ArticlePDF Available

The impact of low-temperature seasonal aquifer thermal energy storage (SATES) systems on chlorinated solvent contaminated groundwater: Modeling of spreading and degradation

Authors:

Abstract and Figures

Groundwater systems are increasingly used for seasonal aquifer thermal energy storage (SATES) for periodic heating and cooling of buildings. Its use is hampered in contaminated aquifers because of the potential environmental risks associated with the spreading of contaminated groundwater, but positive side effects, such as enhanced contaminant remediation, might also occur. A first reactive transport study is presented to assess the effect of SATES on the fate of chlorinated solvents by means of scenario modeling, with emphasis on the effects of transient SATES pumping and applicable kinetic degradation regime. Temperature effects on physical, chemical, and biological reactions were excluded as calculations and initial simulations showed that the small temperature range commonly involved (ΔT<15°C) only caused minor effects. The results show that a significant decrease of the contaminant mass and (eventually) plume volume occurs when degradation is described as sediment-limited with a constant rate in space and time, provided that dense non-aqueous phase liquid (DNAPL) is absent. However, in the presence of DNAPL dissolution, particularly when the dissolved contaminant reaches SATES wells, a considerably larger contaminant plume is created, depending on the balance between DNAPL dissolution and mass removal by degradation. Under conditions where degradation is contaminant-limited and degradation rates depend on contaminant concentrations in the aquifer, a SATES system does not result in enhanced remediation of a contaminant plume. Although field data are lacking and existing regulatory constraints do not yet permit the application of SATES in contaminated aquifers, reactive transport modeling provides a means of assessing the risks of SATES application in contaminated aquifers. The results from this study are considered to be a first step in identifying the subsurface conditions under which SATES can be applied in a safe or even beneficial manner.
Content may be subject to copyright.
2009; Foster, 2001). Today, aquifers are increasingly being
used for aquifer storage and recovery (Dillon et al., 2006; Pyne,
2005), facilitated storm water infiltration (Ferguson, 1990),
CO
2
storage (Steeneveldt et al., 2006), and brine disposal
(Tsang et al., 2008) and are a target for remediation of
contamination (European Environment Agency, 2007). During
the past decade, a rising demand for sustainable energy
sources has led to intensified use of seasonal aquifer thermal
energy storage (SATES), a cost-effective energy technology in
support of ambitions for CO
2
-emission reductions. This
technology provides seasonal heating and cooling of buildings
by means of the alternating injection and recovery of heated
(injection during summer, recovery during winter) and cooled
(injection during winter, recovery during summer) ground-
water via wells in aquifers (Carotenuto et al., 1991; Edworthy
and Puri, 1986; Kim et al., 2010; Molz et al., 1978, 1979, 1981;
Tsang and Hopkins, 1982). Most of those SATES systems
are operated with only limited temperature differences
(ΔTb15 °C) between warm (b20 °C) and cold wells (~5 °C)
in shallow aquifers with an ambient groundwater temperature
of ~11 °C. An increasing number of SATES systems is reported
in European countries and elsewhere (Gao et al., 2009; Sanner
et al., 2003). In the Netherlands, for example, it is assumed that
SATES will be the largest form of groundwater usage by 2020,
pumping 1225 to 6300 million m
3
per year, thereby probably
exceeding the total drinking-, industrial-, and agricultural
groundwater extraction of 1500 million m
3
per year (Bonte et
al., 2011a).
Depending on their location, spatial density, and capacity,
the associated rise in groundwater use by these SATES
systems may cause interference between SATES systems
and groundwater extractions (Bonte et al., 2011a,b). SATES is
also likely to affect soil and groundwater contamination,
since the largest demands for heat and cold exist in
urbanized and industrial areas. The depth at which SATES
systems are typically operated (>10 to b250 m below the
water table) coincides with the depth at which groundwater
contaminant plumes of chlorinated hydrocarbons (CHCs)
are generally present, particularly tetrachloroethene (PCE)
and trichloroethene (TCE) and their daughter products
dichloroethene (DCE) and vinyl chloride (VC). Such CHC
plumes develop through the long-term dissolution of dense
non-aqueous phase liquids (DNAPLs) often used in dry
cleaning and metal degreasing processes and are one of the
most prevalent organic contaminations in urban groundwa-
ter (e.g., Bradley, 2000; Parker et al., 2003; Smidt and de Vos,
2004; Wiedemeier et al., 1999).
The presence of CHC plumes currently hampers the
development of SATES in urbanized areas in the Netherlands
(Slenders et al., 2010; Taskforce ATES, 2009). But besides
potential negative effects associated with pumping of contam-
inated groundwater, several potentially positive effects of
SATES systems on the transport and degradation of contam-
ination plumes have been suggested. These presume the
anticipated stimulation of contaminant degradation due to
the effect of periodically elevated temperatures and enhanced
mixing of microorganisms, nutrients, oxidants, reductants, and
contaminants within SATES systems (Slenders et al., 2010;
Verburg et al., 2010). Since temperature differences in current
SATES systems are relatively limited and pumped volumes are
large, fundamental processes like spreading of (diluted)
contaminants and enhanced DNAPL dissolution increasing
dissolved contaminant mass need special consideration.
However, they have so far not been evaluated in the
assessment of the impacts of SATES on CHC contaminated
aquifers and focus has been mainly on modeling of temper-
ature and carbonate precipitation in relation to well clogging
(e.g., Brons et al., 1991; Molz et al., 1983). This knowledge gap
in water quality development has been a severe constraint in
SATES application in many urban aquifers.
Here, we present a first modeling approach to study the
effects of alternating transient pumping by low-temperature
SATES systems on CHC contaminated aquifers. The aims of
this study are to:
assess the extent to which various factors (hydrogeology,
contaminant degradation kinetics and DNAPL dissolution)
can affect CHC plumes when SATES is applied;
evaluate the conditions under which SATES systems are
likely to have either positive or negative effects on CHC
plumes in the presence or absence of DNAPLs within the
zone of influence of SATES wells.
Temperature dependencies could be ignored due to the
limited temperature differences of current thermally bal-
anced SATES systems and were therefore not incorporated in
the model. This study underlines potential benefits and
identifies the risks of SATES application in CHC contaminated
aquifers, prior to the approval of field implementation. An
existing SATES system with therefore a hypothetical PCE
aquifer contamination was modeled with a customized
version of PHT3D.
2. Material and methods
2.1. Model components
In this study, the reactive multi-component transport
model for saturated porous media PHT3D Version 2
(Prommer and Post, 2010; Prommer et al., 2003) was used
for the simulation of groundwater flow, solute transport, and
degradation of contaminants (Fig. 1). This coupled model of
MODFLOW/MT3DMS (Harbough et al., 2000; Zheng and
Wang, 1999) and PHREEQC-2 (Parkhurst and Appelo, 1999)
enables the simulation of advection, dispersion, diffusion,
and almost any desired chemical reaction. In this study, the
processes of DNAPL dissolution and sequential degradation of
dissolved CHCs were included in a three-dimensional flow
field. The MT3DMS-based version of PHT3D does not allow
simulating the effects of density and viscosity changes due to
the effect of temperature differences in the SATES system on
groundwater flow. Therefore, a version based on SEAWAT
version 4 (Langevin et al., 2008) was used for this purpose,
which was coupled with PHT3D Version 2 in the same way as
was done previously by others (Mao et al., 2006; Post and
Prommer, 2007; Robinson et al., 2009). Temperature effects
were simulated initially, but did not affect the outcomes of
the initial stage of modeling. Therefore, temperature differ-
ences were not subsequently considered.
In order to realistically simulate the distribution of
injected and recovered water over several screened intervals
of the wells, as well as the mixing of water qualities in the
wells during recovery from several model layers by one well,
2K.G. Zuurbier et al. / Journal of Contaminant Hydrology 147 (2013) 113
the multi-node-well (MNW) Version 1 package (Halford et
al., 2002) was used. A modification was made to the code in
order to enable recirculation of the recovered mixed water in
the SATES system to the injection well. This modification
enabled recirculation for the MNW in the same way as for the
existing recirculation for a single-node well (Zheng, 2006).
2.2. Reaction network
2.2.1. Degradation of CHCs
In this study the full degradation sequence from PCE to
ethene as described by Wiedemeier et al. (1999) was
modeled. The kinetics of this process have been described
by various rate models, mainly zero-order, first-order with
respect to contaminant concentration, and Michaelis
Menten type (Bekins et al., 1998; Suarez and Rifai, 1999).
Depending on which factor limits the overall degradation
rate, the applicable type of kinetic rate expression for the
degradation of CHCs varies, thereby affecting overall degra-
dation within SATES systems. In our study we applied
zero-order degradation kinetics for the cases where degra-
dation rates were limited by the slow release of electron
donor from the sediment (e.g., McCarty, 1997; Pavlostathis
and Zhuang, 1993), i.e. by hydrolysis of sedimentary organic
matter (SOM: Fig. 2), typical for most natural aquifer
environments (Hartog et al., 2005; Jakobsen and Postma,
1994; van Helvoort et al., 2007). These simulations will be
referred to as sediment-limited(SL) model runs. In the
SL-scenario it is assumed that organic matter within the
model domain degrades at a constant rate in space and time,
independent of CHC concentrations. In contrast, for condi-
tions with sufficient available reactive dissolved organic
matter, degradation kinetics are described as first-order
with respect to contaminant concentrations (Suarez and
Rifai, 1999). These simulations will be referred to as
contaminant-limited(CL) model runs (Fig. 2).
2.2.2. Dissolution from DNAPL and adsorption
For the dissolution of PCE from the DNAPL in the aquifer,
the following rate-limited expression was used to describe
Fig. 1. Modeling scheme illustrating input data and model codes used to create simulation output. Mixed concentrations during recovery were calculated
(recovered mean concentration) and provided the input for the injection well.
