Available online at www.sciencedirect.

com

ScienceDirect
Energy Procedia 101 (2016) 512 – 519

71st Conference of the Italian Thermal Machines Engineering Association, ATI2016, 14-16
September 2016, Turin, Italy

Numerical Simulation of a Down-hole Heat Exchanger: an

Application to a Case Study in Central Italy
Maurizio Carlinia,*, Elena Allegrinia, Sonia Castelluccia
a
University of Tuscia, via san Camillo de Lellis snc, 01100 Viterbo, Italy

Abstract

According to many recent European and National Directives, ambitious targets are increasingly set so as to ensure a clean and
sustainable future for the forthcoming generations. In order to change from being predominantly fossil fuelled to being fuelled by
locally available sources, low-enthalpy geothermal installations for heating and cooling buildings can be successfully used, such
as Down-hole Heat Exchangers (DHEs). Although in the last decades DHEs have been studied extensively, both numerically and
experimentally, some aspects still need to be further investigated, namely the interaction between aquifer, well, and the DHE
itself. Thus, the present paper investigates the thermal behaviour of a DHE in Central Italy, considering one-year experimental
campaign. The numerical simulation is carried out by the finite-element program Comsol Multiphysics and is validated by the
experimental data concerning the values of temperature within the DHE at a specific depth. In more detail, it is implemented by
a model describing the U-tube and the surrounding aquifer. Although in most of the simulation only a single domain is taken into
account, the adjacent medium is here implemented as a boundary condition. The results for the computational domain are
presented in terms of temperature, highlighting that the numerical simulation is an efficient method to accurately reproduce the
temperature pattern within the DHE.

©
© 2016
2016TheTheAuthors. Published
Authors. by by
Published Elsevier Ltd.Ltd.
Elsevier This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility ofthe Scientific Committee of ATI 2016.
Peer-review under responsibility of the Scientific Committee of ATI 2016.
Keywords:heat transfer; simulation; down-hole heat exchanger.

* Corresponding author. Tel.: +39 0761401343; fax: +39 0761401343.

E-mail address: maurizio.carlini@unitus.it

1876-6102 © 2016 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
Peer-review under responsibility of the Scientific Committee of ATI 2016.
doi:10.1016/j.egypro.2016.11.065
 Maurizio Carlini et al. / Energy Procedia 101 (2016) 512 – 519 513

