Groundwater contamination by nonaqueous phase liquids (NAPLs) such as fuels, wood preserving oils, chlorinated solvents, and PCB oils is a common occurrence throughout industrialized areas of North America. NAPLs which are more dense than water are referred to as dense, non-aqueous phase liquids (DNAPLs) while NAPLs less dense than water are referred to as light, non-aqueous phase liquids (LNAPLs). Although NAPLs exist in the subsurface as a separate liquid phase, they slowly dissolve into flowing groundwater giving rise to aqueous phase plumes. The levels of contamination arising in such plumes is typically far above regulated drinking water standards. The life-span of NAPL in the subsurface is expected to be on the order of several decades to a few centuries depending upon the type of chemicals comprising the NAPL, and site-specific groundwater flow conditions.
Because NAPLs represent potentially toxic, long-term sources of subsurface contamination, much attention has been directed in recent years to the use of aggressive remediation technologies such as surfactant flushing, cosolvent flushing, steam flooding, and air sparging. The majority of these technologies were originally developed in the petroleum industry to facilitate improved hydrocarbon recovery. Their use in environmental applications such as aquifer remediation is relatively new, with most laboratory and field trials having been carried out during the past 8 years. This manual focuses specifically on the use of surfactants and cosolvents to remove NAPL from the subsurface.
As a point of introduction, it is important to note that surfactant/cosolvent flushing is an emerging technology that, to date, has seen no know "full-scale" application.
Surfactants are chemical compounds which have the ability to alter the properties of solution interfaces. There are a number of ways in which surfactants could be utilized for NAPL removal. In a solubilizing surfactant flood, surfactants are injected into the subsurface as an aqueous solution and flushed through the zones containing NAPL. Upon contact with the NAPL, the surfactants bring about an increase in the total aqueous solubility of the chemical components comprising the NAPL through a process referred to as micellar solubilization, thereby accelerating the dissolution process. In a mobilizing surfactant flood, surfactants are again injected as part of an aqueous solution, but the objective is to lower NAPL-water interfacial tension to the point that physical mobilization of the NAPL takes place. This may be undesirable at some DNAPL sites where vertical mobilization may lead to a worsening of the extent of contamination. The degree of solubilization versus mobilization occurring in a surfactant flood can be controlled through appropriate surfactant selection.
Alcohols are similar to surfactants in that they can alter the properties of solution interfaces. Alcohols, such as ethanol, methanol, and isopropyl alcohol can bring about both an increase in aqueous contaminant solubility and a lowering of NAPL-water interfacial tension. Unlike surfactants, alcohols bring about these changes without the formation of micelles in solution, and are therefore often referred to as cosolvents. Because of the tendency of some alcohols to partition significantly into the NAPL phase, the density of the NAPL can be manipulated in-situ through appropriate alcohol selection. Alcohols are often combined with surfactants to improve flood performance. The design of either a surfactant or an alcohol flood represents a significant level of effort, requiring a number of laboratory-based tests and a pilot field trial.
Given the current state of the technology, surfactant or cosolvent floods are most appropriately implemented using a step-wise sequential project approach involving the following:
Whether a surfactant or cosolvent flood should be considered for implementation at a site starts with consideration of remedial objectives. If the objective is to restore an aquifer to typical concentration-based drinking water standards, then surfactant and cosolvent flushing alone will likely not be successful. If the remedial objective is to mitigate risk to human health and the environment, then chemical flooding can be considered, keeping in mind that short-term risks may increase during and shortly after execution of a flood. If the remedial objective is to reduce the amount of mass present at a site, surfactant and cosolvent flooding are capable of attaining this objective if they are appropriately applied.
At each stage of the project, projections of effectiveness, cost, implementability, etc. are progressively refined. Design of the surfactant or cosolvent flood will often be an iterative process. For example, treatability testing of produced fluids may lead to re-formulation of the chemical system in an attempt to strike the appropriate balance between efficiency of the chemical in terms of NAPL removal, produced fluid treatment costs, and ability to reuse the injected chemicals.
In general, a high level of effort is required to properly design a surfactant flood. The choice of surfactant at one site may differ significantly from that at another site given variations in contaminant types, geology, and flow regime. A number of laboratory based tests need to be carried out as part of the design effort, followed by pilot field-scale testing. It is not uncommon to screen the performance of up to 100 different surfactants prior to final selection for a site. Factors which need to be evaluated include degree of enhanced solubilization, interfacial tension lowering, viscosity changes, surfactant degradation, surfactant sorption, emulsion formation, salinity effects, and surfactant precipitation. In addition to chemical design, a high level of effort is required to properly characterize the distribution of contaminants at a site.
Once a particular surfactant, cosolvent, or surfactant-cosolvent combination has been selected to optimize in-situ performance, consideration must be given to deliverability. Deliverability refers to injection of the chemical system such that it properly contacts the targeted NAPL. In general, surfactant/cosolvent flushing will be difficult or impossible to implement in fine grained soils because of the difficulty in delivering fluids. Because the viscosity and density of the delivered chemical system is often different than that of the resident fluids, a hydrodynamically unstable displacement can take place, resulting in poor contact and poor mass removal. Poor contact and mass removal can also result because of subsurface heterogeneity. Techniques to improve deliverability include the use of polymers, foams, and neutral density solutions. At present, one of the largest obstacles to the improvement of in-situ surfactant and cosolvent performance is overcoming the influence of subsurface heterogeneity.
