Accepted Manuscript

A modified mathematical model for spiral coil-type horizontal ground heat exchangers

Jun-Seo Jeon, Seung-Rae Lee, Min-Jun Kim

PII: S0360-5442(18)30597-8
DOI: 10.1016/j.energy.2018.04.007
Reference: EGY 12646

To appear in: Energy

Received Date: 28 January 2017

Revised Date: 15 March 2018
Accepted Date: 2 April 2018

Please cite this article as: Jeon J-S, Lee S-R, Kim M-J, A modified mathematical model for spiral coil-
type horizontal ground heat exchangers, Energy (2018), doi: 10.1016/j.energy.2018.04.007.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT

1
2

3 A modified mathematical model for spiral coil-type horizontal ground heat

4 exchangers
5
6 Jun-Seo Jeon1+, Seung-Rae Lee2*, Min-Jun Kim3

PT
7
8 1,2,3
Department of Civil and Environmental Engineering, KAIST, Daejeon, 305-701, South Korea
9

RI
10
11 ABSTRACT

SC
12
13 Although many studies have investigated spiral coil-type horizontal ground heat exchangers (HGHEs),

14 a reliable mathematical model of them has not yet been developed. Accordingly, this study proposes a

15
U
modification of existing mathematical model, which was originally developed for energy piles and vertical
AN
16 ground heat exchangers, for HGHEs. Based on the solution of the heat equation using Green's function, the new

17 mathematical model was derived by employing the mirror image method and superposition to consider the
M

18 effect of a semi-infinite medium and ground temperature distribution. The prediction of the mathematical model

19 showed a good fit with the results from on-site experiments with a mean absolute percentage error of 0.30 %
D

20 and was perfectly matched with the numerical results. The proposed mathematical model is capable of capturing
TE

21 the structure of the spiral coil exactly, and consequently, it offers a more accurate prediction of ground

22 temperature.
EP

23
24 Keywords: spiral coil-type; horizontal ground heat exchanger; thermal response test; Green’s function;
25 numerical simulation
C

26
AC

27 *Corresponding author
28 e-mail: srlee@kaist.ac.kr
29 phone: +82-42-350-3617
30 fax: +82-42-350-7200
31
32 +
Current address: Department of Civil and Environmental Engineering, KAIST, 291, Daehakro, Yuseong-gu,
33 Daejeon, 34141, South Korea
34
35

1
 ACCEPTED MANUSCRIPT

36
37 1. Introduction

38 As a common form of renewable energy, geothermal energy has increased in important due to the

39 need to reduce fossil energy consumption and to fight climate change. Originating from environmentally

40 friendly sources, geothermal energy originating is primarily used in electric generation and space heating and

PT
41 cooling. Because ground temperature changes with depth, geothermal energy is available in various forms

42 depending on the depth of the source: Heat source can be classified as deep (4-5 km), intermediate (0.1-4 km),

RI
43 or shallow (0-0.1 km) [1]. Geothermal energy from a deep depth, where the ground temperature reaches nearly

44 200 °C , is used to produce electricity. By pushing water into hot fractured rock with an injection well, and

SC
45 collecting hot geothermal water through a production well, a large volume of thermal energy is obtained, and is

46 then converted to electricity in power turbines. Currently operating geothermal power plants can be categorized

U
47 as dry-steam plants, flash-steam plants and binary-cycle plants. Although single- and double-flash power plants
AN
48 are suitable for high-temperature geothermal sources (greater than 150 °C ), binary power plants, which use an

49 organic fluid instead of geothermal water, are chosen due to their ability to generate electricity from medium-
M

50 temperature geothermal sources (between 90 and 150 °C ) [2]. Extensive studies have been performed on these

51 power cycle technologies, examining single-flash power plants [3] to transcritical carbon dioxide Rankine cycle
D

52 plants coupled with the liquefied natural gas process [4, 5].
TE

53 For space heating and cooling, sources of geothermal energy from intermediate and shallow depths are

54 generally used with ground source heat pump (GSHP) systems. Compared with other conventional heating and

55 cooling systems, GSHP systems feature a lower operational cost and have less impact on the environment,
EP

56 among other advantages [6]. Ground temperature is generally lower than the temperature of the atmosphere in

57 summer, but higher in winter. Therefore, GSHP systems utilize the ground as a heat source in winter, and as a
C

58 heat sink in summer. A GSHP system consists of three main components: the ground heat exchanger, heat pump
AC

59 unit and heat distribution system (Fig. 1). By circulating the working fluid through a ground heat exchanger

60 (GHE), the GSHP system can release heat to the ground for space cooling in summer, and extract heat from the

61 ground for space heating in winter. Therefore, a GHE plays an important role in the heat transfer mechanism.

62 Vertical ground heat exchangers (VGHEs) and horizontal ground heat exchangers (HGHEs) are commonly used

63 in closed-loop GSHP systems. In general, a VGHE is installed at an intermediate depth (in Korea, greater than

64 150m), and an HGHE is placed at a shallow depth (in Korea, typically 2-3 m). Due to the effect of groundwater

2
 ACCEPTED MANUSCRIPT

65 flow and higher thermal properties of the surrounding ground, such as rock, the heat transfer performance of

66 VGHEs is higher than that of HGHEs [7, 8, 9]. Although drilling work increases construction cost, low land

67 demand and the high performance of VGHEs enables their wide use in urban areas. Conversely, HGHEs, with

68 the advantage of low excavation costs, are suitable for suburban areas where sufficient area is available. In

69 Korea, closed-loop GSHP systems with VGHEs accounted for 65.1% of all GSHP systems in 2008 and 74.0%

PT
70 in 2011. However, the percentage of closed-loop GSHP systems with HGHEs was 3.5% in 2008 and 6.0% in

71 2011. The remainder consisted of standing column well system and other systems [10, 11]. Although the use of

RI
72 VGHEs was still overwhelming, the rate of increase in HGHE applications was noticeably sharper.

