Energy 162 (2018) 630e644

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

A novel single-well geothermal system for hot dry rock geothermal

energy exploitation
Wenbo Huang, Wenjiong Cao, Fangming Jiang*
Laboratory of Advanced Energy Systems, Guangdong Key Laboratory of New and Renewable Energy Research and Development, CAS Key Laboratory of
Renewable Energy, Guangzhou Institute of Energy Conversion, Chinese Academy of Sciences (CAS), Guangzhou 510640, China

a r t i c l e i n f o a b s t r a c t

Article history: Existing hot dry rock geothermal projects are commonly confronted with some technical issues, such as
Received 30 September 2017 corrosion and scaling, and water loss. To resolve these issues, the present work proposes a novel system
Received in revised form for mining hot dry rock geothermal energy, in which a reservoir is combined with a heat pipe system.
30 July 2018
The new system encompasses a heat pipe placed in a single-well to extract hot dry rock geothermal
Accepted 7 August 2018
energy, while an artificial reservoir is built around the main endothermic region of the well, which is
permeable and saturated with carbon dioxide (CO2). This wellbore structure design may stimulate a
stronger natural convection in the reservoir, resulting in a higher thermal power production. To evaluate
Keywords:
Hot dry rock
the proposed system, an extensive numerical investigation was conducted. The comparison of the pro-
Enhanced geothermal system posed system with the conventional downhole heat exchanger (DHE) system in terms of heat extraction
Supercritical CO2 performance indicates clear superiority of the proposed system primarily due to the associated ther-
Earth-deep geothermal energy mosyphon effect of CO2 fluid in the reservoir. To better understand how operating and design variables
Downhole heat exchanger affect the heat extraction performance of the system, a detailed sensitivity analysis was conducted taking
Heat pipe into consideration a wide range of possible configurations and working conditions. The eventually ob-
tained knowledge will guide the design of the system in practice.
© 2018 Elsevier Ltd. All rights reserved.

1. Introduction in the hot impervious crystalline basement rocks. Investigation

Geothermal energy is an important branch of the so-called
renewable energies and it is typically characterized by good reli-
ability due to its lack of dependence on the weather condition,
which is a major advantage, in particular, for electricity power
generation. The World Energy Assessment (WEA) estimates that
global annual potential of available geothermal energy is about
600,000 EJ [1]. Considering the present annual global energy con-
sumption is about 570 EJ [2], geothermal energy is practically an
infinite energy resource. Present exploitation of geothermal energy
falls primarily into the hydrothermal resource category, in which
natural fracture networks contain a fluid and allow its circulation.
The requirement of a particular type of geologic structures restricts
the construction of hydrothermal electricity power stations to a
limited number of regions in the world.
Hot dry rock (HDR) energy is another form of geothermal en-
ergy, representing a vast store of thermal energy that is contained
* Corresponding author.
E-mail address: fm_jiang2000@yahoo.com (F. Jiang).
shows more than 90% of the total US geothermal resource is stored
in HDR [3]. Aiming to extract the energy stored in HDR, the pro-
totype of an Enhanced Geothermal System (EGS) was designed and
implemented by the Los Alamos National Laboratory in the 1970s
[4]. The initial concept is straightforward: for low-permeability
rocks, a series of rock-fracturing procedures, such as hydraulic
stimulation to create an artificially fractured reservoir is performed.
By circulating water through the stimulated region, heat can be
continuously extracted from the rock, just like a natural hydro-
thermal system. EGS has drawn a considerable amount of interest
over the past 40 years. However, so far significant financial risks
still exist for EGS projects since it remain as a major challenge to
effectively control the quality of the fractures’ network in EGS
reservoir, which to a great extent determines the production po-
tential of the EGS [5]. Some practical issues also lead to operation
and maintenance costs, such as corrosion and scaling in wellbores
and power plant components, and the loss of working fluid while
circulating [6,7].
In this context, some researchers considered alternative systems
to use deep geothermal energy, and particular attention was given

