Chemical flushing for nonaqueous phase liquid solubilization and recovery from porous media

Abstract

Contamination of the subsurface by organic compounds in the form of non-aqueous phase liquids (NAPL) is a major threat to groundwater resources. Many of these compounds are characterized by low water solubility and are highly persistent in the subsurface environment. This means that once the NAPL is inadvertently released into the subsurface, it can act as a secondary source of contamination for many decades to come. Furthermore, although the solubility of most of these compounds is low, the aqueous concentrations encountered in the field are sufficiently high to cause significant adverse effects on human health and the environment. The significance of this problem is exasperated by the wide use of NAPLs in the industry and the fact that many of these compounds are classified as known human carcinogens. Therefore, any cleanup effort must not only attempt to clean the dissolved aqueous plume, but should also focus on the recovery of the released NAPL mass. The remediation of groundwater resources contaminated by NAPL poses a formidable environmental challenge. Currently, there is no known technology that can fully recover the NAPL mass once released into the subsurface. Over the past two decades numerous innovative technologies for the remediation of groundwater with NAPL contamination were developed. One of the most promising is the use of chemical agents, such as cosolvents and surfactants, for the enhanced solubilization, mobilization and removal of NAPL mass. Nonetheless, despite their potential, these technologies have been shown to be insufficient in many field applications in reaching the desired target efficiencies. This is primarily attributed to the complexity of the subsurface system and that the performance of chemical agents can be affected by the interactions between the subsurface properties and chemical agent formulations. The two NAPLs considered in this study are toluene and 1,2-dichlorobenzene. These two compounds are two of the most prevalent contaminants in the subsurface due to their established toxicity and their widespread use. Toluene is a major constituent of petroleum v products such as gasoline, while 1,2-dichlorobenzene is widely used as a degreasing agent for metals, leather and wool. Prior to the application of any in-situ flushing scheme, it is important to understand the phase behaviour of the system, in particular the tendency of the NAPL for enhanced solubilization (increase solubility of the organic phase in the aqueous phase) or partitioning (increase solubility of the aqueous phase in the organic phase). For this purpose a bench scale titration method was used to identify phase compositional changes in the presence of a variety of chemical agents such as alcohols, surfactants and modified sugars. The results of the phase behavior as a function of solution content were presented in the form of ternary phase diagrams which are composed of miscibility curves and tie-lines and are unique in that they show all three components of a reaction system on one plot. The main benefit of the titration based phase change observations was to prepare a quick procedure which enabled rapid evaluation of the effectiveness of different agents for the solubilization and mobilization of NAPLs. Results indicated that the alcohols, in particular ethanol, were effective in solubilizing model NAPLs. However, the effectiveness of alcohols was enhanced when used together with surfactants, provided that the surfactants can be added to the system efficiently. A computer code was also developed to determine the ternary phase behaviour of the water-NAPL-agent system numerically. The code provides a rapid and inexpensive means for the evaluation of different flushing solutions and how the ternary system would behave at different mole fractions. The program solves a system of nonlinear algebraic equations which represent multiphase equilibrium. Because of the nonlinearity of the system, an iterative solution scheme is used to compute the mole factions at equilibrium, the tie lines and plait point. Within the scope of this study, flushing experiments were also conducted in a onedimensional column packed with sand. The target NAPL was toluene, while the solubilizing agent was ethanol. Ethanol content and the flow rate of the flushing solution were evaluated in terms of toluene recovery from the sandpack. The experiments were also conducted at different temperatures to evaluate the overall system efficiency when thermal enhancements were applied in combination with the ethanol flushing solution. The results vi of the column experiments showed that although solubilization was enhanced when the chemical agent content in the flushing solution is increased, equally significant improvements can be achieved with relatively small increases in system temperature. Improved performance was also observed with variable flow regimes that can increase the contact time between the NAPL and the flushing solution. In summary, this study shows that in-situ flushing is a viable remedial scheme for groundwater contaminated with organic compounds in the form of NAPLs. Yet, significant challenges remain for the successful transfer of this technology to the field. Moreover, it is important to recall that while extraction of the NAPL from the porous medium is the primary focus of this study, in real applications such a scheme would produce large volumes of waste water. Efforts must also focus on methods for the separation and recycling of the NAPL and agent volumes.