand was quantitatively captured for the
time period of about 1000 years.
Figure 28 demonstrates the increase of liquefaction resistance with age, making the resist-
ance of very recent soil being equal to unity. The results of the present study is shown by
shaded rectangles whose size corresponds to the uncertainty in Figure 27. Basically the trend
is consistent with results of other studies but may suggest more profound ageing in Japan than
in North America. Towhata et al. (2016a) experimentally studied the mechanism of ageing
and concluded that slow progress of grain movement (dislocation, change of fabric structure,
change of force transmission among grains) is the cause of ageing, It is suggested that grain
dislocation is promoted by minor earthquake shaking which has been called preshaking effects
in other studies (Wichtmann et al., 2005). If this hypothesis is right, higher earthquake activ-
ities in Japan accounts for more progress of ageing than in North America (Figure 29).

4.2 Assessment of liquefaction vulnerability in residential land

The 2011 Tohoku earthquake triggered liquefaction in many residential lands. After the disas-
ter, it became necessary to provide appropriate methodology to assess liquefaction vulnerability
in residential land so that people and local community officers might be able to briefly assess
the risk and undertake mitigation measures, if necessary. The author believed that conservatism
should be eliminated from the methodology because people cannot afford cost for additional
safety. To avoid forcing additional safety cost to people, the factor of safety, FL, was calculated
by using nearby ground motion records and considering the ageing effect (40% increase in resist-
ance after 400-year age) based on many bore hole data that was obtained from both sites of
liquefaction and no liquefaction.
An important issue was that minor liquefaction damage should be tolerated because light tilt-
ing, subsidence and others can be repaired at cost lower than soil improvement to the depth.
Hence, three extents of house damage were defined and are illustrated in Figure 30 in which “A”
means no damage, “B” are minor and “C” should have been avoided by soil improvement. At
each site, thereafter, the factor of safety against liquefaction, FL, was calculated along the bore
hole log for further examination (Towhata et al., 2016b). The point of discussion was that the
house damage is a consequence of the overall soil behavior and the FL value at a single depth
does not account for it. Hence, two indices (PL and Dcy) for the overall behavior were calculated;
Z 20
PL ¼ F ðFL Þð10  0:5zÞdz ð2Þ
0

13
Figure 30. Three extents of liquefaction-induced damage to house foundation (Towhata et al., 2016b).

Figure 31. Significance of surface unliquefied crust resisting against punching failure of house foundation.

in which z is the depth in meter (Highway Bridge Design Code by Japan Road Association,
Tatsuoka et al., 1980) and
 
1  FL ðFL  1Þ
F ðFL Þ ¼ ð3Þ
0 ðFL 41Þ

while Dcy is vertical integration of cyclic shear strain amplitude, which is assessed by data of
standard penetration tests and seismic load and is considered to be equivalent with the displace-
ment amplitude at the ground surface (Recommendations for Design of Building Foundations by
AIJ; Tokimatsu and Yoshimi, 1983; Tokimatsu, 1997). In addition, the thickness of the surface
unliquefiable layer, H1, was taken into account. This parameter stands for the bearing capacity of
surface soil that resists against the punching failure caused by the weight of houses (Figure 31).
Initially, it was difficult to obtain good correlation between the damage classification in
Figure 30 and the combination of assessed indices (H1, PL) or (H1, Dcy). This problem was
overcome by examining the precise location of boreholes and the sites of reported damage
which might be sometimes far from each other. In other extreme case, the reported depth of

14
Figure 32. Correlation between damage extent of houses due to liquefaction and subsoil indices; B1-B3 are
named after location in the graph. There is no difference in damage extent from what was shown in Figure 30.

Figure 33. Sand ejecta from the middle height

of a levee (Yoshida River, Miyagi Prefecture). Figure 34. Mechanism of liquefaction inside a levee.

ground water table was 6 meter (H1 = 6m) in a man-made island where the ground elevation
is only 4 meters above sea level. This unrealistic depth of ground water was reported during
borehole investigation probably when ground excavation was going on near the site and the
ground water was pumped up for temporary purposes.
After detecting and thus removing irrelevant data from the study, a reasonable correlation was
obtained between damage extents and the assessed liquefaction indices; no solid symbol in “A”
category (Figure 32). These diagrams are used for brief evaluation of the quality of residential
land subject to liquefaction risk (Towhata et al., 2016b). The difference between the present study
and the well-known H1-H2 diagram by Ishihara (1985) is the replacement of H2 by PL or Dcy.
The author felt that H2 parameter is determined simply by whether FL is greater than or less than
unity near the ground water level and might be too sensitive to minor difference in the original
SPT data. Also, the mechanical role of H2 layer is unclear. In contrast, PL and Dcy are not so
sensitive and expected to give stable conclusion. Although H1 is sensitive as well, its mechanical
role is clear as played in the resistance against punching failure. Hence, H1 remains here.

5 VULNERABILITY OF RIVER LEVEES TO LIQUEFACTION

Sasaki et al. (2012) summarized the seismic damage of river levees in 2011 and indicated that
most damage was related with liquefaction. Although traditional soil mechanics states that
embankment has to be compacted for stability, the levees that are situated upon very soft soil
cannot be compacted during construction because the sensitive subsoil may be disturbed and
fail due to the impact. Consequently, the levee soil remains loose. This situation is aggravated
by infiltration of rain water into the levee body. Consequently, a body of loose water-satur-
ated sand (silty sand) is formed inside the levee. During the 2011 event, liquefaction occurred

15
Figure 35. Cross section of Tone River levee at Iijima Site.

Figure 36. 1-G shaking model of liquefac- Figure 37. Time history of horizontal shaking in tests
tion-prone ground with vertical columns for on mitigation effects of vertical group columns (10 Hz
restriction of cyclic shear strain. for 15 seconds).

inside the body of levees despite that the local natural subsoil was clay and unlikely to liquefy
(Figure 23). Figure 33 shows sand ejecta from the middle height of a levee. This is an evidence
that liquefaction is possible to occur at elevations higher than local ground water level.
Figure 34 schematically illustrates the mechanism of liquefaction inside a levee. The water is
supplied into loose sand in the levee either by infiltration of rain water or from the surround-
ing subsoil into the subsided part of the levee. Note that levee is often constructed upon soft
clayey soil along a river channel; the clay is unlikely to liquefy but its consolidation settlement
leads to submergence of loose levee bottom that comes down below the water table.
The 2011 Tohoku earthquake caused liquefaction in the levee body at many places.
Although detailed risk assessment was insisted on after the disaster, the lack of information
made it difficult. First, local geomorphological information does not help liquefaction risk
assessment because the problem lies in the levee. Second, the soil condition inside the levee is
hardly known due to the long history of levee construction. Along major rivers, the earliest
and lowest levee was constructed before the 19th Century and the height has been increased
several times without construction records available. Figure 35 illustrates a cross section of
Tone River levee that was excavated for a research purpose. Note that the levee consists of
many layers of past construction. More detailed investigation has to be conducted inside levee
at a reasonable cost. Geophysical attempts of combined use of S wave velocity and electric
conductivity to detect liquefaction-prone sand in a levee could not achieve desired resolution.
In this regard, cone penetration and other probing are promising and desired.