73 Recently, many numerical and experimental studies have investigated the heat transfer mechanisms of

SC
74 the various types of HGHE. Fig. 2 presents a schematic view of the various types of HGHEs. Using CFD

75 simulations, Congedo et al. [12] investigated the configuration effect of HGHE on the heat fluxes transferred to

U
76 and from the ground. Various parameters such as the velocity of the fluid, installation depth, the thermal
AN
77 conductivity of the ground, and pitch size were considered in an investigation of the influence of heat fluxes in

78 linear, spiral coil, and slinky-type HGHEs. These researchers determined that the depth of the HGHE
M

79 installation does not play an important role in the system performance. In addition, they noted that the spiral

80 coil-type HGHE exhibits higher heat transfer performance than line and slinky-type HGHEs. Kim et al. [13]
D

81 performed a numerical simulation and experiments using thermal response tests (TRTs) to determine the factors
TE

82 that influence the performance of HGHEs. They conducted a parametric study based on various values of

83 ground thermal conductivity and pipe diameter. The results showed that pipe diameter does not affect system

84
EP

performance, and that the performance of the spiral coil-type HGHE is 10% better than that of the slinky-type

85 HGHE. Han et al. [14] analyzed the short-term and annual performance of geothermal heat exchangers with

86 varying configurations and geological conditions. Although major conclusions were mainly focused on the
C

87 importance of local geological data utilization, it was also concluded that linear HGHEs with two pipes per
AC

88 trench and spiral coil-type HGHEs exhibited better performance in thermal energy extraction for the identical

89 installation area and similar total length.

90 Certain studies were also conducted to investigate additional influence parameters such as pipe

91 materials and heat pump units. Using CFD simulation, Selamat et al. [15] considered different layouts and pipe

92 materials to optimize the HGHE design. Li et al. [16] investigated the operating characteristics of spiral coil-

93 type HGHEs to consider the effect of the heat pump, subsurface factors, and daily variations. These researchers

3
 ACCEPTED MANUSCRIPT

94 determined that the heat pump should be considered when analyzing the performance of GSHP systems and that

95 the thermal conductivity of the ground and pipe spacing significantly affect the performance of HGHEs.

96 Despite the superior performance of these spiral coil-type HGHEs, mathematical models of spiral coil-

97 type HGHEs are rare. Since the spiral coil-type GHE had initially drawn attention as a suitable GHE for the

98 energy pile, which utilizes the foundation piles of buildings as load supports and as ground heat exchangers,

PT
99 most mathematical models were developed for the VGHE.

100 Man et al. [17] proposed a solid cylindrical heat source model using Green’s function. In this model,

RI
101 the spiral coil was assumed to be a continuous cylindrical heat source, and the inner part of the cylinder was

102 identical to the outer part. Cui et al. [18] proposed a ring-coil heat source model using the cylindrical heat source

SC
103 model by simplifying the geometry of the spiral shape as a series of separated ring coils. Because the

104 discontinuity effects of the heat source and the coil pitch were considered, this model is justifiably described as

U
105 the heat transfer process of the spiral coil when it is compared with both the infinite ring-coil and cylindrical
AN
106 source models. Man et al. [19] also developed a spiral heat source model to capture the exact geometry of the

107 spiral coil pipes; however, this caused a computational difficulty induced by both the singularity and the
M

108 presence of double integrations. By changing the variable into an error function, Park et al. [20] alleviated the

109 singularity problem, and consequently simplified the computation. This mathematical model was also validated
D

110 by comparison with a numerical study and test measurements from a thermal response test conducted in a sand
TE

111 chamber.

112 Recently, Wang et al. [21] suggested a new mathematical model for a spiral coil-type HGHE using a

113
EP

concept similar to the ring-coil heat source model proposed by Cui et al. [18]. Since the temperature of the

114 ground surface significantly affected the performance of HGHEs, it was also considered in this model.

115 Nevertheless, this model still differs from the real geometry of a spiral coil shape.
C

116 The purpose of this paper is to present a mathematical model for spiral coil-type HGHEs that can
AC

117 capture the real geometry of a spiral coil. To this end, a modification was performed of the spiral coil source

118 mathematical model proposed by Park et al. [20], which was originally developed for energy piles and VGHEs,

119 using the mirror image method and superposition. Validation was also conducted using on-site experiments and

120 numerical analysis results implemented in COMSOL Multiphysics. In addition, the proposed mathematical

121 model was compared with the multiple ring-coils source model proposed by Wang et al. [21].

122

4
 ACCEPTED MANUSCRIPT

123 2. Mathematical models for a Spiral Coil-Type Ground Heat Exchanger

124 The modification of the mathematical model for a spiral coil-type HGHE is addressed in this section.

125 The solution of the heat equation in a homogeneous, isotropic solid body with a point heat source was first

126 introduced using Green’s function. With that, the derivation process of the mathematical model for the spiral

127 coil-type VGHE proposed by Park et al. [20] was complete. With the description of the difference in the

PT
128 derivation process between the VGHEs and HGHEs, the mathematical model for a spiral coil-type HGHE was

129 proposed.

