You are on page 1of 14

Applied Energy 151 (2015) 178–191

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Review

Review of analytical models for heat transfer by vertical ground heat


exchangers (GHEs): A perspective of time and space scales
Min Li a,⇑, Alvin C.K. Lai b
a
School of Energy Science and Engineering, Central South University, Changsha, China
b
Department of Architecture and Civil Engineering, City University of Hong Kong, Tat Chee Avenue, Kowloon, Hong Kong Special Administrative Region

h i g h l i g h t s g r a p h i c a l a b s t r a c t

 We provide a review and introduction


to analytic models for ground heat
exchangers.
 We evaluate various models in a time
and space scale framework.
 We compare six analytic G-functions
for ground heat exchangers of single
U-tube.
 We discuss several key problems with
indoor and in situ experimental work.
 We identify several unsolved
problems in the analysis of ground
heat exchangers.

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

Article history: Ground (or geothermal) heat exchangers are attracting a great deal of attention as a way of using shallow
Received 3 September 2014 geothermal energy. This paper provides not only a critical review but also a thorough introduction to the
Received in revised form 3 April 2015 analysis of heat transfer by borehole and foundation pile ground heat exchangers, with an emphasis on
Accepted 17 April 2015
different analytical models. The literature is reviewed in a time-scale framework because of the diversity
of the time and space scales involved in the thermal processes of ground heat exchangers. We summarize,
discuss, and evaluate major advances in this field, including heat-source models, short-time models,
Keywords:
models for energy piles, in situ thermal-response tests, indoor sandbox experiments, and parameter esti-
Ground heat exchanger
Energy pile
mation as an inverse problem. Of particular note is that the unit-step temperature response (i.e., G-func-
Ground-coupled heat pump tion) of a ground heat exchanger with one U-shaped pipe is calculated; and six analytical models are
Ground heat storage compared: an infinite cylinder-source model, two infinite line-source models, two finite line-source mod-
In situ thermal-response test els, and a composite-medium line-source model. This paper closes by identifying several unsolved prob-
lems that require solutions.
Ó 2015 Elsevier Ltd. All rights reserved.

⇑ Corresponding author at: Room 309, Energy Building, School of Energy Science and Engineering, Central South University, Changsha, Hunan 410083, China. Tel.: +86 185
7311 9955.
E-mail address: cnlimin78@gmail.com (M. Li).

http://dx.doi.org/10.1016/j.apenergy.2015.04.070
0306-2619/Ó 2015 Elsevier Ltd. All rights reserved.
M. Li, A.C.K. Lai / Applied Energy 151 (2015) 178–191 179

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
2. Types of ground heat exchangers (GHEs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
3. Uniform framework for analysis of heat transfer by GHEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
3.1. Scale analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
3.2. General assumptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
3.3. Formalization of key problem. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
3.4. Methods of analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
3.4.1. Numerical vs. analytical methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
3.4.2. Analytical methods for heat conduction of GHEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
3.4.3. Duhamel’s theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
4. Key developments in heat-transfer analysis of borehole GHEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
4.1. Heat transfer models for effective thermal resistance of borehole, Rb . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
4.2. Conventional models for G-function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
4.2.1. Effect of ground surface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
4.2.2. Effect of groundwater movement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
4.2.3. Effect of freezing. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 185
4.2.4. Effect of moisture transfer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
4.2.5. Ground heat storage (GHS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
4.3. Short-term responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 186
4.4. A comparison study of various G-functions for borehole GHEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
5. Developments in heat-transfer analysis of pile GHEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
6. Sandbox and in situ thermal-response tests (TRTs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
6.1. In situ TRTs of GHEs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
6.2. Indoor sandbox experiments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
6.3. Estimation of parameters from TRTs as inverse problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
7. Conclusions and future research suggestions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189

1. Introduction of GCHP and GHS systems [9]. It is impractical to develop these


strategies experimentally because a matrix of GHEs is very large
Ground heat exchangers (GHEs), which are also called geother- and is coupled thermally to the ground. Theoretical predictions
mal heat exchangers, have emerged as a promising and globally from heat-transfer models seem to be the only feasible cost-effi-
accepted way of exploiting shallow geothermal energy (Fig. 1). A cient approach to developing these strategies [1]. Additionally, var-
GHE is essentially a channel (e.g., U-, W-, or helical-shaped) in a ious types of GHEs, drilling technologies, and systems are steadily
vertical borehole or a foundation pile of a building, in which a cir- evolving, further highlighting the importance of research on the
culating heat-carrying fluid absorbs (or discharges) heat from (or fundamental thermal process of GHEs.
to) the ground [1]. This heat exchange process, together with some Surprisingly, however, relatively few attempts have been made
special features of the ground, makes the use of GHEs attractive but to improve our understanding of the thermal process. The authors
also difficult. categorized and analyzed a set of representative reports published
Through the use of a cluster of GHEs, the ground offers a feasible during the period 1990–2012 that concern GCHPs and vertical
economical approach to the seasonal storage of a vast amount of GHEs. Among them, system-related studies and in situ experi-
solar energy and industrial waste heat (Fig. 1A and B) [1–3]. The ments account for almost 70%, with the remaining 30% being
seasonal storage of solar energy may greatly reduce the energy devoted to heat transfer analysis by analytical or numerical meth-
consumption of a central-heating system. The seasonal storage of ods (27%) and to laboratory experiments (3%). The lack of funda-
waste heat not only provides additional free heat sources for a cen- mental heat transfer research might stem largely from the
tral-heating system but also improves the energy efficiency of difficulty of the analysis and from the scattered appearance of
industry and reduces the industrial consumption of water. GHEs reports on the subject over decades. The authors believe that a
are also critical components of ground-coupled heat pumps thorough critical review of the literature, accompanied by a sys-
(GCHPs), which use the ground as a heat source or sink to complete tematic introduction to this topic, will help engineers understand
a vapor-compression [4–6] or an absorption [7] heat pump cycle this problem and also help researchers with their investigations.
(Fig. 1C and D). The use of the earth rather than the ambient air There are reviews of GCHPs and GHS, but most of them focus on
provides a lower-temperature sink for cooling, a higher-tempera- applications and system-related work, such as heat pump tech-
ture source for heating, and smaller temperature fluctuations, nologies [10], types of GHEs and GSHP systems [11,12], integrated
thereby yielding higher efficiency for the heat pump. According approaches [13], comparisons with other heating options [14], and
to the US Environmental Protection Agency [8], ground source heat the status of recent applications in different regions and countries,
pumps have the lowest CO2 emissions and the lowest overall including Europe [15], Australia [16], Korea [17], and China [18].
environmental costs of all the technologies analyzed. From the perspective of heat transfer, the pioneering work of
Despite their great energy-saving potential, ground heat storage Ingersoll et al. [4] summarized research up to and including the
(GHS) and GCHPs have not yet been fully developed. This undesir- 1950s. It provides an excellent basis for subsequent developments
able state is mainly due to difficulties in the design and analysis of in this field. Hellstrom reviewed analyses of borehole GHEs, with a
GHEs, which involve complicated heat transfer calculations [4,5]. primary focus on duct GHS systems [1]. Although several recent
Knowledge of the thermal processes in the ground is also a prereq- papers covered aspects of heat transfer calculations, such as the
uisite for creating optimum strategies for the control and operation design method and penalty temperature [19], parameter
180 M. Li, A.C.K. Lai / Applied Energy 151 (2015) 178–191

Fig. 1. Methods of exploiting shallow geothermal energy that employ borehole GHEs with single or double U-shaped tubes.

estimation [20–22], quasi-steady heat transfer around a borehole increasing interest in the GCHP community; and thus, recent
[23], and progress in conventional models for borehole GHEs advances in this field are reviewed in Section 5.
[24], a systematic review of recent developments in the analysis
of heat transfer through pile and borehole GHEs is still unavailable. 3. Uniform framework for analysis of heat transfer by GHEs
This paper aims to fill this gap in our knowledge by providing
not only a critical review but also a systematic introduction to heat 3.1. Scale analysis
transfer analysis for vertical GHEs. This paper attempts to evaluate
existing models within a time- and space-scale framework. This A huge challenge in predicting the thermal response of a GHE is
approach enables us to find enough common threads to make a the diversity of the time and space scales involved. It is necessary
‘‘high pass’’ over the complex details of each model. At the same to analyze these scales because most heat-transfer models make
time, it also provides key references and descriptive details, which assumptions about the time characteristics of the thermal process.
may help novices follow the theoretical development. The time scales important for heat transfer through vertical GHEs
Furthermore, this paper identifies some unsolved issues and can be estimated by analyzing the differential equation (Eq. (1a))
problems associated with heat transfer by GHEs, which deserve for heat conduction. According to Bejan [27], estimates of the
to be explored in the future. orders of magnitude of the terms in Eq. (1a) are given by Eq. (1b)
for a given radial space range, Dr:
 
2. Types of ground heat exchangers (GHEs) @T a @ @T
¼ r ð1aÞ
@t r @r @r
GHEs can have various configurations. This review focuses only
 
on closed loops, or more precisely, borehole and foundation pile DT a 1 DT
GHEs. The borehole type is the most common. It consists of one  Dr ; ð1bÞ
t Dr Dr Dr
or two U-shaped pipes that are inserted into a vertical borehole
and connected to a heat pump or a heating system to form a closed where t denotes the estimated time, and a is the thermal diffusivity
loop (Fig. 1). A U-shaped channel usually comprises two small-di- of the medium. The symbol ‘‘’’ used here means ‘‘is of the same
ameter high-density polyethylene (HDPE) tubes thermally fused to order of magnitude as’’ [27]. When the transient term on the left
form a U-shaped bend at the bottom [5]. The space between the is comparable to the diffusion term on the right, the time t can be
wall of the borehole and the U-shaped tubes is usually grouted estimated by rearranging the scaling relation Eq. (1b):
completely with grouting material or, in some cases, partially filled t  ðDr Þ2 =a: ð2Þ
with groundwater [25]. The depth of the hole (generally from 30 m
to 200 m) depends strongly on local geological conditions and The transient term can be ignored when t is about equal to
available drilling equipment. 10(Dr)2/a. From Eq. (2), it is easy to estimate several time scales
In a foundation pile GHE (or energy pile), the heat transfer tubes that are of practical importance, provided that the orders of mag-
are inside the steel frame of a foundation pile. There are various nitude of the thermal diffusivities of the grouting material, ab,
possible shapes, two of which are shown in Fig. 2 [26]. and the ground, as, are known (e.g., 106 m2/s).
Foundation piles are usually much shallower than boreholes and Four space scales and eight time scales (Fig. 3) are involved in
have a greater radius. These features of building piles and heat- the heat transfer of GHEs. The first space scale having practical
exchange channels make the heat transfer analysis of energy piles importance is the diameter of the borehole, rb, (0.1 m) and the
somewhat different from that of borehole GHEs. Since energy piles associated time (tb  r2b =ab ) is on the order of 1 h, during which
generally require less land area, this technology is evoking the effect of the heat capacity of the backfilling material is
M. Li, A.C.K. Lai / Applied Energy 151 (2015) 178–191 181

Fig. 2. Schematic diagram of energy piles with (a) spiral coil and (b) W-shaped tube. (Source: Authors’ published paper [26].)