Fig. 2. Controlling factors assumed for CHC degradation in this study. SOM is the sedimentary organic matter, C
DOM
is the concentration of dissolved organic
matter (M L
3
), t is time (T), λ
0
is the degradation constant (M L
3
T
1
), C
CHC
is the concentration of chlorinated hydrocarbon species considered (M L
3
) and
λ
1
is the degradation rate constant for first-order degradation (T
1
).
3K.G. Zuurbier et al. / Journal of Contaminant Hydrology 147 (2013) 113
the mass transfer rate between the immobile DNAPL phase
and the dissolved phase as (e.g., Hayden et al., 1992; Miller et
al., 1990; Prommer et al., 1999):
MDNAPL
PCE
t¼ωPCE Csat
PCECPCE
! ð1Þ
where M
PCE
DNAPL
is the mass of DNAPL PCE per unit pore volume
(M L
3
), ω
PCE
is a lumped rate-transfer coefficient (T
1
), C
PCE
sat
is the solubility of PCE (M L
3
) and C
PCE
is the concentration of
PCE in the groundwater (M L
3
). To enhance dissolution and
thereby emphasize the effect of the presence of the DNAPL, no
permeability decrease by the presence of the DNAPL was
modeled, while a relatively high rate-transfer coefficient
(ω
PCE
=1) was chosen, which is in line with Park and Parker
(2005), resulting in equilibrium dissolution at solubility
concentrations. Adsorption of dissolved CHC species was
modeled using a linear sorption isotherm.
2.3. Case study
As the application of SATES in contaminated aquifers is
restricted to date, no field data exist that can be used for the
purpose of this study. The well-monitored existing SATES
system Uithofin an uncontaminated aquifer in the city of
Utrecht (The Netherlands) was therefore used as the basis for
the simulations (Figs. 3 and 4). The local hydrogeology and
one pair of wells of the SATES system were represented in the
model. A fictitious contaminant source was added in the
model by placing a DNAPL at the base of the aquifer, 10 m
from one of the SATES wells (Fig. 4,Section 2.5).
2.3.1. Hydrogeological setting
The hydrogeological schematization of the field site
was derived from the regional geological model REGISII.1
(TNO-NITG, 2009). At the Uithoflocation (Figs. 3 and 4),
the subsurface consists of unconsolidated deltaic and fluvial
sediments forming alternating sand and/or gravel layers
(aquifers) and clay layers (aquitards). The SATES system is
injecting cold water in the western well (well Cin Fig. 4) and
warm water in the eastern well (well W). This ~47 m thick
aquifer with a transmissivity of 1320 m
2
/d is covered by a
2.8 m thick Holocene clay layer (K
v
=0.01 m/d) and underlain
by a 6 m thick aquitard (K
v
=0.005 m/d). In this aquifer, the
regional groundwater flow is determined by topographically
Fig. 3. Location and hydrogeological setting of the Uithof case study area. I Holocene clay cover; II aquifer 1; III aquitard 1; IV aquifer 2; V local clay
layer; VI aquitard 2; VII ice pushed ridge; VIII SATES system Uithof. Local hydrogeology based on the REGIS II.1 hydrogeological model (TNO-NITG, 2009).
Well separation not to scale.
4K.G. Zuurbier et al. / Journal of Contaminant Hydrology 147 (2013) 113
elevated ice pushed ridges in the northeast of the study area
(Fig. 3), resulting in a southwestern flow of ~7 m/year at the
SATES location. This estimate is based on a measured hydraulic
gradient of 0.00024 m/m (Faneca-Sànchez et al., 2010), a
porosity of 0.35, and a hydraulic conductivity of 28 m/d.
2.3.2. SATES conguration
The wells of the SATES system are screened across two
stratigraphic sandy sections of high permeability within
aquifer 1 which are separated by a sandy clay layer (Fig. 4).
The weekly injected volumes in these layers were registered
in 2009 (Fig. 5), of which the last 4 weeks were prefixed to
the front of the data set in order to start with a full cold cycle.
A total of 147,951 m
3
of groundwater was pumped from the
warm well to the cold well during winter, while 132,232 m
3
was pumped from the cold to the warm well during summer.
Due to the climate control requirements of the building, heat
demand in winter was of shorter duration but more intense
than the cold demand in summer, resulting in this volumetric
imbalance.
2.4. Model set-up
2.4.1. Grid design
The local hydrogeology and SATES configuration in Fig. 4
were incorporated into the PHT3D model. For this model a fine
vertical discretization (z= 0.3 m) was used where the DNAPL
was present (at the base of aquifer 1) to allow for steep CHC
concentration gradients in this zone. Since larger gradients of
the CHC concentrations and the temperature were also
expected near both SATES wells, refinement was also made in
the grid near the wells (x=3 m, y=0.5 m), whereas the
cells away from the DNAPL and the SATES wells were enlarged
stepwise to a maximum horizontal size of 120 m× 80 m.
The total model extent was 1760 m (westeast)×1323 m
(northsouth), and the minimum distance between the wells
and the nearest edge of the model was 660 m, which was
sufficient to prevent boundary effects from affecting the
simulated flow regime and to prevent any contaminant mass
from leaving the system.
2.4.2. Groundwater ow and transport parameters
The values of Table 1 were used for the modeling of
groundwater flow and solute transport. The third-order
total-variation-diminishing (TVD) scheme (Leonard, 1988)
was used since it is mass-conservative and provided the most
stable breakthrough curves at the wells during test runs. The
distribution coefficients for the linear isotherm sorption were
based on the K
oc
(Karickhoff, 1981) for each species and an
organic carbon fraction (f
oc
) of 0.005. This resulted in
retardation factors of 7.2, 2.5, 2.2 and 1.04 in aquifer 1 for
PCE, TCE, DCE and VC, respectively. The solubility of PCE was
set to 1.3 mmol/L (215.6 mg L
1
;Montgomery, 2000).
Regional flow at the location of the SATES system was
simulated by specifying interpolated regional hydraulic heads
(Faneca-Sànchez et al., 2010) as constant head boundaries at
the edges of the model. These interpolated heads were also
used for the initial hydraulic heads. Preliminary simulations
Fig. 4. Local hydrogeology and SATES configuration. In winter (situation in this figure), cold water is being injected at well Cand warm water is recovered at well
W. In summer, the pumping directions are reversed.
5K.G. Zuurbier et al. / Journal of Contaminant Hydrology 147 (2013) 113
showed that groundwater flow was not significantly
influenced in the temperature range of the SATES system
simulated (8 °C in the cold bubble to 16 °C in the warm
bubble), and therefore the density and viscosity corrections
were not performed in the final simulations in order to
constrain model runtimes. This approach is supported by Ma
and Zheng (2010), who demonstrated that temperature
effects on fluid flow parameters may be neglected when the
maximum temperature difference is within 15 °C.
A total of 988 stress periods was used, each lasting 7 or
8 d (51 and 1 stress period(s) per year, respectively),
resulting in a total period of 19 years. Within each stress
period, the reaction time step (i.e., a call to PHREEQC using
the concentrations in each cell after transport) was 1 d. The
daily pumped volumes varied per stress period, based on the
measured pumping volumes (Fig. 5).
2.4.3. Degradation rate constants
For this study it was assumed that the CHC degradation
rate is either (i) zero-order sediment-limited (SL) or
(ii) first-order with respect to contaminant concentrations
(contaminant limited: CL). The selected constants for
first-order degradation scenarios (Table 2) are consistent
with those of Suarez and Rifai (1999) and Van Breukelen et
al. (2005). Zero-order degradation rates were less available
but were scaled to the first-order degradation rates. Based on
the degradation constants used, first-order PCE degradation
rates equal zero-order degradation rates at a PCE concentra-
tion of 0.37 mg L
1
, above which first-order degradation
rates are higher. Reaction rates are known to increase
exponentially with temperature (Arrhenius, 1889), and this
has also been shown for microbial dechlorination rates (e.g.,
Friis et al., 2007). However, for a thermally balanced SATES
system, the temperature increase for groundwater in the
warm bubble is balanced by the temperature decrease in the
cold bubble. The expected rate increase in the warm bubble is
therefore counteracted by a rate decrease in the cold bubble.
Although there will be a net increase of the overall rate due to
the exponential dependence of rates on temperature, this
effect is very small (b1%) for the temperature differences
(ΔTb15 °C) between warm and cold bubbles under which
SATES systems are generally operated, as calculated using the
Arrhenius equation (Hartog, 2011). For this reason, the
temperature dependency of degradation rate constants was
neglected in the model.
2.5. Contamination scenarios
A modeled dissolved PCE plume, which was generated
during 40 years of DNAPL dissolution without further PCE
degradation, was used as the initial contaminant condition for
the four SATES scenarios (Table 3). DNAPL is absent in
Fig. 5. Daily mean air temperature at the weather station De Bilt (~2 km
from the UithofSATES system) of the Royal Netherlands Meteorological
Institute (KNMI, 2011) and corresponding injection rates. Week 1 represents
the first week of December, which marked the start of the cold period when
heating was required and cold water was injected.
Table 1
Set of parameters used in model.
