w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 7 6 4 e5 7 7 2

Available online at www.sciencedirect.com

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

Managed aquifer recharge of treated wastewater: Water

quality changes resulting from infiltration through
the vadose zone

Elise Bekele a,*, Simon Toze b,c, Bradley Patterson a,d, Simon Higginson e
a
CSIRO Water for a Healthy Country Flagship, CSIRO Centre for Environment and Life Sciences, Private Bag No 5, PO Wembley,
Western Australia 6913, Australia
b
CSIRO Water for a Healthy Country Flagship, Ecosciences Precinct, 41 Boggo Road, Dutton Park, QLD 4102, Australia
c
School of Population Health, University of Queensland, Herston Road, Herston, QLD 4006, Australia
d
School of Biomedical, Biomolecular and Chemical Sciences, University of Western Australia, Crawley, WA 6009, Australia
e
Water Corporation of Western Australia, PO Box 100, Leederville, WA 6902, Australia

article info abstract

Article history:
Received 8 June 2011
Received in revised form
23 August 2011
Accepted 27 August 2011
Available online 3 September 2011
Keywords:
Managed aquifer recharge
Wastewater infiltration
Natural attenuation processes
Secondary treated wastewater was infiltrated through a 9 m-thick calcareous vadose zone
during a 39 month managed aquifer recharge (MAR) field trial to determine potential
improvements in the recycled water quality. The water quality improvements of the
recycled water were based on changes in the chemistry and microbiology of (i) the recycled
water prior to infiltration relative to (ii) groundwater immediately down-gradient from the
infiltration gallery. Changes in the average concentrations of several constituents in the
recycled water were identified with reductions of 30% for phosphorous, 66% for fluoride,
62% for iron and 51% for total organic carbon when the secondary treated wastewater was
infiltrated at an applied rate of 17.5 L per minute with a residence time of approximately
four days in the vadose zone and less than two days in the aquifer. Reductions were also
noted for oxazepam and temazepam among the pharmaceuticals tested and for a range of
microbial pathogens, but reductions were harder to quantify as their magnitudes varied
over time. Total nitrogen and carbamazepine persisted in groundwater down-gradient
from the infiltration galleries. Infiltration does potentially offer a range of water quality
improvements over direct injection to the water table without passage through the
unsaturated zone; however, additional treatment options for the non-potable water may
still need to be considered, depending on the receiving environment or the end use of the
recovered water.
Crown Copyright ª 2011 Published by Elsevier Ltd. All rights reserved.

1. Introduction either ponds, basins or shallow buried trenches. There is

An essential design question for proponents of managed
aquifer recharge (MAR) schemes using recycled water in
shallow aquifers is whether it is more beneficial to recharge
the aquifer via direct well injection or to use infiltration with
a range of different types of MAR as described in Dillon (2005).
Ideally the design of a MAR scheme in an urban setting should
have minimal surface footprint and minimal exposure
potential for the community. Well injection offers advantages
such as no evaporative loss, algae or mosquitoes, and no loss

