Transport in Porous Media, Vol.108, No.2, 335-354, 2015
Experimental Study of Foam Flow in Sand Columns: Surfactant Choice and Resistance Factor Measurement
Contamination of soil and groundwater with nonaqueous phase liquid represents a major environmental concern because of the negative effect on human health. Traditional soil flushing techniques have been proved ineffective in heterogeneous aquifers, since liquids for remediation flow preferentially through the most permeable layers and residual oil remains trapped in the low-permeability sediments. To improve the sweep efficiency, fluids with lower mobility than resident liquids can be injected, such as polymers or in situ generated foam. In addition, foam has the potential of selectively reducing mobility in higher-permeability layers compared to lower-permeability layers. The use of foams has not been popular because foam mobility is difficult to predict. In this paper, a series of 1D column experiments were conducted to investigate the variation of the resistance factor RF as a function of the permeability. Five types of sand and calibrated glass beads were used to obtain porous media of permeabilities ranging from 250 millidarcy to 100 Darcy. Two commercial nonionic surfactants, X-100 (commonly used but toxic) and sucrose laurate (environmentally friendly surfactant, biodegradable), were used at a concentration equal to 10 critical micelle concentration. Columns were first saturated with water and then flushed with 1.5 PV (pore volumes) of surfactant solution in order to achieve the adsorption of molecules into the soil matrix. Then, co-injection of air and surfactant solution was performed at a constant total rate to generate in situ foam. The quality of the foam was varied from 85 to 99 % to investigate the effect of this parameter on the resistance factor. A graphic showing the dependence of RF and the permeability is presented. Results suggest that for the same total flow rate, the RF decreases as permeability increases, but for permeabilities smaller than 1 Darcy, the RF is weak, probably due to the difficulty to generate stable foam under high capillary pressures. Variations in foam quality and surfactant type do not seem to have an important effect on the results regarding RF. The presented results are assistance for predicting foam behavior in porous media of different permeabilities.