permeable reactive barriers: a sustainable technology for cleaning contaminated groundwater in...
TRANSCRIPT
Permeable reactive barriers: A sustainable technology for
cleaning contaminated groundwater in developing
countries
D.H. Phillips
Environmental Engineering Research Centre, School of Planning, Architecture and Civil Engineering,
Queen’s University of Belfast, Belfast BT9 5AG, United Kingdom
Tel. þ44 (0)28 9097 55; Fax þ44(0)28 9097 42; email: [email protected]
Received 31 January 2008; revised accepted 15 May 2008
Abstract
Permeable reactive barriers (PRBs) are a proven technology for remediating contaminated groundwater, particularly on indus-
trial and mining sites. PRBs are a sustainable technology that can operate over a long time scale with low maintenance. Over the
past 10–15 years, there have been great strides in refining site characterisation techniques (i.e. geophysical techniques), develop-
ing/discovering reactive materials/sorbents (i.e. Fe0 filings), and the installation and design of PRBs (i.e. funnel-and-gate design)
which have increased the cost-effeciveness of this technology. Prior to installation, careful consideration of the ease of removal of
the PRB should be considered as part of the design. This is important as the PRB may eventually need to be decommissioned. PRBs
are a sustainable site specific remediation technology that has the great potential to work well as a part of a larger scale integrated
water resource management programme in developing countries.
Keywords: Permeable reactive barriers (PRBs); Reactive materials; Sorbents; Remediation; Site characterisation; PRB designs
1. Introduction
Permeable reactive barrier (PRB) technology has
been sucessful in remediating a variety of groundwater
contaminants including heavy metals [1], organics [2]
and radionuclides [1,3]. Most PRBs have been installed
on industrial, mining and agricultural sites around the
world [1–3]. PRBs use the natural hydraulic gradient
of the groundwater plume to move the contaminants
through the reactive zone giving it an advantage over
traditional pump-and-treat technologies by being more
cost effective and lower maintenance in the long-term
[3]. Over the past decade, much work has been done on
improving site characterisation techniques, developing
reactive materials/sorbents, and the installation and
design of PRBs. This work has increased the cost-
effeciveness of this technology making it a more viable
remediation option for developing countries.
Presented at the Water and Sanitation in International Development and Disaster Relief (WSIDDR) International
Workshop Edinburgh, Scotland, UK, 28–30 May 2008.
0011-9164/09/$– See front matter © 2009 Published by Elsevier B.V.doi:10.1016/j.desal.2008.05.075
Desalination 248 (2009) 352–359
Geophysical techniques such as magnetic and
ground probing radar are a form of non-invasive site
characterisation that yields valuable information about
site geology, in situ engineering properties, hidden cul-
tural features, and contamination in the shallow subsur-
face. Non-intrusive investigations are a quick and cost
effective means of obtaining data, especially when
combined with old site plans, and are useful in planning
the intrusive site investigations. These techniques are
reducing the need for more expensive trial pitting or
borehole drilling with lower risks and decreasing the
chances of missing buried targets. Some of the instru-
ments (EM units) are hand carried and generally do not
contact the ground. All accessible areas of a site can be
quickly surveyed (up to 2 ha/day) without disturbing
the surface [4]. This is important because there may
be considerable surface and near surface contamina-
tion on former industrial and mining sites. Comparing
site historical plans with their geophysical surveys is a
very beneficial ground truthing method and is gener-
ally part of the protocol of a study.
A variety of reactive materials and sorbents,
which can be used separately or in combination
depending on the groundwater contamination, have
been successful in remediating contaminated
groundwater in PRBs. These materials, such as Fe0
filings [4], peat [3], limestone [4,8], granular acti-
vated carbon (GAC) [5,6] and zeolite [1], are easily
available and some are fairly inexpensive. Bench-
scale treatability studies are carried-out in the initial
screening of the reactive or sorbent material to plan
the design of PRBs using site groundwater. Batch
studes using a number of likely reactive and sorbent
materials are conducted to determine the best per-
forming materials. Then column tests are carried-
out on the best performers. Column tests can give
information towards the design of the PRB and indi-
cations on how an in situ PRB will perform [7].