https://doi.org/10.1016/j.energy.2018.08.055
0360-5442/© 2018 Elsevier Ltd. All rights reserved.
 W. Huang et al. / Energy 162 (2018) 630e644
to the Downhole Heat Exchanger (DHE); DHE typically consists of a
system of U-pipes or coaxial tubes suspended in the well, through
which heat carrier fluid is pumped to extract the thermal energy
from the well. Since the working fluid is circulated in a closed-loop
system, the problems like fluid loss, corrosion and scaling are
essentially avoided. DHEs have been widely used for the direct
utilization of geothermal resources [8e11]. Recently, many studies
presented analytical and numerical models to estimate the pro-
duction potential of DHE geothermal systems used in abandoned
oil or gas wells for mining HDR geothermal energy [12e18]. For
instance, Kohl et al. [12] investigated the behavior of a heat
exchanger installed at 2302 m subsurface in an abandoned bore-
hole in Switzerland and the results indicated the produced thermal
power increases from 40 kW to over 200 kW with a properly set
fluid circulation rate. Nalla et al. [13] conducted a numerical study,
which dealt with a vertical borehole drilled to a total depth of
5593 m; the results indicated that the system hardly supplies suf-
ficient energy to generate 50 kW electrical power. Based on nu-
merical results, Bu et al. [15] estimated the electrical power
production of a DHE system and found the energy production from
abandoned oil wells depends largely on the fluid flow rate and the
geothermal gradient. Nevertheless, most of these studies consid-
ered only heat conduction between rock and well. Since the heat
conduction flux from rock is primarily determined by thermal
conductivity of the rock, heat exchange area and temperature dif-
ference between the average temperature of rock and working fluid
[19], it is difficult to further improve the heat extraction perfor-
mance of such a system, considering the limited number of pa-
rameters that can be altered.
Several conceptual designs about combining a permeable
reservoir with a DHE system for HDR heat extraction were pro-
posed [20e22]. Wang et al. [20] investigated a single-well DHE
with connection to artificial fracture intervals, through which the
working fluid was circulated to take advantage of the resulting
thermosiphon effect. Feng et al. [21] proposed a coaxial DHE
concept, in which a downhole pump was placed inside a horizontal
wellbore to generate fluid convection for better heat transfer effect.
Shi et al. [22] established a 3D unsteady state numerical model to
study the coupling effect of ground water flow and the heat transfer
of DHE system. These studies proved that convective flow and the
buoyancy driven flow in permeable formation around wells can
obviously increase the production capacity of DHE systems.
Another branch research direction regarding to DHEs is the
application of heat pipe system. The heat pipe has the advantage of
transferring heat very effectively from its high to the low temper-
ature end relying on the working fluid phase transition. The heat
pipe, compared with the U-tube or coaxial tube DHEs, has the
advantage of not requiring external pumping power and it can
enhance the heat transfer rate due to very high boiling and
condensation heat transfer coefficients. Moreover, the heat pipe
system is able to maintain a higher temperature difference between
working fluid and reservoir, which will promote the fluid natural
convection in reservoir and certainly lead to an increased heat
extraction rate. Using heat pipe to harvest geothermal heat was
reported by quite a few researchers [23e26]. Thermal performance
of a conventional single-tube thermosyphon is generally limited by
the counter-resistive interaction between the oppositely-flowing
vapor and liquid. Vasil'Ev [23] proposed a variant heat pipe in
1990, in which an additional tube is added to separate the vapor
and liquid flow, enabling its utilization in geothermal energy
exploitation. Kusaba et al. [24] developed a 150 m-long heat pipe, in
which the liquid feeding tube was equipped with showering noz-
zles to form a uniform liquid film over the large evaporator internal
surface. More recently, Ebeling et al. [26] conducted simulation and
experimental validation work of a 400 m-long geothermal heat
631
pipe system, in which CO2 was used as the working fluid of heat
pipe. These studies indicate the heat pipe system has the potential
for mining the earth-deep geothermal energy.
In this study, a novel design for mining hot dry rock geothermal
energy is proposed, which is a combination of a fractured reservoir
and a heat pipe system. The basic configuration is a heat pipe
system installed in a single-well with the purpose of extracting
geothermal energy. In addition, a fractured reservoir is built around
the main endothermic region of the well, which is permeable and
saturated with a second working fluid. A heat pipe, instead of a
pump-assisted flow through a U-tube or coaxial tube, is used to
transfer geothermal heat to the up-ground. During the heat
extraction process, the temperature difference between the evap-
orator of heat pipe and rock can induce a natural flow in the
reservoir, which will lead to an increased heat flux between the
wellbore and reservoir. Thus, the proposed system can have a
higher heat production as compared to the conduction-only DHE
system, while it shares all the advantages of the closed-loop DHEs
mentioned in the previous paragraphs.
The objective of the present work is to assess the production
potential for mining hot dry rock geothermal energy by this com-
bined Reservoir and Heat Pipe System (RHPS). Detailed sensitivity
analyses are conducted taking into consideration a wide range of
possible configurations such as well geometry and depth, working
fluid, and operation conditions, with the aim of understanding how
operation and design variables affect the heat extraction perfor-
mance of the system and providing guidance towards improved the
design of RHPS.
2. The combined reservoir and heat pipe system
The RHPS we propose for mining hot dry rock geothermal en-
ergy combines features of the heat pipe and the fractured reservoir,
as depicted in Fig. 1. A heat pipe, instead of a pump-assisted flow
through a U-tube or coaxial tube, is placed into a vertical well to
extract geothermal energy. An annulus flow channel exists be-
tween the wellbore casing and the heat pipe outer shell. In addi-
tion, a reservoir is built around the main endothermic region of the
well, which is permeable and saturated with a second working
fluid.
The RHPS requires a permeable reservoir saturated by the
working fluid in order that natural convective flow is induced by
the thermosyphon effect. The wellbore casing in the reservoir re-
gion allows the reservoir working fluid flowing across the heat pipe
evaporator surface, as depicted in Fig. 1. During heat extraction, the
temperature difference between well and rock induces a natural
convective flow in the reservoir, which will lead to an increased
heat flux at the heat pipe evaporator outer surface. The working
fluid in the reservoir and that in the heat pipe are segregated. The
proposed system uses CO2 and water (or other suitable fluid) as the
working fluid in the reservoir and in the heat pipe, respectively;
CO2 is an attractive geothermal working fluid because its specific
thermodynamic and fluid dynamics properties suggest that it can
transfer geothermal heat more efficiently and its large thermal
expansion coefficient contributes significantly to the buoyancy
force. Pruess [27] quantitatively evaluated the heat extraction of an
EGS plant at Soultz-sous-Fore ^ts using CO2 as the working fluid. The
simulation results indicate that the heat extraction rate with CO2 is
approximately 50% higher than that for water. It is also noted that
the use of CO2 as the working fluid in the reservoir would have the
ancillary benefit of providing geologic storage for the CO2. Ran-
dolph et al. [28] demonstrated that combining geothermal energy
extraction with CO2 sequestration vastly improves the economic
feasibility of the CO2 capture and storage.
A heat pipe, instead of a pump-assisted flow through a U-tube or
632 W. Huang et al. / Energy 162 (2018) 630e644
Fig. 1. Conceptual design of mining heat from earth-deep hot dry rock by a combined reservoir and heat pipe system. “Blue” and “red” lines signify fluid flow and heat flow,
respectively.
coaxial tube, is used to transfer geothermal heat to the up-ground.
As we have discussed in the introduction part, the heat pipe has the
advantage of not requiring pumping power and it has the potential
to obtain a higher power production rate since the heat transfer by
phase transition is more efficient than the convective heat transfer.
In summary, the advantages of the proposed system are: i)
Avoidance of fluid loss, corrosion and scaling problems as a closed-
loop system, and eventual saving in what concerns costs with
drilling wells, when the system is installed in an abandoned oil or
gas well; ii) No pumping power requirements; iii) Higher produc-
tion performance as compared to the DHE system; iv) Potential for
CO2 sequestration.
3. Numerical model and simulation approach
Equivalent porous medium (EPM) model is widely used for
simulating the flow and heat transfer processes in fractured rocks,
because of its simple data requirements and computational effi-
ciency. The existing EPM models can be roughly classified into two
categories in terms of physical description to the fractured rock
mass: the single porosity model and the dual porosity model. The
single porosity model represents the property of fractures and rock
matrix by a certain macroscopic approximation or average. This
approach may be applicable to modeling flow and heat transfer
processes in fractured media under near-thermodynamic-
equilibrium conditions [29] which, however, is hard to accom-
plish when rapid flow and transport processes occur in fractured
reservoirs [30]. The dual porosity model treats the heat reservoir as
a porous medium consisting of two distinct porous regions: large-
porosity fracture network and rock matrix of relatively smaller
porosity [31]. Global flow in the reservoir occurs mainly through
the fracture network, which is described as an effective porous
continuum. Rock matrix and fractures may exchange fluid (or heat)
locally by means of “interporosity flow”, which is driven by the
difference in pressures (or temperatures) between matrix and
fractures. The concept of dual porosity has been widely adopted in
current geothermal models (e.g. Refs. [32e34]) though an accurate
specification of model parameters may be very difficult as a clear
differentiation of these two porous regions is normally hard, even if
not impossible.
Some recent research work considered another approach, called
local thermal non-equilibrium model [30,35,36]. The model ac-
counts for the actual existence of temperature difference between
the rock matrix and the fluid flowing in fractures and uses two
energy equations to describe heat transfer in the rock matrix and in
the fluid respectively. Heat exchange between fluid and rock matrix
is modeled by Newton's law of cooling assuming the heat exchange
rate is directly proportional to the fluid-rock temperature differ-
ence. The proportionality coefficient, termed as the equivalent
solid-fluid heat transfer coefficient, can be measured
 W. Huang et al. / Energy 162 (2018) 630e644 633

experimentally or calculated from fine-scale models such as
fracture-scale model [30,36].
In the present work, a previously developed transient three-
dimensional numerical model [30,37] is used for the simulation
of heat extraction process in the geothermal reservoir. The model
treats the geothermal reservoir as a porous medium and adopts
local thermal non-equilibrium method to model the heat transfer
process between the solid rock matrix and the fluid in fractures.
The code was successfully used in several studies related to the
simulation of EGS long-term heat extraction processes [38e40]. In
the present work, we further implement in the code the sub-model
dealing with the RHPS system.
reservoir is in super-critical state; (4) the fluid flow in the porous
region is in laminar regime and the Reynolds number is sufficiently
low to allow the use of Darcy law; (5) absence of fluid-structure
interactions, including, among others, chemical dissolution/depo-
sition and physical pressing. The governing equations describing
the conservation of mass, momentum and energy are formulated as
follows:
Conservation of mass (mass continuity equation):
vðεrÞ
þ VðruÞ ¼ 0 (1)
vt
Conservation of momentum:

