Geotextiles and Geomembranes 42 (2014) 428e436

Contents lists available at ScienceDirect

Geotextiles and Geomembranes

journal homepage: www.elsevier.com/locate/geotexmem

Organic acid transport through a partially saturated liner system

beneath a landfill
Tingfa Liu 1, Liming Hu*
State Key Laboratory of Hydro-Science and Engineering, Department of Hydraulic Engineering, Tsinghua University, Beijing 100084, PR China

a r t i c l e i n f o a b s t r a c t

Article history: A one-dimensional model was developed to investigate the transport of organic acids (commonly found
Received 14 April 2014 in landfill leachate) through a partially saturated composite liner system beneath a landfill. Specific
Received in revised form attention was paid to the influence of water content distribution on aqueous-phase diffusion process.
9 June 2014
Composite liner system was investigated, which was consisted of a geomembrane and a compacted clay
Accepted 21 June 2014
liner underlain by three kinds of attenuation layer: sand layer, sandy clay loam layer, and clay layer.
Available online 11 July 2014
Volumetric water content profile in soil layers was obtained by Van Genuchten model, and the Millington
and Quirk model was employed to describe the non-linear relationship between volumetric water
Keywords:
Landfill
content and diffusion coefficient. Three cases were analyzed and compared, i.e., totally saturated con-
Composite liner dition, unsaturated condition without considering unsaturated diffusion model, and unsaturated con-
Organic contaminant dition considering unsaturated diffusion model. The numerical results show that the unsaturated sand
Millington and Quirk model attenuation layer could serve as excellent diffusion barrier to organic contaminant due to its low water
Unsaturated diffusion model retention capacity. When the dependence of diffusion coefficient on volumetric water content is suffi-
ciently considered, the contaminant flux decreases significantly in all the three kinds of attenuation layer.
Unsaturated diffusion model capturing the relationship between water content and diffusion coefficient
enables a more reasonable prediction of contaminant transport and distribution in soils.
© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

1. Introduction unsaturated pore space and their transport will be different to

In modern landfills, liners are crucial horizontal barriers
designed to protect groundwater and soils. Composite liners
involving a geomembrane (GMB) over a compacted clay liner (CCL)
or a GMB over a geosynthetic clay liner (GCL) are widely used in
landfills, and extensive research has been conducted to investigate
their short- and long-term performance (Rowe, 2005; Rowe, 2012).
Intact HDPE GMBs without defects are excellent barriers to the
advective and diffusive migration of many contaminant com-
pounds in landfill leachate. Volatile organic contaminants (VOCs)
can readily diffuse through HDPE GMBs (Haxo, 1990; Park and
Nibras, 1993; Sangam and Rowe, 2001; Edil, 2003; Rowe et al.,
2004; Sangam and Rowe, 2005; McWatters and Rowe, 2010), thus
the penetration of organic contaminant could occur in the whole
scale of composite liner. However, VOCs will volatilize in the
* Corresponding author. Tel: þ86 10 62797416; fax: þ86 10 62785593.
E-mail addresses: liutf11@mails.tsinghua.edu.cn (T. Liu), gehu@tsinghua.edu.cn
(L. Hu).
1
Tel.: þ86 10 62794154.
other less volatile organic contaminants.
Landfill leachate contains a variety of concentrated organics
such as carboxylic acids (Kjeldsen et al., 2002). The transport pro-
cess of these organic compounds through intact composite liners
deserves careful consideration.
Although geomembrane wrinkles are commonly generated
during construction, field observation suggested that wrinkles
could be effectively controlled when geomembranes are covered
at appropriate time and with sufficient restrained area (Chappel
et al., 2012; Rowe et al., 2012). Overburden pressure caused by
solid wastes mass as well as overlying drainage layers greatly
contributes to the reduction of wrinkles (Dickinson and
Brachman, 2006; Brachman and Gudina, 2008a,b). With good
construction quality assurance, the majority of the geomembrane
will be intact and in good direct contacted to subgrade and
diffusion may be the dominated mechanism controlling the
transport of organic contaminant beneath landfill. Therefore, the
ability of geomembrane (GMB) to control the transport of organic
contaminant requires quantification, and the capability of un-
derlying soil layers to block contaminant diffusion needs careful
consideration.