Installation of PRBs is a crucial stage, especially in
the excavation of geological material. Improved equip-
ment and techniques used to excavate geological mate-
rial without obstructing the flow of the contaminated
groundwater plume in and out of the PRB has helped
to increase the success rate of PRB performance. Dur-
ing installation, loose geological material and soil can
be packed, smeared and fill void space that the con-
taminated groundwater flows through adjacent to the
PRB [8,9]. Methods in depositing of the reactive mate-
rial in to the PRB have also been refined to reduce par-
ticle size grading which can alter groundwater flow
through the reactive zone [9]. Prior to installation, care-
ful consideration of the ease of removal of the PRB
should be considered as part of the design. This is
important as the PRB may eventually need to be
excavated due to the completion of remediation; there-
fore, decommissioning could become an issue [7].
Of the two basic designs, i.e. the continuous trench
and the funnel-and-gate, the funnel-and-gate design
with the reactive material placed in single or sequenced
containers is probably the most cost-effective design.
This is because the funnel-and-gate usually uses less
reactive material than the continuous trench. The reac-
tive material(s) is placed in the canister(s) (reactors)
and can be removed if it needs to be replaced. This is
an important consideration as there could eventually
be built-up of contaminant concentrations in the reac-
tive/sorbent material(s) from the remediation process.
The containers can be designed to be reused at other
sites once remediation is finished.
PRBs are a sustainable site specific remediation
technology that has the great potential to work well
as a part of a larger scale integrated water resource
management programe in developing countries. The
objectives of the paper are to illustrate how PRBs are
planned and installed, and highlight cost effectiveness
which may allow them to be installed in developing
countries.
2. Material and methods
If PRB technology is considered one of the
options for the remediation of contaminated ground-
water at a site, there are a series of steps that should
be taken to ensure that the PRB is viable, and that it is
designed and installed properly. One of the first steps
is to collect as much information about the site as
possible, such as blueprints, geological maps, and
records. This will help in ground truthing and deter-
mining contamination. What are the possible con-
taminants? Have there been any buildings on site?
Buried debris may imped PRB installation and
groundwater flow across the site. Knowing what
contaminants may be present at the site will also
allow for better health and safety plans.
D.H. Phillips / Desalination 248 (2009) 352–359 353
2.1. Site characterisation
Non-invasive characterisation of the site using geo-
physical techniques, especially when used along with
test pits and wells improves the detail of the site char-
acterisation and is cost effective and time saving. Gen-
erally, a site is first cleared of surface objects (i.e. scrap
metal, reinforced concrete rubble) that may interfere
with subsurface features. A grid (i.e. 5 m) is surveyed
out (i.e. using a Total Station Lecia 1010 Wild Heer-
brugg surveyor). Further points are measured at the site
boundaries and across the site to give a representation
of height A.O.D.
Electromagnetic (EM) and ground penetrating
radar (GPR) instruments are portable units that can
detect metal, debris, geological features, and contami-
nation in the shallow subsurface. A Geonics EM61
time domain metal detector is used to survey for fer-
rous and non-ferrous metal objects down to 5 m. The
instrument, operating at 75 Hz, is trawled across the
site by the operator. The Geonics EM31 can map geo-
logical variations and any other feature associated with
ground conductivity change down to 6 m. The EM31 is
in the form of a boom that is carried about 1 m above
ground. It can be used in a normal operating orientation
(vertical dipole mode) or turned 90� to its long axis
(horizontal dipole) and used to measure down bore-
holes. The Geonics EM38 also measures conductivity
and is carried 15–20 cm above the ground. It is particu-
larly sensitive to soil salinity which could be a charac-
teristic of contaminated land/groundwater. Similar to
the operation principle of the EM31, the EM38 surveys
to the depth of 1.5 m in the vertical dipole, and 0.75 m
in the horizontal dipole modes [10]. GPR is also used
for high resolution shallow subsurface investigations.
Ground truthing should be part of the protocol in a field
investigation by comparing the geophysical results to
historic plans and site visits.
2.2. Treatability studies
Batch tests are used to screen potential reactor mate-
rials. They can determine whether the contaminants are
amenable to sorption, degradation, or precipitation by
different types of media. They also can be used to com-
pare treatment efficiency of reactive and sorptive mate-
rials. Contaminant degradation rates are measured
quantitatively over time, and various parameters and
several sampling intervals can be tested. Batch tests are
carried-out in bottles or vials capped with inert septa. A
known weight of reactive material and a known concen-
tration and volume of contaminant solution (preferably
the contaminated groundwater) are added to the vials.