* Corresponding author. Tel.: þ61 0 8 9333 6718; fax: þ61 0 8 9333 6211.
E-mail address: Elise.Bekele@csiro.au (E. Bekele).
0043-1354/$ e see front matter Crown Copyright ª 2011 Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.watres.2011.08.058
 w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 7 6 4 e5 7 7 2
of prime land (Pyne, 2006). Well injection, however, can suffer
from clogging issues relating to the initial quality of the
recycled water and loses the benefit of potential treatment
processes provided by the unsaturated zone. In comparison,
surface infiltration offers potential treatment during the
migration of recycled water through the vadose zone, which
can result in improved water quality via biological, chemical
and physical processes before reaching the water table but
can have the drawback of having a larger surface footprint and
potential for exposed water, none of which is ideal in an urban
environment. Improvements to water quality using infiltra-
tion have been demonstrated to reduce organic matter
(Quanrud et al., 2003; Vanderzalm et al., 2010), trace organic
compounds (Montgomery-Brown et al., 2003), nitrogen (Zhang
et al., 2005) and bacteria (Schafer et al., 1998; Toze et al., 2004).
However, a rigorous determination of the magnitudes of
concentration reductions for a range of different chemical and
biological contaminants in treated wastewater has not been
conducted previously in MAR studies using unsaturated,
calcareous sands such as the Spearwood sands of Swan
Coastal Plain of Western Australia.
The aim of this study was to determine the changes in
recycled water quality after infiltrating vertically through
a 9 m-thick vadose zone and migrating laterally through 2.3 m
of aquifer using infiltration galleries as the MAR method
during a 39 month MAR field trial. The focus of this work is on
MAR in calcareous sand and limestone due to the prevalence
of these deposits, which comprise unconfined aquifers in the
Perth metropolitan region of Western Australia where there is
also keen interest in enhancing the role of MAR (Scatena and
Williamson, 1999; Smith and Pollock, 2010). The evaluation of
water quality was based on measured concentrations of major
ions, nutrients, trace metals, organic carbon, pharmaceutical
compounds (carbamazepine, diazepam, oxazepam,
phenytoin, temazepam) and numbers of faecal indicator
microorganisms and selected enteric pathogens (thermoto-
lerant coliforms, enterococci, bacteriophage, adenovirus)
before and after recycled water passed through the
subsurface.
2. Materials and methods
2.1. Facilities
The source of secondary treated wastewater was the Subiaco
Wastewater Treatment Plant in Western Australia. After
passage through a multi-media filtration system (AMIAD),
the treated wastewater was pumped to the recharge site
(Bekele et al., 2009). Two infiltration galleries were used at the
CSIRO Centre for Environment and Life Sciences in Floreat,
Western Australia for a pilot-scale investigation of MAR
(Bekele et al., 2009). The east gallery was filled with 10 mm
graded and washed granite gravel; the west gallery contained
a series of modular polypropylene tanks, referred to as the
Atlantis system by the manufacturer. The dimensions of
each tank were 685 mm  408 mm  450 mm (length by width
by height) and the tanks have a modular design so that they
can be positioned in the trench and clipped together. The
construction of each module resembles a milk crate that has
5765
large void spaces supported by matrix. In this study, each
gallery trench was 25 m  1 m  0.5 m (length by width by
height) and the width of the Atlantis gallery was 816 mm,
consisting of two tank modules positioned side-by-side. Each
gallery was buried to a depth of 1 m below ground and covered
with 0.5 m of sediment backfill composed of Spearwood Dune
sand from digging the trench (Bekele et al., 2009). The recharge
site was located in a grassy paddock where sheep were
allowed to graze. The infiltration galleries operated almost
continuously for 39 months and infiltrated a total of 36.7 ML of
treated wastewater to the aquifer supplied at a daily constant
rate of 17.5 L per minute (Bekele et al., 2009).
2.2. Local geology and hydraulic considerations
The geology of the site consisted of a 7 m-thick top layer of
Spearwood Dune sand, overlying the Tamala Limestone
aquifer. The Tamala Limestone is a calcareous aeolinite which
has been weathered to produce the overlying Spearwood
sands (Tapsell et al., 2003). The aquifer extends to a depth of
31 m below ground and is underlain by sediments from
a regional aquitard. Regional investigations of the Tamala
Limestone reveal high porosity zones from fractures and
cavities (Davidson, 1995).
The sand mineralogy from 0 to 7.5 m below ground was
predominantly quartz (>80%), underlain by a more heteroge-
neous section from 7.5 to 11.6 m below ground with an
average composition of quartz (60%), calcite (30%), microcline
(5%) and anorthite (5%) from XRD analysis. Further mineral
phase characterization using AutoGeoSEM (Robinson et al.,
2000) on a sand sample revealed aluminum and iron oxides.
The aluminum oxides are silicate weathering products (clay
minerals) deposited as coatings on sand grains, which are
colored yellow by the presence of hydrated iron oxides
(Bastian, 1996).
The depth to the water table below the galleries varied
seasonally between 10 and 11 m. The regional groundwater
flow direction was from east to west. For experimental
purposes, an artificial hydraulic gradient was produced by
continual pumping from a well located 50 m west of the
infiltration galleries. The natural hydraulic gradient coincided
with the imposed gradient. A series of monitoring wells were
installed with slotted intervals positioned at different depths
below ground as shown in Fig. 1.
The migration rate of the infiltrated recycled water through
the vadose zone was previously determined based on
a bromide tracer experiment (Bekele et al., 2009). From this
data, a minimum travel time of 3.7 days was estimated
through the unsaturated zone. The travel time to BH1 after
passage through the vadose zone was estimated to be an
additional 0.5 day based on a comparison of the electrical
conductivity of the recycled water relative to groundwater
from BH1, indicating that the substantial time for processes to
occur was in the vadose zone compared to the aquifer.
2.3. Description of monitoring wells
To assess the potential infiltration benefits, improvements in
quality of the recycled water immediately down-gradient
from the infiltration galleries were compared to the quality
5766 w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 7 6 4 e5 7 7 2

general accordance with ASTM D 4448-01 Standard Guide for

Sampling Ground-Water Monitoring Wells (2001). Wastewater
samples were collected from the discharge chamber in the
infiltration galleries.
Water samples were analyzed for major ions, nutrients,
trace metals, dissolved and total organic carbon, trace
organics and faecal indicator microorganisms and pathogens.
Further details regarding water sampling, analytical proce-
dures and detection limits for the water chemistry are given in
Bekele et al. (2009). Faecal indicator microbes E. coli and
enterococci were detected using membrane filtration and
selective media while the presence of enteric pathogens were
determined from a concentrate of 40 L using PCR as described
in Toze and Bekele (2009). Sampling water from the galleries
and the wells for different chemicals and microbes was con-
ducted during the first 25 months of the MAR trial at different
time intervals, but generally on a weekly, fortnightly or
monthly basis.
Physical water parameters measurements were taken
before water sampling occurred. These included pH, dissolved
oxygen, electrical conductivity, temperature, oxidation-
reduction potential and turbidity measurements using
a Troll 9000 multisensory meter housed within a flow cell.

2.5. Criteria to assess changes in recycled water quality

Fig. 1 e The Floreat MAR site showing the positions of the
two galleries, the monitoring wells (BH1, BH2 and BH5) and
the recovery well in map view (a), and relative to the
maximum height of the water table in cross-section (b).
of the infiltrated recycled water. Monitoring wells close to the
infiltration galleries (BH1, BH2 and BH5) were selected to
evaluate the benefits to water quality of infiltrating recycled
water through the vadose zone. Three background ground-
water monitoring wells located hydraulically up-gradient of
the infiltration galleries (BGRND1, BGRND2 and BGRND3) were
also monitored to provide an assessment of changes in
groundwater quality due to infiltration. Monitoring well
details are given in Table 1.
2.4. Collection and analysis of water samples
Details of the groundwater sampling procedure are given in
Bekele et al. (2009). Groundwater sampling procedures were in
To assess the water quality changes as a result of the migra-
tion through the vadose zone, the quality of water samples
collected from the shallow monitoring wells immediately
down-gradient from the infiltration galleries were compared
to the quality of the recycled water prior to infiltration.
Sampling of the vadose zone water directly below the infil-
tration galleries and immediately above the water table was
not undertaken. Thus, only gross changes after passage over
the entire vadose zone could be quantified as changes during
infiltration could not be determined. The significance of the
effects of migration through the subsurface on recycled water
quality was assessed using Student’s two-tailed t-test
(unpaired) applied to datasets of measured nutrients, inor-
ganic compounds, pharmaceutical and microbial pathogens,
comparing the mean concentrations in recycled water prior to
infiltration relative to the mean concentrations in recharged
water sampled from BH1. For the Student’s t-test, the null
hypothesis was that water quality was unchanged despite
passage through the vadose zone. Microsoft Excel was used to

Table 1 e Monitoring well details.

