Bosporus Bridge Traffic Operation Techniques Using Real-time Earthquake Information to Mitigate the Risk Involved

Abstract

One of the two bridges connecting Asia to Europe, Bosporus Bridge in Istanbul/Turkey, will be affected by the expected severe earthquake from the underneath of the Sea of Marmara in near future. As the traffic density on the bridge corresponds to the busiest in Turkey, utmost effort must be paid to keep the lives and casualties at the possible lowest level. No research hitherto has been conducted to explain the concept of risk management with regard to the lives of those people travelling on the bridge to be saved by combining both traffic management techniques and earthquake early warning system technology. This paper investigates the traffic operation techniques on Bosporus suspended bridge when a pre-known time period is available for the earthquake. Furthermore, this paper focused also on the issues of the strategies to manage the traffic by investigating the occurrence probability of the danger zone lengths and manipulating the average speeds of the vehicles on the bridge. Practical guideline and countermeasure strategies are offered through the use of real time earthquake information. The results indicated that to increase current average travel speed, 45 km/h, on the bridge would make tremendous changes to mitigate causalities.

Full text

DISASTER SCIENCE AND ENGINEERING p. 17-24, 1(1), 2015
17
Received: 06.07.2015 Accepted: 13.07.2015
1
Sakarya University, Department of Civil Engineering, Sakarya, Turkey, skuyuk@sakarya.edu.tr
2
Bogazici University. Kandilli Observatory and Earthquake Research Institute, Istanbul, Turkey, serdar.kuyuk@boun.edu.tr
3
Sakarya University, Department of Civil Engineering, Sakarya, Turkey, haslan@sakarya.edu.tr
4
Sakarya University, Department of Civil Engineering, Sakarya, Turkey, muharrema@gmail.com
Abstract: One of the two bridges connecting Asia to Europe, Bosporus Bridge in Istanbul/Turkey, will be affected by the
expected severe earthquake from the underneath of the Sea of Marmara in near future. As the traffic density on the bridge
corresponds to the busiest in Turkey, utmost effort must be paid to keep the lives and casualties at the possible lowest level. No
research hitherto has been conducted to explain the concept of risk management with regard to the lives of those people travelling
on the bridge to be saved by combining both traffic management techniques and earthquake early warning system technology. This
paper investigates the traffic operation techniques on Bosporus suspended bridge when a pre-known time period is available for
the earthquake. Furthermore, this paper focused also on the issues of the strategies to manage the traffic by investigating the
occurrence probability of the danger zone lengths and manipulating the average speeds of the vehicles on the bridge. Practical
guideline and countermeasure strategies are offered through the use of real time earthquake information. The results indicated that
to increase current average travel speed, 45 km/h, on the bridge would make tremendous changes to mitigate causalities.
Index Terms— Traffic management, Bosporus Bridge, earthquake early warning, risk management
I.
I
NTRODUCTION
1
STANBUL, being the most crowded city in Turkey with a
15 million population, faces the major and complicated
traffic problems to be solved. The city is located at a very
high active seismic zone in the Marmara region on two
continents, Asia and Europe (Figure 1). The traffic on and in
the vicinity of Bosporus Bridge; one of the two suspension
bridges that connects Asia and Europe, has probably the
highest congestion level in Istanbul, thus in Turkey. When
the traffic congestion is a matter, Bosporus Bridge-related
traffic might be seen as the one having almost all the negative
aspects of congestion. The high possibility of having a major
earthquake in the near future adds the reliability dimension
to the congestion problem of the bridge traffic already
available.
The reliability of transportation systems mentioned has
two main directions: Connectivity and Travel Time
reliability (Bell, 1998; and Iida, 1999). While connectivity
reliability concerns the physical connections of the nodes in
the networks, travel time reliability deals with the
performance of the networks by investigating the possibility
of making journeys between origin and destination within an
acceptable time limit.
Figure 1. Tectonic models of the Marmara Sea region. (changed from
Laigle et al, 2018)
The reliability of Bosporus Bridge in this research,
however, is different from these two aspects of the
transportation reliability concepts explained. It is basically
related to the possibility of having least life-loses on
acceptance of the fact that the expected earthquake is so
strong that it might demolish the bridge in service to some
extent. The main question remains as; can the traffic on the
bridge be managed in a way that, the road users get the least
possible adverse effects, especially in terms of lives to be
lost, in case earthquake hits the bridge within a pre-known
time period? Altough this question is asked for Bosporus
Bridge in this particular study; it is valid for any bridge,
tunnel, and metro passing a danger zone such as Marmaray
Bosporus Bridge Traffic Operation Techniques
Using Real-time Earthquake Information to
Mitigate the Risk Involved
Huseyin Serdar Kuyuk
1,2
, Hakan Aslan
3
, Muharrem Aktas
4
I

