PREDICTION OF DROPLET SIZE DISTRIBUTION FROM SUBSURFACE OIL RELEASES WITH AND WITHOUT CHEMICAL DISPERSANTS APPLICATION
MetadataShow full item record
Subsurface oil and gas exploration activities in deeper water have increased in the last few decades. During these activities, subsurface oil blowout incidents may occur and consequently, disastrous impacts on the environment and economy have attracted great attention from scientists and the public. Several deep water oil spill models have been developed to simulate the fate and transport of oil in marine environments and provide guidance for emergency responses. During modeling, the prediction of oil droplet size distribution in the subsurface blowout is important because it has direct influences on the estimated fate and transport of oil in the marine environment. However, the capability to predict droplet size distribution from subsurface release is still limited, because there are insufficient experimental studies on droplet size distribution from subsurface release with subsurface chemical dispersant. In order to have a better understanding of oil droplet size distribution from subsurface oil blowout, a series of subsurface oil release experiments were conducted in an outdoor horizontal wave tank, with different release rates and at different ambient water temperatures. Two crude oils, Intermediate Fuel Oil 120 (IFO-120, heavy crude oil) and Alaskan North Slope (ANS, medium crude oil) crude oil, were used. To study the effect of dispersant application on droplet size distribution, a chemical dispersant, Corexit 9500, was applied at four dispersant-to-oil ratios (DORs). The oil droplet size distributions were measured using LISST-100Xs. Based on the measured droplet size distribution data, the corresponding median droplet diameters (d50) and relative droplet size (d50/D) were calculated for each experiment. The value of d50 revealed that the dispersant had a strong influence on reducing droplet size of both ANS and IFO-120. With the same DOR, ANS was dispersed better than IFO-120, with its d50 being much smaller than that of IFO-120. A relationship, Reynolds Number Scaling, between the relative droplet size (d50/D) and the Reynolds Number (Re) was then established and it showed a good fit to the experimental data. It was found that the empirical coefficient (A) in the Reynolds Number Scaling was dependent on DORs, as well as oil types. The study also found that the spreading coefficient, based on the Rosin-Rammler approach, was different for droplets smaller than d50 (d /d50≤1) and those of d /d50 > 1. A two-step Rosin-Rammler approach (using two spreading parameters, α1 for d /d50≤1and α2 for d /d50 > 1) was then proposed to improve the accuracy of prediction of statistical droplet size distributions. A case study was then conducted by using the improved oil droplet size distribution equations to study the effect of chemical dispersant application on the fate/behaviour of oil from a hypothetical deepwater oil blowout on the Scotian Shelf. The result showed that these equations worked well on predicting the fate of oil from subsurface oil blowout, and application of dispersant greatly reduced surface oil for IFO-120 (heavy crude oil) case.