Probabilistic Threshold Conditions for Northern Sea Route Cost Competitiveness: A Scenario-Based Comparison with the Suez Canal Route

This study evaluates the cost competitiveness of the Northern Sea Route (NSR) relative to the Suez Canal Route (SCR) for a representative Asia–Europe voyage, using the Shanghai–Rotterdam route as the reference case. A scenario-conditioned stochastic Monte Carlo framework is developed to estimate the probability that the NSR incremental voyage cost is lower than that of the SCR, PNSR. The analysis identifies the probabilistic threshold conditions under which the NSR can become cost-competitive under seasonal and geopolitical uncertainty. The model decomposes insurance-related costs into route-specific risk components and jointly incorporates seasonal conditions, Arctic geopolitical risk, and Red Sea war-risk exposure under five scenarios. The results show that PNSR remains below 50% in all scenarios, indicating that the NSR does not achieve cost competitiveness in the majority of simulations under the assumed cost structure. The summer-current base case yields PNSR = 16.6%, while geopolitical easing increases it to 23.5%. In contrast, the winter-current scenario reduces PNSR to 1.4%, confirming a strong seasonal constraint on NSR viability. Single-variable sensitivity analysis identifies icebreaker fees and Arctic navigation insurance premiums as the most influential NSR-side cost variables, whereas SCR war-risk exposure has the largest overall effect by increasing SCR-side costs. Further threshold analysis shows that no single variable within a plausible operating range is sufficient to raise PNSR to 50%. A combined two-variable analysis shows that simultaneous reductions in icebreaker fees and Arctic navigation premiums can bring PNSR close to 50% under less extreme individual parameter values than those required in the single-variable analysis. These findings suggest that NSR competitiveness is structurally constrained and is more likely to depend on coordinated changes in multiple cost components than on a favourable change in any single variable.