Analytical equations can be used for analysis of contaminant transport under homogeneous, isotropic conditions, based on site-specific groundwater flow parameters and contaminant characteristics. These analytical models can be programmed in spreadsheet format, facilitating rapid estimation of potential impacts at off-site receptor points as well as back-calculation of applicable soil and groundwater clean-up standards. In this paper, guidelines for proper application of such spreadsheet models are demonstrated by comparing results of the analytical contaminant transport expression developed by Domenico and Robbins (1985) to those generated by the BIOPLUME II numerical modeling program (Rifai et al, 1987). However, the procedures outlined below can be generally applied to extend the utility of any properly validated analytical model:
• Model Selection: The model should be matched to the complexity of natural site conditions and to the degree of resolution provided by the existing site database. Steady-state analytical expressions provide significant advantages in terms of ease of use, speed of operation, and interactive sensitivity analysis. However, given adequate site data, numerical models may be required to properly account for flow field heterogeneities and complex biodegradation processes.
• Source Term Characterization: The source term must be calculated to match current plume concentrations and to account for the presence of non-aqueous phase liquids (NAPLs). Guidelines are provided for source term characterization in the Domenico model based on a 'layer-cake' approximation of the transverse plume cross-section.
• Aerobic Biodegradation: Many analytical contaminant transport models incorporate a first-order decay equation to account for aerobic microbial biodegradation of organic plume constituents. However, by neglecting oxygen limitations, such first-order expressions commonly overestimate the rate and degree of natural biodegradation in groundwater systems (Borden et al, 1986a; Lee et al, 1987). Guidelines are provided for use of an oxygen superposition method with the Domenico analytical model that generates results consistent with those of BIOPLUME II.
• Management of Parameter Uncertainty: For spreadsheet-based analytical models, the effect of parameter uncertainty on modeling results can be readily evaluated using companion statistical packages such as Crystal Ball. For a given site, the degree of variability observed in key modeling parameters will determine whether the deterministic or probabilistic modeling method produces the more conservative result.