6 INNOVATIONS IN MITIGATION OF LIQUEFACTION DISASTER

6.1 Use of group vertical columns on mitigation of liquefaction effects

In view of the significant liquefaction effects on our community, a numerous number of mitigation
technologies have been investigated in the recent years. This chapter addresses some of them.
They are characterized by reduced cost while allowing for a limited extent of ground deformation.

16
 Two kinds of soil deformation associate liquefaction of subsoil. The one is the cyclic shear
deformation during shaking which leads to high excess pore water pressure if its magnitude is
profound. The other is the permanent deformation or flow failure that remains after shaking
and is perceived as “damage” if significant. In the present study, it was aimed by a group of
piles/columns in liquefaction-prone subsoil to reduce the cyclic deformation to avoid liquefac-
tion and, even if liquefaction happens, to mitigate lateral flow.
Figure 36 shows a 1-G shaking model test in which model sandy ground with columns was
shaken in a laminar box (Bahmanpour et al., 2019). The idea herein was that columns reduce
cyclic strain during strong shaking, expecting that columns, which are constructed at cost lower
than underground grid walls (Figure 58) and full grouting, reduce the damage at an appropriate
cost. After the stage of Figure 36, sand was poured into the container and the entire model was
submerged in water. Then the heads of the columns were connected by a template to prevent turn-
over of columns during softening of subsoil. Figure 37 presents a typical time history of base
acceleration.
Figure 38 shows the variation with time of excess pore water pressure that was measured
among columns near the bottom of the model ground (60 cm below surface). The improve-
ment ratio of AR=35% is defined by the cross section of columns divided by the horizontal
area of ground. Although the pore pressure development ratio (max. pressure divided by the
initial effective overburden pressure) is similar whether with or without columns, the rate of
pore pressure rise (before 10 seconds of time) was reduced by columns and the extent of pres-
sure dissipation after the peak value was accelerated. These two features make shorter the dur-
ation of adverse time period in which pore water pressure is very high. In Figure 39, shaking
was repeated on the same model ground by increasing the amplitude of acceleration while
maintaining the shape of the time history (Figure 37). It is shown that pore pressure was
reduced by increasing the area ratio (closer spacing among columns).
Further study was made of the effects of columns on reduction of lateral soil flow as a con-
sequence of liquefaction (Towhata et al., 2015a; Takahashi et al., 2016). The major point of
concern was the benefit of irregular configuration of columns that maintains the same number
of columns and the same cost of construction but achieves better reduction of liquefaction
effects. In Figure 40, the conventional square and triangular column configurations allow rela-
tively free cyclic shaking of soil and easier flow of soil among columns in case of liquefaction.
In contrast, the “irregular” configuration does not have such free direction of motion and is
expected to reduce both cyclic strain amplitude and flow displacement. The irregular configur-
ation is composed of square unit geometry and should not be confused with a “random”
geometry that is difficult to install in field practice.
A series of centrifugal model tests (50G) have been conducted at the Port and Airport
Research Institute in order to verify the mitigative effects of the irregular configuration.
Figure 41 illustrates one of the tested models and Figure 42 shows an example of input shaking.
Figure 43 compares the distribution of lateral displacement. In this figure, Case 0 had no
column and the maximum input acceleration was 150 cm/s2. In contrast, Case 3 employed the

Figure 38. Time history of excess pore water Figure 39. Effect of area ratio of columns on
pressure with and without columns. reduction of excess pore water pressure.

17
Figure 40. Plan view of three column Figure 41. Centrifuge model test on irregular col-
geometries. umns configuration embedded in a liquefiable slope.

Figure 42. Example of input acceleration for Figure 43. Effects of vertical columns on reduced
tests on irregular column configuration. lateral flow of liquefied slope.

Figure 44. Excavated shape of columns with head connection (full-scale model, after Dr. Naoki Takahashi).

column spacing = 2.5*column diameter (12.6% improvement) and the input shaking was as
strong as 225.8 cm/s2. In spite of this strong shaking, the columns successfully reduced the
lateral displacement. Note that the difference between tests with square and irregular config-
uration was not significant. Furthermore, the head of columns has to be connected together in
order to prevent turnover of the entire column group (Figure 44).

6.2 Extensive installation of drainage pipes

Because of the significant liquefaction effects on houses (Figure 18), need was strongly felt for
simple soil improvement under existing houses. Attention was paid, accordingly, to such technolo-
gies as injection of air bubbles into subsoil (Okamura et al., 2011) and sheet pile walls around indi-
vidual house foundation (Yasuda, 2013a & b). In this regard, Razouli et al. (2016) studied the
possibility of installed drain pipes that dissipate the excess pore water pressure quickly under exist-
ing houses.

18
Figure 45. Schematic diagram of typical experiments and position of drains.

Figure 46. Base input motion (a) the whole shaking of 10 Hz and 300 cm/s2; (b) the early stage of shak-
ing for considering weaker earthquakes (equivalent 15 cycles).

Figure 47. Recorded time histories of house Figure 48. Recorded time histories of house settle-
settlement with different mitigation measure(no ment with different mitigation measures (with sheet
sheet pile wall). pile wall).

