Contaminated Sites

Sites,, Ground Water

ater,,
Water
Control and Remediation
 Exhumation of an Unlined Leachate Lagoon – a Case Study
A. Ferrari
Solví Participações S.A., São Paulo, Brazil (aferrari@solvi.com)
J. J. Scaff
Vega Engenharia Ambiental, São Carlos, Brazil (jscaff@vega.com.br)
S.A.N. Toyota
Vega Engenharia Ambiental, São Paulo, Brazil (stoyota@vega.com.br)
W.S.Junior
Vega Engenharia Ambiental, São Carlos, Brazil (wjunior@vega.com.br)

ABSTRACT Unlined leachate lagoons are still a reality in developing countries. Meanwhile, with regards to
tropical soils, the mechanism of leachate transport across a soil layer is still unknown. Full-scale studies are
needed to provide answers to this complex issue. This study involves an unlined lagoon which has been used
for over five years to store leachate. Exhumation was performed, along with the collection of soil samples
from various depths. The lagoon worked like a “full scale permeameter” providing some conclusions
regarding the attenuation mechanisms of specific pollutants, showing the difference in behavior on clayey
and sandy soils.

1 INTRODUCTION the characteristics of tropical soils (contaminant

Only 39% of cities in Brazil have an adequate
domestic waste disposal system (Abrelpe, 2007).
Just as in others developing countries, there are
many factors that contribute to prolonging this
situation such as economic problems, political
priorities, trend of importing technology from
developed countries and few practical studies that
considering the local environment.
There is a large gap between the liner system
design and its real-life application. One of the
problems is the lack of knowledge regarding the
interaction between tropical soils and leachate
from Brazilian landfills. There is no regulation in
Brazil that requires the use of clay liner and
geomembrane for sanitary landfills. The
Environmental Agency for the State of São Paulo
generally follows the U.S. Environmental
Protection Agency (EPA) minimum requirement,
a composite liner consisted of a 0.60m thick
compacted clay liner, followed by a 1.5 or 2.0mm
thick HDPE geomembrane. The permeability of
the clay liner must not exceed 1x10-9m/s.
Studies seeking alternatives to this requirement are
needed mainly for small cities, where there are
fewer economic resources. It is important to know
transport mechanisms), but only laboratories tests
are not sufficient. Full-scale research is necessary.
Practical studies such as the exhumation of old
landfills or leachate lagoons are important for
measuring real leachate contamination conditions.
This paper describes the exhumation of an unlined
leachate lagoon located in the city of São Carlos,
state of São Paulo, Brazil.
2 SITE CHARACTERISTICS
2.1 The city of São Carlos
The city of São Carlos is located in the central part
of the State of São Paulo, Brazil, and has an area
of 1,141 km² and a population of over 220,000
inhabitants. The region of São Carlos is
characterized by the presence of sandstone (sandy
soil) and basalt rocks (clayey soil), Basso (2003).
2.2 The sanitary landfill of São Carlos
The sanitary landfill in São Carlos receives about
220 tons of domestic waste each day. The landfill
began operating in 1996. The location of the
landfill is characterized by sandy soils from a
colluvial surface deposit and sandy soils derived
from sandstone, Almeida (2009).
624 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

2.3 Leachate lagoon characteristics

The landfill leachate is stored in lagoons originally
constructed without a geomembrane. The subject
of this paper is the exhumation of a lagoon that
was constructed in 2000. This lagoon began
receiving leachate of the landfill in January 2001.
The lagoon is roughly 45m long, 25m wide and
3.5m deep. Figure 01 shows the São Carlos
landfill and the exhumed leachate lagoon.

Fig.2 Leachate lagoon profile – “full scale

permeameter”.

In 2007, the local environmental agency forced the

landfill owner to install a 1.5 mm-thick HDPE
geomembrane in this lagoon. First, it was
necessary to empty this lagoon, the leachate was
removed, Figure 3. It was an unique opportunity to
exhume the lagoon and evaluate the soil condition
considering the leachate exposure.

Fig.1 São Carlos landfill and the exhumed lagoon.

For the purposes of this paper, the term “unlined”

refers to the absence of a geomembrane. The base
of this lagoon was constructed above the natural
sandy soil (sandstone), a reference of the grain
size distribution for this type of soil can be found
in Basso (2003), constituting 89.4% sand, 1.9 %
silt and 8.7% clay. An imported clayey sand also
derived from sandstone was spread over the
natural soil. The compaction controls are not
available. The grain size distribution for this
clayey sand is 56% sand, 8% silt and 36% clay,
JMF (2009). It is important to note that obtaining
the information on the past construction was a
difficult mission. Figure 02 shows the profile of Fig. 3 Empty lagoon prior to exhumation.
this lagoon, which worked like a full-scale
permeameter for over five years.
2.4 Leachate characteristics
The leachate in this landfill is stored in the
lagoons and then transported to an external
wastewater treatment plant located in a
neighboring city. The leachate characteristic for
the São Carlos Landfill is shown on Table 1. It is
important to note the high variability of
concentration for each constituent, considering
different leachate landfill samples.
 Ferrari et al. 625

TABLE 1 Composition of São Carlos Landfill leachate.

Concentration of constituents (mg/l) and pH

Constituent Contrera TCA TCA

(2008) (2003) (2006)
BOD5 - 973 480
DQO 6,396 3,440 1420
TOC - 372 -
Total solids 15,167 - -
Ammonia nitrogen as 1,930 73 189
N
pH 8.0 8.3 8.23
Chemical and metals
Arsenic - - 0.027
Boron - - 1.73
Cadmium 0.004 0.04 <0.005
Chloride 2,425 2,800 -
Chromium 0.225 0.44 0.22 Fig. 4 Exhumation soil profile.
Iron 1.676 13.1 2.8
Lead 0.069 - <0.10
Manganese 0.022 - 0.36 Five soil samples were collected and chemically
Nickel 0.080 - <0.01 analyzed. Figure 5 contains the exhumation soil
Potassium 2,107 - - profile and position of each sample.
Sodium 3,084 -
Sulfate - 35 -
Zinc 0.137 - -

3 EXHUMATION OF AN UNLINED
LEACHATE LAGOON

3.1 Soil sampling

Soil sampling began after the lagoon was emptied.

It was observed during the excavation, that the
bottom is formed by a 0.07m thick sludge layer,
followed by a 0.30m layer of clayey sand and then
by natural sandy soil, Figure 4.

Fig. 5 Soil sample points in the exhumation profile.

4 RESULTS

4.1 Chemical analysis

The soils samples were chemically analyzed. The
concentration of constituents in the soil (mg/kg)
and the comparison made using the reference
values for soil in the state of São Paulo are showed
on Table 2 and 3.
626 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

TABLE 2 Reference and intervention values for soil in of all elements in the local sandy soil (points C, D
the State of São Pauloa (mg/kg) and E) accounted for less than the sludge or the
clayey sand. Two situations were observed in the
Constituent Reference Intervention exhumation soil profile. The first situation is when
values values the concentration at point A (sludge) is greater
than the concentration at point B (clayey sandy).
Antimony <0.5 25 Figures 6 to 10 illustrate this situation for some
Arsenic 3.5 150 elements.
Cadmium <0.5 20
Chromium 40 400 Concentration (mg/kg)

Cobalt 13 90 0 500 1000 1500 2000

0
Cooper 35 600 B A
Lead 17 900 50 C
Mercury 0.05 70

Depth (cm)
Nickel 13 130 100
Vanadium 275 - D
Zinc 60 2000 150
a
(Cetesb, 2005). E
200

250
Chloride
TABLE 3 Concentration of constituents in the soil
profile (mg/kg)b. Fig. 6 Profile for chloride

Concentration (mg/kg)
Constituent A B C D E
0 1000 2000 3000 4000 5000
Aluminum (x103) 41.5 66.4 9.4 12.9 14.7 0
A
B
Antimony 6.2 20 1 1.1 2
50
Arsenic <0.4 <0.5 <0.3 <0.4 <0.3 C
Boron 100 217 25 29 33
Depth(cm)

100
Cadmium <0.4 <0.5 <0.3 <0.4 <0.3 D
Chloride 1,788 322 123 174 188 150
Chromium 32 8.6 4.6 4.7 5.2
Cobalt 6.6 11 0.5 <0.4 0.6 200
E
Cooper 31 76 4.8 5.6 6.5
250
Iron (x103) 31.7 72.3 8.2 10.2 11.1 Nitrogen Kjeldal
Lead 14 4 2.9 3.4 3.7
Fig. 7 Profile for nitrogen kjeldal
Manganese 249 277 60 44 43
Mercury 0.093 0.039 <0.02 <0.03 <0.02
Nickel 10 5.9 1.3 1.6 2 Concentration (mg/kg)
Sodium 1,727 394 119 213 277 0 5 10 15
Vanadium 106 269 24 28 33 0
B A
Zinc 59 20 2.2 2.8 6.1
Ammonia nitrogen 64 75 29 56 37 50 C
Nitrogen Kjeldal 4645 2363 289 389 325
Depth (cm)

TOC (x104)c 6.10 23 0.92 1.3 1.2 100

D

b 150
Bioagri (2007)
c
Total organic compounds E
200

For the local sandy soil, other than antimony, all

250
others constituents showed concentrations below Lead
the reference values for soil in the state of São Fig. 8 Profile for lead
Paulo. It was also observed that the concentrations
 Ferrari et al. 627

Concentration (mg/kg) Concentration (mg/kg)

0 500 1000 1500 2000 0 10000 20000 30000 40000 50000 60000 70000
0 0
B A A B
50 C 50
C
Depth (cm)

Depth (cm)
100 100
D D

150 150

E
200 E
200

250 250
Sodium Aluminum

Fig. 9 Profile for sodium Fig. 12 Profile for aluminum

Concentration Concentration (mg/kg)

0 5 10 15 20 25 30 35
0 50 100 150 200 250
0
B A
0
A B
50
C
50
C
Depth (cm)

100
Depth (cm)

D
100
150 D

E 150
200

E
250 200
Chromium Boron
Fig. 10 Profile for chromium Fig. 13 Profile for boron

The second situation is when the concentration at Concentration (104 x mg/kg)

point B (clayey sand) is greater than the - 5.00 10.00 15.00 20.00 25.00
concentrations at point A (sludge). Figures 11 to 0

15 illustrate this situation for some elements. A B

50
Concentration (mg/kg) C

0 5 10 15 20 100
Depth (cm)

0 D
A
B 150
50
C
E
200
Depth (cm)

100
D
250
150 TOC

E
200
Fig. 14 Profile for TOC

250
Antimony

Fig. 11 Profile for antimony

628 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

Future studies of contaminant transport

Concentration (mg/kg) mechanism from a sanitary landfill should be
0 50 100 150 200 250 focused on the implementation of more full scale
0
A
exhumations and laboratory tests, this appear to be
B
very important for defining new design parameters
50 for developing countries.
C
Depth (cm)

100 ACKNOWLEDGMENTS
D

150 The authors would like to thank Vega Engenharia

Ambiental and Solví Participações for their
E support.
200
Vanadium

Fig. 15 Profile for vanadium

REFERENCES

Abrelpe (2007). Overview of Solid Waste in

In cases where the concentration at point A is Brazil. Brazilian Association for Public
greater than the concentration at point B, C(A) > Sanitation and Special Waste (in Portuguese).
C(B), can indicate a great mobility of the element Almeida T.L.(2009). Environmental implications
in the clayey sand layer, especially when the from leachate attenuation processes in urban
concentration at point C (local sandy soil) is close solid waste disposal locations. D.Sc. Thesis,
to the concentration at point B (clayey sand), as University of São Paulo, São Carlos, Brazil (in
can be seen for chloride and sodium. Portuguese).
Basso J.B. (2003). Transport and retention of K+,
For the second situation, C(B) > C(A), it is Cu+2 and Cl-through compacted mixture of
possible to infer that the mobility of the element is lateritic soils for use in liners: homogeneity
lower in the clayey sand layer, demonstrating the procedures and column test equipment
sorption capacity of a clayey soil. Another point is adaptations . M.Sc. Thesis, University of São
the significant difference between concentrations Paulo, São Carlos, Brazil (in Portuguese).
in clayey sand and local sandy soils, Bioagri (2007). Soil Analysis Report number
C(B)>>>C(C). This is the case majority of metals. 110591. Bioagri Ambiental Ltda. (in
Portuguese).
For the local soil, the points C, D and E, Cetesb (2005). Board Decision number 195-2005-
confirmed the lower capacity of sorption for sandy E. -Environmental Sanitation Technology
soils, with low percentage of clay, in this case Company (in Portuguese).
fewer than 8,7%. Contrera R.C. (2008). Landfill leachate treatment
in sequential anaerobic and aerobic batch
5 CONCLUSIONS
reactors systems. D.Sc. Thesis, University of
São Paulo, São Carlos, Brazil (in Portuguese).
Considering the clayey sand derived from JMF (2009). Grain size distribution analysis
sandstone, the grain size distribution with at least Report. JMF Design and Consultant (in
36% clay appears to be responsible for the Portuguese).
sorption capacity of some metals and other TCA (2003). Waste Water Analysis Report
elements. Laboratory tests are also important for number 2580. Technology Support and
adjusting these field results with retardation and Environmental Consulting Company (in
diffusion parameters obtained from tests. Portuguese).
TCA (2006). Waste Water Analysis Report
The lower concentration of some elements in the number 901. Technology Support and
local sandy soil shows the efficiency of this thin Environmental Consulting Company (in
clayey sand layer (0,30m) that was exposed to a Portuguese).
leachate for over 5 years. This is an important
issue for developing countries because, depending
on the soil and leachate, new landfill liner
requirements must be defined, considering the
characteristics of tropical soils.
 Biogeochemistry of Highly Alkaline Waste Disposal Sites

D.I. Stewart
University of Leeds, Leeds, LS2 9JT, UK (d.i.stewart@leeds.ac.uk)
R.J.G. Mortimer
University of Leeds, Leeds LS2 9JT, UK (r.j.g.mortimer@leeds.ac.uk)
R.A. Whittleston
University of Leeds, Leeds LS2 9JT, UK (lec5raw@leeds.ac.uk)
Shanmugam Palani
Central Leather Research Institute, Chennai, India
I.T. Burke
University of Leeds, Leeds LS2 9JT, UK (i.t.burke@leeds.ac.uk)
ABSTRACT Poorly disposed alkaline wastes at industrial legacy sites pose a serious environmental threat.
Due to the cost and hazard of waste removal, in situ management is often the only option.
Geomicrobiological processes occurring in and around the waste are critical to the geochemistry and
therefore the mobility of toxic metals such as Cr(VI). Mechanistic understanding of the role of soil microbes
would allow manipulation of conditions to achieve a desired engineering outcome such as the reductive
precipitation of toxic metals. We present studies that provide insight into the role of soil micro-organisms in
this process.
INTRODUCTION Derbyshire). All three are legacy sites from
Many industrial processes can produce highly industries that closed more than 40 years ago
alkaline wastes (e.g. lime production, chromium (indeed one waste pile has been in place more than
ore processing, cement production and nuclear 100 years). The sites share the characteristic that
reprocessing, to name but a few). These wastes can the wastes were tipped without much concern for
have a seriously detrimental impact on the the environment, and that subsequent remedial
environment if poorly disposed of, and rainwater measures consist principally of soil cover that
leaching downwards from the waste can prevent dusts from being generated. At all three
contaminate any groundwater resources beneath sites water is entering the waste and highly
disposal sites. In the UK, many legacy sites alkaline leachate is emerging.
containing poorly managed alkaline wastes exist
where high pH leachate is released into the Fig. 1 shows an alkaline seep emerging from the
environment. Such sites rarely have an imperm- base of the waste at site 1. The pore water in the
eable barrier underneath the waste. Thus the fate of waste at this site has elevated sulphate, calcium
the alkaline leachate, and particularly any and sodium concentrations, high pH and can
contaminants within that leachate, depends solely contain ca. 600 ȝmol.lí1 of chromate (for further
on the biogeochemistry of the soil beneath the site. details see Stewart et al. 2007). At site 1 a calcium
This paper discusses geomicrobiological factors carbonate based precipitate forms whereever the
that control the geochemistry beneath alkaline leachate emerges to atmosphere.
waste disposal sites, and thus control the impact of
alkaline leachate on the wider environment. At site 2 the waste is placed on the side of a valley
and water is constantly seeping from the down-
ALKALINE WASTE SITES slope face of the waste where it enters a drainage
We are currently studying three alkaline waste ditch which is visibly yellow (Fig. 2). The volume
disposal sites. Two are chromium ore processing of water that emerges is small, but it typically has
residue (COPR) disposal sites (site 1 in N.W. a pH value of 11.4 and can contain up to 293 µM
England, and site 2 in Yorkshire), and the third site Cr(VI) (for details see Stewart et al., 2010). The
is a lime kiln waste disposal site (site 3 in
630 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
Fig. 1 Calcium carbonate precipitate that has formed at
site 1 where leachate from highly alkaline COPR is
emerging near the base of the waste pile.
waste that has been deposited at site 2 is similar to
that dumped at site 1 (COPR produced by the high
lime process), but the waste has been in place since
the 19th century (whereas factory at site 1 closed in
the late 1960’s). Despite the similarity of the waste
at the two sites, far less precipitate is forming at
site 2, possible because the leachate contains less
calcium.
Site 3 is in a small valley adjacent to a former lime
works. A Hoffmann kiln for roasting limestone to
produce lime occupied high ground above the site
in the early part of 20th Century. As part of kiln
operations, limestone unsuitable for burning in the
Hoffman kiln and kiln waste were dumped in a
former quarry at the kiln site (Derbyshire County
Council, 2004). Leachate with a pH value of ~13 is
emerging from an engineered fill (probably the
kiln waste) at the head of the valley. Almost
immediately after emerging from the engineered
fill, the leachate enters a small stream that follows
the valley bottom. Large volumes of what is most
probably calcium carbonate have precipitated in
the valley bottom for a distance of over 250m
along the valley (see Fig. 3).
Fig. 2 Drainage ditch capturing alkaline chromate
contaminated leachate from the COPR waste at site 2.
(For colour figure, refer to CD)
All 3 sites are examples of situations where the
leachate poses an environmental threat, but the
only realistic solution is in situ management as the
cost and hazard of removing the waste are
prohibitive. Consequently it is important to
understand the geomicrobiological processes
occurring at these sites and the potential to
manipulate them in ways that could reduce the
environmental impact of the leachate.
GEOMICROBIOLOGICAL PROCESSES
THAT CONTROL THE GEOCHEMISTRY
Metal Reduction during Anaerobic Respiration
Soil microorganisms have evolved to obtain
energy to support growth from the breakdown of
organic matter. This process of catalysing
thermodynamically favourable chemical reactions
to obtain the energy released is called dissimilatory
metabolism. Soil microorganisms also obtain
carbon for synthesising cell mass from organic
matter in a process called assimilatory metabolism.
Some microorganisms obtain energy by coupling
the oxidation of simple organic compounds to the

Fig. 3 Precipitate that has formed where leachate from
lime kiln waste at site 3 has joined a nearby stream
(For colour figure, refer to CD)
reduction of Fe(III) or Mn(IV) during the process
of anaerobic respiration (see Fig. 4). In such
reactions the organic molecule being oxidised is
often referred to as the electron donor, where as
the substance being reduced is called the electron
acceptor.
The Electron Acceptor
Decay of organic matter is largely an oxidation
reaction (electrons are lost), and this must be
balanced by the reduction of a suitable electron
acceptor. Where sufficient organic matter is
available, progressively more anoxic conditions
develop and a cascade of terminal-electron-
accepting processes (TEAPs) occur in sequence
(Burke et al. 2005). Processes releasing most
energy are favoured, so the sequence in which
electron acceptors are used by micro-organisms
typically follows the decreasing order of redox
potentials shown in Table 1. When oxygen is
present it is the preferred electron acceptor, but
micro-organisms can reduce other substances such
as nitrate, manganese, iron and sulphate during
anaerobic respiration.
Stewart et al. 631
Electron Electron
Donor Acceptor
Oxidized state CO2 Fe(III)
e
Labile organic
molecules derived
from soil organic
matter
Metal
Reduced state Acetate reducing Fe(II)
bacterium
Fig. 4 Schematic diagram illustrating dissimilatory
metal reduction (For colour figure, refer to CD)
Some redox active metals that are usually regarded
as contaminants, such as Cr(VI) and U(VI) can be
reduced either directly, or indirectly by soil
microorganisms (Lovley, 1995). This process can
be of engineering interest if the reduction converts
an environmentally mobile species into an
immobile one. For example Cr(VI) as the divalent
chromate anion is both highly toxic and very
mobile, but Cr(III) is relatively insoluble at
environmental pH values and therefore immobile.
Contaminant reduction can be a direct enzymatic
process, or an indirect microbially mediated
process involving other redox active species (e.g.
microbially reduced Fe(II) can reduce Cr(VI) by
oxidation to Fe(III)). Direct contaminant reduction
requires that the microorganism responsible has an
appropriate terminal reductase to catalyse the
transfer electrons to the specie being reduced. The
redox potentials in table 1 indicate that most of the
common electron acceptors used in the natural
environment yield energy to support life at high
pH when coupled with oxidation of an organic
molecule such as acetate. However high pH is a
challenging environment for microorganisms; they
need to have evolved mechanisms for regulating
cytoplasmic pH and have surface layer enzymes
that function at high pH. Further, the
microorganisms at contaminated sites need, at least,
to be tolerant to any contaminants present.
The Electron Donor
In most soil, the complete oxidation of organic
matter requires the cooperative activity of a
632 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

community of microorganisms collectively by soil organic matter can be simplified by the

exhibiting several different metabolic pathways
(e.g. hydrolysis of complex organic matter,
fermentation of sugars, and oxidation of fatty acids,
lactate, acetate and H2; Lovley, 1993). Different
organic substrates can support different metabolic
functions by the same microorganism, and
different microorganisms can metabolise a given
substrate by different pathways. The particular
group of microorganisms that compete
successfully for each component of organic matter
will depend on the prevailing geochemical
conditions. The aim of any engineering
intervention will be to control conditions to
manage the microbial community to promote a
particular geochemical reaction and therefore a
desired outcome.
In the sub-surface environment microbial numbers
are usually limited by the availability of electron
donors. Thus the complexity of a system supported
addition of an excess of an appropriate electron
donor (usually a simple organic molecule that is a
natural intermediate is the oxidation of organic
matter, such as acetate). The effect is that a sub-
population of microorganisms that can oxidise the
chosen electron donor most efficiently under the
prevailing conditions will thrive. These will
consume the natural electron acceptors in the
system in turn, with the microbial population
evolving as the conditions change. Thus the natural
microbial community of a soil can be manipulated
to encourage it to reduce a particular electron
acceptor, which in turn can result in the reductive
precipitation of a contaminant such as chromium.
MANIPULATING REDOX CONDITIONS AT
ALKALINE CONTAMINATED SITES
Microcosm experiments have been conducted
using the soil and stream water from site 1. These

TABLE 1 Microbially significant half-reaction reduction potentials: Standard Reduction Potential, E0, and redox
potential, Eh, at 10.5 (at 25qC and atmospheric pressure).
Eh @ Eh @
E0 pH 7 pH
Transformation Reaction Assumptions
(V) (V) 10.5
(V)

O2 Depletion+ O2 + 4H+ + 4e = 2H20 1.230 0.687 0.598 PO2 = 0.2 bar

NO3 + 6H+ + 5e [NO3] = 1 mmol L-1

Denitrification+ 1.240 0.571 0.464
= 1/2N2 + 3H2O PN2 = 0.8 bar
Mn reduction + Mn3O4 + 2H+ + 2H2O + 2e–
0.480 - -0.140 -
Mn(III) to Mn(II) = 3Mn(OH)2
Cr reduction* CrO42 + 8H+ + 3e
1.507 0.404 -0.150 [CrO42] = [Cr3+]
Cr(VI) to Cr(III) = Cr3+ + 4H2O
Fe reduction+ Fe(OH)3 + H+ + e
0.257 - -0.364 -
Fe(III) to Fe(II) = Fe(OH)2 + H2O
Sulfate reduction+ SO42 + 10H+ + 8e
0.301 -0.217 -0.476 [SO42] = [H2S]
S(VI) to S(-II) = H2S + 4H2O
Bicarbonate reduction
2HCO3 + 9H+ + 8e [HCO3] = [CH3COO]
to acetateu 0.187 -0.292 -0.525
= CH3COO + 4H2O = 20 mmol L-1
C(VI) to C(0)
+
after (Langmuir 1997)
*
calculated using thermodynamic data from (Stumm & Morgan 1996)
u
calculated using thermodynamic data from (Thauer et al. 1977)
 Stewart et al. 633

consisted of 10 g of soil and 100 ml of water in
120 ml glass serum bottles that were sealed with
butyl rubber stoppers and aluminium crimps. The
microcosms were amended with potassium chrom-
ate to produce a final concentration of 250 ȝmol.lí1
Cr(VI) and 20 mmol.lí1 sodium acetate as an
electron donor. Soils from two different locations
were used, and these produced microcosms with
pH values between 7–8 and 9–10, respectively. For
each pH system three repeat microcosms and a
heat killed sterile control were run.
In the active microcosms the addition of acetate
caused chromate to be removed from solution as
part of a cascade of terminal electron accepting
processes (see Fig. 5), whereas there was very little
geochemical change in the sterile controls (see
Stewart et al. 2007 for details). This was therefore
a microbially mediated process, and due to the
timing of Cr(VI) removal in the redox cascade it
was most probably by reductive precipitation as a
nitrate (aq)
Cr(III) hydroxide. What is notable about Fig. 5 is
that it clearly demonstrates that bio-reduction can
be stimulated in quite alkaline conditions, but the
rate at which the redox cascade develops in
alkaline conditions is slower than at pH7.
At site 2 a soil sample recovered from beneath the
waste pile had a pH value of 10.5 and has accum-
ulated 0.3% chromium by weight. This soil was
the surface soil prior to waste tipping and contains
3.6% total organic carbon. There is no evidence
that the chromium in the soil is due to physical
transport of the waste particles, and it is assumed
that it is the result of exposure to chromate con-
taining leachate. As chromate is unlikely to accum-
ulate in soil to the extent observed either by anion
exchange or precipitation, and a population of soil
microorganisms is present (Fig. 6) it is proposed
that the accumulation of chromium is by microb-
ially mediated reductive precipitation (Stewart et al.
2010). Work is on-going to test this hypothesis.
Cr(VI) (aq) acid extractable iron as Fe(III) (s) sulphate (aq)

100
Relative amount %.

80

60

40

20

0
0 10 20 30 40 50 60 70 80
(b)

100
Relative amount %.

80

60

40

20

0
0 10 20 30 40 50 60 70 80
Time (days)
(a)
Fig. 5 Aqueous concentration of nitrate, Cr(VI) and sulphate, and the proportion of the acid extractable solid phase iron
as Fe(III) in the microcosm experiments on soils from site 1 (a) at pH 7-8 and (b) pH 9-10.
634 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
Alpha-
Unidentified proteobacteria Gamma-
9% proteobacteria
Actinobacteria
3%
3%
3%
Firmicutes
82%
Fig. 6 Microbial community in the soil beneath COPR
waste at site 2 (32 clones). Charts show phylogenetic
affiliation of 16S rRNA gene sequences
At site 3 there is a geochemical profile in the soil
beneath the calcium carbonate that suggests
vertical separation of redox active species in a
sequence of TEAs normally associated with a
redox cascade in order of energy yield. A soil
sample with a pH of 12.5 taken from below a thin
cover of calcium carbonate contained a microbial
community despite the highly alkaline conditions
(Fig. 7). This was a zone where nitrate
concentration is lower that the zone above, nitrite
is considerably enriched, and there is some
evidence for an increase in the proportion of acid
extractable iron in the Fe(II) oxidation state. Our
working hypothesis is that microbial respiration is
responsible for the redox profile at this site,
however work is continuing to prove our
hypothesis (see Burke et al. forthcoming).
SUMMARY
It is shown that a range of alkaliphilic bacteria are
found beneath alkaline waste sites, and these
bacteria are capable of reducing metals, such as
Fe(III), at high pH. In such a reducing environment,
contaminants such as Cr, U, and Tc can be retained
in the soil as insoluble phases, reducing their
environmental impact. It is also shown that a
mechanistic understanding of the metabolic
processes of alkaliphiles is essential to
understanding the pH and redox controls on
leachate migration at such sites. Such a mechan-
istic understanding will also provide the best
available conceptual model of the biogeochemistry
and contaminant transport at, for example, deep
nuclear waste repositories where cement is used as
part of the primary barrier system.
REFERENCES
Burke, IT, Boothman, C, Lloyd, JR, Mortimer,
RJG, Livens, FR and Morris K (2005). Effects
of progressive anoxia on the solubility of
technetium in sediments. Environ. Sci. & Tech.
39(11): 4109-4116.
Derbyshire County Council (2004). Report of the
Director of Environ. Services to the Regulatory
– Planning & Control Committee, 7 June 2004.
Langmuir, D (1997). Aqueous Environmental
Geochemistry. Prentice Hall.
Lovley, DR (1993). Dissimilatory Metal Reduct-
ion. Ann. Rev. of Microbiol. 47:263-290.
Stewart, DI, Burke, IT, Hughes-Berry, DV and
Whittleston, RA (2010). Microbially mediated
chromate reduction in soil contaminated by
highly alkaline leachate from chromium
containing waste. Ecol. Eng., 36, 211–221.
Stewart, DI, Burke, IT and Mortimer, RJG (2007).
Stimulation of microbially mediated chromate
reduction in alkaline soil-water systems.
Geomicrobiology Journal 24(7-8): 655-669.
Stumm, W and Morgan, JJ (1996). Aquatic
Geochemistry. New York: John Wiley & Sons.
Thauer RK, Jungermann K, and Decker, K (1977).
Energy conservation in chemotrophic anaer-
obic bacteria. Bacteriol Rev 41(1):100–180.
Unidentified
26%
Verrucomicrobia
5%
Bacteriodetes
5% Beta-
proteobacteria
64%
Fig. 7 Microbial community in the soil beneath
carbonate deposit at site 3 (19 clones). Charts show
phylogenetic affiliation of 16S rRNA gene sequences.
 The Persistence of the Soil Contamination by Lead and Cadmium Around the
City of Santo Amaro da Purificação, Brazil
S L Machado
Federal University of Bahia, Brazil (smachado@ufba.br)
R B Portella
Federal University of Bahia, Brazil (roberto.portella@ufba.br)
E Cesana
Ecológica ± Tecnologia e Controle Ambiental Ltda, Brazil (ecesana@ecologica.com.br)
T S Rabelo
Federal University of Bahia, Brazil (thayrabelo@yahoo.com.br)

ABSTRACT Several studies since 1970 have demonstrated the severity of the contamination case in Santo
Amaro, Bahia. For many years, the factory installed in the city operated without any emissions control,
impacting directly the local vicinity. This study evaluates the persistence of contamination of the superficial
soil around the company, relating them to expected particles concentration in the air when the company was
still active. In order to do this, the emissions and dispersion of particles into the atmosphere were simulated,
considering the metallurgical process data and local weather and topography parameters. The results obtained
were used to plot isolines of particle concentration in the air. Radial lines were drawn from the main chimney,
defining sampling points at their intersections with isolines. The results showed that the past emissions still
play a very important role in the soil contamination. There is a clear persistence of the lead and cadmium
concentration values and, especially in areas close to the factory, they are far above the reference limits.

INTRODUCTION
During the years 1956 to 1993, intense lead
metallurgic activities took place in the city of
Santo Amaro da Purificação, Bahia, Brazil. These
activities generated several environmental impacts
that have been studied since the 1970´s (Machado
et al. 2003).
Lead metallurgical debris are typically composed
of Si, Ca, Fe, Zn, Pb and S oxides with traces of
Cd, As, Sb, Co e Cr. Many studies evaluating the
effects of the contamination on human health and
the environment have been carried out in Santo
Amaro showing the severity of this contamination
case (Anjos, 1998, 2001, 2003; Brasil 2003;
Carvalho et al., 1984, 1985, 1986, 1996, 2001;
Machado et al., 2003, 2004; Tavares, 1978, 1990;
Tavares & Carvalho 1992).
These studies demonstrate that the Santo Amaro
Environmental Contamination Case can be
considered as one of the most serious heavy metal
contamination cases in the world.
For 34 years, the metallurgical company operated
without any emissions control, having direct
impacts on the immediate surroundings. In 1980, a
few contamination control initiatives were taken,
such as the implantation of a 90m high chimney
with air filters placed at the particle emissions
sources, the prohibition of the use of the lead
debris for landfill purposes and the removal of all
the residences within a radius of 500m of the
industry. (Anjos 2003).
Although there were many sources of
contamination derived from company activities,
the uncontrolled air emissions, the liquid effluents
poured directly into the River Subaé, the
uncontrolled debris disposal and the superficial
drainage water coming from the debris storage
area can be cited as the main causes of pollution.
Moreover, the lead debris, considHUHG DV ³QRQ
GDQJHURXV´ E\ WKH HQWUHSUHQHXU ZDV XVHG E\ WKH
Santo Amaro city authorities and by the
population for several purposes, such as in
636 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
pavements and as a landfilling material in
residential courtyards and public areas (Machado
et al. 2003).
After 16 years of company closure and a huge
amount of papers published on the subject, there
are still many persisting doubts about the residual
contamination still active in the region. This study
aims at evaluating how the contamination levels in
the superficial soil around the company are still
influenced by the previous/past atmospheric
emissions. In order to do this, simulations of
emissions and particle dispersions into the
atmosphere were performed, considering data
from the metallurgical processes, local weather
and topography parameters which were then
compared with the heavy metal concentrations (Pb
e Cd) in the soil surrounding the former mining
company.
MATERIALS AND METHODS
First of all, numerical simulations of the
dispersion of atmospheric emissions during the
operating period of the company were performed,
considering the metallurgical processes used, the
topographical parameters of the region and the
average local climate conditions (the weather data
was obtained from weather stations located at
Salvador airport). Hourly data was collected for
wind direction and speed, temperature, stability,
and mixing height for the rural and urban zones.
The topography was based on data acquired from
SRTM Shuttle Radar Topography Mission,
performed by NASA. The company operation data
was obtained from work performed by The
Environmental Institute of Bahia (Bahia 1992), by
Anjos (1998) and by Tavares (1990). The
modeling results indicated the probable
concentration of the particles and sulphur oxides
within the company surroundings and identified
the regions where maximum concentrations
occurred. For calculation purposes, it was assumed
that SO2 emissions were emitted from the 80m
high chimney and that the particles were emitted
via the seven smaller chimneys distributed
throughout each stage of the metallurgical process.
These calculations were based on the assumption
that gross emissions without control occurred,
which is consistent with the characteristics during
the majority of the operation period of the plant.
Results obtained from the numerical simulation
were interpolated and presented in the form of
isolines representing air particle concentrations, as
shown in Figure 01.
A total of five isolines representing concentrations
of 100, 80, 60, 40, and 20 ȝJP3, were determined
within a radius of 4,5 km from the main chimney.
A sixth contour line was projected at a distance of
 NP IURP WKH  ȝJP3 isoline and was adopted
as a boundary line, beyond which, the influence of
the past emissions can be neglected. The total
study area, covered by the sixth contour line, is
about 7500 hectares in size and corresponds to
nearly 16% of the total area of Santo Amaro City.
Eight radial axis were plotted every 45 °, using the
factory chimney as the center point. The first axis
was intentionally positioned over the urban area of
Santo Amaro. The intersections of the radial axis
with the concentration isolines defined the
location of the 48 sampling points used in this
work.
These points were geo-referenced to enable the
field sample campaign be performed, using GPS.
Superficial soil samples (up to 20 cm) were then
collected at these 48 points in order to determine
the Pb and Cd metal concentration values.
Superficial soil samples were analyzed using AAS
Atomic Absorption Spectrometry.
RESULTS AND DISCUSSION
For each isoline, the results were analyzed taking
into account the average lead and cadmium
concentrations obtained in the superficial soil
samples, as shown in Figure 2. In this figure, the
dashed lines include all the experimental results
and the solid line excludes the DA 01-06 axis,
which passes through the urban area of Santo
Amaro.

Fig. 01 Map of the city of Santo Amaro showing the soil sampling points and the isolines of particle concentrations of
air.
As it can be seen in the Figure 2 there is a clear
correlation between the concentration values in the
superficial soil and the values obtained from the
atmospheric dispersion simulation, indicating that
past air emissions still play an important role in
the contamination of the studied area.
It can also be observed that the exclusion of the
axis that passes through the urban area causes a
considerable decrease in the average concentration
values obtained in field. This is probably due to
the presence of the debris disposed in the urban
area. The authors believe that due to the
circulation of vehicles and people in these
locations and the regular works of excavation,
transportation and re-compaction of superficial
soil performed by the city authorities and
population, there is a considerable amount of dust
Machado et al. 637
with high levels of lead and cadmium that is
deposited on the surface soil, increasing the
concentration values in the urban zone. Another
important aspect to be considered is that the lead
debris are classified as a Class I residue according
to the NBR 10004, NBR 10005 and NBR 10006
(Brazilian Standards).
According to Anjos (2001) and Machado et al.
(2003), it is clear that this material liberates a
significant quantity of lead into the environment
due to the leaching of rainwater.
Table 1 summarizes the lead concentration values
obtained and compares them with the reference
values established by CETESB (2005). It can be
observed that 43 sampling points presented lead
concentration values above the quality limit, 16
638 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
 Machado et al. 639

Fig. 2 Average Lead Concentrations in soil as a function of the simulated concentrations

of the Atmospheric Emissions from the company

Fig. 3 Results obtained in this study compared to other results published by different
authors (Anjos 1998, Brasil 2003, Costa 2001 and Machado et al. 2003)

Anjos J.A.S.A. (1998) Remediation Strategies contaminated with industrial waste: The
for sites contaminated by heavy metals: Plumbum Case, Santo Amaro da
Case study. Dissertation, University of São Purificação/BA. Bahia Análise e Dados,
Paulo, Polytechnic School, São Paulo, 157p. Salvador, Bahia, v 2, p. 306-309. (In
(In Portuguese) portuguese)
Anjos J.A.S.A. and Sánchez L.E. (2001) Bahia (1992) Centre for Environmental
Environmental Management Plan for sites Resources ± CRA/BA. Answers to Attorney
640 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
General Requirements Salvador,
Salvador, 7p. (In Portuguese)
Brasil (2003) FUNASA ±National Foundation of
Health Risk Assessment for Heavy Metals in
Santo Amaro, Bahia, Ministry of Health
Final Report, 268p. (In Portuguese
Carvalho F.M., Silvany-Neto A.M., Tavares
T.M., Costa A.C., Chaves C.E.R.,
Nascimento L.D. and Reis M.A. (2001) The
Persistence of High Levels of Lead in the
Blood of Children from the ages of 1 to 5
years in Santo Amaro da Purificação. In:
Seminar on contamination by heavy metals
in Santo Amaro-BAhia. Santo Amaro ± SP,
Brazil, 37-51. (In portuguese)
Carvalho F.M, Tavares T.M, Souza S.P. and
Linhares P. (1984) Lead and cadmium
concentrations in the hair of fishermen from
the Subaé River Basin, Brazil. Environ. Res.
33, 300-306.
Carvalho F.M., Silvany-Neto A.M., Peres
M.F.T., Gonçalves H.R., Guimarães G.C.,
Amorim C.J.B., Silva Jr J.A.S. and Tavares
T.M. (1996) Lead poisoning:
zincoprotoporfirina in the blood system of
children in Santo Amaro da Purificação and
Salvador, BA. Journal of Pediatry, v. 72, n. 5,
p.295-298. (In portuguese).
Carvalho F.M., Silvany-Neto A.M, Tavares T.M.,
Lima M.E.C. and Waldron H.A. (1986) Lead
Poisoning of children in Santo Amaro, Bahia,
Brazil. Bol. Of Sanit Panam, p. 309-318. (In
portuguese).
Carvalho F.M, Silvany-Neto A.M., Tavares T.M.,
Lima M.E. and Waldron H.A. (1985) Lead
poisoning among children from Santo
Amaro, Brazil. Bull. Pan. Amer. Health
Organization, v. 19, 165-175.
CETESB (2005) Environmental Company of
São Paulo. Board Decision No. 198:
Guidance Values for soils and groundwater
in the State of São Paulo. Available on:
7p. www.cetesb.sp.gov.br>. Accessed on: 2nd of
May 2009. (In Portuguese).
Costa A.C.A. (2001) Evaluation of some aspects
of de environmental liabilities of a lead
metallurgy in Santo Amaro da Purificação,
Bahia. Dissertation. Federal University of
Bahia, 152p. (in Portuguese).
Cunha F.G. da (2003) Human and Environmental
Contamination by Lead in the Ribeira Valley,
in the states of São Paulo and Paraná. Phd
Thesis, State University Of Campinas,
Geosciences Institute, São Paulo, p125. (In
Portuguese).
Machado, S.L., Carvalho M. de F., Delgado
C.W.C., Anjos J.A.S.A., Dourado K.A.,
Amparo N.S. do, Cardoso L.P., Lima A.V.M.,
Santos D.M., Santos D.B. dos, Kiperstok
First A., Sánchez L. Botelho M.A.B. and
Oliveira Neto F.A. (2003) PROJETO
PURIFICA ± Remediation Proposal for
Areas Degraded by Lead Mining Activities
in Santo Amaro. Technical Report. 223 p.
Machado S.L.; Ribeiro L.D., Kiperstok A.,
Botelho M.A.D. and Carvalho M.F. (2004)
Diagnosis of Heavy Metal Contamination in
Santo Amaro- BA. Sanitary and
Environmental Engineering, v 9, n. 2 ±
April/June, 140-155. (In Portuguese)
Tavares T. M. (1978) Integrated Technical
Report of Environmental Analysis.
EDUFBA, Salvador BA, Brazil, 189p. (In
Portuguese).
Tavares T. M. (1990) Effect of Lead and
Cadmium Emissions in Santo Amaro da
Purification. Phd Thesis, IQ/USP, São Paulo-
SP, Brazil, p271. (In Portuguese).
Tavares T.M. and Carvalho F.M. (1992)
Assessment of Human Population Exposure
to Heavy Metal Environments: Examples
from Northeastern Bahia. New Chemistry,
p.147-154. (In Portuguese).
 Leachate Recirculation to Reduce Its Pollution
Potential for Warriana Dumpsite-A Case Study
M. K. Kaushik,
DAVIET, Jalandhar, India, mh_kaushik@ yahoo.co.uk.
Arvind Agnihotri,
NIT, Jalandhar, India, agnihotriak@nitj.ac.in.
Ajay Bansal,
NIT, Jalandhar, India, drajaybansal@gmail.com

ABSTRACT Jalandhar city is facing a big problem of solid waste disposal. In Jalandhar, 400 tones of municipal solid
waste (MSW) is generated daily and dumped mainly in warriana disposal site. Leachate was collected from the site
and analyzed for various chemical parameters. This landfill produces a great amount of leachate which contains
inorganic as well as organic matter. This is especially for the initial leachate from land filled MSW having
concentration of COD and BOD up to 2050 and 1000 mg/l, respectively. COD/DOC ratio was around 1.5. Amm-N
was up to 280 mg/l while chlorides were 392 mg/l. The ground water quality is severely affected by the leachate
contaminant migration in aquifers. But, this leachate could be disposed safely by means of recirculation
techniques, which decomposes the organics through the action of proliferating microorganism placed on gravel
and tire chips filter media and thereby purifies the leachate and simultaneously accelerates organic decomposing
through water saturation control. The scrap tire layer purifies the leachate in the landfill and also prevents buildup
of the hydraulic head within the filter.

can create environmental and health hazards and

INTRODUCTION
Punjab’s various cities are facing
problems due to ever increasing amount of waste
and improper waste management practices. In
Punjab, 3500 tones of municipal solid waste are
generated daily. Out of this, the 4 municipal
corporations (Ludhiana, Jalandhar, Amritsar,
and Patiala) account for 2000 Tons per day
(TPD) generation of MSW in Punjab. All the
other districts of the state generate the remaining
1500 TPD municipal solid waste. Ludhiana
generates 29% followed by Jalandhar 17% of
total MSW generated in the state. Jalandhar city
alone generates about 400 Tons per day of
MSW. Jalandhar is one of the cities which are
facing similar problems of poor waste
management. The indiscriminate dumping of
this municipal solid waste in water bodies and
low lying areas is a common practice followed
by most of the municipalities with no
consideration of its effect on the environment.
This solid waste also contains scrap tires which
requires proper management.
Now, leachate treatment systems are
also extensively used in modern landfills to
control the leachate flow and hence to minimize
contaminant impact on the environment. These
systems commonly comprise leachate treatment
units with anaerobic and aerobic treatment
methods applicable. Leachate collection system
is used for the collection and removal of leachate
from the bottom layers of landfills. These are
commonly comprises of perforated leachate
collection pipes at regular spacing in a
continuous blanket of uniformly graded granular
material. Now, shredded scrap tires have also
been proposed as an alternative drainage
material, based in their high hydraulic
conductivity (Hall 1991, Edil et. al, 1992,
Ahmed and Lovell 1993, Duffy, 1995).
Over 100 million scrap tires are
generated annually in India. But, only 18% of
scrap tires is beneficially and environmental
safely reused or recycled. Even through tires
comprises only 1% of the total waste generated
642 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
in India, they present significant problem due to
their nonbiodegradable, compaction resistant,
and self combustible nature (Moo-Young et.al.
2003). Automobile tires are generally made of
natural rubber, synthetic rubber elastomers,
polymers and other additives. Steel
reinforcement is also provided in tires to
improve its strength. These tires even after use,
maintain their chemical composition, requiring
thousands of years to fully decompose.
In the present study, these scrap tires
have been used as drainage material for landfill
leachate treatment. The scrap tire layer purifies
the leachate in the landfill and also prevents
buildup of the hydraulic head within the filter.
Overall, these types of usages will decrease the
cost of land filling and will be an
environmentally friendly option for the disposal.
Filter layer should be designed to prevent a
leachate head from exceeding 0.3 m above the
top layer. Additionally, a minimum thickness
and hydraulic conductivity of 0.3 m and 1X10-5
m/s respectively, have been generally
recommended for the design of these layers by
various environmental agencies (USEPA,
Subtitle-D regulations), (Reddy and Saichek,
1998 a). Uniform gravel along with scrap tire
shreds has been used in these filters layers.
These materials have been used due to their high
hydraulic conductivity. Several laboratory
studies have been conducted to show the
effectiveness of scrap tires in leachate collection
layers (Hall, 1991, Edil et. al. 1992, Humphrey
and Manion, 1992, Reddy and Saichek 1998 a,
Warith et. al. 2004, Rowe and Mclsaac, 2005).
The drainage performance of these scrap tires as
well as their capabilities in protecting the
underlying liner systems, was also observed in
field studies. (Evans 1997, Reddy and Saichek
1998 a, Park et.al., 2003). According to these
studies, scrap tires have hydraulic conductivity
one or two orders of magnitude, much higher
than the design performance criterion of 1X10-5
m/s, even under significant pressures (45-70
kpa). Most of the existing work showed that
these scrap tires did not pollute groundwater and
surface water reservoirs, since the measured
concentration of many constituents leached from
the tire shreds, were significantly below the
USEPA norms (Humphrey and Katz, 2001,
Park. et al. 2003). Laboratory column studies in
DAV Institute of Engineering and Technology’s
environmental engineering laboratory indicated
that scrap tires may be more prone to clogging
than uniform gravels in critical zones. So, to
reduce the problem uniformly graded large size
tire chips are placed at appropriate design
thicknesses. The objective of this study was to
investigate the performance of the scrap tires
and gravel and their mixer in landfill leachate
treatment systems and also to provide the
additional data to demonstrate their effectiveness
as a drainage medium. To achieve these
objectives laboratory tests were conducted to
determine the treatment potential and
leachability of scrap tires under different
pressure gradient conditions. The drainage
performance and environmental suitability of
tires, as well as their potential to treat leachate
was observed in these test cells.
MATERIAL CHARACTERIZATION:-
The tire chips used in this study were
obtained from the local market of Jalandhar city.
Conventional grain distribution analysis (ASTM
D 422) was not applicable to tire chips used in
the study, due to their irregular sizes. The sizes
of tire chips ranged from 20 to 85 mm in length,
with an average chip length of 40 mm. But, the
tire chips of more than 100 mm were not used in
this study. The portion (§2 %) which was having
size of 100-200 mm in length was eliminated
before the laboratory test. Small size tire chips
of less than 2 mm size were also removed, by
sieving through a 2 mm sieve. The test specimen
was mostly comprised of wire free tire chips,
but, also included small percentage of about
10% of wire containing tire chips. The specific
gravity of tire shreds ranges from 1.02 to 1.36,
depending upon amount of steel wires in the tire
shreds (Edil and Bosscher, 1994: Zimmesman
1997, ASTM 1998).
Gravel was analyzed for particle sizes
distribution studies before experiment. This was
also obtained from local market, mainly
produced by crushing of local limestone

materials. Particle size distribution analysis
indicate that the mean diameter of gravel was
near 38 mm. Field experiments (Brune et. al.
1994, Fleming et. al. 1999, Maliva et. al. 2000,
Bouuchez et. al. 2003) shows that a biological
slime layer (bio film) grows on the materials in
the collection and treatment systems is followed
by biologically induced deposition of inorganic
constituents, on gravel and tire shreds materials
from the leachate flow. The clog that develops
decreases the pore space available to transmit
leachate, reduces the hydraulic conductivity of
the layers. For the better efficiency and to filter
this leachate for extended period of times, it is
important to design these properly to optimize
their long term performance and service life.
One mean of extending this service life of the
leachate drainage system is to use uniformly
graded gravel. (Rowe et. al. 2004). Shredded
scrap tires have also been proposed as an
alternative drainage material, based on their high
hydraulic conductivity (Hall, 1991, Edil et. al.
1992, Ahmed and Lovell, 1993, Duffy, 1995)
METHODOLOGY:-
Characteristics of MSW were analyzed
by sampling of solid waste from the site. A
composite sample was made with collection of
different samples of different areas of waste
piles and mixing it for a total weight of 18.5 kg.
This collected sample is carried to the laboratory
at DAVIET for further analysis. Test samples
have been preserved in plastic bags. These
samples were weighed and manually separated
by hand to get various components of the sample
by weight. Leachate samples have been
collected from the disposal site. Immediately
after sampling, the leachate was tightly sealed in
plastic containers and then transported to the
laboratory in an ice pack and stored to room
prior to testing. The procedure was adopted to
prevent or reduce bacterial growth and calcite
flock formation within the leachate (Rowe et. al.
1995, Fleming. 1999). Various analysis of solid
waste and leachate were conducted according to
the procedures used and has been given in
manual for solid waste analysis published by
NEERI, Nagpur, India and Standard Methods for
Kaushik, Agnihotri and Bansal 643
Examination of Water and Wastewater by
APHA,AWWA (20th edition).
Three polyvinyl chloride (PVC)
columns were prepared, to determine the
treatment potential of the tire chips and its mixer
with gravel with an internal diameter of 20.32
cm and a total length of 160 cm. one column
was filled with tire shred donated as ‘T1 shreds’.
Another column was filled with tire shreds and
gravel mixer at a ratio in layers of 50-50 % each,
donated as ‘TG shreds’ and the third column was
filled with conventional uniformly graded gravel
with a nominal size of 38 mm. These materials
were loosely placed into the columns. The
normal flow rate of 0.4 m3/m2/day was selected
to stimulate an accelerated rate of mass loading
(to allow experiment to be completed in a
reasonable period of time), but no so high, as to
change the processes occurring in the
drainage/filter material. The test specimen was
mostly comprised of wire free tire chips, but also
included small percentage of about 10% of wire
containing tire chips. (1 to 10 % of the chips
included wires greater than 50 mm in length.
The experimental columns commenced
operation in June 2009. Leachate samples were
collected everyday for a six months time period
and analyzed for pH, EC, TS, DS, COD, BOD,
TOC, TKN, and Ammonia-N, Chlorides, Nitrate
as per the Standard Methods. Sulfate, Phosphate,
Sodium, Potassium, Calcium, and Magnesium
were also analyzed for the leachate samples
collected from the outlet of columns. Fresh
leachate samples were collected from the
warriana dump site. Leachate samples were also
collected at the outlet of the columns for 7 days
in succession after an initial lag of 20 days (time
required to develop biofilm), to study the
changes in the leachate with the flow period.
The effect of tire chips on leachate
composition was evaluated by comparing the
results for leachate collected from cells 1,2 and
3. Grab samples of leachate were collected
following the procedures described in ASTMD
6759. The leachate samples were analyzed for
the total concentrations of the following metals:-
Silver (Ag), Arsenic (As), Barium (Ba),
Cadmium (Cd), Chromium (Cr), Copper (Cu),
644 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
Iron (Fe), Mercury (Hg), Manganese (Mn),
Nickel (Ni), Lead (Pb), Zinc (Zn) and Selenium
(Se). These constituents were selected because
they fall in one or more of the following
groups:-
1. Metals, for which a regulatory toxicity
characteristic concentration has been
established as identified in USEPA 40
CFR 261.24 (i.e. Ag, As, Ba, Cd, Cr,
Hg, Pb, Se)
2. Metals contained in the National
Primary Drinking water standards as
identified in USEPA 40 CFR 141.11
(i.e. As, Ba, Cd, Cr, Cu, Hg, Ni, Pb and
Se).
3. Typical metals for which acceptance
criteria have been established by
leachate treatment facilities (i.e. Cu,
Hg, Cd, Ni, Pb and Zn).
Iron (Fe) and Manganese (Mn) were added
to the list of analysis because an increase in the
concentrations of these two metals would likely
indicate that oxidation of the steel wires in the
tire chips was occurring. Additionally, the
leachate samples were also analyzed for nitrates,
sulfates and phosphates which may likely to
leach either from tire chips or gravel. The
organic strength of leachate was also measured
as BOD and COD.
RESULTS AND DISCUSSIONS
Waste samples were collected from
Warriana landfill of Jalandhar city, Punjab
which is owned and operated by Municipal
Corporation Jalandhar and Punjab Grow More
Fertilizer Ltd., Jalandhar. The incoming waste
consisted of approximately 85% MSW, 13%
soils and 2% other types of waste. The waste
components included 51.31%, fruit and
vegetables waste, 7.08%, yard waste, 4.9%
paper, 4.9% plastic and 4% of rags, Less then
1% glass, rubber1%, and other miscellaneous
materials. It consisted about 60% of bio
degradable matter and remaining may be
attributed as non bio degradable materials.
Chemical properties of the waste indicate that
the C/N ratio is around 24.0. The average
moisture content in city waste was around
33.3%, while the average calorific value was
found to be 932 Kcal/kg.
The experiment columns used for the treatment
study were commenced into operation in June
2009. Tire shred samples processed under
different operation conditions do not posses the
same properties. Therefore, it was felt necessary
to characterize each tire shred type before being
considered for use as a leachate drainage
material. Leachate samples were analyzed for its
pH, EC, TS, VS, COD, BOD, TOC, DOC, TKN,
and Ammonia-N, Chlorides, Nitrate as per the
Standard Methods (APHA, 1998). Sulfate,
Phosphate, Sodium, Potassium, Calcium, and
Magnesium were also analyzed for the leachate
samples. The value for COD and BOD were
2030 and 930 mg/l respectively which indicate
presence of high organic matter content in the
leachate samples. A decreasing trend of BOD
and COD concentrations in leachate was noted
over a period of 6 months of leachate
recirculation and treatment. BOD and COD
concentrations varied from 200 to 1000 mg/l and
from 300 to 4000 mg/l over the same period.
The ratio of BOD/COD decreased from about
0.5 to 0.3 over a period of 6 months, which has
illustrated the reduction in the biodegradable
organic compounds and the increase in microbial
activities due to the increase in microbial slime
layer. Since, it takes time to establish the biofilm
(on the drainage material) needed to cause
significant leachate treatment, there is an initial
lag period, during which there is relatively little
difference between influent and effluent
concentrations and the effluent COD
concentrations were 90% or more of influent
concentrations. This accumulation period (lag
time) was similar for all columns and was less
than 20 days for both the gravel and tier shred
materials. Once, the biofilm was established,
there was a significant reduction in COD values,
as the leachate passed through the columns,
irrespective of the type of drainage material (tier
shreds or gravels). The similar reduction patterns
in BOD and dissolved solids were also observed
in all the columns. Effluent COD concentrations
were measured expect at times when, the
columns had much higher than normal influent

organic loading. At these times the effluent
concentrations were elevated (i.e. less relative
treatment) for all the columns. This is likely the
result of the Leachability tests have shown that
low levels of some organic compounds and
metals can be released from the tire shreds
(Zelibor 1991, Eading 1992, Humphrey etc at
1997, Humphrey and Katz 2001), but when
leaching did occur, the leachate compounds
were found at low concentrations (i.e typically
action levels). This indicates that testing and the
effects of tire shreds on landfill leachate quality
under different lab inability of the mixed
population of the bacteria to adapt to the
increased concentration within a short period of
time. The variations in the quality of leachate as
it passed through both the rubber shred and
gravel filled columns are consistent with the
leachate chemistry study by Rittmann et. al
(1996). To provide initial insight regarding
whether metals were being leachate from the
shred material as the leachate from the shred
material as the leachate passed through the
columns, influent and the effluent leachate from
the tire shred column were used for analysis.
Although, numerous studies have indicated that
metals can be released from tire shreds (Zelibor
1991; Ealding 1992; Humphray and Katz 2001),
no study has explained the uptake of these
metals which would potentially prevent them
from being released into the groundwater.
Although, low metal concentrations were
measured in the source leachate as discussed
above, the columns operated under high volumes
of leachate and for a long duration allowing for
potential buildup of metals in the clog material.
Tire shreds operating under these conditions
could both leach metals from the shreds and
accumulate them in the clog material. The total
mass of metals in leachate entering the column
over the period of the experiment was calculated
from the measured volume of leachate and its
chemical composition. It is inferred that the
majority of the mass of the metals Al, Zn, Fe,
Mn, Cu, and Ni originated from other source
other than from the influent leachate (i.e, the tire
shreds). Many studies Rome R. K. and Mclsaac.
R. (2005), Aydilek and Madden. et. al. (2006)
Kaushik, Agnihotri and Bansal 645
suggest that Al, Zn, Fe, Mn, Ni and Cu were
leached from the shreds but most became
immobilized in the clog material and did not exit
the columns in the effluent leachate. Although
the data collected in this study, provides useful
insights additional work is required to define
the leachability of tire shreds used in leachate
collection system
CONCLUSIONS
Samples of solid waste and leachate
were analyzed and the results shows that the
solid waste samples were having maximum of
organic type of waste and it was found nearly 60
% followed by comparatively less amounts of
paper, plastic, rags and glass as well as rubber
waste. The leachate samples analyzed shows
high vales of COD and dissolved solids.
The values obtain for Sulfates, Nitrates,
Sodium, Potassium and Calcium were also
found high which shows high nutrient content of
leachate samples strengthens the possibilities of
biological treatment of this leachate sample.
Iron, Zinc and Manganese values were also high
while all other metal values were comparatively
low in the leachate samples. These results
support the possibilities of biological treatment
of leachate by microbial layer provided on
suitable growth medium. For this purpose, three
columns were permeated with landfill leachate at
a normal flow rate, up to six months were used
in order to investigate the use of tire shreds with
a combination with gravel in the drainage layer
of a landfill. Different types of tire shred with
many exposed wires and uniformly graded
38mm gravel were examined. The following
conclusion was drawn from results of this
experiment. Some metals leached from tire shred
were exposed to typical MSW leachate. These
metals include Zn and Fe. However, they were
not detected in the influent leachate due either to
their very low concentration or due to the
precipitation into the clog material. The results
of these tests suggest that the gravel should
continue to be used in critical zones where there
is a high mass loading. In less critical zones the
results imply that an increased thickness of
compressed tire shreds may be used to give a
646 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

TABLE 01:-Results of leachate analyzed and re-circulated through gravel, tire chips (T 1) and mixer (TG).

Parameters/ Leachate Gravel Tier Mixer in Standards* (Land disposal)

Characteristics Characteristics (a) chip two layers ( as per mode of disposal)
(T1 shreds) (TG shreds) surface Public Land
(b) (c) water sewer disposal
Alkalinity 1580 1400 1650 1500 ----- --- ----
Hardness 4000 4400 3500 3200 ----- --- ----
Chloride 392 450 350 480 1000 1000 600
BOD 930 720 700 640 30 350 100
COD 2030 1840 1800 1640 250 --- ----
TOC 5380 4050 4000 4000 ----- --- ----
Dissolved solids 3100 2850 2800 2820 2100 2100 2100
Suspended solids 520 600 500 520 100 600 200
Turbidity(NTU)* 100 50 45 50 ----- --- ----
pH* 8 8.5 7.2 7.5 5.5 to 9 5.5 to 9 5.5 to 9
EC(ms/cm)* 10.8 11 11.2 8.8 ---- --- -----
NH3- N 230 120 100 105 50 50 -----
TKN 180 140 110 105 100 --- ----
NO3-N 20 18 15 15 --- --- ----
Total P 85 80 40 35 --- --- ----
Potassium 680 540 520 450 ---- --- ----
Calcium 1930 1950 1850 1850 ----- --- ----
Magnesium 150 125 120 130 ----- --- ----
Sodium 1200 1000 850 900 ----- --- ----
Sulfate 490 340 250 280 ---- --- ----
Iron 171 150 280 170 ---- --- ----
Zinc 120 100 140 120 5 15 ---
Manganese 125 110 100 90 ----- --- ---
Copper 0.5 0.5 0.5 0.4 3.0 3.0 ----
Lead 1.2 0.9 0.8 0.7 0.1 1.0 ----
Chromium 1.6 0.9 0.9 0.4 2.0 2.0 ---
Cadmium 0.5 0.4 0.2 0.2 2.0 1.0 -----
Mercury 0.9 0.8 0.6 0.5 0.01 0.01 -----
Arsenic 2.2 2.0 1.5 1.5 0.2 0.2 0.2
Barium 2.0 1.8 2.1 2.2 ----- --- ----
Nickel 0.1 0.1 0.4 0.2 3.0 3.0 -----
Aluminum 0.5 0.5 0.6 0.5 ---- --- ----
Boron 0.8 0.4 0.5 0.5 ----- --- -----
Note: All values other than * are in mg/l,
For a, b, c number of circulations is 10, for this initial level of study,
Standards as given in (Land disposal) as in MoEF notification, 25th September 2000.

service life similar to that of a given thickness
of gravel.
Leachate permeation through the tire was found
to leach certain elements (especially Mn, Zn, and
Fe) from the shreds, although the past studies
suggest that these metals quickly become
captured in the clog material and they were not
detected at elevated levels in the effluent
leachate. This require more investigation,
however the preliminary indication is that it
would not cause a problem with respect to
increased risk of metals getting into
groundwater. An extensive literature review in
regard to evaluation of the environmental
suitability of tire chips was made by Park et. al.
(2003). Their study indicated that Cr, Cd, Fe and
Mn were the only compounds that were
sporadically detected in laboratory leaching tests.
Rowe and Mclsaac (2005) reported possible
leaching of very low concentration of Al, Zn, Fe
and Cu. However they concluded that these
constituents were more likely to precipitate into
the clog material and get trapped in the tire chip
layer in the landfill system. All laboratory test
indicated that the concentration of analyzed
metals were below the drinking water standards
limits, expect the EP-toxicity test study
conducted by Grefe (1989), which suggest that
the Fe and Mn concentrations may be above or at
the safe disposal limits for leachate.
Even tough serval inorganic and organic
compounds were detected in tire chips passed
leachate concentrations were generally below
and were not significantly higher than the safe
Disposal limit’s for landfill leachate given in
MoEF notification of 25th September 2000.
Furthermore, neither of the two field studies
conducted by the Minnesota Pollution Control
Agency (1990) and Humphrey and Katz (1995)
indicated the presence of groundwater or surface
water contamination due to tire chips placement.
Additional field data provided by Park et. al
(2003) and Edil et. al. (2004) along with the
results obtained in the current study suggests that
Kaushik, Agnihotri and Bansal 647
tire chips can be safely used in leachate
treatment facilities.
REFERENCES:-
Ahmed, I., & Lovell, C. W. (1993), Tire
chips as permeable aggregate in landfills. Proc.,
1st Annual Great Lakes
Geotechnical/Geoenvironmental Conf., Univ. of
Toledo, Toledo, Ohio, 83–90.
ASTM (1998), Standard Practice for Use of
Scrap Tires in Civil Engineering Applications-
ASTM D 6270-98, American Society for
Testing and Materials, PA, 19p.
Benson, A., Warith, M., Evgin, E., &
Moore, R. (2002). “Suitability of shredded tires
for use in landfill leachate collection systems.”
Ground and Water: Theory to Practice; Proc.,
55th Canadian Geotechnical and 3rd Joint
IAH–CNC and CGS Groundwater Specialty
Conf., Niagara Falls, Ontario, Canada, 565–572.
Duffy, D. P. (1995). Using tire chips as a
leachate drainage layer. Waste Age, 26(9), 113–
122.
Edil, T. B., & Bosscher, P.J. (1994),
Engineering Properties of Tire Chips and Soil
Mixtures, Geotechnical Testing Journal, 17(4),
453-464.
Edil, T. B., Fox, P. J., & Ahl, S. W. (1992).
Hydraulic conductivity and compressibility of
waste tire chips. Proc., 15th Annual Madison
Waste Conf., Univ. of Wisconsin–Madison,
Madison, Wis., 49–61.
Edil, T. B., Park, J. K., & Kim, J. Y. (2004).
Effectiveness of scrap tire chips as sportive
drainage material. J. Environ. Eng., 130(7),
824–831.
Evans, P. A. (1997). Use of tire shred in
landfill construction. Proc., 3rd Annual Symp on
Environmentally Friendly Technologies in
Geotechnical Engineering, Geotechnical Society
of Edmonton, Alta., Canada.
Fleming, I. R., Rowe, R. K., & Cullimore,
D. R. (1999). Field observations of clogging in a
landfill leachate collection system. Can.
Geotech. J., 36(4), 685–707.
648 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
Grefe, R. (1989). Review of waste
characterization of shredded tires.
Interdepartmental memorandum, Wisconsin
Dept. of Natural Resources, Wis.
Hall, T. J. (1991). Reuse of shredded tire
material for leachate collection systems. Proc.,
14th Annual Madison Waste Conf., Madison,
Wis., 367–376.
Humphrey, D. N., & Katz, L. E. (2001).
Field study of water quality effects of tire shreds
placed below the water table. Proc., Conf. on
Beneficial Use of Recycled Materials in
Transportation Applications, Air and Waste
Management Association, Pittsburgh.
McIsaac, R., & Rowe, R. K. (2005). Change
in leachate chemistry and porosity as leachate
permeates through tire shreds and gravel. Can.
Geotech. J., 42(4), 1173–1188.
Moo-Young, H., Sellasie, K., Zeroka, D., &
Sabnis, G. (2003). Physical and chemical
properties of recycled tire shreds for use in
construction. J. Environ. Eng., 129(10), 921–
929.
Park, J. K., Kim, J. Y., Edil, T. B., Huh, M.,
Lee, S. H., & Lee, J. J. (2003). Suitability of
shredded tires as a substitute for a landfill
leachate collection medium. Waste Manage.
Res., 21(3), 278–289.
Reddy, K. R., & Saichek, R. E. (1998a).
Assessment of damage to geomembrane liners
by shredded scrap tires. Geotech. Test. J., 21(4),
307–316.
Reddy, K. R., Stark, T. D., & Marella, A.
(2008). Clogging potential of tire shred-drainage
layer in landfill cover systems. Int. J. Geotech.
Eng., 2(4), 407–416.
Reddy, K., & Marella, A. (2001). Properties
of different size scrap tire shreds: Implications
on using as drainage material in landfill cover
systems. Proc., 17th Int. Conf. on Solid Waste
Technology and Management, Philadelphia.
Reddy. K., & Saichek, R. (1998a).
Characterization and performance assessment of
shredded scrap tires as drainage materials in
landfills.” Proc., 14th Int. Conf. on Solid Waste
Technology and Management, Philadelphia.
Rittmann, B. E., Fleming, I. R., & Rowe, R.
K. (1996). Leachate chemistry: Its implications
for clogging. Proc., North American Water and
Environment Congress ‘96, CD-ROM,
American Society of Civil Engineers, Anaheim,
Calif.
Rowe, R. K., & McIsaac, R. (2005).
Clogging of tire shreds and gravel permeated
with landfill leachate. J. Geotech. Geoenviron.
Eng., 131(6), 682–693.
Rowe, R. K., Armstrong, M. D., &
Cullimore, D. R. (2000b). Particle size and
clogging of granular media permeated with
leachate. J. Geotech. Geoenviron. Eng., 126(9),
775–786.
Rowe, R. K., Quigley, R. M., Brachman, R.
W. I., & Booker, J. R. (2004). Barrier systems
for waste disposal facilities, E & FN Spon,
London.
Rubber Manufacturers Association. (1990).
“RMA leachate study.” Rep., Radian Inc.
Warith, M. A., Evgin, E., & Benson, P. A.
S. (2004). Suitability of shredded tires for use in
landfill leachate collection systems. Waste
Manage.,24,967–979.
 Sand Filter System: a Solution for Domestic Sewage Disposal Problem
in the Niger Delta
S. U. EJEZIE
Department of Civil & Environmental Engineering, University of Port Harcourt, Box 415
Choba, Port Harcourt, Nigeria (samejezie@yahoo.co.uk)
J. O. M. Amasuomo
Department of Civil & Environmental Engineering, University of Port Harcourt, Nigeria

ABSTRACT The relative effectiveness of the “Soak-away pit” system and the “Sand-filter” system in the
disposal of domestic waste in the riverine area of the Niger Delta has been investigated as presented in this
paper. Field and laboratory test results indicate that the soil underlying the area is silty Clay with low
acceptance rate (5.21litres/m2/day), low percolation rate (218 min/100mm), high ground water table and, in
most cases, prone to flooding above grade level in the rainy season. These results further reveal that the soil
is semi-impervious and hence the soak-away pit system would be of very low efficiency in domestic waste
disposal in this area. The “Sand-filter” system, on the other hand, is portrayed by the results as having a high
potential of providing the required treatment capacity for this soil type and hence of being efficacious for
domestic sewage disposal in the deltaic environment. It is therefore strongly recommended in preference to
the “Soak-away” pit system.

As public awareness on environmental pollution

INTRODUCTION
The Soak-away pit system of disposing domestic
sewage is a common practice in Nigeria. In this
method, the domestic sewage from an individual
residence is piped into an underground septic tank
and allowed to settle to form sludge at the bottom.
The effluent (liquid waste) is discharged into an
adjoining soak-away pit from where it is absorbed
into the subsoil through the side openings (Obande,
1983; Jerus, 1983; Metcalf and Eddy, 1995; and
Henry and Heinke, 1996).
With urbanization and improved condition of
living, there has been proliferation of the soak-
away pit system with attendant environmental
pollution problems. The environmental pollution
arises from inappropriate disposal of the domestic
sewage through the soak-away pit system. The
inappropriately disposed sewage contaminates
both the soil and the underground water harnessed
for drinking, and also destroys the aquatic life of
streams especially where the waste finds its way
into such water bodies.
grows, the effect of wastes released from burst
septic tanks and soak-away pits in polluting land,
streams and air becomes a serious concern to
public health and environmental managers. The
effort to solve the problems brings to focus the
issue of selection and design of domestic sewage
disposal systems that will be appropriate for
various types of soil. This is because the
performance of a particular system may be
satisfactory in a given soil medium but not suitable
for another type of soil with different geotechnical
and hydro-geological characteristics. Therefore,
for any disposal system to be effective in terms of
cost, efficiency and durability, its selection and
design should meet some design criteria, specified
in terms of such parameters as flow-rate, soil type
(as indicated by moisture content and gradation),
groundwater table, soil porosity, acceptance rate of
soil and percolation rate of the effluent in the soil.
These parameters have been exhaustively
discussed by previous researchers. According to
Lui and Evett (1990), moisture content has been
established to be an indicator of the average
amount of water present in the soil. In the same
vein, Jackson and Dahir (1996) conclude that soil
650 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
porosity indicates the proportion of the total
volume of the soil that is not occupied by solid
constituents. Soil gradation is defined by Bowles
(1982) and Lui and Evett (1990) as a soil
identification and classification index which is also
used in the evaluation of soil stability and for other
specific applications. The groundwater table
delineates the depth below ground surface of the
upper limit of the zone of saturation where the
water is free to move or flow under gravity
(Bowles, 1984). The flow rate is generally well-
known, and represents the quantity (in litres) of
domestic sewage generated per person per day. It
is used extensively in the determination of the size
of domestic sewage systems. Metcalf and Eddy
(1995) define acceptance rate of soil as an
indicator of the ability of the soil to absorb or
assimilate water. They further explain that this
parameter is generally used to estimate the amount
of water that can be transported away from the
sewage disposal system. Based on their findings in
the same research work, they state that percolation
rate (relative absorption) of soil is the rate at which
water moves through soil in operational effluent
disposal systems.
Definition of the Problem
There is a general assumption that the soak-away
pit system of domestic sewage disposal is always
very appropriate. Sometimes it is even ranked
above other methods like the disposal field bed
(field tiles) and the sand filter systems. As a result,
the soak-away pit is generally used for domestic
sewage disposal irrespective of the soil type. This
practice is widespread despite the shortcomings of
the system as stated by Jerus (1983). According to
him, the system cannot be used in soils rated as
semi – impervious, and in soil deposits where the
water table is not at least 2.4m below grade level
or 600mm below the bottom of the soak-away pit.
He is also of the view that the system should not
be located within 30m of portable water supply or
within 5m of the building it serves.
This assumption on the appropriateness of the
soak-away pit system is as a result of limited
engineering and scientific information available
with regard to other types of domestic sewage
disposal systems such as field bed and sand filter.
Presently however, extensive background data are
available showing that disposal systems that are
not suitable for a particular soil type may
nevertheless be effective for the disposal of
domestic sewage in other soil types.
In Nigeria, the effective disposal of domestic
sewage has become an important issue both in
terms of public health and good environment.
Consequently, an assessment of the relative
effectiveness or otherwise of the soak-away pit
system in the riverine communities, with their
peculiar geographical terrain, is deemed necessary.
FIELD SURVEY AND DATA COLLECTION
Description of study area
The study site is the Bayelsa State Housing and
Property Development Authority’s Housing Estate,
located along Azikoro Road, Ekeki, Yenagoa. The
estate consists of 152 plots of land, each measuring
38m x 30m (1140m2). Each plot has a four-
bedroom detached bungalow building with a floor
area of 228m2 and occupies about 22% of the plot.
According to Amasuomo (1983) and Makinwa
(2000) the population of the estate is 1064 persons,
made up of members of 152 residential families,
with an average family size of seven, comprising
the father, mother and five children per family unit.
Data Collection
To accomplish the task stated above in the
problem definition, data were collected on
moisture content, porosity, gradation, ground
water table, acceptance rate and percolation rate of
the soils of the study area. These soil properties
were targeted because of their perceived
characteristic influence on the selection and
design of domestic sewage disposal systems.
Moisture content was determined from a
representative soil sample of about 1000g collected
during the dry and wet seasons, while the
gradation was obtained using a representative soil
sample obtained from a trial pit. The determination
of the soil porosity was achieved using a

representative sample of soil with a volume of
about 1000cm3. The ground water table, on the
other hand, was determined by drilling a well of
600mm diameter into the water-bearing formation
and measuring directly by means of a measuring
tape.
The acceptance and percolation (relative
absorption) rates of the soil were obtained from
field tests conducted in the study area, which
yielded data employed in their calculation. The
flow rate, as stated earlier, was evaluated in terms
of the quantity (in litres) of domestic sewage per
person per day.
RESULTS AND DISCUSSION OF FINDINGS
The results of the analysis of data collected in this
study are presented in table 1.
From the results, the moisture content of 72%,
obtained as average value for the dry and wet
seasons, is an indication that the soil is of high
water content and low water absorption. These
features are not suitable for the soak-away pit
system (Metcalf and Eddy, 1995; and Jerus, 1983).
The result of the gradation analysis – the
preponderance of soil grains of less than 0.074mm
diameter – indicates that the soil is silty Clay. This
soil type is associated with a characteristic low
permeability and therefore not suitable for soak-
away pit system. This is despite the relatively
appreciable porosity value of 0.49 because a fine-
grained soil generally has little or no permeability.
This low permeability is primarily owing to the
small and narrow size of the pore throats, with
their associated flow-inhibiting surface tension
forces, typical of fine-grained soils. A suitable soil
type should have a grain size range between
0.074mm and 2mm, which is considered normal
for coarse to fine Sand; and silty Sand to dirty
Sand (Henry and Heinke, 1996; and Metcalf and
Eddy, 1995).
The ground water table (GWT) during the rainy
season is between 0.15mm and 0.30mm below
grade level and in some cases flooding of between
0.10m and 0.5m above grade level. During the dry
season however, the GWT drops to between 2.0m
Ejezie and Amasuomo 651
and 3.0mm below grade level. This relatively high
GWT creates absorption problems for any soak-
away disposal system.
The average percolation rate of the soil was
218min/100mm which is greater than the
acceptable value of 118min/100mm. This also
makes the soak-away pit system unsuitable.
The acceptance rate of 5.21 lit/m2/day was less
than the recommended value of 8 lit/m2/day and
above. It also indicates the non-suitability of the
soil for soak-away pit system.
SELECTION / DESIGN OF APPROPRIATE
DISPOSAL SYSTEM
Selection of disposal system
The findings and results of data analysis
summarised in table 1 of the previous section have
indicated that all the parameters required for a
suitable soak-away pit system were not obtainable
at the study site. This narrows the selection of
appropriate system to either the field disposal or
the sand filter system. However, the field disposal
system is also not suitable because the available
land space is limited, while the soil is semi-
impervious with high ground water table. As a
result, the sand filter system is chosen as the most
suitable in line with the views of Metcalf and Eddy
(1995), and Cooke (1994).
Design of the Sand Filter disposal system
Details of the criteria and specifications for the
design of the sand filter system have been
summarized as presented in Table 2 below. These
are in the form of specific parameters considered
appropriate for achieving optimum performance of
the system in domestic sewage disposal.
Detail Design/Calculations
The design of sand filter involves the
determination of the following:
1. Area of Sand Filter (As)
Fr 1210lit / d
As = = -- (1)
HL 24lit / m 2 / d
§ 50m2
652 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

Therefore recommended size of sand filter 5. Total flow in each lateral for 4 orifices
according to space available in the case study site per lateral
is 54m2 (10,000mm x 5,400mm). This is given by: n x 0.265 lit/sec
= 4x0.265 lit/sec. = 1.06 lit/sec.
2. Layout of Effluent Distribution System.
From the computed value of the area of the sand 6. Head loss through pipes from orifice ( h fp )
filter (10.0 x 5.4) m2, and the parameters in table 2,
the layout recommended shall consist of 16 laterals h fp = 189.18 ( L1,n ) (Q/C)1.85 D-4.87 -- (4), where:
of 2.1m long on each side of the main pipe and
four orifices equally spaced at 600mm centres on
each lateral (Fig 2).
h fp = Actual head loss through pipes from L1 to Ln (m)
L1,n = Effective length of lateral pipe (distance
3. Flow Discharge per Lateral (FL)
between first orifice, L1 and last orifice, Ln ).
Flowrate( Fr )
FL = --- (2) Q = Pipe discharge or total flow rate (1.06 lit/sec).
NumberofLaterals( Ln )
C = Hazen – William discharge coefficient (150
1210 L / d for plastic pipes).
= = 76 lit/day/lateral
16 Laterals 189.18 = Conversion factor.
D = Diameter of orifice (3mm).
4. Flow rate in the Last Orifice ( qn )
Therefore:
= 8.9CD2 2 ghn --- (3)
h fp = 189.18(2.4  0.3)(1.061 )1.85 (0.003) 4.87
150
Where qn = Discharge from orifices in litres per = 0.058m
second.
8.90 = Conversion factor, where discharge is 7. Head loss in actual distribution pipe ( h fdp )
expressed in m3/sec, diameter of
orifice in mm, and velocity in m/s. h fdp = 1/3 h fp - (5)

C = Orifice discharge coefficient for holes = 1/3 x 0.058m

drilled in the field (0.61).
= 0.019m
D = Diameter of orifice (3mm).
g = Acceleration due to gravity (9.81 m/s). 8. Head loss in the first orifice ( h1 )

hn = Orifice discharge head (1.5m). = hn + h fdp - (6)

n = Number of orifices per lateral (4No). Where hn = orifice discharge head (1.5m)
Therefore: h1 = (1.5 +0.019) m = 1.519m
2
qn = 8.90 (0.61)(0.003) 2 x 9.81 x 1.5
9. Decimal value less than 1 (m)
= 2.65 x 10-4m3/s or 0.265lit/sec
hn
m = - (7)
h1

1.50
= ¥( ) = 0.9937
1.519
Note: The calculated decimal value (m) of 0.9937
is above the critical value, 0.9800, below which
the effectiveness of the sand filter system shall be
affected. The m – value is also used to calculate
the difference in discharge of treated effluent
between the first and the last orifices in each
lateral pipe.
10. Percentage Difference in discharge
between the first and last orifices in
each lateral ( Ddfl )
Ddfl (%) = (1 – m) x 100 - (8)
= (1 – 0.9937) x 100
= 0.63 %
Comment: The value of 0.63% calculated above
shows that the difference in time lag in the
discharge of treated effluent between the first and
last orifices in each lateral pipe is well below the
specified maximum value of 2% (Metcalf and
Eddy, 1995). The distribution system therefore
does not require re-sizing and can effectively
discharge the effluent for treatment in the sand
filter system with a flow rate of 1210 lit/day and a
peak factor of 4 for a residential unit for 7 persons.
The working drawings of sand filter are shown in
Figs 1 and 2.
SUMMARY AND CONCLUSION
The sand filter system has been determined to be
suitable for soil described as “impervious with
high ground water table” while the soak-away pit
system was found to be suitable only for soils that
indicated “rapid, medium or slow relative
absorption and low ground water table”. Because
the soils in the riverine areas are predominantly of
silty Clay type, which exhibits semi-
imperviousness, the sand filter was selected as
appropriate domestic sewage disposal system for
the region.
The sand filter system has some significant
advantages over the soak-away pit system in the
Ejezie and Amasuomo 653
effective treatment of domestic sewage onsite.
These include the fact that it does not depend on
low ground water table and the imperviousness of
the soil. The soak-away pit system, on its part, can
only provide some level of effluent treatment in
dry season and little or no treatment in the wet
season. In conclusion therefore, it can be stated
that the sand filter system is the most appropriate
for the study location.
REFERENCES
x Amasuomo, J.O.M. (1983) Development of a
residential estate in rainbow town. Unpublished
B.Tech. Environmental Design Project, Dept of
Architecture, Rivers State University of Science
and Technology, Port Harcourt.
x Bowles, J.E. (1984) Physical and geotechnical
properties of soil. McGraw – Hill International
Book Company Japan.
x Cooke, L. J. Jr. (1994) Sewage and drainage -
septic drainage fields in Ramsey, C.G and Sleeper,
H.R (1994) Architectural graphic standards 9th
edition. 144-146
x Henry, G and Heinke, G.W. (1996)
Environment Science and Engineering. Prentice
Hall Inc. New Jersey
x Jackson, N and Dahir, R.K (1996) Civil
engineering materials. McMillian Press Ltd.
London.
x Jerus, G.R (1983) Plumbing – sewage disposal
In Calendar, J.H (ed) Time saver standards for
architectural design data 6th edition. McGraw-Hill
Book company, New York. 4/241 – 250
x Lui, C and Evett, J. B (1990) Soil properties
testing measurement and evaluation. Prentice Hall
Inc. New Jersey
x Makinwa, M (2000) Finding the ‘Missing’
Nigerians in the year 2001. Saturday Punch
January 29 Pp 4.
x Metcalf and Eddy (1995) Waste water
engineering - treatment, disposal and reuse.
Revised by Tchobanoglous, G and Burton, F.C.
Tata McGraw–Hill Publishing Company, New
Delhi.
x Obande, M.O (1983) Block laying and
concreting. Longman Nigeria Ltd, Lagos.
654 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

Table 1: Summary of results of data analysis

Assessed parameters Results and/or observations Recommended Value Remarks

Flow rate for a family unit (7 1210 litres/day As calculated Used for computing size of
persons) sewage system

Moisture content (average 72% < 50% Result indicates high moisture
value) content and low water
absorption. Soil not suitable for
soak-away pit system.
Porosity 0.49 <0.49 High porosity and low
permeability soil. Not suitable
for soak-away pit system.
Gradation analysis of soil Above 0.074mm: 15.6% Below 0.074mm to 2mm (Coarse to Result indicates silty Clay soil.
0.074mm: 84.4% Classification: fine Sand, silty Sand to dirty Therefore not suitable for soak-
silty Clay. Sand) away pit system.
Groundwater table below grade 2.4m below grade level or High water table.
level: Observations: 600mm below bottom of Not suitable for soak-away pit
(i) Dry season 2.0m - 3.0m. soak-away pit system.
(ii) Wet season 0.15 - 0.3m.
(iii) Flooding Up to 0.5m above grade.
Percolation rate (Average 218 min/100mm < 118min/100mm Result Indicates semi-
value) impervious soil such as silty
Clay. Not suitable for soak-
away pit system.
Acceptance rate 5.21 lit/m2/day ' 8 lit/m2/day Not suitable for soak-away pit.

