You are on page 1of 30

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/266330104

Colloid-Facilitated Contaminant Transport in


Unsaturated Porous Media

Article · January 2010

CITATIONS READS

22 176

2 authors, including:

Arash Massoudieh
The Catholic University of America
102 PUBLICATIONS   1,124 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

A flexible tool for hydraulic and water quality performance assessment of stormwater green infrastructure View project

Bacterial Fate and Transport View project

All content following this page was uploaded by Arash Massoudieh on 11 October 2015.

The user has requested enhancement of the downloaded file.


CHAPTER 8
Colloid-Facilitated Contaminant
Transport in Unsaturated Porous Media

Arash Massoudieh and Timothy R. Ginn

8.1 INTRODUCTION/EVIDENCE

Classical models of contaminant transport in porous media often considered contami-


nants occurring in the two major phases: dissolved, or bound to immobile matrix. It
has been shown in the past two decades that colloidal particles can also play a
significant role in the transport of contaminants in the natural subsurface (McCarthy
and Zachara, 1989), and therefore in order to predict the transport of contaminants in
porous media realistically a third phase (the mobile colloidal phase) should also be
considered.
Colloids are defined here as particles for which Brownian forces matter: they
generally include sizes ranging from 1 nm to 10 m in characteristic diameter. There
are various types of colloid in the groundwater, including colloids produced as a result
of mineral precipitation involving iron, aluminium, calcium and silica precipitates, as
well as colloids produced as a result of rock or sediment fragmentation, organic
colloids, and biocolloids such as bacterial cells and cell components, viruses and
protozoa.
It has been shown that colloids are ubiquitous in groundwater and in the vadose
zone (McCarthy and Degueldre, 1993), and because of their large surface area colloids
can act as agents for enhancing the migration of highly sorbing pollutants in porous
media. In such media colloids can be mobilised as a result of chemical perturbations
that decrease the net attractive force between the solid mineral phases and the
colloidal particles, including a decrease in ionic strength, or an increase in ions that
can adsorb to the mineral surfaces and change their surface charge (Ryan and
Elimelech, 1996). Also, it has been shown that hydrodynamic perturbation – such as
rapid infiltration, episodic wetting and drying, and large increases in shear stress
caused by changes in pore water velocity – can also cause detachment of colloidal
particles from the solid minerals (Saiers and Lenhart, 2003a; Zhuang et al., 2007).

Modelling of Pollutants in Complex Environmental Systems, Volume II, edited by Grady Hanrahan. # 2010 ILM
Publications, a trading division of International Labmate Limited.
260 MODELLING OF POLLUTANTS IN COMPLEX ENVIRONMENTAL SYSTEMS

Ryan and Elimelech (1996) suggested that three criteria should be present in
order for the effect of colloidal particles on the transport of contaminants to be
important:
• colloidal particles should be present in large enough concentrations;
• the contaminant should bind to the colloidal particles; and
• the colloidal particles should be able to move in the porous medium faster than
the contaminant would move in the absence of colloids.
There is strong evidence for the potential role of colloids in accelerating the
movement of metal and transuranic contaminants in the natural subsurface. Nyhan et
al. (1985) observed the migration of plutonium and americium to distances several
orders of magnitude larger than what is expected from their equilibrium partitioning
to the immobile solid phase as quantified by the sorption distribution coefficient.
However, they did not attribute this phenomenon explicitly to the role of colloidal
particles. Short et al. (1988) observed a significant association of uranium (U) and
thorium (Th) with colloidal particles consisting mainly of Fe and Si species, with sizes
ranging from 18 nm to 1 m. Buddemeier and Hunt (1988) found almost 100% of
transition elements (Mn, Co) and lanthanide radionuclides (Ce, Eu) associated with
colloidal particles in a site in Nevada. Ryan and Elimelech (1996) pointed out the
inadequacy of two-phase models for highly sorbing contaminants in the presence of
colloids. Penrose et al. (1990) also observed americium and plutonium moving to
unexpected distances, considering the apparent partitioning coefficients of these
metals, and suggested that this phenomenon is due to irreversible association of these
metals with mobile colloids. However, Marty et al. (1997) did not support the idea that
colloid-facilitated transport played a role in the transport of radionuclides in this site,
and suggested that the ratios of 239 Pu/238 Pu, and the fact that the radionuclides had
been transported faster than the average groundwater flow, showed that they had been
transported through surface water.
Colloid-facilitated transport is more frequently identified as a vehicle, especially
with heavier metals and radionuclides. Most of the plutonium occurring in ground-
water samples from the Nevada test site was associated with colloidal particles
(Reimus et al., 2005), and the concentration of plutonium in those samples was highly
correlated with the number density of colloids. McCarthy et al. (1998) identified
organic matter colloidal fractions as significant facilitators of transuranic radionuclide
transport in the karstic aquifer system at Oak Ridge National Laboratory. Novikov et
al. (2006) noted for the Russian Mayak site that plutonium transport rates of ,1 km/
14 yrs were facilitated by iron hydroxide colloids, which probably arose from the
activity of iron-reducing bacteria. Pang and Simunek (2006) highlighted the role of
bacterial cells themselves in the conveyance of cadmium, and radionuclides can also
occur as biogenic mineral phases associated with cells (e.g., Suzuki et al., 2002).
Acid-mine drainage provides another context for the formation of colloid-
associated contaminant phases involving metals. These solutions, resulting from the
oxidation and dissolution of minerals brought to the surface, undergo slow neutralisa-
tion via mixing with ambient waters, resulting in reprecipitation of iron and other
metal phases as colloids (e.g., Accornero et al., 2005; Hochella et al., 2005), which
COLLOID-FACILITATED CONTAMINANT TRANSPORT IN UNSATURATED POROUS MEDIA 261

can impact on hydrologically linked subsurface waters through flood events (e.g.,
Kretzschmar and Schafer, 2005).
The focus of this chapter is to review various modelling approaches used to
quantify colloid-facilitated transport in the porous media. Section 8.2 focuses on
modelling the processes affecting the movement of colloids, and Section 8.3 on
modelling approaches used in contaminant transport in the presence of colloids. The
focus will be mainly on unsaturated porous media, although models constructed solely
for saturated groundwater are also discussed, and their evolution into models applic-
able to unsaturated media is described. Section 8.4 discusses the methods used to
incorporate the effects of geochemical and physical heterogeneities in the models and
Section 8.5 discusses the unknowns and future directions necessary to build more
realistic contaminant transport models in the presence of mobile colloids.

8.2 COLLOID TRANSPORT IN UNSATURATED MEDIA


The basis for any colloid-facilitated contaminant transport model is a colloid transport
model, so it is important to first review different approaches used to model colloid
transport in saturated and unsaturated media before moving on to the topic of
modelling colloid-facilitated transport. In unsaturated media (Figure 8.1), owing to the
high variability in hydraulic shear stress, ionic strength, pH and moisture content as a
result of wetting and drying cycles, colloid transport is more complicated than in
saturated media. Also, processes such as film straining and entrapment in the air/water
interface may influence the movement of particles in unsaturated media. These
processes are non-existent in the case of colloid transport in saturated porous media.
Although different models have been proposed for the deposition and release of
colloids in porous media, there is no consensus on a single modelling framework
applicable to all circumstances. For further details of modelling approaches to colloid
transport the interested reader is referred to the paper by Ryan and Elimelech (1996),
who extensively reviewed the various processes involved in colloid transport in porous
media, and to that by DeNovio et al. (2004) for unsaturated media.

Figure 8.1: Colloid retardation mechanisms in unsaturated porous media.


de
ch
tta

ile
A

ob
M

ir
A
I
AW
to
ed

ed
d
ch

ne

in
tta

ai

ra
A

str

St
lm
Fi
262 MODELLING OF POLLUTANTS IN COMPLEX ENVIRONMENTAL SYSTEMS

Typically, the colloid phase is treated as a suspension, and an advection–


dispersion equation is considered to model colloid transport in porous media:
 
@ ðŁGm Þ @Gs @Ga @Łvc Gm @ @Gm
þ rB þˆ þ ¼ Dc Ł (8:1)
@t @t @t @z @z @z

where Gm [M/L3 ] is the mass concentration of mobile colloidal particles; Gs [M/M] is


the immobilised (attached, filtered, trapped) colloid concentration; Ga [M/L2 ] is the
concentration of colloids trapped in the air/water interface; ˆ is the surface area of the
air/water interface; rB [M/L3 ] is the dry bulk density of the soil matrix; Ł is the water
content equal to the porosity in a saturated medium; s [Mc/M] is the sorbed
contaminant concentration; vc [L/T] is the advective velocity for colloidal particles;
and Dc [L2 /T] is the dispersion coefficient for colloidal particles. In cases where the
film straining is considered as a separate mechanism from entrapment into the air/
water interface an additional term representing it will be added to the equation:
 
@ ðŁGm Þ @Gs @Ga @G @Łvc Gm @ @Gm
þ rB þˆ þ þ ¼ Dc Ł (8:2)
@t @t @t @t @z @z @z

where G is the concentration of colloids entrapped by film straining, expressed as the


mass of colloids per volume of bulk medium. Most of the colloid-facilitated transport
and colloid transport models have ignored the generation of colloids by erosion of
colloidal materials from the immobile solid phase Massoudieh and Ginn (2007), after
Sen et al. (2004), assumed that colloids irreversibly attach to the collectors, and that
the source of colloids being released is different from the colloids filtered previously.
They therefore divided the attached colloid population (Gs ) into two portions,
irreversibly attached colloids (Gsi ) and immobile colloids available for release (Gsf ):
 
@ ðŁGm Þ @ ð Gsi þ Gsf Þ @Ga @G @Łvc Gm @ @Gm
þ rB þˆ þ þ ¼ Dc Ł (8:3)
@t @t @t @t @z @z @z

Although this approach considers a separate source for colloids being released to the
medium, it still considers a finite quantity of immobile colloids available to be
released in the overall system. Thus it does not explicitly assume colloid generation in
the system; however, by assuming a large value for the initial Gsf compared with the
detachment rate, generation of colloids from the solid matrix can be simulated. To
consider the rate of generation of immobile colloids explicitly, an additional term
representing the generation process (e.g., erosion) can be used:
 
@ ðŁGm Þ @Gs @Ga @G @Łvc Gm @ @Gm
þ rB þˆ þ þ ¼ Dc Ł þ pG rB (8:4)
@t @t @t @t @z @z @z

where pG represents the specific rate of colloid generation from the soil matrix. Since
the source of colloid generation is the solid matrix, to preserve the mass we need also
to have
@rB
¼  pG rB (8:5)
@t
COLLOID-FACILITATED CONTAMINANT TRANSPORT IN UNSATURATED POROUS MEDIA 263

