Those crude simulations were inadequate to determine the feasibility of the NSR. In this simulation, new mathematical models were developed to predict ship speed under various ice conditions. The AARI's historical ice data for the past 40 years were analyzed along each NSR route, for an accurate and precise evaluation of the relation between ice conditions and ship speed. Two approaches were available for description of ice conditions. The first approach was based on probabilistic descriptions of ice thickness, ice concentration and so forth in each region along the routes, applying the Monte Carlo technique or similar methods. The alternative approach was to evaluate ship speeds through long-term virtual voyages, extrapolating ice conditions continuously in time domain for the period of the historical data. In this simulation, the latter method was selected, as the Monte Carlo method cannot give realistic simulations for the negative correlation between the ice conditions of the east and west NSR; namely when the ice conditions in the East Siberian Sea are heavy, those in the western NSR are light, and vice versa (see Section 3.3.3). This simulation made full use of the reliable propulsion performance data for transiting vessels, obtained through model tests in ice model basins, while excluding uncertain predicted data to improve accuracy in the simulation as much as possible. The present simulation appears to be the most comprehensive yet available. Research agencies from Russia, Finland and Japan participated in this simulation. While Russia afforded the data and information of routes, ice and meteorological statistics vital for the simulation, Finland was in charge of conceptual design of the vessels and development of ship speed code in ice. Japan integrated all of the data from the other partners and produced the cost simulation program to yield the final cost calculation. 50,000DWT bulk carrier developed by JANSROP was also evaluated through the simulation. This bulk carrier sacrifices some icebreaking capability in exchange for enhanced open water performance and cargo volume, providing a useful comparison with other types of ship. Using the bulk carrier model, researchers examined the NSR feasibility of less powerful ships with larger cargo volumes.
Selection of routes
As described in Section 4.1, four routes were selected (Table 4.4-3, Figure 4.2-8), consisting of two transit routes and two regional routes. Both transit routes linked Yokohama directly to Hamburg. A northerly route traversed relatively deep waters and was suited for transit ships with drafts up to 12.5m, while a second, southerly route followed the NSR's shallow coastal waters and was chosen for ships with drafts less than 9.0m. Of the regional routes, a regional west route connected Dikson in the western NSR with Hamburg, and a regional east route linked Tiksi with Yokohama. In actual operations, routes will be selected to avoid the most grueling ice conditions, according to the instructions of the MOH as described in Section 4.3, but for simulation purposes these fixed routes were assumed. The routes were plotted for every 20NM on the sea charts published by the Russian Government, taking into account ice conditions and water depths. The regional east and west routes both followed the southern route as soon as they departed their respective ports of Dikson and Tiksi.
Selection of service ships
In the analysis of NSR commercial ships in Section 4.2, we noted that the most modern ice-breaking freighter currently operating in the NSR is the SA-15 class of vessels. Unfortunately its capacity is only 15,000DWT. To reduce operational cost and raise profitability, the largest possible vessel should be introduced. In this simulation, therefore, INSROP conducted conceptual designs of large vessels with drafts of 9m and 12.5m and proposed new hull forms (Table 4.4-4, Figure 4.4-2).