1. Introduction

According to the European Directive 2009/28/EC which was implemented in Italy in 2010, ambitious targets
need to be achieved in order to ensure a clean and sustainable future for the forthcoming generations. In more detail,
its aim is to reduce greenhouses gases emissions by 20%, to produce 20% of energy from renewable sources and to
decrease the consumption by 20% improving the energy efficiency by 2020. Nowadays, 80% of the present energy
is still generated by fossil fuels, although they are finite resources and their combustion is harmful to the
environment because of Greenhouse Gases (GHGs) and pollutants emissions which contribute to the global
warming problem [1, 2].
Since fossil fuels shortages are predicted for the near future, the use of alternative/Renewable Energy Sources
(RES) for heating and cooling buildings or to produce electricity is strongly encouraged in many countries. As a
consequence, public administrations and private stakeholders have led to investigate new technologies which may
face the growing energy demands by the use of renewable sources alternative to fossil fuels, changing from being
predominantly fossil fuelled to being fuelled by locally available sources. In this scenario, low-enthalpy geothermal
plants, such as Down-hole Heat Exchangers (DHEs) for heating and cooling buildings can be successfully used [1,
2].
Basically, ground source heat pump systems can be distinguished in two different types: earth- coupled or closed-
loop and groundwater or open-loop. In the first case, heat exchangers are underground and located horizontally or
vertically -as in a DHE- or obliquely. Heat is transferred from or to the ground thank to a heat-carrying which
circulates within the exchanger [1, 2].
The exploitation of geothermal resources by DHEs is not characterized by mass withdrawal from the aquifer:
more precisely, since they permit heat transfer without extracting any fluid from the ground, the aquifer depletion is
avoided, as requested by legislative restrictions. However, an important aspect has to be considered: the amount of
withdrawable heat may be limited and is usually less than 100 kW because of the interaction between heat
exchanger, well and aquifer. As a consequence, DHEs are mostly used in small applications, such as buildings,
greenhouses and thermal baths [3, 4].
In order to increase vertical circulation of water and the natural mass transfer between aquifer and well,
promoters can be used. The exchangers consist in pipes or tubes which are located within the well. Because of the
low capacity, these systems are successful up to 150 m of depth. Several designs are nowadays available but the
most common one is a U-tube which extends to near the bottom of the well. Multiple tubes can be alternatively used
[5].
The aim of the present paper is to investigate the thermal behaviour of a DHE in Central Italy, considering one-
year experimental campaign by running steady-state simulations and involving the solution of heat transfer Partial
Differential Equations (PDEs). The numerical analysis is carried out by the finite-element program Comsol
Multiphysics and is validated by the experimental data concerning the values of temperature within the DHE at 44
m-depth.
As it is commonly known, PDEs arise in many subject areas, such as fluid dynamics, electromagnetism, and in
many physical, chemical and biological phenomena. Frequently, the related equations are so complicated that
finding their solutions in closed form or by analytical means are impracticable or possible only for very simple
geometries. Thus, one has to resort to seeking numerical approximations to the unknown analytical solution
describing the real and complex object. These notes are devoted to a particular class of numerical techniques for the
approximate solution of PDEs, that is the Finite Element Method (FEM) [6].
Comsol Multiphysics is a commercial software package which performs equation based-modelling for different
physical processes and applications, and consists of several modules concerning different branches. The Heat
Transfer module helps the user in investigating the heat transfer mechanism, that is conduction, convection and
thermal radiation, and a combination of them or in collaboration with other physics [7].
514 Maurizio Carlini et al. / Energy Procedia 101 (2016) 512 – 519

Nomenclature

Cp specific heat at constant pressure ( J / kg ˜ K )

eb blackbody hemispherical total emissive power
F body force vector (N/m3)
G irradiation or incoming radiative heat flux (W/m2)
J radiosity or total outgoing radiative flux (W/m2)
n normal vector of the boundary
k thermal conductivity ( W/ m˜ K )
p pressure (Pa)
q conductive heat flux vector (W/m2)
q0 inward heat flux, normal to the boundary (W/m2)
Q heat source term (W)
Qvh viscous heating (W)
t time (s)
T absolute temperature (K)
T0 temperature at a boundary (K)
u velocity field (m/s)
Wp pressure work
ε surface emissivity (dimensionless)
μ dynamic viscosity (Pa x s)
ρ density of the material (kg/m3)
ρrefl surface reflectivity (dimensionless)

2. Materials and methods

2.1. Governing equations

Broadly speaking, the simulating process in Comsol Multiphysics encompasses several steps, leading to build a
specific model and following a logical sequence within a tree structure. First, the user is required to define the space
dimension (1D, 2D, 2D axisymmetric or 3D) and the physic of the problem. Then, the study type –whose main
items are stationary or time dependent analyses- has to be selected. Global definitions, such as parameters, variables
or functions, may be assessed by the user in order to quickly draw the object or to specify other parameters within
the simulation (e.g. heat source). The geometry is successfully built by importing files or working with drawing
operations within Comsol Multiphysics [7].
The Heat Transfer module comes stocked with an internal material database containing the material properties of
a number of common fluids, gases and solids. This includes thermal conductivity, heat capacity and density.
However, other material data can be imported or modified by the user. In order to solve the PDE describing the
specific phenomenon, the mesh has to be created, choosing among different available options, namely from extra-
fine to extra coarse meshes, leading to different and accurate solutions. The results and convergence plots of the
problem can be visualised in the graphics window [7].
The heat transfer mechanism within a DHE involves conduction and convection. In the first case, the
phenomenon is governed by the following partial differential equation:

