The Great Hanshin Earthquake Disaster
The January 17, 1995
South Hyogo Prefecture Earthquake

Preliminary Investigation Report
February 5, 1995

J.P. Bardet
Civil Engineering Department
University of Southern California, Los Angeles, California

and

F. Oka, M. Sugito, and A. Yashima,
Department of Civil Engineering
Gifu University, Gifu, Japan


TABLE OF CONTENTS

  1. INTRODUCTION
  2. EARTHQUAKE OVERVIEW
  3. OVERVIEW OF STRONG MOTION RECORDS
  4. DAMAGE OVERVIEW
  5. DAMAGE TO INFRASTRUCTURES, AND RESIDENTIAL AND INDUSTRIAL BUILDINGS
  6. DAMAGE TO GEOTECHNICAL STRUCTURES
  7. ACKNOWLEDGMENT
  8. REFERENCES

INSTRUCTIONS

The following document will expand as you click on underlined and/or bolded text. For Instance, if you click on EARTHQUAKE OVERVIEW, then you may read the text of the chapter entitled EARTHQUAKE OVERVIEW. In a similar way, you may display figures and photographs. For instance, click on Photo.A1 if you want to display Photograph A1.


1. INTRODUCTION

This report summarizes the investigation of the 1995 South Hyogo Prefecture Earthquake which occurred in the Hanshin area, Japan, on January 17, 1995 at 5:46 AM (local time). The investigation was carried out by F. Oka, M. Sugito, and A. Yashima of Gifu University; J. Kiyono of Yamaguchi University; J.P. Bardet of the University of Southern California; and a few members of the Hiroshima Institute of Technology, Muramoto Construction Company, and NKK Corporation. The investigation was mainly focused on the damage to ground and civil engineering constructions. Owing to time constraints and limited means of transportation, this report is only preliminary, and may contain some omissions and inaccuracies. A large fraction of this report was first published in Japanese (Oka et al., 1995a and 1995b). The English version was edited and extended by J.P. Bardet, with the help of F. Oka, A. Yashima, and M. Sugito.

2. EARTHQUAKE OVERVIEW

At 5:46:51 AM, on January 17, 1995, a magnitude 7.2 earthquake occurred around Kobe City in Japan. The earthquake magnitude was defined by the Japan Meteorological Agency (JMA). The earthquake was named 'the 1995 South Hyogo Prefecture Earthquake' (or the 1995 Hyogoken Nanbu Earthquake) by JMA. It was also referred to as 'the Great Hanshin Earthquake Disaster' by the Japanese Government.

The Hyogo Prefecture

The Hyogo Prefecture is located about 450 kmto the southwest of Tokyo. In May 1991, the Hyogo Prefecture had 5.3 million inhabitants. Its largest city and capital is Kobe, with 1.4 million inhabitants (Fig.2.1). Nishinomiya City and Amagasaki City are two other major cities with 420,000 and 500,000 inhabitants, respectively. The Port of Kobe is the second largest port in Japan.

Earthquake Hypocenter

The hypocenter of the main shock was located by Kyoto University and Nagoya University (DPRI, Kyoto University,1995) to be slightly to the north of Awaji Island. The exact position of the earthquake epicenter is: Latitude = 34.641 degrees N, Longitude = 135.179 degrees E, and Depth =13.3 km.

Earthquake intensity

Fig.2.2 shows the JMA earthquake intensities throughout Japan which were announced on January 17. On January 20, JMA intensities at Hokutan-cho and Ichinomiya-cho, located both on Awaji Island, and Kobe City were upgraded to 7 based on field observations. Fig.2.3 and Fig.2.4 show the updated maps of JMA intensity in the Kobe area. In Japan, Intensity 7 corresponds to earthquakes that destroy more than 30 percent of Japanese wooden houses. Table 1 gives an approximate relation between JMA intensity and ground acceleration.

Table 1. Relation between JMA intensity and ground acceleration (1 g = 980 gal).

