International Journal

on Marine Navigation

and Safety of Sea Transportation

Volume 3

Number 3

September 2009

295

1 INTRODUCTION

The Turkish Straits (the Straits of Istanbul and

Canakkale), which have narrow and winding shapes

that give them the semblance of a river, are one of

the most strategically important waterway systems

in the world. As the Black Sea's sole maritime link

to the Mediterranean and the open seas beyond, they

are a vital passageway not just for trade but for the

projection of military and political power. Also, their

hard to navigate geographical properties, meteoro-

logical conditions, dense and increasing transit/local

traffic, vessel/cargo characteristics, and physical

hindrances, such as cross continental bridges, energy

transfer lines, make the Straits’ traffic conditions

quite complex and risky. Moreover, this narrow pas-

sage runs through the heart of Istanbul, home to over

12 million people and some of the world’s most cel-

ebrated cultural and historical heritage.

Geographically, the Strait of Istanbul is one of the

narrowest waterways in the world. It has length of

31 kilometers with an average depth of 45 meters

(Ozturk, 1995). Its average width is 1.5 km, where

this width decreases to 700 meters at its narrowest

point (Tan & Otay, 1999). Additionally, frequent

adverse meteorological conditions, such as dense

fogs and high currents and winds, contribute to the

complexity of navigation in the Strait.

There are also some non-natural factors making

navigation through the Strait of Istanbul hazardous.

One of them is the dense local traffic, such as intra-

city passenger boats, fast ferries, fishing boats,

pleasure boats, tugboats etc. (VTS User Guide,

2004). Another important non-natural factor that

negatively effects navigation in the Strait is the fre-

quency and cargo characteristics of transit vessels.

Over 56,600 vessels (10,050 being dangerous mate-

Simulation-Based Risk Analysis of Maritime

Transit Traffic in the Strait of Istanbul

B. Ozbas & I. Or

Bogazici University, Istanbul, Turkey

O. S. Uluscu & T. Altıok

Rutgers, The State University of New Jersey, NJ, U.S.A.

ABSTRACT: In this manuscript, development and preliminary results of a simulation based risk modeling

study for the Strait of Istanbul is presented. The goal of this research is to analyze the risks involved in the

transit vessel traffic in the Strait of Istanbul. In the first step of the study, the transit vessel traffic system in

the Strait of Istanbul has been investigated and a simulation model has been developed. The model gives due

consideration to current traffic rules and regulations, transit vessel profiles and schedules, pilotage and tug-

boat services, local traffic, meteorological and geographical conditions.

Regarding risk assessment, two sets of factors are used to evaluate the risk of accident in the Strait: the proba-

bility of an accident and its potential consequences, as estimated and evaluated at various points along the

Strait. Experience has shown that maritime accident occurrences can be very dissimilar from one another and

therefore, probabilistic analysis of accidents should not be done independent of the factors affecting them.

Thus, in this study, we have focused on the conditional probability of an accident, under a given setting of

various accident causing factors. Unfortunately, historical accident data is by far insufficient for a proper sta-

tistical consideration of all possible settings of these factors. Therefore, subject-expert opinion is relied upon

in estimating these conditional accident probabilities. Assessment of the consequences of a given accident (in

terms of its effects on human life, traffic efficiency, property and environment) was also accomplished using a

similar approach.

Finally, by integrating these assessments into the developed simulation model, the risks observed by each

vessel at each risk slice are calculated in regard to the natural and man-made conditions surrounding. A sce-

nario analysis is performed to evaluate the characteristics of the accident risk as the vessel moves along the

Strait. This analysis allows us to investigate how various factors impact risk. These factors include vessel ar-

rival rates, scheduling policies, pilotage service, overtaking and pursuit rules, and local traffic density. Policy

indications are made based on the results of these scenarios.

296

rial carriers) traveled through the Strait of Istanbul in

2007.