Assessing the performance of a completed surfactant or chemical flood can be done in a variety of ways. Perhaps the most direct method is obtaining a measurement of the amount of mass removed. This can be complicated, however, given the uncertainty associated with estimating the initial amount of contaminant mass present at a site. Other measures of performance include assessing the degree of reduction in soil contaminant concentrations, assessing the reduction in aqueous phase mass flux leaving the site, and assessing the reduction in downstream contaminant concentrations in groundwater. Given the fact that most regulatory standards governing groundwater quality are currently based on aqueous phase concentrations, it would appear that this should be the preferred method of assessment. Measuring groundwater concentrations, however, generally gives little indication of the amount of mass remaining on a site, or the reduction in source life-span that may be associated with application of these technologies.
In addition to subsurface performance and deliverability issues, attention must be given to the design of a produced fluids treatment system. In many cases, surfactant and cosolvent flooding will be significantly less costly if the injected chemical agents can be recovered and managed in a manner that allows re-use. The potential for reusing chemical agents appears greatest in applications of surfactant/cosolvent systems that target the removal of volatile NAPLs, in which the produced fluids can be air stripped to preferentially remove the volatile NAPL constituents. Systems for effective reuse of surfactants are generally in an earlier stage of technology development and demonstration. Many researchers feel that the design of a produced fluids treatment system presents the largest obstacle to surfactant or cosolvent floods being a viable, cost-effective alternative for remediation of subsurface NAPL.
At present, there have been no full-scale applications of surfactant/cosolvent flushing for which information is readily available. Consequently, the technology can be regarded as an emerging and not a proven or "off the shelf" technology. There are currently a significant number of investigations underway to evaluate the technology. For example, 26 field demonstrations are currently ongoing or have been completed. Of these, 12 have been completed and the remaining 14 are in various phases of planning, implementation, data analysis, or reporting. Many of these 14 field demonstrations will be completed in 1997.
The amount of time required to complete a surfactant or cosolvent flood will depend on a number of factors, including the permeability of subsurface sediments, spacing of injection and recovery points, the number of pore volumes required, and the degree of mass removal that is required. In a low-permeability environment such as fine sands or silts, chemical flooding may need to proceed for many years to achieve significant mass removal. For higher permeability media and smaller well spacings, this time may be shortened. The time required to achieve a given degree of mass removal will also depend on the type of chemical flood designed. In general, a surfactant or cosolvent flood designed to achieve accelerated NAPL dissolution will require a longer period of operation than a flood designed to mobilize NAPL.
The performance of a surfactant/cosolvent flushing system should be evaluated in terms of the remedial objectives. At present, neither field-scale demonstrations nor full-scale applications have yet to demonstrate the ability to attain stringent cleanup criteria, such as typical drinking water standards, in the zone of contamination. Likewise, risk mitigation has not yet been demonstrated. Significant mass removal, however, has been demonstrated during most of the field demonstrations. To date, of the 12 field demonstrations, 11 report mass removal data. Three of these demonstrations exhibited mass removals of less than 15 percent, and another two achieved removals from 40 to 60 percent. The remaining six demonstrations had removals greater than 80 percent. The wide range of mass removals in these demonstrations points out the potential success of the technology, but also the poor performance that can occur if the technology is not applied appropriately.
The mobilization of NAPL in response to surfactant or cosolvent flooding can lead to a worsening of the extent of contamination at a site. In the case of DNAPL, any lowering of interfacial tension has the potential to vertically remobilize contaminants. Methods to minimize the risk of vertical DNAPL mobilization include the use of upward hydraulic gradients, the use of swelling alcohols, and the use of tailored surfactant-cosolvent mixtures which promote neutral density emulsions. In addition to NAPL mobilization, it must be recognized that elevated contaminant concentrations in groundwater will occur during a chemical flood, raising the short-term risk of exposure. A related issue pertains to the toxicity of surfactants and cosolvents. Care must be taken to ensure that residual fluids can be removed from the aquifer in an efficient manner. Recent work has focused on the use of food-grade surfactants to address this concern.
The cost of implementing a surfactant or cosolvent flood can vary significantly from site to site. Because costs for actual full-scale implementation do not exist, currently available cost estimates are based on hypothetical examples and extrapolation from field pilot tests. For typical waste sites having contamination limited to the upper 15 m (49 feet) below ground surface, estimated costs range from $1.4 million per hectare ($0.57 million per acre) to $18 million per hectare ($7.5 million per acre). These costs translate to a range of approximately $90 to $990 per cubic meter ($65 to $750 per cubic yard) of treated contaminated soil. Higher costs are generally associated with smaller sites where no economies of scale can be realized, and with lower permeability or heterogeneous soils where flushing needs to be carried out for longer periods of time. The cost of implementing a surfactant or cosolvent flood may vary dramatically from site to site depending on the ability to reuse injected chemicals, methods of waste disposal, and the amount of chemical that needs to be purchased per unit of mass removed.
Whether or not surfactant or cosolvent flooding will be a viable means of achieving the cleanup objectives at a site will depend on the performance and cost of the technology relative to other alternatives. The extent to which surfactant and cosolvent flooding are ultimately used in remediation of NAPL sites will be a function of the success of ongoing technology development and demonstration projects. These projects will hopefully increase the ability to predict performance, and decrease both the uncertainty and magnitude of implementation costs.