3.1. Physical and computational model


v rf uε
u u
u m
þ V rf $ ¼ VP þ V mV$  u þ rf g (2)
An axisymmetric computational region is selected for the un- vt ε ε ε K
derground portion of the RHPS system, which is treated as a single-
Energy equation in the fluid:
domain consisting of multiple sub-regions, as schematically dis-
played in Fig. 2, namely: i) the heat reservoir; ii) the annulus flow  
v εrcp T f h  i


eff
channel between the wellbore casing and the heat pipe outer shell; þ V$ rcp T f u ¼ V$ lf VTf þ hv Ts  Tf (3)
iii) the rock enclosing the heat reservoir. The simulation domain is vt
taken as a porous medium, in which the different sub-regions have Energy equation in the rock or rock matrix in the porous
distinct porosity and permeability. The rock enclosing the heat reservoir:
reservoir is impermeable to fluid, i.e., ε ¼ 0 and K ¼ 0. The annulus
 


flow channel within the wellbore casing is set with higher porosity v ð1  εÞrcp T s
and permeability as compared to those of the reservoir. ¼ V$ leff
s VTs  hv Ts  Tf (4)
vt

3.2. Mathematical model
The mathematical model entails the following assumptions: (1)
only single-phase flow is taken into consideration; (2) the porous
region is fluid (CO2) saturated, consequently, no immiscible gas is
trapped in the porous medium; (3) the CO2 fluid sequestrated in the
u, p, Tf and Ts are the primary variables to be solved, denoting the
superficial fluid velocity vector, fluid pressure, fluid temperature
and rock temperature, respectively. r and cp with subscript “s” and
“f” respectively indicate the density and specific heat capacity of
fluid and rock, respectively. m is the viscosity of CO2 fluid. The heat
conductivity of fluid and rock matrix in the reservoir are

Fig. 2. Schematic of the underground portion of RHPS and its main dimensions.
634 W. Huang et al. / Energy 162 (2018) 630e644

implemented with the effective quantity leff, i.e. leff s ¼ ls(1-ε)

1.5
and to the thermal property of fluid and solid, but also is affected by the
leff
f ¼ lfε 1.5
, where ε denotes the local porosity in the reservoir. Note inner structure of porous medium. For the fractured reservoir, the
that the Bruggemann approximation was implemented here, which value of hv is selected as 1 W/K/m3 referring to Ref. [30]. However, it
assumes the effective transport property can be approximated by is difficult to determine the value of hv in the annulus flow channel,
the intrinsic physical property of the medium multiplied by the since the inner structure of flow channel could be flexible. Thus, we
component volume fraction raised to the power of 1.5. ε denotes the set the value of hv in the flow channel at 1000 W/K/m3 for the base
local porosity in the reservoir. case, which is high enough to bring thermal equilibrium to the fluid
The last term, ±hv(Ts - Tf), of Eqs. (3) and (4) describes the heat and solid in flow channel, and lower values of hv will be considered
exchange between rock matrix and heat transfer fluid, where the in comparison cases (Table 2).
parameter, hv, denotes the volumetric heat exchange coefficient The density, specific heat capacity, thermal conductivity, and
between solid and fluid. viscosity of working fluids (CO2 and water) are taken as functions of
pressure and temperature. Further details about determining the
3.3. Model parameters variable thermophysical properties can be found in Ref. [42].
The mesh used to discretize the whole computational domain
The reservoir is located at the earth-deep, a few kilometers encompasses 2536 structural grid elements. Grid refinement was
below the surface. The dimensions of the computational region made for the annulus flow channel and the regions near the well-
were presented in Fig. 2, where h indicates the vertical length of the bore to ensure sufficient resolution to describe the physical pro-
heat reservoir. rpip, rcas and rres denote the radius of the heat pipe, cesses. Grid-independence tests have conducted by comparing the
wellbore casing and heat reservoir, respectively. The impermeable heat extraction rate (see the definition in Section 4.1.1) of the base
rock has a thickness of 250 m, which is sufficient to shield the case using the current mesh to that using a new mesh with 4438
temperature field within the reservoir from the influence of the grid elements, and the maximum difference within 30 years’ heat
rock boundary condition. The top and bottom boundaries of the extraction process is less than 1%, indicating that the present mesh
impermeable rock are taken as constant temperature boundary, system can give solutions of adequate accuracy.
while the lateral boundary of the impermeable rock is assumed to
be adiabatic; the mass and heat transfer satisfy continuity condi- 3.4. Model validation
tions at the interface of the impermeable rock and reservoir. We
treat the heat pipe modeling with the simplification that all the Validating the present numerical model in actual scale is diffi-
regions of the evaporator are at the same temperature and the heat cult owing to lack of comprehensive studies of the proposed RHPS
transfer between the heat pipe and rock only occurs at the evap- system so far. Instead, the present code is validated through a
orator section. Thus, the boundary conditions for the heat pipe can similar problem, which has been widely studied experimentally
be readily set as: i) the surface in contact with the reservoir is fixed and numerically. As shown in Fig. 3, the problem describes a two-
at a constant temperature; and ii) the surface in contact with the dimensional rectangular enclosure, which is vertically divided
impermeable rock is taken as adiabatic boundary. Initially, the into a fluid layer and a porous layer, with the vertical walls being
temperature of the rock linearly increases with the depth, and fluid held at different temperatures and the connecting horizontal walls
and rock are in thermal equilibrium. The pressure of reservoir are adiabatic. Natural convective flow is driven by the temperature
center is fixed to a constant value which is calculated from reservoir difference between the two vertical walls. The described heat
depth and pressure gradient. transfer and fluid flow process is similar to that between the flow
The parameter settings are listed in Table 2, including configu- channel and heat reservoir of the proposed RHPS.
rations and working conditions considered for the different com- Beckermann et al. [43] did an experimental and numerical study
parison cases. For the base case, the radius of heat pipe and
wellbore casing are 0.1 m and 0.2 m, respectively, resulting in a gap
of 0.1 m for the annulus flow channel. The ground temperature is
30  C. The reservoir locates 4000 m underground with a
geothermal gradient of 0.035  C/m and a pressure gradient of
10 MPa/km. Thus, at the initial stage, the temperature and pressure
of reservoir center are about 170  C and 40 MPa, respectively. The
artificial reservoir has a length of 1000 m and a radius of 50 m,
which is assumed to be homogeneously fractured with constant
and uniform values of porosity and permeability. The porosity was
selected as 0.05. According to the available literature data [27,41],
the permeability of an EGS artificial reservoir with fracture spacing
of 50 m is about 5  1014 m2. Calculated from this value, the
permeability is 1012, 1013, 1014 m2 for the fractures of spacing
2.5, 25, and 250 m, respectively. Since the size of the RHPS artificial
reservoir is only about one tenth of a typical EGS artificial reservoir,
the fracture density and conductivity in a RHPS reservoir could be
higher than those of an EGS artificial reservoir. Thus, the present
takes the RHPS reservoir permeability to be 1012 m2 for the base
case, and the values of 1013 m2 and 1014 m2 are considered in
parametric study to show the influence of reservoir permeability.
Considering that the channel is equipped with different baffles to
enhance the heat exchange between fluid and rock, it will be
treated as porous medium with a porosity of 0.15 and a perme-
ability of 108 m2 (base case). The volumetric heat exchange coef- Fig. 3. Schematic of the natural convective heat transfer process considered for model
ficient hv in the two-equation energy model is not only correlated validation.
 W. Huang et al. / Energy 162 (2018) 630e644 635

of this problem; the following non-dimensional results yield. transfer in free fluid and porous medium combined system.