w § wT · w § wT · w § wT · wT
¨k ¸ ¨k ¸ ¨k ¸Q UC p
wx © wx ¹ wy © wy ¹ wz © wz ¹ wt (1)
 Maurizio Carlini et al. / Energy Procedia 101 (2016) 512 – 519 515

or:

wT
’ ˜  k ’T   Q U ˜ Cp (2)
wt
which is the general form of the heat diffusion equation in Cartesian coordinates. The latter provides the basic
tool for heat conduction analysis, and describes an important physical condition, that is conservation energy. Its
solution leads to determine the temperature distribution T(x, y, z) as a function of time. Hence, this equation states
that at any point in the medium the net rate energy transfer due to conduction into a unit volume plus the volumetric
rate of thermal energy generation must equal the rate of change of thermal energy stored within the volume. For a
steady-state problem, the temperature does not change with time and the first term disappears, that is the amount of
energy storage does not vary and hence Eq. 2 reduces to the following form [6]:

’˜  k’T   Q 0
(3)
The foregoing equations emerge from Fourier’s law, which may be considered as the cornerstone of conduction
heat transfer. In order to determine the temperature distribution in a medium, the heat equation must be
appropriately solved: such a solution depends on the physical conditions existing at the boundaries of the medium
and on conditions existing in the medium at some initial time if the problem is time dependent. The heat equation
accepts two basic types of boundary conditions: specified temperature and specified heat flux. The first condition is
a constraint type –which is commonly known as Dirichlet condition- and corresponds to a situation for which the
boundary surface is maintained at a fixed temperature [6]:

T T0 on w:
(4)
The specified heat flux conditions defines the inward heat flux by the following expression:

n ˜ q q0
(5)
The heat transfer in fluids is given by:

U˜
wu
wt
ª
¬
T 2
3

º
 U u ˜’u ’ ˜ « pI  P ’u   ’u   P  ’ ˜ u  I »  F
¼ (6)
wU
 ’ ˜  Uu  0
wt (7)
UC p ˜ u’T ’˜  k’T   Q  Qvh  Wp
(8)

which represents the general expression of the momentum, the continuity and the heat equation for a fluid
respectively. When the node Fluid is added within the Comsol model, the volume force, all the terms related to the
heat sources and pressure work are neglected [6, 7].

2.2. Description of the experimental plant

The experimental plant is located in Viterbo (northern Latium, Italy), an area which hosts several thermal springs
in its territory. Its hydrogeological settings and shallow groundwater basins seem to be particularly suitable for a
516 Maurizio Carlini et al. / Energy Procedia 101 (2016) 512 – 519

wide implementation of DHE systems [8].

According to the drilling log, the well is 60 m deep below the ground level. Its diameter is 150 mm and reaches
the bottom of the aquifer. Moreover, a steel flanged pipe is located on the top of the well in order to contain the
upflow water, which is generated by gases pressure, and reaches 1,5 m above ground level. The heat exchanger
consists in a single vertical U-pipe: the internal and external diameters are 29,1 mm and 33,7 mm respectively and
the total length is 110 m. The pipe is made of steel and its thickness is 4,6 mm. The hydraulic system is made of
PVC pipes connected to the heat exchanger. Moreover, the plant is set through manual valves in order to allow
series or parallel connections between the heat exchangers. The hydraulic system is equipped with a pressure gauge
and a valve to avoid air accumulation within the pipes which may reduce fluid flow and heat transfer [2, 9-11].
The low-enthalpy geothermal system can be drafted as a reservoir which contains hot water and a pipe. Three
different modes of heat transfers occur: natural convection in the groundwater basin, conduction within the wall of
the pipe, and forced convection inside the heat exchanger because of water pumping. The groundwater within the
well is characterised by convective motions which are due to differences of density and consequently cause heat
transfers. Measurements of the groundwater basin temperature were carried out at different depths, 2, 26 and 44 m
respectively: after a very rapid transitory of 15 minutes, the temperature shows a constant trend, reaching 60°C. The
water temperature was measured by a PTC sensor, while the operating conditions were controlled by a monitoring
system [2, 9-11].

