4th International Conference on

Earthquake Geotechnical Engineering

June 25-28, 2007
Paper No. 1621

DUZCE (M7.2, 1999) EARTHQUAKE: DAMAGE CORRELATION

WITH THE MICROZONATION STUDY AND VULNERABILITY
ASSESSMENT OF THE WATER SYSTEM IN DUZCE, TURKEY

Dimitra MANOY 1 , Maria ALEXOUDI 2 , Dimitrios RAPTAKIS3, Kyriazis PITILAKIS4

ABSTRACT

Duzce, in Turkey with a population of about 80.000 people, was severely damaged during the
November 12th, 1999 earthquake (Mw=7.2), which followed the Kocaeli Mw=7.4 earthquake that
happened only two months earlier in a close distance. In this paper, the main results of a pilot
microzonation study conducted after the earthquake are presented in order to investigate whether the
damages in buildings and lifelines (i.e. water system) are correlated with the local soil conditions. The
vulnerability assessment of the water system in Duzce has been the final output of this study.

Keywords: Duzce earthquake, damages, pilot microzonation study, local soil conditions, peak ground
acceleration, water system, vulnerability assessment

INTRODUCTION

The 12 November 1999 earthquake in Turkey caused considerable damage to residential and
commercial buildings, public facilities and infrastructures with substantial casualties. The epicenter of
the earthquake was located about 6km south of Duzce. Buildings, already moderately and lightly
damaged by 17/8/1999 Kocaeli M7.4 earthquake, have been seriously damaged and many collapsed
(Figure 1a). Office buildings, utility systems accounted additional serious damages The water and the
sewage system were heavily damaged due to ground deformations and shaking. Damage to the
transport infrastructure observed along the 60 km of the Ankara-Istanbul highway crossing the Duzce
Fault; at Bakacak section (between Kaynasli and Bolu) two lanes of the four-lane highway collapsed
due to land slide for a section of about 200m, as it is depicted in Figure 1b (Erdik, 1999).

Focusing in the city of Duzce, the present study aims to investigate the relation between observed
damages and site effects. For this purpose, a pilot microzonation study was implemented in order to
estimate the magnitude and spatial distribution of peak ground acceleration, velocity and displacement
values as well as peak spectral acceleration values. Based on these results, the vulnerability assessment
of the water system in Duzce was conducted and compared to actual reported damages.

1
Civil Engineer, MSc, Department of Civil Engineering, University of Thessaloniki, Greece, Email:
dmanou@civil.auth.gr
2
Dr Civil Engineer, Department of Civil Engineering, University of Thessaloniki, Greece, Email:
alexoudi@civil.auth.gr
3
Lecturer, Department of Civil Engineering, University of Thessaloniki, Greece, Email:
raptakis@civil.auth.gr
4
Professor, Department of Civil Engineering, University of Thessaloniki, Greece, Email:
kpitilak@civil.auth.gr
 Figure 1a. A general view of building damage Figure 1b. Landslide on E5 highway in Duzce
in Duzce (Erdik, 1999). earthquake (Erdik, 1999).

THE NOVEMBER 12TH, 1999 EVENT

November 12th, 1999 Duzce M7.2 earthquake was associated with the North Anatolian fault, and in
particular with the Duzce fault, which forms a morphological boundary at the south of the Duzce
Plain. The Duzce fault, (Figure 2a), extends eastwards about 70km from the main branch of the North
Anatolian Fault, having its terminus near the village of Kaynasli, located to the southeast of Duzce. It
is a right lateral strike-slip fault.

The peak ground acceleration recorded at the Meteorological Station of Duzce was invariably large,
(PGA[EW]=0.51g), something that is not unexpected due to closer proximity of fault rupture. The long
duration and the long period energy content in the time history of the Duzce record, (Figure 2c), is a
strong indication of basin response and effect of soft soil behaviour (Sucuoglu, 2002). Moreover, from
the shape of the corresponding acceleration response spectra, (Figure 2b), it is evident that the
acceleration demands for periods between 0.1 to 1.0 seconds, which is the expected range for periods
of most structures at the area, were considerably large. Due to these considerably high demands, the
structures that survived the first 17/8/1999 Kocaeli quake were severely damaged and many of them
collapsed in the Duzce 12/11/1999 earthquake.