		JMA intensity 		ground
		scale				acceleration (gal)
		_______________________________
		0				< 0.8	
		1				0.8 - t2.5
		2				2.5 - 8
		3				8 - 25
		4				25 - 80
		5				80 - 250
		6				250 - 400
		7				> 400

Location of aftershock epicenters

Fig.2.5 shows the locations of the epicenters of the aftershocks that occurred in the six hours following the main shock. Fig.2.6 shows the major active faults and the aftershock epicenters in the Kobe area. The aftershocks were located along some of these active faults, especially Nojima Fault, Suma Fault, Suwayama Fault and Ashiya Fault. Photo.2.1 shows the fault rupture that emerged on the ground surface in Awaji Island after the main shock. Kobe city is located in the 'Kinki Triangle', which is surrounded by major active faults. The Kinki triangle is located on the Eurasian plate, close to the intersection of the Philippine Sea Plate, Pacific Plate, and Eurasia Plate. The South Hyogo Prefecture Earthquake is not the first large earthquake in the Kinki triangle. In 1916, there was a magnitude 6.1 earthquake almost at the same location as in 1995. However, since 1916 (i.e., in the last 80 years), Kobe has undergone no earthquake of such a magnitude.

3. OVERVIEW OF STRONG MOTION RECORDS

Strong motion recordings in the South Hyogo Prefecture were obtained by various organizations, including the Japan Railway Company, Osaka Gas Company, Port and Harbor Research Institute, Ministry of Construction, Kobe Marine Meteorological Observatory, Kansai Internal Airport Company, Port and Harbor Research Institute (PHRI) and the Committee of Earthquake Observation and Research in the Kansai Area (CEORKA). CEORKA installed 10 strong motion instruments in the Kansai area. These digital instruments were designed to record the time history of ground velocity with long period motion.

Maximum peak ground acceleration and velocity

Table 2 and Fig.3.1 show the maximum peak ground velocities at 11 locations (including Abeno area). At Kobe University, the maximum peak horizontal velocity of ground surface reached 55.1 kine (1 kine = 1 cm/sec). Fig.3.2, and Tables 3 and 4 show the maximum peak ground accelerations which were recorded by CEORKA and other organizations. The maximum peak horizontal acceleration (883 gal) was recorded at Kobe by Osaka Gas Company. At many locations, large vertical acceleration components were observed. The maximum peak vertical acceleration (446.5 gal) was recorded at Kobe University.

Table 2. Summary of maximum peak acceleration (PGA) and velocity (PGV) recorded by CEORKA in the Kansai Area (1 g = 980 gal, 1 cm/s = 1 kine, NS= North-South, EW=East-West, and UD=Up-Down).

    Epicentral Distance (km) PGV, Peak ground velocity (kine) PGA, Peak ground acceleration(gal)
  Station Name NS EW UD NS EW UD
1
Kobe University
23.6
55.1
31
33.2
269.8
305.3
446.5
2 
Kobe Motoyama
(off scale)
26.0
40.0?
40.0?
40.0?
421.0?
774.9?
379.3?
3
Amagasaki
(off scale)
37.5
40.0?
40.0?
26.1
271.4
321.5
327.9
4
Fukushima
37.5
31
29.8
9.6
180
211.5
194.8
5
Morigawachi
49.6
27.1
24.7
6.1
210.1
123.3
158.8
6
Yae
52.9
21.2
21.8
7
154.7
144.9
127.1
7
Toyonaka
46.2
30.6
12.5
8.3
-
-
-
8
Sakai
40.3
15.9
15.7
6.6
150.2
124.7
100.3
9
Tadaoka
34.0
24.4
14.7
5.9
290.4
190.1
136.5
10
Chihaya
57.3
5.2
4.9
2.5
90.6
108.6
73.6
11
Abeno
44.8
21.4
24.9
6.3
217.4
226.4
136.2

Note: Most CEORKA instruments are calibrated for a maximum velocity of 40 cm/s, except for the instrument at Kobe University. A question mark next to a measured value indicates that the instrument went off scale, and that its maximum peak values are unreliable.

Table 3. Additional data on maximum peak horizontal ground acceleration (gal) recorded by Japan Railways Company (JR) and Osaka Gas Company in the Kansai Area (1g = 980 gal). The exact horizontal orientation is not yet available.

