You are on page 1of 11

Water Air Soil Pollut (2014) 225:2009

DOI 10.1007/s11270-014-2009-7

Biogeochemical Process-Based Design and Performance


of a Pilot-Scale Constructed Wetland for Arsenic Removal
from Simulated Bangladesh Groundwater
Jeffrey P. Schwindaman & James W. Castle &
John H. Rodgers Jr.

Received: 11 February 2014 / Accepted: 14 May 2014 / Published online: 29 May 2014
# Springer International Publishing Switzerland 2014

Abstract A pilot-scale constructed wetland treatment


system (CWTS) was designed and built to produce biogeochemical conditions promoting processes targeted for
removal of arsenic from simulated Bangladesh groundwater. Two CWTS series were designed to promote
coprecipitation and sorption of arsenic with iron
oxyhydroxides under oxidizing conditions, and two series were designed to promote precipitation of arsenic
sulfide and coprecipitation of arsenic with iron sulfide
under reducing conditions. Arsenic removal performance
was greater in series with oxidizing conditions than in
series with reducing conditions (mean outflow concentrations of 64 and 108 g L1, respectively). Additions of
zero-valent iron (ZVI) to oxidizing series and to reducing
series enhanced arsenic removal (mean removal efficiencies of 72 and 42 %, respectively) compared to unamended series (27 and 20 %, respectively). Arsenic removal
performance was significantly greater (=0.05) in the
oxidizing series amended with ZVI than in the other
series, with removal extents, efficiencies, and rate coefficients ranging from 6 to 79 g L1, 40 to 95 %, and 0.13
to 0.77 day1, respectively. Results from this pilot-scale
study demonstrate that a CWTS can decrease
J. P. Schwindaman : J. W. Castle (*)
Department of Environmental Engineering and Earth
Sciences, Clemson University,
Clemson, SC 29634, USA
e-mail: jcastle@clemson.edu
J. H. Rodgers Jr.
School of Agricultural, Forest, and Environmental Sciences,
Clemson University,
Clemson, SC 29634, USA

concentrations of arsenic in arsenic-contaminated water


to below the World Health Organization (WHO) drinking
water quality guideline of 10 g L1.
Keywords Constructed wetland . Arsenic . Bangladesh .
Zero-valent iron . Groundwater . Treatment

1 Introduction
The greatest extent of arsenic contamination in the world is
in the Bengal Basin, where approximately 1.5 million
groundwater wells, used for drinking water and irrigation,
contain arsenic concentrations in excess of 50 g L1,
affecting approximately 35 million people in Bangladesh
(Kinniburgh et al. 2002; Smedley and Kinniburgh 2002).
Chronic exposure to arsenic concentrations greater than
10 g L1 can lead to negative health effects, including
cardiovascular, pulmonary, nervous, and endocrine system
disorders; skin lesions; and cancer (Hughes et al. 2009;
WHO 2011). A low-cost, low-maintenance approach is
needed to decrease arsenic concentrations in drinking water. Several strategies exist for removal of arsenic from
water, including precipitation, coprecipitation, sorption,
ion exchange, and membrane filtration (Bahar et al.
2013; Meng et al. 2001; USEPA 2002); however, these
strategies are usually cost-prohibitive. Constructed wetland
systems for treatment of arsenic-contaminated water may
offer a low-cost, low-maintenance alternative. Constructed
wetland treatment systems (CWTSs) target specific biogeochemical processes (e.g., sorption, precipitation,
coprecipitation, and volatilization) to remove constituents

2009, Page 2 of 11

of concern (COCs) from the aqueous phase or transform


COCs into less bioavailable forms (Dorman et al. 2009;
Rodgers and Castle 2008).
Arsenic is present naturally in a variety of geologic
media but is typically associated with sulfide mineral
deposits or bound to iron oxyhydroxides (Henke 2009;
Rahman et al. 2006). Arsenic commonly exists in natural waters as arsenite [As(III)] or arsenate [As(V)]
(Francesconi and Kuehneit 2002). Arsenite is the predominant species of arsenic in groundwater and is more
toxic to humans than arsenate (Sharma and Sohn 2009).
Under reducing conditions and near-neutral pH, arsenite
most often occurs in groundwater as H3AsO3. Arsenite
can bind with sulfide under reducing conditions to form
the insoluble minerals orpiment (As2S3), realgar (As4S4), and arsenopyrite (FeAsS) (Cheng et al. 2009;
Jackson et al. 2013; Lizama et al. 2011). Under oxidizing conditions and near-neutral pH, arsenate most often
occurs as H 2 AsO 4 and HAsO 4 2 (Henke and
Hutchison 2009). As an oxyanion, arsenate can be involved in electrostatic interactions with constituents in
soil and sediment, such as iron oxyhydroxides. Arsenic
can enter groundwater both by oxidation of sulfide
minerals and by reductive dissolution of iron
oxyhydroxides (Bhattacharya et al. 1997; McArthur
et al. 2004; Nickson et al. 1998, 2000, 2005).
Therefore, the hypothesis investigated here is that arsenic can be removed from water in a CWTS by promoting reducing conditions that result in precipitation of
sulfide minerals and by promoting oxidizing conditions
that cause coprecipitation of arsenic with iron
oxyhydroxides. Two CWTS series were designed to
promote reducing conditions (reducing series), and
two series were designed to promote oxidizing conditions (oxidizing series). Zero-valent iron (ZVI) was
added to one oxidizing series and one reducing series.
The objectives of this study were to (1) construct a pilotscale CWTS to compare arsenic removal in the reducing
series with arsenic removal in the oxidizing series and
(2) assess arsenic removal performance in each series by
determining removal extents, efficiencies, and rates.