eff eff 4. Results and discussion

ð1  εÞls  εlf gbðTh  Tc ÞL3
Rk ¼ ; Ra ¼ ;
lf vf af 4.1. Heat extraction by the combined reservoir and heat pipe system
vf K (5)
Pr ¼ ; Da ¼ To analyze the effect of wellbore structure on the geothermal
af L2
heat extraction performance, we consider three different configu-
x y T  Tc rations for the proposed system: i) the RHPS, as depicted in Fig. 1; ii)
X ¼ ;Y ¼ ;q ¼
L L Th  Tc the RHPS without flow channel, i.e. rpip ¼ rcas ¼ 0.2 m; and iii) the
heat pipe system is set as rpip ¼ rcas ¼ 0.2 m, εres ¼ 0 and Kres ¼ 0.
For the experiments in Ref. [43], the parameter settings are:
The setting rpip ¼ rcas indicates that there is no annular channel, and
Rk ¼ 1.397, Ra ¼ 3.028  107，Pr ¼ 6.97，Da ¼ 7.354  107,
the setting εres ¼ 0 and Kres ¼ 0 means that there is no permeable
ε ¼ 0.36. For model validation, we consider cases which has the
heat reservoir. The three cases have the same wellbore size and
same length scales as in the work by Beckermann et al. [43] The
working conditions.
initial temperature is set to 20  C for the fluid and porous layers,
which is taken as the average temperature of the two vertical walls.
The pressure at the center of simulation region is set to 101.3 kPa. 4.1.1. Heat extraction performance
The values of leff
s , K, and (TheTc) are calculated according to the non- The heat extraction rate, which is defined as the integration of
dimensional numbers and the properties of water at 20  C tem- the heat flux over the heat pipe surface, is used to quantify the heat
perature and 101.3 kPa pressure. The unsteady process is simulated extraction performance of the three systems. As it can be observed
until it comes close to the steady state. from Fig. 5, the heat extraction rate for all the three systems ex-
The obtained results are compared with data taken from periences a slight decay with time; the heat extraction rate of RHPS
Ref. [43], as shown in Fig. 4. It is noticed that the temperatures largely outperforms that of the other two systems.
calculated with the present model show very good agreement with After 10 years of operation, the heat extraction rate of RHPS is
the experimental and numerical results of Ref. [43], proving the 573 kW, which is 4.58 times that of the heat pipe system. However,
validity of present model for simulating natural convective heat the heat extraction rate of RHPS without the annular channel is

Fig. 4. Comparison of temperatures simulated with the present model and those taken from Ref. [43].
636 W. Huang et al. / Energy 162 (2018) 630e644

Fig. 5. Comparison of heat extraction performance for the three different cases.

only 129 kW, which is approximately the same value (125 kW) of within the wellbore of the RHP system has a great contribution to
that for the heat pipe system. This fact clearly indicates that the the buoyancy flow in reservoir.
annular channel flow is crucial to the performance of the RHP The pressure distribution in reservoir is demonstrated by Fig. 8.
system, without this flow channel the reservoir serves no purpose. Since the gravity will cause a great static pressure difference in the
reservoir which makes it difficult to observe the dynamic pressure
difference, a new parameter is define, signified by DPD, to show
4.1.2. Temperature field, seepage flow and pressure distribution
distribution of the dynamic pressure at horizontal section. The
The simulated temperature fields of the three systems are re-
definition of DPD can be found at the caption of Fig. 8. As shown in
ported in Fig. 6. The temperature field of the heat pipe system
this figure, a large pressure gradient exits around the pipe for the
presents a typical conduction-only pattern; the influence of
RHPS system, which is significantly higher than that of RHPS with
boundary conditions, in time, declines due to the diffusion process.
no flow channel. The directions of pressure gradients for the RHPS
For the RHPS with no flow channel, the near-wellbore region is
are consistent with the directions of seepage flow shown in Fig. 7.
permeable, and it can be noted from Fig. 6 in the temperature field
The flow channel has higher permeability than that of the reser-
that the cold fluid slowly accumulates at the bottom of the reser-
voir; therefore, the cold fluid has its downward flow facilitated. In
voir, which is caused by the change in density (buoyancy), while
this way, a large amount of cold fluid will gather at the bottom part
this accumulation is relatively weak and cannot make an obvious
of the channel where a large pressure difference between the
difference in the temperature field, when compared against that of
channel and reservoir occurs, which will impel the cold water to
the heat pipe system. However, observation of the temperature
discharge into the reservoir and lead to a thermosyphon effect.
field of RHPS shows that there is considerable amount of cold fluid
For the RHPS case, at the operation time of 10 years, the heat flux
flowing to the bottom of reservoir which is much larger than that of
through the heat pipe surface and the mass flow rate in the annulus
RHP system with no flow channel.
channel along the depth, where the value “0” at the vertical axis
Fig. 7 displays the seepage flow for the RHPS and RHPS with no
denotes the location at the top of the reservoir, are reported in
flow channel. For the RHP system, the fluid is continuously flow
Fig. 9. It can be noted that both curves present an inflexion point at
into the flow channel at the top region and then discharges into
approximately the same depth (~750 m), which, as it will be
reservoir at the bottom region. It can be noted from Fig. 7 that the
explained further in the texts, is caused by the accumulation of cold
maximum velocity of the seepage flow for the RHPS is almost 4
fluid at the bottom of the reservoir. The heat flux curve shows that
order of magnitude higher than that of the RHP system without
above the inflexion point, as the depth increases, the heat flux
flow channel. It is clearly demonstrated that the annular channel
 W. Huang et al. / Energy 162 (2018) 630e644 637

Fig. 6. Temperature distributions in the reservoir for the three different systems.

