journal of contaminant hydrology - western engineering · drainage were not representative of...

12
Wettability contrasts between fresh and weathered diesel fuels Stephanie S. Drake 1 , Denis M. O'Carroll , Jason I. Gerhard Department of Civil & Environmental Engineering, The University of Western Ontario, London, ON, Canada N6A 5B8 article info abstract Article history: Received 19 May 2012 Received in revised form 11 September 2012 Accepted 28 September 2012 Available online 10 October 2012 The remediation of non-aqueous phase liquid (NAPL) contaminated sites is impeded due to subsurface complexities, including wettability. Wettability quantifies which of two immiscible fluids preferentially coats a solid. At most contaminated sites water-wetting conditions are typically assumed despite mounting evidence that this is not always the case. In this study, wettability was examined for two NAPL samples of contrasting origin: a fresh and a field sample. Wettability was assessed through (i) cyclical, cumulative elapsed contact timeintrinsic contact angle measure- ments, (ii) interface jar tests, and (iii) cyclical, pseudo-static capillary pressuresaturation curves. The work as a whole demonstrated that while the fresh diesel sample was consistently water-wet, the field diesel sample exhibited repeatable cycles of wettability reversal between water drainage and imbibition. And while wettability hysteresis increased with contact time for the field diesel, the occurrence of wettability reversal at each change of saturation direction was independent of contact time. Such behavior is not easily assessed by standard wettability indices. Moreover, it contrasts with the permanent wettability alteration observed for complex organics (e.g., coal tar) observed in most studies. It is hypothesized that the cyclical wettability reversal is related to cyclical changes in intermediate pore wettability due to sorption of surface active compounds (causing NAPL-wetting imbibition) and rupturing of the soil grain water film (causing water-wet drainage). The wettability differences between the two NAPLs may be due to additives (i.e., a surfactant) in the original formulation and/or byproducts from subsurface weathering. These results support better characterization of site-specific wettability, improved model development and more realistic site conceptual models for improved remediation efforts. © 2012 Elsevier B.V. All rights reserved. Keywords: Wettability NAPL Remediation Retention functions 1. Introduction Remediation efforts at Non-Aqueous Phase Liquid (NAPL) contaminated sites often do not reduce the contamination to health-based standards (National Research Council, 1999). This may be due to an incomplete understanding of subsurface heterogeneities, including wettability (Dwarakanath et al., 2002). Wettability is the affinity of one fluid for a given solid in the presence of another immiscible fluid (Craig, 1971). Many aquifers are composed primarily of quartz sand or other naturally hydrophilic materials (Powers et al., 1996), therefore it is commonly assumed in conceptual models of contaminated sites that the porous medium is strongly water-wetting. This strongly water-wet condition is herein referred to as ideal wettability. There is mounting evidence, however, that the assumption of ideal wettability may not be appropriate at a significant fraction of contaminated sites (Dwarakanath et al., 2002; Harrold et al., 2001, 2003; Jackson and Dwarakanath, 1999; Powers et al., 1996; Zheng and Powers, 1999). Wettability can be quantified at the scale of a single interface by measuring the intrinsic contact angle (θ int ), which is dependent on the interfacial tension (IFT) between the immis- cible fluids. A surface is typically defined as water-wetting for (approximately) θ int b 60° to 75° (measured through the aqueous phase), NAPL-wetting for θ int > 105° to 120° and intermediate- wet for 75° b θ int b 105° (Anderson, 1986). For air/water/quartz, advancing contact angles (i.e., air displacing water) are small Journal of Contaminant Hydrology 144 (2013) 4657 Corresponding author. Tel.: +1 519 661 2198; fax: 519 661 3942 E-mail address: [email protected] (D.M. O'Carroll). 1 Now at WorleyParsons, Calgary, AB, Canada T3B 5R5. 0169-7722/$ see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jconhyd.2012.09.008 Contents lists available at SciVerse ScienceDirect Journal of Contaminant Hydrology journal homepage: www.elsevier.com/locate/jconhyd

Upload: others

Post on 12-Feb-2020

3 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Journal of Contaminant Hydrology - Western Engineering · drainage were not representative of pressures in the porous medium. Once the diesel had entered the porous medium it was

Journal of Contaminant Hydrology 144 (2013) 46–57

Contents lists available at SciVerse ScienceDirect

Journal of Contaminant Hydrology

j ourna l homepage: www.e lsev ie r .com/ locate / jconhyd

Wettability contrasts between fresh and weathered diesel fuels

Stephanie S. Drake 1, Denis M. O'Carroll⁎, Jason I. GerhardDepartment of Civil & Environmental Engineering, The University of Western Ontario, London, ON, Canada N6A 5B8

a r t i c l e i n f o

⁎ Corresponding author. Tel.: +1 519 661 2198; faxE-mail address: [email protected] (D.M. O'Car

1 Now at WorleyParsons, Calgary, AB, Canada T3B 5

0169-7722/$ – see front matter © 2012 Elsevier B.V. Ahttp://dx.doi.org/10.1016/j.jconhyd.2012.09.008

a b s t r a c t

Article history:Received 19 May 2012Received in revised form 11 September 2012Accepted 28 September 2012Available online 10 October 2012

The remediation of non-aqueous phase liquid (NAPL) contaminated sites is impeded due tosubsurface complexities, includingwettability.Wettability quantifieswhich of two immiscible fluidspreferentially coats a solid. At most contaminated sites water-wetting conditions are typicallyassumed despite mounting evidence that this is not always the case. In this study, wettability wasexamined for two NAPL samples of contrasting origin: a fresh and a field sample. Wettability wasassessed through (i) cyclical, ‘cumulative elapsed contact time’ intrinsic contact angle measure-ments, (ii) interface jar tests, and (iii) cyclical, pseudo-static capillary pressure–saturation curves.The work as a whole demonstrated that while the fresh diesel sample was consistently water-wet,the field diesel sample exhibited repeatable cycles of wettability reversal between water drainageand imbibition. Andwhile wettability hysteresis increased with contact time for the field diesel, theoccurrence ofwettability reversal at each change of saturation directionwas independent of contacttime. Such behavior is not easily assessed by standard wettability indices. Moreover, it contrastswith the permanent wettability alteration observed for complex organics (e.g., coal tar) observed inmost studies. It is hypothesized that the cyclical wettability reversal is related to cyclical changes inintermediate pore wettability due to sorption of surface active compounds (causing NAPL-wettingimbibition) and rupturing of the soil grainwater film (causingwater-wet drainage). Thewettabilitydifferences between the two NAPLs may be due to additives (i.e., a surfactant) in the originalformulation and/or byproducts from subsurface weathering. These results support bettercharacterization of site-specific wettability, improved model development and more realistic siteconceptual models for improved remediation efforts.

© 2012 Elsevier B.V. All rights reserved.

Keywords:WettabilityNAPLRemediationRetention functions

1. Introduction

Remediation efforts at Non-Aqueous Phase Liquid (NAPL)contaminated sites often do not reduce the contamination tohealth-based standards (National ResearchCouncil, 1999). Thismay be due to an incomplete understanding of subsurfaceheterogeneities, including wettability (Dwarakanath et al.,2002). Wettability is the affinity of one fluid for a given solidin the presence of another immiscible fluid (Craig, 1971). Manyaquifers are composed primarily of quartz sand or othernaturally hydrophilic materials (Powers et al., 1996), therefore

: 519 661 3942roll).R5.

ll rights reserved.

it is commonly assumed in conceptual models of contaminatedsites that the porous medium is strongly water-wetting. Thisstrongly water-wet condition is herein referred to as ‘idealwettability’. There is mounting evidence, however, that theassumption of ideal wettability may not be appropriate at asignificant fraction of contaminated sites (Dwarakanath et al.,2002; Harrold et al., 2001, 2003; Jackson and Dwarakanath,1999; Powers et al., 1996; Zheng and Powers, 1999).

