We employ a partial-condition designated probabilistic oil spill model based on a Lagrangian particle technique to simulate the Dalian New Port oil spill accident. By correlating the model with aerial images and field observations of oil slicks on the sea surface, the model is applicable for evaluating the probability spatial scattering of the area polluted by the spilled oil. Evidence shows that multiple simulations under specified conditions fit the real distribution and behavior of oil slicks more accurately than those from fully random initial conditions. The findings suggest that the effects of seasonal variations in wind and current dynamics within tidal cycles on spill distribution should be taken into account for estimating the impact of potential oil spills in the North Yellow Sea. A statistical model that considers condition selection could enhance the reliability of risk estimation and enable the application of appropriate recovery operations in contaminated seas.