Fig. 7. Seepage flow for the RHPS and RHPS with no flow channel (Depth ¼ 0 means at the top of the reservoir; Depth ¼ 1000 m means at the bottom of the reservoir).

decreases at a relatively high rate, which declines beyond the
inflexion point. Also, the mass flow rate in the channel, above the
inflexion point, experiences a quasi-linear increase with increasing
depth; beyond the inflexion point, the mass flow rate presents a
sharp decrease. This particular behavior indicates that there is a
link between the heat flux through the heat pipe surface and the
mass flow rate in the channel, up to the depth of approximately
750 m, the rate of the increase in mass flow rate is in positive
correlation with the heat flux. In combination with the flow di-
rection shown in Fig. 7, the explanation for this behavior is rela-
tively straightforward. The increase of the mass flow rate is due to
the inflow coming from the reservoir. Taking into consideration
that the temperature in reservoir is higher than that in the channel,
the inflow causes an increase of the heat flux.
However, for higher depths beyond the inflexion points, the
mass flow rate sharply decreases, indicating outflow from the
annulus channel. Consequently, no energy is supplied to the
channel and the temperature in the channel will decrease when the
depth increases, which leads to a sharp decline in the heat flux. The
temperature field of RHPS after 10 years of operation (see in Fig. 6)
shows that the inflexion point occurs at the depth where the rock
temperature was obviously brought down by the cold fluid.
Considering that the cold fluid accumulation region grows with the
operation time, as shown in Fig. 6, the depth of the inflexion points
638 W. Huang et al. / Energy 162 (2018) 630e644
Fig. 8. Pressure distribution of horizontal sections at different depth for the RHPS and RHPS with no flow channel. (DPD ¼ P-PD, where, P is the pressure of a monitoring point in
reservoir and PD is the pressure of pipe surface at the same depth with the monitoring point; x is the horizontal distance of monitoring point from the pipe axis).
Fig. 9. Heat flux (solid line) at heat pipe surface and the mass flow rate (dotted line) in
the annulus channel as a function of depth (Time of mining: 10 years).
tends to occur at lower values, which may explain the decline of the
heat extraction rate of RHPS as reported in Fig. 5.
4.1.3. Comparison with downhole heat exchanger system
The production performance of RHPS is compared with that of
downhole heat exchanger system. The data taken from Ref. [15] is
selected for this comparison, since it seems to be a relatively
optimistic estimation for the production of such systems [12e18].
For the RHPS case, the well depth (the depth of reservoir bottom)
and geothermal condition were adapted to the same values as in
Ref. [15]; they are 4000 m and 0.045  C/m, respectively. Other pa-
rameters of the RHPS case are the same as those of the based case.
Besides the heat extraction rate, the electrical power production
is also an important performance indicator. For simplicity, exergy
production, instead of actual electrical power generation, is
employed to represent the potential of electrical power production.
The exergy signifies the maximum electrical power that can be
converted from a certain amount of thermal power, and can be
calculated according to Carnot's theorem, as

T ðSout  Sin Þ
E ¼Q 1 0 (6)
Hout  Hin
where, E is the exergy production rate; Q is the heat extraction rate;
T0 is the environment temperature, considered to be 30  C; S and H
are the entropy and enthalpy, respectively; the subscript “out” in-
dicates the entropy or enthalpy flowing out of the system;
contrarily, the subscript “in” indicates the entropy or enthalpy
injecting into the system.
The comparison of heat extraction rate and exergy production
 W. Huang et al. / Energy 162 (2018) 630e644 639

Table 1
Comparison of heat extraction rate and exergy production for a RHPS and a coaxial-tube DHE system.

Time (years) Heat extraction rate (kW) Exergy production rate (kW)

DHEa RHPS improvement ratio DHEb RHPSc improvement ratio

1 802.1 907.7 13.15% 99.2 170.3 71.60%

2 796.9 902.9 13.29% 98.6 169.4 71.81%
3 793.9 896.8 12.95% 98.2 168.2 71.30%
4 791.8 890.5 12.47% 97.9 167.1 70.56%
5 790.2 884.6 11.95% 97.7 165.9 69.77%
6 788.9 879.1 11.45% 97.6 164.9 69.01%
7 787.7 874.1 10.97% 97.4 164.0 68.29%
8 786.8 869.6 10.53% 97.3 163.1 67.62%
9 785.9 865.5 10.13% 97.2 162.4 67.01%
10 785.2 861.9 9.77% 97.1 161.7 66.47%
a
Directly taken from the Table 2 in Ref. [15].
b
Calculated by Eqn. (6); the production temperature is obtained from the Table 2 in Ref. [15]; the injection temperature is 30  C; the pressure is 2 MPa for production fluid
and 7 MPa For injection fluid.
c
Calculated by Eqn. (6); the entropy and enthalpy are taken the values of vapor (outflow) or liquid (inflow) water at saturation state with temperature of 120  C.

for RHPS and coaxial-tube DHE is presented in Table 1. The heat
extraction rate of RHPS shows a little higher than that of the DHE
system. Nevertheless, the exergy production rate of RHPS is obvi-
ously higher than that of DHE system, which is mainly owing to the
different operational mechanisms between heat pipe and coaxial-
tube. The fluid in the coaxial-tube of DHE is in single phase. After
absorbing thermal energy from rocks, the temperature of the
working fluid in coaxial-tube increases significantly which also
brings a large increase of entropy of the working fluid. By contrast,
during the operation of the heat pipe, the entropy difference be-
tween production steam and injection liquid is relatively small,
leading to a higher usable thermal power production than that of
the DHE system.
4.2. Sensitivity analysis
Detailed sensitivity analyses were conducted regarding a wide
range of possible configurations and working conditions to un-
derstand how the operation and design variables affect the heat
extraction performance of the system. The results may provide
guidance toward design enhancement of RHPS.
The variables considered in this study are organized into three
categories: i) operation variables, including evaporator
temperature of the heat pipe, average reservoir temperature,
geothermal gradient and working fluid in reservoir; ii) wellbore
variables, including wellbore length, radius of heat pipe and well-
bore casing, permeability and porosity of flow channel, volumetric
heat exchange coefficient within flow channel; and iii) reservoir
variables, including radius of the reservoir, permeability of the
reservoir and solid rock heat conductivity within reservoir. For the
sake of convenience, in all the line graphs in this section, the black
solid line is always used to denote the base case, and unless stated
otherwise, the parameters are those set for the base case (Table 2).
4.2.1. Operation variables
Fig. 10 (a) reports the dependence of the heat extraction rate on
the evaporator temperature (Tpip)，while the average temperature
of reservoir (Tave) is fixed at 170  C. The temperature difference
between Tave and Tpip is signified by DT. As expected, the increase of
the temperature difference has an extremely positive effect on the
heat extraction rate of RHPS. At the mining time of 10 years, the
heat extraction rate increases from 179 kW to 1691 kW when Tpip
goes from 145  C to 70  C (e.g. DT goes from 25  C to 100  C). The
heat extraction rate is directly proportional to the product of the
equivalent heat transfer coefficient and the temperature difference.
In fact, since the buoyancy force is stronger, the increase of DT

Table 2
Parameters of simulation cases.