Wettability can be quantified at the scale of a single interfaceby measuring the intrinsic contact angle (θint), which isdependent on the interfacial tension (IFT) between the immis-cible fluids. A surface is typically defined as water-wetting for(approximately) θintb60° to 75° (measured through the aqueousphase), NAPL-wetting for θint>105° to 120° and intermediate-wet for 75°bθintb105° (Anderson, 1986). For air/water/quartz,advancing contact angles (i.e., air displacing water) are small

Page 2: Journal of Contaminant Hydrology - Western Engineering · drainage were not representative of pressures in the porous medium. Once the diesel had entered the porous medium it was

47S.S. Drake et al. / Journal of Contaminant Hydrology 144 (2013) 46–57

(e.g., 2°). Other materials such as shales, coal, graphite, talc,dolomite and calcite exhibit significantly higher air/watercontact angles (e.g., 33° to 82°) (Ryder and Demond, 2008).And, while wettability is partly dependent on porous mediummineralogy, it has been suggested that fluid composition has thegreatest influence on wettability in the saturated zone (Harroldet al., 2001; Ryder andDemond, 2008). For example, the additionof surfactants to an air/water/quartz system can significantlyincrease measured contact angles (e.g., ~47°) (Desai et al., 1992;Lord et al., 2000). Pure NAPL/water/quartz contact anglesreported in the literature range from 5° to 72° (Barranco et al.,1997; Demond and Roberts, 1991; Powers et al., 1996) whileNAPLs containing surfactants exhibit larger contact angles(e.g., 15° to 180°) (Demond et al., 1994; Lord et al., 1997a;Molnar et al., 2011; Powers and Tamblin, 1995). Contact anglesmeasured for field or waste NAPL samples on quartz have alsoshown awide range of contact angles: from 12° to 163° (Harroldet al., 2001; Powers et al., 1996).Wettability is further dependenton the order of fluid contact: a typically water-wet system maybeNAPL-wetting if theNAPL phase contacts the solid first (Ryderand Demond, 2008).

The macroscopic effects of non-ideal wettability (i.e., at thescale of a representative elementary volume, REV), are apparentin the capillary pressure–saturation (Pc–S) relationships. As afluid/fluid/solid system becomes less-water wet, the capillarypressure required to achieve a given saturation decreases,compressing the Pc–S curve (Bradford and Leij, 1995a;O'Carroll et al., 2005; Powers and Tamblin, 1995; Powers et al.,1996). This compression is typically accounted for by modifyingstandard Leverett (1941) Pc–S scaling to include a contact angleterm (Bradford and Leij, 1995a,b; Demond and Roberts, 1991;O'Carroll et al., 2005). This contact angle is defined as theoperative contact angle (θop) and is typically less than theintrinsic contact angle. Pc–S scanning curves are less commonlyexplored, particularly with respect to wettability, but can revealvaluable characteristics associated with governing surfacechemistries and pore-filling behavior on saturation reversals(Gerhard and Kueper, 2003; Kovscek et al., 1993).

NAPLs that are used as solvents and fuels typically containcompounds added to improve their performance. These addi-tives can impact interfacial behavior, specifically those with anamphiphilic structure (i.e., surfactants), which can sorb onto thesolid surface, altering surface charge. For example, polybuteneamine, the active ingredient in products added to fuels to cleanthe carburetor and fuel injector, has rendered quartz surfacesneutral wet in the presence of synthetic fuel and water (Powersand Tamblin, 1995). Furthermore, spent solvents and fuels arelikely composed of a wide range of constituents that they cameinto contact with during their useful life. Chlorinated solventshave been used as degreasers and in the dry cleaning industrywhere surfactant use is common to improve detergency. Theseadditional constituents can have a significant impact on solidphase wettability, rendering the solid surface intermediate toNAPL-wet (Demond et al., 1994; Harrold et al., 2001, 2005; Hsuand Demond, 2007; Jackson and Dwarakanath, 1999; Molnar etal., 2011). Similarly crude oils, coal tars and creosotes, which arecomprised of a wide range of constituents, frequently wetquartz surfaces. This has been attributed to the presence of highmolecular weight asphaltenes with polar functional groups thatreduce adhesive forces (Barranco andDawson, 1999;Dong et al.,2004; Hugaboom and Powers, 2002; Nelson et al., 1996; Powers

and Tamblin, 1995; Powers et al., 1996; Ren et al., 2009; Zhengand Powers, 2003; Zheng et al., 2001).

Weathering processes such as leaching, volatilization andbiological decay are all mechanisms by which a NAPL canchange composition in the subsurface and potentially inducenon-ideal wettability. For example, weathering processesmay degrade lighter compounds of petroleum-derived prod-ucts, leaving behind heavier hydrocarbons that are known toalter wettability (Cuiec, 1990; Powers et al., 1996). As well,oxidation of NAPL systems has been shown to shift water-wetting systems to intermediate- andNAPL-wetting conditions(Cuiec, 1990; Ren et al., 2009).

These studies provide insight into how additives in fuels orweathering of waste/field NAPLs can lead to non-ideal wetta-bility behavior. However a direct correlation between constit-uents present (or absent) in waste/field NAPLs and wettabilityalterations is still unclear. This is complicated by the significantnumber of constituents typically present and the potential forsynergistic effects of these constituents impacting wettability.For example, the presence of both organic acid and basesurfactants in a NAPL/water/quartz system has been found toalter interfacial activity more than the sum of the individualeffects (Hsu and Demond, 2007; Spildo et al., 2001). Withmultiple additives and impurities in field systems, as well ascomplex release histories for many field NAPLs, it is therefore asignificant challenge to a priori predict thewetting behavior of agiven NAPL.

The purpose of this study was to investigate the wettabilityof a diesel/water/quartz system and, in particular, the contrastbetween aweathered field diesel and a fresh diesel.Wettabilitywas investigated at (i) the pore scale via cyclical advancing andreceding intrinsic contact angle measurements, (ii) at the scaleof an interface through time lapse and equilibrium bottle tests,and (iii) at the REV scale through multiple, hysteretic, Pc–Sexperiments including detailed scanning curves. The interface-scale analysis was complemented by testing of fluid andinterfacial matter composition. The REV scale analysis wasaugmented through Leverett scaling of different fluid-fluidpairs and Pc–S curve fitting. This comprehensive analysis of thissingle fluid pair aims to provide a holistic picture of (i) thewettability thatmay be expected for a petrochemical recoveredfrom the field, (ii) the types of measurements that are mostindicative of the observed behavior, and (iii) the potentialunderlying causes of that behavior. The potential applicationsof the information obtained include improving both numericalsimulations and remediation of contaminated sites.

2. Materials and methods

2.1. Materials

Two diesel fuel oils were used as the non-aqueous phases inthe experiments. A field diesel was obtained from a monitoringwell at a contaminated site in London, Ontario. The NAPL wasstored in a tightly sealed container in absence of light at 5 °Cfrom time of sampling until time of equilibration for testing toensure the sample maintained its properties at representativeaquifer conditions. The fresh diesel was obtained from a gasstation (Drummond Fuels Ltd., Ottawa, ON). Equilibration of thediesels with the aqueous phase occurred at room temperature(25°), the same temperature employed for experiments. The

Page 3: Journal of Contaminant Hydrology - Western Engineering · drainage were not representative of pressures in the porous medium. Once the diesel had entered the porous medium it was

48 S.S. Drake et al. / Journal of Contaminant Hydrology 144 (2013) 46–57

specific gravity of the NAPLs was measured with a pycnometer.The aqueous phase used in all experiments was distilled,de-ionized water (Milli-Q filters, Millipore, Billerica, MA). Asaqueous phase ionic strength has been known to impact IFT(Barranco et al., 1997; Lord et al., 1997b; Standal et al., 1999), IFTmeasurements were additionally taken for the field diesel and0.01 M solution of NaCl. The average measured IFT of the fielddiesel/0.01 M NaCl solution was within the 95% confidenceinterval of the average measured IFT for field diesel/Milli-Qwater.