Table 2: Design Criteria and Specifications for Sand Filter System

Parameters Recommended specification

Peak factor (Pf) 4
Flow rate (Fr) for 7 persons 1210 lit/day
Main and under drain pipe size 100mm diameter
Lateral pipe size 50 mm diameter
Orifice size (lateral pipes) 3mm
Orifice size (under drain pipes) 12mm
Orifice spacing 600mm

Orifice discharge head ( hn ) 1.5m

Lateral pipe spacing 1.2m

Hydraulic loading (HL) 24 lit/m2/day
Difference in discharge between the first and last
orifice < 2%
 Ejezie and Amasuomo 655

1:
656 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

2:
 GHIJKLMNOPNQIORKLKSKTNUOKRNLHIVNWRXYKLXYNZ[SVITSKOLN\RILS
_`abc

 
  
    
  
defeghij`kcll`i`m
!"!   #$ !"  

W]U^_W`^ Nnmogpqcriopgpcss`tiuorgiagqmipgk`kotgvotogiaiqi`qorgvmoag`gktolisia`twgpo`txmgyztg
mo`jia{gzygpzilgwiolrorglisiqorgr`q`giagqmogpcttzcaria{gzyg|lcsiacsgo}qt`xqizagkl`aqegnmotoyzto~g
topo`txmgv`pgyzxcporgzagqmogcarotpq`aria{gqmogx`cpogzygmo`jogiagqmogazapvollia{g€`zliaiqixgpzilgrcog
qzg `l€`lie g nmo g yo`qctop g iaxztkzt`qor g szaiqztia{ g iaro} g ktzkotqwg `ar g jzlcso g xm`a{o g om`jizct g zyg
pkoxisoa g xzaq`sia`qor g viqm g ‚arcpqti`l g pkoaq g liƒczte g |azts`l g pvollia{ g m`p g ooa g azqixor g vmoag
pkoxisoagv`pgiacar`qorgviqmg‚arcpqti`lgpkoaqgliƒcztgpzlcqizaegnmogiarcxorgpvollia{gipg`qqticqorgqzg
qmogyzts`qizagzygaovgsiaot`lp~gvmixmgm`pgooagxzayitsorgwg„t`wgriyyt`xqizagk`qqotpeg|qqoskqpgm`jog
`lpzgooagrzaogqzgxzaqtzlgqmogcacpc`lgmo`jogviqmglisogpq`iliu`qizae

ab^_cde`^acb m`jo g `lpz g ooa g rzao g qz g xzaqtzl g qmo g cacpc`lg

‚qgipgycar`soaq`lgrcqwgzyg{ozqoxmaix`lgoa{iaootg
qz g p`yo{c`tr g qmo g xijil g oa{iaootia{ g pqtcxqctopg
`{`iapq g r`s`{o g topclqia{ g ytzs g l`t{o g jzlcsog
xm`a{op g ia g qmo g yzcar`qiza g pzile g …ipqtopp g qzg
pqtcxqctopgrcogqzgckliyqsoaqgzygyzcar`qizapg`lpzg
x`a g zxxct g rco g qz g cao}koxqor g a`qct`l g `arg
s`as`ro g y`xqztpe g †zaq`sia`qiza g zy g qmog
yzcar`qizagpzil grco gqzgqmo gpkill`{o gzygktzxoppg
ylcirpgiagiarcpqti`lgkl`aqpgipgazqg`gt`togiaxiroaqg
ia g rojolzkia{ g xzcaqtiope g ‚aqot`xqiza g zy g pc
{t`rogpzilgviqmgqmogkzllcq`aqpgx`agiaylcoaxogqmog
oa{iaootia{ g om`jizct g zy g pzilp g `ar g x`cpog
ripqtopp g qz g qmo g y`xiliqiop g xzapqtcxqorg
‡htirm`t`a g oq g `le~ g ˆ‰Šˆ‹ g Œ`z g  g Œ`z~ g ˆ‰‰Žeg
†l`wow g pzilp g `to g szto g ktzao g qz g `qq`x€ g wg
xmosix`lpg`pgqmowg`togoloxqtzxmosix`llwg`xqijoeg
nmip g x`a g o g qtco g ia g zqm g o}k`apijo g `ar g aza
o}k`apijogpzilpegnmogpqcrwgv`pgiaiqi`qorgvmoag`g
ktolisia`twgpo`txmgyztgmo`jia{gzygpzilgwiolrorg
lisiqor g r`q`e g nmotoyzto~ g topo`txm g v`p g yzxcporg
zagqmogcarotpq`aria{gqmogx`cpogzygmo`jogiagqmog
pzil g pcttzcaria{ g qmo g |lcsiacs g o}qt`xqizag
kl`aqe g g | g roq`ilor g o}kotisoaq`l g ktz{t`s g v`pg
carotq`€oa g qz g szaiqzt g qmo g pkoxisoa g iaro}g
ktzkotqiop g `ar g xzapzlir`qiza g om`jizct g carotg
riyyotoaqgxzariqizag`arg`a`lwpia{gqmogom`jizctg
iagli{mqgzyg„Œ…g`argnmots`lg`a`lwpipeg|qqoskqpg
mo`jogviqmgxmosix`lgpq`iliu`qizagsoqmzre
fW^Z_aWgUNWbdNfZ^GcdUN
UOKRNILhNiRXKhNejHhN
nmo g l`zt`qztw g pqcriop g tokztqor g ia g qmip g k`kotg
voto g carotq`€oa g za g vo`qmotor g tor g o`tqm g pzilg
zq`iaor g ytzs g qmo g |lcsiacs g o}qt`xqiza g kl`aqg
piqoegnmogpzilgv`pgxzlloxqorgwgzkoago}x`j`qiza~g
ytzsg`grokqmgzygzaogsoqotgytzsga`qct`lg{tzcarg
lojole g nmo g kmwpix`l g ktzkotqiop g zy g pzil g `tog
tokztqor g iag hij`kcll`i`m g  g _`abcg ‡‘‘’e g nzg
pqcrw g qmo g oyyoxq g zy g iarcpqti`l g `l€`liao g pkoaqg
liƒcztgpzlcqizagzagoa{iaootia{gktzkotqiopgzygtorg
o`tqm~gpkoaqgliƒcztgv`pgtzc{mqgytzsg|lcsia`g
o}qt`xqizagdl`aq~g‚“…|”~g‚ari`eghkoaqgliƒcztgipg
m`jia{ g ` g xzaxoaqt`qiza g zy g Že’ g “ g `ar g ipg
xzaq`iaia{g‘•g`lcsia`gwgvoi{mqgzygpzlcqiza~g
ipgcporg`pgiacar`qia{gylcirgiagqmogpqcrwe
fZ^GcdUN
WSSHVkHVMNgKYKSj
hkoxisoap g voto g si}or g viqm g iarcpqti`l g pkoaqg
liƒczt g `p g szclria{ g ylcir g `ar g kl`xor g iag
kzlwoqmwloao g `{p g `ar g pqztor g ia g `itqi{mqg
xzaq`iaot gyzt gxctia{gkotizrgzygzaogr`w~ gpojoag
658 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

r`wp~ g qmtoo g szaqmp g `ar g –g szaqmp g ia g ztrot g qzg hiaxogqmogpkoxiyixg{t`jiqwgzygqmogs`qoti`lpg{oqpg

zpotjo g qmo g oyyoxq g zy g iaqot`xqiza g qiso g kotizreg xm`a{or g rco g qz g xzaq`sia`qiza~ g qmog r`q` g `tog
”iƒcir g lisiq~g kl`pqix g lisiq g `ar g pmtia€`{o g lisiqg toktopoaqorgwgjotqix`lgpqt`ia gHg `{`iapqgjotqix`lg
qopq g m`p g ooa g x`ttior g zcq g `p g kot g qmo g |hn_g xzapzlir`qizagpqtoppgiagqmipgk`koteg
hq`ar`trpe
_ZUeg^UNWbdNdaU`eUUacbN
^VHISYHLS N lKSm N WRnIRK N POV N oHVPOVYKLM N SmHN nmogxm`a{opgiagqmogiaro}gktzkotqiopg`argiagqmog
ompjKTOTmHYKTIRNILhNYKLHVIROMKTIRNSHjS siaot`lz{w g zy g a`qct`l g tor g o`tqm g rco g qzg
Œorgo`tqmgiagqmog`qxmgzyg’‘g{gv`pgoƒcilit`qorg xzaq`sia`qiza g viqm g iarcpqti`l g pkoaq g liƒczt g ipg
viqm g iarcpqti`l g pkoaq g liƒczt g pzlcqizap g ‡pzlir—g ripxcpporgiagqmogyzllzvia{gpoxqizaeg
pzlcqizagt`qiz~ gˆ— ge’ gyzt g˜‘g r`wpe g nmog `l€`lig
qto`qor g tor g o`tqm g v`p g `it g rtior~ g kzvrotor g qzg
WSSHVkHVMNgKYKSjNOPNWRnIRKN`OLSIYKLISHhN
k`ppgŽ’Psg‚ari`aghq`ar`trgpiojog`argpqztorgiag
QIORKLKSKTNUOKRj
kzlwqmoaog`{pegnmipgqto`qorgpzilgv`pgcporgyztg
hzilgqto`qorgviqmgiarcpqti`lgpkoaqgliƒcztgpmzvpg
pkoxiyix g {t`jiqw g qopq~ g v`qot g `rpztkqiza g `arg
`agiaxto`pogiagliƒcirglisiqgj`lcog`pgxzsk`torgqzg
pkoxiyix g pcty`xo g `to`~ g `ar g yzt g sixtz g lojolg
caqto`qorgpzilg‡n`logˆegnmogkl`pqixiqwgiaro}gzyg
pqcriop g li€o g „t`w g riyyt`xqiza g `a`lwpip~ g `arg
pzilgxzaq`sia`qorgviqmgiarcpqti`lgpkoaqgliƒcztgipg
qmots`l g `a`lwpipe g …oq`ilor g ropxtikqizap g zy g qmog
pli{mqlw g iaxto`pore g nmip g ip g rco g qz g pcyyixioaqg
soqmzr g `to g {ijoa g ia g hij`kcll`i`m g  g _`abcg
iaxto`pogiagliƒcirglisiqgzygpzilpg`yqotgqto`qsoaqg
‡‘‘’g`arghij`kcll`i`mgg_`abcg‡‘‘™g
viqm g `l€`lie g ‚q g ip g xzssza g carotpq`aria{ g qm`q~g
mi{motgqmogliƒcirglisiq~gmi{motgipgqmogpmtia€`{oeg
cLHNqNdKYHLjKOLIRN`OLjORKhISKOLN^HjSjN ‚agx`pogzygtorgo`tqmgqto`qorgviqmgiarcpqti`lgpkoaqg
šao g › g risoapiza`l g xzapzlir`qiza g qopq g votog liƒczt~ g azg pmtia€`{o g v`p g zpotjore g hmtia€`{og
xzarcxqorgzagpzil gp`sklop gcpia{gxzajoaqiza`lg om`jizct g ip g `lszpq g p`so g yzt g zao g r`wg xctia{g
zorzsoqotg`kk`t`qcpg`qgdtzxqztœpgs`}iscsgrtwg `argpi}gszaqmgxctia{gzygtorgo`tqmgp`sklopeg
roapiqw g zy g caxzaq`sia`qor g pzile g |ll g qmo g qopqpg
voto gxzarcxqor gckgqzg`gktoppcto gzygŠ‘‘g€d`eg
rISHV N WhjOVoSKOL N `IoITKSp N OP N WRnIRKN
nmogiacar`qia{gylcirgv`pgiarcpqti`lgpkoaqgliƒcztg
`OLSIYKLISHhNQIORKLKSKTNUOKRj
xzlloxqor g ytzs g yiolre g nmip g o}kotisoaq g v`pg
nmo g pzil g szipqcto g xzaqoaq g v`p g roqotsiaor g `qg
ropi{aor g qz g tia{ g zcq g qmo g oyyoxq g rctia{g
qmtoogriyyotoaqgtol`qijogmcsiriqwg‡ŒŸgqzg€azvg
xzaq`sia`qizae g zt g xzsk`tipza~ g qmog
qmogxm`a{opgiagqmogv`qotg`rpztkqizagx`k`xiqwgzyg
xzapzlir`qiza g qopqp g za g pzilp g toszclror g `arg
pzil g pkoxisoa g za g iaqot`xqiza g viqm g iarcpqti`lg
iacar`qorgviqmgv`qotgvotog`lpzgxzarcxqoregnzg
pkoaqgliƒczte
`ppopp g qmo g oyyoxq g zy g liso g `p g ` g pq`iliuot g yztg
xzaqtzllia{ g qmo g mo`jia{ g a`qcto g zy g pzilg
xzaq`sia`qor g viqm g pkoaq g liƒczt~ g pzil g p`sklopg n| ”¡gˆ—g‚aro}gktzkotqiopgzygcaxzaq`sia`qorg`arg
xzaq`sia`qorgŒorgo`tqmgviqmgpkoaqgliƒcztg`yqotgxctia{g
votogsi}orgviqmgpq`iliuia{g`{oaqg`argripqillorg
yztgriyyotoaqgqisogkotizrp
v`qotegnmogsi}qctopgvotogloyqgyztgxctia{gyztg`g
kotizr g zy g zao g voo€e g |yqot g qmo g xzskloqiza g zyg
xctia{ g kotizr~ g p`sklop g voto g toszclror g iag nopq …ct`qiza ¢iqmzcqg hkoaqgliƒczt
`rriqijop
xzapzlir`qizagtia{pgqzgdtzxqztžpgs`}iscsgrtwg
””gg‡•

roapiqw g zy g caxzaq`sia`qor g pzil g `ar g iacar`qorg ˆg…`w ˜Š ŽŽ

viqmgiarcpqti`l gpkoaq gliƒcztegnmo goyyoxqijoaoppg ™g…`w ˜Š ’‘
zy g qmo g pq`iliuia{ g `{oaqp g v`p g o}`siaor g wg ˜gszaqm ˜Š ’
xzsk`tia{ g qmo g xzapzlir`qiza g om`jizct g zy g torg –gszaqm ˜™ ’
o`tqmgxzsk`xqorgviqmgv`qotg`argiacar`qorgviqmg
iarcpqti`lgpkoaqgliƒczte ˆg…`w ˆ‰ ˆ‰
 Manju and Sivapullaiah 659

nopq …ct`qiza ¢iqmzcqg hkoaqgliƒczt `lcsiacsN pilix`qo g mwrt`qo g ‡“|hŸg

`rriqijop “`˜e–Š|l˜e–hiŠeŽšŽ g ‡Ÿše g nmip g ip g ¥ozliqog
siaot`l g ‡i{eˆ g |azqmot g azqixo`lo g kziaq g qm`qg
d‚g‡•

™g…`w ˆ‰ ˆ‰
v`pgzpotjorgv`pg qmogtorcxqizagiagqmogiaqoapiqwg
˜gszaqm ‘ ˆ‰ zygqmogko`€pgqm`qgx`agog`ppi{aorgqzg€`zliaiqoeg
–gszaqm  ‘
h”g‡•

ˆg…`w ˆ’ n| ”¡ g— g ¢`qot g xzaqoaq g j`lco g `q g riyyotoaqg

™g…`w ˆ’
“zg
˜gszaqm ˆŽ hmtia`€{o
–gszaqm ˆ˜
hkoxiyixg{t`jiqwg£p e–Š e‰
tooghvollg‚aro}~g
xx¤{e ˆe ˆe™
n`loggpmzvpgqmogszipqctogxzaqoaqg`qgriyyotoaqg
ŒŸ• g `ar g pkoxiyix g pcty`xo g `to` g zyg
caxzaq`sia`qor g `ar g xzaq`sia`qor g pzilg
pkoxisoape g nmo g tol`qizapmik g oqvooa g pzilg
szipqctogxzaqoaq g‡vg•g`argtol`qijogmcsiriqwg
‡ŒŸ g • g pmzvp g `a g iaxto`po g ia g v`qot g mzlria{g
x`k`xiqw g zy g xzaq`sia`qor g pzile g g ¢milo g qmipg
iaxto`pogipgao{li{ilog`qg‘•gtol`qijogmcsiriqw~g
`q g mi{mot g tol`qijo g mcsiriqw~ g qmo g iaxto`po g ipg
sztoe g nmo g tol`qijolw g l`t{o g `szcaq g zy g v`qotg
`rpztor g toyloxqp g qmo g {to`qot g pcty`xo g `to`eg
hkoxiyixgpcty`xog`to`gx`lxcl`qorg`pgkotgdctig`arg
_ct`ti~gˆ‰–˜‹gpmzvagqm`qgqmogpkoxiyix gpcty`xog
`to`gzygxzaq`sia`qorgp`sklopgipgsztogxzsk`torg
qzgcaxzaq`sia`qorgp`sklopg‡n`logegg‚qgx`agog
zpotjorgqm`qgqmogiaxto`pogiagqmogj`lcogzygliƒcirg
lisiqgzygpzilgqto`qorgviqmgpkoaqgliƒcztgpzlcqizapg
ipgzygp`sogztrotg`pgqm`qgzygqmogv`qotg`pztkqizag
x`k`xiqwg`argmoaxogpkoxiyixgpcty`xog`to`gzygpzilg
`lpzeg
st_Ip N dKPPVITSKOL N USXhKHj N OP N WRnIRKN
`OLSIYKLISHhNQIORKLKSKTNUOKRj
„t`w g riyyt`xqiza g k`qqotap g zy g caxzaq`sia`qorg
pzil g `ar g xzaq`sia`qor g viqm g pkoaq g liƒczt g votog
xzsk`tor g viqm g pq`ar`tr g k`qqota g zy g kzppilog
qwkop g zy g xzskzcarp g qz g iroaqiyw g qmo g xm`a{opg
‡dzvrot g riyyt`xqiza g ilo~ g ˆ‰‰‘e g ‚q g m`p g ooag
azqixor g qm`q g aov g ko`€p g m`jo g `kko`tor g ia g qmog
p`sklo g qto`qor g viqm g pkoaq g liƒczt~ g vmixm g votog
azq g zpotjor g ia g „Œ… g k`qqotap g zy g caqto`qorg
pzilpegnmogaovgko`€gx`agog`ppi{aorgqzgpzricsg
tol`qijogmcsiriqwg`arghkoxiyixghcty`xog`to`gyztg
Œorgo`tqm
Œol`qijogŸcsiriqw ¢`qot hkoaqg
”iƒczt
‡• ¢`qotg†zaqoaqg‡•
‘ ‘e™’ ˆeŠ
’ e‘‰ Še’Š
ˆ‘‘ Še™ˆ –’e’–
hkoxiyixgpcty`xog`to`~g ™’e’’ ˜‘‰e‰™
hg‡s¤{
nmogiaqot`xqizag zygpkoaqgliƒcztgipgmi{mg`argqm`qg
s`wgogqmogx`cpogyztgrip`kko`t`axogzygqmogko`€pg
rcogqzg€`zliaiqogsiaot`lpgiagtorgo`tqmegg_zmazqg
oq g`le‹ gˆ‰Š™g†mzƒcoqqo~goq g`legˆ‰‰ˆ‹ g†mots`€~g
ˆ‰‰‹g¦`yz€cgoqg`leg‘‘Ž‹gm`rgtokztqorgqm`qgqmog
xzajotpiza g pq`tqp g viqm g qmo g yzts`qiza g zy g `ag
`sztkmzcpg{olgktoxctpztg`qg`gt`qogrokoaria{gzag
qmog`l€`ligxzaxoaqt`qizaeg
^mHVYIRNUSXhKHj
nmo g …n| g `ar g n£ g voto g zq`iaor g w g cpia{ g `g
¦ˆ–‘‘gpisclq`aozcpg…h†¤n£|g‡h…ngnmots`lg
`a`lwuot~gwgtcaaia{gqmogp`sklopgiag“gviqmg`g
xzapq`aq g mo`qia{ g t`qo g zy g ˆ‘ g ro{esiaˆ g qz g `g
s`}iscs g qoskot`qcto g zy g ˆ‘‘‘‘†e g nmo g …n|g
xctjo g zy g tor g o`tqm g ip g xm`t`xqotiuor g w g qmtoog
oarzqmotsix g ko`€p g ‡i{ee g itpq g oarzqmotsixg
ko`€ g zxxcttor g `q g n g | g –’‘ g† g viqm g `g
xzttopkzaria{ g s`pp g lzpp g zy g zalw g ˆ•e g nmipg
pmzvpgqm`qgqmogmw{tzpxzkixgv`qotgipgjotwgloppeg
hoxzar~ g pqtza{ g `ar g pm`tk g oarzqmotsix g ko`€g
`kko`torg`qgn g | g Ž™‘g †gviqmg`gxzttopkzaria{g
s`pp g lzpp g zy g –•e g g ¡arzqmotsix g ko`€ g `q g qmipg
qoskot`qcto gsi{mq go grcogqzgktopoaxo gzygytoog
z}iro g pcxm g `p g {zoqmiqoe g |ar g qmo g qmitrg
oarzqmotsix g ko`€ g `q g n | g ’™’‘† g ip g rco g qzg
romwrtz}wl`qiza g zy g qmo g €`zliaiqo g siaot`le g nmog
660 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

o}zqmotsix g ko`€ g `q g n | g ‰ˆ’‘† g ip g rco g qzg ro{t`rorgxzskloqolwg`pgiarix`qorgwgqmog„Œ…g

toxtwpq`liu`qizagzygxl`wgsiaot`lpeg `a`lwpipegšaogps`llgoarzqmotsixgko`€g`qgn g | g
–‰˜‘†gm`pg`lpzgooagazqixoregnmipgoarzqmotsixg
ko`€ g `q g mi{mot g qoskot`qcto g x`a g o g rco g qzg
g romwrtz}wl`qizagzygsiaot`lpegnmots`lg`a`lwpopg
“|hŸg

“|hŸg

m`jo gyctqmot gxzayitsor gqm`q gaovgyzts`qizagzyg

siaot`lgipguozliqopgzalweg
“|hŸg

Œ¡…g¡|ŒnŸg¨ghd¡“ng”‚¦©šŒg
“|hŸˆg

¥ozliqix g siaot`lp g `to g €azva g qz g `rpzt g v`qotg

    

`ar g yzt g m`jia{ g lopp g pkoxiyix g {t`jiqwe g nmog

¦g
roxto`pogiagpkoxiyixg{t`jiqw~giaxto`popgiagv`qotg
Œ¡…g¡|ŒnŸgg `rpztkqizagx`k`xiqw~ g`argliƒcirglisiqgj`lcopgzyg

   


    torgo`tqmgp`sklopgqto`qorgviqmgiarcpqti`lgpkoaqg
§g
¦g
§g liƒczt g ip g rco g qz g qmo g yzts`qiza g zy g uozliqop g ipg
xzttzzt`qoe g¥ozliqopg gv`qot gszloxclopgrzgazqg
{oq gtoszjorgiagtzzsgqoskot`qctoe gnmipgsi{mqg
iKMuNvN„ggt`wgriyyt`xqizagk`qqotagzygŒorgo`tqmg o g qmo g to`pza g yzt g zpotjia{ g az g pmtia€`{o g iag
pzilgqto`qorgviqmgiarcpqti`lgpkoaqgliƒczt qto`qorgpkoxisoaeg
nmogqmotsz{t`kmgzygtorgo`tqmgpzilgqto`qorgviqmg
iarcpqti`lgpkoaqgliƒcztgm`pgpmzvag`gpi{aiyix`aqg v{~

oarzqmotsixgko`€g`qgng | gˆ‘‘†gviqmgs`ppglzppg
^}
v{{ d^W

zy g `zcq g ˆ‘• g ‡i{e˜e g “z g oarzqmotsix g ko`€g z~

¢oi{mqglzppg‡•

`tzcar gqmip gqoskot`qcto gm`p gooa gzpotjor giag z{

w~
qmogcaxzaq`sia`qorgpzilegnmipgoarzqmotsixgko`€g
w{
iagp`sklogqto`qorgviqmgpkoaqgliƒcztgipgrcogqzglzppg
~
zygv`qotgszloxclopgytzsguozliqogx`jiqiope
{

y~
v{|
y{
v{{ ^} { v{{ |{{ €{{ x{{ ~{{ y{{ {{ w{{ z{{ v{{{
d^WN noskot`qctog‡‘†
zw

iKMuN€ N…n|g`argn£gklzqgzygŒorgo`tqmgqto`qorg
¢oi{mqglzppg‡•

zy

zx
viqmgpkoaqgliƒcztgpzlcqiza
z|

z{
ww
wy
wx
{ |{{
iKMuN|N…n|g`argn£gklzqgzygŒorgo`tqmgpzil
nmo g uozliqo g siaot`lp g m`jo g ` g xm`t`xqotipqixg
oarzqmotsixgko`€giagqmogqoskot`qcto gt`a{o gzyg
’‘˜‘‘‘† g rco g qz g ropztkqiza g zy g uozliqo g v`qoteg
“z g oarzqmotsix g ko`€ g xzttopkzaria{ g qzg
€`zliaiqo g siaot`lp g m`jo g ooa g zpotjore g tzsg
qmipgiqgx`agogxzaxlcrorgqm`qgtorgo`tqmgpzilgm`pg
ORXYHN`mILMHN]HmIJKOXVNOPN`OLSIYKLISHhN
QIORKLSKTNUOKRj
i{cto g Ž g pmzvp g qmo g klzq g zy g jotqix`l g pqt`ia gH
`{`iapq g jotqix`l g xzapzlir`qiza g pqtopp g yzt g qmog
x{{ y{{
‘
w{{ v{{{ xzsk`xqor g tor g o`tqm g pzil g viqm g v`qot g `arg
noskot`qctog‡ †
iacar`qorgviqmgv`qot~g`argpkoaqgliƒcztegnmogtorg
o`tqmgpzilgpkoxisoagiacar`qorgviqmgpkoaqgliƒcztg
o}miiqor g azts`l g xzsktoppiza g rctia{ g lz`ria{g
xwxlo g cq g o}miiqor g `azts`l g pvoll g ‡tozcarg
rctia{ gcalz`ria{e gnmo g xzsktoppiza gpqt`ia gm`pg
ooa gxzskloqolwgtoxzjotor gckzagcalz`ria{gzyg
tor g o`tqm g pzile g ‚q g m`p g ojoa g xtzppor g iqp g iaiqi`lg
moi{mqeghvollgrctia{gcalz`ria{gxwxlogpmzvpgqm`qg
xm`a{o g rco g qz g xzaq`sia`qiza g zxxctp g `yqotg
 Manju and Sivapullaiah 661

xotq`iagkotizrgzygiaqot`xqizaeg‚qgipgto`pza`logqzg ˆ‰Š‰g tokztqor g qm`q g a`qct`llwg zxxcttia{ g ¥ozliqog

xzaxlcro g `q g qmip g pq`{o g qm`q g qmo g zjot`llg o}k`arpgvmoagmwrt`qorg`argiygtzx€gipgxzskzporg
xzsktoppiiliqw g ‡opkoxi`llw g pvollia{ g rctia{g zygl`t{olwg¥ozliqop~ gqmogoaqito gtzx€gs`wgpvollg
calz`ria{ g zy g pzil¤€`zliaiqo g ip g ƒciqo g riyyotoaqg pi{aiyix`aqlwg`pgqmogtzx€goxzsopgp`qct`qoreg
ytzsgqmogcaqto`qorgtorgo`tqmegg‚qgv`pgzpotjorgiag g

¦g
qmo g iaro} g ktzkotqiop g pqcrw g qm`q g qmo g ¥ozliqixg Œorgo`tqmg¨gv`qotg¨ghkoaqgliƒcztg

“|hŸg
siaot`l g ip g yztsor~ g vmixm g m`p g mi{m g v`qotg

“|hŸg
“|hŸg

“|hŸg
mzlria{gx`k`xiqwg`argmi{mgpkoxiyixgpcty`xog`to`e

¦g

¦g
¦g

¦g
€
_HhNHIVSmN‚NrISHVN‚NrISHVN
| _HhNHIVSmN‚NrISHVN‚NUoHLSNgKƒXOV Œorgo`tqmg¨g¢`qotg¨g¢`qotg
¦g

¦g

¦g
§g

¦g
v §g
fotqix`lghqt`iag‡•

tv iKMuN~ N„ggt`wgriyyt`xqizagk`qqotagzygtorgo`tqmg
toszclrorgviqmgv`qotg`argiacar`qorgviqmgpkoaqg
t|
liƒcztgrctia{gxzapzlir`qizagqopqpeg¦gªgƒc`tqu~g§g
t€
ªg€`zliaiqo~g“|hŸgªgpzricsg`lcsiacsgpilix`qog
tx mwrt`qog
v v{ v{{ v{{{
†zapzlir`qizagdtoppcto‡€d`
_HYHhKIRNfHIjXVHjNPOVNWRnIRKN`OLSIYKLISHhN
iKMu N x Nhqtopp g hqt`ia g om`jizct g zy g tor g o`tqmg QIORKLSKTNUOKRj
toszclrorgviqmgv`qotg`argiacar`qor |goqqotgcarotpq`aria{gzygpzilg`l€`ligto`xqizapg
m`p g molkor g ia g `kklix`qiza g zy g `a g `lqota`qijog
soqmzrgzygqto`qsoaqegŸoaxo~gqzgxzaqtzlgmo`jogiag
‚a gqmipgx`pog`lpz~gqmipgxzclrgogqmogto`pzagyztg pzilp g ia g qmo g ktopoaxo g zy g `l€`li g pzlcqiza~ g lisog
pvollia{giagpzilgiacar`qorgviqmgiarcpqti`lgpkoaqg qto`qsoaq g m`p g ooa g xmzpoa g `p g `a g oyyoxqijog
liƒcztgpzlcqizapegnzgtoxzayitsgqmopogxm`a{opgiag so`pctoe g | g poq g zy g o}kotisoaqp g m`jo g ooag
qmogpzilgsiaot`lz{wg`yqotgxzapzlir`qorgqopqpgviqmg xzarcxqorgvmotoiagqmogpzilgipgqto`qorgviqmgˆ•g
pkoaq g liƒczt g pzlcqizap~ g qmo g „ g t`w g riyyt`xqizag `arg˜g•gx`lxicsgmwrtz}irogwgvoi{mqgzygpzilg
‡„Œ…gk`qqotapgvotogzq`iaorgzagqmopogp`sklopg `ar g iacar`qor g viqm g pkoaq g liƒczt g pzlcqizapg
`yqotgqopqia{eg rctia{gxzapzlir`qizagqopqeg

     

   
    
  i{cto g – g pmzvp g qmo g jotqix`l g pqt`ia g jotpcpg



  xzapzlir`qiza g ktoppcto g tol`qizapmik g yzt g pzilg
    

 

  
  si}orgviqmgˆg`arg˜g•gzyglisogwgvoi{mqgzygpzilg



 `ar g toszclror g qz g dtzxqztœp g s`}iscs g rtwg
„ g  g t`w g riyyt`xqiza g k`qqota g zy g tor g o`tqmg
toszclror g viqm g v`qot g `ar g iacar`qor g viqmg
iarcpqti`lgpkoaqgliƒcztg`yqotgxzapzlir`qizagqopqpg
ip g pmzva g ia g i{e’ g ‚a g p`sklo g iacar`qor g viqmg
pkoaqgliƒcztgpzlcqiza~gaovgko`€pgxzttopkzaria{g
qzg ¥ozliqo g siaot`l gm`jo g ooa g zpotjor g iag torg
o`tqmeg|qgqmogp`sogqisogiqgx`agogazqixorgqm`qg
vmilo g qmo g ko`€p g xzttopkzaria{ g qz g €`zliaiqog
siaot`lp g m`jog rip`kko`tor gxzskloqolwg iagx`pog
zygtorgo`tqmgiacar`qorgviqmgpkoaqgliƒcztegnmipgipg
rco g qz g k`tqixik`qiza g zy g €`zliaiqo g siaot`l g iag
xmosix`l g to`xqizap g viqm g `l€`lie g §t`au g oq g `le~g
roapiqw g `q g zkqiscs g v`qot g xzaqoaq g vmoag
iacar`qor g viqm g pkoaq g liƒczte g nmo g pzil g qto`qorg
viqmgˆg•glisogpmzvpgmi{motgxzsktoppizagviqmg
iaxto`po g ia g oyyoxqijo g pqtopp g `ar g `azts`lg
tozcaregŸzvojot~gqmogpzilgqto`qorgviqmg˜•gzyg
liso g pmzvp g loppot g xzsktoppiza g `ar g azg
`azts`l g pvollia{ g rctia{ g calz`ria{e g nmipg
iarix`qopgqm`qgˆg•gzyglisogx`aazqgxzaqtzlgmi{motg
jzlcso g xm`a{op g x`cpor g rco g qz g siaot`lz{ix`lg
xm`a{opgiag`l€`ligxzaq`sia`qorgpzileg¢iqmgˆ•gzyg
liso~ g xzapirot`lo g k`tq g zy g ˆ• g zy g liso g si{mqg
662 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

m`jo g ooa g cqiliuor g yzt g x`qiza g o}xm`a{og «†d… g ‡ˆ‰‰‘e g dzvrot g …iyyt`xqiza g ilog
to`xqizapeg |lkm`oqix`l g ‚aro}op~ g iazt{`aix g km`popeg
v€
!    %   #  "##   " ~g
_HhNHIVSmNN‚NNrISHVNN‚NNrISHVN
vv _HhNHIVSmNN‚NNvN„NRKYHNN‚NUoHLSNNRKƒXOV ©h|
_HhNHIVSmNN‚NN€N„NRKYHNN‚NUoHLSNNRKƒXOV
z _HhNHIVSmNN‚NrISHVNN‚NUoHLSNRKƒXOVNNN
§t`au~ g Œe”e~ g  ipm~ g …e”e~ g `ar g  l`xix~ g «e…eg
fotqix`lghqt`ia‡•


~ ‡ˆ‰Š‰e g Ÿwrt`qiza g `ar g …omwrt`qiza g zyg
€ ¥ozliqix g qcyy g ytzs g ¬cxx` g _zcaq`ia~g
v “oj`r`eg* +- eg
tv
t€ _zmazq~g he g _e~g  `o~ g «eg Ÿeg `ar g zlow~ g ¢e g ”eg
t~ ‡ˆ‰Š™eghqcrwgzyg_iaot`l|l€`ligŒo`xqizapeg
t
v v{ v{{ v{{{
$  #   /#0!'/~g
†zapzlir`qizagdtoppctog‡€d`
ˆ‰e
iKMuNyN¡yyoxqgzyg`rriqizagzyglisogzagqmogpqtoppg› dcti~g geŒeg`arg_ct`ti~g§eg‡ˆ‰–˜eghqcriopgiag
pqt`ia g om`jizct g zy g pzil g iacar`qor g viqmg hcty`xo g `to` g so`pctosoaq g zy g pzil~g
iarcpqti`lgpkoaqgliƒczt xzsk`tipza g zy g riyyotoaq g soqmzre g$ 
$~g‰–~g˜˜ˆg˜˜–egg
`cb`geUacbU ¦`yz€c~ g “i€zll` g de~ g |iapvztqm~ g †`ljia g †e~g
ša g xzaq`sia`qizag viqmg iarcpqti`l g pkoaq g liƒcztg huoxpzrw~g«`sopg¡e~g`arg¦`yz€c~gšroq`gheg
pzlcqiza~gqmogkmwpix`l~gxmosix`lg`argoa{iaootia{g ‡‘‘Že g nt`apkztqxzaqtzllor g €iaoqixp g zyg
om`jiztpgzyga`qct`lgtorgo`tqmgpzilg{oqg`lqotoreg rippzlcqiza g `ar g ktoxikiq`qiza g ia g qmog
†m`a{opg`togrcogqzgxm`a{ogiagsiaot`lz{wegnmog porisoaqp g carot g `l€`liao g `ar g p`liaog
siaot`lz{ix`l g xm`a{op g `to g `qqticqor g qz g qmog xzariqizape g    % 
yzts`qizagzyg¥ozliqogsiaot`lpgzagiaqot`xqizagzyg 0 ~g–Š~g‰Šˆ‰‰’e
pzilgviqmg`l€`ligpzlcqizapg`pgxzayitsorgytzsgqmog
„t`wgriyyt`xqizagpqcriop~g`argnmots`lgpqcriopeg Œ`z~ g he g _e~ g `ar g Œ`z~ g §e g h g ‡ˆ‰‰Že g £tzcarg
dtzkotqiopgzyg¥ozliqog`togqmogto`pzagyztgiaxto`pog mo`jogytzsgx`cpqixgpzr`gpzlcqizaglo`€`{opgg
ia g liƒcir g lisiq~ g pmtia€`{o g lisiq g `ar g jzlcsog ` g x`po g pqcrwe g g$   " 
" ~ g ˜Ž~g
xm`a{o g om`jizct g zy g `l€`li g qto`qor g pzilg ˆ˜ˆŠe
xzsk`torgqzgqmogcaqto`qorgpzileg|rriqizagzyglisog hij`kcll`i`m~gdefeg`arg_`abcg‡‘‘’eg§`zliaiqog
x`a g xzcaqot g qmo g ao{`qijo g oyyoxq g zy g `l€`lig › g `l€`li g iaqot`xqiza g `ar g oyyoxqp g za g `pixg
xzaq`sia`qizae ktzkotqiope g   "    
N /~g˜~g–‘ˆ–ˆŽg
_ZiZ_Zb`ZUN hij`kcll`i`m~gdefeg`arg_`abcg‡‘‘™eg‚arcxorg
†mots`€~ g «e|eg ‡ˆ‰‰e g”zv g noskot`qctog hvollia{ g zy g §`zliaiqix g hzil g ia g |l€`lig
¡}kotisoaq`lg‚ajopqi{`qizagzygnmog¡yyoxqgzyg hzlcqizaeg$ "
" ~gŽ™~g’‰––e
Ÿi{mgkŸg“`šŸgpzlcqizapgzagnmogškl`aicpg
hm`lo~ g hviquotl`are g%   "  %   htirm`t`a~g|e~g“`{`t`{~gneghe~g`arghij`kcll`i`m~g
' ~g–~g–’‘–’Šeg degfeg‡ˆ‰ŠˆegŸo`jia{gzygpzilgrcogqzg`xirg
xzaq`sia`qizae g  2  !   
†mzƒcoqqo~g_e~g otco~g_`tx|artog`arg”zx`q~g %#  $  '  " 
«`xƒcop g ‡ˆ‰‰ˆe g  om`jizct g zy g xzsszag
"   /~ g hqzx€mzls~g
tzx€yztsia{gsiaot`lpgiag`gpqtza{lwg`pixg  `l€os`~g~g˜Š˜˜Š–eg
“`šŸ g pzlcqizae g%"  '  ~g
‰~gˆ–˜ˆ™˜e
 Application of a New System for Evaluation of Groundwater
Contamination Hazard Rating of Abandoned MSW Landfills in India
R. K. Singh
Housing & Urban Development Corporation Ltd, New Delhi, India (rksingh3112@gmail.com)
M. Datta
PEC University of Technology, Chandigarh, India (director@pec.ac.in)
A. K. Nema
Indian Institute of Technology, New Delhi, India (aknema@gmail.com)

Abstract
In developing countries, several abandoned MSW landfills (uncontrolled waste dumps) exist adjacent to
large cities, contaminating the groundwater. The polluting abandoned landfills need to be prioritized
according to the degree of the groundwater contamination hazard posed by them, to undertake necessary
control and remedial measures at different sites in a phased manner. This paper presents the application of
a new and simple groundwater contamination hazard rating system at six old municipal solid waste
landfills in four cities of India namely Delhi, Kolkata, Chennai and Nagpur. The study shows that the Delhi
and Kolkata landfills are in urgent need of control and remedial measures.

1. INTRODUCTION systems for evaluating landfill hazards have been

In developing countries, several abandoned
unlined landfills exist adjacent to large cities,
releasing contaminated leachate and thereby
polluting the underlying aquifer. A number of
incidences have been reported in the past, where
leachate had contaminated the groundwater
(Srivastava and Ramanathan, 2008; Mor et al.,
2006; Haq, 2003). Growing concerns about public
health and degradation of groundwater quality
from existing abandoned landfills call for
undertaking appropriate control and remedial
measures at such landfills. However, due to
financial constraints, control measures cannot be
applied to all polluting landfills simultaneously.
Therefore, a sytem that can evaluate groundwater
contamination hazard of landfill sites and can
prioritize them according to the degree of such
hazard is needed.
Numerous models and approaches ranging from
deterministic water balance analyses and stochastic
simulation models to relative hazard assessment
developed. However, owing to their simpler
methodology, relative hazard assessment systems,
also referred to as hazard rating systems, are
considered to be more suitable for priority setting.
A detailed review of the existing groundwater
contamination hazard rating systems for landfills
has been reported elsewhere (Singh et al., 2010).
In the present paper, a new system by developed
by Singh et al. (2009) for groundwater
contamination hazard rating of landfills has been
briefly discussed and then applied to six
abandoned MSW landfills in India.
2. AN OVERVIEW OF THE NEW SYSTEM
In site hazard assessment, three components
namely source, pathway, and receptor need to be
evaluated. The source refers to the candidate
landfill and is characterized by the parameters that
govern the amount and potency of waste
contaminants to be potentially released to the sub-
surface. A pathway refers to the course a
664 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

contaminant takes while migrating from source to hazardous waste (HW), biodegradable waste, and
receptor, and is described by various construction and demolition waste (C&DW)
characteristics that govern contaminant transport. fractions. For a mixed waste, the waste
Receptors are sub-soil, groundwater, and the composition indicator is given by the following
groundwater users in the vicinity of the landfill. equation:
Groundwater users include human beings, ( 25 H  5 B  C ) (4)
Wc i
livestock, crops, and sensitive environment. 5
where, H is HW fraction (%), B is biodegradable
In the new system, the groundwater contamination fraction (%), and C is C&DW fraction (%), and
hazard rating ( H R,GW ) of a landfill site is given by H  B  C d 100 .
the following relationship:
H R , GW v H S u H P u H R (1) 2.1.3 Determination of infiltrating precipitation
indicator
where, H S is source hazard rating, H P is pathway The infiltrating precipitation indicator is a relative
hazard rating, and H R is receptor hazard rating. measure of the portion of annual precipitation that
infiltrates into landfill through its cover. It is given
2.1 Evaluation of Source Hazard Rating by the following relationship:
The source hazard rating is a relative combined Ip i Ps u i s (5)
measure of the amount and toxicity of the where, Ipi is infiltrating precipitation indicator, Ps
contaminants contained in a landfill and the
potential for their leaching out from the fill, and is is precipitation score, and i s is infiltration score.
given by the following relationship:
H S Wq i u Wc i u Ip i (2) The precipitation score is given by the following
where, H S is source hazard rating, Wqi is waste relationship:
P
P (6)
quantity indicator, Wc i is waste composition s
10
indicator, and Ipi is infiltrating precipitation where P is annual precipitation in mm. The best
indicator. and worst values of annual precipitation (rainfall)
are taken as 100 mm and 1000 mm.
2.1.1 Determination of waste quantity indicator
The waste quantity indicator is a relative measure The infiltration score indicates the relative
of total waste quantity disposed at a site in measure of the fraction of rainfall that will
comparison with other sites, and is given by the infiltrate into a landfill through its cover. The
following relationship: infiltration score for different cover conditions of
an abandoned landfill is given in Table 1.
Wq (3)
Wq i
3 TABLE 1. Infiltration scores for different waste site
where, Wq is waste quantity in tons. The best and cover scenarios
worst values of waste quantity for municipal solid Type of cover Infiltration score
waste landfills are taken as 200,000 and 3,000,000 No soil cover, undulating 1
surface
tons respectively.
No soil cover, graded surface 0.9
Nominal (60 cm thick) soil 0.8
2.1.2 Determination of waste composition cover with graded surface
indicator
The waste composition indicator is a relative
2.2 Evaluation of Pathway Hazard Rating
measure of the potency (toxicity) of the waste. A The pathway for leachate contaminants is
municipal solid waste will usually consist of comprised of three distinct media namely
 Singh, Datta and Nema 665

containment zone (base liner) if any, vadose zone, 2.2.2 Vadose zone indicator
and aquifer zone. In each medium, contaminant A vadose zone can be characterized by two
loading is decreased by way of physical barrier or parameters – vadose zone thickness and vadose
attenuation mechanism. Thus, the pathway hazard zone permeability, which determine a contaminant
rating is given by the following expression: travel time and hence the extent of attenuation in
H P C i u Vi u Aq i (7) the vadose zone. The best and worst values of
where, Ci is containment indicator, Vi is vadose vadose zone thickness and vadose zone
permeability are taken as ”5 and t50 m and ”10-9
zone indicator, and Aqi is aquifer zone indicator. and t10-5 m/s, respectively. The vadose zone
indicator ( V i ) is determined using the following
2.2.1 Containment indicator
relationship:
A containment system for a municipal solid waste
landfill is usually comprised of a single composite ª log( Z v , b )  log( Z v ) º (9)
Vi 0 .7  0 .3 « »
liner (clay + geomembrane) and a leachate ¬« log( Z v , b )  log( Z v , w ) ¼»
collection and removal system (LCRS). The best 1 K v and the subscripts
and worst values of leachate containment system
where, Z
v u b and w
2 L
parameters and their relative importance weights represent best and worst conditions, respectively.
are indicated in Table 2. The vadose zone indicator value varies from 0.7
for the best vadose zone to 1 for the worst vadose
TABLE 2. Best and worst values and relative
importance weights of containment system parameters
zone.
Parameter Best Worst Importance
value value weight 2.2.3 Aquifer zone indicator
Thickness of clay t0.9 0 0.4 Aquifer zone can be characterized by four major
layer (m) parameters namely aquifer thickness, aquifer
Thickness of HDPE t1.5 0 0.35 permeability, groundwater gradient, and distance
geomembrane (mm) to nearest groundwater well. The best and worst
Leachate collection Yes No 0.25 values and the relative importance weights of
& removal system aquifer zone parameters are given in Table 3.

The containment system indicator ( Ci ) is Aquifer zone indicator ( Aqi ) is the combined
determined by using the following relationship: value of the above four parameters as determined
­° § Z ct , b  Z ct · § Z gml ,b  Z gml · ½
CI 0 .2  0 .8 ® w ct ¨ ¸  w gml ¨ ¸  w lcr 1  Z lcr °¾ from the following relationship:
¨Z ¸ ¨Z ¸
°̄ © ct , b  Z ct , w ¹ © gml ,b  Z gml , w ¹ °¿
(8) ª § Z at , b  Z at · § log( Z ap , b  log( Z ap ) ·º
«0.30 ¨¨ ¸  0.35 ¨ ¸»
where wct , wgml and wlcr are the relative Aqi
«
0.8  0.2 «
¸
© Z at , b  Z at , w ¹
¨ log( Z
©
¸
ap , b  log( Z ap , w ) ¹ » § Z dw , b  Z dw ·
» u ¨¨ ¸
¸
« § Z gg  Z gg , b ·
¨ ¸ » © Z dw, b  Z dw, w ¹
importance weights of clay layer thickness, « 0.35 ¨ Z
©  Z ¸
¹
»
¬ gg , w gg , w ¼
geomembrane thickness, and LCRS, respectively.
Z ct is clay layer thickness in m, and Z gmt is
(10)
geomembrane thickness in mm, both to be where, Z at , Z ap , Z gg , and Z dw are aquifer
restricted to their respective best values, and the
thickness, aquifer permeability, groundwater
subscripts b and w represent best and worst
gradient, and distance to nearest groundwater well,
values, respectively. Z lcr indicates presence or
respectively; and the subscripts b and w represent
absence of a functional LCRS and is equal to 1 best and worst values. The value of Aqi varies
when a functional LCRS is present, and is equal to
0 when LCRS is clogged or absent. between 0.8 and 1. However, for Z dw 0 , Aqi 1.
666 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

TABLE 3. Best and worst values, and relative where, SG i is indicator for sub-soil/groundwater
importance weights for aquifer zone parameters
Parameter Best Worst Importance
and Gu i , j is indicator for j th groundwater user
value value weight category and m is the number of groundwater
Aquifer thickness •50 ”5 0.3 user categories. In the present study, groundwater
(m) users have been divided into four categories,
Aquifer ”10-6 •10-4 0.35 hence, m 4 . The value of receptor hazard rating
permeability
(m/sec)
varies between 0.25 and 1.
Groundwater ”1 •5 0.35
gradient (%) 2.4 Overall Groundwater Contamination
Distance to nearest •5000 ”500 - Hazard Rating
groundwater well The overall groundwater contamination hazard
(m) rating of a landfill is obtained by the following
relationship:
2.3 Evaluation of Receptor Hazard Rating HS uHP uHR (12)
H R , GW u 1000
The receptors to groundwater contamination SF
hazard include sub-soil, groundwater, and the where H S , H P and H R are the source hazard
groundwater users drawing groundwater from rating, pathway hazard rating and receptor hazard
within 5000 m of the source. Groundwater can be rating, respectively; and SF is a scaling factor
used for various purposes including drinking, (equal to 1,000,000). The scaling factor is equal to
livestock watering, irrigation, commercial and the product of the source, pathway, and receptor
industrial use, and sensitive environment watering. hazard ratings of a landfill having all its
The receptor indicators for sub-soil/groundwater parameters at the worst values. The overall hazard
and different groundwater user categories are score obtained from equation (12) for MSW
given in Table 4. landfills is restricted to a maximum of 1000.
TABLE 4. Receptor indicators
3. APPLICATION OF THE NEW SYSTEM:
Receptor (Sub-soil/groundwater and Indicator
groundwater user category) CASE STUDIES
( SG i / Gu i )
The new system has been applied to six old
A) Sub-soil/Groundwater: municipal solid waste (MSW) landfills located in
i) If aquifer exists 0.5 four cities of India, namely Ghazipur and Okhla in
ii) If aquifer does not exist (sub- 0.25
Delhi, Perungudi Dumping Ground (PDG) and
soil alone is receptor)
Kodungaiyur Dumping Ground (KDG) in
B) Groundwater user category
a) Human population Chennai, Bhandewadi in Nagpur, and Dhapa in
i) Drinking water use 0.5 Kolkata. None of these landfills have covers and
ii) Recreation use 0.5 liners. The site parameters of the various landfills
iii) Industrial use 0.2 are indicated in Table 5.
b) Crops 0.35
c) Livestock 0.35 The groundwater contamination hazard ratings of
d) Sensitive environment 0.35 different landfills, alongwith source hazard,
pathway hazard, and receptor hazard ratings are
The receptor hazard rating is given by the given in Table 6. It is seen from Table 6 that
following relationship: among all the landfills, the Dhapa landfill has the
j m
(11) highest source hazard rating followed by Gazipur
H RSG  iGu ¦j 1
i, j
 Singh, Datta and Nema 667

TABLE 5. Site parameters of six landfills located in four cities of India

No. Landfill site parameters Ghazipur Okhla Bhandewadi KDG PDG Dhapa
landfill landfill landfill Chennai Chennai landfill
Delhi Delhi Nagpur Kolkata

1 Area of landfill (ha) 30 16.2 21.5 55 20 24.7

2 Average waste thickness (m) 16 13 2.0 3 3 17
3 Precipitation (mm)
Annual 721 721 1050 1200 1200 1650
Net 143 143 401 431 431 921
4 Biodegradable waste (%) 60 60 47 41 41 51
5 C&DW fraction (%) 16 16 20 20 20 17
6 HW producing industries in the Yes Yes Yes Yes Yes Yes
vicinity of the site without HW
landfill in the region
7 Leachate COD (mg/l) 27200 23306 2500* 2000 1100 22000*
8 Vadose zone thickness (m) 12 12 3 4 2 40
9 Vadose zone permeability (m/s) 1×10-6 1×10-6 1×10-8 8×10-9 3×10-9 1×10-9
10 Aquifer thickness (m) 20 20 11 15 15 >50
11 Aquifer permeability (m/s) 9×10-5 9×10-5 1×10-5 1×10-5 1×10-5 1×10-6
12 Groundwater gradient (%) 0.2 0.2 0.2 2 2 0.1
13 Distance to G.W. well 500 500 500 1000 1000 500
14 Groundwater used for
a) Human consumption# Yes Yes Yes No No Yes
(Yes/No)
b) Irrigation (Yes/No) Yes Yes Yes No No Yes
c) Livestock watering (Yes/No) Yes Yes Yes Yes Yes Yes
*Assumed values based on comparable waste dumps
**Estimated values
# At all the landfill sites, a human population of 3000 is assumed to be using the groundwater.

TABLE 6. Values of source hazard, pathway hazard, and receptor hazard ratings produced by the new system for
different landfills
Landfill Source hazard rating Pathway hazard Receptor hazard Overall hazard
(0-1000000) rating (0-1) rating (0-1) rating (0-1000)
Ghazipur, Delhi 8847308 0.831 1 704
Okhla, Delhi 561242 0.831 1 466
Bhandewadi, Nagpur 263710 0.752 1 198
KDG, Chennai 508645 0.730 0.85 320
PDG, Chennai 306725 0.724 0.85 189
Dhapa, Kolkata 1079244 0.565 1 609

Landfill and the Bhandewadi landfill the lowest.
The highest source hazard rating of Dhapa landfill
site is due to its large size and high annual rainfall.
Table 6 further shows that the Dhapa landfill
shows the lowest value of pathway hazard rating,
which is due to higher vadose zone and aquifer
thickness, and lower vadose zone permeability and
aquifer permeability in comparison with other
sites. As the groundwater at KDG and PDG sites is
used for irrigation only, these sites return lower
values of receptor hazard rating than the other
sites.
According to the site hazard scores produced by
the new system, the Ghazipur landfill is given top
priority for remediation whereas the PDG
668 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

TABLE 7. Site scores for six Indian landfills produced by the new system and the existing hazard rating systems
Landfill Site score
New System HRS-1990 ERPHRS RSS NCS NPC JENV
system system
Ghazipur, Delhi 704 262 323 360 599 729 657
Okhla, Delhi 466 262 323 360 599 675 633
Bhandewadi, Nagpur 198 71 256 360 616 779 521
KDG, Chennai 320 35 256 180 603 779 624
PDG, Chennai 189 35 256 180 589 779 568
Dhapa, Kolkata 609 152 179 144 596 704 527

site the least priority. The order of priority
generated by the new system for remedial action at
various landfills is Ghazipur, Dhapa, Okhla, KDG,
Bhandewadi and PDG. The site hazard scores
show that the Ghazipur and Dhapa landfills are in
urgent need of remedial measures.
The site hazard scores obtained by using the new
system have been compared with those of nine
existing systems. The site hazard scores
(normalized to 1000-point scale) produced by
different systems are given in Table 7. It is seen
from Table 7 that, whereas all the existing systems
individually produce similar scores for all the
landfills and return same rank for more than one
landfill, the new system returns different ranks to
different landfills. Further, the range of scores for
different landfills produced by the new system is
much wider than those depicted by the existing
systems. This clustering of scores (i.e. similarity in
scores for different sites) exhibited by the existing
systems shows that the existing systems do not
adequately respond to varied site conditions. This
demonstrates that the new system makes a superior
basis for site prioritization.
4. CONCLUSIONS
A new system for evaluation of the groundwater
contamination hazard rating of landfills has been
briefly discussed and applied to six abandoned
landfills in India. The case studies show that in
comparison with the existing systems, which
produce clustered scores for all the landfills and
return same rank to more than one landfills, the
new system produces significantly varying scores
and distinctly differing ranks to all the landfills.
The study also shows that in comparison with
other landfills, the Ghazipur (Delhi) and Dhapa
(Kolkata) landfills pose greater groundwater
contamination hazard and are in urgent need of
remedial measures.
REFERENCES
Haq, I. (2003). Environmental impact assessment
study: Leaching of chemical contaminants from
a municipal dump site Hastsal, Delhi (Capital
of India). International Journal of
Environmental Studies 60, 363-377.
Mor, S., Ravindra, K., Dahiya, R. P., Chandra, A.,
(2006). Leachate characterization and
assessment of groundwater pollution near
municipal solid waste landfill site.
Environmental Monitoring and Assessment
118, 435-456.
Singh, R. K., Datta, M., Nema, A. K. (2010).
Review of groundwater contamination hazard
rating systems for old landfills. Waste
Management & Research 28(2), 97-108.
Singh, R. K., Datta, M. and Nema, A. K. (2009). A
new system for groundwater contamination
hazard rating of landfills. Journal of
Environmental Management 91(2), 344-357.
Srivastava, S. K., Ramanathan, A. L. (2008).
Geochemical assessment of groundwater
quality in vicinity of Bhalswa landfill, Delhi,
India, using graphical and multivariate
statistical methods. Environmental Geology 53,
1509-1528.
 The Effect of Crude Oil Contamination on Geotechnical Properties of
Fine-grained Soils in the Vicinity of Tehran Oil Refinery Site
T EBADI
Amirkabir University of Technology, Tehran, Iran (tebadi@aut.ac.ir)
M KERMANI
Amirkabir University of Technology, Tehran, Iran (kermani-m@aut.ac.ir)

ABSTRACT Soil pollution is not only a matter for environmentalists, but also should be considered as a
geotechnical problem. Oil contamination causes some changes in geotechnical properties of soil. Therefore, a
comprehensive set of laboratory tests has been carried out on uncontaminated as well as contaminated soils
with different amounts of crude oil by dry weight of soil. The soil samples were taken from the lands in the
vicinity of Tehran Oil Refinery site where there is a vast area subjected to the problem.
The results indicated an increase in angle of internal friction, maximum dry density, compression index and
Atterberg limits and a decrease in optimum water content and cohesion by increase in oil content.

Al-Sanad and his colleagues (1995, 1997) carried

INTRODUCTION
Human is contaminating the environment for his
short-term benefits. Air, water and land is being
contaminated by industrial, petrochemical,
constructional and sanitary activities. Considering
land contaminations, Environmentalists are afraid
of subsurface water aquifer contaminations, plant
growth in the contaminated soil and its effects on
life cycle and the life of the fauna which live in the
area. However, geotechnical experts consider the
effects of soil contamination on structures
constructed over or by means of the soil, such as
soil bearing capacity, foundation settlement, shear
resistance, compressibility and plasticity.
Crude oil is one of the most common soil
contaminators. over 2 million tons of oil is being
produced all over the world every day and about
10 percent of it is entering the environment
because of pipeline breakage, leakage from
reservoir tanks, tanker accidents, discharge from
coastal facilities and offshore petroleum
productions [9].
out a laboratory testing program to evaluate the
influence of oil contamination and aging effects on
the geotechnical properties and behavior of
Kuwaiti sand. They artificially contaminated the
sand with different contamination amounts to
determine the effects of oil contamination on
strength parameters, compressibility, permeability,
and compaction characteristics of the Sand [2].
Evgin and Das (1992) carried out triaxial tests on
clean and oil-contaminated quartz sand. They
found that full saturation with motor oil caused a
significant reduction in the friction angle of both
loose and dense sands and a drastic increase of
volumetric strains [4]. Another laboratory testing
program was carried using samples from Persian
gulf beaches by khamehchiyan and his colleagues
(2006) [6].
This paper presents the results of a comprehe`nsive
laboratory testing program carried out to determine
effects of crude oil contamination, on fine-grained
soils in the vicinity of Tehran Oil Refinery site
where there is a vast area subjected to the problem.
670 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

These properties included Atterberg limits,
consolidation, strength parameters and compaction
characteristics.
Properties of the Soil and Crude-Oil
samples were contaminated artificially by the
crude-oil provided from Tehran Oil Refinery with
the properties including viscosity, density,
American Petroleum Institute (API) gravity, flash
point and specific gravity listed in table 2.

Soil samples were taken from 30 to 50 centimeters TABLE 2. Summary of some crude-oil properties
below the ground surface avoiding upper organic Parameter Quantity
soil layers. Basic properties of the soil is listed in Viscosity (cp) 41.3
the table 1. Sieve size analysis test including Density (g/cc) 0.891
hydrometer test (ASTM D422) was conducted on API gravity (at 60˚F) 26.8
the soil provided from the area. Fig 1 shows the Flash poin (˚C) 44.1
grain size distribution curve for the sample. As Spicific Gravity (at 25˚C) 0.89
shown in the figure and table, most particles were
silt-sized. However, according to the Unified Soil SAMPLE PREPARATION
Classification System, the soil is classified as CL
(lean clay). The soil had a specific gravity of 2.65 Because it was recommended in the program tests’
kg/m3. standards, first of all, the soil was pulverized until
all soil aggregations could pass through 4 sieve
TABLE 1. Summary of some soil properties size. The pulverized soil was mixed with sufficient
Parameter Quantity distilled water to reach the sample’s water content
and then crude oil was added with the amount
Clay, Silt, Sand & Gravel 3%, 89%, which was predicted in the testing program. Then
(MIT Categorization) 7%, 1% the samples were put in insulated plastic containers
LL, PL & PI 23.9, 45.5, 21.6 for 1 week to 3 months allowing possible reactions
Water content 3.9 % between soil and oil.
Organic materials content 3.64 %
pH (in 27˚C) 8 - 8.3 Atterberg Limits
Electrical conductivity (ms/cm) 4.16
(in 26.4˚C) Plastic limit, liquid limit and plasticity index was
tested and evaluated for uncontaminated soil as
well as contaminated soil with 4%, 8% and 12%
by weight of dry soil to study the plasticity
behavior of the fine-grained soil by different
amounts of contamination.

As the pore liquid in this test is a mixture of

FIG. 1. Size distibution curve for soil samples
(For colour figure, refer to CD)
In order to have oil content in control, the soil
distilled water and crude oil and Atterberg Limits
are water content percents which are division of
water weight and soil dye weight, after the samples
were put in the furnace and their mass before and
after evaporation of the pore liquid was calculated,
water content can be measured from the equations
below:
* (1)
Z Z  Zc
Where
 Ebadi and Kermani 671

Woil of oil. Moreover, The size of the fabric elements in

Zc (2) the contaminated clay is sufficiently larger than the
Wd soil ones in uncontaminated clay. This is due to the oil
coating on the individual clay particles and the
And clay groups. The flocks so formed seem to be silt
or even sand-sized particles [8]. Therefore, the
Wwater  Wcrude  oil change in shape of compaction curves is
*
Z (3) reasonable. In addition, as the results of Atterberg
Wd soil Limits tests represent, the soil plasticity decreases
Any soil sample was made with an specific Z c . due to oil contamination. A reduction in soil
*
plasticity may cause better compaction.
Z is calculated easily by weighing a soil sample
before and after putting it in oven.
The results of the Atterberg Limits tests is shown
in FIG. 2. Plastic limits tend to ascend distinctively
as the oil content increases, liquid limits ascends
with a very short slope and plastic index descends
consequently.

As crude-oil covers clay particles and does not let

water molecules reach double-layer water, more
water is needed for the soil to gain plastic
properties. Therefore the plastic limit increases. FIG. 2. Influence of oil content on Atterberg Limits and
However as oil surrounds the soil particles, most plasticity index
of the water added to the soil during the test, will (For colour figure, refer to CD)
join free water and because flow characteristics of
clay depend on free water, no distinctive increase
in liquid limit was observed. Finally, as added
water was less reached double-layer water and
adsorbed water, and much joined free water, the
low difference between LL and PL and
consequently low amounts of PI seem reasonable.

Compaction Tests

Standard proctor tests (ASTM-D698) was carried

out on the artificially prepared samples with 4%,
8% and 12% oil content and uncontaminated soil
samples. The water content was calculated from
FIG. 3. Compaction curves for samples with different Oil
the relations given in previous part. The result is content
plotted in FIG. 3 in the form of Dry density versus (For colour figure, refer to CD)
water content curves.
As oil contamination increases from 0% to 12%,
the maximum dry density ascends from 1.65 Direct Shear Tests
gr/cm2 to 1.73 gr/cm3 and optimum water content
decreases from 19.2% to 10.1%. This A comprehensive set of direct shear tests (ASTM
improvement in compaction characteristics can be D3080-72) was conducted with standard
attributed to lubricating effect caused by presence
672 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

100×100×20 mm shear boxes with a rate of shear

equal to 1 mm/min and normal stresses of 0.5, 1
and 1.5 kg/cm2 to examine the conjugated effect of
moisture, contamination and aging on shear
strength of the soil. The samples were prepared by
field dry density of 1.51 gr/cm3 and water contents
ranged from 3% to 15% considering changes in
ground water level and rainfalls. Because of
crucial importance of shear strength, oil contents
were chosen in a wide range, as 3%, 6%, 9%, 12%
and 15% by weight of crude-oil. The samples were
artificially prepared and put in insulated plastic
containers. in order to take aging effects into FIG. 4. The influence of oil and water content on soil’s
cohesion for samples with age of 7 days
consideration, for each test, three samples were
prepared, first sample was tested after 1 week after (For colour figure, refer to CD)
preparation, second was tested after 1 month and
third was tested after 3 months. All tests were done
in dry condition to let the samples preserve their
oil and water content. The results are presented in
two groups of charts:

Internal friction angle and cohesion, versus oil

content for different water contents and an
specific sample age

As shown in FIG. 4-6, for a specific sample age, a

reduction in cohesion occurs by increase in oil
content and an increase is seen in cohesion as
FIG. 5. The influence of oil and water content on soil’s
moisture ascends. Scanning Electron Microscope cohesion for samples with age of 30 days
(SEM) has shown that clay’s fabric converts into (For colour figure, refer to CD)
Aggregated due to oil contamination and as a
result, it’s SSA (specific surface area)
decreases[8,11]. In addition, studies has shown
that being mixed by oil, clay’s Cation Exchange
Capacity (CEC) reduces distinctively [8].

A reduction in CEC of the soil proves the

reduction of SSA. Reduction in SSA leads to
reduction in adsorption of water molecules with
clay particles and also reduction in clay particles’
contact and consequently reduction in cohesion.
As moisture increases, more water molecules will
be attached to clay particles surfaces, more bonds
between water molecules and clay minerals and FIG. 6. The influence of oil and water content on soil’s
also between different clay minerals will be cohesion for samples with age of 90 days
constituted and cohesion will increase. (For colour figure, refer to CD)
Nevertheless, as contamination increases, the
affect of moisture on cohesion decreases.

Moore and Mitchell (1974) and Sridharan and Rao
(1979) cite that When clay soils are contaminated,
a reduction in dielectric constant of pore fluid
reduces the physicochemical interaction in the clay
pore fluid electrolyte system, and hence should
result in a higher strength. Lambe (1958) cites that
any increase in pore liquid’s pH and dielectric
constant, leads to increase in shear resistance [12].
However, it must be taken into consideration that
cohesive soils behave different due to long-term
contamination that causes changes in soil fabric.
FIG. 7. The influence of oil and water content on soil’s
internal friction angle for samples with age of 7 days
(For colour figure, refer to CD)
FIG. 8. The influence of oil and water content on soil’s
internal friction angle for samples with age of 30 days
(For colour figure, refer to CD)
As shown in FIG. 7-9, for a specific sample age,
an increase in internal friction angle occurs by
increase in oil content. Results also indicate that
Ebadi and Kermani 673
internal friction angle decreases as moisture
amount increases.
Diversion of soil fabric to aggregated and better
compaction characteristics as discussed in previous
parts are the reasons for friction improvement. In
saturated fine-grained soils, pore fluid fills the
pores of the fabric and a thin layer of fluid exists
between mineral contacts. Water acts as a lubricant
and makes particles attain a closer packing [13].
FIG. 9. The influence of oil and water content on soil’s
internal friction angle for samples with age of 90 days
(For colour figure, refer to CD)
Internal friction angle and cohesion, versus
water content for different oil contents and an
specific sample age.
According to the test results, cohesion tend to
decrease as sample age increases but the reduction
rate descends as time passes. since changes in
cohesion relate to chemical reactions, as time
passes, more clay particles will be exposed to oil
and more reduction in cohesion occurs. It can be
implied from the charts that the more the water
content, the less the reduction rate of cohesion.
This might refer to catalyzing effect of water in
reactions between oil and soil minerals.
The test results indicate that ascending of sample
age, does not have specific effect on internal
friction angle. Since the changes in internal friction
angles are due to physical matters, such as better
compaction and aggregation of the particles, it
seems logical that no distinct change is observed in
internal friction angle by aging of samples.
674 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
Consolidation Tests
One dimensional consolidation tests were
conducted on clean soil and soil mixed with 4%,
8% and 12% of crude oil (ASTM D2435). The
specimens were 50 mm diameter and 20 mm thick
and were prepared at field dry density of 1.51
gr/cm3 and field water content of 3.87 and then
saturated with distilled water prior to the test. The
results are plotted in FIG. 10 in the form of e-log p
curves. As can be seen, the compressibility
increases distinctively as oil content increases.
CONCLUSIONS
The following conclusions can be asserted from
the extensive laboratory testing program discussed
above:
1. Plastic limits tend to ascend distinctively as the
oil content increases, liquid limits ascends with a
very short slope and plastic index descends
consequently.
2. As oil contamination increases from 0% to 12%,
the maximum dry density ascends from 1.65
gr/cm2 to 1.73 gr/cm3 and optimum water content
decreases from 19.2% to 10.1%.
3. for a specific sample age, a reduction in
cohesion occurs by increase in oil content and an
increase is seen in cohesion as moisture ascends.
However, as contamination increases, the affect of
moisture on cohesion decreases.
FIG. 10. e-logp curves for samples with different
contamination amounts (For colour figure, refer to CD)
4. The compressibility increases distinctively as oil
content increases. This behavior can be attributed
to lubrication effects of oil. Moreover, according
to contamination, abandoned water molecules do
not tend to come back to clay minerals’ surface.
REFERENCES
[1] Alsanad, H.A., Eid, w.k., Ismael, N.F., 1995.
Geotechnical properties of oil contaminated Kuwaiti
sand. Jl. of Geotechnical Engineering, ASCE 121 (5),
407–412.
[2] Alsanad, H.A., Ismael, N.F., 1997. Aging effect on
oil contaminated Kuwaiti sand. Jl. of Geotechnical and
Geoenvironmental Engineering 290–294.
[3] Aiban, A., 1998. The effect of temperature on the
engineering properties of oil-contaminated sand. Jl. of
Environmental International 24, 153–161.
[4] Evgin, E., Das, B.M., 1992. Mechanical behavior of
an oil-contaminated sand. Proc., Mediterranean
Conference. Balkema Publishers, Rotterdam, The
Netherlands, pp. 101–108.
[5] Shin, E.C., Das, B.M., 2001. Bearing capacity of
unsaturated oil-contaminated sand. International Jl. of
Offshore and Polar Engineering 11 (3), 220–227.
[6] Khamehchiyan, M., Charkhabi, A.H., Tajik, M.,
2007. Effects of crude oil contamination on geotechnical
properties of clayey and sandy soils. Engineering
Geology 89, 220–229
[7] Meegoda, N.J., Ratnaweera, P., 1994.
Compressibility of contaminated fine grained soils.
Geotechnical Testing Jl. 17 (1), 101–112.
[8] Ur-Rehman, H., Abduljauwad, S. N., Akram, T.,
2007. Geotechnical behavior of oil-contaminated fine-
grained soils. Electronic Jl. of Geotechnical Engineering,
v 12 A, 15-23
[9] Ewies, J.B., Ergas, S.J., Chang, D.P.V., Schroeder,
E.D., Bioremediation principles, McGaw-Hill, 1998.
[10] Das, B.M., 1994. Principle of Geotechnical
Engineering, 3rd edition. PWS Publishing Company.
436 pp.
[11]. Van Olphen, H., An introduction to clay colloid
chemistry, John Wiley and Sons, New York, 1977. 66
pp.
[12]. Lambe, T.W., “The engineering behavior of
compacted clay”, ASCE Jl. of Soil Mech. and
Foundation Division, Vol. 84(2), 1958, pp. 1-35.
[13] Proctor, R.R., 1933. Fundamental principles of soil.
Jl. of the Aeronautical Society 111 (9), 245–25.
 Contaminated Zone Around Fan River and Their Rehabilitation
L.Bozo
Polytechnic University of Tirana, Albania (lujeta.bozo@yahoo.com)
Gj.Ikonomi
Polytechnic University of Tirana, Albania (Giergi7@hotmail.com)

ABSTRACT In the north part of Albania there are big mine industry for the extraction and
treatment of minerals as iron, copper etc. These mines are situated around river Fan. The zones
around this river are contaminated in different quantities by dangerous chemical elements and
heavy metals. In this paper we would like to present our study about the degree of the
contamination of the zones around Fan river and our measures for their rehabilitation.

INTRODUCTION In t he north part of the country flows in an

accented mountain terrain the Fan river which
Albania is traversed by a dense river network
is the main branch of the Mat river. This zone
(Figure1).
is very rich with minerals. During 1960-1990 this
zone had a very big development in extraction
and elaboration in mineral industry, mainly copper,
zinc, iron, chrome etc. Also for the refute
deposits from mineral enrichment the tailing
dams were constructed with height 30-50m.

Mineral extraction, their enrichment and the

siderurgic industry have grown the pollution
of the zone around Fan river to make a serious
danger for the fauna and flora that is developed
there. We have undertaken many studies for the
determination of the pollution degree and the
danger of the environment of this zone; in the
following paper we are going to present one of
these studies.

THE NATURAL POLLUTION OF FAN RIVER

The Fan river in the main branch of Mat river

which is one of the most important river
Fig.1 The hydrological map of Albania. in Albania (Figure2).
676 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
Fig. 2 River Fan and its branches.
(For colour figure, refer to CD)
The water collect surface of Fan river is 1076km2
and after it is joined with Mat river the water
collect surface grows up to 2441km2. Fan river
is formed by the connection of two branches:
big Fan, 76.9km long which stems from Qafa e
Malit (1397m height) and small Fan, 54.5km long
which stems from Runes (1856m height). After
17.5km of their connection it flow’s in to Mat
river. The medium precipitation for year is
1750mm in the big Fan and 1660mm in small
Fan. The feeds between October-June are
from 14.4m3/s to 59.7m3/s. the erosion activity
of Fan river is big and it transports 208000 T/
year solid material. Also the total quantity of the
solid feeds in suspended and underwater
condition is 119000 T/year. From the chemical
point of view the waters are carbonates and their
mineralization (mg/l) is shown in table 1.
TABLE 1 Jones content in river water.
Jones Ca Mg Na K HCO 3 Cl
Big Fan 25,73 11,1 2,31 0,5 86,08 8,55
Sm all Fan 39,19 9,68 4,19 0,49 126,8 7,66
THE FACTORS THAT INFLUENCED IN THE
CONTAMINATION OF THE ZONE AROUND
FAN
As we said before the zone where the Fan river
flows is rich with minerals, but the Cu-mineral
creates the biggest toxic risk for the water, land
and air when it enters in contact with them. The
factors which have contaminated this zone are
the following:
The mines
On the surface and underground in mine’s zone
there is flowing water which permits to extract
from them the minerals. Water from mine’s
can be extracted directly or by putting pumps
in the pits. This water contents heavy metals
and consequently presents high pollution degree
for flora and fauna.
The liberation of the lixiviates.
After the abandonment of the mines (20 years
ago) began the creation of the percolated
acids or lixiviates from the solid remains outside
of the mines. The lixiviates have a big percentage
of sulfuric acid and heavy metals as Ag, Cd, Co,
Cu, Mg, Hg, Ni, Pb, Zn, As, Se.
The mineral remains have extension along the
entire Fan river contaminating all the zone. Except
the “Fe” all the other metals are toxic.
The tailing dams.
The acid is created rapidly in the tailing dams.
This depends from sulfite content in minerals,
SO 4 Σj
from climatic conditions and necessary oxygen
for the oxidation process. All this phenomena’s
10,57 144,8
made possible the contamination of the surface
and underground waters. Also the oxidation
29,48 217,5
of sulfites can lead to the liberation of the heavy
 Bozo and Ikonomi 677

metals as Ag, Cd, Co, Ni, Pb, Zn, As etc. The University. For the monitoring and determination
zones with great acid generates is positioned of the contaminated degree we built about 30
15km far from tailing dams. This acids cause stations to take the samples from water collect
not only destruction of the flora but even the basin of the Mat river and their branches. The
ravage of the fishes and other micro result of the analysis taken from the
organisms in the river. monitoring stations during period 1998-1999
(10 after mines were closed) we can see in the
The enrichment in Fushe-Arrez, Reps, tables 2,3,4,5.
Rreshen and the foundry in Rubik.
TABLE 3 Continue
All this industries are other factors for the
air and water pollution which pollutes them Station
(when MO Ni Pb V Zn
by dust which comes out from the grinding
take the
process of the minerals. The elaboration samples) Sept Sept Sept Sept Sept
capacity of the minerals reaches 300.000 T/
year and the content of “Cu” goes from 12- 9 5 2 5 2 2
10 5 5 5 2 7
24%. 11 5 6 5 2 23
12 5 15 5 4 210
DETERMINATION OF THE 14 5 2 5 3 2
CONTAMINATION DEGREE OF THE ZONE 15 5 3 5 4 34
24 5 70 5 66 25
AROUND FAN RIVER

For the determination of the pollution degree we

took many water samples from the rivers (Big Station Ag As Cd Co Cr Cu
Fan, Small Fan etc). The samples were (when April Jan April April April Jan
take the
exanimate in chemical laboratory of Tirana’s samples)

TABLE 2 The metal content mg/l for Big Fan 9 0.46 10 0.04 0.5 1.4 7
(1999) 10 1 10 0.1 10 10 20
11 1 10 0.1 10 10 310
Station 12 1 10 0.1 10 10 20
(when As Cd Co Cr Cu 15 1 - 0.1 10 10 -
take the 24 0.1 10 0.12 2.5 4.8 60
samples) 24M 0.05 - 0.06 1.1 3.5 -
Sept Sept Sept Sept Sept

9 10 1 1 2 2
10 10 1 1 2 34
11 10 1 3 2 37
12 10 1 51 2 150
14 10 1 1 2 10
15 10 1 1 2 110
24 10 1 13 50 43
678 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

TABLE 4 The metal content mg/l for Small Fan Station PH Cu mg/l Zn mg/l
(april 1998 –January 1999). (when April Sep April April April Sep
take the ‘98 ‘99 ‘98 ‘99 ‘98 ‘99
Station Ca Fe K Mg Na S samples)
(when April April April April Jan April
take the 16(S) 4.5 3.4 680 2200 610 4200
samples) 17(FV) 8.5 8.5 14 5 5.6 2
18(FV) 7.5 7.6 30 120 40 210
9 7 0.14 0.49 7.5 0.98 2.8 19(FV) 7.3 5.3 120 87 120 160
10 6.5 0.31 0.37 8 1.3 3.3 20(FV) 7.8 7.6 52 20 10 6
11 8.3 0.28 0.39 11 3.7 5.9 21(FV) 7.7 7.9 60 140 60 200
12 29 1.2 0.31 7.9 1.4 30 23(FV) 7.9 7.9 59 150 43 120
15 10 1.2 0.30 7.1 6.85 6.3 25(F) 8.0 8.1 48 18 26 3
24 13 1.2 0.48 4 1.7 5.2 28(F) 7.6 7.9 16 32 11 31
24M 14 0.52 0.43 12 4.9 29(F) 7.8 7.7 24 140 23 170
32(F) 7.9 7.9 30 52 20 50

Station PH Cu mg/l Zn mg/l From the graphics compiled by us, which show the
(when April Sep April April April Sep changes of the concentration of some heavy metals
take the ‘98 ‘99 ‘98 ‘99 ‘98 ‘99 (Cu, Zn ) in the water from the source of Big and
samples) Small Fan until they join together (Figure 3,4), wecan
1(LZ) 7.8 6.2 85 11 86 2 see:
2(LZ) 7.6 7.9 40 16 20 34
4(LZ) 6.4 5.8 730 4200 320 480
5(LZ) 7.6 5.5 47 7 16 2

6(LZ) 7.4 6.2 18 12 7.2 2

8(LZ) 7.0 6.2 130 14 45 6

13(LZ) 7.3 7.2 220 120 110 220

9(FM) 7.6 7.7 16 4 7.7 2

10(FM) 7.6 7.6 30 5 10 7

11(FM) 7.1 7.5 20 9 10 23

14(FM) - 8.4 - 5 - 2

15(FM) 7.5 6.9 90 50 50 34

24(FM) 7.9 7.9 46 14 19 25

Fig.3 Variation of Cu concentration in different zone
Table 5 PH and concentration of Cu and Zn in the
samples took from Stefa (S) Small Fan (FV) And
Fan (F) rivers.

Fig.4 Variation of Zn concentration in different zones.
The most contaminated zone is around Sefta river.
Nevertheless all zones around Fan river and its
branches have a high contamination level by heavy
metals Cu, Zn, Cr, Cd, As.
Usually with pronounced acid characteristics is the
water of Small Fan. Also with high Fe content are
Small Fan, Lumi Zi, Big Fan and Fan.
Highly contaminated by P is Small Fan.
High content of Pb have Small Fan and Fan.
THE MEASSURES FOR THE REHABILITATION
OF THE SITUATION
We think that it is necessary to create a permanent
monitoring system which must observe how the scale
of contamination changes year after year in Fan river
and its branches. It is necessary to neutralize the
remains of dangerous minerals especially in the
zones around Lumi Zi and Sefta river. Another
measure is the maintenance of the 7 existing tailing
dams with intension to impede the slope erosion.
Bozo and Ikonomi 679
Also the tailing dams must be covered by
geomembranes or vegetable layer with goal to
protect the air from dangerous elements. Finally if in
the future the
mining activity restarts they must use only the adapt
technology which can protect the environment from
pollution. Also in the zones were the soils are not
alkaline there must be an intervention with intention
to create the necessary capacity to neutralize the
acids and to precipitate the deleterious metals in
them.
CONCLUSIONS
The Fan river with its branches has a high degree of
contamination in acid water and high content of
dangerous elements as Cr, Cd, As, Zn, Cu etc.
Rehabilitation of the zone around Fan river and its
branches needs to undertake urgent measures as:
reduction of acidity, elimination of the heavy metals
in the underground and surface waters, the
maintenance of the tailing dams etc.
The damage of the flora and fauna in this zone is
with great consequences for the life and health of
the people for that reason the engineering measures
and chemical interventions, must minimize the toxic
matter in the river waters.
REFERENCES
Bozo L. (2009) Geotechnic for the environmental
engineering. Press Natyra, Tirane, Albania.
Nobre R.C.M and Nobre M.M.M and Bonaparte L.B
(2006) The evaluation of soil and ground water
remediation at an industrial site using risk
analysis. Proceedings of 5th ICEG, Cardiff.
Pastarus J.R (2006) Ground surface subsidence risk
assessment by underground mining in Estonia.
680 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

Proceedings of 5th ICEG, Cardiff. Science Academy of Albania (2008) Albanian

Hydrology.
Stratoberdha P. (2007) Hydraulic modeling theory of
the rivers. Science Academy of XAlbania.
 Ground Improvement of Closed Landfill Sites Using
Stabilisation
Hadi Khabbaz
Senior Lecturer in Geotechnical Engineering, School of Civil and Environmental Engineering, University
of Technology, Sydney (UTS), NSW, Australia. E-mail: hadi.khabbaz@uts.edu.au
Behnam Fatahi
PhD Candidate, School of Civil and Environmental Engineering, University of Technology Sydney (UTS),
NSW, Australia. E-mail: behnam.fatahi@student.uts.edu.au

ABSTRACT Construction on the top of closed landfills is generally a challenging task due to complex
behaviour of creep, settlement and weak shear strength of waste. This paper presents shear failure and settlement
mechanism of soils with high content of organic matters representing the municipal solid waste with or without
chemical stabilisation. An array of consolidation and large-scale direct shear tests have been conducted on
organic soils mixed with different percentages of fly ash to determine the characteristics of organic soil samples.
According to the experimental results and data taken from published literature, it is found that the chemical
stabilisation can decrease the settlement, whilst increase the shear strength of the organic soils.