8.2.1 Attachment mechanisms


Quantifying the filtration rate of colloidal particles in a natural porous medium is a
challenging scientific problem involving the physical and chemical characteristics of
the mineral surfaces, the colloids and the pore water. Colloid filtration theory (CFT)
has traditionally been used to model the attachment rate of colloidal particles. The
basic assumption underlying CFT is that the rate of deposition of colloidal particles
onto a porous medium is proportional to the frequency at which the mobile colloids
come into contact with the grain surface, known as the collision efficiency () and the
fraction of collisions that leads to attachment, known as the sticking efficiency (Æ)
(Rajagopalan and Tien, 1976), as in
@Gs
rb ¼ k att ŁGm (8:6)
@t
where
3 ð1  ÞÆvc
k att ¼ (8:7)
4 ac
where  is the porosity of the medium, and the collision efficiency  is defined by
considering an idealised isolated spherical collector covered by a spherical liquid
shell, referred to as the Happel sphere-in-cell model. The collision efficiency is
defined as the fraction of particles entering the sphere-in-cell model that collide with
the sphere surface, and is given by the deposition rate normalised by the total particle
influx rate:
I
¼ (8:8)
a2c vc Gm

where ac is the radius of the collector, and I is the deposition rate on the collector. Yao
et al. (1971) suggested a relationship for the collision efficiency by adding the
collision rates as a result of diffusion, sedimentation and interception onto an isolated
sphere. The deposition rate as a result of each process was calculated analytically
when ignoring the other processes. Expanding on this approach, Rajagopalan and Tien
(1976) used a deterministic numerical trajectory analysis on a wide variety of
conditions, and then performed a regression analysis on their results and obtained a
closed-form approximation for the value of  as a function of a set of non-dimensional
variables defining the geometrical and physical properties of the colloids and the
grain. Tufenkji and Elimelech (2004) included the effects of hydrodynamic inter-
actions and wan der Waals forces, and solved the convection–diffusion equation
directly to obtain a revised relationship for the collision efficiency. Both Rajagopalan
and Tien (1976) and Tufenkji and Elimelech (2004) assumed additivity of the various
processes causing the deposition of particles, including gravity, Brownian diffusion,
and hydrodynamic and wan der Waals forces. Nelson and Ginn (2005) offered a new
equation using stochastic direct particle tracking, and concluded that the effects of
Brownian diffusion and advection cannot be directly added together. Of course,
because of the idealisation of the porous medium by the Happel sphere-in- cell model,
264 MODELLING OF POLLUTANTS IN COMPLEX ENVIRONMENTAL SYSTEMS

it should not be expected that these equations will have a high predictive capability,
especially for porous media with non-uniform particle size (e.g., Ryan and Elimelech,
1996). In a heterogeneous medium where the matrix grains are not monodisperse,
Equation 8.7 can be written in terms of the specific surface area f, defined as the
surface area over the bulk volume of the soil:
k att ¼ 14 f Ævc (8:9)

Deposition of colloids on the grains can either increase (ripening) or decrease


(blocking: e.g., Redman et al., 2001; Tufenkji et al., 2003), the deposition of incoming
particles depending on the presence of attractive or repulsive forces between the
particles. Ripening is reported by Kretzschmar et al. (1995), who observed an increase
in filter efficiency for colloid transport in intact cores of saprolite as increasing
amounts of colloids were deposited. More recently Tong et al. (2008) found ripening
triggered by relatively subtle changes in solution chemistry and fluid velocity. Tong et
al. (2008) point out that the propensity to trigger ripening increased with the number
and length of grain-to-grain contacts.
To incorporate the blocking effect a correction factor, referred to as the blocking
function, is multiplied by the attachment rate coefficient (e.g., Davros and van de Ven,
1982):
k att ¼ 14 f Ævc Bð Gs Þ (8:10)

The isotherm has been used to model the effect of blocking (Kallay et al., 1987). The
mass-based blocking function can be expressed as
Gs
Bð G s Þ ¼ 1  (8:11)
Gs,max

where Gs,max is the maximum mass concentration of colloids that can be attached to a
unit bulk mass of the medium. This approach does not take into account the areal
exclusion effect of particles with finite sizes on the grain surfaces. Schaaf and Talbot
(1989) analytically solved the random sequential adsorption model (RSA) using the
statistical mechanics approach and found a third-order equation for the blocking
function expressed in terms of the mass concentration of attached colloidal particles:
   
Gs Gs 2 Gs 3
Bð Gs Þ ¼ 1  4 þ 3:31 þ 1:407 (8:12)
Gs,ch Gs,ch Gs,ch

where Gs,ch is the characteristic attached concentration, defined as


Gs,ch ¼ 0:728 fap rp (8:13)

where ap is the radius of colloids and rp is the mass density of colloids. Johnson and
Elimelech (1995) compared the Langmuir- and RSA-based dynamic blocking func-
tions, and concluded that the RSA model does a better job in fitting observed
breakthrough curves. Song and Elimelech (1993) offered a double-layer particle
deposition model, and also considered the non-uniform distribution of deposition over
COLLOID-FACILITATED CONTAMINANT TRANSPORT IN UNSATURATED POROUS MEDIA 265

a spherical collector. Song and Elimelech (1994) used a mono-layer model, but
included the effect of heterogeneity of the forces between colloidal particles and
different patches of the porous medium by expressing the distribution of the surface
potentials of surface sites as Gaussian.
CFT is applicable only under favourable conditions for attachment of colloids
when the surface charges of the colloidal particles and the grains are opposite. Under
unfavourable conditions, in theory, CFT predicts no attachment. So other mechanisms
have been suggested to cause filtration under unfavourable conditions. These mechan-
isms include straining, wedging and entrapment of colloids in flow-stagnation regions.
Bradford et al. (2003) considered both attachment controlled by CFT and
straining, assuming that attachment controlled by CFT is reversible whereas straining
is irreversible. They suggested a straining rate coefficient as a function of distance
from the inlet of the porous medium:
 
@G d 50 þ z
¼ Łk str Gm (8:14)
@t d 50

where kstr [T 1 ] is the straining rate coefficient, and d50 is the medium soil particle
size. The problem with this formulation is that in field-scale modelling it is not always
clear what the basis for z should be. Bradford and Bettahar (2006) also took into
account saturation of sites by strained colloids by a Langmuirian-type blocking
function:
  
@G G d 50 þ z
¼ Łk str 1  Gm (8:15)
@t G,max d 50

where G,max is the maximum solid phase concentration of strained colloids. The
reason for the dependence of the straining rate on the travel distance can also be
attributed to the filtration of larger colloids at regions closer, to the inlet and the
reduced susceptibility of the remaining colloid to be trapped in small pores. However,
to model this mechanism, a multi-disperse colloid transport model should be used. Xu
et al. (2006) suggested the following equation for the straining rate, which is an
inverse a function of the strained concentration:
@G
¼ k 0 C Gm =º (8:16)
@t
where k0 and º are empirical coefficients. This relationship does not predict a
declining attachment rate with distance from the inlet, as opposed to the relationship
by Bradford et al. (2003).
In the representative elemental volume (REV) approach the velocity distribution
in the pores with various diameters is replaced by one average velocity. In the case of
straining, in addition to the attachment and detachment rates, which depend on the
flow velocity, the straining is also controlled by the pore diameter, which is homoge-
nised in an REV approach. Various researchers have suggested using a dual porosity
model to capture the effect of the distributions of pore size on colloid transport,
including:
266 MODELLING OF POLLUTANTS IN COMPLEX ENVIRONMENTAL SYSTEMS

• Corapcioglu and Wang (1999) and Bradford et al. (2009) for the general
modelling context;
• Woessner et al. (2001) for virus transport;
• Kim and Corapcioglu (2002) for organic and bacterial colloids;
• Schelde et al. (2002) for colloid release;
• Robinson et al. (2003) for radionuclide transport at the Yucca Mountain site;
• Dusek et al. (2006) for cadmium transport in soils; and
• Harter et al. (2008) for Cryptosporidium oocyst transport.

In spite of the progress made, there are still many challenges in understanding the
processes involved in colloid attachment in real porous media, mainly because of the
difficulties in capturing and upscaling the effect of irregular geometries of the pores,
chemical and physical heterogeneities of the matrix and colloids, and the effect of
water chemical properties on the attachment behaviour of colloids (Figure 8.2). The
relationships offered for film straining in saturated porous media have been used for
colloid transport models in unsaturated media without considering the effect of film
straining (Simunek et al., 2006). Wan and Tokunaga (1997) used a probabilistic
approach, and by assuming a power law relationship between the film-straining rate
and average linear velocity and the ratio between the diameters of particles and the
film thickness, suggested a relationship for the rate of straining as a function of matric
potential.

8.2.2 Detachment mechanisms


Detachment of colloids from the surfaces has not been studied as much as attachment.
Colloid filtration theory determines only rates of attachment of colloids to surfaces.
Early colloid-facilitated models ignored detachment altogether (Yao et al., 1971; Song
and Elimelech, 1993, 1994). In early colloid-facilitated transport models, typically the
detachment is modelled as a rate-limited process with constant rate coefficient
(Corapcioglu and Jiang, 1993; Jiang and Corapcioglu, 1993; followed by Saiers and

Figure 8.2: Various contaminant phases in unsaturated porous media in the presence of
colloids.