To obtain reproducible primary drainage Pc–S data, thefield diesel was pre-filtered through a 0.45 μm filter toremove particulate matter. Initial experiments with the fielddiesel were conducted with unfiltered samples, causing theupper hydrophobic membrane (through which the NAPLmust pass before it enters the porous medium) to act as afilter, removing any particulate matter. This resulted inpartial clogging of the upper membrane, thus NAPL pressurereadings measured by the pressure transducers on primarydrainage were not representative of pressures in the porousmedium. Once the diesel had entered the porous medium itwas essentially filtered and therefore imbibition data wasconsistent for both filtered and unfiltered samples.

Quartz was used as the solid phase as it is a ubiquitousaquifer mineral and has been widely studied. This ensuredthe focus of the investigation was on wettability alterationsdue to NAPL properties alone. Two silica sand size distribu-tions (Opta Minerals Inc., Waterdown, ON) were used in thecapillary pressure–saturation experiments: (i) 50% by weightmixture of F32 and F50 for experimental system validation(mean grain size=0.42 mm, uniformity index=2.3), and(ii) F70 for detailed analysis (mean grain size=0.18 mm,uniformity index=1.6) (Camps-Roach et al., 2010). Contactangles were completed on smooth quartz plates (Alfa Aesar,Ward Hill, MA). All glassware and apparatus were cleanedusing established procedures (Lord et al., 1997b) to eliminatethe presence of impurities that may impact interfacialbehavior.

2.2. Weathering

GC–MS analysis was used to establish the degree ofweathering by determining the presence of diesel constituents.An Agilent 7890 GC with a DB-5 column (30 m×0.25 mm i.d.,0.25 μm thickness) was coupled with an Agilent 5675 MSD.The GC–MS method used was based on the weathered dieselanalysismethod of Lang et al. (2009) for GC–MS analysis. Freshand field diesel samples were diluted with dichloromethane(DCM) using a 5:1 ratio.

2.3. Interfacial analysis

Bottle tests were performed to observe temporal changesin the diesel–water interface in the absence of a porousmedium. Vials were suspended over the camera (Nikon D80)so the interface was directly above the field of view. The 3 mLvials contained 1 mL of de-ionized water and 2 mL of dieselthat was filtered through a 0.2 μm syringe filter. Photographswere taken at two-hour intervals over a period of two weeks.Additional tests were performed to determine if microorgan-isms were accumulating in the vials. An inoculation loop was

used to streak samples from the jar tests onto LB-Agar platesand left to incubate at room temperature for 13 days.

2.4. Interfacial tension and contact angle

Interfacial tension and intrinsic contact angle measurementswere obtained using an Axisymmetric Drop-Shape Analysissystem (First Ten Angstroms, Portsmouth, VA). The systememploys Young–Laplace drop shape analysis applied to highresolution photographs of light non-aqueous phase liquid(LNAPL) drops in water; a j-shaped needle was used to float anLNAPL bubble in water(for IFT) or to introduce an LNAPL bubbleagainst the bottom of a quartz plate (for contact angle). Allmeasurements for both IFT and contact angle were conductedwith systems equilibrated in a consistentmanner. A beaker with30 mL of the aqueous phase containing a quartz plate and aninverted vial with 2 mL of LNAPL was allowed to equilibrate forat least 72 h. This setup allowed equilibration between all threephases while still allowing the quartz plate to be removed fromthe aqueous phase without contacting the NAPL. This setupensured the contact angle measurements were analogous to theinitially water-saturated Pc–S experiments (corresponding toNAPL encountering an uncontaminated aquifer).

LNAPL contact angles were achieved by repeatedly advanc-ing and receding an LNAPL drop on the same area of a quartzplate. Between advancing and receding contact angle mea-surements, LNAPL drops were left to equilibrate on the quartzplate. Advancing and receding contact angle measurementswere taken at multiple equilibration time intervals to investi-gate temporal contact angle hysteresis. At each time interval, atleast fivemeasurements for each of the advancing and recedingcontact angles were taken and averaged. In cases where themeasurements were not consistent (e.g., the receding contactangle varied as the drop receded) or a single value was noteasily determined (e.g., due to imperfect resolution of theintersection of the contact line and surface) the maximum andminimum possible contact angles are reported. This cyclicaland ‘cumulative elapsed contact time’ contact angle measure-ment method was developed specifically to imitate the cyclicalsaturation reversals experienced by the quartz sand in the Pc–Sexperiments (i.e., NAPL contact times in the porous mediavaried from experiment to experiment, particularly whenconducting scanning curve experiments).

2.5. Capillary pressure–saturation experiments

Capillary pressure–saturation experiments were conductedusing a membrane-based stainless steel pressure cell based onan established design (Salehzadeh and Demond, 1999). An airpressure regulator (Model 44-2,Moore Products, SpringHouse,PA)was used to control NAPL pressure in a reservoir connectedto the top of the pressure cell (Fig. 1). Stainless steel porousplates (20 μm pores and 1.6 mm thickness, Mott Corp.,Farmington, CT) served as structural support for the upperhydrophobic membrane (PTFE with 0.45 μm pores, Pall Corp.,Port Washington, NY) and the lower hydrophilic nylonmembrane (0.2 μm pores). The cells were wet-packed inthree layers using Milli-Q water that had equilibrated withthe NAPL for at least 72 h using a method similar to that usedby Chen et al. (2007). 25 pore volumes of the aqueous solution

Page 4: Journal of Contaminant Hydrology - Western Engineering · drainage were not representative of pressures in the porous medium. Once the diesel had entered the porous medium it was

Fig. 1. Experimental setup for running two parallel, pseudo-static capillary pressure–saturation experiments.

49S.S. Drake et al. / Journal of Contaminant Hydrology 144 (2013) 46–57

were then flushed through the packed cell using a syringepump (kd Scientific, Holliston, MA).

Pc–S data was obtained using an automated pseudo-staticsystem (Chen et al., 2007). The NAPL pressure was graduallyincreased by increasing the air pressure above the NAPL surface(Fig. 1) at a slowand constant rate (0.2 cmwater/min). Pressurereadings were collected every 10 s using gauge pressuretransducers (FP2000, Honeywell, Morristown, NJ) connectedto a datalogger (CR3000, Campbell Scientific, Edmonton, AB).The pressure transducers were used to quantify both phasepressure and saturation (Chen et al., 2007). Pc values wereaveraged over 0.5% saturation bins, and each resulting Pc datavalue (composed of, on average, 12–15 measurements) wasreported at themidpoint of each saturation bin. A complete Pc–Sprimary drainage/main imbibition curve was completed inapproximately 15 h. The system was validated by demonstrat-ing that the Pc–S curve for a silicone oil/water/F32-50 sandsystem determined by a standard static experiment lay withinthe 95% confidence interval about the mean of four repeatpseudo-static Pc–S curves (Drake, 2010).

Six experiments consisting of primary drainage and mainimbibition curves were completed for the fresh diesel with F70sand. In addition, two of the six experiments included scanningloops (6 loops over 3 unique saturation paths and 21 loops over9 unique saturation paths, respectively). For the field diesel,eight experiments with F70 sand were completed: five withunfiltered and three with filtered field diesel; this resulted inreliable data for three drainage and eight imbibition Pc–Scurves. One of the eight experiments included scanning curves(13 scanning loops over 11 unique saturation paths).

