Thermal short circuit on groundwater heat pump

Thermal short circuit on groundwater heat pump

Applied Thermal Engineering 57 (2013) 107e115 Contents lists available at SciVerse ScienceDirect Applied Thermal Engineering journal homepage: www.e...

3MB Sizes 329 Downloads 138 Views

Applied Thermal Engineering 57 (2013) 107e115

Contents lists available at SciVerse ScienceDirect

Applied Thermal Engineering journal homepage:

Thermal short circuit on groundwater heat pump Antonio Galgaro*, Matteo Cultrera* Geosciences Department, University of Padua, via G. Gradenigo 6, 35131 Padua, Italy

h i g h l i g h t s  We summarized the problems concerning the groundwater thermal feedback.  The paper suggests the analytical solutions to analyze the “Thermal short circuit” problem.  Two graphical solutions are proposed to check the minimal distance between abstraction and reinjection well.  The paper suggests the right way to design a groundwater heat pump system.  A simple study case is submitted as exemplum.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 25 January 2012 Accepted 3 March 2013 Available online 27 March 2013

This paper presents a study of the feasibility of providing heating and cooling by means of an open-loop groundwater heat pump system for a restored commercial building in Rovigo, located in the Po River Plain (Italy). Results obtained from the modelling confirm the potential hydrogeological capacity of the site to provide the necessary amount of groundwater and associated energy with limited environmental impact. Injection of warmer (or cooler) water into the aquifer creates a thermal plume whose dimensions and geometry depend at first on the properties of the subsurface formations and particularly on working system conditions and by the cooling and heating loads. This study shows the risk of the thermal breakthrough between well doublets and suggests that there are several possible heating/cooling daily timetables that reduce the risk of thermal feedback between extraction and injection wells. These timetables may prevent the GWHP system from becoming uneconomical and energetically inefficient. Thermal breakthrough is common in groundwater heat exchange systems, particularly in historical town districts where the distance between wells is necessarily close due to buildings proximity and the possibility of other group plant in the neighbourhood. Most probably due to modelling difficulties, it is unusual to take into account this type of thermal contamination during an ordinary analysis of the interaction between and consequences of groundwater heat pump and aquifer systems. An approach using complex open loop modelling allows the analysis of a case of thermal feedback in order to obtain the best planning and use of the geoexchange plant. Ó 2013 Elsevier Ltd. All rights reserved.

Keywords: Thermal condition Urban groundwater Numerical modelling Rovigo

1. Introduction A groundwater heat exchange system requires the presence of an aquifer from which water can be extracted via a borehole or well (called a production, withdraw, extraction or abstraction well). The water is abstracted from one part of the aquifer system and is used to heat or cool buildings via a heat pump [1,2]. This simple scheme is known as open-loop well doublet or groundwater heat pump

* Corresponding authors. Tel.: þ393495714159. E-mail addresses: [email protected], [email protected] (M. Cultrera). 1359-4311/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.

(GWHP) system. Consequently, a GWHP system is a doublet well open-loop system that pumps groundwater from an extraction well, exploits the water calories through a heat exchanger and finally dismisses groundwater into an injection well in the aquifer [3e5]. Once the heat exchange via the heat pump has occurred, the water is re-injected into an aquifer (return or injection well), which, according to Italian law (D.Lgs 152/2006) and following the principle of renewability. In this way, the quantitative and qualitative aquifers parameters are preserved. Aquifer thermal energy storage (ATES) exploits groundwater for the storage of cold and warm water in order to use this thermal energy to cool and heat buildings; during the summer season, cold


