A Multi-Component Evaporation Model to Determine the Onset of Crude Oil Emulsification in Seawater

Once crude oil is spilled in an aquatic environment, its emulsification state becomes a key property in combating the spill. The onset of emulsification has been hypothesized to be due to decreases in the solvent strength of bulk crude oil. To better predict when emulsification will occur, a compound-specific evaporation model was developed for crude oil spilled on surface seawater. A simple crude oil composition was constructed from 300 of the most commonly found crude oil constituents. Time varying evaporation rates of these constituents were derived by implementing a filmbased exchange model that utilized evolving bulk crude oil properties and chemical composition. A preliminary comparison between the proposed approach and common practice in oil spill modeling suggests that the component specific approach characterizes the evolution of bulk crude oil properties and composition reasonably well. Model predicted times to achieve empirically observed weathering percentages within the first 24 hours varied between 15 minutes and 5 hours of empirical observations for a subset of crude oils and fuels, with the model under-predicting the evaporation rate for light fuels and over-predicted weathering for heavier fuels. Comparisons of model constructed and empirical density data (N = 17, density range 859 – 936 kg m-3) showed that the simple crude oil construct agreed within 3% of empirical values for the initial oil composition with laboratory weathered oil density values agreeing within 6%. Bulk crude oil viscosity values calculated based on a friction-theory model showed poor agreement with empirical data at low and high viscosity values: over-predicting the empirical viscosity by up to 600% for empirical viscosity values below 50 mPa.s, and under-predicting empirical viscosity results greater than 300 mPa.s by 50%. Model predicted viscosity values best agreed with empirical data in the 100 – 300 mPa.s range. Implementation of model predicted bulk data in an emulsion state calculation, however, showed that the model accurately predicted the emulsification state for 14 of 17 sampled crude oils. The time and weathering state for the onset of emulsification, a critical parameter for response operations, was accurately predicted for each of the 4 oils chosen for model comparison. The observed results suggest that the proposed model, even with the observed discrepancies in viscosity, is useful for predicting the onset and ultimate emulsification state of spilled crude oil