Flow and physical transport parameters Value Unit
Model dimensions l
x
×l
y
×l
z
1760× 1323× 135 m
Discretisation x3120 m
Discretisation y0.580 m
Discretisation z0.379 m
Horizontal hydraulic conductivity K
h
0.135 m d
1
Vertical hydraulic conductivity K
v
0.0117.5 m d
1
Effective porosity η0.35
Specific yield 0.25
Specific storage 0.0001 m
1
Longitudinal dispersivity α
t
1 m
Horizontal transversal dispersivity α
T(h)
0.1 m
Vertical transversal dispersivity α
T(v)
0.01 m
Molecular diffusion coefficient D
m
(PCE/TCE/DCE/VC)
3.76/4.06/4.32/5.79
×10
5
m
2
d
1
Distribution coefficient K
d
(PCE/TCE/DCE/VC)
1.26/ 0.32/ 0.24/ 0.01
×10
3
m
3
kg
1
Bulk density ρ
(clay/sandy clay/sand)
1430/ 1573/ 1716 kg m
3
PCE: perchloroethene, TCE: trichloroethene, DCE: dichloroethene, VC: vinyl
chloride.
Table 2
Degradation rate constants.
Species Sediment-limited (SL)
Zero-order rate
constants
(10
4
mg L
1
d
1
)
Contaminant-limited (CL)
First-order degradation
rate constants
(10
3
d
1
)
PCE 1.86 0.5
TCE 1.74 0.5
DCE 1.86 0.25
Vinyl chloride 0.11 0.125
6K.G. Zuurbier et al. / Journal of Contaminant Hydrology 147 (2013) 113
scenarios 1 and 2. In scenarios 3 and 4, a fictitious (PCE)
DNAPL initial source of 630 L (3 × 0.3 m; 1022 kg) was
positioned at the base of aquifer 1 (Fig. 4). Each of these four
model scenarios considered different combinations of specific
factors: the SATES system, occurrence of CHC degradation, and
presence of DNAPL (Table 3). All scenarios were performed for
SL- as well as CL-limited degradation conditions.
2.6. Model output
The development of the PCE plume volume and total mass
(sum of the dissolved and adsorbed mass) in the aquifer and
remaining DNAPL mass were monitored for each scenario
based on the mass budgets reported by PHT3D. In addition,
the total volume of groundwater exceeding Dutch national
groundwater quality criteria was calculated based on the
modeled concentrations, porosity, and the volume of each
grid cell. The desired target contaminant level (TCL) for PCE is
0.01 μg/L and the maximum contaminant level (MCL, above
which remediation may be required) is 40 μg/L (VROM,
2009).
3. Results
The results are described for the scenarios presented in
Table 3. All scenarios started with a PCE plume developed from
the DNAPL near well W with a total plume mass of 29 kg. In
this initial PCE plume, 13.0 Mm
3
of groundwater exceeded the
national target contaminant level (TCL; 0.01 μg/L PCE), of
which 2.4 Mm
3
also exceeded the maximum contaminant
level (MCL; 40 μg/L PCE).
3.1. Scenario 1: reference scenario: plume with degradation
For the sediment-limited degradation (SL, zero-order)
scenario, the combined effect of degradation and dispersion
led to a decrease of more than 0.6 Mm
3
of the plume volume
in which PCE concentrations exceeded the TCL after 19 years.
In this plume, slightly less than 0.6 Mm
3
also exceeded the
MCL (Fig. 6A). Degradation caused the total PCE plume mass
to decrease by 32.5% (Fig. 7A). These results indicate that
without the SATES system the contaminant plume would
remain present in the aquifer for over several decades.
For the contaminant-limited degradation scenario (CL, first-
order), the decrease of PCE plume mass was 94.6% (Fig. 7B).
Despite this larger overall degradation than in the SL case, the
groundwater volumes exceeding the TCL and MCL for PCE were
respectively larger and comparable to SL degradation (Fig. 6B).
Despite the significant plume mass reduction, the decreasing
degradation rates with decreasing concentrations for this case
mean that several decades are still required to meet the
regulatory concentration limits.
Table 3
Scenarios analyzed in the case study.
Scenario Characteristics Degradation SATES DNAPL
S-1 Degradation present, no SATES
system and no DNAPL
(reference scenario)
X– –
S-2 Addition of SATES system in
absence of DNAPL
X X
S-3 SATES and DNAPL present, no
degradation occurring
X X
S-4 DNAPL and SATES, present,
degradation occurring
(combination scenario)
X X X
Fig. 6. Total PCE plume volume (top values indicate C
PCE
>TCL) for each scenario in case of A) sediment-limited degradation and B) contaminant-limited
degradation. Bars give the volumes only exceeding the TCL (0.01 bC
PCE
b40 μg/L) and volumes exceeding the MCL (C
PCE
>40 μg/L) after 19 years.
7K.G. Zuurbier et al. / Journal of Contaminant Hydrology 147 (2013) 113
3.2. Scenario 2: plume capture within the SATES system with
degradation
In the case of SL degradation, the introduction of the SATES
system resulted in a rapid decrease of the PCE mass in the
plume. A nearly complete reduction of PCE plume mass of
99.1% was achieved after 19 years, reflecting the effect of
increased degradation due to spreading of the contaminants
across a largersediment volume. A concomitant decrease in the
volume of groundwater exceeding the regulatory limits for PCE
was observed: only 0.1 Mm
3
of groundwater still exceeded the
MCL, and an even smaller volume exceeded only the TCL.
Daughter products were largely absent after 19 years in this
scenario. VC for instance, reached its peak in plume mass (3 g)
at t=78 d, which decreased to a total mass of 9 mg (Fig. 8A)
and a maximum concentration of 0.07 μg/L, after 19 years.
Contrary to the case with SL degradation, the PCE plume
mass in the case of CL degradation was equal to the situation
without the SATES system (Fig. 7B, scenarios 1 and 2). This
was because the first-order rate expression with respect to
contaminant concentration did not result in an increased PCE
mass removal by spreading over a larger aquifer volume. The
increased contaminated volume was accompanied by de-
creased degradation rates due to the proportional dilution of
concentrations. Moreover, with the first-order degradation
kinetics for the CL case, concentrations in the center of the
plume were lowered more rapidly due to a high degradation
rate, whereas in the reinjected (and therefore diluted) SATES
water and in groundwater at the plume fringe low degradation
rates led tolonger persistence of concentrations above the TCL.
The latter is shown by the PCE plume volume exceeding TCL
especially in scenario 2 (Fig. 6B), which was more than 20,000
Fig. 7. Plume mass development in the PCE plume per scenario during A) sediment-limited (SL) degradation and B) contaminant-limited (CL) degradation.
Seasonal changes in plume mass are caused by SATES pumping scheme, further explained in Section 3.3. The vertical arrows indicate the direction in which the
curves for scenario 4 would shift if the rate constant would be increased or decreased.
Fig. 8. Plume mass development in the VC plume per scenario during A) sediment-limited (SL) degradation and B) contaminant-limited (CL) degradation.
8K.G. Zuurbier et al. / Journal of Contaminant Hydrology 147 (2013) 113
times larger in this case, compared with the SL degradation
case. VC reached its maximum plume mass of 390 g at the end
of the simulation (Fig. 8B), which is a significantly larger mass
compared to the SL case.
3.3. Scenario 3: presence of DNAPL in absence of degradation
within the SATES system
In this scenario, an increase in PCE plume mass was
caused by enhanced dissolution of the DNAPL. This is due to
the dilution of PCE by mixing in the well during extraction
phases, which allows additional DNAPL to dissolve when the
water with diluted PCE concentrations passes the DNAPL
during the subsequent injection phases. A yearly fluctuation,
caused by the alternating recovery and injection and varying
pumping rates, is superimposed on a nearly linear increase in
PCE plume mass (Fig. 7A,B; scenario 3). This linear increase
indicates that the dilution that occurred by dispersion and
particularly by mixing in the abstraction well was sufficient
to maintain large concentration differences (C
PCE
sat
C
PCE
) near
the DNAPL, which was confirmed by concentration observa-
tions in the model. After 19 years, 560 kg of PCE was found in
the groundwater, present in a groundwater volume of
439.7 Mm
3
exceeding the TCL, and a volume of 254.5 Mm
3
even exceeding the MCL (Fig. 9). This indicates a large
increase in contaminated aquifer volume, compared to the
same scenario without SATES introduced, in which 20.2 Mm
3
and 4.8 Mm
3
exceeded the TCL and MCL for PCE, respectively.
On the other hand, only 48% of the original DNAPL mass
remained, which means that almost 532 kg of PCE was
dissolved, indicating a mean mass discharge from the DNAPL
of 76.6 g/d. Without SATES, an additional PCE mass of only
14 kg would have been dissolved.
A variation on this scenario was modeled by placing the
DNAPL on the edge of the capture zone, 60 m from the
warm well, where increased flow velocities were limited. In
such a case, a minor increase of DNAPL dissolution was found
(5.13 g/d) compared with a situation without SATES (1.95 g/d),
but this increase was significantly lower compared with the
standard scenario 3. Since no contaminants entered the well of
the SATES system in this alternative setting, the role of the
SATES system in plume dilution was limited.
3.4. Scenario 4: combined effect of enhanced DNAPL dissolution
and degradation within the SATES system
As expected, the inclusion of degradation in this scenario
4 resulted in a significantly smaller increase in both the
contaminated groundwater volume and dissolved mass of
PCE compared to scenario 3. The DNAPL mass remaining after
19 years in case of SL degradation was slightly less compared
to scenario 3. In the plume, a groundwater volume of
46.7 Mm
3
exceeded the TCL for PCE, of which 24.5 Mm
3
exceeded the MCL. For comparison, if no SATES had been
introduced, this was only 0.7 Mm
3
(>TCL) and little more
than 0.6 Mm
3
(>MCL).
In Fig. 7A it is shown that the PCE mass increased to 95 kg
in 19 years, whereas 32 kg was reached in a situation
without the SATES system. The rate of PCE mass increase
becomes smaller with time for SL degradation. It is expected
that, ultimately, the mass of PCE would remain approximate-
ly constant in time as long as the DNAPL source remains, with
superimposed seasonal variations caused by varying DNAPL
dissolution in response to alternating recovery and injection
and variable pumping rates. In this situation, the mass of PCE
dissolved each year equals the mass of PCE removed by
degradation. The increased degradation capacity is due to the
increased contaminated aquifer volume. Therefore, the level
of plume mass stabilization depends on the zero-order rate
constant and the spatial extent of the plume.
