ORGANIC GEOCHEMISTRY APPLIED TO ENVIRONMENTAL ASSESSMENTS OF PRINCE-WILLIAM-SOUND, ALASKA, AFTER THE EXXON-VALDEZ-OIL-SPILL - A REVIEW

Organic geochemistry played a major role in the environmental assessme nts conducted following the Exxon Valdez oil spill, which occurred on March 24, 1989, and released about 258,000 bbls (41 million liters) of Alaska North Slope crude oil into Prince William Sound. Geochemical a nalyses of more than 15,000 sediment, tar, and biological samples and about 5000 water samples provide the largest database yet collected on oil-spill chemistry, and we review the results here. The marine envir onment of the Sound has a complex background of petrogenic, pyrogenic, and biogenic hydrocarbons from natural and anthropogenic sources. Geo chemical evaluation of the fate and effects of the spilled oil require d that this oil and its residues be distinguished from the background. A variety of molecular and isotopic techniques were employed to ident ify various hydrocarbon sources and to distinguish quantitatively amon g mixed sources in the samples. Although the specific criteria used to distinguish multiple sources in the region affected by the Exxon Vald ez spill are not necessarily applicable to all spill situations, the p rinciples that governed their selection are. Distributions of polycycl ic aromatic hydrocarbons (PAH) and dibenzothiophenes distinguish Exxon Valdez oil and its weathered residues from background hydrocarbons in benthic sediments. Ratios of C-2-dibenzothiophene/C-2-phenanthrene an d C-3-dibenzothiophene/C-3-phenanthrene were particularly useful. Carb on isotopes and terpane distributions distinguished Exxon Valdez resid ues found on shorelines from tars from other sources. Diesel and diese l soot were identified by the absence of alkylated chrysenes and a nar row distribution of n-alkanes, whereas pyrogenic products were disting uished by the dominance of 4- to 6-ring PAH over 2- to 3-ring PAH acid by the dominance of non-alkylated over alkylated homologues of each P AH series. The presence of 18 alpha(H)-oleanane in benthic sediments, coupled with its absence in Exxon Valdez oil and its residues, confirm another petrogenic source. Results of geochemical studies suggest tha t the petrogenic component in the background of benthic sediments is d erived from oil seeps in the eastern Gulf of Alaska. In 1990 and 1991, Exxon Valdez residues, generally forming a small increment to the pre -spill background, were found to be only sporadically distributed in s ome shallow, near shore sediments adjacent to shorelines that had been heavily oiled in 1989. In 1994, occurrences of Exxon Valdez tars on s horeline surfaces were rare, although residues could be found buried i n shoreline sediments at some isolated locations along the spill path where they were protected from wave action. Spilled oil residues colle cted 16 months after the spill were degraded, on average, by nearly 50 %. Shoreline residues from sources other than the spill were also iden tified and are widespread throughout the Sound. These residues include (1) geochemically distinct tars and oils imported from California oil fields to Alaska for fuel and construction purposes prior to the disc overy of the Cook Inlet and North Slope oil fields, (2) diesel and die sel soot, and (3) more highly refined products. Of the more than 2700 chemical analyses of biological samples of higher life forms (fish, bi rds, and mammals) about 150 (6%) indicate recognizable residues of Exx on Valdez oil, which were identified by their distribution of polycycl ic aromatic hydrocarbons (PAH). Most of these samples (138) were colle cted in 1989 and most were associated with external surfaces or the ga strointestinal tract. Rarely do internal tissues or fluids contain rec ognizable fingerprints of spilled oil. This observation includes sampl es from marine mammals that were visibly oiled externally. Other hydro carbon sources, including diesel and a non-petroleum artifact that occ urs when concentrations of individual PAH are at or near their method detection limit, are also identified in biological samples. Copyright (C) 1996 Elsevier Science Ltd