6
R e c o rd _ E W (L ) D uzc e
Acceleration (m/sec )
2

4 1 2 /1 1 /1 9 9 9

0
0 5 10 15 20 25
-2

-4
t( )
6
Re cord_ NS (T) D uzce
Acceleration (m/sec )
2

4 1 2/1 1/1 99 9

Figure 2a. The Duzce fault which caused the 12-11- 2

1999, Duzce strong seismic event. 0
Normalized Response Spectra Acceleration 0 5 10 15 20 25
(Record Duzce 12/11/1999) -2
6
Response Spectra Acceleration

Duzce_EW (L) -4
5
Duzce_NS (T) 6 t (sec)
Duzce_up
4 Record_V D uzce
Acceleration (m/sec )
(m/sec )

Turkish Seismic Code (Z3) 4

2

12/11/1999
3 EC8 - (C)
EC8-Draft (C) Type1 2
2
0
1 0 5 10 15 20 25
-2
0
-4
0 1 2 3 4 t (sec)

Figure 2b. 12-11-1999 Duzce earthquake response Figure 2c. 12-11-1999, strong motion
spectra at the Meteorological Station. recordings at the Meteor. Station.
 (b)

(a) (c)
Figure 4. (a) Distribution of building damage in Duzce (b) Number of stories vs. damage (c)
Number of stories vs. damage, reported by the municipality (Donmez and Pujol, 2005).

OBSERVED DAMAGES

An extensive survey study was carried out in Duzce to document the sustained damages of the
reinforced concrete structures (Donmez and Pujol, 2005). Based on thse data, an interesting
observation is the correlation between the spatial distribution of the damages observed in buildings
and their location with respect to site conditions and the fault. The majority of the buildings that
suffered great damage or collapsed are located at the centre of the city close to the river and closer to
the fault trace. Soil formations are composed mainly of soft alluvial deposits (Figures 4a). However
there is not a clear conclusion whether or not extensive damages are correlated to the soft soil
conditions as we observe at least equal number of building with low to moderate damages. The quality
and the performance of the building stock in Duzce has to played the critical role. Another interesting
observation of the survey is the fact that taller buildings of 4-6 stories with To=0.3-0.4s according to
Aydan et al., (2000), suffered relatively more damages. The fundamental period of these type of
buildings is very close to the predominant period of the record in Metereological station (see Figure
2b). However based on the available records it is difficult to attribute the higher damages solely to a
resonance effect, as actually the high spectral values are spreading in a wide range from 0.2s to 0.8s
and the fundamental period of the site where the station is located is around 0.75s, as we will see later.

¯
Legend
waterfailure
waterpipes

0 600 1,200 2,400 3,600 4,800

km

Figure 5. Water system damage in Duzce (Kocaeli & Duzce 1999 earthqks.) (Alexoudi, 2005).
The water system in Duzce was also severely damaged due the combined effects of Kocaeli
(17/8/1999) and Duzce (12/11/1999) earthquakes. Tromans, (2004) developed a general database
providing water pipe failures per district for three time periods; before Kocaeli earthquake, between
Kocaeli and Duzce earthquake and after Duzce earthquake. In the present study another water system
damage database is used, created by Alexoudi, (2005) (Figure 5), which provides a limited number of
fully documented failures. As it can be observed, most failures are concentrated at the southern part of
the city, around the river area, where -as already mentioned before- the soil conditions are dominantly
soft and loose.

Finally, although no surface indication of liquefaction was reported in Düzce after the 12/11/1999
strong seismic event, in several buildings settlements were observed, a fact that may be attributed to
partial liquefaction of some thin silty soil layers (Aydan, 2000). Considering the spatial distribution of
structure and lifeline damages in Duzce the present study aims to investigate the role of site effects on
damage distribution and magnitude. The work presented herein is part of he Pilot Microzonation Study
performed by Pitilakis et al., (2006) after the November 12th, 1999 earthquake.