3.2. General assumptions

Several assumptions must be made when developing a mathe-


matical model. The most common ones are given below.

Assumption 1: The ground is infinite or semi-infinite in extent,


depending on whether or not the influence of
the surface is considered.
Assumption 2: The ground has a uniform initial temperature
(effective undisturbed ground temperature). If
the surface is considered, this initial temperature
can be used as a constant-temperature boundary
condition (BC) for the surface.
Assumption 3: The BC for the wall of borehole or heat transfer
pipe is either a constant flux or a constant temper-
ature, with the constant-flux BC being more
convenient.
Fig. 3. Time and space scales involved in thermal process of borehole GHEs. The
thermal process can span four space scales and eight time scales. (Source: Authors’ Assumption 4: If the effect of the seepage of groundwater cannot
published paper [86].) be ignored, the flow is generally assumed to be
homogeneous and parallel to the surface.
Assumption 5: Although the ground is usually layered and inho-
significant. The second important space dimension is the half dis- mogeneous, the ground can be treated as a med-
tance between two adjacent boreholes, L, which is on the order ium with an equivalent thermal conductivity,
of several meters. The corresponding time tL  L2/as is on the order ks,e. Analytical models use Assumption 5 almost
of a month, during which the thermal interaction between adja- exclusively. At first glance, this assumption seems
cent boreholes is important. The largest space scale can be tens to be highly idealized; but several numerical stud-
of meters or more, such as the half length of a borehole, H/2, and ies have confirmed that it is appropriate for pre-
the horizontal scale of a GHE cluster. The time scale involved, dicting the overall temperature response [28–
tH  H2/4as, is as long as the lifetime of a GHE (decades). There is 30]. It should be noted that this equivalence prop-
another time, tr  H/u, (where u denotes the velocity of the fluid erty is also a key parameter used in practical GCHP
in the pipes) called the residence time, which has been discussed system design [5]. In fact, there are several theo-
in some studies [1]. It is generally several minutes long. It is the retical expressions for ks,e [31]. For example, the
time during which the heat capacity of the fluid circulating in a equivalent conductivities are
pipe needs to be addressed.
The short-term hourly temperature response of the ground is 1 XN
vital for analyzing the energy of GCHP and GHS systems and for ks;e ¼ PN k b
i¼1 s;i i
ð3aÞ
i¼1 bi
their optimum control and operation. By contrast, the long-term
response determines the overall feasibility of a system from the for a horizontally layered ground medium and
standpoint of life cycle. Addressing the complete spectrum of time
1 1 XN bi
scales require vast computational resources, particularly in ¼ PN ð3bÞ
ks;e i¼1 bi
i¼1 k
s;i
large-scale applications.
182 M. Li, A.C.K. Lai / Applied Energy 151 (2015) 178–191

8
for a vertically layered ground medium, where ks,i and bi are the < T f;i ðtÞ ¼ T f ðtÞ þ ql H
2qf cf V f
thermal conductivity and thickness of the ith layer of the ground, ; ð6Þ
: T f;o ðtÞ ¼ T f ðtÞ  ql H
respectively, and N is the total number of layers. An expression 2qf cf V f
for ground stratified in an arbitrary direction can also be found in
Ref. [31]. where subscript f denotes fluid; H is the length of the GHE; and q, c,
Assumptions 1–5 are widely used in analytical models of GHEs, and V are density, mass specific heat, and volume flow rate, respec-
and they have been verified to some extent by numerical simula- tively. Eq. (6) is equivalent to qfcfVf(Tf,i  Tf,o) = qlH. More accurate
tions [28–30,32]. expressions for Tf,o can be found by solving energy balance equa-
tions for the circulating fluid based for a given Tf,i and for a specific
3.3. Formalization of key problem point x. A good example of this is the quasi-three-dimensional mod-
els for Rb reviewed in Section 4.1.
The main questions that engineers may ask in the early stages To conclude, the key to heat transfer analysis is first to deter-
of designing a GHE are (a) what the heat transfer rate of a GHE mine G and Rs. Then, Tf, Tf,i, and Tf,o can be determined. Existing
as a function of time is, given a particular temperature difference models, as discussed in Sections 4 and 5, can be evaluated in terms
between the circulating fluid and the ground, and (b) what the of the extent to which the predictions are affected by the constant
temperature difference as a function of time is, given a required thermal resistance, Rs.
heat exchange rate. In the language of heat transfer, the two ques-
tions can probably be expressed as [4] 3.4. Methods of analysis

T f ðtÞ  T s;0 DT 3.4.1. Numerical vs. analytical methods


ql ¼ ¼ ; ð4Þ
RðtÞ RðtÞ Analytical methods and models appear to be more useful than
numerical methods for advancing GHE technology. Although
where Tf is the average temperature of the circulating fluid, Ts,0 is numerical methods, such as finite-difference [28–30,36], finite-
the effective, undisturbed temperature of the ground, ql is the heat volume [37–39], and finite-element methods [40–42], are elabo-
transfer rate of the GHE per unit time per unit length (W/m), and R rate enough to describe all the underlying physical mechanisms,
is the total thermal resistance (mK/W). they are impractical for engineering applications for three reasons:
Eq. (4) provides a general basis for modeling the thermal pro- (1) They are time consuming for year-round and/or life-cycle sim-
cesses of vertical GHEs, including borehole GHEs and energy piles. ulations, particularly in large applications where all the time and
For a GHE, R(t) is often an unknown variable that needs to be deter- space scales are important and must be tackled. (2) It is very diffi-
mined by heat transfer analysis. Despite R(t) being a function of cult for in-house programming to develop a general grid genera-
time, analytical models exclusively decompose it into a time-inde- tion program for various configurations of ground channels; and
pendent part and a time-dependent part to simplify the analysis. it is thus very difficult to develop software for the analysis, design,
Thus, a possible approach to evaluating existing models can rest and simulation of GHS and GCHPs. (3) In the literature, a majority
in a time-scale context, in which the constant part of R is evaluated of numerical models are implemented in commercial software,
according to specific requirements. such as Fluent and Ansys. However, it is very difficult to incorpo-
The unsteady part of R is often denoted by G(t) (called the G- rate commercial software into the pre- and post-processing stages
function). There are minor differences between the G-functions to perform a system simulation for a particular application. It is
proposed by different researchers [26,29]. The G-function1 was first also difficult for designers and engineers to use computational fluid
suggested in a dimensionless form by Ingersoll et al. [4] and was dynamics software. Thus, it would be better to use a general design
subsequently used and further developed by Eskilson [28] and and simulation tool based on analytical heat-transfer models. This
Claesson and Eskilson [29]. The dimensionless form is widely used paper focuses mainly on analytical models, and numerical models
because it facilitates the analysis and summary of results [5,33– are mentioned only when necessary.
35]. By contrast, the G-function used here has a clear physical mean-
ing: It has the same dimension as thermal resistance and is entirely
3.4.2. Analytical methods for heat conduction of GHEs
consistent with the unit-step response function defined in the prin-
The most widely used analytical tools for analyzing heat trans-
ciple of superposition (Section 3.4.3) [26]. In other words, the
fer in GHEs are Kelvin’s theory of heat sources and the Laplace
unsteady thermal resistance, G, can be understood as the tempera-
transform method [4,43]. Kelvin’s theory of heat sources has a very
ture response in the ground due to a unit-step change in the heat
clear simple physical meaning [43]: Taking the solution of an
flux, ql. Using the definition of G, Eq. (4) becomes
instantaneous point source (an abstract concept similar to the
DT ¼ ql RðtÞ ¼ ql ½Rs þ Gðx; tÞ; ð5Þ mass point in mechanics) as a fundamental solution (also called
Green’s function for an infinite medium) enables the solution for
where x denotes the coordinates of the point under consideration; a continuous point source to be obtained by the integration of
and Rs is the steady part of the thermal resistance, which depends the fundamental solution over time. The solutions for line, plane,
on the choice of x. surface, and volume sources can also be obtained by the integra-
The temperature, Tf, in Eqs. (4) and (5) is the average tempera- tion of point-source solutions over appropriate space variables.
ture of the ground loops. Another model or assumption is neces- This method has proven to be extremely successful in analyzing
sary to determine the temperatures of the inlet and outlet fluids. the thermal processes of various GHEs because many configura-
One commonly used assumption is that Tf approximately equals tions of boreholes and ground loops, which are intractable by other
the average of the temperatures of the inlet and outlet fluids of means, can be well approximated by line or surface sources of heat.
the GHE [5]. Thus, the temperatures at the inlet and outlet, Tf,i The Laplace transform method basically has two steps [43,44]:
and Tf,o, respectively, are First, applying the Laplace transform reduces the original problem
to a subsidiary problem, whose solution is easy to find. And second,
the solution to the original problem can be derived by the use of
1
Some GCHP reports mistake the G-function for Eskilson’s G-function. As shown in
the inversion theorem for the Laplace transform. This method is
Section 3.4.3, the G-function is defined in Duhamel’s theorem and was initially often used to derive short-term solutions for borehole GHEs
proposed by Ingersoll et al. [4]. (Section 4.3) and to obtain fundamental solutions for the
M. Li, A.C.K. Lai / Applied Energy 151 (2015) 178–191 183

heat-source method [45]. These two methods are described in 4.1. Heat transfer models for effective thermal resistance of borehole,
great detail in the classical book of Carslaw and Jaeger [43]. Rb