2 Materials and Methods


2.1 Design and Construction of Pilot-Scale CWTS
Targeted biogeochemical processes for arsenic removal
and conditions that promote these processes were

Water Air Soil Pollut (2014) 225:2009

determined from a review of published literature. The


pilot-scale CWTS was specifically designed and constructed to produce these conditions. Processes identified for removal of arsenic included sorption and
coprecipitation of arsenic with iron oxyhydroxides, precipitation of arsenic sulfide, and coprecipitation of arsenic with iron sulfide. Sorption and coprecipitation of
arsenic with iron oxyhydroxides occur under oxidizing
conditions (sediment oxidation-reduction (redox) potential > 50 mV), high dissolved oxygen (DO) concentration (2 mg L1), and near-neutral pH (49), whereas
precipitation of arsenic sulfide and coprecipitation of
arsenic with iron sulfide occur under reducing conditions (sediment redox potential from 50 to 250 mV),
low DO concentration (2 mg L1), and near-neutral pH
(58) (Buddhawong et al. 2005; Cheng et al. 2009;
Huang 2014; Lizama et al. 2011; Rahman et al. 2011).
Four pilot-scale CWTS series, specifically designed
to promote identified processes for arsenic removal,
were built and operated for 210 days (August 17, 2012
to March 15, 2013) in a climate-controlled greenhouse
in Clemson, SC, USA (Fig. 1). Two series were designed to promote oxidizing conditions, and two series
were designed to promote reducing conditions. Each
series consisted of four treatment reactors, with each
reactor contained in a 265-L (70 gal) Rubbermaid
utility tank (92 cm long by 74 cm wide by 61 cm deep).
Each reactor, with the exception of the third reactor in
each of the two oxidizing series, contained a 30-cm
thickness of river sand from 18-Mile Creek in
Clemson, SC, water to a depth of 25 cm, and approximately 20 Typha latifolia plants harvested from an
aquaculture pond in Clemson, SC (Fig. 2). The river
sand, which was characterized by Kanagy et al. (2008),
consisted of 12 % organic matter, 13 % clay and silt,
and the remainder quartz (SiO2) sand.
The third reactor in each of the two oxidizing series
contained a 50-cm thickness of approximately 3-cmdiameter granitic gravel and water to a depth of 5 cm,
and was unvegetated. In the two reducing series, the
upper 5 cm of sediment in each reactor was amended
with gypsum pellets (1 %v/v) as a source of sulfate for
dissimilatory sulfate reduction and with hay and shredded hardwood mulch (5 %v/v) to provide carbon and
nutrient sources for sulfate-reducing bacteria (SRB).
Water was transferred from a 5,678-L (1,500 gal) polypropylene mixing tank into the first reactor of each
series by a piston pump (FMI QG400) at a flow rate
of 128 mL min1, resulting in a nominal hydraulic

Water Air Soil Pollut (2014) 225:2009

Page 3 of 11, 2009

Fig. 1 Schematic diagram of


pilot-scale constructed wetlands
for treatment of arseniccontaminated water. Series 1 and
2 were designed to promote
oxidizing conditions, and series 3
and 4 were designed to promote
reducing conditions. Series 1 and
4 were amended biweekly with
20 g ZVI per reactor

retention time (HRT) of 24 h per reactor or 96 h per


series. Reactors were connected by PVC pipe fittings
located at 4 cm below the top of each reactor. Reactors

Fig. 2 Schematic diagram of a CWTS reactor. Each reactor, with


the exception of the third reactor in each of the two oxidizing
series, contained a 30-cm thickness of sand, water to a depth of
25 cm, and approximately 20 Typha latifolia plants. The third
reactor in each of the two oxidizing series contained a 50-cm
thickness of gravel and water to a depth of 5 cm and was
unvegetated. Each reactor contained one platinum-tipped electrode
in the sediment near the inflow area and one near the outflow area.
Each platinum-tipped electrode was placed at a depth of 2 cm and
remained in situ during the experiment

were arranged with decreasing elevation from the first to


fourth reactor in each series to induce gravity flow.
In bench-scale batch equilibrium experiments, addition of ZVI to oxidizing environments removed >90 %
of total arsenic from water due to rapid coprecipitation
with iron oxyhydroxides (Farrell et al. 2001; Kanel et al.
2005; Li et al. 2011; Manning et al. 2002; Su 2007),
while addition of ZVI to reducing environments promoted dissimilatory sulfate reduction (Karri et al. 2005).
Therefore, to investigate the effects of ZVI on performance of a CWTS, one oxidizing and one reducing
series were amended with ZVI (Peerless Metal
Powders & Abrasive, Detroit, MI) by distributing 20 g
per reactor into the water column approximately every
14 days.
Water for the CWTS was simulated based on composition of actual arsenic-contaminated groundwater in
the Lakshmipur District, southern Bangladesh (Table 1)
(Kinniburgh and Smedley 2001). Use of simulated water provides greater experimental control over water
characteristics, increases repeatability of results, and
eliminates the economic burden of acquiring, shipping,
and storing large volumes of actual groundwater (Alley
et al. 2013). Concentrations of chemical constituents in
groundwater of the Lakshmipur District were within the
range in groundwater of other districts in central and
southern Bangladesh and were therefore considered
representative of arsenic-contaminated groundwater
from the shallow Holocene aquifer, which is the
source of most drinking water in Bangladesh
(Kinniburgh and Smedley 2001).