In order to control and mitigate maritime accident

risks and improve the safety of navigation in the de-

scribed dire environment, The Bureau of Turkish

Strait’s Maritime Traffic Services (BMTS) has set

up a sophisticated Vessel Traffic Control & Moni-

toring System (VTS), (covering not only the Strait,

but also 20 miles into the Black Sea and the Sea of

Marmara) and has established and effected a set of

stringent Maritime Traffic Rules and Regulations

(R&R). The vessels arriving at the northern and

southern entrances of the Strait of Istanbul enter and

then navigate through the Strait according to the di-

rections of the BMTS, which are based on the VTS

inputs and the R&R (VTS User Guide, 2004).

Figure 1. The Strait of Istanbul

The objective of this study is to analyze the risks

involved in the transit vessel traffic in the Strait of

Istanbul. In order to achieve this, a detailed mathe-

matical risk analysis model is developed to be used

in a risk mitigation process (Uluscu et al., 2008).

Firstly, in order to study and better understand the

system, a functional simulation model of the transit

vessel traffic in the Strait of Istanbul is built. In this

simulation, which is based on the mentioned R&R,

in addition to the geographical/meteorological con-

ditions, transit and local vessel traffic in the Strait,

the current vessel scheduling practices are also mod-

eled using a specially designed scheduling algo-

rithm. This scheduling algorithm, which is devel-

oped through discussions with the BMTS

authorities, primarily mimics their decisions on se-

quencing vessel entrances, as well as northbound

and southbound traffic flow time windows (Uluscu

et al., 2009). Finally, by integrating, expert opinion

and historic data based risk assessments into the de-

veloped simulation model, the risks generated by

each vessel, are calculated in regard to the natural

and man-made conditions surrounding it (such as,

vessel characteristics, pilot/tugboat deployment,

proximity of other vessels, current & visibility con-

ditions, location in the Strait etc.), as the vessel

moves along the Strait. Preliminary results obtained

in the application of this procedure are presented and

discussed in later sections.

2 MODELING RISK

The primary objective of this study is to develop a

realistic model to assess and investigate maritime

risk imposed by the transit traffic in the Istanbul

Strait; furthermore, it is expected that such a model

and an accompanying scenario analysis will suggest

and support strategies and operational policies that

will mitigate the risk of maritime accidents that will

endanger the environment, the inhabitants of Istan-

bul and impact the economy, while maintaining an

acceptable level of vessel throughput.

Regarding the modeling of risk, first events that

may trigger an accident are identified and defined as

instigators (for example, there can be a mechanical

failure in the vessel or the captain can make a judg-

mental error, during the transit of the vessel through

the Strait of Istanbul). Through the examination of

the historical accident data and discussions with lo-

cal maritime experts, the occurrences of the follow-

ing incidents have been identified as possible insti-

gators of maritime accidents in the Strait: human

error, rudder failure, propulsion failure, communica-

tion and/or navigation equipment failure, and other

mechanical and/or electrical failure. Clearly, the oc-

currence of an instigator depends on the situation,

which may be represented by a vector of situational

attributes. Given the occurrence of an instigator, typ-

ical accidents that may occur in the Strait have been

considered and classified as, collision, grounding,

ramming, sinking and fire and/or explosion. It is also

possible to have accidents may occurring in chain,

so that a prior (1

tier) accident may cause later (2

tier) one. 1

tier accident types include collision,

grounding, ramming and fire and/or explosion, while

the 2

tier accident types include grounding, ram-

ming, fire and/or explosion, and sinking. Potential

consequences of the 1

and 2

tier accidents include

human casualty, property and/or infrastructure dam-

age, environmental damage and loss of traffic effec-

tiveness and throughput. This framework is present-

ed in Figure 2. Defining situations (factors and their

states) that affect the likelihood and/or impact level

of instigators and accidents is critical for the intend-

ed risk analysis. Such factors as called Situational

Attributes, and are divided into two groups: attrib-

utes influencing accident occurrence (vessel class,

vessel reliability, pilot request, tugboat request, visi-

bility, current, local traffic density, vessel proximity,

zone and time of the day) and attributes influencing

consequences (vessel cargo, length, zone). These

297

two groups of situational attributes (are displayed in

Figure 3 and 4.