As with SL degradation, the total mass of PCE stabilized
over time in the CL degradation case (Fig. 7B) as the mass
dissolved from the DNAPL each year approached the mass
removed by degradation. In this case, this can be explained
by the increase in PCE plume mass, which controls the overall
degradation. The level of stabilization depends solely on the
degradation rate constant. Total dissolved PCE mass in the
aquifer increased to 151 kg, but would have decreased to
6 kg if no SATES had been introduced. An aquifer volume of
560.9 Mm
3
exceeded the TCL and 103.6 Mm
3
exceeded the
A B
Fig. 9. Plan view distribution of PCE concentrations exceeding the TCL (6 ×10
11
mmol/L) and MCL (2.4×10
7
mmol/L) in scenario 3 at the bottom of aquifer 1.
The cold well is indicated by C, the warm well by W. The end of the first cycle at t =1 year (A) and the end of the last cycle (t =19 years, B) is shown.
9K.G. Zuurbier et al. / Journal of Contaminant Hydrology 147 (2013) 113
MCL for PCE. This would have been 12.3 Mm
3
(>TCL) and
0.6 Mm
3
(>MCL) if no SATES was introduced.
4. Discussion
The results of this study provide important insights for the
evaluation of future contaminant conditions at CHC contam-
inated SATES sites. With the relatively simple descriptions of
the DNAPL dissolution and degradation processes, the
disparate results from the different scenarios clearly illustrate
the interaction between a SATES well pair at a (CHC)
contaminated site on plume concentrations and, total
plume volume. Evidently, specific outcomes such as plume
size and mass development will depend on site-specific
conditions such as rate constants and limiting factors for
degradation, degree of mixing within the SATES system and
the presence and location of DNAPL with respect to the SATES
systems. However, clear generic trends and controlling
processes were identified in this study, of which the most
important ones will be discussed below.
4.1. How degradation kinetics affect contaminant degradation
in SATES systems
In a SL case, a higher degree of solute spreading caused by
the SATES system results in increased degradation marked by
contaminant mass removal, but this beneficial effect is absent
in a case with CL degradation. Furthermore, it is found that
SATES can cause a dramatically expanding PCE plume in
which concentrations exceed the TCL for the CL degradation
case compared to the reference scenario (no SATES), albeit
that the plume volume exceeding MCL is reduced. In contrast,
the plume volumes exceeding both TCL and MCL can
decrease for the SL degradation case, provided that zero-
order rate constants are sufficiently high.
For the SATES configuration and contaminant situation
considered, a dissolvedPCE plume degraded almost completely
in 19 years for the SL case (scenario 2). Therefore, the
degradation rate constants obtained from literature are high
enough for clean-up in case of SL and absence of DNAPL
(scenario 2). In case of DNAPL presence, a steady state situation
was attained for both degradation regimes within decades
(scenario 4, SL and CL).
For the purpose of this study it was assumed that the
redox conditions for full CHC degradation were present
throughout the modeled domain and the SATES system did
not affect those conditions. However, SATES is applied in
aquifers having subtle redox gradients within overall anoxic
conditions (Bonte et al., 2011b). In such aquifers, a stratifi-
cation of degradation rates can be expected as particular
CHCs are more easily degraded in specific redox conditions.
The natural redox stratification may become disturbed with
the introduction of SATES, leading to mixing of methane,
sulfate, and/or nitrate with resultant variability in PCE, TCE,
DCE, and VC degradation (e.g., Bradley, 2000; Bradley and
Chapelle, 2011; Suarez and Rifai, 1999; Wiedemeier et al.,
1999). Although temperature effects on degradation rates
could be ignored for the conditions of thermally balanced
SATES systems with a small temperature difference in this
study, these should be taken into account when studying
thermally non-balanced SATES systems with a considerably
larger temperature difference.
4.2. Contaminant dilution and enhanced degradation
When dissolved CHC plumes are present in a homogeneous
target aquifer and DNAPL is absent, the dilution caused by
mixing in the well is controlled by the proportion of the plume
relative to the aquifer thickness, transmissivity stratification
and the proportion of the capture zone of the well that is
contaminated (Einarson and Mackay, 2001). Consequently,
dilution increases when the ratio of plume thickness over
screen length decreases, or when ambient groundwater flow
velocities are higher, as the reinjected CHC plume will partly
move outside the upstream orientated capture zone of the
well. This is followed by mixing of the recoverable part of the
plume with upstream (uncontaminated) groundwater (Bear
and Jacobs, 1965; Ceric and Haitjema, 2005). As illustrated in
the modeling scenarios, the mixing effect increases overall
degradation when the degradation rate is sediment-limited.
This is due to the SATES system, which reinjects mixed
groundwater with diluted contaminant concentrations and
causes enhanced spreading of contaminants in the aquifer,
exposing the contaminants to a greater portion of degradation
capacity present in the sediment. As a consequence, CHC
concentrations will not only decrease by dilution, but also by
the enhanced mass removal due to spreading by SATES, which
is larger in case of stronger dilution. This study shows how
relevant this coupled dilution and spreading is for the
development of the CHC plume.
4.3. Inuence of DNAPL presence and mass ux
In this study, scenarios with and without a DNAPL source
are evaluated. The equilibrium-controlled DNAPL-dissolution
rate was a function of the difference between its solubility
and the groundwater concentration. Therefore, the degree to
which SATES causes mixing of the contaminant plume with
unpolluted groundwater, and thus maintains low concentra-
tions, will largely control enhanced DNAPL dissolution. This
study shows that when DNAPL is present outside the capture
zone of the SATES system, the influence of the SATES system
on DNAPL dissolution becomes strongly reduced even though
enhanced DNAPL dissolution still occurs due to increased
groundwater flow velocities. Neglecting dispersion and
background lateral flow, mixing of CHC dissolved from
DNAPL by SATES can only occur when the DNAPL is within
a radius r
1
, which is controlled by the radius of the capture
zone (r
c
) and the retardation factor R(Eq. (2)):
r1¼ffiffiffiffi
r2
c
R
s:ð2Þ
This implies that the most mobile component dissolved
from DNAPL or generated by degradation in the plume having
the lowest retardation factor (in this case VC) determines
whether dilution of any CHC species in the SATES wells occurs.
The use of a high lumped rate-transfer coefficient (ω
PCE
) in
our simulation of DNAPL dissolution represents the upper-
limit of enhanced mass flux compared to the reference
10 K.G. Zuurbier et al. / Journal of Contaminant Hydrology 147 (2013) 113
condition with no SATES system. However, due to the
configuration and small size of the DNAPL in our model, the
increased mass discharge (up to 76.6 g/d) due to the presence
of the SATES system can be considered intermediate in
comparison to mass discharge at actual DNAPL contaminated
sites (Newell et al., 2011). The model is highlighting the
worst-case potential relevance of DNAPLs near the SATES
wells for the given DNAPL volume and geometry. In case
dissolution from the DNAPL is mass-transfer limited (Miller et
al., 1990; Park and Parker, 2005) within the SATES affected
DNAPL source zone, the increased groundwater flow velocities
in SATES systems will not result in equally increased mass
fluxes (Parker and Park, 2004).
Depending on the source strength, biodegradation capac-
ity, and the extent to which mass-transfer limited dissolution
occurs, the steady-state contaminant situation found in this
study (Fig. 7A,B) may be established more rapidly, resulting
in a smaller aquifer volume becoming contaminated. How-
ever, as DNAPL dissolution will be slower in time, the CHC
plume will persist longer, albeit at a smaller scale.
4.4. Risk-assessment of SATES as a remediation technology
Performance of different remediation technologies is
evaluated based on multiple methods and standards. Rele-
vant criteria for this study are provided by Rao et al. (2001).
Rao et al. (2001) gauged the performance of remediation by
whether (a) the spatial extent of the existing dissolved
plume was stable or decreased; (b) the total contaminant
mass within the plume was constant or diminishing; (c) both
the average concentration and the range in concentrations
were diminishing; and (d) contaminant fluxes decreased at
successive control planes along the dissolved plume. In order
to meet condition (b), the contaminant flux leaving the
source zone must be equal to or less than the potential to
degrade the contaminant in the plume.
In our study, the reference scenario 1 failed to meet
condition (a) as the starting plume was not in equilibrium
after 19 years (Table 4). Only the SATES scenario without
DNAPL (scenario 2) met conditions (b), (c), and (d), but
failed to achieve (a) on the short-term. For SL degradation,
condition (a) was met only over the longer term. Therefore,
this study shows that for a dissolved plume with depleted or
removed DNAPL, beneficial effects of SATES are achieved only
for SL degradation and if a temporary plume size increase is
acceptable.
In scenarios 3 and 4, none of the conditions were met in the
short term, although in scenario 4 conditions (a) and (b) were
met eventually. Although the mixing by SATES lowered the
average concentration, the range in concentrations remained,
so condition (c) could not be satisfied. Where DNAPL occurs,
therefore, this study shows that SATES may only be considered
a successful remediation scheme if (1) DNAPL source-zone
depletion is a remediation target and (2) a continuous larger
contaminated aquifer volume is accepted in the long-term.
This study shows the relevance of (information on)
DNAPL presence, and its mass and distribution in the target
aquifer. Further effects of SATES on the contaminant
distribution will depend on the degradation regime (Section
4.1) and is directly related to the hydrogeological setting and
the configuration of the SATES wells relative to (potential)
DNAPL zones (Section 4.2). This study also highlights the
importance of long-term monitoring during field application
in order to evaluate its impacts, as an increased groundwater
contamination may precede ultimate plume attenuation.
