4.1.4  Volatilization


It is assumed that contaminants in the pond/surface impoundment volatilize into the atmosphere above the pond only from the aqueous phase (i.e., no contaminants volatilize directly from the surface of suspended sediment particles or NAPL-phase globules that may be located at the surface of the pond and that may periodically be in contact with the atmosphere).  This seems to be a reasonable assumption because the vast majority of the pond surface should be an air-water interface rather than a sediment-air or globule-air interface.  With this assumption, and the fact that the contaminants are assumed to be dilute constituents of the aqueous phase, standard theory (USEPA 1988a, 1990) for volatilization from the surface of a water body should apply.  This theory is based on the "two-resistance" model (Thibodeaux 1979; Hwang 1982; Mackay and Leinonen 1975) which assumes that the controlling mechanisms for volatilization from the bulk solution to the atmosphere are the diffusion resistances of thin transition layers in the liquid phase and the gas phase located at the air-water interface.  Based on this theory, the mass flux equation for loss from the pond/surface impoundment by volatilization (either alone or simultaneous with the other processes) can be expressed as



where    Kli is the overall mass transfer coefficient from a liquid phase to the atmosphere for contaminant i, expressed in terms of the liquid-phase concentration (cm yr-1).

When suspended globules of a NAPL phase exist in the pond, the aqueous concentration of the contaminant is controlled by the composition of the NAPL phase.  In this case, the source-term release module calculates the value of Cwi in Equation 4.13 by the phase partitioning theory described in Section 2.2.4 (which uses different methods, depending on the mole fraction [or concentration] of the contaminant in the NAPL phase).

 When no suspended globules of a NAPL phase exist in the pond, the aqueous concentration can be calculated by a simple phase partitioning relation (i.e., Equation 4.9).  Now, substituting Equation 4.9 into Equation 4.13, the mass flux equation for contaminant loss (when no NAPL is present) from the pond/surface impoundment by volatilization (either alone or simultaneous with the other processes) can be expressed as



The expressions in Equations 4.13 and 4.14 represent instantaneous rates of mass loss at a specific time (just like the expressions for all other mass loss term presented thus far).  However, during the course of testing the source-term release module, it was discovered that for certain more volatile compounds, these expressions caused all of the initial contaminant inventory to be lost to volatilization in the firsttime step (which was 1 yr for those runs).  It was hypothesized that the reason for this was that a 1-yr time step was too large for the Runge-Kutta numerical solution for volatilization.  However, applications of RAAS/MEPAS often simulate scenarios that cover an extremely long time period (e.g., 10,000 yr).  So, using a time step significantly smaller than 1 yr would cause undesirably long run times.  In addition, modifying the numerical solution algorithm to use variable-length time steps, the size of which would be determined by limits on the magnitude of contaminant mass that could be lost in any given time step, was deemed beyond the scope of the current version of the source-term release module.

 The approach taken to compensate for these high volatilization loss predictions was to use a mass loss expression that was time-averaged over the time step (rather than instantaneous).  This will reduce the predicted flux because the time-averaged expression accounts for the fact that the flux decreases over time as the aqueous concentration decreases.  One simplifying assumption made is that the time-averaged formula for volatilization loss can be derived by assuming that volatilization is the only process contributing to loss over the time step.  Another simplifying assumption made is that aqueous contaminant is not replenished by desorption from suspended sediment or dissolution from the NAPL phase as it is decreased through volatilization.  This latter assumption must be made because the complex phase partitioning theory required when a NAPL phase is present precludes deriving an analytical expression for the time-averaged mass loss.

 Mackay and Leinonen (1975) began with an expression similar to Equation 4.13, and derived a transient-state expression for volatilization from a pond where only aqueous contaminant is present that was applicable for times t = 0 to t = t. .  The analogous transient-state expression applicable for times between the beginning and end of a time step (of size Dt), is



In Equation 4.15, Cwi should be interpreted as the concentration at the beginning of the time step, and t should be interpreted as the length of time from the beginning of the time step.  As stated above, if we assume that only aqueous mass changes during the time step, Equation 4.15 can be averaged over the time step to obtain



Note that when Kli is small, Equation 4.16 reduces to Equation 4.13; which means that using Equation 4.16 for compounds that are not highly volatile (for which there was no problem with the instantaneous mass loss expressions) is equivalent to using the instantaneous mass loss expressions.

So, when suspended globules of a NAPL phase exist in the pond, the module actually uses Equation 4.16, coupled with phase partitioning theory to determine Cwi.  When no suspended globules of a NAPL phase exist in the pond, the module actually uses



which is the analogue of Equation 4.14 (and reduces to it when Kli is small [or, more precisely, when the argument of the exponential function on the right-hand side of Equation 4.17 is small]).