Effect of weathering and biodegradation on markers used in oil fingerprinting

Abstract

The chemical fingerprinting of crude oils and petroleum products is an important technique in environmental spill investigations. Due to the environmental impact of petroleum spills, there is a need to understand the fate of crude oil hydrocarbons in the environment as well as identifying the source for forensic purposes for blame apportionment litigation and, for cleanup and restoration planning. Once petroleum hydrocarbons are spilled in the environment, they are gradually subjected to increased dispersion and degradation processes, collectively termed weathering. This work investigated the effects of the main weathering processes including evaporation, water washing, aerobic and anaerobic biodegradation, on petroleum hydrocarbon molecular markers. The study focused on aliphatic and aromatic compound composition and concentration changes including biomarkers and non-biomarker geochemical markers, in two crude oils (light and medium) from the Niger Delta region of Nigeria and one from the North Sea. Marine sediment from Whitley Bay, UK and river sediment from the Niger Delta, were used as inocula for the aerobic biodegradation microcosms, while anoxic sediment from the River Tyne was used for the anaerobic microcosms. Gravimetric and chromatographic data was obtained from crude oil evaporation experiments in accordance with the Nordtest methodology, where Nigerian light oil was the most affected by evaporation (63% loss by weight), followed by the North Sea oil (49%), and Nigerian Medium oil (36%) after 30 days evaporation. Only relatively low-molecular weight components (including n-alkanes <C20) and adamantanes were observed to be affected by evaporation while the terpanes, steranes, triaromatic steranes, bicyclic sesquiterpanes, PAHs and their alkylated homologues, were unaffected. Similarly, only relatively low-molecular weight components were observed to be affected by water washing experiments and they were less affected when compared with the evaporation experiments. The aerobic biodegradation experiments with Whitley Bay sediment showed the most effects where n-alkanes were completely degraded for the North Sea oil after 100 days incubation. The degradation extent was ranked using the Peters and Moldowan (PM) scale which progressed from level 0 (intact) for the starting oil, level 1 (very slight) for 5, 10, and 20 days incubated samples, level 2-3 (slight to moderate) for 50 and 100 days periods, and level 4 (moderate to heavy) after 300 days. Many of the weathering parameters unaffected by the evaporation and water washing experiments as shown on the Nordtest diagnostic ratio plots, however biomarkers were observed to be affected by biodegradation. Nordtest diagnostic ratio plots of the Whitley Bay sediment experiments show alteration of the parameters such as those based on the dibenzothiophenes and phenanthrenes from 50 days incubation at PM levels 4-5 (heavy) to PM levels 6-10 after 300 days incubation. The C21 pregnanes and some C15 bicyclic sesquiterpanes were affected after 100 days, whereas the terpanes including novel tricyclic terpanes and tetracyclic terpanes identified in the two Nigerian oils together with the steranes and triaromatic steroids were generally unaffected, indicating the potential usefulness of the diagnostic ratio based on these compounds, even in heavily biodegraded oils. Anaerobic biodegradation experiments for 300 days incubation resulted in much less hydrocarbon removal than aerobic experiments, but n-alkanes were depleted resulting in methane production. Novel bicyclic sesquiterpanes, and unidentified suspected pentamethylnaphthalene peaks were identified in the two Nigerian crudes and their resistance to weathering and their potential for use as markers for forensic and correlation purposes were found to be similar to other bicyclic sesquiterpanes. The correlation efficiency of the Nordtest methodology for oils in sediment is affected by high sediment background levels of the geochemical markers considered. Therefore, the determination of their background concentration is necessary in those cases, and certain parameters with high background levels should be ignored to avoid false negatives in the correlation.