5. Conclusions
This is the first comprehensive study of the interaction
between SATES pumping and CHC in a groundwater system.
Using a reactive transport modeling approach, conditions that
either promote or negate remediating effects of SATES were
demonstrated. Only if the degradation of contaminants is
controlled by sediment reactivity, the spreading of contami-
nants strongly enhances contaminant mass removal compared
with a situation without SATES, provided that (redox)
conditions for degradation do not deteriorate due to the
SATES system. However, in case degradation is controlled by
the contaminant concentration, a larger plume is generated
without any enhanced mass removal. This study also shows
that under conditions of partially increased DNAPL dissolution,
particularly near SATES wells, a potentially long-lasting
increase in both plume volume and mass may be anticipated.
The degree to which these SATES-influenced plumes reach
spatially stable conditions will depend strongly on the mass
discharge from the DNAPL source, in situ geochemistry, and
degradation rates in the aquifer. Overall, the interaction
between the design and operational conditions of SATES
Table 4
Risk-assessment SATES in contaminated aquifers according to Rao et al. (2001) for the sediment-limited (SL) and contaminant-limited (CL) degradation cases.
Scenario Degradation Condition a Condition b Condition c Condition d
Spatial extent Contaminant mass Average and range of concentration Contaminant fluxes
S-1 SL + + +
CL + + +
S-2 SL +/+ + +
CL + + +
S-3 − −
S-4 SL +/+/− −
CL +/+/− −
+: condition is met.
+/: condition is met only on the long term.
-: condition is not met.
11K.G. Zuurbier et al. / Journal of Contaminant Hydrology 147 (2013) 113
systems and site-specific aspects of hydrogeological, biogeo-
chemical, and contaminant characteristics within an aquifer,
determines the quantitative effect of SATES systems on
contaminant degradation and spreading. In the assessment of
further potential benefits or risks of SATES in CHC contami-
nated aquifers, insight into the occurrence of biodegradation
under varying temperatures and mixed redox zones, as well as
the presence and location of DNAPL within the aquifer is
crucial.
Acknowledgments
This study was performed within the Dutch national
research project Meer met Bodemenergie(MMB). The editor
and four anonymous reviewers are thanked for their thought-
ful comments that helped to improve this manuscript.
References
Arrhenius, S., 1889. Uber die Reaktiongeschwindigkeit bei der Inversion von
Rohrzucker durch sauren. Zeitschrift fur Physik Chemique 4, 226248.
Bear, J., Jacobs, M., 1965. On the movement of water bodies injected into
aquifers. Journal of Hydrology 3 (1), 3757.
Bekins, B.A., Warren, E., Godsy, E.M., 1998. A comparison of zero-order, first-
order, and monod biotransformation models. Ground Water 36 (2),
261268.
Bonte, M., Stuyfzand, P.J., Van den Berg, G.A., Hijnen, W.A.M., 2011a. Effects
of aquifer thermal energy storage on groundwater quality and the
consequences for drinking water production: a case study from the
Netherlands. Water Science and Technology 63, 19221931.
Bonte, M., Stuyfzand, P.J., Hulsman, A., Van Beelen, P., 2011b. Underground
thermal energy storage: environmental risks and policy developments
in the Netherlands and EU. Ecology and Society 16 (1), 22 (online).
Bradley, P.M., 2000. Microbial degradation of chloroethenes in groundwater
systems. Hydrogeology Journal 8 (1), 104111.
Bradley, P.M., Chapelle, F.H., 2011. Microbial mineralization of dichloroethene
and vinyl chloride under hypoxic conditions. Ground Water Monitoring &
Remediation 31 (4), 3949.
Brons, H.J., Griffioen, J., Appelo, C.A.J., Zehnder, A.J.B., 1991. (Bio)geochemical
reactions in aquifer material from a thermal energy storage site. Water
Research 25 (6), 729736.
Carotenuto, A., Fucci, F., La Fianza, G., Reale, F., 1991. Physical model and
demonstration of an aquifer thermal energy store. Heat Recovery
Systems and CHP 11 (23), 169180.
Ceric, A., Haitjema, H., 2005. On using simple time-of-travel capture zone
delineation methods. Ground Water 43 (3), 408412.
Dillon, P., Pavelic, P., Toze, S., Rinck-Pfeiffer, S., Martin, R., Knapton, A.,
Pidsley, D., 2006. Role of aquifer storage in water reuse. Desalination 188
(13), 123134.
Edworthy, K.J., Puri, S., 1986. Groundwater and aquifers; an overview of
exoticuses. Quarterly Journal of Engineering Geology & Hydrogeology
19 (2), 8795.
Einarson, M.D., Mackay, D.M., 2001. Predicting impacts of groundwater
contamination. Environmental Science & Technology 35 (3), 66A73A.
European Environment Agency, 2007. State of the Environment. EEA,
Copenhagen.
Evans, D., Stephenson, M., Shaw, R., 2009. The present and future use of
landbelow ground. Land Use Policy 26 (Supplement 1), S302S316.
Faneca-Sànchez, M., Hermeler, H., Huibregtse, E., Mayer, M., Meindertsma,
W., Van Oort, S., Sommer, W., Sun, W., Vermeer, M., Hao, W., 2010. A
Method for Sustainable Water Management Test Case De Uithof
TNO-034-UT-2010-01059. TNO, Utrecht.
Ferguson, B.K., 1990. Urban storm water infiltration. Journal of Soil and
Water Conservation 45 (6), 605609.
Foster, S.S.D., 2001. The interdependence of groundwater and urbanisation
in rapidly developing cities. Urban Water 3 (3), 185192.
Friis, A.K., Heimann, A.C., Jakobsen, R., Albrechtsen, H.-J., Cox, E., Bjerg, P.L.,
2007. Temperature dependence of anaerobic TCE-dechlorination in a
highly enriched Dehalococcoides-containing culture. Water Research 41
(2), 355364.
Gao, Q., Li, M., Yu, M., Spitler, J.D., Yan, Y.Y., 2009. Review of development
from GSHP to UTES in China and other countries. Renewable and
Sustainable Energy Reviews 13 (67), 13831394.
Halford, K.J., Hanson, R.T., Santa Clara Valley Water, D., 2002. User Guide for
the Drawdown-Limited, Multi-node Well (MNW) Package for the US
GeologicalSurvey's Modular Three-dimensional Finite-difference Ground-
water Flow Model,Versions MODFLOW-96and MODFLOW-2000.US Dept.
of the Interior, US Geological Survey.
Harbough, A.W., Banta, E.R., Hill, M.C., McDonald, M.G., 2000. Modflow-2000,
the U.S. Geological Survey Modular Groundwater Model User Guide to
Modularization Concepts and the Groundwater Flow Process. Open-File
Report 0092. U.S. Geological Survey.
Hartog, N., 2011. Anticipated temperature effects on biogeochemical
reaction rates in seasonal aquifer thermal energy storage (ATES)
systems: an evaluation using the Arrhenius equation. 1ste Nationaal
Congres Bodemenergie,Maarssen (http://www.nielshartog.nl/articles/
nhartog_DutchGeothermalCongress2011.pdf).
Hartog, N., Grif fioen, J. , van Bergen, P.F., 2005. Deposi tional an d
paleohydrogeological controls on the distribution of organic matter and
other reactive reductants in aquifer sediments. Chemical Geology 216
(12), 113131.
Hayden, N.J., Voice, T.C., Annable, M.D., Wallace, R.B., 1992. Prediction of
leachate concentrations in petroleum-contaminated soils. Journal of Soil
Contamination 1 (1), 8193.
Jakobsen, R., Postma, D., 1994. In situ rates of sulfate reduction in an aquifer
(Rømø, Denmark) and implications for the reactivity of organic matter.
Geology 22 (12), 11011106.
Karickhoff, S.W., 1981. Semi-empirical estimation of sorption of hydrophobic
pollutants on natural sediments and soils. Chemosphere 10, 833846.
Kim, J., Lee, Y., Yoon, W.S., Jeon, J.S., Koo, M.-H., Keehm, Y., 2010. Numerical
modeling of aquifer thermal energy storage system. Energy 35 (12),
49554965.
KNMI, 2011. Weather database: Station De Bilt. KNMI, De Bilt.
Langevin, C.D., Thorne, D.T., Dausman, A.M., Sukop, M.C., Guo, W., 2008b.
SEAWAT version 4: a computer program for simulation of multi-species
solute and heat transport. in: U.S.G. Survey (Editor), Reston, Virginia.
Leonard, B.P., 1988. Universal limiter for transient interpolation modeling of
the advective transport equations: the ULTIMATE conservative differ-
ence scheme. NASA Technical Memorandum 100916 ICOMP-88-11.
Ma, R., Zheng, C., 2010. Effects of density and viscosity in modeling heat as a
groundwater tracer. Ground Water 48 (3), 380389.
Mao, X., Prommer, H., Barry, D.A., Langevin, C.D., Panteleit, B., Li, L., 2006.
Three-dimensional model for multi-component reactive transport with
variable density groundwater flow. Environmental Modelling & Soft-
ware 21 (5), 615628.
McCarty, P.L., 1997. Breathing with chlorinated solvents. Science 276 (5318),
15211522.
Miller, C.T., Poirier-McNeil, M.M., Mayer, A.S., 1990. Dissolution of trapped
nonaqueous phase liquids: mass transfer characteristics. Water Re-
sources Research 26 (11), 27832796.
Molz, F.J., Warman, J.C., Jones, T.E., 1978. Aquifer storage of heated water:
part Ia field experiment. Ground Water 16 (4), 234241.
