Remediation of NAPLs in groundwater using in situ air sparging

A comprehensive investigation has been conducted to study the different parameters that affect the performance and effectiveness of in situ air sparging. To achieve the target goal synchronized laboratory testing coupled with assessment of full-scale field systems and mathematical model simulation performed. The laboratory tests were performed to investigate the effect of different contaminant type, operational parameters and porous media on mass removal rate during air sparging. The laboratory experimental testings were performed with both one and two dimensional test apparatus. The one dimensional tests were conducted to study different spatially independent variables that affect mass removal during air sparging. The spatially independent variables considered for this investigation are contaminant NAPL type, initial concentration and phase of contamination. The two dimensional tests on the other hand were conducted to investigate spatially dependent variables that have impact on mass removal and system performance during air injection. These variables include effect of media heterogeneity, effect of groundwater flow and air injection flow rate. In addition to the laboratory investigation full and pilot scale field system assessments were performed to investigate for possible correlation between mass removal and different system and site-specific variables. The data for this study was extracted from two well organized air sparging databases. The investigation indicated that there is no clearly defined correlation between the variables considered and mass removal.; The last topic in this research was mathematical modeling of air sparging. TOUGH2 multiphase three dimensional models have been applied for this investigation. Both one and two dimensional laboratory experiments were modeled using this numerical mode. The model successfully simulated the laboratory experiments with some discrepancies. The discrepancies between the laboratory tests and model simulation was due to assumptions and model input variation. The model assumes a constant hydraulic permeability and porosity while in reality horizontal and vertical permeabilities are different. The second reason for the observed discrepancies was the assumed mode of airflow. The model assumes the porous media to have a constant hydraulic permeability therefore uniform airflow with respect to one axis of flow. But the real observation during laboratory experiment showed either bubble/channel mode of airflow depending on particle size of the porous media. Modifying the source code of the TOUGH2 module will help to tailor the software for modeling air sparging laboratory test.