Figure 45 demonstrates one of the tested models in which a house model was placed at the
center and three kinds of mitigation were installed; diagonal drain pipes at shallow (SD) or
deep (BD) elevations, sheet pile walls around the house foundation with vertical drains and
lowering of ground water level. Because of the existing house at the surface, drains have to be
installed in the diagonal directions. In the series of tests, three technologies were investigated
separately or in combination. Figure 46 plots the input acceleration time history. Because the
study was made after the gigantic earthquake of magnitude = 9, the number of shaking cycles
is many. To take account of earthquakes of smaller magnitude as well, test data will be inter-
preted at 16 seconds after 15 cycles of 300 cm/s2 as well as at the end of the whole shaking.
Figure 47 compares the time history of house settlement with different kinds of mitigation
measure. The most effective mitigation was lowering of ground water by 10 cm which is about

19
25% of the house size. Noteworthy is that lowering of ground water level (GWL) for a single
house requires underground walls around the house which increases the cost in practice. If
lowering is executed for an entire community, care must be taken of possible consolidation
settlement in cohesive subsoil. For this reason, the conclusion from this figure is that the com-
bination of drains (SD, BD and vertical) without GWL lowering is the most effective. Only
vertical and BD drains without GWL lowering was not effective at all. Further comparison
was made in Figure 48 in which cases with sheet piles (SP) around the house foundation are
plotted. In this figure, SP with vertical drains attained the best performance; see Case 15 in
which settlement is small in spite of shallow GWL (-5cm).

7 PROTECTON OF EXISTING RESIDENTIAL AREAS FROM FUTURE

EARTHQUAKE DISASTER

7.1 Tsunami
One of the lessons learnt from the 2011 tsunami disaster was that sea wall is not a fully reliable
safety measure. If tsunami is higher than wall, overtopping results in a catastrophic situation.
Therefore, the new principles of safe community are “evacuation” and “high ground level” that
are combined with reasonably high sea wall. The high ground elevation prevents catastrophic
inundation in case of extremely high tsunami, while sea wall stops more frequent lower tsunamis.
Figures 49 and 50 illustrate the on-going reconstruction of Rikuzen-Takata town; see
Figure 8 for the damage. In Figure 49, a new town center together with the new town hall
have been relocated to the top of a hill. Following this stage, re-construction of the former
town area, which was inundated by tsunami water, resumed by elevating the ground level to
9-12m above sea level. Relocation and elevating are executed over an area of 186.1 ha. Its
budget is 53.69 Billion Yen for population of 1450 families and 4300 people (Rikuzen-Takata
City, 2016). Thus, the cost is approximately 34.4 million Yen/family or 12.5 million Yen/
person (1 Euro = 134 Yen on Dec. 10, 2018). Moreover, the railways in this region that were
washed away by tsunami are reconstructed as a “bus rapid transit” (Figure 51) in which bus
can run on the designated road and the maintenance cost is lower than that of railway.

7.2 Liquefaction in residential district

Tens of thousands of existing houses were affected by subsoil liquefaction in 2011 because they
were situated on young loose cohesionless deposits. Since the effect was as vast as that of tsunami
damage (Section 7.1), similar public supports for reconstruction of the liquefaction-prone com-
munity was considered necessary. However, the public principle is that the damage restoration of
private properties is the responsibility of individual people and that the public money should not
be spent on it. Moreover, the previous regulation assumed the responsibility of seismic safety of
residential land to the owners who purchased the land; not to the developers or the contractors.
After a long negotiation with the financial ministry, this difficulty was overcome by
pointing out that liquefaction in private lands affects the public properties such as roads and
lifelines (Figure 52). Because liquefaction in subsoil occurred uniformly irrespective of the

Figure 49. New town center at top of hill Figure 50. Raising ground level by 15 m in former
(Rikuzen-Takata, March 2017). town area (Rikuzen-Takata).

20
Figure 51. Reconstruction of local public trans- Figure 52. Heaving of road induced by subsid-
port by bus rapid transit (Kesen-numa). ence in private land (Urayasu, Chiba Prefecture).

border between private and public properties, it was decided that liquefaction mitigation
works in private land should be installed by 50% public and 50% private expenses. Because
people are not familiar with technical issues, the ground improvement projects have been con-
ducted under the initiative of local governments. Another point is that the ground improve-
ment has to be conducted on a community basis in which tens or hundreds of families in town
blocks unanimously agree to join the project. This requirement was found very difficult to ful-
fill later. Note that relocation of liquefaction-prone community to safer places, which was the
case in Christchurch of New Zealand, was not practiced in Japan because people preferred to
stay in the same place and local governments did not want to lose tax payers.
Several municipalities started to prepare for this public-private framework. However, some
people did not want to spend money on ground improvement because they could not afford
it, they were afraid of future maintenance cost (operation of ground water pumping in case of
lowering water level) or they had already installed expensive mitigation measure under their
own houses.
The choice of liquefaction mitigation technology was entrusted by people to local govern-
ments. Under this situation, only those technologies that had been verified or had been used
in practice were eligible. In other words, technologies under development or research such as
those referred to in Chapter 6 had to be ruled out. Accordingly, only two remained which
were ground water lowering and installation of underground grid walls.
Ground water lowering was previously employed in Tsukiji District of Amagasaki City which
is the western neighbor of Osaka. The 1995 Kobe earthquake caused liquefaction here and most
houses were demolished. Earth was then filled by 1.5 m on the surface and pipes were installed
to drain ground water (Suwa and Fukuda, 2014). Earth filling was a compensation for the long-
term consolidation settlement that used to be serious due to industrial water pumping until
1960s and had made the ground surface even lower than the sea level. Since industrial pumping
was banned in 1960s, the ground water level had been rising, making the local soft clay over-
consolidated. Because of this, earth filling and limited lowering of ground water level did not
cause consolidation settlement (Figure 53). The liquefaction mitigation project in Tsukiji was
completed in 2004 which is 9 years after the Kobe earthquake (Amagasaki City Government,
2007). After 2011, several municipalities successfully achieved unanimous agreement of people
on ground water lowering and the measures were installed (Figures 54 and 55); refer to Itako
City Government (2013) and Kokusho (2014) as well as Yasuda and Hashimoto (2016).

7.3 Liquefaction in Urayasu City, Chiba Prefecture

9154 houses were affected by liquefaction in Urayasu City. Most parts of this city (Figures 14
and 24) were constructed in 1960s and 70s by dredging of marine bed and reclamation of the
dredged silty sand in shallow sea water. The liquefaction resistance of this sand was obtained

21
 Figure 54. Hinode District of Itako City where
Figure 53. No settlement problem in Tsukiji ground water is drained by gravity flow through
of Amagasaki after ground water lowering. installed pipes.

Figure 55. Discharge of ground water into river in Kuki City (photo provided by Prof. J. Koseki).