RI
130 The general heat equation with a heat source in a homogeneous, isotropic solid body Ω ⊂ 3
can be

131 expressed as follows:

SC
∂θ
− α ∆θ = Q (u , t ) in Ω
∂t
132 θ (u , 0) = j (u , 0) in Ω ( x, y , z ) ∈ Ω (1)

U
θ (u , t ) = h (u , t ) on Ω
AN
133 where 3
is three-dimensional Euclidean space, θ is the temperature, t denotes time, ∆ is the Laplace

134 operator, α is the thermal diffusivity, Q(u, t ) is the source density, u is the position vector, j (u, 0) is a
M

135 known function defined in the boundary, and h(u , t ) is a known function defined on the boundary.

136 The second and third equations in Eq. (1) represent the initial condition and the Dirichlet boundary
D

137 condition, respectively.

TE

138 Based on Green’s function theory [22, 23], the solution of Eq. (1) can be expressed in terms of the boundary

139 condition, the initial condition and Green’s function by solving Eq. (2).
EP

∂G
140 − α∆G = δ (u − u ')δ (t − τ ) (2)
∂t
C

141 where u ' denotes the position of the source, and x ' , y ' , and z ' are the coordinates of the source in a

142
AC

three-dimensional space.

143 Eq. (2) expresses the physical meaning of the temperature distribution within a medium when a unit of

144 heat is applied at u ' at time τ . If the time t is equal to τ , the temperature is zero at all points except

145 u '( x ', y ', z ') , where it becomes infinite.

146 The temperature variation for the general heat equation using Green’s function can be described as

147 follows:

5
 ACCEPTED MANUSCRIPT

t t ∂G
148 θ (u, t ) = ∫ ∫ G (u, t | u ',τ )Q(u ',τ )du ' dτ − α ∫ ∫ h dSdτ + ∫ j ⋅ G (u, t | u ',τ )du ' (3)
0 Ω 0 ∂Ω ∂n Ω

149 where S is the boundary of domain Ω , and the second term on the right-hand side denotes the surface

150 integral over the boundary of domain Ω .

151

PT
152 The effects of the heat source and the boundary condition are included in the first and second terms,

153 respectively, and the final term includes the effect of the initial condition. Within the free space, which is a

RI
154 domain of infinite extent, Green’s function for heat conduction in Cartesian coordinates is expressed as follows

155 [22, 23]:

SC
 ( x − x ')2 + ( y − y ')2 + ( z − z ')2 
− 
1 4α ( t −τ )
156 G (u, t | u ',τ ) = e  
(4)
8 [πα (t − τ )]
3/ 2

U
157
AN
158 2.1 Spiral coil source mathematical model for a VGHE

159
M

160 Park et al. [20] proposed the mathematical model for the heat conduction problem for a spiral coil-

161 type VGHE with the following conditions:

D

∂θ
= α∆θ + Q(u, t ) for z > 0
∂t
TE

162 θ (u, 0) = 0 for z > 0 (5)

θ (u , t ) = 0 for z = 0

163 Because the spiral coil trajectory can be expressed using the sine and cosine functions, the heat source
EP

164 was defined as follows:

ql
C

165 Q (u , t ) = δ (u − s ( z )) for 0 ≤ z ≤ hS (6)

ρc
AC

166 where s ( z ) = (r0 cos(2π z / pS ), r0 sin(2π z / pS ), z ) is the spiral coil trajectory, pS is the pitch size, r0 is the

167 radius of the coil.

168 To consider the semi-infinite medium, the mirror method was applied as shown in Fig. 3. Green’s

169 function was replaced by the following:

170 G% (u , t | u ',τ ) = G (u , t | u ',τ ) − G (u , t | u * ', τ ) (7)

171 where u *' = ( x ', y ', − z ') is the position of the image spiral coil source, which is a reflection of the spiral coil

6
 ACCEPTED MANUSCRIPT

172 source in the z-axis.

173 According to Eq. (5), the effects of the boundary and initial conditions on the temperature variation

174 were not considered. Finally, the temperature variation for the spiral coil source can be obtained by integrating

175 the point heat source along a spiral coil:

ql t ∞ %
θ (u, t ) = G(u, t | u ',τ )du ' dτ
ρ c ∫0 ∫0

PT
176 F ( x , y , z ')  ( z − z ') 2
( z + z ')  2 (8)
ql t 1 hS − − −
4α (t −τ ) 4α (t −τ ) 4α (t −τ )
dz ' dτ
8(πα )3/ 2 ρ c ∫0 (t − τ )3/ 2 ∫0
= e  e − e
 
 

RI
177 with F ( x, y, z ') = x2 + y 2 + r02 − 2 xr0 cos(2π z '/ pS ) − 2 yr0 sin(2π z '/ pS ) .

SC
178 Using the complementary error function, Eq. (8) was modified to address the integration difficulty

179 caused by a singularity:

U
 − ( z − z ') ( z + z ') 
2 2
F ( x , y , z ')
ql t 1 hS − −
θ (u , t ) = ∫ ∫ e 4α ( t −τ )  e 4α (t −τ ) − e 4α ( t −τ ) dz ' dτ
8(πα ) ρ c 0 (t − τ )
3/ 2 3/ 2 0  
AN
180   (9)
ql hS erfc ( A− (u , z ') / 2 α t ) erfc ( A+ (u , z ') / 2 α t )
=
4πλ ∫0 A− (u , z ')
−
A+ (u , z ')
dz '
M