http://dx.doi.org/10.1016/j.geotexmem.2014.06.007
0266-1144/© 2014 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).
 T. Liu, L. Hu / Geotextiles and Geomembranes 42 (2014) 428e436
Several analytical and numerical methods have been proposed
to investigate the contaminant transport in composite liners
(Foose, 2002, 2010; Foose et al., 2002; Rowe et al., 2004; Rowe and
Booker, 2005; El-Zein and Rowe, 2008; Chen et al., 2009; El-Zein
et al., 2012; Xie et al., 2013). A widely used assumption is that
soils beneath GMB are totally saturated. This assumption leads to
conservative prediction of contaminant transport in soils, and the
results are applied in engineering design. However, field research
results (Bonaparte and Gross, 1990; King et al., 1993) showed that a
zone of partial saturation is a commonplace occurrence beneath a
landfill. In western areas of China, groundwater level is low all the
year. The unsaturated zone beneath the surface is quite thick and
the soil layers are in low water content (Yang et al., 2012). Even for a
saturated compacted clay liner, Celik et al. (2009) demonstrated
that failure to consider the presence of an unsaturated zone
beneath a clay liner could significantly underestimate the leakage
through the liner. Normally, in landfill site selection and liner sys-
tem design, a distance between landfill liner and groundwater level
is required in order to minimize the contamination risk to
groundwater. The potential effect of unsaturated zone on contam-
inant transport beneath landfills requires adequate evaluation.
Typical analytical and numerical methods mainly focus on perfor-
mance of composite liners, but fail to address the effectiveness of
the composite liners in terms of ultimate impact on groundwater
quality (Foose, 2010). Computational tools, such as POLLUTE (Rowe
and Booker, 2005) and SPAS (El-Zein et al., 2012), are widely
recognized as robust tools in dealing with the complexities and
variations in landfill liner system. However, the potential effect of
unsaturated condition for liner soils is not systematically
investigated.
The effect of water content on aqueous-phase diffusion in soils
was initially studied by soil scientists concerning nutrient supply,
but it remained largely unexamined by geotechnical engineers in
the context of contaminant transport and waste disposal (Hu and
Wang, 2003). Several theoretical models were developed to quan-
titatively describe the diffusion behavior of solute in unsaturated
soils (Lim et al., 1998; Revil and Jougnot, 2008). Although diffi-
culties in matrix control are often encountered in laboratory tests
on unsaturated soils, extensive laboratory studies concerning so-
lute transport in unsaturated soils have been conducted (Rowe and
Badv, 1996; Hutchison et al., 2003; Badv and Faridfard, 2005;
Hamamoto et al., 2009, 2012; Savoye et al., 2010, 2012; Chou
et al., 2012). Theoretical and experimental progress greatly helps
to quantify and harness the slow diffusion process in unsaturated
soils. A review of the unsaturated diffusion models describing
relation between effective diffusion coefficient and volumetric
water content was presented by Hu and Wang (2003). Chou et al.
(2012) evaluated the solute diffusion models with tortuosity fac-
tor based on experiments on sand, sandy loam, and clay under
varied degree of saturation, and concluded that the Millington and
Quirk (1961) model fitted best with the experimental results. It is
clear that the diffusion coefficient in aqueous phase would be
significantly reduced as soils become unsaturated. Factors attrib-
uted to this reduction may include: (1) increase in diffusion path
length; (2) increased viscosity of liquid next to the soil surface; (3)
increased ionic interaction along small pores and water films; (4)
interconnectivity between water filled pores and water films (Lim
et al., 1998). The pure diffusion process is highly related to the
variation of water content and diffusion coefficient, thus the
dependence of unsaturated diffusion coefficient on water content
needs to be sufficiently considered in the contaminant prediction.
A numerical model was developed to investigate the influence
of water content variation and the unsaturated diffusion on the
transport of organic contaminant with low volatility and high
aqueous solubility beneath a landfill. Acetic acid, an organic
429
compound with relatively high concentration in landfill leachate in
China, was chosen as the typical organic compound (Kjeldsen et al.,
2002; Yue et al., 2007). A typical composite liner system consisted
of a geomembrane and a compacted clay liner underlain by an
attenuation layer was investigated. Three kinds of attenuation soil
layer: sand layer, sandy clay loam layer and clay layer were sepa-
rately studied. The Millington and Quirk model was employed to
describe the non-linear relationship between effective diffusion
coefficient and volumetric water content. Three calculation condi-
tions, i.e., totally saturated, unsaturated condition with or without
considering unsaturated diffusion model, were investigated.
2. Mathematical model
The governing geometry of the problem is shown in Fig. 1. A
compacted clay liner and an attenuation layer are beneath the
geomembrane. The groundwater level is at the base of the atten-
uation layer, and the flow is assumed to be in horizontal direction.
Lg, Ls1, Ls2 and h are thickness of geomembrane, compacted clay
liner, attenuation layer, and aquifer, respectively. The origin of the
coordinate system is at the interface between the geomembrane
and the underlying compacted clay liner. The positive z-axis is
assumed downward. The thickness of the compacted clay liner (Ls1)
is defined as 0.75 m, which is the minimum thickness prescribed in
the technical code by Chinese government (Chinese Ministry of
Construction, 2007). The thickness of the attenuation layer (Ls2) is
3 m, so that the total distance between the top of liner and un-
derlying receptor aquifer is 3.75 m, which is the minimum distance
allowed under Ontario Ministry of the Environment (1998). The
aquifer is assumed to be 1.0 m deep.
In engineering practice, soils at landfill site would vary in a wide
range. In order to investigate the influence of soil characteristics on
the transport and tempo-spacial distribution of contaminant, three
kinds of underlying attenuation layer, i.e., sand layer, sandy clay
loam layer, and clay layer, are separately investigated and are
denoted as Scenarios 1, 2, and 3, respectively, in the later discussion.
Acetic acid was chosen to represent the organic compound in
landfill leachate. The concentration of acetic acid (and other fatty
acids or carboxylic acids) in leachate could be significant, especially
in the leachate produced by MSW with high organic content and in
the relatively early stage of leachate production (Kjeldsen et al.,
2002). Yue et al. (2007) investigated the organic component in
young landfill and presented that concentration of pentanoic acid,
hexanoic acid and several other carboxylic acids could exceed
1000 mg/L, and the concentration of acetic acid could even reach
4286 mg/L. The volatility of acetic acid is relatively low. Under
Fig. 1. Schematic diagram of diffusion through intact composite soil liner.
430 T. Liu, L. Hu / Geotextiles and Geomembranes 42 (2014) 428e436

