Degradation and aquatic toxicity of oil sands naphthenic acids using simulated wetlands
Oil sands process-affected waters (OSPW) from the Athabasca oil sands (AOS) located in northern Alberta, Canada, are toxic to aquatic organisms due to the presence of organic and inorganic constituents. Much of this toxicity is related to a group of dissolved organic acids known as naphthenic acids (NAs). Naphthenic acids are a natural component of bitumen and are released into process water during the caustic hot water extraction process used to separate the bitumen from the oil sand ore. This complex mixture of non-cyclic and mono- and poly-cyclic alkanes containing carboxyl groups are characterized by the general formula CnH2n+zO2, where n indicates the carbon number, and Z represents the number of fused rings in the structure. Currently, all process-affected waters are stored within large holding ponds and settling basins on the oil sands mining lease sites with the understanding that eventual reclamation of this water must be undertaken. Successful reclamation of OSPW is expected to require a reduction in total NAs concentrations in the OSPW and the removal of the toxic character of the water. Natural or enhanced bioremediation in lakes and wetlands within the lease closure landscapes will play a critical role in meeting these two requirements. This research investigated the potential for the reduction of total NAs concentrations in OSPW due to biotic (e.g., biodegradation) and abiotic (e.g., sorption) processes, and its relationship to the overall toxicity of OSPW. The specific goals of this research were to determine if natural degradation of NAs in simulated wetland environments could be enhanced by manipulating various physical and chemical factors of the environment, to describe and quantify the selective biodegradation rates of NAs congeners, and to correlate observed changes in total NAs concentration and composition with changes in the aquatic toxicity of OSPW. The complexity of both OSPW and NAs mixtures presented an unusual set of challenges. A preliminary investigation was used to determine the contributions of salinity and NAs to the total aquatic toxicity of OSPW in order to identify a suitable test organism that would respond to NAs concentrations while tolerating the high ionic content of OSPW for the main simulated wetland microcosm study. Seven-day Ceriodaphnia dubia chronic toxicity tests, using both un-manipulated (containing NAs) and manipulated (substantially reduced NAs) samples of OSPW, identified salinity as a potential contributing factor to the overall toxicity of this complex water. Only a 5% reduction in acute toxicity and an 11% reduction in chronic toxicity was observed with a 91% reduction in total NAs concentration (from 67.2 to 5.9 mg/L; removed by solvent extraction). However, when the same samples were tested using the salt tolerant bacteria Vibrio fischeri in the Microtox® bioassay system, the 91% reduction in total NAs concentration, the toxicity was removed (EC50 changed from 57.8 to >100%). These results suggested that salts in OSPW may drive the toxicity of OSPW to some freshwater invertebrates, such as C. dubia, and that the Microtox® bioassay was better suited to track the overall toxic potential of NAs in OSPW. Using flow-through, laboratory microcosms to mimic natural wetlands, it was demonstrated that the reduction in total NAs concentration, based on the Fourier Transform Infrared (FTIR) spectroscopy analysis, was dependent upon hydraulic retention time (HRT), but appeared to be unaffected by nutrient addition (nitrogen and phosphorus). Microcosms with a longer HRT (for two OSPW types; Syncrude and Suncor) showed higher reductions in total NAs concentrations (64¬ to 74% NAs reduction) after the 52-week test period, while nutrient enrichment appeared to have little effect. While the total NAs concentrations decreased in the waters from the microcosms, a 96-hr static acute rainbow trout (Oncorhynchus mykiss) bioassay showed that the initial acute toxicity of Syncrude OSPW (LC50 = 67% v/v) was reduced (LC50 >100% v/v) independent of HRT. However, EC20s from the Microtox® bioassays were relatively unchanged when comparing the input and output microcosm waters maintained at both HRTs over the 52-week study period, indicating that some sub-lethal toxicity persisted under these experimental conditions. The study demonstrated that given sufficiently long HRTs, simulated wetland microcosms containing OSPW significantly reduced total NAs concentrations and acute toxicity, but left behind a persistent component of the NAs mixture associated with residual toxicity. Further investigations aimed to describe and quantify the selective biodegradation of NAs congeners and correlate the observed changes in total NAs concentration and composition (i.e., NAs fingerprint profile) with the aquatic toxicity of OSPW. High performance liquid chromatography/quadrupole time of flight-mass spectrometry (HPLC/QTOF-MS) analysis was used to track the changes in NAs mixture profiles or ‘fingerprints’ in each experimental treatment over time. Based on first-order degradation kinetics, rapid degradation was observed for NAs that had lower carbon numbers (11 to 16) and fewer degrees of cyclization (Z series -2 to -4; half-lives between 19 to 28 weeks). Within the NAs mixture fingerprint, the two most persistent groups of NAs homologues were identified (NAs with carbon numbers 17 to 20 and Z series -6 to -12; half-lives between 37 to 52 weeks). Their persistence may have resulted in the residual chronic toxicological response as measured by the Microtox® bioassay (EC20). An additional study was conducted to characterize potential changes in the total concentration and composition of NAs in OSPW due to sorption to organic wetland sediments. The batch-reactor investigation showed a rapid (<1 day) and significant reduction in total NAs concentrations in OSPW when mixed with the wetland sediment at a ratio of 2:1 v/v (OSPW:sediment). The mean percent reduction of NAs in OSPW was 67% during the 14-day test period, suggesting a significant influence of sorption on the removal of NAs than previously expected. However, no preferential sorption was observed based on the distribution of NAs congeners with respect to carbon number, Z series, and arbitrarily defined clusters. The potential sorption of OSPW NAs as a result of using substrates with high organic carbon content (e.g., 27.6% total organic carbon content) in designed wetlands may enhance the mitigative capabilities of these reclamation landscapes at the AOS. Further investigations into understanding NAs sorption kinetics without substrate agitation are warranted before these results can be extrapolated to the field. Finally, to test the hypothesis that persistent components of an OSPW NAs mixture (e.g., NAs congeners with higher carbon numbers and degrees of cyclization) may be responsible for the observed residual chronic toxicity identified in the previous simulated wetland microcosm study, the fractionation of OSPW NAs was attempted using both off-line anion exchange chromatography and batch-wise co-polymer filtration and elution. Although complete separation was not achieved in this investigation, the results suggested that specific variations of the co-polymer were most effective and showed the most promise for separating the NAs mixtures based on polarity and size. With further refinements to the procedure, future investigations may be able to achieve adequate separation of the NAs mixture into fractions with compositions different enough to conduct toxicity bioassays.
DegreeDoctor of Philosophy (Ph.D.)
DepartmentGraduate Studies and Research
CommitteeBlakley, Barry R.; Siciliano, Steven D.; Headley, John V.; Wilson, Lee D.
Copyright DateJanuary 2012
Oil sands process water