						Epicentral	PGA
						distance	horizontal					
Station Name					(km)		(gal)
____________________________________________________________________
JR Nishi-Akashi					11.8		481
JR Shin-Kobe					21.9		561	
JR Takatori					13.2		616
JR Shin-Osaka					45.5		243
JR Amagasaki					41.6		307
Kobe Marine Meteorological Observatory		26.0		818
Osaka Gas Kobe					27.8		833
Osaka Gas Nishinomiya				32.6		792
JR Takarazuka					37.5		601

Table 4. Additional data on maximum peak ground acceleration (gal) and velocity (kine) recorded by Port and Harbor Research Institute and Ministry of Construction in the Kansai area (1g = 980 gal, 1 cm/s = 1 kine, NS= North-South, EW=East-West, UD=Up-Down, LG=longitudinal, TR=transverse, and H=horizontal).

	Station name		Maximum Peak Ground Acceleration(gal)[velocity, (kine)]
__________________________________________________________________________________									
1 Kakogawa Ozeki		NS	114	LG	136	UD	139
2 Amagasaki highway bridge	LG	310[52]	TR	274[45]	UD	336[30]
3 Yodo River Embankment		LG	144	TR	253	UD	57
  Hirakata
4 Port and Harbor Research	NS	475	EW	320	UD	312
  Institute Amagasaki
5 Port and Harbor Research	NS	394	EW	169	UD	275
  Institute Kobe
6 Yodo River Embankment		LG	150	TR	200	UD	300
  Oyodo
7 Kansai International 		H	160	-	 -	UD	247
  Airport
__________________________________________________________________________________
Number in brackets indicate maximum peak ground velocity (kine)

Attenuation of maximum peak ground acceleration with distance

Fig.3.4 shows the attenuation characteristics of recorded peak horizontal accelerations (the larger horizontal components). It also compares recorded accelerations to the attenuation equation (Sugito, 1986) which was previously developed for hard soil condition on the basis of major Japanese strong motion records. The recorded accelerations appear to decrease rapidly with distance.

Time history and response spectra of acceleration, velocity and displacement

Fig.3.5a, Fig.3.5b, Fig.3.5c, Fig.3.6a, Fig.3.6b, Fig.3.6c, Fig.3.7a, Fig.3.7b, Fig.3.7c, Fig.3.8a, Fig.3.8b, Fig.3.8c, Fig.3.9a, Fig.3.9b, Fig.3.9c, Fig.3.10a, Fig.3.10b, Fig.3.10c, Fig.3.11a, Fig.3.11b, Fig.3.11c, Fig.3.12a, Fig.3.12b, Fig.3.12c, Fig.3.13a, Fig.3.13b, Fig.3.13c, Fig.3.14a, Fig.3.14b, and Fig.3.14c show the uncorrected time histories of acceleration, velocity and displacement of CEORKA's recordings. The time history of acceleration was calculated by differentiating the digital velocity record. The time history of the ground displacement was obtained by integrating the velocity record.

4. DAMAGE OVERVIEW

Casualties

On February 7, 1995, 5,273 persons were reported to have been killed by the earthquake. This is the largest casualty number since the Great Kanto Earthquake which occurred on September 1, 1923, and killed 142,807 persons. In 1995, most victims were crushed to death by the collapse of their houses, and burned to death by the fires that followed the earthquake. Fig.4.1 and Table 5 shows the distribution of deaths by age group. Half of the quake dead were over age 60. Among the victims, 59% were women, and 41% were men. Kobe's city officials attributed the large number of deaths among the elderly to the growing number of younger people living in the suburbs, the fact that many elderly people live alone in the quake-stricken areas, and the fact that a large number of homes in the area were built before and immediately after World War II. In the Kobe area, the old traditional Japanese wooden houses have heavy ceramic tile roofs and clay filling. Designed to resist typhoons, they presented a poor resistance to the earthquake forces.

Table 5. Distribution of deaths per age group.

		Age group	Percent of total(%)
		___________________________________
		0-9				4.7
		10-20			6.3
		20-29			8.7
		30-39			4.8
		40-49			8.2
		50-59			15.2
		60-69			19
		70-79			18.5
		80-89			13.1
		90-				1.7

Assessment of total damage

On Saturday February 4, 1995, the Hyogo Prefecture Government estimated the cost of rebuilding areas hit by the earthquake to some 9.63 trillion yen, approximately 96.3 billion dollars ($1 = about 100 yen). Hyogo Prefecture officials said that 5.8 trillion yen would be necessary to reconstruct buildings and a little more than 1 billion yen to rebuild port facilities. These numbers have been steadily increasing since January 17, 1995, and are not likely to stop at these values.