Figure 56. Liquefaction resistance of undis-

turbed samples collected at two places in Urayasu
by gel push sampling developed by Mr. K. Sakai Figure 57. Consolidation settlement in recent
of Kisojiban Consultants, Tokyo. land reclamation of Urayasu.

by using undisturbed samples collected by gel push sampling (Figure 56). Because the area is
on the mouth of big Tone River, whose channel was here until new channel was constructed
in late 17th Century, the man-made island is underlain by 40-meter deposit of soft alluvial
clay. Hence, the man-made island in the city has been annoyed by consolidation settlement
for decades (Figure 57). The memory of settlement is still vivid among people and repetition
of settlement due to ground water lowering was not appreciated. Under the democratic frame-
work, therefore, the city government could not pursue ground water lowering.

22
Figure 58. Conceptual illustration of the role Figure 59. Narrow space between two adjacent
played by underground grid wall. houses.

Figure 60. Jet grouting by a small machine

placed on an elevated platform (August, Figure 61. Construction of underground wall under
2018). street by mechanical mixing.

The alternative technology was installation of grid-type underground walls that were con-
structed by mixing cement and local soil by mechanical mixing or jet grouting. Certainly, this
option was more expensive than ground water lowering. Figure 58 schematically shows the
grid wall that are installed in a narrow space among houses. The rigidity of the grid wall
reduces the cyclic shear deformation of subsoil during earthquakes and mitigates the risk of
liquefaction. This technology became popular after its success in Kobe Harbor in 1995. Never-
theless, Urayasu was its first application to subsoil under “existing” houses that are fragile
and do not allow very small distortion or tilting that affect health of residents. Very narrow
space for underground wall construction was another big issue (Figure 59). After discussion,
it was decided to select the grid wall as the measure and design started. The goal of design was
“little damage in houses under the same strong earthquake as the one in 2011”. It was not the
goal to insure zero damage in any future earthquakes.
Accordingly, the cost evaluation on ground improvement in the private land was such as 6
million Yen per house. As stated before, its 50% was paid by the national government and the
local government offered supports of about 17% so that the people’s payment might be not
more than 33%. Based on this condition, many meetings were held between the administration
and the residents, aiming at unanimous agreement of people in 16 districts (4103 houses) in
the city. Finally, three districts with 471 houses (only!) reached unanimous agreement and pre-
liminary works started on sites. Figure 61 illustrates a local street under which wall was suc-
cessfully constructed by powerful mechanical cement mixing.
The works, however, had to be stopped in April, 2017. The first reason was non-uniform
soil condition in the city. Jet grouting did not work well in clay parts. The second reason was
detecting many embedded plastic drains that were not expected to be there during design.
They had been installed at an interval of several meters for promotion of consolidation settle-
ment probably in 1970s when the land reclamation took place. Rotating jet grout machines
caught drains (Figure 62) and grout could not be injected into subsoil. Consequently, stable
and continuous grid wall was not able to be constructed.

23
Figure 62. Plastic drain caught by rotation of Figure 63. Successful grouting shown by solidi-
jet grouting machine. fied soil cores.

Nowadays, construction engineering is not respected by people. Its public image is far from
that of the advanced science and technology. Moreover, media frequently reports construction
related with political corruption. The author has been advocating that more efforts are required
to improve the public image. In this regard, what the Urayasu project had to do was to over-
come the technical difficulty irrespective of the possible increase of cost and keep the promise
with people. Therefore, the author, who was the Chair of the technical advisory committee of
the project, visited the sites more than 30 times in summer of 2017, free of charge, and encour-
aged engineers to overcome the problem. Consequently, it was decided to increase the grouting
pressure, run the jet grouting procedure twice at sites and inspect possible drain effects so that
drains might be removed from the machine and grouting might be run one more time. After this
decision, jet mixing worked well and Figure 63 shows a core sample of solidified clay that was
continuous throughout the clayey subsoil. The negative effect of the solution was elongated con-
struction time and cost. The national government decided to provide additional financial sup-
ports and people were not required to spend more expense than the original plan.
Considering the elongated project period, people’s unanimous agreement was required
again in early 2018 and the rate of agreement dropped to 78%. Although this still meant that
absolute majority supported the project in the three districts, the project had to be cancelled in
two districts with 393 and 45 houses, respectively. Only one district with 33 houses accepted
the project and ground improvement was carried out in 2018 (Figure 60). The total construc-
tion cost in this district is 1.147 Billion Yen for both streets and private lands. The difficulty in
other districts was that people did not want to live in “construction site” any more. Further,
seven years had passed since the earthquake and people’s concern had decreased.

8 PROTECTION OF BURIED LIFELINE FROM LIQUEFACTION DAMAGE

The 2011 earthquake and induced liquefaction destroyed embedded life lines at many places.
In the modern times, lifeline is an essential component of urban community and its seismic
vulnerability is a big threat to the resilience of cities in seismically active regions. The author
paid special attention to liquefaction resistance of sewage pipelines because possible change of
the pipe gradient due to backfill liquefaction significantly affects the flow of waste water
under gravity.
The damage of sewage lifelines is classified into that of manhole (Figure 16) and floating of
pipes (Figure 17). Because the mitigation technology for manhole has been investigated by
Orense et al., (2003), Kiku et al. (2006), Tobita et al. (2011) and Yoshida et al. (2008) among
others, the author put emphasis on existing embedded pipelines (Otsubo et al., 2016a and b).
It was attempted therein to avoid overall excavation and improvement of liquefaction-prone
backfill soil around the pipe and thereby to reduce the cost.
mitigation.
Figure 64 illustrates one of the innovations by which floating of pipe is prevented by insert-
ing a “horn” between the pipe and the surface pavement. Herein, the backfill soil is supposed

24
Figure 64. “Horn mechanism” to Figure 65. Input acceleration record in 1-G model tests on
prevent floating of pipe. pipeline.