normal temperature and air pressure (298.5 K, and one standard
atmospheric pressure), the ratio of gas-phase and aqueous-phase
concentration Cg/Ca (based on Henry’s Law) is in the order 106,
which means the amount of acetic acid volatilizing in air is small
compared to the amount dissolved in water. Thus this paper con-
siders the transport in the aqueous phase while acknowledging
that there will also be some transport in the air phase that may
become more significant at low degrees of saturation. Diffusion in a
geomembrane can be described by Fick's second law:
An aquifer with a constant horizontal groundwater velocity was
set beneath the attenuation layer. It is assumed that the concen-
tration of contaminant in the aquifer does not vary in vertical or
horizontal direction, and the contaminant transport in this layer is
dominated by advection. The net change of contaminant mass in
the aquifer at any time t will be equal to the difference between the
contaminant mass flux entering the aquifer from above and that
leaving the aquifer from the side. The contaminant concentration in
the aquifer CBN(t) is given by (Rowe and Booker, 2005):

vCg v2 C g Zt Zt
¼ Dg 2 ; Lg < z < 0 (1) fBN ðc; tÞ vA cBN
vt vz CBN ðtÞ ¼ dt  dt; z ¼ H (6)
nA h nA L
0 0
where Cg is the contaminant concentration in the geomembrane; t
is time; Dg is the diffusion coefficient. Biological decay and vCr
adsorption is neglected. fBN ðc; tÞ ¼ D*e ; z¼H (7)
vz
For the direct contact situation analyzed here, continuity of
contaminant concentration and mass flux exists at the interface of where fBN(c, t) is the contaminant mass flux into the aquifer from
the geomembrane and soil liner. Continuity of contaminant flux can the soil above; vA is the advection flow velocity in the base aquifer;
be described using Fick's first law: h is the height of the aquifer; nA is the porosity of the aquifer, and L
is the length of landfill. Parameters involved in the calculation are
vCg vCr summarized in Table 1.
Jd ¼ Dg j  ¼ De j þ; z ¼ 0 (2)
vz z¼0 vz z¼0

where Jd is the mass flux via diffusion. 3. Water content distribution in composite liner and
Continuity of concentration interface can be described by: attenuation layer
Cg
j  ¼ Cr jz¼0þ ; z ¼ 0 (3) For landfill liners constructed with adequate care and supervi-
Sgf z¼0
sion, geomembrane could provide sufficient separation between
landfill leachate and underlying soils. Fityus et al. (1999) demon-
where Cr is the contaminant concentration in the pore water of the
strated that when leakage through geomembrane defects was ab-
soil liner; Cg is divided by the partition coefficient of the geo-
sent, time required for moisture equilibration in compacted clay
membrane (Sgf) because the solute partitions out of the geo-
membrane at the interface. In this study, a normalization method
proposed by Foose (2002) is employed to deal with the change of Table 1
concentration at the surface of geomembrane. The partition coef- Geometric and material parameters involved in the simulation.
ficient Sgf is defined as 0.021, referring the experiment results Landfill leachate (see Eq. (5))
compiled in Rowe et al. (2004). Initial concentration, C0 (mg/L) 4000
Chemical transport process is attributed to Fickian pure diffu- Half-life, t1/2 (year) 2
sion process and biological decay, and sorption is assumed to be Geomembrane (GMB) (see Eq. (1))
negligible in all the soil layers. The diffusion of the organic con- Thickness, Lg (mm) 1.5
taminants through soil liners considering first-order degradation Partition coefficient, Sgfa 0.021
Diffusion coefficient, Dg (m /s) 1.1  1012
2 a
can be described by Fick's second law:
Compacted clay liner (see Eq. (4))
vCr v2 C r Thickness, Ls1 (m) 0.75
q ¼ De 2  qlCr ; z > 0 (4) Effective porosity, n1 0.38
vt vz Half-life, t1/2 (year) 50

where q is the volumetric water content; l is the first-order decay Attenuation layer (see Eq. (4))
Thickness, Ls2 (m) 3
coefficient. D*e is defined as actual diffusion coefficient here, and is
Effective porosity, n2 0.43 (sand) 0.39 (sandy clay loam) 0.38 (clay)
used to describe the diffusive ability of contaminant through Half-life, t1/2 (year) 50
porous media.
Base aquifer (see Eq. (6))
A decaying concentration is applied as the top boundary, and is Thickness h (m) 1
given by (Rowe and Booker, 2005): Effective porosity nA 0.3
Darcy velocity vA (m/s) 3.17  108
Zt
Landfill (see Eq. (6))
CðtÞ ¼ C0  l CðtÞdt (5) Landfill length, L (m) 200
0 Unsaturated conditionb (see Eq. (10))
Van Genuchten model, a(1/m) 14.5 (sand) 5.9 (sandy clay loam)
where C0 is the initial source concentration in leachate; l is the 0.8 (clay)
first-order decay coefficient, and l ¼ ln2/t1/2, in which t1/2 is the Van Genuchten model, n 2.68 (sand) 1.48 (sandy clay loam )
1.09 (clay)
half-life of contaminant. In a realistic situation, the contaminant
Van Genuchten model, m 0.627 (sand) 0.324 (sandy clay loam)
concentration in leachate fluctuates widely over time, but there is a 0.083 (clay)
lack of field data from a wide variety of modern landfills to develop Residual water content, qr 0.068 (sand) 0.1 (sandy clay loam)
a more complex boundary condition. For the organic contaminant 0.045 (clay)
studied here, the decaying concentration applied would partially a
Rowe et al. (2004).
b
represent the variation of concentration with time. Carsel and Parrish (1988).
 T. Liu, L. Hu / Geotextiles and Geomembranes 42 (2014) 428e436
liner was considerably shorter than the time scale normally
considered for contaminant transport beneath a landfill. Therefore,
a steady volumetric water content profile is adopted in this anal-
ysis. Even when geomembrane defects are involved, which is
inevitable in the short and long service life of composite liner, the
leakage of landfill leachate through geomembrane defects could be
limited into a relatively small area compared to the whole scale of
the geomembrane. Therefore, the assumption of stable water
content profile in the soil liner and attenuation layer is a reasonable
approximation for the condition analyzed here. The suction head in
the soil is determined by the distance between the soil and the
groundwater level. The volumetric water content of soil between
geomembrane and groundwater level is obtained by the van Gen-
uchten model described as follows (Van Genuchten, 1980):