DISASTER SCIENCE AND ENGINEERING p. 17-24, 1(1), 2015
18
in Istanbul, Bay Area Rapid Transit (BART), Bay Bridge,
and Golden Gate Bridge in San Fransisco or Rainbow Bridge
in Tokyo etc (Aktas et al. 2010, Aslan et al. 2011, Kuyuk et
al. 2011).
A. Seismicity of Marmara Region And Earthquake
Early Warning System
The tectonic processes forming the Sea of Marmara and
its surrounding area have been controlled by the North
Anatolian Fault Zone (NAFZ). NAFZ is a dextral strike-slip
fault zone, extending from the Karlıova region to the Gulf of
Izmit along Anatolia, and south of Thrace as the Ganos Fault
(Sengor A, 1985). There have been many micro and macro
tectonic earthquakes occurred along NAFZ and its segments
in the vicinity of Istanbul.
The occurrence and the hypocentral information of huge
events in the area, one of the largest disasters affected the
region was 17 August 1999 Golcuk earthquake with Mw 7.4,
indicate that a likely earthquake is forthcoming in near future
from the underneath of the Sea of Marmara (Figure 1). It is
expected that the next earthquake would result in enormous
undesirable consequences. On the other hand, fortunately,
this seismically vulnerable city developed a reliable
Earthquake Early Warning System (Figure 2, Erdik M.,
2003) that could lessen the adverse consequences of the
threatening earthquake.
Figure 2. Distribution of early warning stations and 2.4 GHz spread-
spectrum radio modem transmission through repeater stations (filled
rectangles). (Alcık et al 2011)
Early warning systems give warnings of upcoming danger
by rapid estimation of the earthquake source parameters
(Kuyuk et al, 2014, Kuyuk and Allen, 2013a, 2013b). To do
so, systems use the capability of modern real-time systems
to process and transmit information faster than seismic
wave’s propagation (up to 8 km/s). The possible warning
time is usually in the range of up to 70 seconds, depending
on the distances among seismic sources, seismic sensors and
user sites. As the city is getting now ready for further North
Anatolian Fault System (NAFS) earthquakes, it is also vital
to prepare traffic management plans and take precautionary
actions to moderate the casualties.
Kuyuk (2010, 2015), investigated available time assuming
possible hypocentres underneath of Marmara Sea. Available
time is defined as the time provided by EEWS before strong
ground motion hits a place. He considered 486 simulated
earthquakes with three different depths in the region. The
available times to take action in case of earthquake range
from 2.4 to 31.1 seconds. The average travel time based on
available time in terms of time difference between S-wave
arrival to Bosporus Bridge and P-wave arrival to front station
is calculated according to Eqn. 1.
 