base case other cases

operating parameters D depth of reservoir center (m) 4000 2000, 7000

dP/dy pressure gradient along depth (MPa/km) 10 \
dT/dy geothermal gradient ( C/m) 0.035 0.02, 0.07
Tpip temperature of the evaporator in heat pipe ( C) 120 70e145
DT Tave- Tpip ( C) 50 25,100
working fluid in reservoir CO2 water
Wellbore parameters h wellbore length within reservoir (m) 1000 500, 2000
rpip radius of the evaporator in heat pipe (m) 0.1 0.05, 0.2
w width of flow channel (m) 0.1 0.05, 0.15
εDHE porosity of the porous medium within flow channel 0.15 0.3, 0.99
KDHE permeability of the porous medium within flow channel (m2) 10e8 107, 109, 1010
hv,DHE volumetric heat exchange coefficient within flow channel (W/K/m3) 1000 200, 5000
reservoir parameters rres radius of reservoir (m) 50 25, 100
εres porosity of reservoir 0.05 \
Kres permeability of reservoir flow channel (m2) 10e12 1011, 1013, 1014, 0
hv,res volumetric heat exchange coefficient within flow channel (W/K/m3) 1 \
other parameters εrock porosity of impermeable rock 0 \
Krock permeability of impermeable rock (m2) 0 \
Cps specific heat of solid (J/kg/K) 1000 \
rs density of solid (kg/m3) 2650 \
ls thermal conductivity coefficient of the solid (W/K/m) 2.5 \
640 W. Huang et al. / Energy 162 (2018) 630e644
(a)
(c)
Fig. 10. Heat extraction rate with different operating variables: a) evaporator temperature; b) reservoir depth and geothermal gradient; c) working fluid in reservoir.
would also lead to an increase at the equivalent heat transfer co-
efficient. As a result, if we augment DT by a factor of two, the heat
extraction rate will increase to a value higher than two times of the
original value. Since some issues may arise in geothermal power
generation if the production temperature is too low, there should
be an optimal production temperature with consideration of the
utilization efficiency of available work for a geothermal generator.
Fig. 10 (b) presents the dependence of the heat extraction rate
on the reservoir depth (D). Different geothermal gradients are set
for the three cases in order to achieve the same reservoir average
temperature (170  C). Therefore, the change of the heat extraction
rate shown in this figure should be due mainly to the variation of
the CO2 properties at differing pressure and geothermal gradients.
The average pressure in the reservoir changes from 20 MPa to
70 MPa when the reservoir depth increases from 2000 m to
7000 m, which leads to obvious changes at the density, viscosity
and thermal expansion coefficient values of the working fluid, CO2.
As shown in Fig. 10 (b), the heat extraction rate of the case with
D ¼ 4000 m is higher than the other two cases with D ¼ 2000 m
and D ¼ 7000 m, respectively. There is a visible increase in the heat
(b)
extraction rate of the case with D ¼ 2000 m as the progressing of
the heat mining process, which is quite different from other cases.
As we have discussed in the first paragraph of Section 4.1.2, the cold
fluid will accumulate at the bottom of the reservoir with the pro-
gressing of the heat mining process, forcing the fluid originally at
the bottom of the reservoir moves towards the reservoir top and
yielding an increased temperature in the top half of the reservoir,
which will be noticeable at a high geothermal gradient. As shown in
Fig. 10 (b), for the case with a geothermal gradient of 0.07  C/m, the
heat extraction rate is higher than that of the case with a
geothermal gradient of 0.02  C/m after decades of heat mining,
even though it is the lowest among that of the three cases at the
beginning. Overall, for a same reservoir temperature, the reservoir
depth and geothermal gradient would have a relatively small in-
fluence on the RHPS total heat mining amount considering the
whole heat mining process.
Fig. 10 (c) compares the heat extraction rates of RHPS using
different working fluid in the reservoir. As we have mentioned in
the second paragraph of Section 2, as compared to water, CO2 have
smaller viscosity coefficient and larger volumetric expansion
 W. Huang et al. / Energy 162 (2018) 630e644
coefficient which could contribute significantly to the buoyancy
forces and the transmission of geothermal heat. Seen from Fig. 10
(c), at the operation time of 10 years, the heat extraction rate of
the case using water as reservoir working fluid is 298 kW, which is
only a half value comparing to that of the case using CO2 as reser-
voir working fluid.
4.2.2. Wellbore variables
Fig. 11 (a) and (b) describe the dependence of the heat extraction
rate on the length of the wellbore within the reservoir and the
radius of heat pipe, respectively. For the cases in Fig. 11 (b), the
width of the flow channel is fixed at 0.1 m. The increase of the
wellbore length and heat pipe radius both augment the heat ex-
change area of the heat pipe evaporator, which will enhance the
heat extraction rate of RHPS. It can be noted in Fig. 11 (a) and (b)
that the heat extraction rate is more sensitive to the heat pipe
radius than to the wellbore length for the same increase ratio at the
heat exchange area. At the mining time of 0.1 years, the heat
extraction rate increases by 174% when the heat pipe radius in-
creases from 0.05 m to 0.2 m, while the heat extraction rate in-
crease is by 55.4% when the wellbore length increases from 500 m
to 2000 m. As already discussed, in Fig. 5 the heat flux on the heat
pipe evaporator surface is not uniformly distributed and it de-
creases sharply along the reservoir depth; therefore, the increase of
the wellbore length may not significantly increase the heat
extraction rate, as the heat flux on the extra evaporator area
resulting from the added wellbore length is relatively small. On the
other hand, for the case with h ¼ 2000 m in Fig. 11 (a) and the case
with rpip ¼ 0.2 m in Fig. 11 (b), the heat extraction rates increase by
13% and decrease by 22%, respectively, after 30 years of heat min-
ing. It indicates that increasing the wellbore length (reservoir
height) will be helpful to maintain the performance of RHPS; this
finding may be related to the fact that the extra reservoir volume
will allow more cold water being held at the bottom of the reservoir
instead of moving to its upper region.
Fig. 11 (c) illustrates the dependence of the heat extraction rate
on the width of the annulus flow channel. When the width of the
flow channel increases from 0.05 m to 0.1 m, the heat extraction
rate increases by 39%. However, the heat extraction rate only in-
creases by 3% when the width of flow channel increases from 0.1 m
to 0.15 m. This observation may indicate that a flow channel with
width of 0.1 m is large enough for generating a well-developed
natural convective flow and it would be unnecessary to enlarge
the width of flow channel over 0.1 m.
As it can be observed in Fig. 11 (d), the permeability of the
annulus flow channel in the wellbore has significant influence on
the heat extraction rate of RHPS. At the heat mining time of 10
years, in comparison with an impermeable (zero permeability) flow
channel, the heat extraction rates increase by 3%, 61%, 425% and
1117% for the cases with permeability values of 1010 m2, 109 m2,
108 m2 and 107 m2, respectively. It indicates that only when the
permeability of flow channel is larger than 109 m2, the heat
extraction performance of RHPS can show more significant
improvement. For the case with a permeability value of 107 m2,
the heat extraction rate is as high as 1.65 MW at the mining time of
0.1 years, but after 30 years of mining time, it drops to 0.89 MW. It
should be noted that a higher heat extraction rate will yield faster
rising cold fluid accumulation region, which, in turn, results in
accelerated heat extraction rate decline of RHPS.
The influence of the porosity of the annulus flow channel on the
heat extraction rate of RHPS is depicted in Fig. 11 (e). Considering
the heat transfer coefficient of CO2 is much smaller than the solid
heat conductivity used in this study, the thermal energy extracted
by the evaporator is mainly transferred from the solid phase of the
flow channel porous medium by conduction. Thus, the
641
enhancement of the solid heat conductivity coefficient is important
to achieve a higher heat extraction rate. As defined in the last
paragraph in Section 3.2, the heat conductivity of solid phase is
correlated with porosity, i.e. leff
s ¼ ls(1-ε) . Therefore, the increase
1.5
of the porosity within flow channel would lead to negative effects
on the heat extraction performance of RHPS. We obtain at 10 years
of heat mining time a heat extraction rate decline from 573 kW to
102 kW, when the porosity of flow channel augments from 0.15 to
0.99. It should be noticed that if the flow channel has no porous
medium inside, two opposite influence on the heat extraction rate
will be generated by the higher permeability and higher porosity
(see in Fig. 11 (d) and (e)). Thus, equipping the flow channel with
some fins which can acquire a high permeability and a high
equivalent heat conductivity both may be the best choice.
Fig. 11(f) describe the dependence of the heat extraction rate on
the volumetric heat exchange coefficient within the flow channel.
The increase of hv will propel the fluid and solid within the porous
medium locally close to thermal equilibrium. As we have analyzed
in the previous paragraph, the thermal energy extracted by the
evaporator is mainly transferred from the solid component of the
flow channel porous medium. Thus, the heat extraction perfor-
mance will also benefit from the increase of hv,DHE. As shown in
Fig. 11 (f), at the mining time of 10 years, the heat extraction rate
increases by 68% when hv,DHE increases from 200 W/K/m3 to
1000 W/K/m3 and only increases by 4% when hv,DHE increases from
1000 W/K/m3 to 5000 W/K/m3. This observation indicate the heat
exchanger between solid and fluid in flow channel is sufficient
enough when hv,DHE is higher than 1000 W/K/m3.
4.2.3. Reservoir variables
Fig. 12 (a) illustrates the dependence of the heat extraction rate
on the reservoir radius. As already discussed, the heat extraction
rate decreases with heat mining time is primarily due to the
accumulation of cold fluid in the reservoir, consequently, an in-
crease at the reservoir volume will slow down the cold fluid
accumulation process. Thus, the decline of the heat extraction rate
with heat mining time can be alleviated by increasing the reservoir
radius. In Fig. 12 (a) we can note that during the 30 years of heat
mining process, the decrease of the heat extraction rate is only 1.6%
for the reservoir radius of 100 m, while the decrease amount is 32%
for the reservoir radius of 25 m.
The permeability of reservoir affects the heat extraction per-
formance of RHPS as a higher permeability would reduce the flow
resistance when the cold fluid flows into the reservoir, and the heat
extraction rate benefits from a lower cumulative height of cold fluid
within the annulus flow channel. As indicated by in Fig. 12 (b), the
permeability should be larger than 1013 m2 to ensure that the cold
fluid can leave the flow channel at an appropriate rate.
5. Summary and conclusions
A novel design for mining hot dry rock geothermal energy is
proposed in this work. The combined reservoir and heat pipe sys-
tem (RHPS) encompasses a heat pipe in its interior installed in a
single well with the aim of extracting geothermal energy and an
artificial reservoir built around the main endothermic region of the
well. The reservoir is permeable and saturated with CO2 in super-
critical state. This system has the advantage of not having problems
associated with corrosion and scaling, and water loss etc., which are
commonly encountered in the conventional EGS technology. The
proposed system, mainly due to the thermosyphon effect of CO2
fluid in the reservoir, offers superior heat extraction performance as
compared to the conduction-only geothermal system. Moreover,
RHPS has the additional advantage of providing active CO2
sequestration.
642 W. Huang et al. / Energy 162 (2018) 630e644