Well Ground elevation Distance and direction Total depth Screened depth interval below the
(m AHDa) from west gallery below ground (m) maximum height of the water table (m)

BH1 12.97 2.3 m west 12.01 0.0e2.0

BH2 13.01 2.5 m west 12.04 0.0e2.0
BH5b 12.86 8.8 m east 10.81 0.0e1.0
BGRND1 15.94 185 m northeast 15.00 0.0e2.05
BGRND2 12.00 75 m east 12.23 2.23e3.23
BGRND3 12.00 75 m east 20.11 10.11e11.11

a Australian Height Datum.

b BH5 is located up-gradient relative to the groundwater flow direction and a distance of 4 m east of the east gallery.
 w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 7 6 4 e5 7 7 2 5767

calculate the probability (P) and the null hypothesis was
rejected if P < 0.05.
3. Results and discussion
3.1. Characterization of recycled water and ambient
groundwater
To understand what water quality changes were potentially
achievable during passage of the recycled water through the
vadose zone, it was also necessary to characterise and
compare the ambient groundwater. Both the ambient
groundwater and the recycled water were consistently
aerobic. The average water temperature, pH, electrical
conductivity, dissolved oxygen concentrations and sulfate
concentrations of the ambient groundwater were similar to
that of the recycled water (Table 2). The concentration of total
dissolved solids was highly variable in the recycled water and
the range of variation overlapped with that of the ambient
groundwater. The major ions defining the water types were
CaeNaeCleHCO3 for ambient groundwater and NaeCleHCO3
for recycled water.
The recycled water and ambient groundwater had low
concentrations of several inorganic species, namely arsenic,
boron, cadmium, cobalt, chromium, copper, mercury,
manganese, molybdenum, nickel, lead, selenium, uranium,
vanadium and zinc. The impact of infiltration on these inor-
ganic chemicals was not investigated as their measured
concentrations were either very low or were below detection
limits.
3.2. MAR impacts on groundwater quality near the
galleries
Site groundwater monitoring was undertaken to confirm that
the shallow groundwater collected closest to the infiltration
galleries did represent infiltrated recycled water and not
ambient groundwater. Average concentrations of potassium,
chloride and sodium in the recycled water were similar to the
average concentrations from water samples collected from
the monitoring well BH1 (Table 3). The recycled water had an
average potassium concentration that was greater than four-
fold more enriched in potassium compared to the ambient
groundwater and therefore was an effective tracer for the
infiltrated recycled water as demonstrated by others in
previous studies (Rueedi et al., 2009; Wolf et al., 2004). The
recharged water quality sampled from monitoring wells BH1
and BH2 was indistinguishable from the recycled water prior
to infiltration, but substantially different from ambient
groundwater on the basis of potassium and chloride concen-
trations (Fig. 2).
Water samples collected from the monitoring well BH5,
4 m up-gradient from the galleries, showed a greater propor-
tion of ambient groundwater mixed with recycled water
compared to the water samples from the monitoring wells
down-gradient of the galleries. A theoretical line of mixing
between the average concentrations of chloride and potas-
sium in ambient groundwater and that of recycled water was
generated (Fig. 2). During the first three months of infiltration,
the composition of groundwater from BH5 was predominantly
ambient groundwater (80%); thereafter, the composition of
groundwater from BH5 had a higher proportion of recycled
water relative to ambient groundwater. Since water sampled
from well BH1 did not show temporal transition and it was
frequently monitored, water collected from BH1 was used for
evaluating influent water quality improvements as a result of
its migration through the vadose zone.
3.3. Water quality changes during infiltration
3.3.1. Inorganic chemicals
The recycled water prior to infiltration had higher average
concentrations of most species of nitrogen, phosphorous, and
fluoride compared to ambient groundwater (Table 3).
Recharged water sampled down-gradient from the galleries
from BH1 had average concentrations for these chemicals that
were generally between those for ambient groundwater and
recycled water or within one standard deviation of these two
water sources.
To assess improvements due to infiltration, total nitrogen
concentrations and a relatively non-reactive analyte (i.e.
Table 3 e Mean concentrations (mg/L) of selected
chemicals in recycled water and groundwater sampled
down-gradient from the galleries from BH1 compared to
ambient groundwater. Standard deviations are provided
in parentheses.
Analtye Recycled Ambient BH1
water groundwater

Chloride 245 (62) 162 (27) 248 (50)

