Source apportionment in oil spill remediation

Abstract

A pipe rupture during unloading led to a spillage of 350–700 tonnes of Caño Limon, a light sweet crude oil, into San Vicente Bay in 2007. Initial clean-up methods removed the majority of the oil from the sandy beaches although some oil remained on the rocky shores. It was necessary for the responsible party to clean the spilled oil even though at this location there were already crude oil hydrocarbons from previous industrial activity. A biosolvent based on vegetable oil derivatives was used to solubilise the remaining oil and a statistical approach to source apportionment was used to determine the efficacy of the cleaning. Sediment and contaminated rock samples were taken prior to cleaning and again at the same locations two days after application of the biosolvent. The oil was extracted using a modified USEPA Method 3550B. The alkanes were quantified together with oil biomarkers on a GC-MS. The contribution that Caño Limon made to the total oil hydrocarbons was calculated from a Partial Least Squares (PLS) analysis using Caño Limon crude oil as the source. By the time the biosolvent was applied, there had already been some attenuation of the oil with all alkanes <C₁₆ removed. After treatment, more of the mid-chain length compounds had been removed leading a skew towards the longer chain length compounds in the samples. The steranes provided a good indication of the source of the oil in this case and the contribution that Caño Limon made to the total oil ranged from 0% to 74%. The total hydrocarbon concentrations were lower after cleaning indicating an efficacy of 90% although the reduction in Caño Limon oil was smaller. This was sufficient to make further remediation unnecessary.