qr þ Se ðqs  qr Þ Hp < 0
q¼
qs Hp  0
8 Fig. 3. Actual diffusion coefficient profile in Case C in three scenarios.
> 1
<  n m H < 0
Se ¼ 1 þ aHp  p
>
: Hp  0
1
where Hp is the pressure head, Se is the effective degree of satu-
ration, and a, m, n are van Genuchten model parameters. The soil
parameters used in the calculation are given in Table 1.
4. Unsaturated diffusion model
When dealing with solute transport process in unsaturated soils,
a common approach is to substitute the effective porosity (n) with
the volumetric water content (q). The underlying assumption is that
the diffusion coefficient varies linearly with volumetric water
content. However, extensive research has proven that the effect of
water content on diffusion is not linear, and the applicability of
diffusion coefficient measured in saturated systems within unsat-
urated media is questionable (Hu and Wang, 2003). In this analysis,
Millington and Quirk (1961) model is employed to quantitatively
describe the relationship between diffusion coefficient and volu-
metric water content due to its simplicity and satisfying fitness
with experimental results for a wide range of soils (Chou et al.,
2012). Relationship between the actual diffusion coefficient (D*e)
and volumetric water content (q) can be described as follows.
Fig. 2. Volumetric water content profile in three scenarios.
431
(8)
(9)
D*e ¼ q10=3 q2
s D0 (10)
where D*e is the actual diffusion coefficient in soil; D0 is the solute
diffusion coefficient in pure water; qs is the saturated water content.
Three conditions are analyzed here to investigate the effect of
volumetric water content and unsaturated diffusion on contami-
nant transport and distribution. In Case A, the totally saturated
condition is analyzed, and the actual diffusion coefficient (D*e) is
derived by substituting q with qs in Eq. (10). In Case B, the variation
of volumetric water content along depth is considered, and the
actual diffusion coefficient (D*e) is obtained by multiplying the D*e in
Case A with the degree of saturation (q/qs). In Case C, the non-linear
relationship between diffusion coefficient and volumetric water is
considered, and the actual diffusion coefficient is calculated directly
from Eq. (10).
Volumetric water content profile in the composite liner and
underlying attenuation layer in three scenarios is presented in
Fig. 2. Comparison of the actual diffusion coefficient in Case C is
shown in Fig. 3. The results show that in the sand attenuation
layer (Scenario 1), the volumetric water content decreases
sharply from the saturated water content (qs ¼ 0.43) at the bot-
tom (z ¼ 3.75 m) to the residual water content (qr ¼ 0.068) at the
top (z ¼ 0.75 m). This leads to a sharp decrease in the actual
diffusion coefficient D*e, from 2.75  1010 m2/s at the bottom
(z ¼ 3.75 m) to 6.07  1013 m2/s at the top (z ¼ 0.75 m). While in
the clay attenuation layer (Scenario 3), the volumetric water
content decreases slightly, from 0.380 at the bottom (z ¼ 3.75 m)
to 0.346 at the top (z ¼ 0.75 m), and the corresponding actual
diffusion coefficient D*e decreases from 2.33  1010 m2/s to
1.77  1010 m2/s. In the sandy clay loam attenuation layer
(Scenario 2), the distribution of the volumetric water content and
the actual diffusion coefficient lies in between the above two
scenarios.
5. Results and analysis
In order to investigate the effect of water content variation and
unsaturated diffusion process on contaminant transport beneath
landfills, three types of attenuation layer (Scenarios 1, 2, and 3) and
three kinds of calculation condition (Cases A, B, and C) are calcu-
lated and analyzed below. Detailed description of the scenarios and
cases is summarized in Table 2.
432 T. Liu, L. Hu / Geotextiles and Geomembranes 42 (2014) 428e436
Table 2
Detailed description of the soil layers and calculation conditions.
Case A Case B
Scenario 1 Sand AL, totally saturated condition Sand AL, D*e varies linearly with q
Scenario 2 Sandy clay loam AL, totally saturated Sandy clay loam AL, D*e varies linearly
condition with q
Scenario 3 Clay AL, totally saturated condition Clay AL, D*e varies linearly with q
5.1. Scenario 1: composite liner underlain by a sand layer
Fig. 4 shows the comparison of acetic acid concentration in soil
layers in the three cases. The influence of water content variation
and unsaturated diffusion process on acetic acid transport is sig-
nificant. In Case A, when both the compacted clay liner and the
sand attenuation layer are assumed saturated, the acetic acid dis-
tributes evenly along the depth and the peak concentration reaches
7.20 mg/L at depth z ¼ 1.97 m. In Case B, when the actual diffusion
coefficient varies linearly with water content, the peak concentra-
tion reaches 6.71 mg/L at depth z ¼ 2.12 m. A slight concentration
gradient discontinuity exists at the interface of the compacted clay
liner and the attenuation layer due to the difference of volumetric
water content. In Case C, when the non-linear relationship between
actual diffusion coefficient and the volumetric water content are
considered, the peak acetic acid concentration reaches 9.46 mg/L at
depth z ¼ 0.94 m. In this case, the contaminant is largely confined
within the depth z < 1.0 m, and the concentration of acetic acid at
the top of aquifer is zero at t ¼ 100 years. As shown in Figs. 2 and 3,
both the volumetric water content and effective diffusion coeffi-
cient in most part of the sand attenuation layer stay at relatively
low values, and the unsaturated zone serves as an effective barrier
to contaminant diffusion. Since the degradation of acetic acid is
considered in leachate source and soil layers, the acetic acid con-
centration decreases quickly with time. Thus, the acetic acid con-
centration at the bottom of attenuation layer keeps zero even when
the calculation time is extended to several hundreds of years, which
means that acetic acid will not reach the aquifer in this case.
The acetic acid concentration profile presented above could also
be well explained by the contaminant mass flux rate at depth
z ¼ 0.75 m, as shown in Fig. 5. In all the three cases, the mass flux
rate increases from zero to a peak value at about 6.5 years. Obvi-
ously, the mass flux in Case A is higher than that in Case B and Case
C. The peak mass flux rate reaches 71.2 kg/ha/year in Case A, and
that is 3.5 times higher than that in Case B and is about 30 times
Fig. 4. Calculated acetic acid concentration at t ¼ 100 years in Scenario 1.
Case C
Sand AL, unsaturated diffusion model considered
Sandy clay loam AL, unsaturated diffusion model considered
Clay AL, unsaturated diffusion model considered
higher than that in Case C. In all cases, the mass flux reduces to
values below zero at about 40 years since degradation of organic
contaminant is considered in leachate and soil layers. The results
show that unsaturated zone significantly impedes contaminant
diffusion and effectively reduces contaminant flux into underlying
soil layers.
The comparison of cumulative mass in the three cases is shown
in Fig. 6. At depth z ¼ 3.75 m, the calculated acetic acid mass within
100 years in Case A and Case B is 25.8 kg/ha and 12.5 kg/ha,
respectively. In Case C, the cumulative mass flux at z ¼ 3.75 m is
zero within 100 years, indicating that the contaminant transport is
effectively blocked in the upper unsaturated zone. At depth
z ¼ 0.75 m, the cumulative acetic acid mass flux increases with
time, and reaches peak values at about 22 years, then gradually
decreases values below zero due to organic degradation in the
leachate source and soil layers. The cumulative mass discharged
within 100 years in Case A, Case B, and Case C is 435.7 kg/ha,
102.8 kg/ha, and 12.4 kg/ha, respectively. The comparison of Case B
and Case C indicates that the consideration of the effect of water
content on aqueous diffusion will lead to 8.3 times difference in
cumulative mass flux, and the saturated assumption significantly
overestimate the mass flux.
5.2. Scenario 2: composite liner underlain by a sandy clay loam
layer
As shown in Fig. 2, the volumetric water content distribution
profile in the sandy clay loam layer is in-between the sand and clay
layer. The volumetric water content decreases from 0.346 at the
bottom of the attenuation layer to 0.173 at the top, and the decrease
is much slighter than that in Scenario 1. Calculated acetic acid
concentration in the soil layers at 100 years is presented in Fig. 7,
and the variation of cumulative mass flux with time at depth
z ¼ 0.75 m and z ¼ 3.75 m is shown in Fig. 8.
Fig. 5. Mass flux rate with time at depth z ¼ 0.75 m in Scenario 1.
 T. Liu, L. Hu / Geotextiles and Geomembranes 42 (2014) 428e436 433