Figure 66. 1-G shaking model for pipeline tests Figure 67. Details of prevention mechanism by
(pipe diameter=6cm). overlying unliquefied layer.

to liquefy but pipe stays at the original position because floating is prevented by the horn.
Note that this idea requires excavation only at the places of horn installation, which is less
expensive than other measures such as overall excavation and cement mixing of backfill soil,
or placement of better backfill materials.
1-G shaking tests were conducted on prevention of floating of a pipe embedded in liquefiable
backfill sand. Figure 65 shows the input motion which has the amplitude of 3.5 m/s2 and 10 Hz
frequency. The strong acceleration was chosen by considering the social demand after the big
earthquake. The high frequency was employed from the similitude viewpoint. The relative density
of sand was 30%, which produces similar dilatancy as the prototype in spite of reduced stress level.
The overall configuration of a tested model is illustrated in Figure 66. The pipe model was embed-
ded in liquefiable sand and its two ends were connected to the soil container by hinge (manhole
connection in reality). In the middle, the pipe floating was prevented by a column (horn in prac-
tice) and a top plate (square or circular in shape; see Figure 67) which is overlain by an unliquefi-
able layer of gravel or sand (road pavement in practice) or a stationary beam (Case H1 with zero
displacement as the control case). Friction angle of the medium dense gravel, loosely packed
gravel and sand are chosen to be 40°, 35° and 30°, respectively. As illustrated in Figure 67, the
prevention of floating is developed by the weight of unliquefied soil and the shear resistance along
the (punching) failure plane. Herein, values of A and H were varied and many tests were run. The
oblique angle of the failure plane (θ) will be determined by referring to the test results.
Figure 68 shows an example record of the force that developed in the horn structure sup-
ported by a rigid beam for zero floating. By eliminating the high-frequency component, the
maximum uplift force was determined at 74.3 N. Another case was run without horn and the
ultimate floating was 3.1 cm. In case of sandy or gravelly surface layer with horn, the floating
was smaller and the horn was compressed by less extent of force (<74.3N). The maximum pos-
sible force of 74.3 N was compared against the uplift resistance (combination of weight and
shear resistance in the surface layer) in order to assess the factor of safety against floating.
Figure 69 illustrates the relationship between the factor of safety and the reduction of float-
ing displacement in which 100% stands for null floatation. The greater factor of safety results

25
Figure 68. Example time history of uplift resist- Figure 69. Relationship between the factor of
ance measured in hone structure. safety and the reduction of floating.

Figure 70. Flow-in of liquefied backfill

sand into separated sewage pipes (photo-
graph by Urayasu City). Figure 71. Sheath pipe inside a PVC pipe model.

Figure 72. Configuration of model ground for test with Figure 73. Situation inside an embedded
inserted sheath pipe. pipe during shaking.

in more reduction of floatation. In this figure, the factor of safety obtained by θ=’ gives 100%
reduction at the factor of safety = 1.0. Thus, the design of punching resistance is assessed by
θ=’ in the surface layer of road pavement.
One of the frequent damage mechanism of sewage pipeline during earthquakes is the flow-
in of liquefied backfill into pipe through disconnection at pipe joints (Figure 70). To avoid

26
this, slip lining of a sheath pipe inside an existing pipe (Figure 71) seems promising (Matthews
et al., 2013) because sheath pipe prevents flow-in of external sand and provides sufficient
space to flow of water. Sheath is a useful measure also for rehabilitation of aged buried pipes.
Shaking tests were run on the model in Figure 72 where two segments of pipes were loosely
connected so that breakage might be induced there and liquefied sand might flow in.
Figure 73 illustrates the situation inside the pipe that was captured by a video camera during
shaking. Obviously, liquefied sand could not come into the sheath pipe.

9 PROFESSIONAL ENGINEER FOR GEOTECHNICAL EVALUATION

The 2011 earthquake exhibited that many residential lands are prone to seismic risk (Figures 18,
19 and 30). Residential land is vulnerable to further risk due to heavy rain, debris flow and land-
slide as well (Figure 74). People are living there without recognizing the situation. They purchase
land considering only the price, comfort to live and convenience in life. After disasters, they
lament that there was no way to be informed of the risk, although many municipalities provide
hazard maps to residents. In this situation, the Japanese Geotechnical Society established the Jap-
anese Association for Geotechnical Evaluation together with several other institutions in order to
promote the new qualification of geotechnical evaluator (Towhata and Nakamura, 2015b).
The title of “Professional Engineer for Geotechnical Evaluation (PEGE)” is awarded to those
who wish to work for the safety of people and have sound knowledge of geotechnical engineering
as well as practice of residential land construction. Knowledge on local geomorphological history
is important as well. Their basic qualification is verified by professional titles that they have
obtained such as “Professional Engineer”, “Registered Architect” and “Civil Engineering Works
Execution Managing Engineer” among several others. The PEGE title is awarded to those who
have passed additional examination on practice of residential land construction and is valid for
five years, followed by further extension. In early 2018, this title was officially certified by the
national government. Note that PEGE is not a voluntary title. PEGEs are expected to work as
professional engineers who earn reasonable income from clients.

10 ROLES OF GEOTECHNICAL ENGINEERING TOWARDS SOLVING THE

PROBLEM OF FUKUSHIMA NO.1 NUCLEAR POWER PLANT

One of the severest disasters caused by the 2011 Tohoku earthquake was the tsunami disaster
in the Fukushima No.1 Nuclear Power Plant. Note that the present paper calls this plant

Figure 74. Consequence of rain-induced debris flow in Hiroshima, 2014.

27
Figure 75. Piles of plastic bags to store radio-
actively contaminated surface soil around the Figure 76. Super heavy mud water for radio-
power plant (January 15th, 2019). active shielding (study at Waseda University).