2009, Page 4 of 11

Simulated Bangladesh groundwater was formulated


by mixing chemical constituents with municipal water
from Clemson, SC, in the polypropylene tank. Arsenic
was added as arsenic(III) oxide (As2O3 99 %) at
236 g L1 and arsenic(V) oxide (As2O5 99 %) at
518 g L1. Major cations and anions (Ca2+, Mg2+, K+,
Na+, Cl, HCO3, PO42, SO42, SiO32) were added as
calcium chloride (CaCl2) at 140 mg L1, magnesium
chloride (MgCl2) at 170 mg L1, potassium chloride
(KCl) at 24 mg L1, sodium bicarbonate (NaHCO3) at
two concentrations, magnesium sulfate (MgSO4) at
15 mg L1, iron(II) sulfate (FeSO4) at 15 mg L1, sodium metasilicate (Na2SiO3) at 60 mg L1, and potassium
phosphate (K2PO4) at 1.4 mg L1. Arsenic removal in
the CWTS was investigated at bicarbonate concentrations of ~300 and ~30 mg L1.
2.2 Measurement of CWTS Conditions
and Performance
DO concentration, pH, and sediment redox potential were
measured biweekly in each reactor (Fig. 1). DO concentration and pH were measured using YSI (model 55)
and Orion (model A325) field instruments, respectively.
Alkalinity was measured using standard methods (APHA
2005) and converted to bicarbonate concentration using
the carbonate-speciation method (Rounds 2006). To measure sediment redox potential, one platinum-tipped electrode was placed at a sediment depth of 2 cm in the inflow
area and one in the outflow area of each reactor (Fig. 2).
Electrodes remained in situ for the duration of the experiment. Sediment redox potential was measured using a
GDT-11 Multimeter connected to in situ electrodes and
an Accumet calomel reference electrode (Faulkner et al.
1989). Statistical differences were determined by
ANOVA and post hoc t tests with =0.05.
One sample from the upper 2 cm of sediment was
collected from near the inflow, one from the middle, and
one from near the outflow of the first reactor of each
series on day 1 of the experiment (August 31, 2012), day
80 (November 19, 2012), day 126 (January 4, 2013),
and day 174 (February 21, 2013). Samples were
scooped with an acid-washed metal spatula into a
50-mL centrifuge tube and sealed underwater. The three
samples from each reactor were mixed and analyzed for
acid volatile sulfide (AVS) using a modified diffusion
method and ion-selective electrode (Leonard et al.
1996). AVS is defined as the sulfide extracted from
sediment by 1 N HCl (Di Toro et al. 1992; Leonard

Water Air Soil Pollut (2014) 225:2009

et al. 1996) and interpreted as the reactive fraction of


sulfide available to bind metals (Di Toro et al. 1992;
Keon et al. 2001; Wilkin and Ford 2002). External
standards for AVS analysis, ranging from 104 to
101 M sulfide, were created by serially diluting a
101 M sulfide stock solution. A four-point calibration
curve, with R2 >0.997, was produced for each AVS run
to ensure precision. Detection limit was 0.1 mol g1.
Aqueous samples were collected biweekly over 14
sampling periods (August 31, 2012 to March 15, 2013)
from each treatment reactor sequentially according to
the HRT (24 h per reactor). For example, samples of
inflow water entering the first reactor were collected at
t=0 h, and samples of outflow water from the first
reactor were collected at t=24 h, from the second reactor
at t=48 h, from the third reactor at t=72 h, and from the
fourth reactor at t=96 h. In this way, a single volume of
water was theoretically sampled as it passed through a
series. Samples were collected in acid-washed 50-mL
polypropylene centrifuge tubes from sampling ports
between each reactor. Twenty-five milliliters of sample
was centrifuged for 20 min at 8,000 rpm. One milliliter
of supernatant was transferred gravimetrically to an
acid-washed 15-mL polypropylene centrifuge tube and
brought to a volume of 10 mL with 2 % trace metal
grade nitric acid solution. Samples were analyzed for
arsenic using inductively coupled plasma mass spectrometry (ICP-MS) (Thermo Scientific X Series) according to EPA Method 200.8 (USEPA 1994) with a
detection limit of 0.01 g L1. External standards, ranging from 0.01 to 100 g L1, were created by serially
diluting a 100-g L1 arsenic stock solution. A ninepoint calibration curve with R2 >0.99 was produced for
each ICP-MS run to ensure precision. Gallium and
yttrium internal standard recoveries were within 80 to
120 % according to instrument quality assurance/quality
control protocol.
Treatment performance of each series was assessed in
terms of removal extent, removal efficiency, and removal rate coefficient. Removal extent is defined as concentration of arsenic in the outflow (g L1). Removal
efficiencies were calculated using Eq. (1):
removal efficiency %

C 0 C
 100
C 0

where [C]0 is the inflow concentration (g L1) and [C]


is the outflow concentration (g L1). Removal rate
coefficient (k, day1) was calculated using Eq. (2):

Water Air Soil Pollut (2014) 225:2009

ln C =C 0
HRT

Page 5 of 11, 2009


2

where HRT is the nominal hydraulic retention time (i.e.,


the time between sampling inflow and outflow). Firstorder kinetics, assumed in Eq. (2), is a standard approach for calculating treatment rates in constructed
wetland systems (e.g., Crites 1994; Johnson et al.
2008; Jou et al. 2008; Kadlec 1997; Kadlec and
Wallace 2009; Knight et al. 1999; Lizama et al. 2011;
Reed and Brown 1995; Siracusa and La Rosa 2006;
Wong et al. 2006; Wood 1995). By comparing different
rate models, Rousseau et al. (2004) confirmed the applicability and utility of first-order kinetic models to
treatment in constructed wetland systems.

3 Results
Sediment redox potential was within the targeted range (>
50 mV for oxidizing series, 50 to 250 mV for
reducing series) in 63, 80, 55, and 64 % of measurements
for series 1, 2, 3, and 4, respectively (n=56). Mean
Table 1 Chemical characteristics of actual Bangladesh groundwater and simulated Bangladesh groundwater (units are mg L1
except as noted)
Constituent

Bangladesh
groundwatera

Simulated Bangladesh
groundwaterb

AsTotal

0.16

0.30

As(III)

0.09

0.18

As(V)

0.07

0.12

Ca

55

55

Mg

50

50

Na

180

180

13

13

HCO3

220

30300

Cl

220

220

SO4

18

18

Si

15

15

1.2

1.2

Fe

3.0

3.0

pH (S.U.)