Fire

It was reported that there were 164 independent fires in Hyogo Prefecture, and that more than 50 percent of fires broke out 24 hours after the earthquake. A large number of fires were attributed to gas leaks, which were sparked by electrical short-circuits. More than 1,300,000 square meters were burned down. Fig.4.2 and Table 6 show the distribution of areas destroyed by fire in the Hyogo Prefecture.

Table 6. Distribution of deaths, areas destroyed by fire, and homes damaged or destroyed (as of February 3, 1995).

							Areas		Homes
Areas			Deaths		destroyed by	destroyed or
							fire(m2)	damaged
_____________________________________________________________________
Akashi			5			100			535
Ashiya			405			5600			4062
Awajishima Island	56			120			5451
Chuo District		206			21700		6544
Higashi-Nada District	1216			39215		10800
Hyogo District		403			12420		4373
Itami				10			295			887
Kakogawa			2		
Kawanishi			1			0			1523
Kita District		2			60			8
Kyoto Prefecture	2			0			1
Miki				5			13			138
Nada District		819			27600		9220
Nagata District		712			503350		15994
Nishi District		2			13			138
Nishinomiya		938			9600			13931
Osaka Prefecture	24			2310			2342
Suma District		333			305600		7811
Takarazuka		86			148			5057
Tarumi District		2			50			193

5. DAMAGE TO INFRASTRUCTURES, AND RESIDENTIAL AND INDUSTRIAL BUILDINGS

The general location of all the observations (i.e, photographs) made during the earthquake investigation are referred to by using letter A to W in Fig.5.1.

Residential dwellings and buildings

About 170,000 houses were destroyed or severely damaged in Hyogo Prefecture and Osaka Prefecture (Photo.J2, Photo.J3, Photo.J4, Photo.P3, and Photo.P4). Many office and apartment buildings were also severely damaged. The number of refugees is reported to be in excess of 300,000.

Industrial buildings and factories

A large number of 7-12 floor reinforced concrete office buildings were damaged at middle or top floors (Photo.E1, Photo.E2, Photo.E3, Photo.E4, Photo.E5, Photo.E6, Photo.E7, and Photo.E8). A large number of factories have been rendered inoperative by the earthquake, either directly by the destruction of their facilities, or indirectly by the interruption of their supply lines. The port of Kobe, which is the second largest port of Japan, was shut down. In Kobe, in contrast to Los Angeles, supplies are difficult to move through alternate routes. The damage to the traffic network in that small and narrow area where the mountains close in on the sea cannot be compared with Los Angeles where detours were possible when expressways collapsed during the 1994 Northridge Earthquake. The earthquake has been called Japan's greatest postwar disaster. At the present, it is difficult to assess the total amount of damage to the Japanese economy.

Railways lines and bridges

There are four major railways companies that operate in the Hyogo Prefecture. Their lines are referred to as Sanyo Shinkansen line, Japan Railways (JR) line, Hankyu line, and Hanshin line. Railway lines, including the fast train (Shinkansen) line, were severely damaged. Many bridge decks and girders of railway lines moved transversely or dropped down to the ground (Photo.E9, Photo.J1, Photo.J5, Photo.J6). A large number of reinforced concrete bridge piers were destroyed or severely damaged (Photo.L1, Photo.L3, Photo.L5, Photo.P1, and Photo.P2). These damaged piers severely deformed the railway tracks above them (Photo.J1). At 5:46 AM on January 17, 1995, most trains were fortunately not yet moving. However, derailment still took place at Sumiyoshi Station of Hanshin railway line. At least 500 m of railways were damaged between Shukugawa Station and Nishinomiya-Kitaguchi Station of Hankyu line. The Sanyo Shinkansen line was damaged in the Kamioichi area in Nishinomiya and Ashiya City (Photo.L1, Photo.L2, Photo.L3, Photo.L4, Photo.L5, Photo.L6, Photo.L7, and Photo.L8).

Highway bridges

Highway bridges sustained severe damage during the South Hyogo Prefecture Earthquake. The locations of representative bridge failures are shown in Fig.5.2. The main failures are summarized in the following section.

