Derivation of site-specific remediation goals by incorporating the bioaccessibility of polycyclic aromatic hydrocarbons with the probabilistic analysis method.

Incorporating bioaccessibility into human health risk assessment is recognized as a valid way to reduce the conservative properties of conventional results, where the total concentration of a contaminant analysed by exhaustive chemical extraction is applied. Taking a coke production site in Beijing as an example, a mild chemical extraction technology was employed to profile the bioaccessibility of benzo[a]pyrene (BaP), indeno[1,2,3-cd]pyrene (IcP) and dibenz[ah]anthracene (DBA) in soils. The results that were regressed using two bi-phase desorption models (Karickhoff and Weibull) revealed that the rapid desorption fractions of BaP, IcP and DBA, which are taken for bioaccessible fractions, were basically less than half of the total contents in the soils. Probabilistic analysis (PA) was carried out with pre-set distributions of the exposure parameters to characterize the uncertainty in the assessment. The results incorporating bioaccessibility and PA were several times higher than the generic remediation goals which equal to national screening levels, and orders of magnitude higher than the baselines of the region and nation. The results of the Weibull fit were finally recommended as site-specific remediation goals (SSRGs) (10.59 mg/kg, 95.48 mg/kg and 9.24 mg/kg). Over-remediation was avoided while contributing to considerable economic and environmental benefits.