PARENT SESSION
PM20 - Credibility of Regulatory Environmental Models
Monday, 18 November 2002
8:00 AM to 6:30 PM
Exhibit Hall
(P397) Comparison of aquatic toxicity experimental data with EPA/OPPT/SAR prediction on PPG polymers.
Chun, Jean*,1, Nabholz, Joseph2, Wilson, Maggie2, 1 PPG Industries, Inc., Allison Park, PA, USA2 US EPA, Washington, DC, USA
ABSTRACT- Experimental aquatic toxicity data on several different types of polymers have been compared with the data generated by structure activity relationship (SAR) predictions using EPA's Pollution Prevention (P2) Framework. The P2 Framework is a series of computer-based models that estimate hazard information based on a chemical's structure. These models provide screening-level information by estimating physico-chemical properties, fate in the environment, toxicity to aquatic organisms, and carcinogenicity. The validation study was conducted to evaluate the U.S. EPA SAR methods used to predict aquatic toxicity of polymers. The experimental data were obtained from an algae growth inhibition test, a Daphnia acute immobilisation test, or an acute toxicity test to Rainbow trout. The studies were conducted in four different contract laboratories in the USA or in the UK following OECD guidelines and good laboratory practice standards. The data from studies on a total of 38 polymers (55 actual test data points) including 7 polyesters, 5 polyurethanes, 8 acrylates, 4 acids and anhydrides, 13 epoxies, and one melamine were compared against SAR predictions. Data were considered to be in agreement if the SAR predictions were within the same order of magnitude (less than a ten-fold difference) as the experimental data. Data were also considered to be in agreement if the SAR prediction indicated that the material caused no effects at saturation (NES) or a low concern for toxicity and the experimental data showed no effects at the maximum attainable concentration or an EC50/LC50 value of >100 mg/L. Forty-eight out of the 55 SAR predictions were in agreement when compared to the experimental data. Five SAR predictions were not in agreement with the experimental data. Two SAR predictions were considered as inconclusive. Overall, the reliability of SAR predictions as compared to the experimental data was 87 to 90 percent depending on whether or not two inconclusive results were included.
Key words: Pollution Prevention Framework, Structure Activity Relationship, Aquatic Toxicity, PPG Polymers