7.1

7.1

Conductivity (S/cm)

890

890

Values from shallow depth (<150 m) wells in Lakshmipur District (n=59) (Kinniburgh and Smedley 2001)

Targeted values

sediment redox potential was lower in reducing series 3


and 4 (77 and 79 mV, respectively) compared to
oxidizing series 1 and 2 (45 and 51 mV, respectively).
Dissolved oxygen concentration was within the targeted
range (2 mg L1 for oxidizing series, 2 mg L1 for
reducing series) in 100, 98, 70, and 66 % of measurements for series 1, 2, 3, and 4, respectively (n=56). Mean
DO concentration was less in reducing series 3 and 4 (2.0
and 1.8 mg L1, respectively), compared to oxidizing
series 1 and 2 (6.6 and 7.4 mg L1, respectively).
Measured pH was within the targeted range (49 for
oxidizing series, 58 for reducing series) in 93, 82, 89,
and 93 % of samples for series 1, 2, 3, and 4, respectively.
In oxidizing series 1 and 2, mean sediment redox
potential (344 and 227 mV, respectively) and mean DO
concentration (9.3 and 11.5 mg L1, respectively) were
significantly greater (=0.05) in the third oxidizing reactor than in any other reactor in the series (range from 78
to 38 mVand 5.6 to 6.3 mg L1). pH values in the planted
reactors of series 1 and 2 (range of 6.98.4 and 6.98.7,
respectively) were less than in the unplanted third reactor
of these series (7.19.6 and 7.79.9, respectively).
Throughout the experiment, AVS concentrations
were greater by approximately two orders of magnitude
in reducing series 3 and 4 (27.5199.9 mol g1) than in
oxidizing series 1 and 2 (<0.12.2 mol g1) (Fig. 3).
AVS concentrations in the reducing series amended with
ZVI (31.6199.9 mol g1) were greater than in the
unamended reducing series (27.564.8 mol g1).
Concentrations of arsenic decreased between inflows
and outflows in the pilot-scale CWTS with very few
exceptions (Fig. 4). Arsenic removal performance was
significantly greater (=0.05) in the oxidizing series
amended with ZVI than in any other series, with removal extents and efficiencies ranging from 6 to 79 g L1
and 40 to 95 %, respectively, compared to 49
162 g L1 and 079 %, respectively, in the oxidizing
series not amended with ZVI (Table 2). Addition of ZVI
significantly improved the removal extent and efficiency of oxidizing series (p=1.7106 and 1.2107, respectively) and reducing series (p=3.9103 and 1.7
102, respectively). Outflow concentrations of arsenic
were less than the World Health Organization (WHO)
drinking water quality guideline of 10 g L1 in the
oxidizing series amended with ZVI during the October
28, 2012; November 13, 2012; and November 29, 2012
sampling periods. Mean removal extent and removal
efficiency were 89 g L1 and 42 %, respectively, in
the reducing series amended with ZVI compared to

2009, Page 6 of 11

Water Air Soil Pollut (2014) 225:2009

128 g L1 and 20 %, respectively, in the reducing


series not amended with ZVI. The removal efficiency
of series 1, 2, 3, and 4 ranged from 70 to 95, 20 to 49, 3
to 36, and 31 to 69 %, respectively, during the period of
low sodium bicarbonate loading (mean bicarbonate concentration ranged from 39 to 44 mg/L), compared to
removal efficiencies of 4489, 079, 076, and 054 %,
respectively, during periods of high sodium bicarbonate
loading (mean bicarbonate concentration ranged from
278 to 294 mg/L). Removal rate coefficients in oxidizing series 1 and 2, ranging from 0.13 to 0.77 and 0.00 to
0.39 day1, respectively, were greater than removal rate
coefficients in reducing series 3 and 4, ranging from
0.00 to 0.36 and 0.00 to 0.29 day1, respectively.

4 Discussion
Results from this pilot-scale study demonstrated that a
CWTS can be designed and effectively utilized to produce conditions that promote targeted processes for
removal of arsenic from water. Conditions (i.e., redox
potential and DO concentration) were favorable for
sorption and coprecipitation of arsenic with iron
oxyhydroxides in the series designed to be oxidizing
(series 1 and 2) and favorable for dissimilatory sulfate
reduction in the series designed to be reducing (3 and 4).
Arsenic removal performance was maintained when
conditions were temporarily beyond the targeted ranges
identified from published literature, indicating that once

AVS concentration (mol g -1)

250.0
Series1
Series2
Series3
Series4

200.0
150.0
100.0
50.0
0.0
8/31/2012

11/19/2012
1/4/2013
Sampling date

2/21/2013

Fig. 3 AVS concentrations (mol per gram dry sediment) measured in the first reactor of each series. AVS concentrations in the
reducing series 3 and 4 were approximately two orders of magnitude greater than those in the oxidizing series 1 and 2. AVS
concentrations in the reducing series amended with ZVI (series
4) were greater than in the unamended series (series 3) during the
November 19, 2012; January 4, 2013; and February 21, 2013
sampling periods. Error bars represent standard deviation of three
replicates