A. Galgaro, M. Cultrera / Applied Thermal Engineering 57 (2013) 107e115

water is pumped from the aquifer with a closed piping and heat exchange system, into the buildings that require cooling [6e8]. After heat exchange, the warmer water is transferred to a second location within the same aquifer, which consequently will rise in temperature. In the winter, the process is reversed leading to the heating of the building and the lowering of the temperature in the aquifer, which returns to its initial thermal condition [8]. Therefore, the aquifer can be defined as a thermal storage system; when the flow velocity of groundwater is quite low, the aquifer can be used to store calories (or frigories) to be used in the opposite season. In other setups, ATES restores heat storage via solar collectors or even excess heat from power stations [9,10]. Mainly in Europe and North America, but also in the rest of the world, this technique is being used more and more frequently, caused by the increased demand for sustainable technologies. This increase is due to the increase costs of conventional energy and to improve the use of renewable energy whit low carbon dioxide production. The ATES systems are coupled with a geothermal heat pump or groundwater heat pump (GWHP). GWHP is a central heating and/or cooling system that pumps heat to or from the ground and needs a smaller amount electric energy, instead of air conditioners and combustion heat, than a traditional heating/cooling plant [11,12]. The environmental impact is slighter than a traditional heating/ cooling plant, thanks to the high GSHPs energy-efficiency environmentally clean and cost-effective characteristics. Heat pumps offer significant emission reduction potential, particularly where they are used for both heating and cooling and produce zero emissions locally, but their electricity supply almost always includes components with high greenhouse gas emissions. Their environmental impact therefore depends on the characteristics of the electricity supply [13,14]. GWHP have unsurpassed thermal efficiencies and produce zero emissions locally, but their electricity supply almost always includes components with high greenhouse gas emissions. Their environmental impact therefore depends on the characteristics of the electricity supply [14]. The ATES system may cause an excessive variation in temperature within the aquifer from which the groundwater is exploited, thus causing the GSHP to become inefficient or even unsuitable. The aim of this paper is to analyse differential temperature generated in the exploited aquifer and the consequences on the GSHP system of this induced variation and the ways to reduce them. 2. The open loop groundwater heat exchange system The GSHP presents several advantages in the Mediterranean area, with respect to colder climates, because the heat exchange system is used both in winter and summer for heating and cooling purposes respectively. In some cases, cooling system (and/or heat system) is obtained by a simple heat exchange, without the use of heat pump system (e.g. free cooling). If the aquifer is large enough, the boundary conditions and parameters (mainly groundwater gradient and permeability) can permit energy storage. As a result, the hydrogeological layer is used as an ATES [15,16]. In some cases, a good strategy is to switch the direction of the doublet scheme, so the coefficient of performance (COP) can be increased. In heating mode the COP is derived by the equation:

COPheating ¼



where H is the heating supplied from the reservoir, and M is the work consumed by the heat pump system.

In the cooling mode, the COP is defined as:

COPcooling ¼



2.1. Governing equation The heat flux through a porous media in the aquifer was simulated by means of a deterministic model. The differential equations describing groundwater flow are coupled with the equations describing heat flow in porous media [17]. The groundwater velocity or specific discharge qi in x position in Cartesian coordinates can be derived from Darcy’s law [18]:

qi ¼ kij

vh vxj


where kij is the hydraulic conductivity tensor and h is the hydraulic head. Assuming that hydraulic head h is conservative and the fluid density variations rw are negligible, the equation describing the groundwater flow through a non-homogeneous porous aquifer may be written:

v vxi

vh LKij vxj

! þR ¼ 0


where L is the aquifer thickness and R is the abstraction or injection rate of the wells per unit area (source term). In addition, the instantaneous thermal equilibrium must be reached, and assuming that the heat capacity is constant, the partial equation describing the movement of heat in a groundwater system through a non-homogeneous porous media may be written:

vC vðq CÞ v rs cs ¼ rw cw i þ vt vxi vxi

vC lij vxj

! þW


where C is the temperature, rscs the volumetric heat capacity of the porous media, rwcw the water volumetric heat capacity, and lij is the effective thermal conductivity according to the properties of both fluid and solid phases. The second expression of equation (4) includes a convective heat transport term, a heat conduction term and a heat source term. The groundwater heat transport equation is analogous to the mass balance equation in groundwater solute transport problems [19]. 2.2. The thermal short circuit In general, abstracting and injecting groundwater encounter considerable difficulty in modelling the groundwater temperatureetime variations field, especially in an area where multiple open loop doublets operate for thermal purposes [20,21]. Moreover, a geologic and hydrogeologic model is required to avoid unexpected problems. Three elements are necessary to analyze the plant functionality and the groundwater thermal footprint evolution: (1) the detailed (hydro)geological model of the study area; (2) an accurate thermotechnical design; (3) the correct management of physical, analytical or numerical solutions to solve the problem [22]. Considering a simple open loop well-doublet system consisting of one abstraction well and one injection well, under different conditions of regional groundwater assessment, the groundwater breakthrough time caused by heat exchange processes can be calculated. The breakthrough time is the length of time to produce a