Table 2 e Mean values of water quality parameters of the Potassium 22.9 (3.5) 4.96 (0.83) 21.7 (4.1)
recharge water that were similar to ambient groundwater Sodium 194 (44) 92.6 (15) 194 (42)
with standard deviations in parentheses. Ammonia as Nitrogen 0.64 (0.85) <0.01 0.037 (0.046)
Nitrate as Nitrogen 2.16 (1.41) 0.16 (0.26) 3.90 (1.61)
Parameter Recycled Ambient
Total Kjeldahl Nitrogen 1.86 (0.98) 0.066 (0.030) 0.95 (0.51)
(Unit of Measure) water Groundwater
Total Nitrogen 4.27 (1.9) 0.302 (0.36) 4.93 (1.8)
Water temperature ( C) 24 (4) 22 (1) Soluble reactive 6.31 (3.32) 0.0126 (0.007) 2.19 (1.53)
pH 7.33 (1.11) 7.04 (0.88) phosphorus as P
Electrical Conductivity (mS m1) 143 (53) 124 (63) Total organic carbon 9.98 (3.8) 2.66 (2.72) 6.42 (3.57)
Eh (mV-SHE) 385 (184) 321 (189) Fluoride 0.73 (0.2) 0.15 (0.1) 0.25 (0.05)
Dissolved Oxygen (mg L1) 2.15 (1.82) 4.02 (2.56) Iron 0.14 (0.20) 0.44 (0.72) 0.053 (0.054)
Sulfate as S (mg L1) 64.1 (8) 64.5 (18) Calcium 28.6 (9.49) 98.8 (23.5) 60.4 (7.28)
Total Dissolved Solids (mg L1) 755 (179) 644 (25) Aluminum 0.018 (0.011) 0.22 (0.48) 0.045 (0.048)
5768
30
25
Potassium (mg/L)
20
15
10
5 BH5, samples during first 3 months
0
0 100 200
Chloride (mg/L)
Fig. 2 e Cross plot showing concentrations of chloride and
potassium in ambient groundwater sampled from the
background monitoring wells, recycled water and in water
recovered from the monitoring wells (BH1, BH2 and BH5). A
mixing line is shown between the average composition of
recycled water and ambient groundwater.
potassium) were plotted (Fig. 3). No improvement in total
nitrogen was demonstrated as shown by the cluster of data for
BH1 in Fig. 3, which overlaps with that of recycled water and
indicates a higher range of total nitrogen in contrast to
ambient groundwater (<1.2 mg/L). Migration of recycled water
through the vadose zone and aquifer to reach BH1 did not
improve total nitrogen concentrations (P ¼ 0.24, Student’s t-
test).
Although passage through the vadose zone did not signif-
icantly alter the concentration of total nitrogen in the recycled
water, there were changes to different nitrogen species
(Table 3). The mean concentration of total Kjeldahl nitrogen in
the recharged water sampled from BH1 was 49% lower relative
to the mean concentrations in recycled water prior to infiltra-
tion by (P < 0.01, Student’s t-test), whereas the mean concen-
tration of ammonia was 94% lower (P < 0 0.01). In comparison,
the mean concentration of nitrate was 77% higher (P < 0.001) in
the recharged water sampled from BH1 relative to the mean
concentrations in recycled water prior to infiltration. This data
suggests that nitrification was occurring through the vadose
10
9 14
Total Nitrogen (mg/L)
8 BH1
7
6 10
5
4
3 6
2
1
0 2
0 5 10 15 20
Potassium (mg/L)
Fig. 3 e Cross plot showing concentrations of total nitrogen
and potassium in recycled water, ambient groundwater
and recharged water sampled down-gradient from the
galleries from monitoring well BH1.
w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 7 6 4 e5 7 7 2
zone with nitrate production from the ammonium and total
Kjeldahl nitrogen, and provides a plausible explanation for the
lack of change in total nitrogen concentrations.
As conditions in the vadose zone were not conducive for
reducing nitrate concentrations, potential treatment options
to consider include manipulation of the oxygen content to
drive denitrification and the addition of carbon amendments
Ambient groundwater to promote nitrate removal within the aquifer (Patterson et al.,
BH1
BH2 2004, 2011).
BH5, samples after first 3 months
Migration of the recycled water through the vadose zone
Recycled water and 2.3 m of aquifer aided removal of phosphorous (P < 0.0001,
300 400 500 Student’s t-test). Further investigation revealed a time-
dependence for phosphorous removal. During the first 124
days of infiltration, phosphorous in recharged water sampled
from the monitoring well BH1 down-gradient from the
galleries remained low, similar to ambient groundwater
(<0.02 mg/L), whereas the average phosphorous in the
influent recycled water during this initial period was 10 mg/L
(Fig. 4). After 271 days of infiltration, the average phosphorous
concentration in the influent recycled water was 4.7 mg/L and
more variable than previously (standard deviation of 2.1 mg/
L). In comparison, the concentration of phosphorous in
recharged water collected from BH1 reached a maximum on
day 271 and then stabilized with a mean concentration of
3.2 mg/L (standard deviation of 0.68 mg/L).
Comparison of average phosphorous concentrations in
BH1 and the influent recycled water reveals a reduction of 31%
after 271 days of infiltration.
Adsorption and precipitation are reported to be the main
causes of phosphorous retention in calcareous sands and soils
(von Wandruszka, 2006; Whelan, 1988; Whelan and Barrow,
1984). However, calcareous sands have a limited capacity to
store phosphorous, thus caution may be warranted in antici-
pating consistent levels of phosphorous removal, particularly
at sites recharging higher volumes of recycled water and after
prolonged recharge. Desorption and remobilization of phos-
phorous should be considered in dealing with long-term MAR
systems. Previous studies have demonstrated the transient
nature of phosphorous retention in carbonate soils receiving
wastewater. Examples include the release of sorbed phos-