INTRODUCTION • type of waste,

Many closed municipal solid waste (MSW)
landfills are located near urban areas, even though
originally established away from residential or
commercial communities. There is a high
prospective to reuse these sites for redevelopment
in spite of potential risk for human health and
environment. Landfill design and construction
technology has progressed rapidly during the past
two decades in reaction to more strict controlling
requirement and demands. A solid waste landfill
must be able to prevent groundwater pollution,
collect leachate, permit gas venting and provide for
groundwater and gas monitoring. In this paper, an
experimental program has been carried out on
engineering properties of organic soils (with
characteristics in common with MSW) after
stabilising with various percentages of fly ash.
BACKGROUND
Accurate prediction of landfill compressibility is a
challenging task because of the large number of
variables involved in the settlement process.
Elagroudy and Abdel-Razik (2008) showed that
the following factors have significant effects on
waste settlement ratio:
• organic and moisture contents,
• compaction density,
• porosity,
• compressibility,
• biodegradation rate,
• mode of landfill operation
Compression
Categories of Secondary Settlement Based on the
type of the applied loading, Sharma and De (2007)
showed the secondary settlement in a MSW
landfill can be classified into the following two
categories:
¾ Settlement under self-weight:
This type of settlement is caused by the load
imposed due to the weight of waste material on the
underlying waste layers. Long-term settlement
estimation due to self-weight can be estimated by
the following equation:
t2
∆H S = ∆H ( SW ) = Cα ( SW ) H1 log (1)
t1
where, ǻH (SW) = settlement at time t2 after fill
placement;
t1 = time for primary settlement;
682 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

t2 = time of interest, since the self-weight was Common Characteristics between MSW
applied; Landfills and Organic Soils
H1 = thickness of refuse fill at the end of the Organic soils have many characteristics in
primary settlement; common with MSW in landfills. Organic soil, a
CĮ (SW) = coefficient of secondary compression due major group of soft soils, is considered
to self weight. problematic due to its low shear strength and high
Typical values for CĮ(SW) range between 0.1 and amount of moisture content and compressibility.
0.4 Same as landfill sites engineers are not interested
to construct over lands contain organic soils due to
¾ Settlement under external loads: lack of bearing capacity and uncertainties to
The time-dependent secondary settlement occurs estimate their engineering properties.
over a long period of time in response to externally
applied loads. This settlement can be estimated by In this research, due to common characteristics,
the following equation: some mechanical, physical and chemical tests such
t as permeability, compaction, uniaxial, direct shear
∆H S = ∆H ( EL ) = Cα ( EL ) H 1 log 2 (2)
and consolidation were performed on organic soils.
t1
The results, provided in this paper, are currently
where: ǻH(EL) =settlement at time t2 after external
being compared with samples taken from the
load application;
MSW landfills in the NSW, Australia; and the
t1= time for primary settlement;
findings will be published in the follow up papers.
t2= time of interest, since the external load
application;
EXPERIMENTAL PROGRAM
H1= thickness of refuse fill at the end of the
primary settlement; Permeability Tests (Constant Head)
CĮ(EL)= coefficient of secondary compression due A number of permeability tests on samples of
to external loads. organic soil with and without mixing with fly ash
were conducted by the authors. Physical
TABLE 1 Recommended Values of CĮ Parameters characteristics of the employed organic soil are
(After Sharma and De 2007) given in Khabbaz and Fatahi (2010). Figures 1 and
CĮ under external load, CĮ(EL) 2 show a permeability test apparatus and a sample
Overall range of CĮ 0.01–0.07 of permeability test, respectively. The results in
1
Table 2 present the coefficient of permeability of
With pre-treatment using DDC 0.014 organic soil with and without fly ash stabilisation.
2
With pre-treatment using RC 0.03
With pre-treatment (surcharge) 0.045
CĮ under self weight CĮ(SW)
Fresh waste 0.014–0.06
Waste during active decomposition 0.1–0.34
CĮ for bioreactor landfills
under self-weight CĮ(Bioreactor)SW 0.1–0.34
1
DDC = deep dynamic compaction.
2
RC = roller compaction.

Table 1 showed by Sharma and De (2007) Fig. 1 Permeability test Fig. 2 A sample of organic soil
apparatus. used for permeability test.
summarises the recommended CĮ values for
(For colour figure, refer to CD)
various conditions.
 Khabbaz and Fatahi 683

The results show that the flow rate reduces and

950
consequently, the coefficient of permeability of
organic soil decreases after mixing with fly ash. 900

Dry density (kg/m3)

This reduction in permeability can be attributed to 850
the chemical reaction of fly ash and organic soil.
800

TABLE 2 Parameters and results of permeability test 750

for organic soil (Khabbaz and Fatahi 2010) 700

Parameter Description Value 650

2 -3 25 35 45 55 65 75
A Cross sectional area (m ) 9.49×10 Moisture content %
h Head difference (m) 2.10
L Length of sample (m) 0.173 Organic soil Organic soil + 7.5% Fly ash
t Time (sec) 2700
Q1 Flow (organic soil) (cm3) 1020 Fig. 3 Dry density versus moisture content – organic
Flow(organic soil+fly ash) soil and organic soil with 7.5% fly ash. (Khabbaz and
Q2 323
(cm3) Fatahi 2010)
Coefficient of
K1 permeability organic soil 3.27× 10-06
only (m/s)
Coefficient of
K2 permeability organic 1.03× 10-06
soil + 7.5% fly ash (m/s)

Compaction Tests
The authors also carried out a number of
compaction tests on samples of organic soil with
and without mixing with fly ash. The goal of these
tests was to find the maximum dry density and the
optimum moisture content of organic soil with and
without fly ash.
The compaction results are depicted in Fig. 3 in the
form of dry density versus water content.
According to this figure, by mixing fly ash with
organic soil the maximum dry density increases
from 885 kg/m3 to 935 kg/m3 and the optimum
moisture content decreases from 55% to 46%.
Increase in the unit weight and reduction in the
moisture content are two important parameters in
stabilisation of organic soils and peat.
Figure 4 presents the effect of soil type on the dry
density versus the moisture content. Comparison
of Figs. 3 and 4 reveals that the trend of reduction
in the moisture content and increase in the dry
density by stabilisation of organic soil acts similar
to the effect of soil type on dry density and
moisture content.
Fig. 4 Effect of soil type on dry density versus water
content. (For colour figure, refer to CD)
Unconfined compressive strength of organic soil
with and without stabilisation
A number of uniaxial tests on organic soil were
performed by the authors. The results showed
unconfined compressive strength of organic soil
with and without chemical stabilisation by fly ash.
Tests were performed by various percentages of
fly ash (0%, 5%, 7.5% and 10%). Figure 5 shows
two samples of organic soil for unconfined
compressive strength test mixed with 5% and 7.5%
of fly ash.
Figure 6 shows displacement versus compression
strength of organic soil with various percentage of
fly ash. Based on this figure, unconfined
684 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

compressive strength (UCS) of the organic soil is Shear Strength

increased by increase in percentage of fly ash. It At present, limited data on in situ MSW shear
shows that unconfined compressive strength of behaviour are available. The shear strength of
organic soil has been increased 20% from 50 kPa MSW is generally defined based on the Coulomb
for organic soil to 60 kPa for organic soil mixed failure criterion. Cohesion between particles is a
with 7.5% or 10% fly ash. function of curvature of the failure envelope,
variation between samples, measurement errors,
and an indication of tensile strength. Waste
contains particles that are compressible, can
sustain large tensile strains (e.g. plastic) and can
change with time (e.g. through degradation). An
outcome of using the Coulomb criterion is that it
gives an increase in shear strength with increasing
stress level, based on the slope of the failure
envelop, and hence with depth of burial.

Direct shear Direct shear testing appears to be a

popular test for the evaluation of MSW shear
Fig. 5 Uniaxial test for organic soil with various strength. Zeccos (2005) expressed that the
percentage of fly ash. (Khabbaz & Fatahi 2010) performance of large direct shear tests was
(For colour figure, refer to CD) important for the following reasons:
• The direct shear test is a popular test in the
In addition, Fig. 6 shows that the optimum evaluation of MSW shear strength.
percentage of fly ash to increase the compressive • Direct shear tests generally appear to yield lower
strength of organic soils can be around 7% of estimates of shear strength than triaxial tests.
organic soil’s weight. No further increase would be • Because the failure surface is necessarily
observed in the compression strength of the horizontal and well defined, specimens prepared
organic soil if the amount of fly ash was more than with different particle orientations with respect to
7%. the shearing surface can provide significant insight
regarding the mechanics of waste response and the
70
anisotropic nature of strength in MSW.
60
Compression Strength (kPa)

50
Sample preparation and experimental methods
Soil samples were prepared by adding of fly ash in
40
the range of 7.5% of organic soil’s weight to the
30 samples. Following sample preparation, the
20
specimens were accurately mixed. Then a period
of 7 days was allocated for curing the samples
10 under chemical reactions.
0
0 5 10 15 20 25 30 35 After preparation of specimen, the conventional
Displacement(mm) geotechnical experiments such as consolidation
Organic soil Organic soil + 5% Fly ash
and direct shear test were then performed. The soil
Organic soil + 7.5% Fly ash Organic soil + 10% Fly ash particles under 20 mm were used for tests.
Experimental procedures including consolidation
Fig. 6 Compression strength versus displacement - with
and direct shear test were performed according to
various percentages of fly ash (0%, 5%, 7.5% and 10%).
(Khabbaz and Fatahi 2010) (For colour figure, refer to CD) the Australian standard AS 1289.
 Khabbaz and Fatahi 685

Direct Shear Test RESULTS AND DISCUSSION

Large-scale (300 mm × 300 mm × 180 mm) direct
Direct Shear Tests
shear testing program was performed on organic
A numbers of direct shear tests were performed on
soil. The results show shear strength parameters of
samples of organic soil with and without mixing
organic soil with and without chemical
with fly ash. The results in Figs. 9 and 10 present
stabilisation by fly ash.
the variation of shear stress with normal stress of
organic soil with and without fly ash stabilisation.
Tests were performed with 3 different normal
forces (225, 450 and 710 kg) with the natural
The results show that by mixing fly ash with
moisture content. The speed of applying shear
organic soil the cohesion value increases whereas
force was 0.15 mm/min. Specimens included
the friction angle decreases.
100% smaller than 20 mm materials only. Figure 7
shows a direct shear test apparatus and a sample
preparation of direct shear test, respectively. 100
y = 0.7359x + 32.658

Shear Stress (kPa)

80

60

40

20

0
0 20 40 60 80 100

Normal Stress (kPa)

Fig. 9 Shear stress versus normal stress (organic soil

without fly ash).
Fig.7 Direct shear test apparatus and sample preparation
for the test. (For colour figure, refer to CD)
100
y = 0.6622x + 35.545
Consolidation Tests
Shear Stress (kPa)

80
Consolidation tests were performed on organic
soil. 60
Fig.8 shows samples of consolidation test with and
40
without chemical stabilisation. The results indicate
the changes of void ratio in organic soil with fly 20
ash stabilisation.
0
0 20 40 60 80 100

Normal Stress (kPa)

Fig. 10 Normal stress versus shear stress (organic soil

with 7.5% fly ash).

Fig. 8 Samples of consolidation test with and without
chemical stabilisation. (For colour figure, refer to CD)
Consolidation Tests
Figures 11 and 12 show the applied stress against
the void ratio of organic soil with and without
employing fly ash. As can be seen, by mixing fly
ash with organic soil the void ratio and the total
settlement decreased.
686 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

1.4
attributed to the chemical reaction of fly ash
1.2
and soil.
¾ Organic soils have lower values of unit weight
1
compare to inorganic soils. Through chemical
0.8
stabilisation the maximum dry density of
e

y = -0.3883x + 1.2386
0.6 organic soil is increased and the optimum water
0.4 content is decreased.
0.2 ¾ According to previous investigations, high
0 amount of moisture content is one of the most
0 0.5 1 1.5 2 important problems with organic soils. Fly ash
Log P'
to some extent absorbs moisture content of
e-logp' Linear (e-logp') organic soil and lead to increase the strength of
Fig. 11 Void ratio of versus applied stress in kPa organic soils.
(organic soil with 7.5% fly ash). ¾ Compressive strength of organic soil can
increase by chemical stabilisation using fly ash.
However, this increase is not significant for
1.2
percentages of fly ash greater than 7% of
1 organic soil weight.
0.8

0.6
REFERENCES
e

y = -0.2734x + 0.9736 AS 1289 (2008) Methods of Testing Soil for

0.4 Engineering Purposes, Australian Standard.
0.2
Elagroudy S.A. and Abdel-Razik M.H. (2008)
0 Waste settlement in bioreactor landfill models.,
0 0.5 1 1.5 2 Waste Management 28 ., 2366–2374.
Log P'
Khabbaz H. and Fatahi B. (2010) Chemical
e-logp' Linear (e-logp')
Stabilisation of Organic Soil Using Fly ash,
Fig. 12 Void ratio versus applied stress in kPa (organic The 17th Southeast Asian Geotechnical Conf.
soil with 7.5% fly ash). May 10-13, 2010, Taipei, Taiwan.

CONCLUSIONS Sharma H.D. and De A. (2007) Municipal Solid

In order to stabilise soils with high organic
content, various percentages of fly ash were
applied. An array of tests was conducted. The main
findings of this experimental program can be
summarised as follows:
¾ The changes in shear strength parameters of
organic soil by mixing chemical agent such as
fly ash were examined. These changes include
increase in cohesion and decrease in friction
angle values.
¾ by mixing fly ash in organic soil the void ratio
and total amount of settlement decrease
¾ Coefficient of permeability of organic soil by
mixing a chemical agent such as fly ash is
reduced. This reduction in permeability can be
Waste Landfill Settlement: Postclosure
Perspectives, Journal of Geotechnical and
Geoenvironmental Engineering, Vol. 133, No.
6, 619-629.
Sharma H.D. (2000) Solid waste landfills:
Settlements and postclosure perspectives,
Proc., ASCE National Conf. on Environmental
and Pipeline Engineering, R. Y. Suranpalli,
ed., ASCE, Reston, VA. USA, 447–455.
Zeccos D.P. (2005) Evaluation of Static and
Dynamic Properties of Municipal Solid
Waste, A dissertation for the degree of Doctor
of Philosophy In Engineering - Civil and
Environmental in the graduate division of the
University of California, Berkeley,USA.
 Interdisciplinary Three-Level Approach to Study Impact of
Contamination on Public Health in Puerto Rico
Michael B. Silevitch (m.silevitch@neu.edu), Roger Giese (r.giese@neu.edu),
David Kaeli (kaeli@ece.neu.edu), Thomas Sheahan (tsheahan@coe.neu.edu),
Mariah Nobrega (m.nobrega@neu.edu), Akram Alshawabkeh (aalsha@neu.edu)
Northeastern University, Boston MA
Jose F. Cordero (jose.cordero6@upr.edu)
University of Puerto Rico Medical Sciences Campus, San Juan PR
Ingrid Padilla (padillai@uprm.edu)
University of Puerto Rico Mayagüez, Mayagüez PR
Rita Loch-Caruso (rlc@umich.edu), John Meeker (meekerj@umich.edu)
University of Michigan, Ann Arbor MI

ABSTRACT This article summarizes an interdisciplinary approach to study groundwater contamination and
its impact on public health. Preterm birth is a major health problem in Puerto Rico and there is sufficient
evidence that exposure to environmental contamination is a factor. A three-level approach is adopted to
study this public health issue in Puerto Rico. The approach defines a Grand Challenge (Level 3), Enabling
Systems (Level 2) and Fundamental Science (Level 1). A top-down approach is followed for research
program development and a bottom-up approach is used for implementation. The program involves five
interdisciplinary (Level 1 Fundamental Science) research projects.

INTRODUCTION
Geoenvironmental engineering research and
development, specifically restoration of hazardous
waste contaminated sites to protect public health,
is an application-driven (i.e. focused on a problem
rather than a discipline) science in which basic
research is defined by the need for (a) assessment
of the effect and risk on human health presented by
hazardous substances, and (b) development of
technologies to detect hazardous substances in the
environment, reduce their toxicity, and protect
public health. However, a major challenge for any
application-driven environmental research is to
understand the complex dynamics that arise from
interactions between multiple biomedical and non-
biomedical disciplines and to translate the basic
research beyond the specific disciplines and
academic audience.
In this paper we describe our research program
“Puerto Rico Testsite for Exploring Contamination
Threats (PROTECT)” as an interdisciplinary
approach to study groundwater contamination and
protection of public health.
RESEARCH NEEDS
Preterm birth is a major health problem in the US.
Despite a steady increase in the preterm birth rate
over the past few decades, the issue has received
relatively little attention from the public or the
research community. The increase in preterm birth
is concerning because of the associated rise in the
infant and maternal morbidity and mortality rates,
and other related societal and economic costs. A
report published by the Institute of Medicine
(Behrman & Butler 2007) identifies various
complex factors that can contribute to preterm
birth, including a combination of “behavioral and
psychosocial factors, neighborhood characteristics,
environmental exposures, medical conditions,
infertility treatments, biological factors, and
genetics”. The report recommends establishing
multidisciplinary research centers to address the
complexity of interrelated biological,
psychological, social, and environmental factors
that are involved in preterm birth, and concludes
that “the potential contribution of environmental
688 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
chemical pollutants to preterm birth is poorly
understood.”
To study the impact of soil and groundwater
contamination on preterm birth as a public health
concern and develop prevention strategies, a
holistic approach should be adopted for a specific
community within a well-defined system of
contamination sources, pathways and receptors.
Puerto Rico, as a controlled island environment
with relatively stable population, provides a unique
research opportunity. Compared to other
jurisdictions in the US, Puerto Rico has the highest
rate of preterm births. Furthermore, the increase
has been dramatic over the past few years,
increasing by more than 50% between 1997
(12.8%) and 2004 (19.3%).
Puerto Rico has 14 Superfund sites; phthalates and
chlorinated ethenes are among the most frequently
encountered contaminants at Superfund sites, and
are reported in most of these sites (ATSDR 2002;
HazDat 2008). Recent studies show increasing
and widespread exposure to phthalates in the US
population (CDC 2005), yet little is known about
their toxicity and the reproductive consequences of
exposure in humans. Some phthalates are
considered endocrine disruptors. Among the few
pollutants studied for effects on human parturition,
phthalates were recently associated with decreased
gestation length (Latini 2003). In laboratory
animal tests, they have been shown to cause
reproductive and neurological damage and there is
sufficient evidence that exposure to environmental
contamination such as phthalates may contribute to
the rise of preterm birth (NTP-CERHR 2006). The
toxicity of chlorinated ethenes, particularly to
children, is well-established. They are considered
endocrine disruptors and carcinogens. Some
epidemiologic studies of women exposed
occupationally to TCE and other solvents have
reported increased risk for spontaneous abortion
(Lipscomb and Fenster 1991) and lower birth
weight (Lipscomb and Fenster 1991, Khattak
1999, Ha and Cho 2002).
Environmental exposure to phthalates can occur
through air, water, or food due to phthalates’
ubiquity (they are present in toys, detergents, food
packaging, pharmaceuticals, blood bags and
tubing, and personal care products, for example).
Phthalates are also present in landfill leachate and
are reported at most of the Superfund sites
(ATSDR 2002). Exposure to chlorinated solvents
can occur through contaminated groundwater,
particularly when they are trapped as source zone,
non-aqueous phase liquids (NAPLs). Karstic
aquifers (i.e., those in geologic settings
predominated by dissolution of soluble rock such
as limestone and dolomite) present dynamic
pathways for exposure to contamination due to
presence of fissures, sinkholes, underground
streams, and caverns. Groundwater flow in karstic
aquifers is vulnerable to pollution from either
landfill or Superfund origin, because of the lack of
filtration that is typical of alluvial aquifers. The
dynamics of the solubility, flow and exposure to
phthalates and chlorinated ethenes through karstic
groundwater is made more complex by their
presence as NAPLs and their possible interactions
and transverse partitioning. Seven of the 14
Superfund sites in Puerto Rico exist within karstic
aquifers in the northern part of the island.
Innovative and affordable intervention methods are
required to confront and prevent exposure under
such dynamic flow conditions. The EPA’s
strategic plan for compliance and environmental
stewardship strives for “green” cleanup programs
that use natural resources and energy efficiently,
reduce negative impacts on the environment, and
minimize pollution at its source. Zero valent iron
(ZVI) is an example of a natural resource that is
commonly used for remediation of contaminated
soil groundwater. However, the use of ZVI is
challenged by their uncontrolled reactivity and
passivation. Development of “green” technologies
that can control reactivity of ZVI while using
natural resources fits within EPA’s strategic plan
for compliance and environmental stewardship.

Fig. 1 Three level holistic approach for (top-down) development
and (bottom-up) implementation of PROTECT
THREE-LEVEL APPROACH
In this program, we define a three-level holistic
approach to study environmental contamination
and preterm birth using “Puerto Rico as an
enabling system”. Our approach defines
“Fundamental Science” as Level 1, “Enabling
Systems” as Level 2 and Public Health “Grand
Challenge” as Level 3, as shown in Fig. 1.
The three levels of the proposed program are
defined as:
Level 3 – Grand Challenge:
Preterm birth correlation to “Superfund”
environmental contamination: assessing exposure
and risk and developing intervention strategies.
Level 2 – Enabling System:
Puerto Rico, a community specific, controlled
island environment to study contamination
sources, pathways, receptors, risks and protection.
Level 1 – Fundamental Science:
Research framework and thrusts identified within
the scope of assessing exposure to environmental
contamination; its effect on preterm birth, and
development of intervention strategies.
Silevitch et al. 689
A top-down methodology (Fig. 1, Steps 1 – 3)
informs the basic research program development.
By specifying the use-inspired “Grand Challenge”
(Level 3, Step 1), we can identify an Enabling
System (Level 2, Step 2) such as Puerto Rico, a
community-specific, well-defined, relatively
controlled system that can be used for scientific
research on exposure, outcome and risk. From the
system identified in Step 2, we can then identify
cross-disciplinary research barriers, thrusts and
goals at the fundamental science level (Level 1,
Step 3). Conversely, a bottom-up implementation
(Fig. 1, Steps 4 – 6) of the research program will
conduct fundamental science in cross-disciplinary
projects (Level 1, Step 4). Basic science will lead
to development of models, testbeds,
demonstrations, geographic information systems
(GIS) and other research translation systems
(Level 2, Step 5). Finally, system-level validation
is implemented at the Grand Challenge level
(Level 3, Step 6) to engineer prevention strategies
to the problems identified in Step 1.
FUNDAMENTAL SCIENCE THRUSTS AND
PROJECTS
The Program’s “Fundamental Science” (Level 1)
defines several critical research thrusts. Cross-
690 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
Fig. 2 Three-level Strategic Approach to the PROTECT Program (For colour figure, refer to CD)
disciplinary fundamental science research is
required for understanding contamination fate
(sources, transport and exposure), threats (response
at the cell, organ and body scales) and
interventions (characterization, risk assessment,
remediation and clinical treatment) in a holistic
approach. Our framework is based on phthalates
(primarily di-2-ethylhexyl phthalate or DEHP) and
chlorinated ethenes (specifically trichloroethylene
or TCE) as the primary contaminants of interest for
many reasons, including:
x they are predominant groups of contaminants
at Superfund sites in Puerto Rico and the US
mainland;
x phthalates are also present as a plasticizer in a
wide range of personal care products – e.g.
perfumes, lotions, cosmetics – and are in the
class of “emerging contaminants,” a category
of increasing concern with regard to
groundwater contamination and related human
health effects;
x some phthalates were recently associated with
decreased gestation length (Latini, 2003); and
x chlorinated ethenes are the primary
contaminants of concern at recently added
(e.g., Maunabo, listed 2006; Cidra, listed
2004; and San German, listed March 19,
2008) Superfund sites in Puerto Rico.
The ultimate strategic goals of the PROTECT
program are to translate knowledge gained from
fundamental research.
Figure 2 presents the proposed bottom-up
implementation of our strategic plan. To facilitate
the translation of the fundamental science (Level
1), we define the Research Translation within
enabling systems (Level 2) in order to meet the
requirements of ultimate use (our grand challenge
– Level 3). In other words, the Research
Translation Core (as enabling system) will serve to
advance both useful knowledge (Level 1
fundamental science) and informed action (Level
3) by creating a dynamic bridge between the two.
The PROTECT Program includes five integrated
biomedical and non-biomedical research projects
as displayed in Figure 3:
x Project 1: Discovery of Xenobiotics
Associated with Preterm Birth
x Project 2: Pollutant Activation of Cell
Pathways in Gestational Tissues
x Project 3: Phthalate Exposure and Molecular
Epidemiologic Markers of Preterm Birth
among Women in Puerto Rico

x Project 4: Dynamic Transport and Exposure
Pathways of Contaminants in Karst
Groundwater Systems
x Project 5: Green Remediation by Solar Energy
Conversion into Electrolysis in Groundwater
The projects are supported by the following cores:
x Core A: Administrative Core
x Core B: Research Translation Core
x Core C: Human Subjects and Sampling Core
x Core D: Data Management and Modeling
Core
x Core E: Training Core
The Program is multidisciplinary and
interdisciplinary, since it involves significant
interaction and sharing of samples, testing and
results among the disciplines of analytical
chemistry, epidemiology, engineering and
toxicology. The Program will be greater than the
Silevitch et al. 691
sum of its projects because the projects are highly
integrated, as shown below. The integration is
enhanced by mutual use of the cores, including
Core D, a centralized data repository for sharing
and mining results to achieve the Program Goals
and Objectives.
ACKNOWLEDGEMENT
The project described was supported by Award
Number P42ES017198 from the National Institute
of Environmental Health Sciences. The content is
solely the responsibility of the authors and does
not necessarily represent the official views of the
National Institute of Environmental Health
Sciences or the National Institutes of Health.
692 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
REFERENCES
ATSDR. 2002 Toxicological Profile for Di(2-
ethylhexyl)phthalate (DEHP). Atlanta, GA:
Agency for Toxic Substances and Disease
Registry.
Behrman, R. E., and Stith Butler, A., eds. (2007).
Preterm Birth: Causes, Consequences, and
Prevention. Committee on Understanding
Premature Birth and Assuring Healthy
Outcomes, Institute of Medicine of the
National Academies, The National Academies
Press, Washington, DC.
CDC. 2005. Laboratory Procedure Manual for
Phthalates. Atlanta, GA: US Centers for
Disease Control and Prevention.
http://www.cdc.gov/nchs/data/nhanes/nhanes_
03_04/l24ph_c_met.pdf Accessed March 25,
2008.
NTP-CERHR. National Toxicology Program-
Center for the Evaluation of Risks to Human
Reproduction. Draft Update Monograph on
the Potential Human Reproductive and
Developmental Effects of Di(2-ethylhexyl)
Phthalate (DEHP). Research Triangle Park
(NC). 2006 [online]. Available at URL:
http://cerhr.niehs.nih.gov/chemicals/dehp/deh
p-eval.htmlExternal Web Site Icon.
Ha, E., S. I. Cho, et al. (2002). "Parental exposure
to organic solvents and reduced birth weight."
Arch Environ Health 57(3): 207-14.
HazDat. Query Search at
http://www.atsdr.cdc.gov/hazdat.html, March
12, 2008.
Khattak, S., G, K. M., McMartin, K., Barrera, M.,
Kennedy, D., and Koren, G. (1999).
Pregnancy outcome following gestational
exposure to organic solvents: a prospective
controlled study, Jama 281, 1106-1109.
Latini G, De Felice C, Presta G, Del Vecchio A,
Paris I, Ruggieri F, Mazzeo P. 2003. In utero
exposure to di-(2-ethylhexyl) phthalate and
duration of human pregnancy. Environ Health
Perspect 111:1783-1785.
Lipscomb, J. A., L. Fenster, et al. (1991).
"Pregnancy outcomes in women potentially
exposed to occupational solvents and women
working in the electronics industry." J Occup
Med 33(5): 597-604.
 Monitoring Nanoiron Transport in Porous Media Using Magnetic
Susceptibility Sensor
Kenneth Darko-Kagya (kdarko2@uic.edu) and Krishna R. Reddy (kreddy@uic.edu)
University of Illinois at Chicago, Department of Civil and Materials Engineering,
842 West Taylor Street, Chicago, Illinois 60607, USA

ABSTRACT Nanoiron has recently shown to be effective for reduction of recalcitrant organic contaminants
such as PCP, DNT and TCE in the environment. This paper investigates the transport of nanoiron through
porous media (natural sand) without and with lactate surface modification. Magnetic susceptibility (MS) or
κ sensor system was used to monitor the real-time transport of nanoiron in sand column experiments. A
series of laboratory experiments was conducted using a Plexiglas® column with an inside diameter of 38 mm
and length of 135 mm. Bartington MS2C magnetic susceptibility meter sensor was used to monitor the
transport of bare and lactate-modified nanoiron at two different slurry concentrations of 1 g/L and 4 g/L
under two different flow rate (hydraulic gradient) conditions. The results showed a linear correlation between
the iron concentration and κ. Overall, this study showed that MS can be used for real-time monitoring of
nanoiron transport in porous media, and the 4 g/L lactate-modified nanoiron transported better than bare
nanoiron.

INTRODUCTION
Nanoscale iron particles (NIPs) have recently
been shown to possess the potential to
decontaminate soils and groundwater
contaminated by dinitrotoluene (DNT),
trichloroethylene (TCE), pentachlorophenol
(PCP) and chlorophenol (CP). This is due to
their unique properties such as infinitesimal
small size and high surface to volume ratio.
Reddy and Karri (2008) effectively treated PCP
contaminated soils using NIPs. Although the
reactivity of NIPs has been established by
researchers, there have been limited studies on
the mobility or transport of NIPs. USEPA-NSF
workshop held in 2006 concluded that there are
limited studies on the transport and fate of NIPs
(Reddy, 2006). Delivery of NIPs uniformly in
required amounts is essential for a successful in-
situ remediation of soils and groundwater.
Agglomeration and fast settling of NIPs
impedes transport of NIPs in soils during in-situ
applications. Researchers have tried to modify
the surface of the NIPs particles with different
types of dispersants such as polymers,
surfactants and cyclodextrins (Cameselle et al.,
2008; Saleh et al., 2007) to improve transport in
subsurface. The dispersants are believed to form
a light film around the surface of the particles
thereby reducing sticking coefficient and
attachment efficiency. The modification
changes the hydrodynamic radius and can also
alter the surface charge of the particles reducing
aggregation (Cameselle et al., 2008, Saleh et al.,
2007).
Magnetic susceptibility (MS) is a fast,
inexpensive, nondestructive and straightforward
system that has been used to monitor and detect
iron bearing minerals in rock cores and metal
pollution in soils and sediments. It has also been
used in mapping and distribution of pollutants in
the environment. This is because most of the
polluted particles in the environment contains or
either attached by magnetic particles. MS is the
ratio of magnetization to the applied magnetic
field and it measures how a magnetic material
will respond to the presence of magnetic field.
When the material comes near magnetic field, it
694 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
gets magnetized according to the amount of
iron-bearing minerals available. A material can
be paramagnetic, diamagnetic and
ferromagnetic due to how they respond to the
presence of a magnetic field. The presence of
more iron bearing minerals per unit volume in a
sample will indicate high MS values. Paradelo
et al. (2009) used magnetic susceptibility to
determine the presence of metal contamination
in eight different composts. The study found a
high correlation between MS and total
concentration of Cd, Zn, Pb, Cr and Ni present.
No studies have been reported to investigate to
monitor the mobility of NIPs in soils with
magnetic susceptibility system.
NIPs are magnetic and they can respond to the
presence of applied magnetic field. In this study,
the transport of bare and lactate modified NIPs
was examined in column experiments by
monitoring the movement of NIPs in natural
field sand using Bartington MS2C magnetic
susceptibility meter with 130 mm diameter
sensor. Bare and lactate-modified NIPs were
investigated at two different slurry
concentrations of 1 g/L and 4 g/L. Different
hydraulic gradient (0.5, 1.0) were examined to
find the effect of flow rate. The extent and
transport of NIPs were tracked using the
magnetic susceptibility as they move through
the soil.
MATERIALS AND METHODS
Materials
NIPs used in this study were obtained from
Toda Kogyo (Japan). The particles had an
average diameter of 70 nm (with a range of 50-
300 nm), pH of 10.7, and BET surface area of
37.1 m2/g. Based on X-ray diffraction methods,
the NIPs are found to consist of an elemental
iron core (α-Fe) and a magnetite shell (Fe3O4).
The approximate composition of NIPs is 50
wt.% α-Fe core and 50 wt.% Fe3O4. The density
of the aqueous NIPs particle suspension is 1.27
g/mL at solids concentration of 25.6 wt.%. Due
to the presence of magnetic properties of NIPs,
the particles will exhibit magnetic susceptibility
in the presence of a magnetic field. This
indicates that the movement or mobility of NIPs
can be tracked or monitored with the help of a
magnetic susceptibility sensor. Therefore the
use of the magnetic system is to take advantages
of the magnetic properties of NIPs to monitor
their mobility.
Native field sand was used for this study to
represent typical field soils. Most transport
studies conducted used glass beads which do not
represent real world soil conditions. The sand
used is classified as poorly graded sand (SP).
The aluminum lactate was used for the surface
modification of NIPs and electrolyte was used
in order to simulate groundwater conditions.
The electrolyte contained 0.006 M of sodium
bicarbonate, 0.002 M of calcium chloride and
0.001 M of magnesium chloride.
Test Set-up, Variables and Procedures
Series of transport experiments were performed
in a one dimensional bench-scale horizontal
column set up. A Plexiglas® column of inside
diameter of 38 mm and length of 135 mm was
used. The set-up consisted of two porous stones,
a cell, two 0.45 micron cellulose filter papers, a
reservoir, and Tygon tubing. A Marriott tube of
diameter 40 mm was used as the inlet reservoir
to maintain the desired constant hydraulic head
throughout the experiment. The height of the
reservoir could be adjusted to apply different
constant hydraulic head conditions.
Two different concentrations (1 and 4 g/L) of
NIP- slurries were prepared using electrolyte
and additional two slurries were prepared with 1
and 4 g/L NIPs containing 10% aluminum
lactate (w/w NIPs). Two different gradients (HG)
(0.5 and 1) were investigated to check the effect
of flow rate. The sand was packed into the cell
in uniform layers and compacted using a tamper
to ensure uniform initial soil density. Before the
start of each experiment, 200 ml of electrolyte
alone was flushed to fully saturate the sand
column. The electrolyte in the inlet reservoir
 Darko-Kagya and Reddy 695

was replaced with the selected NIP-slurry and
allowed to flush through the sand. One liter of
NIPs and Lactate-modified NIPs (LM-NIPs)
with different concentration solutions were
prepared just prior to the start of the testing and
were introduced into the inlet reservoir. Care
was taken to keep the NIPs in suspension by
shaking the inlet solution regularly. The effluent
samples were collected in 120 mL bottles (i.e.
approximately every 3 pore volumes) for
analysis. At the end of each experiment, the soil
was extruded from the column and sectioned
into four parts. Soil sample from each section
RESULTS AND DISCUSSION
Effect of Aluminum Lactate on the
Transport of NIPs
The variation of the magnetic susceptibility
along the length of the soil column is attributed
to changes in concentrations of magnetic
materials (in this case NIPs). There was no
significant variation or change in magnetic
susceptibility from the inlet to the outlet for the
baseline condition during which there was no
introduction of NIPs into the soil (Fig.1).

Change in Magnetic Susceptibility (X10 )

-5
140
was visually observed and photographed, and 10min
120
iron concentrations were measured using acid Baseline 15 min
25 min
i=1
100
digestion and atomic absorption spectrometry 35 min
45 min
(AA). 80 55 min
75 min
60 85 min
95 min
During the experiments, Bartington MS2 40 100 min
105 min
magnetic susceptibility system was used for 20 125 min

monitoring real-time movement of the NIPs and 0

LM-NIPs in the soil. The system consisted of a
-20
MS meter and it was equipped with 130 mm 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
diameter sensor. The sensor was mounted on a No rmalized Distan ce from Inlet

specially fabricated retractable stand that can be
moved manually freely back and forth.
Magnetic susceptibility measurements were
taken at 10 mm distance interval with a
resolution of 1.0 SI units along the length of the
column starting from the inlet. This equipment
can measure at lower sensitivity (1.0) or higher
sensitivity (0.1) controlled by a switch. The
system can measure from 1 to 9999 x 10-5(SI) or
1 to 9999 x 10-8 (cgs). Before starting the
experiment, the susceptibility meter reading was
zeroed. At different locations of the sample, MS
measurement was then made to serve as the
baseline values in absence of NIP. The
measurements were then taken at different time
intervels while the NIP-solution was flowing
through the soil. Different tests were conducted
under two different hydraulic gradient (HG) (0.5
and 1.0) to assess the effects of varying flow
rate on NIPs transport. The extent and transport
of NIPs were inferred based on the MS variation
across the soil sample with respect to the
baseline values.
Fig.1 MS values along the column at different time
periods for the baseline test without NIPs and HG=1
(For colour figure, refer to CD)
Figure 2 shows the MS values at different time
periods for the test with NIPs and LM-NIPs at
HG of 1. In these experiments, high MS values
were observed at the inlet sections of the
column and then they reduced near the outlet. It
can be inferred that NIPs were able to transport
pass the mid point of the LM-NIPs column than
that of the bare NIPs column from the point of
inlet. For example, at normalized distance of
0.5, the MS value measured for 4g/l LM-NIPs
test was 45x10-5 SI units compared to 24x10-5 SI
units for 4 g/l NIPs test.
Similarly for HG of 0.5, the MS value at
normalized distance of 0.5 for 4 g/L LM-NIPs
was 36x10-5 SI units whereas that of 4 g/L NIPs
was 8x 10-5 SI units (results not shown). High
MS values at a point is an indication that more
NIPs have traveled to that point, while low MS
measurements implies that less amount of NIPs
696 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

transported to that point. Aluminum lactate This can be observed in Fig 3 as the MS value
enhanced the transport or mobility of NIPs near the inlet at HG of 1 for 4 g/L NIPs was
through the soil. This can be attributed to the 118x10-5 SI units and that of 1 g/L NIPs was
fact that the lactate forms a film around the 38x10-5 SI units. This shows that MS system
NIPs, thereby reducing sticking coefficient, can be used to quantify the extent of transport of
aggregation and attachment to soil particles. For NIPs in soils. It can also be seen that the MS
10% Aluminum lactate, the surface charge of values for 4 g/L NIPs were high near the inlet,
the NIPs also reduces from a positive value to but the values drop drastically just before the
negative value. The negative charge reduced the mid section of the soil column. However in the
sorption/absorption of NIPs to the negatively case of 4 g/L LM-NIPs, the MS at the inlet was
charged soil particles. initially lower than that of the NIPs test but
Change in Magnetic Susceptibility (X10 )

there was gradual decrease of MS measurement

-5

120 3min
10 m in
100
4g/l NIP
i= 1
20 m in away from the inlet indicating that the particles
30 min
40 m in have been transported relatively more and
80 50 m in
60 m in uniformly than that of NIPs. There was not
65 min
60 75 m in much difference between the two lower NIP
85 m in
40 95 m in
105 min
concentration tests in-terms of their MS values.
Change in Magnetic Susceptibility (X10 )
110 min -5 120
20
4 g/L LM-NIP, HG=1
100 4 g/L NIP, HG=1
0 1 g/L LM-NIP, HG=1
Inlet O utlet
1 g/L NIP, HG=1
-20
80
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
60
N o rm a liz ed Di sta n ce fr o m Inl et
Fig.2a MS values for 4 g/L NIP and HG=1 test along
40
the column at different time periods
(For colour figure, refer to CD) 20
Change in Magnetic Susceptibility (X10 )
-5

100
6min
0
15 min Inlet Outlet
80
4g/l LM-NIP 25 min -20
35 min
HG = 1
45 min
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
60
65 min Normalized Distance from Inlet
80 min
40 90 min
100 min
Fig. 3 MS values along at the end of the testing
105 min
(For colour figure, refer to CD)
20

0 Figures 4 and 5 show the MS measurements for

-20
0.0
Fig. 2b MS values for 4 g/L LM-NIP and HG=1 test
along the column at different time periods
(For colour figure, refer to CD)
The MS of NIPs was measured at the end of
each test and all the results are plotted together
(Fig.3). It can be seen that all the higher NIPs
concentration (4g/l NIPs and LM-NIPs)
conditions yielded higher magnetic
susceptibility values than the lower NIP
concentrations (1g/l NIPs and LM-NIPs) which
indicates the fact that the presence of higher
amount of NIPs caused higher values of MS.
both 1 g/L NIPs and LM-NIPs tests. It can be
Inlet Oulet
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 inferred from the graph that at normalized
Normalized Distance from Inlet distance of 0.5 for HG of 1, the MS value for 1
g/L NIP test was 10x10-5 SI units which is not
significantly different than that of 1 g/L LM-
NIP test with a value of 11x10-5 SI units. This
may be attributed to the fact that aggregation is
less at lower concentrations; therefore, their
mobility is somehow similar. For HG = 0.5, it
was observed that the MS values of 1 g/L LM-
NIP test are slightly higher than that of 1 g/L
NIP test, but the same cannot be said for the test
with HG=1 . This can be also attributed to the
high flow rate in the case of HG=1, causing
similar NIPs transport.
 Darko-Kagya and Reddy 697

be observed at lower concentrations. The MS

Change in Magnetic Susceptibility (X10 )
-5

1 20
3m in

1 00
40

30
10 m in
20 m in
values for HG of 0.5 at normalized distance of
20
1 g /l N IP
HG = 1
1 g/ l B a r e N I P
25 m in
30 m in
0.5 for 1 g/L NIP test and 1 g/L LM-NIP test are
80
6x10-5 SI units and 15x10-5 SI units,
i = 1
10 35 m in
4 5 m in
0
60 50 m in
-1 0 55 m in respectively. This indicates that lactate modified
0 .0 0 .1 0 . 2 0 . 3 0 . 4 0 .5 0 .6 0 .7 0 . 8 0 . 9 1 .0 65 m in
40 75 m in NIPs were better, but at higher gradient the
20 mobility of 1 g/L NIPs is almost the same as 1
0
g/L LM-NIPs.
In le t Ou tle t

-20
0 .0 0 .1 0 .2 0.3 0.4 0.5 0 .6 0 .7 0 .8 0 .9 1 .0 Correlation of Magnetic Susceptibility and
No rm alize d Dis tan ce fro m In le t Iron Concentration
Fig.4 MS values for 1 g/L NIP and HG=1 test along At the end of the experiments, the soil samples
the column (For colour figure, refer to CD) were extruded from the columns, divided into
6 m in
four sections and chemical analysis was
120
1 0 m in
performed to determine the concentration of the
Change in Magnetic Susceptibility (X10 )
-5

1 5 m in
2 0 m in
100
40

30
1 g / l + A l L a c ta te
i = 1
3 0 m in
4 0 m in
iron in each section based on acid digestion and
80
20
5 0 m in
6 0 m in
atomic absorption (AA) analysis.
10 6 5 m in
7 5 m in
0 8 5 m in
60
-1 0
I n le t

0 .0 0 .1 0 .2 0 . 3 0 . 4 0 . 5 0 . 6 0 . 7 0 . 8 0 . 9 1 .0
O u tl e t

Figure 6 shows the iron concentrations at the

40
20
0
-2 0
0 .0
Fig. 5 MS values for 1 g/L LM-NIP and HG=1 test
along the column at different time period
(For colour figure, refer to CD)
Effect of Hydraulic Gradient (HG)
In assessing the transport of NIPs in subsurface
for effective remediation, one parameter such as
the flow rate (or HG=1) cannot be overlooked.
Increase in flow rate or HG also increases the
extent of transport of NIPs. For example, for 4
g/L NIP tests, at a normalized distance of 0.6,
the MS value for HG=1 was 18x10-5 SI units as
compared to a lower MS value of 6x10-5 SI
end of each test in different sections from inlet
to outlet. It can be observed that higher
concentration of iron at the inlet section and
then reduced with distance away from the inlet.
0 .1 0.2 0 .3 0 .4 0 .5 0 .6 0 .7 0 .8 0.9 1 .0
For the 4 g/L NIP tests, almost all the particles
N orm alised Distance from Inlet
accumulated in the section close to the inlet and
then reduced drastically away from the inlet
indicating that the NIPs were not able to
transport far away from the inlet. However, in
the case of 4 g/L LM-NIP tests, the iron was
higher throughout the column, indicating high
transport. This was attributed to the presence of
lactate. Results in Fig. 6 follows similar trend
as the results shown in Fig 3. This shows that
magnetic susceptibility can be used as a proxy
measurement to estimate the transient
transport/mobility of NIPs.
250 00
Concentration of Iron (mg/kg)

4g/l LM-N IP
units at the same point with a HG of 0.5. 200 00 4g/l N IP
1g/l LMN IP
Similarly for HG=1, 4 g/L LM-NIP test, MS 150 00
1g/l N IP

value at normalized distance of 0.6 yields 32e-5

SI units, whereas it is 20x10-5 SI units for 100 00

HG=0.5. 50 00

This indicates that the particles are able to 0

0 1 2 3 4 5
Sec tions
transport more distance at higher flow rate than Fig. 6 Iron concentrations at end of testing (section 1
lower flow rate and this was affirmed by Kanel near inlet) of the column tests with HG=1
and Choi (2007). The effect of lactate can also (For colour figure, refer to CD)
698 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

The iron concentrations and the MS values just NIPs was found to be higher in the 4 g/L
before the termination of the tests were used LM-NIP test.
determine any correlation between iron 3. A significant correlation was observed
concentration and MS values in the soil. between magnetic susceptibility and iron
Significant correlation was observed for each concentration in both NIPs and LM-NIPs
test with a correlation coefficient (R2) greater tests. Therefore, MS monitoring can provide
than 0.76 (data not shown). Figure 7 shows real-time data on the transport and
results from all of the tests, also demonstrating distribution of NIPs in soils.
even better correlation between MS and iron 4. Increase in flow rate (hydraulic gradient)
concentrations. caused increase in the transport of NIPs.
25000
ACKNOWLEDGEMENT
Concentration of Iron (mg/kg)

20000 All Conditions

Gradient = 1
Financial support for this project is provided by
the U.S. National Science Foundation (Grant
15000
CMMI #0727569), which is gratefully
5
y = 136.36x10 MS + 7334.4
2
R = 0.8459
acknowledged.
10000
1 g/L LM-NIP

REFERENCES
4 g/L NIP
1 g/L NIP
5000 Regression Line
4 g/L LM-NIP Cameselle, C., Darko-Kagya, K., Khodadoust,
0 A., and Reddy, K. R. (2008). “Influence of
0 10 20 30 40 50 60 70 80 90 100 110 120 13 Type and Concentration of Dispersants on
0
Change in Magnetic Susceptibility (MS)(x10 -5 ) the Zeta Potential of Reactive Nanoiron
Fig. 7 Correlation of MS and iron concentrations in Particles,” Proc. International
soil based on all column tests Environmental Nanotechnology Conference,
(For colour figure, refer to CD) USEPA, Chicago.
CONCLUSIONS Kanel,S.R and Choi, H. (2007). “Transport
This study investigated if magnetic Characteristics of Surface-Modified
susceptibility can be used as a rapid, simple, Nanoscale Zero-Valent Iron in Porous
reliable and nondestructive method to estimate Media.” Water Sci & Techn Vol 55 (2 ):
the existence and movement of NIPs through 157–162.
soils. Laboratory tests were conducted using Paradelo, R., Moldes, A.B., and Barral,
NIPs and LM-NIPs in sand columns under M.T.(2009). “Magnetic Susceptibility as an
different hydraulic gradient conditions. MS Indicator of Heavy Metal Contamination in
measurements were made during the testing, Compost.” Waste Management & Research:
and the iron concentrations were determined at 27: 46–51.
the termination of testing and were correlated Reddy, K.R. (2006). “Transport of Modified
with the MS measurements made just before the Reactive Nanoscale Iron Particles in
test termination. Subsurface Soils.” Proc. Nanotechnology
for Site Remediation Workshop, EPA
Based on the experiments performed, the 905K07001, USEPA, Region 5 Superfund
following conclusions can be drawn: Division, pp.11-11.
1. Tests using aluminum lactate modified NIPs Reddy, K.R., and Karri, M.R. (2008). “Removal
showed enhanced transport of NIPs through and Degradation of Pentachlorophenol in
the sand. Clayey Soil Using Nanoscale Iron
2. Aluminum lactate reduces agglomeration of Particles.” Geotechnics of Waste
NIPs, thereby enhancing dispersivity and Management and Remediation,
transport of NIPs in sand. The transport of Geotechnical Special Publication, No.177,
ASCE , pp.463-469.
 Role of Colloids in the Transport of Radionuclides in the Ground
Environment

R. Deepthi Rani, P. Sasidharand S.E. Kannan

Safety Research Institute, Atomic Energy Regulatory Board, Kalpakkam 603 102, India
(sasidhar@igcar.gov.in)

ABSTRACT Earlier approaches treated the transport of radionuclides as a two-phase system comprised of
immobile aquifer media and mobile aqueous phase. Many hydrophobic contaminants are considered to be retarded
due to strong interactions with immobile aquifer material. Experimental evidence now exists that radionuclides are
transported not only in the dissolved state by water but also by sorption to colloids. Failure to account for colloid
facilitated transport underestimates the distance radionuclides can migrate. The present study highlights the need
and relevance of detailed colloid characterization to assess its role in the migration of radionuclides from near
surface disposal facility. This paper discusses the sampling methods of colloids; determination of their size and
charge; effect of ionic strength on colloid stability and their role in the transport of radionuclides in the ground
environment.

INTRODUCTION
The following criteria were proposed (Ryan &
Many hydrophobic contaminants are considered
Elimelech, 1996) for the transport of
to be retarded due to strong interactions with
radionuclides by colloids: a) Colloids must exist
immobile aquifer material. But recent studies in
in an appreciable concentration, b) colloids must
groundwater at the Nevada Test Site (Kersting et
be mobile and stable in the environment of
al. 1999) detected plutonium at a large distance
interest, c) radionuclides must have high sorption
than predicted by existing models. Also at Los
affinity for colloids, d) colloids and the associated
Alamos National Laboratory both plutonium and
radionuclides must be transported through the
americium were detected in monitoring wells as
aquifer. Key factors affecting the stability of
far as 3390 m downgradient from the discharge
colloids are pH, redox potential, salinity and
and also were found to be associated with
hardness (Degueldre et al., 1996a; Degueldre et
colloidal material in the groundwater (Penrose et
al., 1996b). It is necessary to characterize the
al. 1990). Minhan Dai et al (2005) also reported
colloids for its size, size distribution, zeta
the detection of plutonium in the groundwater at
potential and surface charge, because these are the
the Hanford Site 100K-Area. In subsurface
parameters that determine the stability and
environment, colloids having similar chemical
mobility of these colloidal particles. The extent of
composition as that of immobile aquifer material
radionuclide sorption on to colloids also depends
can sorb the radionuclides and can facilitate the
on nature of clay and its exchange capacity. For
transport (McCarthy & Zachara, 1989).
proper risk assessment and remediation
Consequently, the association between the
measurements, the role of colloids in radionuclide
contaminant and colloid affects the transport of
transport in geo environments should be
the contaminant. Failure to account for colloid-
examined in greater detail.
facilitated transport can underestimate the
distances the radionuclides will migrate. There is
This paper discusses the sampling methods of
increasing evidence for the association of
colloids and the effect of ionic strength on colloid
radionuclides with inorganic and organic colloids
stability. This study also discusses the
and that such an association can influence the
characterization of groundwater colloids for its
mobility of radionuclide (Degueldre et al. 1989).
concentration, size, size distribution and zeta
700 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
potential to assess its role in the migration of
radionuclides from a near surface disposal facility.
STUDY AREA
The radioactive waste disposal facilities at
Kalpakkam require a proper understanding of the
possible mechanisms of radionuclide transport in
the groundwater. The study area is situated about
60 km south of Chennai and is bounded on the
east by Bay of Bengal, on the north by Edaiyur
backwater and on the west by Buckingham Canal
(Fig 1). The entire area is covered with alluvial
sands and weathered charnockite underlain by
charnockite rock. Lenses of clays and clay
pockets were encountered at some locations in the
alluvial formations (Singh et al. 2003).
Fig 1. Location of borewells in the study area
MATERIALS AND METHODS
Sampling Technique
In the present study, a peristaltic pump was used
to collect samples at a flow rate of 100 ml/ min. A
multi-parameter instrument (YSI 556 MPS, YSI
Inc., USA) with flow-through-cell was used to
monitor water quality parameters. Turbidity of
unfiltered groundwater was measured in
nephelometric turbidity units (NTUs) using a
Hach 2100P portable turbidimeter. Field
parameters and turbidity were recorded after
stabilization. Groundwater samples for physico
chemical characterization were collected in
properly labeled bottles.
Laboratory analysis
Groundwater analysis for major cations and
anions
Chemical composition of the groundwater
samples was obtained by atomic absorption
spectroscopy, turbidimetry and titrimetric
techniques. The major ions such as Ca2+, Mg2+,
Cl-, CO32- and HCO3- were determined
titrimetrically (APHA, 1976). Na+ and K+ were
determined using GBC Avanta Model atomic
absorption spectroscopy. Turbidity and sulphates
were determined using turbidimeter (Hach
2100AN Desktop turbidimeter).
Colloid characterization
Colloid concentration
The abundance of natural colloids was measured
by turbidity both in the field and in the laboratory.
Colloid concentration was measured by filtering
groundwater samples through previously weighed
220nm filters and measuring the weight of the
particles deposited on the filters. Colloid
concentration is expressed in mg/L. The number
of particles (Np, particles/L) was calculated from
the mass of colloidal particles (Mp, mg/L) using
Eq-1
Mp= Np πd3ρ/ 6 (Eq-1)
where Mp is the mass of colloidal
particles in mg per liter, Np is the number of
particles per liter, d is the diameter of the particle
in cm and ρ is the particle density. The particle
density of 2600 mg/cm3 was assumed
(Randall.J.Schaetzl, 2005) for clay minerals. The
particle diameter was measured using the
procedure explained in the following section and
the mean diameter was computed from the
 Rani, Sasidhar and Kannan 701

diameter of different colloids encountered in the as EC, TDS, salinity, DO, pH, turbidity and
study area. colloid concentration. In addition detailed
Particle size and size distribution characterization of colloids was attempted for
selected groundwater samples where colloid
The colloid size distribution was measured using
concentration was appreciable.
Mastersizer Micro (Malvern Instruments). The
Mastersizer Micro comprises a Helium-Neon The physicochemical characteristics of the
laser as a light source, which illuminates the groundwater samples during January 2007-
dispersed particles in the measuring zone. This is January 2008 are given in Table-1.The EC values
then focused by a Fourier lens to a detector. The of groundwater samples was high near
intensity of the scattered light is measured and a Buckingham Canal (western side of study area)
volumetric particle size distribution is calculated. compared to groundwater samples near the shore
The average colloid size was measured using (eastern side of the study area) and is attributed to
Zetasizer Nano. constant saline water charging of the aquifer by
Buckingham Canal.
Zeta potential Colloid characterization
The zeta potential measurement was made using a Colloid concentration
Zetasizer Nano (Malvern Instruments) that uses a
Colloids are present in every groundwater sample
clear disposable zeta cell for zeta potential
although there was a variation in their
measurement. An average of 12 runs was taken to
concentration. The colloid concentration was
represent the potential. The instrument calculates
varying between 0.05mg/L and 6mg/L. The
the electrophoretic mobility of the particle and
number of particles per liter was calculated from
converts it to zeta potential using the
colloid concentration using Eq-1. The number of
Smoluchowski equation. The Smoluchowski
particles was found to be varying between 3 х 109
equation (Eq-2) gives the direct relationship
particles/L to 3 х 1011 particles/L. Results are
between zeta potential and electrophoretic
given in Table-2. It was observed that colloid
mobility.
concentration was less in borewells BW6, BW7,
BW8 and PBW1 (Fig 1) that are close to saline
ζ = 4π η u/ ε (Eq-2)
water bodies like Buckingham canal and Bay of
Bengal.
TABLE 1. Physicochemical characteristics of the
where u is electrophoretic mobility, η is viscosity
groundwater samples
of the suspending liquid in centipoise, ε is
dielectric constant, and ζ is zeta potential in mV. Temp EC TDS Salinity
pH
(oC) (µS/cm) (g/L) (ppt)
Mean 31.21 3748.8 2.19 1.87 7.73
RESULTS AND DISCUSSION
Std dev 0.99 1024.3 0.62 0.55 0.36
Groundwater analysis
Minimum 29.56 1991.1 1.14 0.94 7.18
The groundwater samples collected every month
Maximum 32.42 5020.0 3.04 2.64 8.29
were analysed for the following parameters such

Ca2+ Mg2+ Na+ K+ CO32- HCO3- Cl- SO42-

(ppm) (ppm) (ppm) (ppm) (ppm) (ppm) (ppm) (ppm)
Mean 36.4 94.0 3693.2 58.6 28.3 302.3 2803.9 300.9
Std dev 9.4 26.9 1380.8 14.9 15.8 68.9 763.9 36.7
Minimum 23.9 63.2 1815.3 30.6 9.2 179.1 1760.4 257.6
Maximum 53.5 152.5 6049.9 75.7 53.0 397.8 4043.2 347.5
702 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

TABLE 2. Colloid concentration for groundwater

samples. 450

400
Colloid Number of
Salinity
Location concentration colloid
(ppt) 350
(mg/L) particles/L
10
BW1 0.11 0.21 1 x 10
300
BW2 0.03 2.40 1 x 1011
BW3 0.14 0.06 3 x 109 250

BW4 0.31 0.23 1 x 10 10

200
BW6 7.86 0.07 4 x 109
BW7 4.44 0.05 3 x 109 150

BW8 5.75 0.09 5 x 10 9

0.08 100
PBW1 3.35 4 x 109
PBW2 0.26 1.50 8 x 1010
50
PBW4 0.44 1.10 6 x 1010
PBW6 0.28 1.41 7 x 1010 50 100 150 200 250

PBW8 0.04 5.79 3 x 1011

Fig 2. Colloid concentration contour map

Colloid concentration contours are shown in Fig 2.

As can be seen from the figure, the colloid
concentration was higher in the central part of the
site. This colloid level contour map provides a
broad guideline regarding location of waste
disposal trenches. As can be seen from Fig 2, the
higher concentration of colloids in the central part
of the study area has a potential to mobilize the
radionuclide if adsorbed onto these colloid
particles. They can migrate towards west and east
direction due to advective-dispersive flow
(UmaMaheswaran et al. 2008). In view of the
above, location of disposal facility demands a
careful consideration.
Effect of Ionic Strength on Colloid Stability
The chemistry of the groundwater plays a very
crucial role in determining the stability of
colloidal particles. The ability of the colloidal
particles to remain suspended in the medium
depends on the interaction between particles when
they approach each other. Colloids acquire
stability by the formation of an electric double
layer of sufficient strength to prevent
agglomeration (Verwey & Overbeek, 1948). The
counter ions present in the medium can cause the
coagulation of colloidal particles by destroying
the double layer thickness and the valency of
counter ions determines the flocculation value.
Colloidal particles can approach each other more
closely at high TDS than at low TDS, allowing
van der Waals attraction to be more effective. It
 Rani, Sasidhar and Kannan 703

was observed that borewells with high ionic

strength showed low colloid concentration, on the 10
9
other hand borewells with low ionic strength 8
showed high colloid concentration (Fig-3). The 7
6
PBW4
borewells BW6, BW7, BW8 and PBW1 were 5
PBW8
having low turbidity levels (0.05 to 0.09) and 4
3
high ionic strength. Higher turbidity levels were 2

observed in wells BW2, PBW2, PBW4, PBW6 1

0
and PBW8 (1.1 to 5.79) with low ionic strength 0.05 0.08 0.13 0.2 0.31 0.49 0.78 1.24

levels. The remaining three borewells (BW1,

BW3 and BW4) showed low turbidity levels and
could not be correlated with ionic strength data.
This phenomenon is attributed to the higher pH
values encountered at these locations. At higher
pH, adsorption of ions or macromolecules can
take place that alter the mineral surface charge
and reduces stability.
Fig 4. Particle size distribution of colloids from
the borewells
(For colour figure, refer to CD)
Zeta potential
The zeta potential measurements showed that the
groundwater colloids were negatively charged at
groundwater pH due to the presence of clays. The
zeta potential values were varying between –25.5
mV and –34.0 mV. The zeta potentials of
groundwater colloidal particles at six borewell
locations are given in Table-3.

TABLE 3. Zeta potentials of groundwater

colloids

Location Zeta potential (mV)

BW2 -20.2
Fig. 3. Variation of turbidity with (Na+K) and (Ca+Mg) BW4
concentration (mol L-1 × 103) in the study area
-23.9
PBW2 -34.0
PBW4 -29.5
PBW6 -24.4
Particle size and size distribution PBW8 -25.5

Colloids from selected borewells were analysed SUMMARY AND CONCLUSIONS

for particle size and particle size distribution. All
the samples had monomodal size distribution. For
a typical groundwater sample (PBW4) the
average colloid size was 267.2 nm and for a
typical borewell (PBW8) the average colloid size
was 235.9 nm. The particle size distribution of
colloids from the borewells is given in Fig 4.
The stability of colloids is found to be controlled
by salinity. Colloids are more abundant in wells
with low salinity. It is observed that the colloid
concentration reduces as the groundwater
approaches freshwater-saline water interface.
Understandably, there is a strong correlation
between colloid abundance and ionic strength of
704 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

groundwater. The wells in which the ionic and characterisation. Appl. Geochem.
strength is high tend to have low colloid 11,677-695.
concentration compared to wells with low ionic
Kersting A.B. Efurd D. W. Finnegan D. L.
strength.
Rokop D. J Smith D. K Thompson J. L (1999)
Colloids were present in all the groundwater Migration of plutonium in ground water at
samples but are relatively higher concentration in the Nevada Test Site. Nature 397, 56– 59.
the center part of the study area and the
Mc Carthy J.F. Zachara J.M. (1989) Subsurface
concentration was varying between 0.05mg/L and
transport of contaminants, Environ. Sci.
6mg/L and the number of colloidal particles were
Technol. 23, 496-502.
varying between 3 х 109 particles/L to 3 х 1011
particles/L. The size distribution of colloids is Minhan Dai, Ken.O.Buesseler. Steven.M.Pike.
monomodal consisting of particles less than 350 (2005) Plutonium in groundwater at the
nm. Zeta potential measurements showed that the 100K-Area of the U.S.DOE Hanford Site. J.
colloids are stable at groundwater pH. The colloid Contam. Hydrol.76, 167-189.
concentration contour map and the dependence of
Penrose W.R. Polzer W.L. Essington E.H. Nelson
colloid concentration on salinity have
demonstrated their utility in locating the disposal D.M. Orlandini K.A. (1990) Mobility of
plutonium and americium through a shallow
facility in the study area and in assessing the
aquifer in a semiarid region. Environ. Sci.
hazard potential due to colloid facilitated
Technol. 24, 228– 234.
radionuclide transport. These findings assist
decision makers while selecting the location of Randall.J.Schaetzl and Sharon Anderson. (2005)
engineered disposal facilities to choose between Soils-Genesis and Geomorphology.
the relative hazard potential of colloid facilitated Cambridge University Press. p 18
transport and proximity of disposal facilities to
the water bodies. Ryan J.N. and Elimelech M. (1996) Colloid
mobilization and transport in groundwater.
Colloids Surf. A: Physicochem.Eng.Aspects
107, 1-56.
REFERENCES
Singh, V.S. Saxena V.K. Jain S.C. Anjaneyulu
APHA: (1976) Standard methods for the
G.R. Prakash B.A. Mondal N.C. (2003)
examination of water and wastewater (14th
Hydrogeological and geophysical
edition). American Public Health Association,
investigations at PFBR site, Kalpakkam,
Washington, DC.
Tamilnadu.Tech.Rept.No.NGRI-2003-GW-
Degueldre C. Baeyens B. Goerlich W. Riga J. 396.
Verbist J. Stadelmann P. (1989) Colloids in
water from a subsurface fracture in granitic Uma Maheswaran T. Sasidhar P. Kannan S.E.
(2008) Modeling of unconfined aquifer at
rock, Grimsel Test Site, Switzerland.
Kalpakkam plant site. Proc. 16th Nat. Symp.
Geochim. Cosmochim.Acta 53, 603-610.
Environment (NSE 16), 70-74.
Degueldre C. Grauer R. Laube A. Oess A Silby H.
Verwey E.J.W. and Overbeek J.T.G. (1948)
(1996) Colloid properties in granitic
Theory of stability of lyophobic colloids.
groundwater systems. II: Stability and
Elsevier, Amsterdam.
transport study. Appl. Geochem. 11,697-
710. Tan K.H. (2000) Environmental Soil Science.
Marcel Dekker, Inc.
Degueldre C. Pfeiffer H.R. Alexander W. Wernli
B. Bruetsch R. (1996) Colloid properties in
granitic groundwater systems. I: Sampling
Selected Characteristics of Clay Soils Polluted by Petroleum Substances
in the Context of Their Barrier Properties
D. Izdebska-Mucha
Faculty of Geology, University of Warsaw, Warsaw, Poland (dim@uw.edu.pl)
E. Korzeniowska-Rejmer
Faculty of Engineering of Environment, Technical University of Cracow, Cracow, Poland
(elakorz@poczta.onet.pl)

The paper presents effects of petroleum contamination on selected physical properties of clay soils. The
soils tested represented various types: silts, clays, sandy clays. The tests were carried out on soils polluted
in situ in industrial areas, petrol stations and areas of fuel storages and on soils polluted in laboratory. The
analysed properties included: grain size distribution, plasticity, linear shrinkage, volumetric shrinkage,
compaction characteristics. Petroleum pollution of clay soils caused significant changes in grain size
distribution, reduction of plasticity, decrease of shrinkage and compaction parameters. Knowledge of these
effects is important in the evaluation of barrier properties of soils.

INTRODUCTION soils subjected to a long-term pollution proved

Some engineering constructions such as landfills,
sewage treatment plants, storage and distribution
facilities of liquid fuels are a potential source of
soil and groundwater pollution. They should be
located on a subsoil which in a long time range
would effectively protect the ground against
pollutant migration.
Character and spatial range of the pollution
depend on pollutant properties, hydrogeological
conditions, soil type and its physical properties.
An interaction between chemical substances and
soil particles depend on the substance type (acid,
base, polar or non-polar compound) and soil
microstructure (fabric, forces, composition).
Deficiency of proper protection and unfavorable
hydrogeological conditions can cause pollutant
migration into the soil ground, that results in a
degradation of their natural state (changes of
physicochemical and mechanical characteristics).
In case of mineral sealing barriers changes of soil
properties caused by a long-term pollutant
infiltration may result in a significant reduction of
geotechnical parameters responsible for soil
permeability and deformation properties. Changes
in physical and mechanical properties of cohesive
their limited suitability for construction of sealing
barriers designed for long time range petroleum
pollution (Korzeniowska-Rejmer, 2007).
This paper discusses the results of tests carried out
on selected physical properties of soils exposed to
petroleum substances. The parameters presented
should be taken under consideration along with
the criteria described in legal regulations while
evaluating suitability of soils for mineral sealing
barriers. Those are: particle size distribution,
compaction, Atterberg limits, shrinkage index and
linear shrinkage.
PARTICLE SIZE DISTRIBUTION
Conducted analyses of silty and clayey soils,
sampled from subsoils of fuel storage facilities
and oil-refinement waste tanks, showed
significant changes in their particle size
distribution in comparison to natural (clean)
reference samples (Tab. 1).
The soils tested were: 1) contaminated in-situ with
petroleum substances in the time range of a few
decades: loess and glacial till “1”; 2) model soils
contaminated in laboratory with diesel oil for the
time period of 12 months: glacial tills “2” and
706 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
TABLE 1 Changes in Particle Size Distribution of Cohesive Soils Polluted with Petroleum Substances
Loess Glacial till ”1”
1 2
clean P clean P clean
Sand [%] 27 10 2 2
Silt [%] 64 87 47 67
Clay [%] 9 3 51 31
Soil type4 silt silt clay silty clay
P – soil polluted with petroleum substances
1
hydrocarbon content 25%
2
hydrocarbon content 600 mg/kg
3
diesel oil content 10%
4
acc. to BS1377, after Head (1992)
“3”. The hydrometer method was applied
according to BS1377: Part 2: 9.5.
The results, although they concern soils that differ
in terms of genesis, lithology, occurrence and
pollution conditions, showed similar tendencies of
particle size distribution changes caused by
petroleum pollution. It was revealed that in all
polluted samples the content of sand and clay
decreased and the content of silt increased. In
some cases (glacial till “1” and “3”) the change
was enough to modify the soil type (Tab. 1). The
range of alterations observed depends on the soil
type. The highest drop of clay content occurred in
clay soil (glacial till “1”), while the sand content
decreased at most in silty soil (loess) - by 17%,
probably as a result of disintegration of natural
sandy aggregates.
Observed changes may result from the aggressive
influence of chemical substances - admixtures
occurring in petroleum products (sulphur, acids,
nitrogen compounds, heavy metals, and other
additives improving their quality) that fill pores -
on the soil grains and particles. They may cause a
destruction of structural bonds and minerals that is
similar to weathering process. Clays are most
sensitive to chemical aggressiveness, therefore in
cohesive soils some major changes in physical,
chemical and mechanical properties may occur.
The redistribution of particle size in polluted soil
is also related to physicochemical properties of
petroleum substances. Hydrocarbons are non-
polar compounds of a low dielectric constant.
They cause reduction of thickness of electric
double layer and interparticle forces (Kaya &
Glacial till ”2” Glacial till ”3”
3
P clean P3
7 4 57 54
78 84 28 35
15 12 14 11
clayey silt clayey silt clayey sand clayey & sandy silt
Fang 2000, 2005). Reduction of repulsive forces
is higher than of attractive ones. As a result, clay
particles tend to flocculate and form bigger (silty)
assemblages, which explains increase of silt
content in polluted samples.
A very important problem in polluted soils testing
is choosing appropriate methods that would the
most efficiently reveal processes and changes that
take place in a soil due to pollution. Suspension of
glacial till “1” was prepared for hydrometer
measurements without adding dispersing agent.
As hydrocarbons influence the force system and
aggregation processes in a clay, such procedure
seems to reflect the most accurately the real
aggregation of solids in a soil. Glacial till “1”
exhibited the highest changes in clay content,
which might be partly resulting from
methodology applied.
Particle size distribution is one of the basic
criteria to evaluate the soil suitability for
construction of mineral sealing barriers. Clay
content determines its sorption capacity,
permeability and sealing characteristic. A
decrease of clay content in soil may significantly
enhance its permeability and reduce its barrier
properties.
COMPACTION
The studies on the influence of the petroleum
waste on the physico-mechanical properties of
soils at several oil refineries in southern Poland
revealed significant changes in compaction
characteristic of soils used in construction of
settling ponds embankments and their subsoil.
Long-term saturation of soils with petroleum
waste caused reduction of bulk density and dry
density of about 10 to 15% at a corresponding
pollutant concentration of 20 to 25%. As a result
of long-term pollutant infiltration through the
 Mucha and Rejmer 707

TABLE 2 Influence of Petroleum Pollution on Compaction of Sandy Till

Relative compaction
Relative compaction* Change of
of soil polluted with diesel oil
of non-polluted soil relative
(10%)
Test no. compaction
Number Relative Average Number Relative Average
[%]
of blows compaction value of blows compaction value
I 10 0.948 10 0.927
II 10 0.947 0.947 10 0.930 0.928 2.12
III 10 0.946 10 0.926
I 15 0.955 15 0.940
II 15 0.953 0.954 15 0.944 0.942 1.26
III 15 0.954 15 0.943
I 25 0.985 25 0.963
II 25 0.981 0.983 25 0.959 0.962 2.14
III 25 0.981 25 0.964

embankments and their subsoil, failure situations PLASTICITY

were reported in some storage facilities for fluid
wastes. They have been caused by the loss of
embankments stability as well as significant and
non-uniform settlement of plasticized subsoil
(Pađka & Korzeniowska 1990).
Table 2 shows the results of the model research on
the compaction of sandy till subjected to diesel oil
contamination (10%) over 1 year. Proctor
compaction tests were conducted using method I
according to polish standard PN-88/B-04481
(mould: 1±0.003 cm3, hammer weight: 2.5±0.005
kg, height of drop: 32.0 cm, compactive effort:
0.6 Jcm-3, number of layers: 3).
As a result of the pollution of tested soils a
relative compaction decreased by about 2%. A
decrease of compaction properties of soils
polluted with organic compounds was also
reported by e. g. Khamehchiyan et al. 2007, Singh
at al. 2009. In natural conditions, with long-term
contamination involved, changes in compaction of
soils used as a construction material for sealing
embankments at waste storage facilities or fluid
waste tanks, may result in a decrease of their
long-term barrier properties. In case of
embankments of open fluid waste storage tanks,
loss of overall stability may occur.
Substantial changes due to petroleum pollution
were observed in consistency limits of soils. In a
two-phase system: cohesive soil – petroleum-
derived liquid, loss of cohesion and soil plasticity
was observed as well as a decrease of liquid limit
value. Table 3 presents results of plastic and
liquid limit measurements of miopliocene clay,
glacial till, bentonite, kaolinite and illitic clay.
The soils were tested with distilled water, petrol
and diesel oil. The tests were carried out
according to BS1377: Part 2: 1990. All analysed
soils after mixing with petrol and diesel oil
showed entirely different behaviour in comparison
with distilled water. Soils saturated with
petroleum substances lost cohesion, which made it
impossible to define plastic limit. All analysed
soils polluted with petrol and diesel oil were
defined as non-plastic (NP). Liquid limit values of
soils saturated with petroleum-derived liquids
were lower than for distilled water.
The highest decrease of liquid limit was observed
in most hydrophilic soils, such as bentonite, illitic
clay, miopliocene clay, the lowest in kaolinite and
glacial till. Identified changes in consistency
limits cause reduction of plasticity index.
Cohesive soils polluted by petroleum substances,
gained features of fine non-cohesive soils, which
raises the question of their usability as insulating
barriers.
708 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

TABLE 3 Plastic Limit and Liquid Limit of Cohesive Soils Contaminated by Petroleum Substances
Plastic limit [%] Liquid limit [%]
Soil
Water Petrol Diesel oil Water Petrol Diesel oil
Miopliocene clay 25 NP NP 65 36 34
Glacial till 13 NP NP 24 24 23
Bentonite 55 NP NP 257 34 37
Illitic clay 26 NP NP 46 24 26
Kaolinite 40 NP NP 51 45 37
NP – non-plastic soil
*
relative compaction = achieved dry density/max dry
capacity of the soil when changing from solid to
density
liquid consistency, as well as lower plasticity of
the soil.
SHRINKAGE
The analysis of linear shrinkage in function of
The analysis of shrinkage properties of soils
pollutant concentration (Fig. 3) revealed that the
polluted with petroleum substances was carried
higher the pollutant content the lower linear
out for two soils: glacial till and miopliocene clay,
shrinkage. Previous studies (Izdebska-Mucha
that were classified as sandy clay with silt and
2005) showed that soils saturated with petroleum
clay (of high clay content 85%) respectively. The
liquids only did not exibit volumetric changes or
soils were polluted in laboratory with diesel oil.
just minor ones, as for example illustrated in the
Several soil specimens containing 0, 4, 8, 12, 16,
photo in Fig. 4.
20, 24, 30% of diesel oil by dry weight of the soil
were prepared and stored over 3–5 months before
shrinkage limit [%]

testing.
20
Shrinkage limit and linear shrinkage were
determined according to BS1377: Part 2: 1990.
10
Before shrinkage limit measurements the polluted
soil specimens were mixed with distilled water to
0
a stiff plastic consistency. The initial moisture 0 4 8 12 16
content of soil pastes prepared for linear shrinkage diesel oil [%]
measurements was of about the liquid limit. miopliocene clay glacial till
Shrinkage index was calculated as a difference
Fig. 1 Shrinkage Limit of Miopliocene Clay and
between liquid and shrinkage limit (Ranganathana
Glacial Till vs. Diesel Oil Pollution.
& Satyanarayana 1965). The tests results are
presented in Figs. 1-3.
For glacial till containing 8, 16% of pollutant, it 80
shrinkage index [%]

was impossible to determine shrinkage limit due

60
to the loss of soil cohesion – samples crushed and
disintegrated. It is worth pointing out that in the 40
clay this effect did not occurred at the analysed 20
range of pollution.
The results obtained showed that shrinkage limit 0
0 4 8 12 16
values initially decreased for the lowest pollutant diesel oil [%]
concentration and then increased for higher miopliocene clay glacial till
concentrations (Fig. 1). Increase of shrinkage
limit values determines reduction of shrinkage Fig. 2 Shrinkage Index of Miopliocene Clay and
index (Fig. 2). This indicates lower sorption Glacial Till vs. Diesel Oil Pollution.

linear shrinkage [%]
R2= 0.9326
21
17
13
9 worsen significantly deformation, sorption and as
5 a result barrier properties of the soil.
0 5 10 15
diesel oil [%]
miopliocene clay
Fig. 3 Linear shrinkage of Miopliocene Clay and
Glacial Till vs. Diesel Oil Pollution.
A
B substances that are released from the stored waste
C
Fig. 4 Samples of miopliocene clay saturated with: A)
distilled water, B) petrol and C) diesel oil after linear
shrinkage test. (For colour figure, refer to CD)
Shrinkage parameters of cohesive soils, such as
shrinkage index and linear shrinkage, can be
applied in evaluating usability of the soil as
barrier material. According to some authors
(Gawriuczenkow 2001) shrinkage index is a
parameter which better reflects plastic properties
of soil than commonly used in criteria plasticity
index.
CONCLUSIONS
The results presented in this paper showed that
petroleum substances significantly change
physical properties of clay soils. In polluted soils,
sand and clay content decreased in favor of silt
content. Change in clay content affects soil
permeability which is one of the basic criteria in
Mucha and Rejmer 709
evaluating barrier properties of soils. According
to polish and european regulations coefficient of
permeability should be k ≤ 1x10-9 ms-1.
Petroleum substances eliminated or reduced
cohesion, plasticity, liquid limit, linear shrinkage,
R2= 0.878 shrinkage index of clay soils. Such changes
20 25 30 35 Relative compaction of soil polluted with
petroleum substances decreased in comparison to
glacial till unpolluted soil. In a soil subjected to a long-term
pollution with petroleum substances may reduce
its stability, bearing capacity and sealing
effectiveness.
Since many possibilities of the interactions
between soils and chemical substances should be
taken into consideration, designing of multilayer
mineral seals is recommended, as they provide an
effective protection of the subsoil against the
pollutants. Consideration should be also given to
the changes in composition and amount of
material, the influence of microorganisms, and
time-related reactions.
Petroleum pollution of the soils used in mineral
barrier results in the fact, that some criteria, which
are important at the evaluation of the soil long-
term sealing properties, may be no longer
satisfied.
REFERENCES
BS (1990) British Standard Institution, London.
1377: Part 2.
Gawriuczenkow I. (2001) Evaluation of Poznań
silts as potential, insulating geological
barriers. Doctoral dissertation, Faculty of
Geology, Warsaw.
Head K.H. (1992) Manual of soil laboratory
testing. Vol. 1: Soil classification and
compaction tests. London.
Izdebska-Mucha D. (2005) Influence of oil
pollution on geological-engineering
properties of clay soils. Przegląd
Geologiczny, 53, 766-769. (In polish with
english abstract).
Kaya A. and Fang H.-Y. (2000) The effects of
organic fluids on physicochemical
710 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
parameters of fine-grained soils, Canadian
Geotechnical Journal, 37,943-950.
Kaya A. and Fang, H.Y. (2005) Experimental
evidence of reduction in attractive and
repulsive forces between clay particles
permeated with organic liquids. Canadian
Geotechnical Journal, 42, 632–640.
Khamehchiyan M., Charkhabi A.H. and Tajik M.
(2007) Effects of crude oil contamination
on geotechnical properties of clayey and
sandy soils. Engineering Geology. 89 (3-4),
220-229.
Korzeniowska-Rejmer E. (2007) The effects of
persistent anthropogenic on the
geotechnical properties of soil. In: Sarsby
R.W., Felton A.J Geotechnical and
Environmental Aspects of Waste Disposal
Sites, Proc. GREEN 4, 157-162.
Pađka J. and Korzeniowska E. (1990) Influence of
post-refining wastes on physico-
mechanical properties of embankment
soils and drainage corrosion in Oil Refinery
in Trzebinia. Cracow University of
Technology. Institute of Geotechnics
Archives, Cracow.
PN (1988) Polish Standard. Polski Komitet
Normalizacji Miar i Jakości, Warszawa.
PN-88/B-04481.
Ranganathan B.V. and Satyanarayana B. (1965)
A rational method of predicting swelling
potential for compacted expansive clays.
Proc. of the Sixth Intern. Conf. Soil
Mechanics and Foundation Engineering, 1,
92-96.
Singh S.K., Srivastava R.K. and John S. (2009)
Studies on soil contamination due to used
motor oil and its remediation. Canadian
Geotechnical Journal, 46, 1077–1083.
 The Study of Permeability of Incompressible Fluids in
Infiltrating into the Unsaturated Compacted Soil

F. Soltani
Kerman University of Technology, Iran (soltani.fazlollah@gmail.com)
I. Naeimifar
Isfahan University of Technology, Iran (I.Naeemifar@cv.iut.ac.ir)

ABSTRACT In comparison with saturated soils, lesser researches have done about fluids flow in
unsaturated soils. In this paper, it is attempted to examine the permeability of water in infiltrating into the
unsaturated soils in laboratory. The results of permeability tests on the above specimens show that if the fluid
that infiltrates into the soil (while infiltrating) have the steady characteristics (viscosity, temperature etc) the
coefficient of permeability is a function of moisture and soil suction matrix. Regarding obtained results from
the semi-empirical tests, it was suggested semi-empirical equations for determining the coefficient of
permeability in the certain range of moisture.