established, the targeted processes for removal of arsenic from water were not inhibited by natural variations
in biogeochemical conditions. Greater AVS concentrations in the reducing series than in the oxidizing series
are attributed to dissimilatory sulfate reduction and to
gypsum and organic matter amendments to the sediment. Greater AVS concentrations in the reducing series
amended with ZVI than in the unamended reducing
series are attributed to ZVI acting as an electron donor
for dissimilatory sulfate reduction (Karri et al. 2005).
Lower mean DO concentrations and lower sediment
redox potential in the reducing series compared to the
oxidizing series are interpreted to be the result of oxygen
consumption by aerobic microorganisms during biodegradation of organic matter in the reducing series. In the
two oxidizing series, greater mean DO concentrations
and sediment redox potential in the third reactor of each
series compared to the other reactors are attributed to
lower organic matter content in substrate of the third
reactor (unplanted granitic gravel) compared to other
reactors (river sand planted with T. latifolia) and to
greater atmospheric oxygen diffusion as the result of
shallower water depth in the third reactor (5 cm) than in
the other reactors (25 cm). Lower pH values in the
planted reactors of series 1 and 2 than in the unplanted
third reactor of these series may be the result of organic
acids in the planted reactors.
In this study, arsenic removal under oxidizing conditions was more effective than arsenic removal under
reducing conditions (mean removal extent of 64 and
108 g L1, respectively). Addition of ZVI significantly
improved the performance of both oxidizing and reducing series. First-order arsenic removal rate coefficients
in the oxidizing series amended with ZVI were consistent with rapid removal associated with coprecipitation
of arsenic with iron oxyhydroxides and sorption of
arsenic observed in previous bench-scale batch reactor
studies (Bang et al. 2005; Lackovic et al. 2000; Su and
Puls 2001a, 2001b). Results of this research indicate that
ZVI is an important addition to a CWTS designed to
treat arsenic-contaminated groundwater. ZVI in the
form of cast iron filings is a waste product of many
industrial processes and therefore can typically be obtained at a low cost (Bang et al. 2005).
Dissolved anions such as bicarbonate, phosphate,
sulfate, and silicate can negatively affect sorption and
ion exchange processes by competing with arsenic for
binding sites (Henke 2009; Leupin et al. 2005;
Stollenwerk 2002; USEPA 2002). Therefore, it was

Water Air Soil Pollut (2014) 225:2009

300

Inflow
Outflow

250
200
150
100
50

As concentration (g L-1)

As concentration (g L-1)

Page 7 of 11, 2009

300
250
200
150
100

50
0

Date

d
As concentration (g L-1)

As concentration (g L-1)

c 300

Date

250
200
150
100
50
0

300
250
200
150
100
50
0

Date

Date

Fig. 4 Inflow concentrations (white bars) and outflow concentrations (black bars) of arsenic in a series 1, b series 2, c series 3, and d series
4. Outflow concentrations were lower in series 1 than those in the other series

hypothesized that high bicarbonate loading could result


in competition of anions with soluble arsenic, primarily
arsenate (H2AsO4), for iron oxyhydroxide sorption
sites. However, no negative effect of bicarbonate loading on arsenic removal performance was observed during this experiment. Results of this study indicate that a
CWTS can effectively treat arsenic in the presence of
competitive anions.
CWTSs have proven to be a low-cost treatment option and are widely accepted for treating a broad range
of contaminants in water for various uses including
drinking, irrigation, surface water discharge, and industrial purposes (Kadlec and Wallace 2009; Mooney and
Murray-Gulde 2008; Nelson and Gladden 2008;
Nijhawan and Myers 2006). The rates and extent of
treatment (e.g., arsenic removal to below the WHO
drinking water quality guideline of 10 g L1) achieved
in our study suggest the suitability of CWTSs for removing arsenic from groundwater to provide safe drinking water. Based on the results from this investigation

and previous studies, several considerations are important in moving forward to designing and building an onsite demonstration or full-scale CWTS for treating arsenic in groundwater. CWTSs are designed for accumulation and sequestration of potentially toxic materials in
the sediment. Both accumulation rate and retention capacity (mass of a constituent retained in the sediment)
are part of the design for a CWTS. A CWTS can be
constructed to have a certain retention capacity by designing to produce a specific sediment accretion rate and
accumulation level (Murray-Gulde et al. 2008). With
this design strategy, a system can be built to maintain
concentrations of constituents below a toxic level in the
sediment, and the sediment can be left in place. If a
CWTS is specifically designed to accumulate high concentrations of the contaminants retained, the closure
plan typically includes capping the sediments. Based
on previous studies, full-scale CWTSs analogous to
our pilot-scale system will function for at least 40
50 years (Kadlec and Wallace 2009), with the arsenic

56.5

15 Mar. 2013

66.9

40.3

43.6

50.7

83.2

71.0

83.5

94.2

95.4

91.8

69.9

52.6

88.7

76.8

Removal
efficiency
(%)

0.28

0.13

0.14

0.18

0.45

0.31

0.45

0.71

0.77

0.63

0.30

0.19

0.55

0.37

Rate
coefficient
(day1)

No decrease in concentration from inflow to outflow

78.7

1 Mar. 2013

60.0

45.4

15 Feb. 2013

24.1

18 Jan. 2013

1 Feb. 2013

29.2

33.3

4 Jan. 2013

29 Nov. 2012

16 Dec. 2012

5.5

6.9

13 Nov. 2012

27.2

58.9

26 Sept. 2012

7.6

25.2

9 Sept. 2012

28 Oct. 2012

32.2

31 Aug. 2012

14 Oct. 2012

Removal
extent
(g L1)

Sampling period

Series 1

126.8

130.6

79.0

161.8

123.1

97.3

136.6

63.9

60.5

49.2

61.6

113.0

50.0

65.1

Removal
extent
(g L1)

Series 2

0a

0a
25.4

9.2

0.07

0.02

0.06

0.06

0.06

0.14

0.17

0.16

0.08

0.02

0.39

0.19

Rate coefficient
(day1)

19.9

21.1

22.7

41.9

49.2

47.9

28.3

9.2

79.2

52.9

Removal
efficiency (%)

Table 2 Removal extents, removal efficiencies, and first-order rate coefficients

223.0

151.0

77.8

148.5

146.0

117.8

165.1

85.7

113.5

79.2

98.2

149.4

103.9

128.1

Removal
extent
(g L1)

