Probabilistic approach for collision risk analysis of powered vessel with offshore platforms
Introduction
Over time, the ocean has become busy with various types of vessels and offshore platforms. To meet the continuous increase in the demand for hydrocarbons, exploration and drilling have increased, several offshore structures have been constructed, and new ships and shipping routes have been introduced. Because marine transportation remains the cheapest way to transport cargo, there is increasing concern about collisions between vessels and offshore structures.
The risk of collision with offshore platforms is increased in areas of dense ship traffic, such as near coastal areas or naval bases. Effective planning of ship traffic is needed, along with more stringent rules and regulations for marine activities around platforms, from the planning and exploration stages to the operational stage of platform deployment.
The ships that collide with offshore platforms can be generally classified as visiting vessels or passing vessels (Vinnem, 2007, Moan et al., 2002). The vessels that visit platforms, such as supply boats, service vessels, and shuttle tankers that routinely berth at the offshore facilities, contribute significantly to collision accidents. Fig. 1 (a) shows the total loss of the Mumbai High North complex platform in July 2005 after a collision with a multipurpose support vessel, and Fig. 1 (b) shows the capsize of a monopod platform after a boat collision. Passing vessels may also pose a serious threat, especially if the offshore facility is located within a frequently sailed route (see Fig. 1 (c)) (Paik and Thayamballi, 2007). Fig. 1 (d) shows the distribution of the types of vessel involved in collision accidents according to the Worldwide Offshore Accident Databank (WOAD), obtained from OGP (2010). It shows that nearly 23% of the total collision accidents involved a passing vessel. A study of the United Kingdom continental shelf (UKCS) found that passing-vessel collisions occur an average of once every 2 years (Oil & Gas UK, 2010).
The estimation of collision risk requires the quantification of credible collision frequency and associated consequences, which is integral to the safe design and robust building of a platform (Bai and Wei-Liang, 2015). DNV (2002) classified risk assessment techniques into qualitative, semi-qualitative and quantitative techniques. Quantitative risk assessment (QRA) is considered to be the most sophisticated numerical technique that can provide useful guidance for predicting collision accidents, but it is associated with a large degree of uncertainty and requires expert judgement.
Based on the extent of damage to the structures, a collision event can be categorised as minor or major collision. A minor collision is characterised by only repairable damage of the structure and probably will not call for cease of operation. On the other hand, a major collision will damage the platform globally and most certainly require a cease of operation.
However, it seems extremely uneconomical to design a platform to withstand a major collision and remain operational. Therefore, in order to practically while at the same time economically solve the offshore collision problems the probability of major collisions should be kept at a low level by defining adequate preventive measure and minor ones should be considered in the design stage of the platform. This is the design concept of offshore structures against collision adopted by many classification societies.
Most studies on the estimation of collision frequency have taken a scenario-based approach, which uses historical accident databases such as WOAD (DNV.GL). Extensive accident reports and statistical analyses are recorded for the UKCS by the UK Health and Safety Executive (HSE) (see DNV, 2007a, DNV, 2007b, Robson, 2003). Muncer (2003) analysed accident statistics for floating production storage and offloading (FPSO) and floating storage unit (FSU) structures from 1996 to 2002 and compared them with fixed installations in the UKCS area. The study revealed a 5% increase in the number of FPSO/FSU structures. The UK Offshore Operators Association (UKOOA) (2002) concluded that the collision probability for an FPSO subjected to passing traffic is increased due to the increased length of the FPSO, combined with a shuttle tanker, compared to fixed platforms. Furnes and Amdahl (1980) developed a drifting vessel collision-risk model for a shuttle tanker colliding with an offshore platform using Monte Carlo simulation techniques.
Ship traffic databases are also used for estimating passing-vessel collisions. Automatic identification systems (AIS) are considered to be the most advanced and efficient tools for tracking vessel movements, providing up-to-date information on location, heading, course and other details of the ship. Since 2002, IMO regulations have required new ships and all larger seagoing vessels (greater than 300 gross tons) and all passenger vessels to carry AIS on board (IMO, 2001). The AIS messages are transmitted from ship to ship and ship to port using very high frequency (VHF) radio wave signals in a limited geographical space (Eriksen et al., 2006). There are two methods by which AIS tracks ship movements: terrestrial and satellite. Terrestrial AIS is cheaper, but satellite AIS is more useful when a vessel is in open seas and out of range of the network of terrestrial AIS receivers.
Most studies have used AIS marine traffic information to study ship-ship collision probability. Several researchers have used AIS information to analyse safety and the risk of ship collisions in busy sea areas such as the Singapore Strait (Qu et al., 2011), the Gulf of Finland (Goerlandt and Kujala, 2011) and the Malacca Strait (Zaman et al., 2015). Xiao et al. (2015) analysed and compared AIS data for narrow and wide waterways. Zhang et al. (2016) proposed an advanced method to detect possible near-miss ship collisions using AIS data. All of these researchers studied ship-ship collisions; there have been very few studies of ship–platform collision frequency using AIS information (Haugen, 1998). Recently Hassel et al. (2017) used AIS data to study change in passing vessel traffic pattern found before and after platforms were installed and concluded that the current risk assessment practices are overly conservative.