Introduction into the unsaturated soil are different from

Researches in unsaturated soil mechanics
considerably developed in the past decades. In an
unsaturated soil other than solid grains of soil
(stone minerals) there are two phases of
compressible fluid (usually air) and
incompressible fluid (usually water). The
available air in the soil regarding to the degree of
saturation can be diffused in soil in the forms of
continuous space or independent air-bubbles.
Naturally, with increase of degree of saturation,
the continuity of air phase decrease and the
available air in the soil diffuse into the soil in the
form of independent air-bubbles. On the contrary,
with decrease of degree of saturation, the
continuity of air phase increase and the available
air in the soil shapes in the form of continuous
space. Soil compaction is another effective factor
that can independently make this change on the
continuity of air phase of soil. In the compaction
process, air is the first element that exists from
soil. With the air movement from the soil, the
volume of available air in the soil and the
continuity of air phase decrease, and consequently
the more the compaction energy is, as much is the
decrease of this continuity. Therefore, it is obvious
that dominant laws on flow of water infiltrating
saturated soils. In 1928, Richards developed the
water potential laws for liquid water flow in
unsaturated soils [10]. Gardner (1956) and Corey
(1957) suggested a transient and a steady state
method of measuring the water permeability
coefficient of the degree of saturation. They
presented some techniques to determinate suction
matrix. Most of their methods are presently
commonly used [2, 9]. According to researches,
several phenomena such as moisture gradient,
suction matrix gradient and hydraulic gradient can
be the factors of the incompressible fluid (water)
movement in unsaturated soil that dominant law
on the flow of fluid is different for every
phenomenon. In relation to the moisture gradient
factor, frequently the water phase movement is
from the point that more moisture toward the area
with less moisture. Nevertheless, this rule does not
hold for all conditions, for example for soils with
different layers the water movement can done
from the areas with little water content toward the
areas with much water content. Therefore, the
moisture gradient cannot be considered as the
main excitation of movement from the inside of
unsaturated soil [5, 6].
712 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
According to researches of Fredlund and Rahardjo
(1993) and Daniel (1983), the suction matrix
gradient can also play as excitation factor for
water movement from the inside of unsaturated
water. However, it is observed that in the direction
of water flow movement, the suction matrix
gradient can be positive, negative, or zero.
Therefore, with regarding to these observations
can obtain this result that neither suction matrix
gradient can be counted as a main excitation of
movement in relation to the water movement from
the inside of the unsaturated soil. In particular
state when the air pressure gradient is zero the
suction matrix gradient in scalar mode becomes
equal to the water pressure gradient, which this
fact often occurs in nature. [3, 4, 5].
Concerning third phenomenon namely hydraulic
gradient that consists of fluid pressure gradient
and height gradient can consider this factor as
principal excitation factor of water movement
from the inside of unsaturated water [1, 7, 8].
Test Theory
As we know that, the suction matrix in unsaturated
soils is parallel to the pore air pressure different
with pore water pressure, which, each of these
pressures changes the degree of soil saturation
sensibly. In unsaturated soils, the amount of water
in soil increases with the increase of soil moisture
and consequently the water pore pressure
increases that results in the decrease of soil
suction matrix. On the other hand, the increase of
soil moisture (or increase of the degree of
saturation) culminates in the decrease of soil
suction matrix.
As it is mentioned before, regarding moisture of
the specimen positioned in the cell, it is possible
to determine the negative pore water (Uw) of
sample using tensimeters. Moreover we can
operate a certain and fixed amount of pore air
pressure (Ua) on sample using air capsule.
Regarding the certain amounts of above-
mentioned two pressures, we can calculate the
relevant suction matrix (Ua-Uw).
According to the above certain amount of suction
matrix, the coefficient of permeability of specimen
in continuous flow status using Darcy's
generalized law is calculated through following
equation:
 − Q  d t 

kw =   (1)
 A∆t  ht1 − ht 2 
It is necessary to mention that amounts of ht1 and
ht2 are the functions of soil suction matrix. In the
other word, with the change of air pore pressure,
the above amounts change. Therefore, it can be
concluded that in unsaturated soils, the coefficient
of permeability is a function of pore air pressure
of soil or is a function of soil suction matrix more
generally. In order to determine the air pore
pressure the manometer was connected to the
specimen. In this way, the amount of pore air
pressure operated on the specimen can be
measured as follows:
Ua=ρmghm (2)
In order to determine the water pore pressure, two
tensimeters connected to the specimen were used.
In this state, the water height in the tubes of
tensimeters indicates the water pressure head.
Average value of water pore pressure of sample
can be determined according to the following
equation. It is necessary to mention that in
unsaturated soils the water pore pressure is
negative.
 h p1 + h p 2 
(u w )ave =   ρ w g (3)
 2 
Regarding to the amount of pore air pressure and
average value of water pore pressure, we can
determine the average of soil suction matrix from
the subtraction of these two pressures as follows:
 h p1 + h p 2 
(u a − u w )ave = ρ m ghm −   ρ w g (4)
 2 
In order to facilitate the explanation, average
value of pore water pressure and the average value
of soil suction matrix are shown u w and
( u a − u w ) respectively. As was indicated before,
the coefficient of permeability of water is

determined according to the average values of soil
suction matrix using Darcy's generalized law.
Preparing specimens for tests
In performed researches, two different sorts of
fine-grained soils were examined which,
concerning the primary moisture of soils, the
necessary amount of water added to the specimens
was measured in order to reach the expected
moisture and then this volume of water was added
to the samples in several phases. After kneading
samples, a homogeneous soil mixture (with
expected moisture) was provided. This mixture
was rammed in compaction mold using normal
Proctor Method and then was put in the unit cell.
The specifications of selected soils in this research
after performing primary tests are shown briefly in
table (1).
TABLE (1): specifications of primary samples of
unsaturated soil
Soil sort Soil (A) Soil (B)
(%)primary moisture 3.5 2.6
(%)liquid limit 43.8 32.4
Plastic limit 18.2 15
Plasticity limit 25.6 17.4
maximum bulk density 1640 1739
(%)optimum moisture content 16.9 14.3
Results of tests
In order to analyze the obtained results, Figures 1
through 7 are drawn. Fig. 1 indicates the changes
of average soil suction matrix according to the
pore air pressure. As it is perceived from these
diagrams while the moisture of soil is constant, the
soil suction matrix approximately increases in a
linear function with increasing of pore air pressure.
Parallel to stable pore air pressure, with increase
of moisture the negative value of pore water
increases that result in the increase of the suction
matrix of soil.
For two selected samples Fig. 2 indicates the
changes of average soil suction matrix according
to the different values of pore air pressure based
on the test expired time. According to these
diagrams, it is obvious that with operating stable
pore air pressure on sample, the average suction
matrix of soil (or average suction of pore water)
increases gradually while it reaches to balance. In
Soltani and Naeimifar 713
addition, the above diagrams suggest that in little
pore air pressure, time to reach balance is less than
other ones. Moreover, regarding to curves slopes
can be concluded that the changes of average
suction matrix of soil in primary times are more
than other ones.
For two sorts of soils under study (A and B), the
semi-logarithmic diagrams (3) through (6) indicate
the changes of coefficient of permeability
according to soil suction matrix. Every curve in
these diagrams is relevant to a compacted soil
sample with primary moisture (W). According to
these diagrams with increase of soil suction matrix,
the coefficient of soil permeability decreases that,
this behaviour can considered as a general law.
The cause of this behaviour can be explained in
this way that the increase of soil suction matrix
means the increase of suction of sample pore
water which, this pressure increase also means the
air departure from specimen that finally results in
increase of degree of specimen saturation. It is a
crucial factor in faster transfer of water from the
soil.
The above diagrams indicate that in moistures
lower than 10% according to the equal values of
suction matrix; the soil (A) has the less
permeability than soil (B). However, with increase
of sample moisture over than 10%, this difference
practically becomes insignificant. This result
suggests that with increase of moisture, the
behaviour of soils under study in permeability
become similar to the way that the semi-empirical
relation between the changes of soil suction matrix
and coefficient of permeability with high
coefficient of correlation can be obtained.
In the continuous of discussion and reviewing
results, the diagram (7) have been drawn in order
to indicate the changes of coefficient of soil
permeability according to soil suction matrix with
different moistures. In accordance with these
diagrams, with increase of this moisture of
specimen with the similar suction matrix, the
permeability decreases. The cause of this
phenomenon can be pertinent to the compaction
effect on soil permeability. In the other word, due
to increase of moisture, the soil compaction is
performed well and the coefficient of soil porosity
decreases.
714 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

SOIL A 1.00E-05
150

k (m /sec)
1.00E-06
ua-uw (kPa)

100
1.00E-07
50

0 1.00E-08
0 10 20 30 40 50 60 30 50 70 90 110 130
ua (kPa) ua-uw (kPa)
w1 = 7.34% w2 = 9.53% w3 = 11.64% w4 = 13.87%
SOIL A ; w = 7.34% SOIL B ; w = 7.12%

SOIL B
1.00E-05
-0.0277(ua-uw )
150 k = 2E-06e
ua-uw (kPa)

1.00E-06 2
R = 0.499

k (m/sec)
100
1.00E-07
50
1.00E-08
0
30 50 70 90 110 130
0 10 20 30 40 50 60 ua-uw (kPa)
ua (kPa)
SOIL A, w = 7.34% ; SOIL B, w = 7.12%
w1 = 7.12% w2 = 9.81% w3 = 11.79% w4 = 13.16%

Fig. 1 Changes of Average Soil Suction Matrix Fig. 3 Changes of Coefficient of permeability
According to Air Pore Pressure According to the Average Soil Suction Matrix

SOIL A ; w = 13.87% 1.00E-06

140
k (m/sec)

120 1.00E-07
ua - uw (kPa)

100
80
60
1.00E-08
40
20
1.00E-09
0
0 100 200 300 400 500 600 700 800 900 1000 1100 40 60 80 100 120 140
ua-uw (kPa)
t (min)

SOIL A ; w = 9.53% SOIL B ; w = 9.81%

ua=10 kPa ua=20 kPa ua=30 kPa ua=40 kPa ua=50 kPa

SOIL B ; w =14.7 % 1.00E-05

-0.0356(ua-uw )
1.00E-06 k = 6E-07e
k (m/sec)

140
2
120 R = 0.896
1.00E-07
ua-uw (kPa)

100
80
1.00E-08
60
40
1.00E-09
20
0 40 60 80 100 120 140
0 200 400 600 800 1000 1200
ua-uw (kPa)
t (m in) SOIL A, w = 9.53% ; SOIL B, w = 9.81%
ua=10 kPa ua=20 kPa ua=30 kPa ua=40 kPa ua=50 kPa

Fig. 2 The Changes of Average Soil Suction Matrix Fig. 4 Changes of Coefficient of Permeability
According to Expired Time to Experiment According to Average Soil Suction Matrix
 Soltani and Naeimifar 715

1.00E-07 SOIL A
1.00E-06
1.00E-08
k (m /s ec )

1.00E-07

k (m /s e c )
1.00E-08
1.00E-09
1.00E-09
1.00E-10 1.00E-10
40 60 80 100 120 140
ua-uw (kPa) 20 40 60 80 100 120 140
ua-uw (kPa)
SOIL A ; w = 11.64% SOIL B ; w = 11.79%
w1 = 7.34% w2 = 9.53% w3 = 11.64% w4 = 13.87%

1.00E-07
k = 1E-07e
-0.0396(ua-uw ) SOIL B
2 1.00E-05
k (m/sec)

1.00E-08 R = 0.9781
1.00E-06

k ( m /s e c )
1.00E-09
1.00E-07
1.00E-10 1.00E-08
30 50 70 90 110 130 150
ua-uw (kPa) 1.00E-09
SOIL A, W = 11.64% ; SOIL B, w = 11.79%
1.00E-10
20 40 60 80 100 120 140
Fig. 5 Changes of Coefficient of permeability
According to Average Soil Suction Matrix ua-uw (kPa)
w1 = 7.12% w2 = 9.81% w3 = 11.79% w4 = 13.16%
1.00E-07
k (m /sec )

1.00E-08 Fig. 7 Changes of Coefficient of Permeability

According to Average Soil Suction Matrix
1.00E-09
Conclusion
1.00E-10 Regarding to this research, it can be concluded
40 60 80 100 120 140 that the behaviour of unsaturated soils has a basic
ua-uw (kPa)
and significant difference with the behaviour of
SOIL A ; w = 13.87% SOIL B ; w = 13.16% saturated soils against infiltration of
incompressible fluids. This difference derives
1.00E-07 from this issue that the air pressure in the
k = 2E-07e
-0.0488(ua-uw )
unsaturated soil has a crucial role in the
k (m/sec)

1.00E-08 2
R = 0.9874 mechanism of fluid flow in the way that with
1.00E-09
increase of this pressure, the soil strength against
fluid infiltration increases. Whereas, the increase
1.00E-10 of air pressure in soil (pore air pressure) results in
30 60 90 120 150 the decrease of existing water pressure in the soil
ua-uw (kPa)
(water pore pressure). Therefore, it is predictable
SOIL A, w=13.87% ; SOIL B, w=13.16%
that if the water pore pressure changes in
unsaturated soil, this change has direct effect on
Fig. 6 Changes of Coefficient of Permeability According the soil permeability. In saturated soils because of
to Average Soil Suction Matrix the lack of air, the water flow can infiltrate into
716 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
the soil easily and without trouble. Consequently,
in similar conditions, the permeability of saturated
soil against liquid infiltration is more than the
permeability of unsaturated soil.
In addition, the result of this research indicates the
effects of compaction on the permeability of
unsaturated soils. In this way, it should not always
be expected that the increase of moisture (or the
increase of the degree of saturation) cause the
increase of soil permeability. On the contrary,
sometimes in spite of increase of moisture or
increase of degree of saturation it is perceived that
permeability decreases which, compaction effects
on soil structure can be the cause of this
phenomenon. This fact is shown in Fig. 7 clearly.
References
1. Barden L., and Pavlakis, G., (1971), Air and
Water Permeability of Compacted Unsaturated
Cohesive Soil, J. Soil Sci., vol. 22, no. 3, pp. 302-
317.
2. Corey, E. C., (1957), Measurement of Air and
Water Permeability in Unsaturated Soil.
Proceedings of the Soil Science Society America
21, 7-11.
3. Childs E. C., and Collis-George E. C., (1950)
The Permeability of Porous Materials, Porc. Royal
Soc. vol. 201 A, pp. 309-405.
4. Daniel, D. E., (1983), Permeability Test for
Unsaturated Soil, Geotech. Testing J. (ASTM),
vol. 6, no. 2, pp. 81-86.
5. Fredlund, D. G., (1993), Rahardjo, H., Soil
Mechanics for Unsaturated Soils, ( ISBN 0-471-
85008-X, U.S.A)
6. Fredlund, D. G., (1991), How Negative Can
Pore-Water Pressure Get?, Geotechnical News,
vol. 9, no 3, Can. Geot. Society, pp. 44-46.
7. Fredlund, D. G., (1979), Second Canadian
Geotechnical Colloquium: Appropriate Concepts
and Technology for Unsaturated Soils, Can.
Geotech. J., vol. 16, no. 1, pp. 121-139.
8. Fredlund D. G., and Rahardjo, H., (1986),
Unsaturated Soil Consolidation Theory and
Laboratory Experimental Data, Consolidation of
Soils : Testing and Evaluation, ASTM STP 892, R.
N. Yong and F. C. Townsend, Eds., ASTM,
Philadelphia, PA, pp. 154-169.
9. Gardner, R., (1956), Calculation of Capillary
Conductivity from Pressure Plate Outflow Data,
Proc. Soil Science Society of America 20, 317-
320.
10. Richards, L. A., (1928), The Usefulness of
Capillarity Potential to Soil Moisture and Plant
Investigation, Journal of Agricultural Research 37,
719-742.
 Interaction of Vinase and a Sample of Soil

A.A., Cunha
State University of North Fluminense Darcy Ribeiro, Campos, Brazil (cunhabr@hotmail.com)
S. Tibana
State Univetsity of North Fluminense Darcy Ribeiro, Campos, Brazil (tibana@uenf.br)
R. Nascentes
Federal University of Viçosa, Rio Paranaíba , Brazil (rejane.nascentes@ufv..br)
I.D., Azevedo
Federal University of Vicosa, Vicosa, Brazil (iazevedo@ufv.br)
F. T. de Almeida
Federal University of Mato Grosso, Sinop, Brazil (fredterr@gmail.com)

ABSTRACT This paper presents results of an experimental program that has been carried out
by the Civil Engineering Laboratory of UENF in cooperation with UFMT and UFV concerning
of “contaminant” behavior. The main aim of this research is the assessing of interaction between
the vinase, one of the effluents from alcohol industry, and one sample of a typical soil from
Campos dos Goytacazes area (BRAZIL). In the course of this research, it has been observed that
the vinase shows an increase degradation process due to the microbiological action.
Geotechnical tests and physical-chemical analysis were done, and the results are presented
herein.

INTRODUCTION shallow, what promote faster eutrophysation of

In Brazil, many sugar-cane plantations are
irrigated and fertilize with the residue of alcohol
industry, the vinase. It is used due to its
chemical composition that is rich of
micronutrients like potassium (K), phosphorus
(P), nitrogen (N) and organic nutrients.
Therefore, if for one hand it is suitable to soil
improvement by the other one it may cause a
serious environmental problem when the
micronutrients are transported to the water table
and, consequently, to the water bodies like
lagoons and river, leading to a proliferation of
plant life, especially algae. These organisms
may reduce the dissolved oxygen content
causing the extinction of other way of life. The
eutrophysation of water bodies closed to the
ferti-irrigated areas were already observed in the
past. Besides the chemical composition of the
vinase the water table in this area is very
groundwater resources.
The soil in sugar-cane plantation works like a
natural barrier to the transport of vinases´
minerals and organic nutrients. It is well known
that as greater as the content of clay particles
and its superficial specific better barrier the soil
will be. But considering that the vinase have
been use to ferti-irrigate the soil for a long term
it probably can be saturated by nutrients.
So, the main objective of this research program
is the assessment of soil capacity to retain
nutrients from vinase. Recently, two master
degree dissertations were done (Cunha, 2007
and Fonseca, 2007), and some results of it are
presented herein.
718 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

VINASE AND SOIL COMPOSITION (from 3.45 kg/m3 to 3.08 kg/m3) and nitrogen
The vinase is an effluent from sugarcane juice content (from 0.39 to 0.35).
fermentation. It is composed predominantly by Characterization tests were carried out with a
water (93%) and organic and mineral solids sample of soil used in this research. The results
(7%). It has high level of potassium (K), total are show in Table 2 and Table 3. The high
nitrogen (N), calcium (Ca), magnesium (Mg) content of fines (silts and clays), around 53% of
and phosphorus (P) in lower concentration. The the sample, and the optimum compaction
complexity of this effluent is the high micro- parameters, optimum water content of 16.6%
biological activity that degraded the micro- and dry unit weight of 17.0kN/m3, were
nutrients increasing the organic matter content. observed. The results of Atterberg Limits were
Chemical analysis, pH and density 35% for LL and 20% for LP. So, the IP of this
determination and micro-biological analysis soil, IP, is 15%.
were carried out in other to have characterized
the vinase from one of the sugar-cane industry TABLE 2 Grain size distribution
of Campos dos Goytacazes, BRAZIL. Table 1 Fine Medium Coarse
Clay Silt Gravel
shows the concentration of nitrogen, phosphorus sand sand sand
and potassium as soon as the pH of one sample 42.7 10.3 14.0 21.0 10.2 1.9
of vinase prepared by two different ways.
Sample 1 was a sample of vinase in natura, TABLE 3 Atterberg limits and compaction test
while sample 1a was the same sample of vinase
but filtrated. These results confirm the high Gs LL LP IP hot d
content of potassium, 3.75 kg/m3. The pH of (kN/m3)
both samples were 4.0. 2.78 35.0 20.0 15.0 16.6 17.5

The chemical and mineralogical analysis of

TABLE 1 Vinase chemical composition
pH
samples of the soil pointed out that the soil is a
Sample N P2O K2O
kg/m3 Kg/m3 Kg/m3 composed predominantly by kaolinite. The pH
1 4.0 0,.8 0.16 3.75 is 5.9, the CEC is 31,3 mmolc/dm3, and a high
1a 4.0 0.52 0.14 3.52 concentration of SiO2 and Al2O3 in the clay
fraction were observed, 43.2% and 40.9%,
During seven days, the vinase density and pH respectively.
were measured in a proveta. The density
changes from 1.015 g/cm3 to 1.018 g/cm3 while PERMEABILITY TESTS
pH ranges from 3.9 to 4.2. In this stage a white Falling head permeability tests were carried out
thick layers were observed on the surface of the percolating water and vinase through specimens
effluent, suggesting that a high microbiological of soil compacted at optimum water content.
activity taken placed.
A rigid wall permeability device was used to
In order to check the microbiological carried out the falling head permeability test.
degradation, samples of vinase were collected The specimens degree compaction was 94%.
and chemical analysis were carried out. The The coefficient of permeability to water was
results show a little variation of a potassium 1.85 x10-7 cm/s.
 Cunha et al. 719

Following the falling head tests with water, the

permeability of the soil with vinase was
measured. During 18 days the permeability
parameter changed from 3.0x10-7cms to 2.0x10-
7
cms. Considering that the specimen has been
saturated, the decrease of permeability
coefficient may be attributed both to the soil and
the apparatus clogging up.

DIFUSION PARAMETERS
In order to evaluate the soil and vinase
interaction the diffusion molecular coefficient
and dispersion coefficient were investigated.
Three flexible wall columns were assembly with
Mariotte vases that were used to percolate
water, during saturation procedure of soil
specimens, and also to percolate the vinase.
These apparatus allow carry out three tests
simultaneously.

Three different hydraulic gradients were applied

20, 15 and 10, in order to define these transport
parameters. Samples of effluent were collected
to chemical analysis. The reference values of
concentration were 0,189 kg/m3 to the nitrogen
and 3.375kg/m3 to the potassium. These values
were used to calculate the breakthrough curves.
Figure 1 and 2 present the breakthrough curves
for nitrogen and potassium, respectively.

The permeability coefficients of the specimens,

measured during these tests, were greater than
that found with the falling head test. These
values were around 10-4 cm/s and 10-5 cm/s. Fig. 1 Breakthrough cures of Nitrogen
These high permeability coefficients, measured
during saturation step, may influence the The dispersion and retard coefficients were
adsorption of nitrogen and potassium. The obtained using the Ogata (1971) solution. The
breakthrough curves shows that the retard molecular diffusion and the dispersion are
calculated by relating the velocity against
coefficients is too low, ranging from 1.2 to 1.6.
dispersion coefficient. These average results are
These values suggested that the soil did not presented in Table 4.
retain these micronutrients.
720 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
Fig. 2 Brealthrough curves of potassium
CONCLUSIONS
The permeability coefficients of compacted
specimen found with falling head test and
constant head tests were considerably different.
Probably, it is due to the different compaction
degree of the specimens.
The results of column tests proved that the
potassium and nitrogen retention is small. The
retard coefficients ranged from 1,2 to 1,6 in all
tests, suggesting that the soil used in this
research is not a good barrier to these nutrients.
The dispersion and molecular diffusion
parameters were 0.156 and 0.22x10-5cm/s,
respectively, for the nitrogen, and, 0.765 and
0.10x10-5 cm/s, respectively for the potassium.
TABLE 4. Dispersion and molecular diffusion
Molecular
Dispersion
diffusion
(cm)
(cm2/s)
N 0.156 0.22x10-5
K 0.765 0.10x10-5
ACKOWLEDGE
The authors are grateful to the FAPERJ, UFV
and UENF.
REFERENCES
A. Ogata (1971) Theory of dispersion in a
Granular Médium. U.S. Geol. Surv. Proc.
Paper 411-I
Cunha, A. A. da (2007). Estudo da interação do
vinhoto e um solo utilizado o método das
colunas de lixiviação. Master Degree Thesis
in Civil Engineering, Campos dos
Goytacazes, RJ, Brazil, State University of
North Fluminense - UENF.
Fonseca, F. D. (2007). Avaliação da mobilidade
da vinhaça através de estudos em colunas de
solo. Master Degree Thesis in Civil
Engineering, Campos dos Goytacazes, RJ,
Brazil, State University of North Fluminense
– UENF.
Nobre, M.M. (1987). Estudo Experimental do
Transporte de Poluentes em Solos Argilosos
Compactados. Master Degree Thesis in Civil
Engineering, DEC da PUC-RJ, Brazil. 214p.
1987
 Sortion and Leaching Characteristics of Heavy Metals
Through Clay Mixed with Waste Materials
S.K Singh
PEC University of Technology, Chandigarh, India ( sksingh@pec.ac.in)

ABSTRACT Generally, 10-15% bentonite are added in locally available soil to achieve the desired
permeability (<10-9 m/s) as clay barrier. An experimental program were undertaken to investigate the
hydraulic conductivity and sorption behavior of locally available clay samples mixed with some
waste materials like wood saw dust and fly ash. Heavy-metal solutions (Pb, Cu and Cd) were
permeated to evaluate locally available soil mixed with reduced percentage of bentonite (5%) and
inert additives (5%) as potential clay barrier. The sorption of heavy metals was found to be highest
for soil + 5% bentonite +5% fly ash whereas hydraulic conductivities are found to be comparable.
Retention capacity was found to be highest for lead and lowest for cadmium.

INTRODUCTION metals most frequently found in the identified

Now-a days contamination of soil by human activities
has emerged as a widespread problem. Among the
most commonly encountered soil contaminants are
heavy metals, petroleum hydrocarbon and
halogenated organics. Mining, smelting,
containment facilities such as landfills, surface
impoundments and industrial activities have caused
extensive heavy-metal contamination of the
environment by directly introducing heavy metals
into the surrounding atmosphere, waterways, and
soil. The presence of heavy metals in the sub-soil and
ground water poses a potential threat to human health
and ecosystems. Engineered clay barriers with
permeability lower than 1 x 10 -9 m/s are commonly
used in both landfills and surface impoundments to
prevent the migration of heavy metals. For areas
afflicted with heavy-metal–contaminated soil or
ground water, vertical clay barriers may be used to
prevent heavy metals from reaching drinking water
sources. Migration of contaminants present in the
leachate through clay barriers are retarded due to its
low permeability and sorption of contaminants in
clay barrier. It is desired that liner system should be
able to restrict discharge of leachate to admissible
concentration level of contaminants in sub-soil and
ground water. The mobility studies with respect to
those contaminants are required to be studied. The
hazardous sites under Superfund programs are: As,
Cr, Pb, Zn, Ni, Cd, Cu, and Hg (USEPA 2004). Elevated
concentrations of heavy metals in soils are of
potential long-term environmental and health
concerns because of their persistence and
cumulative tendency in the environment, and their
associated toxicity to biological organisms (Nriagu
1979, 1988; Nriagu and Pacyna 1988). The clay
barriers are practical and economical, especially
when clay sources are locally available. Because
clay barriers of desired permeability are not available
everywhere, generally 10-15% bentonite are added
to achieve the desired permeability. The ability to
sorb heavy metals is also an important characteristic
to judge the suitability for liner material. Thus, the
compatibility of a clay barrier to a specific
contaminant depends on two factors i.e.
permeability of the clay barrier and its capacity to
retard the migration of contaminants through
sorption. The characteristics of locally available clay
as a barrier possibly could be improved by adding
some waste materials with high heavy-metal sorption
capacities. The paper deals with an experimental
program based on batch adsorption tests and
column leaching tests to investigate the hydraulic
conductivity and sorption behavior of locally
available clay samples mixed with some waste
materials like wood saw dust and fly ash in addition
722 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

to conventional bentonite. Heavy-metal solutions TABLE 2 Properties of Soil- Mix

(Pb, Cu and Cd) were permeated to evaluate locally
Soil- Mix LL PL Sp.Gr OMC 0D[Ȗd
available soil mixed with additive materials as (%) (%) (G) (%) (kN/m3)
potential clay barrier.
SB10 80 32.8 2.71 19.5 15.6
SB5F5 58 37.0 2.60 17.6 14.7
MATERIALS
SB5W5 52 35.0 2.61 16.8 14.1
The soil for the study was collected locally from
open pits at Mohali in the outskirts of Chandigarh. Heavy Metal Solutions
The locally available soil was mechanically sieved
to remove coarse fractions grater than 4.75mm sieve. Heavy-metal solutions (Pb, Cu and Cd) were
Physico-chemical properties of local soil were permeated through soil-mixes in batch and column
determined and reported in Table-1. Three types of studies. All the solutions were prepared by
soil mixes were prepared (by weight) for the study in dissolving calculated weight of each solid crystals
order to make its less permeable that can be used as of Pb(NO3)2 , Cu(NO3)2 and Cd(NO3)2 into distilled
clay barrier. water separately for known concentration of the
∑ Soil + 10% Ca-bentonite referred as SB10 metals in the solutions. Diluted nitric acid was used
to adjust the pH of all the solutions as 5 for both
∑ Soil + 5% Ca-bentonite + 5% Fly ash batch and column experiments.
referred as SB5F5
∑ Soil + 5% Ca-bentonite + 5% Wood saw METHODS
dust referred as SB5W5
Batch Adsorption Tests
Ca-bentonite (CaOAl2O35SiO22H2O) was procured
from commercial shop. Fly ash used was taken from Batch adsorption tests involve mixing a batch of
thermal power plant Bhatinda. The grain-size solutions of Pb(NO3)2 , Cu(NO3)2 and Cd(NO3)2, each
composition of fly ash was sand 9.5%, silt 89% and with the same volume with known initial solute
clay 1.5% and it was non-plastic. The wood saw dust concentration, with a fixed mass of adsorbent in each
(d50 = 0.25 mm) was obtained from local wood saw mill. reaction vessel. The resulting change in solute
Index properties and compaction characteristics of concentrations after contact with the adsorbent
the soil- mixes are given in Table 2. provides the basis for the construction of isotherms
(EPA 1992).The procedure of these tests was based
TABLE 1 Physico-Chemical Properties of Local Soil
on ASTM D4319 guideline, with soil-solution ratio
Properties Values Chemical 1:4 (25 g of dry soil to 100 mL of solution).
Values
Content
Sand (%) 10 SiO2 (%)
Equilibration time, defined as the minimum time
62.50
needed to establish a rate of change of the solute
Silt (%) 67 Al2O3 (%) 26.00 concentration in solution that is equal to or below
Clay (%) 23 Fe2O3 (%) 2.50 5% for a 24 h interval (EPA 1992), was determined as
LL (%) 32.7 CaO (%) 24 hrs. The suspensions were placed inside 250 mL
1.60
inert flasks at an orbital shaker for 24 h. Liquid and
PL (%) 19.8 MgO (%) 3.20 solid phases were separated by filtration, and the
SL (%) 16.4 Na2O (%) 0.07 concentration in the liquid phases was analyzed by
CEC 12.7 K2O (%)
0.04
atomic Adsorption spectrometer (Make ECI, EC AAS
meq/100gm 4103). Lead, copper and cadmium were tested
pH Organic
8.4
Matter (%)
2.87 separately. Duplicate were conducted for each test.
 Singh 723

Leaching Column Tests intervals (usually two days) and the concentrations
of metals were measured. The test duration was 15-
The experimental set up consisted of a cylindrical 20 days.
column made of PVC and measuring 7.73 cm inner
diameter and 10 cm in height. End diffusers made of RESULTS AND DISCUSSION
perforated PVC plates of 1.5 cm were placed on the
top and bottom of compacted soil specimen in the Batch Tests
column for uniform discharge through soil column.
The height of the soil column in the PVC cell was 7 The sorption capacity of the soil-mixes for heavy
cm. The other major components of the experimental metals under consideration (Pb,Cd and Cu) in form
setup were Influent Reservoir, 10 L capacity made of of isotherm curves are shown through Figures 1 to
polypropylene, pressure system and effluent 3. Freundlich isotherms, expressed as in Eq. (i), were
collector. fitted to the experimental data using nonlinear least-
squares curve fittings.
The oven dried soils of required quantity was mixed
with necessary amount of water separately as to S = Kf (Ce)b (i)
prepare samples of required density. The soils were
where S = adsorption degree (mg/g); Ce=equilibrium
mixed thoroughly and kept in polythene bag in humid
concentration (mg/L); Kf =Freundlich partition
desiccators overnight to achieve uniform moisture
coefficient (L/g); b=empirical constant of the
content. The soil was then compacted in the PVC
Freundlich isotherm;
cell in three equal layers with dynamic compaction
to achieve 0.85 times of Proctor’s maximum dry
The coefficient of regression fit ranges from 0.94-
density at water content 2% wet side of optimum
0.98 denoting that sorption of metals on the soil can
water content. The soil in the column is saturated
be best described as Freundlich isotherms. The
by passing distilled water through influent reservoir
Freundlich equations for isotherms are presented in
consisting of with two opening, one at the top for
Table 3 to compare the relative sorption capacities
transferring the source solution of interest with
of soil-mixes for metals under consideration.
known concentration (100 ppm) into it and the other
at the bottom to allow it to migrate through the soil TABLE 3: Equations of Freundlich Isotherms
specimen. After saturation with water for 24 hours,
the solution of interest is placed in this influent Sorbate SB10 SB5F5 SB5W5
reservoir and is stirred at frequent intervals so as to Cu S = 0.058 S = 0.080 S = 0.056
0.52 0.45 0.421
maintain constant initial concentration. The solution (Ce) (Ce) (Ce)
is then passed through the soil compacted in the Pb S = 0.120 S = 0.135 S = 0.075
0.405 0.387 0.500
(Ce) (Ce) (Ce)
column at constant hydraulic gradient of 25 to reduce
Cd S = 0.053 S = 0.083 S = 0.046
the testing duration to reasonable period. Pressure (Ce)
0.358
(Ce)
0.255
(Ce)
0.340
gauge is connected to the influent reservoir and to
the column assembly. A uniform pressure of 17.5 The sorption ranking deduced from the Table 3 is
kPa is maintained throughout the experimental period SB5W5 < SB10 < SB5F5 among soil mixes and Cd <
by controlling the flow rate from the influent Cu < Pb among heavy metals. Therefore, soil mixed
reservoir. The effluent is collected in the effluent with 5% of bentonite and 5% of fly ash is capable to
collector consisting of a measuring jar covered at sorb heavy metals more than that of soil mixed with
the top so as to avoid evaporation of collected 10% bentonite. Mobility of cadmium and lead will
leachate. The volume of the effluent that comes out be highest and lowest, respectively, among the
of the column with time was monitored at regular considered heavy metals.
724 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
Leaching Column Tests
The equal volume of solutions of Pb(NO 3) 2 ,
Cu(NO3)2 and Cd(NO3)2 into distilled water were mixed
and permeated in column. Concentration of metal in
each solution was 100 ppm. The effluent was
collected in the effluent collector consisting of a
measuring jar covered at the top. The volume of the
effluent that comes out of the column with time was
monitored at regular intervals (usually two days).The
permeability of the soil-mixes with respect to
discharge is presented in Figure 4. From Figure 4 it
can be observed that the permeability increases with
discharge but tends to be leveled off after permeation
of 5-7 pore volumes. The leveled off co-efficient of
permeability (m/s) of heavy metal solutions were
found to be 1.15 x 10-9, 1.25 x 10-9 and 1.47 x 10-9 for
the soil mixes SB10, SB5F5 and SB5W5, respectively.
Hydraulic conductivities of the clay barriers should
be determined with respect to permeation of the
contaminating fluids for which they are designed to
contain as physico-chemical alteration of the soil
environment, soil acidification and heterogeneous
preferential flow may affect the heavy metal mobility
through the sub-soil system (Kabata-Pendias and
Pendias 1991; Dube and Galvez-Cloutier 1998). The
co-efficient of permeability (m/s) of tap water were
found to be 0.85 x 10-9, 1.11 x 10-9 and 1.27 x 10-9 for
the soil mixes SB10, SB5F5 and SB5W5, respectively.
Thus, hydraulic conductivities through the soil mixes
get augmented upon permeation of heavy metals
through them.
The local soil mixed with 10% of bentonite (SB10)
just qualifies the criterion of hydraulic conductivity
less than 10-9 m/s but other two soil-mixes fails. But
at the same time, it was found that soil mix SB5F5
had the greater sorption capacity than soil mix SB10
which compensates favourly. On account of higher
sorption capacity, the soil SB5F5 could be selected
as barrier material at heavy metal contaminated site
in spite of slightly higher value of hydraulic
permeability than required. Therefore, replacement
of 5% of bentonite with fly ash will make the liner
material economical and at the same time recycling
of waste material. The soil mixed with bentonite with
wood saw dust could not be a potential material for
clay barrier on account of lower permeability and
sorption capacity.
CONCLUSIONS
The compatibility of a clay barrier to retard the
migration of contaminants should be judged taking
into consideration of permeability and its sorption
capacity both. The performance of locally available
clay with 5% of bentonite and 5 % of fly ash are
comparable with clay mixed with 10% bentonite as a
liner material for heavy metal retardation. Sorption
characteristics of the bentonite enriched (10%) soil
with heavy metals increased by replacing 5% of
bentonite with same percentage of fly ash. Moreover,
high percentage of bentonite is susceptible of cracks
due to swelling and shrinkage. Reduced bentonite
along with flyash will result into a more stable barrier
material. Relative mobility of the metals are in the
order of Cd > Cu > Pb.
REFERENCES
Dube,J.S. and Galvez-Cloutier, R (1988) Sequential
extractions and SEM for Pb behaviour analysis
under soil pH alteration, In proceedings of the
3 rd Int Congress on Environmental
Geotechnics, CSCE 28th annual conf, London
Ont, 169-179.
Environmental Protection Agency (EPA). (1992)
Batch-type procedures for estimating soil
adsorption of chemicals, Technical Resource
Document No. EPA/530-SW-87-006-F,
Washington, D.C.
Ernst, W. H. O. (1995) Decontamination or
consolidation of metal contaminated soils by
biological means, Heavy metals: Problems and
solutions, W. Salomons, U. Forstner, and P.
Mader, eds., Springer, Berlin, 141–149.
Kabata-Pendias, A. and Pendias, H. (1991) Trace
elements in soil and plants, CRC Press, Boca
Raton Fla
Nriagu, J. O. (1979) Global inventory of natural and
anthropogenic emissions of trace metals to the
atmosphere, Nature (London), 279, 409–411.
 Singh 725

Nriagu, J.O. (1988) A silent epidemic of environmental Shackelford, C. D., and Rowe, R. K. (1998)
metal poisoning? Environ. Pollut., 50, 139–161. Contaminant transport modeling, Proc., 3rd Int.
Nriagu, J. O., and Pacyna, J. M. (1988) Quantitative Congress of Environmental Geotechnics,
assessment of worldwide contamination of air, Lisbon, Portugal, 939–956.
water, and soils by trace metals, Nature USEPA. (2004) Cleaning up the nation’s waste sites:
(London), 333, 134–139. Markets and technology trends, 4th Ed., EPA
Rowe, R. K. (1987) Pollutant transport through 542-R-04-015, Office of Solid Waste and
barriers, Proc., Geotechnica Practice for Waste Emergency Response,Washington,D.C.
Disposal ’87, ASCE, New York, 159–181.

Fig. 1 Sorption Isotherms for Cu Fig. 2 Sorption Isotherms for Pb

(For colour figure, refer to CD) 0.7

0.6
Adsorbed Conc. (mg/l)

0.5

0.2
0.4
1.25

0.3 SB 10
SB5F5
SB5W5
K (m/s) X 10-9

0.15
Adsorbed Conc (mg/l)

1 0.2

0.1
0.1 SB10
SB10
0.75 SB5F5
SB5F5 0
0 10 20 30 40 50 60 70 SB5W5
80 90
SB5W5
Equilibrium Concentration (mg/l)
0.05

0.5

0
0 10 20 30 40 50 60
Equilibrium Concentration (mg/l) 100 200 300 400 500 600 700 800 900 1000 1100 1200
Discharge Volume (ml)

Fig. 3 Sorption Isotherms for Cd Fig. 4 Hydraulic Permeability Curves

(For colour figure, refer to CD) (For colour figure, refer to CD)
 Arsenic Contamination of Ground Water in Bangladesh
Mohammad Nurul Islam
Rajshahi University of Engineering & Technology, Rajshahi, Bangladesh (nislamce@yahoo.com)
Mohd. Abdus Sobhan
Rajshahi University of Engineering & Technology, Rajshahi, Bangladesh (msobhan@yahoo.com)
Ayesha Siddika
Rajshahi Polytechnic Institute, Rajshahi, Bangladesh (shompace98@yahoo.com)
H.M Rasel
Rajshahi University of Engineering & Technology, Rajshahi, Bangladesh (hmrruet@yahoo.com

ABSTRACT Arsenic is highly toxic and carcinogenic and Bangladesh is at risk of arsenic contaminated
drinking water. Groundwater is particularly vulnerable to contagion. Arsenic pollution of groundwater
recognized as a serious threat to human health. The purpose of this paper is to highlight the different sources,
effects and remedial measures of arsenic infectivity in ground water in Bangladesh. It becomes essential to
take necessary measures to mitigate the contamination

INTRODUCTION Manmade Origin

Contamination of ground water from arsenic in
Bangladesh is judged as one of the recent
environmental disaster in the world (Smith et.
al.2000). According to the WHO (World Health
Organization) Report, September, 2000,
Bangladesh is snatching with the largest mass
poisoning of a population in history as ground
water that consumed for drinking purpose has been
tainted with arsenic. Arsenic contamination of
ground water and its adverse effect on health has
now been established as a public health problem.
Now a day it is revealed that arsenic has entered in
food chain, which becomes a great threat for
human being .There are different theories about the
mobility of arsenic contamination, but the actual
reason is still elusive. The effect of contamination
is possible to minimize by applying short and long
term remedial measures.
MOBILITY OF ARSENIC
Several theories have been put forward to explain
the origin of dissolved arsenic in groundwater of
Bangladesh. The summary for the mobility of
arsenic may be divided under two headings:
manmade disasters and natural cataclysm.
Mobility of arsenic contamination from different
points of view may be explained as follows:
Over withdrawal of ground water
Due to India’s unilateral diversion of Ganges water
at Farakka Barrage, irrigation becomes necessary
for crops. On the other hand, to get
bacteriologically safe drinking water installation of
tube well spread all over the Bangladesh. For
irrigation and drinking intention, by last few
decades, huge amount of ground water withdrawn,
which affect severely the ground water aquifer and
changed the ground water chemistry. On the other
hand MIT news reported that arsenic in Bngladesh
is interrelated to crop irrigation (MIT News 2002
a,b). Neuman et. al. (2009) also reported that crop
irrigation play an important role in arsenic
contamination of ground water.
Agricultural Activity
The economy of Bangladesh is primarily
dependent on agriculture .The agriculture sector
plays a very important role in the economy of the
country. Agricultural production in Bangladesh has
increased a lot through last three decades due to
irrigation and use of chemical fertilizer. But now
from research it comes to light that chemical
compounds containing fertilizers have direct
impact on the arsenic contamination. High nitrate
concentration from chemical fertilizers increased

arsenic attentiveness. Use of phosphate fertilizers
on lead arsenate-contaminated soils may lead to
arsenic contamination of groundwater (Davenport
and Peryea 1991).
According to Sciences News (2009) human
alteration to the landscape, the construction of
villages with ponds and the adoption of irrigated
agriculture are responsible for the current pattern
of arsenic concentration underground. On the other
hand Neuman et. al. 2009, reported that water with
high arsenic content originates from the human
built ponds and water with lower arsenic contents
from the rice field.
Probable Natural Origin
Current research emphasis on the geogenic source
of arsenic and its release in ground water is a
natural process (Bhattacharya et al., 1997).
Oxidation of pyrite (FeS2) or arsenopyrite (FeAsS)
was postulated as the dominant process for arsenic
mobilization due to lowering of the water table
following excessive pumping of groundwater that
was widely accepted at the beginning (Chakraborti
et al. 1996 and Mallik and Rajagopal, 1996). An
alternate theory, known as the Fe oxyhydroxide
reduction hypothesis, is now widely accepted as
the principal mechanism of Arsenic mobilization
in the groundwater of the alluvial aquifers of the
GBM Delta (Bhattacharya et al., 1997,
Bhattacharya et al., 2001, Ahmed et al., 1998,
Nickson et al., 2000, Routh et al., 2000, McArthur
et al., 2001, Dowling et al., 2002 and Anawar et
al., 2003).
EFFECTS OF ARSENIC
Effect on Human Body
The toxic effects of arsenic on the human body are
brutal. Humans appear to be terribly susceptible to
arsenic than animals. Diseases caused due to
arsenic is classified under three headings (Azad
2001, Hokkaido University, Japan) : i) Initial
Stage: Desmatities, keratities, conjunctivities,
bronchities and gastroenterities. ii) Second Stage:
Peripheral neuropathy, hepatopathy, melanosis,
depigmentation and hyperkeratosis. iii) Final
Stage: Gangrene in the limbs, malignant neoplasm,
Islam et al. 727
and cancer. In Bangladesh, the majority of patients
are suffering from the initial and second stages.
Social Effect
The arsenic infected diseases increase the anxiety
of the people. The native people consider the
arsenic diseases contagious. In many instances, the
people suffering from arsenic diseases have been
ostracized by neighbors, friends and relatives
(Safiuddin Karim 2001). The situation is worse for
women. The women suffering from arsenic
diseases are increasingly facing ostracization and
discrimination. Young women suffering from
arsenicosis are often compelled to stay unmarried.
Thus, the unaffected parents and children are also
suffering socially with the affected females. Above
all, the affected people are losing their usual social
relations with neighbors and relatives.
Financial Effect
Bangladesh is one of the poorest countries in the
world. To mitigate the arsenic problem it needs
huge financial support, which becomes a burden.
Every year in the annual budget a part is allocated
for arsenic combating. International agencies
extend their hand, but it is not sufficient to mitigate
the problem.
REMEDIAL MEASURES
There are good numbers of treatment technologies,
which are efficiently used to remove arsenic from
water like Iron (and Manganese) Oxidation,In-situ
oxidation, Soalr oxidation, Arsenic and iron
removal plant, Bucket treatment unit, Steven
institute technology, pitcher treatment,activated
alumina, ion exchange,membrane technique etc.
(Rahman M.H. et al.).But socio-economic point of
view, all technologies are not suitable for
Bangladesh. We need cost effective technologies
which may adopt for affected area. For Bangladesh
environment some suitable treatment procedure for
arsenic were developed by private as well as
government finance like Bangladesh Council of
Scientific and Industrial Research (BCSIR) Filter
Unit, DPHE-Danida Fill and Draw Unit, Read-F
Aresenic Removal Unit, Sapla Filter, Granet
Home-made Filter, Adarsha Filter, Safi Filter,
Bijoypur Clay Filter, Several Cartridge Filter,
728 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
Iron coated sand, Granular ferric hydroxide,
Tourmaline mineral. Except remediation in
arsenic contaminated area people may use
alternative sources of drinking water like ground
water from deep arsenic free aquifer, deep tube
well, shallow shrouded tube well and very shallow
shrouded tube well, dug well, Rainwater
harvesting, Solar densification.
CONCLUSION
There are the different hypotheses about source of
arsenic contamination, but the actual source is still
now concealed. It is very essential to assess the
risk of arsenic contamination. Some researchers
found arsenic entered into the food chain, which
becomes a threat for human life. It is now the vital
issue to find out the long term suitable remedial
measures against contamination. On the other hand
treatment for the infected person is an important
question. National and international agencies may
play an important role in favor of the mitigation of
arsenic contamination and in future it may be
possible to reduce the effect of contamination.
REFERENCE
Ahmed, K.M., Imam, M.B. Akhter, S.H., Hasan,
M.A., Alam, M.M., Chowdhury,S.Q., Burgess,
W.G., Nickson, R., McArthur, J.M., Hasan, M.K.,
Ravenscroft P. and Rahman,M.M., 1998.
Mechanism of arsenic release to groundwater:
geochemical and mineralogical evidence. Internat.
Conf. on Arsenic Pollution of Groundwater in
Bangladesh: Causes, Eơects and Remedies.
Dhaka, February 1998, pp. 125–126.
Ahmed K.M. et al., (2004), Arsenic enrichment in
groundwater of the alluvial aquifers in Bangladesh:
an overview, Applied Geochemistry Volume 19(2),
181-200.
Anawar, H.M., Akai, J., Komaki, K., Terao, H.,
Yoshioka, T., Ishizuka, T., Safiullah, S., Kato, K.,
2003. Geochemical occurrence of arsenic in
groundwater of Bangladesh: sources and
mobilization processes. J. Geochem. Explor. 77,
109– 131.
Azad M. (2001), Diseases suspected to be caused
by Arsenic, (Online).
http://www.sustainablefuture.se/arsenic/diseases.html
Bhattacharya, P., Chatterjee, D., Jacks, G., 1997.
Occurrence of arsenic-contaminated groundwater
in alluvial aquifers from delta plains, eastern india:
options for safe drinking water supply. J. Water
Resour. Dev. 13, 79–92.
Bhattacharya P., Jacks G., Jana J., Sracek A.,
Gustafsson J.P.,Chatterjee D., 2001. Geochemistry
of the Holocene Alluvial sediments of Bengal
Delta Plain from West Bengal, India: implications
on arsenic contamination in groundwater. In: Jacks
G., Bhattacharya P., Khan A.A. (Eds.),
Groundwater Arsenic Contamination in the Bengal
Delta Plain of Bangla- desh. KTHSpecial
Publication. TRITA-AMI Report 3084, pp. 21–40.
Bhattacharya et al. (2004), Arsenic enrichment in
groundwater of the alluvial aquifers in
Bangladesh:an overview” Applied Geochemistry
Volume 19(2), 181-200.
Chakraborti D. et al. 1996, Arsenic in groundwater
in seven districts of West Bengal, India: the
biggestarsenic calamity in the world, Current
Science , Volume 70 (11), 976–986.
Davenport J.R. and Peryea F.J. (1991), Phosphate
Fertilizers Influence Leaching of Lead and Arsenic
in a Soil Contaminated with Lead Arsenic, Water,
Air, and Soil Pollution, 57-58:101-I I0.
Dowling, C.B., Poreda, R.J., Basu, A.R., Peters,
S.L., 2002. Geochemical study of arsenic release
mechanisms in the Bengal Basin groundwater.
Water Resour. Res. 38, 1173– 1190.
McArthur, J.M., Ravencroft, P., Safiullah, S.,
Thirlwall, M.F., 2001. Arsenic in groundwater:
testing pollution mechanism for sedimentary
aquifers in Bangladesh. Water Resour. Res. 37,
109–117.

Mallik, S., Rajagopal, N., 1996. R, Groundwater
development in the arsenic-aơected alluvial belt of
West Bengal— Some questions. Curr. Sci. 70,
956–958.
MIT News 2002 (a), Arsenic in Bangladesh
drinking wells may be linked to crop irrigation,
(Online)
http://web.mit.edu/newsoffice/2002/bangladesh.
MIT News 2002 (b), Arsenic in Bangladesh
drinking water from wells may be linked to crop
irrigation, (Online)
http://web.mit.edu/newsoffice/2002/bangladesh-1204.
Nickson, R.T., McArthur, J.M., Ravenscroft, P.,
Burgess, W.G., Ahmed, K.M., 2000. Mechanism
of arsenic release to groundwater, Bangladesh and
West Bengal. Appl. Geochem. 15, 403–413.
Neuman, R.B. et al (2009) Anthropogenic
influences on groundwater arsenic concentrations
in Bangladesh, Nature Geoscience
http://censam.mit.edu/publications/nature-geo.pdf
Rahman M.H. et al. () Arsenic Contamination of
Groundwater in Bangladesh and Its Remedial
Measures (online)
http://www.cepis.ops-
ms.org/bvsacd/arsenico/arseniccont/arsenic.pdf
Islam et al. 729
Routh, J., Bhattacharya, P., Jacks, G., Ahmed,
K.M., Khan, A.A., Rahman, M.M., 2000. Arsenic
geochemistry of Tala groundwater and sediments
from Satkhira District, Bangladesh. Eos Trans Am.
Geophys. Union 81, 550.
Safiuddin M. and Karim M.M. (2001), Ground
Water Arsenic Contamination in Bangladesh:
Causes,Effects and Remediation, Proc. 1stT IEB
international conference and 7 th annual paper
meet, Bangladesh, 1-14.
Science News (2009), Dissolved Arsenic in
Bangladesh Drinking Water Is from Human
Alteration of Landscape, (online)
http://www.sciencedaily.com/releases/2009/11/0911151
34130.
Smith A.H. , Lingas E.O. and Rahman M. (2000)
Contamination of drinking-water by arsenic in
Bangladesh: a public health emergency, Bulletin of
the World Health Organization, Volume 78(9),
1093-1103.
 Effects of Friability and Grain Size on the Leaching of Heavy Metals
in Excavated Rock Materials
T. Inui
Kyoto University, Kyoto, Japan (inui.toru.3v@kyoto-u.ac.jp)
T. Katsumi
Kyoto University, Kyoto, Japan (katsumi.takeshi.6v@kyoto-u.ac.jp)
M. Katayama
Kyoto University, Kyoto, Japan (mariko-110@t06.mbox.media.kyoto-u.ac.jp
A. Dejima
Kyoto University, Kyoto, Japan (akanedejima@moueeyaro.mbox.media.kyoto-u.ac.jp)

ABSTRACT A series of leaching tests were conducted for five different rock materials from excavation
works at several construction sites in Japan, which contained natural derived lead and/or arsenic. The focus
was placed on the effect of their friability and grain size on the leaching amount of trace metals as well as
chemical properties of leachate (pH, Eh). Typical hard rocks such as black shale and andesite were subjected
to the specific leaching trend that leaching concentrations of more friable and finer fractions were higher
than those of less friable and course fractions. However, there are no significant effects of friability and grain
size on the leaching amounts for relatively softer mudstones.

INTRODUCTION influence the leaching amount of trace elements,

In Japan, soils and rocks generated from
excavation works often contain natural-derived
heavy metals and other harmful substances such as
lead and arsenic with relatively low concentrations
(Geological Survey of Japan, AIST 2004). These
constituents are fundamentally stable and
immobile under the depositional environment.
However, in some cases, leaching of these
constituents is triggered by the exposure with
water and oxygen after excavation. Acid drainage,
which is resulted mainly from the interactions
between water, oxygen, and sulphide minerals in
rocks, often increases the leaching concentration of
these elements. Thus, development of a testing
methodology to evaluate their long-term leaching
behaviour has been a great issue in Japan.
Conventional batch tests have been widely
employed to evaluate the leaching potential of
contaminated soils. However, when they are
employed for rock materials, sample preparation
including sampling and crushing is more likely to
compared with the cases of soils and sediments. In
this study, a series of batch leaching tests were
conducted for five rock materials with various
hardness and compositions, which were generated
from excavation works at several construction sites
in Japan. They contain certain concentrations of
natural-derived lead and arsenic, which are
expected to be released in the case they are
exposed to water and oxygen. The main focus was
placed on the effects of their friability and crushing
grain size on the leaching amount of trace metals
as well as chemical properties of the leachate such
as pH and electrical conductivity.
MATERIALS AND METHODS
Rock Samples
Five different rocks materials were used. They
were generated in construction works including
excavation and tunnelling, which were executed in
mountainous areas in Japan. Accordingly, it can be
considered that heavy metals contained in them are
derived from natural resources. Five rock samples

consist of black shale, andesite, and three
mudstones (MS-1, 2 and 3). MS-1 was more
solidified and much harder than the other two
mudstones. They were all stored with their natural
water content under the temperature of
approximately 20 ºC in sealed condition to prevent
the exposure to oxygen and water.
Crushing and Classification
Each rock block was subjected to two crushing
processes to prepare the samples with various
friability and grain sizes for the leaching test. The
detailed procedure is shown in Fig.1. In the first
crushing process, each rock block was hit
manually by a non-metallic hammer until the
maximum grain size was less than 9.5 mm.
Thereafter, crushed rock was sieved and classified
into two fractions; 2 to 9.5 mm in grain size
(Fraction A) and < 2 mm (Fraction B). In addition,
fraction A was subjected to the second crushing
process, in which 0.7 kg of sample was placed in
the compaction mould (I = 100 mm) and crushed
using a rammer basically according to JIS A 1210
(Japan Industrial Standards Committee 2009).
Compaction energy applied to the sample was
equal to approximately 1300 kJ/m3, which is
ranging between the Standard Proctor energy and
the Modified Procter energy.
Fig.1 Crushing and classification process employed in
this study.
Fractions A and B were re-classified into sub-
fractions according to the grain size as shown in
Fig.1. Fraction B consists of more friable elements.
Inui et al. 731
The first crushing process simulates the rock
crushing due to the excavation and blasting, and
the second one simulates the compaction of the
excavated material when used for filling or
embankment.
Chemical Tests
Arsenic (As) and lead (Pb) contents of each sub-
fraction were determined using energy-dispersity
X-ray fluorescence spectrometer (Shimadzu EDX-
720) according to the Japanese standard testing
method (Japan Industrial Standards Committee
2008). In addition, a conventional batch leaching
test was conducted for each sub-fraction according
to the official batch leaching test for soil quality,
regulated by the Japanese Environmental Agency
notification No.46, 1991. The test conditions were
shown in Table 1. Leachate was filtered using a
0.45 ȝm-opening membrane filter after the
centrifuge separation, and then analyzed to
determine As and Pb concentrations using the
atomic adsorption spectrometer with the
electrothermal atomization system.
Table 1 Testing conditions of the batch leaching test
Sample weight (g) Variable
Solvent Distilled water (pH = 5.8 to 6.3)
Liquid to solid ratio 10 mL/g-sample
Duration 6 hours
Mixing Horizontal shaking at 200 rpm
RESULTS AND DISCUSSIONS
Friability and Hardness of Rock Samples
Figure 2 shows the percentage of fraction B in wet
weight-basis after the first crushing process.
Fig.2 Percentage of fraction B after the first crushing
process.
732 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

Figs.3 Grain size distribution curves of fractions A before and after the crushing process using a rammer.

Larger percentages ranging from 25 to 30% were are 40% and 22%, respectively. Mudstone-1 is
observed in relatively harder rock samples expected to have the friability and hardness which
(Andesite and MS-1). This was because more is ranging between those of both groups.
energy was applied to andesite and MS-1 to make
their maximum grain size less than 9.5 mm and Distribution and Leaching of Heavy Metals
relatively larger portions were accordingly crushed Figure 4 shows the As and Pb concentrations of
into <2mm in grain size. sub-fractions A of andesite. For As, no significant
effect of the grain size and friability was observed
Figures 3 show the grain size distribution curves in andesite and the other four samples. For Pb, the
for fractions A of five rock samples before and finer fractions tend to contain the more Pb,
after the second crushing process. Considering that compared with As, especially in black shale and
a constant energy was applied in the second andesite.
crushing, how the curve after the crushing is
shifting in left-upper direction from the original
curve is a good index of friability and hardness of
the rock sample of concern. Judging from the
typical compression strengths of these rock
samples, black shale and andesite are harder and
less friable compared with mudstones. Grain size
distribution curves of black shale and andesite in
Figs. 3 are consistent with this presumption.
Percentages finer than 2 mm and 850 Pm are 30%
and 12%, respectively, for black shale and andesite.
On the other hand, mudstones-2 & 3 are more
friable: percentages finer than 2 mm and 850 Pm Fig.4 Effect of the grain size on As and Pb contents.
 Inui et al. 733

Figures 5 shows the As leaching concentrations of

each fraction of five rock samples (Fractions B of
black shale were not tested). Leaching
concentrations of fractions B of andesite and
mudstone-1 were much higher than those of
fractions A. This friability-dependency was also
observed in the Pb leaching concentrations and
electric conductivity values. As for fractions A,
leaching concentrations of finer fractions were
higher than those of courser fractions. Unlike these
harder rocks, a definite friability-dependency was
not observed in the leaching properties of
mudstones-2 and 3. For mudstone-3, fractions finer
than 2 mm had similar leaching concentrations
regardless of friability.

From these observations, it can be concluded that

spatial heterogeneity in physical and chemical
properties exists more significantly within black
shale, andesite and harder mudstone, and more
friable fractions of them have the higher leaching
potential as well as the chemical reactivity. On the
other hand, there are less significant effects of
friability and grain size on the leaching amounts
for relatively softer mudstones. Thus, sample
preparation including sampling and crushing prior
to the batch leaching test should be carefully
considered in the case of harder rock materials.

ACKNOWLEDGEMENT
This research was financially supported by the
Grant-in-Aid for Scientific Research (B) (No.
20360211, Project Leader: Dr. Masashi Kamon
(Kagawa National College of Technology)), Japan
Society for the Promotion of Science.

REFERENCES
Geological Survey of Japan, AIST (2004):
Elemental Distribution in Japan –Geological
Map of Japan, National Institute of Advanced
Industrial Science and Technology.
Japan Industrial Standards Committee (2008):
Determination of Arsenic and lead in Clay and
Sand Using Energy-dispersive X-ray
Fluorescence Spectrometry, JIS K 0470.
Japan Industrial Standards Committee (2009): Test
Methods for Soil Compaction Using a Rammer,
Figs.5 Effect of grain size on the leaching of arsenic. JIS A 1210.
 Long Term Column Experiments for Testing the Landfill Leachate
Diffusion Through 0.5 m Thick Compacted Clay
J Hervás, J M Martínez Santamaría
Laboratorio de Geotecnia, Centro de Estudios y Experimentación de Obras Públicas,
Alfonso XII, 3-5. 28014 Madrid, Spain. (jesus.hervas@cedex.es,
jose.m.martinez@cedex.es)
A I Ruiz, M Regadío, I de Soto, M Rodríguez, J Cuevas
Dpto de Geología y Geoquímica, Universidad Autónoma de Madrid, Campus de
Cantoblanco. 28049 Madrid, Spain. (anai.ruiz@uam.es, mercedes.regadio@uam.es,
isabel.desoto@uam.es, manuel.rodriguez.rastrero@uam.es, jaime.cuevas@uam.es)
ABSTRACT Two clay materials, were compacted in a cylinder of 1m length in which 0.50 m was occupied
by clay and the upper 0.45 m was filled with a landfill leachate (acetate-chloride (pH 5; 30 mS/cm and urban
landfill (pH = 8.5; 40 mS/cm). After 10 and 48 months clays were characterized for hydraulic conductivity
(HC), mineralogy, and physical-chemical parameters. The clay experimented softening to a 5-10 cm depth.
The specific surface area decreased in contact with leachate due to bio-film formation around clay particles.
However, HC was lowered but maintained around the initial order of magnitude (10-9- 10-10 m/s).
Mineralogical changes were limited to the dissolution of carbonates (2 cm depth) at pH 5. NH4+, Na+ and K+,
penetrated as exchangeable cations (20 cm depth) but chloride reached more than 30 cm in 4 years. SO42-
was partially reduced, by means of the precipitation of traces of dark patches of sulfides.

predominance of clay under a landfill vessel and

INTRODUCTION
The technical requirements for landfill of
municipal wastes in the European Union (EU) are
given in the Council Directive 1999/31/EC and
decision (2003/33/CE). In this legislation it is
established that a geological barrier of at least 1 m
thick with a hydraulic conductivity of 1·10-9 m·s-1
is required. A suitable geological barrier with the
ability to confine waste and buffer the hazardous
leachates is a key factor for the protection and
safety objectives in waste disposal. (Bilitewski et
al., 1997; Savage, 1995; Astudillo, 2001).
the natural pollution attenuation. Ammonium
attenuation as a function of depth is a good
example illustrating the role of clays (National
Groundwater and Contaminated Land Center,
2003). However, some biogeochemical processes
should induce physical effects, as the clogging of
the porous clay matrix. Redox reactions as well,
will drive to the precipitation of interstitial
minerals as calcite, siderite or sulphides (Islam et
al., 2001; Steefel et al., 2005; Castelló et al., 2008;
Cuevas et al., 2008). The scope of this work is to
characterize the effects of leachate-clay
interactions in terms of permeability changes and
chemical reactions produced in a 0.5 m clay
thickness and 1-4 years scale.

Clay minerals are the natural materials with lowest
permeability due to their small particle size and
complex porous structure. Their high specific
surface area allows strong physical and chemical
interactions with fluids and dissolved species.
These interactions are responsible for the retention
of leachate components (Davis and Kent, 1990;
Sposito, 1990; Stumm, 1992; Rowe et al., 1995;
Sawney, 1996; Michael et al., 2002). This can be
shown by the correlation between the
MATERIALS AND METHODS
Two Spanish clays composed mainly by illite with
accessory smectite (TABLE 2), used as ceramic
raw materials (Cuevas et al., 2008), were
compacted in a cylindrical tube of 1m length in
which 0.50 m was occupied by clay and the upper
0.45 m was filled with a leachate solution. The
clays were compacted at water saturated
conditions (Proctor test: 1.7-2.0 g/cm3 dry density;
14-10 % H2O). After 10 months in contact with,
 Hervás et al 735

either an acetate-chloride (L1: leachate pH 5; 30 synthetic acid (pH=5), L4: El Garráf leachate. k:
mS/cm) solution or a “true” leachate solution, as Hydraulic conductivity. *: clay disaggregates, HC non
sampled in an active old (>10 years) urban landfill measured.
(L2: pH = 8.5; 40 mS/cm), TABLE 1, Pantoja
(TABLE 2) clay was characterized for hydraulic RESULTS AND DISCUSSION
conductivity (HC; edometer), mineralogical (X-
Ray Diffraction) and physical-chemical properties The clay at the leachate interface experimented
(BET specific surface area (N2-gas adsorption), swelling and softening in a 5-10 cm depth region
Na, K, NH4+ exchange capacity and soluble measured from the leachate interface. In this
anions, Cl-, SO42-: measured by ion region specific surface area decreased due to bio-
chromatography (IC) in a distilled water aqueous film formation around clay particles and
extract: 1solid:10 water). In addition, a 4 years aggregation processes. (FIGURE 1)
similar experiment was performed using the
Bailén Clay (TABLE 2), representative of the
usual complex mineralogy for the clay materials
under some landfills in Spain.
TABLE 1 Synthetic leachate and landfill leachate (El
Garráf, Barcelona, Spain)
-----------------------------------------------------------------
Acidic (L1) Landfill (L2)
-----------------------------------------------------------------
pH 5.0 8.6
(mg/L)
TOC 8772 2272
(CH3-COOH) 3508
NH4+ 819 4575
K+ 538 2300
Na+ 973 4500
Cl- 791 8966
SO42- 209 76
---------------------------------------------------------------- Fig.1 Setup of the column experiment, marks of
TOC: Total Organic Carbon sampling and biofilm formation after 10 months
with landfill leachate (right). M01 from TABLE 2
TABLE 2 Mineralogy of clay materials (mass %), is the 2 cm sample.(For colour figure, refer to CD)
surface area (BET) and hydraulic conductivity (HC).
------------------------------------------------------------------- Although, HC was maintained around the initial
F Q fd Ca Do Py BET k
order of magnitude (10-9- 10-10 m/s) at a depth >20
(m2/g) (m/s)
------------------------------------------------------------------- cm, it is increased significantly at long term, when
Pantoja 75 17 7 <1 <1 <1 42 2·10-10 the highly organic landfill leachate was used (L2).
(10 months) The decrease of specific surface area was very
M01 – L1 78 15 7 - - - 44 1·10-10 important in the case of 48 months Bailén L2 test.
M01 – L2 81 15 4 <1 <1 - 31 5·10-10 In this column it was not possible to obtain a
------------------------------------------------------------------- usable probe for HC measure because of it
Bailén 51 30 2 11 5 0.7 30 2·10-10 disaggregated when was cut. The effect of
(48 months) aggregation of clay agglomerates, that are bonded
M01 – L1 55 30 2 9 4 0.3 30 1·10-9
by organic matter, seems to be responsible for the
M01 – L2 40 38 2 13 5 1.0 16 --*
------------------------------------------------------------------- decrease of the accessible surface area (Cuevas et
F: total amount of sheet-silicates, Q: quartz, fd: al., 2008) and probably for the increase of HC.
feldspars, Ca: calcite, Do: dolomite, Py: Pyrite. L3:
736 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

Bulk mineralogical changes were limited to the

dissolution of carbonates in the first 2 cm with the TABLE 3. pH; Exchangeable Cations (cmol(+)/kg
pH 5 leachate, which was completely neutralized clay); Soluble anions (mmol/L) in the extract
(pH 7.0-8.0; TABLE 3). At 48 months it can be (1:10 clay/distilled water).
followed the preferential dissolution of dolomite --------------------------------------------------------
pH Na+ K+ NH4+ Cl- SO42-
to form calcite (Fig.2). This dedolomitization has --------------------------------------------------------
to do with Ca-sulphate transport and common ion Pantoja 7 1.9 0.7 - 0.9 5.2
effect, and also with sulfate reduction (Castelló et
al., 2008). In this case, the excess of calcium can M01 – L1 8 1.3 5.5 4.6 8.5 4.0
precipitate calcite with the inorganic carbon M01 – L2 9 3.0 4.7 2.9 15.6 5.2
released as a by-product of sulfate-reducing --------------------------------------------------------
microorganisms metabolites. A dark region was Bailén 7 0.1 0.5 - 0.1 4.4
developed at 20-30 cm from the leachate contact
in 10 months and at 40-50 cm in 48 months (Fig. M01 – L1 7 0.2 0.9 2.1 0.4 3.1
M01 – L2 9 1.2 1.7 8.1 0.9 3.7
3), due to sulphate reduction and the precipitation --------------------------------------------------------
of traces of sulphides.
Ammonium, sodium and potassium present in the
6.0
leachate penetrated in the clay at a depth of 10-20
L2
5.5 L1 (2 cm) Delolomitization in L4 cm (35 cm height in Fig. 4) and occupied cation
5.0
Bailén
exchange positions. In contrast, the diffusion of
Dolomite (%)

chloride, which can be considered a conservative

4.5
anion, reached more than 30 cm, virtually the
4.0 Calcite dissolves in acidic conditions whole column.
3.5

60
3.0
9 10 11 12 13 14 15 16
Calcite (%)
50
Fig.2 Calcite-dolomite relationship in Bailén 48 months
clay column height (cm)

experiments.
40

30 NH4
Cl
SO4
20

10

0
0 2 4 6 8
Cl,SO4 mmol/L; NH4 (cmol(+)/kg clay)

Fig.4 Exchangeable ammonium in clay and soluble

Fig.3 Dark patches for reduction of sulfates in the clay anions profiles in 48 months. Note the minumum in
columns. (For colour figure, refer to CD) sulfate related to dark areas in Fig 3 (Bailén).

ACKNOWLEDGEMENTS
This work is supported by the Spanish
Environmental Ministry (MMA: I+D+i
A113/2007/3_02.6) in cooperation with Geotecnia
y Cimientos S.A. (GEOCISA); Centro de
Experimentación de Obras Públicas (CEDEX), and
the Instituto de Ciencias de la Tierra Jaume
Almera (ICTJA).
REFERENCES
Astudillo, J., (2001). El almacenamiento geológico
profundo de los residuos radiactivos de alta
actividad:principios básicos y tecnología.
ENRESA, Madrid, 200 p.
Bilitewski, B., Hardtle, G., Marek, K., Weissbach,
A., Boeddicker, H., (1997). Waste
Management, Springer, Berlin, 699 p.
Brun, A., Engesgaard, P., 2002. Modelling of
transport and biogeochemical processes in
pollutionplumes: literature review and model
development. Journal of Hydrology,256, 211-
227.
Castelló, R., Recio, C., Morillas, P., y Vizcayno,
C., (2008). Pyrite formation driven by MSW
landfill leachate in the Madrid Basin, Spain.
Environmental Geology, 54, 679-688.
Cuevas, J., Leguey, S., Garralón, A., Rodríguez
Rastrero, M., Procopio, J.R., Sevilla, MªT.,
Sánchez, Jiménez, N., Rodríguez Abad, R.,
Garrido, A., (2009). Behavior of kaolinite and
illite-based clays as landfill barriers. Applied
Clay Science 42, 497-509.
Islam, J., Singhal, N. and O’Sullivan, M.O. (2001).
Modeling Biogeochemical processes in
leachate-contaminated soils: A review.
Transport in porous media, 43, 407-440.
Davis, J.A., Kent, D.B., 1990. Surface
complexation modelling in aqueous
geochemistry, in: Hochella, M.F., White, A.F.
(Eds.), Mineral-Water interface geochemistry.
Reviews in Mineralogy 23, pp.177-248.
Hervás et al 737
Michael, A. Malusis, M.A., Shackelford, D.,
(2002). Theory for reactive solute transport
through clay membrane barriers. Journal of
Contaminant Hydrology 59, 291-316.
National Groundwater and Contaminated Land
Center, (2003), Review of ammonium
attenuation in soil and groundwater. NGWCLC
report NC/02/49.
Pivato, A., Raga, R., (2006). Tests for the
evaluation of ammonium attenuation in MSW
landfill leachate by adsorption into bentonite in
a landfill liner. Waste Management, 26, 123-
132.
Rowe, R., K., Quigley, R.M., Booker, J.R., (1995).
Clayey barrier systems for waste disposal
facilities. Spon Press, Abingdon, Oxon. 390 p.
Sawney, B.L., (1996). Sorption and desorption of
organic contaminants by clays and soils, in:
Sawney, B.L. (Ed.): Organic pollutants in the
environment, CMS workshop lectures 8, 45-69.
Savage, D., (1995). The scientific and regulatory
basis for the geological disposal of radioactive
waste. John Willey and Sons Ltd. West Sussex,
437 p.
Steefel, C.I., De Paolo, D.J. and Lichtner, P.C.
(2005). Reactive transport modelling: An
essential tool and a new research approach for
the Earth Sciences. Earth and Planetary Science
letters, 240, 539-588.
Sposito, G., (1990). Molecular models of ion
adsorption on Mineral Surfaces, in Hochella,
M.F., White,A.F. (Eds.), Mineral-Water
interface geochemistry. Reviews in Mineralogy
23, 261-278.
Stumm, W., 1992. Chemistry of the Solid-Water
Interface, John Wiley & Sons, New York, 428
p.
Williams, P.T., 1998. Waste treatment and
Disposal. John Willey and Sons, Chichester,
380 p.
 Adsorptive Chromium Removal Study by Some Clayey Soils
for Abatement of Tannery Waste Pollution
S Ghosh
Jadavpur University, Kolkata, India (sghosh56@yahoo.com)
S N Mukherjee
Jadavpur University, Kolkata, India (snm_ju@yahoo.com)
D Dey Tarafder
Jadavpur University, Kolkata, India (deytarafder.dipankar@rediffmail.com )

ABSTRACT An investigation of laboratory scale was carried out to examine the suitability of different
types of clayey soils collected from a leather tannery site as well as from other nearby area located in
West Bengal, India as a liner material. The leaching characteristics of Cr6+ with respect to different pH
revealed that at lower pH the leachability of Cr6+ is maximum. Batch adsorption kinetics results showed
a reasonable Cr6+ uptake capacity of soil in equilibrium condition. Breakthrough adsorption study also
was carried out in a vertical column which also showed a good metal (Cr6+) adsorption capacity of soil.

INTRODUCTION suitability of different types of clayey soils

One of the major environment problems of tanning
industry is generation of chromium-laden
chemical sludge that is produced as a bye product
of waste water treatment. Since scientific solid
waste disposal practices are seldom employed,
large amount of chromium contaminated sludge is
usually be disposed either in unlined or ill
maintained solid waste dumping grounds/landfills.
The fate of chromium in such disposal sites is a
cause of concern for the leaching effect under
favorable condition. Soil composition (electron
donor availability, soil texture, competitive ions,
adsorption capabilities etc) and conditions in the
soil (pH, moisture content, temperature and
presence of vegetation) are the principal factors
affecting adsorption, leachability and mobility of
chromium (Zachara et al. 1989, Hanson et al.
1993). These literatures entail that the clay can be
used as a liner material for containment of heavy
metals generated in leachates. In this context there
is a need to design a properly compacted clay liner
for the containment of the waste at disposal sites
so that leachate generated from that source,
containing appreciable amount of chromium, can
be retarded while migrating through fine grained
soil from waste disposal site. In the view of this
concerning problem as stated the present study
was undertaken with the objective to asses the
collected from different areas of plant and
neighbouring areas of a tannery industry for using
as a liner material in a landfill near tannery waste
disposal site to prevent chromium contamination
in the surrounding lithospheric environment.
MATERIAL AND METHODS
Leather complex at Bantala in South 24 Parganas
district of West Bengal, India was considered the
proposed site for the present investigation. Soil
was also collected from the campus of Jadavpur
University, Kolkata, India for carrying out the
necessary studies.
Leachate Characteristics
Waste water and chemical sludges were collected
from the plant site and relevant physico-chemical
parameters were analyzed in the civil engineering
laboratory in accordance with Standard methods
(APHA et al. 1995). All heavy metals were
analyzed using the instrument atomic absorption
spectrophotometer (Per kin Elmer, make USA,
model Varian 50 BIO). The samples were filtered
and digested with concentrated HNO3 for 4 hours
for getting accurate result. The concentration of
chromium found both in waste water and sludge
were found to be very high compared to other
heavy metals and exceeding the permissible limit.

Thus in the present study, chromium was used as
the test adsorbate metal. The following laboratory
tests were conducted for determining the
characteristics of soil samples collected at 2m
depth from the ground level. Following
geotechnical parameters were determined for
characterization of soil, hence for the adsorbing
medium too. i) Moisture content, ii) Bulk density,
iii) Particle size distribution, iv) Atterberg limits,
v) Optimum moisture content, vi) Permeability
(Falling Head), vii) Cat ion exchange capacity.
Batch Adsorption Studies
The Batch Kinetic test was carried out to
determine the kinetics and equilibrium of
adsorption process. The rate at which adsorption
equilibrium takes place dictate the contact time
between the adsorbent and solution which is
required to be provided for optimum adsorption.
In batch adsorption test, each 100ml of sample
containing 1 mg/L of chromium and 10 mg/L of
field soil was taken in six different polythene
bottles and kept in a horizontal shaker (speed 100
rpm). Every half an hour interval, one bottle was
taken out from the shaker and the supernatant was
analyzed for the residual concentration of
chromium (Cr 6+) after filtering it through filter
paper (Whatman 42). Such tests were repeated
with similar time intervals and with different
initial concentrations of chromium and with
different soil quantities.
Isotherm Studies
The adsorption of substance from one phase to the
surface of other in a specific system leads to a
thermodynamically defined distribution of those
substances between the phases when the system
reaches equilibrium i.e. when ceasing of sorption
occurs. The common manner, in which it is
expressed, is the amount of sorbate adsorbed per
unit weight of sorbent (X/M) as a function of
residual equilibrium concentration C of substances
remaining in solution phase. An expression of this
type, termed as adsorption isotherm defined the
functional equilibrium distribution of adsorption
with concentration of adsorbate in the solution at
constant temperature. Experimental isotherms are
Ghosh, Mukherjee and Tarafder 739
useful for describing sorption capacity to evaluate
the suitability of adsorbent. Sorption kinetics were
studied for chromium concentration of different
initial concentrations for 2, 3, and 4 mg/L, with
different amounts of soil ( 300, 400, 500, 600, 700
g/L) , and different time intervals e.g. 30 , 60 , 90 ,
120 , 150 , and 180 min. The specimens were
placed in a 250 ml glass bottle in which the first
soil sample and then chromium solution was
added for sorption mechanism. The 250 ml glass
bottle was placed on a rotary shaker for necessary
agitation @ 100rpm speed and samples were taken
at different time intervals and the supernatant was
analyzed for chromium adsorption. The minimum
time after which there was no increase in the
adsorption rate of chromium was taken as
equilibration time, and the concentration of
chromium remaining at that time in the
supernatant is considered equilibrium
concentration. The isotherms were drawn with
equilibrium concentration (mg/L) in x-axis and the
ratio of amount of chromium removed to amount
of soil in y-axis for a specified concentration, both
in Langmuir and Freundlich isotherm models.
Column Breakthrough Studies
Column study was carried out to obtain
breakthrough curves by evaluating some time
concentration parameters. In this test a column of
diameter 115 mm and height 580 mm made of
steel was used (Fig.1). Soil more or less
homogeneous in nature was brought from the site
and was placed in the column and compacted in a
layer of height of 550 mm with relative
compaction 85% with respect to standard proctor
density and was saturated with distilled water for
24 hours. The chromium spiked solution of
concentration 7.5 mg/L, was introduced to column
from the top. The solution was allowed to pass
through the soil at a steady flow of 0.27 ml/min.
The samples were collected from the outlet of the
column installed at the bottom at different
intervals for measuring the residual Cr6+
concentration. The effluent concentration was also
measured with respect to time intervals at middle
point of soil column.
740 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

RESULTS AND DISCUSSION Soil – B removal adsorption occurred within first

90 minutes, thereafter the rate of uptake decreased
sharply. After a contact period of 120 minutes the
Test Results on Soil Quality curves become almost asymptotic; therefore 120
The results of some relevant physical properties of minutes of contact was considered as the
two soil samples namely Soil – B (collected from equilibrium time for batch experiment. For Soil –
plant site at Bantala) & Soil – J (collected from J, the removal kinetics show approximately 75-
university campus) are presented in Table 1. Both 90% of the total adsorption occurred within first
these soils have been collected from a depth of 100 minutes, thereafter the rate of removal was
2m from the existing ground level. The data reveal decreased.
that the soils were primarily of cohesive in nature,
with high silt content and low permeability value TABLE 1 Test results on relevant physico-
which is very much closer to the value given by chemical properties of soils.
Benson (1994) (1 x 10-7 cm/s).Hence the soils
possess a reasonable property against the Parameter Soil -B Soil - J
transmission of leachate to the surrounding ground Bulk density (kN/m3) 18.3 18.7
water resources due to low permeability. Optimum moisture
17.20 15.80
content (%)
Maximum dry density
16.80 17.00
(kN/m3)
Water content (%) 23 29
Liquid limit (%) 44 42
Plastic limit (%) 19 19
Co-efficient of
2.7x10-7 3.85x10-7
permeability (cm/s)
Sand(%) 6 7
Grain size
Silt(%) 62 64
distribution
Clay(%) 32 29
Cation-exchange
1.9 41.144
capacity (meq/100gm)
pH 7.64 7.73

After 120 minutes the curve become asymptotic;

Fig.1 Experimental Set-up for Column Kinetics
Study.
Batch adsorption studies
Fig.2 and Fig.3 show the typical batch kinetics
profile on the removal of Chromium (hexavalent)
during the time course study against different
initial concentration of soil suspensions for Soil –
B and Soil – J respectively. The above figures
demonstrate a set of identical patterns Cr6+ uptake
kinetics, and show approximately 75-90% of the
total adsorption. by varying amount of soil. For
therefore 120 minutes contact time was considered
to be as the equilibrium time for batch experiment.
It may be concluded that both types of soils were
able to remove Cr6+ in significant amount to the
extent of 75-90%. The plot also shows that the
variation of percent removal of chromium with
time. Fig.4. and Fig.5 show the plot of equilibrium
% removal of Cr 6+ with soil dose for Soil – B and
Soil – J respectively. In both the cases the
optimum dose of soil was found to be 700 g/L for
initial concentration range of chromium between 2
to 4 mg/L. It is also noted that from Fig.6 and
Fig.7, that the removal of chromium decreases
 Ghosh, Mukherjee and Tarafder 741

100 100

E q u ilib r iu m % R e m o v a l o f C r 6 +
90 Initial Concentration of Cr 6+ = 3 mg/L. 90
80 80
% R em ov al of C r 6+

70 70
60 60
50 50
40 40
Soil Dose 300 g/L
30 Soil Dose 400 g/L 30 Initial Concentration of Cr 6+ = 2 mg/L
20 Soil Dose 500 g/L 20
Soil Dose 600 g/L Initial Concentration of Cr 6+ = 3 mg/L
10 Soil Dose 700 g/L 10 Initial Concentration of Cr 6+ = 4 mg/L
0 0
0.00 20.00 40.00 60.00 80.00 100.00 120.00 140.00 160.00 180.00 200.00 0 100 200 300 400 500 600 700 800 900
Contact Time, Minute Soil Dose ( g/L)

Fig.2 Batch Kinetics Test of Fig.4 Equilibrium % Removal of Cr6+ vs Soil Dose of
Soil - B. Soil – B

100 90

E q u ilib riu m % R e m o v al o f C r6 +
90 Initial Concentration of Cr 6+= 3 mg/L. 80
80 70
% R em o v al o f C r 6 +

70 60
60 50
50
40
40 Soil Dose 200 g/L
30 Soil Dose 300 g/L 30
Soil Dose 400 g/L Initial Concentration of Cr 6+ = 2 mg/L
20 Soil Dose 500 g/L 20
Initial Concentration of Cr 6+ = 3 mg/L
10 Soil Dose 600 g/L 10
Soil Dose 700 g/L Initial Concentration of Cr 6+ = 4 mg/L
0 0
0.00 20.00 40.00 60.00 80.00 100.00 120.00 140.00 160.00 180.00 200.00 0 100 200 300 400 500 600 700 800 900
Contact Time, Minute Soil Dose ( g /L)

Fig.3 Batch Kinetics Test of Fig.5 Equilibrium % Removal of Cr6+ vs Soil Dose of
Soil - J. Soil - J.

sharply for change in initial concentration from 2. 100

to 3 mg/L, thereafter only a marginal removal is
% R e m o v a l o f C h ro m iu m

90
observed both for Soil – B & Soil – J respectively.
However, the rate of removal had been continued 80

even the initial concentration of chromium 70

exceeds 4 mg/L for both the soils with lesser 60
uptake rates. The percent removal was found to be
50
varied between 75 - 90% for both these soils. The
experimental data also reveals that, initially all of 40
0 1 2 3 4 5
the adsorption sites of the soil media were vacant Initial concentration of Cr 6+ (mg/L)
and active, for which, because of the high solute
concentration gradient, there was high affinity Fig.6 Influent Concentration vs % Removal of
between the solute and the soil mass interaction as Chromium for Soil - B.
as an adsorbent. The predominant phenomenon of
742 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

R e m o v e d /A m o u n t o f S o il, m g /g
100 1
6+
Initial Concentration of Cr = 3 mg/L.
% R e m o v a l o f C h r o m iu m

A m o u n t o f C h r o m iu m
90
y = 0.0033x + 0.0019
80 0.1 2
R = 0.9694
70

60 0.01

50

40 0.001
0 1 2 3 4 5 0.1 1 10
Initial concentration of Cr 6+ (mg/L) Equlibrium Concentration(Ce), mg/L.