All data was plotted as apparent saturation (Bradford et al.,1998) to accommodate curve-fitting and comparison betweendata sets; this is reasonable because the focus in this work is onsignificant differences in capillary pressure values characteriz-ing the majority of the Pc–S main branches and not on the

details of the residual water or NAPL at the curve endpoints. Amodified form of the van Genuchten (1980) equation, incor-porating a translation parameter (η) that satisfies (Pc+η)≥0for all Pc , enabled fitting of both the positive and negative Pcvalues observed (Bradford and Leij, 1995a). The best-fit vanGenuchten parameters were determined by the method ofO'Carroll et al. (2005), which minimized the root mean squaredifference of saturation values, putting importance on theintermediate saturation range.

Operative contact angles, quantifying the hysteresis betweendrainage and imbibition for the measured Pc–S imbibitioncurves, were determined through modified Leverett scaling.The fresh diesel/water drainage curve, taken as the referencecurve, was scaled to the imbibition curves via the modifiedLeverett–Cassie equation (O'Carroll et al., 2005):

Pαβc Sappα� � ¼ γαβ

γrefcos θαop

� �Prefc Sappw

� �� �ð1Þ

where Pcαβ(Sαapp) is the Pc for a given imbibition curve (fluids α

and β) at apparent saturation of the α phase corresponding tothe reference Pc at the same saturation (Pcref(Swapp)),γαβ is the IFTbetween fluids α and β, γref is the reference system IFT and θopα isthe operative contact angle through the α (wetting) phase. TheLeverett–Cassie equation was developed for systems withheterogeneous wettability, however it reduces to Eq. (1) whenthe porous medium is homogeneous.

3. Results and discussion

3.1. Weathering analysis of diesels

The GC/MS chromatograms for the two diesel samples areconsistent with expectations given their origins. The freshdiesel chromatogram (Fig. 2) reveals the strong presence of

Page 5: Journal of Contaminant Hydrology - Western Engineering · drainage were not representative of pressures in the porous medium. Once the diesel had entered the porous medium it was

Fig. 2. Fresh diesel GC/MS chromatogram illustrating major alkane species occurring at regular intervals across chromatogram.

50 S.S. Drake et al. / Journal of Contaminant Hydrology 144 (2013) 46–57

straight-chained alkanes (n-alkanes) that peak at regular in-tervals over the chromatogram range (Christensen and Larsen,1993); in particular, the major components present are thetwelve dominant alkanes ranging from n-C10 to n-C21, whichare typical of a diesel fuel (Galperin and Kaplan, 2008). Incontrast, the field diesel chromatogram (Fig. 3) exhibitsprominent peaks for pristane and phytane while the n-C17peak is absent and the n-C18 peak is relatively small. Whilen-alkanes such as n-C17 and n-C18 are subject to biodegrada-tion, their closely related isoprenoid compounds – pristane andphytane– are not; for this reason, the ratio of n-C17 to pristane istypically correlated to the age of a diesel spill (Christensen andLarsen, 1993). Conservative estimates suggest that n-C17 isexpected to be completely degraded after two decades in thesubsurface, although this can occur in as little as 5–6 years underideal, aerobic conditions (Hurst and Schmidt, 2005). The absenceof n-C17 in the field diesel chromatogram (Fig. 3) suggests thatthis dieselwas significantlyweathered. Oxidation productswerenot observed in either chromatogram. Refined petroleumproducts commonly have stabilizer additives, included to inhibit

oxidation of the fuels during storage (Song et al., 2000). Theseadditives may be present in either or both diesel samples.

3.2. Interfacial analysis

Bottle tests conducted in the absence of porous mediarevealed that a white matter accumulated at the interfacebetween diesel and water for both diesel types. No interfacialmatter was observed for control samples (i.e., diesel not incontact with water) at the same temperature. Samples werefiltered prior to each test and no interfacial matter was initiallypresent. Moreover, results revealed that the field diesel sampledeveloped significantly more interfacial matter than the freshdiesel and the matter appeared sooner. The first appearanceof the interfacial matter was at 7.5 h for the field diesel,contrasted with 5 days for the fresh diesel. The interfacialmatter developed at the lowest point of the interface (visible asa bright white dot) but spread out over the interface with time(appearing as a white film and resulting in the diesel lookingcloudy). On the timescale of a Pc–S experiment (20 h to

Page 6: Journal of Contaminant Hydrology - Western Engineering · drainage were not representative of pressures in the porous medium. Once the diesel had entered the porous medium it was

Fig. 3. Field diesel GC/MS chromatogram illustrating major alkanes at regular intervals across the chromatogram with C17 notably absent, and pristine andphytane exhibiting significant peaks.

51S.S. Drake et al. / Journal of Contaminant Hydrology 144 (2013) 46–57

3 days), and specifically the primary drainage curve (7 h), it isexpected that only the field diesel Pc–S resultsmay be impactedby the development of this interfacial matter.

The formation of interfacial matter has also been observedin other studies. Field samples of black diesel and coal tarformed semi-rigid interfacial films impeding accurate contactangle measurements (Powers et al., 1996). Similar films havebeen observed for waste trichloroethylene (Harrold et al.,2001), crude oils and field coal tars (Nelson et al., 1996).Nelson et al. (1996) concluded that the binding of watermolecules to coal tar molecules created this film which actedas a very thin emulsion layer. They did not see any functionalgroups typical of surfactant compounds in GC/MS, NMR andIR spectra, however they note that very small concentrationswould not be detected. The white interfacial matter in thisstudy did not form a continuous, semi-rigid film as observedwith the heavier NAPL samples discussed above. Consider-able effort was expended attempting to identify the precip-itate. Isolating the matter from the diesel proved to bevery difficult. Neither FTIR nor SEM/EDX analyses provided adefinitive answer regarding composition.

The diesel, interfacial matter, and water from the bottletests were additionally streaked on LB-Agar plates to investi-gate the presence of microorganisms. For both diesels, therewas microbial growth for those samples taken from both thewater phase and the interfacial area but not the diesel. Thenumbers of colonies from thewater and interfacialmatter fromthe field diesel sample were substantially higher than thosefrom the fresh diesel sample, possibly indicating that thepresence and accumulation rate of the interfacial matter maybe related to microorganisms.

There are at least two possible other explanations for themore significant rate of accumulation of interfacial matterobserved in the field diesel. First, some difference in theoriginal composition of the two diesels may play a significantrole. Composition variations may have been due to differentintended uses of each diesel and/or the formula of additivesused. Hundreds of additives, which vary between supplier,may be used to improve performance and these often havesurface active behavior (Song et al., 2000). Analysis for thesecompounds is challenging as the target compound(s) must beknown in advance. Second, the final composition of the diesels

Page 7: Journal of Contaminant Hydrology - Western Engineering · drainage were not representative of pressures in the porous medium. Once the diesel had entered the porous medium it was

Table 1Physical properties of non-aqueous phases.

NAPL Density(g/mL)

IFT with water(25 °C) (mN/m)a

θAdvancing(°)a

θReceding(°)

Fresh diesel 0.808 28.1±1.5 10.7±0.3 b33b

Field diesel 0.825 17.2±1.0 29.5±2.2 180c

a Error reported as 95% confidence interval about the mean.b Visual obstruction made measurement difficult therefore the maximum

possible angle is reported here.c Pinning of contact angle taken as 180°.

52 S.S. Drake et al. / Journal of Contaminant Hydrology 144 (2013) 46–57

(i.e., weathering) may have been significant. As previouslydiscussed, GC–MS chromatograms indicated the field dieselwas significantly weathered. In such cases, the biodegradationand oxidation of hydrocarbons may produce more polarcompounds (Cuiec, 1990; Lang et al., 2009) reducing hydro-phobicity. Therefore, when the diesel was left in contact withwater these compounds could preferentially partition to thewater phase, changing the diesel phase chemistry and resultingin heavier, non-polar molecules that precipitated out. Inaddition, the weathering of the field diesel may have resultedin NAPL components binding with water molecules asobserved for Nelson et al. (1996).

