ABSTRACT
In this study, we focus on the development and validation of a deep learning (long short-term memory, LSTM)-based algorithm to predict the accidental spreading of LSFO (low sulfur fuel oil) on the water surface. The data for the training was obtained by numerical simulations of artificial geometries with different configurations of islands and shorelines and wind speeds (2.0-8.0 m/s). For simulating the spread of oils in O(102) km scales, the volume of fluid and discrete phase models were adopted, and the kinematic variables of particle location, particle velocity, and water velocity were collected as input features for LSTM model. The predicted spreading pattern of LSFO matched well with the simulation (less than 10 % in terms of the mean absolute error for the untrained data). Finally, we applied the model to the Wakashio LSFO spill accident, considering actual geometry and weather information, which confirmed the practical feasibility of the present model.
Subject(s)
Fuel Oils , Sulfur/chemistry , Petroleum Pollution , Water Pollutants, Chemical , Algorithms , Models, Theoretical , Computer SimulationABSTRACT
Effective countermeasures against the marine pollution caused by spilled oil are enabled based on the understanding of its physical and weathering characteristics. In that sense, our knowledge of the newly enforced low-sulfur fuel oil (LSFO) needs to be secured urgently. First, we show that the oil viscosity increases with decreasing temperature, following the William-Landel-Ferry law developed for bunker oil. The meso-stable emulsion is achieved from the emulsion test, of which the viscosity is 10-100 times larger than the normal one. On the other hand, the portion of the evaporation of LSFO was insignificant (less than 3%), and thus, its effect on the oil properties is not substantial except the increase of the viscosity. In addition, we experimentally examine the spreading features (e.g., spreading area and rate) of LSFO on the water surface in the circulating water bath. We find that initially, the oil spreading area increases quite fast but saturates, of which the details are explained in terms of the driving and retarding forces involved in the spreading processes. Finally, considering the procured properties of the LSFO, we performed a numerical simulation of spreading LSFO on the water surface with a scale of hundred meters, which shows that our analysis can be extended to larger scales.