Molz, F.J., Parr, A.D., Andersen, P.F., Lucido, V.D., Warman, J.C., 1979. Thermal
energy storage in a confined aquifer: experimental results. Water
Resources Research 15 (6), 15091514.
Molz, F.J., Parr, A.D., Andersen, P.F., 1981. Thermal energy storage in a confined
aquifer: second cycle. Water Resources Research 17 (3), 641645.
Molz, F.J., Melville, J.G., Parr, A.D., King, D.A., Hopf, M.T., 1983. Aquifer
thermal energy storage: a well doublet experiment at increased
temperatures. Water Resources Research 19 (1), 149160.
Montgomery, J.H., 2000. Groundwater Chemicals Desk Reference. CRC
Press, Boca Raton, Florida, USA.
Newell, C.J., Farhat, S.K., Adamson, D.T., Looney, B.B., 2011. Contaminant
plume classification system based on mass discharge. Ground Water 49
(6), 914919.
Park, E., Parker, J.C., 2005. Evaluation of an upscaled model for DNAPL
dissolution kinetics in heterogeneous aquifers. Advances in Water
Resources 28 (12), 12801291.
Parker, J.C., Park, E., 2004. Modeling field-scale dense nonaqueous phase
liquid dissolution kinetics in heterogeneous aquifers. Water Resources
Research 40 (5), W05109.
Parker, B.L., Cherry, J.A., Chapman, S.W., Guilbeault, M.A., 2003. Review and
analysis of chlorinated solvent dense nonaqueous phase liquid distribu-
tions in five sandy aquifers. Vadose Zone Journal 2 (2), 116137.
Parkhurst, D.L., Appelo, C.A.J., 1999. User's guide to PHREEQC (version 2): a
computer program for speciation, batch-reaction, one-dimensional
transport, and inverse geochemical calculations. Water-Resources In-
vestigations Report. : Open-File Reports Section [Distributor]. U.S.
Geological Survey: Earth Science Information Center, Denver, Colorado,
pp. 994259.
Pavlostathis, S.G., Zhuang, P., 1993. Reductive dechlorination of chloroalkenes
in microcosms developed with a field contaminated soil. Chemosphere 27
(4), 585595.
12 K.G. Zuurbier et al. / Journal of Contaminant Hydrology 147 (2013) 113
Post, V.E.A., Prommer, H., 2007. Multicomponent reactive transport simula-
tion of the Elder problem: effects of chemical reactions on salt plume
development. Water Resources Research 43 (10), W10404.
Prommer, H., Post, V.E.A., 2010. A Reactive Multicomponent Model for
Saturated Porous Media, Version 2.0, User's Manual. http://www.pht3d.
org.
Prommer, H., Davis, G.B., Barry, D.A., 1999. Geochemical changes during
biodegradation of petroleum hydrocarbons: field investigations and
biogeochemical modelling. Organic Geochemistry 30 (6), 423435.
Prommer, H., Barry, D.A., Zheng, C., 2003. MODFLOW/MT3DMS-based
reactive multicomponent transport modeling. Ground Water 41 (2),
247257.
Pyne, R.D.G., 2005. Aquifer Storage and Recovery A Guide to Groundwater
Recharge Through Wells. ASR Systems LLC, Gainesville, Florida (608
pp.).
Rao, P.S.C., Jawitz, J.W., Enfield, G.E., Falta, R.W., Annable, M.D., Wood, A.L.,
2001. Technology integration for contaminated site remediation: clean-
up goals and performance criteria. Groundwater Quality.Sheffield, U.K.
Robinson, C., Brovelli, A., Barry, D.A., Li, L., 2009. Tidal influence on BTEX
biodegradation in sandy coastal aquifers. Advances in Water Resources
32 (1), 1628.
Sanner, B., Karytsas, C., Mendrinos, D., Rybach, L., 2003. Current status of
ground source heat pumps and underground thermal energy storage in
Europe. Geothermics 32 (46), 579588.
Slenders, H.L.A., Dols, P., Verburg, R., de Vries, A.J., 2010. Sustainable
remediation panel: sustainable synergies for the subsurface: combining
groundwater energy with remediation. Remediation Journal 20 (2),
143153.
Smidt, H., de Vos, W.M., 2004. Anaerobic microbial dehalogenation. Annual
Review of Microbiology 58 (1), 4373.
Steeneveldt, R., Berger, B., Torp, T.A., 2006. CO2 capture and storage: closing
the knowingdoing gap. Chemical Engineering Research and Design 84
(9), 739763.
Suarez, M.P., Rifai, H.S., 1999. Biodegradation rates for fuel hydrocarbons and
chlorinated solvents in groundwater. Bioremediation Journal 3 (4),
337362.
Taskforce ATES, 2009. Green Light for Underground Thermal Energy Storage.
(in Dutch) Ministry of Housing, Spatial Planning and Envrironment
(VROM), Den Haag (downloadable from: www.vrom.nl).
TNO-NITG, 2009. REGIS II.1 Regional Hydrogeological Model. www.
dinoloket.nl.
Tsang, C.F., Hopkins, D.L., 1982. Aquifier Thermal Energy Storage: A Survey
(Medium: X; Size: Pages: 427441 pp.).
Tsang, C.F., Birkholzer, J., Rutqvist, J., 2008. A comparative review of
hydrologic issues involved in geologic storage of CO
2
and injection
disposal of liquid waste. Environmental Geology 54 (8), 17231737.
Van Breukelen, B.M., Hunkeler, D., Volkering, F., 2005. Quantification of
sequential chlorinated ethene degradation by use of a reactive transport
model incorporating isotope fractionation. Environmental Science &
Technology 39 (11), 41894197.
van Helvoort, P.-J., Griffioen, J., Hartog, N., 2007. Characterization of the
reactivity of riverine heterogeneous sediments using a facies-based
approach; the RhineMeuse delta. (The Netherlands) Applied Geo-
chemistry 22 (12), 27352757.
Verburg, R., Slenders, H.L.A., Hoekstra, N., Van Nieuwkerk, E., Guijt, R., Van
der Mark, B., 2010. Manual BOEG: Underground Thermal Energy and
Groundwater Contamination (in Dutch).
VROM, 2009. Remediation Circular 2009. in: Ministry of Housing, Spatial
Planning and Environment (VROM) (Editor), BWBR0025649, The
Hague.
Wiedemeier, T.H., Rifai, H.S., Wilson, J.T., Newell, C.J., 1999. Natural
attenuation of fuels and chlorinated solvents in the subsurface. John
Wiley and Sons, New York.
Zheng, C., 2006. MT3DMS v5. 2 Supplemental User's Guide. Department of
Geological Sciences, University of Alabama, Tuscaloosa, AL.
Zheng, C., Wang, P.P., 1999. MT3DMS: a modular three-dimensional
multispecies transport model for simulation of advection, dispersion
and chemical reactions of contaminants in groundwater systems;
Documentation and user's guide. in: U.S. Army Corps of Engineers
(Editor), Tuscaloosa, Alabama.
13K.G. Zuurbier et al. / Journal of Contaminant Hydrology 147 (2013) 113
... Similarly, during winter season, the pumping direction of groundwater and energy exchanger between building and groundwater are reversed with the function of heat-pump (Ni et al., 2020). The injection temperature is usually around 20-25 C depending on the weather conditions (Slenders et al., 2010;Zuurbier et al., 2013). Due to the rapid development of ATES in urban areas, where groundwater contamination is often found at similar depths to the locations where ATES is applied (McCarty, 2010;Zuurbier et al., 2013;Ni et al., 2016), the combination of enhanced bioremediation with ATES represents a promising way to reduce contamination (Fig. 5a). ...
... The injection temperature is usually around 20-25 C depending on the weather conditions (Slenders et al., 2010;Zuurbier et al., 2013). Due to the rapid development of ATES in urban areas, where groundwater contamination is often found at similar depths to the locations where ATES is applied (McCarty, 2010;Zuurbier et al., 2013;Ni et al., 2016), the combination of enhanced bioremediation with ATES represents a promising way to reduce contamination (Fig. 5a). Ni et al. (2016) conducted an experiment to inoculate Dehalococcoides in the ATES warm wells (20-25 C) and found that ATES created the optimal conditions for bio-reductive dechlorination. ...
Article
Thermally enhanced bioremediation (TEB), a new concept proposed in recent years, explores the combination of thermal treatment and bioremediation to address the challenges of the low efficiency and long duration of bioremediation. This study presented a comprehensive review regarding the fundamentals of TEB and its applications in soil and groundwater remediation. The temperature effects on the bioremediation of contaminants were systematically reviewed. The thermal effects on the physical, chemical and biological characteristics of soil, and the corresponding changes of contaminants bioavailability and microbial metabolic activities were summarized. Specifically, the increase in temperature within a suitable range can proliferate enzymes enrichment, extracellular polysaccharides and biosurfactants production, and further enhancing bioremediation. Furthermore, a systematic evaluation of TEB applications by utilizing traditional in situ heating technologies, as well as renewable energy (e.g., stored aquifer thermal energy and solar energy), was provided. Additionally, TEB has been applied as a biological polishing technology post thermal treatment, which can be a cost-effective method to address the contaminants rebounds in groundwater remediation. However, there are still various challenges to be addressed in TEB, and future research perspectives to further improve the basic understanding and applications of TEB for the remediation of contaminated soil and groundwater are presented.
... Temperature changes induced by UTES can directly influence the physical and chemical properties of organic contaminants. It has been demonstrated that solubility (Koproch et al., 2019), dissolution kinetics (Imhoff et al., 1997), sorption equilibrium (Ngueleu et al., 2018), biodegradation (Sommer et al., 2013;Zuurbier et al., 2013), etc. are susceptible to temperature changes. Alternatively, temperature changes can also affect the transport of contaminants by altering the density and viscosity of water. ...