(a) (b)

(c) (d)

(e) (f)
Fig. 11. Heat extraction rate with different wellbore variables: a) length of wellbore within reservoir; b) heat pipe radius; c) width of flow channel; d) permeability of flow channel;
e) porosity within flow channel; f) volumetric heat exchange coefficient within flow channel.
 W. Huang et al. / Energy 162 (2018) 630e644
(a)
Fig. 12. Heat extraction rate with different reservoir variables: a) reservoir radius; b) permeability of reservoir.
In the numerical study conducted in this work to assess the heat
extraction performance of RHPS, a base case, in which the reservoir
is at an average temperature of 170  C, and it has height of 1000 m
and radius of 50 m, was considered; the system has a continuous
heat production of 570 kW with a vapor temperature of 120  C. The
comparisons indicate that the proposed system has better heat
extraction performance and significantly higher exergy production
rate than that of the convectional DHE system with the same
operation conditions.
Detailed sensitivity analyses are conducted with reference to a
wide range of possible configurations and working conditions to
better understand how the operating and design variables affect
the heat extraction performance of the system. This acquired
knowledge will provide guidance toward the design enhancement
of the RHPS system.
Main conclusions drawn from the present work can be sum-
marized as follows:
 An increase of the temperature difference between the reservoir
and the heat pipe evaporator generally yields growths of heat
extraction. Using the lower production temperature in normal
permission scope for geothermal power generation tends to be
optimal.
 When the average temperature of reservoir is fixed, the
geothermal gradient and the depth of the reservoir has little
influence on the long-term heat mining performance of RHPS.
 The heat extraction rate of RHPS using CO2 as reservoir working
fluid will be nearly twice over that for water.
 The increase of the heat pipe radius will be more effective in
terms of heat extraction performance of RHPS than the increase
of the heat pipe length for the same heat exchange area increase.
 The annular channel is a critical component of the proposed
system, and the channel should have a width of about 0.1 m to
avoid restrictions on the natural convective flow occurring in it.
 The permeability of the flow channel has significant influence
on the heat extraction rate of RHPS; only for flow channel
permeability values larger than 109 m2, the heat extraction
performance of RHPS is obviously better than that for the con-
ventional DHE system.
 Using an empty flow channel to improve the permeability in the
annulus flow channel may not be perfect. Equipping the flow
643
(b)
channel with some fins which can acquire a high equivalent heat
conductivity may be the better choice.
 Enhancement promoters of the heat exchange between fluid
and rock in the annulus channel yield major improvement of the
heat extraction performance of RHPS as compared to that
without the promoters.
 Higher heat extraction rate values lead to rapid decline of the
heat extraction performance with the time advancement of the
heat mining process.
 Volume enlargement of the artificial reservoir can slow down
the temporal decline of the heat extraction rate.
 The reservoir permeability should have a value larger than
1013 m2 to ensure an adequate cold fluid outflow from the
channel.
Acknowledgments
Financial support received from the China National Science
Foundation and Guangdong-Province Joint Project (U1401232), the
Key Scientific Development Project of Guangdong Province
(2014A030308001), the Guangdong Key Laboratory of New and
Renewable Energy Research and Development Foundation
(Y709jf1001), and the China National Science Foundation
(51406213) is gratefully acknowledged.
References
[1] Goldemberg J. World energy assessment: energy and the challenge of sus-
tainability. United Nations Pubns; 2000. p. 165.
[2] Scenario N, East M. World energy outlook 2014. 2014.
[3] Brown DW, Duchane DV, Heiken G, Hriscu VT. Mining the earth's heat: hot dry
rock geothermal energy. Springer Science & Business Media; 2012.
[4] Robinson E, Potter R, McInteer B, Rowley J, Armstrong D, Mills R. Preliminary
study of the nuclear subterrene. N. Mex: Los Alamos Scientific Lab.; 1971.
[5] Breede K, Dzebisashvili K, Liu X, Falcone G. A systematic review of enhanced
(or engineered) geothermal systems: past, present and future. Geoth Energy
2013;1(1):1e27.
[6] Baticci F, Genter A, Huttenloch P, Zorn R. Corrosion and scaling detection in
the Soultz EGS power plant, upper Rhine Graben, France. In: Proceeding world
geothermal congress; 2010.
[7] Clark CE, Harto CB, Schroeder JN, Martino LE, Horner RM. Life cycle water
consumption and water resource assessment for utility-scale geothermal
systems: an in-depth analysis of historical and forthcoming EGS projects.
Argonne, IL (United States): Argonne National Laboratory (ANL); 2013.
[8] Florides G, Kalogirou S. Ground heat exchangers - a review of systems, models
and applications. Renew Energy 2007;32(15):2461e78.
644 W. Huang et al. / Energy 162 (2018) 630e644
[9] Balbay A, Esen M. Experimental investigation of using ground source heat
pump system for snow melting on pavements and bridge decks. Sci Res Essays
2010;5(24):3955e66.
[10] Ozgener O, Hepbasli A. Performance analysis of a solar-assisted ground-source
heat pump system for greenhouse heating: an experimental study. Build