“Fukushima No.1” in place of the official name of “Fukushima Daiichi” because “Daiichi”
means “No.1” and better understanding is achieved by the latter. While there is a wide range
of arguments about the future of nuclear power technology, the author understands that the
solution of the Fukushima No.1 problem is of unanimous support from people. It is a
national goal to solve this nuclear disaster and return the area safely to local people.
As is well known, the nuclear disaster was caused by the unexpected height of tsunami. Shak-
ing toppled electric transmission towers and the external power supply stopped. However,
internal power (battery and diesel generator) was supposed to survive the earthquake because
they were placed in the basement where shaking effect was least. Unfortunately, the tsunami
was higher than the design height and the internal power supply was destroyed by inundation in
the basement. Cooling of reactors stopped, the nuclear fuel melted down, the containers of
reactor were damaged, the hydrogen gas leaked out of the containers and got exploded. Fur-
thermore, air, ground water and sea water were contaminated radioactively, local people had to
evacuate and, consequently, the community was lost. Fortunately, other nuclear power plants in
the same coastal area were able to survive. Within a few years after the disaster, it became evi-
dent that the problems to be solved are so complicated that many different fields of science and
technology have to collaborate on this for the next 40 years or more.
Geotechnical engineering is expected to play such important roles as cleaning of contaminated
soils around the power plant (Figure 75), control of radioactivity in and around the foundation of
power plants and final treatment of debris (molten fuel and other radioactive wastes in the des-
troyed reactors) and radioactive wastes. Moreover, removal of debris from the reactors is similar
to excavation or rock coring in geotechnology. The whole efforts to solve the problem is called
decommission. Accordingly, the Japanese Geotechnical Society resumed efforts to provide core
engineers to the decommission of Fukushima No.1 Plant for the next 40 years (Komine, 2017).
One of the technical innovation of the said activity is super heavy mud water (Figure 76). Simi-
lar to well-known bentonite water, this liquid is produced by mixing special clay with water.
Because of its high mass density (exceeding 2.5 Mg/m3), it can shield (reduce the intensity of) γ
ray, while its high water content (50-90% of volume occupied by water) helps shield neutron
beam. Moreover, its fluid nature and low permeability make it possible for the mud water to flow
into small parts of the damaged reactors and stop leakage of radioactive water (Komine et al.,
2016; Yoshikawa et al. 2017). It is noteworthy that this material can be produced and supplied in
a massive scale to meet the needs of real construction project. Thus, the super heavy mud water
will contribute to maintain the radioactive safety in the decommission site for 40 years.
Development of human resource who will be engaged in decommission during the next 40
years is an essential part of the activity. It is aimed therein to produce civil/geotechnical engin-
eers who have reasonable knowledge of nuclear engineering. They need to know well about
radioactivity and structure of nuclear power plants. In 2018, the first version of teaching
material was completed for one-semester subject “geotechnical engineering for decommission”

28
that is to be used in university education. It puts emphasis on construction practice and con-
sists of the following topics:
basics of radioactivity, overall scope of geotechnical engineering for decommission, geomaterials,
geoenvironmental engineering, waste treatment, super heavy mud water, ground water control,
final treatment of nuclear wastes, effects of nuclear accident on human health and brain storming.
All the contents emphasize contribution to decommission. Moreover, a technology map
was made in which information on technologies that facilitate decommission project was col-
lected and engineers can search details according to their needs (Towhata et al., 2018).

11 TREATMENT OF DISASTER WASTE

The earthquake left behind a huge amount of waste that were destroyed buildings, house-
holds, radioactively-contaminated soil etc. (Inui et al., 2012). Prior to treatment, they had to
be classified in accordance with their size and properties. Although detailed discussion is
desired on this issue, there is no more space in this paper.

12 EMERGING PROBLEMS

Civilization changes with time. Life style of people are not same as before. The recent trend is
increased desire of people for safety. Because of this, new earthquake problems become
important one after another.
The 2016 Kumamoto earthquake sequence, Japan, of Mw=6.2 and then 7.0 demonstrated
the difficulty in coping with two consecutive strong earthquake shakings; the first one caused
partial damage in structures and slopes, while the second one 28 hours later triggered their
total collapse. Should we prevent any partial damage during such a rare first event for the
safety during the second event? It is too costly. The same earthquakes produced fault rupture
in a populated area of Mashiki Town.
The 2018 Hokkaido Eastern Iburi earthquake of Mw=6.6 revealed the weakness of private
residential fill in which collapse has been repeated during many past earthquakes. Failure of
slopes that comprised of volcanic ash is another problem. Sapporo City experienced liquefac-
tion problem in its residential area during this earthquake (Figure 77). After learning negative
lessons from Urayasu and other areas, the Sapporo City Government put emphasis on quick
execution of ground improvement project under existing houses. Accordingly, the design was
completed within 6 months after the earthquake and project will start in April, 2019, after
snow melt. The subsoil under houses will be improved by injecting colloidal silica.
Further, the 2018 Sulawesi earthquake of Mw=7.5 in Indonesia was associated with sub-
marine landslides that induced tsunami. Also, gentle slopes flowed hundreds of meters at 4
places and claimed hundreds of casualties. People’s desire for safety is far from satisfaction.

Figure 77. Liquefaction-induced ground distortion in residential area of Sapporo in 2018.

29
13 CONCLUSION

This paper summarizes a range of geotechnical efforts and contributions towards the recovery
from the seismic disaster in 2011. In contrast to what was done in the 20th Century, the activ-
ities after 2011 are more related with people’s property (liquefaction under houses), less expen-
sive but long structures (river levee and lifeline) and completely new problems (nuclear
accident). Efforts so far made are not complete and are required to be continued in the
coming decades.

ACKNOWLEDGMENT

The contents of this paper are based on many efforts of colleague geotechnical engineers.
First, the contributions of many former civil engineering students of University of Tokyo are
appreciated. Collaboration with and supports provided by Ministry of Land, Infrastructure,
Transport and Tourism was essential. Centrifugal model tests at the Port and Airport
Research Institute provided many valuable findings. Earthquake acceleration records col-
lected by K-NET is always useful. Development of geotechnical engineering for decommission
is supported by a project named “Human Resource Development and Research Program for
Decommissioning of Fukushima Daiichi Nuclear Power Station” under the auspices of Minis-
try of Education, Culture, Sports, Science and Technology. The author expresses his sincere
appreciation to those efforts, supports and collaboration.