phate caused by dissolution of carbonate, e.g. from the acid-
ification of leachate during nitrification of ammonium
Ambient groundwater
12 Recycled water
Phosphorous (mg/L)
8
4
25 30 35
Ambient groundwater 0
BH1 0 100 200 300 400 500 600 700
Recycled water Time since start of infiltration (days)
Fig. 4 e Temporal changes in concentrations of
phosphorus in recycled water, ambient groundwater and
water sampled down-gradient from the galleries from
monitoring well BH1.
 w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 7 6 4 e5 7 7 2
(Whelan, 1988), and soil saturated with phosphate becoming
a source of phosphate if concentrations in the wastewater
recharge decrease (Lin and Banin, 2005).
Migration through the subsurface reduced fluoride
concentrations in the infiltrated water (P < 0.0001, Student’s t-
test). The comparison of average fluoride concentrations in
BH1 and the influent recycled water reveals a reduction of 66%
(Table 3). Similar to the reduction in phosphorous, reductions
in fluoride concentrations were likely due to adsorption and
precipitation in calcite-bearing layers (Turner et al., 2005).
Adsorption of fluorite was more likely as both the recycled
water and ambient groundwater were under-saturated with
respect to fluorite.
Concentrations of soluble iron in the ambient groundwater
sampled from the three background wells were quite variable
(mean of 0.44 mg/L; standard deviation of 0.72 mg/L) making it
difficult to interpret changes to recycled water quality from
passage through the subsurface. Infiltration of the recycled
water sustained aerobic conditions favorable for the removal
of soluble iron near the gallery. This may explain the nearly
three-fold difference between the average concentration of
soluble iron in water sampled from well BH1 and recycled
water prior to infiltration (P ¼ 0.013, Student’s t-test).
Comparison of average iron concentrations in BH1 and the
influent recycled water reveals a reduction of 62%.
While the pH of the recycled water varied over time during
the trial, the pH of the groundwater at BH1 was consistently
higher than the recycled water. During infiltration, the
average pH of water sampled from well BH1 was 7.61 (stan-
dard deviation of 0.80), whereas the average pH of recycled
water prior to infiltration was 7.33 (standard deviation of 1.11;
Table 2). The increase in pH of the recycled water as it infil-
trated through the calcareous sand and limestone was likely
due to calcite dissolution in the vadose zone.
Calcium and aluminium concentrations in water collected
from well BH1 were higher relative to their concentrations in
the recycled water; however ambient groundwater had the
highest concentrations of these ions (Table 3). The average
calcium concentration in the water sampled from BH1 was
twice the average concentration in the recycled water prior to
infiltration. The recycled water was under-saturated with
respect to calcite whereas ambient groundwater was in
equilibrium with calcite; hence conditions were favourable for
calcite dissolution thereby increasing calcium in water
sampled down-gradient from the galleries. The average
concentration of aluminium in groundwater sampled from
well BH1 was also twice the average aluminium concentration
in the recycled water. It is likely that the mobility of
aluminium is linked to dissolution of calcium carbonate
coated with aluminium oxides.
3.3.2. Organic carbon
Migration of recycled water through the subsurface to reach
well BH1 reduced the total organic carbon concentrations
(P < 0.01, Student’s t-test). Comparison of average TOC
concentrations in BH1 and the influent recycled water reveals
a reduction of 51% after 400 days of infiltration (Fig. 5). Prob-
able mechanisms for the removal of organic carbon included
filtration, adsorption and biodegradation (Drewes et al., 2003;
Fox et al., 2001; Vanderzalm et al., 2010).
5769
30
Ambient groundwater
BH1
Total Organic Carbon (mg/L)
25 Recycled water
20
15
10
5
0
0 100 200 300 400 500 600 700
Time since start of infiltration (days)
Fig. 5 e Temporal changes in concentrations of total
organic carbon in recycled water, ambient groundwater
and down-gradient from the galleries in water sampled
from monitoring well BH1.
3.3.3. Pharmaceuticals
All of the pharmaceuticals tested revealed concentrations in
ambient groundwater that were below detection limits (i.e.
<0.05 mg/L for carbamazepine; <0.01 mg/L for diazepam,
oxazepam, phenytoin and temazepam). In comparison, diaz-
epam and phenytoin were the only pharmaceuticals below
the detection limit in the recycled water (Table 4). Migration of
recycled water through the vadose zone produced reductions
in oxazepam and temazepam (P < 0.01, Student’s t-test);
however, concentrations of these pharmaceuticals in water
collected from BH1 were highly variable. Additional treatment
may be required to produce consistent levels of concentration
reduction for oxazepam and temazepam, depending on the
end use of the extracted water and the potential for large
volumes to be consumed intentionally or unintentionally by
the community. Sorption and biodegradation are common
mechanism for the removal of pharmaceuticals under aerobic
conditions in the vadose zone (Amy and Drewes, 2007; Conn
et al., 2010; Scheytt et al., 2006; Tiehm et al., 2011), but
further investigation of removal mechanisms for the phar-
