Article Cite This: Environ. Sci. Technol. 2018, 52, 4256−4264
pubs.acs.org/est
Transport and Retention of Concentrated Oil-in-Water Emulsions in Porous Media Katherine A. Muller,*,†,§ Somayeh G. Esfahani,†,‡ Steven C. Chapra,† and C. Andrew Ramsburg† †
Department of Civil and Environmental Engineering, Tufts University, 200 College Avenue, Room 204 Anderson Hall, Medford, Massachusetts 02155, United States ‡ Department of Civil, Architectural and Environmental Engineering, University of Texas at Austin, Austin, Texas 78712, United States S Supporting Information *
ABSTRACT: Oil-in-water emulsions are routinely used in subsurface remediation. In these applications, high oil loadings present a challenge to remedial design as mechanistic insights into transport and retention of concentrated emulsions is limited. Column experiments were designed to examine emulsion transport and retention over a range of input concentrations (1.3−23% wt). Droplet breakthrough and retention data from low concentration experiments were successfully described by existing particle transport models. These models, however, failed to capture droplet transport in more concentrated systems. At high oil fraction, breakthrough curves exhibited an early fall at the end of the emulsion pulse and extending tailing. Irrespective of input concentration, all retention profiles displayed hyper-exponential behavior. Here, we extended existing model formulations to include the additional mixing processes occurring when at high oil concentrationswith focus on the influence of deposited mass and viscous instabilities. The resulting model was parametrized with low concentration data and can successfully predict concentrated emulsion transport and retention. The role of retained mass and viscous instabilities on mixing conditions can also be applied more broadly to systems with temporal or spatially variant water saturation or when viscosity contrasts exist between fluids.
■
INTRODUCTION Oil-in-water emulsions have been used in environmental remediation for a variety of applications including: enhanced contaminant recovery (e.g., refs 1,2); contaminant stabilization (e.g., ref 3); fermentable substrate delivery (e.g., refs 4−6); mobility control (e.g., refs 7−9); and to deliver active ingredients to the subsurface (e.g., refs 10−12). Utilization of emulsions for in situ remediation requires a balance between retention of remedial amendments and ease of introducing and distributing the amendment in the subsurface. Concentrated emulsions (e.g., in excess of 10% wt dispersed phase content) or neat edible oils can be used to reach remediation outcomes.4,6,13 However, emulsion transport and retention, especially at high concentration, remains largely empirical. Remediation design can be improved through a greater emphasis on the processes controlling emulsion mobility and retention in porous media. A critical aspect of emulsion transport modeling is the need to incorporate changes to the droplet retention characteristics over the course of a retention event (e.g., through a reduction of the filter coefficient,14 introduction of surface capacity,15,16 or conceptualization of film flow component17). The influence of input concentration, emulsion viscosity, and droplet retention on mixing and tailing is largely absent in emulsion transport models, despite observations of emulsion tailing (e.g., refs 15,18,19). Increased colloid mobility with increasing input concentration has been demonstrated within the colloid literature, yet the influence of input concentration in the emulsion © 2018 American Chemical Society
literature is limited. Neglecting input concentration is particularly troubling, given that droplet concentration (or oil fraction) can cause emulsion viscosities to vary over orders of magnitude. For example, Soo & Radke14 exclude emulsions greater than 1% from their modeling effort, and Coulibaly et al.15 suggest degraded model performance when emulsion density and viscosity deviate from that of water. Viscosity contrasts between invading and resident fluids, regardless of miscibility, can result in flow instabilities. Specifically, conditions are favorable for viscous fingering when a less viscous fluid displaces a more viscous fluid.20 This is a relevant, but often overlooked, aspect of amendment delivery, as groundwater re-enters a treatment zone following a period of amendment injection (amendments have viscosities greater than that of water). The resulting instability may or may not be negligible. But the influence of viscous fingering on transport during miscible displacements such as those experienced during concentrated emulsion delivery, needs to be accounted for using averaged models or direct numerical simulations of the physical fingering process (e.g., refs 20,21). Viscous fingering degrades the applicability of standard mechanical dispersion assumptions;22 however, when the dispersed phase is conceptualized as Received: Revised: Accepted: Published: 4256
November 23, 2017 March 1, 2018 March 8, 2018 March 9, 2018 DOI: 10.1021/acs.est.7b06012 Environ. Sci. Technol. 2018, 52, 4256−4264
Article
4257
98 101 102 104 98 100 100 0.65 0.75 0.89 0.94 5.91 5.73 7.44 1.42 1.36 2.77 2.65 28.3 30.5 38.1 8.2 7.8 7.8 (±0.1) 23.0 (±0.1)
7.5 7.9b 22.7 (±0.4)
7.9
1.6 1.6 (±0.1) 2.32 (±0.0)
1.7
2.6 2.6 2.3 2.3 2.4 2.5 3.0 1.5 1.4 n.m. 1.25 (±0.1)
0.028 0.034 0.042 0.052 0.084 0.072 0.021 1.05 1.03 1.06 1.02 1.03 1.03 1.00 67.7 67.2 68.1 66.6 71.9 67.8 68.6 9.8 9.8 10.4 10.3 10.4 10.2 10.4 1.65 1.65 1.70 1.71 1.64 1.68 1.69
(g) (PV) (−)
M μe
(mPa·s) (mPa·s)
μ C0
(% wt)
α0
(cm) (mL·min−1)
Q PV
a
FF FF FF FF FF FF FF 1A 1B 2A 2B 3A 3B 4
L
(cm)
ρb
(g· cm−3)
n
(−)
(mL)
dispersed phase content initial dispersivity measured flow rate pore volume packed length bulk density porosity porous media experiment
Table 1. Experimental Column Parameters and Results
measured viscosity
modeled viscosityc
mobility ratio
emulsion introduced
MATERIALS AND METHODS Materials. Soybean oil (SBO, MP Biomedicals, Laboratory grade), Gum Arabic (GA, > 99% purity), sodium bromide (NaBr, 99.9% purity, ACS grade), and sodium chloride (NaCl, 99.5% purity, ACS grade) were purchased from Fisher Scientific. High purity water, denoted Milli-Q water, (resistivity >18.2 mΩ-cm and total organic carbon