REFERENCES

Amagasaki City Government. 2007. Project for restoration of Tsukiji Community after earthquake disas-
ter (in Japanese).
Bahmanpour, A., Towhata, I., Sakr, M., Mahmoud, M., Yamamoto, Y. & Yamada, S. 2019. The effect
of underground columns on the mitigation of liquefaction in shaking table model experiments, Soil
Dynamics and Earthquake Engineering, 116(1):15-30.
Endo, K. & Mazereeuw, M. 2015. Historical study on geographical features of shrines in town of
Minami Sanriku, Miyagi Prefecture, Journal of the Japanese Institute of Landscape Architecture, 78
(5):693-696 (in Japanese).
Hatori, T. 1975 Tsunami magnitude and wave source regions of historical Sanriku tsunamis in Northeast
Japan, Bulletin of Earthquake Research Institute, University of Tokyo, 50: 397-414 (in Japanese).
Hayati, H. & Andrus, R.D. 2009. Updated liquefaction resistance correction factors for aged sands.J.
Geotech. Geoenviron. Eng., ASCE 135(11): 1683-1692.
HERP (Headquarters for Earthquake Research Promotion) 2009. Long-term evaluation of seismic activ-
ities in the region from Off-Sanriku to Off-Boso (revised), Report by Geological Investigation Com-
mittee (in Japanese).
Housner, G.W. 1963. An engineering report on the Chilean earthquakes of May 1960, Bulletin of the
Seismological Society of America, 53(2): 219-223.
Inui, T., Yasutaka, T., Endo, K. & Katsumi, T. 2012. Geo-environmental issues induced by the 2011 off
the Pacific Coast of Tohoku Earthquake and tsunami, Soils and Foundations, 52(5),856-871.
Ishihara K. 1985. Stability of Natural Deposits During Earthquakes, Theme Lecture at 11th ICSMFE,
San Francisco, Vol. 1:321-376.
Ishihara, M. & Sasaki, T. 2012. Relationship between age of ground and liquefaction occurrence in the
2011 Great East Japan earthquake. Proc. Int. Symp. Engineering Lessons Learned from the 2011
Great East Japan Earthquake, Tokyo, 771-776.
Itako City Government. 2013. Report on liquefaction mitigation project in Hinode District, http://www.
city.itako.lg.jp/page/page002163.html retrieved on Dec. 11, 2018 (in Japanese).
JMA (Japanese Meteorological Agency). 2011. On The 2011 off the Pacific coast of Tohoku Earthquake
(Ver.15), Press release, Retrieved Nov. 28 2018 (in Japanese).
Kawasumi, H. & Sato, Y. 1949. Characteristics of ground deformation, tsunami, seismic intensity and
damage extent caused by the 1946 Nankai earthquake, Record of 1946 Nankai earthquake disaster,
Ed. Kochi Prefectural Government, 18-52 (in Japanese).

30
Keino, C., Kohiyama, M. & Sonobe, T. 2011. Health damage of residents caused by liquefaction damage
of houses - Examples in Mihama of Chiba during the East Japan Earthquake Disaster, Annual Con-
vention of JAEE, 80-81 (in Japanese).
Kiku, H., Fukunaga, D., Matsumoto, M. & Takahashi, M. 2006. Shaking table tests on countermeasures
against uplift of manhole due to liquefaction. Annual Conference of JGS, 2:1889-1890 (in Japanese).
Kitahara, M. & Uno, R. 1965. Relationship between tilting of floor and dizziness - Case history study in
tilting building in Niigata, Practica Oto-Rhino-Laryngologica, Society of Practical Otolaryngology, 58
(3):145-151 (in Japanese).
Kohiyama, M. & Keino, C. 2012. Health disturbance of residents in houses tilted due to liquefaction
during the Great East Japan earthquake. 15th World Conf. Earthq. Eng., Lisbon.
Kohiyama, M., Keino, C. & Sonobe, T. 2012. Questionnaire Survey on Health Disturbance of Residents
in Mihama Ward, Chiba City Who Suffered Housing Damage due to Liquefaction, Journal of Insti-
tute of Social Safety Science, 17: 1-7 (in Japanese).
Kokusho, T. 2014. Liquefaction damage and restoration in Itako City, Report on the Great Est Japan
Earthquake Disaster, Fundamental Aspect 3 “Geohazards”, 261-267 in DVD publication　(in
Japanese)
Komine, H. 2017. Creation of decommission technology and collaboration with nuclear engineering,
ATOMOΣ, Atomic Energy Society of Japan, 59 (4):11-13 (in Japanese).
Komine, H., Towhata, I. & Narushima, S. 2016. Environmental geotechnics and education initiatives for
recovery from the Fukushima I Nuclear Power Plant Accident, Japanese Geotechnical Society Special
Publication, 2:1982-1985.
Kotoda, K. & Wakamatsu, K. 1988. Liquefaction and Resulting Damage to Structures during the Chiba-
ken-Toho-Oki Earthquake, Monthly Magazine, Japanese Society of Soil Mechanics and Foundation
Engineering, 36 (12):19-24 (in Japanese).
Lewis, M., Arango, I. & McHood, M. 2008. Site characterization philosophy and liquefaction evaluation
of aged sands — a Savannah River site and Bechtel perspective, Proc. Conference “From Research to
Practice in Geotechnical Engineering”, Geotechnical Special Publication, New Orleans, USA, 180:
540-558.
Matthews, J., Condit, W., Sinha, S., Maniar, S., Sterling, R. & Selvakumar, A. 2013. State of Technology
for Rehabilitation of Water Distribution Systems, Report to EPA/600/R-13/036, p.31.
Matsuura, S. 2013. Tsunami effect on shrines during the Tohoku earthquake in March 2011, Hydro-
logical Science, 57 (3):168-186 (in Japanese).
Mulilis J.P., Mori K., Seed H.B. & Chan C.K. 1977. Resistance to liquefaction due to sustained pressure.
Proc. ASCE, 103(GT7): 793-797.
Nagase, H., Shimizu, K., Hiro-oka, A., Mochinaga, S. & Ohta, M. 1999. Increase in liquefaction strength
of sandy soils due to aging, Proc. 34th Japan National Conf. Geotechnical Engineering, 987-988 (in
Japanese).
Okamura, M., Takebayashi, M., Nishida, K., Fujii, N., Jinguji, M., Imasato, T., Yasuhara, H. & Naka-
gawa, E. 2011. In-situ desaturation test by air injection and its evaluation through field monitoring
and multiphase flow simulation, Journal of Geotechnical and Geoenvironmental Engineering, ASCE,
137(7): 643-652.
Orense, R.P., Morimoto, I., Yamamoto, Y., Yumiyama, T., Yamamoto, H. & Sugawara, K. 2003. Study
on wall-type gravel drains as liquefaction countermeasure for underground structures,Soil Dynamics
and Earthquake Engineering23: 19-39.
Otsubo, M., Towhata, I., Hayashida, T., Shimura, M., Uchimura, T., Liu, B., Taeseri, D., Cauvin, B. &
Rattez, H. 2016a. Shaking table tests on mitigation of liquefaction vulnerability for existing embedded
lifelines, Soils and Foundations, 56(3): 348–364.
Otsubo, M., Towhata, I., Hayashida, T., Liu, B. & Goto, S. 2016b. Shaking table tests on liquefaction
mitigation of embedded lifelines by backfilling with recycled materials, Soils and Foundations, 56(3):
365-378.
Plafker, G. 1969. Tectonics of the March 27, 1964, Alaska earthquake, USGS Professional Report “The
Alaska Earthquake, March 27, 1964: Regional effects”, Report No. 543-I.
Razouli, R., Towhata, I. & Akima, T. 2016. Experimental evaluation of drainage pipes as a mitigation
against liquefaction-induced settlement of structures, Journal of Geotechnical and Geoenvironmental
Engineering, ASCE, 142(9).
Rikuzen-Takata City 2016. Reconstruction of Rikuzen-Takata (in Japanese).
Sasaki, Y., Towhata, I., Miyamoto, K., Shirato, M., Narita, A., Sasaki, T. & Sako, S. 2012. Reconnais-
sance report on damage in and around river levees caused by the 2011 off the Pacific coast of Tohoku
Earthquake, Soils and Foundations, 52(5): 1016-1032.