(Case C), the cumulative mass flux will be about 200 times less
compared to those in Case A and Case B. At depth z ¼ 0.75 m, the
cumulative mass flux in Case A, Case B, and Case C is 406.2 kg/ha,
226.6 kg/ha and 94.2 kg/ha, respectively. In the three cases, the
volumetric water content and actual diffusion coefficient in the
upper compacted clay liner is quite close due to its high water
retention capacity. However, when different calculation conditions
are applied to the underlying attenuation layer, the mass flux
through the upper clay liner varies greatly. Therefore, the proper-
ties of underlying soils could greatly influence the transport and
distribution of contaminant in the upper liner, and that requires
adequate consideration in the performance evaluation of compos-
ite liners.

5.3. Scenario 3: composite liner underlain by a clay layer

In this scenario, all the calculation parameters for the underlying

clay attenuation layer are the same as the upper compacted clay
liner. The acetic acid concentration profile at 100 years is shown in
Fig. 9. Compared with Scenarios 1 and 2, the acetic acid distribution
profile in this Scenario is quite smooth in all the three cases, and the
influence of water content distribution and unsaturated diffusion
process is much less obvious.

Fig. 6. Cumulative mass discharged with time in Scenario 1: (a) bottom of the com-
pacted clay liner, z ¼ 0.75 m; and (b) bottom of the attenuation layer, z ¼ 3.75 m.

At depth z ¼ 3.75 m, the cumulative mass flux within 100 years

in Cases A, B and C is 24.2 kg/ha, 19.2 kg/ha and 8.79  102 kg/ha,
respectively. Consideration of the variation of water content in the
soil layers only produces a slight difference between Case A and
Case B. However, when the unsaturated diffusion model is used