Mobile – available
Solid
matrix

Sorbed to
solid matrix Sorbed to
colloids
COLLOID-FACILITATED CONTAMINANT TRANSPORT IN UNSATURATED POROUS MEDIA 267

Hornberger, 1996; van de Weerd and Leijnse, 1997; van de Weerd et al., 1998;
Bradford et al., 2003):
@Gs
rb ¼ k att ŁGm  k det rb Gs (8:17)
@t

Ryan and Elimelech (1996) categorise the factors influencing the mobilisation of
colloids into physical and chemical perturbations. The physical causes of mobilisation
are associated with hydraulic transients and colloid–colloid collisions. The aqueous
chemical perturbations that can cause colloid mobilisation include changes in ionic
strength and pH. In addition, the colloid–surface bond perturbations that can lead to
mobilisation include bond ageing or weakening, and colloid–colloid bond interactions
on surfaces (Davros and van de Ven, 1982).
Several different hydraulic changes have been observed to be able to cause
colloid mobilisation, including an increase in shear stress (Sen et al., 2002; Shang et
al., 2008), wet and dry cycles (Saiers and Lenhart, 2003b), and movement of the air/
water interface (Saiers and Lenhart, 2003b) . Owing to the complexities in colloid
mobilisation, in most colloid-facilitated transport models either the colloids are
assumed to be irreversibly attached to the porous medium, or a constant detachment
rate coefficient is used. Govindaraju et al. (1995) and Sen et al. (2002) used the
threshold relationship originally suggested by Arulanandan et al. (1975) and Khilar et
al. (1985) for evaluating the erosion of fine particles by seepage flow in earth dams:
f
pG ¼ Æh H ð w   c Þ (8:18)
rs

where Æh is the release coefficient, H is the Heaviside function, w is the shear stress
exerted by the flow, and c is the critical shear stress. This relationship assumes that all
the colloidal particles have the same critical shear stress, and also that the shear stress
exerted by the flow is a uniform function of the velocity. In reality it is expected that
both the critical and the actual shear stress will be spatially variable. In that case
Equation 8.18 should be modified to
f
pG ¼ Æh ðw Þ (8:19)
rs

where  is a function that depends on the joint distribution of flow shear stress and the
critical shear stress, and w is a hypothetical quantity representing the average shear
stress. Schelde et al. (2002) hypothesised that the release of colloidal particles in
unsaturated media does not depend on the flow shear stress, but is mainly a diffusion-
limited process: they based this on the results of the column studies done by Jacobsen
et al. (1997). Based on this hypothesis they developed a dual-domain mobile/immobile
diffusive transport model for colloid transport with constant colloid exchange rates
between the domains. Jarvis et al. (1999) considered the effect of raindrop impact on
the mobilisation of colloids close to the surface. They considered a limited source of
colloids available that can replenish to reach to a maximum level as a result of
268 MODELLING OF POLLUTANTS IN COMPLEX ENVIRONMENTAL SYSTEMS

processes during the wet and dry periods, including freezing/thawing and wetting/
drying cycles.
Majdalani et al. (2007) proposed a detachment model in which the detachment
rate is a function of the cumulative detached mass at any location. They proposed the
following relationship for the detachment rate:
k det ¼ kdet Eð z, tÞŁv (8:20)

where kdet is the detachment rate constant, and E(z,t ) is a function that expresses the
effect of the past history of detachment at location z, and is related to the cumulative
detached concentration through the function
1 þ º1 X ð z, tÞ
Eð z, tÞ ¼ (8:21)
eº2 X ð z,tÞ
where º1 [M/L3 ] and º2 [M/L3 ] are model parameters, and X(z,t ) is the cumulative
detached concentration of particles from the surface at location z:
ðt
X ð z, tÞ ¼ pG ðÞrb d (8:22)
0

This model is based loosely on the consideration that detachment may modify surface
structure (pore space geometry) in ways that makes the frequency of detachment
depend on the cumulative mass removed from a particular location in the porous
medium. In subsequent work, Majdalani et al. (2008) showed that wetting–drying in
unsaturated porous media generates erosive-type release of colloids that increases with
‘‘time dried’’: they hypothesised that this is related to weakening of the soil matrix
tension during longer drying periods, and that this effect may be greater than those
caused by ionic strength or rainfall intensity (e.g., Tong et al., 2008).
Saiers and Lenhart (2003a, 2003b) found that the source of colloids available to
be detached depends on the moisture content and the flow velocity, so that at higher
flow rates more colloids become available for release. Based on this observation they
considered attached colloids to be in multiple compartments, so that release of
colloids from each compartment is dictated by the relationship between the moisture
content and a critical moisture content assigned to that compartment:
X
NC
Gs ¼ Gs,i (8:23)
i¼1

and
@Gs,i Ł 1
¼ k att Gm  k det ,i Gs,i (8:24)
@t rb NC

where NC is the number of compartments, and the release rate from each compart-
ment is controlled by the moisture content and the average linear pore water velocity
according to the following relationship:
COLLOID-FACILITATED CONTAMINANT TRANSPORT IN UNSATURATED POROUS MEDIA 269

(
0 for Ł , Łcr i
k det ,i ¼ kdet vk for Ł . Łcr i (8:25)

The DLVO force plays an important role in holding the colloidal particles on the
surfaces: thus any change in this force caused by factors such as a change in the
chemical composition of the pore water can cause the release of colloidal particles.
This role has been observed in several studies (Ryan and Gschwend, 1994; Roy and
Dzombak, 1996, 1997); however, owing to the complicated nature of the dependence
of colloid mobilisation on the chemical properties, it is less thoroughly quantified.
Saiers and Lenhart (2003b) addressed this within their multiple compartments model
by specifying that each compartment have a different threshold for release of colloids
as a function of ionic strength (Equation 8.23) with
 
@Gs,i Ł 1 Gs,i
¼ k att  Gm  k det ,i Gs,i (8:26)
@t rb NC Gs max

where katt and Gs max depend on the ionic strength and the detachment rate coefficient
from compartment i, and kdet, i is controlled by the relationship between the ionic
strength IS and the critical ionic strength IScr assigned to compartment i:
(
0 for IS . IScr i
k det ,i ¼ (8:27)
kdet eð NCªGs,i =Gs max Þ for IS , IScr i

More general conceptual models build from the way that DLVO theory specifies
the interaction forces between colloids and surfaces under variably favourable condi-
tions. Smets et al. (1999) found that the DLVO secondary energy minimum plays a
role in the detachment frequency of Pseudomonas fluorescens cells. In particular, they
found that DLVO-predicted energy barriers underestimated cell collision efficiencies,
suggesting that secondary energy minimum interactions governed the initial attach-
ment of cells. The partial reversibility of adhesion upon ionic strength reduction
supported the secondary minimum interaction hypothesis. Redman et al. (2004)
further this hypothesis, and report a sizeable increase in attachment with fluid ionic
strength despite DLVO calculated repulsive electrostatic interactions, and hypothesise
that deposition is likely occurring in the secondary energy minimum, which DLVO
calculations indicate increases in depth with ionic strength.
The general application of DLVO theory to colloid and/or microbial transport in
porous media is critically reviewed in Grasso et al. (2002), in Bostrom et al. (2001), in
Hermansson (1999) with particular reference to microbial colloids and their adhesion
to surfaces, and in Strevett and Chen (2003), who introduced extended DLVO theory
to account for hydration forces.
More general models of the way in which detachment frequency depends on
residence time while attached have a basis in the works of van de Ven and co-workers
(e.g., Dabros and van de Ven, 1982; Adamczyk et al., 1983). Dabros and van de Ven
(1982) specify a convolution form for detachment rate as follows:
270 MODELLING OF POLLUTANTS IN COMPLEX ENVIRONMENTAL SYSTEMS

J net ð tÞ ¼ J att ð tÞ  J det ð tÞ (8:28)

where Jnet (t ) is the net rate of transfer from aqueous to attached phases, Jatt (t ) is the
rate of attachment, and Jdet (t ) is the rate of detachment, which involves a dependence
on residence time while attached:
ðt
J det ð tÞ ¼ ð t  ÞJ net ðÞd (8:29)
0

where (t ) is the desorption rate coefficient per residence time sorbed, t. Dabros and
van de Ven (1982) suggested that (t ) is a decreasing function of t, so that it represents
bond strengthening, and wrote a model for it that drops exponentially from an initial
value to a lower asymptotic value. Meinders et al. (1992) recognised this generality
and put it to use in the analysis of data from a microflow (parallel-plate) chamber,
hypothesising the same form for (t ). This approach was subsequently used in
Johnson et al. (1995), who used a step-function form for (t ). Ginn (2000) generalised
this approach further to make attachment/detachment a function of residence time in
any phase with different memories of residence time effects. Talbot (1996) also
provided a general mathematical framework for residence-time-dependent desorption.

8.2.3 Air/water interface


Unsaturated conditions can affect the movement of colloids in several ways, including
the formation of discontinuous capillary fringes, prevention of the movement of
colloids, attachment to the air/water interface (AWI), forced deposition onto the
surfaces owing to the presence of thin films of water (referred to as film straining), and
the effect of wetting and drying on colloid mobilisation (Powelson and Gerba, 1994;
Chu et al., 2001; McCarthy and McKay, 2004). Several researchers have found a
higher rate of colloid removal in unsaturated conditions than in saturated conditions
(Powelson and Gerba, 1994; Jewett et al., 1999; Sirivithayapakorn and Keller, 2003;
Keller and Sirivithayapakorn, 2004; Gargiulo et al., 2008). Sirivithayapakorn and
Keller (2003) observed that the attachment of colloids to the AWI is irreversible as
long as the AWI exists; however, during the imbibitions and disappearance of the
AWI, colloids can become mobilised into the aqueous phase. Similar behaviour was
observed by Torkzaban et al. (2006) in their experiments on the transport of viruses
under variably saturated conditions. Wan and Wilson (1994) suggested that the
presence of the air phase can decrease the mobility of colloids owing to the
irreversible attachment of the colloid to the air/water interfaces of stationary air
phases. However, they also pointed out that attachment to the air/water interface can
increase mobilisation in the cases of moving interfaces during imbibitions, or bubbling
processes. Chu et al. (2001) attributed the larger retardation in unsaturated condition
mainly to higher attachment rates to the solid/water interface as a result of the
presence of the air phase (film straining) rather than attachment to the AWI. On the
other hand Lazouskaya and Jin (2008) observed a higher concentration of colloids
near AWI relative to the bulk solution for relatively hydrophilic colloids, suggesting
COLLOID-FACILITATED CONTAMINANT TRANSPORT IN UNSATURATED POROUS MEDIA 271

the favourability of AWI for colloid attachment. Corapcioglu and Choi (1996) and
Choi and Corapcioglu (1997) modelled the entrapment into the AWI as a reversible
process with linear kinetics using the volumetric concentration of colloids per air
volume. Sim and Chrysikopoulos (2000) expressed the concentration of colloids
entrapped in AWI as mass of colloids per area of the AWI, as calculated using the
relation suggested by Cary (1994):
@Ga
¼ Łk wa Gm  Łk aw ˆGa (8:30)
@t
and
 
2Łsb Łb
s Ł
b
Ł1b  Ł1b
ˆ¼ Łr þ s
(8:31)
r0 b 1b
where Łs and Łr are the saturated and residual water contents, respectively; r0 is the
effective pore radius; and and b are empirical parameters. Massoudieh and Ginn
(2007) used the equation suggested by Cary (1994); however, they assumed that the
colloid exchange between the AWI and the bulk aqueous phase is fast enough that an
equilibrium partitioning can be assumed:
Ga ¼ K aw Gm (8:32)

where Kaw is the distribution coefficient of colloids between the water and air phases.
Simunek et al. (2006) used a modified version of the kinetic mass exchange model
considering the availability of space on the AWI for colloid attachment:
@ˆGa
¼ Łk wa łaca f a Gm  k aw ˆGa (8:33)
@t
where łaca is a dimensionless colloid retention function, and fa (dimensionless) is the
fraction of the air/water area available for attachment. The AWI area relationship
suggested by Bradford and Leij (1997) is
ð
r g n
ˆ¼ w łðŁÞdŁ (8:34)
 aw Łw

where ł(Ł) is the matric potential,  is the porosity, and aw is the surface tension at
the air/water interface.