1. Route 43 (Iwaya Viaduct-Kobe, Nada district). Reinforced Concrete columns of Iwaya Viaduct collapsed, and the superstructure fell down (Photo.I1, Photo.I2, Photo.I3, and Photo.I4).
2. Route 171 (Mondo Viaduct, Nishinomiya). Mondo Viaduct is a long span bridge which overpasses an Hankyu Railway line. The upper deck of the crossover fell down onto the railway (Photo.K1, Photo.K2, and Photo.K3).
3. Route 2 (Hamate Bypass). The upper deck of this double deck bridge shifted transversely to the north, and almost fell down (Photo.D5).
4. Hanshin Expressway Route 5, Wangan Line (Nishinomiya-Harbor Bridge, east closure span, Nishinomiya). The span length of the Nishinomiya Harbor bridge is 252 m. The closure span on the eastern side of the bridge collapsed, the bridge support failed, and the girders partially buckled (Photo.5.1, Photo.C1, Photo.C2, Photo.C3, Photo.C4, Photo.C5, Photo.C6, Photo.C7, Photo.C8 and Photo.C9).
5. Hanshin Expressway Route 3, Kobe line (Kobe, Nada district, Fukae-Honmachi-Ashiya, Hirata) In the Fukae-Homachi area, a 500 meter section of Route 3 completely collapsed in the transverse direction onto National Road Route 43.
6. Hanshin Expressway Route 3, Kobe line (Nishinomiya, Tateishi junction). A steel bent buckled during the earthquake.
7. Hanshin Expressway Route 3, Kobe line (Nishinomiya, Hamawaki-cho, Satsuba area). The restrainer of the girder failed, and two spans between Bents P40 and P42 collapsed.
8. Hanshin Expressway Route 3, Kobe line (Nishinomiya, Koshien Takashio-cho). A reinforced concrete Bent P267 collapsed (Photo.A1, Photo.A2, Photo.A3, Photo.A4, and Photo.A5).
9. Hanshin Expressway Route 3, Kobe line (Kobe, Chuo District, Hatoba-cho). The roadway deck moved downwards due to the failure of reinforced concrete bents.
10. Hanshin Expressway Route 3, Kobe line (Minatogawa Ramp Bridge- Kobe, Nagata District, Higashi-Shiriike-cho). The upper deck collapsed at three locations.
11. Hanshin Expressway Route 3, Kobe line (Emergency Parking Strip -Kobe, Chuo District, Hatoba-cho). A girder of an emergency parking strip fell to the side of route 2, together with the slab, walls and railings.
12. Meishin Highway, Kawaragi Nishi viaduct (Nishinomiya, Oya-cho). A bent of three-span continuous reinforced concrete bridge collapsed.
13. Harbor Highway, Access to Kobe bridge. Columns of a frame-type pier underwent severe damage.
14. Akashi strait bridge. This cable supported bridge connects the main land to Awaji Island. The bridge, which is still under construction, has a total span length of 3,910 m, and two towers and anchorages. The relative east-west displacement between towers was 1.3 m. The relative east-west displacement between the northern tower and the southern anchorage was 1.4 m. However, no damage was observed in the bridge, including the two towers and anchorages.

In spite of the recent improvement of design standards in Japan, many bridge girders were constructed based on the old design standards. However, the bridge girder of Nishinomiya-Harbor bridge of the Hanshin Expressway, which was opened to traffic in 1994, also fell down.

6. DAMAGE TO GEOTECHNICAL STRUCTURES

Embankments and slopes

A total length of 2800 m of the left levee of Yodo River, located to the southeast of Amagasaki City and to the northwest of Osaka Prefecture, was severely damaged by liquefaction (Photo.Q1). An 800 m length of the right levee was also damaged. Yodo River is the largest river in the Kansai area. Sand boils were seen almost everywhere along the failed levees (Photo.Q2, Photo.Q3). The concrete parapet at the top of the levees settled the a maximum of 3 m, and at some location leaned flat on the ground. The areas where ground had been improved performed slightly better.

The embankments of Niteko Reservoir in Nishinomiya City were severely damaged by liquefaction. This reservoir is made of three separate units which are contained by four parallel embankments in the east-west direction (Fig.5.3). The reservoir covers a 450 m by 80 m area. The three embankments to the north completely failed (Photo.O5), and the one to the south underwent major cracking, and settled at least 1m (Photo.O1, Photo.O2, Photo.O3, and Photo.O4).