Fig. 8. Cumulative mass discharged in Scenario 2: (a) bottom of the compacted clay
Fig. 7. Calculated acetic acid concentration at t ¼ 100 years in Scenario 2. liner, z ¼ 0.75 m; and (b) bottom of the attenuation layer, z ¼ 3.75 m.
434 T. Liu, L. Hu / Geotextiles and Geomembranes 42 (2014) 428e436
Fig. 9. Calculated acetic acid concentration at t ¼ 100 year in Scenario 3.
The cumulative mass discharged in the three cases is presented
in Fig. 10. At depth z ¼ 0.75 m, the cumulative mass flux in Case A
and Case B is close, and is about 40% higher than that in Case C. The
volumetric water content declines from 0.33 at the bottom of the
Fig. 10. Cumulative mass discharged in Scenario 3: (a) bottom of the compacted clay
liner, z ¼ 0.75 m; and (b) bottom of the attenuation layer, z ¼ 3.75 m.
attenuation layer to 0.30 at the top of the compacted clay liner. The
cross-sectional area available for mass transfer is much larger than
that in unsaturated sand layer due to high water content in the
compacted clay liner. The results show that in this scenario, the
assumption of totally saturated condition will not result in obvious
deviation in contaminant transport prediction beneath landfill. For
composite liner underlain by clay attenuation layers, the design and
performance evaluation of landfill liners based on saturated
assumption is convenient and acceptable.
5.4. Comparison and discussion of three scenarios
The concentration profiles along the depth at 100 years are
shown in Figs. 4, 7 and 9, and the peak concentration values and the
corresponding depths in different scenarios and cases are sum-
marized in Table 3. In Case A, when all soil liners are assumed to be
saturated, the concentration profiles in all three scenarios are quite
close, and the peak concentration in Scenarios 1, 2, and 3 is 7.19 mg/
L, 7.43 mg/L, and 7.47 mg/L, respectively. However, in Case C, when
the dependence of actual diffusion coefficient on volumetric water
content is sufficiently considered, the peak concentration in Sce-
narios 1, 2, and 3 is 9.45 mg/L, 8.40 mg/L, and 7.36 mg/L, respec-
tively. In this case, contaminant diffusion is largely blocked in the
upper part of the attenuation layer, and the corresponding depth of
peak concentration in Case C are relatively shallower than those in
the other two cases.
The difference of acetic acid transport in soil types could be
further illustrated by the comparison of mass flux through typical
depths, as shown in Figs. 6, 8 and 10. The cumulative mass flux
within 100 years is presented in Table 4. In saturated condition
(Case A), the values of cumulative contaminant mass in the three
scenarios are close, and the difference is caused by the difference in
the saturated water content of the three soil types. When unsatu-
rated condition is considered (Case B), the reduction of cumulative
contaminant mass at depth z ¼ 0.75 m in Scenario 1 and 2 could
reach 76.4% and 44.2% respectively, and a further 97.2% and 76.8%
reduction could be reached when unsaturated diffusion model is
applied (Case C). Comparison between Cases B and C in all the three
scenarios indicates that the assumption of a linear relation between
volumetric water content and actual diffusion coefficient would
greatly deviate from the actual condition, especially for soil layers
with low water retention capacity. While for a clay attenuation
layer, this deviation is much smaller, and the contaminant transport
could be reasonably predicted based on the linear relationship
assumption.
6. Discussion
In the above part, acetic acid was chosen as a representative of
the organic compounds in landfill leachate. Many carboxylic acids,
such as acetic acid, have very low volatility, and their diffusion in
gas-phase could be neglected since there is little mass transfer from
aqueous phase to air phase. In developing countries such as China
and India, the municipal solid waste (MSW) contains a significant
Table 3
Peak concentration values and corresponding depths.
Peak concentration Depth for Peak concentration
(Unit: mg/L) (Unit: m)
Case A Case B Case C Case A Case B Case C
Scenario 1 7.19 6.71 9.45 1.96 2.11 0.94
Scenario 2 7.43 7.11 8.40 2.08 1.93 1.31
Scenario 3 7.47 7.54 7.36 2.08 1.89 1.66
 T. Liu, L. Hu / Geotextiles and Geomembranes 42 (2014) 428e436
Table 4
Cumulative mass flux in 100 years at depth z ¼ 0.75 m and depth z ¼ 3.75 m (Unit:
kg/ha).
Depth z ¼ 0.75 m Depth z ¼ 3.75 m
Case A Case B Case C Case A Case B Case C
Scenario 1 435.7 102.8 12.4 25.8 12.5 0
Scenario 2 406.2 226.6 94.2 24.2 19.2 8.79  102
Scenario 3 398.5 359.4 299.4 23.9 21.5 14.7
proportion of organic compounds and moisture, which provides a
good environment for leachate production from biodegradation
(Chen et al., 2013). The anaerobic acid phase of leachate production
could last for quite long period of time, and the carboxylic acids
could remain in significantly high concentration (could reach
several orders of magnitude in mg/L). The analysis results pre-
sented above show that concentration and accumulated mass of
acetic acid in composite liner and underlying soils is very high, and
large amount of pollutants could reach the aquifer. Therefore, the
high source concentration of carboxylic acids needs to be consid-
ered, In fact, this high organic compound concentration in leachate
has become one of the major concerns in the design of liner system