SEISMIC HAZARD ASSESSMENT OF DUZCE

Seismic hazard analysis of Duzce has been estimated applying a probabilistic approach, using code
CRISIS 99 and considering the seismic zones affecting Duzce (Kayabali, 2002). The probabilistic
seismic hazard analysis has been performed using three different attenuation relations proposed by
Ambraseys, (1996), Sadigh et al., (1997) and Ozbey, (2004), estimating the distribution of PGA values
in case of rock conditions for a mean return period of 475 years. The resulting PGA maps (Figure 6)
have been validated with existing results from previous studies already performed in the area. PGA
values for outcrop conditions according to Ambraseys, (1996) relation give higher values compared to
Ozbey, (2004) and Sadigh et al., (1997) relations, as they are presented in Table 1. However, they
could be used as the peak input motion values for the site response analysis. It is also estimated that
the 475 year mean return period earthquake is similar to the November 1999 earthquake.

Table 1. Comparison between different relations (Probabilistic seismic hazard) – Rock site
ROCK475
Ambraseys, (1996) 0.630-0.690 g
Sadigh et al., (1997) 0.550-0.600 g
Ozbey, (2004) 0.523-0.566 g

PGA (gal) for 475y return period PGA (gal) for 475y return period
PGA (gal) for 475y return period
(Atten.Relation: Sadigh et al, 1997) (Atten.Relation: Ozbey et al, 2004)
(Atten.Relation: Ambrasey et al, 1996)

Legend Legend
road road
mahalle_ Legend mahalle_
PGA (gal) PGA (gal)
road
High : 690 High : 566
mahalle_
PGA (gal)
0 900 1,800 3,600 0 875 1,750 3,500 0 900 1,800 3,600
Low : 630 Km 550-600 Km Low : 523 Km

(a) (b) (c)

Figure 6. PGA (gals) values for Duzce - Probabilistic scenario for mean return period of 475
years. Attenuation relations (a) Ambraseys (1996), (b) Sadigh et al., (1997), (c) Ozbey (2004).
Soil conditions: Rock.
 AVAILABLE GEOPHYSICAL AND GEOTECHNICAL DATA

Duzce is situated in a tectonic basin filled over time with river and lake sediments. The plain (Figure
2a) is now consist of layers of clay, sand, and gravel. The basin sediments, which at the centre of the
plain attain up to 250 m, are generally looser and soft when reaching the surface.

Figure 7. Available geophysical and geotechnical data in Duzce.

Following the 12/11/1999 earthquake, a number of geophysical surveys were conducted in Duzce.
Kudo et al., (2000) carried out array measurements of microtremors aiming to determine the S-wave
velocity structures at the DZC (Duzce) strong motion site (Figure 7); based on his results, the site is a
sedimentary basin of thick and soft sediments at the surface (Vs~260 m/sec) and intermediate depths
(Vs~460-510 m/sec), while the depth of the bedrock reaches -750m. Rosenblad et al., (2001) applied
the spectral-analysis-of surface-waves (SASW) to characterize small-strain shear wave velocity
profiles, using in situ seismic measurements. The SASW profile indicates that the velocity of the near-
surface materials is less than 200 m/sec, but the velocity quickly increases to 400 m/sec at 15m depth.
Kudo et al., (2000) profile shows a constant shear wave velocity of 260 m/sec extending from the
surface to a depth of 35m. At depths below about 40m, both profiles show a shear wave velocity
between 400 and 450 m/sec. Yamanaka et al., (2002) conducted single-point measurements and array
measurements of microtremors at 25 sites in Duzce along a line from the south to the north. He
proposed a 7-layer model according to which the shear-wave velocity is Vs=769 m/sec at a depth of
approximately 100m. Finally, a site specific microtremor survey has been conducted by Tromans,
(2004) applying the horizontal-to-vertical spectral ratio technique. According to this study the
predominant HVSR frequency fp varies between 0.5–1.5 Hz in a rather wide area which covers the
southern part of the city of Duzce, while the value of fp at the record site (Meteorological Station) was
defined at fp = 0.73 Hz.

In addition to the geophysical surveys, borehole investigations (10-15m and rarely 20m depth) have
been performed at 185 proposed locations by the Ministry of Settlement and Public Works, General
Directorate of Disaster Affairs, (2000) in the provincial centre of Duzce (Figure 7). Moreover, the
Department of Geology and Geophysical Engineering of Ankara University (Kayabali et al., 2001)
performed in the research area, approximately 60 drillings with a maximum depth of 15m each. Within
the framework of the Pilot Microzonation Study of Duzce (Pitilakis et al., 2006) eight deep boreholes
were drilled which are presented in Figure 7; in specific, six boreholes of 40m depth and two of 90m.
Six boreholes (four of 40m and the two of 90m ) were drilled at the southern part of the city, where as
already mentioned before, the major damages were observed after November 12th, 1999 earthquake.
The two remaining boreholes of 40m depth were drilled at the northern part of the city. In every
borehole, Standard Penetration Test (SPT) was applied at equal intervals of 1.5m and disturbed
samples were collected in order to determine the soil properties. The dynamic properties of typical
soils in Duzce have been determined trough resonant column tests perform by Assoc. Professor
Th.Tika in Aristotle University (Pitilakis et al 2006).