A. Galgaro, M. Cultrera / Applied Thermal Engineering 57 (2013) 107e115

thermal effect on the pumping well due to discharge in the injection ones (Fig. 1). Considering a common scenario of GWHP in the summer season (Fig. 1), the water is extracted from the abstraction well at the temperature DQ, which generally is quite constant. This water is exploited by the geoexchange system and is injected with a reference temperature of Q þ DQx. The pumping activity modifies the natural hydraulic gradient and triggers a local inversion of groundwater flow from the injection towards the abstraction well. In this case, the injected hot water reaches the abstraction well after a time Dt from the start of pumping. When the breakthrough time occurs on the abstraction well, the temperature of the water from the abstraction well progressively increases, rising to a temperature of DQab. Consequently, in order to continue cooling the building, the GWHP must increase the water abstracted temperature to a value of Q þ DQx þ DQab (Fig. 1). This phenomenon is known as “thermal feedback” and causes the increasing of energy supplies of the GWHP, while the COP will decrease. By an analogy with the electric network, the thermal feedback can be defined as a “Thermal Short Circuit” (TSC), because the proper connection is interrupted by a wrong connection and the system is consequently damaged (Fig. 1). When the regional velocity v is negative and its magnitude is:


2W DH fp


the injected groundwater will never reach the abstraction well, where W is the flow rate, f is the porosity, H the aquifer thickness. In this simple equation the heat conduction is ignored and the heat transport is due only to the groundwater flow. If equation (5) is not respected, the thermal breakthrough time (tb) can be calculated using the analytical solution proposed by


Lippmann and Tsang (1980) and Banks (2009). The equation is [19,23]:

 tb ¼




B Lns B B1 þ B @ Ki @ 0

C 2Q rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffitan1 C A zQ pHDiL 1  pHDiL 11

B CC 1 CC B @rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi zQ AA 1  pHDiL


when i is non zero value. If the regional flow is not parallel to the line between the open loop doublet wells, the angle between the regional groundwater velocity vector and the line from the abstraction well to the injection one must be considered [24]. When the hydraulic gradient i is too low, very frequent hydraulic feedback and TSC occur [25]. If equation (5) is respected, the abstraction and the injection wells are far enough apart. In this case, the GWHP system works correctly; otherwise, the doublet pumping system triggers the local inversion of the hydraulic gradient. In ordinary practice, TSC is very frequent [1,24,26]. Only when the injection and abstraction wells are quite far apart can hydraulic feedback be avoided. Consequently, the TSC modifies the functioning of the open loop well doublet system and the heat pumps can reduce the COP or even make the geothermal plant inoperable [27]. As mentioned above, the use in urban areas of an open loop well doublet system is increasing. In historical city centres, it is difficult to respect the double well distance and interference often occurs between two or more wells of an open loop doublet well system or nearby GWHP systems [28].

Fig. 1. Scheme for an abstraction and injection well. The thermal system is controlled by the earth’s geothermal gradient from the bottom, by weather influence at the top, and by the GWHP.


A. Galgaro, M. Cultrera / Applied Thermal Engineering 57 (2013) 107e115

As suggested by equation (5), the TSC problem can be analysed using an analytical approach [24]. In this case, the homogeneity, isotropic and simple conditions are assumed to be:


2W T pi


Figs. 2 and 3 allow a simple evaluation of distance between abstraction and injection wells; the well distance must be above the line which represents the well double system. In the case of Fig. 2, a constant hydraulic gradient is assumed (i ¼ 0.05). The knowledge of flow rate and hydraulic transmissivity allow the calculation of distance D. In the case of Fig. 3, a constant hydraulic transmissivity is assumed (T ¼ 2.5E-03 m2/s). The knowledge of flow rate and hydraulic gradient allow the calculation of minimum distance D. A preliminary analysis of Figs. 2 and 3 allows the planning of an optimal design of a doublet well system, where the geologic and hydrostratigraphic conditions are isotropic and homogeneous. In the confined aquifers of the Veneto and Po Plains, transmissivity is not very high (generally less than 1E-03 m2/s). The hydraulic gradient, moreover, is often less than 0.05, as is shown in Fig. 4. In fact, Fig. 4 shows the hydraulic gradient of shallow groundwater in the Venetian plain. Also in the Po Plain, the groundwater gradient is generally quite low, as evidenced in Fig. 4, where in the orange to green area the gradient is lower than 0.005. In the Middle and Low Venetian Plain, the well spacing for a classical GWHP sketch must be more than 200e250 m. In urban areas like the Venetian Area and the Po Plain, this distance is difficult to obtain because of population increase and the reduction of green areas (For interpretation of the references to colour in this paragraph, the reader is referred to the web version of this article.). The analytical solution suggested by Clyde et al. (1983) and by Banks (2009) requires the assumption of the following: a constant flow rate from the abstraction/injection well, an isotropic and homogeneous aquifer and a regular regional groundwater flow [19,24]. 2.3. Feflow’s OpenLoop module A different way to solve the TSC is the numerical approach which considers the geological and hydrogeological heterogeneity and anisotropy.