3.3. Interfacial tension and contact angle

The fresh diesel exhibited a greater IFT than thefield diesel (Table 1), however both IFT values were relativelylow compared to an analogous, single component NAPL;for example, the IFT between pure o-xylene and water is39 mN/m (Lord et al., 2000). The measured IFT values areconsistent with diesel fuels commonly containing additivesfor improved performance. It is possible that the reducedIFT of the field diesel is associated with the white interfacialmatter observed in the jar tests. However, weatheredNAPLs do not necessarily exhibit IFT values less than that ofun-weathered samples. Harrold et al. (2003) found thatsurface active constituents present in NAPL sorbed to soil,

0

15

30

45

60

75

90

105

120

135

150

165

180

0 200 400Co

nta

ct a

ng

le (

thro

ug

h w

ater

ph

ase)

Elapsed contact time

Fig. 4. Advancing and receding contact angles plotted against total (cumulative) elwithout error bars represent average values with deviation between the measuremeentire range of possible values for a given contact time, where the data point serveline at 90° represents the distinction between water-wetting and NAPL-wetting reg

thereby decreasing their concentration in the NAPL phaseand increasing interfacial tension.

Intrinsic contact angles measured with the fresh dieselwere consistent with strongly water wet conditions (Table 1,Fig. 4). No significant hysteresis was observed betweenadvancing and receding contact angles and no sensitivity toelapsed contact time (up to 300 min.) was observed (Fig. 4).The maximum θint=33° (with mean θint=10.7±0.3) indi-cated that fresh diesel remained water-wetting even aftersignificant contact time with the solid and repeated reversalsof contact line direction. In contrast, the field diesel exhibitedconsiderable contact angle hysteresis that was time-dependent.The field diesel advancing θintwas consistent at 29.5°±2.2° over700 min. of total elapsed contact time (Table 1, Fig. 4). The fielddiesel receding contact angles, however, exhibited intermediatewetting behavior immediately (θint=146° at 0 min.) andincreasing hydrophobicity with elapsed contact time. From20 min to 110 min of elapsed contact time, observed recedingcontact angles were not consistent since erratic, “temporarypinning” led to NAPL drop recession that was not smooth andcontinuous. In Fig. 4, the range between the maximum andminimum observed receding contact angles are indicated by theerror bars. After 110 min of elapsed time, the contact line pinnedcompletely – i.e., the baseline of the drop remained static as thedrop volume was reduced – and the receding intrinsic contactangles for the field diesel approached 180°.

Harrold et al. (2001) also observed contact angles thatincreased with the time that the quartz plates were immersedin the NAPL phase prior to measurement. Immersion for 1 minwas sufficient to change an initially strongly water-wet plate toweakly NAPL-wet (Harrold et al., 2001). The impact of priorexposure in their study was more significant for waste NAPLs(i.e., from a solvent recovery company) than for pure NAPLswith surfactant additives. Molnar et al. (2011) also observedsimilar tetrachloroethylene/water contact angle hysteresis dueto sorption of a cationic surfactant onto an iron oxide-coatedplate at neutral pH. Lam et al. (2002) identifies a number offluid-fluid–solid systems that exhibited similar “stick/slip”behavior, attributed to chemical bonding between the receding

600 800

(minutes)

Fresh Advancing

Fresh Receding

Field Advancing

Field Receding

apsed time of contact between the NAPL drop and solid surface. Data pointsnts less than the size of the symbol. Data points with error bars represent thes as the median value between the maximum and minimum. The horizontalions of the plot.

Page 8: Journal of Contaminant Hydrology - Western Engineering · drainage were not representative of pressures in the porous medium. Once the diesel had entered the porous medium it was

- 10

0

10

20

30

40

50

60

70

80

0 0.2 0.4 0.6 0.8 1

Cap

illar

y P

ress

ure

(cm

wat

er)

Apparent Water Saturation

Drainage pseudo-static average (6 curves)

Imbibition pseudo-static average (6 curves)

Drainage van Genuchten fit

Imbibition van Genuchten fit

Fig. 5. Fresh diesel/water/F70 pseudo-static capillary pressure–saturationprimarydrainage andmain imbibition curves (each anaverage of 6 experiments).95% confidence intervals plotted for every fifth point (averaged over that pointand the 4 nearest neighbors). Also plotted are best-fit van Genuchten functions(parameters in Table 2).

-10

0

10

20

30

40

50

60

0 0.2 0.4 0.6 0.8 1

Cap

illar

y P

ress

ure

(cm

wat

er)

Apparent Water Saturation

Drainage pseudo-static average (3 curves)

Imbibition pseudo-static average (8 curves)

Drainage van Genuchten fit

Imbibition van Genuchten fit

Fig. 6. Field diesel/water/F70 pseudo-static capillary pressure–saturationprimary drainage and main imbibition curves (averages of 3 drainage and8 imbibition experiments). 95% confidence intervals plotted for every fifthpoint (averaged over that point and the 4 nearest neighbors). Also plottedare best-fit van Genuchten functions (parameters in Table 2).

53S.S. Drake et al. / Journal of Contaminant Hydrology 144 (2013) 46–57

drop and the surface. That work further quantified the timedependence of receding contact angles with liquid moleculepenetration of the surface and/or sorption and, when the liquidmolecules anchored themselves sufficiently firmly on the solid,the dropwould not recedewhatsoever (Lam et al., 2002). Thus,the time-dependent and pinning behavior observed here islikely due to surface active constituents present in the fielddiesel sorbing to the solid surface.

3.4. Capillary pressure–saturation main curves

The average of the six fresh diesel/water system primarydrainage and imbibition experiments, with 95% confidenceintervals, is presented in Fig. 5. The curves are consistent withstrongly water-wet conditions, with positive capillary pres-sures exhibited on both Pc–S pathways at almost all satura-tions. The imbibition curve reaches negative Pc values atwater saturations above 96.5% This is in agreement with themeasured advancing and receding intrinsic contact angles(Table 1, Fig. 4). van Genuchten fits reproduced the measuredmean behavior except for a slight discrepancy near residualand high water saturation (parameters in Table 2). A similarstudy on a diesel fuel obtained from a retail outlet also showed

Table 2van Genuchten best-fit parameters for LNAPL/water/F70 sand.

Experiment N [ ] α [cmwater]−1 η [cmwater]

Fresh diesel primary drainage 14.04 0.032 0Fresh diesel main imbibition 5.18 0.071 5.88Field diesel primary drainage 14.90 0.061 0Field diesel main imbibition 5.69 0.024 10.84

strongly water-wetting conditions through advancing contactangle quantification, qualitative bottle tests and USBM/Amott–Harvey indices calculated from Pc–S curves (Powers et al.,1996).

Fig. 6 presents the averaged field diesel Pc–S data, includingthree primary drainage and eight main imbibition curves. Thefield diesel primary drainage data was consistent with intrinsic(advancing) contact anglemeasurements that indicated stronglywater-wet behavior (Table 1, Fig. 4). Capillary pressures onimbibition, however, were negative above an apparent watersaturation of 5%, consistent with intermediate- or NAPL-wettingconditions. This indicates a reversal of the fluid phase thatpreferentially wetted the quartz surfaces once the field dieselcontacted the porous medium. This wettability reversal is alsoconsistentwith themeasured (receding) intrinsic contact angles(Table 1, Fig. 4).

Fig. 7 plots the primary Pc–S data for both diesels, withthat of the fresh diesel and its 95% confidence intervals scaledby the ratio of their interfacial tensions. The scaled freshdiesel data lies slightly above that of the field diesel, howeverthe 95% confidence intervals overlap suggesting that bothsystems are strongly water-wet and the operative contactangle is approximately 0°.