Fig. 1. Temperature measurements at different depths.

After defining the temperature profile within the heat exchanger by programming in Matlab and implementing
an iteration loop, a simulation in Comsol Multiphysics was carried out in order to analyse the thermal behaviour of
the well surrounding at 44 m-depth. Three different modes of heat transfers occur within the system: natural
convection in the groundwater basin, conduction within the wall of the pipe, and forced convection inside the heat
exchanger. The latter is due to the pumping of water. The groundwater within the well is characterised by
convective motions which are due to differences of density and consequently cause heat transfers [11]. The
simulation involved the following stages:

x selection of the space dimension, namely 2D to represent the horizontal section at 44 m-depth;
x drawing the geometry of the problem after defining parameters in the Global Definition branch, as shown
in table 1. The geometry consists of 6 domains;
x choice of the material and its thermophysical properties for water and steel according to the Material
Library in Comsol Multiphysics, and for the basalt according to the value given in table 1;
x selection of the physics and definition of the boundary conditions, as shown in figure 2, in order to solve
the differential equation;
x meshing the control volume, following the physical-controlled mesh. A normal size was chosen for the
simulation, consisting of 1852 and 208 domain and boundary elements respectively;
x results visualization in terms of isothermal contours.
 Maurizio Carlini et al. / Energy Procedia 101 (2016) 512 – 519 517

Table 1. Global Definitions.

Parameters Description
D_int_p = 29,1 mm Internal diameter of the pipe
D_ext_p = 33,7 mm External diameter of the pipe
D_w = 150 mm Diameter of the well
D_s = 80 cm Diameter of the surrounding volume
T_44_wat = 60°C Water temperature at 44 m
T_44_p_in = 28°C Inlet water temperature at 44 m
T_44_p_out = 37°C Outlet water temperature at 44 m
k_b_2 = 2,5 W/m/K Thermal conductivity of basalt

Fig. 2. Boundary conditions and meshing phase.

3. Results

The implementation in Matlab shows that the water temperature on the outgoing cross section of the DHE is
equal to 45°C, thus increasing up to 30°C if compared with the initial value at the entrance (15°C). The temperature
difference is only due to the heat transfer occurring between terrain, groundwater basin, pipe and water within the
DHE. The calculation led to define the water temperature trend along the U pipe, namely the spatial coordinate x, as
shown in figure 2. It can be noted that the temperature value rapidly increases as far as the fluid reaches the end of
the DHE. The water exiting the DHE has a limited value if compared with Chiasson et al. (2004) and Culver and
Lund (1999), who considered a 123-m deep well cased to 112 meters with a bottom hole temperature of 160 °C.
However, this value is in good agreement with Steins et al. (2008) who investigated two wells drilled to 55 m. The
survey demonstrated that the outlet temperature ranges between 42 and 77°C, the temperature difference from 2 to
33°C and the heat output from 2 to 49 kW. Moreover, the heat output generated from the low-enthalpy geothermal
system can be successfully used for civil purposes, such as district heating, which provides energy to supply thermal
needs [12].
The result in Comsol Multiphysics is shown in figure 3 in terms of isothermal contours within the DHE and the
well, and considering the surrounding. The maximum temperature of the system -when the depth is equal to 44 m- is
reached within the groundwater basin (59,2°C) because of the influence of the surrounding hot terrain. Moving
closer to the DHE, the temperature decreases: then, if we consider the arrow surface, the direction of the total heat
flux is given and is oriented from the terrain to the DHE. The temperature difference between experimental (figure
1) and simulated results (figure 4) is equal to 0,8°C, i.e. approximately 1,3 %. PTC measurement errors can be
518 Maurizio Carlini et al. / Energy Procedia 101 (2016) 512 – 519

neglected in the present case study because of the limited temperature range. Figure 5 represents the convergence
plot: the convergence is reached very quickly, after 3 iterations.

Fig. 3. Temperature profile of the water within the heat exchanger emerging from Matlab code.