Samples are extracted at regular intervals and the con-
taminant concentration measured. Degradation curves
can be made from this data, and degradation products
can also be determined [7].
Column tests are used to collect detailed informa-
tion on the degradation, precipitates, removal and/or
sorption of contaminants by reactive/sorptive media.
Removal rate data under a range of flow conditions,
especially those that mimic groundwater flow velocity
in the field can be determined with column tests. Con-
taminated or simulated groundwater is passed through
the column at a known flow rate. The column design
allows changes in contaminant composition and other
parameters (e.g. major ions, pH) to be determined at
the influent, effluent and along the column length
(Fig. 1) [7]. Column dimensions are generally 10–
100 cm long, with a 2.5–3.8 cm inside diameter [8].
Glass columns are generally the least reactive or
adsorptive with chlorinated organic compounds. How-
ever, no significant loss of organics have been found
using Perspex columns [11]. Sampling ports along the
length of the column should be constructed of stainless
steel fittings or inert stoppers (i.e. Teflon coated or
Viton). The ports should allow the sampling needle
to be inserted into the centre axis of the column or a
needle to be fixed in place in each port. A three-way
port should also be positioned in the influent and efflu-
ent lines. All tubing and fitting for the influent and
effluent lines should be composed of an inert material.
Information from the column study can be used along
with the site characterisation and modeling to help to
design the field-scale PRB [7].
2.3. Excavation
In conventional excavation of continuous trenches,
the soil is removed and the trench is backfilled with
reactive/sobent materials. When the emplacement of
the reactive materials is completed, temporary retain-
ing structures which support the walls will be removed
from the ground. A backhoe comprised of a digging
bucket on the end of an articuated arm is frequently
used for rapid digging of shallow trenches [8] less that
354 D.H. Phillips / Desalination 248 (2009) 352–359
10 m deep. A continuous trenching machine can also
be used to create a narrow trench less than 7 m deep.
It is comprised of a chain saw cutting belt and a trench
box on the boom. The backfill of reactive/sorbent
materials is carred-out with a hopper on the top of the
trench box. The removal of soil and the backfilling of
reactive materials are simutanously carried-out. This
elminates the need for dewatering and temporary
retaining structures [11,12]. An excavation of up to
15 m depth can be carried-out with a caisson for
funnel-and-gate PRBs. Caissons with a circular
cross-section are frequently used. After it has been
installed, a large auger can be used to remove the soil
within the caisson. The reactive/sorbent materials can
be then backfilled into the caisson [11,12].
2.4. PRB design
The two common PRB designs are the continuous
trench PRB and the funnel-and-gate system (Fig. 2a
and b). The continuous trench PRB does not contain
any structures, so the contaminant plume flows through
the treatment zone using the natural hydraulic gradient.
This PRB, which is perpendicular to groundwater flow
direction, needs to be slightly larger than the cross-
sectional area of the contaminated groundwater inorder
to capture the contaminants in both vertical and hori-
zontal directions [9]. The top of the PRB should be at
least 0.60 m above the water table and the bottom of the
PRB should be extended at least 0.30 m into a low per-
meability zone (i.e. clay), if it is present. The PRB
thickness should be designed to provide sufficent resi-
dence time for the contaminants within the treatment
zone to be completely treated. The funnel-and-gate
system is composed of impermeable walls and at least
one reactive zone. The funnel structure could be sheet
piles or slurry walls. The function of the funnel is to
intercept the contaminated groundwater and lead it to
the treatment zone. The bottom of the funnel and reac-
tive zone needs to be extended at least 0.30 m into the
less permeable soil layer, while the top of the funnel-
and-gate needs to be set at least 0.60 m above the water
table [10]. The reactive material is directly implaced or
filled into the reaction vessel(s) [9]. Multi-sequenced
reactive barriers are also being installed, especially
on sites with multiple groundwater contaminants such
as gas works sites. Multi-sequenced PRBs use muliti-
ple reactive materials in more than one reactive zone
to treat the contaminated groundwater [13] (Fig. 2c).