The most known commercial computer models used to simulate coupled flux and heat transport in aquifers are HST3D, FEFLOW and Shemat [29e31]. To solve the TSC problems, the authors applied the Feflow code. This 3D finite element method (FEM) code allows the simulation of the well doublet geothermal system, coupling the groundwater flow analysis with heat transport [33,34]. A very helpful tool is the application of a Feflow module, called “OpenLoop” ( This tool employs the temperature resulting from the nodal computation in the injection well; this resultant value is increased or decreased by another value which has been defined before computation and is derived from the planned thermotechnical analysis. Also, the temperature computed from the abstraction well node may be influenced by neighbouring heat sources/sinks. The OpenLoop module runs defining the timevarying temperature differential between groups of abstraction and injection wells; the advantage to this module is the capability of handling more than one doublet of abstraction and injection wells. This FEM code is very versatile and it is possible to simulate and provide analyses for any real scenario. In the case studied, the aquifers exploited generally present a very low hydraulic gradient, equal to 0.001. In the Padana Plain and in the Middle and Low Venetian Plain, the average gradient is no more than 0.005(Fig. 4). 3. A study case: Rovigo (Italy) 3.1. Geology of the study area The study area is located in North-east Italy, where the Adriatic microplate represents the most northern area of the African Plate [35]. The Southern Alpine system is the result of a polyphasic evolution, active in the Tertiary period. The oldest structural system coincides with the Mesoalpine (Eocene) and the following phases of Neoalpine compression (OligoceneeMiocene), which produced the Dinaric System (NWeSE) and follows a NE direction in the eastern part of the Southern Alps [35e37]. The Po basin progressed into the foreland basin with respect to the Southern Alpine area in the Oligocene period, while in more

Fig. 2. Analysis of minimum distance between abstraction and injection wells, when aquifer transmissivity and flow rate are known (i ¼ 0.05).

A. Galgaro, M. Cultrera / Applied Thermal Engineering 57 (2013) 107e115


Fig. 3. Analysis of minimum distance between abstraction and injection wells, when hydraulic groundwater gradient and flow rate are known.

recent times it progressed into the foreland basin also with respect to the Apennine system (Fig. 5) [38,39]. The eastern Po Plain is set, from the structural point of view, between the northern border of the Adriatic monoclinal and the Pedealpine one. This boundary is represented by the Schio-Vicenza to the West, a sinistral transform fault whose direction is resumed by the neotectonic lineaments recognizable by the NWeSE fault system. The whole flood basin has been affected by land subsidence from the Pliocene period onwards, mainly connected to the collisional force produced by the Apennine chain to the SW. The

Fig. 4. Groundwater hydraulic gradient in the Venetian plain (Veneto country, Italy) in longitudinal to vertical ratio (head loss to horizontal distance).

Venetian area is influenced by Apennine tectonics whose tilling presents a dip direction towards south (Fig. 5) [40]. The prevailing direction of the fault proves how the whole region’s tectonic and evolving structure is influenced by the SchioVicenza, following the NNWeSSE direction. This direction is also recognizable in neotectonic structures that influence the whole quaternary morphological evolution [41]. 3.2. Hydrogeological setting From a geological point of view, the study area is located in the Eastern Po alluvial plain entirely covered by Quaternary fluvial sediments [42]. In the Pliocene and Pleistocene, the Adige River’s alluvial megafan built this plain, characterized by marine, deltaic, and alluvial materials [43]. Some palaeorivers flow through the Eastern Po and Atesian Plain, which are filled in with sand and silty sand sediments. Clay, silty clay, organic soil and peat with few sandy layers represent the nearby areas. The leakage and drainage rates through present rivers are negligible because of the low hydraulic permeability of sediments filling on their bed. The shallow aquifer system is a multilayer confined aquifer system where the groundwater flows to E, according to the regional flownet. The mean value of groundwater gradient is very low, locally less than 0.0005 (see Fig. 4) and in the summer period, local inversion of the gradient direction may take place due to the use of drainage pumps used in agriculture [44]. Two palaeorivers more than 200 m wide, characterized by sandy sediments flow through the centre of Rovigo, crossing each other close to the study area (Fig. 6). The palaeo river system can influence the local flownet, like in 2003, when groundwater extraction from a wellpoint system induced land subsidence in the centre of Rovigo [45]. In this case, the land subsidence was locally very different due to heterogeneities and anisotropy in both sediments and hydraulic conductivity [46]. Two drill holes were carried out in opposite corners of the study area. The core drill hole permitted core recovery for both geological and geotechnical detection. The subsoil geotechnical analysis allowed the development of the geological model. Fig. 7 shows the