Series 3

0.04
0a

0a

0.02

0.02

0.01

0.03

0.11

0.05

0.11

0.07

0.36

0.23

0.03

Rate
coefficient
(day1)

15.9

7.6

7.5

2.6

12.0

36.1

18.3

34.5

23.5

76.2

59.8

10.5

Removal
efficiency
(%)

68.2

120.4

34.3

117.1

68.7

83.4

123.4

42.6

68.3

38.1

86.0

151.9

110.5

136.4

Removal
extent
(g L1)

Series 4

60.2

36.8

55.0

12.4

59.1

34.8

31.2

60.0

48.2

69.3

28.0

38.7

0.23

0.11

0.20

0.03

0.22

0.11

0.09

0.23

0.16

0.29

0.08

0.12

0.20

0a

0a
54.4

Rate
coefficient
(day1)

Removal
efficiency
(%)

2009, Page 8 of 11
Water Air Soil Pollut (2014) 225:2009

Water Air Soil Pollut (2014) 225:2009

sequestered in low-solubility, non-bioavailable solid


phases (Stottmeister et al. 2006; Swash and
Monhemius 2005). In addition to providing testing and
confirmation of treatment processes and conditions for
arsenic removal using CWTSs, our study provides rate
coefficients for designing size and inflow rates of demonstration and full-scale CWTSs for removal of arsenic
from groundwater. Components of the pilot-scale system used in this study could be adapted for implementation in a full-scale CWTS. For example, existing handpump tube wells in Bangladesh could be used to deliver
arsenic-contaminated groundwater directly to an inground detention basin, while natural topography could
be utilized to induce gravity flow of water through a
CWTS.

5 Conclusions
This study represents a step in the process of developing
a low-cost, passive, long-term solution to arsenic contamination of groundwater. Results demonstrate that a
pilot-scale CWTS can be designed and built to produce
biogeochemical conditions that promote processes for
removal of arsenic from simulated Bangladesh groundwater to below the WHO drinking water quality guideline of 10 g L1. In this study, arsenic removal in the
CWTSs was more effective under oxidizing conditions
than under reducing conditions. The addition of ZVI
enhanced arsenic removal performance in oxidizing
series and in reducing series (mean removal efficiency
of 72 and 42 %, respectively) compared to unamended
series (27 and 20 %, respectively). The design of a fullscale CWTS for treatment of arsenic-contaminated water would benefit from incorporation of features from
this pilot-scale study including oxidizing conditions and
amendment with ZVI. Results indicate that a CWTS is
suitable for treatment of arsenic-contaminated water
containing competitive anions.

References
Alley, B. L., Willis, B., Rodgers, J., Jr., & Castle, J. W. (2013).
Seasonal performance of a hybrid pilot-scale constructed
wetland treatment system for simulated fresh oil field produced water. Water, Air, and Soil Pollution, 244, 1639. doi:
10.1007/s1127001316395.

Page 9 of 11, 2009


APHA. (2005). Standard methods for the examination of water
and wastewater (21st ed.). Washington D.C: American
Public Health Association.
Bahar, M. M., Megharaj, M., & Naidu, R. (2013). Bioremediation
of arsenic-contaminated water: recent advances and future
prospects. Water, Air, and Soil Pollution, 244, 1722. doi:10.
1007/s112700131722y.
Bang, S., Korfiatis, G. P., & Meng, X. (2005). Removal of arsenic
from water by zero-valent iron. Journal of Hazardous
Materials, 121(13), 6167.
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. International Journal of Water Resources
Development, 13, 7992.
Buddhawong, S., Kuschk, P., Mattusch, J., Wiessner, A., &
Stottmeister, U. (2005). Removal of arsenic and zinc using
different laboratory model wetland systems. Engineering in
Life Sciences, 5(3), 247252.
Cheng, H., Hu, Y., Luo, J., Xu, B., & Zhao, J. (2009).
Geochemical processes controlling fate and transport of arsenic in acid mine drainage (AMD) and natural systems.
Journal of Hazardous Materials, 165, 1326.
Crites, R. W. (1994). Design criteria and practice for constructed
wetlands. Water Science and Technology, 29(4), 16.
Di Toro, D. M., Mahony, J. D., Hansen, D. J., Scott, K. J., Carlson,
A. R., & Ankley, G. T. (1992). Acid volatile sulfide predicts
the acute toxicity of cadmium and nickel in sediments.
Environmental Science & Technology, 26, 96101.
Dorman, L., Castle, J. W., & Rodgers, J. H., Jr. (2009). Performance of
a pilot-scale constructed wetland system for treating simulated
ash basin water. Chemosphere, 75(7), 939947.
Farrell, J., Wang, J., O'Day, P., & Conklin, M. (2001).
Electrochemical and spectroscopic study of arsenate removal
from water using zero-valent iron media. Environmental
Science & Technology, 35(10), 20262032.
Faulkner, S. P., Patrick, W. H., Jr., & Gambrell, R. P. (1989). Field
techniques for measuring wetland soil parameters. Soil
Science Society of America Journal, 53, 883890.
Francesconi, K. A., & Kuehneit, D. (2002). Arsenic compounds in the
environment. In W. T. Frankenberger Jr. (Ed.), Environmental
Chemistry of Arsenic. New York: Marcel Dekker, Inc.
Henke, K. R. (2009). Arsenic in natural environments. In K.
Henke (Ed.), Arsenic Environmental Chemistry, Health
Threats, and Waste Treatment. West Sussex, UK: Wiley.
Henke, K. R., & Hutchison, A. (2009). Arsenic chemistry. In K.
Henke (Ed.), Arsenic Environmental Chemistry, Health
Threats, and Waste Treatment. West Sussex, UK: Wiley.
Huang, J. H. (2014). Impact of microorganisms on arsenic biogeochemistry: a review. Water, Air, and Soil Pollution, 225,
1848. doi:10.1007/s112700131848y.
Hughes, M. F., Thomas, D. J., & Kenyon, E. M. (2009). Arsenic in
human history and modern societies. In K. Henke (Ed.),
Arsenic: Environmental Chemistry, Health Threats and
Waste Treatment. West Sussex, UK: Wiley.
Jackson, C. K., Koch, I., & Reimer, K. J. (2013).
Mechanisms of dissolved arsenic removal by biochemical reactors: a bench- and field-scale study. Applied
Geochemistry, 29, 174181.
Johnson, B. M., Kanagy, L. E., Rodgers, J. H., Jr., & Castle, J. W.
(2008). Feasibility of a pilot-scale hybrid constructed wetland