Fig.7 Influent Concentration vs % Removal of Fig.9 Freundlich Isotherm from Batch Kinetics of
Chromium for Soil - J. Soil - B.

the adsorption of Cr6+ is perhaps due to the 0.02

R e m o v e d /A m o u n t o f S o il,m g /g
resulting fact of exchange adsorption or ion Initial Concentration of Cr6+= 3 mg/L.

A m o u n t o f C h ro m iu m
exchange. 0.015 y = 0.0032x + 0.0014
R2 = 0.9757
Isothermal studies 0.01

The adsorption isotherm is plotted using the

0.005
experimental result as obtained from time
concentration study using various doses of soil 0
with a typical single initial concentration and 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
equilibrium concentration obtained after a Equilibrium Concentration(Ce), mg/L.
specified time period. The test results are plotted
in both Langmuir and Freundlich isotherm models, Fig.10 Langmuir Isotherm from Batch Kinetics of
as shown in Fig. 8 through Fig. 11 for Soil – B & Soil - J.
Soil – J respectively. However Langmuir isotherm
models fit the experimental data more reasonably The average value of ‘n’ for Soil – B is 0.0029 and
(r2 = 0.97). Both these models show reasonably for Soil – J is 0.0028. Low value of ‘n’ (Slope of
good agreement as evident from the above figures. the line) reveals that both the soil has high rate of

0.02 1
Initial Concentration of Cr 6+ = 3 mg/L.
R e m o v e d /A m o u n t o f S o il, m g /g
R e m o v e d /A m o u n t o f S o il, m g /g

Initial Concentration of Cr6+ = 3 mg/L.

A m o u n t o f C h r o m iu m
A m o u n t o f C h r o m iu m

0.015 y = 0.0033x + 0.0019 y = 0.0032x + 0.0014

0.1
R2 = 0.9694 R2 = 0.9757
0.01

0.01
0.005

0 0.001
0 0.2 0.4 0.6 0.8 1 1.2 1.4 0.1 1 10
Equlibrium Concentration(Ce), mg/L. Equlibrium Concentration(Ce), mg/L.

Fig.8 Langmuir Isotherm from Batch Kinetics of Fig.11 Freundlich Isotherm from Batch Kinetics of
Soil – B. Soil - J.
 Ghosh, Mukherjee and Tarafder 743

chromium adsorption capacity. Because of low indicates good resistance against transmission
value of ‘n’ (Slope of the line) the retardation co- of leachate to ground water.
efficient Rd, will also be migration also would be x The mechanism of chromium attenuation was
less which corroborates high chromium adsorptive primarily due to adsorption by the soil as
capacity of the test soil. indicated by adsorption isotherm studies.
x From adsorption isotherm model (Langmuir),
Breakthrough study from Vertical Column it was found that r2 = 0.97 and low value of
Test ‘n’ (slope of the best fitted line) i.e. n =
Fig.12 shows the variation of concentration ratio 0.0029 and n = 0.0028 for Soil – B and Soil –
(i.e. C/C0) with time for 7.5 mg/L of initial J respectively, which reveals good adsorption
chromium concentration (C0), for a constant head capacity of these soils.
of solution containing Cr6+ in a vertical column. x Vertical column test results show that the
Initial breakthrough is observed after 700 hours, equilibrium time is achieved after 900 hours.
when 7.1 mg/L (C/C0 = 0.9) of chromium is found
in the effluent. The equilibrium concentration has REFERENCES
reached after 900 hours at which approximately
98% of the initial concentration has been traced in American Public Health Association, American
the effluent solution, which indicates exhaustion Water Works Association, Water Environment
of the soil column. Federation (1995) Standard Methods for the
Examination of Water and Wastewater. 15th ed;
Initial concentration of Cr 6+ = 7.5 mg/L American Public Health Association:
1
Washington, D.C.
0.8

0.6
Effluent Point Benson C., Zhai H. and Wang X. (1994)
C /C o

Middle point
Estimating Hydraulic Conductivity of
0.4
Compacted Clay Liners. Journal of
0.2 Geotechnical Engineering, ASCE, 120, 366-387.
0
0 100 200 300 400 500 600 700 800 900 1000 Chen J., Anandarajah A. and Inyang H. (2000)
Time (hour)
Pore fluid properties and compressibility of
kaolinite. Journal of Geotechnical and
Fig.12 Breakthrough Curve from Vertical Column Test
with Soil – B.
Geoenvironmental Engineering, ASCE, 126(4),
798-807.
CONCLUSIONS
Following conclusions are drawn on the basis of Hanson A. T., Dwyer B. D., Samani Z. A. and
experimental test results and analytical exercises. York D. (1993) Remediation of chromium
containing soils by heap leaching: a column
x In the present work Cr6+ was found as toxic
study. Jl.of Environ. Eng. 119, 824-841.
metal in the sludge and waste material
emanated from a tannery unit using chromium
Kozuh N., Stuper J. and Gorenc B. (2000)
as tanning agent, which is satisfactorily
Reduction and oxidation processes of chromium
attenuated by the soil sample taken from the
in soils. Environ. Sci. Technol. 34, 112-119.
two sites.
x The field soil samples have very low
Zachara J. M., Ainsworth C. C., Cowan C. E. and
permeability and possess a marginally, higher
Resch C. T. (1989) Adsorption of chromate by
value than the recommended one, which
subsurface soil horizons. Soil Sci. Soc. Am. J,
53, 418-428.
 Thermally Obtained Carbon as a Potential Cr(VI) Adsorbent
Zhenze LI
(GSGES, Kyoto University, JAPAN; lazyhero@live.cn)
Shigeyoshi Imaizumi
(Dep. Advan. Interdiscip. Sci., Utsunomiya University, JAPAN; imaizumi@cc.utsunomiya-u.ac.jp)
Takeshi Katsumi
(GSGES, Kyoto University, Japan; katsumi.takeshi.6v@kyoto-u.ac.jp)
Xiaowu Tang
(Dep. Civil Eng., Zhejiang University, China; tangxiaowu@zju.edu.cn)

Abstract: Chromium (VI) is a common toxicant existed in fly ashes of coal, municipal solid waste and
hazardous waste etc. The leaching potential of Cr(VI) in these ashes is so high that the leachate collected
from the landfill site has caused intensive interests and attention to get rid of the contained pollutant. This
paper attempted to change pine wood chip into a stable and durable carbon whose applicability in Cr(VI)
adsorption was further testified by Batch adsorption tests. Preliminary results showed that the preparation
conditions had a significant influence on the final Cr(VI) adsorption performance. This study provides a
novel and in-depth perspective on available management techniques for Chromium attenuation and
retardation in the soil strata.

1 INTRODUCTION 2 MATERIALS AND METHODS

Cr(VI) is widely used in industries including
leather tanning, furnace cooling, plating, rinse waters,
wood preservation (Jorge et al., 1999) etc.
Hexavalent chromium could be reduced by
sulfur-containing proteins and then combined with
nucleic acids as a mutagenic toxicant. (Mali et al.,
1964; Samitz et al., 1965) Cr could stablely exist in
the environment at both Cr(III) and Cr(VI) species.
Cr(III) is susceptible to precipitating in soil strata
while Cr(VI) is mobile and is always leached out by
ground water.
Carbons obtained from waste bio-resource provide
an economic and sustainable alternative for
wastewater treatment (Nakajima & Baba, 2004;
Chauhan et al., 2008; Bansal et al., 2009). Although
promising in wastewater treatment, the high tendency
of biological decomposition of biosorbents limits
their application into the in situ treatment of Cr(VI)
polluted ground water. Activated carbon is
chemically stable and thus could serve as an
alternative in permeable reactive barriers.
This paper aims to prepare Cr(VI) adsorbents by
transforming natural wood into carbons through
different routines under the lowest energy
consumption. The Cr(VI) adsorption performances of
various obtained carbons were evaluated using Batch
test, considering pH and initial Cr(VI) concentration.
The relevant mechanisms were also discussed using
scanning electron microscopy (Hitachi FE-SEM
S-4500, Japan).
2.1 Preparation and Characterization
Shaved pine wood chip was oven dried at 105 oC
for 24 h, cooled to room temperature and then stored
in sealed plastic bags. A piece of wood chip was
filled in a glass bottle whose temperature was
increased to target temperature with an increment of
about 20 oC per minute and then pertained for 0.5 h.
The obtained pine char was pulverized after cooled to
room temperature. The fine particles passing through
a 0.25 mm mesh was collected for further test.
Morphologies of the carbons were obtained using a
SEM facility (Hitachi SEM- S 4500). The surface
functional groups were analyzed by EDX facility.
2.2 Isothermal Adsorption Test
A piece of sorbent (0.02 ± 0.001 g) was transferred
to a plastic centrifuge tube. Cr(VI) solutions (20 mL,
dissolved K2CrO4 in analytical grade) with initial
concentrations at 1, 2, 5, 10, 25 mg/L were separately
poured into the centrifuge tubes. The tube was tapped
and placed in a shaking bed (150 round per minute)
and then equilibrated for 24 h. The slurry mixture
was filtered with a 0.45 um membrane filter and the
clear water was subject to Inductively Coupled
Plasma test (ICP) to determine the equilibrium
concentration of Cr in total.
2.3 Effect of pH on Cr(VI) adsorption
Similar to the isothermal adsorption test, the
dosage and initial Cr(VI) concentration were adjusted
at 1 g/L and 5 mg/L, respectively. The volume of
 Li et al. 745

solution was first supplied as 90 % of the total

required volume. Then the solution pH was adjusted
by adding 0.01-0.1 M HNO3 or NaOH solutions. The
total volume was controlled at 20 mL after the pH
was adjusted to the target. The mixtures of sorbent
and Cr(VI) solution with increasing pHi from 2 to 6
at an increment of 1.0, were equilibrated for 24 h
before being filtered for ICP test.

3 RESULTS AND DISCUSSION

3.1 Characterization

The surface chemical composition can be

quantified from O/C atomic ratio by X-ray
photoelectronic spectroscopy (XPS) (Buchert et al., (a)
1996). SEM-EDX could also work like XPS and has
been utilized to characterize the chemical
composittion of wood (Butterman & Castaldi, 20 007).
Table 1 showed thaat the percen ntage of oxy ygen
determineed by differeent method deeviated obvio ously
from eacch other. Bassically, the deecomposed wood
w
char has less oxygen percent
p than the natural wood.
w
With incrreasing activaating temperaature from 2000 to
300 oC, the
t oxygen percent
p decreaased continuo
ously
from 2.54
4 to 1.73 %, correspondinglly.

Table 1 Change of oxygen

o ratio on
o the adsorbeents’
surface
Oxygen
atomic Acttivation
Sorbent Analysis
A
concentratio tem
mperatur Referrence
s f
facility (b)
n at e oC
Figure 1 SEM photos of prepareed carbons (a: C1
1; b: C2)
surface %
SEM-ED
S
C1 2.54 0
200 This study Heat
H treatmennt upon woood by Gerard din et al.
X
S
SEM-ED resu
ulted in a signnificant decreaase of the O/C
C ratio for
C2 1.73 300
0 This study
X bothh species. Affinity
A of wood for water is
Wood
(Buttterma connsiderably reduuced as the reesults of hemiicelluloses
chips
SEM-ED
S n & deggradation. Coonsequently, the electron n-donating
(maple 1.8-4.5
X Castaaldi,
and commponent is sttrongly modiffied by heat treatment
2007))
poplar)
(Geerardin et al., 2007). Carboons produced at higher
Pine (Geraardin
wood 34.6 X
XPS et al., tem
mperature will show strongeer hydrophobiicity. This
chip 2007)) is also the reasonn why the carbbon always neeeds to be
Heated activated by steamm and salt.
(Geraardin
pine
wood
28.6 X
XPS et al., Figure
F 1 shows the SE EM of the obtained
2007)) adsoorbents. Diffferent tempperature ressulted in
chip
sign
nificant differrence in micrrostructure off carbons.
746 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

C1 has macroporre with diameeter at 4-5 um m. In these improving the adsorption performance of natural
maccropores, abun ndant mini-rod sized at 400 0 nm long biomasses. Our previous studies have proved this
andd diameter 100 0 nm. This feeature is not present in speculation. By transforming natural biomaterials
anyy other sorbentts prepared inn this study. C2
C appears such as leaf, grass etc. into chars we could obtain
to be
b a complex of o constituentss with differennt features: enhanced adsorption capacity and affinity towards
one kind is brok ken particles sized less thaan 10 um, heavy metal cations (Li et al., 2009). This proposal
andd the other is wedge-shapped plates with w sharp needs further in-depth investigations to get validated
edgges, sized largeer than 100 umm. The conneccted pores and optimized for engineering application.
willl give rise to its pore voluume and the adsorption
a
capacity. However, no obviouus pores weree observed 3.3 Effect of pH
in C3,
C which will w limit its ssorption behaavior. The Figure 3 shows the effect of pH on Cr(II)
minni-rod in C1 will
w work as aan reductant for f Cr(VI) adsorption percentage on the adsorbents. From this
as it is formed by the preccipitated orgaanics that figure, the decreasing change of Cr(VI) removal
evap porate from the inside of wood. Th his could percentage could be observed with increasing
explain the featurre that C1 couuld adsorb Crr(VI) even equilibrium pHs. The obtained sample C1 appeared
at pHs
p between 6 and 7. Otherr reported dataa is absent to show the best performance among all the samples.
or leess than C1 within
w this rangge, as shown in
i Fig. 3. The reported carbon shown in Fig. 3, cited from the
study of Mohan et al. (Mohan et al., 2005), was the
8
best sorbent obtained by activating coconut fibers at
C1 600 oC for 1 h.
C2 Figure 3 also shows the species distribution curves
6 W
Wood chip of Cr(VI) at pHs < 7.0. It is obvious that HCrO4-
dominates at pHs < 5.0, while percent of CrO42-
Cs mg/g

4 increased steadily from 0 % at pH 4.0 to 75 % at pH

7.0. All the reported data was approaching 0 at pH >
6.0, except those of C1 in this study.
2
100 100
-
HCrO4
Cr(VI) Removal Percentage %

C1
0
80 C2 80
0 5 10 15 20 25 Reported Tannin gel
Reported carbon
Ce mg
g/L
60 60
Fiigure 2 Adsorpttion isotherm of C
Cr(VI) on various adsorbents
(pHi | 3.33) (For colour figure, refer to CD) 40 40
2-
CrO4

3.2 Adsorption isotherm
F
Figure 2 showws the adsorpttion isotherm of Cr(VI)
on various adsoorbents. The isotherm forr C1 was
hypperbolic curves, while thosee for C2 and wood
w chip
werre linear. As too the locations of the isotheerms, they
couuld be sequencced as C1>C22>Wood chip p. Overall,
the affinity of vaarious sorbents towards Cr((VI) could
be sorted as C1> >C2 >>woodd chip. The wood w char
obtaained at comp
mparatively low w temperaturre such as
200 oC appeared to be better than that at high
temperature. Besides, no matter what the treatment
temperature was, the chars showed increased
adsorption affinity towards Cr(VI) in comparison of
the natural wood chip. This behavior confirmed the
effectiveness of the activation treatment method on
20 20
0 0
2 3 4 5 6 7
pHe
Figure 3 Variation of Cr(VI) removal percentage with solution
pHe (Tannin gel was reported by Nakajima and Baba (Nakajima &
Baba, 2004); the cited activated carbon was the best result reported
by Mohan et al. (Mohan et al., 2005); dashed line indicating the
species distribution of Cr(VI) was predicted by Visual MINTEQ
software) (For colour figure, refer to CD)
The Cr(VI) adsorption affinity of C1 was proved to
be stronger than any other adsorbents. At pH > 5.0,
the oxidizing potential of HCrO4- has greatly
decreased due to the lack of H+ (Nakajima & Baba,
2004; Mohan & Pittman Jr, 2006). However, C1
could still retain Cr(VI) and even maintain a stable

( )
level of Cr(VI) removal around 10 % at pH > 4.0.
This feature could be attributed to the special
structure as shown in Fig. 1.
Figure 4 shows the variation of pHe with pHi after
adsorbent slurry equilibrated with Cr(VI). The pHe
is larger than pHi at pH < 3, but turns to be less than
pHi at pHs • 3.0. The difference is not so large that
all the data distributed near the equal line between
pHi and pHe.
7 on pine needles based hydrogels. Bioresource
6 P. Gerardin, M. Petric, M. Petrissans, J. Lambert &
5
pHi
4
3 1999. Reactions between Cr(VI) and wood and its
2 33, 487-499.
2 3 4 5 6 7
pHe
Figure 4 Variation of pHe with pHi after adsorbent slurry
equilibrated with Cr(VI)
(For colour figure, refer to CD)
4 CONCLUSIONS
The thermal treatment decreased the particle size
and the atomic oxygen percentage of the pine wood
chip while producing macropores of various shapes
and dimensions. Carbons prepared at low
temperatures (200 - 350 oC) showed enhanced Cr(VI)
adsorption capacity from aqueous solution at low
solution pHs. The carbon prepared at much lower
temperatures proved to be a better choice for Cr(VI)
removal (10 % at neutral pH condition) than the
reported carbons activated at higher temperatures
(carbonized at > 600 oC, activated by steam with
specific salts).
References
M. Bansal, D. Singh & V.K. Garg. 2009. A
comparative study for the removal of hexavalent
chromium from aqueous solution by agriculture
wastes' carbons. Journal of Hazardous Materials,
171, 83-92.
Li et al. 747
J. Buchert, G. Carlsson, L. Viikari & G. Stream. 1996.
Surface Characterization of Unbleached Kraft
Pulps by Enzymatic Peeling and ESCA.
Holzforschung, 50, 69-74.
H.C. Butterman & M.J. Castaldi. 2007. Influence of
CO2 Injection on Biomass Gasification. Industrial
& Engineering Chemistry Research, 46,
8875-8886.
G.S. Chauhan, S. Chauhan, S. Kumar & A. Kumari.
2008. A study in the adsorption of Fe2+ and NO3-
Technology, 99, 6464-6470.
J.J. Ehrhrardt. 2007. Evolution of wood surface
free energy after heat treatment. Polymer
Degradation and Stability, 92, 653-657.
F.S. Jorge, T.M. Santos, J.P. de Jesus & W.B. Banks.
model compounds. Wood Science and Technology,
Z. Li, X. Tang, Y. Chen, L. Wei & Y. Wang. 2009.
Activation of Firmiana Simplex leaf and the
enhanced Pb(II) adsorption performance:
Equilibrium and kinetic studies. Journal of
Hazardous Materials, 169, 386-394.
J.W.H. Mali, W.J. Van Kooten & F.C.J. Van Neer.
1964. Some Aspects of the Behavior of
Chromium Compounds in the Skin. Journal of
Occupational and Environmental Medicine, 6,
241.
D. Mohan & C.U. Pittman Jr. 2006. Activated
carbons and low cost adsorbents for remediation
of tri- and hexavalent chromium from water.
Journal of Hazardous Materials, 137, 762-811.
D. Mohan, K.P. Singh & V.K. Singh. 2005. Removal
of Hexavalent Chromium from Aqueous Solution
Using Low-Cost Activated Carbons Derived from
Agricultural Waste Materials and Activated
Carbon Fabric Cloth. Industrial & Engineering
Chemistry Research, 44, 1027-1042.
A. Nakajima & Y. Baba. 2004. Mechanism of
hexavalent chromium adsorption by persimmon
tannin gel. Water Research, 38, 2859-2864.
M.H. Samitz, S. Katz & R. A. 1965. A Study of the
Chemical Reactions Between Chromium and Skin.
Journal of Occupational and Environmental
Medicine, 7, 243.
 Leaching Property of Cr(VI)-Contaminated Soil by Mixing Useful
Microorganisms

K. Omine
Kyushu University, Fukuoka, Japan (oomine@civil.kyushu-u.ac.jp)
N. Yasufuku
Kyushu University, Fukuoka, Japan (yasufuku@civil.kyushu-u.ac.jp)
K. Tamura
Kyushu University, Fukuoka, Japan (Former student)
C. Edward Raja
Kyushu University, Fukuoka, Japan (edumku@yahoo.co.in)

ABSTRACT This paper describes the effect of microorganisms on leaching property of Cr(VI)-
contaminated soil. First of all, effect of mixing microorganisms such as lactic acid bacteria, yeast or Bacillus
natto for Cr (VI)-contaminated water was discussed in the present study. It was found that concentration of
Cr (VI) from the contaminated water decreases with increases in the amount of microorganisms. Secondly,
leaching property of Cr (VI)-contaminated soil by fermentation of organic matter was confirmed by the test
results. It was also concluded that the anaerobic fermentation of organic matter was effective for reducing
concentration of Cr (VI) from the contaminated soil.

INTRODUCTION the Soil Contamination Countermeasures Law. It

Hexavalent chromium is a well-known highly toxic
metal, considered a priority pollutant. Industrial
sources of Cr (VI) include leather tanning, cooling
tower blowdown, plating, electroplating and
anodizing baths. Hexavalent chromium dissolves
in water as unstable material and oxidizes an
organic matter in an acid condition, and then it
resolves itself into trivalent chromium. It is said
that a humus material affects an oxidation-
reduction of chromium (Nakayasu et al.1999). It is
also reported that concentration of hexavalent
chromium in a chrome slag ruin was restrained
largely by disseminating wood vinegar or bran
vinegar (Watanabe 2008).It is therefore considered
that harmlessness of hexavalent chromium may be
possible by the reduction action of various organic
matters.
Many contaminated soils containing heavy metals
have been detected in Japan, because the pollution
investigation on vacant sites of factories and
groundwater investigation are promoted based on
is said from the investigation of the soil
environmental center that 320,000 places of
contaminated sites are needed to purify in our
country. A brownfield place remaining untouched
due to high cost of soil remediation becomes
social problem. It is therefore that a low cost
purification technique with low environmental
load is required for the remediation of heavy
metals.
In this study, the effect of mixing microorganisms
for Cr (VI)-contaminated water is investigated
based on the test results. Furthermore, from a
viewpoint of composting of organic matter, the
aerobic and anaerobic fermentations are used for
reducing a concentration of Cr (VI) from the
contaminated soil (Omine et al. 2009). The
reducing effect of Cr (VI) was clarified from the
leaching test.

EFFECT OF MIXING MICROORGANISMS ON
CR(VI)-COMTAMINATED WATER
Samples and Test Method
Heavelant chromium (2 mg/L) was dissolved in
distilled water and used for the preparation of
artificial contaminated water. The culture
solutions were prepared by mixing yogurt in 25 g,
dry yeast in 2 g or a grain of fermented soybeans
individually into water in 450mL with sugar in 25
g were incubated at 35ºC for a week. The
commercially produced yogurt, dry yeast and
fermented soybeans are used and selected for the
production of microorganisms such as lactic acid
bacteria, yeast and Bacillus natto are incubated as
mentioned above. Each culture solution with lactic
acid bacteria, yeast or Bacillus natto in 20 mL was
mixed into the Cr (VI)-contaminated water in 200
mL. The mixtures are wrapped under anaerobic
condition for 6 or 14 days. After that incubation,
the values of pH, ORP and ATP (adenosine
triphosphate) and concentration of Cr (VI) in the
solutions were measured. ATP uptake and
concentration of Cr (VI) were measured by using
Kikkoman’s Lumitester C-110 with luciferase
(enzyme derived from fireflies) and portable
spectrophotometer (DR2800), respectively.
Effect of Microorganisms
Lactic acid bacteria and yeast are facultative
anaerobes that can survive in the absence as well
as in the presence of atmospheric oxygen. Bacillus
natto is a type of Bacillus subtilis which are
aerobic microbe. These are useful and
environmental-friendly microorganisms.
The test results on the Cr (VI)-contaminated
Table 1 Test results on the Cr(VI)-contaminated water with the culture solutions
pH ORP (mV)
Sample
6 days 14 days 6 days
Lactic acid bacteria 4.00 4.07 165
Yeast 4.52 5.69 135
Bacillus natto 4.26 4.49 151
Omine et al. 749
water by mixing the culture solutions are shown in
Table 1. The pH of each sample indicates acidity
in a range of 4~5 and ORP is in the range of 70 to
170 mV. According to the results, these samples
exist as trivalent chromium in an oxidation
reduction potential was determined and shown in
figure.
Relationship between the reduction effect of Cr
(VI) and a quantity of microorganisms on each
sample was discussed. From the results ATP is a
multifunctional nucleotide, and plays an important
role in cell biology. It acts as a coenzyme that is
"molecular unit of currency" of intracellular
energy transfer. Relationship between ATP and
curing time on the Cr (VI)-contaminated water by
mixing the culture solutions are shown in Fig.1.
ATP in the contaminated water with the culture
solutions increases with curing time and the value
is larger in order of yeast, lactic acid bacteria and
Bacillus natto at the same curing days.
The change of concentration of Cr (VI) for curing
time on the contaminated water with the culture
solutions are shown in Fig. 2. The maximum
concentration of Cr (VI) in the contaminated water
is 2 mg/L. It is found that the concentration of Cr
(VI) have been decreased by mixing the
microorganisms such as lactic acid bacteria or
yeast. Yeast has higher effect on decrease in
concentration of Cr (VI). There is a tendency that
the concentration of Cr (VI) decreases with
increase in curing time independent on the
selected microorganisms. Figure 3 shows
relationship between the concentration of Cr (VI)
and ATP on the contaminated water. There is a
linear relation between Cr (VI) and ATP. It is
ATP (mol/L) Cr(VI) (mg/L)
14 days 6 days 14 days 6 days 14 days
-10 -9
161 6.309×10 7.407×10 1.11 0.79
72 9.197×10-10 2.327×10-8 0.85 0.437
-11 -10
144 2.169×10 1.649×10 1.396 1.024
750 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

-7
10 1.4

Concentration of Cr(VI) (mg/L)

-8
1.2
10 Yeast
ATP (moL/L)

1
-9
10 Lactic acid bacteria
0.8

-10
10 Lactic acid bacteria
0.6
Yeast
Bacillus natto Bacillus natto
-11 0.4
10
-11 -10 -9 -8 -7
6 8 10 12 14 16 10 10 10 10 10
Curing time (days) ATP (mol/L)

Fig. 1 Relationship between ATP and curing time on Fig. 3 Relationship between Concentration of Cr(VI)
the contaminated water and ATP on the contaminated water
(For colour figure, refer to CD) (For colour figure, refer to CD)

1.6 LEACHING PROPERTY OF CR(VI)-

COMTAMINATED SOIL BY FERMENTAION
Concentration of Cr(VI) (mg/L)

1.4
Bacillus natto Samples and Test Method
1.2
Lactic acid The effect of mixing the microorganisms into
1 bacteria Cr(VI)-contaminated artificial soil was
investigated. The artificial contaminated soil was
0.8 prepared by mixing a hexavalent chromium
Yeast
solution (K2Cr2O7) into granite decomposed soil.
0.6
The standard solution with concentration of 50
0.4 mg/L was used for the analysis. The hexavalent
chromium 5 mg was weighed, and added into 500
0.2 g of granite decomposed soil.
6 8 10 12 14 16
Curing time (days) Test conditions of samples with aerobic and
Fig. 2 Relationship between Concentration of Cr(VI) anaerobic fermentations are listed in Tables 2 and
and curing time on the contaminated water 3. Mowed weed is used as an organic matter. Case
(For colour figure, refer to CD) 1 is the artificial contaminated soil without
treatment and Case 2 is a soil sample mixed with
humic leaf soil only into the contaminated soil.
found that the concentration of Cr (VI) decreases
Case 3 is a soil sample with the aerobic
when ATP becomes larger. It is considered that
fermentation carried out by mixing the mowed
the concentration of Cr (VI) decreases due to
weed and rice bran into the contaminated soil.
effects of adsorption and reduction of the
Case 4 is a soil sample with the anaerobic
microorganisms.
fermentation carried out by mixing the mowed
weed and EM (effective microorganisms)
resources offered commercially into the
 Omine et al. 751

Table 2 Composition of samples

Humic leaf Mowed Rice EM resources

Sample Soil (g) Remarks
soil (g) weed (g) bran (g) (g)

Case 1 500 - - - - Contaminated soil

Case 2 500 50 - - - Soil with humic leaf soil
Case 3 500 50 50 25 - Soil with aerobic fermentation
Case 4 500 50 50 - 5 Soil with anaerobic fermentation

Table 3 Test conditions on content of hexavalent

chromium for samples with 500 g
according to Notification No. 46 of the Japan
 Environmental Agency. The test fluid was
Hexavalent prepared with a solid-liquid ratio of 1:10 and
C1 C2 C3
chromium agitated for 6 h to leach the contaminants in the
Weight (mg) 0.25 0.50 0.75 specimen. The mass of the hexavalent chromium
Content (mg/kg) 0.5 1.0 1.5 released from the soil sample was measured using
portable spectrophotometer
(DR2800).

Humic
Decomposed leaf soil Mowed
granite soil weed
Photo 1 Each sample used in the tests
(For colour figure, refer to CD)
 slightly due to the condition of
contaminated soil in a sealed plastic bag incubated
for one week. The EM resources contain various
useful microorganisms such as lactic acid bacteria,
yeast, and phototrophic bacteria. After that, this
sample was fermented by opening the plastic bag
in the aerobic condition. Photo 1 shows the state
of each sample before mixing.
After the fermentation, a sample of 50 g was taken
from each soil and leaching test was performed
Effect of Microorganisms
Table 4 shows the results of
leaching and pH tests on the
contaminated soil with aerobic
or anaerobic fermentation for
the different content of Cr(VI).
In comparison, the results in the
Rice EM case of the contaminated soil
bran resources without treatment or mixed
with humic leaf soil only are
also shown in the table.
Although pH in Case 1 and
Case 2 is in a range of 7~8, the
value in Case 4 decreases
anaerobic fermentation. Figure
4 shows the concentration of Cr(VI) for each soil
sample under the initial condition of Cr(VI)
content of 1.0 mg/kg in which the maximum
concentration of Cr(VI) is 0.10 mg/L. The
concentration of Cr(VI) in Case 1 without
treatment exceeds 0.10 mg/L and it is considered
that the value has a margin of error. The value
decreases in sequence of humic leaf soil, aerobic
fermentation, and anaerobic fermentation. The
concentration in Case 4 with the anaerobic
752 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

contaminated soil with hexavalent chromium. In
the further study, effects of organic content and
fermentation period should be also clarified.
A quantity of microorganisms is estimated by
measuring ATP. The values of ATP on each
sample are shown in Fig. 6. The ATP value was
measured on the sample with the aerobic or
anaerobic fermentation only for five days and on
the sample with both fermentations from the
anaerobic condition to the aerobic condition for
more five days. As shown in the figure, amount of
ATP increases considerably after the
fermentations. Particularly, the sample with the
aerobic fermentation indicates highest value of
ATP. This sample contains the rice bran. It is
considered that this is due to the effect of rice bran
with rich nutrition. However, the anaerobic
fermentation is more effective for the reduction of
Cr(VI). It is suggested that the effect depends on
not only the amount of ATP but also kind of the
microorganisms. In other words, the anaerobic
microorganisms are effective for the reduction of
Cr(VI) in comparison with the aerobic
microorganisms.
CONCLUSIONS
A fundamental study for effect of microorganisms
on concentration of Cr(VI) was performed by the
laboratory test. Especially, effects of mixing
microorganisms for contaminated water and
composting with aerobic or anaerobic
fermentation were discussed based on the test
Omine et al. 753
results. Main conclusions obtained from this study
are as follows:
1) Concentration of Cr(VI) in the contaminated
water can be decreased by mixing useful
microorganisms such as lactic acid bacteria or
yeast. The reduction effect of Cr(VI) increases
with increase in the amount of microorganisms.
2) Aerobic and anaerobic fermentations with
organic matter is effective for reducing the
concentration of Cr(VI) from the contaminated
soil.
3) It is confirmed that the reduction effect of
Cr(VI) using the anaerobic microorganisms is
much higher than that of the aerobic
microorganisms.
REFERENCES
Nakayasu K., Fukushima M., Sasaki K., Tanaka S.
and Nakamura H. (1999) Comparative studies
of the reduction behavior of Chromium (VI) by
humic substances and their precursors,
Environmental Toxicology and Chemistry,
Vol.18, Issue 6, 1085–1090.
Omine K., Yasufuku N. and Tamura K.: (2009)
Purification Effect of Cr(VI)-Contaminated
Soil by Fermentation of Organic Matter, Proc.
of the International Symposium on
Geoenvironmental Engineering, 755-759.
Watanabe N. (2008) Strong reduction effect of
wood vinegar and barn vinegar -purification of
hexavalent chromium contaminant-, Reader for
utilization of wood, bamboo and bran vinegars,
Noubunkyou, 186-187 (in Japanese).
 Analysis of Groundwater Flow in Underground Storage Caverns

A. Usmani, A. Nanda & G. Kannan

Sub Surface Projects Divison, Engineers India Limited, New Delhi, India. (altaf.usmani@eil.co.in)

S.K. Jain
Department of Civil Engineering, Jaypee University of Information Technology, Solan, India

ABSTRACT The paper discusses the hydrogeological design aspects of an underground storage scheme.
This includes the various investigations necessary to study the hydrogeological properties of the site followed
by finite element analysis of the underground storage system, necessary to establish hydraulic containment
and to prevent any flow of contaminants into the subsurface. The study shows that distribution of ground
permeability can significantly influence the overall seepage estimation and a water curtain is necessary for
the hydraulic confinement.

INTRODUCTION fractures needs to be filled by water which inflows

The safe and economical storage of large
quantities of hydrocarbons in today’s world is
ensured by locating it deep below the ground water
table along with directing flows towards the cavity
so that any outward migration of the product is
barred. The design of an underground storage
scheme involves a specific methodology which
takes into account results of numerical
computations, comparisons with standard
approaches and feed back of site monitoring
results. Such an approach of design of theses
structures is suited to avoid any alarming
conditions associated with uncertainties of these
projects. Various design requirements considered
during design stage includes minimum and
maximum pressure conditions along with
consideration of construction stage under
atmospheric conditions. In order to ensure
hydraulic confinement, any upward movement of
gas (vapour phase) bubbles needs to be prevented
through any of the rock fractures. Thus these micro
in the downward direction with such a velocity that
there is a hydraulic gradient in the rock, i.e., the
value of which exceeds 1.0 when measured in the
vertical direction (Aberg, 1977). This requirement
is an important feature to be considered in the
design, especially in a condition when a rock is
crossed by fractures in more than one direction. In
order to maintain required hydrostatic gradient a
water curtain system is provided consisting of
galleries located above the crown of the cavern.
Boreholes are drilled from these water curtain
galleries so as to intersect the joints of rock mass.
Horizontal/Inclined boreholes of 50-100 m length
are drilled perpendicular to main joint set to obtain
a saturated rock mass. A saturated rock mass and
ground water flowing into caverns ensure proper
sealing of the stored product. During construction
boreholes are drilled at least 40 m ahead of the
cavern excavation face to maintain statured
conditions in and around the cavern area. This
paper discusses the investigations, hydrogeological

model and analysis for an underground storage
scheme.
TOPOGRAPHY AND INVESTIGATIONS
Maximum ground surface elevation within the
study area is +130 m mean sea level (msl). To the
south of the study area, there is a hill slope where
the elevation reaches more than +200 m. To the
north, there is an E-W valley, ground surface
elevation +20 to +40 m, having a width of more
than 200 m and a length of 1500 m. To the east of
the study area there is a narrow valley of N-S,
width 50-100 m, stretches some 300 m to the
south. The elevation of the ground surface in the
narrow valley is mainly +15 to +20 m. The study
area falls within the hill side.
Investigations
The following investigations have been carried out
in the study area:
x Geological mapping
x Core drilling
x Laboratory testing of core samples
x Seismic refraction survey
x ҏIn situ stress measurement
x Water pressure tests
x Groundwater level monitoring
x Groundwater analysis
x Interference Well Testing
Based on all the above investigations a hydro-
geological model was prepared.
Hydro-geological Model
The study area falls in the Eastern Ghat Mobile
belt and comprises of typical rock types of Eastern
Ghat Supergroup: Khondalite Group and
Charnockite Group. Laterite capping exists on the
hill tops for a thickness between a 0.2 m to 6 m.
This laterite grades into weathered Khondalite. At
surface depth three joint sets makes Khondalite a
blocky mass. At greater depth the joints sets
appear quite tight. This results in an almost
massive rock mass of Khondalite.
A conceptual three-layer model, layer A, layer B
and layer C, showing separate hydraulic properties
is developed based on the investigations. Layer A,
Usmani et al. 755
consists of the loose top soil, layer B of the
weathered upper part of the rock mass, and layer C
of the fresh part of the rock mass. Layer A is very
conductive, and the groundwater level is in general,
found well below this layer at the bottom of layer
B. The water curtain and the cavern area are
mainly situated within layer C. The study area falls
within the hill side, and the average groundwater
level here is in the range of +60 m msl. The water
pressure test results from the surface investigation
boreholes of the hill side and in layer C represent
the main rock mass of the study area. The
hydraulic conductivity from this area is estimated
as 5.1x10-8 m/sec. At the eastern part of the cavern
area, close to the valleys for about 30-35m, the
bedrock cover is about 25 m. The hydraulic tests
here indicate significantly higher conductivity
values of 10-5 to 10-7 m/s for at least up to a depth
of 20-30 m. The groundwater level (GWT) here is
10 m below the ground surface which is about +10
msl.
During construction of the water curtain boreholes,
water pressure tests were conducted in the
boreholes (apprx. 50-75 m length each). The water
curtain is mainly situated within layer C, except
near the eastern valley. The testing of water curtain
boreholes in Layer C indicates hydraulic
conductivity of 7.6 x 10-9 m/s. The water curtain
boreholes in the eastern side were showing high
conductivity in the range of 10-6 to 10-7 m/s. Based
on the test results an upper bound hydraulic
conductivity of 2.4 x 10-8 m/s and a lower bound
hydraulic conductivity of 1 x 10-8 m/s has been
established for the main study area. For the eastern
side, it has been assumed that the rock mass shall
be grouted to a hydraulic conductivity of at least
1 x 10-7 m/s.
ANALYSIS
Geological, geo-technical, geo-physical and
hydro-geological investigations were initially
carried out to establish the rock mass
characteristics of the storage area. Investigation
results were then realized to carry out numerical
studies of the proposed underground storage
system. Some similar earlier studies reported by
Usmani et al. (2009 and 2010) reflected that
provision of water curtain system significantly
756 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

improves the flow conditions around the storage
units while maintaining a strong flow conditions
all along the cavern section. The study also
advocated the importance of considering the
actual conditions of operation in numerical
modeling in order to find out any possible flaws of
the designed system.
Numerical Analysis
Finite element ground water analysis of an
underground storage systems were carried out
using finite element code Phase 2 (RocScience,
2007). The model as shown in Fig. 1 consists of
two cavern units A and B which are 60 m apart.
Unit A consists of three legs, each of 840 m length
while unit B has two cavern legs of 360 m each
with a pillar width of 30 m. Caverns are D shape
in cross-section, located at an elevation of 30 m
below mean sea level (-30 m msl) with a width of
20 m and height 30 m. The distance between
cavern and right side model boundary is taken as
ten times the width of the excavation to eliminate
the effect of flow conditions on the cavern
boundary.
of the project. Two operational conditions; i.e.
“when both the units are completely filled with
oil” and in other case “both units are empty” were
studied. The maximum operational pressure
considered during the study is 1.3 bars. Due to
large variation in topographical levels of the area
(hill vs. valley), two different ground water tables
of +60 m msl for rock mass on hill side and
+10msl on the valley side were considered. In the
numerical analysis seepage for Unit A and B were
computed taking into account the combined effect
of rock mass having an average permeability of
2.4X10-8 m/s for Case 1 and 1.0X10-8 m/s for Case
2 along with a weak zone which was assumed to
be grouted to permeability of 10-7 m/s for a length
of 35 m.
Analysis of Results
The tabular representation of gradients along with
the discharge quantities from the numerical
analysis is shown in Table 1. This table presents
four cases of operation; two in full condition and
two in empty condition. Number 1 and 2 denotes
permeability combinations of 2.4X10-8 m/s or
1X10-8 m/s along with grouted permeability of 10-7
m/s, to take into account two interpretations of bed
rock permeability’s. The table also gives
quantities of makeup water which is the artificial
quantity supplied to water curtain system over unit
A and B to maintain saturated conditions. The
flow patterns for case Full-1 is shown in Fig.2.

Fig.1 Numerical Model of Underground Storage Table 1 Seepage analysis results

(For colour figure, refer to CD) Case Gradient in Seepage Makeup
Roof Water
All other boundaries were considered as no flow A B (Total)
m3/s m3/s m3/s
boundaries. The vertical distance between water
Min Max
curtain gallery and cavern is kept equal to 20 m
Full-1 1.95 2.1 19.1 7.0 12.5
which comes from the requirement of hydraulic Full-2 1.95 2.1 9.6 4.2 6.54
gradient of greater than 1.0 at the cavern roof level. Empty-1 2.0 2.8 27.2 10.5 17.0
The water curtain tunnel was charged with a Empty-2 2.0 2.8 14.0 6.10 8.53
hydraulic head equivalent to minimum natural
ground water table of +10 m msl. The two units It is observed from the results that while the
were separated with a central access tunnel which essential containment criterion of hydraulic
was charged to +10 m msl level during operations. gradient of t 1 is satisfied in all of cases, the
During this study interaction effect of two cavern proper selection of permeability is found to have
units A and B were analyzed taking into account significant impact on the overall seepage and
the pressure variations during the operation phase
 Usmani et al. 757

artificial makeup water quantities. The above

analyses were done assuming full recharge
conditions.

Fig. 3 Flow patterns with no recharge

(For colour figure, refer to CD)

Fig. 2 Flow patterns with full recharge (Full -1)

(For colour figure, refer to CD)

In addition to the pressure variations and ground

water table variation, further studies were also
conducted to assimilate effects of no recharge of
the water table on seepage. All other conditions of
analysis were same as Case Full 1 of Table 1. The Fig.4 Flow patterns in full-empty case
results for the same are tabulated in Table 2. It is (For colour figure, refer to CD)
observed that a significant drop in water level can CONCLUSIONS
occur as seen in Fig. 3 along with appreciable The present study shows that permeability
effect on seepage. distribution along the underground storage site
significantly effects overall seepage estimation
Table 2 Seepage analysis results along with the additional water requirements. The
Case Gradient in Seepage Makeup need of water curtain is also highlighted in the
Roof Water paper. The study also highlights that detailed
A B (Total) hydrogeological design considering different
m3/s m3/s m3/s
pressure conditions and maintenance requirement
Min Max should be considered before finalizing the
Full 1.95 2.1 19.1 7.0 12.5 underground storage scheme.
recharge
No 1.95 2.1 17.3 6.0 7.50 REFERENCES
recharge Aberg, B. (1977). Prevention of gas leakage from
unlined reservoir in rock. Proc. RockStore,:
The effect of one unit completely empty 399-413
(atmospheric) and other unit full under operational RocScience, (2007). Phase2, Version 6.023.
condition were also analyzed during this exercise Toronto, RocScience Inc.
as shown in Fig. 4. During these full-empty Usmani, A., Nanda, A. & Sharma, K.G. (2009).
conditions, however no fall of gradient is observed Stress and seepage analysis of underground
below minimum value of 1, but level to which rock caverns. Proc. 17th ICSMGE, pp 1753-
water curtain is charged is found to be sensitive to 1756, Alexandria, Egypt.
hydraulic gradient requirements. Therefore it was Usmani, A., Nanda, A., Kannan. G. & Jain, S.K.
inferred from the numerical study that lowering of (2010). Hydraulic confinement of
the potential at the water curtain are required to be hydrocarbons in unlined rock caverns. 6
properly addressed to define operation and ARMS, New Delhi, India.
maintenance requirements of such underground
storage system.
 In Situ Evaluation of a Shallow Soil Bentonite Slurry Trench Cutoff
Wall
Daniel Ruffing
Bucknell University, Project Engineer, Geo-Solutions Inc, New Kensington, PA 15068 USA
druffing@geo-solutions.com
Jeffrey C. Evans, Ph.D., P.E.
Bucknell University, Department of Civil and Environmental Engineering, Lewisburg, PA 17837 USA
evans@bucknell.edu

ABSTRACT Soil bentonite (SB) slurry trench cutoff walls have been widely used in the USA to control
ground water flow and the migration of contaminants in the ground water. While substantial laboratory
testing has been conducted, field studies are limited. Researchers at Bucknell University were afforded the
opportunity to conduct in situ tests on a SB cutoff wall constructed during the summer of 2008. Cutoff
wall properties were measured in situ employing cone penetration tests (CPT), Marchetti dilatometer tests
(DMT), vane shear tests (VST), and ground water level monitoring on both sides of the wall. Tests were
conducted during construction and at times of 3 months, 6 months and 9 months after construction to
evaluate the change in wall properties with time. In addition, bulk samples and a Shelby tube SB backfill
sample were obtained during construction for laboratory testing which included water content, grain size
distribution, consolidation and rigid wall hydraulic conductivity. The field and laboratory data were
analyzed to develop a consistent understanding of the in situ properties of the cutoff wall backfill. The
VST and CPT showed an increase in backfill shear strength over the time-frame of the study. A slight
increase of shear strength with depth was also found. However, a comparison of shear strength measured
compared with that predicted using typical ratios of strength to consolidation stress indicated that the in
situ stress is less than geostatic. Laboratory testing revealed a decreasing hydraulic conductivity with
increasing consolidation stress demonstrating the importance of a reliable estimation of the stress state in
the wall.
control ground water flow beneath a municipal
INTRODUCTION wastewater facility in Birdsboro, PA. A flood
Soil bentonite (SB) slurry trench cutoff walls control dike and the SB wall were built to limit
are commonly used as vertical barriers for the inflow of ground water such that below grade
control of ground water and subsurface wastewater tanks do not become buoyant during
contaminant flow. Despite some research efforts flooding events of a nearby river. The barrier wall
in the laboratory, limited field studies have been is approximately 1400 m2 and 4.5 m deep at its
conducted to determine how these walls behave in deepest point. Design studies showed that the
situ. (National Research Council 2007). Previous desired hydraulic conductivity (1·10-6 cm/s) could
research indicates that the stress is less than that be achieved by simply blending the excavated
which a geostatic (triangular) pressure distribution soils with bentonite-water slurry due to the high
would predict (Evans et al. 1995, Filz 1996, percentage of natural fines along the trench
Ruffing et al. 2010). alignment. However, during construction,
Late in the summer of 2008, researchers at approximately one percent dry bentonite was
Bucknell University were afforded the opportunity added. A clay core dike was installed on top of
to conduct a suite of laboratory and in situ tests on the wall to prevent surface water flooding of the
a shallow SB cutoff wall constructed in eastern site.
Pennsylvania. This wall was constructed under Quality control testing was conducted by
the technical guidance of Geo-Solutions Inc. to Geo-Solutions Inc. to ensure that the finished wall
 Ruffing and Evans 759

had met the design specifications. A summary of consolidation effective stress to determine the
the properties collected during this QC testing is relationship between hydraulic conductivity and
shown on Table 1. consolidating stress for this backfill.

Table 1. QC Properties of Birdsboro Wall METHODS AND MATERIALS

QC Laboratory testing was conducted in Bucknell
Site Specific Results University’s Geotechnical Engineering laboratory.
Requirement Range The Atterberg limit tests and sieve test were
Fresh Slurry conducted according to ASTM D4318-05 and
Marsh Viscosity (sec) > 36 37 - 38 ASTM D422-63 respectively. Classification of
10.4 – the backfill was conducted according to ASTM
Mud Density (kN/m3) > 10.1 10.6 D2487-00. Consolidation and hydraulic
Filtrate Loss (ml) < 20 16 - 19 conductivity testing followed closely the
pH 6.5 - 10 9 - 9.5 procedures in ASTM D2435-04 and Yeo et al.
(2005) respectively. The CPT testing followed the
Trench Slurry procedure presented in ASTM D5778-07 and the
Marsh Viscosity (sec) > 40 41 - 45 data was analyzed using principles from Powell
10.6 – and Lunne (2005), and more specifically, the
Mud Density (kN/m3) 10.1 – 13.4 12.1
effective cone method with two different
Filtrate Loss (ml) < 30 17 - 24 correction factors. DMT testing and data analysis
Sand Content (%) < 15 1 - 11 followed the procedures in Schmertmann (1988)
Backfill Properties combined with recommendations from ASTM D
Backfill Unit Weight 17.3 – 6635-01. Finally VST testing followed the
(kN/m3) > 15.8 21.0 procedures in ASTM D2573-01. More detailed
Percent Fines (%) > 30 41 - 55 procedures can be found in Ruffing (2009).
Slump (mm) 76 - 127 51 - 146
RESULTS
Moisture Content (%) varies 24 - 51
The liquid and plastic limits of the backfill
were found to be 28 % and 17%, respectively,
Field testing was conducted on the wall resulting in a plasticity index of 11%. The
immediately after construction and after the wall moisture content of the freshly placed backfill
had aged 3, 6, and 9 months. This field testing ranged from 24% - 31% with an average of 28%
included cone penetration tests (CPT), Marchetti indicating that the backfill was placed in a viscous
dilatometer tests (DMT), and vane shear tests liquid condition (it was placed at a slump of 76 to
(VST). Groundwater monitoring wells were 127 mm as shown on Table 1). The grain size
installed inside and outside of the wall for long- distribution from the sieve and hydrometer tests is
term monitoring of the ground water on both sides shown on Fig. 1. The grain size distribution and
of the wall. Along with field testing, both an Atterberg Limits were used with the Unified Soil
“undisturbed” backfill sample using a thin-walled Classification System (USCS) to classify the SB
sampler and bulk backfill samples were taken on backfill as a SC, a silty clayey sand with gravel.
the day of construction. Laboratory tests Figure 2 shows the results with error bars of one-
performed on the field mixed backfill samples dimensional consolidation tests conducted on
included Atterberg limits, moisture content, and triplicate samples. These remoulded samples were
grain size distribution by sieve and hydrometer. taken from a Shelby tube sample of the backfill
Along with these physical property tests, one- collected immediately after construction. The
dimensional consolidation tests were conducted on compression index of the backfill was 0.16.
three remoulded samples of the backfill. The
hydraulic conductivity was measured at each
760 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

100 10-4

Conductivity, k (cm/s)
80 10-5
Percent Passing, %

Fine-Grained

FInal Hydraulic
Soil
60
10-6

40
Granular Soil 10-7
20

10-8
1 10 100 1000
0 Log of Effective Stress, Vv (kPa)
100 10 1 0.1 0.01 0.001
Opening Size (mm) Fig. 3 Results of Hydraulic Conductivity Tests
Fig. 1 Grain Size Distribution for Birdsboro
Backfill Monitoring wells installed inside and outside
of the wall were used to determine the
0.8
effectiveness of the wall to slow ground water
migration into the site. Figure 7 shows the
y = 0.77425 - 0.1441log(x)
0.7 R2= 0.99483 measured water levels inside and outside of the
Final Void Ratio, e

wall at the times that the field tests were

0.6
conducted.
0.5
0
Cc = 0.16 0
0.4 3 Months
Construction
1
1
0.3 N=9
1 10 100 1000
Avg. N
Log of Effective Stress, Vv (kPa) 2
2 Depth (m)
Depth (m)

3 3
Fig. 2 Results of Consolidation Tests
4 4
Figure 3 presents the results of hydraulic
conductivity tests (rigid wall) that were conducted 5 5
at each consolidation load in order to evaluate the 0 5 10 15 20 25 30 35 40 0 5 10 15 20 25 30 35 40
CPT Shear Strength, S u (kpa) CPT Shear Strength, S u (kpa)
effect of consolidating pressure (over the range 0
expected in an SB wall) on the hydraulic 0
9 Months
conductivity of the backfill. 6 Months
1
1

Figures 4, 5, and 6 show the shear strength 2

Depth (m)

2
Depth (m)

values in Ruffing (2009) predicted from the CPT

results (using two different correction factors 3
3
called N-values), VST, and DMT, respectively,
during construction, and after the wall had aged 3, 4 4
6, and 9 months. The first N-value was chosen
based on recommendations in Powell and Lunne 5
0 5 10 15 20 25 30 35 40
5
0 5 10 15 20 25 30 35 40
(2005). The second was predicted using the vane CPT Shear Strength, S u (kpa) CPT Shear Strength, S u (kpa)
shear results of this study. Fig. 4 Shear Strength from CPT
Ruffing and Evans 761
762 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
While geosynthetics were used at the base of the
dike to span the trench, the shear strength gain
exhibited in the CPT and DMT data between
construction and 3 months may have been
influenced by the placement of the earthen
embankment along with the self-weight
consolidation of the backfill. (The vane shear
strength at 3 months is inconsistent with all other
strength data and is considered anomalous; these
anomalous results could be caused by the large
rocks in the backfill or possibly because the vane
had strayed into the sidewall of the trench.) The
wall’s shear strength at all ages and depths
classifies the wall materials as a very soft material
(Terzaghi and Peck 1967).
CONCLUSIONS
The findings of this study demonstrated that
the hydraulic conductivity decreases as the
effective confining stress increases, even at the
low stress ranges found in these walls. The
increase in shear strength between the end of
construction and the three-month data measured
using three different field investigation tools
indicated that there is some consolidation and
strength gain soon after construction, but the
backfill does not show any significant change in
shear strength beyond this initial gain.
ACKNOWLEDGEMENTS
Financial support for this work was provided,
in part, by the National Science Foundation (NSF)
through Grant No. CMMI-0722584. The opinions
and recommendations in this paper are solely
those of the authors. The authors would also like
to thank the municipality of Birdsboro, PA for
allowing the research and Brent Estock for
coordinating the access to the site. The time and
effort of James Gutelius, Director of Laboratories
for Civil Engineering at Bucknell University, is
greatly appreciated.
REFERENCES
ASTM D2435-04, Standard test methods for one-
dimensional consolidation properties of soils
using incremental loading. ASTM International,
West Conshohocken, PA.
ASTM D2487-00, Standard classification of soils
for engineering purposes (Unified Soil
Classification System). ASTM International,
West Conshohocken, PA.
ASTM D2573-01 Standard Test Method for Field
Vane Shear Test in Cohesive Soils. ASTM
International, West Conshohocken, PA.
ASTM D422-63, Standard test method for
particle-size analysis of soils. ASTM
International, West Conshohocken, PA.
ASTM D4318-05, Standard test methods for
liquid limit, plastic limit, and plasticity index of
soils. ASTM International, West Conshohocken,
PA.
ASTM D5778-07, Standard Test Method for
Mechanical Cone Penetration Test of Soils.
ASTM International, West Conshohocken, PA.
ASTM D6635-01 Standard Test Method for
Performing the Flat Plate Dilatometer. ASTM
International, West Conshohocken, PA.
Barben, E.J. (2008) Enhanced Clay Barriers: Soil-
Bentonite Cutoff Wall Backfill Amended with
Activated Carbon. MS thesis, Dept. of Civil and
Env. Engineering, Bucknell Univ., Lewisburg,
PA.
Evans, J.C., (1994). “Hydraulic conductivity of
vertical cutoff walls.” Hydraulic Conductivity
and Waste Contaminant Transport in Soil,
ASTM STP 1142, 79-94.
Evans, J.C., Costa, M.J., and Cooley, B. (1995).
“The state of stress in soil-bentonite slurry
trench cutoff walls.” GSP No. 46, ASCE,
Reston, VA: 1173-1191.
Filz, G.M. (1996). “Consolidation stresses in soil-
bentonite back-filled trenches.” Proc., 2nd Int.
Congress on Env. Geotechnics, M. Kamon, Ed.,
Osaka/Japan, 497-502.
Filz et al. (2001). “Determining hydraulic
conductivity of soil-bentonite using the API
filter press.” ASTM Geotechnical Testing
Journal, v. 24, no. 1, pp 61-71.

Kezdi, A. (1975). “Lateral Earth Pressures.” in
Foundation Engineering Handbook, Von
Nostrand Reinhold, Winterkorn H.F. and Fang,
H.Y., eds.
National Research Council (2007). Assessment of
the Performance of Engineered Waste
Containment Facilities, National Academics
Press, Washington/DC: 121 p.
Powell, J., and Lunne, T., (2005) “Use of CPT
Data in Clays/Fine Grained Soils.” Studia
Geotechnica et Mechanica, Vol. XXVII, No. 3-
4, 53-63.
Ruffing, D.G., Evans, J.C. and Malusis, M.A.,
(2010). “Prediction of Earth Pressures in Soil-
Bentonite Cutoff Walls” GeoFlorida 2010, GSP
No. 199, ASCE, Reston, VA.
Ruffing, D. (2009). A reevaluation of the state of
stress in soil-bentonite slurry trench cutoff
walls. MS thesis, Dept. of Civil and
Ruffing and Evans 763
Environmental Engineering, Bucknell Univ.,
Lewisburg, PA.
Schmertmann, J.H. (1988). "Guidelines for Using
the CPT, CPTU and Marchetti DMT for
Geotechnical Design". Rept. No. FHWA-PA-
87-022+84-24 to PennDOT, Office of Research
and Special Studies, Harrisburg, PA, in 4
volumes with the 3 below concerning primarily
the DMT: Vol. III - DMT Test Methods and
Data Reduction (183 pp.)
Terzaghi, Karl, and Ralph B. Peck. (1967). Soil
Mechanics in Engineering Practice,
International Edition A Wiley.
Yeo, S.-S., Shackelford, C.D. and Evans J.C.
(2005) Consolidation and Hydraulic
Conductivity of Nine Model Soil-Bentonite
Backfills. J. Geotech Geoenv. Eng. 131 (10),
1189-1198.
 Multiswellable Bentonite for Soil-Bentonite Vertical Barriers
M.A. Malusis
Bucknell University, Lewisburg, Pennsylvania, USA (michael.malusis@bucknell.edu)
M.D. McKeehan
Bucknell University, Lewisburg, Pennsylvania, USA (mdm021@bucknell.edu)
R.A. LaFredo
Bucknell University, Lewisburg, Pennsylvania, USA (richard.lafredo@gmail.com)

ABSTRACT The viability of multiswellable bentonite (MSB) for soil-bentonite (SB) vertical barriers was
investigated. Bentonite-water slurries and model SB backfills containing MSB were compared against
similar slurries and backfills containing natural bentonite (NB) or “salt-resistant” bentonite (SW101).
Results illustrate that MSB slurry exhibited similar viscosity, density, and filtrate loss as NB slurry. For
backfills exhibiting similar hydraulic conductivities (k) to water, MSB backfill exhibited a lower increase in
k than NB or SW101 backfill when the permeant was changed from water to a 50-mM CaCl2 solution. The
findings are sufficiently promising to warrant further investigation of MSB for vertical barrier applications.

INTRODUCTION swelling of the bentonite (e.g., Shackelford 1994).

Soil-bentonite (SB) slurry trench vertical barriers
(cutoff walls) are commonly employed to provide
in situ containment of contaminated groundwater.
The performance of an SB barrier in these
applications depends largely upon the ability of
the SB backfill to maintain a low hydraulic
conductivity (k), such that spreading of the
contaminant plume is minimized while the
contamination may be addressed through
treatment or natural attenuation. In many cases,
the k of the backfill is governed by the bentonite
fraction, which typically ranges from 1-5 wt %
depending upon the amount of native fines present
in the backfill matrix (Sharma & Reddy 2004).
Research has shown that various factors can
negatively influence the swelling capacity of
bentonite, thereby causing an increase in k of
bentonite-rich barrier materials such as SB
backfills. One such factor involves the chemical
properties of liquid permeating through the
barrier, most notably the concentration and
valence of inorganic cations that inhibit osmotic
The potential for an increase in k of SB backfill
due to interactions between bentonite and
groundwater constituents, particularly divalent
and/or polyvalent cations, has significant
implications for SB vertical barriers used in
geoenvironmental containment applications.
Thus, a primary design objective in these
applications is to ensure that the SB backfill and
the groundwater are chemically compatible, such
that interactions between the bentonite and cations
in the groundwater do not result in an
unacceptable increase in k of the backfill.
One potential approach for enhancing the
chemical compatibility of SB backfill is to use
modified bentonite clay designed to maintain high
swelling and low k in the presence of electrolyte
solutions (e.g., Patton et al. 2007). For example,
modified bentonite clay known as multiswellable
bentonite (MSB) has attracted considerable recent
interest in the geoenvironmental engineering
community due to the ability of MSB to exhibit
swelling in both fresh water and electrolyte
 Malusis, McKeehan and LaFredo 765

solutions (Shackelford et al. 2000, Mazzieri et al.
2005, Katsumi et al. 2008). As described by
Onikata et al. (1996), MSB is created by
compounding natural bentonite (NB) with
propylene carbonate (PC) that expands the clay
lattice and forms a hydration shell around the
interlayer cations. The resulting NB-PC complex
has exhibited osmotic swelling at NaCl
concentrations higher than that in sea water
(Onikata et al. 1999).
The physical and chemical properties and the
mineralogical compositions of the three bentonites
are summarized in Table 1. All three bentonites
classify as high plasticity clays (CH; ASTM D
2487) and contain predominantly montmorillonite.
The cation exchange capacity (CEC) and distilled
water swell index of MSB are lower than those of
NB and SW101. However, MSB exhibits the
highest swell index in a solution containing 50
mM NaCl or CaCl2.

Studies have shown that MSB exhibits greater TABLE 1 Properties and mineralogy of bentonites
swelling and lower k than NB when exposed to tested in study.
electrolyte solutions with ionic strengths up to 1 M Value
Property
(Shackelford et al. 2000, Katsumi et al. 2008). MSB NB SW101
Also, results of long-term k tests (conducted for up Soil Classification1 CH CH CH
to seven years in some cases) illustrate that MSB Swell Index (mL/2 g):2
Distilled water 28.5 35.0 43.0
offers the potential for long-term resistance
50 mM NaCl 39.5 33.0 35.0
against chemical attack in the presence of high 50 mM CaCl2 19.5 12.0 18.5
concentrations of monovalent and/or divalent Principal Minerals (%):3
cations (Katsumi et al. 2008). Montmorillonite 74 69 86
Cristobalite 10 14 6
Based on the above considerations, MSB may be Quartz 2 12 3
effective for SB vertical barriers in groundwater Other 14 5 5
with high electrolyte concentrations. However, CEC (cmolc/kg)4 49.8 83.4 85.7
use of MSB in SB vertical barriers has been given EM (cmolc/kg):4
Ca 7.7 4.9 6.1
little consideration to date. Therefore, the
Mg 6.1 8.8 10.2
objective of this study was to perform a Na 33.3 73.4 64.1
preliminary assessment of MSB for slurry trench K 0.5 1.1 0.2
barrier applications based on evaluation of slurry Sum 47.6 88.2 80.6
properties and backfill hydraulic conductivity. CEC = Cation Exchange Capacity; EM = Exchangeable Metals
1
Based on ASTM D2487
2
Based on ASTM D5890
MATERIALS AND METHODS 3
X-ray diffraction tests by Mineralogy, Inc. (Tulsa, OK)
4
Procedures given by Shackelford & Redmond (1995)
Materials
Three types of powdered bentonite clays were In addition to the bentonites, locally supplied
tested in this study, viz., (1) MSB (Hojun Corp., mortar sand was used to make model SB backfills
Japan), (2) a natural (untreated) sodium bentonite, for k testing. The mortar sand is a poorly graded,
or NB (NaturalGel Wyo-Ben, Inc., Billings, MT), predominantly fine sand with <5 % fines (see
and (3) SW101 (Wyo-Ben, Inc., Billings, MT), a Malusis et al. 2009). The sand was chosen to
treated, “salt-resistant” bentonite developed for simulate SB barrier construction in a clean sand
use in drilling and cutoff wall applications where aquifer.
exposure to seawater is expected. The treatment
process used to create SW101 is unknown Slurry Preparation and Testing
(proprietary). Nonetheless, SW101 was selected Construction of a SB slurry trench barrier requires
for this study based on recent testing of SW101 the use of bentonite-water slurry (typically 4-6 wt
for an SB vertical barrier at the Rocky Mountain % bentonite) to maintain stability of the excavated
Arsenal site in Colorado (see Patton et al. 2007). trench. The viscosity and density of the slurry
766 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
must be sufficiently high to maintain trench
stability, yet sufficiently low to be easily displaced
by the backfill. The slurry also should form an
adequate filter cake to minimize slurry loss from
the trench during construction. Typical properties
of prepared slurry include a Marsh viscosity of 32-
40 s, a mud density on the order of 1.03 Mg/m3,
and a filtrate loss of <25 mL (Millet & Perez
1981, Evans 1993). In addition, a slurry pH of
6.5-10 is desirable to minimize flocculation and
subsequent settlement of the bentonite (Millet &
Perez 1981).
In this study, slurries were prepared by blending
the bentonites with tap water in a high-speed
colloidal shear mixer (i.e., a Hamilton-Beach 7-
speed blender at the highest speed) for 5 min. The
slurries were allowed to hydrate for 24 h prior to
testing for Marsh viscosity (API RP 13B), mud
density, filtrate loss (ASTM D 5891), and pH.
Mud density was measured using a mud balance
(NL Baroid, Houston, TX), and pH was measured
using an Oakton Instruments (Vernon Hills, IL)
bench pH meter. The influence of bentonite
content on these properties was investigated by
testing slurries with bentonite contents ranging
from 2 to 5 wt %.
Backfill Preparation and Testing
Model SB backfills were prepared by combining
mortar sand, slurry (5 wt % bentonite), and
additional dry bentonite (3 or 4 % by dry wt) in a
Hobart mixer (Model #N50, Hobart Corp., Troy,
OH) to ensure uniformity. The dry bentonite was
added with the goal of creating backfills that
exhibited a hydraulic conductivity to water, kw ≤
10-9 m/s, in accordance with typical requirements
for geoenvironmental containment. Slurry was
added incrementally until the backfills exhibited a
slump (ASTM C 143-00) of 125±12.5 mm, a range
consistent with typical field specifications (Evans
1993). After addition of slurry, the total bentonite
contents of the backfills ranged from 4.5 to 5.7 %
by dry weight.
Each backfill was subjected to flexible-wall k
testing, in accordance with ASTM D 5084-03
Method C (falling headwater-rising tailwater).
Backfill specimens were prepared using the
procedures described by Malusis et al. (2009) and
were consolidated under an effective confining
stress of 34.5 kPa prior to permeation with tap
water (pH = 6.65, electrical conductivity, EC =
1.67 mS/m). A hydraulic gradient of 30 or less
was maintained during permeation in all tests.
Each backfill specimen was permeated with tap
water until steady-state conditions were observed
and ASTM D 5084-03 termination criteria were
achieved. The permeant solution then was changed
from tap water to a 50-mM CaCl2 solution in
selected tests, and permeation continued in these
tests until ASTM D 5084-03 termination criteria
were re-established and the ratio of effluent EC to
influent EC was within 1.00±0.05. The latter
criterion has been suggested by Jo et al. (2005) as
a practical criterion for ensuring that final
measured k values are reasonably representative of
long-term conditions in which chemical
equilibrium has been achieved in the test
specimens (i.e., chemical reactions between the
permeant liquid and the bentonite are complete).
This criterion is more stringent than the current
criterion for GCLs permeated with chemical
solutions (i.e., ASTM D 6766), which requires a
ratio of effluent EC to influent EC of 1.0±0.1.
RESULTS
Bentonite-Water Slurry
Values of Marsh viscosity, density, filtrate loss,
and pH for slurries prepared with MSB, NB, and
SW101 are shown as a function of bentonite
content in Fig. 1. The results in Fig. 1a illustrate
that values of Marsh viscosity for the slurries
containing MSB and NB were within the typical
range of 32-40 s and were relatively insensitive to
the percentage of MSB or NB in the slurry. In
contrast, the Marsh viscosity of the SW101-water
slurry increased from 36 s to 200 s as the SW101
content was increased from 2 wt % to 5 wt %.
Mud densities increased with increasing bentonite
content, as expected (Fig. 1b). The mud density of
MSB slurry was slightly lower than that of slurry
containing the same percentage of NB or SW101.
 Malusis, McKeehan and LaFredo 767

Slurries with 5 wt % bentonite all exhibited mud

200 densities between 1.02 and 1.03 Mg/m3, regardless
(a) of bentonite type. In addition, all slurries
Marsh Viscosity (s)

160 MSB
exhibited acceptable values of filtrate loss (Fig.
NB
1c) and pH (Fig. 1d).
120 SW101
The results in Fig. 1 indicate that slurry containing
80
5 wt % MSB or NB exhibited properties generally
40
considered acceptable for slurry trench barriers.
In contrast, the viscosity of SW101 slurry would
0 be excessively high for slurry trench barriers
1.03 unless the slurry is prepared with a relatively low
(b) percentage of SW101 (i.e., 2-3 wt %), which may
not be desirable. Alternatively, the viscosity of
Density, ρ (Mg/m3)

1.02 the slurry may be reduced by adding a thinning

agent. For example, values of Marsh viscosity for
1.01 SW101 slurries containing 4 wt % and 5 wt %
MSB SW101 were reduced to ~40 s and ~50 s,
1.00 NB respectively, by amending the slurries with a small
SW101 amount (0.3 wt %) of chromium-free lignosulfate
thinner (Spersene® CF, M-I SWACO, Houston,
0.99 TX), as shown in Fig. 2a. Although addition of
30 thinner reduced the slurry pH, the pH values were
(c) within the desirable range of 6.5-10 (Fig. 2b). The
MSB
Filtrate Loss (mL)

25 thinner had no significant effect on slurry density

NB
or filtrate loss (data not shown). Based on these
20 SW101
results, 0.3 % thinner was added to the slurry used
15 to prepare the model backfill containing SW101.

10 Backfill Hydraulic Conductivity

Results of the hydraulic conductivity (k) tests on
5 the model SB backfills are summarized along with
10 relevant properties of the test specimens in Table
2. The results show that the final hydraulic
(d) conductivity to water, kw, for the specimen
9 containing 4.6 % SW101 (2.5x10-10 m/s) was well
below 10-9 m/s, whereas the final kw was only
pH

MSB slightly below 10-9 m/s for the specimen

NB containing 4.6 % NB (7.6x10-10 m/s) and above
8
SW101 10-9 m/s for the specimen containing 4.5 % MSB
(5.0x10-9 m/s). However, kw for the MSB and NB
backfills decreased with increasing bentonite
7
2 3 4 5 content (Fig. 3), such that the final kw values for
Bentonite Content (wt %) the specimens containing 5.6 % MSB, 5.7 % NB,
and 4.6 % SW101 all were within the relatively
Fig.1 Properties of bentonite-water slurries as a narrow range of 2.2x10-10 ≤ kw ≤ 2.7x10-10 m/s.
function of bentonite content: (a) Marsh viscosity; (b) Given the similar final kw values for these three
density; (c) filtrate loss; and (d) pH. specimens, the permeant liquid was changed from
768 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

TABLE 2 Summary of data from hydraulic conductivity tests.

Bentonite Water Porosity, Final Final
Bentonite
Content Content, w n kw1 kc1 kc /kw
Type
(dry wt %) (%) (---) (m/s) (m/s)
MSB 4.5 38.5 0.51 5.0x10-9 --- ---
MSB 5.6 40.0 0.54 2.2x10-10 4.2x10-10 1.9
NB 4.6 40.0 0.54 7.6x10-10 --- ---
NB 5.7 43.0 0.55 2.7x10-10 6.9x10-10 2.6
SW101 4.6 39.9 0.56 2.5x10-10 1.0x10-9 4.0
1
kw and kc are hydraulic conductivities to tap water and 50-mM CaCl2 solution, respectively.

200 kc/kw computed based on the final kc and kw values.

The tests were terminated after 2.4-3.0 PVF, once
Marsh Viscosity (s)

160 4 wt % SW101
5 wt % SW101

Hydraulic conductivity, kw (m/s)

120 10-7
MSB
80
NB
10-8
40 SW101
(a)
0
0 0.1 0.2 0.3 0.4 0.5
10-9
Thinner Content (wt %)
10
4 wt % SW101
10-10
9 5 wt % SW101 4 5 6
Bentonite Content (% by dry wt)
pH

8
Fig.3 Final values of hydraulic conductivity to tap
water, kw, for model SB backfill specimens as a function
7 of bentonite content.
(b)
6 10-8
Hydraulic Conductivity, kc (m/s)

0 0.1 0.2 0.3 0.4 0.5

Thinner Content (wt %)
5.6 % MSB
5.7 % NB
Fig.2 Influence of lignosulfate thinner on (a) Marsh 4.6 % SW101
viscosity and (b) pH of slurries containing 4 wt % and 5
wt % SW101. 10-9

tap water to the 50-mM CaCl2 solution in these

three tests so that changes in k due to chemical
incompatibility could be compared relative to a
similar baseline kw. 10-10
0 0.5 1.0 1.5 2.0 2.5 3.0
Pore Volumes of Flow
Values of hydraulic conductivity to the CaCl2
solution, kc, are plotted versus pore volumes of Fig.4 Hydraulic conductivity to a 50-mM CaCl2
flow (PVF) in Fig. 4. Final values of kc also are solution, kc, versus pore volumes of flow for selected
listed in Table 2 along with values of the ratio backfill specimens.

all of the aforementioned termination criteria had
been achieved. The results illustrate that all three
specimens were susceptible to an increase in k
upon permeation with the CaCl2 solution. The
specimen containing 4.6 % SW101 exhibited the
highest kc (1.0x10-9 m/s) and the highest value of
kc /kw (4.0), whereas the specimen containing 5.6
% MSB exhibited the lowest kc (4.2x10-10 m/s) and
the lowest kc/kw (1.9). The better performance of
the MSB backfill relative to the SW101 backfill
may have been influenced by the higher content of
MSB (5.6 %) relative to that of SW101 (4.6 %).
Nonetheless, the results suggest that, for backfills
exhibiting similar kw, backfill containing MSB
may offer greater resistance to chemical attack
than backfill containing NB or SW101.
CONCLUSIONS
Based on the results of this study, MSB appears to
be viable for use in SB vertical barriers. The
physical properties of MSB slurry (viscosity,
density, and filtrate loss) are similar to those of
slurry containing the same percentage of NB.
Model SB backfills containing 5.6 % MSB, 5.7 %
NB, and 4.6 % SW101 exhibited similar k to water
and maintained acceptable k (i.e., k ≤ 10-9 m/s)
when the permeant liquid was changed to a 50-
mM CaCl2 solution. However, the MSB backfill
exhibited the lowest k upon permeation with the
CaCl2 solution. The findings indicate that a more
comprehensive study on the performance of MSB
in SB vertical barriers is warranted.
REFERENCES
Evans J.C. (1993) Vertical Cutoff Walls.
Geotechnical Practice for Waste Disposal, D.E.
Daniel, Ed., Chapman & Hall, London, 430-454.
Jo H., Benson C., Shackelford C., Lee J., and Edil
T. (2005) Long-term Hydraulic Conductivity of a
Geosynthetic Clay Liner Permeated with
Inorganic Salt Solutions. Jl. of Geotech. Geoenv.
Engrg., 131(4), 405-417.
Katsumi T., Ishimori H., Onikata M., and
Fukagawa R. (2008) Long-term Barrier
Performance of Modified Bentonite Materials
against Sodium and Calcium Permeant Solutions.
Geotex. Geomem., 26, 14-30.
Malusis, McKeehan and LaFredo 769
Malusis M.A., Barben E.J., and Evans J.C. (2009)
Hydraulic Conductivity and Compressibility of a
Model Soil-Bentonite Backfill Amended with
Activated Carbon. Jl. of Geotech. Geoenv.
Engrg., ASCE, 131(5), 664-672.
Mazzieri F., Van Impe P.O., and Di Emidio G.
(2005). Chemico-Osmotic Behaviour of
Modified “Multiswellable” Bentonite. Proc., 16th
Int. Conf. Soil Mech. Geotech. Engrg., 2297-
2300.
Millet R.A. and Perez J.Y. (1981) Current USA
Practice: Slurry Wall Specifications. Jl. Of
Geotech. Engrg. Div., ASCE, 107, 1041-1056.
Onikata M., Kondo M., and Kamon M. (1996)
Development and Characterization of a
Multiswellable Bentonite. Proc., 2nd Int. Cong.
Environ. Geotech., Osaka, Japan, 587-590.
Onikata M., Kondo M., Hayashi N., and
Yamanaka S. (1999) Complex Formation of
Cation-Exchanged Montmorillonites with
Propylene Carbonate: Osmotic Swelling in
Aqueous Electrolyte Solutions. Clay Clay Miner.,
47(5), 672-677.
Patton P., Day S., and Byle M. (2007)
Compatibility Evaluation of Groundwater Cutoff
Wall using Salt-Resistant Bentonite and
BFS/Cement for Deep-Mix Barrier Wall. Proc.,
Geo-Denver 2007, ASCE GSP 172, CD-ROM.
Shackelford C.D. (1994) Waste-Soil Interactions
that Alter Hydraulic Conductivity. Hydraulic
Conductivity and Waste Contaminant Transport
in Soil, ASTM STP 1142, 111-168.
Shackelford C.D. and Redmond P. (1995) Solute
Breakthrough Curves for Processed Kaolin at
Low Flow Rates. Jl. Of Geotech. Engrg., ASCE,
121(1), 17-32.
Shackelford C.D., Benson C.H., Katsumi T., Edil
T., and Lin L. (2000) Evaluating the Hydraulic
Conductivity of GCLs Permeated with Non-
Standard Liquids. Geotex. Geomem., 18, 133-
161.
Sharma H. and Reddy K. (2004) Geo-
environmental Engineering, Wiley, Hoboken, NJ,
968 p.
 Migration of Fine Particles Through Porous Media: Application for
Filters Clogging of Permeable Reactive Barriers

Ghizlane Benosman
Ecole Centrale Paris, Laboratoire MSS-Mat,France (ghizlane.benosman@ecp.fr)

Arezou Modaressi-Farahmand-Razavi
Ecole Centrale Paris, Laboratoire MSS-Mat, France (arezou.modaressi@ecp.fr)

ABSTRACT Permeable reactive barriers (PRBs) are now an established approach for groundwater
remediation. However, the in-situ experience shows that the deterioration of filter’s performance is mostly
due to the reduction in pore space caused by mineral precipitation in the reactive area and deposition of fine
particles in the porous media. This study will focus only on the clogging phenomena of PRB material by the
fine particles. Therefore sand filters of different grain size distribution have been tested. A gamma
densitometer has been used to measure the evolution of particle capturing in time and space along the
column.
Keywords: filter, clogging, particle capture, grain size distribution

INTRODUCTION the grain size of infiltrated particles that controls

In environmental problems related to human
activities, the treatment of contaminated soils and
groundwater becomes inevitable. Use of
Permeable Reactive Barriers (PRB) is a passive
rehabilitation technique among others which
permits to treat in-situ the groundwater pollutants
(Blowes et al, 1995). In this technique the barriers
are composed of permeable and reactive filters
(Scherer et al., 2000). Thus, when the pollutants
are in contact with the filter, different mechanisms
take place, e.g. degradation, precipitation and the
sorption (Thiruvenkatachari et al, 2008). However,
the fine particles and the mineral precipitates in the
porous media affect permeability by clogging
pores (Lin et al, 2005 and McDowell et al, 1986).
Particle transport in porous media has been studied
by several authors in different applications. For
example: deep bed filtration (Herzig et al, 1970),
oil production (Moghadasi et al., 2004 and Ochi et
al 1998), soil pollution (Ryan et al, 1995 and
Kretzschmar et al 1997). In general, the transport
and the deposition of fine particles in the porous
media depend on several factors; among others (i)
the kind of forces involved in each process (i.e.
physico-chemical and/or hydrodynamics forces)
(Gohr, et al 1998), (ii) the grain size distribution of
the porous media and (iii) the flow rate conditions
(Reddi et al, 2005).
This work is focused on the fine particle capture
and clogging phenomena in the PRB filters. The
other phenomena as chemical reactions will be
treated in further work. So as to understand the
migration behaviour of pollutants in porous media
as well as the factors influencing transport and
deposition of particles infiltration tests on filters
with different grain sizes have been performed.
The pollutant is simulated by kaolinite clay and the
filter is composed of a sandy material. A gamma
densitometer has been used in order to study the
evolution of particle capturing at the pore scale in
both time and position in the filter. The
permeability evolution and thus clogging is
evaluated thanks to the pore pressure
measurements.
 Benosman, Modaressi and Razavi 771

LABORATORY PROCEDURES AND mean reference permeability of the filter is

MATERIAL computed, applying Darcy’s law and using the
recorded differential pore water pressures for three
Experimental device and procedure different flow rates. The properties of the packing
A scaled model has been constructed to assess material are given in Table 1.
experimentally the clogging threshold on the filters. Then, the suspension is introduced into the filter
Figure 1 shows schematically the principal under investigation. Both, the pressure-time and
components of developed device. density-position data are recorded to yield
information about permeability reduction with
time and density evolution in the filter. This
procedure is repeated for different initial
3 conditions. The fluid flow rate and the
7 concentration of fine particles are kept constant. In
4
this study the flow rate is defined as 100 cm3 min− 1.
Each experience was usually achieved after 5 to 13
1
days.

