Currently, there is a provision in the draft Alberta Tier 2 soil and groundwater remediation guidelines for development of a Tier 2 site-specific remedial objective(s) (SSRO(s)) for PHCs in soil for the protection of ecological receptors exposed via the direct contact exposure pathway; however, in practice only the pass/fail approach applied to remediated PHC-contaminated soils has been used successfully to secure site closure. The current pass/fail process involves the use of a post-remediation ecotoxicity assessment to demonstrate minimal risk to ecological receptors via the soil contact exposure pathway for in situ contamination. The objective of Phase I (Y1) of this project (CAPP Ref. # 09-9193-50) was to review the results of the Tier 2 pass/fail approach and to investigate possible alternative approaches for deriving SSROs for sites that failed to meet the Tier 2 criteria for a particular land use. To that end four approaches were investigated and compared and, although the project is on-going, two of the approaches can be excluded from further investigation and two show promise and should be investigated further.
The various approaches investigated for the development of SSROs included a :1) geometric mean approach, 2) data exploration, reduction and modeling (ERM) approach; 3) partial least squares and multiple-regression modeling approach (PLSM); and 4) structural equation modeling approach (SEM). The aim was to investigate and assess relationships among the physico-chemical characteristics (e.g., pedological variables), contaminant variables, and the biological response variables measured for different soils contaminated with petroleum hydrocarbons. The data set comprised those acquired over the last five years. The idea was to developed predictive models built using critical explanatory variables (or synthetic variables) which would explain a significant portion of the variability in the response data and which, in turn, would be predictive of the biological responses at similar sites with similar PHC contamination. Once the constraints and uncertainty of the predictive model(s) are identified and quantified, respectively, these models could be applied to PHC-contaminated sites to derive Tier 2 remedial values.
Depending on the size and complexity of a site, members of the oil and gas industry can pay up to 160K for an ecotoxicity assessment of a remediated site with PHC contamination. Should the ecotoxicity assessment indicate that the soils on the site do not satisfy the “pass/fail” criteria, further remedial activities are required for those soils with the hope that they will then pass a second ecotoxicity assessment. Alternatively, the data generated from the Tier 2 ecotoxicity assessment could be used to feed into a predictive model that would then predict a remedial benchmark to guide remediation at the site. A Tier 2 ecotoxicity assessment generates between 11 (minimum requirement) and 36 endpoints depending primarily on the number of site soil samples tested. Phase 1 of this project developed a process to derive SSROs for the PHC fractions of concern using A) endpoint sensitivity distributions (SSDs) of the geometric means for all endpoints; and B) general linear or non-linear models to integrate data and predict impairment concentrations for sites with multiple contaminants of concern (COCs). The first approach using the SSD of the geometric means proved undesirable in that the remedial values were marginally higher and remained as conservative as the current Tier 1 values. The second approach (i.e., the predictive modeling approach) proved both interesting and useful in that a remedial benchmark could be derived for aggregate species responses. Once the models have been applied to site soils that are independent from the model development process, and verified by toxicity testing, they could be applied to any site with any type of PHC contamination. This will mitigate the need for a second ecotoxicity assessment and should overcome some of the current limitations inherent in the current Tier 2 pass/fail approach and reduce remedial costs. If accepted and effective, the process could impact the current regulatory remediation guidelines.