Several commercial software programs are available for estimating collision risk, such as COLLIDE (Safetec), COLLRISK (Anatec UK Ltd.), Computerised Risk Assessment of Shipping Hazards (CRASH) (DNV), SOCRA (MARIN) and COLWT (GL). With the objective of harmonising various assumptions followed in the models, Safeship (2005) compared the models of MARIN, GL and DNV.
The existing software is based on model assumptions; however, improvements taking into account the advanced technologies now in use and the stricter rules and regulations have not been made. For instance, Hassel et al. (2014) highlighted improvements required in the collision-risk model, which was introduced about 20 years ago for the Norwegian continental shelf (NCS), and noted that the ability of the platform to physically get out of the way of a vessel on a collision course was not considered and that there was also inaccurate modelling of the failure factors considered for both the ship and offshore platform.
Geijerstam and Svensson (2008) also reviewed various risk models and concluded that ship watchkeeping failure is the main factor in collision risk. Flohberger (2010) concluded that passing-vessel accidents have not been reduced, despite the introduction of modern technology, because the majority of collisions reported are caused by watchkeeping failures. Upgrading collision-risk software to take account of technological advancements remains a major challenge.
The main objective of this paper is to present a simple probabilistic collision-risk analysis method for offshore platforms subjected to a powered collision by a passing vessel, using a real ship traffic data. A detailed statistical assessment of a traffic database is presented and a simple probabilistic method is used by following a real industrial procedure to estimate the collision frequency, considering various causal factors and the collision mitigation measures in place.
A case study uses ship traffic data in the vicinity of the Busan coast. The database provides information regarding the general trends for the various types of ship traversing the sea, which in turn provides an estimation of all possible threats to the platform. Finally, collision frequency and the corresponding impact energy curves are derived for the various categories of vessel identified in the database.
Section snippets
Collision risk assessment
Offshore platforms located in heavy ship traffic regions demand a comprehensive quantitative collision-risk model. Fig. 2 shows the general procedure for performing a QRA of ship to offshore platform collision following IMO (2002). Although the concept of ‘risk’ is referred to in the paper, the current study mainly focuses on risk analysis, i.e., the quantification of collision frequency and impact energy, using marine traffic data to obtain a frequency vs. impact energy exccedance curve for
Assumptions followed in the study
The following main assumptions are followed in this paper:
- 1.
Offshore platforms considered in the study are assumed to be stationary, i.e. the position of the platform remains unaffected by mean wind, current and steady wave drift force.
- 2.
There has been an accident record on a collision between German submarine U27 and Oseberg B jacket platform in the Norwegian Continental Shelf (NCS) (Vinnem, 2007) with a little damage was caused to the platform. However, the probability of collision by this vessel
Collision impact energy
A collision event can be divided into external dynamics and internal mechanics (Pill and Tabri, 2011, Zhang, 1999). External impact energy influences the time-dependent rigid-body motion of the ship and the associated hydrodynamic effects (Paik and Thayamballi, 2003). Since Minorsky (1958), a number of studies have estimated collision impact energy and associated damage to offshore structures (see DNV, 2010, NORSOK, 2004). Table 2 shows the representative collision energies for powered passing
Application
The ship traffic database forms the basis for probabilistic modelling of a ship's collision with a platform. It not only provides the longitudinal and latitudinal positions of all vessels but also includes maritime mobile service identity number (MMSI), vessel type, vessel dimensions, cruising speed, heading and course.
Fig. 9 shows the geographical region over which the ship movements are measured, with red marks indicating the position of a ship recorded at every instant. The database is
Results and discussion
Fig. 13 shows the probability distribution of course deviation for the six major routes identified in the database. The course deviations are plotted between −90° and +90°, with the assumption that the ships will not sail in the opposite direction. The maximum value of standard deviation is 14.60° for Route A. This falls within the 15° assumed by MARIN in the estimation of course deviation (Ellis et al., 2008). Table 6 summarises the heading range of each route banding and its properties. Among
Conclusions
Offshore platforms installed in high-density ship traffic areas are prone to a high risk of collision by passing ships. This study describes a simple probabilistic approach to risk analysis for powered ship collisions with offshore platforms, using a marine traffic database. The primary focus of the study was to estimate the collision frequency from AIS data for different categories of vessel identified in the database, accounting for various collision mitigating factors such as enhanced ship
Acknowledgements
This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (NRF-2017R1A2B4004891).
References (62)
- et al.
Maritime traffic monitoring using a space-based AIS receiver
Acta Astronaut.
(2006) - et al.
Computer simulation study of offshore collisions and analysis of ship-platform impacts
Appl. Ocean Res.
(1980) - et al.
Traffic simulation based ship collision probability modeling
Reliab. Eng. Syst. Saf.
(2011) - et al.
Evaluation of damage to offshore platform structures due to collision of large barge
Eng. Struct.
(2005) - et al.
Probability modelling of vessel collisions
Reliab. Eng. Syst. Saf.
(2010) - et al.
Study on collision avoidance in busy waterways by using AIS data
Ocean Eng
(2010) - et al.
Ship collision risk assessment for the Singapore Strait
Accid. Anal. Prev.
(2011) - et al.
Design of offshore structures against accidental ship collisions
Mar. Struct.
(2014) - et al.
Energy dissipation in high-energy ship-offshore jacket platform collisions
Mar. Struct.
(2015) - et al.
Numerical analysis of steel tubular member response to ship bow impacts
Int. J. Impact Eng.
(2014)