Heat transfer in a borehole depends on the configuration of the


3.4.3. Duhamel’s theorem flow channels, the thermal properties of the grouting materials,
Duhamel’s theorem, which is also known as the principle of and the surrounding ground. This local thermal process has three
superposition [43,44], is an efficient theoretical tool for developing components:
solutions to heat conduction with time-dependent BCs and/or
time-varying energy source terms. In actual applications, the heat (1) convective heat transfer between the circulating fluid and
transfer rate of GHEs varies continuously due to variations in the the inner surface of the U-shaped pipes,
cooling or heating loads of buildings. According to Duhamel’s the- (2) conductive heat transfer through the wall of the U-shaped
orem, the problems of time-varying loads can easily be tackled by pipe, and
using the solution to the problem for a unit-step load [43,44]: (3) conductive heat transfer through the backfill material.
Z t
@Gðx; t  sÞ In a strict sense, if t is greater than or equal to 5tb, the thermal
T 1 ðx; tÞ ¼ T s;0 þ ql ðsÞ ds; ð7Þ
0 @t process in the borehole can approach a steady-flux state (not a
steady state), which is a state in which the temperature difference
where T1 is the temperature at the point under consideration, x; the between the fluid and the wall of the borehole (not the tempera-
time-dependent function ql(t) is the heat exchange rate of a GHE; tures themselves) is constant. Under this steady-flux condition,
and s is an integral variable with the dimension of time. The func- the three processes can be characterized by three constant thermal
tion G is the temperature response to the unit-step load of a GHE; it resistances; and their sum yields the effective fluid-to-ground
is the same as in Eq. (5). Since the heating and cooling loads of a thermal resistance, Rb (mK W1) [1].
building are commonly available as step-wise constant values (on The thermal resistances associated with the first two processes
an hourly, daily, or other basis), it is more convenient to write Eq. are referred to as the thermal resistance of the pipe, Rp. It is easy to
(7) as Eq. (8) [44]: calculate from the formula [1]
1 ro 1
X
N1 Rp ¼ ln þ ; ð9Þ
T 1 ðx; tÞ ¼ T s;0 þ Dql;j Gðx; t  jDtÞ: ð8Þ 2pkp ri 2pr i a
j¼0
where kp is the thermal conductivity of the U-shaped pipe
Here, Dql,j is the step-wise change in ql at the beginning of the (W m1 K1); ro and ri denote the outer and inner radii of the legs
jth time interval. Eqs. (7) and (8) imply that the temperature of the U-shaped pipe, respectively; and a is the convective heat
response of any time-varying ql is readily determined once the transfer coefficient (W m2 K1). There are many formulas for cal-
response to the unit-step load, that is, the G-function, is available. culating or estimating a. The most common formula for turbulent
Thus, a heat transfer analysis can focus only on the development of flow in a circular pipe is the one proposed by Dittus and Boelter
the G-function. [27]:
2ar i
Nu ¼ ¼ 0:023 Re4=5 Pr n ; ð10Þ
kf
4. Key developments in heat-transfer analysis of borehole GHEs
where kf is the thermal conductivity of the fluid; Re is Reynold’s
The borehole GHE is the most common type of GHE and has number, which is defined in terms of the diameter of the pipe; Pr
been studied extensively. The conventional way in which the ther- is Prandtl’s number, which is determined at the mean temperature
mal processes of a borehole GHE are decomposed (i.e., R) is in of the bulk fluid; and n is equal to 0.4 when the fluid is being heated,
terms of the radius of the borehole: The process in the borehole and to 0.3 when it is being cooled. Eq. (10) is a good approximation
is assumed to be in a steady state that is described by the effective when 0.7 6 Pr 6 120, 2500 6 Re 6 124,000, and the ratio of length
thermal resistance of the borehole, Rb. The process outside the to diameter is greater than 60. Although an empirical correlation
borehole depends on time and is represented by a G-function. simply yields a rough estimate of a due to the inherent complexity
This decomposition is widely accepted and used; and for simplic- of convective heat transfer, the influence of such a rough estimate
ity, all the models based on this assumption are called conven- on the calculation of Rb is negligible because convective thermal
tional models in this paper. While Yang et al. have reviewed resistance accounts for only 2–3% of Rb in most situations (except
recent developments in this field [24], in Sections 4.1 and 4.2 we in the case of laminar flow) [46].
discuss these conventional models in the time-scale framework, Models for Rb can be divided into empirical and theoretical
which yields new insights. groups, and some of them are listed in Table 1. Typical empirical

Table 1
Models of effective thermal resistance of borehole of GHE with single U-shape pipe.

Expressions for Rb Comments


Empirical models Rb ¼ 1
b
This expression uses the shape-factor concept in heat conduction. The empirical
b0 kb ðrb =r p Þ 1
coefficients are obtained by fitting experimental data [49]
Rb ¼ 2p1k ln prffiffinbr This is derived from the equivalent-diameter assumption. n denotes the number of pipes in
b p
a borehole [48]
 qffiffiffiffi
rb rp This is for a GHE with a single U-shaped pipe. It is also derived from the equivalent-
Rb ¼ 2p1k ln rp D
b
diameter assumption [47]
  r 

Two-dimensional models r2 r4b R The influence of ks is represented by the dimensionless ratio r. g is equal to 0 or calculated
Rb ¼ 4p1k ln 2Drb o r4 D  g þ 2p
b
4
b
from Eq. (11), which is derived by the multipole method [1]
 
Quasi-three-dimensional models Rb ¼ q cHf V f
T f;i T b
 12 Tf,o is obtained by solving energy equations for up- and down-flow channels [53]
f T f;i T f;o
184 M. Li, A.C.K. Lai / Applied Energy 151 (2015) 178–191

models are one-dimensional models that take the U-shaped pipe to proposed a mathematically rigorous boundary condition at rc
be a pipe of ‘‘equivalent’’ diameter [47,48]. This equivalent-diame- based on a matched-asymptotic-expansion technique [52].
ter assumption simplifies a two-dimensional geometric region into Empirical and two-dimensional theoretical models consider the
a concentric annular region, thereby reducing a complicated multi- local heat exchange at a given depth but ignore the variation in the
dimensional problem to a simple one-dimensional one. Empirical temperature of the fluid in the downward and upward channels. To
models may contain several empirical constants, which may be address this variation, Hellstrom proposed two quasi-three-dimen-
determined by fitting experimental or computational data to the sional models for Rb, one employing a uniform-flux BC for the bore-
model for a particular geometric arrangement [49,50]. Empirical hole wall and one employing a uniform-temperature BC [1]. The
models are widely used because of their simplicity; but they pro- final expressions include additional correction factors for the
vide little insight into the underlying heat transfer processes, two-dimensional models. Based on Hellstrom’s models, Zeng
which makes them difficult to generalize. et al. also devised quasi-three-dimensional models for GHEs with
Theoretical models can be broken down into two-dimensional single and double U-shaped tubes. They employ functions of the
and quasi-three-dimensional models. A Swedish group proposed temperature of the fluid at the outlet (Table 1); the expressions
two two-dimensional models for Rb [1], one derived from the for the outlet temperature are rather complex and can be found
steady-state line-source assumption and one based on a multipole in their paper [53].
method. The difference between them lies in the dimensionless Last but not least, all the models for Rb are applicable only for
variable g (defined in Table 1). For the configuration with a single time t > 5tb, which means that the steady-flux state is attained in
U-shaped tube, either g equals 0, if the steady-state line-source the borehole. These models are conventionally used together with
assumption is used, or it is calculated from Eq. (11), which is a G-function for the thermal process outside the borehole
derived from the first-order multipole expression [1] (Section 4.2). Therefore, no matter which G-function is used, the
  2 conventional approach is unsuitable for t < 5tb when a rapidly-
ro 4rD4 varying heat flux must be addressed. To describe the thermal
2D
1  r 4 D4
response to a high-frequency heat flux, we must deal with the
g¼ b ; ð11Þ
1þ2pkb Rp r 2 16rD4 r 4 effect of the heat capacities of the backfilling materials in the bore-
12pkb Rp
þ 4Do2 1 þ 4 4 b2 hole. Work on this is reviewed in Section 4.3.
ðrb D Þ

where D is the half spacing between the legs of the U-shaped tube; 4.2. Conventional models for G-function
the dimensionless variable r is defined to be (kb  ks)/(kb + ks); and
ks and kb denote the thermal conductivities of the ground and back- The conventional models for a G-function have been inspired by
filling materials, respectively. The difference between the line- the seminal work of Ingersoll et al. [4], who presented an infinite
source model and the first-order multipole model (i.e., g) is usually line-source model and a cylindrical-source model for heat transfer
less than 15% in most situations [1]. through the ground. Even though they did not develop concrete
It should be noted that ks enters the two-dimensional models expressions, they did propose ideas for dealing with additional
via the dimensionless variable r, highlighting the fact that the complicated factors, including the method of images of heat
thermal conductivity not only of the backfill material but also of sources used to account for the influence of the ground surface,
the surrounding ground can affect the steady-flux thermal process the moving heat-source method for determining the effect of the
within the borehole. All empirical expressions incorrectly ignore movement of groundwater, and the principle of superposition for
the influence of ks by arbitrarily imposing a constant-temperature addressing the issue of variable heating rates. They also discussed
BC on the wall of the borehole [47–50]. Theoretical models address relevant topics such as the effect of ice formation, moisture trans-
the influence of ks by imposing a constant-temperature BC at a cer- fer, temperature recovery, and heat storage [4]. Subsequently, the
tain distance rc from the wall [1]. Thus, the predictions of the mod- ideas suggested by Ingersoll et al. were extended to the analysis
els depend somewhat on the choice of rc. To completely eliminate of borehole GHEs. Table 2 lists some typical G-functions, all of
the dependence on rc, Claesson and Hellstrom have recently refor- which are for the point rb on the borehole wall because the thermal
mulated the multipole model and derived a physically and mathe- resistance, R, is split depending on the radius of the borehole in the
matically correct definition of Rb [51]; Hermanns and Pérez also conventional approach. These functions are discussed below.

Table 2
Model developments for G-function.