Used material
5 2 Three different particle size distributions of the
6 same sand are used for simulated porous media
(Figure 2). Kaolinite P300 whose particles’ size
Figure1. Schematical presentation of the infiltration distribution is given in Figure 3 is used for the
column experimental device: 1, gamma ray source; 2, suspension. It is to be noted that 80% of particles
probe; 3, sand column; 4, differential pressure sensors; 5,
have a diameter under 3µm. A predetermined
tank of stirred water and particles; 6, peristaltic pump; 7,
evaluation unit. amount of particles is mixed with water, so that, a
concentration equal to 0.5g/l is obtained. This
The experiments were performed in a 6280 cm3 concentration remains constant during the
sand filled column (length 80 cm and internal infiltration process.
diameter 10cm). A differential pressure sensor is
placed between the inlet and outlet of the column Table1: Physical properties of material
to follow the evolution of pore water pressure Material Particle Porosity Permeability
(item 4 in Figure 1). A peristaltic pump supplies Size mm m²
fluid (mixture of water and kaolinite) into the
column from a reservoir to simulate several flow Sand 0.315-0.63 0.41±0.01 2.09e-11
0.315-1 0.39±0.02 6.72e-11
conditions (items 5 and 6). A bench gamma
1- 4 0.42±0.01 3.73e-9
densitometry moving along the column measures
the spatial and temporal density of the material
(items 1 and 2). It lets obtain different density
profiles. The temperature in the experimental room
is controlled during the tests to reduce
measurement fluctuations. All transducers have
been connected to a computer-controlled data
acquisition system. So as to study the particle
deposition within the sand filter, a well defined
granular material, with fixed porosity is packed in
the column. Once the column is fully packed,
distilled water without fine particles is pumped
Particle size(mm)
into the column in order to saturate the filter. The Figure 2: Grading curve of the different sands
772 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

(0.315-1) mm sand is greater than that of the

(0.315-0.63) mm sand filter. Others authors
reported the similar tendency of permeability drop
under the same conditions of the sand and fine
particles of kaolinite (Alem et al, 2009) .

Figure 3: Grading curve of the Kaolonite P300

RESULTS AND DISCUSSIONS

The clogging of filters by fine particles is
evaluated by the permeability reduction and
variation of the pore space. The clogging
phenomenon is affected by solid particles
concentration, flow rate and initial permeability of
porous media among others. This experimental
study is focused on various grain size distributions.
The experimental data obtained from the
differential pressure sensor are processed to
calculate the initial permeability (see Table 2) and
the evolution of the permeability drop using
Darcy's law. The pressure gradient is proportional
to the flow rate, by the following expression:
Figure 4: Variation of permeability ratio as a function
of injected volume
To study the capture and release of particles over
time, the gamma densitometry gives density
profiles throughout the column over time (Figure
5) and the density profiles evolution along the
column (Figure 6). In following, Figures 5 to 8
present results only for the (0.315 -0.63) mm sand
filter.

q k  ∆p 
v= = −  (1)
s µ  ∆z 
Where the permeability coefficient k depends only
on the geometry of the porous media.

These data are obtained for two grain size

distributions. Figure 4 shows the permeability
evolution in two porous media as a function of
solution volume injected. As it can be observed,
the permeability ratio decreases along the test. In
this study, the overall permeability between the
bottom and the top of the sand column decreased
down to 20% of the initial permeability. Figure 5: Evolution of density with time in different
Surprisingly, the permeability decrease in the heights of the column

Figure 5 shows the evolution of the density with
time in different points of the column profile.
After 2 days of filtration the density increases only
in the first 10 cm of the column. After 2 filtration
days the deposition front advances in the column.
The deposition of kaolinite is stabilized in the
bottom after 6 days but the capture continues in
the middle until the top of the column for 12 days
of filtration.
Figure 6: Evolution of density along the column during
the test
Figure 6 shows the density profile before filtration
(humid density of only the sandy filters) and along
the test (humid density of the sandy filter
including captured fine particles). This data is
used to calculate the evolution of the porosity ratio
(n/n0) of the filter.
The initial porosity n0 is calculated by:
 ρ − ρ1 
n0 =  sf 

(2)
ρ
 sf − ρ w 
where ρ1 is the humid specific mass of the water
saturated filter before filtration of the Kaolinite
suspension and ρ w and ρ sf are the specific mass
of water and sand particles of the filter
respectively.
Assuming that for a unit volume, the change in the
mass is due to the replacement of the porous water
with Kaolinite particles, the change in the porosity
is given by:
Benosman, Modaressi and Razavi 773
 
 
 ρ 2 − ρ1 
∆n =   (3)
 ρ 1 − ρ w  
 sp  ρ  
  sp  
where ρ 2 is the specific mass of the humid filter
containing Kaolinite and ρ sp is the specific mass
of Kaolinite particles.The previous relationship
can be used to calculate the instantaneous porosity
(Figure 7).
Figure 7: Evolution of porosity at different heights of
the column along the filtration experiment
This figure shows porosity evolution over the
height of the column. As it can be observed, the
porosity in the bottom decreases 20% after 2 days
but the pores space remains constant in the middle
and the top of the column. Following this,
particles may migrate and deposit along the
column in the porous media. After 6 days of
filtration the deposition of particles is stopped in
the bottom and it increases along the column.
Knowing the evolution of porosity (Figure7),
using the basic Kozeny-Carman model porosity
given below which takes into account only the,
porosity we can estimate the decline of
permeability and compare with the measured
774 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
values (Figure4). Kozeny (1927) and Carman
(1937) suggested
3
K n  1 − n0 
2
=   
K 0  n0   1− n 
It is found (Figure 8) that though the values
corresponding to the beginning and the end of the
test are approximately in the same ranges, the
intermediary results are not satisfactory. The
discrepancy is even larger for the (0.315-1) mm
sand.
Figure 8: Measured and predicted evolution of
permeability as a function of height of column at
different times
Several phenomena govern the permeability of a
porous media such as the structure and shape of
the pores, the specific surface of the material.
There is no surprise if the Kozeny-Carman model
in its elementary shape can not explain the
permeability evolution obtained in these filtration
tests where the above mentioned parameters may
change.
Figure 9 shows the evolution of density along the
column for different s andy filters. The final
porosity ratios at the end of filtration tests for the
three grain size distribution are given. We notice
that the filter with larger grain sizes result in more
homogeneous porosity change than the small grain
filters in which the probability of the particle
capturing is higher near the inlet.
(4)
Figure 9: Density profiles at the eleven of filtration as a
function of height of column for different filters
CONCLUSIONS
A series of filtration tests are performed to
simulate PRB clogging. Kaolinite suspensions are
passed through sand filters of different grain size
distribution. Thanks to the measurements made by
gamma densitometry the evolution of the particle
capture through the column during the infiltration
is measured. In all tests, the specific mass reduces
monotonously in the whole column. The mean
permeability of the column estimated using the
differential pore pressure between the inlet and
outlet of the column, shows that the (0.315-1) mm
sand which contains a certain amount of bigger
particles, results in higher mean permeability drop
than the (0.315-0.615) mm sand. It was noticed
that the considering only the porosity change in the
Kozeny-Carman relation does not permit to
explain this permeability drop. Other observations
and tests are necessary to better understand this
phenomenon.
REFERENCES
Alem A, Elkawafi A-G and Benamar A, (2009).
Transport et rétention de particules d’argile dans
un milieu poreux saturé, colloque international
sols non saturés et environnement, p303-313

Blowes D.W., Ptacek C.J., Cherry J.A., Gillham
and R.W., Robertson W.D.(1995). Passive
remediation of groundwater using in situ treatment
curtains. In Lin Li ,Craig H. Benson; Elizabeth M.
Lawson (2005).,Modeling porosity reductions
caused by mineral fouling in continuous-wall
permeable reactive barriers, Journal of
Contaminant Hydrology Volume 83, Issues 1-2, 1
February 2006, p89-12
Carman, P.C.(1937). Fluid flow through granular
beds. Trans. Inst.Chem. Eng,15, p150–166.
Gohr Pinheiro I, Schmitz P and Houi D (1998).
Particle capture in porous media when physico-
chemical effects dominate, journal of chemical
engineering science 54, p3801-3813
Herzig J. P., Leclerc D and Le Goff P. (1970).
Flow of suspensions through porous media,
application to deep filtration, Industrial and
engineering chemistry. Vol. 62, No. 5, p. 8-35.
Kozeny, J., (1927). Ueber kapillare Leitung des
Wassers. im Boden, .wien.Akd.wiss,p136-271
Li L, Benson C.H and Lawson E. M.
(2005). ,Modeling porosity reductions caused by
mineral fouling in continuous-wall permeable
reactive barriers, Journal of Contaminant
Hydrology Volume 83, Issues 1-2, p89-12.
McDowell-Boyer L.M, Hunt J.R and Sitar N
(1986). Particle through porous media, journal of
water resources research, vol 22,No 13,p 1901-
1921.
Benosman, Modaressi and Razavi 775
Moghadasi J, Muller-Steinhagen H, Jamialahmadi
M and Sharif A (2004). Theoretical and
experimental study of particle movement and
deposition in porous media during water injection,
Journal of petroleum science and engineering, 43
p 163-181
Ochi J and Venoux J-F (1998). Permeability
decrease in sandstone reservoirs by fluid injection:
hydrodynamique and chemical effects, journal of
hydrology, V 208, p237-248.
Reddi N, Xiao M, Hajra M G and Lee I M (2005).
Physical clogging of soil filters under constant
flow rate versus constant head, Candian
Geotechnique J, vol 42, p 804-811.
Ryan J.N and Elimelech. M (1995). Colloid
mobilization and transport in groundwater. Journal
of colloids and surfaces, Volume107 , p56-1995.
Scherer M., Richter S., Valentine R.L. and Alvarez
P.J.J (2000). Chemistry and microbiology of
permeable reactive barriers for in situ groundwater
clean up, Crit. Rev. Microbiol. 26 p. 221–264.
Thiruvenkatachari.R, Vigneswaran S and Naidu R,
(2008). Permeable reactive barrier for groundwater
remediation, journal of industrial and engineering
chemistry ,p145-156.
 Design Procedure for Permeable Reactive Barrier
with Zeolite-sand Mixture
J Fronczyk
Department of Geotechnical Engineering WULS, Warsaw, Poland (joanna_fronczyk@sggw.pl)
K Garbulewski
Department of Geotechnical Engineering WULS, Warsaw, Poland (kazimierz_garbulewski@sggw.pl)

ABSTRACT The removal of pollutants from the groundwater using permeable reactive barriers (PRB)
is an innovative remediation technology. Liquid contaminants are removed from the groundwater by geochemical
processes (reduction, biodegradation, sorption etc.) occurring during groundwater flow through the reactive material
filling PRB. This paper presents the design procedure recommended for PRB with zeolite-sand mixture. PRB design
involves three steps: investigation tests, qualification tests and verification tests. The criteria proposed for selection
of suitable reactive material were divided into three groups: hydraulic, mechanical and sorption.
Hydraulic and sorption criteria are equivalents and complementary to one another in the process of reactive material
selection. Only a material that fulfills both criteria should be used. The third criterion – mechanical that is related
to the deformation and strength properties of the reactive material should be considered in designing the technology
and evaluation of safety conditions.

INTRODUCTION (Gavaskar et al. 2000, Roehl et al. 2005) is

Waste products and materials resulting
from industrial, commercial, mining,
and agricultural operations placed in landfills
are generally the most dangerous sources
of pollution to the environment. In the case
of landfill location in unfavorable conditions,
unfulfilling the design recommendations,
or damage to the synthetic or clay liners
and drainage, a groundwater protection system
against the infiltration of contaminants
from landfills should be installed. Most systems
contain vertical impermeable barriers insulating
the contaminated soil under the landfill from
its surroundings. However, this method, although
limiting the flow of groundwater contaminated
by leachates from the landfill, does not reduce,
but rather increases the concentration of pollutants
directly under the landfill. In addition,
the impermeable barrier causes changes
in groundwater flow, which is not without
significance for the natural environment.
One of the innovative methods of environmental
protection introduced in the recent years
the method of permeable reactive barriers – PRB
(Fig. 1), which improves the natural attenuation
processes. Presently, studies are undertaken
to identify a reactive material that is inexpensive,
but effectively prevents the migration of leachate
from landfills.
DESIGN REQUIREMENTS
The design of PRB should provided a general
framework for: - identification of the geometrical
parameters, - qualification and selection
of reactive material to guarantee the optimal
conditions for barrier functioning by established
period of time. To determine the geometrical
parameters of PRB analysis of the investigations
carried out in accordance with the diagram shown
in Figure 2 are recommended. The concept
of geometrical parameters of barrier mean width,
thickness, depth as well as barrier orientation
in relation to contaminants plum in groundwater
(Gavaskar et al. 2000). Orientation and depth
of barrier are dependent from hydrogeological
conditions of contaminated subsoil of landfill,
 Fronczyk and Garbulewski 777

where: b – barrier thickness, tPRB – PRB “working

landfill PBR time”, va – groundwater velocity, R – retardation
factor.
PRB design involves the following steps:
investigation tests which include site
groundwater
level and contaminants characterization, qualification
leachate tests during which reactive material is selected
and verification tests, which are necessary
unpolluted for modeling and engineering design.
groundwater flow groundwater
direction
IDENTIFICATION TESTS
Fig. 1 The concept of permeable reactive barriers.
(For colour figure, refer to CD)
To answer the question, is whether or not
the site suitable for PRB application
however width and thickness from dimensions the hydrogeochemical conditions of contaminated
of contaminants plum and degree of groundwater site must be recognized.
contamination. The PRB thickness (b) can be To the most important factors that need
estimated from the equation (Czurda et al. 2002): to be considered to determine the suitability
of a site to PRB application belong contaminant
t PRB ˜ v a type, plum size and distribution, aquifer depth and
b (1)
R also groundwater velocity (Gavaskar et al. 2000).

Fig. 2 Diagram of determining the geometrical parameters of PRB.

(For colour figure, refer to CD)
778 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
Part of identification tests includes drilling
and sampling by conventional drilling, cone
penetrometer testing (CPT), Geoprobe sampling
or installing groundwater monitoring wells.
These test methods can be used for determination
of local hydrogeology, contaminant distribution
in the groundwater, groundwater geochemical
composition and geotechnical features.
Detail information about site characterization
is given in Gavaskar et al. (2000) and Roehl et al.
(2005).
QUANTIFICATION TESTS
Once site characterization information have been
obtained, a suitable reactive material must
be selected for use in PRB. Investigations carried
out in the Department of Geotechnical
Engineering, Warsaw University of Life Sciences,
allowed to propose criteria for the selection
of an optimal reactive material. Analysis
of the test results and numerical calculations
has led to the conclusion that sorption
and hydraulic criteria are equivalents to one
another and complementary in the selection
of the material. The third criterion – mechanical
properties related to the compressibility
and strength properties, is important with regard
to technology and safety conditions. These criteria
have been described in this paper in two aspects:
geotechnical and environmental requirements.
Sorption tests
Evaluation of zeolites or zeolite-sand mixtures
used to fill gates in the PRB with regard
to sorption capacity should be made taking
into account the characteristics of the materials
and the physical-chemical properties
of contaminated groundwater (quantitative
and qualitative chemical composition, pH
and redox potential). For materials, in which
sorption is the main retention mechanism
of groundwater contaminants, the maximum
sorption capacity Ca,max should be taken into
account. In general, it can be assumed that
the sorption capacity of materials should be higher
when a lower ratio of the value of the pollutant
concentration from outflow (CK)
to the concentration of the substance from
inflow to the gates (CD) is required.
In the qualification tests it is recommended
to carry out batch tests of the sorption capacity
of reactive materials with application
of groundwater taken from the polluted site.
An assessment of the reactive material should
be made based on calculations of the time
of proper functioning of the barriers, taking
into account local hydrogeological conditions
(velocity of groundwater flow va, concentration
of the substance in groundwater CD),
and environmental requirements (required
concentrations after treatment CK, usually adopted
in accordance with the limit concentrations
of groundwater purity grade (Polish decree
no. 32.284). Application of a reactive material
is justified when one of the conditions is fulfilled:
Ca,max > 5 mg/g or tPRB > 4 years (2)
where: tPRB – estimated “working time” (proper
functioning) of PRB. Based on batch tests,
Fronczyk (2008) showed that the Slovak zeolite
ZS and zeolite-sand mixtures ZS80 (80%
of zeolite) are materials, which selectively
decrease pollution occurring in the leachate
of municipal landfills. The most intensive
ion-exchange – sorption processes on the surface
of studied materials take place in inorganic
cationic substances (Cu2 +, NH4+, Pb2+), with lower
intensity organic compound dissociated to cations
(e.g. methylene blue). Organic compounds that
are non-dissociated or dissociated to anions
(e.g. phthalic acid, benzene) are not retained
by ZS and ZS80 regenerated to sodium.
Hydraulical and mechanical tests
The hydraulic criterion in the evaluation
of reactive materials such as zeolites and zeolite-
sand mixtures is expressed as the ratio
of hydraulic conductivity of the reactive material
(ks) to the hydraulic conductivity of the aquifer
(kg) in the form:
ks
t 10 (3)
kg

Criterion (3) results from the necessity of taking
into account changes in the hydraulic properties
of materials during PRB exploitation. Laboratory
studies showed a decrease of hydraulic
conductivity values of zeolite and zeolite-sand
mixtures under the influence of landfill leachate.
In the case of zeolite-sand mixtures with a zeolite
content Z (dimensionless value), estimation
of hydraulic conductivity ks in m/s at relative
density ID requires the application of the following
equation (Fronczyk 2008):
ks A ˜ Z ˜ ID
0 , 328
(4)
For mixtures made of zeolite with particle size
of 1.0–2.5 mm, A = 3.0 · 10-4, whereas at particle
size of 0.5–1.0 mm, A = 2.6 · 10-5.
Selection of the reactive material from
the geotechnical point of view requires
the knowledge of the strength and compressibility
characteristics of reactive materials. Test results
(Fronczyk 2008, Katzenbach et al. 2008) provide
a proposal that the application of appropriate
technologies for PBR fulfillment with zeolite
and zeolite-sand mixtures does not cause increased
subsoil softening. As specified by Fronczyk
(2008) and Katzenbach et al. (2008),
the compressibility and strength characteristics
(Table 1) may be used in the design of the PRB
with zeolite (ZS) and zeolite-sand mixtures with
80-percentage of zeolite in the mixtures (ZS80).
Table 1. Strength and compressibility parameters of ZS
and ZS80
Parametr Unit ZS ZS80
Oedometer
compressibility
modulus M MPa 26.9 42.8
(for stress range of
0.10-0.45)
Angle of internal
º 44 44
friction ij’
Cohesion c’ kPa 28 17
VERIFICATION TESTS
Verification tests (if only one or two reactive
materials have been identified) are conducted
Fronczyk and Garbulewski 779
to evaluate the performance of the reactive
material with groundwater from a specific site.
The purposes of verification tests are estimation
of retardation factor and changes of hydraulic
properties of reactive materials.
For the final selection of the reactive material
and retardation factor determination column tests
are recommended (Fig.3). Column tests are more
representative of dynamic field conditions
than batch tests and provide more accurate design
information (Gavaskar et al. 2000, Roehl et al.
2005).
outflow
pump
leachate
Fig. 3 Laboratory stand up for column tests
(Katzenbach et al. 2008). (For colour figure, refer to CD)
Interpretation of column tests results allows
specifying the value of the retardation factor (R),
which can be used in the calculations of the tPRB
value from the equation (1). Using this equation
the relationship between proper working time
of barrier and barrier thickness for zeolite-sand
mixture as a reactive material was drawn (Fig. 4).
In order to control the proper operation of PRB,
the term critical hydraulic conductivity (kcr)
is introduced, which is defined as the value
(decreasing under the influence of the leachate)
of the hydraulic conductivity of the aquifer
(ks = kg) = kcr. A situation when ks/kg<1
is unacceptable, as it means a change
of the direction of groundwater flow
and the possibility of flow across the barrier.
Changes in the hydraulic conductivity
for leachates (k) in relation to the hydraulic
conductivity for water (k0) depending on the pore
780 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

60
NH4
[rok]

1
50 Cu
PRB

40 3
Propere working time t

3a
30 2 3b

20

10 2a 4
2c
2b
0
0 1 2 3
Barrie r thickne ss [m]
Fig. 4 Correct working time and barrier thickness
relationship for zeolite-sand mixture.
(For colour figure, refer to CD)
Fig. 5 Scheme of a flexible-wall permeameter:
1-control panel, 2-bladder accumulator, 2a-water,
volume of flow (PVF) may be 2b-liquid other than water, 2c-elastic membrane,
determined by using the function: 3-chamber, 3a-sample, 3b-latex membrane, 4-
measuring cylinder (acc. to Trautwein Soil Testing
k / k0 504 ˜ ( PVF) 2,34 (5)
Equipment Co.). (For colour figure, refer to CD)
The time elapsing from the start of the operation
of the barriers to reach the value of kg = ks CONCLUSIONS
is termed the time of proper working of the barrier Results of laboratory tests indicate that the use
tPRB. Both kcr and tPRB depend on local of PRB with zeolite and zeolite-sand mixtures
hydrogeological conditions of infiltration (ZS80) for groundwater protection
into the PRB. After this time, regeneration in the neighborhood of landfills is justified
of the reactive material is needed to dissolve in the following cases:
inorganic compounds, mainly CaCO3, before x location of particularly disadvantageous
the creation of solids (VanGulck and Rowe 2004). conditions or improperly protected
From these prerequisites it can be concluded landfill
that the choice of the reactive material for PRB x damage of composite liners of landfills
requires determining the hydraulic conductivity (e.g. geomembrane).
of reactive materials with regard to the relative In the selection of reactive materials for PRB
density and strain conditions (effective stress) the hydraulic, sorption, as well as mechanical
in the barrier. In addition, it is necessary to carry criteria should be taken into account.
out “compatibility tests”, i.e. test of contaminated These criteria should be analyzed based
groundwater influence on changes of the hydraulic on the results of investigations carried out
properties of the selected material. in three stages. Qualification tests permit assessing
Due to the character of these studies, the properties of reactive materials on the basis
it is necessary to use equipment made of the maximum sorption capacity Ca,max,
from materials not reacting with contaminated hydraulic conductivity, as well as compressibility
groundwater. The analysis, the results of which and strength parameters.
was the basis of the criteria presented PRB is an inexpensive method and
in this paper, was carried out in the apparatus does not convey technological difficulties.
shown in Figure 5. The effectiveness of the method depends

on the properties of reactive materials. In the case
of zeolites and zeolite-sand mixtures the method
should apply the principles given in this study.
Acknowledgments: This research was supported
by Grant no. 2P04G 088 29 from the Ministry
of Science and Higher Education, Warsaw, Poland.
REFFERENCES
Czurda K.A., Haus R., 2002: Reactive barriers
with fly ash zeolites for in situ ground water
remediation. Applied Clay Science, 21: 13-20.
Fronczyk J., 2008: Zeolite-sand mixtures
in permeable reactive barriers in the landfill
surroundings. PhD thesis, Faculty
of Engineering and Environmental Science
WULS.
Fronczyk and Garbulewski 781
Gavaskar A., Gupta N., Sass B., Janosy R., Hicks
J., 2000: Design guidance for application
of permeable reactive barriers for groundwater
remediation. Battelle.
Katzenbach R., Fronczyk J., Garbulewski K.,
2008: Evaluation of zeolite-sand mixtures
as a reactive material towards landfill leachate.
11th Baltic Sea Geotechnical Conference.
15.-18.09., GdaĔsk.
Roehl K.E., Meggyes T., Simon F.-G.,
Stewart D.I., 2005: Long-term performance
of permeable reactive barriers. Elsevier B.V.
VanGulck J.F., Rowe R.K., 2004: Influence
of landfill leachate suspended solids on clog
(biorock) formation. Waste Management, 24:
723-738.
 Analysis of the Efficiency of a Zeolite Permeable Reactive Barrier
as Part of the Treatment of the Leachate of Muribeca Landfill, Brazil

M C M Alves
Federal University of Rio de Janeiro/Federal University of Pernambuco, Brasil
(tina@poli.ufrj.br)
C M M Lins
Federal University of Pernambuco, Brasil (ceciliammlins@yahoo.com.br)
J F T Jucá
Federal University of Pernambuco, Brasil (jucah@ufpe.br)

ABSTRACT Among the compounds present in leachate of Municipal Solid Waste (MSW) Landfills in
Brazil one can highlight ammoniacal nitrogen due to its high concentration. The present work has the aim of
assessing the efficiency of a natural zeolite to be applied in permeable reactive barriers (RPB) in the
treatment of the leachate in order to reduce its ammoniacal nitrogen concentration. This investigation
consisted of kinetic studies and batch equilibrium tests on the natural zeolite to build the sorption isotherms.
The contaminant liquids were the raw leachate, leachate pre-treated with calcium hydroxide and an aqueous
solution of ammonia chloride.

carbon for the tertiary treatment for heavy metals

INTRODUCTION
The leachate used in the experiments came from
the Leachate Treatment Plant of the Muribeca
landfill 16 km from the city of Recife, in the
county of Jaboatão dos Guararapes, Pernambuco
State, Northeast Brazil. This is the largest landfill
in operation in the Recife Metropolitan Region,
covering an area of 64 ha.
Permeable reactive barriers (PRB) are successfully
used in various projects as an onsite treatment
technique mainly to rehabilitate soils and
underground water. These barriers are built
perpendicular to the contamination plumes, so that
the contaminated water or effluent is treated as its
goes through the zone with reactive material, in
which the concentration is reduced and/or
contaminants are removed by biological, chemical
or physical processes (US-EPA,1998). Regard to
leachate treatment, Alves et al. 2006, carried out a
study on the behaviour of a permeable reactive
barrier consisting of clay, the prevailing mineral
of which was kaolinite, mixed with activated
of the leachate from the Muribeca landfill, situated
in the Metropolitan Region of Recife, Pernambuco,
Brazil. In this work ammoniacal nitrogen is
studied because of its high concentration found in
most leachates from Brazilian landfills, which
can cause serious environmental problems. This
contaminant contributes to the increase in the
DBO due to its biological oxidation by nitrifying
bacteria, thus requiring a high concentration of
dissolved oxygen for the occurrence of this
biochemical reaction. Moreover, the presence of
the nitrate in the leachate, the end and principal
product of nitrification, stimulates the growth of
algae in the eutrophisation of water resources,
Wang et al., 2006.
Previous technical literature show that the zeolites
can be a beneficial substitute for the resins for
cation exchange in effluent treatment, and are
capable of causing ionic exchange between the
ammonium ion (NH4+) and other biologically
acceptable cations such as sodium (Na+), calcium
(Ca+2), potassium (K+) or hydrogen (H+) (D.
KARADAG et al. 2008; SALTALI et al. 2007;
SARIOGLU 2005; WANG et al. 2005). Also, they
have a high capacity for cation exchange, good
chemical and physical strength and an excellent

compatibility with the environment. Accordingly,
the zeolites are a class of adsorbents of major
economic and social interest, principally with
regard to removal of metal cations and ammonium
ions. Another major benefit in relation to the use
of zeolites is the possibility of regeneration that
these minerals present (Leitão, 2006).
This study was intended to assess the sorption
capability of the natural zeolite to be applied in
permeable reactive barriers in the tertiary
treatment of the leachate from Municipal Solid
Waste Landfill of Muribeca, in Pernambuco State,
in order to enhance its treatment with a view to
reducing the ammoniacal nitrogen concentration
in the leachate.
METHODOLOGY
This investigation consisted of two stages. The
first stage involved characterising the natural
zeolite and raw leachate used in the study. In the
second stage, kinetic studies and batch equilibrium
tests were carried out to build the sorption
isotherms for three different liquids: the raw
leachate, the treated leachate and an aqueous
solution of ammonia choride.
Characterization of natural zeolite
The natural zeolite used in the experiments was
extracted from a deposit located in the county of
Calingasta, San Juan Province, Argentina. The
chemical composition of the oxides in the studied
zeolite, was determined by the semi-quantitative
analysis using a Rigaku X-ray fluorescence
spectrophotometer model RIX 3000. It is mostly
composed of SiO2 (58.16%), followed by Al2O3
(7.67%), CaO (9.9%), Fe2O3 (6.63%), K2O
(3,07%). The other compounds, TiO2, P2O5, BaO,
SrO, MgO, ZrO2, MnO add up to 2.725% . A lost
to fire of 11.62% has occurred. Additionally,
cation exchange capacity (CEC) of the zeolite was
found 130 meq/100g, whose value fits the range
obtained by other researchers (Jorgensen et al,
1976 apud Qi Du et al. 2005), (Hlavay, 1982 apud
Qi Du et al. 2005), (Kesraoui-Oukl et al, 1993),
(Booker et al, 1996), (Sarioglu, 2005). The
mineralogical characteristics of the zeolite were
determined by analyses of X-ray diffraction using
Alves, Lins and Jucá 783
Siemens equipment model D 5000, graphite
monochromatiser and nickel filter (Ni). These data
together with the analysis of the composition of
the oxides present in the zeolite, lead to the
conclusion of being classified as clinoptilolite.
From the results of the physical characterization of
the zeolite it was possible to note that the zeolite
in the study has a specific mass of 2.25g/cm3, a
hygroscopic moisture content of 6%, characteristic
grading of coarse sand and did not present limits
of consistency.
Characterization of leachate
The leachate samples were collected in the flow
just before coming into the treatment plant.
Samples went to laboratory for physicochemical
characterization in accordance with the Standard
Methods for the Examination of Water and
Wastewater, ALPHA,1998. Table 1 shows the
complete characterization of the raw leachate.
TABLE 1 Characterization of raw leachate
Parameters Values
pH 8.59
Alkalinit (mg L-1 CaCO3) 7,243
DBO5 (mg L-1) 2,781
DQO (mg L-1) 3,951
DBO5 /DQO 0.70
Colour (Hz) 9,558
Turbidity (NTU) 170.0
Conductivity (mS cm-1) 19.90
Ammoniacal Nitrogen (mg L-1) 1,453
Chloride (mg L-1) 411.0
Total Solids (mg L-1) 10,341
Total Dissolved Solids (mg L-1) 8,530
Total Suspended Solids (mg L-1) 1,811
Preparation of Liquid Solutions
Three different contaminants were prepared for
experiments using natural zeolite: raw leachate
(L1), leachate pre-treated with calcium hydroxide
(L2) and an aqueous solution of ammonium
chloride (AS). The raw leachate (L1) was
prepared as follows: After determining the
ammoniacal nitrogen concentration dilutions with
de-ionised water dilutions were made in the raw
leachate to obtain different concentrations of
784 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
NH4+. The L2 was obtained from pre-treatment of
the L1 with calcium hydroxide and brought
ammoniacal nitrogen concentration down to
values around 1,000 mg L-1 (Santana-Silva, 2008).
Is is important to remark that the above mentioned
leachate treatment was used mainly to diminish
color values of the leachate. For the aqueous
ammonium chloride solution (AS) a solution was
prepared with de-ionised water and ammonium
chloride to obtain samples with different
concentrations of NH4+ for the experiments.
Batch Tests
Kinetic Studies
The kinetic tests for assessing the ammoniacal
nitrogen sorption by the zeolite were performed in
order to determine the optimum point so that the
zeolite–ammoniacal nitrogen system would reach
equilibrium and also to assess the effect of the
time for shaking the samples. In this way, six
grams of zeolite were weighed and placed in 125
ml Erlenmeyer flasks, to which 100 ml of
contaminants (L1, L2 and AS) was added. For
each contaminant 11 Erlenmeyer flasks were
prepared (zeolite + contaminant) that were placed
in an incubator under shaking of 120rpm and
constant temperature of 28oC. At preset intervals,
varying from three minutes to 24 hours, the
supernatant in the Erlenmeyer flasks
corresponding to the preset time was separated
from the zeolite and analysed immediately with
regard to the concentration of ammoniacal
nitrogen concentration, pH variation, conductivity
and DQO.
Batch Equilibrium Tests
The sorption of the zeolite was assessed using
batch equilibrium tests in accordance with the
standard EPA/530/SW-87/006-F (US EPA 1992).
At this stage in experiments a series of
suspensions was prepared in 125ml Erlenmeyer
flasks, each consisting of six grams of zeolite and
100 ml of contaminants (L1, L2 and AS), with
ammoniacal nitrogen concentrations varying
between 30 and 1700 mg of NH4+ L-1. Control
samples were prepared together with these
samples, corresponding to 100ml of the
contaminant with different ammoniacal nitrogen
concentrations, without the presence of zeolite, in
order to assess a possible removal of the
ammoniacal nitrogen only by the shaking process.
Next, the samples were placed in the incubator
and shaken 120 rpm at 28oC. After 24 hours the
solutions were removed, the supernatant was
separated from the zeolite in each flask and the
ammoniacal nitrogen concentration, pH variation,
conductivity and DQO were determined in the
supernatant.
RESULTS AND DISCUSSION
Kinetic Studies
As shown in Fig 1 it is found that the absorption
of the ammoniacal nitrogen occurs quickly,
approximately 4 hours or less of shaking for the
three contaminants. From this point the removal
efficiency begins to gradually diminish as the time
of contact increases, and then steadies after 6
about hours of shaking. After 24 hours of shaking
it is still possible to see a small reduction in the
concentration of ammoniacal nitrogen, which can
be considered negligible.
Fig.1 Time of contact versus efficiency of removal of
the ammoniacal nitrogen by zeolite.
(For colour figure, refer to CD)
The decrease in the removal of ammoniacal
nitrogen by the zeolite over time occurs probably
due to the more available “exchangeable places”
of the zeolite in the beginning of the experiment
and by the fact that the concentration gradient of
the solute is high at first, facilitating the fast
 Alves, Lins and Jucá 785

removal of the ammoniacal nitrogen in the first expected because of the lack of competition
few minutes. On the other hand, the removal of other ions besides NH4+.
efficiency of the ammoniacal nitrogen by zeolite
reached about 60% in six hours for L1 being more
significant than 45% removal in L2. The removal
efficiency obtained with AS was above 75% after
six hours, reaching a little more than 80% after 24
hours. Wang et al. 2006, also find that higher rates
of adsorption were obtained with the aqueous
ammonium chloride solution compared to the
leachate. It certainly has a strong influence of the
presence of other elements within the leachate
leading to ion competition for adsorption. Twenty
four hours batch equilibrium tests were made for
construction of sorption isotherms from here on. Fig. 2 Sorption isotherm with zeolite and raw
leachate (L1) (For colour figure, refer to CD)
Batch Equilibrium Tests
For all contaminants (L1, L2 and AS) the sorption
isotherms did not show a linear behaviour so they
were adjusted to the experimental data according
to the non-linear Freundlich and Langmuir models
as shown in Fig. 2 for L1, Fig. 3 for L2 and Fig. 4
for AS. The experiment with aqueous solution was
undertaken to check the influence of other
components existing in L1 and L2 in the capacity
of adsorption of the ammoniacal nitrogen by the
zeolite, since this solution does not have ions other
than the ammonium ion. Fig. 2, 3 and 4 show that
the Langmuir isotherm was better adjusted to the Fig 3 Sorption isotherm with the zeolite and
experimental data. From these data, the sorption leachate pre-treated with calcium hydroxide (L2)
parameters were calculated for the two models in (For colour figure, refer to CD)
question, namely: partition coefficient (Kp),
Freundlich sorption constant (N), maximum
sorption rate (E), Langmuir sorption constant (D)
and the adjustment coefficient (R2). The values of
these parameters are described in Table 2. It was
possible to find that the maximum sorption rates
obtained were slightly higher to L1 when
compared with the L2, probably, due to the
treatment of the L2 resulting in a larger quantity of
calcium ions compared to L1. This may therefore
have contributed to increasing the competition for
the exchangeable places of the zeolite reducing the
adsorption rate by the ammoniacal nitrogen. When Fig.4 Sorption isotherm with zeolite and an
analysing the data obtained with AS it is found aqueous ammonium chloride solution (AS)
that a maximum sorption rate of 13.69 mg NH4+ (For colour figure, refer to CD)
g-1 was obtained. This value is the largest one as
786 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

TABLE 2 Sorption parameters obtained from batch
equilibrium tests
Models Parameters
Raw Treated Aqueous
Leachate Leachate Solution
(L1) (L2) (AS)
kp = 0.54 kp = 0.647 kp = 1.04
Freundlich N = 0.446 N = 0. 413 N = 0.363
R2 = 0.939 R2 = 0.962 R2 = 0.99
E = 11.36 E = 10.52 E = 13.69
Langmuir D = 0.0075 D= 0.0092 D = 0.008
R2 = 0.987 R2 = 0.993 R2 = 0.97
kp , E (mg g ) N, D (L g-1)
-1
The values of the sorption parameters obtained in
the study herein for the Freundlich model for both
L1 and L2 and for AS are close to the results found
in the literature, as in Wang et al. Wang et al.
[2006] and D. Karadag et al. [2006] for the
leachate tests, and in Qi Du et al. [2005], Sarioglu
[2005] and Saltali et al. [2007] for the aqueous
solution tests. When analysing the results from
using the Langmuir model, which gave a better
adjustment of the experimental data (R2), it is
found that the values of the maximum sorption
capacity (E) for L1 and L2 are close to the results
obtained by by Saltali et al. [2007], Englert and
Rubio [2005], Lei et al. [2008] for an aqueous
solution, as in the results obtained with the
aqueous solution in the study herein. This
proximity of the values of the sorption rates
obtained for L1, L2 and AS, which were close to
12 mg/g, can be justified by the close affinity of
the zeolite with the ammonium ions, in accordance
with the order of selectivity described by Ames in
Sarioglu. 2005.
CONCLUSION
Considering the tests performed on a bench scale,
the results show a promising usage of zeolite for
ammonium attenuation in MSW landfill leachates
for two main reasons. Initially, from the kinetic
tests, it was possible to find that the removal of
ammoniacal nitrogen was fast, in the first six
hours of shaking, and later a decrease in behaviour
tending to stabilisation. In the second place, with
regard to the batch tests it showed high sorption
capacity of ammoniacal nitrogen present in the
leachates, whose values were close to the sorption
rates found for the ammonium chloride aqueous
solution confirming selectivity of NH4+ ions by
the natural zeolite used. Complementary column
tests in structured material close to field
conditions were run in the Environmnental
Geotechnical Laboratory in Federal University of
Pernambuco. These results are being analysed and
shall be published soon. Natural zeolite shall
played a good role on attenuation of ammoniacal
nitrogen when used as filling material for the PRB
of the leachate treatment plant of Muribeca MSW
Landfill.
ACKNOWLEGMENTS
The authors are grateful to the National Council
for Research, CNPq, for financial support. We are
also grateful to Solid Research Group at Federal
University of Rio de Janeiro, GRS/UFPE for
technical support.
REFERENCES
ALVES, M. C. M.; BELTRÃO, K. G. Q.; JUCÁ,
J. F. T. (2006). Assessment of Metals Sorption
in the Permeable Reactive Barrier of the
Leachate Treatment System of Muribeca
Landfill, Recife, Brazil. In: 5th International
Congress on Environmental Geotechnics,
Cardiff - Wales, United Kingdom.
APHA - AMERICAN PUBLIC HEALTH
ASSOCIATION. (1998). Standard Methods
for the Examination of Water and Wastewater.
20. ed. Washington: American Public Health
Association. 1220 p.
D. KARADAG; KOC, Y.; TURAN, M.;
ARMAGAN, B. (2006). Removal of
ammonium ion from aqueous solution using
natural Turkish clinoptilolita. Journal of
Harzardous Materials, v. 136, p. 604-609.
D. KARADAG; TOK, S.; AKGUL, E.; TURAN,
M.; OZTURK, M.; DEMIR, A. (2008).
Ammonium removal from sanitary landfill
leachate using natural Gordes clinoptilolite.
Journal of Harzardous Materials, v. 153, p. 60-
66.

ENGLERT, A.H.; RUBIO, J. (2005).
Characterization and environmental
application of a Chilean natural zeolite.
International Journal of Mineral Processing. v.
75, p. 21-29.
JORGENSEN, T. C.; WEATHERLEY, L. R.
(2003). Ammonia removal from wastewater by
ion exchange in the presence of organic
contaminants. Water Research, v.37. p. 1723 –
1728.
KESRAOUI-OUKL, S.; CHEESEMAN, C.;
PERRY, R. (1993). Effect of conditioning and
treatment of chabazite and clinoptilolite prior
to lead and cadmium removal, Environmental
Science Technology. n. 27, p. 1108–1116.
LEITÃO, R.L.F. (2006). Remoção de nitrogênio
amoniacal de efluente de reator hidrolítico de
lodo primário de estação de tratamento de
esgotos utilizando adsorção em zeólita visando
à produção de fonte de carbono para
desnitrificação. Dissertação Mestrado – Escola
de Engenharia de São Carlos, Universidade de
São Paulo, São Carlos-SP.
LINS, C. M. M. (2008). Avaliação da zeólita
natural para aplicação em barreiras reativas
permeáveis no tratamento do lixiviado do
Aterro da Muribeca/PE. Dissertação de
Mestrado - Centro de Tecnologia e
Geociências, Universidade Federal de
Pernambuco, Recife-PE. p. 130.
Alves, Lins and Jucá 787
QI DU; LIU, S.; CAO, Z.; WANG, Y. (2005).
Ammonia removal from aqueous solution
using natural chinese clinoptilolite. Separation
purification technology, v. 44, p. 229 – 234.
SALTALI, K.; SARI, A.; AYDIN, M. (2007).
Removal of ammonium íon from aqueous
solution by natural turkish (Yildizeli) zeolite
for environmental quality. Journal of
Harzardous Materials, v. 141, p. 258-263.
SANTANA-SILVA, F. M. (2008). Avaliação do
método de precipitação química associados so
dtrippind de amônia no tratamento do lixiviado
do Aterro Muribeca-PE. Dissertação Mestrado,
Centro de Tecnologia e Geociências,
Universidade Federal de Pernambuco. Recife.
133 p.
SARIOGLU, M. (2005). Removal of ammonium
from municipal wastewater using natural
Turkish (Dogantepe) zeolite. Separation and
Purification Technology, v. 41, p. 1-11.
UNITED STATE ENVIRONMENTAL
PROTECTION AGENCY - US EPA (1998).
Permeable Reactive Barrier Technologies for
Contaminant Remediation. Office of Research
and Development, Washington.
WANG, Y.; LIU, S.; XU, Z.; HAN, T.; CHUAN,
S.; ZHU, T. (2006). Ammonia removal from
leachate solution using natural Chinese
clinoptilolite. Journal of Hazardous Materials, v.
136, p. 735–740.
 Time Dependant Differential Consolidation of Slurry Wall Backfill

Rahul V. Mukherjee
University of Saskatchewan, Saskatoon, Canada (rahul.mukherjee@usask.ca)

Moir D.Haug
University of Saskatchewan, Saskatoon, Canada (mhaug@mdhsolutions.com)

ABSTRACT: This paper presents the preliminary results of a laboratory testing program to investigate time-dependant
differential consolidation of slurry wall backfill material. The program was conducted to supplement the design and
installation of an 11,000 m long slurry wall. This slurry wall is installed through intermediate aquifers within low permeable
glacial till to a final depth of nearly 50 m. The testing program involves large diameter large strain consolidation testing of
selected field backfill samples. The apparatus was also modified to measure lateral pressure during loading. The heads
across the large strain consolidation apparatus were measured continuously. These results were supplemented by direct
shear tests performed on the different backfill materials. The results of this program show that there is significant retardation
in vertical stress in the backfill caused by arching. It also shows that the narrower the trench the greater the loss of vertical
effective stress. Future testing and analysis may show that the loss of this stress increases the potential for differential
consolidation around intermediate aquifers.

INTRODUCTION
The potential for slurry wall backfill arching and hang-up
was examined in a laboratory testing program. The
permeability of clay till slurry backfill is highly dependent
on the load placed on the backfill. As a result permeability
usually decreases with depth as the backfill consolidates
slowly under increasingly high loads. If the width of the
backfill trench is narrow and the depth great, there is a
potential for arching and backfill hang-up. In addition, if
the slurry wall crosses multiple aquifers it is possible that
differential consolidation around these aquifers could
result in the creation of more permeable zones with depth.
In order to examine this potential a laboratory program
was conducted to examine large-strain consolidation of
clay-till slurry backfill.
BACKGROUND
Soil bentonite (SB) slurry walls vertical barriers
constructed by excavating a narrow trench through
permeable material or zones and backfilling with a
mixture of soil, bentonite and water to achieve a low
permeability barrier (Xanthakos 1979, Haug 1983, Evans
1995). The soil either comes from the excavated trench or
from a source nearby, if the soil is deemed to be
unsuitable for use. The slurry in the SB wall is a mixture
of water and dry bentonite (4% - 7%) by weight (Evans
1994). Before placement in the trench the backfill is
mixed with desired amount of slurry from the trench to
provide a desired consistency. Since SB walls are mainly
used as low permeability barriers and constructed as a part
of polluted site remediation process, hydraulic
conductivity value plays a very important role. In general,
they range from 10-7 m/s to 10-11 m/s (D’Appolonia 1981,
Haug 1987, Evans 1995, Yeo et al 2005).
The value of permeability for the backfill mixture depends
upon many factors such as (i) the amount of dry bentonite
added to the mix (Ryan 1987, Evan 1994, Yeo et al 2005),
(ii) the amount of fines in the mixture (Xanthakos 1979,
D’Appolonia 1981, Yeo et al 2005), (iii) resistance of the
mix against the contaminant being contained and (iv) the
state of stress in the slurry wall (Evans 1995, Filz
1999).The performance assessment of these walls in terms
of hydraulic behaviour is generally done through
laboratory testing of the backfill mixture.
Laboratory model and field study on the slurry walls have
shown that the stress in the wall does not increase linearly
with depth as assumed earlier (McCandless and Bodocsi
1987, Evans 1995).The friction at the trench/backfill
interface governs the consolidation behaviour of the
backfill-slurry mix, which is fairly compressible. This
phenomenon can be explained by arching mechanism.
THEORY
Arching (Terzaghi 1943) can be described as “transfer of
stress from a yielding mass of a soil onto adjoining
stationary parts”. This theory was first described by
Marston and Anderson in 1913 to show the state of stress
on pipes in ditches, which was further described by
Terzaghi in his paper in 1945. Figure 1 shows the force
mobilization around a small element of backfill dh in
thickness in a trench of width B due to arching. According
to this theory, the trench walls are considered as rigid and
 Mukherjee and Haug 789

the backfill material is considered as compressible. backfill and potentially the sequence and spacing of
Consolidation and settlement of backfill with time cause permeable zones cut-off by the slurry wall.
shear stresses to be mobilized along the trench walls (Fig
In 2009, construction was started on an 11,000 m long,
1), which act as a partial support for the backfill and hence
over 45 m deep slurry wall through glacial till at a
reduces the effective vertical stress in the trench below the
Saskatchewan potash mine. This slurry wall is intended to
overburden pressure.
control the migration of chloride impacted water. The
According to static equilibrium conditions, considering
permeability and the potential for osmotic consolidation is
the vertical force equilibrium (™ FY=0) of the horizontal
a function of the stress environment of the trench backfill.
slice in Fig1.
As a result, the potential for arching or wall hang-up is a
v concern. Further complicating the situation is the presence
dV 2K dh Bdh (a)
B of stacked aquifer which will result in more rapid backfill
Where, consolidation at various depths in the trench. Figure 2
shows the site geology.
µ = co-efficient of friction between the trench wall and
backfill
K = co- efficient of earth pressure (ratio of horizontal
stress to vertical stress)
Solving the partial differential equation in (a), we get
K n
B2 2
V (1 e B ) (b)
2K
Computing the average vertical stress from equation b,

K n
V B 2
B ) (c)
av (1 e
B 2K

Fig 2: Schematic diagram of site geology

Fig 1: Arching mechanism after (Marston and Anderson, 1913)
FOCUS OF THE STUDY
Backfill hang-up and arching is of particular concern for
deep slurry walls. This potential is a function of the width
of the trench, the depth of the trench, shear strength of the
EXPERIMENTAL PROGRAM
The experimental program involves large strain
consolidation testing of the backfill mixes. A 150 mm
diameter consolidation cell was modified to increase its
depth and fitted with ports (A to E) to enable pore
pressures to be measured. Marriott bottle system was also
used to enable continuous monitoring of permeability
during loading. Miniature load cells were installed on the
walls of the mold at different depth to measure the stress
profile. Figure 3 shows a sketch of the mold.
Direct shear tests were also conducted on the different
backfill mixes to measure their frictional characteristics.
TEST MATERIALS
Till that was used in the experimental study was collected
from the site. Table 1 shows the index properties of the till
and the backfill. Additional backfill mixes were prepared
790 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

for this study by adjusting the fine content to 10%, 25% load cell and potentiometer reading stabilized. The whole
and 50%. setup was then loaded using Karol-Warner Conbel. Load
cell and deflection readings were recorded and monitored
continuously. A constant head hydraulic conductivity test
was run simultaneously, with a use of a Marriott bottle
arrangement (Fig 3). Once the equilibrium was attained,
next incremental load was applied to the sample.

DISCUSSION OF TEST RESULTS

Impact of fine content on shear strength

Fig 3: Modified mould
TABLE1. Index properties of Till and clay-till backfill
Figure 4 shows the variation of shear strength with
percent of fine in the backfill mix. The bold line
represents the 50% fine content mix and the dotted line
represents the 25% fine mix. The angle of internal friction
was not highly dependent on the % of fine. The friction
angle for these clay-till mixes ranged from 23 to 26
degrees. These values are considerably lower than the 30
to 35 degrees previously reported in the literature.

Property Till Backfill

Specific Gravity 2.72 2.68

Liquid Limit (%) 42.5 37.5
Plastic Limit (%) 24 20
Plasticity Index (%) 18.5 17.5
Gravel (%) 30 20
Sand (%) 40 52
Silt (%) 14.5 19
Clay (%) (<0.002 mm) 15.5 9
Fig 4: Shear strength of clay-till
backfill
TESTS PERFORMED AND TEST PROCEDURE

Direct shear test

Samples from different mix were allowed to consolidate
in large consolidation molds at different specified normal
loads. After, consolidation was complete; the samples
were extruded and trimmed to fit into 150 mm x 150 mm
diameter direct shear box. The samples were left for 1 -3
days depending upon the normal load. The shear box test
was then carried out according to ASTM D-3080-04.

Large strain consolidation and hydraulic conductivity

testing
All the saturated samples in the molds were compacted at
a water content corresponding to a slump of 12 cm and at
a constant dry density. The depths in the field were
simulated corresponding to the load applied onto the
sample. The samples once compacted were left to
consolidate under the weight of the loading plate, till the
Fig 5: Variation in vertical hydraulic conductivity and
coefficient of lateral earth pressure with calculated depth
Figure 5 shows the variation of coefficient of lateral earth
pressure (k) and hydraulic conductivity measured during
one of the large strain consolidation tests. The hydraulic

conductivity decreases. This is because, as the load is
increased the void ratio decreased, gradually with depth,
however the rate of decrease slowed under increased
loads. A sharper decrease was measured for the
coefficient of lateral earth pressure.
Stress distribution in slurry wall
Figure 6 shows the stress distribution profile in deep clay-
till backfilled deep slurry trench with width of 0.6 to 2.1
m using the results of the test program according to
equation (c). The unit weight, coefficient of lateral earth
pressure and shear strength of the backfill are: Ȗ = 17.8
kNm3, k=0.10 and = 23 .The stress in the backfill is
always less than the geostatic stress (Ȗ`h) and becomes
constant with depth. This is because at large depths, the
shear stresses at the sides of the trench are then large
enough to balance the extra weight of each additional
layer of backfill. However, for any given depth, it is
evident from the figure that with the increase in the width
of the wall, the stress increases.
Fig 6: Stress profile in SB walls
CONCLUSION
The results of this study show that there is a significant
loss of effective vertical stress in the backfill with depth,
and that this loss increases for narrower trench. Future
testing will be required to show if such loss in stress
increases differential consolidation across aquifers. The
hydraulic conductivity of the clay-till backfill was found
to range from 10-5 m/s at a depth of 0.5 m to 10-8 m/s at a
depth of 2m. The hydraulic conductivity was still
decreasing with load. The coefficient of lateral earth
pressure was also found to decrease ranging from near 0.3
Mukherjee and Haug 791
near ground surface to 0.1 approaching 2.5 m depth. In the
case of a 1m wide trench only a small proportion of the
backfill weight acts to consolidate backfill at depth.
REFERENCES
ASTM D 3080 – 04. Standard test method for direct shear test of
soils under consolidated drained conditions
D’Appolonia, D. (1980) Soil-bentonite slurry trench cut-offs.
Journal of Geotechnical Engineering Division, 106(4), 399-417.
Evans, J.C., Costa, M.J. and Cooley, B. (1995) The state-of-
stress in soil-bentonite slurry trench cut-off walls, Proc.,
Geoenvironment 2000, ASCE Geotechnical Special Publication
No.46, Y.B. Acar and D.E. Daniel, eds., New Orleans,
Louisiana,1173-1191
Filz, G.M. (1996) Consolidation stresses in soil-bentonite
backfilled trenches, Proc.,Environmental Geotechnics,
Kamon(ed), Balkema,Rotterdam,497-502.
Haug, M.D. (1983) Selection criteria for slurry trench cut-offs.
Canadian Journal of Civil Engg, 10(3),527-537.
Marston, A. and Anderson, A.O. (1913) The theory of loads on
pipes in ditches and tests of cement and clay drain tie and sewer
pipe. Iowa Engineering Experimental station Bulletin (31),
Ames, Iowa.
McCandless, R.M. and Bodocsi, A. (1988) Hydraulic
characteristics of model soil-bentonite slurry cut-off walls. In
Proceedings of the 5th National Conference on Hazardous
Wastes and Hazardous Materials.
Evans, J.C. 1994. Hydraulic conductivity of vertical cut-off
walls. Hydraulic conductivity and waster contaminant transport
in soil, ASTM STP 1142, D.E. Daniel and S.J Trautwein, eds.,
Chapman and Hall, London,430-454.
Ryan, C.R. 1987. Soil bentonite cut-off walls, Geotechnical
practice for waste disposal’87, R.D Woods, ed., ASCE, New
York, 182-204.
Terzaghi, K. (1945) Stability and stiffness of cellular
cofferdams. Transactions, ASCE, 110(2253), 1083-1119.
Xanthakos, P.P. (1979) Slurry walls, McGraw Hill, New York,
NY.
Yeo, S.S., Shackelford, C.D. and Evans, J.C (2005)
Consolidation and hydraulic conductivity of nine model soil-
bentonite backfills. Journal of Geotechnical and
Geoenvironmental Engineering, 131(10), 1189-1198.
Migration of Sulphate Solutions Through Cement-bentonite Diaphragms

E Fratalocchi
Technical University of Marche, Ancona, Italy (e.fratalocchi@univpm.it)
S Giorgini
Technical University of Marche, Ancona, Italy (simone.giorgini@hotmail.it)
E Pasqualini
Technical University of Marche, Ancona, Italy (e.pasqualini@univpm.it)

ABSTRACT The paper concerns the performance of cement-bentonite diaphragms as barriers against
migration of sulphate solutions, frequently met in pollutant soils. Preliminary experimental results of column
tests and batch tests are presented and interpreted aimed at evaluating the long term performance and
investigating diffusion and sorption, on the basis of the interaction mechanisms between cementitious
materials and sulphates. Column tests have been analyzed taking into account the variation in hydraulic
conductivity with time typically occurring in cement-bentonite mixtures.

INTRODUCTION concentrations are planned to be analyzed taking

Although cement-bentonite (CB) diaphragms are
often used in contaminated soils as barriers against
migration of pollutants, few data are available in
literature on their long term performance and on
migration parameters through this kind of barriers
(e.g. Manassero et al., 1995; Visudmedanukul et
al., 2007). In particular, referring to the effective
diffusion coefficient, D*, values similar to those of
compacted clays are generally assumed (2x10-10
m2/s to 2x10-9 m2/s for most chemical species;
Mitchell, 1991). Rarely D* is deduced from the
literature of cement pastes because it is typically
drawn from cured samples (D* of the order of 10-
11
m2/s for Na+, K+, Ca2+, SO42-, Cl-). To be on the
safe side, solute migration analyses through CB
barriers are sometimes performed with the
diffusion coefficient in free solution (D0) of the
species of interest and a high tortuosity factor, W
(D* = W D0).
A research program was set up aimed at
investigating interaction, diffusion and sorption of
sulphate solutions through CB mixtures. Both
curing time and a wide range of sulphate
into account the reduction of hydraulic
conductivity with time typically occurring in
cement-bentonite mixtures. The paper presents the
preliminary results available on interaction and
migration of highly concentrated aqueous
solutions of K2SO4 through a typical CB mixture.
MATERIALS AND TESTING METHODS
CB mixture
Composition of the CB mixture was selected to
minimise hydraulic conductivity (in aggressive
environment) and increase workability as much as
possible. The mixture consists of slag cement =
17.4% by weight (about 70% slag), activated Na-
bentonite = 3.6%, tap water = 79%. A special
additive (2 l/m3 of mixture) were added in order to
reduce its hydraulic conductivity and increase
workability. The mixture was prepared according
to the procedures recommended by ETC8 (1993),
poured in plastic moulds and submerged in tap
water. Properties of the fresh mixture met practice
requirements: volume weight = 1.16 g/cm3, Marsh
viscosity = 50 s (1000 ml), bleeding = 1.0 %.
 Fratalocchi, Giorgini and Pasqualini 793

Sulphate solutions Table 2 shows the composition of the solution at

Aqueous solutions of K2SO4 at 95 g/l, 50 g/l and equilibrium of the batch tests. The high alkalinity
27.5 g/l were used as permeants. Table.1 shows is mainly due to dissolution of calcium dioxide
the composition of the solutions for column tests (solubility =1.6 g/l at T=20°C), to the reaction of
and for batch tests. Distilled water and tap water sulphates with calcium dioxide to form gypsum:
were also used for batch tests to evaluate release Ca(OH)2 + SO42- + 2H2O => CaSO4ǜ2H2O + 2OH-
of K+.
and, minimally, also to dissolution of alkali (Na2O
Batch sorption tests and K2O) resulting from both bentonite and clayey
Since there are no standard procedures for batch components of cement:
tests on CB mixtures, tests were performed on the Na2O + H2O =>2Na(OH) => 2Na+ + 2OH-
basis of the guidelines for batch tests of soil (EPA, K2O + H2O => 2K(OH) => 2K+ + 2OH-
1992; ASTM D4319-93), with some differences. In particular, the higher the initial concentration of
CB samples were crumbled by forced passing sulphate, the lower the calcium concentration and
through the N.10 ASTM sieve. Mixture-solution the higher the pH at equilibrium. The quantity of
ratio of 1:2 (50 g of mixture : 100 ml of solution) gypsum that precipitates increases with the
were tested instead of 1:4, considering the high sulphate concentration.
water content of the mixture (about 250%). The
samples weren’t dried to avoid increasing of With reference to K+, of the theories presented in
curing and changes in microstructure. The mixture literature (Taylor, 1987; Stade, 1989) to explain its
was tested after 30-35 days of curing in water at sorption by cement pastes, the most probable is
room temperature. A contact period of 72 hours that binding occurs at relatively acidic Si-OH sites
resulted sufficient in all batch tests to reach (Hong & Glasser, 1999). In this hypothesis,
chemical equilibrium (EPA, 1992), according to bindings occurs as the acidic sites are neutralised
results in literature of batch tests on cement pastes by reaction with strong bases (KOH).
tested in similar conditions of temperature and
soil/water ratio (Hong & Glasser, 1999). Liquid TABLE 1 Chemical composition of the solutions used
and solid phases were separated by centrifuge and for batch tests and column tests.
concentration in the liquid phase was analysed by Solution SO42- K+ Ca2+ pH
ionic chromatography. (mg/l) (mg/l) (mg/l)
distilled water 0.00 0.0 0.8 6.75
Column tests tap water 71.35 7.9 86.1 7.35
Column tests were performed in flexible wall K2SO4 0.2 g/l 110 90 3.5 6.94
permeameters with bladder accumulators. Samples K2SO4 0.9 g/l 496 404 0.1 7.30
were consolidated after 14-17 days of curing at the K2SO4 5.0 g/l 2756 2244 1.3 6.80
effective confining pressure of 45 kPa; negligible K2SO4 27.5 g/l 15159 12341 1.6 7.02
K2SO4 50 g/l 27562 22438 10.4 7.34
reduction in volume (< 0.1%) were measured.
K2SO4 95 g/l 52368 42632 5.0 8.03
Samples were then permeated with the sulphate
solution under a constant hydraulic gradient of 30. TABLE 2 Chemical composition of the solutions at
Effluent samples were taken by a sampling port on equilibrium of batch tests.
the effluent line close to the sample effluent base.
Concentration, C, of SO4= and K+ were measured Test SO42- K+ Ca2+ pH
to get their breakthrough curves. solution (mg/l) (mg/l) (mg/l)
distilled water 4.57 620 428 12.75
tap water 5.47 697 412 12.77
RESULTS AND DISCUSSION
K2SO4 0.2 g/l 3.97 236.2 419 13.01
K2SO4 0.9 g/l 7.97 391.0 339 12.92
Batch tests K2SO4 5.0 g/l 18.73 1352 167 13.05
794 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

K2SO4 27.5 g/l 3718.0 7220 8.6 13.52 sulphate solutions and by a net of diffuse fissures
K2SO4 50 g/l 9121.2 12960 8.0 13.58 observed on the samples at the end of the test
K2SO4 95 g/l 25622 26520 12.2 13.70 (Fratalocchi & Pasqualini, 2007).
A linear isotherm (Fig. 1): 0.12

S = kd C e (1) 0.10

was found to well fit the data from batch tests of 0.08

S (-)
K+ (S = sorption degree [-], Ce = equilibrium 0.06
concentration of solute [ML-3], kd = distribution 0.04 kd = 3.90 ml/g
coefficient [L3M-1]). The same kind of isotherm 0.02
2
R = 0.98
was found for C-S-H (Hong & Glasser, 1999) up 0.00
to K+ concentrations of 12 g/l and values of kd 0 0.005 0.01 0.015 0.02 0.025 0.03
Ce (g/ml)
were found to range from 0.5 to 4 ml/g (depending +
on Ca/Si ratio). The kd value of the tested CB Fig.1 Linear sorption isotherm of K .
mixture belongs to this range and corresponds to 0.30
low Ca/Si ratio (=0.85) typical of blended cements, S = 27.81 Ce / (1+130.55 Ce)
0.25 2
such as the slag cement of the mixture investigated. S = 1.828 Ce 0.583
R = 0.98
0.20 2
R = 0.87
As far as SO4+ is concerned, results of batch tests S (-) 0.15
were fitted with linear (eq.1), Freundlich (eq. 2)
0.10
and Langmuir (eq. 3) isotherms, as shown in Fig.2: S = 7.301 Ce
S = kf CeH   
(2) 0.05 2
R = 0.80
S = SmbCe/(1+bCe) (3) 0.00

where kf = Freundlich partition coefficient [L3M-1], 0 0.005 0.01 0.015

Ce (g/ml)
0.02 0.025 0.03
2-
H = constant of the Freundlich isotherm [-], Sm = Fig.2 Sorption isotherms of SO4 .
constant of the Langmuir isotherm [-]b = rate of
0.20
sorption [L3M-1]. The Langmuir equation was 0.18
found to give the best fitting, but it could not be 0.16
0.14
used to interpret the breakthrough data and a bi-
0.12
linear isotherm was considered (Fig. 3).
S (-)

kd = 3.94 ml/g
0.10
0.08
0.06
Column tests 0.04 kd = 19.27 ml/g
Figure 4 shows the hydraulic conductivity, k, with 0.02 2
R = 0.93
curing time of the samples permeated with the 0.00
0 0.005 0.01 0.015 0.02 0.025 0.03
sulphate solutions. A decrease in k followed by an Ce (g/ml)
increase is evident in all the samples permeated Fig.3 Bi-linear sorption isotherm of SO42-.
with the K2SO4 solutions; in particular, the initial
reduction in the hydraulic conductivity exhibited
by the mixture during the first month of curing is
equal or even higher that the reduction of k in
water: this is probably be due to a partial clogging
as a consequence of gypsum formation. As soon
as the pore water is saturated of gypsum, ettringite
starts to form (Gollop & Taylor, 1992; 1995) and
such an expansive reaction is able to invert the
hydraulic conductivity trend with time up to a
rapid increase. The expansive reaction of ettringite
was confirmed by increase in volume (7-10%)
measured on the samples permeated with the

1.E-05
95 g/l
k (cm/s)
50 g/l
1.E-06 27.5 g/l
1.E-07
1.E-08
tap water
1.E-09
0 30 60 90 120 150
curing time (days)
Fig.4 Hydraulic conductivity with curing time of the
mixture permeated with K2SO4 solutions at different
concentrations (see the legend) and with tap water.
From Fig. 4 it is also evident that the trend of k
versus time seem to depend on the concentration
of sulphate: the highest the concentration of
sulphate, the more rapid the reduction of k at brief
curing; the subsequent increase of k occurs earlier
and is more rapid for the most concentrated
sulphate solution. After a fast increase, the
hydraulic conductivity of all samples permeated
with the sulphate solutions rises more and more
slowly up to almost the same constant value with
time; this trend should be due to the close net of
fissures (created by the expansive reaction) in all
the samples, that give rise to preferential pathways.
With reference to migration through the mixture,
Fig. 5, 6 and 7 show the breakthrough curves of
K+ and SO4= (time is from the beginning of the
test). Due to cement hydration and chemical
interactions (Fig.4), the seepage velocity changes
during the tests (performed at constant hydraulic
gradient) as well as the diffusion and dispersion.
The experimental data were fitted by the
POLLUTEv7 (2004) program, which solves the
one dimensional contaminant migration equation
subjected to different boundary conditions and
with time varying properties. Breakthrough data of
Fratalocchi, Giorgini and Pasqualini 795
SO4= and K+ were fitted assuming constant
concentration at the top boundary (Co) and infinite
thickness of the sample. Different time periods
were considered at constant v equal to the average
seepage velocity in each period.
Both kd and the coefficient of longitudinal
dispersivity, DL [L], were specified for each period.
The values of DL was chosen on the basis of
laboratory values measured on fine grained soils,
which commonly range between 0.1 and 10 mm
180 210 (e.g. Freeze & Cherry, 1979; Lee et al., 2009).
Fitting were tried with DL = 0.1 L (L = samples
thickness = 4.6-4.8 cm), but good fittings were
obtained with DL = 0.01 L (unless unrealistic kd
values). Values of kd higher than those from
isotherms of Fig.1 and 3 were required for a best
fitting. For each time period D* was got from the
best fitting of the breakthrough data.
Tables 3 and 4 summarise the results of the
column tests for K+ and SO4=, respectively, with
Dh [L2T-1] = hydrodynamic dispersion coefficient:
Dh = DL v n + D* (4)
where n = porosity, assumed equal to 0.6.
Further analyses will be conducted in the near
future (in particular, time periods of fitting will be
increased and effluent concentrations of several
ions will be analysed) for a better understanding
of reactions and migration mechanisms. However,
some general remarks can be drawn.
1.0
0.8
K2SO4 = 95 g/l
C/Co (-)
0.6
SO4=, measured
0.4
SO4=, fitting
K+, measured
0.2
K+, fitting
0.0
0 50 100 150 200 250
time (days)
Fig.5 Sulphates and potassium breakthrough curves of
the column test with K2SO4 at concentration = 95 g/l.
796 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

1
more favourable condition occurs in column tests
due to the continuous sulphate supply thus
0.8 promoting gypsum and then ettringite formation.
K2SO4 = 50 g/l Curing time and testing time being equal, the
C/Co (-)

0.6
lower the sorption rate observed, the higher the
0.4 SO4=, measured concentration level, according to the literature.
SO4=, fitting
0.2 K+, measured
K+, fitting
With reference to dispersion and diffusion, data in
0 Tables 3 and 4 must be analysed considering the
0 50 100 150 200 250 seepage velocity and the hydraulic conductivity
time (days)
trend with time. The first line of Tables 3 and 4
Fig.6 Sulphates and potassium breakthrough curves of
refers to k values < k with water, that is, clogging
the column test with K2SO4 at concentration = 50 g/l.
effect due to gypsum precipitation is still
equivalent or prevalent on the expansion due to
1 ettringite. In this period, both for K+ and SO4=, D*
0.8
tends to be lower the higher the seepage velocity;
K2SO4 = 27 g/l this trend is well evident if we consider that in
C/Co (-)

0.6 general diffusion decreases when C increases. In

this period the mechanical dispersion is negligible
SO4=, measured
0.4
SO4=, fitting
(except for the SO4= at Co = 27.56 g/l).
0.2 K+, measured When the effect of expansion due to ettringite
K+, fitting starts prevailing (values of the second line of
0 Tables 3 and 4), the k values of the samples
0 50 100 150 200 250 permeated with the K2SO4 solutions are more and
time (days)
more higher than the k values in water. The micro
Fig.7 Sulphates and potassium breakthrough curves of
the column test with K2SO4 at concentration = 27.5 g/l.
and macro-structure of the samples starts to be
Breakthrough curves of Fig. 5, 6 and 7 confirm a significantly modified, but dispersion is still
higher retardation of SO4= with respect to K+, due negligible and D* was found to be 2-3 x 10-6 cm2/s
to the formation of gypsum and ettringite; the for both K+ and SO4=. The diffusion coefficients
difference between the breakthrough curves of K+ maintain about the same values even when k starts
and SO4= tends to be more remarkable the lower increasing, at least up to about 69-79 days of
the influent concentration and the lower the curing continuous permeation (corresponding to 2-3 pore
time, that is, when the mixture has not yet volumes of flow). In this case, v is of the order of
significantly change its macrostructure. 10-6 cm/s (3-8 x 10-6 cm/s).
Only when the seepage velocity is of the order of
Referring to sorption capacity, Table 3 and 4 show 10-5 cm/s the mechanical dispersion starts to be
that to get a good fitting of the experimental data, influent. In this case, D* increases up to values of
kd values higher than those from batch tests had to 7x10-6 cm2/s, due to significant reduction of the
be used for K+ and for SO4= at the lowest tortuosity factor as a consequence of the sample
concentrations. The best performance in terms of expansion and fissures due to ettringite formation.
sorption of the samples in column tests respect to
TABLE 3 Sulphate migration parameters from column
batch tests could be due to a greater availability of tests of Figures 5, 6 and 7 (ve = effective velocity)
adsorption sites because of the lower curing of the period Co ve kd D* Dh
samples of column tests. Indeed, as found by (days) g/l cm/s ml/g cm2/s cm2/s
Visudmedanukul et al. (2007) for chloride, kd 0-25 52.368 7.2E-06 19 2.9E-06 3.2E-06
decreases as curing time increases. Moreover, for 26-48 52.368 1.8E-06 4 2.9E-06 2.9E-06
49-79 52.368 1.3E-05 4 2.9E-06 3.5E-06
sulphate reaction with hydrated cement products a 80-98 52.368 2.3E-05 4 7.5E-06 8.7E-06
 Fratalocchi, Giorgini and Pasqualini 797

99-113 52.368 3.6E-05 4 7.5E-06 9.3E-06 cm/s. This performance is quite poor for a CB
114-161 52.368 1.2E-04 4 7.5E-06 1.3E-05
mixture, considering that at the same curing time
0-10 27.562 9.8E-06 19 3.2E-07 7.8E-07
11-40. 27.562 3.4E-06 19 1.6E-06 1.8E-06 the k value in water is 3 orders of magnitude lower.
41-69 27.562 5.0E-06 8 1.6E-06 1.8E-06
69-107 27.562 3.9E-05 8 7.5E-06 9.4E-06 A linear isotherm was found to well fit the data
108-358 27.562 1.5E-04 8 7.5E-06 1.5E-05 from batch tests with reference to K+, with kd =
0-11 15.159 2.0E-06 19 3.2E-07 4.1E-07
12-50. 15.159 4.6E-06 19 2.2E-06 2.4E-06 3.90 ml/g. This value corresponds to data from
51-77 15.159 8.3E-06 8 3.2E-06 3.6E-06 literature on sorption of K+ by C-S-H (kd from 0.5
78-194 15.159 6.2E-05 8 3.2E-06 6.0E-06 to 4 ml/g depending on Ca/Si ratio).
195-271 15.159 1.3E-04 8 7.5E-06 1.4E-05 For sulphates the Langmuir equation was found to
give the best fitting of data from batch tests, with
TABLE 4 Potassium migration parameters from column
Sm = 4.69 ml/g and b = 130.55.
tests of Figures 5, 6 and 7 (ve = effective velocity).
period Co ve kd D* Dh Sorption of K+ and SO4= was found to be higher in
(days) g/l cm/s ml/g cm2/s cm2/s column tests than in batch tests; this difference
0-25 42.632 7.2E-06 3.8 2.2E-06 2.5E-06 was probably due to a longer curing time (30-35
26-48 42.632 1.8E-06 3.8 3.2E-06 3.3E-06 days) of the samples in batch tests than the curing
49-79 42.632 1.3E-05 3.8 3.2E-06 3.8E-06
80-98 42.632 2.3E-05 3.8 7.5E-06 8.6E-06
time at the beginning of the column tests (15-18
99-113 42.632 3.6E-05 3.8 7.5E-06 9.2E-06 days). If this will be confirmed, the curing time is
114-161 42.632 1.2E-04 3.8 7.5E-06 1.3E-05 a fundamental parameter to consider also in terms
0-10 22.438 9.8E-06 5 2.2E-06 2.7E-06 of sorption capacity for the design of CB barriers.
11-40. 22.438 3.4E-06 5 2.2E-06 2.4E-06
41-69 22.438 5.0E-06 5 2.2E-06 2.5E-06
As long as k of the CB mixture permeated with the
69-107 22.438 3.9E-05 5 7.5E-06 9.4E-06 sulphate solutions is equal to k measured in water
108-358 22.438 1.5E-04 5 7.5E-06 1.5E-05 or greater no more than one order of magnitude,
0-11 12.341 2.0E-06 5.5 3.2E-07 4.1E-07 D* of both K+ and SO4= were found to be < 3.2 x
12-50. 12.341 4.6E-06 5.2 2.2E-06 2.4E-06
51-77 12.341 8.3E-06 5.2 3.2E-06 3.6E-06
10-6 cm2/s. Higher values of D* were found on
78-194 12.341 6.2E-05 5.2 3.2E-06 6.0E-06 each sample after full reaction of expansion due to
195-271 12.341 1.3E-04 5.2 7.5E-06 1.4E-05 ettringite. Further analyses are necessary to
CONCLUSIONS confirm these values.
On the basis of results preliminary and analyses of Both for K+ and SO4= D* was found to be lower the
batch tests and column tests on a typical CB higher the seepage velocity. The seepage velocity
mixture with highly concentrated K2SO4 solutions, at which mechanical dispersion prevails on
the following conclusions can be drawn. diffusion was found to be of the order of 10-4 cm/s,
according to what found by Visudmedanukul et al.
Continuous permeation with sulphate solutions at (2007) for chloride ions through CB mixture.
C > 15 g/l increases the k values after just one
pore volume of flow since expansion due to REFERENCES
ettringite starts prevailing on clogging by gypsum EPA, US Environmental Protection Agency (1992)
and producing fissures. The effects of clogging Batch-type procedures for estimating soil
and subsequent expansion on the hydraulic sorption of chemicals. EPA/530/SW-87/006-F.
conductivity trend versus time are more rapid the ETC8, European Technical Committee N.8 (1993)
higher the sulphate concentration. Technical recommendation ‘GLC’. Geotechnics
After a fast increase, the k values of all the samples of landfills and contaminated land. German
permeated with the sulphate solutions rise more Geotechnical Society for the ISSMFE.
and more slowly up to almost the same constant Fratalocchi E. & Pasqualini E. (2007) Vertical
value with time, due to close net of fissures that barriers for side containment. Proc. XXI Turin
give rise to preferential pathways. The k value in Geotechnical Conference, CGT, Torino (Italy).
the long term (150 days of curing, corresponding Freeze R.A. & Cherry J.A. (1979) Groundwater.
to about 7-8 pore volumes of flow) is 1-3 x 10-6 Prentice-Hall Inc.
798 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
Gollop R.S. & Taylor H.F.W. (1992) Micro-
structural and microanalytical studies of sulphate
attack. I. Ordinary Portland cement paste. Cem.
& Concr. Res., 22(6): 1027-1038.
Gollop R.S. & Taylor H.F.W. (1995) Micro-
structural and microanalytical studies of sulphate
attack. III. Sulphate-resisting Portland cement:
reactions with sodium and magnesium sulphate
solutions. Cem. & Concr.Res., 25(7):1581-1590.
Hong S.-Y. & Glasser F.P. (1999) Alkali binding
in cement pastes. Part I. The C-S-H phase. Cem.
& Concr. Res., 29: 1893-1903.
Manassero M., Fratalocchi E., Pasqualini E.,
Spanna C., Verga F. (1995) Containment with
vertical cut-off walls. ASCE GSP46:1142-1172.
Mitchell J.K. (1991) Conduction Phenomena: from
theory to geotechnical practice. Geotechnique,
41 (3): 299-340.
Taylor H.F.W. (1987) A method for predicting
alkali ion concentrations in cement pore
solutions. Advanced in Cement Res., 1 (1): 5-16.
Stade H. (1989) On the reaction of C-S-H with alkali
hydroxides. Cem. & Concr.Res., 19: 802-810.
Visudmedanukul P., Kamon M., Katsumi T. (2007)
Chloride transport through cement-bentonite
barriers. J. Geot. Geoenv. Eng.,133(2): 175-185.
 Effect of Filter Cake on Evaluation of Hydraulic Conductivity of
Cutoff Wall in Analyzing Slug Test Results
H. Choi
Korea University, Seoul, Republic of Korea (hchoi2@korea.ac.kr)
T.B. Nguyen
Korea University, Seoul, Republic of Korea (nguyenthebao@korea.ac.kr)
C. Lee
Korea University, Seoul, Republic of Korea (cryfreer@korea.ac.kr)

Abstract: A 3D numerical model was developed to simulate the slug test in a vertical cutoff wall
with the presence of filter cake on the interface boundary between the cutoff wall and the natural soil
formation. The experimentally obtained characteristics of the filter cake were considered in a series
of simulations. The result of the simulations was then compared with the numerical simulations that
consider constant-head and no-flux boundary conditions applied to the interface boundary. The
comparison shows significance of the filter cake in the estimation of the hydraulic conductivity of
cutoff wall backfill with performing a slug test.

INTRODUCTION formation. Formation of bentonite filter cake

The hydraulic conductivity of vertical cutoff
walls can be estimated by the slug test using a
single well set up in the cutoff wall (Choi &
Daniel 2006a, b, Choi 2007). Recently, Choi &
Daniel (2006a, b) and Choi (2007) developed a
numerical program Slug_3D that is able to
simulate the slug test in the vertical cutoff wall
with consideration of compressible materials,
of different boundaries between the vertical
cutoff wall and the surrounding soil formation,
and of variable hydraulic properties with a
change of effective stress.
In a vertical cutoff wall construction,
bentonite-water slurries have been used to
effectively prevent the collapse of the trench
excavations, which are constructed for vertical
cutoff walls. The stability of trench excavation
is maintained by the lateral pressure difference
exerted by the slurry. Further, a thin and
impervious bentonite layer called filter cake
can be formed on the interface between the
vertical cutoff wall and the natural soil
during the construction of vertical cutoff walls
has been observed and reported by Henry et al.
(1998). However, the filter cake has not been
rigorously considered in slug test analyses for
vertical cutoff walls. Because the filter cake is
supposed to have a very low hydraulic
conductivity and exists between the cutoff wall
and the natural soil formation, it directly
influences the boundary conditions of the wall.
Therefore, the role of a filter cake in slug test
analyses to evaluate the hydraulic conductivity
of a cutoff wall needs to be considered
carefully.
In this study, the authors first experimentally
measured the hydraulic conductivity of the
filter cake in laboratory and then examined the
effect of filter cake on the slug test result in
vertical cutoff walls with the aid of the
modified program Slug_3D. Based on the
examination of the effect of filter cake, it is
found that the no-flux boundary condition on
the interface between the cutoff wall and the
800 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

natural soil formation can represent the filter
cake in the slug test simulations. The type
curve method (Cooper et al. 1967) was utilized
to reanalyze the case study that had been
performed by Choi & Daniel (2006a). The
comparison between previous and current
results highlights the importance of
considering filter cake in practice.
HYDRAULIC CONDUCTIVITY
FILTER CAKE
A range of hydraulic conductivity from
8.16x10-12 m/s to 3.51x10-10 m/s was reported
by Chung & Daniel (2008). A representative
thickness of 0.5 cm was reported and used for
interpretation of the in situ hydraulic
conductivity test results by Britton et al. (2004).
10-9
Hydraulic Conductivity k (m/s)
2.98u10-10 m/s (Nguyen et al. 2008). This
range is very similar to the ranges reported by
Chung & Daniel (2008).
EFFECT OF FILTER CAKE IN SLUG
TEST
The program Slug_3D was modified to include
the filter cakes on the interface boundaries.
The regulatory hydraulic conductivity required
OF for the backfill is specified equal to 10-8 m/s.
The thickness of the filter cake is chosen as 0.5
cm. Thus, the ratio of the hydraulic
conductivity of the backfill (kwall) to that of the
filter cake (kcake) is from 28 to 1225 (Chung &
Daniel 2008) and from 34 to 476 (Nguyen et al.
2008). The typical hydraulic conductivity of
the filter cake under examination is adopted as
100 to 1000 times smaller than that of the
backfill. This range corresponds to the typical
Tixoton hydraulic conductivities of the filter cake

Bentonil GTC4 presented previously. Given typical well and

DY - 100S
10-10
10-11
100
Overall Pressure p0 (kPa)
Fig.1 Hydraulic conductivity – overall pressure
relation of Tixoton, Bentonil GTC4, and DY – 100S
(Nguyen et al. 2008)
Chung & Daniel (2008) modified the fluid loss
test (ASTM D 5891 - 2002) to estimate the
hydraulic conductivity of filter cake effectively.
Three types of bentonite including Tixoton,
Bentonil GTC4, and DY – 100S produced by
different manufacturers were used for the
modified fluid loss test in this study. Results of
the experiment are presented in Fig. 1. The
range of hydraulic conductivities of the three
bentonite filter cakes is from 2.10u10-11 m/s to
cutoff wall geometries, the dimensionless
compressibility parameter Dp (= SsLwrw2/rc2,
where Ss = the specific storage, Lw = the length
of the filter pack, rw = the outside radius of the
filter pack, and rc = the inside radius of the
well casing) of 10-3 and 3.16u10-2 are
corresponding to the specific storage (Ss) of
4u10-5 m-1 and 1.264u10-3 m-1, respectively.
These values of specific storage properly
1000
represent the typical range of the
compressibility of soil-bentonite backfill
materials (Khoury et al. 1992). The slug test
simulations were performed with various
interface boundary conditions using Slug_3D.
The type curves are proposed by plotting the
normalized head recovery (H(t)/H0, where H(t)
= the hydraulic head drop at a certain time, H0
= the initial hydraulic head drop) versus
logarithm of a dimensionless time parameter,
ȕp (= kLwt/rc2, where k = the preset hydraulic
conductivity of the backfill, and t = time).
Figure 2 shows that the type curve for the case
of filter cake has a tendency getting close to
the no-flux-boundary type curve especially
when the specific storage increases.
Accordingly, if a filter cake forms on the

Fig. 2 Type curves for slug test simulations with various interface boundary conditions in cutoff walls
interface of the cutoff wall with highly
compressible backfill material, the no-flux
boundary condition can be a conservative
alternative to represent a cutoff wall boundary
condition in slug test analyses. The type curves
of the case of no-flux boundary condition were
built using the program Slug_3D. For the case
of less compressible backfill material, using
the no-flux boundary condition provides a
conservative approach on analyzing the slug
test data.
Choi, Nguyen and Lee 801
CASE STUDY
A case study from EMCON (1995) had been
performed by Choi & Daniel (2006a) without
the consideration of a filter cake (see Table 1).
In this study, the filter cakes were assumed to
exist. The type curve method for estimating
hydraulic conductivity of a cutoff wall with the
consideration of the filter cake was applied to
reanalyze the slug test data of the case study.
The type curves built with the no-flux
boundary condition were used. The estimates
of hydraulic conductivities in the modified
line-fitting method with the consideration of a
filter cake are about 65 to 165% higher than
802 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

TABLE 1 Case study results of vertical cutoff walls

Analysis without consideration of filter cake (Choi &
Analysis with consideration of filter cake
Daniel 2006a)
t1.0 Ssa kb t1.0 Ssa kb
Case Dp Dp
at ȕp =1.0(s) (mí1) (m/s) at ȕp =1.0(s) (mí1) (m/s)
93-1
7.9 x 10í3 8.0 x 103 3.9 x 10í4 6.8 x 10í8 1.3 x 10í1 2.2 x 104 6.7 x 10í3 2.5 x 10í8
(M-11/15)
94-15
7.1 x 10í3 2.2 x 104 6.2 x 10í3 5.2 x 10í8 3.9 x 10í2 4.0 x 104 3.5 x 10í2 3.0 x 10í8
(M-17/21)
a 2 b
D p § rc · 1.0rc2
Ss ¨ ¸ k
Lw ¨© rw ¸¹ Lwt1.0

the previous results presented by Choi & Methods. and II: Applications., J. Geotech.
Daniel (2006a), which do not consider a filter Geoenviron. Eng., 132(4): 429–447.
cake. This discrepancy shows the necessity to Chung, J., and Daniel, D. E. (2008). Modified
consider the filter cake in a slug test analysis. fluid loss test as an improved measure of
hydraulic conductivity for bentonite.,
CONCLUSION Geotechnical Testing Journal, 31(3), 243-
251.
The results from the case study prove the Cooper, H. H., Bredehoeft, J. D., and
significance of a filter cake in estimating Papadopulos, I. S. (1967). Response of a
hydraulic conductivity through a slug test. The finite-diameter well to an instantaneous
hydraulic conductivity of the cutoff wall will charge of water., Water Resour. Res., 3(1),
be underestimated in a slug test analysis if the 263–269.
filter cake is not properly considered. With the EMCON. (1995). M-11/15, M-17/21, and M-
measured hydraulic conductivities of the filter 26/E-29 slurry walls postconstruction
cakes and of the backfill material, the performance evaluation, West Contra Costa
equivalent hydraulic conductivity of the cutoff Sanitary Landfill, Richmond, Ca.
wall construction can be calculated for cutoff Henry L B, Filz, GM, and Davidson RR (1998).
wall performance evaluation. Formation and properties of bentonite filter
cakes., GSP No. 78, ASCE: 69–88.
REFERENCES Khoury, M. A., Fayad, P. H., and Ladd, R. S.
ASTM D5891 (2002). Standard Test Method (1992). Design, construction and
for Fluid Loss of Clay Component of performance of a soil-bentonite cutoff wall
Geosynthetic Clay Liners., American constructed in two stages, STP 1129, ASTM,
Society for Testing and Materials. Philadelphia, 289–308.
Britton, J. P., Filz, G. M., and Herring, W. E. Nguyen T.B, Lee, C, Yang J, and Choi, H
(2004). Measuring the hydraulic (2008). Evaluation of hydraulic
conductivity of soil-bentonite backfill., J. conductivity of bentonite filter cake using
Geotech. Geoenviron. Eng., 130(12), 1250– modified fluid loss test. Proc., KGS Fall
1258. Conference, Kwangju, Korea.
Choi H (2007). Numerical model for analyzing
slug tests in vertical cutoff walls., J.
Geotech. Geoenviron. Engrg. ASCE.
133(10): 1249–1258.
Choi H and Daniel DE (2006a, b). Slug test
analysis in vertical cutoff walls. I: Analysis
 CSM Cutter Soil Mixing and Technique for the Construction of Cut-
Off Walls
Wolfgang G. Brunner
BAUER Maschinen GmbH, Schrobenhausen, Germany, (Wolfgang.Brunner@bauer.de)
Holger Itzeck
BAUER Equipment of Canada Ltd., Edmonton, Canada, (Holger.Itzeck@bauer.de)
Renato Fiorotto
BAUER Maschinen GMBH, Germany

ABSTRACT The new Cutter Soil Mixing can largely replace the more conventional auger methods of soil
mixing or jet grouting. This system allows much deeper walls to be installed and enables hard soil layers to
be penetrated. The CSM technique differs from the traditional deep mixing method in so far as it makes use
of two sets of cutting wheels, as used on a trench cutter. As the cutter wheels rotate and penetrate into the
ground, they break up and loosen the soil. The CSM tool can be mounted on the Kelly bar of a drilling rig
or operated suspended from the rope of the base machine in order to achieve deep walls.

