The Istanbul Strait, the narrow waterway separating Europe from Asia, holds a strategic importance in maritime transportation as it links the Black Sea to the Mediterranean. It is considered one of the world’s most dangerous waterways to navigate. Over 50,000 transit vessels pass through the Strait annually, 20% of which carry dangerous cargo.

In this research, we have developed a mathematical risk analysis model to analyze the risks involved in the transit vessel traffic system in the Istanbul Strait. In the first step of the risk analysis, the transit vessel traffic system is analyzed and a simulation model is developed to mimic and study the system behavior. In addition to vessel traffic and geographical conditions, the current vessel scheduling practices are modeled using a scheduling algorithm. This algorithm is developed through discussions with the Turkish Straits Vessel Traffic Services (VTS) to mimic their decisions on sequencing vessel entrances as well as coordinating vessel traffic in both directions. Furthermore, a scenario analysis is performed to evaluate the impact of several parameters on the system performance.

Risk analysis is performed by incorporating a probabilistic accident risk model into the simulation model. A mathematical model is developed based on probabilistic arguments and historical data and subject matter expert opinions. We have also performed a scenario analysis to evaluate the characteristics of the accident risk. This analysis allows us to investigate how various factors impact risk. These factors include vessel arrivals, scheduling policies, pilotage, overtaking, and local traffic density. Policy indications are made based on results.

Finally, complexity of the operations at the Strait has motivated us to model congestion at the waterway entrances through queueing analysis. We have developed queueing models subject to various operation-independent interruptions. We have used waiting time arguments and service completion time analysis to approximate the expected waiting time of a vessel in the aforementioned queue for various cases of service interruptions. These cases include the single-class models with non-simultaneous and possibly simultaneous interruptions, the multi-class priority queueing model with k possibly simultaneous class-independent interruptions, and the two-class priority queueing model with k possibly simultaneous class-dependent interruptions.