Expression for G-function Comments


Infinite cylindrical-surface source model [4]
R1
Gðt; r b Þ ¼ p2 r1 ks 0 ðeas u t  1Þ J0 ðurbuÞY
2
1 ður b ÞY 0 ður b ÞJ 1 ður b Þ
du It is more complicated than the line-source model, and it is difficult to extend to handle other influences
2 J 2 ður ÞþY 2 ður Þ
b ½1 b 1 b 
(surface of ground, seepage flow, etc.)
Infinite line-source model [4]
R1  2 
Gðt; r b Þ ¼ 4p1ks r2 =4as t expðuÞ
r
du ¼ 4p1ks E1 4abs t It has the same performance as the infinite cylindrical surface model, and it is easy to extend to handle other
u
b relevant factors
Finite line-source model [55]
2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffi2ffi3
r2 þðzz0 Þ
bpffiffiffiffi
2
r 2 þðzþz0 Þ
bpffiffiffiffi
It is suitable for predicting long-term responses because the influence of the ground surface is considered.
R H 6erfc 2 as t
erfc
2 as t 7 0 This G also depends on z, but the value at z = H/2 is usually used as the average G-function
Gðt; rb ; zÞ ¼ 4pk 0 4 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi2  pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi2 7
1 6
5dz
s r2b þðzz0 Þ r 2b þðzþz0 Þ

Infinite moving line-source model [59]


  R r2 =4a t  
U 2 r 2b g This model accounts for the effect of the characteristic velocity of groundwater, U, on the responses of
Gðt; r b Þ ¼ 4p1ks I0 Ur
s 1 1
g exp  g  16a2 dg
b b
2as 0 s borehole GHEs
Infinite phase-change line-source model [43]
h  r2 i
Gðt; rb Þ ¼ Tq1 þ 4p1k E1 ðb2 Þ  E1 4abfr t This is the only analytical solution that can address the issue of ice formation. T1 is the melting-point of
l fr
groundwater; afr and kfr are the effective properties of frozen soil; and b is the root of Eq. (18)
M. Li, A.C.K. Lai / Applied Energy 151 (2015) 178–191 185

Practically speaking, the ground can be thought of as an infinite we compare the complicated cylindrical-surface source model
medium. Thus, heat transfer outside a borehole can be presumed to with the line-source model, the former seems unnecessary and
be heat conduction in an infinite region bounded internally by the impractical because the difference between the two is less than
borehole wall, which is subject to the boundary condition of a con- 10%.
stant flux or temperature. Carslaw and Jaeger employed the Second, Eq. (16) (including Eq. (12)) cannot be used to predict
Laplace transform method [43] to solve this problem. Then, the response of GHEs for very long times (t  tH) because it ignores
Ingersoll et al. used that solution to model the thermal processes the influence of the ground surface. This is a common disadvantage
of buried pipes [4], and Deerman and Kavanaugh used it to model of all the models that assume the ground to be an infinite medium.
U-tube GHEs [54]. This model is often referred to as the infinite
cylindrical-surface source model. It is given mathematically by 4.2.1. Effect of ground surface
8 2  The ground surface has an impact on long-term temperature
> @ Ts 1 @T s
> qs cs @t ¼ ks @r2 þ r @r
@T s
> variations in the ground, especially when the heating and cooling
>
>
< loads differ from each other. Excess heating or cooling loads can
r ¼ r b  2pr b ks @T
@r
s
¼ ql : ð12Þ
> accumulate in the ground, causing the average ground tempera-
>
> r ! 1 T s ¼ T s;0
>
> ture to gradually increase or decrease over time to such an extent
:
t ¼ 0 T s ¼ T s;0 that heat transfer through the ground surface becomes significant.
If the time is long enough, a quasi-steady-state can be approached
Here, r denotes radial distance. Carslaw and Jaeger’s solution is
whereby the excess heat is transferred to the air through the
[43]
ground surface [55–57].
T s ðt; r b Þ ¼ T s;0 þ ql Gðt; rb Þ; ð13Þ As discussed in Section 3.2, the effect of the ground surface is
generally simplified and taken to be a constant-temperature BC.
where the response function G(t, rb) is given in Table 2. For large val-
While the influence of the ground surface cannot be addressed
ues of ast/r2b, the G-function can also be written as
by the infinite cylindrical-source model, it can be treated analyti-
  
1 4as t r2 4as t cally by the method of images if Kelvin’s theory of heat sources
Gðr; tÞ ¼ ln 2  c þ ln 2  c þ 1 ; ð14Þ
4pks r 2as t r is used: The ground surface is thought of as a reflecting plane,
and a mirror-image line of heat sinks (strength: ql) exists [58].
where c = 0.5772 is Euler’s constant. The symmetrical distribution of line source and sink can keep the
The radius of the borehole of a GHE is usually several centime- temperature of the ground surface at a constant value. The addition
ters. It is very small compared to the length, which is tens of of the temperature responses of the heat sources and sinks yields
meters. Thus, this very thin borehole can be thought of as a line finite line-source models [55–57]; the final expression is given in
of infinite length with no volume that releases (or absorbs) heat Table 2. The finite line-source model requires a double integral to
to (or from) the surrounding soil [1,4,55–57]. This approach is obtain the integral average temperature of the borehole wall,
called the infinite line-source model and is given mathematically which leads to a marked increase in computational cost.
by Lamarche reformulated the finite line-source model so as to avoid
8 2 
use of the double integral [57].
>
>
> qs cs @T@ts ¼ ks @@rT2s þ 1r @T@rs
>
>
<
r ! 0  2pks limr @T s
¼ ql 4.2.2. Effect of groundwater movement
r!0 @r : ð15Þ
>
> The movement of groundwater affects heat transfer by involv-
> r ! 1 T s ¼ T s;0
>
>
: ing gross heat convection, which is significant for the long-term
t ¼ 0 T s ¼ T s;0 temperature response of GHEs [2]. Whereas a conservative design
The difference between Eqs. (15) and (12) is the way in which assumes no benefit from this flow [25], an analytical estimate of
the BC on the borehole wall (r = rb) is dealt with. The solution to the influence of groundwater flow is desirable. Most analytical
Eq. (15) is [4,43] models solve this diffusion-convection problem by using the mov-
Z ing heat-source method, which was first proposed by Ingersoll
1
ql expðuÞ et al. [4] and was further explored or extended by Diao et al.
T s ðt; r b Þ ¼ T s;0 þ du ¼ T s;0 þ ql Gðt; r b Þ: ð16Þ
4pks r 2 =4as t u [59], Sutton, et al. [60], Molina-Giraldo [61,62], Chiasson and
b
O’Connell [63], and Tye-Gingras and Gosselin [64]. All these mod-
For large values of as t=r 2b , the exponential integral E1 defined in els are theoretically identical to each other, and that of Diao et al. is
Table 2 can be approximated by [43] given in Table 2. The seepage flow of groundwater may be very
 2    complex, being vertical, horizontal, or both. All these analytical
1 rb 1 4as t
E1 ¼ ln 2  c ; ð17Þ models, however, use Assumption 4 in Section 3.2. Therefore, their
4pks 4as t 4pks rb
accuracy depends heavily on how closely this assumption repre-
with a maximum error of 2% for t > 5tb. Eq. (17) is also equal to the sents the actual seepage conditions. A further extension of analyt-
first two terms on the right-hand side of Eq. (14). The infinite line- ical solutions to consider more complicated seepage flows is
source model is widely used in GHE simulation because of its sim- extremely difficult. The movement of groundwater in a porous
plicity. Nevertheless, it suffers from two limitations on the time aquifer may also cause heat transfer by thermal dispersion [31].
scale. Several authors have addressed the effect of thermal dispersion
First, Eq. (16) is unsuitable for small periods because of the line- through the concept of modified thermal conductivity [61,65].
source assumption. This assumption can initially delay the temper-
ature response at the borehole wall if a hollow borehole is 4.2.3. Effect of freezing
employed, as in the infinite cylindrical-surface source model (Eq. ASHRAE recommends that all ground-source heat pumps use
(12)). The relative error between these two models is less than extended-range heat pump units for most water-to-air configura-
about 10% when t  5tb [1] and 2% when t  20tb [4]. An interesting tions [5]. An extended-range heat pump is a unit specifically
observation is that t > 5tb is required when using any of these G- designed to operate when the entering water has a temperature
functions in conjunction with any of the models for Rb reviewed (leaving the GHE) of 4 °C in the heating mode and 38 °C in the
in Section 4.1 (i.e., the conventional approach). In this sense, if cooling mode. A temperature of 4 °C for the entering water
186 M. Li, A.C.K. Lai / Applied Energy 151 (2015) 178–191