Figure 2. The framework of the risk model

Given the above described framework, the fol-

lowing questions need to be answered in order to

quantify risks:

− How often do the critical situations occur?

− For a particular situation, how often do instigators

occur?

− If an instigator occurs, how likely is an accident?

− If an accident occurs, what would the damage to

human life, property, environment and infrastruc-

ture be?

Figure 3. Situational attributes influencing accident occurrence

Figure 4. Situational attributes influencing the consequences

In this study, answers are provided to these ques-

tions (and risk quantification accomplished) based

on historical data, expert judgment elicitation and

simulation model generated output regarding the

state of the situational attributes. The 21 slice divi-

sion of the Istanbul Strait, depicted in Figure 5 (each

slice being 8 cables long) assumed in the simulation

model, is also pursued for risk analysis purposes.

The risk at a slice is calculated based on the snap-

shot of the traffic in that slice every time a vessel en-

ters it.

Figure 5. Risk slices at the Strait of Istanbul

In order to calculate risk, the product of two sets

of factors is sought for associated with each transit:

the probability of an accident and the potential con-

sequences of this accident, during that particular

transit. Since two groups of accidents are considered

and 2

tier accidents), the expected slice risk can

be calculated accordingly.

(1)

( )

Pr 1 st tier accident type

is obtained using condi-

tional probabilities of all possible accidents given

situations (e.g. visibility) and instigators (e.g. human

error); conditional probabilities of instigators given

situations; and finally probabilities of situations.

( )

Pr 2nd tier accident type

is obtained using condi-

tional probabilities of all possible 2

tier accidents

given 1st tier accidents and probabilities of 1

tier

accident occurrences.

E Consequence type Accident type 



is obtained

using the consequence impact levels, conditional

probabilities of all possible consequences given ac-

cidents and situations and finally probability of situ-

ation.

To be able to calculate the expected risk, R, as

shown above, most of the accident and consequence

298

probabilities (conditioned on the occurrence of insti-

gators and/or state of situational attributes) are ob-

tained via elicitation of expert judgments; other

probabilities (e.g. instigator and 2

tier accidents

probabilities) are obtained from the historical data.

The specific states of the many situational attributes

are obtained from the simulation model (as the ves-

sels generated in the model move through the Strait,

in the environment also generated by the model)

Experience has shown that maritime accidents

can be quite different from one another in terms of

factors causing them. As introduced above, various

conditional probabilities of accidents are sought af-

ter in this study. Unfortunately, historical data has

been insufficient for a proper statistical analysis of

these probabilities. Therefore, expert opinion has

been relied upon in their estimation. Expert opinion

on accident probabilities is obtained through an

elicitation process using questionnaires focusing on

pairwise, uni-dimensional (one at a time) compari-

sons of factor (situational attribute) settings (while

keeping the remaining factors at pre-determined

fixed levels).

Conditional probabilities of accident consequenc-

es (in terms of low, medium or high effects on hu-

man life, traffic efficiency, property, infrastructure

and environment) are also determined through a sim-

ilar elicitation process. On the other hand, quantifi-

cation of these qualitatively defined impact levels is

accomplished through parameterization. One such

set of parameters assumed (for different levels of

consequence impacts) is presented in Table 1. These

values do not represent the actual consequence of an

accident in specific units (e.g. dollars or number of

casualties). Instead, index values representing the

experts’ perceptions of low, medium and high con-

sequences are utilized. As a result, the calculated

risk values are meaningful when compared to each

other in a given context.

Table 1. Consequence impact levels

___________________________________________________

Impact Level Value

___________________________________________________

Low Uniform(0-1,000)

Medium Uniform(4,000-6,000)

High Uniform(8,000-10,000)

___________________________________________________

Finally, these assessments are integrated into the

simulation model such that the risks observed by

each vessel, at each slice are calculated and com-

piled considering all the natural and man-made con-

ditions surrounding the slice and the vessel (such as,

vessel characteristics, pilot/tugboat deployment,

proximity of other vessels, current and visibility

conditions, location in the Strait etc.), as the vessels

moved along the Strait.

