Assessing Lead Risks at Non-Residential Hazardous Waste Sites

ABSTRACT

In 1996, the U.S. Environmental Protection Agency (USEPA) developed the Adult Lead Methodology (ALM) to provide an interim approach to assessing risks from non-residential exposures to lead. Because such exposures often involve occupational activities of adults, the ALM was directed at assessing soil-related lead risks to adults. Consistent with other approaches used in Superfund risk assessment, the ALM was designed to predict quasi-steady state blood lead concentrations (PbB) that might result from soil exposure. These predictions are converted to a risk estimate, expressed as the probability of exceeding a PbB level of concern. To examine the assumptions and variables in the ALM that have become available since 1996, a comparison was made of the attributes of seven alternative research models for which adequate documentation is available to understand and implement each approach. Several of these models have been used in regulatory decision-making; however, the USEPA has officially embraced none for general use. This analysis suggests that the ALM can continue to serve as a reasonable tool for assessing risks associated with non-residential exposures to soil. Under certain circumstances other models may be more applicable (i.e., for assessing acute or intensive exposures); however, the ALM is recommended for the majority of risk assessment applications.

Key words:

• adult lead methodology
• lead
• non-residential
• risk assessment
• slope factor.

ACKNOWLEDGMENTS

The authors acknowledge helpful discussions with members of the USEPA Technical Review Workgroup for Lead; as well as the following people who reviewed and commented on earlier versions of the ALM: Susan Griffin (USEPA Region 8) and Chris Weis (USEPA Region 8). In addition, the authors would like to thank the authors of all the models reviewed herein.

This work is a product of the U.S. government and is not copyrighted.

Notes

^a GSD_{i, adult} ^1.645 and PbB_adult,0 are for women of ages 17–45 years as reported in the combined NHANES III Phases 1 and 2 (USEPA 2002).

¹Throughout the manuscript, significant digits were rounded to tenths for practical purposes of running the models. The data reported from elsewhere, such as NHANES, are reported as stated in tables and figures.

^a Statistics include negative estimates; 3/6 estimates were negative for Si and Zr while 1/6 was negative for Y.

^a Implemented in Excel/@Risk. The model predicts a soil concentration corresponding to a given increase in PbB.

^b Implemented in Excel and ASCL.

^c Implemented in Excel.

^d Fortran version 5/15/97, provided by Dr. Joel Pounds.

^e ASCL version 6/96, provided by Dr. Ellen O'Flaherty.

^a Not a parameter in the model.

^b Simulated as an increment in daily uptake of 6 μ g/day (i.e., 1000 × 0.05 × 0.12) above baseline.

^c Simulated as an increment in daily intake of 50 μg/day (i.e., 1000 × 0.05) above baseline.

^d Default value are shown in parenthesis.

^e A daily uptake of 4.1 μ g/day yielded a quasi-steady state PbB of 2 μ g/dL.

^f A daily intake of 38.7 μ g/day yielded a quasi-steady state PbB of 2 μ g/dL.

^g A daily intake that varied from 12 to 25 μ g/day yielded pre-adult PbBs within the ranges reported from Phase I of NHANES III (Brody et al. 1994) and an adult PbB of 2 μ g/dL.

^h Setting all lead concentrations and intakes to null and food lead ingestion by adults born in 1980 to 25 μ g/day yielded pre-adult PbBs within the ranges reported from Phase I of NHANES III (Brody et al. 1994) and an adult PbB of 2 μ g/dL.

ⁱ The default exposure frequency for the ALM is 219 days/year; however, the assumption of 260 days/year in the simulations would not change the outcome of the model comparisons.

^j Adults born in 1980.

^k Predictions based on the default value for the AF are shown in parenthesis.

^a Time interval over which paired PbBs and lead uptakes were sampled for regression analysis.

^b Linear regression of dependence of PbB (μ g/dL) on lead uptake (μ g/dL) and estimate of the BKSF.

^a Default values from Bowers et al. (1994).

^b ALM default.

^c Number in parentheses is fraction of days per year (219/365).

^d Assumes 16-hour waking day, 8-hour workday, 5 days/week, 50 weeks per year.

^e Midpoint of range based on NHANES III Phases 1 and 2. See Table 2 and text (USEPA 2002).

^f Default values from Carlisle and Wade (1992).

^g USEPA Maximum Contaminant Level.

^h Corresponds to the biokinetic slope factor of 0.4 (see Equation [5] in text).

ⁱ Based on analysis of Pocock et al. (1983) described in text.

^j 5 days/week, 52 days per year; ALM default is 219 days/year. NA = variable not available in the default model.

^a Stern (1994).

^b Revision of Stern model default values based on ALM.

^c Stern (1996).