maceuticals was not undertaken in this study.
An unexpected observation was that concentrations of car-
bamazepine in the water collected from well BH1 were consis-
tently higher (80% higher on average; standard deviation of 37%)
than carbamazepine detected in the recycled water. The higher
carbamazepine concentrations from BH1 relative to recycled
water may be due to changes in concentration in the recycled
water over time, which varied between 0.13 and 0.33 mg/L
(average of 0.21 mg/L, Table 4). In column experiments using
aquifer sediment and recycled water from the MAR project and
under fully saturated, aerobic conditions, no sorption or
degradation were observed for carbamazepine and oxazepam
(Patterson et al., 2009). The persistence of carbamazepine in the
environment has been demonstrated previously and it has been
suggested that it could be used as a potential anthropogenic
marker in aquatic environments (Clara et al., 2004).
These results may be important for wastewater reuse near
wetlands or ecologically-sensitive areas where endocrine-
disrupting chemicals could impact negatively on aquatic
species in the receiving environment (Lister et al., 2009;
Saaristo et al., 2010; Sulleabhain et al., 2009).
5770 w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 7 6 4 e5 7 7 2
Table 4 e Mean concentrations (mg/L) of pharmaceuticals
in recycled water, water sampled from BH1, and ambient
groundwater. Standard deviations are given in
parentheses.
Analtye Recycled water Ambient BH1
(n ¼ 25) groundwater (n ¼ 16)
(n ¼ 33)
Carbamazepine 0.21 (0.067) <0.05 0.38 (0.089)
Diazepam <0.1 <0.1 <0.1
Oxazepam 0.31 (0.090) <0.1 0.21 (0.14)
Phenytoin <0.1 <0.1 <0.1
Temazepam 0.31 (0.084) <0.1 0.17 (0.12)
3.3.4. Faecal indicator microorganisms and enteric pathogens
It is worth noting that thermotolerant coliforms and entero-
cocci were detected in ambient groundwater from wells up-
gradient from the recharge site in the sheep paddock. With
regard to thermotolerant coliforms, 13% of ambient ground-
water samples tested (n ¼ 38) and 19% of water samples from
BH1 (n ¼ 27) had numbers greater than 10 cfu 100 mL1. It is
hypothesized that surface contamination was most likely due
to excreta from the grazing sheep at the recharge site as the
source of these microbes and not the treated wastewater
being recharged to the aquifer.
Despite the additional source of faecal indicator micro-
organisms from sheep excreta, migration of recycled water
from the galleries through the vadose zone to well BH1
produced a reduction in thermotolerant coliforms
(P < 0.0001, Student’s t-test). Thermotolerant coliforms were
routinely detected in the recycled water with 80% of the
samples tested (n ¼ 46), exceeding 10 colony forming units
(cfu) 100 mL1.
High numbers of enterococci were present in the recycled
water with 79% of the water samples tested (n ¼ 48) exceeding
100 cfu 100 mL1, whereas 19% of ambient water samples
(n ¼ 42) exceeded 100 cfu 100 mL1. Despite the high counts in
the recycled water, there was some reduction in enterococci
numbers comparing recycled water and water from BH1 with
only 28% of samples from the BH1 (n ¼ 29) exceeding 100 cfu
100 mL1 (P < 0.001, Student’s t-test).
Reductions in microbial pathogen numbers in the recycled
water were shown by fewer detections of adenovirus and Fþ
bacteriophage after migration through the subsurface.
Adenovirus was detected in 68% of the samples of the recycled
water (n ¼ 19) prior to infiltration but in only 6% of the samples
from BH1 (n ¼ 18) and in none of the wells further down
gradient. Adenovirus was not detected in any of the ambient
groundwater samples. Fþ bacteriophage, commonly used as
a surrogate enteric virus, were detected in 94% of recycled
water samples tested (n ¼ 36). In comparison only 4% of the
water samples from BH1 (n ¼ 24), and 6% of the ambient
groundwater samples (n ¼ 33) were positive for this
bacteriophage.
Reductions in microbial pathogens in treated wastewater
recharged to the Tamala Limestone aquifer over time were
demonstrated using survival experiments conducted with
selected faecal indicators (Cryptosporidium, adenovirus, rota-
virus, coxsackievirus, MS2, E. coli, S. enterica and E. faecalis) in
in-situ diffusion chambers (Toze et al., 2010). The reductions in
microbial pathogens were attributed to a combination of
physical removal processes during filtration and the activity of
indigenous groundwater microorganisms (Gordon et al., 2002;
Toze and Hanna, 2002). The aquifer has an active treatment
capacity to remove pathogens (Toze and Bekele, 2009), but
a longer period of aquifer residence may be needed to allow for
more inactivation of microbial pathogens (Toze et al., 2010).
4. Conclusions
 Water quality benefits were achieved by infiltrating
secondary treated wastewater through weathered calcar-
eous sands and limestone of the vadose zone in an urban
area using infiltration galleries.
 Reductions in the average concentrations of several
constituents in the recycled water before and after MAR
were identified as follows: 30% for phosphorous, 66% for
fluoride, 62% for iron and 51% for total organic carbon.
Phosphorous removal declined over time, implying the
“maximum” P adsorption capacity was reached. Reductions
in the average concentrations of oxazepam and temazepam
and the numbers of thermotolerant coliforms and entero-