CUTTER SOIL MIXING CSM

In the year 2003 Bauer developed the CSM
method by exploiting its experience in the
manufacture and use of the trench cutter systems
to excavate diaphragm walls panels. The CSM
method differs from the traditional DMM method
in so far as it makes use of two sets of cutting
wheels that rotate around a horizontal axis to
produce rectangular panels of treated soil rather
than one or more vertical rotating shafts that
produce circular columns of treated soil. Two
cutter gear boxes are connected to a special
mounting that is connected to a robust kelly bar.
The kelly bar is connected to the mast of a drill
rig by two guide sledges that steer and provide
crowd and extraction forces (Fig. 1). As the
cutting wheels rotate and penetrate into the
ground they break up and loosen the soil. During
this phase a fluidifying agent or the binder itself
is injected into the area between the two cutting
wheels. In the extraction phase the cutting wheels
rotate in a mixing mode and blend the binder and
soil to form a rectangular panel of treated material Fig. 1. CSM head mounted to Kelly bar
in a cut-off wall. (For colour figure, refer to CD)
804 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
The parameters of the machines we have
designed and built to-date are:
Length of panel 2200, 2400 & 2800 mm
Width of panel 500 mm to 1200 mm
Torque output 30 kNm, 50 kNm, 80 kNm
Weight of cutter unit 3700 kg, 5100 kg, 7400 kg
Maximum depth capability with single tube Kelly
35 m
Maximum depth capability with cable suspension
70 m
There is a much lower risk of leakage through the
CSM panel (Fig. 3).
There are a number of other advantages that the
CSM method and machinery offers when
compared to the traditional rotating augers or
paddles, notably: the only moving parts in the
CSM method are the cutting wheels, as a result
we can mount instruments inside the cutter
gearbox support frame that give real time
information throughout the treatment depth,
information such as verticality, deviations, excess
pressure build-up in the surrounding soil etc.
Fig. 2. Wire rope-suspended CSM unit
(For colour figure, refer to CD)
The CSM tool as rope suspended version is used
when greater depth has to be achieved (Fig. 2).
The capacity to reach great depths offers an
enormous potential for the construction of deep
cut-off walls for dams or the encapsulation of
contaminated sites.
Fig. 3. Top of excavated CSM wall
(For colour figure, refer to CD)
CSM CUTTER SOIL MIXING PROJECT
EXAMPLES
Cut-off Wall MOSE Flood Control Project
Venice, Italy
In a highly sensitive environment like the Venice
Lagoon the new CSM method has proven highly
effective in producing an important cut-off wall
and respecting the stringent environmental
constraints.
During the last century, “high waters” have
invaded the built up areas in the Venice lagoon
with increasing frequency and intensity. The
worsening of this phenomenon is due to the
difference in the height of the land with respect to
the sea which, between the beginning of the 20th
century and today is of more than - 23 cm as a
result of the rise of the level of the sea and
subsidence of the land.
As well as causing considerable inconvenience to
the population and businesses, high waters lead to
the slow but relentless degradation of the
environment and of the artistic and architectural
heritage of Venice. There is also a very high risk
 Brunner, Itzeck and Fiorotto 805

of a catastrophic event such as the 4 November the gates. In order for the pit to be dewatered an
1966 flood where built up areas in the lagoon impermeable cut-off needs to be installed to
were completely submerged by up to a metre of circumscribe the dock and to extend to a depth of
water. 28 m where it will socket into a relatively
In 1984, after many discussions and fiery contro- impermeable clayey layer (Fig. 6).
versies work began on a General Feasibility Plan A number of solutions to produce the cut-off wall
covering all the measures necessary to preserve the were evaluated by the contractor, amongst these
hydro-geological balance of the lagoon and to were sheet piles, diaphragm walls, plastic
mitigate high waters in the historic city and town diaphragm walls, secant piles and CSM. The CSM
centres. Finally in 1989 a conceptual design was alternative proved to be the optimum solution.
completed for an integrated system of various types The specifications required the cut-off wall to
of measures that would provide complete protect- have two main characteristics: a permeability
tion for all built-up areas in the lagoon from high better than 10-6 cm/sec and a strength better than
waters of all levels. This included mobile barriers, 500 kPa.
that is, rows of gates to be installed at the three The first step in the operation was to take a
lagoon inlets that are able to isolate the lagoon from number of representative samples of soil from the
the sea during high tides above an established level; site to a laboratory where they were mixed with
a series of complementary structures such as break- different quantities of water + cement and with
waters outside the inlets designed to attenuate the plastic binders. The cement used was a
levels of the most frequent tides and the raising of pozzolanic cement and all samples were prepared
quaysides and paving in the lowest lying urban with potable water but left to hydrate under a
areas. head of water sampled from the lagoon.
In 2003 the MOSE project to save Venice was Quantities and proportions were chosen to
given the go-ahead to start. The expected duration represent cement contents in the final soil/binder
of construction is 8 years (Fig. 4). mix of 200, 250 and 300 kg/m3 of soil.
The mobile barriers are the heart of the defence
system and consist of a total of 79 gates, each 5 m
thick, 20 m wide and 30 m long, placed in the inlet
channels, they are designed to withstand a
difference in level of up to 2 metres between the
sea and the lagoon. When at rest the gates are full
of water and lie in caissons in the sea bed. When a
high tide (above +110 cm) is forecast compressed
air is introduced into the gates, expelling the water,
they swing upwards until they emerge above the
water level and block the tidal flow entering the
lagoon. The mobile barriers remain in position for
the duration of the high water only.
To allow port activities to continue and pleasure
craft, emergency vessels and fishing boats to
shelter and pass through when the barriers are in
operation, a lock will be constructed at the
Malamocco inlet for the larger ships. Small craft
harbours and locks will also be constructed at the
Lido and Chioggia inlets (Fig. 5). Fig. 4. Venice Lagoon
One of the small craft harbours at the Lido inlet (For colour figure, refer to CD)
will serve initially as a construction pit for the
precast caissons that will form the foundations for
806 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

Plastic diaphragm
wall
Sea
+4 Lagoon
0

-5
A
-10

-15 B
-20 C

-25 D

Fig. 6. Cross Section through Cut-off Wall

The Two phase method – Primary and secondary

Fig. 5. Malamocco Ship Lock
(For colour figure, refer to CD)
Permeability and compressive strength tests were
carried out. These gave good results with
compressive strengths well over the 500 kPa limit.
Average permeabilities at 28 days were of the
order of 4.03 x 10-7 cm/sec.
Given the depth of the wall we chose to use the
cable suspended CSM machine mounted to a
Bauer BG 28 base carrier.
The following alternative construction methods
were tested:
The One Phase method – Primary and secondary
panels and a series of fresh-in-fresh panels were
mixed using a plastic mix on both the down and
up strokes. In this case the intersection between
the adjacent primary panels and the secondary is
700 mm.
panels were mixed using a pure bentonite mix on
the down stroke and a water + cement binder,
with a w/c ratio of 0.8 on the up-stroke. In this
case the intersection between the adjacent
primary panels and the secondary is 300 mm.
During this work, spoil from the panels was
collected in a pit and its volume was carefully
measured as the specifications required this to be
minimized. The one phase system produced the
lesser amount of spoil – 40% of the total panel
volume compared to 47% for the two phase
system.
Upon analysis of the test results, values of
permeability and strength were very similar for
the two methods of execution but even though the
one phase method produced less spoil the two
phase system was chosen for the project. This
choice was based upon the condition that when
carrying out CSM it is important that at the end of
the down stroke the soil/binder mix above the
CSM be fluid enough to facilitate extraction of
the machine. On some areas of this site however
the salt content in the soil is such that when work
was carried out using the one phase method, it
causes the soil/cement mix to lose fluidity very
rapidly and thereby to make extraction of the
machine difficult. In the two phase method
however the soil/bentonite mix above the
machine remains very fluid and the extraction
process becomes easy thereby increasing
productivity and causing the soil/binder mix
below the machine to be well blended and very
homogenic.

Verticality of the panels was another important
parameter on this site as it is fundamental that the
wall be impermeable throughout its length and
depth. Deviation readings from the machine’s
computer however indicated that the maximum
deviations were less than 0.3%, ie. 84 mm on the
“X” axis and less than 0.2%, ie. 50 mm on the
“Y” axis (Fig. 7).
Fig. 7. Operator Cabin with Data Computer
(For colour figure, refer to CD)
Given that the work is being carried out using the
two phase method the contractor carried out a
series of primary panels for 3 days and then
returned to carry out the secondary, closure
panels. This work programme optimised produc-
tivity as after three days the machine was able to
intersect the primary panels with some ease.
Average daily productivity was 3 panels or 235
m2. On peak production days they were able to
mix 4 panels or 310 m2.
Cut-off Wall, Flow Path Protection Project
Leuna, Germany
The town of Leuna is situated approximately 25
km south of the City of Halle (Saale) in the State
of Saxony-Anhalt. Leuna was and still is one of
the most important synthetic chemical industry
sites in Germany. In 1916 Leuna began its
chemical industrial development and became
Germany’s largest synthetic oil factory and
producer of fuel from liquefied coal during World
War II. As the Leuna plant played such a key
role in the German war economy, the plant was
Brunner, Itzeck and Fiorotto 807
the target of numerous air attacks in which around
80% of the Old Refinery was destroyed. During
the 1950s fuel production was switched
completely to the processing of mineral oils.
After its closure in 1995/96, the Old Refinery was
demolished completely. Left behind were large
quantities of contaminants that had seeped into
the subsoil and groundwater as a result of the
effects of the war, accidents, leakages and
spillages during processing and handling. It is
estimated that fuel in excess of 10 million litres
seeped into the ground, contaminating
approximately 1.8 million tonnes of soil with fuel
at the site of the Old Refinery in the form of
petroleum hydrocarbons, together with other
problematic compounds such as the fuel additive
MTBE (methyl-tertiary-butyl-ether) manufac-
tured on the site as an anti-knock agent. Over a
period of some 80 years, the Leuna chemical
complex had developed into a mega-
contaminated site and was having a significant
impact on the area’s ecological and human
environment by discharging approximately 1,400
kg of MTBE and 800 kg of benzene annually in
the direction of the River Saale. By 2004, the
flow of contaminated groundwater in the form of
an extensive "contamination plume" had reached
a length of several hundred metres and was
threatening the town’s Water Works some 2000
m east of the Old Refinery.
In view of the extend of the contamination and
the imminent thread this caused to the local
community, the commercial, technical, social and
political stakeholders of the site finally agreed to
act. After approval of the overall remediation
strategy, both the concept and engineering design
was completed and construction works could
finally begin.
As it had been acknowledged by all stakeholders
that the site could not be “cleaned up“ in the short
term and a complete removal of the soil and
groundwater contamination was practically
impossible, the remediation strategy focused on
the containment of this legacy of industrial
pollution. Central to the containment strategy
was the construction of an impermeable cut-off
wall at the southern flank of the contaminant
source zone that would protect the downstream
808 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
groundwater flow path from further
contamination. The contract for the design and
supervision of the construction of the scheme was
awarded to the “ARGE Abstromsicherung“, a
joint venture between GUT Gesellschaft für
Umwelttechnologie GmbH and GuD Consult
GmbH.
The core element of the Southern Downstream
Flow Path Protection Scheme was the
construction of an impermeable subterranean
barrier in the shape of a single-phase cut-off wall
that would prevent the outward flow of
groundwater and thus the continued migration of
contaminants. The contract for the construction of
the cut-off wall was awarded to BAUER of
Schrobenhausen. The successful bid was based
on the proposal to employ the Cutter-Soil-Mixing
(CSM) technique developed by BAUER
Maschinen.
Based on a geological and hydro-geological
subsoil investigation, the effectiveness of the cut-
off wall was designed for the contaminant-rich
water bearing formations in the Quaternary
deposits and the upper aquifers in the Tertiary
formations, in order to prevent contaminant
transport via hydraulic windows between the
Quaternary and the Tertiary (Fig. 8).
Fig. 8. Basic design concept of the Southern
Downstream Flow Path Protection Scheme with
Groundwater isohypses (For colour figure, refer to CD)
To prevent groundwater from banking up in front
of the cut-off wall, the contaminated groundwater
is collected in a deep collector drain system
upstream of the cut-off wall, from where it is
pumped to a groundwater treatment plant and
cleaned by a multi-stage system of chemical and
physical remediation processes. The treated
groundwater is then discharged via a pipeline into
infiltration basins located downstream of the cut-
off wall, from where it is reintroduced into the
downstream groundwater flow path through
infiltration or recharge wells.
The volume of water being treated annually
currently amounts to 90,000 m³. After an initial
pilot operation, the plant went into permanent
operation on schedule. In line with the current
state of technology, the Southern Downstream
Flow Path Protection Scheme is expected to be in
operation over a period of some 15 years (Fig. 9).
Fig. 9. Excavated top of Cut-off Wall
(For colour figure, refer to CD)
CONCLUSION
CSM is an extremely competitive and useful Soil
Mixing method. In both CSM projects MOSE
and Leuna, where site constraints and ground
conditions were particularly difficult, the method
proved to be technically superior and more cost
effective than other common methods.
Meanwhile many projects using the CSM-System
were carried out successfully in Germany,
Netherlands, Belgium, Italy, Switzerland, Portugal,
Japan, USA, Canada, Australia and New Zealand.
 Electromagnetic Stimulation of Air Sparging for
Geoenvironmental Applications
Arvin Farid
Boise State University, Idaho USA (ArvinFarid@BoiseState.edu)
Harlan Sangrey
Boise State University, Idaho USA(harlansangrey@u.BoiseState.edu)
Jim Browning
Boise State University, Idaho USA (jimbrowning@BoiseState.edu)

ABSTRACT Cleaning contaminated soil/groundwater is important to federal and state agencies. Traditional
contaminant removal methods are costly and impractical for large sites. Less disruptive remediation techniques
(e.g., air sparging) are attractive but limited by the restriction of airflow in soil.
The use of electromagnetic (EM) stimulation to expedite airflow and diffusion, and control air channel formation
is investigated. The stimulation of water is expected to increase the diffusion of air between air channels and
expand the air channels. EM fields can be emitted by a dipole antenna into the medium to oscillate water
molecules to enhance diffusion and enlarge air channels.

INTRODUCTION levels of effectiveness. Excavation can cause

Large numbers of underground chemical,
petrochemical and gasoline storage tanks around
the country and world are aging and beginning to
leak or have been leaking for years. The
contamination due to these spills can spread into
the ground. Without any treatment, the problem
can spread to drinking water and harm living
organisms in the area. If left untreated, the spills
will take tens or hundreds of years to dilute and
degrade to an accepted level by drinking-water
standards. The need for cleanup of environmentally
hazardous spills has been increasing and is not
expected to diminish. The ability to treat large
areas in an efficient cost-effective manner is
necessary for feasibly faster treatment of
contaminated sites. Removal of contaminated soils
can be very intrusive and expensive.
There are many cleanup techniques for
environmental spills. Gasoline and other volatile
chemicals have proven very difficult to remove
from soil, whether leaked from a tank or seeped
from surface spills. The ex-situ method of
excavation and removal of contaminated soils is
very expensive and destructive but has been used
for many years. Other alternatives to excavation
and treatment have been developed with different
further propagation of mobile contamination.
Ex-situ remediation can be cost-effective and
efficient only on small spills. Ex-situ
remediation is very destructive, expensive,
and, hence, ineffective for large spills,
especially when the contaminants are mobile.
Bioventing is a method used in unsaturated
soils for remediation (Dupont 1993).
Bioventing utilizes low-volume airflow in soil
to promote microbial activities and
biodegradation. The transfer of oxygen to in-
situ organisms enables aerobic metabolization
of contaminants. Bioventing injection is often
used in conjunction with an extraction well to
help to remove the vapors from the soil.
Slurping is another method, which is used to
remove contaminants, usually oil or LNAPLs
(light non-aqueous phase liquids), from the top
of the groundwater table. In slurping, a
vacuum is used to “slurp” the layer of less
dense and water-phobic oil or non-aqueous
phase liquid (LNAPLs) that are in the region
above the water table beneath the unsaturated
soil above (Weingardt & Gonzales 2002).
Normally, large amounts of water and vapors
810 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
are removed and treated as parts of this
remediation process. The hydrocarbon
contaminants are removed before the water is
returned to the ground.
Steam injection is also used as another method of
in-situ remediation. In this method, steam is
pumped into the ground helping the volatile
compounds to vaporize. The vapor can, then, be
removed more easily with an extraction well (Falta
2001). Although less effective, vacuum extraction-
wells are used as standalone remediation tools.
Air sparging is an increasingly popular method for
remediation of sites contaminated with NAPL
spills. The method of pumping air into the
contaminated groundwater is shown to be effective
in reducing contamination to acceptable levels.
This process can take months or years, and can,
therefore, be very costly. Air sparging is limited by
air channel formation (Elder & Benson 1999),
which limits the mass transfer zone (MTZ) (Braida
& Ong 2001). The process can be expedited by
promoting air saturation between the air channels,
enlarging air channel size, and increasing diffusion
of contaminants through the media. One of the
main existing methods to enhance air sparging is
pulsating, which is time-consuming and
uncontrolled. Radio frequency (RF) bioheating has
been examined (Vermelulen & McGee 2000) and
shown successful for the enhancement of some of
these methods. However, only preliminary work
was done, and the process was not controlled and
increased the temperature to 100oC+, which harms
bacteria present in soil.
Several methods of remediation are in use today.
The efficiency and effectiveness of these methods
are being improved constantly. This paper focuses
on improving airflow to enhance several of these
methods (that use air injection or vacuums, such as
air sparging or bioventing) using EM-stimulation.
This will help the cleanup to become more
effective and less-costly.
METHODOLOGY
To experimentally model the stimulation of air
channels and groundwater, series of bench-top tests
with several apparatuses are conducted. The first
objective is to prove that EM waves stimulate
air sparging in a different mechanism beyond
heating. In other words, the main motivation
behind the selection of EM stimulation is to
prove that the frequency of EM waves can be
controlled to reduce the emitted power to
maintain a low temperature, and at the same
time, stimulate air sparging. The effect of
temperature increase and heating is compared
with EM stimulation. The first series of tests
use a non-reactive dye to compare the effect of
temperature and EM stimulation on diffusion.
The tests, in which thermal heating is used, are
herein called thermal tests. The comparison
between the thermal and EM stimulation
effects is studied using two different
approaches: (i) by comparing the thermal
temperature increase that can create the same
level of diffusion stimulation as EM waves;
and (ii) by comparing the level of diffusion of
the EM-stimulated tests with the non-
stimulated tests at the same medium
temperature. The EM-stimulated tests use a
submerged PVC-cased monopole antenna as
the source of EM stimulation. The other series
of tests that use the submerged PVC-cased
monopole antenna, as well as the thermal tests
are performed in a 1000 mL volumetric flask.
The volumetric flask is used to enable good
preliminary visualization and stimulation area.
I. Thermal Tests
The thermal tests were performed in the lab
where a water bath was used to maintain the
temperature of the test volumetric flask
constant. During preliminary tests, and
because the dye is denser than water, it was
observed that the dye rapidly descended and
went through a splash-like turbulence
propagation. Therefore, the dye is injected at
the bottom of the medium to minimize any
mixing mechanism except diffusion. The
temperature of the test media and water bath
were monitored closely using thermometers to
ensure no rapid thermal changes to ensure
stability of the temperature. The water used as
the test media is from the tap and the dye used
is McCormick® green food color. The dye was

introduced into the water in the volumetric flask
and the time was recorded. The time was stopped
when the dye in the volumetric flask reached a
steady state (uniform concentration).
II. EM-Stimulated Diffusion Tests
The experiments of this task are performed in a
volumetric flask stimulated with relatively high-
power EM waves. The equipment for creating this
stimulation is discussed later in the paper. The
water in the volumetric flask is stimulated by a
PVC-cased monopole antenna submerged in the
medium.
Fig.1 Experimental setup and devices
(For colour figure, refer to CD)
The entire system is placed in a large Faraday cage
to contain the EM waves, to prevent interference
from outside the experiment, and to protect
researchers. High-power EM waves can be
harmful, if not contained (Jamieson et al. 2007).
The Faraday cage comprises of a wooden frame
with an aluminum screen stretched across and with
a removable door for access, to create a uniformly
conducting cage. All seams of the box were taped
using aluminum tape to ensure conduction. The test
apparatuses for the Thermal and first EM
Stimulation tests are shown in Fig.1.
III. EM-Stimulation Experimental Setup
Preparation for In-Situ Simulation
The simulated air sparging tests are performed in
an acrylic container stimulated with relatively
high-power EM waves (33 Watts). These tests are
performed in a 60 cm × 40 cm × 15 cm acrylic box.
The walls and all dividers are made of 0.625 cm
thick acrylic plates, while the base is made of a
1.25 cm thick plate for higher strength. All pieces
were assembled and water sealed using a clear
Farid, Sangrey and Browning 811
cement adhesive. All hose fittings and flow-
controlling ball valves are made of
nonmetallic, nonmagnetic material (plastics)
so as not to interfere with the stimulating EM
field. Removable dividers were added to the
box to create two side reservoirs to control
lateral flow and a bottom reservoir for
saturation from the bottom. Glass bead media
is used in this box to simulate in-situ
conditions. With a homogeneous glass-bead
medium, used to simulate air sparging in-situ
tests, the pluviation method (Vaid & Negussey
1998; Yegian et al. 2007) is used. The dividers
are perforated with 0.625mm holes and
covered with geo-textile fabric to contain the
glass beads.
The diffusing matter, in this task, is air. The
water is filled from the bottom to help to
eliminate any initial air bubbles in the
medium. Then, air is injected through a PVC
pipe at a depth of 20 cm, which has a filter
over the end to keep out the glass beads. Glass
beads (used to simulate a transparent soil
medium) are Ballotini “A” spherical blast
media, which are 600-850 microns in
diameter. The injected air is supplied using a
central compressed air system, controlled with
a pressure regulator, and monitored with a
flowmeter and a thermometer.
The test medium is stimulated by two different
methods. Source 1 is a monopole antenna (an
N-type coaxial cable, with 25.4 mm of the
conducting shield stripped at the end) placed
in a PVC casing, submerged in the medium;
and Source 2 is a plane wave generated within
the test medium using two parallel plates.
The cage used in all the experiments was built
large enough not to create a path for the EM
field more preferential than the one between
the two plates (Source 2). In other words, the
air-filled distance between the Faraday cage
and the feed plate is larger than the distance
between the feed and ground plate normalized
to the square root of the dielectric constant of
812 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

the stimulated medium between the two parallel perforated PVC pipe. The PVC pipe is
plates. perforated ensuring enough flow to achieve an
accurate reading of the surrounding medium.
The amplifier used to increase the signal strength is The PVC pipe is fixed in the box for the full
a Model 100LMB amplifier manufactured by test to ensure sufficient consistency in readings
Amplifier Research (AR). A matching network is without disturbing the testing medium.
used to match the impedance of the medium and
the system to the one of the amplifier. The network
added was an L-shaped matching network shown
in Fig.2. The L-shaped matching network circuit is
built in a “BUD” box (i.e., aluminum box to
contain the EM waves). This is to minimize the
reflection at the interface between the medium of
interest and amplifier back into the amplifier, and
hence, ensure maximum power transfer into the
medium as well as for the safety of the amplifier.
Fig.2 Schematic of network
The matching network includes a 10 µH inductor
and a 14-380 variable two-gang capacitor in
Digital images are taken of the diffusing phase
parallel to tune the impedance of the testing
(dye in the first series of experiments; and air
medium to the impedance of the amplifier (50 ȍ).
channels during the air sparging in the second
The components are connected with 50ȍ RG8
series of test) with and without stimulation.
coaxial cables. This impedance matching is
The images are compared to analyze the effect
performed for a single frequency within the
of the oscillating H20 molecules on air channel
frequency range of optimum performance of the
formation and air or dye diffusion. All images
amplifier (50 MHz to 230 MHz). For different
are taken with a Cannon Rebel T2i digital SLR
frequencies, the inductor can be removed if the
camera (without use of a flash and controlled
load is too inductive. At some frequencies, the
background lighting using LED lights)
variable capacitor is utilized alone to match the
controlled with a Linux-based code for
network. By varying the capacitance, signal power
accurate image timing.
is maximized to a 50ȍ inductance. The matching
network is shown in Fig.2.
Air is to be injected at a pressure that will be
determined experimentally. The pressure will
To measure the circuit impedance for network
vary to maximize the visibility of
matching, the system (the antenna submerged in
channel/bubbles formation.
the medium), the cables, and the matching network
are connected to an Agilent Technologies E5071C
SUMMARY AND RESULTS
vector network analyzer (VNA). Once the network
Tests were performed to compare the
is matched, the signal at the selected frequency is
stimulating effects of temperature and EM
created using an Agilent E4400B function
waves on diffusion, as well as to minimize and
generator run through the amplifier. The signal is,
monitor the temperature change during EM
thereafter, sent through a dual-directional coupler,
stimulation. The relationship between
which uses an Agilent N9320A Spectrum Analyzer
diffusion and temperature was observed to be
to monitor the forward power and the reflection
exponential. As the temperature is increased,
back toward the amplifier. The dissolved oxygen
the diffusion rate increases exponentially. The
concentration is measured at intervals through the
temperature versus the natural logarithm of the
test by inserting a dissolved oxygen (DO) sensor.
inverse of time required to reach a uniformly
The total dissolved gas pressure is also measured
diffused concentration (a steady state) is
using a total dissolved gas (TDG) sensor cased in a
shown in Fig.3.
 Farid, Sangrey and Browning 813

Average Temp. vs ln(Delta Time^-1)

During testing, the temperature was monitored
0
and recorded so that any effect of heating on
-0.5
0 5 10 15 20 25 30 35 40 45 50 diffusion could be evaluated and monitored.
-1
There was no measurable increase in the
ln (Delta T im e^-1) (1/m in )

-1.5
y = 0.1192x - 6.2873
temperature during EM-stimulated testing.
-2
R2 = 0.907 Therefore, no microwave heating effects were
-2.5 present, and yet the diffusion was considerably
-3 stimulated.
-3.5

-4
Temp (deg C)

Fig.3 Temperature vs. natural logarithm of

inverse of time to reach a completely diffused
state

Figure 3 shows the increase of the diffusion rate

due to thermal stimulation. The time to reach a
steady state (uniformly diffused concentration) is
proportional to the diffusion rate and so is used in
Fig.3 for the comparison.

The initial tests have been performed to show the Fig.4 Comparison between stimulated and
diffusion of a non-reactive dye in water for non-stimulated diffusion
stimulated and non-stimulated cases. Time to reach
the steady state was significantly reduced when Using a broadcast dual-directional coupler, the
stimulated. Looking at the simple diffusion partial forward and reflected power to the antenna is
differential equation: constantly measured. These tests are
wU performed at 89.6 MHz where the direct
’ 2 ( DU ) (1)
coupler is measured to have an attenuation
wt
where U is the concentration, t is the time, and D is factor of -40 dBm. The measured power from
the diffusion coefficient; the stimulated time to the attenuated port was 5 dBm, which implies
reach the steady state is on average 52% of the that the 32 Watts of EM power was emitted
time required to reach the steady state in the non- into the medium. Using a probe connected to
stimulated condition. Through a simple the spectrum analyzer and inserted into the
approximation, it can be observed that D2 (the medium, the radiation pattern of the EM power
stimulated diffusion coefficient) has increased over emission into is mapped at only 89.6 MHz.
92% over D1 (the non-stimulated diffusion However, without calibrating the probe, the
coefficient). This supports the hypothesis that the exact power collected by the probe cannot be
EM power introduced into the medium has measured. Further calibration of the probe or
significantly increased the diffusion rate, and acts use of power-meters is needed for accurate
like a stimulation source for diffusion without power measurements.
heating, as expected. The plotted test results are
shown in Fig.4. The solid line is the average time CONCLUSIONS
to reach the steady state for the non-stimulated The initial tests have shown that diffusion
tests. The dashed line shows the average time to through the test medium is stimulated using
reach the steady state for the stimulated tests. EM waves. The EM power, at the selected
frequency, does not measurably increase the
temperature but does increase the diffusion
814 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

rate, which confirms the hypothesis of EM- Jamieson, K.S., ApSimon, H.M., and Bell,
stimulation without any considerable microwave J.N.B. (2007). “Electrostatics in the
heating. The diffusion rate was increased by environment: how they may affect
approximately 92%. This increase can be justified health and productivity,” Journal of
by the EM power absorbed by the dipole water Physics: Conference Series, pp. 1-7.
molecules from the oscillating EM field. The
increase in the diffusion rate can help airflow Sahuquet, B.C., Spreux, A.M., Corre, B.,
between air channels formed through air sparging Guittard, M.P., and Aquitaine, E.
and, hence, expedite remediation. (1990) “Steam injection in a low-
permeability reservoir through a
Further investigation is being pursued to observe horizontal well in Lacq Superieur
the effect of the stimulation on air channel field,” Proceedings – SPE Annual
formation. The changes in the air channels will be Technical Conference and Exhibition,
studied using measurement of channels’ diameter September 23-26, New Orleans,
and tortuosity. More efforts will focus on studying Louisiana, Paper #: 20526-MS, pp.
the correlation among the frequency, power level 559-568.
and the stimulation, to minimize the level of
introduced energy and, hence, microwave heating. Vaid, Y.P., and Negussey, D. (1998).
“Preparation of reconstituted sand
ACKNOWLEDGEMENTS specimens,” Advanced triaxial testing
This project is supported by the National Science of soil and rock, ASTM STP 977,
Foundation (NSF) through IDR program (CBET Philadelphia, pp. 405-417.
Award Number 0928703).
Vermeulen, F., and McGee, B. (2000). “In Situ
REFERENCES Electromagnetic Heating for
Braida, W., and Say Kee, O. (2001). “Air sparging Hydrocarbon Recovery and
effectiveness: laboratory characterization Environmental Remediation” Journal
of air-channel mass transfer zone for VOC of Canadian Petroleum Technology,
volatilization,” Journal of Hazardous Vol. 39, No. 8, pp. 25-29
Materials, B87, pp. 241-258. Weingardt, K., and Gonzales, M. (2002). “Soil
Dupont, R.R. (1993). “Fundamentals of bioventing and Groundwater Remediation at an
applied to fuel contaminated sites,” Abandoned Fuel Farm (Site 7) using
Environmental Progress, Vol. 12, No. Vacuum-Enhanced Pumping, Naval
1, pp. 45-53. Air Facility, El Centro, California,”
Elder, C., and Benson, C.H. (1999). “Air channel Soil and Sediment Contamination: An
formation, size, and spacing, and International Journal, Vol. 11. No. 3,
tortuosity during air sparging,” Ground pp. 390-391.
Water Monitoring & Remediation, pp.
171-181. Yegian, M.K., Eseller-Bayat, E.,
Alshawabkeh, A.N., and Ali, S.
Falta, R.W. (2001). “Steam flooding for (2007). “Induced partial saturation for
environmental remediation,” Manuals and liquefaction mitigation: experimental
Reports on Engineering Practice, investigation,” Journal of
American Society of Civil Engineers Geotechnical and Geoenvironmental
(0734-7685), (100), 153 p. Engineering, ASCE, Vol. 133, No. 4,
pp. 372-380.
 Natural Attenuation and Biostimulation Techniques in Soil
Contaminated with Biodiesel

A. Thomé
University of Passo Fundo, Passo Fundo – RS, Brazil (thome@upf.br)
L. R. R. Meneghetti
Federal University of Rio Grande do Sul, Porto Alegre – RS, Brazil (lilianerebechi@yahoo.com.br)
F. Schnaid
Federal University of Rio Grande do Sul, Porto Alegre – RS, Brazil (fernando@ufrgs.br)
P. D. M. Prietto
University of Passo Fundo, Passo Fundo – RS, Brazil (pdmp@upf.br)

ABSTRACT: Ground contamination caused by fuel leak is one of the main research subjects in
the geoenvironmental area. In this study, the decontamination of a basalt residual soil from
southern Brazil contaminated with biodiesel was evaluated through two bioremediation
techniques: natural attenuation and biostimulation (nutrients added: Nitrogen, Phosphorus, and
Potassium). The soil was characterized physically, chemically and microbiologically, and
contaminated by adding 4% of biodiesel in relation to the dry weight of soil. Monitoring of
biodegradation was carried out through the measure of CO2 evolution and by performing
gaseous chromatography. It was possible to conclude after 120 days that the biostimulation was
more effective than the process of natural attenuation in biodiesel degradation.

INTRODUCTION
Several methods can be used to remove
hydrocarbons from soil and groundwater.
Processes of remediation such as air spraying,
soil vapor extraction, and in-situ or ex-situ
bioremediation are examples of such methods
(Spinelli 2005).
Bioremediation is a suitable process of cleaning
up soils with petroleum hydrocarbons and
organics contaminants, as it is simple to
maintain, applicable over large areas, cost-
effective, and leads to the complete destruction
of the contaminant (Frankenberger 1992). It is
also a process where the microorganisms
degrade organics contaminants and immobilize
inorganic contaminants (USEPA 1991). From
the biodegradation of organic pollutants, like
biodiesel, for instance, microorganisms use the
carbon (C) present in the soil-contaminat system
as a source of energy and growth. Microbial
degradation of hydrocarbons results in the
release of innocuous products such as CO2,
H2O, and cellular biomass.
Many factors are important to the
biodegradation process. However, basic
information is needed, such as the residual oil
concentration, the population density of the oil-
degrading microorganisms, chemical, physical,
and microbiological properties of the
contaminated soil, and the biodegradation
816 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
potential when optimum rates of fertilizers are
applied (Frankenberger 1992; Bento 2005).
Efficient and low-cost methods for in-situ
bioremediation of contaminated soil include
natural attenuation, biostimulation (with the
addition of nutrients), bioventing (with the
addition of oxygen), and bioaugmentation (with
the introduction of specific microorganisms).
These methods are reviewed in numerous recent
publications reporting on the bioremediation of
gasoline (Solano-Serena et al. 1999; Cunha &
Leite 2000; Passman et al. 2001; and Spinelli et
al. 2002), diesel oil (Richard & Vogel 1999;
Olson et al. 1999; Capelli et al. 2001; Spinelli et
al. 2005; Bento et al. 2004; and Meneghetti
2007), and biodiesel (Fernández-Alvarez et al.
2007; De Mello et al. 2007; and Meneghetti,
2009).
In the natural attenuation process, indigenous
microorganisms occurring in the soil (fungi and
bacteria), degrade the contaminants for their
survival. However, depending on the type of
contaminant and its toxicity levels, specific
microbes might be introduced into the soil
(Sharma 2004).
The biostimulation technique involves the
addition of nutrients, electron acceptors (or
electron donors), and, sometimes, auxiliary
substrates to stimulate the growth and activity of
the indigenous microbial population (Alvarez
2006).
To warrant practical application, any
bioremediation process should demonstrate that
removal of contaminants is the primary effect of
biodegradation, and that the degradation rate is
greater than the natural rate of decontamination.
However, the main difficulties lie in found as
good as or better results in the field than those
found in laboratory scale tests (Juhasz et al.
2000).
In this study, the decontamination of a basalt
residual soil from southern Brazil contaminated
with biodiesel was evaluated in the laboratory
through two bioremediation techniques: natural
attenuation and biostimulation, with the addition
to soil of Nitrogen, Phosphorus, and Potassium
(NPK).
MATERIALS AND METHODS
Soil
The soil utilized in this study is a B-horizon
basalt residual soil, commonly found in
southern Brazil. The samples were retrieved at a
depth of 1.20m from ground surface and
submitted to physical-chemical and
characterization, which are summarized in
Table 1.
Bioremediation treatments
The bioremediation comparative experiment
was carried out on 24 samples divided in four
treatments: T1 – control (natural soil), T2 –
natural attenuation (contaminated natural soil),
T3 – biostimulation (natural soil plus nutrients),
and T4 – biostimulation (contaminated natural
soil and nutrients), as shown in Table 2.

The contaminant used was pure biodiesel, which
composition is shown in Table 3. For soil
contamination, it was used 25mL of biodiesel
per kg of dry soil, corresponding to 2,5% mass
contamination.
 Thomé et al. 817

TABLE 1 Physical and chemical characteristics The bioremediation tests were performed by
of the basalt residual soil placing each sample (500g of dry soil,
Properties Values contaminant, and nutrients) into a hermetic 2L
Liquid limit (%) 53 capacity glass vessel (see Figure 1) during a
Plasticity limit (%) 42 period of 120 days.
Plasticity index (%) 11
Real density of grains (Ȗs) 2,7
Clay (%) 57
pH (H2O) 5,4
Ind. SMP 5,4
P (mg/dm3) 3
3
K (mg/dm ) 35
Organic Matter (%) 0,8
Al (cmolc/dm3) 2,0
Ca (cmolc/dm3) 0,6 Fig.1 System used to capture CO2.
Mg (cmolc/dm3) 0,3 (For colour figure, refer to CD)
a 3
CEC (cmolc/dm ) 9,6
3 TABLE 3 Composition of Biodiesel.
S (mg/dm ) 8
B (mg/dm ) 3
0,2 Fatty acid
Numeric Trivial
3 esters
Mn (mg/dm ) 4 symbol name
(%)
3
Zn (mg/dm ) 0,1 C14:0 Myristic 0,06
3
Cu (mg/dm ) 0,7 C16:0 Palmitic 11,58
a C18:0 Stearic 5,1
cation exchange capacity.
C18:1 Trans Oleic 0,0
TABLE 2 Bioremediation Treatments. C18:1 Cis Oleic 24,14
Treatments Addition Values C18:2 Cis Linoleic 52,2
T1 C18:3 Cis Linolenic 7,47
- - C20:0 Arachidic 0,46
(Control)
T2 Biodiesel 25mL/kg of soil
T3 NPK Solution 2,5mL/kg of soil Microbial activity
Biodiesel 25mL/kg of soil The indigenous soil microorganisms were
T4 + + qualitatively and quantitatively characterized
NPK Solution 2,5mL/kg of soil through the methods described by MacFaddin et
al. (2000) and MAPA (2003), respectively.
The soil-contaminant system was biostimulated
by adding a solution of NPK (2.5 mL/kg of dry
The indigenous microorganisms Bacillus sp.
soil) containing NH4NO3 (0.275g/50mL of
and Psudomonas sp. were found in the natural
distilled water) and KH2PO4 (0.875g/50mL of
soil at the concentrations of 1.93 x 106 UFC/g.
distilled water).
These microorganisms are referred in the
818 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
literature as being very effective at the
biodegradation of organics pollutants (Alvarez
& Illman 2006).
The microbial activity was monitored by the
evolution of CO2, as described by Ölinger et al.
(1996).
Chromatographic analysis
The chromatographic profiles of the samples
submitted to the bioremediation treatments were
analyzed at 120 days and compared with the
chromatographic profile of pure biodiesel. The
chromatographic analysis was performed by
using a gas chromatograph equipped with an
ionization detector flame.
The percentage of contaminant degradation was
calculated by the following expression:
Percentage of degradation = [(esters in control –
esters in treatment)/ester in control] x 100.
RESULTS
Microbial activity by CO2 evolution
The CO2 evolution, monitored throughout the
testing period at specified time intervals (4, 6, 8,
10, 12, 15, 20, 25, 30, 50, 70, 90, and 120 days),
is shown in Figure 2, in which the CO2
quantification can be observed for the four
different treatments: (T1) control, (T2) natural
attenuation (natural soil with biodiesel), (T3)
biostimulation (NPK), and (T4) biostimulation
(NPK plus biodiesel). The CO2 accumulated
during the tests is summarized in Figure 3, in
which the columns and bars indicate the means
and the standard deviations of the accumulated
CO2 at 120 days.
It can be readily noticed that the quantity of CO2
increased throughout the testing period. At the
early stages of the tests, it was observed an
evolution of CO2 similar between treatments
(phase adaptation of microorganisms to the
environment). After the 8th day, however, the
growth phase started and the treatment T4
differentiate from the other treatments,
presenting a higher microbial activity until the
end of the tests. The addition of biodiesel and
nutrients (NPK solution) to the soil produced
changes in the cumulative evolution of CO2
when compared with the control treatment.
It was also observed that the lower microbial
activity occurred for the control treatment T1
(92 mg/kg of CO2), while the treatments
containing biodiesel (T2) or NPK (T3), showed
similar CO2 evolution, of approximately 112
mg/kg.
The treatment T4 (biodiesel and NPK) was the
one with higher microbial activity over the
period investigated, with approximately
200 mg/kg of CO2.
 Thomé et al. 819

250
T1 control (natural soil)

T2 natural attenuation (natural soil + biodiesel)

200
CO2 evoluid (mg.kg-1)

T3 biostimulation (NPK)
150 T4 biostimulation (NPK + biodiesel)

100

50

0
4 6 8 10 12 15 20 25 30 50 70 90 120
Time (days)

Fig.2 Accumulated evolution of CO2.

nutrients or other improvements in the

120 environmental conditions, can occur continuously
CO 2 evoluid (mg.kg-1 )

100
due to the adaptation of the native soil
80
microorganisms impacted.
60
40
Chromatographic Analysis
20
Figure 4 shows the percentage of total degradation
0
T1 T2 T3 T4
of methyl esters present in biodiesel by the
Treatments processes of natural attenuation and biostimulation
Fig.3 Means and standard deviations of the at the end of the tests (120 days).
accumulated CO2 at 120 days.
100
degradation (%)

These results indicate that the C:N:P:K 80

proportion used in treatment T4 was effective 60
for the biodegradation of the organic pollutant 40
(biodiesel) by the indigenous soil 20

microorganisms. 0
natural attenuation biostimulation
bioremediation process
According to Baptista & Rizzo (2004), the
process of natural attenuation of organic Fig.4 Percentage of total degradation of methyl
pollutants in soils, without any additional esters in biodiesel at 120 days.
820 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
Among all the treatments investigated in this
study, the highest degradation rate of esters
was observed when a nutrient solution was
added to the soil-contaminat system.
In fact, the biostimulation proved to be the
most efficient process for the
decontamination of a residual clayey soil
contaminated with biodiesel, when compared
to the process of natural attenuation. The
effect of adding nutrients to the soil promoted
a biodegradation of about 90% of the esters in
biodiesel, after 120 days of monitoring. In the
same period, the process of natural
attenuation achieved only 17% of biodiesel
biodegradation.
On the other hand, the natural attenuation
treatment showed a low percentage of
biodiesel biodegradation, probably because
the time needed for this process to be
completed is usually longer when compared
to other bioremediation processes.
CONCLUSIONS
In the present work, a study was carried out to
evaluate the potential for bioremediation of
biodiesel in a basalt residual soil by two
bioremediation process: natural attenuation
and biostimulation. The main conclusion
drawn from the results are:
x The microbial activity was constant during
the period investigated (120 days);
x The biostimulation technique was found to
be the most effective for the bioremediation
of biodiesel, when compared to the natural
attenuation process.
ACKNOWLEDGEMENTS
The authors wish to express their gratitude to
CAPES and CNPq – Brazilian Council of
Scientific and Technological Research for their
financial support to the research group.
REFERENCES
Alvarez P.J.J. and Illman W.A. (2006)
Bioremediation and Natural Attenuation
Process. Fundamentals and Mathematical
Models. New Jersey. John Wiley & Sons. 609.
Baptista P.M. and Rizzo A.C.L. (2004)
Monitoring the natural attenuation of soil
contaminated by oil. XII Journey of Scientific
Initiation of CETEM/MCT. (In Portuguese)
Bento F.M. et al. (2005) Comparative
bioremediation of soils contaminated whit diesel
oil by natural attenuation, biostimulation and
bioaugmentation. Bioresource Technology, v.96,
1049-1055.
Capelli S.M., Busalmen J.P. and Sanches S.R.
(2001) Hydrocarbon bioremediation of a
mineral-base contaminated waste from crude oil
extraction by indigenous bacteria. International
Biodeterioration and Biodegradation, v. 47,
233-238.
Cunha C.D. and Leite S.G.F. (2000) Gasoline
biodegradation in different soil microcosms.
Brazilian Journal. of Microbiology, v.31, 45-49.
USEPA – U. S. Environmental Protection Agency
(1991), Understanding Bioremediation: A
Guidebook for Citizens, EPA/540/2-91/002,
Office of Research and Development, USEPA,
Washington, DC.

Fernández-Alvarez P. et al. (2007) Evaluation of
biodiesel as bioremediation agent for the
treatment of the shore affected by the heavy oil
spill of the Prestige. Journal of Hazardous
Materials, v.147, 914-922.
Frankenberger W.T. (1992) The need for a
laboratory feasibility study in
bioremediation of petroleum hydrocarbons.
In: Calabrese, E.J., Kostecki P. T (Eds),
Hydrocarbon contaminated soils and
groundwater. Lewis, Boca Raton, FL, 237-
293.
Juhasz A. et al. (2000) Degradation of high
molecular weight PAHs in contaminated
soil by a bacterial consortium: effects on
microtox and mutagenicity bioassays.
Bioremediation Journal, v. 4, 271-283,
MAPA – Ministério de Estado da Agricultura,
Pecuária e Abastecimento e da Reforma
Agrária. (2003). Instrução Normativa n° 62
de 18 set. Métodos Analíticos Oficiais para
Análises Microbiológicas para Controle de
Produtos de Origem Animal e Água.
Brasília. (In Portuguese)
Meneghetti L.R.R. (2007) Bioremediation in
the descontamination of a residual soil of
basalt contaminated with diesel oil and
biodiesel. (Engineering Master Theses).
University of Passo Fundo. RS. Brazil. (In
Portuguese)
Meneghetti L.R.R. et al. (2009). Monitoring
of bioremediation in clay soil contaminated
with biodiesel. V Seminar of Geotechnical
Engineering of Rio Grande of Sul. Pelotas,
Brazil. (In Portuguese)
Thomé et al. 821
Öhlinger R. et al. (1996). Methods in Soil
Biology. Springer- Verlag: Berlin,
Olson J.J. et al. (1999) Biodegradations rates
of separated diesel components.
Environmental toxicology and chemistry, v.
18, n.11, 2448-2458.
Passman F.J. et al. (2001) Oxygenated gasoline
biodeterioration and its control in laboratory
microcosms. International Biodeterioratin &
Biodegradation, v.47, 95-106.
Richard J.Y. and Vogel T.M. (1999)
Characterization of a soil bacterial consortium
capable of degrading diesel fuel. International
Biodeterioration & Biodegradation, v. 44, 93-
100.
Solano-Serena F. et al. (1999) Biodegradation of
gasoline: Kinetics, mass balance and fate of
individual hydrocarbons. Journal of Applied
Microbiology, n.86, 1008-10016.
Spinelli L. F. (2005) Bioremediation, toxicity and
cell damage in spills of gasoline. Thesis (PhD in
Engineering), UFRGS. (In Portuguese)
Spinelli L.F. et al. (2005) Enhancing
bioremediation of diesel and gasoline in soil
amended with an agroindustrial sludge. Journal
of the air & Waste Management Association,
v.55, 421-429.
 Soil Remediation by High Vacuum Extraction Techniques
M. Alferi
DITAG, Politecnico di Torino, Torino, Italy (manuela.alferi@polito.it)
A. Dominijanni
DITAG, Politecnico di Torino, Torino, Italy (andrea.dominijanni@polito.it)
M. Manassero
DITAG, Politecnico di Torino, Torino, Italy (mario.manassero@polito.it)

ABSTRACT: A laboratory experimental activity has been carried out in order to evaluate the advantages to
apply high vacuum to extract volatile and semi-volatile pollutants from the subsoil. A new apparatus,
consisting of a cell connected to a vacuum pump and two differential pressure transducers, has been built.
The vacuum extraction has been simulated at different average gas pressures lower than the atmospheric
one, maintaining a constant mass flux. The obtained results show a significant (about 35%) decrease of the
clean-up time when the average gas pressure is reduced from atmospheric pressure (# 101 kPa) to a
pressure equal to 10 kPa.

INTRODUCTION 5) last but not list, to increase the contaminant

Recently, a new contaminant extraction system
(SGIBITI) has been proposed by Manassero
(2003) and Dominijanni et al. (2007). The
system is fundamentally based on the well
known technology of the vacuum consolidation
that is currently used in the traditional
geotechnical field for the improvement of the
mechanical behaviour of fine grained soft soils.
It consists of a confinement of the polluted
volume of soil by a capping geomembrane and
side cut-off walls providing a series of injection
and extraction wells for flushing fluids (Fig. 1).
The extraction of pollutants is enhanced by the
high vacuum application that promotes the
contaminant volatilisation and with the other
components of the system is able 1) to reduce
the drainage path length; 2) to widen the
contaminant extraction surface including the
ground surface; 3) to squeeze the contaminant
from the pore fluids of fine grained soft soils; 4)
to reverse the flushing fluid flow direction and
concentration in the effluent gas phase.
In order to study the effects produced by the
application of high vacuum to gas flow in soils,
a series of experimental investigations have
been carried on small scale soil samples in the
laboratory. The preliminary results of this study
are presented in this paper together with some
practical consideration related to field
applications.
MATERIALS AND METHODS
In order to simulate the effects of the
application of vacuum on soil at the laboratory
scale, a new apparatus has been set up (Fig. 2).
The equipment consists of a cell containing the
soil sample, a vacuum pump (Laboxact Vacuum
System, KNF Neuberger) and two differential
pressure transducers placed near the top and the
bottom of the cell. The effluent mass flow of air
is measured by means of a flow meter.
 Alferi, Dominijanni and Manassero 823

Fig. 1 High vacuum extraction system (SGIBITI) (For colour figure, refer to CD)

The pump used is a diaphragm vacuum pump and are characterized by a dry density, rd, of
connected with a system for controlled and 1.63 Mg/m3 for the first one (sand) and 1.7
gentle distillation. Mg/m3 for the second one (silty sand).
Thanks to a valve placed at the bottom of the The contaminant used in the work is 2,2,4-
cell (inlet valve), the inlet line to the sample is Trimethylpentane, also know as Iso-octane, a
at atmospheric pressure. Opening or closing the volatile hydrocarbons (VOC – Volatile Organic
aforementioned valve and controlling the pump Compound) with density lower than water
power, the pressure inside the sample and the (LNAPL – Light Non Aqueous Phase Liquid).
effluent air flow can be regulated. Table 1 summarized the main physical
In this study, tests were carried out applying characteristics of this liquid.
different pressures lower than the atmospheric
pressure (from 90 kPa to 10 kPa) and
maintaining constant the effluent mass rate of
air flow.
Two type of tests were carried out: the first one
considering only the steady state air flow in the
soil and the second one aimed at simulating the
extraction of a volatile contaminant varying the
average pressure.
Samples used in the experimental activity were
prepared using two different type of natural soil.
The first sample is made only with a natural
sand (Ticino Sand), while the second one is silty
sand obtained mixing 75% of Ticino Sand and
25% of Pontida Silt. Samples are artificially
Fig. 2 Experimental apparatus for the application of
compacted in dry condition directly in the cell vacuum (For colour figure, refer to CD)
824 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

TABLE 1. Physical properties of Iso-octane L = distance between the transducers (m).

Dynamic Vapor Under the assumption of constant hydraulic
Density,
Viscosity, Pressure, conductivity, Eq. 1 suggests that the plot of
r (Mg/m )3
m (Pa·s) Pv (kPa) Qm˜L/'P versus the average pressure, Pavg,
at 20°C at 20°C
at 22°C should be given by a linearly increasing
0.69 5.10·10-3 1.5 relation. The plot of the experimental data of the
sand sample, reported in Figure 3, is in good
agreement with this theoretical result.
RESULTS The test has been successively repeated on two
The experimental activity has been aimed at samples of silty sand, a soil mixture
investigating the air flow under steady state characterized by a higher deformability than
conditions and evaluating the clean-up time clean sand. In this case, a significantly different
under different average air pressures. The behaviour has been observed. In fact, when the
obtained results are reported in the next average pressure has been reduced inside the
paragraphs. silty sand samples, the hydraulic conductivity
changed in response to suction stress.
Steady state air flow
In order to evaluate the air flow under steady
state conditions, a series of preliminary tests
3
have been carried out on a sand sample (Table SAND
2), at constant mass rate of air (Qm # 3˜10-2 g/s) 2.5
(a)
and different average pressures. The average
 (10-5 g*m*s-1Pa-1)

pressure has been reduced by partial closing of 2

the inlet valve and the mass rate has been

1.5
maintained constant by increasing the pump
vacuum. The pressures, measured by the 1
QmL
DP

transducers inside the sand sample, have been

reported in Table 2. 0.5

Under steady state conditions, the mass rate of

0
an ideal gas flow, Qm (g/s), can be expressed as
0 20 40 60 80 100
follows: Average Pressure, Pavg (kPa)
3
SILTY SAND
K Pavg M 'P 2.5
Qm A (1) (b)
 (10-6 g*m*s-1Pa-1)

P RT L
2

where: 1.5
A = transversal area of the sample (m2);
K = intrinsic hydraulic conductivity (m2); 1
QmL
DP

P = air dynamic viscosity (Pa˜s); 0.5

Pavg = (Pin + Pout)/2 = average air pressure (Pa);
Pin = air pressure at the inlet transducer (Pa); 0
Pout = air pressure at the outlet transducer (Pa); 0 20 40 60 80 100
Average Pressure, Pavg (kPa)
R = 8.314 J˜mol-1˜K-1 = universal gas constant;
T = absolute temperature (K); Fig. 3 Experimental data relative to steady state air
M = 28.9644 g/mol = molar mass of air; flow for clean sand (a) and silty sand (b).
'P = Pin  Pout = pressure difference (Pa);
 Alferi, Dominijanni and Manassero 825

TABLE 2. Experimental Data

Mass
Pressure Volume
Inlet Outlet Average Pump Rate of Pump
Difference, Gas
Sample Test Pressure, Pressure, Pressure, Pressure, Gas Power,
Flux,
Pin (kPa) Pout (kPa) DP (kPa) Pavg (kPa) Pp (kPa) Flow, Wp (W)
q (m/s)
Qm (g/s)
1 97.66 97.46 0.20 97.56 90.4 3.01·10-2 4.99·10-3 0.15
-2
2 88,97 88.69 0.28 88.83 81.6 2.91·10 4.70·10-3 0.40
3 70.33 69.76 0.57 70.04 62.6 3.01·10-2 6.55·10-3 1.21
SAND
4 36.30 35.25 1.05 35.77 28.4 3.11·10-2 1.32·10-2 4.96
5 22.90 20.35 2.55 21.62 13.6 3.11·10-2 2.19·10-2 10.01
6 13.47 8.03 5.44 10.75 2 2.91·10-2 4.13·10-2 21.49
-2
1 94.76 89.86 4.90 92.31 82.40 3.01·10 4.97·10-3 0.37
2 72.55 67.50 5.05 70.03 60.30 3.01·10-2 6.55·10-3 1.30
SILTY 3 56.87 51.72 5.15 54.30 44.40 2.91·10-2 8.17·10-3 2.42
SAND 4 35.83 30.60 5.23 33.22 22.80 2.91·10 -2
1.33·10 -2
5.60
Sample 5 17.19 13.19 4.00 15.19 5.80 2.81·10-2 2.82·10-2 15.09
1 6 10.55 7.72 2.83 9.14 2.90 2.91·10-2 4.85·10-2 25.90
-2 -2
7 10.12 7.42 2.70 8.77 2.60 2.91·10 5.06·10 27.07
8 17.23 12.75 4.48 14.99 6.80 3.01·10-2 3.06·10-2 15.12
9 15.56 11.86 3.70 13.71 4.50 3.01·10-2 3.35·10-2 16.96
SILTY
SAND 10 29.19 24.05 5.14 26.62 16.40 2.96·10-2 1.69·10-2 7.60
-2 -3
11 58.28 53.16 5.12 55.72 45.80 3.01·10 8.23·10 2.29
Sample
12 68.35 62.99 5.36 65.67 56.00 3.01·10-2 6.98·10-3 1.55
2
13 95.19 90.1 5.09 92.65 83.60 3.01·10-2 4.95·10-3 0.34

In particular, the observed increase of Qm˜L/'P
for a decreasing average pressure is indicative
of an increase of the hydraulic conductivity (see
Eq. 1), that can be attributed to the shrinkage
effect undergone by the sample under higher
values of applied vacuum.
The different behaviour of clean sand and silty
sand samples can be explained by the different
deformability of the two soils. The shrinkage
effect, in fact, is more relevant in the case of
silty sand due to its higher deformability with
respect to clean sand. This effect determines the
volumetric compression of the sample and
promotes the formation of a lateral flow of air at
the induced gap between the sample and the cell
wall.
The test results point out the relevance of soil
deformability and suggest care for the
evaluation of gas flow in field application of
vacuum techniques, especially with reference to
fine grained soils where the shrinkage effect can
determine crack formation.
Clean-up time
The main advantage expected from the
application of high vacuum is the reduction of
time necessary to extract the pollutant from the
subsoil (clean-up time). This advantage is
particularly attractive for the remediation of fine
grained soils that can not be cleaned in
acceptable time using conventional soil venting
techniques. For this reason, the tests on silty
sand samples have been carried out with an
initial contamination of Iso-octane, assumed as
an analogue of volatile oils, in order to evaluate
the reduction of clean-up time obtained by
reaching low average pressures. In these tests,
the evaluation of gas flow has been referred to
826 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

100
the steady state conditions, in correspondence of
Silty Sand - SAMPLE 1
that the Iso-octane was completely removed. Silty Sand - SAMPLE 2

During the first phase of the tests, the 80

progressive extraction of Iso-octane, initially

Average Pressure, Pavg (kPa)

injected at the middle of the samples, has been 60

monitored through a visual observation of the

reducing wet area of the sample.
40
Figure 4 shows a series of photos at different
times from the beginning of the tests at different
average vacuum pressures. It is possible to 20

immediately observe that for a given time after

the beginning of the test, the sample with the 0

higher applied vacuum shows a reduced residual 4000 5000 6000

Clean-up Time, tc (s)
7000 8000

oil content. Based on the visual observation, an
assessment of the clean-up time has been
obtained.
Figure 5 shows the time to the end of the
treatment (i.e. in correspondence of the total
extraction of contaminant) versus the average
vacuum pressure applied. It is possible to
observe a reduction of more than 35% in
extraction time from traditional low vacuum
technologies (90 kPa) to the high vacuum
proposed system (10 kPa), at a given mass gas
flow rate.
The reduction of clean-up time decreasing the
average pressure can be related to two different
phenomena.
Fig. 5 Clean-up time for total contaminant extraction
versus average absolute pressure
The first one is the increase of advective
transport of Iso-octane gas phase, due to the
condition of constant air mass flux and the
simultaneous increase of volumetric flux due to
the rarefaction of air. In fact the advective gas
flux of Iso-octane, Jad, can be expressed as
follows:
J ad q ˜C (2)
where:
q = volumetric flux of air (m3˜m-2˜s-1);
C = pollutant concentration in the air (g/m3).
The volumetric flux of air can be estimated
from the mass rate of air flow through the
following relation:

Qm
q (3)
U avg

where
Uavg = average density of air (g/m3).
The average density of air is related to average
pressure, previously defined, through the law of
ideal gas:
Fig. 4 Experimental contaminant extraction
visualization at different times and absolute Pavg M
U avg (4)
pressures. (For colour figure, refer to CD) RT

Based on Eq. 2-4, the advective transport
increases when the average pressure of air
decreases. The increase of the volumetric flux
decreasing average pressure in the tests on silty
sand samples is shown in the data reported in
Table 2 and is plotted in Fig. 6.
The second phenomenon that can contribute to
reduce the clean-up time is the increase of
volatilisation rate at lower average pressures.
The volatilisation rate, Jvol, can be expressed as
follows:
J vol k oa (C eq  C)
where:
koa = mass transfer coefficient between oil and
air phases (1/s);
Ceq = equilibrium gas concentration of pollutant
(g/m3);
C = gas concentration of pollutant (g/m3).
The mass transfer coefficient is a lump
parameter that is proportional to the pollutant
gas diffusion coefficient, Dg.
0.06
0.05
Volume Air Flux, q (m/s)
0.04
0.03
0.02
0.01
0 XXI Ciclo, Geosintetici e Ambiente, Patron
0 20 40
Average Pressure, Pavg (kPa)
Fig. 6 Volumetric air flux as a function of average air
pressure in the tests on silty sand samples.
Based on the kinetic theory of ideal gas, the gas
diffusion coefficient is inversely proportional to
the average gas pressure (Bird et al., 2002):
Alferi, Dominijanni and Manassero 827
1
Dg v (6)
Pavg
As a result, the volatilisation rate is inversely
proportional to the average air pressure. When
the average air pressure is reduced, the increase
of volatilisation rate promotes the phase change
from liquid to gas, speeding up the pollutant
extraction through the mobilization of the
gaseous phase.
(5) CONCLUSIONS
The preliminary results of a laboratory
experimental activity aimed at evaluating the
effects of high vacuum application for the
remediation of soil polluted by volatile and
semi-volatile contaminants have been reported
and interpreted. The advantage to apply high
vacuum is the reduction of clean-up time due to
an increase of the advective flux and the
volatilisation rate. However, it should be
stressed out the relevance to account of soil
deformability and crack formation for a suitable
evaluation of gas flow both in laboratory and
field application.
SAND
SILTY SAND
REFERENCES
Bird, R.B., Stewart, W.E., Lightfoot, E.N.
(2002) Transport Phenomena, second ed.
Wiley, NewYork.
Dominijanni A., Jones C.J.F.P. and Manassero
M., (2007). Sistemi di lavaggio/estrazione, in
Manassero M. e Dominijanni A. (Eds.), Atti
delle Conferenze di Geotecnica di Torino,
60 80 100 Editore, Bologna (CD-ROM).
Manassero M. (2003). Integrated systems for
the remediation of polluted soils. Progetto di
rilevante interesse nazionale (Project of
rilevant national interest). Ministero
dell’Istruzione, dell’Università e della
Ricerca.
 Remediation Options for Contaminated Canal Sediments

Talib Mahdi
Geoenvironmental Research Centre, Cardiff University, UK (MahdiTA@cardiff.ac.uk)
Hywel Thomas
Geoenvironmental Research Centre, Cardiff University, UK (ThomasHR@ cardiff.ac.uk)
Robert Francis
Geoenvironmental Research Centre, Cardiff University, UK (FrancisRW1@ cardiff.ac.uk)
Guodong Zheng
Institute of Geology and Geophysics, Chinese Academy of Sciences , China
(gdzhuk@hotmail.com)

ABSTRACT A 7 km stretch of a canal in South Wales was affected by a pollution incident due
to a major minewater discharge. The incident resulted in the blanketing of the canal bed with an
orange-yellow (rust coloured) layer that adversely affected the aquatic life of the canal and the
regeneration of the area. This paper presents the research work undertaken to assess the effect of
the incident on the sediments of the canal and to identify the potential factors that would affect the
remediation/final disposal options available for the sediments.

INTRODUCTION and removing the contaminated dredged

A 7 km stretch of a canal in South Wales was
affected by a pollution incident due to a major
minewater discharge. Peak contaminant loading
of 400 mg.Fe/L at flow of 36 L/sec has been
recorded during that incident (Ranson, 1999). The
incident resulted in the blanketing of the canal
bed with an orange-yellow (rust coloured) layer
that adversely affected the aquatic life of the
canal and the regeneration of the area. A mine
water treatment project, which opened in 2002,
has resolved the problem of the contamination of
the mine water discharged to the canal.
The canal company had proposed to fully restore
the polluted sections of the canal to overcome the
environmental and regeneration consequences of
the pollution incident. They propose to achieve
the restoration by completely dredging the canal
sediments to landfill (Dig and Dump strategy).
The Canal Company had based their strategy on
very limited characterisation data. Cardiff
University’s Geoenvironmental Research Centre
(GRC) was approached to undertake a research
programme to further characterise the sediments
including its leachability, assess the viability of
alternative remediation strategies including soil
washing/separation process, examine the
dewatering potential of the sediments and
comments of the potential beneficial use of
sediments. This paper presents the research work
undertaken to assess the effect of the incident on
the sediments of the canal and to identify the
potential factors that would affect the
remediation/final disposal options available for
the sediments

SITE AND SAMPLING
The affected 7 km section of the Canal runs
north-east to south-west and starts at about 400m
upstream the constructed minewater treatment
plant. This Canal section is confined within a
narrow valley corridor that is also occupied by a
river and three roads (Figure 1).
Five locations were selected along the canal for
sampling as shown in Figure 1. For each location
a bulk sediment sample of about 20kg was
obtained from the top 400mm layer of the canal
bed. In location (S5) an 80cm core of undisturbed
sample was obtained for vertical characterisation
of the sediments. Canal water samples and river
water sample at the overflow discharge point to
the river at (S1) were also collected.
S5
S4
S3
S2
S1
Affected
Section
Fig. 1 Site and sampling locations
(For colour figure, refer to CD)
RESULTS AND DISCUSSION
1. Geotechnical Characterisation
Moisture content vary horizontally between 56%
(S4) and 245% (S1) and vertically between 27%
(bottom layers) to 683% (top 5 cm) indicating
difficulties in transporting the sediments if dig
and dump is to be used without dewatering.
Particles size analysis results presented in Table 1
indicates that the sediments are predominantly
well-graded sandy silt or clay (S1, S4 and S5) to
silty sand (S2 and S3). Results show that
Mahdi et al. 829
sediments consist of an average of about 34%
fines 40% sand and 26% fine-medium gravel.
Results also show a great variability in the
fractions content between the different locations
along the canal. This suggests that if soil
washing/separation is to be considered 66% of the
sediment can be recovered as sand and gravel.
TABLE 1 Summary of PSD results
Particle S1 S2 S3 S4 S5 Mean
% % % % % %
Gravel 14 53 36 18 8 26
Sand 49 32 45 37 36 40
Fines 37 15 19 45 56 34
2 Chemical Characterisation
Results for bulk and that of its fraction are
presented in Table 2 together with CLEA soil
guidance values (SGV) (Defra/Environment
Agency, 2002) and ICRCL values (ICRCL,
1987). Results show that concentration of heavy
metals such as As, Ni, Cu and Zn are exceeding
the threshold vales of guidelines for certain
potential uses. Values of the sulphides in the bulk
are higher than the ICRCL threshold values. This
suggest that the sediment is slightly to moderately
contaminated with As, Cu, Ni, Zn and sulphide
and cannot be used as it is as a fill in residential
areas, allotments, parks, playing fields and open
spaces. However the sediment can be used as a
fill for industrial/commercial area.
It is worth mentioning that the sand portions used
in these tests are the by-products of the wet
sieving analysis. This will allow comment to be
made on the viability of the soil washing process
with water only in separating the contaminants
from the coarse sized particles. Results of fines
and sand fractions confirm that most of the
contaminants are associated with the fines
fraction of the sediments. However, some
contaminant concentrations for the washed sand
are still higher than the threshold values at some
locations. This is not completely unexpected
given the high concentration of iron and sulphides
830 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

in the sediments which might encapsulate the (kaolinite, illite, chlorite, montmorillonite)
heavy metals and make them less mobile when minerals. However, the relative contents of each
washed with water only (Zheng et al, 2001). On mineral vary along the canal and with depth. The
the basis of the above results, it can be concluded distribution of clay minerals in the sediment is
that soil washing with water only may not kaolinite>illite>chlorite montmorillonite. The
produce a clean coarse fraction and a results also indicate that sediment at S1 contains
contaminated fines fraction. the highest amount of quartz and pyrite, and the
lowest amount of montmorillonite whereas
TABLE 2 Chemical characterisation results sediments at S5 contain the lowest amount of
quartz and highest content of clay minerals. For
As Cu Ni Zn Sulphide
ppm ppm ppm ppm ppm
the vertical variation samples there is no clear
Bulk
trends for the quartz and pyrite content, but the
45 254 372 576
content of clay minerals increases with depth.
S1 Sand 27 95 149 159
Fines 43 320 322 437
4. Leachability Tests
Bulk 51 294 469 620
Leachability test are required to classify the
S3 Sand 25 171 189 212 material for disposal (Environment Agency,
Fines 54 296 340 812 2001). It is also required in assessing the
Bulk 52 181 367 464 6492 sediment for potential beneficial uses. To cover
S5 Sand 24 74 140 162 2.4 all the possible scenarios of disposing and reusing
Fines 59 190 384 596 28.1 the sediments the leachability was assessed for
CLEA Res 20 50 three cases; Sediment as it is (fresh), Dewatered
SGV Com 500 5000 sediment (centrifuged) and dried sediment.
ICRCL 10-40 130 70 300 1000 Leaching test was carried out according to
National Rivers Authority test method mentioned
Pore water chemical analysis is vital in deciding in R&D Note 301 (Lewin et al, 1994). Results
the type of treatment required, if any, for the are presented in Table 3 together with the EQS
dewatered water if solid-liquid separation values and the threshold values of Table 1 of the
(dewatering) process is to be employed. Results "guidance on the disposal of contaminated soil".
for surface water indicates that heavy metals
concentration of the river and canal surface water The results presented in Table 3 show that the
are below EQS value for List 2 dangerous concentrations of heavy metals in the leachate of
substances. Results for Pore water bulk and core the dewatered sample are slightly higher than
samples show similar trend as that of the surface those of the fresh sample (as it is). This can be
water except for the S5 sample which shows a explained by considering the higher amount of
slightly higher level of As than the EQS value. contaminated solids in the dewatered sample. The
The results suggest that if mechanical dewatering results also show a reduced concentrations of As
is to be used as a size reduction measure, most of and Zn and an increased concentrations of Ni and
the output water (pore water) may need no Cu in the leachate of dried sample compare to
treatment before discharge in the canal or the both the fresh and dewatered sample. Heavy
nearby river. metals concentrations in the leachate of all
samples are below the EQS values. This suggest
3. Mineralogical Characterisation that leachate from the sediments will pose no
Sediment’s mineralogical composition were threat to the controlled surface water if it is
assessed using X-Ray Diffraction (XRD) method. disposed in a landfill or used beneficially. Table 3
Results show that the Canal sediments main also shows that As level in the leachate of all the
constituents are quartz, pyrite, carbonate and clay samples and the Ni and Cu level in the leachate of
 Mahdi et al. 831

the dried samples are higher than the threshold effective in reducing the moisture content of the
values specified in the disposal guidance. This mixture than both the hydrated lime and the
may suggest that the sediment cannot be Portland cement by about 6% and 12% for the
considered as inert for disposal classification. addition of 4% and 8% respectively. The
differences increase with time for up-to 3 hours
TABLE 3 Concentration of HMs in the leachate
after mixing. Quick lime behaviour can be
As Cu Ni Zn Fe attributed to the reaction of the lime with water
Sample
ppb ppb ppb ppb ppb and the high heat produced during this reaction
S5 - As it is 38.1 13.4 39.6 16.4 <6.2 that may have helped further reducing the
S5 - Dewatered 44.6 18.6 46.2 34.9 <6.2 moisture content by evaporation. On the basis of
the above results it can be concluded that quick
S5 - Dried 11.5 65.8 103.9 19.8 <6.2
lime out-perform the cement and hydrated lime in
EQS Values 50 100 200 500 1000 dewatering the sediment and therefore it would be
Disposal Guidance 10 20 50 500 100 the recommended choice if chemical dewatering
is to be considered.
4. Dewaterability Tests
Dewatering is required to help in the handling of Blank 8% Q Lime
the sediment during transportation to landfill. 8% Cement 8% H Lime
Mechanical dewatering may reduce the amount of 150
Moisture Content - %

the material to be transported to landfill and

hence reduce its environmental impact. Chemical
dewatering may cause slight increase in the
volume and weight of the final product, but it has
the added advantage of stabilising, reducing
odour and for some additives pasturing the
material to be dewatered.
Chemical Dewaterability
Chemical dewatering or “conditioning” is the
process of adding certain materials to reduce the
water content of a slurry or sludge. Portland
cement, hydrated lime and quick lime are selected
as the additives in these tests due to its cost
effectiveness and local availability. For each
additive two mixes were tried; 4% and 8% by
weight of the wet sediment. The moisture content
was monitored with time as a measure of
effectiveness. Measured amount of additive was
mixed with a 1kg sample of the sediment in an
open top plastic container. Samples for moisture
content measurement were taken from the
container at pre-specified time intervals. Results,
which show the moisture content - time relation
for the 8% mixes are presented in Figure 2.
The results indicates that for the same mix
proportion, the addition of quick lime is more
130
110
90
70
0 1000 2000 3000
Time - m in
Fig. 2 Moisture content – time relation
(For colour figure, refer to CD)
Mechanical Dewaterability
Sediment Characteristics such as Specific
Resistance to Filtration (SRF), sediments solid
content and moisture quality and content, strongly
influence the sediments drainage rate. Its
determination is prerequisite for the rational
selection of the most cost-effective volume
reduction process (Cristensen, 1983). The SRF is
the main parameter to characterise the mechanical
dewaterability of any sludge or slurry. SRF
provides an empirical measure of the resistance
applied by the solid material to the release of
water (Karpuzcu et al 1996).
832 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

SRF was determined using Buchner Funnel- the soil washing process can be enhanced greatly
vacuum filtration unit similar to the one described by the addition of biosurfactant with NaOH. The
by (Besra et al., 2000). Results indicates that the results indicate that a solution of 1% NaOH and
SRF for the bulk sediment of S1 and S5 and that 0.5% biosurfactant has cleaned the sediments
of the fines of S5 are in the range of 6.64E+11 to from all the Arsenic, 59% of the Nickel and 9%
1.66 E+12. Canal sediment’s SRF values of the Zinc. This is in agreement with the finding
represent poor dewatering characteristics since of Mulligan et al (2001b). On the basis of the
good dewatering characteristics are associated above results it can be concluded that washing
with SRF value of 1E+10 and lower (Karpuzcu et water need to be enhanced with the addition of a
al. 1996). small percentage of biosurfactant with NaOH if
soil washing is to be used to clean the sediment.
5. Soil Washing
Soil washing with pH manipulation and/or the TABLE 4 Heavy metals removal of the
addition of biosurfactant was carried out to assess different solutions
the viability of this technique for the
decontamination of the sediments. Three types of % Removal
Solution pH
additives were investigated in different As Cu Ni Zn
concentration. These additives are: Hydrochloric 1.5% NaOH 12.7 30 0 1 0
Acid (HCl), Sodium Hydroxide (NaOH) and
1% NaOH 12.6 30 0 1 0
Biosurfactant with and without NaOH. Soil
washing tests were carried out on S5 bulk 0.5% NaOH 12.4 29 0 1 0
samples using similar procedure as that used by 0.25% NaOH 12.1 25 2 2 0
Mulligan et al (2001). The percentage metal 1.5% HCl 0.8 26 16 43 12
removal was determined based on the initial 1% HCl 1.2 23 9 41 8
metal content in the sediment and all results are
presented as percent metal removal. Distilled 0.5% HCl 2.0 19 0 35 3
water alone was used to account for removal of 0.25% HCl 5.2 0 0 7 0
contaminants by physical mixing. 0.5% B* 5 0 1 0
1.0% B 6 0 1 0
The results of the soil washing tests are presented
2.0% B 6 0 1 0
in Table 4. The results show that increasing the
alkalinity of the solution by adding NaOH of 1% NaOH + 0.5%B 12.7 100 0 59 9
different concentration helped in removing 24- 1%NaOH + 1.0%B 12.6 93 0 54 8
30% of the Arsenic content of the sediments. 1% NaOH + 2.0%B 12.4 91 0 46 7
However, increasing the alkalinity had no effect
on removing Copper, Nickel and Zinc. On the * B = Biosurfactant
other hand, the results suggest that acidic
solutions with concentration of more than 1% CONCLUSIONS
HCl are effective in removing Nickel (>41%) and Canal sediments are predominantly well-graded
Arsenic (>23%) and less effective in removing sandy silt or clay to silty sand can be classified as
Copper (<16%) and Zinc (<13%). slightly to moderately contaminated with As, Ni,
Cu, Zn and sulphide. Most contaminants are
Results also show that the addition of associated with the fines fraction of the sediment
biosurfactant only to the washing water, at although at some locations the sand fraction of
percentages of up to 2%, has a very little to no the sediments may contain contaminants loading
effect on the heavy metals removal capability of higher than some threshold vales. High moisture
the solution. Contaminants removal capability of content of the top layer of the sediments may

cause some difficulties in handling the sediment
during transportation if dig and dump is to be
used without dewatering. The sediment has poor
mechanical dewatering characteristics, as for its
chemical dewaterability, Quick lime is more
effective in dewatering/conditioning the sediment
than Portland cement or Hydrated lime. If soil
washing/separation is to be used effectively more
than 65% of the sediment can be recovered as an
aggregate, however, the only effective additive to
the washing water that can improve its ability to
remove the different contaminants is a mixture of
(0.5%Biosurfactant+1%NaOH).
REFRENCES
Besra, L, Sengupta D. K. and Roy S. K. (2000).
Particle characteristics and their influence on
dewatering of kaolin, calcite and quartz
suspensions, International Journal of Mineral
Processing, 59 (2), pp. 89-112
Cristensen, G.L. (1983). Units for Specific
Resistance, Journal WPCF, 55, pp. 416-419.
Defra/Environment Agency (2002). Full Set of
CLR, TOX and SGV reports.
Environment Agency (2001). Guidance on the
Disposal of "Contaminated Soil", Version 3
April 2001.
Interdepartmental Committee for the
Redevelopment of Contaminated Land
(1987). Guidance on the assessment and
redevelopment of contaminated land, ICRCL
Mahdi et al. 833
Paper 59/83, 2nd Ed., Dept. of Environment,
London.
Karpuzcu, M., Buktel D. and. Aydin, Z.S. (1996).
The dewaterability, heavy metal release and
reuse characteristics of golden horn surface
sediment, Water Science and Technology, 34
(7-8), pp. 365-374
Lewin, K., Bradshaw, K., Blackley, N.C., Turrell,
J., Hennings S.M. and Flaving, R.J. (1994).
Leaching tests for assessment of
contaminated land: Interim NRA guidance.
National Rivers Authority R&D Note 301.
Mulligan, Catherine N., Yong R.N. and Gibbs,
B.F. (2001). Heavy metal removal from
sediments by biosurfactants, Journal of
Hazardous Materials, 85 (1-2), pp. 111-125
Mulligan, Catherine N., Yong R.N. and Gibbs,
B.F. (2001b). An evaluation of technologies
for the heavy metal remediation of dredged
sediments, Journal of Hazardous Materials,
85 (1-2), pp. 145-163
Ranson, C.M. (1999). Minewater treatment in
Neath Port Talbot. Proceedings of the
Institution of Civil Engineers. 133, pp. 183-
193
Zheng, G., Takano, B., Kuno, A., Matsuo, M.
(2001). Iron speciation in modern sediment
from Erhai Lake, Southwestern China––
redox conditions in an ancient environment,
Applied Geochemistry, 16 (9-10), pp. 1201–
1213.
 Electrokinetic Remediation of Soil Contaminated With Copper

S. Sahoo
Indian Institute of Technology, Roorkee (rubygt01@yahoo.com)
P. Bala Ramudu
Institute of Technology, Banaras Hindu University, Varanasi (balaramudu_p@yahoo.com)
R. K. Srivastava
Motilal Nehru National Institute of Technology, Allahabad (rksciv@yahoo.com)

ABSTRACT Electrokinetic soil remediation is most promising soil decontamination processes. In the
present study an experimental work has been carried out for electrokinetic remediation process in saturated
copper contaminated soil. This study presents the removal efficiencies for copper which is artificially
contaminated in clay and silt soils by electrokinetic remediation with constant voltage of 36 V dc was
applied to induce the movement of electrolyte solution in the soil system for 5 days using deionized water as
anode flushing solution. The results obtained form study indicated that removal rates are significantly
influenced by applied voltage, current, pH value of soil.

