Collision and Grounding of Passenger Ships – Risk Assessment and Emergency Evacuations

This paper considers collision and grounding of passenger ships. Subsequent to a collision or grounding, there might be water ingress into the compartments and the risk of sinking exists. In such circumstances, timely and effective evacuation of all passengers and crew will be extremely important, and failure to evacuate in time may lead to catastrophic consequences. This paper presents probabilities of collision and grounding and investigates possible events subsequent to an incident, e.g. possibilities of flooding, sinking and capsizing, expected time to sink, etc. Evacuations in case of collision and grounding are also considered and the consequences are estimated in terms of expected loss of lives. The risk associated with collision and grounding is hence estimated and compared to other types of accidents such as onboard fire. 1. BACKGROUND AND INTRODUCTION Any ship that experiences flooding of one or more of its compartments is exposed to the risk of losing its stability and thus the risk of sinking. Collision and grounding are considered to be the most relevant accident scenarios that may cause flooding of passenger ships, and will thus be the topic of this paper. Existing passenger ships will generally carry a large number of people when they are in normal operation and modern cruise ships can carry passengers in the order of several thousands. Although accidents involving such large passenger ships are very rare, if a serious accident should occur, its consequences could be disastrous. Even though a lot of effort is constantly being made to keep passenger ships safe and measures are always taken to avoid serious accidents, one can never completely eliminate the probability of a serious accident to occur on board a ship. If an incident takes place, one can try to prevent it from evolving into a serious accident by for example intentionally beaching a ship that is taking in water and thus keep it from sinking. If such measures fail however, an evacuation provides a last opportunity to minimize the consequences of the accident by reducing the number of fatalities. In such situations, the evacuation performance will be very important and an orderly and timely evacuation can save the lives of many people on board. Due to the conviction that an effective evacuation will be important in case of a maritime accident and realising that modern passenger ships will normally carry a large number of people, new requirements from IMO states that an evacuation analysis shall be performed early in the design process for new ro-ro passenger ships [SOLAS (2001)]. The Maritime Safety Committee has produced guidelines for such evacuation analyses [MSC/Circ. 1033 (2002)]. 2. THE FIRE EXIT PROJECT One of the goals of FIRE EXIT is to equip the marine industry with a Ship Evacuation Simulator that represents a quantum leap of present software regarding the level of reliability, realism and design utility [Galea et al. (2002), Galea et al. (2003)]. The simulator being developed within the FIRE EXIT project, the maritimeExodus, should be able to account for issues such as mustering, dynamic ship movement, static trim and heel, fire and smoke and abandonment. Data on passengers’ behaviour in different conditions of movement and with or without the presence of smoke are collected and used in the simulations, and the evacuation simulator software is integrated with an interpretation and optimisation module. The propagation of fire and smoke are to be modelled through a link to the SMARTFIRE software [Taylor et al. (1996), Galea et al. (2003)]. An XML based interface is provided as a link to concept design software, enabling conceptual designs to be tested at an early stage. The outputs from the software simulations are to be validated against full-scale trials on board a real passenger ship in operation. When performing evacuation simulations, it is important to have a basic understanding of what a typical evacuation scenario might be. In addition, the different probabilities associated with the scenarios will be useful. Because of this, an evacuation risk assessment has been carried out within the FIRE EXIT project [Vanem (2003)] that has investigated different accident scenarios. This study suggests that collision and grounding is among the most critical accidents involving passenger ships. The expected frequency of such accidents are comparable with that of fire [Vanem and Skjong (2004)], but the expected time available for evacuation is much less, and the expected consequences in terms of loss of lives are thus higher. Emergency evacuations are hence considered extremely crucial for collision and grounding of passenger ships. 3. EVACUATION ANALYSIS REQUIREMENTS FOR PASSENGER SHIPS The evacuation analysis requirements do not state that advanced simulation software must be used for the analyses. There is an option to perform a simplified analysis using only pen and paper, considering evacuating crew and passenger as a flow of people through the escape routes in much the same way as water flowing through a pipe system. Performing a simplified evacuation analysis with no advanced tools is a very time consuming task however, and running advanced simulations can save several man-days of work for every evacuation analysis performed. It is thus expected that most evacuation analyses carried out in the future will utilize sophisticated tools such as the maritimeExodus to perform an advanced analysis. It should be noted, however, that regardless of whether the simplified or the advanced approach is taken, the objective is to assess the 1 International Maritime Organization, http://www.imo.org 2 FIRE EXIT – project website, http://www.bmtproject.net/fireexit evacuation process through a set of well defined benchmark scenarios rather that to model the evacuation in a real emergency situation. P collision ≈ 5.16 x 10 (1) Furthermore, the probability of flooding given collision was estimated to be 0.38 although the statistical basis for this conclusion was rather weak. Only 16 collision accidents were involved in the study and the uncertainty associated with the estimate is hence considered to be high. An alternative way of estimating the frequency of flooding was therefore sought. According to IMO guidelines, an evacuation analysis performed over a set of benchmark scenarios are required to demonstrate that possible emergency evacuations can be completed within 60 minutes for all new ro-ro passenger ships. The analysis should thus ensure that evacuation arrangements are appropriately adequate for swift evacuation of all passengers. This time limit corresponds to the requirements regarding confinement of fires within main fire zones of passenger ships. Passenger ships are required to be divided by thermal and structural boundaries into main vertical fire zones and horizontal zones, and these boundaries should be able to confine a fire for at least one hour within the fire zone where it originated. Historically, evacuations have primarily been considered in relation to fires and it has been assumed that evacuations would be successful if it is completed by the time the fire spreads from the zone of origin. The HARDER project has collected a set of data on previous maritime accidents, primarily collision and grounding accidents, and studying these data will provide an alternative estimate of the probability of flooding conditioned on collision [Mains (2001)]. The database contains records of 2946 causalities including 1851 collisions involving different types of ships. The 1851 records of collision included ship-ship collisions as well as collisions between ships and other objects, and it included records regarding both the struck and the striking ship. Further examination of the data revealed 801 records of struck ships in ship to ship collisions. It was not obvious from all the records which ship experienced flooding subsequent to the collision, but after studying the data in more detailed it was assumed that 508 of the 801 records involved flooding of one or more compartments, i.e. a probability of flooding of 0.63. The uncertainty of this estimate is also considered to be rather high due to the somewhat incomplete information contained in the database. This paper indicates, however, that fire is not the only relevant scenario for evacuation of passenger ships and that collision and grounding are highly relevant accident scenarios that will impose even stricter requirements on the evacuation performance on passenger ships. 4. COLLISION OF PASSENGER SHIPS Collision between two ships at sea is always a serious incident and depending on the extent of the impact, the ships involved in a collision may or may not sink. The striking ship will normally not be in any great danger of sinking, as it will receive the impact of the collision at the bow, and the bow in front of the collision bulkhead will normally receive all the collision energy. Damages restricted to this part of the ship will normally not affect the stability of the ship. The struck ship, however, if receiving a blow to its side, has a high risk of loosing its stability and will thus be in danger of sinking. If the struck ship is a passenger ship with many people on board, effective and orderly evacuation of these people will be crucial to the outcome of the incident. The two estimates of the probability of flooding conditioned on collision must both be considered as uncertain. However, it might be appropriate to assume the actual value lies somewhere between these values. As an approximate value, it is thus assumed that the probability of flooding given collision is the mean value of these two estimates, i.e. P Flooding | Collision ≈ 0.5 (2)