... Furthermore, temperatureinduced density variations may lead to the appearance of buoyant flow and change the mass transport significantly, as demonstrated in simulations of thermal remediation (Krol et al., 2014). Besides temperature changes, ATES wells can also cause dilution and spreading of contaminants (Zuurbier et al., 2013;Phernambucq, 2015) due to the displacement of groundwater. ...
Article
Full-text available
Underground thermal energy storage is an efficient technique to boost the share of renewable energies. However, despite being well-established, their environmental impacts such as the interaction with hydrocarbon contaminants is not intensively investigated. This study uses OpenGeoSys software to simulate the heat and mass transport of a borehole thermal energy storage (BTES) system in a shallow unconfined aquifer. A high-temperature (70 C) heat storage scenario was considered which imposes long-term thermal impact on the subsurface. Moreover, the effect of temperature-dependent flow and mass transport in a two-phase system is examined for the contaminant trichloroethylene (TCE). In particular, as subsurface temperatures are raised due to BTES operation, volatilization will increase and redistribute the TCE in liquid and gas phases. These changes are inspected for different scenarios in a contaminant transport context. The results demonstrated the promising potential of BTES in facilitating the natural attenuation of hydrocarbon contaminants, particularly when buoyant flow is induced to accelerate TCE volatilization. For instance, over 70% of TCE mass was removed from a discontinuous contaminant plume after 5 years operation of a small BTES installation. The findings of this study are insightful for an increased application of subsurface heat storage facilities, especially in contaminated urban areas.
... Therefore, the use of MT3D and MT3DMS is acceptable for the modelling of conventional GSHP systems, but not for UTES systems. The package SEAWAT was derived from MT3D to perform density and viscosity-dependent flow modelling (Langevin and Guo, 2006) and it was recently applied in the design of ATES systems Zuurbier et al., 2013). ...
... Possible positive effects are also observed, such as the enhanced degradation of pollutants, such as chlorinated hydrocarbons. This led some researchers to consider ATES as a possible way to remediate contaminated sites (Zuurbier et al., 2013). ...
Chapter
The expression Underground Thermal Energy Storage (UTES) identifies shallow geothermal systems where heat from external sources (solar thermal collectors, industrial processes, combined heat and power systems) is stored seasonally into the ground to be used during periods of higher demand. UTES is performed as closed-loop Borehole (BTES) or open-loop Aquifer Thermal Energy Storage (ATES). This article presents UTES techniques with relevant case studies, the software used for modelling energy needs and underground heat transport, some peculiar aspects affecting the storage efficiency, the typical operating issues, and the possible subsurface impacts of UTES installations.
... In urban areas, anthropogenic impacts such as a dense building development, underground car parks, open geothermal systems and injections of thermal wastewater from industry result in local thermal alteration of groundwater by up to several degrees (e.g. Taylor and Stefan, 2009;Zhu et al., 2010;Menberg et al., 2013b;Tissen et al., 2019). According to Brielmann et al. (2011), annual temperature fluctuations in aquifers caused by shallow geothermal energy systems range between 4 • C in winter and ≤ 20 • C in summer. ...
... Moreover, it is assumed that groundwater fauna can usually cope well with short-term changes in physical-chemical parameters (Griebler et al., 2016). Previous studies showed that some species can even benefit from pollutants (Matzke, 2006;Zuurbier et al., 2013). In case of nitrate, numerous studies emphasise that nitrate at concentrations below 50 mg/L does not directly affect groundwater fauna (Fakher el Abiari et al., 1998;Mösslacher and Notenboom, 2000;Di Lorenzo and Galassi, 2013;Di Lorenzo et al., 2020a). ...
Article
Full-text available
In Germany, 70 % of the drinking water demand is met by groundwater, for which the quality is the product of multiple physical–chemical and biological processes. As healthy groundwater ecosystems help to provide clean drinking water, it is necessary to assess their ecological conditions. This is particularly true for densely populated urban areas, where faunistic groundwater investigations are still scarce. The aim of this study is, therefore, to provide a first assessment of the groundwater fauna in an urban area. Thus, we examine the ecological status of an anthropogenically influenced aquifer by analysing fauna in 39 groundwater monitoring wells in the city of Karlsruhe (Germany). For classification, we apply the groundwater ecosystem status index (GESI), in which a threshold of more than 70 % of crustaceans and less than 20 % of oligochaetes serves as an indication for very good and good ecological conditions. Our study reveals that only 35 % of the wells in the residential, commercial and industrial areas and 50 % of wells in the forested area fulfil these criteria. However, the study did not find clear spatial patterns with respect to land use and other anthropogenic impacts, in particular with respect to groundwater temperature. Nevertheless, there are noticeable differences in the spatial distribution of species in combination with abiotic groundwater characteristics in groundwater of the different areas of the city, which indicate that a more comprehensive assessment is required to evaluate the groundwater ecological status in more detail. In particular, more indicators, such as groundwater temperature, indicator species, delineation of site-specific characteristics and natural reference conditions should be considered.
... There are also processes associated with potential changes in groundwater quality, such as i) the release of dissolved organic carbon (DOC) and associated redox processes (Bonte et al., 2013a;Brons et al., 1991;Jesußek et al., 2013a;; ii) the release and fixation of trace elements and heavy metals (M. Bonte et al., 2013b;García-Gil et al., 2016;Lüders et al., 2020;Saito et al., 2016) and iii) increased or decreased solubility (Koproch et al., 2019), volatilisation (Schwardt et al., 2021) and degradation (Men and Cheng, 2011;Němeček et al., 2018;Zuurbier et al., 2013) of organic contaminants. ...
Article
Full-text available
Despite their potential in heating supply systems, thus far high-temperature aquifer thermal energy storages (HT-ATES) currently lack widespread application. Reducing the potential risks by improving the predictability of hydrogeochemical processes accelerated or initiated at elevated temperatures might promote the development of this technology. Therefore, we report the results of a short-term hot water infiltration field test with subsurface temperatures above 70°C, along with associated laboratory batch tests at 10, 40 and 70°C for 28 sediment samples to determine their usability for geochemical prediction. Most groundwater components had lower maximal concentrations and smaller concentration ranges in field samples compared to the batch tests. This indicates that the strongest geochemical effects observed in laboratory tests with sufficient site-specific sediment samples will likely be attenuated at the field scale. A comparison of field measurements with predicted concentration ranges, based on temperature induced relative concentration changes from the batch tests, revealed that the predictive power was greatest, where the hot infiltrated water had cooled least and the strongest geochemical effects occurred. The batch test-based predictions showed the best accordance with field data for components, with significant temperature-induced concentration changes related to ion exchange and (de)sorption processes. However, accurate prediction of concentration changes based on other processes, e.g. mineral dissolution, and downstream reversals in concentrations, requires further investigation. The here presented procedure enables the prediction of maximal expectable temperature-dependent concentration changes for most environmentally relevant ancillary groundwater components, e.g. As, with limited effort.
... One of the first applications of NSGE systems in polluted sites was carried out in the Netherlands [33]. Other researches were carried out at modelling level (see [34] and [35]). ...
Technical Report
Full-text available
Basel, and Regione Lombardia. This deliverable deals with the large-scale mapping of geological features and other factors (environmental issues, bans, law restrictions, etc.) which may interfere with the installation of Borehole Heat Exchangers and/or water wells for Ground Water Heat Pumps. For this assessment, the involved partners collected and processed geo-referenced data of such features, which were put together in a Web GIS available on line. Also, partners provided sources for other kinds of data (books, articles, papers, web-viewer, pdf maps etc.) which are not in GIS formats. This report therefore serves as a reference guide of the Web GIS and as a complementary source of information about risks, interferences, issues and possible solutions for installers, designers, and public authorities involved in the design, approval and installation of NSGE systems. Deliverable D.4.1.1-Assessment and mapping of potential interferences to the installation of NSGE systems in the Alpine Regions 16/12/2015-07/09/2018: A large-scale web GIS map of possible underground interferences with Near-Surface Geothermal Energy systems will be published online, together with a reference guide explaining how the input data were collected, homogenized and processed. Assessment and mapping of potential interferences to the installation of NSGE systems in the Alpine Regions GRETA is co-financed by the European Regional Development Fund through the Interreg Alpine Space programme. See more about GRETA at www.alpine-space.eu/projects/greta.
Article
Full-text available
Low entropy shallow ground heat resources are gaining importance in recent years owing to their availability compared to difficult-to-reach geothermal energy sources. In the last decades, aquifer thermal energy storage (ATES) systems have begun to be utilized increasingly since they can provide one of the cleanest and most energy efficient heating and cooling system alternatives for buildings. One of the main problems in the design phase of ATES systems is the correct estimation of thermal interference distance between extraction and injection side of the system. In order to investigate this problem, an extensive modeling study was carried out to constitute a preliminary approach for the numerical model calibration of heat transport and storage in aquifers. For this purpose, actual site data available for a well-studied coastal aquifer in Mediterranean region of Turkey was used in the analysis and performance assessment of a conceptual open-loop ground source heating well doublet. Three-dimensional coupled numerical model of groundwater flow and heat transfer was produced to constitute an approach for the thermo-hydraulic model calibration by estimating the duration of thermal breakthrough between the abstraction and injection wells. Simulation results were compared with the analytical solution for doublet well breakthrough time with finite hydraulic gradient. The comparison of results indicated that the numerical model is able to represent the thermal behavior in the field. Therefore, calibration methodology established in this paper could be followed in the pre-feasibility study and the design phase of low-temperature aquifer thermal energy storage systems worldwide.