Fig. 4. Isothermal contours on the DHE surrounding.

Fig. 5. The convergence plot.

 Maurizio Carlini et al. / Energy Procedia 101 (2016) 512 – 519 519

4. Conclusion

The implemented model can be successfully used in those areas where shallow thermal ground basins occur and
if their temperature is close to 60°C. Since it was created from measurements in situ, future improvements may be
reached by the validation with more case studies. Besides, considering that the numerical simulation in Comsol
Multiphysics was carried out using geometrical parameters, other situations – different in terms of material and
boundary conditions- may be rapidly solved: the temperature water within a DHE at different depths may be
calculated before drilling, thus predicting the amount of withdrawable heat generated by the system itself and
decreasing the probability of unsuccessful well exploitations.
It is important to underline that the implemented model highlighted the importance of exploiting local available
renewable sources, as shown in [13]. More horizontal sections can be further analysed by developing a Thermal
Response Test, which is commonly used for measuring the ground thermal conductivity and the thermal resistance
R [14].

References

[1] Lo Russo S, Boffa C, Civita MV. Low-enthalpy geothermal energy: An opportunity to meet increasing energy needs and reduce CO2 and
atmospheric pollutant emissions in Piemonte, Italy. Geothermics 2009; 38: 254-62.
[2] Carlini M, Castellucci S, Allegrini E, Tucci A. Down-hole heat exchangers: Modelling of a low-enthalpy geothermal system for district
heating. Math Probl Eng 2012, art. no. 845192.
[3] Carotenuto A, Casarosa C, Vanoli L. Optimizing the position of the tube casing slotted section for geothermal wells with a down-hole
exchanger. Geothermics 2001; 30: 133-57.
[4] Carotenuto A, Massarotti N, Mauro A. A new methodology for numerical simulation of geothermal down-hole heat exchangers. Appl Therm
Eng 2012; 48: 225-36.
[5] Lund JW. The use of down-hole heat exchangers. Geo-heat Center, Oregon Institute of Technology, Klamath Falls, OR, USA, 1999.
[6] Incropera FP, DeWitt DP, Bergman TL, Lavine AS. Fundamentals of heat and mass transfer. 6th ed. John Wiley & Sons Inc.; 2007.
[7] Comsol Multiphysics User's Guide.
[8] Piscopo V, Barbieri M, Monetti V, Pagano G, Pistoni S, Ruggi E, Stanzione D. Hydrogeology of thermal waters in Viterbo area, Central Italy.
Hydrogeol J 2006; 14: 1508-21.
[9] Carlini M, Castellucci S. Efficient energy supply from ground coupled heat transfer source. ICCSA 2010, Part II, LNCS 6017, pp. 177-9.
Springer-Verlag Berlin Heidelberg, Germany.
[10]Carlini M, Monarca D, Biondi P, Castellucci S, Honorati T. A simulation model for the exploitation of geothermal energy for a greenhouse in
Viterbo province. International Conference Ragusa SHWA September 16-18, 2010, Ragusa Ibla Campus, Italy.
[11]Carlini M, Castellucci S. Modelling the vertical heat exchanger in thermal basin. ICCSA 2011, Part IV, LNCS 6785, pp. 277-86. Springer-
Verlag Berlin Heidelberg, Germany.
[12]Muñoz M, Garat P, Flores-Aqueveque V, Vargas G, Rebolledo S, Sepùlveda S, Daniele L, Morata D. Estimating low-enthalpy geothermal
energy potential for district heating in Santiago Basin-Chile. Renew Energ 2015; 76: 186-95.
[13]Lei HY, Dai CS. Heat transfer analysis of centric borehole heat exchanger with different backfill materials. Proceedings World Geothermal
Congress 2015, Melbourne, Australia.
[14]Acuna J. Distributed thermal response tests – New insights on U pipe and coaxial heat exchangers groundwater-filled boreholes. Doctoral
Diss., KTH School of Industrial Engineering and Management, Stockholm, Sweden, 2013.