implies a lower average temperature of the fluid in the ground it probably produces errors when used to model the high-fre-
loop, which can cause problems with groundwater freezing. quency (short-term) thermal response of a GCHP.
Only one analytical expression is available for estimating the
influence of freezing. It combines the infinite line-source solu-
tion with Neumann’s treatment for the phase-change interface 4.3. Short-term responses
[4,43]. This expression may be suitable for a porous aquifer if
proper modifications are made. Table 2 gives a G-function In contrast to GHS, short-term temperature fluctuations are
derived from this solution, in which parameter b is the root of important in the design, optimization, and energy analysis of
Eq. (18) [43]: GCHPs and hybrid GCHPs [9,68]. The heating and cooling loads
of buildings vary continuously with time due to changes in the
ql ks h1 expðb2 afz =as Þ weather, the number of occupants, their activities, etc. In fact,
expðb2 Þ  ¼ b2 afz nqf Q L ; ð18Þ
4p E1 ðb2 afz =as Þ cooling loads are generally calculated on an hourly basis. Hourly
variations can cause corresponding high-frequency fluctuations
where E1 is the exponential integral defined in Table 2; h = Ts,0  Ts in the temperatures of a ground loop. According to Yavuzturk
is the excess temperature; the subscript fz denotes frozen ground; and Spitler [9], the supply and return temperatures of a ground
the subscript 1 denotes the freezing-point of water; n is the porosity loop typically fluctuate up to 5.6–10 °C over a given day.
of a porous aquifer; QL is the latent heat of freezing; and qf is the Predicting the high-frequency response, however, is more difficult
density of the groundwater. Ingersoll et al. conducted a simple case than predicting the low-frequency response because it requires
study using this solution to explore the effect of groundwater freez- handling the effect of heat capacity in a borehole, which involves
ing [4]. Their preliminary conclusion is that, among the factors transient heat conduction in a composite medium together with
affecting heat transfer, freezing is much less important than mois- taking the configuration of various U-shaped tube installations
ture transfer. into consideration [26]. Common analytical methods cannot meet
this challenge satisfactorily.
One way to tackle this challenge is by numerical methods. Since
4.2.4. Effect of moisture transfer
this review focuses only on analytical models, common numerical
Heat transfer in a porous medium is often accompanied by
methods (such as the finite-volume and finite-element methods)
moisture transfer [66]. Even though heat transfer due to moisture
are not discussed here, except for the thermal-network method
transfer by GHEs may be a minor contribution, the accompanying
[69–74]. The simplified thermal-network approach is based on
variations in the thermal properties of the ground can be large,
the concept of lumped capacitance [73] and uses an analogy to
thereby significantly affecting the thermal process in the ground.
electricity to solve transient heat conduction problems. In this
An increase in the amount of moisture in dry soil can lead to a large
method, a GHE and the surrounding ground are divided into incre-
increase in thermal conductivity, while diffusivity generally rises
mental layers in the depth direction, and each layer is subdivided
to a maximum (2–3 times the dry value) at a moisture content of
into annular regions in the radial direction [69–71].
5–10% [43]. Moisture transfer coupled with heat transfer may
Consequently, the domain is discretized into a number of thermal
involve the evaporation (or condensation) of water and diffusion
nodes. The law of energy balance is applied to each node, which
driven by temperature (Soret/Luikov effect), or pressure differ-
yields an ordinary differential equation (referred to as a lumped-
ences [66]. These processes make the analysis more complicated.
capacitance representation) for the temperature of each node.
Analytical solutions seem to be impossible. Very little investigative
The simultaneous solution of these representations yields the tran-
work has been done on a coupled heat and moisture process asso-
sient response of both the GHE and the ground. Moreover, this
ciated with the temperature levels of a GCHP system. Reuss et al.
approach can analyze several configurations of GHEs in both the
used the finite-difference method to investigate coupled heat and
long and short term; and it can also be coupled to a heat pump
moisture transfer in high-temperature ground heat storage [67],
model to simulate an entire GCHP system [72]. The thermal-net-
and found that a deep moisture gradient was detected and drying
work method is essentially a simplified numerical method and
out and cracking out may occurred in some ground media.
may have acceptable accuracy and computational speed.
Another way to model the short-term response of a GHE is to
4.2.5. Ground heat storage (GHS) simplify the geometrical arrangement in the borehole. Most
A GHE used for GHS has a different purpose from one used for a short-time analytical models opt for the equivalent-diameter
GCHP. In a GCHP, heat exchange between GHEs and the surround- assumption [75–83], which simplifies a problem involving compli-
ing ground should be maximized; while in GHS, it should be min- cated geometry to one with a relatively simple hollow cylindrical
imized to reduce heat loss during storage. Ingersoll et al. concluded composite region. It is possible to solve this simplified heat con-
that using just one GHE to store heat holds little promise because duction problem by various methods, such as the generalized
the heat put into the ground dissipates rather quickly [4]. For a orthogonal expansion technique [75] and the Laplace transform
GHS system, therefore, a cluster of many GHEs should be installed method [77,82]. This simplification also makes it possible to tackle
so compactly that there is thermal interaction between adjacent the transient effect of fluid circulating in a U-shaped channel [76].
heat exchangers. Once the interaction is fully developed, the ther- But the equivalent-diameter approach is an empirical method and
mal process around the bores approaches a steady-flux state [1], in cannot handle GHEs with ground channels of other forms. A pile
which the temperature difference between the circulating fluid GHE with a W- or helical-shaped tube is a case in point.
and the local average temperature in the store is constant. More The authors have proposed an alternative method of modeling
details of heat transfer calculations associated with GHS can be the short-term response of a GHE that is based on Jaeger’s instan-
found in Hellstrom’s book [1]. taneous line-source solution for a cylindrical composite medium
So far, Sections 4.1 and 4.2 have provided an overview of the [26,84,85]. It can deal with difficulties involving not only a com-
conventional approach to modeling the heat transfer of borehole posite medium but also the geometric configuration of heat
GHEs. This approach assumes the thermal process in a borehole exchange channels, including single and double U-shaped tubes,
to be in a steady state, and it is only applicable to cases in which W-shaped channels, and helical-coils. It has, however, a compli-
the rate of heat transfer varies slowly. The original purpose of this cated form. Moreover, this model belongs to the group of infinite
approach was to design a GHE cluster for ground heat storage, and line-source models and thus cannot be used to make predictions
M. Li, A.C.K. Lai / Applied Energy 151 (2015) 178–191 187

for a long-term thermal process if there is an imbalance between overestimation of the temperature variation, which leads to a more
heat injection and extraction in the GHE. conservative design for a borehole GHE, despite long-term experi-
mental data being unavailable. There is also a discrepancy between
4.4. A comparison study of various G-functions for borehole GHEs the two finite line-source models: The mid-point temperature is
higher than the integral average value. The impact of this differ-
In this section, six analytical G-functions are used to calculate ence on GHE design, particularly in large-scale applications,
the temperature response of the unit-step heat transfer rate deserves to be explored.
(ql = 1 W/m) for a single-U GHE: the infinite cylindrical-surface Finally, and perhaps most importantly, Fig. 4 clearly reveals the
source model Eq. (13), the infinite line-source model Eq. (16), the marked difference between the calculated short-term responses.
simplified infinite line-source model Eq. (17), two finite line-source The conventional models using the steady-state thermal resis-
models, and the composite-medium line-source model. Except for tance, Rb, (infinite cylindrical-surface model, and conventional infi-
the composite-medium line-source model, the five conventional nite and finite line-source models) give temperatures noticeably
models for G-functions are used in the equation larger than that yielded by the composite-medium line-source
model. The difference arises because the heat capacity of the grout-
T f ðtÞ  T s;0 ¼ Rb þ Gðt; r b Þ; ð19Þ
ing material in the borehole is neglected. Comparison of the com-
where the effective borehole thermal resistance, Rb, is calculated posite-medium model with a set of reference sandbox data and an
using the two-dimensional model (see Table 1). The finite line- elaborate finite-volume model has demonstrated that the compos-
source model (Table 1) is used in two ways to determine the aver- ite-medium line-source model is suitable for times as short as sev-
age temperature: The first one is the computationally intensive eral minutes [85]. Thus, the discrepancies shown herein imply that
integral average, and the second one uses the temperature at all the conventional models fail to accurately predict the short-
z = H/2 as an approximation. term (or high-frequency) temperature responses of GHEs.
By contrast, if the composite-medium line-source model is used To conclude, all the models examined here can be used to pre-
for the G-function, the unit-step temperature response is given by dict the medium-term temperature responses. We should use the
composite-medium line-source model (or similar models that
Rp
T f ðtÞ  T s;0 ¼ þ Gðt; r p Þ; ð20Þ address the effect of heat capacity in the borehole) to predict the
2
short-term responses, and we should use the finite line-source
where the thermal resistance, Rp, is defined in Eq. (9). The expres- models to calculate the long-term responses. Very recently, an ana-
sion for the G-function can be found in the references [26,84]. lytical model applicable to time scales from several minutes to dec-
Fig. 4 shows the calculated temperature responses and the ades has been built based on the idea of matched asymptotic
input parameters that were used. The time range is as long as expansion, which combines the composite-medium, the infinite,
100 years and is divided into three subintervals: short range (0– and the finite line-source solutions [86]. But this model was devel-
5tb), medium range (5tb < t < 0.1tH), and long range (>0.1tH). oped for a one-borehole GHE and cannot handle the thermal inter-
These responses are in reasonable accord with each other in the action among GHEs.
medium range (5tb < t < 0.1tH), which is roughly from several hours
to one year. But the match is poor for the short and long ranges. For
times larger than 0.1tH, all the models assuming that a borehole is 5. Developments in heat-transfer analysis of pile GHEs
infinitely long yield temperature responses that increase to infin-
ity, whereas the finite line-source models show temperatures Pile GHEs, or energy piles, use the foundation piles of buildings
approaching a steady state. The results for finite-length models to house heat transfer tubes. Various forms of pipes can be used,
are thus more reasonable and can be explained by the following e.g., U-, W-, and helical-shaped tubes. Unlike borehole GHEs,
physical insight: Heat transferred through the GHE is balanced energy piles may involve complex thermomechanical processes
by that through the ground surface. The comparison results indi- [87–89], a topic that is far from understood. To grasp the thermo-
cate that the use of an infinitely long model results in an mechanical process of a pile GHE, the thermal response must first
be addressed.
Foundation piles have much larger radii than borehole GHEs do,
but they are generally shorter. Thus, length-to-diameter ratio could
be an important parameter in evaluating energy piles [90,91]. The
heat capacity of a foundation pile must also be considered in con-
junction with the more complicated forms of heat transfer pipes
[26]. The problem of understanding transient heat conduction in
a cylindrical composite medium together with various types of
tube installations presents a huge challenge.
In addition to numerical methods [92–96], several simplified
methods have been used. First, some researchers applied the con-
cept of steady-state thermal resistance to energy piles [97,98]. As
discussed, foundation piles generally have relatively large diame-
ters. Thus, it should take longer for them to reach a steady state
than for borehole GHEs, which may take on the order of a day.
Second, empirical expressions are derived from numerical simula-
tions for the temperature responses of one pile GHE [91] and a field
of pile GHEs [99]. Third, the difference in thermal properties
between a foundation pile and the surrounding ground is ignored
[100–103]. In this way, the heat-source method is highly likely to
be the only viable theoretical method because various arrange-
ments of heat transfer channels exist. Modeling energy piles with
Fig. 4. Comparison of various analytical models for G-function. U- and W-shaped channels is relatively direct if the channels are
188 M. Li, A.C.K. Lai / Applied Energy 151 (2015) 178–191