181 with A± ( u , z ') = F ( x , y , z ') + ( z ± z ') 2

182
D

183 2.2 Spiral coil source mathematical model for a HGHE

TE

184

185 The derivation of the spiral coil source solution for a HGHE can be performed using a process similar
EP

186 to that used in the spiral coil source solution for the VGHE. The differences between the two processes include

187 the location of the spiral coil source and the effect of the ground surface temperature. As in the derivation of the
C

188 spiral coil source mathematical model for the VGHE, a semi-infinite medium was considered, using the mirror
AC

189 image method and Green’s function in Eq. (4), which is valid in a domain of infinite extent. Therefore, the effect

190 of the spiral coil source location was also considered with conditions identical to those in Eq. (5). The effect of

191 the ground surface temperature was addressed through the superposition method and the mathematical model

192 for undisturbed ground temperature. Fig. 4 shows the schematic view of a spiral coil source for the HGHE. The

193 spiral coil source for the HGHE is located mainly along the y-axis and not the z-axis, as in the VGHE. Therefore,

194 Green’s function, which represents temperature variation due to a point source, must be integrated along the y-

195 axis. In addition, parallel transport of the spiral coil source is required to consider the location of HGHE buried

7
 ACCEPTED MANUSCRIPT

196 below the ground surface since the origin of the z-axis in Cartesian coordinates was set as the ground surface.

197 Consequently, the spiral coil heat source can be construed as follows:

ql
198 Q (u , t ) = δ (u − s ( y )) for 0 ≤ y ≤ hS (12)
ρc

199 where s( y ) = (r0 cos(2π y / pS ), y, r0 sin(2π y / pS )) is the spiral coil trajectory.

PT
200 If the spiral coil source for the HGHE is considered in an infinite medium, the temperature variation

201 can be obtained by the following:

RI
ql t ∞
θ (u, t ) = G (u , t | u ',τ )du ' dτ
ρ c ∫0 ∫0
202 F0 ( x , y ', z ) + ( y − y ')2
(13)
hS −

SC
ql t 1
= ∫ 3/ 2 ∫0
e 4α ( t −τ )
dy ' dτ
8(πα ) ρ c (t − τ )
3/ 2 0

203 where Fo ( x, y ', z ) = x2 + ( z − d )2 + r02 − 2 xr0 cos(2π y '/ pS ) − 2( z − d )r0 sin(2π y '/ pS ) .

204
U
Based on the mirror image method shown in Fig. 5, the boundary condition for the ground surface can
AN
205 be assumed to be an adiabatic surface condition or isothermal ground surface condition, depending on the sign

206 of the source in the image component [24]. A positive sign indicates an image source with a strength equal to
M

207 that of the original source, and a negative sign denotes an image sink with counter strength. In the same manner,

208
D

the temperature variation due to the image spiral coil source for HGHE can be obtained by the following:

ql t ∞
θ (u , t ) = G (u , t | u ',τ )du ' dτ
ρ c ∫0 ∫0
TE

209 Fim ( x , y ', z ) + ( y − y ') 2

(14)
ql t 1 hS −
= ∫ 3/ 2 ∫0
e 4α ( t −τ )
dy ' dτ
8(πα ) ρ c (t − τ )
3/ 2 0
EP

210 where Fim ( x, y ', z) = x2 + ( z + d )2 + r02 − 2 xr0 cos(2π y '/ pS ) − 2( z + d )r0 sin(2π y '/ pS ) .
C

211 Applying the superposition method, the temperature variation due to the spiral coil source for the
AC

212 HGHE in the semi-infinite medium can be expressed as follows:

ql t ∞
θ (u , t ) = G (u , t | u ',τ )du ' dτ
ρ c ∫0 ∫0
213 Fo ( x , y ', z ) + ( y − y ')2 Fim ( x , y ', z ) + ( y − y ')2
(15)
ql t 1 h − −
= ∫ 3/ 2 ∫0
e 4α ( t −τ )
−e 4α ( t −τ )
dy ' dτ
8(πα ) ρ c (t − τ )
3/ 2 0

214 Using the complementary error function, the temperature variation can be written as follows:

ql h erfc( Ao (u , y ') / 2 α t ) erfc( Aim (u , y ') / 2 α t )

θ (u , t ) =
4πλ ∫0
215 − dy ' (16)
Ao (u , y ') Aim (u , y ')

8
 ACCEPTED MANUSCRIPT

216 where Ao / im (u , y ') = Fo / im ( x, y ', z ) + ( y − y ') 2 .

217 Unlike for the VGHE, the temperature effect of the ground surface should be considered for the HGHE.

218 The temperature of the ground surface varies with time, and this change has a significant effect on the

219 performance of the HGHE because it is generally installed at a shallow depth [25, 26]. Of the various

220 mathematical models for estimating ground temperature when the ground surface temperature is subject to

PT
221 variation, the most commonly used is the simple harmonic model [27]. Relying on pure heat conduction theory,

222

RI
a mathematical model for estimating ground temperature that varies with time and depth was suggested [28]:

223

SC
−z  
2α  2π z 
224 θ ( z, t ) = θ s , avg + As ⋅ e 2π /365
⋅ cos  ⋅ (t − ϕ ) −  (17)
 365 2α 
 2π / 365 

U
where θ ( z , t ) is the ground temperature at depth z and time t of the year, θs,avg is the annual average ground
AN
225

226 temperature, As is the surface amplitude, z is the ground depth, α is the thermal diffusivity of the ground,
M

227 and ϕ is the phase change of the annual soil temperature cycle.

228
D

229 3. Numerical Simulation Procedure

TE

230 Numerical modeling of heat transfer for the spiral coil-type HGHE was performed using the

231 commercial finite element code in COMSOL Multiphysics. As in the mathematical model, the ground is
EP

232 assumed to be a solid medium.