8.2.4 Plugging
Plugging is the reduction in porosity and hydraulic conductivity of the porous medium
due to entrapment of the colloidal particles. Analysis of plugging effects has particular
implications in quantifying the performance of sand filters in water treatment
facilities. In contaminant fate and transport models plugging can become important
when large quantities of colloidal particles are present. This condition can occur in
stormwater infiltration basins, for example, when large quantities of particles are
entrapped in the surface layers of the sediments, causing the permeability to decrease.
272 MODELLING OF POLLUTANTS IN COMPLEX ENVIRONMENTAL SYSTEMS

A few researchers have studied plugging using quantitative approaches. Sen et al.
(2002) suggested calculating the porosity changes via a mass balance on the entrapped
colloid concentration and then used the equation suggested by Khilar et al. (1985) to
obtain the hydraulic conductivity as a function of porosity:
@ r @Gs
¼ b (8:35)
@t rf @ t

and
 2

K ¼ K0 (8:36)
0

where K is the hydraulic conductivity, rf is the average density of colloidal particles,


and K0 and 0 are the reference hydraulic conductivity and porosity, respectively.
Alternative approaches exist, all of which involve linking the changes to the
hydraulic conductivity through the pore space reduction to the colloid deposition.
Because the change in porosity resulting from deposition of a specified number of
colloids with known geometry is a relatively reliable calculation, the robustness of
these approaches is dependent primarily on the robustness of the linkage between the
porosity and conductivity. Bedrikovetsky et al. (2001) used the following empirical
hyperbolic relationship to predict the decrease in hydraulic conductivity as a result of
colloid filtration:
K0
K¼ (8:37)
1 þ Gs

where  is an empirical parameter.


Shapiro et al. (2007) employed a stochastic approach, by considering a distribu-
tion of pore sizes and colloid sizes and using a population balance model that takes
into account the dynamics of the distributions as a result of colloid entrapment. Using
several approximations they showed that this method can lead to the following
relationship for permeability, which is a generalisation of Equation 8.36:
 

K ¼ K0 (8:38)
0

where  is a parameter that depends on the statistical properties of the pore and colloid
sizes.

8.3 COLLOID-FACILITATED TRANSPORT

Colloid-facilitated transport models have mostly been developed based on the mass
balance of species dissolved in the aqueous phase, sorbed to mobile and immobile
colloids and sorbed solid matrix. Different models arise from: (1) the different levels
of complexity considered in the colloid transport model; and (2) the different levels of
COLLOID-FACILITATED CONTAMINANT TRANSPORT IN UNSATURATED POROUS MEDIA 273

complexity considered in the treatment of mass exchange between the aqueous phase
and the mobile and immobile solid phases (e.g., equilibrium/kinetic, single site/
multiple sites).

8.3.1 Fully equilibrium models


Early approaches to modelling of colloid-facilitated transport involved adjusting the
retardation factor for the transport of contaminants to reflect their additional mobility
due to association with mobile colloids (e.g., Magee et al., 1991) or by adjusting the
effective advective velocity and dispersion coefficients (e.g., Enfield and Bengtsson,
1988). More recent models have been placed on an improved mechanistic basis by
considering dissolved, sorbed (immobile) and sorbed to colloid (mobile) phases. The
associated three-phase equation for colloid-facilitated transport in the presence of
homogeneous colloids can be written as
 
@C @Gm  c rB @ s rB @Gs  cs @C @Gm  c @ @C @Gm  c
þ þ þ þv þ vc ¼ D þ Dc
@t @t Ł @t Ł @t @z @z @z @z @z
(8:39)

where C [Mc /L3 ] is the mass concentration of dissolved contaminants in the pore
water (Mc refers to the dimension for mass of contaminants); s [Mc /M] is the
chemical concentration sorbed to immobile surfaces, c [Mc /M] is the chemical
concentration sorbed to mobile colloids, and cs [Mc /M] is the chemical concentration
sorbed to immobile colloids; v [L/T] is the advective velocity for dissolved species;
and D [L2 /T] is the dispersion coefficient for dissolved contaminants. Other terms are
as previously defined. Assuming equilibrium between sorbed and mobile colloids as
well as aqueous and all sorbed contaminant concentrations,

Gs ¼ K s Gm (8:40)

and

K Dc C ¼  c (8:41)

K Ds C ¼  s (8:42)

K Dsc C ¼  sc (8:43)

where Ks is the equilibrium colloid distribution coefficient between the attached and
mobile colloid phases, and KDc , KDs and KDsc are the contaminant equilibrium
distribution coefficients between the bulk aqueous phase and mobile colloidal
particles, the solid matrix, and immobile colloidal particles, respectively. Incorporat-
ing Equations 8.40 to 8.43 into Equation 8.39, and also assuming that the advective
velocities and dispersion coefficients are equal for colloidal particles and aqueous
species, yields
274 MODELLING OF POLLUTANTS IN COMPLEX ENVIRONMENTAL SYSTEMS

@C @Gm C rB @C rB @Gm C @C @Gm C


þK Dc þ K Ds þ K s K Dcs þv þ vK Dc
@t @t Ł @t Ł @t @z @z
  (8:44)
@ @C @Gm C
¼ D þ DK Dc
@z @z @z

Now, assuming no change in the total concentration of colloidal particles with time,
one can write
 
r r @C @C
1þK Dc Gm þ B K Ds þ B K s K Dcs Gm þ vð1 þ K Dc Gm Þ
Ł Ł @t @z
  (8:45)
@ @C
¼ ð1 þ K Dc Gm Þ D
@z @z

This expression can be expressed either as a simple advection–dispersion equation


(ADE) with adjusted retardation factors (e.g., Magee et al., 1991), or with adjusted
advective velocities and dispersion coefficients, as in Enfield and Bengtsson (1988).
Mills et al. (1991) also used this approach, but considered multi-component contami-
nant transport in variably saturated media. The problem with this approach is that it
assumes equilibrium between mobile and immobile colloids, and also ignores the
spatial and temporal variations of colloid concentration in the medium.

8.3.2 Dynamics of colloid transport and kinetic solid–water


mass transfer
Additional complexity in colloid-facilitated transport is achieved by using more
sophisticated colloid transport models as well as more sophisticated chemical mass
exchange models between different phases. Corapcioglu and Jiang (1993) and Jiang
and Corapcioglu (1993) developed a colloid-facilitated transport model in granular
porous media considering contaminants to reside at three phases: attached to the soil
matrix; attached to mobile and immobile colloids; and dissolved in the aqueous phase.
They considered colloid entrapment and release to be kinetically controlled, but
instantaneous equilibrium to be attained between contaminants and all solid phases. In
this case the relationship expressing the rate of change of attached colloidal particles
with time should replace Equation 8.40:
@Gs
rb ¼ k att ŁGm  k det rb Gs (8:46)
@t

In the case where all the chemical mass exchange processes between the solid and
aqueous phase are considered to be kinetically controlled, the mass balance equations
for concentrations sorbed to soil matrix, mobile and immobile phases can be written
as
@ s
¼ k sa ð s  K Ds C Þ (8:47)
@t
COLLOID-FACILITATED CONTAMINANT TRANSPORT IN UNSATURATED POROUS MEDIA 275

 
@Gm  c @Gm  c @ @Gm  c
þ vc ¼ Dc  k ca Gm ð c  K Dc C Þ
@t @z @z @z
(8:48)
1 
 k att ŁGm  c  k det rb Gs  sc
Ł
@Gs  cs 1 
¼ k csa Gs ð cs  K Dcs C Þ þ k att ŁGm  c  k det rb Gs  sc (8:49)
@t rb

where ksa [1/T], kca [1/T] and kcsa [1/T] are the mass exchange rate coefficients
between the bulk aqueous phase and soil matrix, mobile colloids and immobile
colloids, respectively. The first terms in Equations 8.48 and 8.49 represent the kinetic
contaminant exchange between mobile and immobile colloids and the bulk aqueous
phase, and the second terms represent the effect of deposition and detachment of
colloids to/from the surfaces.
Saiers and Hornberger (1996) implemented a so-called ‘‘two-box’’ approach by
categorising the sorption sites on the colloidal particles into two types, and consider-
ing aqueous–solid mass exchange for the first type of site to be at equilibrium, with
Langmuir isotherms, and kinetically controlled with a linear isotherm for the second
type of site. So the mass balance for the sorbed phases c , s and sc was written as
 c1 ¼ K Dc C (8:50)
 s1 ¼ K Ds C (8:51)
 sc1 ¼ K Dsc C (8:52)
@ s2
¼ k sa ð s2  K Ds C Þ (8:53)
@t
 
@Gm  c2 @Gm  c2 @ @Gm  c2
þ vc ¼ Dc  k ca Gm ð c2  K Dc C Þ
@t @z @z @z
(8:54)
1 
 k att ŁGm  c2  k det rb Gs  sc2
Ł
@Gs  cs2 1 
¼ k csa Gs ð cs2  K Dcs C Þ þ k att ŁGm  c2  k det rb Gs  sc2 (8:55)
@t rb