3 OBSERVATIONS

Experimentation with the aggregate simulation/risk

model described above has been accomplished

through a scenario analysis. In this regard, first the

parameter values reflecting the current situation in

the Strait, based on year 2005-2006 data (such as,

vessel arrival rates, overtake and pursuit distances,

vessel entrance schedules, local traffic density etc.)

is compiled into a “base scenario”. The risk profiles

of this “base scenario” (in terms of average slice

risks and average maximum risks), obtained using

25 replications (simulation runs) - each of one year

length, are displayed in Figure 6. The average slice

risk profile exhibits a steady behavior from the north

entrance all the way down to the Bogazici Bridge,

where the effects of the high local traffic activity in

these highly populated and busy regions of the Strait

start becoming significant. Interaction of the transit

and local traffic patterns generates a large spike in

the average risk in Slice 19 (this is the Strait region

corresponding to downtown Istanbul and including

the main harbor area) and somewhat tapers off

around the south entrance. The average maximum

risk profile also exhibits a similar behavior but fea-

turing 200 to 850 fold increases from average risks

levels observed at various points along the Strait.

This remarkable observation indicates how risky the

maritime traffic in the Strait of Istanbul can get at

specific instances. That is, depending on random re-

alizations of accident causing factors, ordinary and

safe appearance of the Strait maritime activity could

swiftly change into a very risky environment. For

example, a rare realization observed in Slice 1 (cor-

responding to risk value 12210) involved an exces-

sive level of fog during nighttime and two D-class

vessels that just entered the slice before the Strait is

closed. Another rare realization, observed in Slice 19

(corresponding to risk value 10710), involved an A-

vessel that was about to leave the Strait just after the

night schedule started, a D-vessel and an E-vessel

along with 10 local vessels. Such potentially highly

dangerous situations may be rare, but a rare disaster

is a disaster too many. So, high risks indicated by

the maximum risks should be taken seriously.

Next, a series of scenarios has been constructed

and compared against the base scenario (through the

aggregate model), in order to investigate the charac-

teristics of accident risks in the Strait under different

settings and conditions. In Scenarios 1 and 2, arrival

rate of hazardous cargo vessels are increased and

decreased. In Scenarios 3-9, vessels are scheduled

with lesser and greater pursuit distances. In Scenario

10, pilot captain service is turned off. Scenario 11

represents the case where overtaking is not allowed

within the Strait. Finally, local traffic density in the

Strait is decreased by 50% in Scenario 12. An aver-

age maximum slice risk profile is given in Figure 7.

299

This analysis has provided us with the ability to ob-

serve and predict how changes in various policies

and practices impact the risk profile of the Strait.

The results and important observations accom-

plished are summarized below.

Figure 6. Current risk profiles of the Strait of Istanbul

3.1 Observation 1

The accident risks in the Strait and the average

vessel waiting times exhibit a tight and sensitive

balance. For instance, a small increase in arrival

rates may result in rather high waiting times at the

entrances (an increase of 60% for some vessel clas-

ses). Furthermore, scheduling changes made to re-

duce vessel waiting times increase risks in the Strait

substantially. Conversely, one has to be very careful

in revising the scheduling mechanism for the pur-

pose of risk mitigation, since the waiting times are

highly sensitive to entrance rules. The benefits ob-

tained in risks may not justify the resulting waiting

times. In the future, scheduling changes may be jus-

tified, if significant reductions occur in the transit

vessel traffic, perhaps due to alternative oil transport

modes such as pipelines and other routes. Thus,

scheduling decisions to balance out delays vs. risks

should be made based on extensive experimentation

with the model developed in this study.

Figure 7. Maximum Slice Risk in Scenarios 10, 11, and 12

compared to the Base Scenario

3.2 Observation 2

The model indicates that pilots are of utmost im-

portance for safe passage, and lack of sufficient pi-

lotage service significantly increases the risks in the

Strait. Currently, vessels longer than 250 m. are

mandated to take a pilot, and it is voluntary for the

rest. As a result of our experimentation, we have

recommended mandatory pilotage for vessels longer

than 150 m. This will reduce the average risk by

7%, the average of maximum risk by 11% in Slice

19 and the observed maximum risk is 11114 ob-

served in Slice 3 (almost 7,000-fold of its average).

