Use of statistical modeling for BTEX prediction in cases of crude oil spill in seawater.

Cases of oil spillage and leakage in marine environments are increasing, and generating a need to quickly assess the presence of these contaminants in seawater. This work aims to estimate the concentrations of benzene, toluene, ethylbenzene and xylenes (BTEX) dissolved in seawater in cases of oil spillage using experimental factorial planning. The study factors were oil °API and oil/seawater contact time after spillage. The models obtained were able to satisfactorily estimate BTEX concentrations, with accuracy greater than 99.3% within the ranges studied, with R² correlation coefficients ranging from 0.992 to 0.997. The models presented forecast efficiency higher than 88%, with low relative errors, ranging from 0.1% to 12%. The concentrations of benzene dissolved in seawater found experimentally with only one hour of spillage, for the two types of oils studied, were higher than allowed by Brazilian legislation, demonstrating real environmental risk in cases of spillage of these types of oil into the sea. These results can corroborate the development of a risk assessment in oil spills within the studied ranges and serve as a useful analytical tool for emergencies.