Roy and Dzombak (1998) applied this ‘‘two-box’’ approach to model transport of
hydrophobic organic compounds in the presence of colloids. Turner et al. (2006) used
the two-box model by Saiers and Hornberger (1996) to investigate the effect of the
desorption rate from the colloidal particles by observing the transport of caesium, with
slower desorption, and strontium, with faster desorption, in the presence of colloids,
and found that the desorption rate has a significant impact on the role of colloids in
the transport of contaminants.
Choi and Corapcioglu (1997) developed the first colloid-facilitated model in
unsaturated porous media for non-volatile contaminants. They introduced a new
colloidal phase to represent colloids captured in the air/water interface, they assumed
276 MODELLING OF POLLUTANTS IN COMPLEX ENVIRONMENTAL SYSTEMS

the air/water interface to be proportional to the volumetric air content, and they used a
linear kinetic mass-exchange relationship to model the colloidal entrapment and
release to the air/water interface. So for colloid-facilitated transport in the unsaturated
zone there are three distinct phases in which the colloids themselves can occur:
aqueous mobile colloids, colloids attached to the solid phase, and colloids attached to
the air/water interface.
Corapcioglu and Wang (1999) developed a dual-porosity (mobile–immobile)
model for colloid and colloid-facilitated transport using an equilibrium sorption
assumption, aiming at capturing the effects of pore-scale heterogeneity and preferen-
tial flow. Later Bekhit and Hassan (2005b) used this as the basis for their two-
dimensional colloid-facilitated contaminant transport model.
As noted above, several researchers have found that the process of colloid
entrapment can be considered irreversible as long as the physical and chemical factors
affecting mobilisation of the colloids, including ionic strength, pH, pore water velocity
and moisture content, are constant. At the same time, colloids might be produced in
the porous medium through weathering, wet–dry cycles, or decementation (Ryan and
Elimelech, 1996). Therefore in cases where none of these chemical or physical
perturbations takes place, the source of the colloids being mobilised is different from
that for the colloids previously captured. Sen et al. (2004) developed an equilibrium
three-phase model for colloid-facilitated transport in saturated media considering an
irreversible filtration of colloids to the soil matrix. They considered the source of
colloids being released to be different from the colloids captured by the medium by
defining two immobile colloidal phases: irreversibly captured colloids Gsi and
immobile colloids available for release Gsf (Equation 8.3), controlled by the following
mass balance equations:
@Gsi
rb ¼ k att ŁGm (8:56)
@t
@Gsf
rb ¼ k det rb Gs (8:57)
@t

They therefore had five contaminant phases: dissolved; sorbed to solid matrix; sorbed
to mobile colloids; irreversibly attached colloids; and immobile available colloids.
Massoudieh and Ginn (2007) used the same approach for colloid-facilitated transport
of multiple compounds in unsaturated media, but considered the mass exchange
between different phases to be kinetically controlled while the colloidal exchange
between the air phase and the aqueous phase was considered to be in equilibrium,
leading to the following solute transport equation:
@C @Gm  c rB @ s rB @ ð Gsi  csi þ Gsf  csf Þ ˆ @Ga  ca
þ þ þ þ
@t @t Ł @t Ł @t Ł @t
  (8:58)
@C @Gm  c @ @C @Gm  c
þv þ vc ¼ D þ Dc
@z @z @z @z @z

where ca is the sorbed concentration to the colloids entrapped in the air/water
COLLOID-FACILITATED CONTAMINANT TRANSPORT IN UNSATURATED POROUS MEDIA 277

interface, assumed to be equal to the concentration sorbed to the aqueous phase


colloids owing to the high rate of colloid exchange between the air/water interface and
the pore water (i.e., ca ¼ c ); and csi and csf are the sorbed concentrations to the
irreversibly captured colloids and available immobile colloids, respectively, as con-
trolled by the following mass balance equations:
@Gsf  csf 1
¼ k csa Gs ð csf  K Dcs C Þ  k det rb Gs  sc (8:59)
@t rb
@Gsi  csi 1
¼ k csa Gs ð csi  K Dcs C Þ þ k att ŁGm  c (8:60)
@t rb

Simunek et al. (2006) incorporated their colloid-facilitated model in a colloid


transport model that considers both attachment and straining. They used a two-site
model for sorption to the solid matrix and a non-linear single site model for adsorption
and desorption to the colloidal phases. The model is implemented in the HYDRUS-
1D software package.

8.4 HETEROGENEITY

Physical and chemical heterogeneity is shown to play a significant role in the transport
of colloidal particles in porous media. The existence of preferential flow paths can
facilitate the migration of colloidal particles in groundwater and the vadose zone
(Corapcioglu and Wang, 1999). Pore-scale heterogeneity, such as variability in the
pore sizes, geometries and surface charges, can also influence the dynamics of colloid
capture, as pointed out by several researchers (Song and Elimelech, 1994). The
electrochemical heterogeneities in porous media also can create barriers as well as
paths for colloidal transport, depending on their spatial distribution. In unsaturated
media, rapid infiltration of water in distinct paths (known as ‘‘fingering’’), seen
particularly in relatively dry soils, can enhance the migration of large amounts of
colloidal particles into deep layers in a relatively short time (McCarthy and McKay,
2004). These heterogeneities are overlooked in models where the chemical and
physical properties are considered to be homogenised in an REV. Song and Elimelech
(1994) incorporated the effect of surface charge heterogeneity by considering the
deposition rate coefficient to be randomly distributed over the surfaces with a given
frequency distribution:
ð
1
k att ¼ kð
Þłaca ð
Þd
(8:61)
ˆ ˆ

where k(
) is the transfer rate associated with the surface area element
, and łaca (
)
is the surface availability for
. By considering a clean bed initial condition and no
detachment conditions, they obtained an exponential function for łaca (
), and there-
fore were able to express the integral in Equation 8.61 with a correction factor
multiplied by the average deposition rate. So the resulting governing equation looks
like the conventional colloid-filtration theory, with the difference being that the
278 MODELLING OF POLLUTANTS IN COMPLEX ENVIRONMENTAL SYSTEMS

deposition rate coefficient is multiplied by a correction factor. Johnson et al. (1996)


categorised surface sites into favourable and unfavourable (patchwise geochemical
heterogeneity), and essentially used a two-site model to incorporate the effect of
charge heterogeneity. Sun et al. (2001) incorporated the patchwise geochemical
heterogeneity of Johnson et al. (1996) into a two-dimensional advection–dispersion
model while taking into account detachment of colloids from the surfaces. They
randomly generated the chemical heterogeneity factor º, which represents the ratio of
the surface area associated with favourable sites to the total area of the sites over the
two-dimensional grid, to model the geochemical heterogeneities with scales larger
than the grid size.
Saiers (2002) considered the effect of chemical heterogeneity of the colloidal
particles, as well as the soil matrix, by considering a distributed aqueous phase–
colloid mass-exchange rate and partitioning coefficient, as well as a distributed
colloid-deposition rate on surface sites implemented using a multiple-rate model
(Equations 8.23 and 8.26):
Xh X i X Xh X i
@C @ G m, j  c,ij r @  s,i rB @ G s, j  cs,ij
þ þ B þ
@t @t Ł @t Ł @t
( Xh X i) (8:62)
@C @Gm  c @ @C @ Gm, j  c,ij
þv þ vc ¼ D þ Dc
@z @z @z @z @z

with
@ s,i
¼ k sa,i ð s,i  K Ds,i C Þ (8:63)
@t
@Gm, j  c,ij 1
¼ k ca,ij Gm, j ð c,ij  K Dc,i C Þ  ð k att, j ŁGm, j  c,ij  k det, j rb Gs, j  sc,ij Þ
@t Ł
(8:64)
@Gs, j  cs,ij 1 
¼ k csa,ij Gs, j ð cs,ij  K Dcs,i C Þ þ k att, j ŁGm, j  c2  k det , j rb Gs, j  sc,ij
@t rb
(8:65)

where the subscript i refers to sorption site and subscript j refers to colloid type
category. The gamma distribution was assumed to govern the distribution of contami-
nant mass transfer rates to sorption sites, and the distribution coefficient was assumed
to depend on the transfer rate through a power law equation suggested by Pedit and
Miller (1994).
No method has been suggested so far to consider the effect of physical
heterogeneity at scales smaller than the computational grid size. The two-domain
model by Corapcioglu and Wang (1999) is designed to capture the effects of macro-
and micropores or preferential flow paths using a dual-porosity model; however, it is
not based on any relationship between the small-scale properties of the medium and
the parameters used in the homogenised Darcy-scale representation. Some attempts
COLLOID-FACILITATED CONTAMINANT TRANSPORT IN UNSATURATED POROUS MEDIA 279

have been made to incorporate the effect of heterogeneities at scales larger than the
computational grid by using Monte Carlo simulation and by randomly generating the
physical properties affecting the transport of colloids and contaminants (Sun et al.,
2001; Bekhit and Hassan, 2005a). Sun et al. (2001) and Bekhit and Hassan (2005a)
considered the heterogeneity of the porous medium by incorporating a spatially
variable parameter hydraulic conductivity, a colloid deposition coefficient and a
contaminant distribution coefficient. They considered the colloidal particles to be
uniform. By using this approach the effect of heterogeneity at a scale larger than the
grid scale can be captured.
When a wide variety of surface characteristics as well as a range of colloid
properties is present in the system, development of the conventional advection–
dispersion equations based on the REV concept becomes cumbersome, non-unique
and computationally burdensome. Also, it can become impossible to develop upscal-
ing techniques that can link the small-scale physical and chemical properties and their
variations to homogenised representations on a grid-scale using conventional stochas-
tic methods. In such cases particle-tracking and microfluidic approaches are deemed
useful, because they pose no limitations on the number of different distinct categories
of sites and colloids considered in the model. The drawbacks in using such models
include the extensive amount of computational resources they require and, sometimes,
difficulties in estimating the necessary parameters. Marseguerra et al. (2001a, 2001b)
developed a discrete particle colloid-facilitated model based on the Kolmogorov–
Dmitriev theory branching processes (Marseguerra and Zio, 1997). This method is
based on a series of colloidal particles representing colloidal particles and radio-
nuclide molecules that generated in a grid structure. Probabilities are assigned to each
of the processes, such as attachment to or detachment from the solid matrix, sorption
of contaminants to colloidal particles, or advective-diffusive transport of particles
from one grid cell to another. The difference between this method and conventional
particle-tracking approaches is that the exact positions of particles and molecules are
not needed; only the grid cell they are located in at each time step is recorded.

8.5 UNKNOWNS AND FUTURE DIRECTIONS


Although an enormous amount of work has been conducted to quantify colloid-
enhanced contaminant transport in porous media, our knowledge is still not sufficient
to make predictions of this phenomenon in natural settings, and particularly at the
scales of interest in risk assessment and remediation applications. In this section the
various potential directions that may eventually lead us to the construction of more
useful colloid-facilitated contaminant transport models will be presented.