INTRODUCTION

Soils can be contaminated with heavy metals
derived from various sources including abandoned
mining wastes, improper treatment of industrial
wastes, incomplete collection of used batteries,
leakage of landfill leachate, accidental spills and
military activities (Adriano, 1986). The
contamination often affects a large volume of soil
underlying several acres of the surface area. There
are also various types of contaminated lands such
as paddy fields, farms, factory sites, mine fields
and residential districts. Contaminants migrating
from these sources threaten the human health in
the local area and the ground-water supply.
However, technologies for decontaminating these
sites have not been well developed. In addition, it
has been recently reported that soil contamination
is increasing in various sites such as residential
areas near industrial complexes and reservoirs of
drinking water.
Electrokinetic soil processing is also called
electrokinetic remediation, electroreclamation,
electrochemical decontamination. It needs low-
-2
level direct current in the order of mA cm of
cross-sectional area between the electrodes to
remove contaminants from soils (Acar and
Alshawabkeh, 1993). The low-level direct current
results in physico-chemical changes in the soil
mass, leading to species transport by coupled
mechanisms. Electrolysis of water produces
hydrogen ions in the anode compartment and the
acid front migrates across the soil cell and makes
contaminants desorbed from soil surfaces,
resulting in an initiation of electro-migration, i.e.
the transport of ions and polar molecules in
electric field. Electric potential also leads to
electro-osmosis, i.e. the flow of an ionic liquid
under the action of an applied electric field
relative to a charged surface (Reuss, 1809; Acar et
al., 1995). Electro-migration and electro-osmosis
are the important mechanisms in the electro-
kinetic soil processing which remove
contaminants from soil. Electrokinetic soil
processing is an effective technology for removal
of contaminants in low-permeability soils ranging
from clay to clayey sand. The advantages of this
technology are its low cost of operation and its

potential applicability to a wide range of
contaminant types (Pamukcu and Wittle, 1994).
Electrokinetic soil processing is envisioned for the
removal/separation of organic and inorganic
contaminants and radionuclides. The potential of
the technique for waste remediation has resulted in
several studies (Putnam, 1988; Khan et al., 1989;
Thompson, 1989; Mitchell and Yeung, 1991).
The objectives of this study are to experimentally
investigate the feasibility of electrokinetic soil
processing to examine variation of current and
electroosmotic flow with time, variation of
moisture content, soil pH, and contaminant
concentrations with distance, pH variation of
effluent with time at anode and cathode ends.
MATERIALS AND METHODS
Experimental Apparatus
Fig. 1 shows schematic sketch of the electro-
kinetic test set up used for this study. The set up
consists of an electro-kinetic cell, two electrode
compartments, two electrode reservoirs, a power
source (Battery) and digital multimeter. The
Perspex tube has an inside diameter of 69 mm and
3 mm thickness of a total length is 35 cm used for
electro-kinetic cell. Each electrode compartment
consisted of a slotted graphite electrode, filter
paper (Whatman No. 42), filter net and cork
washer. Two Perspex caps containing the graphite
electrodes 66 mm in diameter and 10 mm in
thickness. A Hole is drilled in the centre of
graphite electrodes for the passage of ions from
the soil to electrode compartments and to
accommodate the bulk electrolyte flow by electro-
osmosis. The electrode compartments were
connected to both ends of the cell using screws.
Electrode reservoirs were made of 38 mm inner
diameter, 3 mm thickness and of capacity 150 ml
each.
Sahoo, Ramudu and Srivastava 835
Fig. 1 Schematic diagram of overall Electro-kinetic test
set-up
Exit ports were created in the electrode
compartments, and the tube was attached to these
ports to allow the gases generated due to the
electrolysis of water to escape. The other end of
these gas tubes was connected to the reservoirs to
collect any liquid that was removed along with the
gases. A power source was used to apply a
constant voltage to the electrodes, and a
multimeter was used to monitor the voltage and
current flow through the soil sample during the
test. Five dummy electrodes were inserted in the
cell to determine the voltage distribution along the
length of the reactor or cell.
Physical Properties of the Two Different Soils
and Preparation of Soil Samples
Soils used in this experiment were clay and silt
soil contaminated with copper (Cu) artificially.
Clay soil has been taken from outside of institute
campus, at Meza road, Allahabad district. Based
on geotechnical properties, according to Indian
Standard Soil Classification System this soil is
classified as inorganic clay of medium plasticity
(CI). The silt soil has been taken from institute
campus near Geotechnical Engineering Laboratory
of MNNIT, Allahabad. Based on geotechnical
properties, according to Indian Standard Soil
Classification System this soil is classified as
inorganic silt of low plasticity (ML).
836 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

The physical properties of clay and silt soil used in TABLE 2 : Electro-kinetic testing program
this study are summarized in Table 1.
Test Designation EKT–1 EKT–2
TABLE 1 Geotechnical Properties of Clay Soil and Silt (Clay) (Silt)
soil Contaminant Cu++ Cu++
Concentration(mg/kg) 2000 2000
Properties Clay soil Silt soil Initial Moisture Content (%) 40 25
Grain size analysis Initial Soil pH 8.126 7.5
Clay (%) 68.2 14.9 Voltage (V) 36 36
Silt (%) 18.4 66.8 Purging solution DI DI
Sand (%) 13.4 18.3 water water
Consistency limits Time in(hr) 120 120
Liquid limit (%) 48 31.5
Plastic limit (%) 17 22.5 Measurement, Sampling and Analysis
Shrinkage limit (%) 13 20
Specific Gravity 2.65 2.67 Both the clay and silt soils used in this experiment
compaction were passed through 425 micron IS sieve and their
characteristics physical properties are summarized in Table 1.
Maximum dry unit 18.7 16.7 Two soil samples were prepared for the EK
weight (kN/m3) removal experiments of heavy metals in soils.
Optimum moisture 14.4 14.6 Prepared influent is added to the soil and mixed
content ( % )
thoroughly with a spatula in a proper container.
Coefficient of 4x10-9 1.085x10-7
permeability (k) in
The contaminated soil slurry is then placed in the
m/sec electro-kinetic cell in layers. The weight of soil
required in the reactor is determined such as 1.7
Two soil samples were prepared for the EK kg for clay soil and 2.0 kg for silt soil placed in
experiments of heavy metals in soils. Here for the the cell and approximately 30 g soil is kept for
source of Cu++ the solution of CuSO4.5H2O has checking initial pH of soil. The electrode
been taken as a source of Cu. The solution for the compartments are then connected to the cell and
soil contamination should be prepared as such as connected to the anode and cathode reservoirs
to get the required concentration of Cu++ in the using tubing. The reservoirs are then filled with
soil. For concentration of Cu++, 2000 mg/kg of dry de-ionized water and the soil-water-contaminant
soil the calculated amount of CuSO4.5H2O 15.75 g mixture is equilibrated for 24 hrs. Throughout the
was taken in 2.0 kg of dry soil. Then to get the test duration, the elevation of water in both
required moisture content to attain the required reservoirs is monitored and adjusted to prevent a
density of the soil, deionised water (DI) added hydraulic gradient forming across the specimen.
such that the moisture in soil should be 40% (800
ml) & 25% (525 ml) for 2.0 kg of clay soil and silt The electro-kinetic electrode is then connected
soil respectively. The solution is prepared by to the required power supply. Each test is
dissolving CuSO4.5H2O with the DI water (Table terminated after operating for 120 hrs or as
2). required. Voltage drop across soil is observed
from section to section. There are dummy
electrodes placed across the soil to observe voltage
drop in various section. The voltage & current are
measured with digital multimeters. The electro-
osmotic water flows are collected in the measuring
jar and amount of that are noted at specific time
interval as required. Due to electro-osmosis the
water level in both the reservoirs changes with

time. 5 ml samples are collected from time to time
to analyze kinetics of the process from the
compartments. Samples are taken 5-6 times during
the 120 hrs experiment, generally at least once 24
a day. The analysis gives the idea at which the Cu
is removed.
At the completion of each test, aqueous solutions
from the anode and cathode reservoirs and the
electrode assemblies are collected and the volumes
measured (subsequence the concentration & pH
are measured from respective reservoirs). Then,
the reservoirs and electrode assemblies are
disconnected, and the soil specimen is extruded
from the cell using a mechanical extruder. The soil
specimen is sectioned into six parts at distance of
0 to 20mm (section 1), 20 to 97.5mm (section 2),
97.5 to 175mm (section 3), 175 to 252.5 mm
(section 4), 252.5 to 330mm (section 5) and 330 to
350mm (section 6) from anode respectively. Each
part is weighed and preserved in a container.
To determine total amount of copper removed
from soil and the amount of copper remaining in
soil specimen along the length after electro-kinetic
process. This has been done by two steps- (a) soil
has been digested (b) digested soil solution is
filtered and that filtered solution is tested for
copper concentration by Orion meter.
RESULTS AND DISCUSSION
The results of the electro-kinetic experiments were
analyzed to assess the effect of variation in
moisture content and contaminant migration and
removal. As given in table 2, EK Test 1(EKT-1)
and EK Test 2 (EKT-2) possess different initial
moisture contents, same contaminant
concentrations. The parameters analyzed included
the current and cumulative electro-osmotic flow,
pH, moisture content and contaminant
concentration in the soil after the experiments
were terminated. The results of Tests 1 and 2 were
plotted together to analyze the effect of the
different initial moisture contents.
Sahoo, Ramudu and Srivastava 837
Electrical Current
Fig.2 shows that the initial moisture content
significantly affected the current; as the initial
moisture content increased, the current increased.
In clay soil, the current decreased at a slightly
more gradual rate up to 48 hours then constant up
to 72 hour. Finally current decreased till end of the
test. In silt soil the current increased up to 24 hour
after that decreased up to 48 hr. After 48 hour the
current is constant up to 56 hour and finally
decreased till the end of the test.
160
Clay soil
140
120 Silt soil
100
Current (mA) 80
60
40
20
0
0 50 100 150
Time(hr)
Fig. 2 Variation of current with time
Moisture Content
Fig.3 illustrates that the moisture content of the
soil at the conclusion of testing was compared to
the initial moisture contents of 40% and 25% for
clay and silt soils respectively. Slight changes in
moisture contents are evident which can be
attributed to the variations in the electro-osmotic
flow that occurred as a result of the changes in
parameters such as the ionic strength,
conductivity, and/or electrical gradient. These
results suggest that in clay soil near anode and
cathode section moisture content is higher but in
all other sections the moisture content is slightly
lower than the initial moisture content. In silt soil
moisture content is higher near cathode region but
in rest of section more or less equal to initial
moisture content.
838 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
Moisture Content(%)t (%)
50 Electroosmotic Flow
45 Clay Soil
Silt Soil Fig.5 shows that the flow rate in clay soil is
40 gradually increases up to 48 hour. After 48 hours
35 flow rate become constant up to 72 hour, there
30 after trend showed increasing trend. The flow rate
was relatively low in silt soil as compared to clay
25
soil. In silt soil flow rate increases steep up to 24
20 hour after that as time passes flow rate gently
15 increased up to end of the test.

0 02 04 06 08 300

Clay Soil

Cumulative Volume (mL)

Fig.3 Variation of moisture content with distance 250
Silt Soil
200
Soil pH
150

Fig. 4 illustrates that an acidic front of solution, 100

which was generated by the electrolysis reaction at 50

the anode, migrated from the anode towards the
0
cathode, and this solution significantly lowered 0 20 40 60 80 100 120 140
the pH through the first three soil sections from Elapsed Time (hrs)
the anode. Clay soil produces the highest
cumulative flow, and also noted highest pH in the Fig. 5 Variation of electro-osmotic flow with time
soil adjacent to the cathode. Silt soil produce low
cumulative flow and similar pH profiles like clay Contaminant migration
soil but overall pH values of treated silt soil is
below the pH values of clay soil. Total copper concentrations in each section of the
treated soil were determined and are showed in
12 Fig.6. The concentration profile after EK tests, the
10 silt soil at different sections is observed to be
consisting of high concentration of copper as
8
compared to concentration profiles of clayey soil.
pH

6 Final concentration of copper was lower near the

4 cathode and higher at the anode in silt soil.
Clay Soil
2 Silt Soil Increased copper concentrations were observed
0 near the anode. This raised concentration also
justifies high electric conductivity near to anode
0 0.2 0.4 0.6 0.8 1
both in soil and anolyte.
Normalized Distance from Anode
Fig. 4 Variation of Moisture Content with Distance

250
Residual copper in mg/L
200
150
100
50
0 Air and Soil Pollution, 125,259-272.
0 0.5
Normalised distance from anode
Fig. 6. Variation of contaminant Concentrations with
distance
CONCLUDING REMARKS
The results of studies on copper extraction from
artificially contaminated clay and silt soils by
electro-kinetic remediation process in which
deionised water was used as purging agent were
presented.
The residual copper profiles in EK treated silt soil
were higher than in residual copper profiles of EK
treated clayey soil. It may be explained by the
facts that the dissolution and desorption of heavy
metals on the soil surface. Higher accumulation of
copper in the anolyte and at anode region was
observed in experiments done on treated silt soil.
This indicates that major copper species were
anions (at the beginning of experiment), which
migrated toward the anode when the electric field
was induced. Among the major transport
mechanisms for copper remediation, the electro-
osmosis seems to be dominant in the EK process.
References
Acar, Y. B. and Alshawabkeh, A. N. (1993)
Principles of Electrokinetic Remediation. Environ.
Sci. & Tech, 27(13), 2638-2647.
Acar, Y.B., Gale. R.J., Alshawabkeh, A.N.,
Marks, R.E., Puppala, S., Bricka, M., and Parker,
R. (1995) Electrokinetic Remediation: Basics and
Sahoo, Ramudu and Srivastava 839
Technology Status, Jl. Hazardous Mater., 40, 117-
137.
Clayey soil
Silty soil
Adriano, D.C. (1986) Trace Elements in the
Terreestrial Environment, Springer Verlag, 7-14.
Kim, S., Moon, S.H. and Kim, K.W. (2001)
Removal of Heavy metals from Soils Using
Enhanced Electrokinetic Soil Processing, Water,
1
Khan, L. I., Pamukcu, S. and Kugelman, I. (1989),
Electro-Osmosis in Fine-Grained Soil. 2nd
International Symposium on Environmental
Geotechnology, Shanghi, China, Envo Publishing,
Bethlehem, PA, 1, 39–47.
Lageman, R. (1993) Electro-Reclamation, Jl. Env.
Sci. & Tech. 27(13), 2648-2650.
Mitchell, J. K. and Yeung, T. C. (1991)
Electrokinetic Flow Barriers in Compacted Clay,
Transportation Research Records, No. 1288, Soils
Geology and Foundations, Geotechnical
Engineering 1990, pp. 1--10
Pamukcu, S. and Wittle, J. K.. (1994)
Electrokinetically Enhanced in situ Soil
Decontamination, in Remediation of Hazardous
Waste Contaminated Soils, Marcel Dekker, Inc.,
New York, U.S.A., pp. 245–298.
Probstein, R.F. and Hicks, R.E. (1993) Removal
of Contaminants from Soils by Electric Fields,
Science, 260, 498-504.
Putnam, G. (1988) Development of pH Gradients
in Electrochemical Processing of Kaolinite, M.S.
thesis, Louisiana State University, Baton Rouge,
U.S.A.
Reuss, F. F. (1809) Memoires de la Societe
Imperiale des Naturalistes de Moscou 2, 327.
Shapiro, A. P. (1990) Electroosmotic Purging of
Contaminants from Saturated Soils, Ph.D. thesis,
Massachusetts Institute of Technology,
Cambridge, U.S.A.
 Removal of Arsenic from Groundwater by Sand and Zero Valent Iron
Mixture: Batch and Column Study
M. A. Abedin
Graduate Student, Kyoto University, Japan (masumagriculture@yahoo.com)
T. Inui
Associate Professor, Kyoto University, Japan (inui.toru.3v@kyoto-u.ac.jp)
T. Katsumi
Professor, Kyoto University, Japan (katsumi.takeshi.6v@kyoto-u.ac.jp)
M. Kamon
President, Kagawa National College of Technology, Kagawa, Japan (koucho@t.kagawa-nct.ac.jp)

ABSTRACT This paper focuses a series of batch and column studies to assess the performance and
effectiveness of zero valent iron for arsenic (As) remediation in groundwater and evaluate future implication
in the filtration system. Neutral pH (pH 7.0) showed the best performance over control (without adjusting
pH) and pH 5.0. Arsenic removal capacity of ZVI was determined to be approximately 4.3 mg As/g ZVI
through adsorption and coprecipitation. Both batch and column tests results showed that 99% of arsenic was
removed successfully. HCl-extractable arsenic from the iron samples was 31% indicated that a greater
portion of arsenic is present in a form that is not extractable with 1 N HCl.

INTRODUCTION As (III) to As (V) including high cost and

Water quality will become the principal limiting
factor for sustainable development in many
countries early in this century (Ongley 1999).
“Everything living is created from water” is an
ancient quotation, which closely describes the
importance of water (Anon. 1977). Alarming
information has emerged in recent decades about
the widespread presence of arsenic (As) in
groundwater used to supply drinking water in
many countries on all continents. Of these regions,
Bangladesh and West Bengal are most seriously
affected in terms of the size of the population at
risk and the magnitude of the health problems
(Khan et al. 2000; ANL 2003).
Different techniques have been developed and
tried out for the removal of As from groundwater.
They are the coprecipitation (eg. With iron and
aluminum salts), ion exchange, adsorption by
activated carbon and membrane processes
(Kartinen & Martin 1995). Most of these tested
methods required pre-oxidation step to transform
difficulty of maintenance.
The coprecipitation of arsenic with naturally
present Fe (II) (passive removal) is the simplest
conceivable arsenic removal technique. Passive
removal is applied by allowing pimped
groundwater to stand for periods ranging from few
hours to several days after pumping, allowing the
Fe (II) to oxidize and forming hydrous ferric
oxides (HFO) with adsorbed arsenic to precipitate.
Because As (V) has a higher affinity to HFO than
As (III) (Meng et al. 2000), passive removal works
best if the arsenic is present as As (V). However,
groundwater in Bangladesh contains
predominantly As (III), and the concentrations of
Fe (II) are variable and often too low for sufficient
arsenic removal. To improve removal efficiencies,
oxidants such as NaOCl (Meng et al. 2001) or
KMnO4 (Borho 1996) and iron or aluminum salts
are typically added to oxidize As (III) and
subsequently adsorb As (V) on iron (III) or
aluminum (hydr)oxides (Zouboulis & Katsoyiannis
2002). In this connection, arsenic removal by ZVI

has received special attention due to its higher
removal efficiency. Recent studies have reported
that, As (III) is partially oxidized to As (V) during
iron oxidation process, which results higher As
removal efficiency (Berg et al. 2006). However,
fate of long term application of ZVI, arsenic
removal capacity and pH range has not been well
documented in the developing countries. Therefore
a series of bath and column studies were conducted
to monitor these effects for future application in
the field.
The objectives of this study are to i) assess the
performance and effectiveness of zero valent iron
for arsenic (As) remediation in groundwater, (ii)
determine arsenic removal capacity of zero valent
iron (ZVI), and (iii) evaluate future implication in
the filtration system in the developing countries
like Bangladesh, Nepal and Vietnam etc.
EXPERIMENTAL SECTION
The chemicals used in this study were analytical
grade and all the stock solutions were prepared
with deionized water (DI). The arsenic stock
solution was prepared by Na2HAs4.7H2O in DI
water. The ZVI-A material was obtained from JFE
Steel Corporation, Japan. The properties of the
zero valent iron are shown in Table 1. Toyoura
sand, with particle density of 2.60 g/cm3 was used
in column experiment.
Batch Experiment
The batch experiments utilized ZVI and were
conducted on 500 mL conical flasks. In all cases,
500 mL of arsenic contaminated water was loaded
with different levels of iron. Then the conical flask
containing different levels of arsenic contaminated
TABLE 1 Properties of Zero Valent Iron
Type
Physical Chemical
of
property composition %
iron
Particle
Total > 250
density Metal Fe
Carbon µm
g/cm3
KB-
7.2 93.5 0.482 0 0.2
90
Abedin et al. 841
water and different levels of iron were placed on
an orbital shaker at room temperature (24 ºC) for
different time periods. In case-I, 1 mg/L arsenic
contaminated water was mixed with 1.0 g/L iron
(ZVI-A). In this experiment, the control pH was
adjusted to 5.0 or 7.0 using 0.1 M NaOH and HCl.
The sampling periods were 1, 2, 3, 4, 5, 10, 15, and
24 hours later. In cases-II, 0.2 mg/L arsenic
contaminated water was loaded with 2.0 g/L iron
(ZVI-A) with three levels of pH: viz. control, 5.0,
and 7.0. The sampling periods were similar to
case-I.
A separate batch adsorption study was also
conduced to evaluate arsenic adsorption capacity
for ZVI-A. Duplicate samples of 1.0 g ZVI were
mixed in 200 mL conical flask for 1, 5 and 12
hours at initial arsenic concentration 1, 5, 10, 15,
20, 25, 30 and 40 mg/L in a shaker at 100 rpm and
allowed to settle except 1 and 5 hours. After 1 and
5 five hours and 1, 3, 5 and 7 days settlement, 10
mL of the suspension was collected with the help
of syringe sampler and then filtered through
syringe filter with 0.20 ȝm pore size before As
determination.
Column Experiment
Column experiments were conducted using acrylic
columns packed with ZVI and sand mixture. The
test columns were 5 cm in inner diameter by 30 cm
in height. The column contained inlet, outlet and
three side sampling ports at 8.0, 16.0 and 24 cm
from the bottom end (Fig.1). A well-homogenized
ZVI and sand mixture was funneled into the
column in five increments to ensure that the
columns were packed homogeneously. After
funneling each increment, 10 mild strokes were
Particle size distribution %
>180 >150 >106 >75 >63 >45 <45
µm µm µm µm µm µm µm
0.9 25.5 32.9 8.6 13.9 18.0
842 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

TABLE 2 Characteristics of Each Column 4.0, 7.0 and 10) before each sampling time. Then
suspension was filtered through syringe filter with
ZVI Initial As 0.20 ȝm pore size and were kept in the refrigerator
Flow
content in Porosit conc. at 4ºC until analysis of the total As. The Arsenic
Test case rate
dry wt. y (µg/L) concentrations were measured by Graphite furnace
(ml/hr)
basis (%)
atomic absorption spectrophotometer.
Column 1 62.2 0.49 200 30
Column 2 35.3 0.43 200 30 Solid Phase Characterization
Column 3 0.0 0.41 200 30 The sample was morphologically analyzed before
and after reaction by SEM (Scanning Electron
Column 4 62.5 0.49 1000 30
Microscopy). Iron samples from the column were
Column 5 62.5 0.49 200 100 extracted with 1 N HCl to dissolve hydroxide
precipitates. Total arsenic in the extractant
solutions were determined using the methods
described above.

RESULTS AND DISCUSSION

Fig.1 Setup for the column experiment
applied with a rubber hammer to ensure
homogeneous packing of materials in the column.
The column containing sand was filled with sand
only. The artificial arsenic spiked water was
passed through the columns upward direction at
different flow rate using peristaltic pump. Flow
rates were measured periodically throughout the
experiments to ensure consistency. The testing
conditions for each column are summarized in
Table 2.
Measurements and Chemical Analysis
Immediately after collection of samples in each
time, the redox potential (Eh), Dissolved Oxygen
(DO) and pH were measured by combined pH and
Eh meter and DO meter, respectively. The pH
meter was calibrated with three buffer solution (pH
Batch Experiment
In all the ZVI system, arsenic concentration
decreased exponentially with time. From Fig.2, it
is revealed that after 15 hrs elapsed time 99% of
arsenic is removed, where the initial arsenic
concentration was 0.2 mg/L and ZVI 2 g/L,
respectively and in the same time 97.5% of arsenic
is removed using 1.0 g/L zero valent iron. After
10-hourr elapsed time the result meets the
Bangladesh drinking water standards (0.05 mg/L)
using 1.0 g/L ZVI where WHO (0.01 mg/L) as
well as Bangladesh Standards meets using 2.0 g/L
ZVI. These results indicated that with the increase
in iron doses the arsenic removal efficiency also
increases.
The effect of pH on arsenic removal from arsenic
contaminated water using zero valent iron is
illustrated in Fig.3. The removal of arsenic
occurred prominently at pH 7 than other two cases.
More than 99% of the arsenic was removed after
15 hours of reaction at pH 7. Similar, results have
been reported from other investigators; Biterna et
al. (2007) reported over 99% of As(V) was
removed in 6 hrs when initial pH was controlled at
7..2 and nearly 90% at pH 4..2 and Bang et al.
(2005) found 99.8% removal of arsenate under
oxic conditions at pH 6.0 after 9 hrs of reactions.

Fig.2 Arsenic removal by ZVI. As conc. 0.2 mg/L
without adjusting pH and ZVI-A.
Fig.3 Effect of pH on arsenic removal by ZVI. As 0.2
mg/L & ZVI-A 2.0 g/L
An attempt was also made to determine the highest
capacity of ZVI to adsorb arsenic from
contaminated water through batch experiment.
This experiment helped to estimate each gram of
ZVI can adsorb how much amount of arsenic. This
implies that the arsenic adsorption capacity should
be a better design parameter for the estimation of
the amount of ZVI required for field application of
filtration system.
It is clear from Fig.4, the arsenic adsorption by
zero valent iron increases with arsenic
concentration in solution up to certain limit and
Abedin et al. 843
Fig. 4 Arsenic adsorption capacity (ZVI (KB-90) 1.0 g/
200 mL)
then showed steady behavior. The total arsenic
adsorption by 1 gram ZVI reaches peak with 1.66
mg arsenic when shaking for 5 hours. Moreover, 1,
3, 5 and 7 day’s settlement shows an excellent
variation with without settlement which supports
the coprecipitation of arsenic with iron. Therefore,
the total arsenic removed from the solution
through adsorption and coprecipitation was
increased to 4.33 mg/g (Fig.4).
Column Experiment
The column studies were designed to evaluate
long-term arsenic removal performance (180 days)
from simulated groundwater running at different
flow rates using ZVI-A and sand mixture. In all
cases, the total arsenic concentration at different
positions of the up-flow mode column were in the
following order: column influent > bottom port
effluent > middle port effluent > top port effluent >
column effluent (Figs. 5-6). The effluent arsenic
concentration remained below 6 µg/L in column 5
up to approximately 1475 pore volumes (432 L)
when the initial arsenic concentration was 200
µg/L. Similarly, the arsenic concentration in the
treated water of Columns 1 and 2 (flow rate 720
mL/day) remained below 2 and 5 µg/L up to
approximately 442 and 508 pore volumes,
844 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
Fig.5 Arsenic conc. as a function of pore volume for the
whole column. Column 4: Initial As conc. 1000 µg/L.
Fig.6 Arsenic conc. as a function of pore volume for the
whole column. Column 5: Initial As conc. 200 µg/L and
flow rate 2.4 L/day.
Fig.7 Arsenic removal by ZVI
respectively (Fig.7). The arsenic concentration of
Column 3, which consisted of sand only, increased
with time (Fig.7). After 100 days of elapsed time,
the arsenic concentration in the sand column
reached close to the breakthrough point because
sand can remove only approximately 16.17 mg
arsenic from the water, indicating that the ability
of sand itself to remove arsenic from contaminated
water decreases with time. Thus to improve arsenic
removal performance, additional iron must be
supplied.
The fourth column test evaluated the arsenic
removal efficiency from water contaminated with a
high level of arsenic (Fig.5). After 180 days
elapsed (442 pore volumes), the arsenic
concentration of treated water was 35 µg/L, which
is still below the drinking water standard of
Bangladesh (Arsenic conc. < 50 µg/L).
Again considering the ZVI and sand mixture ratio,
the results for columns 1 and 2 demonstrate that
column 1 performed better than column 2. In
addition, the side sampling ports results revealed
that the lower 8 cm adsorbent medium is sufficient
to ensure that the arsenic concentration remains
below 10 µg/L, i.e. about 30% ZVI is sufficient to
treat arsenic.
Fig8. SEM image of (a) pristine ZVI, (b) pristine ZVI
and sand mixture, (c) ZVI corrosion product, and (d)
enlarged view of corrosion product. Reaction conditions:
1000 µg/L As water passed through the column for 180
days at a flow rate 720 mL/day.

Solid Phase Studies
Figure 8 shows the SEM images for solid samples
collected from pristine ZVI, ZVI and sand mixture,
and the arsenic-treated ZVI corrosion products
after 180 days with an initial arsenic concentration
of 1000 Pg/L. Both arsenic adsorption and reaction
time cause an increase in particle aggregation due
to iron (III) oxide/hydroxide precipitate. In some
areas, mineral precipitates with the iron grains; in
others the iron grains remain free of corrosion
products. SEM pictures clearly indicate the growth
of a fine needle-like crystallite, which transforms
to an apparent amorphous phase. Kanel et al.
(2005) have observed similar results.
Iron samples from column 4 were extracted in 1 N
HCl at room temperature for 2 hr. The supernatant
was analyzed for total arsenic. The recovery of the
1 N HCl leach was greatest near the inlet region of
the column and progressively decreased towards
outlet region. Near the inlet region, the recovery of
arsenic was approximately 62% of the total
amount recovered based upon integration and the
recovery near the outlet region of the column was
about 16%. This trend suggests that the conditions
near the outlet region favor more stable arsenic
removal, i.e., a greater portion of arsenic is present
in a form that is not extractable with 1 N HCl.
Hsing-Lung and Richard (2005) have found
similar results.
CONCLUSION
The efficiency of ZVI for the removal of arsenic
from groundwater was studied. The arsenic
concentration in the treated water remained below
10 µg/L. Solid phase analysis revealed that these
ZVI still can work well for longer time. Finally,
batch and column experiment results suggest that a
filtration system having sand and ZVI mixture can
be a suitable solution for treatment of arsenic
contaminated groundwater in the developing
countries due to its high reactivity.
REFERENCES
Anon. (1977) Aqua. United Nations Water
Conference, vol. 1(1), 7.
Argonne National Laboratory (ANL). (2003)
Review of arsenic removal technologies for
Abedin et al. 845
contaminated groundwaters. Chemical
Engineering Division, Argonne, USA.;
http://www.doe.gov/bridge
Bang, S.; Johson, M.D.; Kortiatis, G.P. and Meng,
X. (2005). Chemical reactions between arsenic
and zero valent iron in water. Water Res., 39,
763-770.
Berg, M.; Luzi, S.; Trang, P.T.K.; Viet, P.H.;
Giger, W. and Stuben, D. (2006) arsenic
removal from groundwater by household sand
filters: Comparative field study, model
calculations and health benefits, Environ. Sci.
Technol., 40, 5567-5573.
Biterna, M.; Arditsoglou, A.; Tsikouras D. (2007).
Arsenate removal by zero valent iron. Batch and
column tests. Journal of Hazardous Materials,
149, 548-552.
Borho, M. Ber. Wasserguete-Abfallwirysch. Tech.
Univ. Meuenchen. (1996) 127, 1-119.
Hsing-Lung, L. and Richard T, W. (2004). High
level arsenite removal from groundwater by
zero valent iron. Chemosphere, 59, 377-386.
Kartinen, E.O. and Martin, C.J. (1995) An
overview of arsenic removal processes,
Desalination, 103, 79-88.
Khan, A.H.; Rasul, S.B.; Munir, A.K.M.;
Habibuddowla, M.; Alauddin, M.; Newaz, S. S.;
Hussam, A. (2000) Appraisal of a simple
arsenic removal method for groundwater of
Bangladesh. J Environ. Sci. Health., 35, 1021-
1041.
Meng, X.G.; Bang, S.B.; Korfiatis, G.P. (2000)
Effect of Silicate, Sulfate and Carbonate on
Arsenic Removal by Ferric Hydroxide, Water
Res., 34, 1255-1261.
Meng, X.G.; Korfiatis, G.P.; Christodoulatos, C.;
Bang, S.B. (2001) Treatment of arsenic in
Bangladesh well water using a household co-
precipitation and filtration system, Water Res.,
35, 2805-2810.
Ongley, E.D. (1999) Water quality: an emerging
global crisis. In: Trudgill, S.T., Walling, D.E.,
Webb, B.W. (Eds.), Water Quality: Processes
and Policy. London.
Zouboulis, A.; Katsoyiannis, I. (2002) Removal of
arsenates from contaminated water by
coagulation–direct filtration, Sep. Sci. Technol.
37, 2859-2873.
 Surfactant Enhanced Decontamination of Petroleum Grease
Contaminated Soil
Ganesh W. Rathod
Research Scholar, Deptt. of Civil Engg., IIT Delhi, New Delhi - 110016, India.
(ganesh.rathod@yahoo.co.in)
K. S. Rao
Professor, Deptt. of Civil Engg., IIT Delhi, New Delhi - 110016, India. (raoks@civil.iitd.ac.in)
K. K. Gupta
Associate Professor, Deptt. of Civil Engg., IIT Delhi, New Delhi - 110016, India. (kkg@civil.iitd.ac.in)
R. K. Srivastava
Professor, Deptt. of Civil Engg., MNNIT , Allahabad - 211004, Uttar Pradesh, India. (rksciv@yahoo.com)

ABSTRACT: Contaminant modifies the engineering properties of soil, thereby restricting its further use either as a
construction material or supporting medium. Due to the scarcity of land, it becomes imperative to reuse the land for
infrastructure developmental activities. Out of numerous soil contaminants, automobile grease is one of the prominent
soil contaminant. Considering the cumulative impact of contamination by grease on soil properties, its remediation has
become inevitable. An experimental programme was undertaken to reclaim the soils contaminated with grease in this
study. The performance of surfactant to remove the petroleum grease from soil and its effect on restoring the
geotechnical properties was quantitatively evaluated. Locally available soil (clay with medium plasticity) was
contaminated with varying percentages of grease i.e. 3%, 6% and 9% by dry weight of soil. Geotechnical properties of
virgin (0% grease), contaminated and decontaminated soils were determined. Surfactant namely LOC (Liquid Organic
Cleaner) was employed to decontaminate the soil.

1. INTRODUCTION has resulted in a major technical opportunity for

Population growth, urbanization, and industrial
revolution lead to the formation of various kinds
of wastes. In case of sites formerly occupied by
some waste generating industries or subjected to a
variety of chemicals, inorganic or organic waste;
careful site evaluation and design measures are to
be devised for its safe use. In this direction, geo-
environmental studies play a major role in
predicting the behaviour and performance of a
contaminated soil to be used as a construction or
support material for any structure. Soil and ground
water contamination affecting housing, industry,
agriculture etc. and involving large amount of
money and time. Waste reduction is a goal that
will be driven forward by economic conditions. As
the cost of waste disposal increases, new methods
of waste reduction will become economically
attractive. Several case studies reported the
degradation of geotechnical properties and
subsequently significant subsidence due to
petroleum hydrocarbons (Cheng et al., 2004). This
research and development in the areas of soil
contamination and remediation.
Petroleum and its refined products (grease,
gasoline diesel, kerosene, engine oil etc.) are
major global resources used in industries for
energy requirement. The important sources of land
contaminations are; waste dumps of industrial and
municipal waste, fuel stations, refineries,
accidental spills and leakage, scrap metal storage
and residues of pesticides etc. Petroleum
hydrocarbons are one of the most frequently
encountered pollutants in soil habitats. Numerous
studies has been carried out to decontaminate such
soils (Cheng et al., 2004; Zhu et al., 2004).
The main aim of the study is to investigate the
effect of contamination of grease on locally
available soil with reference to their engineering
behaviour and its reclamation. More specifically
the objectives addressed in this study are i) Study
 Rathod et al. 847

of the changes in engineering properties and
characteristics of the contaminated soil ii)
Elucidation of mechanism of attachment of
contaminants in the soil system iii) Proposal for an
effective strategy for decontamination iv)
Evaluation of reclamation potential of the
contaminated soils.
The study will be helpful in deciding the
suitability of contaminated land for its intended
use keeping in view of modified behaviour of
contaminated soil. This study would also facilitate
in planning and decision making in the processes
like containment of site with design of
barrier/liner, selection of disposal sites and
remediation of contaminated sites.
Table 1. Depending on these properties, the soil is
classified as clay with medium compressibility.
Relevant parts of IS code 2720 were referred for
geotechnical tests and their description is given in
Table 2. X-ray diffraction (XRD) and Scanning
Electron Microscopic (SEM) studies were
conducted to estimate mineralogical constituents
and morphological nature of the soil. The XRD
pattern and SEM image of the soil are shown in
Fig. 1 and Fig. 2 respectively. The major
constituents of soil are quartz, felspar, dolomite,
calcite, kaolinite and illite. The main elements of
the soil microstructure are voids and aggregates of
particles.
TABLE 1. Physico-chemical properties of soil used

2. MATERIALS Particle Size Distribution

The materials used in this study are experimental Sand (%) 12.0
soil, petroleum hydrocarbon contaminants and Silt (%) 70.7
surfactants. The description of these materials and Clay (%) 17.3
their basic properties are given in the following Atterberg Limits
sections. Liquid Limit (%) 45
Plastic Limit (%) 17
2.1 Soil
To eliminate the soil characteristics as a variable, Shrinkage Limit (%) 13
soil used in this study was acquired from a single Plasticity Index (%) 28
and previously undisturbed local site. The sample Soil Classification (IS: 1498-1970) CI
was then hand sorted to remove any pebbles and Activity 1.87
vegetative matters present in it. Sieve analysis and Specific Gravity 2.62
sedimentation analysis by hydrometer were Cation Exchange Capacity (meq/100 g) 22
carried out on oven-dried samples. Atterberg Maximum Dry Unit Weight (kN/m3) 19.04
limits for the soil were also determined in order to
Optimum Moisture Content (%) 14.4
classify the soil. The results for grain size
Hydraulic Conductivity (m/sec) 4 x 10-9
distribution and Atterberg limit are presented in

TABLE 2. Description of Geotechnical Tests

S.
Properties Studied
No.
1 Grain Size Distribution
2 Atterberg’s Limit
3 Density Index
4 Specific Gravity
5 Compaction
6 Permeability
7 Consolidation
Strength Parameters
8 Part 11: 1993, Part 12: 1981 Reaffirmed 1992
Triaxial Shear Test
9 pH
Relevant Parts of IS 2720 Equipment Used
Part 4: 1985 Reaffirmed 1990 Standard sieves, shaker, hydrometer
Part 5: 1985 Reaffirmed 1990 Cassagrande apparatus
Part 14: 1983 Reaffirmed 1990 Compaction mould, vibrating table
Part 3/Section 1 and 2: 1980 Reaffirmed 1992 Density bottle
Part 7: 1980 Reaffirmed 1992 Standard proctor test apparatus
Part 17: 1986 Reaffirmed 1992 Variable head
Part 15: 1986 Reaffirmed 1992 Fixed ring odeometer
Triaxial test apparatus without pore
pressure measurement
Part 26 : 1987 pH meter
848 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
Fig. 1 XRD pattern of virgin soil
Fig. 2. Scanning electron image of virgin soil
2.2 Contaminant (Grease)
Based on the literature contaminant belonging to
the most widely used petroleum hydrocarbon
group, grease was chosen for the study. The soil
was contaminated in the laboratory artificially
with varying i.e. 0, 3, 6, and 9 percentages of
contaminants.
Castrol Ultratack AP3 is premium quality
multipurpose grease suitable for all types of
vehicles and industrial machinery, is selected for
the study. This particular type of grease is selected
for the study because these greases are considered
as combining many of the desirable features of
calcium and sodium greases, due to which these
greases may be used for all grease applications.
The physical properties of grease are provided by
the manufacturer and are presented in Table 3.
TABLE 3: Physical Properties of Grease
Properties Value
Colour Cream
Temperature -45 0C to 200 0C
Dropping point 304 0C
Viscosity 45.8 centiStoke (at 40 0C)
2.3 Surfactant
Surfactants are used for soil washing or flushing
due to their ability to mobilize contaminants. They
contain a hydrophobic portion with little affinity
for the bulk medium and a hydrophilic group that
is attracted to the bulk medium. Surfactants can
lower the surface and interfacial tension by
disrupting hydrogen bond between water
molecules, and thus facilitate the wetting of
surface and the transport of organic contaminants
from soils into a washing solution. For the present
study Liquid Organic Cleaner popularly known by
its trade name LOC was chosen.
3. DEGREE OF CONTAMINATION
Degree of contamination was defined as the
percentage weight of contaminants with respect to
dry weight of soil. Literature (Kirov, 1989 and
Sunil et al., 2006) has shown that some types of
contaminants change the properties of their host
soils and this behaviour has been shown to be
dependent on the concentration of the
contaminant. The soil can be regarded as
contaminated when total petroleum hydrocarbon
content exceeds 0.2g/kg. Baptista et al., (2005)
reported oil and grease and total petroleum
hydrocarbon (TPH) content, 27.5±1.5 g/kg and
9.7±0.1 g/kg, respectively in the crude oil
contaminated soil. Pala et al., (2002) found 8% of
contamination levels of organic carbon and oil and
grease in the soil selected for their study. Initial
degree of contamination was fixed at 3% because
the state of New Jersey classifies soils with oil
concentration above 3% as hazardous waste
(Meegoda and Ratanweera, 1995). Based on this
consideration the soils were contaminated at 3%,
6% and 9% degree of contamination for studies.
4. RECLAMATION STRATEGY
A number of techniques are now available for the
remediation of contaminated land. The most cost

effective option depends upon engineering factors
such as soil characteristics (e.g. soil type, grain
size, void ratio, degree of saturation etc), site
geology, lateral extent of contamination location,
climate, land usage, surface and ground water
conditions, accessibility and type and amount of
chemicals (Mulligan et al., 2001). Techniques like
bio-remediation are slow and require large time
spans. Electro-kinetics is an effective remedial
technology for removal of heavy metals from all
type of soils but yields poor results in case of
petroleum hydrocarbon. Chemical treatment
technologies such as solidification or stabilization
techniques could be applied in the permeable
formation only and are irreversible whereas soil
matrix may be chemically changed. Among
physical methods, thermal treatment though
suitable for most of the type of contaminants
require high thermal energy making the process
expensive, hence, it is less practical. Air sparging
technique can be well applicable to coarse-grained
soil (k > 10-5 m/sec) only where airflow is
relatively uniform and predictable. The surfactant
enhanced washing techniques could be a viable
solution for the removal of hydrophobic organic
such as non-aqueous phase liquids present in the
soils and aquifers. The surfactant enhanced
washing method seems to be effective in removing
petroleum hydrocarbon from all type of soils.
Reclamation potential of the surfactant LOC is to
be evaluated by conducting tests on
decontaminated samples.
5. METHODOLOGY
The various operations involved in the study
consist of evaluation of geotechnical properties of
selected soil, contamination of the soil with
petroleum hydrocarbons (grease), ascertaining the
geotechnical properties of contaminated soils
followed by decontamination studies.
Decontamination studies included soil washing
with surfactant and ascertaining the effectiveness
of this method in the removal of contaminants
from soil matrix. The contaminated samples were
decontaminated with application of optimum
concentration of LOC. Finally, the geotechnical
properties were evaluated for decontaminated
samples in order to ascertain the reclamation
Rathod et al. 849
potential of soils with the used surfactant. Index
and engineering properties of selected soil were
determined in the laboratory using the standard
methods recommended in relevant parts of IS code
2720.
5.1 Methods of Contamination
Experimental soil was characterized by grain size
analysis and Atterberg limit tests and classified as
clay with medium compressibility (classified as CI
as per IS 1498). To the oven dried and air cooled
soil, contaminant (grease) were mixed in
proportion of 3%, 6% and 9% by weight of dry
sample and allowed to come to equilibrium for
seven days. For this, calculated weight of
contaminating grease as per degree of
contamination was mixed to 5 kg of soil. The soil
contaminant mixture was pulverized with hands
and thoroughly mixed for one hour in a covered
tray. The mixture was placed in a covered
container for a week so that contaminating grease
came to equilibrium with the soil. During this
equilibrium period of one week, soils were mixed
from time to time. After one week of interaction
between soil and contaminant, geotechnical
properties of contaminated soils were evaluated
again.
5.2 Decontamination of Soil
For decontamination of soil, optimum
concentration of LOC used is 0.4%. The optimum
dose of surfactant (LOC) was taken with the
presumption that index properties of contaminated
soils i.e. Atterberg limit and grain size distribution
could be restored at this dose. Then, 5 kg of
contaminated soil were batch washed in a 30 liter
steel container. The amount of water taken was in
the ratio of 1:4 (Weight of soil to volume of
water). The required amount of LOC was
thoroughly mixed with the water in container.
After that, soil was added in LOC mixed water in
the container and was vigorously stirred for half
an hour and then left standing still for two hours.
Subsequently, water along with foam produced
during mixing was decanted. The soil in the
container was washed again with same amount of
fresh tap water and decanted. During the process
of decantation, care was taken that fines present in
850 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

the soil would not drain along with the liquid. The the OMC can be reclaimed in the range of 86% to
decanted solution was kept standing still in a 99%.
separate container for 24 hours. After that, it was
filtered to retain the soil fines, which could have 100
0%
gone with decant. The filtered soil was remixed 3%
6%
with washed soil. The washed soil was left to be 80 9%

air-dried for 6 to 7 days. Subsequently,

geotechnical properties were evaluated again in 60

% Finer
order to ascertain the reclamation potential and
presented in Table 4. 40

20
6. RESULTS AND DISCUSSION
The percentage of degree of contamination i.e.
0
grease is expressed as weight by weight (w/w) 0.001 0.01 0.1 1 10

with respect to the weight of soil, while the Particle Size (mm)

percentage of surfactant i.e. LOC is expressed as Fig. 3. Grain Size Distribution Curve for Virgin and
volume by volume (v/v) with respect to the Contaminated soil
1.95
volume of water. The results of geotechnical tests 0 %
3 %
on virgin (0% grease), contaminated and 1.90 6 %
9 %
decontaminated soil samples have been presented 1.85
Dry Density (gm/cc)

in Table 4. Table 4 also shows the % improvement

1.80
over contaminated soil after decontamination and
% regain over virgin soil after decontamination 1.75

with LOC. Figure 3 and Fig.4 represents the grain 1.70

size distribution and compaction behavior of

1.65
virgin and contaminated soils. Insignificant
changes are observed in grain size after 1.60
8 10 12 14 16 18 20
contamination. It has been observed from the W ater Content (%)

results that engineering properties such as
maximum dry density, optimum moisture content,
angle of internal friction, cohesion, compression
index and coefficient of consolidation of the soil
are recovered up to 88% (average) after washing
with the surfactant LOC.
An attempt has been made to capture the changes
in the engineering properties of soil through
experimental programme. The effect of
contamination and decontaminated with petroleum
grease and LOC respectively, on properties of soil
can be explained here.
Optimum Moisture Content (OMC): A decrease
in OMC with the contamination is observed and
shown in Fig. 5(a). Due to addition of grease, the
pore spaces in soil get reduced and the water
proofing quality of grease may be responsible for
reduction in OMC values. After decontamination
Fig.4. Compaction curves for virgin and contaminated
soils
Maximum Dry Density: Due to addition of
foreign particles (grease), there must be an
increase in MDD due to mobility and decrease in
pore spaces. The lower specific gravity of grease
and the opposite charge on the soil particles and
grease might have resulted into the strong edge to
edge attractive forces and hence decrease in MDD
(Fig. 5 b). It can be inferred that, due to addition
of grease, the rearrangement of soil skeleton takes
place and tend towards the more flocculated array
rather than dispersion for clay particles. After
decontamination the MDD can be reclaimed in the
range of 96% to 99%.
Angle of Internal Friction and Cohesion: Due to
grease the depression of double layer around the
soil particle might have taken place and resulted
 Rathod et al. 851

Table 4. Comparison of properties of Original, Contaminated and Decontaminated Soil

% % Improvement
%
Degradation Over % Regain Over
% of Removal
Virgin Contaminated Decontaminated of Virgin Soil Contaminated Virgin Soil after
Contamination of
after Soil after Decontamination
Grease
Contamination Decontamination
OMC
0.0 14.40 14.40 14.40 -- -- -- --
3.0 14.40 13.78 14.25 4.306 3.411 98.958 0.6
6.0 14.40 13.13 13.51 8.819 2.894 93.819 1.2
9.0 14.40 10.49 12.45 27.153 18.684 86.458 2.3
Ȗd max (kN/m3)
0.0 19.04 19.04 19.04 -- -- -- --
3.0 19.04 18.49 18.98 2.889 2.650 99.685 2.8
6.0 19.04 18.36 18.74 3.571 2.070 98.424 4.0
9.0 19.04 17.85 18.31 6.250 2.577 96.166 3.6
cu (kg/cm2)
0.0 0.507 0.507 0.507 -- -- -- --
3.0 0.507 0.554 0.513 9.339 7.403 98.756 1.7
6.0 0.507 0.612 0.583 20.908 4.735 84.817 1.3
9.0 0.507 0.825 0.692 62.804 16.032 63.297 1.9
ĭ (Degree)
0.0 13.0 13.0 13.0 -- -- -- --
3.0 13.0 10.8 12.4 16.923 14.815 95.385 2.2
6.0 13.0 7.3 8.8 43.846 20.548 67.692 1.2
9.0 13.0 5.8 6.6 55.385 13.793 50.769 1.2
Cc
0.0 0.245 0.245 0.245 -- -- -- --
3.0 0.245 0.247 0.246 0.816 0.405 99.592 1.4
6.0 0.245 0.254 0.251 3.673 1.181 97.551 1.1
9.0 0.245 0.261 0.258 6.531 1.149 94.694 1.0
Cv (*10-5) (cm2/sec)
0.0 3.306 3.306 3.306 -- -- -- --
3.0 3.306 2.108 2.462 36.237 16.793 74.471 1.0
6.0 3.306 2.230 2.925 32.547 31.166 88.475 3.1
9.0 3.306 2.870 3.137 13.188 9.303 94.888 1.7

into the decrease in angle of friction. The
lubricating action of grease also helped to
decrease the angle of internal friction. Its behavior
is shown in Fig. 5 (d). An increase in cohesion
value with the increase in degree of contamination
is observed and shown in Fig. 5 (c). The observed
percentage regain over virgin soil after
decontamination for angle of friction and cohesion
is 50% to 95% and 63% to 98% respectively.
Compression Index and Coefficient of
Consolidation: A sudden decrease in coefficient
of consolidation is observed due to the addition of
grease but an increasing trend of compression
index and coefficient of consolidation has been
observed after contamination and the values are
given in Table 4. The continuous loading may lead
to the depression of double layer and expulsion of
voids.
Reasonably precise and accurate estimate of the
oil and grease content of the contaminated soils
are possible. The accurate estimates are more
useful for evaluating the environmental and
chemical aspects rather than engineering
geotechnical properties. The focus of the study
was to reclaim the soil as a construction material
and hence a simple, indirect and rational
methodology has been adopted to find out the
grease retained in the soil and grease removed
during the washing process. The changes in the
engineering properties of the soil were used for
852 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

4
2%
(a) (b)

(c) (d)

Fig. 5. Variation of properties with amount of contamination (a) maximum dry density (b) optimum moisture content
(c) cohesion (d) angle of internal friction (For colour figure, refer to CD)

this purpose, while the chemical processes can be
used for estimating the exact amount of grease
removed. The procedure is shown in Fig. 5 (a) for
the case of MDD for 6% contamination and
decontamination with LOC. The Fig. 5 (a) shows
that for this particular case still 2% grease is
present in the soil i.e. a 4% removal of grease was
possible during the soil washing. Likewise the
percentage removal is calculated for all degree of
contamination for each soil property and presented
in Table 4. The results show that, the changes in
engineering properties are not proportionate with
each other for a particular degree of
contamination. This clearly shows the effect of
contaminant on each soil property is different.
And this is the reason that percentage
degradation/improvement of soil properties after
contamination/decontamination is different for a
particular degree of contamination. By considering
all the soil properties for a particular degree of
contamination, the average percentage removal of
grease from the original amount of grease added is
calculated as 1.61%, 1.98% and 1.95% for 3%,
6% and 9% contamination respectively. On
average the engineering properties of soil
contaminated with petroleum grease can be
restored upto 88% by soil washing with the LOC.
Modification of soil properties are observed in soil
contaminated with petroleum grease due to
physico-chemical interaction between soil and
contaminant. Following indirect inferences has
been drawn to explain the observed variations.
x Important parameters in the physical
interaction are viscosity, density and surface
tension of contaminating soil.
x Physical interaction is predominant in the
case of granular soils. Whereas physico-
chemical interaction may take place in fine-
 Rathod et al. 853

grained soils. For soils with high plasticity of Indian Standards, New Delhi, India.
chemical interaction overrides the physical
Kirov B. (1989). Influence of waste water on soil
interaction.
deformation, Proc. of 12th ICSMFE — 1989,
x For non-polar organic contaminating fluid,
Brazil, 1881-1882.
the dielectric constant is the important
parameter to bring the changes in the Meegoda, N.J. and Ratanweera, P. (1995).
thickness of diffused double layer in clayey Treatment of oil contaminated soils for
soils. Smaller value of dielectric constant for identification and classification. Geotech
petroleum oils may change the thickness of Testing J. ASTM, Vol. 18(1), 41-49.
diffuse layer and consequently structure of
Mulligan C.N., Yong R.N. and Gibbs B.F. (2001).
clays tend to be more flocculated. The
Surfactant-enhanced remediation of
lubrication is also responsible for flocculated
contaminated soil: a review. Engg. Geology,
array.
Vol. 60, 371-380.
7. CONCLUSIONS Pala D.M., Freire D.D.C., Sant Anna Jr. G.L.
Reclamation of soil is a practical necessity not (2002). Bioremediation of clay soils impacted
only due to modified behaviour of soil but also by petroleum. Engenharia Térmica, Edição
from the environmental considerations. Liquid Especial, 29-32.
Organic Cleaner (LOC) enhanced washing is an
Sunil B.M., Sitaram Nayak and Shrihari S. (2006).
effective method for decontamination of soil
Effect of pH on the geotechnical properties of
contaminated with petroleum grease. Although the
laterite, J. of Engg. Geology, Vol. 85, 197-203.
100% removal of contaminants and reclamation of
soil properties to its original is not possible but the Zhu K., Chen H. Yang R. and Zhou W. (2004).
soil washing with LOC can prove an effective Enhancing attenuation of petroleum pollutants
methodology. The reclamation potential of LOC by modifying natural soil with surfactants.
in decontaminating the clay with medium Practice Periodical of Hazardous, Toxic and
compressibility could be gauged as 88 %. Radioactive Waste Management, ASCE, Vol.
8(2), 128-135.
REFERENCES
Baptista S.J., Cammarota M.C., Carvalho D.D.
(2005). Evaluation of the biostimulation of
crude oil contaminated clay soil. 2nd Mercosur
Congress on Chemical Engg. and 4th Mercosur
Congress on Process Systems Engg., Village
Rio das Pedras, Club Med, Rio de Janeiro, 1-9.
Cheng D., Hadfield P. and Gamette S. (2004).A
Successful Approach to Site Remediation and
Redevelopment of Property Contaminated with
Oil Field Wastes. Ports 2004: Port
Development in the Changing World, ASCE, 1-
10.
IS 1498: (1970). Classification and identification
of soils for general engineering purposes.
Bureau of Indian Standards, New Delhi, India.
IS 2720: (1980). Methods of test for soils. Bureau
 Centrifuge Modeling of Air-Sparging Technique for Groundwater
Remediation
Liming HU, Jianting Du, Wu Xiaofeng
State Key Laboratory of Hydroscience and Engineering, School of Civil Engineering, Tsinghua University,
Beijing 100084, China (gehu@tsinghua.edu.cn)
J. N. Meegoda
New Jersey Institute of Technology, Newark, USA (Meegoda@njit.edu)

ABSTRACT Air sparging (AS) is one of the groundwater remediation techniques to clean up voltaic organic
compounds. Based on the centrifuge tests employing glass beads as soil model, the process of air sparging is
simulated under a wide range of sparging pressure and centrifuge g levels. The test results show the increase
of the sparging pressure improves the air flow velocity; the zone of influence (ZOI) expands with the increase
of sparging pressure. The ZOI decreases with the increase of g level. The ZOI is cone-shaped and can be
described by use of the intrusion length and cone angle.

INTRODUCTION The nature and extent of the air pathways

Soil/groundwater contamination is currently one
of the major environmental problems in the world.
Many remediation technologies have been
developed to clean the contaminant in soils and
groundwater. There are numerous cleanup
technologies for gasoline- contaminated sites such
as soil vapor extraction (SVE), air sparging (AS),
biosparging, land farming, biopiles, bioventing,
thermal desorption, and monitored natural
attenuation (USEPA 1995). Air sparging is an
in-situ groundwater remediation technology,
injecting pressurized air into the contaminated
subsurface to displace the existing pore fluid, and
creating a transient state to extract the trapped
contaminants into air flow. Air sparging can
remove volatilized contaminants by advection and
enhanced biodegradation in subsurface. Air
sparging systems must be designed with adequate
pressure gradient and air flow for the effective
removal of contaminants.
determine the zone of influence during air
sparging remedial process, and the efficiency
depends on microscopic effects such as air flow
patterns as well as macroscopic effects such as the
spatial distribution of air paths. Laboratory testing
was conducted to investigate the efficiency of air
sparging and its influencing factors (Ji et al. 1993,
Johnson et al. 1993, Elder & Benson 1999, Reddy
& Adams 2001, Kirtland et al. 2001, Adams &
Reddy 2003, Waduge et al. 2004, Hu & Liu 2008).
The efficiency of in situ sparging system is
controlled by the extent of contact between
injected air and contaminated soil and pore fluid.
Lundegrad & Andersen (1996) reported that the
ZOI is cone-shaped, and the parabolic shape of
ZOI was demonstrated by several researchers (Ji
et al. 1993, McCray & Falta 1997). However, the
conical zone can be used to estimate the volume
of the ZOI. With larger particle size of soil (e.g.
gravels), the size of ZOI is smaller than that in the

small particle size of soil (e.g. silty sand). The
angle between the vertical axis and the cone
boundary ranges from 15° for gravel to 56° for
fine sand (Nyer & Suthersan 1993, Reddy &
Adams 2008). The influence of heterogeneity
(Reddy & Adams 2001, Ji et al. 1993, Lundegrad
&Andersen 1996, Waduge et al. 2004) and
groundwater flow (Reddy & Adams 2001) on the
ZOI has also been studied by many researchers.
Reddy & Adams (2008) stated the ZOI did not
vary in size during air injection and was
independent of air injection rate. However, most
of the experiments were conducted under
traditional ground test. The air flow
characteristics during air sparging were
investigated using centrifuge tests under a wide
range of gravitational accelerations by Marulanda
et al. (2000). The effect of injector geometry on
the breakthrough velocity and air plume shape
was studied under a specific injection pressure.
Hu et al. (2010) investigated the effect of
particle size and air injection pressure on size
and shape of the zone of influence (ZOI), and
reported that ZOI can be expressed by two
components: the horizontal expansion due to
preferential intrusion around the injection point
and the angle of ZOI which is the angle
between the vertical line and the boundary of
ZOI. Also, there exists a limited angle of ZOI
for each type of porous media.
In this study, centrifugal modeling tests were
performed to investigate the air flow features
during the air sparging process under different
g-levels and a wide range of injection pressure.
The ZOI was observed, and the air flux rate was
measured during air sparging. The effect of air
sparging pressure and centrifugal g-levels on air
flow feature and ZOI was discussed.
MATERIALS AND FACILITIES
Testing Materials
The opacity of natural soil prevents the direct
observation of air flow patterns during air
Hu et al. 855
sparging process. The transparent granular
material is often used to simulate the natural soil
and facilitate visual observation. In this study,
water was used as pore fluid to represent the real
viscosity and unit weight. A kind of fused silica
glass bead was used to simulate sand, since it is
transparent and has almost the same physical
properties as sand particle.
A coarse glass bead was used in the physical
modeling tests. The physical properties are shown
in Table 1, and the testing cases are illustrated in
Table 2. Preliminary test showed that the air
movement behavior in the sample could be
adequately caught by video recorder.
TABLE 1 Physical Properties of Glass Beads
Particle Effective Uniformity Curvature
Size size d10 Coefficient Coefficient
(mm) (mm) Cu Cc
1.5-2.0 1.55 1.16 0.98
TABLE 2 Cases for Centrifuge Tests
Sample Height Test type Porosity
(cm) (g-level) (%)
30 15, 25, 40, 50, 67, 83 36
Test Setup for Air Sparging
A soil strongbox, which has internal dimensions
of 0.6 m (length) u 0.8 m (height) u 0.04 m
(width), was used to carry out the experiments.
The strong box is made of plexiglass to achieve
the visibility of samples during testing. As shown
in Fig. 1, there is an air sparger at the bottom of
the strongbox, which is connected to a supply
system for compressed air. The air sparging
system consisted of an air compressor, regulator,
pressure gauge, flow meter, valve, and sparger
(Hu et al. 2010). The sparger is a 10 mm high
Plexiglas pipe with the internal diameter of 2 mm,
connected to the valve through a PVC pipe.
Centrifuge modeling tests were conducted in the
geotechnical centrifuge laboratory at Tsinghua
University, which has a payload capacity of 50
g-ton with a maximum acceleration of 250 g. The
image data can be collected by the data collecting
856 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
system mounted in the centrifuge and transferred
to the control room via slipping rings. The
compressed air is injected into samples via fluid
conduit in the slipping ring.
80 Bolt
Air Sparger
60
Fig. 1 Strongbox for centrifuge tests (Unit: cm)
Soil Model Preparation
The sand raining method was adopted to obtain
the uniform sample. Then the sample was
connected to a water tank and saturated by
applying a small hydraulic head difference. The
water table was kept 2 cm above the sample
surface, and stayed for 12 hours before
experiment.
Experiment Procedure
Based on the centrifuge tests adopting glass beads
as soil model, the process of air sparging was
simulated under a wide range of sparging
pressures and centrifuge g level.
The soil model was spun at desired g-level for 30
min for further compaction and homogenization
before air sparging test. The water table was
maintained as constant during all the tests, and the
pore water pressure can be measured by means of
miniature transducer close to the sparger point.
Then the air pressure was gradually increased
until air injected into the sample. The pressure
was considered as minimum sparging pressure.
After air entering, the pressure was increased step
by step at a very small rate. For each step, the
pressure was kept for 10 min to stabilize the air
flow and to facilitate the full development of ZOI.
The air flow rate and the development of the ZOI
were monitored and recorded by video camera.
The test stopped when the ZOI did not expand
any further after several steps of air pressure
increase.
TEST RESULTS
Air Flow Patterns
Bubble flow pattern was observed in all the
centrifuge tests. The quantity and velocity of the
bubbles increased with the increase of sparging
pressure. The velocity of the bubbles increased
during the movement. The distribution of the
bubbles was not uniform, and the bubble density
was larger close to the sparging point. The
obvious preferential flow was observed that
bubble moved upwards along the least resistance
route.
Minimum Injection Pressure
The sparging pressure must overcome the
hydrostatic pressure before air flow occurs, so
“effective sparging pressure”, which is the
pressure difference of sparging pressure and
hydrostatic pressure, was investigated in this
study. At this pressure, there is a stable air flow
entering the sample. The minimum effective
sparging pressure is 26, 30, 41, 27, 22 and 33 kPa
for the centrifuge tests of 15, 25, 40, 50, 67 and
83 g levels. It seems that the minimum effective
sparging pressure is much less than the static pore
water pressure, especially in the high g-level tests.
As a reuslt, it seems reasonable to estimate the
minimum injection pressure in the field as equal
to hydrostatic water pressure.
Air Flow Rate
The stable air flow can be indicated by the
constant air flow rate under a specified sparging
pressure. The effective sparging pressure and
corresponding air flow rate Q were monitored
during the tests, and the relationship is illustrated
in Fig. 2. The air flow rate Q increases
monotonously with the increase of injection

pressure when it exceeds the minimum sparging
pressure.
Fig. 2 Relationship between air flow rate and
sparging pressure
ZOI Development Process with Increase of
Sparging Pressure
Fig. 3 shows the boundary of ZOI under different
effective sparging pressures in various g-level
centrifuge tests. The injection pressure has
obvious effect on the air route way development
when the injection pressure was not very high,
and the ZOI expanded with the increase of the
inject air pressure. The expansion of ZOI is
attributed to the lateral expansion around the air
injection point and the cone angle between the
vertical axis and boundary of ZOI. However, it
increased slowly and even remained constant after
the air sparging pressure reached a certain level,
which is consistent with the finding by Hu et al.
(2010). It can be observed that at different
injection pressures, the boundaries of ZOI were
almost parallel to each other, which means the
angles of ZOI were almost the same. After it
reaches the limited angle, the further expansion of
ZOI is attributed to the lateral expansion around
the injection port due to pneumatic fracture or
preferential intrusion.
ZOI Description
The ZOI was truncated-cone-shaped under
various g-levels. The ZOI angle decreased with
Hu et al. 857
the increase of g-levels. Moreover, there existed
lateral intrusion of air close to the sparging point,
which indicated preferential flow occurred. The
length of lateral intrusion increased with the
increase of air pressure, and decreased with the
increase of g-level. The ZOI can be described by
the cone angle and the length of lateral intrusion
close to the sparging point (Hu et al. 2010).
Furthermore, it is related to the g-level. The larger
the g-level is, the smaller the lateral intrusion
length is. Fig. 4 shows the ultimate ZOI angle and
lateral intrusion length for various g-level tests.
The cone angle decreases with the increase of
g-level except the 83 g-level test. The effect of
g-level needs further investigation.
The above findings from centrifuge tests can
provide useful references for design and
application of air sparging for groundwater
remediation.
CONCLUSIONS
This paper presents the centrifugal tests to
investigate the characteristics of air flow through
saturated porous media during air sparging. The
ZOI is cone-shaped during air sparging process.
There exists lateral intrusion of air close to
sparging point, which indicates the preferential
flow of air through least resistance pore in
micro-structure. The ZOI expands with the
increase of the injection pressure, mainly due to
the increase of lateral intrusion length. The ZOI
can be adequately described by means of
combination of the lateral intrusion length and
cone angle. The length of lateral intrusion and the
cone angle decrease with the increase of g-level.
Centrifuge modeling technique provides a
powerful physical modeling technique to
investigate the mechanism of air sparging in
porous media since a larger range of injection
pressure can be applied due to the high stress
level of the sample under high g-level. Under
higher centrifuge g-level, the air flow rate is faster,
and the ZOI is smaller. The scaling law for air
sparging process needs to be further investigated.
858 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

15 g 25 g

40 g 83 g
Fig. 3 Stable ZOI boundary under different air pressure at various g-levels

Fig. 4 Ultimate ZOI angle and lateral intrusion length for various g-level tests

ACKNOWLEDGEMENTS
The research work reported in this paper was
financially supported by Beijing Scientific
Committee (D07050601510000), MWR Key
Laboratory of Geotechnical Mechanics and
Engineering (G07-04), and National Natural
Science Foundation of China (50879038).
REFERENCES
Adams J.A. and Reddy K.R. (2003) Extent of
benzene biodegradation in saturated soil
columns during air sparging. Ground Water
Monitoring and Remediation, 23(3), 85-94.
Elder C. and Benson C. (1999) Air channel
formation, size, spacing and tortuosity during
air sparging. Groundwater Monitor. Remediat.
19, 171–181.
Hu L. and Liu Y. (2008) Physical Modeling of
Air-Sparging Technique for Groundwater
Remediation. Chinese Journal of
Geotechnical Engineering, 30(6): 835-839.
Hu, L., Wu, X., Liu, Y., Meegoda, J. N., Gao, S.
Physical Modeling of Air Flow During Air
Sparging Remediation. Environ. Sci. Technol.
2010, 44(10): 3883–3888.
Ji, W. et al. (1993) Laboratory study of air
sparging:air flow visualization. Groundwater
Monitor. Remediat., 13(4), 115–127.
Johnson, R. et al. (1993) An overview of in situ
air sparging. Groundwater Monitor. Remediat.
13(4), 127–134.
Kirtland, B.C. et al. (2001) Long-term AS/SVE
for petroleum removal in low-permeability
Piedmont saprolite, J. Environ Eng, ASCE.
127(2), 134-144.
Hu et al. 859
Lundegard, P.D. and Andersen, G. (1996)
Multiphase numerical simulation of air
sparging performance. Ground Water., 34(3),
451-460.
Marulanda, C. et al. (2000) Centrifuge modeling
of air sparging – a study of air flow through
saturated porous media. Journal of Hazardous
Materials., 72, 179-215.
McCray, J.E. and Falta, R.W. (1997) Numerical
simulation of air sparging for remediation of
NAPL contamination, Ground Water., 35(1),
99-100.
Nyer, E.K. and Suthersan, S.S. (1993) Air
sparging: savior of ground water remediation
or just blowing bubbles in the bathtub.
Ground Water Monitoring Review. 13(4),
87-91.
Reddy, K.R and Adams, J.A. (2001) Effect of soil
heterogeneity on airflow patterns and
hydrocarbon removal during in situ air
sparging, J. Geotech. Geoenviron. Eng., 127
(3), 234–247.
Reddy, K.R. and Adams, J. (2008) Conceptual
modeling of air sparing for groundwater
remediation, Proceedings of the 9th
international symposium on environmental
geotechnology and global sustainable
development, Hong Kong, June 2008.
USEPA. (1995) How to evaluate alternative
cleanup technologies for underground storage
tank sites. EPA 510-95-007.
Waduge, W.A.P., Soga, K. and Kawabata, J.
(2004) Effect of NAPL entrapment conditions
on air sparging remediation efficiency, J.
Hazard. Mater., 110, 173–183.
 Electrochemical Degradation of MTBE using
Alternating and Direct Currents
M. V. Khire
Michigan State University, United States of America (khire@egr.msu.edu)
Cheonyong Seo
Michigan State University, United States of America (seocheon@msu.edu)

ABSTRACT Methyl tert-butyl ether (MTBE) is an organic compound, which is highly soluble in water and
is a common ground water (GW) contaminant that is typically released to the GW from gasoline spills
Electrochemical degradation of MTBE was investigated by subjecting MTBE dissolved in deionized (DI)
water to direct and alternating currents (DC and AC) at current densities ranging from 0.5 to 2 mA/cm2. The
initial concentration of MTBE was 250 mg/L, supporting electrolyte was sodium sulfate, and polished
titanium electrodes were used in the electrochemical cell. Both AC and DC resulted in electrochemical
degradation of MTBE but the rate of degradation for DC was greater than that for AC at an equivalent current
density. The rate of degradation increased as the current density increased for DC as well as for AC.

INTRODUCTION µg/L. There are thousands of MTBE contaminated

Methyl tertiary-butyl ether (MTBE) is a chemical
compound that is manufactured by the chemical
reaction of methanol and isobutylene. MTBE was
produced in relatively large quantities (over
200,000 barrels per day) in the U.S. and in late 90s
was almost exclusively used as a fuel additive in
motor gasoline. It is a group of chemicals
commonly known as "oxygenates" because it
raises the oxygen content of gasoline. At room
temperature, MTBE is volatile, flammable and
colorless liquid that dissolves rather easily in water.
MTBE could improve fuel combustion and reduce
carbon monoxide but it is a risk to human health as
a possible carcinogen. It is classified by the United
States Environmental Protection Agency (USEPA)
(Hong et al., 2007) as a toxic chemical. MTBE has
relatively high solubility (~ 42 g/L), low soil-water
partitioning coefficient (KOC = 11; log10 KOC =
1.04), and relatively low retardation factor when
compared with other gasoline additives such as
benzene (EFOA, 2002). Thus, MTBE is highly
mobile, and is relatively difficult to clean up.
USEPA drinking water advisory (USEPA, 1997)
recommended concentration of MTBE ” 20 to 40
ground water sites all over the world that are
waiting for cleanup using cost-effective
remediation system.
Electrochemical Degradation
Organic compounds can be degraded and
transformed by electrochemical degradation
(Pepprah and Khire, 2008; Pepprah, 2007;
Alshawabkeh et al., 2005). Electrochemical
degradation of organic compounds in aqueous
phase occurs due to direct oxidation and/or indirect
oxidation or reduction. During the indirect
oxidation processes, the agents that oxidize
organic compounds are generated. Previous studies
have reported hydrogen peroxide (Brillas et al.,
1995), hypochlorite (Rajkumar et al., 2005), ozone
(Stucki et al., 1987) and hydroxyl radicals
(Yangqing et al., 2007) as strong reactive agents
during electrochemical degradation.
Equations 1.1 and 1.2 show electrochemical
breakdown of water at anode and cathode and
resulting formation of oxygen at anode and
hydrogen at cathode.

Anode: 2H2 o O2 + 4H+ + 4e- (1.1)
Cathode: 2H2O + 2e- o H2 + 2OH- (1.2)
During electrochemical conversion, hydroxyl
radicals are also formed at the anode (Eq. 1.3 and
1.4).
H2O + e- o Hǜ + OH- (1.3)
OH- o OHǜ + e- (1.4)
Direct Current (DC)
A majority of published work on electrochemical
breakdown of organic compounds has been
focused on the use of direct current (DC). Wu and
Lin (2007) used DC electrolysis for degradation of
MTBE using iridium electrodes with supporting
electrolytes that included sulfuric acid (H2SO4),
hydrochloric acid (HCl), and nitric acid (HNO3).
They used initial concentration of MTBE equal to
20 mg/L. Removal rate of MTBE ranged from
27% to 92% after electrolysis for 3 hr for various
values of input voltage, which ranged from 0.5 to
3.0 V for the various electrolytes. Degradation rate
of MTBE increased as a function of applied
voltage.
While the use of DC in the lab-scale studies have
demonstrated acceptable rates of breakdown of
many organic chemicals, use of DC for field
application poses many challenges. Pepprah and
Khire (2008), Khire and Pepprah (2008), and Lee
(2008) have indicated that power usage during DC
is relatively high because the resistance of the
electrolyte increases as the time of application of
DC increases. Increased resistance requires
proportionately greater voltage to maintain the
current which, among other parameters, dictates
the rate of breakdown of organic molecules in the
solution. In addition, application of DC results in
decrease in the pH near the anode and increase in
the pH near the cathode. This results in relatively
rapid fouling of the electrode which causes
decrease in the rate of breakdown and affects the
physical integrity of the electrode.
Khire and Seo 861
Alternating Current (AC)
Alternating Current (AC) is an electrical current
which changes its direction periodically. AC wave
form can have well defined shapes such as
sinusoidal, triangular, square, or it can follow an
arbitrary shape. While the frequency of sinusoidal
AC available in north American power grid is 60
Hz, AC function generators typically can produce
AC at frequencies ranging from 0.1 Hz to 1 MHz.
Relatively few published studies where AC has
been used for electrochemical breakdown of
organic compounds exist. Nakamura et al (2005)
presents degradation of Trichlorobenzene (TCB)
using AC electrolysis in aqueous solution.
Decomposition rate of TCB was 29.6% after
electrolysis for 0.5 hr due to hydrogen and
hydroxyl radicals produced during AC electrolysis.
Pepprah (2007), Pepprah and Khire (2008), and
Khire and Pepprah (2008) have compared the use
of DC and AC for electrochemical breakdown of
naphthalene, pyrene, phenanthrene, and salicylic
acid in aqueous solutions. For a given current
density, these studies concluded that the rate of
breakdown of the organic compounds was greater
for DC compared to AC. For AC, the rate of
breakdown decreased as the frequency of the AC
was increased from 0.1 Hz to 1,000 Hz. However,
the extent of electrode fouling and electrode mass
loss was greater for DC compared to when AC was
applied.
Lee (2008) evaluated the power consumption for
AC versus DC when naphthalene was degraded in
electrochemical cells having volumes equal to 1 L
and 3.5 L. Lee (2008) concluded that the power
consumption per unit volume of the
electrochemical cell was about 2 to 5 times greater
when DC was used. Lee (2008) also found that as
long as the current density (defined as total current
divided by the total area of the electrodes that
faces each other) was constant, the size of the cell
did not influence the rate of degradation of
naphthalene in aqueous solution.
MATERIALS AND METHODS
The experimental setup is presented in Fig. 1. The
test cell, which was an undivided cell, consisted of
862 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India
a 1 liter Pyrex glass beaker with a Teflon cap.
These materials of the cell were selected to
eliminate or reduce sorption of the contaminants
onto the walls of the reaction vessel and the cap.
Fig. 1 Experimental Setup.
Electrodes
Titanium plates were used as electrodes in
electrochemical tests of this study. Alshawabkeh
and Sarahney (2005), Li et al. (2005), and Goel et
al. (2003) have used titanium as a base metal and
coated it with mixed metal oxides such as RuO2,
IrO2, and SnO2. The titanium plates used in this
study were not coated. Uncoated titanium is
cheaper than commonly used electrodes such as
platinum or titanium coated electrodes. Also,
titanium offers resistance to corrosion in a wide
range of aggressive conditions. The titanium sheets
used as electrodes in this study were 12.0-cm-long
by 5.0-cm-wide by 1.72-mm-thick and the distance
between them was maintained at 8.0 cm. The
purity of the titanium was about 99%. After each
experiment, in order to remove deposits, the
electrode surface was cleaned using a power
sander.
Electrical Equipment
A 0.1 Hz to 2 MHz Sweep Function Generator was
used to generate a square-wave AC signal. The
power supply consisted of a bipolar operational
amplifier/power supply capable of producing up to
200 V AC/DC and 1 A AC/DC output or a
maximum electrical power of 200 W. Voltage and
current measurements were obtained with the aid
of an oscilloscope with accessories.
Spiked Aqueous Solution
Aqueous solutions of MTBE were prepared by
adding MTBE in liquid form to one-liter of
deionized (DI) water in a capped Pyrex bottle to
achieve target MTBE concentration of 250 mg/L.
The suspension was allowed to stand for about
three days. This was followed by the addition of
0.5 g of anhydrous sodium sulfate (Na2SO4) to
each suspension, to increase its electrical
conductivity. Sodium sulfate was selected as the
supporting electrolyte because all aqueous solution
tests without sand were carried out using Na2SO4
as the supporting electrolyte. The target
concentration of Na2SO4 was approximately 500
mg/L.
Testing Procedure
Tests were conducted in batch mode in 1-liter
undivided cells. Prior to applying AC or DC to the
cells, samples were collected for gas
chromatography (GC) analyses. For each sampling
event, samples were collected from the center of
the cell. During the tests, samples were collected
for GC analyses at elapsed time intervals of 1, 2, 4,
8, and 24 hours. Calibration curves with an
average R2 value of 0.998 were used to determine
the concentration of MTBE in the solutions as the
tests progressed.
During the test, pH, temperature, voltage, and
electrical resistance, R (for DC) and impedance, Z
(for AC) of the aqueous solution were monitored.
Temperature and voltage were measured at the
sampling time, which was 0, 1, 2, 4, 8, and 24
hours using microprocessor thermometer (ŸE
Omega) and digital oscilloscope, respectively.
Electrical resistance and pH were monitored just
before the test started and just after the test ended
using LCR ESR meter and pH meter, respectively.
In addition, the resistance of the solution was
estimated from the voltage readings that were
taken throughout the test and the fixed current
which was maintained throughout the test. Further
details related to the testing and analyses are
presented in Seo (2009).
Khire and Seo 863
864 6th International Congress on Environmental Geotechnics, 2010, New Delhi, India

by AC for a given current density during the 24-hr Groundwater spiked with MTBE was subjected to
testing period. AC and DC at constant current densities in a 1L
laboratory cell. It resulted in electrochemical
Cumulative Electrical Energy Consumed (kJ/L)

1200
C = 250 mg/L degradation of MTBE. However, the rate of
o

Initial / Final Voltage (Volts)

degradation for DC was greater than that for AC at
1000
2
AC - 2 mA/cm : 49.0/ 61.0
an equivalent current density. The rate of
2
DC - 1 mA/cm : 28.0 / 70.0
degradation increased as the current density
800 2
2
DC - 0.5 mA/cm : 18.0 / 44.0
AC - 2 mA/cm increased for DC as well as for AC.
2
600 AC - 1 mA/cm : 18.5 / 30.0
ACKNOWLEDGEMENT
400 DC - 1 mA/cm
2 Financial support for this project has been
provided by the U.S. National Science Foundation
200 DC - 0.5 mA/cm
2 (Grant No. BES-0402772).
2
AC - 1 mA/cm
0
REFERENCES
0 5 10 15 20 25 Alshawabkeh, A. N., and Sarahney, H. (2005)
Elapsed Time (hrs) “Effect of current density on enhanced
Fig. 4 Cumulative electrical energy consumption for transformation of naphthalene,” Environ. Sci.
AC and DC at various current densities. Tech., 39, 5837-5843.
Brillas, E., Bastida, R.M., Llosa, E. and Casado, J.
1.4 (1995) “Electrochemical destruction of aniline
total Electrical Energy Consumed (mg/kJ)

C = 250 mg/L
o and 4-chloroaniline for wastewater treatment
Cumulative Mass Converted per

1.2
using a carbon-PTFE O2-fed cathode,” Journal
1 DC - 1 mA/cm
2 of the Electrochemical Society, 142, 1733-1741.
EFOA (2002) “MTBE resource guide.”
0.8 Hong, S., Zhang, H., Duttweiler, C.M. and
Lemley, A.T. (2007) “Degradation of methyl
0.6
tertiary-buthyl (MTBE) by anodic Fenton
0.4 DC - 0.5 mA/cm
2 treatment,” Journal of Hazardous Materials,
AC - 1 mA/cm
2 144, 29-40
0.2 Khire, M., and Pepprah, E. (2008) “Degradation of
2
AC - 2 mA/cm
Naphthalene in Aqueous Phase of Saturated
0
0 4 8 12 16 20 24 Ottawa Sand using Alternating and Direct
Elapsed Time (hrs) Currents," GeoCongress 2008, American
Fig. 5 Cumulative MTBE mass converted per total
Society of Civil Engineers, GSP No.177.
electrical energy consumed for AC and DC tests Lee, D.G. (2008) “Effect of Scale During
(For colour figure, refer to CD) Electrochemical Degradation of Naphthalene
and Salicylic Acid,” Master’s Thesis, Michigan
State University.
Hence, the cumulative MTBE mass converted Panizza, M. and Cerisola, G. (2009) “Electro-
during electrolysis when normalized with respect fenton degradation of synthetic dyes,” Water
to the cumulative electrical energy, it indicates that Research, 43, 339-344.
DC was more energy efficient for degrading Pepprah, E. (2007) “Degradation of Polycyclic
MTBE because it resulted in greater mass Aromatic Hydrocarbons (PAHs) in aqueous
converted per unit energy consumed (Fig. 5). Media using Alternating Current,” Ph.D.
Dissertation, Michigan State University.
SUMMARY Pepprah, E., and Khire, M. V. (2008)
“Electroremediation of naphthalene in aqueous

solution using alternating and direct currents,”
J. Environ. Eng., 134(1), 32-41.
Rajkumar, D., Kim, J.G. and Palanivelu, K. (2005)
“Indirect electrochemical oxidation of phenol
in the presence of chloride for waste
treatment,” Chem. Eng. Technol, 28, 1.
Seo, C. (2009) “Electrochemical Degradation of
MTBE using Alternating and Direct Currents,”
Master’s Thesis, Michigan State University.
Stucki, S., Baumann, H., Christen, H.J. and Kotz,
R. (1987) “Performance of a pressurized
electrochemical ozone generator,” Journal of
Applied Electrochemistry, 17, 773–778.
Khire and Seo 865
pp y, ,
USEPA (1997) “Drinking Water Advisory:
Consumer Acceptability Advice and Health
Effects Analysis on Methyl Tertiary-Butyl
Ether (MTBE),” EPA-822-F-97-009.
Wu, T.N. and Lin, Z.C. (2007) “Electrochemical
behavior of methyl tert-butyl ether (MTBE)
oxidation on an iridium (IrO2) coated
electrode,” J. Environ. Eng. Manage., 17(1),
49-56.
Yangqing, C., Zucheng, W. and Yuqing, L. (2007)
“Hydroxyl radical electrochemically generated
with water as the complete atom source and its
environmental application,” Chinese Science
Bulletin, 52, 10, 1432-1435.