Impact of Athabasca oil sands operations on mercury levels in air and deposition

Abstract

Oil sands upgrading facilities in the Athabasca oil sands region (AOSR) in Alberta, Canada, have been reporting mercury (Hg) emissions to public government databases (National Pollutant Release Inventory (NPRI)) since the year 2000, yet the relative contribution of these emissions to ambient Hg deposition remains unknown. The impact of oil sands emissions (OSE) on Hg levels in and around the AOSR, relative to contributions from global (anthropogenic, geogenic and legacy) emissions and regional biomass burning emissions (BBE), was assessed using a global 3D-process-based Hg model, GEM-MACH-Hg, from 2012 to 2015. In addition, the relative importance of year-to-year changes in Hg emissions from the above sources and meteorological conditions to inter-annual variations in Hg deposition was examined. Surface air concentrations of Hg species and annual snowpack Hg loadings simulated by the model were found comparable to measured levels in the AOSR, suggesting consistency between reported Hg emissions from oil sands activities and Hg levels in the region. As a result of global-scale transport and the long lifetime of gaseous elemental Hg (Hg(0)), surface air concentrations of Hg(0) in the AOSR reflected the background Hg(0) levels in Canada. By comparison, average air concentrations of total oxidized Hg (efficiently deposited Hg species) in the AOSR were elevated up to 60% within 50km of the oil sands Hg emission sources. Hg emissions from wildfire events led to episodes of high ambient Hg(0) concentrations and deposition enrichments in northern Alberta, including the AOSR, during the burning season. Hg deposition fluxes in the AOSR were within the range of the deposition fluxes measured for the entire province of Alberta. On a broad spatial scale, contribution from imported Hg from global sources dominated the annual background Hg deposition in the AOSR, with present-day global anthropogenic emissions contributing to 40% (<1% from Canada excluding OSE) and geogenic and legacy emissions contributing to 60% of the background Hg deposition. In contrast, oil sands Hg emissions were responsible for significant enhancements in Hg deposition in the immediate vicinity of oil sands Hg emission sources, which were g1/410 times larger in winter than summer (250%-350% in winter and g1/435% in summer within 10km of OSE, 2012-2013). The spatial extent of the influence of oil sands emissions on Hg deposition was also greater in winter relative to summer (g1/4100km vs. 30km from Hg-emitting facilities). In addition, inter-annual changes in meteorological conditions and oil sands emissions also led to significantly higher inter-annual variations in wintertime Hg deposition compared to summer. In 2015, within 10km of major oil sands sources, relative to 2012, Hg deposition declined by 46% in winter but 22% annually, due to a larger OSE-led reduction in wintertime deposition. Inter-annual variations in meteorological conditions were found to both exacerbate and diminish the impacts of OSE on Hg deposition in the AOSR, which can confound the interpretation of trends in short-term environmental Hg monitoring data. Hg runoff in spring flood, comprising the majority of annual Hg runoff, is mainly derived from seasonal snowpack Hg loadings and mobilization of Hg deposited in surface soils, both of which are sensitive to Hg emissions from oil sands developments in the proximity of sources. Model results suggest that sustained efforts to reduce anthropogenic Hg emissions from both global and oil sands sources are required to reduce Hg deposition in the AOSR.