SYNTHESIS OF GEOPHYSICAL AND GEOTECHNICAL DATA

Based on all the available geological, geotechnical and geophysical data and considering especially on
the deep boreholes drilled in Duzce and the results of the dynamic laboratory tests, it was possible to
estimate the soil stratigraphy of the area by producing 2D geotechnical cross-sections along Duzce
(Figure 7). Representative 2D geotechnical cross-sections, along the NS and EW direction, are
illustrated in Figure 8. Results from Kudo et al., (2000) and Rosenblad et al., (2001), as well as results
from the microtremor measurements campaign conducted by Yamanaka et al., (2002) and Tromans,
(2004) were also taken into account for the construction of the 2D geotechnical cross-sections (Figure
8). According to geological and geotechnical evidence and relative geophysical information (Kudo et
al., 2000) the rock basement should be very deep (a few hundred meters), while Aydan et al., (2000)
states that the seismic bedrock’s depth varies between 50m to 250m. Yamanaka et al., (2002) have
come to the same conclusion, since according to the assumed 7-layer model which he is proposing, the
seismic bedrock is defined around 200-250m depth at the centre of Duzce basin, where the shear wave
velocity values are estimated around 1000 m/sec.

In the frame of the Pilot Microzonation Study, the compilation of a number of aftershocks at the
record site (Meteorological Station of Duzce) and the estimation of the HV spectral ratio has been
conducted in order to define the outcropping bedrock’s depth, the results of which are analytically
presented in Figure 9b. As a result of the HV spectral ratio computation, the predominant HVSR
frequency was estimated at the value of fo = 0.75 Hz, and it was possible to estimate the depth of the
“seismic bedrock” which should be approximately at -120m. Taking under consideration that the
estimated value of the predominant frequency fo is in full agreement with the fp value estimated by
Tromans, (2004) who performed array measurements of microtremors in the field and also validating
this results with data from topographic maps of the area as well as with published reports [Kudo et al.,
(2000), Rosenblad et al., (2001) and Yamanaka et al., (2002)], we have finally accepted that the
“seismic bedrock” for the site effects analyses, (Vs≥750m/sec), lies in this depth of around 120m
(Figure 9a).

From the 2D geotechnical cross-sections (Figure 8) it is evident that at the southern part of Duzce the
shear wave velocity at the surface (10m) varies between 250-300 m/sec and the soil configurations

Figure 8. NS-left and WE 2D geotechnical cross-sections along Duzce.

 14/2/2000 Spectral Ratio (NS/V) Meteorological Station

Record 12/11/1999
100
22/3/2000
23/3/2000

27/3/2000

0.0m Vs=200m/s2
14/6/2000

ML NSPT=5-12 10
6/7/2000

-3.0m 7/7/2000

SC-SM Vs=250m/s2

Amplitude
-10.0m NSPT=12 1

SM-GM
0,1

Vs=350m/s2 0,1 Frequency (fo) 1 10

NSPT=>50 10
AVERAGE ± STEV. (NS/V)

-30.0m M.T.
M.T.+T.A.
M.T.-T.A.

Amplitude
GM 1
Vs=450-520m/s2
NSPT>50

-120.0m
Rock 0,1
0,1
Fre quency (fo)
1 10

Figure 9a. 1D soil profile at the Figure 9b. H/V spectral ratio of the aftershocks
Meteorological Station of Duzce. recorded at the Meteorological Station of Duzce.

consist from soft alluvial clay, loose sand and gravel river sediments. At depths between 10-30m, the
shear wave velocity values increase up to 300 m/sec, and the soils are composed of alterations of silt
and clay with generally soft–medium consistency and loose–moderately compact fine-grained sand.
Finally, at the depth between 30-120m, the shear wave velocity values (Vs) have been estimated
around 400-450 m/sec, and the soil gradually change into relatively moderately stiff-stiff clay or silt
and moderately compact to very compact sand, clay and gravel. Based on all the above data and
information we estimated 1D soil profiles, (Figure 9a), to perform a set of 1D equivalent linear
analyses.