and leachate collection system in Chinese landfills, and aqueous-
phase diffusion of such organic compounds deserves adequate
evaluation.
It is noted that there are many volatile organic compounds
(VOCs) in landfill leachate such as mononuclear aromatics, ketones,
and aliphatic compounds with chlorine, etc. These VOCs diffuse
readily in geomembrane, and have high volatility from aqueous
phase to air phase. Hence, considerable VOCs will enter into air
phase when underlying soils beneath geomembrane are in unsat-
urated condition. Since the diffusion coefficient in air phase is many
orders of magnitude higher than that in aqueous phase, most of
these VOCs will migrate in air phase, and have little impact on
groundwater quality. For such case, their migration in aqueous
phase is trivial, and the transport in air phase should be adequately
addressed. The mathematical model and numerical simulation
should be very different from the above-mentioned analysis, which
is suitable for low volatile and highly soluble organics. However, it
is out of the scope of this paper.
7. Summary and conclusions
Numerical model was developed to analyze organic contami-
nant transport through partially saturated soils beneath a landfill.
The main conclusions are as follows.
(1) Characteristics of soil layer beneath composite liners signif-
icantly influence the diffusive transport of organic contami-
nant. Unsaturated sand attenuation layer could serve as
excellent diffusion barrier due to its low water retention
capacity. Organic contaminant could be effectively retarded
in the upper part of soil layer, and thus contaminant flux into
the bottom aquifer is reduced significantly.
(2) The consideration of volumetric water content variation in
soil layers significantly influences the contaminant transport
in both compacted clay liner and underlying attenuation
layer. For the sand attenuation layer and the sandy clay
attenuation layer (Scenarios 1 and 2), when the variation of
water content is considered (Case B), the contaminant flux
through depth z ¼ 0.75 m decreases by 76.4% and 44.2%,
respectively, comparing to those in the totally saturated
condition (Case A).
435
(3) Actual diffusion coefficient is highly dependent on volu-
metric water content. Unsaturated diffusion models capable
of capturing the relationship between diffusion coefficient
and volumetric water content enables a more reasonable
prediction of contaminant transport and distribution in un-
saturated porous media.
Acknowledgments
The authors would like to express their sincere gratitude to the
financial supports from National Basic Research Program of China
(Grant No. 2012CB719804), State Key Laboratory of Hydro-Science
and Engineering (SKLHSE-2012-KY-01, 2013-D-01), and National
Natural Science Foundation of China (Project No. 51323014).
References
Badv, K., Faridfard, M.R., 2005. Laboratory determination of water retention and
diffusion coefficient in unsaturated sand. Water Air Soil Pollut. 161, 25e38.
Bonaparte, R., Gross, B.A., 1990. Field behavior of double-liner systems, waste
containment systems: construction, regulation, and performance. In:
Geotechnical Special Publication, pp. 52e83.
Brachman, R.W.I., Gudina, S., 2008a. Geomembrane strains from coarse gravel and
wrinkles in a GM/GCL composite liner. Geotext. Geomembr. 26 (6), 488e497.
Brachman, R.W.I., Gudina, S., 2008b. Gravel contacts and geomembrane strains for a
GM/CCL composite liner. Geotext. Geomembr. 26 (5), 448e459.
Carsel, R.F., Parrish, R.S., 1988. Developing joint probability distributions of soil
water retention characteristics. Water Resour. Res. 24, 755e769.
Celik, B., Rowe, R.K., Unlu, K., 2009. Effect of vadose zone on the steady-state
leakage rates from landfill barrier systems. Waste Manage. 29, 103e109.
Chappel, M., Rowe, R., Brachman, R., Take, W., 2012. A comparison of geomembrane
wrinkles for nine field cases. Geosyn. Int. 19, 453e469.
Chen, Y.M., Xie, H.J., Ke, H., Chen, R.P., 2009. An analytical solution for one-
dimensional contaminant diffusion through multi-layered system and its ap-
plications. Environ. Geol. 58, 1083e1094.
Chen, Y.M., Zhan, L.T., Xu, X.B., Liu, H.L., 2013. Geo-environmental problems in
landfills of MSW with high organic content. In: Delage, P., Desrues, J., Frank, R.,
Puech, A., Schlosser, F. (Eds.), 18th International Conference on Soil Mechanics
and Geotechnical Engineering, Paris, France, pp. 3014e3017.
Chinese Ministry of Construction, 2007. Technical Code for Liner Systems of
Municipal Solid Waste Landfill (CJJ 113-2007). China Building Industry Press,
Beijing.
Chou, H.Y., Wu, L.S., Zeng, L.Z., Chang, A., 2012. Evaluation of solute diffusion tor-
tuosity factor models for variously saturated soils. Water Resour. Res. 48.
Dickinson, S., Brachman, R.W.I., 2006. Deformations of a geosynthetic clay liner
beneath a geomembrane wrinkle and coarse gravel. Geotext. Geomembr. 24,
285e298.
Edil, T.B., 2003. A review of aqueous-phase VOC transport in modern landfill liners.
Waste Manage. 23, 561e571.
El-Zein, A., McCarroll, I., Touze-Foltz, N., 2012. Three-dimensional finite-element
analyses of seepage and contaminant transport through composite geo-
synthetics clay liners with multiple defects. Geotext. Geomembr. 33, 34e42.
El-Zein, A., Rowe, R., 2008. Impact on groundwater of concurrent leakage and
