Congener-specific intake fractions for PCDDs/DFs and Co-PCBs: modeling and validation

Intake fractions (iFs) for emissions to air, water, and soil for 17 PCDDs/DFs and 12 Co-PCBs were calculated with a level III multimedia model and a food-chain exposure model in succession. The two integrated models were tested by comparing the predicted and measured concentrations in the environment and by comparing intakes through food. Measurement-based iFs were also calculated and compared with the model-based iFs. The air concentrations predicted by the fate model were close to the median of the observed concentrations, whereas the predicted soil and water concentrations were one-third to one-tenth the observed concentrations. This difference was large in case of PCDDs and Co-PCBs, which was explained by the past pollution such as commercial PCB products and PCDD impurities in chloronitrofen (CNP) and pentachlorophenol (PCP). For fish, the predicted and observed exposures agreed well each other. For meat and milk, the predicted exposures were about 10 times the observed exposures for PCDDs/DFs, whereas the predicted and observed values agreed well for Co-PCBs. When the model was modified to consider feeding of fish meal to livestock and geographic bias in feed-grass production, the predicted congener profile was comparable to the measured profile. The comparison also suggested that chickens should be modeled separately from other terrestrial livestock. The model-based iFs for air emission of OCDD and 2378-TCDD were 0.001% and 0.1%, respectively. The iFs of most Co-PCBs were higher than those of PCDDs/DFs. These iF differences suggest the importance of the fate factor in assessing emissions of the 29 congeners.