Fig. 8 illustrates the Leverett scaling of imbibition curvesfrom the reference fresh diesel drainage curve. An operativecontact angle of 71.7° for the fresh diesel imbibition curve wasdetermined using Eq. (1). Stronglywater-wet systems typicallyexhibit capillary pressures on imbibition that are 40–60% oftheir magnitude on drainage (Gerhard and Kueper, 2003;O'Carroll et al., 2005). In this case, cos(71.7°)=0.31 indicatingslightly greater hysteresis than typical, but suggestive thatrelatively strong water-wet conditions prevailed. Eq. (1)determined an operative contact angle of 107.7° for the field

Page 9: Journal of Contaminant Hydrology - Western Engineering · drainage were not representative of pressures in the porous medium. Once the diesel had entered the porous medium it was

-10

0

10

20

30

40

50

60

0 0.2 0.4 0.6 0.8 1

Cap

illar

y P

ress

ure

(cm

wat

er)

Effective Water Saturation

Fresh Diesel Drainage Scaled by IFT

Field Diesel Drainage

Fig. 7. Fresh diesel average pseudo-static drainage curve and error bars, bothscaled by the ratio of field diesel IFT to fresh diesel IFT, compared to the fielddiesel experimental average pseudo-static drainage curve.

-10

0

10

20

30

40

50

60

70

80

0 0.5 1

Cap

illar

y P

ress

ure

(cm

wat

er)

Effective Water Sautration

2

5

7

6

Fig. 9. Fresh diesel/water/F70 capillary pressure–saturation experimentwith scanning curves; 21 scanning curves over 9 unique saturation paths.Numbers near the curves represent the number of overlapping curves at thatlocation.

54 S.S. Drake et al. / Journal of Contaminant Hydrology 144 (2013) 46–57

diesel imbibition curve, indicating intermediate-wetting toNAPL-wetting conditions. The receding contact angle mea-sured on the quartz slide for the fresh diesel system (b33°)wasless than the operative contact angle scaled based on the Pc–Sexperiments (71.7°). This possibly indicates that the intrinsiccontact angle contact angle, measured on a quartz slide, maynot adequately quantify the degree of hysteresis present inporous media. In the field diesel system the intrinsic contactangle (~180°) was considerably larger than that found basedon scaling the Pc–S curves (107.7°). The reason for thisdiscrepancy could be that the residency time of the NAPL in

-10

0

10

20

30

40

50

60

70

80

0 0.2 0.4 0.6 0.8 1

Cap

illar

y P

ress

ure

(cm

wat

er)

Effective Wetting Phase Saturation

Fresh Diesel Drainage Experimental Curve

Fresh Diesel Imbibition Experimental Curve

Fresh Drainage Scaled to Fresh Imbibition

Field Diesel Imbibition Experimental Curve

Fresh Drainage Scaled to Field Imbibition

Fig. 8. Leverett–Cassie scaling for fresh and field diesel imbibition curvesThe fresh diesel main drainage curve is scaled to the imbibition curves by anoperative contact angle of 71.7° for fresh diesel imbibition and 107.7° forfield diesel imbibition.

-20

- 10

0

10

20

30

40

50

0 0.2 0.4 0.6 0.8 1

Cap

illar

y P

ress

ure

(cm

wat

er)

Effective Water Saturation

232

5

Fig. 10. Field diesel/water/F70 capillary pressure–saturation experimentwith scanning curves; 13 scanning curves over 11 unique saturation paths.Numbers near the curves represent the number of overlapping curves at thatlocation.

.

the Pc–S experiments may not have been adequately long torender the porous media extremely hydrophobic.

3.5. Capillary pressure–saturation scanning curves

Two representative experiments, one for each of the freshand field diesel (Figs. 9 and 10, respectively) that includedscanning curves demonstrate the reproducibility of the data.The number of repeat scanning curves that overlay each

Page 10: Journal of Contaminant Hydrology - Western Engineering · drainage were not representative of pressures in the porous medium. Once the diesel had entered the porous medium it was

55S.S. Drake et al. / Journal of Contaminant Hydrology 144 (2013) 46–57

other, the variety of saturation pathways explored, and notingthat each complete data set was measured in approximately89 h speaks to the strength of the pseudo-static method.

On primary drainage both system exhibited water wettingbehavior since they scale well based solely on interfacialtension. It is important to note that both porousmedia systemshave been flushed with numerous pore volumes of aqueousphase that was pre-equilibrated with either fresh or fielddiesel. As such, water/diesel interfacial tensions were notequivalent in both systems and soluble diesel constituentswould have been exposed to the surface of the quartz grains.Upon saturation reversal, either at residual water saturation orat some intermediate saturation, the shape of the imbibitioncurves were distinctly different (Figs. 9 and 10). Moreover, thefield diesel exhibited cyclical, repeatable wettability reversals(Fig. 10). Wettability reversal phenomena are discussed insignificant detail by Kovscek et al. (1993) for a crude oil system.Similar phenomena are likely occurring in the field dieselsystem however wettability reversal is not permanent as is thecase with crude oil. Permanent wettability reversal in crude oilis due to the presence of asphaltenes. On water drainage dieseldisplaces the aqueous phase as capillary pressure increases andthe thickness of the water film surrounding the porous mediadecreases. When the critical capillary pressure is achieved in apore the water film surrounding the quartz grains ruptures,leaving monolayer coverage of water molecules surroundingthe sand grains (Kovscek et al., 1993). At this point constituentsfrom the field diesel can partition to the solid surface renderingthe sand organic-wetting. As suggested by Kovscek et al.(1993), this wettability reversal occurs primarily in theintermediate pore size class as the critical capillary pressuremay not be achieved in the largest pores and no NAPL invadesthe smallest pores.Wettability reversalwas not observed in thefresh diesel system, consistent with measured contact angleand the main Pc–S curves.

With decreased capillary pressure little water imbibedinto the field diesel system initially as the intermediate poresize range was organic-wet and the largest pores requireda significant decrease in capillary pressure prior to waterreinvasion. As illustrated in Fig. 10, some water entered thelargest pores at positive capillary pressures but negativecapillary pressures are required to displace themajority of theNAPL. In the fresh diesel system on water imbibition, waterwould enter the intermediate sized pores followed by thelargest pores as there was no wettability reversal (Fig. 9). Forthis reason, the water imbibition curves in the fresh dieselsystem exhibited a reduced slope relative to the field dieselsystem (i.e., dPc/dSw for field diesel>dPc/dSw for fresh dieselat saturation reversal). In the field diesel system, however, theintermediate sized pores were organic wetting at the end ofdrainage. As such, it is postulated that, similar to waterdrainage in an initially water wet system, a critical (mini-mum, in this case) capillary pressure must be achieved torupture the NAPL film surrounding the quartz sand grains. Atthis point the NAPL molecules on the quartz sand grains aredisplaced by water and pores where this critical capillarypressure has been achieved in the field diesel system becomewater wetting again.

Upon subsequent saturation reversal (i.e., main drainagescanning curve) the water-wet pores in the field diesel onceagain followed the main drainage path consistent with a

water-wet system (Fig. 10). The shapes of the water drainagescanning curves between the fresh and field diesels arerevealing. In general, at the start of a drainage event, a highPc–S slope and the presence of an entry capillary pressure isevidence that the NAPL was nonwetting and discontinuous inthe largest pore bodies (Gerhard and Kueper, 2003). For thefresh diesel (Fig. 9), typical water-wetting behavior wasobserved in that the primary drainage curve exhibited adistinct entry pressure, but as scanning curve reversalpoints approach Sw=0.5, Pc–S slopes were observed toreduce (i.e., become shallower) (Haines, 1930; Morrow,1976; Parlange, 1976). This is expected since, at intermediatesaturations the NAPL was continuous and no entry pressureneeded to be overcome. However, in the field diesel system,all Pc–S drainage curves were equally steep, exhibiting adistinct entry pressure for any NAPL invasion regardlessof reversal saturation, further underscoring the cyclic wetta-bility reversal (Fig. 10).