SITE EFFECT ANALYSES

A series of 1D EQL analysis has been performed in Duzce on thirty (30) typical soil profiles (Figure
10). 1D analysis must be quite suitable for the case of Duzce as the plain is very large and there are not
apparently strong lateral discontinuities except at the souther part at the fault area. The dynamic soil
properties and the soil profiles were derived from the synthesis of previously mentionned surveys and
studies (Pitilakis et al., 2006). Considering the fact that there was no record available of the 12-11-
1999 Duzce earthquake in “outcropping ” conditions in or near Duzce, in order to be used as the input
motion, the selection of the appropriate input motion was a major problem. This problem was
encountered by proceeding at first in the deconvolution of the available record at the Meteorological
Station of Duzce in order to estimate the 12-11-1999 record at the “seismic bedrock”. Except the
deconvoluted time history of the Duzce earthquake record, four (4) other seismic motions recorded in
“outcropping ” conditions were also selected, among them the Gebze record of the 17-8-1999 Kocaeli
earthquake, two records from the 1994 Northridge earthquake (Pacoima Dam and Wonderland
records) and the Sturno record of the 1980 Campano Lucano earthquake in Italy. The specific
selection of these five input motions was based on the processing and analysis of all aftershocks
recorded at the Meteorological Station, and in other sites around Duzce (Pitilakis et al., 2006).

The computed normalized acceleration response spectra in each site were compared to the design
response spectra acceleration of the Turkish Seismic Code and EC8. A set of thematic maps were
produced illustrating the spatial variation of peak ground acceleration PGA (g), (Figure 11), and peak
spectral acceleration PSA (g) for various period values (To=0,3 sec and To=0,6 sec), (Figure 12).
These maps led to the seismic zonation for the city of Duzce, and aim to investigate whether the parts
of the city in which peak ground motion values were observed, suffered as well extensive damage.
 Figure 10. Location of 1D soil profiles sites in Duzce.

As it is observed in Figure 11a, the computed PGA values vary between 0.37–0.50g when 12-11-1999
Duzce deconvoluted time history is used as input motion, and between 0.39-0.56g, when the average
values of the five input motions used are depicted (Figure 11b). The maximum PGA values are
observed at the south-west and south-east part of the city, mainly around and close the Metereological
Station (DZC station) and around the river. Comparing these results with the effective ground
acceleration coefficient (Ao=0.40g) that applies for Duzce according to the Turkish Seismic Code, we
conclude that they are generally in good agreement, although both recorded and computed PGA values
reach 0.50g at the southern parts of the city where the soft river sediments are dominant. The average
value for the whole city of Duzce is effectively of the order of 0.40g. Some spots of higher values are
observed at the rivers zone where actually more damages were observed. Considering now the
computed PSA values they vary between 0.75g and 1.5g for To=0,3 sec (Figure 12a) and between
0.85g and 1.3g for To=0,6 sec (Figure 12b). The highest spectral values at these two characteristic
periods are concentrated at the area of major damages and certainly this is a persuasive observation of
the role of local soil conditions on the magnitude and the spatial distribution of damages.

(a) (b)
Figure 11. Spatial distribution of the computed PGA (g) values, EW component
(α) Deconvoluted 12/11/1999 Duzce record (b) Average values of the 5 input motions used.
 (a) (b)
Figure 12. Spatial distribution of the average computed PSA (g) values (EW component)
(a) To=0,3 sec (b) To=0,6 sec.

VULNERABILITY ASSESSMENT OF THE WATER SYSTEM IN DUZCE

The aim of a typical vulnerability analysis of a water system is to assess its seismic performance, to
identify and localize the weak links-pipes and to make a pre-assessment of the serviceability reduction
in case of a strong earthquake, like the 12/11/1999 that struck the city of Duzce. A detailed inventory
of the network is the basic parameter defining the accuracy of the assessment and it has to be admitted
that this is a very difficult task almost in all over the world. However in Duzce we had a reasonably
good inventory and the whole network has been made in a GIS format. The vulnerability is then
calculated using appropriate relations that correlate the seismic load with the repair rate (RR/km).
Fragility curves describe the probability of reaching or exceeding each damage state given the level of
ground motion or ground permanent displacements caused for example from liquefaction.

