Executive SummaryAmendment solutions with or without surfactants have been used to remove contaminants from soil. However, they have drawbacks in that amendment solutions often mobilize the plume, and its movement is controlled by gravity and preferential flow paths. Foam is an emulsion-like, two-phase system in which gas cells are dispersed in a liquid and separated by thin liquid films called lamellae. The potential advantages of using foams in sub-surface remediation include providing better control on the volume of fluids injected, uniformity of contact, and the ability to contain the migration of contaminant-laden liquids. It is expected that foam can serve as a carrier of amendments for vadose zone remediation, e.g., at the Hanford Site. As part of the U.S. Department of Energy's EM-20 program, a numerical simulation capability will be added to the Subsurface Transport Over Multiple Phases (STOMP) flow simulator. The primary purpose of this document is to review the modeling approaches of foam transport in porous media. However, as an aid to understanding the simulation approaches, some experiments under unsaturated conditions and the processes of foam transport are also reviewed.Foam may be formed when the surfactant concentration is above the critical micelle concentration. There are two main types of foams -ball foam (microfoam) and polyhedral foam. The characteristics of bulk foam are described by properties such as foam quality, texture, stability, density, surface tension, disjoining pressure, etc. Foam has been used to flush contaminants such as metals, organics, and nonaqueous phase liquids from unsaturated soil. Ball foam, or colloidal gas aphrons, reportedly have been used for soil flushing in contaminated site remediation and was found to be more efficient than surfactant solutions on the basis of weight of contaminant removed per gram of surfactant. Experiments also indicate that the polyhedral foam can be used to enhance soil remediation.The transport of foam in porous media is complicated in that the number of lamellae present governs flow characteristics such as viscosity, relative permeability, fluid distribution, and interactions between fluids. Hence, foam is a non-Newtonian fluid. During transport, foam destruction and formation occur. The net result of the two processes determines the foam texture (i.e., bubble density). Some of the foam may be trapped during transport. According to the impacts of the aqueous and gas flow rates, foam flow generally has two regimes -weak and strong foam. There is also a minimum pressure gradient to initiate foam flow and a critical capillary for foam to be sustained. Similar to other fluids, the transport of foam is described by Darcy's law with the exception that the foam viscosity is variable.Three major approaches to modeling foam transport in porous media are the empirical, semi-empirical, and mechanistic methods. Mechanistic approaches can be complete in principle but may be difficult for obtaining reliable parameters, whereas empirical and semi-emp...

CH2M Hill Plateau Remediation Company (CHPRC) requested the services of the Pacific Northwest National Laboratory (PNNL) to provide technical support for the Remediation Decision Support activity within the Soil and Groundwater Remediation Project. A portion of the support provided in fiscal year 2009 was used to develop an alternative approach to estimating the soil unsaturated hydraulic conductivity. This alternative approach uses the interfacial-area-based relative permeability (k r) model presented by Embid 1 rather than the more traditional permeability models of Burdine 2 and Mualem. 3 The expectation was that the Embid k r model would improve the estimation of unsaturated conductivity for at least a subset of soil types. Three retention functions (Brooks and Corey, 4 van Genuchten, 5 and modified van Genuchten) were successfully combined with the Embid k r model. The k r relationship from the Brooks-Corey-Embid combination for the wetting phase is identical to that from the Brooks-Corey-Burdine combination. The general performance of the combined models is shown using typical hydraulic parameters. The relative permeability models for the wetting phase were further examined using two datasets from the literature. The results indicate that the interfacial-area-based model can describe the relative permeability of the wetting phase reasonably well. However, the comparison of the k r relationship from the van-Genuchten-Embid combination with that from the van-Genuchten-Mualem combination shows mixed performance results. Further tests are needed with a larger data set.