233 The generalized energy equation for a solid medium is given as follows:
C

∂T
234 ρ g C p, g + ρ C p v g ⋅ ∇T + ∇ ⋅ q = Q (18)
∂t
AC

235 q = − k g ∇T (19)

236 where ρg is the solid density, C p , g is the solid heat capacity at a constant pressure, T is the temperature of

237 the solid, t is the time, v g is the velocity field, q is the heat flux, Q is the heat source, and k g is the

238 thermal conductivity of the solid.

239 The generalized mass, momentum and energy equations for an incompressible fluid flowing in a pipe

240 are given as follows:

9
 ACCEPTED MANUSCRIPT

241

∂Ap ρ f
242
∂t t ( t )
+ ∇ ⋅ Aρ f v p e = 0 (20)

∂v p 1 ρf
243 ρf = −∇ t p ⋅ et − fD v p v p + R ⋅ et (21)
∂t 2 dh

∂T ρ f Ap
+ ρ f Ap C p , f v p et ⋅ ∇ t T = ∇ t ⋅ ( Ap k f ∇ t T ) + f D

PT
1
244 ρ f Ap C p , f v p v 2p + Q + Qwall (22)
∂t 2 dh

245 where ρ f is the density of the fluid, v p is the velocity of the fluid in a pipe, p is the pressure of the fluid, et

RI
246 is the unit tangent vector to the pipe axis, C p , f is the fluid heat capacity at a constant pressure, Ap is the pipe

SC
247 cross-sectional area, k f is the thermal conductivity of the fluid, f D is the Darcy friction factor, d h is the

248 mean hydraulic diameter, R is the volume force term, Q is the general heat source, and Qwall is the

U
249 external heat exchange through the pipe wall.
AN
250 The Darcy friction factor was obtained using the empirical equation proposed by Churchill [29], which

251 can be used in all fluid flow regimes:

M

252

1/12
 8 12 
253 f D = 8   + (c A + cB ) −1.5 
D

(23)
 Re  
TE

( )
+ 0.27 ( e / d h )  , cB = ( 37530 / Re ) , Re is the Reynolds number, e is
16
where c A =  −2.457 ln ( 7 / Re )
0.9 16
254
 

255 the surface roughness, and d h is the hydraulic diameter.

EP

256 To evaluate the Nusselt number for internal turbulent forced convection, the empirical equation

257 proposed by Gnielinski [30] was applied as follows:

C

258
AC

 Nulam = 3.66 

259 Nu = max  ( f D / 8)( Re− 1000) Pr  (24)
Nuturb =
 1 + 12.7 ( f D / 8) ( Pr 2/3 − 1) 
1/ 2


260 where Pr is the Prandtl number.

261 The heat transfer coefficient on the inside of the pipe and the external heat exchange through the pipe

262 wall are given as follows:

10
 ACCEPTED MANUSCRIPT

kf
263 hint = Nu (25)
dh

264 Qwall = ( ht Z )eff (Text − T ) (26)

2π
265 where ( ht Z )eff = , rin is the inner radius of the pipe, rout is the outer radius of
1 / rin hint + ( ln ( rout / rin ) / k wall , pipe )

PT
266 the pipe, hint is the heat transfer coefficient on the inside of the pipe, and k wall is the thermal conductivity of

267 the pipe wall.

RI
268 Fig. 6 shows a three-dimensional model for the spiral coil HGHE with a domain of 10 m (width) ×

269 24 m (length) × 6 m (height). The spiral coil-type HGHE with pitch size (0.05 m) and coil radius (0.15 m) was

SC
270 located 2.0 m below the ground surface. The material of the spiral coil-type HGHE was assumed to be a

271 polyethylene circular pipe with an outer diameter of 0.02 m, thickness of 0.002 m, thermal conductivity of 0.38

272

U
W / m ⋅ K and surface roughness of 0.00015 mm. A total of 293,794 tetrahedral elements were used to obtain
AN
273 an accurate solution. The minimum element size was chosen as smaller than the pitch size of the coil (<0.05m)

274 to capture the change in the temperature of the ground at any location. To reduce the computational cost of the
M

275 three-dimensional model, a dense mesh was applied in the vicinity of the spiral coil-type HGHE, and a coarse

276 mesh was used in areas far from the spiral coil-type HGHE. Based on the climatic data measured from the 1st of
D

277 January to the 31st of December in 2015 in Daejeon City (36°22.17' N, 127°56.17') by the Korea
TE

278 Meteorological Administration, the initial and boundary conditions for the ground temperature were applied

279 using Eq. (17).

EP

280

281 4. Experimental Investigation Procedure

C

282 As shown in Fig. 7, the spiral coil-type HGHE was installed on the campus of the Korea Advanced
AC

283 Institute of Science and Technology (KAIST), Korea, in 2016. The spiral coil-type HGHE, made of

284 polyethylene circular pipe with an outer diameter of 0.02 m and 0.002 m thick, with a thermal conductivity of

285 0.38 W / m ⋅ K , was installed 2.0 m below the ground surface. The dimensions of the radius of the coil and

286 pitch size for the spiral coil-type HGHE are 0.15 m and 0.05 m, respectively. Fig. 8 shows the schematic

287 diagram of the on-site TRT. Many researchers have employed TRTs to validate mathematical models because of

288 its simple test control and reliability [20, 21, 31]. Since TRTs are generally used to measure the ground thermal

289 properties by injecting constant heat into the ground heat exchanger, the ground temperature near GHE keeps

11
 ACCEPTED MANUSCRIPT

290 increasing with time. This heat transfer mechanism corresponds to that of space cooling in the real application.

