Assessing The Public Health Significance Of Subsurface-Contaminant Vapors Intruding Into Indoor Air

To assess the public health significance of exposures via the vapor intrusion pathway for exposure a risk assessment was conducted for a common VOC trichloroethylene (TCE) in an unbiased sample of all the contaminated sites within a large geographic setting (northern New Jersey). Probabilistic methods were used to minimize the impact of single point-estimate input values and to help assess the impact of variability and uncertainty in input parameters. Central-tendency probabilistic methods were used to provide an estimate of the most likely exposure point concentrations. The exposure assessment involved 709 TCE-contaminated groundwater sites with 29,856 groundwater samples from 11,210 monitor wells in the state’s Hazsite database. The groundwater mapping component focused on the 78 sites with one or more TCE-contaminated wells located on land classified as residential. The extent of groundwater contamination beyond the monitor well locations was estimated (mapped) using generic GIS-based Inverse Distance Weighted methods on a natural-log scale and additional hypothetical ‘clean wells.’ The risk assessment focused on the 38 sites with one or more hypothetical residences overlying groundwater with a house-plot averaged concentration greater than 2.7 ug/L. The attenuation of vapors generated from the upper-most groundwater was estimated for the 883 hypothetical overlying residential structures using the USEPA’s national empirical database of vapor attenuation factors. Receptor characteristics based on county-level statistics are used to estimate adult individual and childhood age-specific exposures using probabilistic “age at move in” techniques and with possible in-utero and lactation exposures. The exposure estimates are combined with central-tendency probabilistic estimates of toxicity (primarily based on NYDOH, 2006a) to estimate central-tendency risks for the cancer and non-cancer outcomes under study (Non-Hodgkin’s Lymphoma and Central Nervous System effects). In general the risks are low and highly skewed. Only those few individuals at the highest level of exposure are estimated to be subject to risks of typical concern. However, the methods used include limitations and these results are not likely to be representative of some other areas of the country. Sensitivity and two-dimensional analyses indicate the inputs for vapor attenuation and groundwater concentration dominate the risk estimates