Had pilotage been obligatory for vessels longer than

100 m., this would reduce the average risks by 46 %

and the average of maximum risks by 33 % at Slice

19.

3.3 Observation 3

Even though current regulations discourage overtak-

ing anywhere in the Strait, results indicate that over-

taking a vessel is less riskier as opposed to requiring

a pursuing faster vessel to slow down behind a slow-

er vessel, where the average slice risk and the aver-

age of maximum risk are increased by 28 % and 21

% in Slice 19, respectively. In the latter case, the

maximum observed risk is 23030 (almost 13,000-

fold of its average) observed in Slice 1. Therefore, in

the regions where the geography of the Strait toler-

ates it, overtaking seems to be a safe practice (as al-

so suggested by expert opinion).

3.4 Observation 4

The most significant contributor to risk appears to be

the juxtaposition of the transit vessel traffic and the

local traffic. When the local traffic density in the

Strait is decreased by 50% during daytime, it results

an 83% decrease in the average risk and 31% de-

crease in the average maximum risk of Slice 19. Ac-

cordingly, for potential risk mitigation, the schedul-

ing procedure maybe revised to enable a more

effective night-time traffic at which time there is al-

2000

4000

6000

8000

10000

12000

14000

16000

Base Scenario

Scenario 10

Scenario 11

Scenario 12

300

most no local traffic. However, this issue requires

further research regarding the kind of modifications

that can be done to the scheduling practice to ac-

commodate a larger volume of night-time traffic,

hopefully without increasing overall vessel delays or

other risks.

4 CONCLUSION

The nature of the global economy and international

politics dictates that the maritime transit traffic in

the Strait of Istanbul cannot be greatly reduced nor

eliminated. Nonetheless, the economic/political real-

ities and environmental awareness and risk man-

agement need not to be mutually exclusive goals in

the Strait. The risks regarding the transit traffic can

be mitigated by operational policies and rules that

adequately regulate and guide the transit traffic,

while maintaining the freedom of passage. Until

then, the environment, the priceless histori-

cal/cultural heritage and the health and safety of the

city’s residents will be at jeopardy.

In this paper, a comprehensive analysis of safety

risks of the maritime transit traffic in the Strait of Is-

tanbul is discussed. This analysis is carried out

through the development and deployment of a de-

tailed hybrid mathematical/simulation model. This

model, which is based on extensive objective and

subjective data from a large number of sources, pro-

vides a realistic and valid representation of the mari-

time traffic operations and their impacts at the Strait

of Istanbul with many interesting results.

Our primary conclusions are in the direction of

maintaining the current scheduling/sequencing pro-

cedures to let transit vessels enter the Strait, while

enforcing pilotage service on a larger scale and seek-

ing more efficient and heavier deployment of night-

time conditions, where the local traffic activity is

almost negligible.

REFERENCES

Ozturk, B. 1995. Turkish Straits. New Problems, New Solu-

tions, Chapter: The Strait of Istanbul, A Closing Biological

Corridor. ISIS, 145-154.

Tan, B. & Otay, E. N. 1999. Dynamic Determination of the

Safest Navigation Route for Transit Vessels in the Strait of

İstanbul. Naval Research Logistics, 46, 871-892.

Ulusçu S. O., Özbaş B., Altıok T., Or I. 2008. Risk Analysis of

the Transit Vessel Traffic in the Strait of Istanbul. LPS-08-1

Technical Report. (Online) Available:

http://dimacs.rutgers.edu/port_security_lab/Reports.html

Ulusçu, S. O., Özbaş, B., Altıok, T., Or, I., Yılmaz, T. 2009.

Transit Vessel Scheduling in the Strait of Istanbul, Journal

of Navigation. 62, 1-19.

VTS Users Guide. 2004. Turkish Straits Vessel Traffic Service.

General Management of Coastal Safety and Salvage Ad-

ministrations, 3

edition, May, Istanbul.