Forensic Identification and Quantification of Oil Sands-based Diluted Bitumen Released into a Complex River Environment – The Kalamazoo River Oil Spill

Oral Presentation

Prepared by G. Douglas, J. Hardenstine
NewFields Environmental Forensics, 300 Ledgewood Place Suite 305, Rockland, MA, 02730, United States


Contact Information: gdouglas@newfields.com; 781-681-5040


ABSTRACT

Accurate identification and quantitation of spilled petroleum in marine and aquatic sediments is a critical component of the Natural Resource Damage Assessment process. The Kalamazoo Oil Spill represents a unique opportunity to investigate the transport and fate of low API gravity Canadian tar sands bitumen / condensate mixture within a river environment that has a substantial background hydrocarbon signature from other petroleum releases, atmospheric deposition, biogenic sources, and land runoff. Once released into the Kalamazoo River, the light condensate was rapidly lost through evaporation and the remaining oil formed small droplets or globules in the water column that settled out in the river bed sediments. Poling, a process of disturbing the sediment bottom to assess the level of sediment contamination produced light, moderate, and heavy surface sheens at water temperatures greater than 60°F, a temperature below which the oil forms a semi-solid. The ability to identify and quantify this oil in the river, however, was confounded by the high level of background hydrocarbons and the remediation method of mixing the upper several feet of sediment to release the trapped oil for recovery. In this work, we document a technical approach to 1) identify the chemical signature of the spilled oil in the mixed sediment layer, and 2) quantify the amount of spilled oil in the sediment. Three types of samples were collected to perform this analysis, oil globules in the sediment core, and whole sediment and sheen from the surface water. These sediment samples were analyzed by high-resolution gas chromatography with mass spectrometry (GC/MS) and flame ionization detection (GC/FID) for polycyclic aromatic hydrocarbon (PAH) /sulfur heterocyclics /biomarkers and alkanes/total petroleum hydrocarbons (TPH) respectively. The chemical signatures were evaluated using a normalized mixing model relative to the source oil, oil globule, sheen and whole sediments where the sediment specific oil signature was extracted from the whole sediment signature, generating a residual hydrocarbon distribution consistent with sediment background. These results will be used to support the NRDA and to accurately delineate the extent of oil transport and degradation in the river sediments.