A. Galgaro, M. Cultrera / Applied Thermal Engineering 57 (2013) 107e115

Fig. 5. Geological map of North-east Italy; the study area is indicated by the red circle (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.).

stratigraphy of the study area, which is assumed locally very homogeneous. Hydraulic tests allowed the identification of main hydrogeological parameters such as transmissivity, specific capacity and well distance radius. Like the geological analysis in the study area, the performed hydraulic tests allowed the verification of the homogeneity of the hydrogeologic setting. Recently, the city of Rovigo was affected by an impressive ground settlement induced by the abstraction of groundwater from a wellpoint system located in the core, during the building phases of an underground car park. The settlement was propagated and increased by the geological setting; the palaeo river sediment

Fig. 6. The two sandy palaeorivers crossing the core of the city of Rovigo (Italy) and the study area (blue circle) (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.).

facilitated the head loss with an anisotropic geometry [45,47]. Thus the Authorities were careful about the problem connected to the abstraction activity from the first aquifer. 3.3. The numerical modelling To verify the effects of the GHPS on the hydrogeological system the numerical Feflow code was used [48]. The three dimension numerical model was carried out by a previous hydrogeological model and it is developed by 582.000 mesh elements and 310.000

Fig. 7. Geological and hydrogeological model used for the groundwater and thermal modelling of study area.

A. Galgaro, M. Cultrera / Applied Thermal Engineering 57 (2013) 107e115

mesh nodes, according to the Delaunay Principle. Fig. 7 shows the framework, characterized by 3 hydrogeological layers. The first is an impermeable layer whose hydraulic conductivity is 1E-08 m/s in accordance with the lithology; the second layer is an aquifer whose assigned hydraulic conductivity derived from the pumping tests and calibration (4.04 E-4 m/s); finally, the third layer is an impermeable clay, here used to fit the optimal thermal boundary conditions (1E-08 m/s). This aquifer is confined and both thickness and hydrogeological parameters are assumed constant in the model domain. The regional and local hydraulic gradient was employed to set the flow boundary conditions. The hydraulic head is assumed constant at 6.9 m asl on the west side of the model and 6.4 m asl on the east side; this configuration determines a correspondent hydraulic gradient fitting the measured groundwater flownet (0.00025, see Fig. 4). A sensitivity analysis shown that longitudinal and transversal dispersivity, porosity, heat conductivity of materials are affecting the results of proposed model very less, if compared with both hydraulic conductivity and gradient, too. Therefore the standard input parameters obtained by bibliography for the investigated materials were carried out, according to the property of materials and the in situ investigations, such as pumping tests [32,49e51]. The geothermal boundary conditions are the following. No geothermal gradient is considered a priori because the thin thickness of layers; at the first slice a yearly average temperature constriction at the ground level is fixed (13.2  C); the aquifer temperature is considered in correspondence of the in and out groundwater borders of the model (17.0  C). Similar results are obtained using the normal regional geothermal flux at the bottom and the monthly average temperature at the top. Pumping test is used for determination of aquifer parameter. The measured groundwater flownet was used to calibrate the flow modelling of the study area. The pumping tests were used to calibrate and validate the numerical flow model, too.


3.4. Results The analysis with the numerical code allowed the demonstration of the serious interference between injection and abstraction wells of the thermal doublet system. The consequence of this interference is the impracticality of the GWHP system in the hydrogeological system simulated. The preliminary analysis of results is the evaluation of the well radius of influence on the well doublet system, as Fig. 8 shows. At 30 m the head perturbation is almost 0.3 m and at 70 m the influence of the well doublet is negligible. The induced head changes on the aquifer are quite negligible because the abstraction and injection wells cancel each other when a few metres apart and the balance of water is zero. 4. Discussions The finite element numerical simulation obtained a great deal of important information for a good analysis of the effects of heat exchange on the local hydrogeological system. Both influence radius and seepage depth are not very large outside the heat exchange site because the abstraction well condition is compensated by the near injection well (21 m); over 60 m from the barycentre of the heat exchange system, the groundwater head is negligible (Fig. 9). The thermal boundary from injection wells reaches the abstraction wells after a few weeks from the start of the heat exchange process. The geological and hydrogeological setting is quite uniform and in this case we can try to use the suggested analytical solution (6). When applying the hydrogeological parameter obtained from the hydrogeological tests and used for the implementation of the numerical model, the minimal distance D must be more than 200 m! This simple analysis using analytical solutions lead to understanding that the planned distance of 21 m was too small to use the

Fig. 8. The drawdown isolines in a steady condition; the influence of doublet well is negligible at 70 m.