2009, Page 10 of 11
treatment system for simulated natural gas storage produced
waters. Environmental Geosciences, 15, 91104.
Jou, C. J., Chen, S. W., Kao, C. M., & Lee, C. L. (2008). Assessing
the efficiency of a constructed wetland using a first-order
biokinetic model. Wetlands, 28, 215219.
Kadlec, R. H. (1997). Deterministic and stochastic aspects of
constructed wetland performance and design. Water Science
and Technology, 35, 149156.
Kadlec, R. H., & Wallace, S. D. (2009). Treatment wetlands (2nd
ed.). Boca Raton, FL: CRC Press.
Kanagy, L. E., Johnson, B. M., Castle, J. W., & Rodgers, J. H., Jr.
(2008). Hydrosoil conditions in a pilot-scale constructed
wetland treatment system for natural gas storage produced
waters. Environmental Geosciences, 15, 105113.
Kanel, S. R., Manning, B., Charlet, L., & Choi, H. (2005).
Removal of arsenic(III) from groundwater by nanoscale
zero-valent iron. Environmental Science & Technology,
39(5), 12911298.
Karri, S., Sierra-Alvarez, R., & Field, J. A. (2005). Zero valent iron
as an electron-donor for methanogenesis and sulfate reduction in anaerobic sludge. Biotechnology and Bioengineering,
92(7), 810819.
Keon, N. E., Swartz, C. H., Brabander, D. J., Harvey, C., &
Hemond, H. F. (2001). Validation of an arsenic sequential
extraction method for evaluating mobility in sediments.
Environmental Science & Technology, 35, 27782784.
Kinniburgh, D. G., & Smedley, P. L. (2001). Arsenic contamination of groundwater in Bangladesh. British Geological
Survey Technical Report WC/00/19, Volume 2, Final
Report, British Geological Survey, Keyworth. http://
www.bgs.ac.uk/downloads/start.cfm?id=2222, Accessed 17
January 2014.
Kinniburgh, D. G., Smedley, P. L., Davies, J., Milne, C. J., Gaus,
I., Trafford, J. M., et al. (2002). The scale and causes of the
groundwater arsenic problem in Bangladesh. In A. H. Welch
& K. G. Stollenwerk (Eds.), Arsenic in Groundwater:
Geochemistry and Occurrence. New York: Springer.
Knight, R. L., Kadlec, R. H., & Ohlendorf, H. M. (1999). The use
of treatment wetlands for petroleum industry effluents.
Environmental Science & Technology, 33, 973980.
Lackovic, J. A., Nikolaidis, N. P., & Dobbs, G. M. (2000).
Inorganic arsenic removal by zero-valent iron.
Environmental Engineering Science, 17(1), 2939.
Leonard, E. N., Cotter, A. M., & Ankley, G. T. (1996). Modified
diffusion method for analysis of acid volatile sulfides and
simultaneously extracted metals in freshwater sediment.
Environmental Toxicology and Chemistry, 15(9), 14791481.
Leupin, O. X., Hug, S. J., & Badruzzaman, A. B. M. (2005).
Arsenic removal from Bangladesh tube well water with filter
columns containing zerovalent iron filings and sand.
Environmental Science & Technology, 39(20), 80328037.
Li, H., Ye, Z. H., Wei, Z. J., & Wong, M. H. (2011). Root porosity
and radial oxygen loss related to arsenic tolerance and uptake
in wetland plants. Environmental Pollution, 159(1), 3037.
Lizama, A. K., Fletcher, T. D., & Sun, G. (2011). Removal
pro cess es fo r arsen ic in c ons truc ted wetlan ds .
Chemosphere, 84(8), 10321043.
Manning, B. A., Hunt, M. L., Amrhein, C., & Yarmoff, J. A.
(2002). Arsenic(III) and arsenic(V) reactions with zerovalent
iron corrosion products. Environmental Science &
Technology, 36(24), 54555461.