291 The TRT equipment includes a heater, a water bath, a pump, a flow meter, an ampere meter, and temperature

292 sensors. A total of five Pt100 resistance temperature detectors (RTDs) were used to measure the inlet fluid

293 temperature, the outlet fluid temperature, and the ground temperature at the locations corresponding to 0 cm, 15

294 cm, and 35 cm from the central line of the spiral coil-type HGHE. All of the Pt100 RTDs have permissible

PT
295 deviations of ± 0.3 K in compliance with the International Electrotechnical Commission Standards (IEC 751)

296 and Korea Standards (KS C1603). The water flow rate was measured, using the instrument SITRANS F M

RI
297 MAG 6000, with a measurement accuracy of ± 0.2%. The measured data were collected every 30 seconds

298 using the data acquisition instrument 34972A LXI. The TRTs were conducted on November 7th, 2016, and

SC
299 continued for 23 hours.

300

U
301 5. Comparison and Discussion
AN
302

303 5.1 Comparison of the numerical simulation and on-site experiment

M

304 The inlet fluid temperature and inlet mass flow of the spiral coil-type HGHE monitored during the on-

305 site experiment are given in Fig. 9. These measured data were used as the Dirichlet and Neumann boundary
D

306 conditions at the inlet point of the spiral coil-type HGHE. Table 1 lists the physical properties used in the
TE

307 numerical simulation. The properties of water, such as thermal conductivity, specific heat, and density, changed

308 with the temperature [32]. Because the soil properties at shallow depth are sensitive to the water content,
EP

309 porosity, and water flow, which are affected by the infiltration, freezing, compaction, etc. [33, 34, 35, 36], it is

310 difficult to obtain reliable soil thermal properties. Therefore, the soil properties were obtained using the back-
C

311 calculation method through comparison of the ground temperature at the monitoring point in between for the

312
AC

numerical simulation and the on-site experiment. Table 2 shows the thermal properties of the soil at shallow

313 depths for various sites in Korea. These values were measured using the KD2 Pro instrument with a

314 measurement accuracy of 10% for all thermal properties. The ranges of thermal conductivity, thermal diffusivity,

315 and volumetric heat capacity were from 0.68 W / m ⋅ K to 1.39 W / m ⋅ K , from 0.27 mm 2 / sec to 0.54

mm 2 / sec , and from 1725000 J / m ⋅ kg to 4069000 J / m ⋅ kg , respectively. These results are also in good
3 3
316

317 accordance with the study of Sohn and Choi [37], who evaluated the results of the thermal diffusivity of soils

318 used as backfilling material for the HGHE. Hence, it appears reasonable to use the soil properties obtained by
12
 ACCEPTED MANUSCRIPT

319 back-calculation. Equal to the data acquisition time in the on-site experiment, the time step for the numerical

320 simulation was set to 30 seconds. Fig. 10 shows the ground temperature measured using RTDs. The ground

321 temperature increased with the time elapsed as the heat injected into the spiral coil-type HGHE was released to

322 the surrounding ground. Although RTD 1 was originally placed at the location corresponding to 0 cm from the

323 central line of the spiral coil-type HGHE, the location of RTD 1 moved between the pipes during soil backfilling

PT
324 (Fig. 11). Consequently, the rapid rise of the ground temperature was detected at RTD 1, due to the thermal

325 accumulation between the pipes. Compared with RTD 2, the slow increase in ground temperature was detected

RI
326 because RTD 3 was located farther from the spiral coil-type HGHE. The initial operating of TRTs for

327 equipment checking could exert a slight effect on variation in ground temperature near the spiral coil-type

SC
328 HGHE. To minimize this effect, the farthest position, which corresponds to a location 35 cm from the central

329 line of the spiral coil-type HGHE, was selected as the monitoring point of the ground temperature.

U
330 Fig. 12 illustrates the ground temperatures measured at the on-site experiment, the temperatures
AN
331 simulated with the numerical model, and their absolute errors. The results show reasonable agreement between

332 the on-site experiment and the numerical model figures. The value of maximum absolute error, mean absolute
M

333 error, and mean absolute percentage error was 0.14 °C , 0.039 °C , and 0.16 %, respectively. After 15 hours of

334 elapsed time, the absolute error increased constantly and showed a maximum value at the end of the test. These
D

335 errors occurred as the rate of increase in ground temperature slowed because rainwater infiltration affected the
TE

336 release of heat energy [21]. According to the meteorological records, the rain began at 23:00 on November 7th

337 and stopped at 05:00 on November 8th, and the precipitation totals on November 7th and 8th were 2.7 mm and 5.0

338
EP

mm, respectively.

339

340 5.2 Comparison of the modified mathematical model and on-site experiment
C

341 Fig. 13 shows the injection heat used to predict the ground temperature with the proposed mathematical
AC

342 model. The injection heat was calculated by multiplying the mass flow, the specific heat, and the temperature

343 difference between the inlet and outlet fluid. The calculated injection heat shows a high value of approximately

344 700 W / m throughout the initial 150 seconds because a fluid temperature difference occurred between the

345 water bath and the spiral coil-type horizontal ground heat exchanger (HGHE) prior to the fluid circulation. After

346 150 seconds, the injection heat decreased steadily to the end of the experiment. In general, the heat exchange

347 rate decreased over time as the temperature in the ground surrounding the spiral coil-type HGHE increased due

13
 ACCEPTED MANUSCRIPT

348 to the circulation fluid [38, 39]. Considering the computational effort and accuracy of the proposed

349 mathematical model, step-wise injection heat, which was used in the proposed mathematical model, was applied.