A. Galgaro, M. Cultrera / Applied Thermal Engineering 57 (2013) 107e115

Fig. 10. In this scenario, the abstraction wells (lines a and b) reach breakthrough time only after 3.5e4 days, when the local temperature increases with respect to the regional reference temperature (green line); the oscillatory behaviour of the curve is due to the eating/cooling daily timetables (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.).

5. Conclusions and future works

Fig. 9. Scenarios after 1 day (A), 4 days (B) and 80 days (C) from starting of activity. a and b are the extraction wells, c and d the injection ones. Obs_1 is the observation well; obs_2 is out of view (Fig. 8).

GWHP system correctly [24,52]. In fact, the mathematical model predicted the breakthrough time after only few days starting from the exploitation of groundwater heat, depending by the parameters used (Figs. 9 and 10).

The authors had a great amount of evidence from many application case studies that thermal short circuits (TSCs) occur when the groundwater hydraulic is too low in relationship to the abstractioneinjection well system. TSCs might occur quite frequently in the Venetian-Po plain area because of the low hydraulic groundwater gradient and the lack of space for a doublet well system in urban area. A prior analytical evaluation is the first step when studying the impact of a GWHP on the aquifer system; this work contributes to performing this preliminary analysis by identifying the most important analytical solutions used for this purpose. In Italy, the use of the GWHP is still limited due to a lack of information on the advantages offered by these systems and because of their high initial costs. In some cases, these costs may be the result of an oversized design of the GWHP, caused by a lack of reliable estimates for heat exchange rates. Even worse, it may occur, as the case study highlights, that the GWHP could be erroneously evaluated and the payback time could increase or the GWHP could be unexploited. If these rates and the hydrogeologic sustainability could be predicted with high accuracy during the design stage of GWHP systems, their initial costs can be reduced and optimized [53]. Thus, a detailed analysis is strongly recommended to avoid a very serious problem when using the GWHP. In this case, a finite element model could carried out using a package software. When

