You are here:
Statistical comparison of performance reference compound (PRC)-based methods for calculating Cfree
Citation:
Burgess, R. AND A. Joyce. Statistical comparison of performance reference compound (PRC)-based methods for calculating Cfree. Society of Environmental Toxicology and Chemistry North America (SETAC-NA) 39th Annual Meeting, Sacramento, CA, November 04 - 08, 2018.
Impact/Purpose:
The freely dissolved concentration (Cfree) is a useful surrogate for the bioavailable concentration of nonionic organic contaminants in surface waters and sediments. Currently, there are four methods for calculating Cfree based on passive sampling data. This variety of methods can create confusion for inexperienced federal, state and tribal environmental managers as to which one to apply. In this analysis, the four methods were used to calculate Cfree for several PCB congeners and total PCBs in the water column of a contaminated sediment site. In general, two of the methods resulted in statistically similar estimates of Cfree while two other methods appeared to underestimate Cfree. Based on studies from the scientific literature, the first two methods which showed good agreement are believed to generate accurate Cfree values. This analysis will be useful to federal, state and tribal environmental managers needing to know how to analyze and interpret their passive sampling data for calculating Cfree.
Description:
Passive sampling is increasingly used for scientific and regulatory applications. Freely dissolved concentrations (Cfree) have proven useful for estimating exposure concentrations of nonionic organic contaminants in the water column and sediment interstitial waters. It is critical that Cfree be determined under actual or modelled equilibrium conditions. Performance reference compounds (PRCs) are applied to model equilibrium conditions and estimate fractional equilibrium (feq) used to calculate Cfree. Currently, there are four methods for using PRC data to calculate feq. This variety of methods can create confusion for inexperienced users as to which one to apply. The methods include assuming equilibrium (PRCs are not used), a first order model, a diffusion-based model, and an exchange rate-based model. In this analysis, the four methods were used to calculate Cfree for several PCB congeners and total PCBs in the water column of a contaminated sediment site. The resulting Cfree values were then compared statistically to identify trends between the methods. Low-density polyethylene (PE) in four thicknesses was used as a passive sampler and 27 PCB congeners served as the target chemicals. The deployment was performed in triplicate in the water column of three stations at the New Bedford Harbor Superfund site in New Bedford, Massachusetts (USA). Following the 30 day deployments, the four methods were used to estimate Cfree and these values were then compared statistically with ANOVA followed by Fisher’s test. This analysis resulted in twelve comparisons of the multiple PCB congeners and total PCBs. In general, the diffusion-based and exchange rate-based methods resulted in statistically similar estimates of Cfree while assuming equilibrium often under-estimated Cfree and using the first order model generated Cfrees statistically-different from the diffusion and exchange rate methods. For example, at the most contaminated site (NBH2), for the 12, 25, 51 and 76 µm PE thicknesses, the diffusion-based and exchange-rate methods Cfree values were statistically identical 68%, 36%, 68% and 32% of the time, respectively. In contrast, Cfree for the assumed equilibrium and first order methods were very often significantly less than the other two methods. Similarly, the total PCBs comparison found few differences between the diffusion and exchange rate methods. Based on the comparability of the diffusion method Cfrees to conventional measures of Cfree from literature studies, this statistical comparison indicates the diffusion-based and exchange rate-based models are the preferred methods to use for calculating accurate water column Cfrees.