350 After the total elapsed time was divided into 277 steps in 300-second intervals, the average injection heat in

351 each step was considered as the step-wise injection heat. The computed ground temperature increases in each

352 step were accumulated to obtain the ground temperature profile over the entire period. Fig. 14 illustrates the

PT
353 ground temperatures measured from the on-site experiment and calculated using the proposed mathematical

354 model and their absolute errors. Similar to the results of the numerical solution, the prediction from the proposed

RI
355 mathematical model showed good agreement with the results of the on-site experiment. Compared with the

356 results of the numerical solution, the results of mathematical model showed a somewhat higher mean absolute

SC
357 error, i.e., 0.0743 °C . The mean absolute percentage error was also 0.30 %. Because of the material properties

358 of the spiral coil-type HGHE, the presence of a straight-line pipe for the outlet and the heat convection due to

U
359 fluid flow in the pipe were neglected in the mathematical model, and these discrepancies occurred in the ground
AN
360 temperature.

361
M

362 5.3 Comparison of the modified mathematical model and simplified numerical simulation

363 To validate the proposed mathematical model, the numerical model was simplified by treating the heat
D

364 transfer by fluid flow in the pipe as simple heat conduction with a line source that discharges heat constantly.
TE

365 Table 3 lists the input parameters and the values used in the simplified numerical simulation. As shown in Fig.

366 15, the difference in the increase of ground temperature for the different sides occurred because of the spiral

367
EP

shape. The positive side indicates the locations lying on the straight line connecting point (0.3, -6, -2) and point

368 (0.3, 6, -2), whereas the sign of the x-coordinates is reversed on the negative side. The ground temperature

369 increases computed by the simplified numerical model and calculated by the proposed mathematical model are
C

370 shown in Fig. 16. Within the elapsed time, the results of the mathematical model matched the results of the
AC

371 simplified numerical simulation almost perfectly. Although the ground temperature increased consistently with

372 elapsed time, the ground temperature at both ends of the spiral coil-type HGHE displayed lower values than

373 those around the other parts of the spiral coil-type HGHE. In addition, the highest value of the ground

374 temperature increase was observed at the center of the spiral coil-type HGHE. With an increase in the elapsed

375 time, the injected heat accumulated around the spiral coil-type HGHE, but it was easily released to the

376 surrounding ground at both ends. The proposed mathematical model for the spiral coil-type HGHE was also

14
 ACCEPTED MANUSCRIPT

377 compared with the multiple ring-coil source model, which was proposed fairly recently for the spiral coil-type

378 HGHE. Because a spiral coil-type HGHE with a length of 4.0 m, pitch of 0.5 m, and total injection heat of 400

379 W was considered, eight ring sources with an injection heat of 50 W per ring were used, as shown in Fig. 17.

380 As portrayed in the top view, each ring source was assumed to be located at the center of each pitch for the

381 spiral coil-type HGHE. The ground temperature increases were calculated on the locations lying on the straight

PT
382 line connecting point (-0.3, -6, -2) and point (-0.3, 6, -2). Fig. 18 compares the ground temperature increases

383 calculated by the proposed spiral coil source and those calculated on the multiple ring-coil source models.

RI
384 Because the multiple ring-coil source model simplified the structure of the spiral coil-type HGHE, the

385 discrepancy in the increases of temperature rose over time [21]. Compared to the results of the multiple ring-coil

SC
386 source model, the proposed spiral coil source model provided more accurate predictions. Even though the

387 validation was confined mainly to the heat release conditions of space cooling, the proposed mathematical

U
388 model can predict the ground temperature variation in the heat extraction conditions of space heating by
AN
389 employing the negative value of heat rate per unit length (Fig. 19).

390
M

391 6. Conclusions

392 Ground heat exchangers are the primary components of the ground source heat pump (GSHP) system
D

393 used for heating and cooling spaces. Despite the superior heat transfer performance of spiral coil-type horizontal
TE

394 ground heat exchangers (HGHEs), a reliable mathematical model has not yet been developed. In this paper, a

395 mathematical model for spiral coil-type HGHEs was proposed based on a modification of the mathematical
EP

396 model proposed by Park et al. [20]. Although the derivation processes of the two mathematical models are

397 similar, the effect of the different locations of the spiral coil source and the effect of the ground surface
C

398 temperature must be considered for the spiral coil-type HGHE. Therefore, based on Green’s function theory, the

399
AC

point sources were superimposed along the spiral coil structure in an infinite medium, and the infinite medium

400 was treated as a semi-infinite medium using the mirror image method. The prediction results obtained from the

401 proposed mathematical model were compared with and validated by the numerical simulation results from

402 COMSOL Multiphysics, and by the results of an on-site thermal response test. In addition, the thermal

403 properties of the soil at shallow depths for various sites in Korea were measured to validate the input parameters

404 obtained by back-calculation. Although a slight discrepancy was observed because of the simplification of the

405 ground and heat transfer mechanism, the proposed mathematical model for the spiral coil-type HGHE shows

15
 ACCEPTED MANUSCRIPT

406 good agreement with the results from the numerical simulation and on-site experiment. Moreover, compared

407 with the multiple ring-source model, the proposed mathematical model provides a more accurate prediction of

408 ground temperature as elapsing time progresses. However, vulnerability still exists due to the simplification of

409 the real spiral coil-type HGHE. For a more accurate mathematical model, the presence of a straight outlet pipe

410 and its heat exchange rate should be considered in future studies.