A. Galgaro, M. Cultrera / Applied Thermal Engineering 57 (2013) 107e115

analysing a thermal short circuit, the authors suggest using a numerical code such as Feflow, which calculates the injection temperature through the calculations of abstraction temperatures. Acknowledgements Special thanks go to the geologists Dr. F. Zambon of SIGEO and Dr. P. Zangheri for the helpfulness for some hydrogeological data and for the aquifer tests. Thanks also go to Dhy-Wasy for the numerical support; thanks go to Dr. P. Shaltz of Dhy-Wasy for kind suggestions. Moreover the authors are very grateful to SINCRONO s.r.l. for the use of investigation data. References [1] C.B. Andrews, The impact of the use of heat pumps on ground-water temperatures, Ground Water 16 (1978) 437e443. [2] F. Ampofo, G.G. Maidment, J.F. Missenden, Review of groundwater cooling systems in London, Applied Thermal Engineering 26 (2006) 2055e2062. [3] I.B. Fridleifsson, Present status and potential role of geothermal energy in the world, Renewable Energy 8 (1996) 34e39. [4] J.-Y. Lee, J.-H. Won, J.-S. Hahn, Evaluation of hydrogeologic conditions for groundwater heat pumps: analysis with data from national groundwater monitoring stations, Geosciences Journal 10 (2006) 91e99. [5] R. Ooka, Y. Nam, Groundwater use for thermal energy, in: S. Takizawa (Ed.), Groundwater Management in Asian Cities, vol. 2, Springer, Japan, 2008, pp. 193e206. [6] T.E. Dwyer, Y. Eckstein, Finite-element simulation of low-temperature, heatpump-coupled, aquifer thermal energy storage, Journal of Hydrology 95 (1987) 19e38. [7] S.M. Hasnain, Review on sustainable thermal energy storage technologies, part I: heat storage materials and techniques, Energy Conversion and Management 39 (1998) 1127e1138. [8] H.O. Paksoy, Z. Gürbüz, B. Turgut, D. Dikici, H. Evliya, Aquifer thermal storage (ATES) for air-conditioning of a supermarket in Turkey, Renewable Energy 29 (2004) 1991e1996. [9] H.O. Paksoy, O. Andersson, S. Abaci, H. Evliya, B. Turgut, Heating and cooling of a hospital using solar energy coupled with seasonal thermal energy storage in an aquifer, Renewable Energy 19 (2000) 117e122. [10] H.E.S. Fath, Technical assessment of solar thermal energy storage technologies, Renewable Energy 14 (1998) 35e40. [11] V.A. Fry, Lessons from London: regulation of open-loop ground source heat pumps in central London, Quarterly Journal of Engineering Geology and Hydrogeology 42 (2009) 325e334. [12] G.W. Huttrer, Geothermal heat pumps: an increasingly successful technology, Renewable Energy 10 (1997) 481e488. [13] P. Blum, G. Campillo, W. Münch, T. Kölbel, CO2 savings of ground source heat pump systems e a regional analysis, Renewable Energy 35 (2010) 122e127. [14] D.P. Jenkins, R. Tucker, R. Rawlings, Modelling the carbon-saving performance of domestic ground-source heat pumps, Energy and Buildings 41 (2009) 587e595. [15] E. Morofsky, History of thermal energy storage, in: H.Ö. Paksoy (Ed.), Thermal Energy Storage for Sustainable Energy Consumption, vol. 234, Springer, The Netherlands, 2007, pp. 3e22. [16] Y. Shaw-Yang, Y. Hund-Der, An analytical solution for modeling thermal energy transfer in a confined aquifer system, Hydrogeology Journal 16 (2008) 1507e1515. [17] J. Bear, Hydraulics of Groundwater, McGraw-Hill, United States, 1979. [18] H.J.G. Diersch, O. Kolditz, Coupled groundwater flow and transport: 2. Thermohaline and 3D convection systems, Advances in Water Resources 21 (1998) 401e425. [19] D. Banks, Thermogeological assessment of open loop well doublet schemes e an analytical approach, in: Proceedings of the IAH (Irish Group) 27th Annual Groundwater Conference, International Association of Hydrogeologists (Irish Group), Tullamore, Offaly County (Ir), 2007. [20] F. Grant, Heterogeneity and thermal modeling of ground water, Ground Water 45 (2007) 485e490. [21] L. Ni, H. Li, Y. Jiang, Y. Yao, Z. Ma, A model of groundwater seepage and heat transfer for single-well ground source heat pump systems, Applied Thermal Engineering 31 (2011) 2622e2630. [22] G. Bodvarsson, Thermal problems in the siting of reinjection wells, Geothermics 1 (1972) 63e66. [23] M.J. Lippmann, C.F. Tsang, Ground-water use for cooling: associated aquifer temperature changes, Ground Water 18 (1980) 452e458. [24] C.G. Clyde, G.V. Madabhushi, Spacing of wells for heat pumps, Journal of Water Resources Planning and Management 109 (1983) 203e212. [25] D.H. Freeston, H. Pan, The application and design of downhole heat exchanger, Geothermics 14 (1985) 343e351.


