(2) Accidental oil spills
Despite the difficulties of NSR navigation, no large-scale oil spills from tankers have ever occurred in the NSR, although minor spills have occurred during bunkering operations at sea and loading oil from temporary pipelines. The volume of such minor spills does not exceed 100-200 liters.
The factors that increase the danger of ship and spilling accidents are many, but the principal factors are hull structure and the state of maintenance, actual operations and human factors. Despite decades of progress in the engineering aspect of navigation, human factors remain the hardest to pin down and the likeliest to cause accidents. The most common cause of ship accidents causing oil spill is collision and grounding on shoals. According to IMO data gathered on a worldwide basis, the average annual frequency for accidents involving tankers of 6,000 register tonnes or more is 31% for collision with other ship or ice and 41% for groundings.
GESAMP (1993) and Engelhardt (1985) report that numerous small-scale oil spills and blow-outs of natural gas have occurred in the Arctic Ocean during loading of oil into tankers and development and production in offshore oil and gas fields. Even these statistics, however, are insufficient to calculate the precise probability of oil spills in the NSR. The estimates for the Baltic Sea indicate that the spill probability on the high seas is about 0.005%, and about 0.25% in high-risk areas (Maisson and Forsman, 1995). Table 4.5-3 shows an estimate of the spill quantity in the NSR on the assumption that an average of 1/48 of the quantity of oil carried on a voyage is spilled in each accident, taking into account the frequency of accidents of collisions and groundings. If these figures were applied to tankers plying the NSR, the spill quantity due to accidents would be 207t in the case of the Ventspils and 503t for the Samatlor.
The above estimate is strictly a reference value indicating the frequency of accidents and spill volumes likely to occur according to probability theory; it does not take account of extraordinary events that can and do take place. For example, if serious inherent cracks on a ship hull remain undetected prior to loading or unloading, they could rapidly extend over a considerable length of the side shell plating. Failure to notice such a crack can lead to the spilling of 500-1,000t of oil in a single hour. The massive 35,000t spill that occurred in the well-known case of the Exxon Valdez, and the even more catastrophic 85,000t spill in the Braer, were caused by such cracks. In other words, there is a yawning gap between the record of actual incidents and theoretical estimates. The GESAMP statistics cited above also show that oil spill quantity fluctuates wildly from year to year, reflecting a function of the number of accidents and the type of ships involved; the same can also be said of environmental impact. Each spill has its own nature, and the environmental impact cannot be addressed by a simple function of the spilled volume.