Environ 2005;40(8):1040e50.
[11] Esen M, Yuksel T. Experimental evaluation of using various renewable energy
sources for heating a greenhouse. Energy Build 2013;65:340e51.
[12] Kohl T, Brenni R, Eugster W. System performance of a deep borehole heat
exchanger. Geothermics 2002;31(6):687e708.
[13] Nalla G, Shook GM, Mines GL, Bloomfield KK. Parametric sensitivity study of
operating and design variables in wellbore heat exchangers. Geothermics
2005;34(3):330e46.
[14] Kujawa T, Nowak W, Stachel AA. Utilization of existing deep geological wells
for acquisitions of geothermal energy. Energy 2006;31(5):650e64.
[15] Bu X, Ma W, Li H. Geothermal energy production utilizing abandoned oil and
gas wells. Renew Energy 2012;41:80e5.
[16] Templeton JD, Ghoreishi-Madiseh SA, Hassani F, Al-Khawaja MJ. Abandoned
petroleum wells as sustainable sources of geothermal energy. Energy
2014;70:366e73.
[17] Alimonti C, Soldo E. Study of geothermal power generation from a very deep
oil well with a wellbore heat exchanger. Renew Energy 2016;86:292e301.
[18] Nian Y-L, Cheng W-L. Insights into geothermal utilization of abandoned oil
and gas wells. Renew Sustain Energy Rev 2018;87:44e60.
[19] Claesson J, Eskilson P. Conductive heat extraction to a deep borehole: thermal
analyses and dimensioning rules. Energy 1988;13(6):509e27.
[20] Wang Z, McClure MW, Horne RN. Modeling study of single-well EGS config-
urations. Bali, Indonesia: World Geothermal Congress; 2010.
[21] Feng Y, Tyagi M, White CD. A downhole heat exchanger for horizontal wells in
low-enthalpy geopressured geothermal brine reservoirs. Geothermics
2015;53:368e78.
[22] Shi Y, Song X, Li G, Li R, Zhang Y, Wang G, Zheng R, Lyu Z. Numerical inves-
tigation on heat extraction performance of a downhole heat exchanger
geothermal system. Appl Therm Eng 2018;134:513e26.
[23] Vasil'Ev L. Geothermal energy utilization with heat pipes. J Eng Phys Ther-
mophys 1990;59(3):1186e90.
[24] Kusaba S, Suzuki H, Hirowatari K, Mochizuki M, Mashiko K, Nguyen T,
Akbarzadeh A. Extraction of geothermal energy and electric power generation
using a large scale heat pipe. In: Proceedings of world geothermal congress;
2000. p. 3489e94.
[25] Franco A, Vaccaro M. On the use of heat pipe principle for the exploitation of
mediumelow temperature geothermal resources. Appl Therm Eng
2013;59(1e2):189e99.
[26] Ebeling J-C, Kabelac S, Luckmann S, Kruse H. Simulation and experimental
validation of a 400 m vertical CO 2 heat pipe for geothermal application. Heat
Mass Tran 2017;53(11):3257e65.
[27] Pruess K. Enhanced geothermal systems (EGS) using CO2 as working fluidda
novel approach for generating renewable energy with simultaneous
sequestration of carbon. Geothermics 2006;35(4):351e67.
[28] Randolph JB, Saar MO. Coupling carbon dioxide sequestration with
geothermal energy capture in naturally permeable, porous geologic forma-
tions: implications for CO2 sequestration. Energy Procedia 2011;4:2206e13.
[29] Wu YS, Haukwa C, Bodvarsson GS. A site-scale model for fluid and heat flow in
the unsaturated zone of Yucca Mountain, Nevada. J Contam Hydrol
1999;38(1e3):185e215.
[30] Jiang F, Chen J, Huang W, Luo L. A three-dimensional transient model for EGS
subsurface thermo-hydraulic process. Energy 2014;72:300e10.
[31] Warren J, Root PJ. The behavior of naturally fractured reservoirs. Soc Petrol
Eng J 1963;3(03):245e55.
[32] Sanyal SK, Butler SJ. An analysis of power generation prospects from enhanced
geothermal systems. Trans Geoth Resour Counc 2005;29.
[33] Taron J, Elsworth D, Min K-B. Numerical simulation of thermal-hydrologic-
mechanical-chemical processes in deformable, fractured porous media. Int J
Rock Mech Min Sci 2009;46(5):842e54.
[34] Gelet R, Loret B, Khalili N. A thermo-hydro-mechanical coupled model in local
thermal non-equilibrium for fractured HDR reservoir with double porosity.
J Geophys ResSolid Earth 2012;117.
[35] Shaik AR, Rahman SS, Tran NH, Thanh T. Numerical simulation of fluid-rock
coupling heat transfer in naturally fractured geothermal system. Appl
Therm Eng 2011;31(10):1600e6.
[36] Gelet R, Loret B, Khalili N. Thermal recovery from a fractured medium in local
thermal non-equilibrium. Int J Numer Anal Meth GeoMech 2013;37(15):
2471e501.
[37] Jiang F, Luo L, Chen J. A novel three-dimensional transient model for sub-
surface heat exchange in enhanced geothermal systems. Int Commun Heat
Mass Tran 2013;41:57e62.
[38] Chen J, Jiang F. A numerical study of EGS heat extraction process based on a
thermal non-equilibrium model for heat transfer in subsurface porous heat
reservoir. Heat Mass Tran 2015:1e13.
[39] Chen J, Jiang F. Designing multi-well layout for enhanced geothermal system
to better exploit hot dry rock geothermal energy. Renew Energy 2015;74:
37e48.
[40] Huang W, Cao W, Jiang F. Heat extraction performance of EGS with hetero-
geneous reservoir: a numerical evaluation. Int J Heat Mass Tran 2017;108:
645e57.
[41] Rowland J, Wallis I, McNamara DD, Massiot C. The nature of fracture
permeability in the basement greywacke at Kawerau Geothermal Field, New
Zealand. In: 37th workshop on geothermal reservoir engineering. Stanford
University; 2012.
[42] Cao W, Huang W, Jiang F. Numerical study on variable thermophysical
properties of heat transfer fluid affecting EGS heat extraction. Int J Heat Mass
Tran 2016;92:1205e17.
[43] Beckermann C, Ramadhyani S, Viskanta R. Natural convection flow and heat
transfer between a fluid layer and a porous layer inside a rectangular enclo-
sure. J Heat Tran 1987;109(2):363e70.