3. Results and discussion
3.1. Cost-effective groundwater clean-up
with PRBs
Over the past 10–15 years since the first PRBs were
installed to remediate contaminated groundwater, there
Fig. 1. Diagram and photograph of a column test.
D.H. Phillips / Desalination 248 (2009) 352–359 355
have been great strides in reducing their costs. Com-
pared to traditional pump-and-treat remediation, PRBs
can be more costly in the initial stages, especially dur-
ing installation. However, since PRBs are a passive
system, which relies on the natural hydraulic gradient
of the groundwater plume to move the contaminated
groundwater through the treatment zone, the long-
term costs are lower than traditional pump-and-treat
operations and maintenance [14]. Many PRB reac-
tive/sorbent materials can remediate a range of con-
taminates, while others are selective. PRBs can be
used as part of a treatment train with other technologies
to clean groundwater. PRBs are generally below
ground and ‘out of sight’ so the site can be used for
other uses. However, there are some disadvantages of
PRBs. Sometimes the PRB require more maintenance
than originally planned. PRBs can also become
clogged and the reactive material coated causing them
to become less effective.
Non-invasive site characterisation has also
decreased the cost of PRB installation. Many of the
magnetic and ground probing radar units are portable
and can easily be transported into sites that may be
remote. Data on the characteristics of a contaminated
site can be gathered rapidly and can help in planning
the invasive characterisation (trial pitting and bore-
holes). Data collected on buried features, geology and
hydrogeology can be used to model or better plan for
Fig. 2. Diagrams of PRBs. (A) Elevation view of a continuous trench or wall (after [14]), (B) plain view of a funnel-and-gate,
and (C) elevation view of a multi-barrier.
356 D.H. Phillips / Desalination 248 (2009) 352–359
monitoring and remediation of contaminated land and
groundwater. The suite of chemicals contributing to
subsurface groundwater contamination may be electri-
cally conductive, either because the contamination is
acidic or contains salt. Therefore, the groundwater
plume is also electrically conductive [4]. Importantly,
the non-invasive site characterisation can also give
indications if the site is not geologically and hydrogeo-
logically fit for a PRB to be installed (i.e. absence of
low permeability material to secure the bottom of the
PRB), which can reduce money and time spent on a
remediation option that may fail.
Treatability studies may be carried-out at a number
of stages of PRB selection including remediation
options screening, PRB design and operation [7]. Treat-
ability studies are used to determine the best reactive/
sorbent media, to help design the PRB, and to predict
weaknesses in the PRB design. The cost of treatability
studies can vary greatly, and should be proportionate to
the total cost of remediation. Generally, treatability
studies should cost <10% of the total remediation costs.
To cut costs, reactive material from local sources can be
screened in a treatability study. For example, activated
carbon from coconut husk is used to treat arsenic
contaminated groundwater [6]. Other media such as
Fe0 [3], sawdust [1], plant material [1], and zeolite from
volcanic rock [1] are also used in PRBs as well as
microbes [1] in biobarriers. Using local sources may
save money on purchasing the media and also renewing
it. Additionally, the use of recycled materials makes
the PRB a sustainable remediation option.
The conventional excavation is a common PRB
construction technique and is more suitable for shallow
PRB systems such as a continuous trench <10 m deep.
Operational cost can increase with the depth of excava-
tion. The conventional excavation can be carried-out
with common excavation equipment such as a back-
hoe, a clamshell and a caisson [12]. Because the trench
is located in shallow strata, the whole installation of the
PRB can be easily monitored. Some potential problems
can occur with the conventional excavation. In addition
to a decrease in the permeability of the PRB [11] from
smearing, contaminated soil and water can potentially
be exposed during the excavation raising health and
safety issues. Nevertheless, the operational costs and
time are low for continuous trenching [12]. Although,
the funnel-and-gate PRB is physically easier to
decommission, the continuous trench PRBs can still
remain an option if the geology and hydrogeology is
favorable for installation, and if there is a lack of funds
and resources to install a funnel-and-gate PRB. In the
funnel-and-gate PRB, reactive material is stored in
canisters or reactors which can be removed after the
PRB has completed the remediation or if the PRB has
problems. This allows for removing of potentially
further contamination from the site as a result of con-
taminant build-up in the reactive material. Addition-
ally, funnel-and-gate PRBs generally require less
reactive material. This is a great cost saver, if the reac-
tive material is pricey. For these reasons, the funnel-
and-gate design should be considered if funds are
available. Therefore, it is important to take the design
of the PRB in to consideration and how this will affect
its long-term operation.