assumed to be line heat sources in a homogenous medium [100]. results in self-verification. In this context, well-controlled sandbox
But the modeling of energy piles with helical-shaped coils is some- experiments are necessary.
what indirect. Three models have been developed. The ring-coil-
source model takes a helical-shaped coil to be a series of separated
6.2. Indoor sandbox experiments
rings of heat that release heat in an infinite or semi-infinite
homogenous medium [101]. The solid cylindrical-surface source
The advantages of indoor sandbox experiments include well-
model takes the helical pipe to be a cylindrical-surface source
controlled parameters, which enable the experimental model to
[102]. Finally, the helical-line-source model employs point sources
resemble the mathematical model, and independently determined
distributed along a helical line [103], which is the most physically
thermal properties, which enable model validation to be complete
reasonable assumption for a coil pipe. The homogenous-medium
and thorough. Unfortunately, very few indoor sandbox experi-
assumption used in these three heat-source models can cause
ments have been reported [67,123–127], and even worse is that
unacceptable errors because energy piles also involve a cylindrical
only one is satisfactory for model validation [127].
composite medium, resembling that of borehole GHEs. Thus, the
Yu et al. [123] and Park et al. [124] reported indoor sandbox
composite-medium line-source approach is a better alternative
experiments on U- and helical-shaped ground loops, respectively,
[26] because it can address the difference in properties between
for validating heat-transfer models. Reuss et al. also performed
the ground and a concrete foundation pile.
several sandbox experiments on different soil materials to validate
As in borehole GHEs, the movement of groundwater may be an
their numerical model for simulating coupled heat and moisture
influencing factor. One way of handling this factor is the moving
diffusion in borehole GHS [67]. These three tests were performed
heat-source method, which assumes seepage flow to be homoge-
in small-scale sandboxes and employed a homogeneous medium,
neous. It was used to extend the conventional heat-source solutions
rather than a composite medium, such as that used in common
for modeling energy piles with helical tubes [104–108]. Since foun-
GHEs. Erol and François performed small-scale sandbox experi-
dation piles are composed of dense concrete, groundwater cannot
ments to compare the thermal performance of several grouting
flow through them; that is, there should be no groundwater flow
materials [125]. Shirazi and Bernier also built a small-scale labora-
in foundation piles. The homogenous-flow approximation will not
tory tank to investigate heat transfer by borehole GHEs [126].
be as satisfactory for energy piles as for borehole GHEs, and the
These two small-scale experiments used backfilling material that
accuracy needs to be further verified and validated.
was different from the sand used in the sand box.
Beier et al. built a medium-scale sandbox with an 18-m-long
borehole [127,128]. This length permits the construction of a bore-
6. Sandbox and in situ thermal-response tests (TRTs)
hole GHE with a radius-to-length ratio close to what is used in an
actual situation. The borehole was grouted with a material other
6.1. In situ TRTs of GHEs
than sand, and the medium resembles what is used in actual
GHEs. In their first paper [127], Beier et al. did not report the heat
Lack of knowledge of the thermal properties of the ground is
capacities of sand and backfilling material, although they did so in
another serious obstacle to advances in GCHPs and GHEs. The
subsequent work [128]. The sandbox experiment of Beier et al.
properties and composition of underground materials vary greatly
may be the best so far for validating heat transfer models and
with location and depth. Thus, it is extremely complicated to deter-
parameter estimation procedures. The authors have used the
mine the spatial thermal properties of the ground. Practically, it is
reported data to validate the short-term performance of the com-
recommended that the equivalent (average) thermal properties of
posite-medium line-source model for GHEs with a single U-shaped
the ground be determined from in situ thermal-response tests
pipes [84,85].
(TRTs) [5]. In situ tests require a mobile testing facility. Such appa-
ratus was first developed and reported in Europe and the USA
[109,110]. In addition, many researchers have developed or 6.3. Estimation of parameters from TRTs as inverse problem
improved similar testing equipment in different regions of the
world, including Latin America [111], China [112], Japan [113], The design of GHEs and heat transfer calculations require the
Korea [114], Turkey [115], Cyprus [116], Saudi Arabia [117], thermal properties of the ground. The determination, or estima-
Tunisia [118], and Canada [119]. Most testing rigs use electrical-re- tion, of those properties from in situ TRTs or sandbox test data is
sistance heaters to increase the temperature of a circulating fluid, an inverse problem of heat conduction.
while some use heat pumps, which can both heat and cool the cir- Both analytical and numerical models have been used to esti-
culating fluid. There are, however, intrinsic uncertainties and prob- mate key parameters [20,21] (e.g., the thermal conductivities and
lems with the complexity and dynamic behavior of the heat pumps diffusivities of soil and grouting material). In general, estimation
used. procedures based on analytical models are simpler, faster, and
In our opinion, the usefulness of in situ TRTs is overrated. A sim- more robust than those based on numerical methods. But few of
ple statistical analysis that the authors performed indicates that them can determine the heat capacity of grouting material because
over 30% of the reports on GCHPs are associated with in situ tests of the steady-state assumption about heat conduction in a bore-
based on pilot or actual installations. Admittedly, in situ TRTs may hole [111–113]. For the same reason, these methods can only use
provide an overall evaluation of the energy performance of a par- late-stage testing data, and thus require a relatively long testing
ticular GCHP installation and yield a practical estimate of the effec- time to ensure accuracy [122]. On the other hand, procedures using
tive thermal properties of a particular location. But, from the numerical models, either one- or multi-dimensional [129–133],
viewpoint of model validation, the data from in situ TRTs is inap- can estimate the heat capacity of grout; and the field testing time
propriate. Some researchers used in situ test data to validate theo- may be short. However, they require more input parameters and
retical models [120,121]. Such validations are inadequate because are more sensitive to the accuracy of the input parameters than
the thermal properties of the ground are actually unknown, and algorithms based on analytical models.
because of various uncontrollable testing uncertainties and esti- Rainieri et al. [20] and Raymond et al. [21] have reviewed many
mated equivalent properties that are generally obtained by aspects of the estimation of parameters from in situ TRTs. One
curve-fitting approaches [122]. Clearly, a comparison of in situ data more issue emphasized here is that almost all the scientific effort
with model predictions based on curve-fitting properties only has been directed at parameter estimation, and little attention
M. Li, A.C.K. Lai / Applied Energy 151 (2015) 178–191 189

has been paid to the accuracy of the parameters or to the estima- References
tion of the error [134].
Error, or accuracy estimation, is an important issue that goes [1] Hellstrom G. Ground heat storage – thermal analysis of duct storage systems
I. Theory. Lund: University of Lund; 1991.
beyond the mere finding of the best-fit parameters [135]. To be [2] Pavlov GK, Olesen BW. Thermal energy storage—a review of concepts and
genuinely useful, an estimation procedure should provide estima- systems for heating and cooling applications in buildings: Part 1—Seasonal
tions of both a parameter and the error. Error estimates provided storage in the ground. HVAC&R Res 2012;18(3):515–38.
[3] Bauer D, Marx R, Nußbicker-Lux J, Ochs F, Heidemann W, Müller-Steinhagen
by linear fitting procedures fail in cases where the normality H. German central solar heating plants with seasonal heat storage. Sol Energy
assumption is violated or nonlinear curve-fitting is required. This 2010;84(4):612–23.
is especially true if early-stage data from in situ TRTs are used to [4] Ingersoll LR, Zobel OJ, Ingersoll AC. Heat conduction with engineering,
geological, and other applications. Revised ed. Madison: The University of
shorten the testing period. In practice, uncertainties that arise in
Wisconsin Press; 1954.
the field are difficult to control. What is needed to improve the [5] ASHRAE. ASHRAE handbook: HVAC applications. Atlanta: ASHRAE, Inc.; 2011.
confidence level of parameter estimation is a general method for [6] ASHRAE. ASHRAE handbook: HVAC systems and
evaluating the influence of uncertainties on the accuracy of param- equipment. Atlanta: ASHRAE, Inc.; 2008.
[7] Wu W, Wang B, You T, Shi W, Li X. A potential solution for thermal imbalance
eter estimation algorithms. of ground source heat pump systems in cold regions: Ground source
absorption heat pump. Renew Energy 2013;59:39–48.
[8] United States Environmental Protection Agency. Space conditioning: the next
7. Conclusions and future research suggestions frontier – the potential of advanced residential space conditioning
technologies for reducing pollution and saving consumers money. EPA 430-
R-93-004; 1993.
This paper provides an overview of GHE heat transfer analysis, [9] Yavuzturk C, Spitler JD. A short time step response factor model for vertical
which is a central issue in the design of GHEs. It evaluates existing ground loop heat exchangers. ASHRAE Trans 1999;105(2):475–85.
[10] Staffell I, Brett D, Brandon N, Hawkes A. A review of domestic heat pumps.
analytical models on the basis of time scale, with an emphasis on Energy Environ Sci 2012;5:9291–306.
various heat-source models. The geometric characteristics of [11] Florides G, Kalogirou S. Ground heat exchangers – a review of systems,
GHEs make Kelvin’s theory of heat source very attractive and effi- models, and applications. Renew Energy 2007;32:2461–79.
[12] Omer AM. Ground-source heat pumps systems and applications. Renew Sust
cient at solving related heat conduction problems.
Energy Rev 2008;12:344–71.
While considerable advances have been made, several funda- [13] Zhai XQ, Qu M, Yu X, Yang Y, Wang RZ. A review for the applications and
mental problems remain unsolved: integrated approaches of ground-coupled heat pump systems. Renew Sust
Energy Rev 2011;15:3133–40.
[14] Self SJ, Reddy BV, Rosen MA. Geothermal heat pump systems: status
(1) The first is how to efficiently solve the problem of heat con- review and comparison with other heating options. Appl Energy
duction in the ground over diverse time and space scales. 2013;101:341–8.
Numerical methods require vast computational resources [15] Bayer P, Saner D, Bolay S, Rybach L, Blum P. Greenhouse gas emission savings
of ground source heat pump systems in Europe: a review. Renew Sust Energy
and are thus impractical for tackling this challenge. Current Rev 2012;16:1256–67.
analytical models all have limitations with regard to time [16] de Moel M, Bach PM, Bouazza A, Singh RM, Sun JLO. Technological advances
scales, and an efficient analytical model is still unavailable and applications of geothermal energy pile foundations and their feasibility
in Australia. Renew Sust Energy Rev 2010;14:2683–96.
for calculating the temperature response of a large matrix [17] Lee JY. Current status of ground source heat pumps in Korea. Renew Sust
of GHEs on scales from less than an hour to decades. Energy Rev 2009;13:1560–8.
(2) The second is how to precisely and efficiently evaluate the [18] Yuan YP, Cao XL, Sun LL, Lei B, Yu NY. Ground source heat pump system: a
review of simulation in China. Renew Sust Energy Rev 2012;16:6814–22.
influence of coupled heat and moisture transfer in the [19] Capozza A, De CarliM, Zarrella A. Design of borehole heat exchangers for
ground on the performance of GCHP and GHS systems. ground-source heat pumps: a literature review, methodology comparison
Since different systems have different operating tempera- and analysis on the penalty temperature. Energy Build 2012;55:369–79.
[20] Rainieri S, Bozzoli F, Pagliarini G. Modeling approaches applied to the thermal
tures in the ground, moisture transfer should affect heat
response test: a critical review of the literature. HVAC&R Res
transfer to different extents. 2011;17(6):977–90.
(3) The third point is how to improve the accuracy and reliabil- [21] Raymond J, Therrien R, Gosselin L, Lefebvre R. A review of thermal response
test analysis using pumping test concepts. Ground Water 2011;49(6):
ity of parameter estimation based on in situ thermal-re-
932–45.
sponse tests (TRTs). Existing methods based on analytical [22] Zhang CX, Guo ZJ, Liu YF, Cong XC, Peng DG. A review on thermal response
models require somewhat long testing times (at least test of ground-coupled heat pump systems. Renew Sust Energy Rev
2 days); the longer the testing time is, the greater the possi- 2014;40:851–67.
[23] Lamarche L, Kajl S, Beauchamp B. A review of methods to evaluate borehole
bility is that accidents will occur in situ. Moreover, general thermal resistances in geothermal heat-pump systems. Geothermics
error or accuracy estimation methods based on analytical 2010;39:187–200.
or numerical models are not readily available yet. [24] Yang HX, Cui P, Fang ZH. Vertical-borehole ground-coupled heat pumps: a
review of models and systems. Appl Energy 2010;87:16–27.
(4) To validate heat transfer models and verify parameter esti- [25] Kavanaugh SK, Rafferty K. Ground-source heat pumps: design of geothermal
mation algorithms, we need very accurate sandbox data sets, systems for commercial and institutional buildings. Atlanta (GA): American
together with independently measured parameters. Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.; 1997.
[26] Li M, Lai ACK. New temperature response functions (G functions) for pile and
However, they are very rare. For borehole GHEs, the refer- borehole ground heat exchangers based on composite-medium line-source
ence data of Beier et al. can be used [126,127], despite some theory. Energy 2012;38:255–63.
improvements in the experiments being desirable. For pile [27] Bejan A. Convection heat transfer. 4th ed. Hoboken: John Wiley & Sons, Inc.;
2013.
GHEs, however, there is no reliable sandbox data for pillars
[28] Eskilson P. Thermal analysis of heat extraction boreholes [PhD Thesis].
with various types of heat exchange tubes. Sweden: University of Lund; 1987.
(5) The complex thermomechanical processes that occur in [29] Claesson J, Eskilson P. Conductive heat extraction to a deep borehole –
thermal analyses and dimensioning rules. Energy 1988;13:509–27.
energy piles are far from understood. Recent progress in
[30] Lee CK. Effects of multiple ground layers on thermal response test analysis
the heat transfer analysis of energy piles provides a good and ground-source heat pump simulation. Appl Energy
starting point for solving this problem. One critical process 2011;88(12):4405–10.
is how the periodic expansion and contraction of concrete [31] Bear J. Dynamics of fluids in porous media. New York: American Elsevier
Publishing Company Inc.; 1972.
pillars caused by heating and cooling processes affect the [32] Lee CK, Lam HN. A modified multi-ground-layer model for borehole ground
frictional force of foundation pillars, especially in the long heat exchangers with an inhomogeneous groundwater flow. Energy
term. 2012;47(1):378–87.
190 M. Li, A.C.K. Lai / Applied Energy 151 (2015) 178–191