[26] D. Banks, An Introduction to Thermogeology: Ground Source Heating and Cooling, first ed., Wiley-Blackwell, 2008. [27] M. Gropius, Numerical groundwater flow and heat transport modelling of open-loop ground source heat systems in the London Chalk, Quarterly Journal of Engineering Geology and Hydrogeology 43 (2010) 23e32. [28] M. Bonte, P.J. Stuyfzand, A. Hulsmann, P. Van Beelen, Underground thermal energy storage: environmental risks and policy developments in the Netherlands and European Union, Ecology and Society 16 (2011) 15. [29] U.S. Geological Survey, HST3D, in: A Computer Code for Simulation of Heat and Solute Transport in Three-dimensional Ground-Water Flow Systems (2006). [30] C. Clauser, J. Bartels, M. Kühn, D. Motthagy, V. Rath, R. Wagner, A. Wolf, SHEMAT and Processing SHEMAT: Numerical Simulation of Reactive Flow in Hot Aquifers, Springer, HeidelbergeBerlin, 2003. [31] C.J. Gandy, L. Clarke, D. Banks, P.L. Younger, Predictive modelling of groundwater abstraction and artificial recharge of cooling water, Quarterly Journal of Engineering Geology and Hydrogeology 43 (2010) 279e288. [32] H.-J.G. Diersch, D. Bauer, W. Heidemann, W. Ruhaak, P. Schätzl, in: D.W. GmbH (Ed.), Finite Element Formulation for Heat Exchanges in Modelling Geothermal Heating System by FEFLOW, vol. 5, White Papers, Berlin, 2010, p. 93. [33] H. Fujii, T. Inatomi, R. Itoi, Y. Uchida, Development of suitability maps for ground-coupled heat pump systems using groundwater and heat transport models, Geothermics 36 (2007) 459e472. [34] H.-J.G. Diersch, Diersch, Discrete feature modeling of flow, mass and heat transport processes by using FEFLOW, in: WASY-GmbH FEFLOW FiniteElement Subsurface Flow and Transport Simulation System, User Manual/ White Papers, Release 5.4, Berlin, 2009. [35] E. Mantovani, D. Babbucci, C. Tamburelli, M. Viti, A review on the driving mechanism of the TyrrhenianeApennines system: Implications for the present seismotectonic setting in the Central-Northern Apennines, Tectonophysics 476 (2009) 22e40. [36] C. Barbieri, G. Bertotti, A. Di Giulio, R. Fantoni, R. Zoetemeijer, Flexural response of the Venetian foreland to the Southalpine tectonics along the TRANSALP profile, Terra Nova 16 (2004) 273e280. [37] A. Castellarin, R. Nicolich, R. Fantoni, L. Cantelli, M. Sella, L. Selli, Structure of the lithosphere beneath the Eastern Alps (southern sector of the TRANSALP transect), Tectonophysics 414 (2006) 259e282. [38] D. Babbucci, C. Tamburelli, M. Viti, E. Mantovani, D. Albarello, F. D’Onza, N. Cenni, E. Mugnaioli, Relative motion of the Adriatic with respect to the confining plates: seismological and geodetic constraints, Geophysical Journal International 159 (2004) 765e775. [39] F. Bleibinhaus, H. Gebrande, Crustal structure of the Eastern Alps along the TRANSALP profile from wide-angle seismic tomography, Tectonophysics 414 (2006) 51e69. [40] E. Carminati, G. Martinelli, P. Severi, Influence of glacial cycles and tectonics on natural subsidence in the Po Plain (Northern Italy): Insights from 14C ages, Geochemistry, Geophysics, Geosystems 4 (2003). [41] F. Antonioli, L. Ferranti, A. Fontana, A. Amorosi, A. Bondesan, C. Braitenberg, A. Dutton, G. Fontolan, S. Furlani, K. Lambeck, G. Mastronuzzi, C. Monaco, G. Spada, P. Stocchi, Holocene relative sea-level changes and vertical movements along the Italian and Istrian coastlines, Quaternary International 206 (2009) 102e133. [42] A. Amorosi, M.C. Centineo, M.L. Colalongo, F. Fiorini, Millennial-scale depositional cycles from the Holocene of the Po Plain, Italy, Marine Geology 222-223 (2005) 7e18. [43] A. Fontana, P. Mozzi, A. Bondesan, Alluvial megafans in the Venetiane Friulian Plain (north-eastern Italy): evidence of sedimentary and erosive phases during Late Pleistocene and Holocene, Quaternary International 189 (2008) 71e90. [44] Regione del Veneto, Carta Idrogeologica della Pianura Veneta, in: Piano Regionale attività di cava (2003) (in Italian). [45] A. Ferretti, G. Franchinoni, L. Jurina, Valutazione degli effetti di scavi in falda sui cedimenti strutturali di edifici mediante utilizzo di tecniche satellitari SAR, in: Crolli e affidabilità delle Strutture, Napoli (2003) (in Italian). [46] D. Raucoules, C. Colesanti, C. Carnec, Use of SAR interferometry for detecting and assessing ground subsidence, Comptes Rendus Geosciences 339 (2007) 289e302. [47] L. Jurina, A. Ferretti, A Rovigo il caso emblematico di un parcheggio interrato, Permeabilità 3 (2003) (in Italian). [48] H.-J.G. Diersch, Feflow, finite element subsurface flow & transport simulation system e reference manual, in: W. GmbH (Ed.), Finite Element Subsurface Flow & Transport Simulation System, Institute for Water Resources Planning and System Research, Berlin, German, 2009. [49] J. Bear, M.S. Beljin, R.R. Ross, Fundamentals of Ground-Water Modeling, United States Environmental Protection Agency, 1992. Ground Water Issue. [50] C.W. Fetter, Contaminant Hydrogeology, Macmillan, New York, 1992. [51] K. Spitz, J. Moreno, A Practical Guide to Groundwater and Solute Transport Modeling, Wiley, New York, 1996. [52] D. Banks, Thermogeological assessment of open-loop well-doublet schemes: a review and synthesis of analytical approaches, Hydrogeology Journal 17 (2009) 1149e1155. [53] Y. Zhao, Z. Shigang, L. Xun, Cost-effective optimal design of groundwater source heat pumps, Applied Thermal Engineering 23 (2003) 1595e1603.