Article
There has been a worldwide interest in renewable energy technologies, as a means of decreasing reliance on fossil fuels, minimizing climate change effects, and reducing greenhouse emissions. One such technology is geothermal heating where the constant subsurface temperature is used to cool or heat building interiors via heat pumps. In Canada, the use of geothermal heating has become a popular option for heating and cooling buildings, and it is anticipated that, in the near term, most large buildings will include geothermal heating as part of their climate control strategy. However, little is known about the environmental impacts of geothermal heating on the subsurface environment. The present review will examine the effect of geothermal heating on groundwater flow and remediation efforts, whereby the heat generated by geothermal systems may help with urban pollution. "Geothermal Remediation" could leverage the subsurface heating resulting from geothermal systems to accelerate biodegradation of certain petroleum-based pollutants at brown-field sites, while providing building(s) with sustainable heating and cooling. This idea coincides with the rising momentum towards sustainable and green remediation in Europe and the United States. To ensure that Geothermal Remediation is achievable, the effect of heat on bioremediation needs to be examined. This review provides an insight into what we know about heat effects on bioremediation activities and subsurface transport.
Article
Chemical and phase equilibrum (CPE) calculations are indispensable to the geochemical simulation of natural systems, such as freshwater, groundwater, and mineral formations. In natural environments, many chemical reactions occur, both at the interface between rocks and fluids (e.g., dissolution reactions) and in bulk phases (e.g., aqueous speciation). Moreover, the typical ionic species found in nature require thermodynamic models for electrolytes, increasing the calculation complexity. In this work, we present RAND-based algorithms, which originate from Gibbs energy minimization, for geochemical CPE problems for closed systems and open systems. The open system problems specify fixed chemical potentials for non-charged species, as in a constant partial CO2 pressure problem, or for charged ionic species, as in a constant pH problem. We showcase the robustness and efficiency of these algorithms using the mineral system Mg-Si-Ca-CO2-H2O because of its relevance to CO2 underground geological storage. The algorithms show quadratic convergence and the coupled tangent plane distance analysis can determine the most stable phase assemblage. The algorithms generate results in agreement with PHREEQC, whereas they are much faster than PHREEQC. The described algorithms can be potentially used for reliable and fast simulation of CO2 storage in geological formations, and of other processes in natural environments involving geochemical equilibrium.
Article
The use of colloidal biliquid aphron (CBLA) as density modifier to reduce the density of dense nonaqueous phase liquids (DNAPLs) irreversibly is an efficient strategy to control the migration of DNAPLs in contaminated aquifers. However, the process and mechanism of the density regulation using CBLA is still not clear and there is still a big gap in the application of CBLA in actual contaminated sites. In this study, we carried out density modification of 5 DNAPLs (nitrobenzene (NB), dichloromethane (DCM), trichloroethylene (TCE), carbon tetrachloride (CTC), perchloroethylene (PCE)) using CBLA and studied the effect of co-existing ions by 3D response surface method. We found that DNAPLs changed to light nonaqueous phase liquids (LNAPLs) and float up after interaction with light organic liquid from CBLA. The density modification process is limited by the demulsificaiton of CBLA and the density of DNAPL itself. Density regulation of DNAPLs followed pseudo-second-order kinetics. The co-existing ions affected the stability of CBLA and the demulsification ability of the demulsifier. Aquifer materials and low temperature did not influence the density control effect of CBLA. This research advances the practical application of density control of DNAPLs using CBLA, and makes important contributions for subsequent combined remediation approach.
Technical Report
Full-text available
The SEAWAT program is a coupled version of MODFLOW and MT3DMS designed to simulate three dimensional, variable-density, saturated ground-water flow. Flexible equations were added to the program to allow fluid density to be calculated as a function of one or more MT3DMS species. Fluid density may also be calculated as a function of fluid pressure. The effect of fluid viscosity variations on ground-water flow was included as an option. Fluid viscosity can be calculated as a function of one or more MT3DMS species, and the program includes additional functions for representing the dependence on temperature. Although MT3DMS and SEAWAT are not explicitly designed to simulate heat transport, temperature can be simulated as one of the species by entering appropriate transport coefficients. For example, the process of heat conduction is mathematically analogous to Fickian diffusion. Heat conduction can be represented in SEAWAT by assigning a thermal diffusivity for the temperature species (instead of a molecular diffusion coefficient for a solute species). Heat exchange with the solid matrix can be treated in a similar manner by using the mathematically equivalent process of solute sorption. By combining flexible equations for fluid density and viscosity with multi-species transport, SEAWAT Version 4 represents variable-density ground-water flow coupled with multi-species solute and heat transport. SEAWAT Version 4 is based on MODFLOW-2000 and MT3DMS and retains all of the functionality of SEAWAT-2000.SEAWAT Version 4 also supports new simulation options for coupling flow and transport, and for representing constant-head boundaries. In previous versions of SEAWAT, the flow equation was solved for every transport timestep, regardless of whether or not there was a large change in fluid density. A new option was implemented in SEAWAT Version 4 that allows users to control how often the flow field is updated. New options were also implemented for representing constant-head boundaries with the Time-Variant Constant-Head (CHD) Package. These options allow for increased flexibility when using CHD flow boundaries with the zero-dispersive flux solute boundaries implemented by MT3DMS at constant-head cells. This report contains revised input instructions for the MT3DMS Dispersion (DSP) Package, Variable-Density Flow (VDF) Package, Viscosity (VSC) Package, and CHD Package. The report concludes with seven cases of an example problem designed to highlight many of the new features.
Article
Full-text available
We present an overview of the risks that underground thermal energy storage (UTES) can impose on the groundwater system, drinking water production, and the subsurface environment in general. We describe existing policy and licensing arrangements for UTES in the Netherlands, as well as the capability of the current and future Dutch policy and legal framework to minimize or mitigate risks from UTES on groundwater resources. A survey at the European Union member state level indicates that regulation and research on the potential impacts of UTES on groundwater resources and the subsurface environment often lag behind the technological development of and ever-growing demand for this renewable energy source. The lack of a clear and scientifically underpinned risk management strategy implies that potentially unwanted risks might be taken at vulnerable locations such as near well fields used for drinking water production, whereas at other sites, the application of UTES is avoided without proper reasons. This means that the sustainability of UTES as a form of renewable energy is currently not fully understood, and the technology may be compromising the natural resilience of the subsurface environment. We recognize three main issues that should be addressed to secure sustainable application of UTES: Scientific research is required to further elucidate the impacts of UTES on groundwater; cross-sectoral subsurface planning is required to minimize negative conflicts between UTES and other subsurface interests; and EU-wide guidelines and standards are required for quality assurance and control when installing UTES systems.
Article
Full-text available
There is a need to develop and field-test integrated remediation technologies for cost-effective treatment of contaminated sites to achieve risk-based and rational endpoints. Aggressive technologies designed for rapid source-zone remediation must be linked to technologies for achieving enhanced clean up of the dissolved plume. Remediation technology integration should minimize the cost of achieving risk-based endpoints by selecting treat­ ment trains or technology combinations that, when coupled together, work in a synergistic manner. Contaminant flux across a control plane immediately downgradient from the source, rather than contaminant concentration, should be used as the basis for evaluating the effectiveness or success of remediation. The acceptable threshold contaminant flux should be set equal to, or less than, the natural attenuation capacity within the dissolved plume. Simulation results show that significant contaminant-flux reductions can be achieved by partial removal of contaminant mass from DNAPL source zones.
Article
In addition to the flooding addressed by detention systems, infiltration can help support groundwater recharge, stream base flows and water quality. Infiltration involves capturing stormwater in a closed basin where it is stored while filtering into the surrounding soil. The author discusses the various types of basin and the environments to which they are best suited. The integral hydraulic design must be able to contain the flow of the design storm, in addition to all background flows. There is a review of several examples from different regions, where stormwater infiltration has been used to address a number of issues, with varying degrees of success. -M.Z.Barber
Article
To aid in testing the idea of storing thermal energy in aquifers, an experiment was performed by Auburn University in which 54,784 m3 of water was pumped from a shallow supply aquifer, heated to an average temperature of 55°C, and injected into a deeper confined aquifer where the ambient temperature was 20°C. After a storage period of 51 days, 55,345 m3 of water were produced from the confined aquifer. Throughout the experiment, which lasted approximately 6 months, groundwater temperatures were recorded at six depths in each of 10 observation wells, and hydraulic heads were recorded in five observation wells. In order to prevent errors due to thermal convection, most of the observation wells recording temperature had to be backfilled with sand. During the 41-day production period, the temperature of the produced water varied from 55° to 33°C, and 65% of the injected thermal energy was recovered. At no time was an appreciable amount of free thermal convection observed in the storage formation. The dominant heat dissipation mechanisms appeared to be hydrodynamic thermal dispersion and possible mixing of cold and hot water induced by clogging and unclogging of the injection-production well. On the basis of laboratory and field studies, it was concluded that clogging of the injection well, which constituted the major technical problem during the experiment, was caused by the freshwater-sensitive nature of the storage aquifer. Due to the relatively low concentration of cations in the supply water, clay particles would swell, disperse, and migrate until they became trapped in the relatively small pores connecting the larger pores. Surging the pump and back washing the injection well would dislodge the clogging particles and temporarily improve the storage formation permeability. The phenomenon seems largely independent of temperature because it was reproduced in the laboratory with unheated water. It may, however, depend on pore velocity. Future research should be directed toward procedures for selecting storage aquifers that will have minimal susceptibility to clogging and other geochemical problems. Procedures for overcoming such difficulties are needed also because clogging and related phenomena will be more the rule than the exception. Designing an aquifer thermal storage system for maximum energy recovery would involve selecting an appropriate aquifer, analyzing the effects of hydrodynamic thermal dispersion and thermal convection if it is predicted to occur, anticipating geochemical problems, designing the optimum supply-injection-production well configuration and injecting a sufficiently large volume of heated water to realize economies of scale related to increasing volume-surface area ratio.