31
Suwa, S. & M. Fukuda 2014. Case Study on Countermeasure to Liquefaction by Dewatering Method,
Journal of the Society of Materials Science, Japan, 63(1): 21-27.
Takahashi, H., Takahashi, N., Morikawa, Y., Towhata, I. & Takano, D. 2016. Efficacy of pile-type
improvement against lateral flow of liquefied ground, Geotechnique, 66(8): 617-626.
Tatsuoka, F., Iwasaki, T., Tokida, K., Yasuda, S., Hirose, M., Imai, T. & Kon-no, M. 1980. Standard
penetration tests and soil liquefaction potential evaluation, Soils and Foundations, 20(4): 95-111.
Tatsuoka, F., Kato, H., Kimura, M. & Pradhan, T.B.S. 1988. Liquefaction strength of sands subjected
to sustained pressure. Soils and Foundations – JGS 28(1): 119-131.
Tokimatsu, K. & Yoshimi, Y. 1983. Empirical correlation of soil liquefaction based on SPT N-value and
fines content, Soils and Foundations, 23(4): 56-74.
Tokimatsu, K. 1997. Earthquake resistant design for buildings and SPT-N value, The Foundation Engin-
eering & Equipment, Monthly, 25 (12):61-66 (in Japanese).
Tobita, T., Kang, G.C. & Iai, S. 2011. Centrifuge modeling on manhole uplift in a liquefied trench, Soils
and Foundations, 51(6): 1091-1102.
Towhata, I., Maruyama, S., Kasuda, K., Koseki, J., Wakamatsu, K., Kiku, H., Kiyota, T., Yasuda, S.,
Taguchi, Y., Aoyama, S. & Hayashida, T. 2014. Liquefaction in Kanto Region during the East-Japan
gigantic earthquake on March 11, 2011, Soils and Foundations, 54(4): 859-873.
Towhata, I., Morikawa, Y., Takahashi, H., Takahashi, N. & Sugawa, T. 2015a. Centrifuge model tests
on mitigation against liquefied-soil lateral flow by using cement treated soil columns, 16th European
Regional Conference on Soil Mechanics and Geotechnical Engineering, Edinburgh, 1199-1204.
Towhata, I. & Nakamura, H. 2015b. Qualified evaluator of subsoil quality for safety of residential land,
Int. Conf. Soft Ground Engineering (ICSGE2015), Singapore, Paper No. 180.
Towhata, I., Taguchi, Y., Hayashida, T., Goto, S., Shintaku, Y., Hamada, Y. & Aoyama, S. 2016a.
Liquefaction perspective of soil ageing, Geotechnique, 67(6): 467-478.
Towhata, I., Yasuda, S., Yoshida, K., Motohashi, A., Sato, S. & Arai, M. 2016b. Qualification of resi-
dential land from the viewpoint of liquefaction vulnerability, Soil Dynamics and Earthquake Engin-
eering Journal, 91:260-271.
Towhata, I. 2018. Progress of Liquefaction Ageing in Seismically Active Japan. Proc. GEESDV Geotech-
nical Earthquake Engineering and Soil Dynamics V, GeoInstitute of ASCE, Austin, Texas. GSP 290:
473-483.
Towhata, I., Komine, H., Goto, S., Suzuki, M., Hishioka, S. & Watanabe, Y. 2018. Geotechnical engin-
eering for nuclear decommission, Journal of Nuclear Fuel Cycle and Environment, Atomic Energy
Society of Japan, ATOMOΣ, 25 (2):85-90 (in Japanese).
Uda, T., San-nami, T., Hoshigami, Y. & Sakai, K. 2012. Damage due to the 2011 great tsunami and
shrines escaped from being damaged, Proc. JSCE B3 68 (2):I_43-I_48 (in Japanese).
Yamazaki, K. 2013. Disater and Shrine, Proc. Annual Meeting of the Association of Japanese Geograph-
ers, p. 100038 (in Japanese).
Yasuda, S. & T. Hashimoto 2002. Subsidence of houses due to liquefaction during the 2000 Tottoriken
Seibu earthquake, 57th Annual Conf. JSCE, Paper III-515: 1029-1030 (In Japanese).
Yasuda S. 2004. Damage in residential district caused by Tottori-ken Seibu earthquake, Journal of Com-
prehensive Papers, Architectural Institute of Japan, 119(2):45-46.
Yasuda, S., Sasaki, S., Noguchi, C. & Ozawa, N. 2013a. Shaking table tests to study the possibility to
prevent liquefaction-induced damage of houses by installing sheet piles (Part 1), Proc. 48th JGS
National Conference, 1759-1760 (in Japanese).
Yasuda, S., Sasaki, S., Noguchi, C. & Ozawa, N. 2013b. Shaking table tests to study the possibility to
prevent liquefaction-induced damage of houses by installing sheet piles (Part 2), 48th JGS National
Conference, 1761-1762 (in Japanese).
Yasuda, S. & Hashimoto, T. 2016. New project to prevent liquefaction-induced damage in a wide existing
residential area by lowering the ground water table, Soil Dynamics and Earthquake Engineering 91:
246-259.
Yoshida, M., Miyajima, M. & Kitaura, M. 2008. Experimental study on mitigation of liquefaction-
induced floatation of sewerage manhole by using permeable recycled materials packed in sandbags.
Proc. 14th World Conference on Earthquake Engineering. Beijing, China.
Yoshikawa, E., Komine, H., Goto, S., Yoshimura, M., Suzuki, M., Narushima, S., Arai, Y., Ujiie, S.,
Sakoda, Y. & Nagae, Y. 2017. The quantitative evaluation for radiation shielding capabilities of soil
materials.Proc JSCE C, 73(4): 342-354 (in Japanese).
Wichtmann, T., Niemunis, A. Triantafyllidis, Th. & Poblete, M. 2005. Correlation of cyclic preloading
with the liquefaction resistance,Soil Dynamics and Earthquake Engineering, 25: 923-932.

32