cocci in the infiltrated water were also identified, but not as
easily quantified.
 The removal rates determined for chemical and biological
species are particular to the MAR conditions in this study.
The impact of a thicker unsaturated zone, anoxic conditions
and different flow rates and contact times were not inves-
tigated and would give rise to different removal rates.
 The aerobic conditions present in the vadose zone were not
conducive for denitrification to reduce nitrate concentra-
tions in the water recharged to the aquifer, revealing that
secondary treated wastewater recharged via infiltration
cannot rely entirely upon processes in the vadose zone for
nitrate removal.
 Geochemical conditions were favorable for calcite dissolu-
tion, suggesting that porosity in the calcareous vadose zone
may increase over time and could cause faster breakthrough
of contaminants to the water table.
 Future MAR in this aquifer type should consider the effects
of prolonged recharge, higher rates of recharge and
concentration-dependency of adsorption and precipitation
reactions (e.g. controlling phosphorous mobility) to deter-
mine the long-term sustainability of recharging recycled
water to a calcareous aquifer in urban environments.
Acknowledgements
This research was funded by the Western Australian
Government through the Water Foundation (gs1), the Water
Corporation of WA and the CSIRO Water for a Healthy
Country Flagship Program. Chemical analyses were
conducted by the Chemistry Centre of Western Australia.
The authors would like to thank Mr. Mark Shackleton and
Mr. Sebit Gama from the CSIRO for their assistance with
water sampling and microbial analyses. Drs Joanne
 w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 7 6 4 e5 7 7 2
Vanderzalm and Grant Douglas (CSIRO) are gratefully
acknowledged for discussions on the geochemistry. The
content of this paper benefited from internal reviews by Drs
Declan Page and Warish Ahmed at the Commonwealth
Scientific and Industrial Research Organization.
references
Amy, G., Drewes, J., 2007. Soil aquifer treatment (SAT) as a natural
and sustainable wastewater reclamation/reuse technology:
fate of wastewater effluent organic matter (EfOM) and trace
organic compounds. Environmental Monitoring and
Assessment 129 (1e3), 19e26.
ASTM, 2001. Standard Guide for Sampling Ground-water
Monitoring Wells D 4448-01. edited. American Society for
Testing and Materials (ASTM).
Bastian, L., 1996. Residual soil mineralogy and dune subdivision,
Swan coastal plain, Western Australia. Australian Journal of
Earth Sciences 43 (1), 31e44.
Bekele, E., Toze, S., Patterson, B., Devine, B., Higginson, S.,
Fegg, W., Vanderzalm, J., 2009. Design and operation of
infiltration galleries and water quality guidelines - Chapter 1.
In: Toze, S., Bekele, E. (Eds.), Determining the Requirements
for Managed Aquifer Recharge in Western Australia (A Report
to the Water Foundation).
Clara, M., Strenn, B., Kreuzinger, N., 2004. Carbamazepine as
a possible anthropogenic marker in the aquatic environment:
investigations on the behaviour of Carbamazepine in
wastewater treatment and during groundwater infiltration.
Water Research 38 (4), 947e954.
Conn, K.E., Siegrist, R.L., Barber, L.B., Meyer, M.T., 2010. Fate of
trace organic compounds during vadose zone soil treatment in
an onsite wastewater system. Environmental Toxicology and
Chemistry 29 (2), 285e293.
Davidson, W.A., 1995. Hydrogeology and groundwater resources
of the Perth region, Western Australia, Bulletin. Western
Australia Geological Survey 142.
Dillon, P., 2005. Future management of aquifer recharge.
Hydrogeology Journal 13 (1), 313e316.
Drewes, J.E., Reinhard, M., Fox, P., 2003. Comparing microfiltration-
reverse osmosis and soil-aquifer treatment for indirect potable
reuse of water. Water Research 37 (15), 3612e3621.
Fox, P., Narayanaswamy, K., Genz, A., Drewes, J., 2001. Water
quality transformations during soil aquifer treatment at the
Mesa Northwest water reclamation plant, USA. Water Science
and Technology 43 (10), 343e350.
Gordon, C., Wall, K., Toze, S., O’Hara, G., 2002. Influence of
Conditions on the Survival of Enteric Viruses and Indicator
Organisms in Groundwater. In: Paper Presented at
Proceedings of the 4th International Symposium on Artificial
Recharge of Groundwater Isar-4-Management of Aquifer
Recharge for Sustainability. Balkema Publishers, Adelaide SA.
Lin, C.Y., Banin, A., 2005. Effect of long-term effluent recharge on
phosphate sorption by soils in a wastewater reclamation
plant. Water Air and Soil Pollution 164 (1e4), 257e273.
Lister, A., Regan, C., Van Zwol, J., Van Der Kraak, G., 2009.
Inhibition of egg production in zebrafish by fluoxetine and
municipal effluents: a mechanistic evaluation. Aquatic
Toxicology 95 (4), 320e329.
Montgomery-Brown, J., Drewes, J., Fox, P., Reinhard, M., 2003.
Behavior of alkylphenol polyethoxylate metabolites during
soil aquifer treatment. Water Research 37 (15), 3672e3681.
Patterson, B., Pearce, J., Spadek, T., Linge, K., Busetti, F.,
Shackleton, M., Furness, A., Blair, P., Heitz, A., 2009. Fate of
trace organics using laboratory column experiments - chapter
5771
2. In: Toze, S., Bekele, E. (Eds.), Determining the Requirements
for Managed Aquifer Recharge in Western Australia (A Report
to the Water Foundation).
Patterson, B.M., Grassi, M.E., Robertson, B.S., Davis, G.B., Smith, A.