n
i
n
iiavrg
itt
11
(1)
where
avrg
t

is average available time, i is the earthquake
number and
i
t

is available time depending on each i. The
average available time is calculated as 13.45 seconds on
assumption of the fact that the occurrence probability of each
earthquake is equal. Although this average time seems small
estimated by difference between P- and S-wave velocities,
there would be additional couple of seconds before taking
actions considering arrival of peak ground acceleration or
total/partially collapse time of bridge.
B. Structural and Traffic Properties of The Bosporus
Bridge
The Bosporus Bridge, in service since 1973, has a length
and height of 1071 m and 165 m, respectively and 6 traffic
lanes (3+3). The average daily traffic on both directions is
about 190.000 veh/day throughout a year (www.kgm.gov.tr)
The Figure 3 illustrates the fluctuating nature of the traffic
volumes available on the bridge for different days of the
months.
Figure 3. Light gray line shows average number of vehicles per day for
one direction for twelve months. Red line shows average.
Red line shows the average daily traffic for one direction.
One way traffic drops 70000 veh/day and rose above 100.000
veh/day during the different parts of the year. The same
unstable nature of the traffic can be observed as far as the
hourly volumes are concerned as shown in the Figure 4.
Although the average hourly traffic (around 3850 veh/h) is
already quite high, the peak hour traffic volume of 6000-
6200 veh/h (1700 - 1800 ) corresponds the main proportion
of the figure causing unbearable queues with extremely high
travel times.

DISASTER SCIENCE AND ENGINEERING p. 17-24, 1(1), 2015
19
Figure 4. Typical hourly traffic available on the Bosporus Bridge.
Number of vehicles per hour for one direction. The average speed estimated
from daily traffic volumes is seen the same figure from 8.00 to 24.00.
Apaydin, (2010) studied on the dynamic properties of the
bridge and maximum transverse and vertical displacements
of the bridge are calculated as 1.36 and 1.154 meter
respectively at mid-span. The distribution of the bending
moment at the apron was also calculated by Gundaydin, et.al
(1997) and plotted in Figure 5. These findings verify that
possible maximum damage will happen at the mid-span as
expected. Thus expectable danger zone would be between
100 to 200 m in both directions from mid-span.
II. M
ETHODLOGY
A. Setting up The Mathematical Structure and Cost
Matrix of the Model
As to determine the best strategies in terms of the speeds
on the bridge, a cost matrix was set up to illustrate the
number of people to be affected if they are the ones on the
danger zone when the earthquake hits the bridge. This matrix
has the possible speeds of the vehicles as its rows and lengths
of the danger zone as its columns. Danger zone describes the
critical sections of the bridge in terms of failure and collapse.
Determination of the values of the matrix is based on the
very well-known model suggested by Greenshields (1935).
This model, being one of the macroscopic approaches to
relate the speed and density of the traffic, hypothesized that
a linear relationship existed between the two parameters of
speed and density. The speed that is used in the algorithm is
the space-mean speed which is the harmonic mean of the
speeds of the vehicles passing a point on a highway during
an interval of time. This speed is obtained through the
division of the total distance on a section of highway by the
total time required for two or more vehicles to travel this
distance. The density, on the other hand is the number of
vehicles travelling over a unit length of highway at an instant
in time.
With these explanations, the mathematical structure
is expressed by Greenshields as follows.
k
k
u
uu
j
f
fs

(2)
where;
s
u
is the space-mean speed of the vehicles
corresponding the density of k
f
u
is the maximum speed when the density is at its
minimum, i.e., 0
j
k
is the jam density
The values of this matrix are determined through a
design vehicle of 4.5m length with four (4) occupants
travelling. Thus, the original matrix represents the numbers
when related speed and corresponding length of danger zone
are the case to represent the real cases as if the earthquake hit
and those speed and length values occurred in real life. In
other words, each cell value is determined by assuming that
the probability of speed and danger zone for that specific
value being one (1).
As one of the main objectives of this research is to
investigate and establish the best set of possible strategies to
manage the traffic to minimize the possible numbers of dead
and injured people, the probabilistic distribution of both
speed and length values was employed to include all
different possibilities and scenarios. Hence, a new cost
matrix was suggested to model these situations. Each
expected cost element (EC) of this matrix is calculated
through
ij
j
i
l
j
i
v
CPPEC 
(3)
where;
i represents the number of the speeds
j represents the number of the danger zones
j
i
v
P
is the probability of i.th speed when the length
of j is the case
j
i
l
P
is the probability of length of the danger zone
when the speed of the vehicle is i is the case
C
ij
is the cost matrix values.
The cost matrix elements of the model of this
research, thus, are related to the probabilistic distribution of
the speeds and length of the danger zones. Four different
types of cases were employed to represent the probabilistic
occurrence of the length of the danger zones as shown in
Figure 6. Probability distribution of danger zone distance is
assumed to be an exponential function. This is due to the
historical data related to the occurrence and magnitude of
earthquakes and realistic evaluation of the fact that stronger
earthquakes cause higher damages. The occurrence
probability of 200m danger zone, for instance, is higher than
500 m danger zone because the probability of the occurrence
of earthquake with Magnitude 5 is higher than the occurrence
of earthquake with magnitude 7.5.