Many slope failures occurred in the embankments and levees along Sumiyoshi River, in Higashi-Nada District. Similar failures were observed next to the mouth of Ashiya River in Ashiya City, along Shukugawa River to the northwest of Niteko Reservoir, and along Nikawa River to the north of Nikawa-Yurino Filtration Plant. The largest slope failure occurred next to the filtration plant in Nikawa-Yurino town in Nishinomiya City (Photo.6.1 , Photo.M1, Photo.M2, Photo.M3, Photo.M4, Photo.M5, and Photo.M6). This landslide originated from the filtration plant, and flowed over a 140 m by 50 m area. It killed 34 persons.

Artificial Islands and Waterfronts

Massive liquefaction occurred on the manmade islands in Osaka Bay, especially at Port Island and Rokko Island.

Rokko Island, to the southeast of Kobe City, is reclaimed land which was constructed between 1973 and 1992. It has a 3.4 km by 2 km rectangular shape, and covers a 580 ha area. This island contains apartment housing, amusement parks, container yards, wharves, and port facilities ( Photo.6.2). The maximum subsidence of ground surface was reported to be 3 m just behind the quaywall on the west side of Rokko Island (Photo.F1, and Photo.F2). At some locations, the quaywall laterally moved 2 m toward the sea. There were many deep ground crackings parallel to the quaywall. Most of the container cranes derailed from their tracks, and were damaged as the result of liquefaction induced ground motion. One crane completely collapsed down (Photo.F4), while others displayed local buckling and plastic hinges. Aerial photographs (Photo.6.2) taken on the day of the earthquake clearly show that deep ground cracking is parallel to and concentrated within 50 m of the shoreline. In the container yards, liquefaction sand deposits littered areas of asphalt pavement as large as 1 ha. Some sand boils contained a large amount of gravel particles and boulders with 10 cm diameter (Photo.F3, Photo.F5, and Photo.F6). Liquefaction also damaged the infrastrucure. A section of the overpass of the Rokko Liner, which is a new traffic system that connects Kobe City and Rokko Island, fell at the north end of Rokko Island (Photo.G1, Photo.G2, Photo.G3, Photo.G4, and Photo.G5).

Port island is an artificial island to the west of Rokko Island (Photo.6.3). This island was constructed in two phases. During the first phase, between 1967 and 1981, a 436 ha area was constructed. This initial island was extended by reclaiming 319 ha. This addition has now completely been reclaimed, but is not yet developed. Aerial photographs(Photo.6.3) taken on the day of the earthquake clearly show that Port Island underwent a more massive liquefaction than Rokko Island. More than 50% of the asphalt pavement surface of the initial island was littered with liquefaction sand deposit (Photo.U2). Ground settlement in the order of 50 to 100 cm could also be observed on the entire island, especially around the buildings and bridges supported by deep foundations (Photo.T1, Photo.T2, and Photo.U1). In the elongated parking lot of Portopia Land Amusement Park, at the southern border of the initial island, the ground surface was completely covered by a 15-20 cm thick layer of sand and mud (Photo.V1, Photo.V2, and Photo.V3). Concrete blocks as large as 30 cm were uniformly scattered throughout the parking lot, and emerged from the sand and mud deposits (Photo.V2). Their presence clearly indicates that liquefaction induced a violent flow of water. The mudline on the walls surrounding the parking lot (Photo.V1), which covers a 1,200 m by 40 m area, was about 1.1 m above the ground surface. Therefore, at least 53,000 m3 of water was ejected in this area. Approximately in the middle of the parking lot, there was a large sand boil that covered a 17 m by 5 m area, and had a height of 1 m (Photo.V4). In the section of Port Island that underwent ground improvement, the ground performed much better than on the rest of the island. For instance, there were no visible sign of ground settlement or lateral motion on the ground beneath the rollercoasters of Portopia Land Amusement Park (Photo.W1). This ground had been improved by using the vibro rod method.

Meriken Park is located on the mainland to the northwest of Port Island (Photo.6.4). It covers a 500 m by 250 m area. The 160 m long waterfront to the south of Meriken Park completely failed by liquefaction, and disappeared into the sea (Photo.D3). Aerial photographs (Photo.6.4) taken on the day of the earthquake show 150 m long traces of muddy water in the sea. Some sections of the wharf on the eastern side of Meriken Park partially failed. The ground settled behind the wharf in excess of 1 m (Photo.D1, and Photo.D2). Liquefaction left behind a mudline that could clearly be observed on the brick wall of the passenger waiting room building (Photo.D4). This mudline indicates that ground water was ejected as high as 40-50 cm during (or after) the earthquake.

