Abstract
A numerical model named 'B-Oil' is developed, based on the solution of the governing partial differential equations of flow and immiscible pollutants to backtrack oil slick incidents in both Kuwaiti offshore regions and the Arabian Gulf waters. The model was coupled with the KGulf Model, a hydrodynamic numerical model, which supplies the necessary 2-D flow fields of the surface waters. The Lagrangian discrete parcel algorithm was adopted to backtrack oil spills, and the Monte Carlo statistical technique was applied to overcome the random walks in the Lagrangian discrete parcel algorithm. The adopted technique was investigated with three previous oil incidents, documented in literature, which occurred in the Arabian Gulf. The model performed adequately in predicting the source of the oil spill, with general higher accuracy, when the distance between the transported and original oil spill locations were shorter, compared with larger distances. This was mainly attributed to the lack of representation of surface flow fields, particularly rising from the application of the average wind conditions. However, conducting larger test numbers, that in principal increases the computational times, have shown to improve the model predictions. Having said that, limitations and sensitivity of the model was addressed by means of numerical scenarios and testing the significance of effective numerical parameters. Such technologies could be used on a real-time basis to predict the source of an oil spill (or any floating matter following lagrangian mechanics) that helps decision makers in responding on timely basis to oil spill accidents in the offshore regions.