Water Air Soil Pollut (2014) 225:2009


McArthur, J. M., Banerjee, D. M., Hudson-Edwards, K. A.,
Mishra, R., Purohit, R., Ravenscroft, P., et al. (2004).
Natural organic matter in sedimentary basins and its relation
to arsenic in anoxic ground water: the example of West
Bengal and its worldwide implications. Applied
Geochemistry, 19, 12551293.
Meng, X., Korfiatis, G. P., Christodoulatos, C., & Bang, S. (2001).
Treatment of arsenic in Bangladesh well water using a household coprecipitation and filtration system. Water Research,
35(12), 28052810.
Mooney, F. D., & Murray-Gulde, C. (2008). Constructed treatment
wetlands for flue gas desulfurization waters: full-scale design, construction issues, and performance. Environmental
Geosciences, 15, 131141.
Murray-Gulde, C. L., Bridges, W. C., & Rodgers, J. H., Jr. (2008).
Evaluating performance of a constructed wetland treatment
system designed to decrease bioavailable copper in a waste
stream. Environmental Geosciences, 15, 2138.
Nelson, E. A., & Gladden, J. B. (2008). Full-scale treatment
wetlands for metal removal from industrial wastewater.
Environmental Geosciences, 15, 3948.
Nickson, R., McArthur, J., Burgess, W., Ahmed, K. M.,
Ravenscroft, P., & Rahman, M. (1998). Arsenic poisoning
of Bangladesh groundwater. Nature, 395, 338.
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. Applied
Geochemistry, 15, 403413.
Nickson, R. T., McArthur, J. M., Shrestha, B., Kyaw-Myint, T. O.,
& Lowry, D. (2005). Arsenic and other drinking water quality
issues, Muzaffargarh District, Pakistan. Applied
Geochemistry, 20, 5568.
Nijhawan N., & Myers J. E. (2006). Constructed treatment
wetlands for the treatment and reuse of produced water
in dry climates. Society of Petroleum Engineers Paper
98567.
Rahman, M. M., Sengupta, M. K., Chowdhury, U. K., Lodh, D.,
Das, B., Ahamed, D. S., et al. (2006). Arsenic around the
worldan overview. In R. Naidu, E. Smith, G. Owens, P.
Bhattacharya, & P. Nadebaum (Eds.), Managing arsenic in
the environment. Collingwood. Victoria, Australia: CSIRO
Publishing.
Rahman, K. Z., Wiessner, A., Kuschk, P., van Afferden, M.,
Mattusch, J., & Mller, R. A. (2011). Fate and distribution
of arsenic in laboratory-scale subsurface horizontal-flow constructed wetlands treating an artificial wastewater. Ecological
Engineering, 37(8), 12141224.
Reed, S. C., & Brown, D. (1995). Subsurface flow wetlandsa
performance evaluation. Water Environment Research, 67,
244248.
Rodgers, J. H., Jr., & Castle, J. W. (2008). Constructed
wetland systems for efficient and effective treatment
of contaminated waters for reuse. Environmental
Geoscience, 15, 18.
Rounds, S. A. (2006). Alkalinity and acid neutralization capacity
(ver. 4): U.S. Geological Survey Techniques of WaterResources Investigations, book 9, chap. A6, sec. 6.6.
Rousseau, D. P. L., Vanrolleghem, P. A., & De Pauw, N. (2004).
Model-based design of horizontal subsurface flow constructed treatment wetlands: a review. Water Research, 38, 1484
1493.

Water Air Soil Pollut (2014) 225:2009


Sharma, V. K., & Sohn, M. (2009). Aquatic arsenic: toxicity,
speciation, transformations, and remediation. Environment
International, 35, 743759.
Siracusa, G., & La Rosa, A. D. (2006). Design of a constructed
wetland for wastewater treatment in a Silician town and
environmental evaluation using the emergy analysis.
Ecological Modelling, 197, 490497.
Smedley, P. L., & Kinniburgh, D. G. (2002). A review of the
source, behaviour and distribution of arsenic in natural waters. Applied Geochemistry, 17, 517568.
Stollenwerk, K. G. (2002). Geochemical processes controlling transport of arsenic in groundwater: a review of adsorption. In A. H.
Welch & K. G. Stollenwerk (Eds.), Arsenic in Ground Water:
Geochemistry and Occurrence. New York: Springer.
Stottmeister, U., Buddhawong, S., Kuschk, P., Wiessner, A., &
Mattusch, J. (2006). Constructed wetlands and their performance for treatment of water contaminated with arsenic and
heavy metals. In I. Twardowska, H. E. Allen, & M. H.
Hggblom (Eds.), Soil and Water Pollution Monitoring,
Protection and Remediation. New York: Springer.
Su, C. (2007). Utilization of zero-valent iron for arsenic removal
from ground water and wastewater. In I. M. C. Lo, R. Y.
Surampalli, & K. C. K. Lai (Eds.), Zero-Valent Iron Reactive
Materials for Hazardous Waste and Inorganics Removal.
Reston, Virginia: ASCE Publications.
Su, C., & Puls, R. W. (2001a). Arsenate and arsenite removal by
zerovalent iron: kinetics, redox transformation, and implications for in situ groundwater remediation. Environmental
Science & Technology, 35, 14871492.

Page 11 of 11, 2009


Su, C., & Puls, R. W. (2001b). Arsenate and arsenite removal by
zerovalent iron: effects of phosphate, silicate, carbonate, borate,
sulfate, chromate, molybdate, and nitrate, relative to chloride.
Environmental Science & Technology, 35, 45624568.
Swash, P. M., & Monhemius, A. J. (2005). Characteristics
and stabilities of residues from the Wheal Jane constructed wetlands. Science of the Total Environment,
338, 95105.
USEPA. (1994). Method 200.8: determination of trace elements in
waters and wastes by inductively coupled plasma-mass spectrometry. United States Environmental Protection Agency,
Office of Research and Development, Cincinnati, Ohio.
USEPA. (2002). Arsenic treatment technologies for soil, waste,
and water. Solid Waste and Emergency Response (5102G).
EPA-542-R-02-004. United States Environmental Protection
Agency. Washington, D.C.
WHO. (2011). Guidelines for drinking water qualityfourth edition. World Health Organization. ISBN 978-92-4-154815-1.
Geneva, Switzerland
Wilkin, R. T., & Ford, R. G. (2002). Use of hydrochloric acid for
determining solid-phase arsenic partitioning in sulfidic sediments. Environmental Science & Technology, 36(22), 4921
4927.
Wong, T. H. F., Fletcher, T. D., Duncan, H. P., & Jenkins, G. A.
(2006). Modelling urban stormwater treatmenta unified
approach. Ecological Engineering, 27, 5870.
Wood, A. (1995). Constructed wetlands in water pollution control:
fundamentals to their understanding. Water Science and
Technology, 32, 2129.

You might also like