DISASTER SCIENCE AND ENGINEERING p. 17-24, 1(1), 2015
20
Figure 6. The probabilistic distribution of the Cases
While Case 1 covers this approach, other cases with
different standard deviation and mean values, shown in
Figure 6, are also assessed for comparison purposes and to
investigate the boundaries of this approach. The Figure 7
depicts the lognormal structure of the averaged space-mean
speed of the vehicles within the scope of this research. This
approach finally produced the concept of “Total Expected
System Cost” formulated in the following equation.
Figure 7. Lognormal distributions of the speeds
ij
j
i
l
j
i
v
n
i
m
j
CPPTEC  

 1 1
(4)
TEC, here is the total expected cost of the whole system.
III. R
ESULTS
The Table 1 indicates the calculated cost values for these
real case scenarios estimated by Eq 2. Rows designate the
speeds and columns depict possible danger zones. For
instance, the number of causalities, cost value, is 291, for a
200 m danger zone with a 50 km/h speed. This means that
when the average speed ascends to 90 km/h for the same
danger zone, causalities will drop to 97, clearly diminishing
the casualties by 67%.
In Figure 8, four different graphics sum the general
pictures of the expected cost values of the problem in detail
regarding the four different cases respectively. The darker
red the cells get, the higher the expected cost of the system
is obtained as far as the combinations of the speed and danger
zone distance probabilities are concerned.
Figure 8. Illustration of the expected cost matrix values for each cell and
scenario
The red cells are located in the left–bottom corner of the
first item of the graph indicate the most dangerous
combination of the probabilities in terms of Case 1. The red
section moves to the down part of the right corner of the last
item representing Case 4. The blue parts of the graphs are
the sections with the safest combinations of the speed and
danger-zone length probabilities. Therefore, the safest
strategies to be implemented and operated by the engineers
lie among the darker sections of the graphs. The total
expected cost of the system can be shown as in Figure 9 for
different probabilistic distribution of danger zone lengths. As
this figure implies, Case 4 represents the worst case scenario
with the total cost of 360 deaths. This was expected as Case
4 represents boundary condition assuming the probabilistic
variation in danger zone length is almost the same regardless
of the magnitude of the earthquake.
Figure 9. Total expected cost of the system for different cases of the
danger zone distance probabilistic distribution
The other cases along with the most realistic approach of
Case 1 resulted in lower expected cost values (total number
of people in danger). Case 1, having the minimum value of
expected cost, is in fact what is expected in real life due the
fact that the probability distributions are determined using
real data. The following Table 2 illustrates the probabilistic
cost values of the matrix for the distribution type of Case 1.