8.5.1 Upscaling to field scale


The final goal of contaminant transport modelling is generally to perform risk
assessment or make predictions at the scale of contaminated subsurface ecosystems.
So far, most of the studies have been conducted in laboratories, at scales ranging from
280 MODELLING OF POLLUTANTS IN COMPLEX ENVIRONMENTAL SYSTEMS

a few centimetres to a few metres. Because of the large timescales of the problem, it is
not possible to conduct field studies to quantify the long-term behaviour of contami-
nants at field scale. No systematic techniques have been suggested to upscale the
outcomes of these studies to larger scales considering the physical and chemical
heterogeneities present in real-world applications. Current computational power does
not make it possible to incorporate heterogeneities at the scales captured in laboratory
experiments directly into field-scale models. The stochastic methods using Monte
Carlo simulation techniques to address the large-scale transport of contaminants in the
presence of colloids have typically used grid sizes much larger than the length scales
at which the laboratory experiments have been performed (e.g., Bekhit and Hassan,
2005a). Models fitted to laboratory experiments are not necessarily appropriate for
application to larger-scale problems without significant alterations of the model
parameters and, in some cases, the model structure, to capture the effect of hetero-
geneities. Although inverse modelling techniques have been used to estimate the
parameters of contaminant transport models at the field scale, there are a few problems
with this approach in the case of contaminant transport in the presence of colloids.
One such problem is the fact that in many cases it is impractical to characterise the
transport of colloids and contaminants at relevant timescales; also, this approach does
not provide reliable information on the actual physical processes at small scales, and
therefore does not provide any relationship between the small-scale characteristics of
the medium and the overall transport. The advection–dispersion equations that are
applicable to small-scale problems are not necessarily applicable to large-scale
heterogeneous media (McCarthy and McKay, 2004). In addition, since the models will
be calibrated using the data obtained for a limited period, there is no guarantee that
this approach will perform well in making long-term predictions.
For all the reasons mentioned above, one vital step in making colloid and colloid-
enhanced transport models applicable to real-world applications is to build systematic
approaches to upscaling the governing equations developed for laboratory-scale
situations to larger scales, and eventually to scales useful for addressing field-scale
contaminant transport problems.

8.5.2 Quantifying geochemical effects


Many studies have been conducted to identify geochemical effects (e.g., aqueous
chemistry, organic matter including ligands, multicomponent speciation and surface
complexation). As a result of these studies we now have an insight into the various
ways in which changes in the chemistry of the system can affect colloid mobilisation
and attachment. For instance, it appears that changes in, as opposed to the absolute
magnitude of, basic geochemical properties (especially ionic strength, concentration
of organic matter, and pH) can lead to episodic colloid release or erosion events in
natural conditions, on a par with the effects of hydrodynamic transients. However, our
ability to couple multicomponent geochemical conditions and transients with colloid
and colloid-facilitated transport to build predictive models for long-term simulation of
contaminant transport remains limited, in part because of the computational chal-
lenges in coupling models of multicomponent geochemical conditions with colloid
COLLOID-FACILITATED CONTAMINANT TRANSPORT IN UNSATURATED POROUS MEDIA 281

transport and colloid–surface interactions, and in part because of the constitutive


theory challenges in the same constructions: that is, we do not know exactly how
different geochemical environments will result in colloid surface properties favourable
or not to transport, in natural (not contrived) conditions. Geochemical fluctuations as a
result of episodic infiltrations, or the movement of contaminants themselves, can
affect the pH, ionic strength and chemical composition of the pore water, each of these
influencing deposition, mobilisation of colloids, and decementation of the solid matrix
leading to colloid release. Biological factors such as biofilm growth are also shown to
affect colloid transport by changing the chemistry of the system or the pore water
hydrodynamics. The development of general models to incorporate geochemical
variations in colloid-facilitated transport seems a difficult goal to reach, at least in the
short term; targeting the construction of specific models to capture the integrated
effects of geochemical variations in certain studies appear to offer a more pragmatic
way to bridge the knowledge gaps.

8.5.3 Heterogeneous colloids characteristics


In any aquifer or vadose zone, colloids are present with various sources and a
spectrum of properties, in terms both of chemical composition (e.g., surface charge,
sorption capability) and of transport behaviour (e.g., size, weight). Lumping all the
different categories of colloids into one homogeneous group is an oversimplification
that can lead to an inability to explain observed migration of contaminants in the
presence of colloids. Other than the model by Saiers (2002), which suggested a way to
incorporate multi-disperse colloids, no other studies have been found that address this
challenge. Two reasons can be given for the rarity of such efforts. The first is that
determination of the range of colloids and estimation of the parameters that control
their transport in porous media, such as the deposition and release rates, is a
challenging task. The second reason is the large computational burden of such an
effort, particularly in view of the wide range of contaminants bound to colloids. One
approach for taking the variability in colloid properties (such as parameters controlling
mobility and sorption) into account is to express these properties as joint distributions,
but quantification of such joint frequency distributions is a difficult task. Particle-
tracking approaches have been shown to be more effective in handling heterogeneous
colloid properties (Sun et al., 2001), but they typically require large amounts of
computer memory in order to be applied to real cases.

8.5.4 Pore-scale heterogeneity


Classical colloid transport models are constructed using idealisation of the pore-scale
geometry, by assuming that the medium consists either of uniformly placed spherical
grains (Happel sphere-in-cell model) or of uniform, mono-sized (e.g., cylindrical or
spherical) pores. A natural porous medium is neither of these. It consists of non-
uniformly distributed, non-spherical and multi-disperse grains, or a range of variably
sized, tortuous pores. The question is: how does either idealisation of the natural
porous medium affect the capability of models that result from it? Several ‘‘micro-
282 MODELLING OF POLLUTANTS IN COMPLEX ENVIRONMENTAL SYSTEMS

fluidic’’ techniques are now available to model hydrodynamic and colloid transport in
complicated and irregular geometries, such as the lattice Boltzmann method (Basa-
gaoglu et al., 2008) and the smoothed particle hydrodynamic approach (Yamamoto et
al., 2007; Tartakovsky et al., 2009). Micro-tomography techniques (Spanne et al.,
1994; Auzerais et al., 1996) and Monte Carlo spherical packing simulation methods
(Yang et al., 2000) have also been used to construct more realistic representation of
porous media. These techniques have provided tools to test various hypotheses
regarding the mechanism of colloid release and entrapment, and can help in gaining
insight into the important processes affecting the transport of colloids. However,
mainly because of limitations in the available computational power, pore-scale models
are applicable only to very small domains, with sizes of the order of centimetres, and
therefore it is not yet possible to evaluate the effects of heterogeneities at larger scales
by direct application of these models.

REFERENCES
Accornero, M., Marini, L., Ottonello, G. and Zuccolini, M.V. (2005). The fate of major constituents
and chromium and other trace elements when acid waters from the derelict Libiola mine (Italy)
are mixed with stream waters. Applied Geochemistry. 20:1368–1390.
Adamczyk, Z., Dabros, T., Czarnecki, J. and van de Ven, T.G.M. (1983). Particle transfer to solid-
surfaces. Advances in Colloid and Interface Science. 19: 183–252.
Arulanandan, K., Longanathan, P. and Krone, R.B. (1975). Pore and eroding fluid influences on
surface erosion on soil. Journal of Geotechnical Engineering. ASCE. 101: 51–57.
Auzerais, F.M., Dunsmuir, J., Ferreol, B.B., Martys, N., Olson, J., Ramakrishnan, T.S., Rothman, D.H.
and Schwartz, L.M. (1996). Transport in sandstone: a study based on three-dimensional
microtomography. Geophysical Research Letters. 23: 705–708.
Basagaoglu, H., Meakin, P., Succi, S., Redden, G.R. and Ginn, T.R. (2008). Two-dimensional lattice
Boltzmann simulation of colloid migration in rough-walled narrow flow channels. Physical
Review E. 77: 031405.
Bedrikovetsky, P., Marchesin, D., Shecaira, F., Souza, A.L., Milanez, P.V. and Rezende, E. (2001).
Characterisation of deep bed filtration system from laboratory pressure drop measurements.
Journal of Petroleum Science and Engineering. 32: 167–177.
Bekhit, H.M. and Hassan, A.E. (2005a). Stochastic modeling of colloid-contaminant transport in
physically and geochemically heterogeneous porous media. Water Resources Research. 41,
W02010.1–W02010.18.
Bekhit, H.M. and Hassan, A.E. (2005b). Two-dimensional modeling of contaminant transport in
porous media in the presence of colloids. Advances in Water Resources. 28: 1320–1335.
Bostrom, M., Williams, D.R.M. and Ninham, B.W. (2001). Specific ion effects: why DLVO theory
fails for biology and colloid systems. Physical Review Letters. 87, 168103/168101–168104.
Bradford, S.A. and Bettahar, M. (2006). Concentration dependent transport of colloids in saturated
porous media. Journal of Contaminant Hydrology. 82: 99–117.
Bradford, S.A. and Leij, F.J. (1997). Estimating interfacial areas for multi-fluid soil systems. Journal
of Contaminant Hydrology. 27: 83–105.
Bradford, S.A., Simunek, J., Bettahar, M., van Genuchten, M.T. and Yates, S.R. (2003). Modeling
colloid attachment, straining, and exclusion in saturated porous media. Environmental Science
& Technology. 37: 2242–2250.
Bradford, S.A., Torkzaban, S., Leij, F., Šimunek, J. and van Genuchten, M.T. (2009). Modeling the
coupled effects of pore space geometry and velocity on colloid transport and retention. Water
Resources Research. 45, W02414, doi:10.1029/2008WR007096.
COLLOID-FACILITATED CONTAMINANT TRANSPORT IN UNSATURATED POROUS MEDIA 283