[33] Cimmino M, Bernier M. A semi-analytical method to generate g-functions for [68] Yavuzturk C, Chiasson AD, Nydahl JE. Simulation model for ground loop heat
geothermal bore fields. Int J Heat Mass Transf 2014;70:641–50. exchangers. ASHRAE Trans 2009;115(2):45–59.
[34] Zanchini E, Lazzari S. New g-functions for the hourly simulation of double U- [69] Zarrella A, Scarpa M, De Carli M. Short time step analysis of vertical ground-
tube borehole heat exchanger fields. Energy 2014;70:444–55. coupled heat exchangers: the approach of CaRM. Renew Energy
[35] Marcotte D, Pasquier P. Unit-response function for ground heat exchanger 2011;36(9):2357–67.
with parallel, series or mixed borehole arrangement. Renew Energy [70] Zarrella A, Scarpa M, De Carli M. Short time-step performances of coaxial and
2014;68:14–24. double U-tube borehole heat exchangers: modeling and measurements.
[36] Mottaghy D, Dijkshoorn L. Implementing an effective finite difference HVAC&R Res 2011;17:959–76.
formulation for borehole heat exchangers into a heat and mass transport [71] Pasquier P, Marcotte D. Short-term simulation of ground heat exchanger with
code. Renew Energy 2012;45:59–71. an improved TRCM. Renew Energy 2012;46:92–9.
[37] Yavuzturk C, Spitler JD, Rees SJ. A transient two-dimensional finite volume [72] Zarrella A, Capozza A, De Carli M. Performance analysis of short helical
model for the simulation of vertical U-tube ground heat exchangers. ASHRAE borehole heat exchangers via integrated modelling of a borefield and a heat
Trans 1999;105:475–85. pump: a case study. Appl Therm Eng 2013;61:36–47.
[38] Koohi-Fayegh S, Rosen MA. Examination of thermal interaction of multiple [73] Mitchell JW, Braun JE. Principles of heating, ventilation, and air conditioning
vertical ground heat exchangers. Appl Energy 2012;97:962–9. in buildings. Hoboken: Wiley; 2013.
[39] Rees SJ, He M. A three-dimensional numerical model of borehole heat [74] Ruiz-Calvo F, De Rosa M, Acuňa J, Corberán JM, Montagud C. Experimental
exchanger heat transfer and fluid flow. Geothermics 2013;46:1–13. validation of a short-term borehole-to-ground (B2G) dynamic model. Appl
[40] Bauer D, Heidemann W, Diersch HJG. Transient 3D analysis of borehole heat Energy 2015;140:210–23.
exchanger modeling. Geothermics 2011;40:250–60. [75] Gu Y, Oneal DL. An analytical solution to transient heat-conduction in a
[41] Choi JC, Lee SR, Lee DS. Numerical simulation of vertical ground heat composite region with a cylindrical heat-Source. J Sol Energy Eng – Trans
exchangers: intermittent operation in unsaturated soil conditions. Comput ASME 1995;117:242–8.
Geotech 2011;38:949–58. [76] Beier RA, Smith MD. Minimum duration of in-situ tests on vertical boreholes.
[42] Wołoszyn J, Gołaś A. Modelling of a borehole heat exchanger using a finite ASHRAE Trans 2003;109:475–86.
element with multiple degrees of freedom. Geothermics 2013;47:13–26. [77] Lamarche L, Beauchamp B. New solutions for the short-time analysis of
[43] Carslaw HS, Jaeger JC. Conduction of heat in solids. 2nd ed. Oxford: Claremore geothermal vertical boreholes. Int J Heat Mass Transf 2007;50:1408–19.
Press; 1959. [78] Bandyopadhyay G, Gosnold W, Mann M. Analytical and semi-analytical
[44] Özisik MN. Heat conduction. 2nd ed. New York: John Wiley & Sons, Inc.; 1993. solutions for short-time transient response ground heat exchangers. Energy
[45] Jaeger JC. Some problems involving line sources in conduction of heat. Build 2008;40:1816–24.
Philosophical Magazine Series 7 1944;35(242):169–79. [79] Bandyopadhyay G, Kulkarni M, Mann M. A new approach to modeling ground
[46] Li M, Lai ACK. Thermodynamic optimization of ground heat exchangers with heat exchangers in the initial phase of heat-flux build up. ASHRAE Trans
single U-tube by entropy generation minimization method. Energy Conv 2008;114:428–39.
Manage 2013;65:133–9. [80] Claesson J, Javed S. An analytical method to calculate borehole fluid
[47] Gu Y, O’Neal DL. Development of an equivalent diameter expression for temperatures for time-scales from minutes to decades. ASHRAE Trans
vertical U-tubes used in ground-coupled heat pumps. ASHRAE Trans 2011;117:279–88.
1998;104:347–55. [81] Javed S, Claesson J. New analytical and numerical solutions for the
[48] Shonder JA, Beck JV. Field test of a new method for determining soil formation short-term analysis of vertical ground heat exchangers. ASHRAE Trans
thermal conductivity and borehole resistance. ASHRAE Trans 2011;117:3–12.
1999;106:843–50. [82] Li M, Lai ACK. Analytical solution to heat conduction in finite hollow
[49] Paul ND. The effect of grout thermal conductivity on vertical geothermal heat composite cylinders with a general boundary condition. Int J Heat Mass
exchanger design and performance [MSc Thesis]. USA: South Dakota Transf 2013;60:549–56.
University; 1996. [83] Shirazi AS, Bernier M. Thermal capacity effects in borehole ground heat
[50] Sharqawy MH, Mokheimer EM, Badr HM. Effective pipe-to-borehole thermal exchangers. Energy Build 2013;67:352–64.
resistance for vertical ground heat exchangers. Geothermics 2009;38:271–7. [84] Li M, Lai ACK. Analytical model for short-time responses of ground heat
[51] Claesson J, Hellström G. Multipole method to calculate borehole thermal exchangers with U-shaped tubes: model development and validation. Appl
resistances in a borehole heat exchanger. HVAC&R Res 2011;17(6):895–911. Energy 2013;104:510–6.
[52] Hermanns M, Pérez JM. Asymptotic analysis of vertical geothermal boreholes [85] Yang Y, Li M. Short-time performance of composite-medium line-source
in the limit of slowly varying heat injection rates. SIAM J Appl Math model for predicting responses of ground heat exchangers with single U-
2014;74(1):60–82. shaped tube. Int J Therm Sci 2014;82:130–7.
[53] Zeng HY, Diao NR, Fang ZH. Heat transfer analysis of boreholes in vertical [86] Li M, Li P, Chan V, Lai ACK. Full-scale temperature response function (G-
ground heat exchangers. Int J Heat Mass Transf 2003;46:4467–81. function) for heat transfer by borehole ground heat exchangers (GHEs) from
[54] Deerman JD, Kavanaugh SP. Simulation of vertical U-tube ground-coupled sub-hour to decades. Appl Energy 2014;136:197–205.
heat pump systems using the cylindrical heat source solution. ASHRAE Trans [87] Brandl H. Energy foundations and other thermo-active ground structures.
1990;97:287–95. Géotechnique 2006;56(2):81–122.
[55] Zeng HY, Diao NR, Fang ZH. A finite line-source model for boreholes in [88] Amatya BL, Soga K, Bourne-Webb PJ, Amis T, Laloui L. Thermo-mechanical
geothermal heat exchangers. Heat Transf Asian Res 2002;31:558–67. behavior of energy piles. Géotechnique 2012;62(6):503–19.
[56] Cui P, Yang HX, Fang ZH. Heat transfer analysis of ground heat exchangers [89] Bourne-Webb PJ, Amatya B, Soga K. A framework for understanding energy
with inclined boreholes. Appl Therm Eng 2006;26:1169–75. pile behavior. Proc ICE-Geotech Eng 2012;166(2):170–7.
[57] Lamarche L, Beauchamp B. A new contribution to the finite line-source model [90] Loveridge F, Powrie W. Pile heat exchangers: thermal behaviour and
for geothermal boreholes. Energy Build 2007;39:188–98. interactions. Proc ICE-Geotech Eng 2012;166(2):178–96.
[58] Eckert ERG, Drake RM. Analysis of heat and mass [91] Loveridge F, Powrie W. Temperature response functions (G-functions) for
transfer. Washington: Hemisphere Publishing Corporation; 1987. single pile heat exchangers. Energy 2013;57:554–64.
[59] Diao NR, Li QY, Fang ZH. Heat transfer in ground heat exchangers with [92] Gao J, Zhang X, Liu J, Li K, Yang J. Numerical and experimental assessment of
groundwater advection. Int J Therm Sci 2004;43:1203–11. thermal performance of vertical energy piles: an application. Appl Energy
[60] Sutton MG, Nutter DW, Couvillion RJ. A ground resistance for vertical bore 2008;85(10):901–10.
heat exchangers with groundwater flow. J Energy Resour Technol – Trans [93] Xing L, Cullin JR, Spitler JD, Im P, Fisher DE. Foundation heat exchangers for
ASME 2003;125(3):183–9. residential ground source heat pump systems — numerical modeling and
[61] Molina-Giraldo N, Bayer P, Blum P. Evaluating the influence of thermal experimental validation. HVAC&R Res 2011;17(6):1059–74.
dispersion on temperature plumes from geothermal systems using analytical [94] Ghasemi-Fare O, Basu P. A practical heat transfer model for geothermal piles.
solutions. Int J Therm Sci 2011;50:1223–31. Energy Build 2013;66:470–9.
[62] Molina-Giraldo N, Blum P, Zhu K, Bayer P, Fang ZH. A moving finite line [95] Lee CK, Lam HN. A simplified model of energy pile for ground-source heat
source model to simulate borehole heat exchangers with groundwater pump systems. Energy 2013;55:838–45.
advection. Int J Therm Sci 2011;50:2506–13. [96] Zarrella A, De Carli M, Galgaro A. Thermal performance of two types of energy
[63] Chiasson A, O’Connell A. New analytical solution for sizing vertical borehole foundation pile: helical pipe and triple U-tube. Appl Therm Eng
ground heat exchangers in environments with significant groundwater flow: 2013;61(2):301–10.
parameter estimation from thermal response test data. HVAC&R Res [97] Loveridge F, Powrie W. 2D thermal resistance of pile heat exchangers.
2011;17(6):1000–11. Geothermics 2014;50:122–35.
[64] Tye-Gingras M, Gosselin L. Generic ground response functions for ground [98] Hu PF, Zha J, Lei F, Zhu N, Wu TH. A composite cylindrical model and its
exchangers in the presence of groundwater flow. Renew Energy application in analysis of thermal response and performance for energy pile.
2014;72:354–66. Energy Build 2014;84:324–32.
[65] Erol S, Hashemi MA, François B. Analytical solution of discontinuous heat [99] Loveridge F, Powrie W. G-functions for multiple interacting pile heat
extraction for sustainability and recovery aspects of borehole heat exchangers. Energy 2014;64:747–57.
exchangers. Int J Therm Sci 2015;88:47–58. [100] Bozis D, Papakostas K, Kyriakis N. On the evaluation of design parameters
[66] Luikov AV. Systems of differential equations of heat and mass transfer in effects on the heat transfer efficiency of energy piles. Energy Build
capillary-porous bodies. Int J Heat Mass Transf 1975;18(1):1–14. 2011;43:1020–9.
[67] Reuss M, Beck M, Müller JP. Design of a seasonal thermal energy storage in [101] Cui P, Li X, Man Y, Fang ZH. Heat transfer analysis of pile geothermal heat
the ground. Sol Energy 1997;59(4):247–57. exchangers with spiral coils. Appl Energy 2011;88:4113–9.
M. Li, A.C.K. Lai / Applied Energy 151 (2015) 178–191 191