DISASTER SCIENCE AND ENGINEERING p. 17-24, 1(1), 2015
21
Table 1
:
System Cost Matrix Values
Velocity
(km/h)
Danger Zone Length ( m )
125
150
175
200
225
250
275
300
325
350
375
400
425
450
475 500
20 218
273
327
382
436
491
545
600
654
709
764
818
873
927
982
1036
1091
30 194
242
291
339
388
436
485
533
582
629
678
727
775
824
872
921 969
40 170
212
255
297
339
382
424
467
509
551
594
636
678
721
763
805 848
50 145
181
218
254
291
327
364
400
437
473
509
546
582
619
655
692 728
60 121
151
182
212
242
273
303
333
364
394
425
455
485
516
546
576 607
70 97
121
145
170
194
218
242
267
291
315
339
364
388
412
436
461 485
80 73
90
109
127
145
164
182
200
218
236
255
273
291
309
328
341 364
90 48
61
73
85
97
109
121
133
146
158
170
182
194
206
218
231 243
100 24
30
36
42
48
55
61
67
73
79
85
91
98
104
110
116 122
110 8 10
12
14
16
18
20
22
24
26
28
30
32
34
36
38 40
Table2. System cost matrix values in terms of probabilistic approach
Velocity
(km/h)
Danger Zone Length ( m )
100
125
150
175
200
225
250
275
300
325
350
375
400
425
450
475
500
20 0.95
1.06 1.14 1.19 1.22 1.23 1.23 1.22
1.20 1.18 1.16
1.13 1.10 1.07 1.04
1.02 0.99
30 2.87
3.19 3.43 3.58 3.67 3.71 3.71 3.68
3.63 3.55 3.48
3.40 3.31 3.23 3.14
3.06 2.99
40 3.26
3.63 3.90 4.07 4.16 4.21 4.20 4.18
4.11 4.04 3.95
3.86 3.76 3.66 3.57
3.47 3.39
50 2.73
3.04 3.27 3.41 3.50 3.54 3.54 3.51
3.47 3.40 3.32
3.25 3.16 3.09 3.00
2.93 2.86
60 2.00
2.22 2.40 2.50 2.56 2.59 2.59 2.56
2.53 2.48 2.43
2.37 2.31 2.26 2.20
2.14 2.09
70 1.33
1.49 1.59 1.67 1.71 1.72 1.72 1.71
1.69 1.65 1.62
1.58 1.54 1.50 1.46
1.43 1.39
80 0.82
0.90 0.97 1.01 1.04 1.05 1.05 1.04
1.03 1.01 0.99
0.96 0.94 0.91 0.89
0.86 0.85
90 0.43
0.49 0.52 0.54 0.56 0.56 0.56 0.56
0.55 0.54 0.53
0.52 0.50 0.49 0.48
0.47 0.46
100 0.17
0.19 0.21 0.22 0.22 0.23 0.23 0.22
0.22 0.22 0.21
0.21 0.20 0.20 0.19
0.19 0.18
110 0.05
0.05 0.05 0.06 0.06 0.06 0.06 0.06
0.06 0.06 0.06
0.05 0.05 0.05 0.05
0.05 0.05
In this table, a lognormal - probability distribution of the
vehicle speeds when crossing the bridge is assumed in
accordance with the probability of occurrence of earthquake
for the calculation of each cell.
IV. D
ISCUSSIONS
R
EGARDING
T
RAFFIC
M
ANAGEMENT OF THE
B
RIDGE
Once the earthquake information is released by EEWS, the
proper messages are conveyed to the drivers on the bridge to
guide them if they are required to stop in case they are in so-
called safe zone area or to drive as fast as they can to reach
the safest part of the bridge when they are in danger zone.
Because the queues normally occur behind the danger zone,
this seems to provide a quite good chance for the drivers in
danger zone to have relatively free and fast forward
movement opportunity for quick evacuation from the danger
zone.
It should, however, be mentioned that the success of this
strategy on the one hand depends heavily on the fact that the
information obtained from EEWS is displayed on the
Variable Message Signs existing on the bridge without losing
even a single second, on the other hand the reasonable and
effective responses of the drivers to the messages they get.
This second one, without any doubt, requires teaching and
training of the bridge users about how they should behave if
they are the ones on the bridge when the earthquake hits.
During the peak hours, the queues occur right before the
toll points back to the kilometers behind. Once the road users
(no pedestrian access is available since 1978) get on the
bridge, the average travel speed is at about 45km/h (12.5
m/sn) make the bridge to be passed completely at 86 seconds.
As the pre-time available through EEWS system to evacuate
the bridge entirely is not long enough (15-20 seconds), the
basic strategy that can be applied to bridge traffic right before
the destructive earthquake is to make sure that the vehicles
close to the foots of the bridge stay where they are and the
vehicles in the danger zone (Figure 10) move as quick as
possible to pass the middle section possibly the weakest part
of the bridge to leave or get the nearest possible most strong
part of the bridge. The possible back turning movements just
before the bridge seem to be infeasible as the traffic at the
back is positioned bump-to-bump in the busiest times giving