3.2. Case studies of PRBs
There are over 100 PRBs in operation around the
world at present; however, only a few are installed in
developing countries. Examples are pilot-scale PRBs
used to treat contaminated groundwater and leachate
from uranium mines in Hungary and Bulgaria, respec-
tively. The PRB in Pecs Hungary which was installed
in 2002, is a continuous trench containing shredded
Fe0 which removes uranium from the groundwater.
Sand layers have been added to the up-gradient and
down-gradient sides of the PRB to allow for better
groundwater flow through the reactive zone. The PRB
was emplaced in an underlying clay and geosynthetic
clay liner. Groundwater U concentrations were
reduced to <1% of influent concentrations after passing
through the reactive zone in year 2003. Uranium con-
centrations were reduced from *1000 mg/L to *100
mg/L in monitoring wells near the PRB and <10 mg/L
within the PRB. A negative performance issue is that
a high amount of precipitates has been estimated to
have formed in the PRB which may reduce ground-
water flow through the PRB and reduce the reactivity
of the Fe0. However, only a 1.6% loss in porosity was
calculated and the PRB is predicted to have a 62 year
lifespan [3]. The PRB in Western Bulgaria (installed
2004) treats acid drainage (pH 2.5–3.7) runoff from a
uranium mine contaminated with radionuclides
(mainly uranium and radium), heavy metals, arsenic
D.H. Phillips / Desalination 248 (2009) 352–359 357
and sulphates. The PRB is part of a treatment train
(connected series) in the form of an alkalizing lime-
stone drain that removes Fe as hydroxides, the PRB for
biosorption and microbial reduced sulphates, and a nat-
ural wetland. The barrier is a ditch/continuous trench
with a reactive material consisting of a mixture of solid
biodegradable organic material (plant and mushroom
compost, wood chips, straw and cow manure), crushed
limestone and zeolite saturated with ammonium phos-
phate. A mixed microbial community consisting of
sulphate-reducing bacteria and other microorganisms
is also present. Uranium, arsenic and non-ferrous
metals were mainly removed in the PRB by the indi-
genous sulphate-reducing bacteria. Portions of the pol-
lutants, and most of the radium, were sorbed onto the
dead plant material in the PRB [1].
Another example of a PRB, but not in a developing
country, is a funnel-and-gate PRB in Monkstown,
Northern Ireland, United Kingdom. This PRB, estab-
lished in 1995, is a field-scale PRB developed to
remediate a small but highly concentrated accumula-
tion of a degreaser pollutant, trichlorethene (TCE), in
gravely glacial till on an industrial site. TCE is a dense
nonaqueous phase liquid (DNAPL) that is denser than
water and has a tendency to settle in the subsurface as
immiscible accumulations (slugs). The PRB consists of
a long tube-like structure that holds Fe0 filing reactive
material. This structure is emplaced in naturally occur-
ring clay at the site. The TCE contaminated ground-
water is funneled from the up-gradient position into
the gate of the PRB where the vessel that holds the
reactive material is present. The groundwater flows
down through the Fe0 reactive zone in the PRB where
it is treated and exits at the down-gradient. There is a
decrease in up-gradient TCE concentrations in this
Fe0 funnel-and-gate PRB over time which suggest that
Fig. 3. PRBs as part of a larger scale integrated water resource management programme in developing countries.
358 D.H. Phillips / Desalination 248 (2009) 352–359
the PRB is remediating the TCE contaminated ground-
water at the site. However, there is also a possibility
that a great portion of the TCE slug was removed while
installing the PRB which would cause the concentra-
tions to be lower over time (Fig. 3). There is decrease
in TCE as it moves through the PRB indicating that
it is being remediated (dehalogenated) by the Fe0. A
pre-existing TCE contamination occurs adjacent to the
down-gradient portion of the PRB possibly due the
clipping of the end of the TCE contamination slug
when the PRB was being installed which hampers the
down-gradient monitoring of this PRB; however, data
shows that it is dramatically decreasing over time [2].