[102] Man Y, Yang H, Diao NR, Liu J, Fang ZH. A new model and analytical solutions [118] Naili N, Attar I, Hazami M, Farhat A. First in situ operation performance test of
for borehole and pile ground heat exchangers. Int J Heat Mass Transf ground source heat pump in Tunisia. Energy Conv Manage 2013;75:292–301.
2010;53:2593–601. [119] Raymond J, Lamarche L. Development and numerical validation of a novel
[103] Li M, Lai ACK. Heat-source solutions to heat conduction in anisotropic media thermal response test with a low power source. Geothermics
with application to pile and borehole ground heat exchangers. Appl Energy 2014;51:434–44.
2012;96:451–8. [120] Michopoulos A, Kyriakis N. Predicting the fluid temperature at the exit of the
[104] Zhang W, Yang H, Lu L, Fang ZH. Investigation on heat transfer around buried vertical ground heat exchangers. Appl Energy 2009;86(10):2065–70.
coils of pile foundation heat exchangers for ground-coupled heat pump [121] Hu P, Yu Z, Zhu N, Lei F, Yuan X. Performance study of a ground heat
applications. Int J Heat Mass Transf 2012;55:6023–31. exchanger based on the multipole theory heat transfer model. Energy Build
[105] Zhang W, Yang HX, Lu L, Fang ZH. The analysis on solid cylindrical heat 2013;65:231–41.
source model of foundation pile ground heat exchangers with groundwater [122] Li M, Lai ACK. Parameter estimation of in-situ thermal response tests for
flow. Energy 2013;55:417–25. borehole ground heat exchangers. Int J Heat Mass Transf 2012;55:2615–24.
[106] Zhang W, Yang HX, Lu L, Cui P, Fang ZH. The research on ring-coil heat [123] Yu Y, Ma Z, Li X. A new integrated system with cooling storage in soil and
transfer models of pile foundation ground heat exchangers in the case of ground-coupled heat pump. Appl Therm Eng 2008;28(11):1450–62.
groundwater seepage. Energy Build 2014;71:115–28. [124] Park HK, Lee SR, Yoon S, Shin H, Lee DS. Case study of heat transfer behavior
[107] Zhang W, Yang HX, Cui P, Lu L, Diao NR, Fang ZH. Study on spiral source of helical ground heat exchanger. Energy Build 2012;53:137–44.
models revealing groundwater transfusion effects on pile foundation ground [125] Erol S, François B. Efficiency of various grouting materials for borehole heat
heat exchangers. Int J Heat Mass Transf 2015;84:119–29. exchangers. Appl Therm Eng 2014;70:788–99.
[108] Go GH, Lee SR, Kang HB, Yoon S, Kim MJ. A novel hybrid design algorithm for [126] Shirazi AS, Bernier M. A small-scale experimental apparatus to study heat
spiral coil energy piles that considers groundwater advection. Appl Therm transfer in the vicinity of geothermal boreholes. HVAC&R Res
Eng 2015;78:196–208. 2014;20(7):819–27.
[109] Austin III WA, Yavuzturk C, Spitler JD. Development of an in-situ system and [127] Beier RA, Smith MD, Spitler JD. Reference data sets for vertical borehole
analysis procedure for measuring ground thermal properties. ASHRAE Trans ground heat exchanger models and thermal response test analysis.
2000;106:365–79. Geothermics 2011;40:79–85.
[110] Gehlin S. Thermal response test: method development and evaluation [PhD [128] Beier RA. Transient heat transfer in a U-tube borehole heat exchanger. Appl
Thesis]. Sweden: Luleå University of Technology; 2002. Therm Eng 2014;62(1):256–66.
[111] Roth P, Georgiev A, Busso A, Barraza E. First in situ determination of ground [129] Shonder JA, Beck JV. Field test of a new method for determining soil
and borehole thermal properties in Latin America. Renew Energy formation thermal conductivity and borehole resistance. ASHRAE Trans
2004;29:1947–63. 2000;106(1):843–50.
[112] Yu M, Peng X, Li X, Fang ZH. A simplified model for measuring thermal [130] Gehlin S, Hellström G. Comparison of four models for thermal response test
properties of deep ground soil. Exp Heat Transf 2004;17:119–30. evaluation. ASHRAE Trans 2003;109:131–42.
[113] Fujii H, Okubo H, Nishi K, Itoi R, Ohyama K, Shibata K. An improved thermal [131] Wagner R, Clauser C. Evaluating thermal response tests using parameter
response test for U-tube ground heat exchanger based on optical fiber estimation for thermal conductivity and thermal capacity. J Geophys Eng
thermometers. Geothermics 2009;38:399–406. 2005;2(4):349–56.
[114] Hwang S, Ooka R, Nam Y. Evaluation of estimation method of ground [132] Marcotte D, Pasquier P. On the estimation of thermal resistance in borehole
properties for the ground source heat pump system. Renew Energy thermal conductivity test. Renew Energy 2008;33:2407–15.
2010;35:2123–30. [133] Bozzoli F, Pagliarini G, Rainieri S, Schiavi L. Estimation of soil and grout
[115] Esen H, Inalli M. In-situ thermal response test for ground source heat pump thermal properties through a TSPEP (two-step parameter estimation
system in Elazig, Turkey. Energy Build 2009;41:395–401. procedure) applied to TRT (thermal response test) data. Energy
[116] Florides G, Kalogirou S. First in situ determination of the thermal 2011;36(2):839–46.
performance of a U-pipe borehole heat exchanger, in Cyprus. Appl Therm [134] Javed S, Spitler J, Fahlén P. An experimental investigation of the accuracy of
Eng 2008;28:157–63. thermal response tests used to measure ground thermal properties. ASHRAE
[117] Sharqawy MH, Said SA, Mokheimer EM, Habib MA, Badr HM, Al-Shayea NA. Trans 2011;117(1):13–21.
First in situ determination of the ground thermal conductivity for borehole [135] Box GEP, Hunter JS, Hunter WG. Statistics for experimenters: design,
heat exchanger applications in Saudi Arabia. Renew Energy 2009;34: Innovation, and Discovery. 2ed. Hoboken: John Wiley & Sons, Inc.; 2005.
2218–23.

You might also like