Studies employing laboratory grade NAPLs and controlledconcentrations of surfactant additives have observed wettingphase reversal (Demond et al., 1994; Desai et al., 1992; Powersand Tamblin, 1995). In these studies, cationic surfactants sorbedto the quartz surface, reducing hydrophilicity and, at highsurfactant concentration, reversed wettability. Lenhard andOostrom (1998) discuss how a NAPL with a sorbing surfactantcan createmixedwettability –where the pores exposed toNAPLbecome NAPL-wetting and the pores continuously filled withwater remain water-wetting. These previous studies provideinsight but do not adequately explain the cyclical wettabilityreversal behavior observed in this study. The interfacial matterobserved in the field diesel after a short time, potentiallyassociated with diesel additives (Song et al., 2000), may play arole. In addition, the weathered composition of the diesel fuelmay play a role as may any compounds added to the diesel toinhibit degradation during storage. Typical antioxidants includealkylated phenols and secondary amines (Song et al., 2000),both of which may act as surfactants. The a priori knowledge ofspecific analytes combined with the extremely low detectionlimits required for these compounds made the identification ofany additives not practical in this study.

4. Conclusions

Examination of wettability at the pore scale, interface scaleand REV scale suggests that NAPLs recovered from a field sitemay not be adequately represented by similar fresh NAPLs dueto wettability differences. Increased contact time between afield diesel and quartz in the presence of water significantlyincreased the hydrophobicity of the solid, from intermediatewetting to NAPL wetting. Moreover, a field diesel exhibitedcyclical reversals of wettability, with water-wetting behavioron drainage and NAPL-wetting behavior on imbibition; suchbehavior at the REV-scale is linked to cyclical hysteresis of theintrinsic contact angle. The shape of the Pc–S scanning curves insuch cases reveals that the invading fluid must overcome theporous medium entry pressure at every saturation reversal,unlike an identical system that is strongly water-wettingirrespective of saturation history.

It is not surprising that a diesel fuel may exhibit NAPL-wetting behavior after contact with the soil, e.g., due to thepresence of a surfactant additive in the formulation or polar

Page 11: Journal of Contaminant Hydrology - Western Engineering · drainage were not representative of pressures in the porous medium. Once the diesel had entered the porous medium it was

56 S.S. Drake et al. / Journal of Contaminant Hydrology 144 (2013) 46–57

compounds produced via weathering. Such compounds, evenin small quantities, can precipitate at the interface, sorb tosolid surfaces, and alter the soil–liquid interfacial forces. It ishypothesized that the cyclical wettability reversal is relatedto cyclical changes in intermediate pore wettability due torupturing of the film surrounding the soil grain at a criticalcapillary pressure. It is possible that cyclical wettabilityreversals have not previously been observed because of thechallenge of conducting multiple, rapid Pc–S experimentswith the small volumes of NAPL typically available from fieldsites. Towards this end, the pseudo-static method for Pc–Scurves combined with the “elapsed contact time” method formeasuring cyclical intrinsic contact angles, may be a valuableapproach to determine the wettability of field soil/NAPLsystems. The wettability observed for the field system wouldneed to be incorporated into numerical models to properlypredict NAPL migration and remediation at the field site.

Remediation plans must be developed with knowledge offield NAPL wettability to achieve source zone remediation goals.Wettability changes will impact source zone architecture —

sub-regions of pools and residual may have significantlydifferent spatial distribution and NAPL/water interfacial areathan expected. Thus, at the field scale, thismayhave implicationson local mass flux and, in turn, the potential longevity of a NAPLrelease under natural or engineered conditions. Additionally,remediation techniques that involve waterflooding are based onunderstanding and manipulating Pc–S relationships; wettabilityimpacts on Pc–S imbibition curves must be understood in orderto adequately manipulate capillary pressure for NAPL displace-ment by water.

It is noted that the conclusions in this study are basedupon a single detailed comparison of two similar NAPLs. Thefresh sample examined in this study exhibited water-wettingbehavior however other retail formulations will vary andmay possess compounds that will create non-ideal wettabil-ity. In addition, only quartz was employed which neglects thepotential impacts on wettability of other minerals or organiccoatings. The aqueous phase chemistry was simplified byusing deionized water at neutral pH. However, this studydoes isolate and illustrate the impacts of a field NAPL onwettability and the significant differences that may existbetween a field and a fresh NAPL.

Acknowledgments

This research was supported by Natural Sciences andEngineering Research Council of Canada (NSERC) and OntarioGraduate scholarships to the first author. Additional supporthas been provided by an NSERC Strategic Grant and theCanadian Foundation for Innovation. The authors would liketo thank Dr. Kela Weber, at the Royal Military College, for hishelp with the microbial analysis and Drs. Tohren Kibbey andLixia Chen, at the University of Oklahoma, for their significantsupport related to the Pc–S apparatus.

References

Anderson, W.G., 1986. Wettability literature survey .2. Wettability measure-ment. Journal of Petroleum Technology 38 (12), 1246–1262.

Barranco, F.T., Dawson, H.E., 1999. Influence of aqueous pH on the interfacialproperties of coal tar. Environmental Science & Technology 33 (10),1598–1603.

Barranco, F.T., Dawson, H.E., Christener, J.M., Honeyman, B.D., 1997.Influence of aqueous pH and ionic strength on the wettability of quartzin the presence of dense non-aqueous-phase liquids. EnvironmentalScience & Technology 31 (3), 676–681.

Bradford, S.A., Leij, F.J., 1995a. Fractional wettability effects on 2-fluid and 3-fluid capillary pressure–saturation relations. Journal of ContaminantHydrology 20 (1–2), 89–109.

Bradford, S.A., Leij, F.J., 1995b. Wettability effects on scaling 2-fluid and 3-fluid capillary pressure–saturation relations. Environmental Science &Technology 29 (6), 1446–1455.

Bradford, S.A., Abriola, L.M., Rathfelder, K.M., 1998. Flow and entrapment ofdense nonaqueous phase liquids in physically and chemically heteroge-neous aquifer formations. Advances in Water Resources 22 (2), 117–132.

Camps-Roach, G., O'Carroll, D.M., Newson, T.A., Sakaki, T., Illangasekare, T.H.,2010. Experimental investigation of dynamic effects in capillary pressure:grain size dependency and upscaling. Water Resources Research 46.

Chen, L.X., Miller, G.A., Kibbey, T.C.G., 2007. Rapid pseudo-static mea-surement of hysteretic capillary pressure–saturation relationships inunconsolidated porous media. Geotechnical Testing Journal 30 (6),474–483.

Christensen, L.B., Larsen, T.H., 1993. Method for determining the age of dieseloil-spills in the soil. Ground Water Monitoring and Remediation 13 (4),142–149.

Council, N.R., 1999. Groundwater and Soil Cleanup: Improving Managementof Persistent Contaminants. National Academy Press, Washington, D.C.

Craig, F.F., 1971. The Reservoir Engineering Aspects of Waterflooding.Monograph Series, 3. Society of Petroleum Engineers, Richardson, TX.

Cuiec, L.E., 1990. Evaluation of Reservoir Wettability and its Effect on OilRecovery. In: Morrow, N.R. (Ed.), Interfacial Phenomena in PetroleumRecovery. Surfactant Science Series. Marcel Dekker, Inc., New York, N.Y.

Demond, A.H., Roberts, P.V., 1991. Effect of interfacial forces on 2-phasecapillary pressure–saturation relationships. Water Resources Research27 (3), 423–437.

Demond, A.H., Desai, F.N., Hayes, K.F., 1994. Effect of cationic surfactants onorganic liquid water capillary-pressure saturation relationships. WaterResources Research 30 (2), 333–342.

Desai, F.N., Demond, A.H., Hayes, K.F., 1992. Influence of surfactant sorptionon capillary-pressure saturation relationships. ACS Symposium Series491, 133–148.

Dong, J.F., Chowdhry, B., Leharne, S., 2004. Investigation of the wettingbehavior of coal tar in three phase systems and its modification bypoloxamine block copolymeric surfactants. Environmental Science &Technology 38 (2), 594–602.

