Modeling of Crude Oil Evaporation: A Bottom-Up Approach to Prediction of Potential Secondary Organic Aerosol Formation Following Oil Spills

Abstract:

Releases of hydrocarbons from oil spills can have large environmental impacts in both the ocean and atmosphere. While evaporation of oil following a spill is mainly modeled simply as a mass loss mechanism, the resulting production of atmospheric pollutants can also be a major concern, particularly for continental releases, such as wrecks of train-tanker or river barges, and near-shore rig releases. Both may occur near population centers. The Deepwater Horizon (DWH) spill in the Gulf of Mexico in 2010 presented a unique opportunity to observe significant secondary organic aerosol (SOA) production due to a large oil spill. Following on these observations, we have conducted a series of measurements on evaporation of oil while explicitly accounting for changes in chemical composition occurring as a function of evaporation time. In this work we use GC×GC-VUV-HRTOFMS to achieve unprecedented characterization of oil composition from the Deepwater Horizon (DWH) oil spill, and how it changes with time following release. Roughly 75% of the total mass of the alkane mixture comprising the oil was classified according to degree of branching, number of cyclic rings, aromatic character, and molecular weight. Such detailed and comprehensive characterization of the DWH oil allows for bottom-up estimates of the relationship between oil volatility and composition. We developed an evaporative model, based solely on our composition measurements and thermodynamic data (vapor pressure, enthalpy of vaporization), rather than common boiling point parameterizations, which is in excellent agreement with published mass evaporation rates and allows for prediction of potential SOA production as a function of both wind speed (evaporation rate) and oil composition. Our measurements yield different oil volatility distributions than previously inferred; this suggests accurate prediction of SOA formation requires detailed oil composition measurements. A wind tunnel was used to verify model predictions. Sustained SOA yields on the order of a gram m^-2 day^-1 are predicted under calm sea conditions.