Moreover, most of the existing models are based on

historical data from past oil spills obtained from the IOPCF statistics, which by definition is passive, for the detailed discussion the reader is referred to Psarros et al. (2011). Furthermore, such models are developed with the use of data about spill sizes falling in a certain range, usually with small median value for a spill, see Kontovas et al. (2010), thus applying such models for extrapolation beyond this range is very questionable. In the scientific literature Crizotinib supplier there are only two models allowing for the estimation of oil spill clean-up costs. One has been proposed by Etkin – Etkin, 1999 and Etkin, 2000 – is deterministic but allows rather wide interpretation of the cost factors considered. Another model has been proposed by Shahriari and Frost (2008) it is

also deterministic, but with no room for interpretation. Predictions of both models hold in the context of global oil spill costs, but they have rather low geographical resolution. Therefore, it is not possible to use the models for the purpose of oil-combating fleet optimization or detailed risk management, as the local conditions are not properly reflected. Moreover, the unique nature of the analyzed sea area of the Gulf of Finland, being classified by the IMO as a Particular Sensitive Sea Area (PSSA), makes it possible for the oil to reach the shore

in a very short time with devastating consequences, Selleck GSK-3 inhibitor see for example Lecklin et al. (2011). This means that once the oil spill at sea has occurred, it is almost impossible to prevent it from reaching the coast, see Hietala and Lampela, 2007 and Aps et al., 2009. What makes the clean-up operations even more demanding is the fact that the coastline is filled with small islands; making it impossible for the clean-up vessels to navigate in some places even though the sea depth would allow it. Another selleck factor that separates the Gulf of Finland from the larger sea areas is that, according to the HELCOM agreement, use of chemical dispersants or in situ burning are not permitted as oil combating techniques, and the clean-up is mainly performed mechanically, see HELCOM (2012). All these show the complexity of the subject and limitations of existing clean-up cost estimation models. Hence, it is desirable to go to the sources of each of the costs, which together make the total cost of oil spill clean-up operation. This paper introduces a probabilistic model for accidental oil spill cleanup-cost estimation for the Finnish response area of the Gulf of Finland – see Fig. 1.