A set of well documented water system failures in Duzce caused by the 12-11-1999 M7.2 earthquake
have already been presented in Figure 5 (Alexoudi, 2005). In Figure 13a, the above set of failures are
plotted with the average reported damages per district (mahalle). This set of data is one of the rare
cases of reported damages in water systems in Europe and Turkey. In the following paragraph we
examine if the available vulnerability relationships and the associated damage states could reproduce
accurately the reported damages in Duzce. The necessary spatial distribution of PGV for the M7.2
earthquake in Duzce has been estimated from the pilot microzonation study using as outcrop input
motion the deconvolution of the recorded ground motion at the Metereological Station. After proper
validation it was decided to apply in this study the fragility function of O‘Rourke & Ayala (1993)
proposed also in HAZUS 2004. This is an empirical relationship based on data collected from actual
pipeline damages observed in four USA and two Mexican earthquakes. Two damage states for
pipelines were considered, leaks and breaks. The Repair Rate according to O‘Rourke & Ayala (1993)
for wave propagation is given as follows:

Repair Rate [Repairs/Km] ≅ K1*0.0001* (PGV) 2.25 (1)

• PGV (cm/sec) peak ground velocity values

• K1: coefficient depending on the type of pipeline (brittle, ductile)
For ductile pipelines (steel, ductile iron and PVC), K1=0.3, while for brittle pipeline (asbestos cement,
concrete and cast iron pipes) K1=1.0. Welded steel pipes with arc-welded joints are classified as
ductile, while welded steel pipes with gas-welded joints and pre-1935 steel pipes are classified as
brittle.
 Legend

REEBRU
waterfailure
1
¯ Legend

REHAZDUZ
waterfailure
1
¯
Low Low
Moderate Moderate
High 3 5 5
High 3
2 2
4 4

7 8 7 8

6 10 6 10

12 12
9 11 9 11
13 13
18 14 18 14

16 17 [ 19 20 21
16 17 [ 19 20 21

24 22 15 24 22 15
27 23 27 23
25 25
28 29 28 29
26 26

0 600 1,200 2,400 3,600 4,800 0 600 1,200 2,400 3,600 4,800
(a) km
(b) km

Figure 13. Mahalle with low (grey), moderate (green) and high (red) damages as result of
Duzce earthquake (a) Marmara damage database, (2005), (b) Fragility curve: O’Rourke &
Ayala, (1993) (With red color are illustrated the documented specific failures/ mahalle).

In Figure 13b the computed damage states (low, moderate and major damages) are illustrated per
district, applying as already mentioned before the O‘Rourke & Ayala (1993) fragility relation for
pipeline failures.

In Figure 14a the number of water pipe damages per district are presented, while in Figure 14b the
computed damages in the water system of Duzce are depicted. In general, it is observed that
O’Rourke & Ayala (1993) relation presents a good correlation of damages both in regards of the total
number of damages and their spatial distribution, as it is illustrated in Figure 13. Most damages are
observed at the southern district and the western part of Duzce which is in good agreement with the
actually reported damages. Soil conditions at these area are rather soft and loose which produced
higher peak ground velocities and consequently damages as well.
g

¯
0-4 [ katagrafi
5 - 10 PGV_(cm/sec)_DuzceEW
11 - 15 High : 85.92cm/sec
16 - 19 Medium : 75.56cm/sec
Low : 65.20cm/sec

¯
20 - 34
FINAHAZDAM
35 - 64 break
waterpipes leak
full-function

[ [

0 600 1,200 2,400 3,600 4,800 0 500 1,000 2,000 3,000 4,000
(a) km (b) km

Figure 14 Vulnerability assessment of water system in Duzce after 12-11-1999 earthquake

(a) Average number of damages per district (mahalle), O’Rourke & Ayala (1993), (b)
Specific spatial distribution of damages using the O’Rourke & Ayala 1993 relation and the
results of the pilot microzonation study in terms of ground velocity.
 CONCLUSIONS

Duzce, located in north-west Turkey, suffered great damage during the 12/11/1999 strong earthquake
(Mw=7.2) that hit the area. The study presents the main results of the pilot microzonation study
performed in Duzce after the earthquake and investigates the role of soil conditions to the spatial
distribution and the intensity of reported damages in buildings and the water system. The correlation
of the damage intensity and spatial distribution of damages for the residential buildings and the water
system with the recorded ground motion and the computed spatial distribution of ground motion is in
general very good. This fact is encouraging concerning the reliability of methodology applied and the
fragility relationship used.