Buddemeier, R.W. and Hunt, J.R. (1988). Transport of colloidal contaminants in groundwater:
radionuclide migration at the Nevada test site. Applied Geochemistry. 3: 535–548.
Cary, J.W. (1994). Estimating the surface-area of fluid-phase interfaces in porous-media. Journal of
Contaminant Hydrology. 15: 243–248.
Choi, H.C. and Corapcioglu, M.Y. (1997). Transport of a non-volatile contaminant in unsaturated
porous media in the presence of colloids. Journal of Contaminant Hydrology. 25: 299–324.
Chu, Y., Jin, Y., Flury, M. and Yates, M.V. (2001). Mechanisms of virus removal during transport in
unsaturated porous media. Water Resources Research. 37: 253–263.
Corapcioglu, M.Y. and Choi, H. (1996). Modeling colloid transport in unsaturated porous media and
validation with laboratory column data. Water Resources Research. 32: 3437–3449.
Corapcioglu, M.Y. and Jiang, S.Y. (1993). Colloid-facilitated groundwater contaminant transport.
Water Resources Research. 29: 2215–2226.
Corapcioglu, M.Y. and Wang, S. (1999). Dual-porosity groundwater contaminant transport in the
presence of colloids. Water Resources Research. 35: 3261–3273.
Dabros, T. and van de Ven, T.G.M. (1982). Kinetics of coating by colloidal particles. Journal of
Colloid and Interface Science. 89: 232–244.
DeNovio, N.M., Saiers, J.E. and Ryan, J.N. (2004). Colloid movement in unsaturated porous media:
recent advances and future directions. Vadose Zone Journal. 3: 338–351.
Dusek, J., Vogel, T., Lichner, L., Cipakova, A. and Dohnal, M. (2006). Simulated cadmium transport
in macroporous soil during heavy rainstorm using dual-permeability approach. Biologia. 61:
S251-S254.
Enfield, C.G. and Bengtsson, G. (1988). Macromolecular transport of hydrophobic contaminants in
aqueous environments. Ground Water. 26: 64–70.
Gargiulo, G., Bradford, S.A., Simunek, J., Ustohal, P., Vereecken, H. and Klumpp, E. (2008). Bacteria
transport and deposition under unsaturated flow conditions: the role of water content and
bacteria surface hydrophobicity. Vadose Zone Journal. 7: 406–419.
Ginn, T.R. (2000). On the distribution of multicomponent mixtures over generalized exposure time in
subsurface flow and reactive transport: theory and formulations for residence-time-dependent
sorption/desorption with memory. Water Resources Research. 36: 2885–2893.
Govindaraju, R.S., Reddi, L.N. and Kasavaraju, S.K. (1995). A physically-based model for mobiliza-
tion of kaolinite particles under hydraulic gradients. Journal of Hydrology. 172: 331–350.
Grasso, M., Catara, F. and Sambataro, M. (2002). Boson-mapping-based extension of the random-
phase approximation in a three-level Lipkin model. Physical Review C. 66: 064303.
Harter, T., Atwill, E.R., Hou, L., Karle, B.M. and Tate, K.W. (2008). Developing risk models of
Cryptosporidium transport in soils from vegetated, tilted soilbox experiments. Journal of
Environmental Quality. 37: 245–258.
Hermansson, M. (1999). The DLVO theory in microbial adhesion. Colloids and Surfaces B:
Biointerfaces. 14: 105–119.
Hochella, M.F., Moore, J.N., Putnis, C.V., Putnis, A., Kasama, T. and Eberl, D.D. (2005). Direct
observation of heavy metal-mineral association from the Clark Fork River superfund complex:
implications for metal transport and bioavailability. Geochimica et Cosmochimica Acta. 69:
1651–1663.
Jacobsen, O.H., Moldrup, P., Larsen, C., Konnerup, L. and Petersen, L.W. (1997). Particle transport in
macropores of undisturbed soil columns. Journal of Hydrology. 196: 185–203.
Jarvis, N.J., Villholth, K.G. and Ulen, B. (1999). Modelling particle mobilization and leaching in
macroporous soil. European Journal of Soil Science. 50: 621–632.
Jewett, D.G., Logan, B.E., Arnold, R.G. and Bales, R.C. (1999). Transport of Pseudomonas
fluorescens strain P17 through quartz sand columns as a function of water content. Journal of
Contaminant Hydrology. 36: 73–89.
Jiang, S.Y. and Corapcioglu, M.Y. (1993). A hybrid equilibrium-model of solute transport in porous-
media in the presence of colloids. Colloids and Surfaces A: Physicochemical and Engineering
Aspects. 73: 275–286.
Johnson, P.R. and Elimelech, M. (1995). Dynamics of colloid deposition in porous media: blocking
based on random sequential adsorption. Langmuir. 11: 801–812.
284 MODELLING OF POLLUTANTS IN COMPLEX ENVIRONMENTAL SYSTEMS

Johnson, P.R., Sun, N. and Elimelech, M. (1996). Colloid transport in geochemically heterogeneous
porous media: modeling and measurements. Environmental Science & Technology. 30: 3284–
3293.
Johnson, W.P., Blue, K.A., Logan, B.E. and Arnold, R.G. (1995). Modeling bacteria detachment
during transport through porous-media as a residence-time dependent process. Water Re-
sources Research. 31: 2649–2658.
Kallay, N., Tomic, M., Biskup, B., Kunjasic, I. and Matijevic, E. (1987). Particle adhesion and
removal in model systems. XI. Kinetics of attachment and detachment for hematite–glass
systems. Colloids and Surfaces. 28: 185–197.
Keller, A.A. and Sirivithayapakorn, S. (2004). Transport of colloids in unsaturated porous media:
explaining large-scale behavior based on pore-scale mechanisms. Water Resources Research.
40: W12403.1–W12403.8.
Khilar, K.C., Fogler, H.S. and Gray, D.H. (1985). Model for piping-plugging in earthen structures.
Journal of Geotechnical Engineering, ASCE. 111: 833–846.
Kim, S.B. and Corapcioglu, M.Y. (2002). Contaminant transport in dual-porosity media with
dissolved organic matter and bacteria present as mobile colloids. Journal of Contaminant
Hydrology. 59: 267–289.
Kretzschmar, R., Robarge, W.P. and Amoozegar, A. (1995). Influence of natural organic matter on
colloid transport through saprolite. Water Resources Research. 31: 435–445.
Kretzschmar, R. and Schafer, T. (2005). Metal retention and transport on colloidal particles in the
environment. Elements. 1: 205–210.
Lazouskaya, V. and Jin, Y. (2008). Colloid retention at air/water interface in a capillary channel.
Colloids and Surfaces A: Physicochemical and Engineering Aspects. 325: 141–151.
Lenhart, J.J. and Saiers, J.E. (2003). Colloid mobilization in water-saturated porous media under
transient chemical conditions. Environmental Science & Technology, 37: 2780–2787.
Magee, B.R., Lion, L.W. and Lemley, A.T. (1991). Transport of dissolved organic macromolecules
and their effect on the transport of phenanthrene in porous media. Environmental Science &
Technology. 25: 323–331.
Majdalani, S., Michel, E., Di Pietro, L., Angulo-Jaramillo, R. and Rousseau, M. (2007). Mobilization
and preferential transport of soil particles during infiltration: a core-scale modeling approach.
Water Resources Research. 43(5): W05401.1–W05401.14.
Majdalani, S., Michel, E., Di-Pietro, L. and Angulo-Jaramillo, R. (2008). Effects of wetting and
drying cycles on in situ soil particle mobilization. European Journal of Soil Science. 59(2),
147–155.
Marseguerra, M., Patelli, E. and Zio, E. (2001a). Groundwater contaminant transport in presence of
colloids. I: A stochastic nonlinear model and parameter identification. Annals of Nuclear
Energy. 28: 777–803.
Marseguerra, M., Patelli, E. and Zio, E. (2001b). Groundwater contaminant transport in presence of
colloids. II: Sensitivity and uncertainty analysis on literature case studies. Annals of Nuclear
Energy. 28: 1799–1807.
Marseguerra, M. and Zio, E. (1997). Modelling the transport of contaminants in groundwater as a
branching stochastic process. Annals of Nuclear Energy. 24: 625–644.
Marty, R.C., Bennett, D. and Thullen, P. (1997). Mechanism of plutonium transport in a shallow
aquifer in Mortandad Canyon, Los Alamos National Laboratory, New Mexico. Environmental
Science & Technology. 31: 2020–2027.
Massoudieh, A. and Ginn, T.R. (2007). Modeling colloid-facilitated transport of multi-species
contaminants in unsaturated porous media. Journal of Contaminant Hydrology. 92: 162–183.
McCarthy, J.F., Czerwinski, K.R., Sanford, W.E., Jardine, P.M. and Marsh, J.D. (1998). Mobilization
of transuranic radionuclides from disposal trenches by natural organic matter. Journal of
Contaminant Hydrology. 30: 49–77.
McCarthy, J.F. and Degueldre, C. (eds) (1993). Environmental Particles (Environmental Analytical
and Physical Chemistry). Lewis Publishers, Boca Raton, FL.
McCarthy, J.F. and McKay, L.D. (2004). Colloid transport in the subsurface: past, present, and future
challenges. Vadose Zone Journal. 3: 326–337.
COLLOID-FACILITATED CONTAMINANT TRANSPORT IN UNSATURATED POROUS MEDIA 285