J., McKinley, A.J., 2004. Use of polymer mats in series for
sequential reactive barrier remediation of ammonium-
contaminated groundwater: field evaluation. Environmental
Science & Technology 38 (24), 6846e6854.
Patterson, B.M., Shackleton, M., Furness, A.J., Bekele, E., Pearce, J.,
Linge, K.L., Busetti, F., Spadek, T., Toze, S., 2011. Behaviour and
fate of nine recycled water trace organics during managed
aquifer recharge in an aerobic aquifer. Journal of Contaminant
Hydrology 122 (1e4), 53e62.
Pyne, R.D.G., 2006. Groundwater Recharge and Wells: a Guide to
Aquifer Storage Recovery, second ed. CRC Press, Boca Raton,
Florida, 401 pp.
Quanrud, D., Hafer, J., Karpiscak, M., Zhang, H., Lansey, K.,
Arnold, R., 2003. Fate of organics during soil-aquifer
treatment: sustainability of removals in the field. Water
Research 37 (14), 3401e3411.
Robinson, B.W., Hitchen, G.J., Verrall, M.R. 2000, The
AutoGeoSEM: A programmable fully-automatic SEM for rapid
grain-counting and heavy mineral characterisation in
exploration, paper presented at International Mineralogical
Association Commission on Ore mineralogy, Proceeding of the
Modern Approaches to Ore and Environmental mineralogy,
Short Course, p. 71e74.
Rueedi, J., Cronin, A.A., Morris, B.L., 2009. Estimation of sewer leakage
to urban groundwater using depth-specific hydrochemistry.
Water and Environment Journal 23 (2), 134e144.
Saaristo, M., Craft, J.A., Lehtonen, K.K., Lindstrom, K., 2010. An
endocrine disrupting chemical changes courtship and parental
care in the sand goby. Aquatic Toxicology 97 (4), 285e292.
Scatena, M.C., Williamson, D.R. 1999, A potential role for artificial
recharge within the Perth region: a pre-feasibility study.
Center for Groundwater Studies report No. 84.
Schafer, A., Ustohal, P., Harms, H., Stauffer, F., Dracos, T.,
Zehnder, A., 1998. Transport of bacteria in unsaturated porous
media. Journal of Contaminant Hydrology 33 (1e2), 149e169.
Scheytt, T.J., Mersmann, P., Heberer, T., 2006. Mobility of
pharmaceuticals carbamazepine, diclofenac, ibuprofen, and
propyphenazone in miscible-displacement experiments.
Journal of Contaminant Hydrology 83 (1e2), 53e69.
Smith, A.J., Pollock, D.W., 2010. Artificial Recharge Potential of the
Perth Region Superficial Aquifer: Lake Preston to Moore River.
CSIRO (Water for a Healthy Country National Research
Flagship Report).
Sulleabhain, C.O., Gill, L.W., Misstear, B.D.R., Johnston, P.M., 2009.
Fate of endocrine-disrupting chemicals in percolating
domestic wastewater effluent. Water and Environment
Journal 23 (2), 110e118.
Tapsell, P., Newsome, D., Bastian, L., 2003. Origin of yellow sand
from Tamala limestone on the Swan Coastal Plain, Western
Australia. Australian Journal of Earth Sciences 50 (3), 331e342.
Tiehm, A., Schmidt, N., Stieber, M., Sacher, F., Wolf, L., Hoetzl, H.,
2011. Biodegradation of pharmaceutical compounds and their
occurrence in the Jordan Valley. Water Resources
Management 25 (4), 1195e1203.
Toze, S., Bekele, E., 2009. Determining Requirements for Managed
Aquifer Recharge in Western Australia. A Report to the Water
Foundation. CSIRO, Canberra, AU. http://www.csiro.au/files/
files/py7j.pdf.
Toze, S., Bekele, E., Page, D., Sidhu, J., Shackleton, M., 2010. Use of
static Quantitative Microbial Risk Assessment to determine
pathogen risks in an unconfined carbonate aquifer used for
Managed Aquifer Recharge. Water Research 44 (4), 1038e1049.
Toze, S., Hanna, J., 2002. The Survival Potential of Enteric
Pathogens in a Reclaimed Water ASR Project. In: Paper
5772 w a t e r r e s e a r c h 4 5 ( 2 0 1 1 ) 5 7 6 4 e5 7 7 2
Presented at Proceedings of the 4th International Symposium
on Artificial Recharge of Groundwater ISAR-4-Management of
Aquifer Recharge for Sustainability. Balkema Publishers,
Adelaide, SA, pp. 139e142.
Toze, S., Hanna, J., Smith, T., Edmonds, L., McCrow, A., 2004.
Determination of water quality improvements due to the
artificial recharge of treated effluent. In: Steenvoorden, J.,
Endreny, T. (Eds.), International Symposium on Wastewater
Re-Use and Groundwater Quality. Int Assoc Hydrological
Sciences, Wallingford, UK, pp. 53e60.
Turner, B., Binning, P., Stipp, S., 2005. Fluoride removal by calcite:
evidence for fluorite precipitation and surface adsorption.
Environmental Science & Technology 39 (24), 9561e9568.
Vanderzalm, J., Page, D., Barry, K., Dillon, P., 2010. A comparison
of the geochemical response to different managed aquifer
recharge operations for injection of urban stormwater in
a carbonate aquifer. Applied Geochemistry 25 (9), 1350e1360.
von Wandruszka, R., 2006. Phosphorus retention in
calcareous soils and the effect of organic matter on its
mobility. Geochemical Transactions 7 (6). doi:10.1186/1467-
4866-7-6.
Whelan, B.R., 1988. Disposal of septic tank effluent in calcareous
sand. Journal of Environmental Quality 17 (2), 272e277.
Whelan, B.R., Barrow, N.J., 1984. The movement of septic-tank
effluent through sandy soils near Perth .2. Movement of
phosphorus. Australian Journal of Soil Research 22 (3),
293e302.
Wolf, L., Held, I., Eiswirth, M., Hotzl, H., 2004. Impact of leaky
sewers on groundwater quality. Acta Hydrochimica Et
Hydrobiologica 32 (4e5), 361e373.
Zhang, J., Huang, X., Liu, C., Shi, H., Hu, H., 2005. Nitrogen removal
enhanced by intermittent operation in a subsurface
wastewater infiltration system. Ecological Engineering 25 (4),
419e428.