A multiple lines of evidence approach for the ecological risk assessment of an accidental bitumen release from a steam assisted gravity drainage (SAGD) well in the Athabasca oil sands region

Abstract

To assess the ecological impacts of two independent accidental bitumen releases from two steam assisted gravity drainage (SAGD) wells in the Athabasca oil sands region, a multiple lines of evidence (LOE) approach was developed. Following the release in 2010, action was taken to minimize environmental impact, including the selective removal of the most highly impacted vegetation and the use of oil socks to minimize possible runoff. An ecological risk assessment (ERA) was then conducted based on reported concentrations of bitumen related contaminants in soil, vegetation, and water. Results of biological assessments conducted at the site were also included in the risk characterization. Overall, the conclusion of the ERA was that the likelihood of long-term adverse health effects to ecological receptors in the area was negligible. To provide evidence for this conclusion, a small mammal sampling plan targeting Southern red-back voles (Myodes gapperi) was carried out at two sites and two relevant reference areas. Voles were readily collected at all locations and no statistically significant differences in morphometric measurements (i.e., body mass, length, foot length, and adjusted liver weight) were found between animals collected from impact zones of varying levels of coverage. Additionally, no trends corresponding with bitumen coverage were observed with respect to metal body burden in voles for metals that were previously identified in the source bitumen. Hepatic ethoxyresorufin-O-deethylase (EROD) activity was statistically significantly elevated in voles collected from the high impact zones of sites compared to those collected from the reference areas, a finding that is indicative of continued exposure to contaminants. However, this increase in EROD was not correlated with any observable adverse population-wide biological outcomes. Therefore the biological sampling program supported the conclusion of the initial ERA and supported the hypothesis of no significant long-term population-wide ecological impact of the accidental bitumen releases.

Response of a Small Mammal Population Postremediation for an Accidental Bitumen and Steam Release from a Wellhead in the Athabasca Oil Sands

“…The higher Cr in voles from the most heavily affected sites in 2014 was a finding mirroring that of Berger et al (); this mineral reverted to background levels similar to that of the reference site by 2017. The positive association seen with higher Cr and higher body condition may have been affected by other unknown, unmeasured factors.…”

“…All presented methods and results are for the present study exclusively, with reference to Berger et al 2016 only as a comparison or to provide background information or comparison.…”

“…The sampling sites in 2017 were used to replicate areas previously sampled in 2014 (detailed in Berger et al 2016). Sherman live traps baited with a mixture of peanut butter and rolled oats placed on a cracker were used to capture the specimens in impacted and reference sites (Supplemental Data, Figure S2).…”

“…Activity of the hepatic enzyme EROD is commonly used to determine the detoxification efforts by organisms, which is assumed to reflect the level of contaminant challenge being dealt with by the animal. The EROD assays for the voles in the present study (performed by the Smits lab) were carried out using the methods detailed in Berger et al (2016; which were also carried out by the Smits lab). All tissue handling and analyses were conducted on ice, with chilled equipment and reagents.…”

Ecological risk assessment (ERA) based on contaminated groundwater to predict potential impacts to a wetland ecosystem

“…The 'triad' integrates chemical, ecotoxicological and ecological lines of evidence (LoE) to estimate the risk. Multiple lines of evidence (up to five) have been proposed to assess the ecological risks from anthropogenic activities (e.g., Berger et al 2016). Nevertheless, the integration of only two LoEs (e.g., chemical and ecotoxicological) has been considered to be sufficient to predict risk as a screening phase, particularly in situations where it is not possible to assess in situ ecological parameters (ISO 19204 2017) such as for groundwater, where knowledge about ecosystem biodiversity is still precarious.…”