Underground structures

Many cracks were observed in the lining of the Rokko tunnel of the Sanyo Shinkansen line, and three other tunnels of the same line. Ground cracking was also observed on the ground above the east entrance of Rokko tunnel.

Many reinforced concrete pillars of the Daikai underground station of the Kobe Rapid Railway Transportation collapsed. This underground station was constructed by the cut-and-cover method. A 95 m by 28 m area of the asphalt road above the station subsided a maximum of 3 meters, and is still settling (Photo.R1, and Photo.R2). In Uezawa Station and Sannomiya Station, the central reinforced concrete pillars also failed under shear.

Rock falls

In Kurakuen, 4-Bancho, in Nishinomiya City, rocks came down from the hill, and fell on the houses and road (Photo.N1, and Photo.N2). The total volume of rocks was grossly estimated to be 20 m3. The largest blocks were 80 cm in diameter (Photo.N3). In Kentani, to the east of the previous site, a rock block as large as 3.5 m toppled down from the hill.

Lifelines

On the first day of the earthquake, there were about one million houses without electricity. However on January 23, 1995, six days after the earthquake, electricity had been restored to almost 100% (Central Research Institute of Electric Power Company, 1995). As a result of gas duct breakage caused by the earthquake, 850,000 houses still did not have gas on January 24, 1995. On January 24, 1995, 660,000 houses also had no water (Photo.E10 , and Photo.E11). Drinking water was provided to the inhabitants of damaged areas by water tanks. During the first five or six days following the earthquake, telephone communication was difficult in Kobe City. Nippon Telegraph and Telephone Corporation (NTT) said a loss of electrical power cut 285,000 lines, while the collapse of telephone poles and fires that burned cables put another 193,000 lines out of service (The Japan Times, Saturday, February 11, 1995). Before the quake, there were about 1.44 million telephone lines for calls within the damaged area, and 60,000 others for long distance calls. Telecommunications conditions were worsened because 50 times more calls than usual inundated the area immediately after the quake, according to NTT. However, by February 1, 1995, telephone services were restored to almost 100%. The earthquake generated an enormous volume of trash and debris, which could not be evacuated by regular means of disposal. Several parks were temporarily requisitioned to dispose of domestic refuse and earthquake debris.

7. ACKNOWLEDGEMENT

The following report was made possible through a SGER grant of the US National Science Foundation and a grant of the Research Committee on the South Hyogo Earthquake Disaster sponsored by the Japanese Society of Soil Mechanics and Foundation Engineering. The authors are also thankful to Prof. T. Adachi, S. Kobayashi and F. Zheng of Kyoto University, and Y. Taguchi of Tasei Corporation for their help during the reconnaissance, and to the Committee for Earthquake Observation and Research of the Kansai Area, Hiroshima Institute of Technology, Muramoto Construction Company, Konoike Construction Company, Osaka Gas Company, Japan Ministry of Construction, and NKK Corporation for their respective contribution.

9. REFERENCES

Central Research Institute of Electric Power Company, 1995, "Preliminary report on the 1995 South Hyogo Earthquake," Report of Central Research Institute of Electric Power Industry, in Japanese, 18 p.

Disaster Prevention Research Institute, DPRI, 1995, Kyoto University, Private communication with Prof. M. Sugito, January 17.

Oka, K., M Sugito, and A. Yashima, 1995a, "The Great Hanshin Earthquake Disaster; The January 17, 1995, South Hyogo Prefecture Earthquake, Preliminary Investigation Report, Part 1, 25 January 1995," Report of Civil Engineering Department, in Japanese, Gifu University, Gifu, Japan, 56 p.

Oka, K., M Sugito, and A. Yashima, 1995b, "The Great Hanshin Earthquake Disaster; The January 17, 1995, South Hyogo Prefecture Earthquake, Preliminary Investigation Report, Part 2, 2 February 1995," Report of Civil Engineering Department, in Japanese, Gifu University, Gifu, Japan, 24 p.

Sugito, M., 1986, Earthquake Motion Prediction, Microzonation, and Buried Pipe Response for Urban Seismic Damage Assessment, PhD Thesis, Kyoto University, Department of Civil Engineering.