Drake, S.S., 2010. Wettability Characterization for Diesel Fuels. WesternUniversity, London.

Dwarakanath, V., Jackson, R.E., Pope, G.A., 2002. Influence of wettability on therecovery of NAPLs from alluvium. Environmental Science & Technology 36(2), 227–231.

Galperin, Y., Kaplan, I.R., 2008. Forensic environmental geochemistry indispute resolution-case history 2: differentiating sources of diesel fuel ina plume at a fueling station. Environmental Forensics 9 (1), 55–62.

Gerhard, J.I., Kueper, B.H., 2003. Capillary pressure characteristics necessaryfor simulating DNAPL infiltration, redistribution, and immobilization insaturated porous media. Water Resources Research 39 (8).

Haines, W.B., 1930. Studies in the physical properties of soil. V. Thehysteresis effect in capillary properties, and the modes of moisturedistribution associated therewith. The Journal of Agricultural Science 20(1), 97–116.

Harrold, G., Gooddy, D.C., Lerner, D.N., Leharne, S.A., 2001. Wettabilitychanges in trichloroethylene-contaminated sandstone. EnvironmentalScience & Technology 35 (7), 1504–1510.

Harrold, G., Gooddy, D.C., Reid, S., Lerner, D.N., Leharne, S.A., 2003. Changesin interfacial tension of chlorinated solvents following flow through UKsoils and shallow aquifer material. Environmental Science & Technology37 (9), 1919–1925.

Harrold, G., Lerner, D.N., Leharne, S.A., 2005. The impact of additives found inindustrial formulations of TCE on the wettability of sandstone. Journal ofContaminant Hydrology 80 (1–2), 1–17.

Hsu, H.L., Demond, A.H., 2007. Influence of organic acid and organic baseinteractions on interfacial properties in NAPL-water systems. Environ-mental Science & Technology 41 (3), 897–902.

Hugaboom, D.A., Powers, S.E., 2002. Recovery of coal tar and creosote fromporous media: the influence of wettability. Ground Water Monitoringand Remediation 22 (4), 83–90.

Hurst, R.W., Schmidt, G.W., 2005. A forensic geochemical technique for estimatingrelease dates of petroleum products. Geochimica Et Cosmochimica Acta 69(10), A206.

Jackson, R.E., Dwarakanath, V., 1999. Chlorinated degreasing solvents:physical–chemical properties affecting aquifer contamination and reme-diation. Ground Water Monitoring and Remediation 19 (4), 102–110.

Page 12: Journal of Contaminant Hydrology - Western Engineering · drainage were not representative of pressures in the porous medium. Once the diesel had entered the porous medium it was

57S.S. Drake et al. / Journal of Contaminant Hydrology 144 (2013) 46–57

Kovscek, A.R., Wong, H., Radke, C.J., 1993. A pore-level scenario for thedevelopment of mixed wettability in oil-reservoirs. AICHE Journal 39(6), 1072–1085.

Lam, C.N.C., Wu, R., Li, D., Hair, M.L., Neumann, A.W., 2002. Study of theadvancing and receding contact angles: liquid sorption as a cause ofcontact angle hysteresis. Advances in Colloid and Interface Science 96(1–3), 169–191.

Lang, D.A., et al., 2009. Polar compounds from the dissolution of weathereddiesel. Ground Water Monitoring and Remediation 29 (4), 85–93.

Lenhard, R.J., Oostrom, M., 1998. A parametric model for predicting relativepermeability-saturationcapillary pressure relationships of oil-watersystems in porous media with mixed wettability. Transport in PorousMedia 31 (1), 109–131.

Lord, D.L., Demond, A.H., Salehzadeh, A., Hayes, K.F., 1997a. Influence oforganic acid solution chemistry on subsurface transport properties .2.Capillary pressure-saturation. Environmental Science & Technology 31(7), 2052–2058.

Lord, D.L., Hayes, K.F., Demond, A.H., Salehzadeh, A., 1997b. Influence oforganic acid solution chemistry on subsurface transport properties .1.Surface and interfacial tension. Environmental Science & Technology 31(7), 2045–2051.

Lord, D.L., Demond, A.H., Hayes, K.F., 2000. Effects of organic basechemistry on interfacial tension, wettability, and capillary pressurein multiphase subsurface waste systems. Transport in Porous Media38 (1–2), 79–92.

Molnar, I.L., O'Carroll, D.M., Gerhard, J.I., 2011. Impact of surfactant-inducedwettability alterations on DNAPL invasion in quartz and iron oxide-coatedsand systems. Journal of Contaminant Hydrology 119 (1–4), 1–12.

Morrow, N.R., 1976. Capillary-pressure correlations for uniformly wettedporous-media. Journal of Canadian Petroleum Technology 15 (4), 49–69.

Nelson, E.C., Ghoshal, S., Edwards, J.C., Marsh, G.X., Luthy, R.G., 1996.Chemical characterization of coal tar-water interfacial films. Environ-mental Science & Technology 30 (3), 1014–1022.

O'Carroll, D.M., Abriola, L.M., Polityka, C.A., Bradford, S.A., Demond, A.H., 2005.Prediction of two-phase capillary pressure–saturation relationships infractional wettability systems. Journal of Contaminant Hydrology 77 (4),247–270.

Parlange, J.Y., 1976. Capillary hysteresis and relationship between dryingand wetting curves. Water Resources Research 12 (2), 224–228.

Powers, S.E., Tamblin, M.E., 1995. Wettability of porous-media after exposureto synthetic gasolines. Journal of Contaminant Hydrology 19 (2), 105–125.

Powers, S.E., Anckner, W.H., Seacord, T.F., 1996. Wettability of NAPL-contaminated sands. Journal of Environmental Engineering-Asce 122 (10),889–896.

Ren, S.L., et al., 2009. Effect of weathering on surface characteristics of solidsand bitumen from oil sands. Energy & Fuels 23 (1), 334–341.

Ryder, J.L., Demond, A.H., 2008. Wettability hysteresis and its implicationsfor DNAPL source zone distribution. Journal of Contaminant Hydrology102 (1–2), 39–48.

Salehzadeh, A., Demond, A.H., 1999. Pressure cell for measuring capillarypressure relationships of contaminated sands. Journal of EnvironmentalEngineering-Asce 125 (4), 385–388.

Song, C., Hsu, C.S., Mochida, I., 2000. Chemistry of Diesel Fuels. AppliedEnergy Technology Series. Taylor & Francis, New York.

Spildo, K., Blokhus, A.M., Andersson, A., 2001. Surface and interfacialproperties of octanoic acid-octylamine mixtures in isooctane-watersystems: influence of acid: amine molar ratio and aqueous phase pH.Journal of Colloid and Interface Science 243 (2), 483–490.

Standal, S.H., Blokhus, A.M., Haavik, J., Skauge, A., Barth, T., 1999. Partitioncoefficients and interfacial activity for polar components in oil/watermodel systems. Journal of Colloid and Interface Science 212 (1), 33–41.

van Genuchten, M.T., 1980. Closed-form equation for predicting thehydraulic conductivity of unsaturated soils. Soil Science Society of AmericaJournal 44, 892–898.

Zheng, J.Z., Powers, S.E., 1999. Organic bases in NAPLs and their impact onwettability. Journal of Contaminant Hydrology 39 (1–2), 161–181.

Zheng, J.Z., Powers, S.E., 2003. Identifying the effect of polar constituentsin coal-derived NAPLs on interfacial tension. Environmental Science &Technology 37 (14), 3090–3094.

Zheng, J.Z., Behrens, S.H., Borkovec, M., Powers, S.E., 2001. Predicting thewettability of quartz surfaces exposed to dense nonaqueous phaseliquids. Environmental Science & Technology 35 (11), 2207–2213.