6. SUMMARY AND CONCLUSIONS

6.1 Summary

A vadose zone leaching, saturated zone mixing, and groundwater flow (VG) computer model was developed to analyze movement of contaminant in a heterogeneous aquifer. A review of pertinent literatures provided insight concerning the factors and processes transporting the contaminant concentration migration through the vadose zone and subsequent mixing within the saturated zone. Previous efforts in aquifer analyses of contaminant movement have provided useful deterministic relationships in homogeneous aquifer system, but also have revealed a lack of knowledge in the heterogeneous aquifer system with linkage of vadose zone and saturated zone mixing.

The VG model developed the program capability to handle simulation results graphically with a Graphic User Interface (GUI) for user friendliness without any commercial programs. The GUI used in this model has made it easy to create the input data file and view the simulation results. Because the model can deal with vertically heterogeneous soil columns which require more parameters to describe the soil properties of each layer, such an easy-to-use interface is even more important than in the previous homogeneous models, such as VLEACHSM 1.0a. The VG model performed with almost instantaneous simulation time on a 200 MHz 80586 PC system to get the final results and graphs in the model described in this dissertation.

The VG model has three major components: a vadose zone leaching, saturated zone mixing, and groundwater flow to the discharge area. The VG model uses input parameters by the user to estimate field contaminant concentration. The purposes of the VG model are to determine the contaminant concentration of heterogeneous vadose zone and saturated zone mixing, to determine how and when contaminants move within the heterogeneous subsurface, and to evaluate the natural rate of groundwater flow and the risk of contamination.

In the first (vadose zone leaching) part of the VG model, the transport processes in all the phases are combined together in one equation using the linear equilibrium partitioning relationships. This not only provides a more consistent formulation, but also is computationally more efficient since only one differential equation needs to be solved. Furthermore, the first (vadose zone leaching) part of the VG model allows the specification of two different types (Dirichlet's and Cauchy's) of boundary conditions at the top of each layer of the soil column.

Using a mass-balance technique, the second (saturated zone mixing) part of the VG model estimates the concentration of contaminants following the mixing of leachate from the vadose zone with groundwater. This sub-model uses the effluent concentration at the bottom of the soil column, which is estimated from the vadose zone leaching sub-model. A complete mixing of the leachate with the groundwater is assumed.

The last (groundwater flow) part of the VG model was developed for vertical heterogeneous two-dimensional steady state flow by using the finite differences method. This sub-model was solved for the distribution of head, flow direction, flow velocities, and estimation of travel time of contaminants.

Code verification of the VG model was identified through reviewing of the same input data of sample problems in previous homogeneous models (LEACHM, VLEACH, and VLEACHSM 1.0a). The calculations for code verification clearly demonstrate that total soil porosity and its water-filled porosity have a strong effect on the liquid-phase contaminant concentration profile.

The VG model performance was evaluated by comparing the results of the soil lab-column studies with the simulation results. The soil property parameters necessary for the model calculations have been estimated through sample testing methods. The sensitivity analysis of the VG model simulation was performed to evaluate the impact of the input parameter of longitudinal dispersivity (a_L = 0.02 ft) on the liquid-phase of the vadose zone and groundwater loading concentration. Also, the approximate time until the contaminant reaches surface water has been estimated from flow rates in the saturated zone beneath the vadose zone.

The comparisons of lab measured data and simulated data clearly demonstrated that the longitudinal dispersivity has a strong effect on the liquid-phase contaminant concentration profile. The experimental data of vadose zone contaminant concentration have mismatched a small bottom portion of the column which may indicate the capillary zone. This mismatch zone suggests that the capillary fringe holds more water and liquid contaminant concentration. However, the measured data just out of the bottom of soil column have been matched with the simulated data. The column study for heterogeneous layers did not show dramatic change of concentrations in layer boundaries because of the similar porosity of samples.

The vadose zone leaching, saturated zone mixing, and groundwater flow (VG) model was applied to a regional aquifer contaminated site at Wood River Junction, Rhode Island. The model simulated the nitrate contaminant leaching in the vadose zone and saturated zone mixing. The bottom of the vadose zone in the column area shows the same historical data (250 mg/l) as 6 days simulated data. The simulated data (60 mg/l) of saturated zone mixing at 6 days indicates close match with observation data (50 mg/l). 8 days later, the simulated data of the saturated zone is the same as the maximum historical data (580 mg/l). The flow pattern shows the route of contaminants within the saturated zone. The flow rate calculated travel time compares with the maximum historical data (580 mg/l) 171 days later, and the total travel time of 458 days to the discharge area.

6.2 Conclusions

The developed model, VG model, proved to be superior to simple homogeneous models, for example, LEACHM (Hutson and Wagenet, 1992), VLEACH (Ravi and Johnson, 1993), and VLEACHSM 1.0a (Lee, 1996). This model has greatly enhanced the VLEACHSM 1.0a capability because it can now account for vertical heterogeneity of soil columns, simulating more complex soil contaminant profiles. This will assist in evaluating the soil and groundwater contamination potential even under complicated geologic situations.

The first and second parts (the heterogeneous vadose zone and saturated zone mixing sub-models) of the developed model present the great improvement of the vadose zone leaching method in modeling leachate contamination. These sub-models established that the heterogeneous vadose zone model, despite its upgrading of existed homogeneous model (VLEACHSM 1.0a), the developed heterogeneous model (VLEACHSM 2.0) reasonably depicts the real complex vadose zone transport processes, including liquid-phase advection, liquid- and vapor-phase dispersion, sorption, and decay of the contaminant.

The measurements of the soil column studies for homogeneous and heterogeneous layers compared favorably with the model simulation for verification purposes. In the lab study, controlling the steady state recharge rate was most important to collect good data. The sensitivity analysis of the longitudinal dispersivity concluded that the parameter of longitudinal dispersivity had significant impact on liquid-phase contaminant concentration. Overall, the verification of this model was successful by comparison with the results of laboratory column experiments.

The VG model can simulate porosity changes by assigning different porosity values for each layer. These simulations gave the same type of graphs in different shorter time steps. However, the program can also assign different values for recharge rates and organic carbon contents, which may also affect the results, although there is no clear justification to make the recharge rate different among the soil layers because water-filled porosity is assumed to be constant.

This model is useful for the evaluation of a regional aquifer contaminant site. In the field application, the results of this model presented a good match with the historical contamination data. Overall, the developed model provides the contaminant leaching profile in the heterogeneous vadose zone and saturated zone mixing, determines how and when contaminants move within the heterogeneous subsurface, and evaluates the natural rate of groundwater flow and the risk of contamination.