diffusion of dichloromethane through geomembranes in landfill liners. Geosyn.
Int. 15, 55e71.
Fityus, S.G., Smith, D.W., Booker, J.R., 1999. Contaminant transport through an un-
saturated soil liner beneath a landfill. Can. Geotech. J. 36, 330e354.
Foose, G.J., 2002. Transit-time design for diffusion through composite liners.
J. Geotech. Geoenviron. Eng. 128, 590e601.
Foose, G.J., 2010. A steady-state approach for evaluating the impact of solute
transport through composite liners on groundwater quality. Waste Manage. 30,
1577e1586.
Foose, G.J., Benson, C.H., Edil, T.B., 2002. Comparison of solute transport in three
composite liners. J. Geotech. Geoenviron. Eng. 128, 391e403.
Hamamoto, S., Moldrup, P., Kawamoto, K., Komatsu, T., 2012. Maxwell's law based
models for liquid and gas phase diffusivities in variably-saturated soil. Soil Sci.
Soc. Am. J 76, 1509e1517.
Hamamoto, S., Perera, M.S.A., Resurreccion, A., Kawamoto, K., Hasegawa, S.,
Komatsu, T., Moldrup, P., 2009. The solute diffusion coefficient in variably
compacted, unsaturated volcanic ash soils. Vadose Zone J. 8, 942e952.
Haxo, H.E., 1990. Determining the transport through geomembranes of various
permeants in different applications. In: ASTM Special Technical Publication, vol.
1081, pp. 75e94.
Hu, Q.H., Wang, J.S.Y., 2003. Aqueous-phase diffusion in unsaturated geologic me-
dia: a review. Crit. Rev. Environ. Sci. Technol. 33, 275e297.
Hutchison, J.M., Seaman, J.C., Aburime, S.A., Radcliffe, D.E., 2003. Chromate trans-
port and retention in variably saturated soil columns. Vadose Zone J. 2,
702e714.
436 T. Liu, L. Hu / Geotextiles and Geomembranes 42 (2014) 428e436
King, K.S., Quigley, R., Fernandez, F., Reades, D., Bacopoulos, A., 1993. Hydraulic
conductivity and diffusion monitoring of the Keele Valley Landfill liner, Maple,
Ontario. Can. Geotech. J. 30, 124e134.
Kjeldsen, P., Barlaz, M.A., Rooker, A.P., Baun, A., Ledin, A., Christensen, T.H., 2002.
Present and long-term composition of MSW landfill leachate: a review. Crit.
Rev. Environ. Sci. Technol. 32, 297e336.
Lim, P., Barbour, S., Fredlund, D., 1998. The influence of degree of saturation on the
coefficient of aqueous diffusion. Can. Geotech. J. 35, 811e827.
McWatters, R., Rowe, R., 2010. Diffusive transport of VOCs through LLDPE and two
coextruded geomembranes. J. Geotech. Geoenviron. Eng. 136, 1167e1177.
Millington, R., Quirk, J., 1961. Permeability of porous solids. Trans. Faraday Soc. 57,
1200e1207.
Ontario Ministry of the Environment, 1998. Landfill Standards: A Guideliner on the
Regulatory and Approval Requirements for New or Expanding Landfill Sites.
Ontario Ministry of the Environment. Ontario Regulation 232/98.
Park, J.K., Nibras, M., 1993. Mass flux of organic chemicals through polyethylene
geomembranes. Water Environ. Res. 65, 227e237.
Revil, A., Jougnot, D., 2008. Diffusion of ions in unsaturated porous materials.
J. Colloid Interf. Sci. 319, 226e235.
Rowe, R.K., 2005. Long-term performance of contaminant barrier systems. Geo-
technique 55, 631e677.
Rowe, R.K., 2012. Short- and long-term leakage through composite liners. Can.
Geotech. J. 49, 141e169.
Rowe, R.K., Badv, K., 1996. Advective-diffusive contaminant migration in unsatu-
rated sand and gravel. J. Geotech. Eng. ASCE 122, 965e975.
Rowe, R.K., Booker, J.R., 2005. POLLUTEv7: Pollutant Migration Through Nonho-
mogeneous Soil, 1993e2005. Distributed by GAEA Environmental Engineering
Ltd, 87 Garden Street, Whitby, Ontario, Canada LIN 9E7.
Rowe, R.K., Chappel, M.J., Brachman, R.W.I., Take, W.A., 2012. Field study of wrinkles
in a geomembrane at a composite liner test site. Can. Geotech. J. 49, 1196e1211.
Rowe, R.K., Quigley, R.M., Brachman, R.W., Booker, J.R., 2004. Barrier Systems for
Waste Disposal Facilities. E & FN Spon, Taylor & Francis Books Ltd, London,
p. 587.
Sangam, H.P., Rowe, R.K., 2001. Migration of dilute aqueous organic pollutants
through HDPE geomembranes. Geotext. Geomembr. 19, 329e357.
Sangam, H.P., Rowe, R.K., 2005. Effect of surface fluorination on diffusion through a
high density polyethylene geomembrane. J. Geotech. Geoenviron. Eng. 131,
694e704.
Savoye, S., Michelot, J.L., Matray, J.M., Wittebroodt, C., Mifsud, A., 2012. A laboratory
experiment for determining both the hydraulic and diffusive properties and the
initial pore-water composition of an argillaceous rock sample: a test with the
Opalinus clay (Mont Terri, Switzerland). J. Contam. Hydrol. 128, 47e57.
Savoye, S., Page, J., Puente, C., Imbert, C., Coelho, D., 2010. New experimental
approach for studying diffusion through an intact and unsaturated medium: a
case study with callovo-oxfordian argillite. Environ. Sci. Technol. 44, 3698e3704.
Van Genuchten, M.T., 1980. A closed-form equation for predicting the hydraulic
conductivity of unsaturated soils. Soil Sci. Soc. Am. J. 44, 892e898.
Xie, H., Lou, Z., Chen, Y., Jin, A., Zhan, T.L., Tang, X., 2013. An analytical solution to
organic contaminant diffusion through composite liners considering the effect
of degradation. Geotext. Geomembr. 36, 10e18.
Yang, L., Wei, W., Chen, L.D., Mo, B.R., 2012. Response of deep soil moisture to land
use and afforestation in the semi-arid Loess Plateau, China. J. Hydrol. 475,
111e122.
Yue, D.B., Xu, Y.D., Zhu, Y., Miao, H.M., Nie, Y.F., 2007. Vaporization of organic acids
from young landfill leachate during evaporation. Environ. Sci. (Huan Jing Ke
Xue) 28 (4), 897e901 (in Chinese).