REFERENCES

Alexoudi M, “Contribution to seismic assessment of lifelines in urban areas. Development of holistic

methodology for seismic risk”, PhD Thesis, Civil Engineer Department, Aristotle University of
Thessaloniki, 2005 (in Greek)
Ambraseys NN, “The prediction of earthquake peak ground acceleration in Europe”, Publ. Earthquake
Engineering and Structural Dynamics, 24, 467- 490, 1996
ArcGIS 9.0, ESRI
Aydan O, Ulusay R, Kumsar H and Tuncay E, “Site investigation and engineering evaluation of the
Duzce- Bolu earthquake of November 12, 1999”, TDV (Turkish Earthquake Foundation), Report
No DR 09-51, 220p., 2000
Crisis 99, M. Ordaz, Institudo de Ingenieria de la U.N.A.M., 1999
Cyberquake, Modaressi H, BRGM, 1998
Donmez C and Pujol S, “Spatial Distribution of Damage Caused by the 1999 Earthquakes in Turkey”
Earthquake Spectra, 21, No.1, 53–69, February 2005
Erdik M, “Report on 1999 Kocaeli and Duzce (Turkey) Erathquakes”, Bogazici University, Dept. of
Earthquake Engineering, Istanbul, 1999
Kayabali K and Akin M, “Modeling of seismic hazard for Turkey using the recent neotectonic data”,
Engineering Geology 63, 221-232, 2002
Kayabali K, Ulugergerli E and Tokgoz E, “Geological-geophysical-geotechnical survey report of 246
hectare area in provincial centre of Duzce based on municipal plan”, vol.1, Ankara University,
Faculty of Science, Dep. of Geology and Geophysical Engineering, 2001
Kudo K, Kanno T, Okada H, Ozel O, Erdik M, Takahashi M, Sasatani T, Higashi S and Yoshida K,
“Site specific issues on strong ground motion during the Kocaeli, Turkey Earthquake of August 17,
1999, as inferred from array observations of microtremors and aftershocks”, BSSA, Special Issue,
2000
Ministry of settlement and public works of turkey, general directorate of disaster affairs, “Duzce city –
No.1 Section (3444 hectare–central business area and close surroundings) in the scope of
reconstruction works after 12 November earthquake”, Geological–geotechnical investigation
report, 2000
O'Rourke MJ and Ayala G, “Pipeline damage due to wave propagation”, Journal of Geotechnical
Engineering, ASCE. 119 (9), 1490-1498, 1993
Ozbey C, “Empirical peak horizontal acceleration relationship for northwestern Turkey”, MSc Thesis,
Bogazici University, Istanbul, 2000
Pitilakis K et al, Final Report Marmara Earthquake Rehabilitation Project - An Integrated Intervention
for the Rehabilitation of Duzce, 2006
Rosenblad BL, Rathje EM and Stokoe KH, “Shear Wave Velocity Profiling by SASW Method at
Selected Strong-Motion Stations from the 1999 Turkey Earthquakes”, Final Report to Pacific
Earthquake Engineering Research Center, September 2001
Sadigh et al, “Attenuation Relationships for Shallow Crustal Earthquakes Based on California Strong
Motion Data”, Seismological Research Letters, 68, January/February 1997
Sucuoglu H, “Engineering characteristics of the near-field strong motions from the 1999 Kocaeli and
Düzce Earthquakes in Turkey”, Journal of Seismology, 6, 347 – 355.3, 2002
Tromans J, “Behaviour of buried water supply pipelines in earthquake zones”, PhD Thesis, Imperial
College of Science, Department of Civil and Environmental Engineering, University of London,
2004
Turkish ministry of public works and housing, Map of Earthquake Zones, General Command of
Mapping Publication, 1996
Yamanaka H, Kato M, Hashimoto M, Gulerce U, Iyisan R and Ansal AM, “Assessment of Seismic
Local-Site Effects at Plural Test Sites”, Final Report to the Ministry of Education, Science, Sports
and Culture, Research Grant No.11694134, 2002