DISASTER SCIENCE AND ENGINEERING p. 17-24, 1(1), 2015
22
no opportunity for vehicles to move backwards in order to
make more space for the vehicles already on the foots of the
bridge.
Figure 10. Illustration of the traffic movements on the bridge
The movement of the vehicles will be managed through
variable sign messages located critical points of the bridge.
The critical points correspond to the locations where the
danger zone of the bridge starts. However, beyond the
application of these signals on top of bridge, we advise to
control entrance of bridge, too. This might be managed, as
done Bay Bridge in California through Oakland to San
Francisco, by stopping vehicles before entering bridge.
Although one of the purposes in Bay Bridge is to merge lines
before entering by red and green lights, this approach also
increases the speed. This has two benefits 1) EEWS info can
be conveyed via signalling, thus preventing any further
vehicle movement into bridge 2) most critically, to increase
average speed on the bridge so that danger zone will be
passed faster. Unlike from Bay Bridge we think that it would
be much advantageous to put signals 3-4 km before entrance.
By doing this, the reaction time will be increased and there
will be fewer cars between signals and bridge foots
increasing the possibility to drive backward from the bridge.
However this might not valid in busy hours.
The Figure 11 shows the required time to travel the
corresponding distance in terms of different travel speeds.
This figure clearly indicates the fact that with average travel
speed of 12.5 m/s (45 km/h), the available time of 13.45
seconds provides 76.7 % of the vehicles on danger zone to
get the safe sections of the bridge. If the average travel speed
is assumed at about 8.33 m/s (30km/h) representing the rush
hour 50.9 % of the vehicles can still be on the safe part of the
bridge. On the other hand, with the speed of 19.44 m/s (70
km/h) assumed to be the average speed at the off-peak
period, all vehicles can safely evacuate the danger zone.
Nevertheless, all the proper measures should be taken to
prevent the vehicles at the stay-zone from entering the
danger zone.
Moreover, it should not be forgotten that, there will be
more time than the estimated one for the evacuation of
danger zone due to the fact that the bridge will not collapse
with the arrival of initial S wave since the bridge will
continue to oscillate even though the earthquake stops. On
top of that, additional time for evacuation of danger zone
would be available by installing high-tech sensor closer to
the fault line underneath of the Sea of Marmara in future.
Figure 11, gives possible seconds required to evacuate the
bridge in terms of speeds and corresponding lengths of
danger zones. Assuming the danger zone to be 200 m, the
required time to evacuate this zone with the current average
speed, (45 km/h) is about 16 seconds. On the other hand
increasing speed of vehicle to 90 km/h would decrease this
required time by 50%.
Figure 11. Distances vs travel times of the vehicle in terms of average
speed
Another aspect of a warning system application to the
bridge is to educate the bridge users through media tools. For
instance, special pamphlets can be distributed when drivers
buy or renew KGS, OGS electronic tooling cards. Moreover,
drivers having email addresses in the system could be
provided with easy and effective explanations regarding how
to behave when they receive alerting massages. Special
programs could also be broadcasted by radio channels
dedicated to bridge itself.
Although various strategies could also be developed for
different traffic scenarios for different times of the day, this
is not directly related to the concept of this paper as in this
research only the applicability and importance of EEWS to
Bosporus Bridge traffic management concept is tried to be
highlighted mainly for the worst case scenario. The
guidelines presented here are thought to be applicable not
only for Bosporus Bridge. Similar approaches are valid for
other bridges in Japan, California, especially Bay Area,
Golden Gate Bridge, tunnels; Bolu Tunnel, metros; BART,
and Marmaray. Actions and precautions need to be taken are
listed below;
1) Main strategy should concentrate on to increase average
speed of vehicles on the bridge. This is crucially important
because 10% increase would save hundreds of lives.
2) Re-arrangement of the position of tolls: Distance
between tolls and bridge foots need to be as long as possible,
to increase the entrance speed of vehicles to bridge as well
as give vehicles an opportunity for possible backward
manoeuvre.
3) Application of a signalized vehicle stopping system on
both sides of the bridge entrances. Presently 7 approaching
lanes reduce to 3 lanes right after the tolls on the entrance of
bridge. This leads congestions to occur right before the feet
of bridge. By merging lanes couple of km before foots will
indisputably make the traffic flow faster and smoothly.
4) Education is the fundamental part of early warning
system applications on traffic management. Without proper
knowledge, it is almost impossible for drivers to act fast

