In recent years, The U.S. Environmental Protection Agency (EPA) has employed groundwater contaminant fate and transport models to set chemical-specific concentration limits that dictate how solid wastes are managed. One such model, the EPA Composite Model for Landfills (EPACML) (Woodward-Clyde) was first published in 1990 and enhanced in 1992. EPA used EPACML to set Toxicity Characteristic (TC) concentration limits which determine whether a waste is hazardous under the Resource Conservation and Recovery Act.
A key component of the proposed 1992 Hazardous Waste Identification Rule (since withdrawn, but expected to be re-proposed) was to establish regulatory levels (i.e., constituent-specific health-based levels multiplied by a dilution/attenuation factor (DAF)) to be used as a threshold for excluding wastes from hazardous waste regulation. EPACML was used to determine appropriate DAFs under an infinite waste source scenario.
EPA recognizes that the infinite source scenario may not be appropriate for groups of chemicals which are either disposed of in small quantities or have high leaching potentials. As a result, EPA modified the source term in the original EPACML model with a finite, constant concentration source.
API performed an initial review (API, 1992) of this enhanced EPACML model that indicated flaws in the finite source methodology. This initial review also suggested that the model had potential problems with conserving mass.
To confirm the results of the initial review, API undertook a more rigorous analysis of EPACML. The objective of the analysis, documented in this report, is to provide technical information on the model's problems, and where possible, to propose some alternative solutions to EPA and the regulated community affected by the use of the model. Specifically, this report documents a technical evaluation of the finite source methodology and the analytical solutions that comprise the enhanced EPACML model. The report contains an overview of the following model components:
• Unsaturated zone transport module
• Finite source methodology
• Saturated zone transport module
• Coupling of the unsaturated zone and saturated zone transport modules
• Estimation of retardation factors and decay coefficients (Appendix A)
Each overview is immediately followed by a critique of the component. Flaws in the model are demonstrated with model runs and simple calculations where possible.
GENERAL DESCRIPTION OF THE EPACML MODEL
The model consists of two major parts: (1) unsaturated transport simulator, and (2) saturated transport simulator. Operated in the deterministic mode, the model calculates a constituent concentration at a designated drinking water (receptor) location for a given set of input parameters. To obtain a statistical distribution of possible receptor well concentrations for ranges of waste, soil, aquifer properties, the model utilizes a Monte Carlo method, which randomly generates a large number of possible combinations of these parameters. For each set of these properties, the model calculates the ratio of the source concentration to the receptor concentration (dilution/attenuation factor - DAF). Finally, a cumulative probability distribution for DAF is estimated based on the results from the Monte Carlo simulations. This cumulative distribution is subsequently used to estimate an allowable leachate concentration. In order to obtain a realistic cumulative probability distribution for DAF, it is important to ascertain whether the deterministic part of the model, i.e. the part that calculates the receptor concentration for each Monte Carlo-generated set of input data, is correct.
This report focuses on the deterministic part of EPACML. Each major component of the EPACML model is described in the report, followed immediately by a critical analysis of the component.