Because rate-limited partitioning models generally depend on a driving force, the difference between ambient and equilibrium concentration, equilibrium data are important whether LEA is invoked or not. Kerfoot [1991] derived equations for the dependence of gas-water-solid concentrations of contaminants on temperature and pore-water content, assuming an ideal gas, isobaric conditions, and complete mixing. Cline et al. [1991] studied the effects of complex composition on gasoline-water partitioning coefficients and found that an ideal mixture (Raoult's law) offered good predictions. Lane and Loehr [1992] contaminated field soils with 16 polycyclic aromatic hydrocarbon (PAH) solvents, obtaining aqueous concentrations independently by direct measurements, predictions from Raoult's law, and predictions with organic cosolvents. Good agreement among these three methods suggested that Raoult's law was valid, and strong dependence of solubility on organic composition was observed. PAH experiments of Lee et al. [1992a,b] also suggested that nonideality was sufficiently small to allow use of Raoult's law in most field-scale applications. However, the work on tailing by Wise et al. [1992] suggests that limitations exist.
Additionally, multiphase equilibria have been studied in the petroleum industry for decades, and a state-of-the-art summary of the best correlations, obtained by comparing published correlations to a Core Laboratories data base, is available [ McCain, 1991]. Because components (e.g., refined as opposed to crude), phases (NAPL and water instead of liquid and vapor hydrocarbon), and conditions (e.g., much lower pressures) are different in groundwater modeling, these equilibria cannot be applied directly, but an analogue would be desirable.