DISASTER SCIENCE AND ENGINEERING p. 17-24, 1(1), 2015
23
within a limited time. Dedicated radio channels can surely be
used for this purpose. Moreover evacuation drills for test
purposes and simulations will help managers to act smartly
in emergencies.
6) Real-time earthquake information has to be conveyed
through panel boards. This might be done by different color
schemes. Bridge currently has night-time lighting system all
over for the purpose of scenery. This lighting might be used
to convey the messages. A red light would indicate a threat
and warn drivers not to enter the bridge. There might also be
another option to pass this information to the drivers via loud
speakers.
7) Evacuation plan is another part of the precaution
measures. The alert system need to be designed by also
considering testing and education. While testing is for health
monitoring of the whole system, education mode is for
simulation and educational purposes to which all media
might involve increasing the awareness of the system.
V. C
ONCLUSION
Istanbul, the biggest metropolitan city of Turkey, is
expecting one of the most devastating earthquakes of its
history with a huge number of buildings to be collapsed and
damaged. The expected earthquake will affect the city from
a wide spectrum of daily life, causing many people to die and
get wounded. The Bosporus suspended bridge, without any
doubt, is one of the most critical structures of the city. This
bridge is not just important as being one of the most
important connecting elements of the European and Asian
sides of the city; it also represents the highest volumes of
traffic of the city as a whole. Determination of proper traffic
management strategies will have utmost importance in order
to minimize the total number of dead and/or injured people
using the bridge for their daily travel purposes.
In this study, the Bosporus suspended bridge is
investigated and proper traffic management techniques are
suggested against the possible upcoming Marmara
Earthquake. This research is believed to be one of the pioneer
studies in the application of EEWS to bridge traffic
management. Although, the assumptions made are quite
reasonable to get the broad picture of the problem,
consideration of analysis of collapse mode along with the
probability of earthquake occurrence in each possible
location would make the solution approach more powerful
and effective.
A
CKNOWLEDGMENT
We thank Bosporus Bridge engineers, Mr. Baki Erdogan
and Ertac Celikel, and Mr. Ramazan Yuksel from General
Directorate of Turkish Highways for providing data.
R
EFERENCES
[1] Gunaydin M., Adanur S., Altunisik A.C., Sevin B., “construction stage
analysis of Fatih Sultan Mehmet suspension bridge”, Structural
Engineering And Mechanics, vol. 42, pp. 489-505, 2012.
[2] Aktas M., Kuyuk H. S., Aslan H., “traffic management for bosphorus
suspended bridge, Istanbul information based on earthquake real-time
information” 14th European Conference on Earthquake
Engineering, Ohrid, Republic of Macedonia, 30th August – 3rd
September, 2010.
[3] Alcik, H., Ozel, O., Wu, Y. M., Ozel, N. M. and Erdik, M., “An
alternative approach for the Istanbul earthquake early warning
system”, Soil Dynamics and Earthquake Engineering, 31(2),
pp. 181-187, 2011.
[4] Apaydin N. M., “Earthquake performance assessment and retrofit