When PRBs were first being inplemented about 15
years ago, it was suggested that they become a reme-
diation option for developing countries. This is because
they are a sustainable remediation technology that
requires little maintenance and operational cost over
a long duration. Addtionally, reactive/sorbent materi-
als such as peat, sawdust, activated carbon, zeolites,
and limestone can be aquired easily and may be from
a local source. Generally, PRBs are used for site speci-
fic remediation. However, in areas where adequate
water supply is limited and there is contamination from
industrial, agricultural or mining sites, if properly
planned they could become part of a larger scale inte-
grated water resources management programme. For
example, an International Water and Sanitation Center
(IRC) project DREAM (Drainage and Reuse of Efflu-
ents for Agricultural Management) has planned PRBs
to be part of an integrated water resources management
programme as a low cost approach for wastewater
treatment for reuse in agricutural areas in developing
countries where water supply is low [15].
References
[1] S. Groudev, P. Geprgiev, I. Spasova, M. Nicolova, A.
Angelov and L. Diels, Cleanup of acid mine drainage
by means of a pilot-scale passive system, Annu. Univ.
Min. Geol. ‘St. Ivan Rilski’ Part. 1. Geol. Geophys.,
48 (2005) 217-220.
[2] P. Beck, N. Harris and R. Sweeney, Design, Installation
and Performance Assessment of a Zero Valent Iron
Permeable Reactive Barrier in Monkstown, Northern
Ireland, CL:AIRE Technology Demonstration Report,
TDP3, 2001, 120pp.
[3] K.E. Roehl, K. Czurda, T. Meggyes, F. Simon and
D.I. Stewart, Long-term Performance of Permeable
Reactive Barriers, Elsevier, New York, 2005.
[4] US Army Corps of Engineers, Geophysical Exploration
for Engineering and Environmental Investigations, EM
1110-1-1802, 1995.
[5] V. Plagentz, M. Ebert and A. Dahmke, Remediation of
ground water containing chlorinated and brominated
hydrocarbons, benzene and chromate by sequential
treatment using ZVI and GAC, Environ. Geol., 49
(2006) 684-695.
[6] G.N. Manju, C. Raji and T.S. Anirudhan, Evaluation of
coconut husk carbon for the removal of arsenic from
water, Water Res., 32 (1998) 3062-3138.
[7] Environment Agency, Guidance on the Use of Perme-
able Reactive Barriers for Remediating Contaminated
Groundwater, National Groundwater and Contami-
nated Land Centre Report NOC/01/51, Environment
Agency, Solihull, UK, 2002.
[8] United States Environmental Protection Agency,
Permeable Reactive Barrier Technologies for Contami-
nant Remediation, EPA 600-R-98-125, 1998.
[9] A. Gavaskar, N. Gupta, B. Sass, R. Janosy and J. Hicks,
Design Guidelines for Applications of Permeable
Reactive Barriers for Groundwater Remediation, Bat-
telle, Columbus, OH, Report No. AFRL-ML-WP-TR-
2000-4546, NTIS: ADA380005, 2000, 399pp.
[10] US Department of Transportation, Determination of the
Depth/Structure/Fracture of Bedrock, 2005, http://
www.cflhd.gov/agm/engApplications/Index.htm.
[11] A. Gavaskar, N. Gupta, B. Sass, R. Janosy and D.
O’Sullivan, Permeable Barriers for Groundwater
Remediation: Design Construction and Monitoring,
Battelle, Columbus, Ohio, 1998.
[12] US Air Force, Design Guidance for Application
of Permeable Barriers to Remediate Dissolved
Chlorinated Solvents, DG 1110-345-117, 1997.
[13] L. Dries Bastiaens, J. Vos, Q. Simons, M. De Smet and
L. Diels, Comparison of different multibarrier concepts
designed for treatment of groundwater containing
mixed pollutants, in: A. Boshoff, B. Bone, eds., Perme-
able Reactive Barriers, International Symposium on
Permeable Reactive Barriers, March 14, 2004, Queens
University Belfast Northern Ireland, vol. 298, 2005, pp.
45-51.
[14] Environmental Protection Agency, A Citizen’s Guide
to Permeable Reactive Barriers, EPA 542-F-01-005,
2001, 2pp.
[15] International Water and Sanitation Centre, Annual
Report, The Netherlands, 2004, 24pp.
D.H. Phillips / Desalination 248 (2009) 352–359 359