McCarthy, J.F. and Zachara, J.M. (1989). Subsurface transport of contaminants: mobile colloids in the
subsurface environment may alter the transport of contaminants. Environmental Science &
Technology. 23: 496–502.
Meinders, J.M., Noordmans, J. and Busscher, H.J. (1992). Simultaneous monitoring of the adsorption
and desorption of colloidal particles during deposition in a parallel plate flow chamber. Journal
of Colloid and Interface Science. 152: 265–280.
Mills, W.B., Liu, S. and Fong, F.K. (1991). Literature review and model (COMET) for colloid metals
transport in porous media. Ground Water. 29: 199–208.
Nelson, K.E. and Ginn, T.R. (2005). Colloid filtration theory and the Happel sphere-in-cell model
revisited with direct numerical simulation of colloids. Langmuir. 21: 2173–2184.
Novikov, A.P., Kalmykov, S.N., Utsunomiya, S., Ewing, R.C., Horreard, F., Merkulov, A., Clark, S.B.,
Tkachev, V.V. and Myasoedov, B.F. (2006). Colloid transport of plutonium in the far-field of
the Mayak Production Association, Russia. Science. 314: 638–641.
Nyhan, J.W., Drennon, B.J., Abeele, W.V., Wheeler, M.L., Purtymun, W.D., Trujillo, G., Herrera, W.J.
and Booth, J.W. (1985). Distribution of plutonium and americium beneath a 33-yr-old liquid
waste-disposal site. Journal of Environmental Quality. 14: 501–509.
Pang, L.P. and Simunek, J. (2006). Evaluation of bacteria-facilitated cadmium transport in gravel
columns using the HYDRUS colloid-facilitated solute transport model. Water Resources
Research. 42, W12S10, doi:10.1029/2006WR004896.
Pedit, J.A. and Miller, C.T. (1994). Heterogeneous sorption processes in subsurface systems. 1: Model
formulations and applications. Environmental Science & Technology. 28: 2094–2104.
Penrose, W.R., Polzer, W.L., Essington, E.H., Nelson, D.M. and Orlandini, K.A. (1990). Mobility of
plutonium and americium through a shallow aquifer in a semiarid region. Environmental
Science & Technology. 24: 228–234.
Powelson, D.K. and Gerba, C.P. (1994). Virus removal from sewage effluents during saturated and
unsaturated flow through soil columns. Water Research. 28: 2175–2181.
Rajagopalan, R. and Tien, C. (1976). Trajectory analysis of deep-bed filtration with the sphere-in-cell
porous media model. AIChE Journal. 22: 523–533.
Redman, J.A., Estes, M.K. and Grant, S.B. (2001). Resolving macroscale and microscale hetero-
geneity in virus filtration. Colloids and Surfaces A: Physicochemical and Engineering Aspects.
191: 57–70.
Redman, J.A., Walker, S.L. and Elimelech, M. (2004). Bacterial adhesion and transport in porous
media: role of the secondary energy minimum. Environmental Science & Technology. 38:
1777–1785.
Reimus, P., Murrell, M., Abdel-Fattah, A., Garcia, E., Norman, D., Goldstein, S., Nunn, A., Gritzo, R.
and Martinez, B. (2005). Colloid Characteristics and Radionuclide Associations with Colloids
in Near-Field Waters at the Nevada Test Site, FY 2005 Progress Report, Los Alamos National
Laboratory, Los Alamos, NM, 2005.
Robinson, B.A., Li, C.H. and Ho, C.K. (2003). Performance assessment model development and
analysis of radionuclide transport in the unsaturated zone, Yucca Mountain, Nevada. Journal
of Contaminant Hydrology. 62–63: 249–268.
Roy, S.B. and Dzombak, D.A. (1996). Naþ -Ca2(þ) exchange effects in the detachment of latex colloids
deposited in glass bead porous media. Colloids and Surfaces A: Physicochemical and
Engineering Aspects. 119: 133–139.
Roy, S.B. and Dzombak, D.A. (1997). Chemical factors influencing colloid-facilitated transport of
contaminants in porous media. Environmental Science & Technology. 31: 656–664.
Roy, S.B. and Dzombak, D.A. (1998). Sorption nonequilibrium effects on colloid-enhanced transport
of hydrophobic organic compounds in porous media. Journal of Contaminant Hydrology.
30:179–200.
Ryan, J.N. and Elimelech, M. (1996). Colloid mobilization and transport in groundwater. Colloids
and Surfaces A: Physicochemical and Engineering Aspects. 107: 1–56.
Ryan, J.N. and Gschwend, P.M. (1994). Effects of ionic strength and flow rate on colloid release-
relating kinetics to intersurface potential energy. Journal of Colloid and Interface Science.
164: 21–34.
286 MODELLING OF POLLUTANTS IN COMPLEX ENVIRONMENTAL SYSTEMS

Saiers, J.E. (2002). Laboratory observations and mathematical modeling of colloid-facilitated


contaminant transport in chemically heterogeneous systems. Water Resources Research. 38:
3.1–3.4.
Saiers, J.E. and Hornberger, G.M. (1996). The role of colloidal kaolinite in the transport of cesium
through laboratory sand columns. Water Resources Research. 32: 33–41.
Saiers, J.E. and Lenhart, J.J. (2003a). Colloid mobilization and transport within unsaturated porous
media under transient flow conditions. Water Resources Research. 39: 33–42.
Saiers, J.E. and Lenhart, J.J. (2003b). Ionic-strength effects on colloid transport and interfacial
reactions in partially saturated porous media. Water Resources Research. 39(9): 10.1–10.11.
Schaaf, P. and Talbot, J. (1989). Surface exclusion effects in adsorption processes. Journal of
Chemical Physics. 91: 4401–4409.
Schelde, K., Moldrup, P., Jacobsen, O.H., de Jonge, H., de Jonge, L.W. and Komatsu, T. (2002).
Diffusion-limited mobilization and transport of natural colloids in macroporous soil. Vadose
Zone Journal. 1: 125–136.
Sen, T.K., Nalwaya, N. and Khilar, K.C. (2002). Colloid-associated contaminant transport in porous
media: 2. Mathematical modeling. AIChE Journal. 48: 2375–2385.
Sen, T.K., Shanbhag, S. and Khilar, K.C. (2004). Subsurface colloids in groundwater contamination:
a mathematical model. Colloids and Surfaces A: Physicochemical and Engineering Aspects.
232: 29–38.
Shang, J.Y., Flury, M., Chen, G. and Zhuang, J. (2008). Impact of flow rate, water content, and
capillary forces on in situ colloid mobilization during infiltration in unsaturated sediments.
Water Resources Research. 44, W06411, doi:10.1029/2007WR006516.
Shapiro, A.A., Bedrikovetsky, P.G., Santos, A. and Medvedev, O.O. (2007). A stochastic model for
filtration of particulate suspensions with incomplete pore plugging. Transport in Porous Media.
67: 135–164.
Short, S.A., Lowson, R.T. and Ellis, J. (1988). U-234/U-238 and TH-230/U-234 activity rations in the
colloidal phases of aquifers in lateritic weathered zones. Geochimica et Cosmochimica Acta.
52: 2555–2563.
Sim, Y. and Chrysikopoulos, C.V. (2000). Virus transport in unsaturated porous media. Water
Resources Research. 36: 173–179.
Simunek, J., He, C.M., Pang, L.P. and Bradford, S.A. (2006). Colloid-facilitated solute transport in
variably saturated porous media: numerical model and experimental verification. Vadose Zone
Journal. 5: 1035–1047.
Sirivithayapakorn, S. and Keller, A. (2003). Transport of colloids in unsaturated porous media: a
pore-scale observation of processes during the dissolution of air/water interface. Water
Resources Research. 39: 6.1–6.10.
Smets, B.F., Grasso, D., Engwall, M.A. and Machinist, B.J. (1999). Surface physicochemical proper-
ties of Pseudomonas fluorescens and impact on adhesion and transport through porous media.
Colloids and Surfaces B: Biointerfaces. 14: 121–139.
Song, L. and Elimelech, M. (1993). Dynamics of colloid deposition in porous media: modeling the
role of retained particles. Colloids and Surfaces A: Physicochemical and Engineering Aspects.
73: 49–63.
Song, L. and Elimelech, M. (1994). Transient deposition of colloidal particles in heterogeneous
porous media. Journal of Colloid and Interface Science. 167: 301–313.
Spanne, P., Thovert, J.F., Jacquin, C.J., Lindquist, W.B., Jones, K.W. and Adler, P.M. (1994).
Synchrotron computed microtomography of porous-media-topology and transports. Physical
Review Letters. 73: 2001–2004.
Strevett, K.A. and Chen, G. (2003). Microbial surface thermodynamics and applications. Research in
Microbiology. 154: 329–335.
Sun, N., Elimelech, M., Sun, N.Z. and Ryan, J.N. (2001). A novel two-dimensional model for colloid
transport in physically and geochemically heterogeneous porous media. Journal of Contami-
nant Hydrology. 49: 173–199.
Suzuki, Y., Kelly, S.D., Kemner, K.M. and Banfield, J.F. (2002). Radionuclide contamination:
nanometre-size products of uranium bioreduction. Nature. 419: 134–134.
COLLOID-FACILITATED CONTAMINANT TRANSPORT IN UNSATURATED POROUS MEDIA 287

Talbot, J. (1996). Time dependent desorption: A memory function approach. Adsorption: Journal of
the International Adsorption Society. 2: 89–94.
Tartakovsky, A.M., Meakin, P. and Ward, A.L. (2009). Smoothed particle hydrodynamics model of
non-aqueous phase liquid flow and dissolution. Transport in Porous Media. 76: 11–34.
Tong, M.P., Ma, H.L. and Johnson, W.P. (2008). Funneling of flow into grain-to-grain contacts drives
colloid-colloid aggregation in the presence of an energy barrier. Environmental Science &
Technology. 42: 2826–2832.
Torkzaban, S., Hassanizadeh, S.M., Schijven, J.F. and van den Berg, H. (2006). Role of air/water
interfaces on retention of viruses under unsaturated conditions. Water Resources Research. 42,
W12S14, doi:10.1029/2006WR004904.
Tufenkji, N. and Elimelech, M. (2004). Correlation equation for predicting single-collector efficiency
in physicochemical filtration in saturated porous media. Environmental Science & Technology.
38: 529–536.
Tufenkji, N., Redman, J.A. and Elimelech, M. (2003). Interpreting deposition patterns of microbial
particles in laboratory-scale column experiments. Environmental Science & Technology. 37:
616–623.
Turner, N.B., Ryan, J.N. and Saiers, J.E. (2006). Effect of desorption kinetics on colloid-facilitated
transport of contaminants: cesium, strontium, and illite colloids. Water Resources Research.
42, W12S09.1–W12S09.17.
van de Weerd, H. and Leijnse, A. (1997). Assessment of the effect of kinetics on colloid facilitated
radionuclide transport in porous media. Journal of Contaminant Hydrology. 26: 245–256.
van de Weerd, H., Leijnse, A. and Van Riemsdijk, W.H. (1998). Transport of reactive colloids and
contaminants in groundwater: effect of nonlinear kinetic interactions. Journal of Contaminant
Hydrology. 32: 313–331.
Wan, J.M. and Tokunaga, T.K. (1997). Film straining of colloids in unsaturated porous media:
conceptual model and experimental testing. Environmental Science & Technology. 31: 2413–
2420.
Wan, J.M. and Wilson, J.L. (1994). Colloid transport in unsaturated porous-media. Water Resources
Research. 30: 857–864.
Woessner, W.W., Ball, P.N., DeBorde, D.C. and Troy, T.L. (2001). Viral transport in a sand and gravel
aquifer under field pumping conditions. Ground Water. 39: 886–894.
Xu, S., Gao, B. and Saiers, J.E. (2006). Straining of colloidal particles in saturated porous media.
Water Resources Research. 42, W12S16, doi:10.1029/2006WR004948.
Yamamoto, R., Kim, K. and Nakayama, Y. (2007). Strict simulations of non-equilibrium dynamics of
colloids. Colloids and Surfaces A: Physicochemical and Engineering Aspects. 311: 42–47.
Yang, R.Y., Zou, R.P. and Yu, A.B. (2000). Computer simulation of the packing of fine particles.
Physical Review E. 62: 3900–3908.
Yao, K.-M., Habibian, M.T. and O’Melia, C.R. (1971). Water and waste water filtration. Concepts and
applications. Environmental Science & Technology. 5: 1105–1112.
Zhuang, J., McCarthy, J.F., Tyner, J.S., Perfect, E. and Flury, M. (2007). In situ colloid mobilization in
hanford sediments under unsaturated transient flow conditions: effect of irrigation pattern.
Environmental Science & Technology. 41: 3199–3204.

View publication stats

You might also like