investigations of two suspension bridges in Istanbul”, Soil Dynamics
and Earthquake Engineering, 30, 702–710, 2010.
[5] Aslan H., Kuyuk H. S., Aktas M., “Transportation risk analysis (TRA)
of Bosphorus Suspended Bridge, using probabilistic approach” First
International Conference on Vulnerability and Risk Analysis and
Management (ICVRAM) – Fifth International Symposium on
Uncertainty Modeling and Analysis, Maryland, USA, 2011.
[6] Bell, M.G.H., “Transportation networks: recent methodological
advances”, Selected Proceedings of the 4th Euro Transportation
Meeting, Pergamon, 1988.
[7] Erdik M., Fahjan Y., Ozel O., Alcik H., Mert A. and Gul M. “Istanbul
earthquake rapid response and the early warning system” Journal of
Earthquake Engineering vol. 1, n: 1, 2003.
[8] Kuyuk H. S., “Available warning time for emergency response in
Sakarya city, Turkey against possible Marmara earthquake” 9th
International Congress on Advances in Civil Engineering, Trabzon,
Turkey, 27-30 September 2010.
[9] Kuyuk H. S. “Warning Time Analysis for emergency response in
Sakarya City, Turkey against possible Marmara earthquake”
Challenge Journal of Structural Mechanics, 1 (2), pp. 56-64, 2015.
[10] Kuyuk H. S., Aktas M. Aslan H., “Risk analysis and real time traffic
management techniques of Bosphorus bridge by early warning
systems information” Seventh National Conference on Earthquake
Engineering, Istanbul, Turkey, 30 May-3 June 2011.
[11] Kuyuk, H.S. and Allen R.M., “Optimal seismic network density for
earthquake early warning: A case study from California”,
Seismological Research Letters, 84 (6), pp. 946-954, 2013.
[12] Kuyuk, H.S. and Allen R.M., “A global approach to provide
magnitude estimates for earthquake early warning alerts”, Geophys.
Res. Lett., 40, 6329–6333, 2013, doi:10.1002/ 2013GL058580, 2013b.
[13] Kuyuk, H.S., Allen R.M., Brown H., Hellweg M., Henson I.,
Neuhauser., “ElarmS-2: Designing a network-based earthquake early
warning system for California” Bulletin of the Seismological Society
of America, 104 (1), pp. 162-173, 2014, doi:10.1785/0120130146.
[14] Laigle, M., Becel, A., de Voogd, B., Hirn, A., Taymaz, T., &
Ozalaybey, S., “A first deep seismic survey in the Sea of Marmara:
Deep basins and whole crust architecture and evolution”, Earth and
Planetary Science Letters, 270 (3), 168-179, 2008.
[15] Iida,Y, “Basic concepts and future directions of road network
reliability analysis”, Journal of Advanced Transportation, 33 (2), 125-
134, 1999.
[16] Greenshields . B.D. “A study of traffic capacity” Highway Research
and Board Proceedings, 14, 448-477, 1935.
[17] General Directorate of Turkish Highways, http://www.kgm.gov.tr
[18] Sengor A.M.C., Gorur N., and Saraoglu F., “Strike-slip faulting and
related basin formation in zones of tectonic escape: Turkey as a case
study”, Society of Economic Paleontologists and Minerologists,
Special Publication, 37, pp. 228–264, 1985.
First Author: is now with the Department of Civil
Engineering, Sakarya University, 54187 Sakarya, Turkey (e-
mail: serdarkuyuk@gmail.com).
Second Author: is now with the Department of Civil
Engineering, Sakarya University, 54187 Sakarya, Turkey (e-
mail: haslan@sakarya.edu.tr).
Third Author: is now with the Department of Civil
Engineering, Sakarya University, 54187 Sakarya, Turkey (e-
mail: muharrema@gmail.com).