PT
411

412 ACKNOWLEDGEMENT

RI
413 This research was supported by the Regional Development Research Program (16RDRP-B076564-03)

414 of the Ministry of Land, Infrastructure and Transport of the Korean Government.

SC
415

416 REFERENCES

U
417 [1] Johnston IW, Narsilio GA, Colls S. Emerging geothermal energy technologies. KSCE Journal of Civil
AN
418 Engineering 2011;15(4):643-53. http://dx.doi.org/10.1007/s12205-011-0005-7.

419 [2] Unverdi M, Cerci Y. Performance analysis of Germencik Geothermal Power Plant. Energy
M

420 2013;52:192-200. http://doi.org/10.1016/j.energy.2012.12.052.

421 [3] Jalilinasrabada S, Itoi R, Valdimarsson P, Saevarsdottir G, Fujji H. Flash cycle optimization of Sabalan
D

422 geothermal power plant employing exergy concept. Geothermics 2012;43:75-82. http://
TE

423 doi.org/10.1016/j.geothermics.2012.02.003.

424 [4] Ahmadi MH, Mehrpooya M, Pourfayaz F. Exergoeconomic analysis and multi objective optimization
EP

425 of performance of a Carbon dioxide power cycle driven by geothermal energy with liquefied natural

426 gas as its heat sink. Energy Conversion and Management 2016;119:422-34.

427
C

http://doi.org/10.1016/j.enconman.2016.04.062

428 [5] Ahmadi MH, Mehrpooya M, Pourfayaz F. Thermodynamic and exergy analysis and optimization of a
AC

429 transcritical CO2 power cycle driven by geothermal energy with liquefied natural gas as its heat sink.

430 Applied Thermal Engineering 2016;109(25):640-52.

431 http://doi.org/10.1016/j.applthermaleng.2016.08.141.

432 [6] Lu Q, Narsilio GA, Aditya GR, Johnston IW. Economic analysis of vertical ground source heat pump

433 systems in Melbourne. Energy 2017;125:107-17. http://dx.doi.org/10.1016/j.energy.2017.02.082.

434 [7] Zhang W, Yang H, Fang L, Cui P, Fang Z. Study on heat transfer of pile foundation ground heat

16
 ACCEPTED MANUSCRIPT

435 exchanger with three-dimensional groundwater seepage. International Journal of Heat and Mass

436 Transfer 2017;105:58-66. http://dx.doi.org/10.1016/j.ijheatmasstransfer.2016.09.066.

437 [8] Florides G, Theofanous E, Iosif-Stylianou I, Tassou S, Christodoulides P, Zomeni Z, Tsiolakis E,

438 Kalogirou S, Messaritis V, Pouloupatis P, Panayiotou G. Modeling and assessment of the efficiency of

439 horizontal and vertical ground heat exchangers. Energy 2013;58:655-63.

PT
440 http://dx.doi.org/10.1016/j.energy.2013.05.053.

441 [9] Lous ML, Larroque F, Dupuy A, Moignard A. Thermal performance of a deep borehole heat exchanger:

RI
442 Insights from a synthetic coupled heat and flow model. Geothermics 2015;57:157-72.

443 http://dx.doi.org/10.1016/j.geothermics.2015.06.014.

SC
444 [10] Lee J. Current status of ground source heat pumps in Korea. Renewable and Sustainable Energy

445 Reviews 2009;13:1560-8. http://dx.doi.org/10.1016/j.rser.2008.10.005.

U
446 [11] Kwon KS, Lee JY, Mok JK. Update of current status on ground source heat pumps in Korea (2008-
AN
447 2011). Journal of the Geological Society of Korea 2012;48(3):193-9.

448 [12] Congedo PM, Colangelo G, Starace G. CFD simulations of horizontal ground heat exchangers: A
M

449 compassion among different configurations. Applied Thermal Engineering 2012;33-34:24-32.

450 http://dx.doi.org/10.1016/j.applthermaleng.2011.09.005.
D

451 [13] Kim MJ, Lee SR, Yoon S, Go GH. Thermal performance evaluation and parametric study of a
TE

452 horizontal ground heat exchanger. Geothermics 2016;60:134-43. http://dx.doi.org/10.1016/

453 j.geothermics.2015.12.009.

454
EP

[14] Han C, Ellett KM, Naylor S, Yu X. Influence of local geological data on the performance of horizontal

455 ground-coupled heat pump system integrated with building thermal loads. Renewable Energy

456 2017;113:1046-55. http://dx.doi.org/10.1016/j.renene.2017.06.025.

C

457 [15] Selamat S, Miyara A, Kariya K. Numerical study of horizontal ground heat exchangers for design
AC

458 optimization. Renewable Energy 2016;95:561-73. http://dx.doi.org/10.1016/j.renene.2016.04.042.

459 [16] Li C, Mao J, Zhang H, Xing Z, Li Y, Zhou J. Numerical simulation of horizontal spiral-coil ground

460 source heat pump system: Sensitivity analysis and operation characteristics. Applied Thermal

461 Engineering 2017;110:424-35. http://dx.doi.org/10.1016/j.applthermaleng.2016.08.134.

462 [17] Man Y, Yang H, Diao N, Liu J, Fang, Z. A new model and analytical solutions for borehole and pile

463 ground heat exchangers. International Journal of Heat and Mass Transfer 2010;53(13-14):2593-601.

17
 ACCEPTED MANUSCRIPT

464 http://dx.doi.org/10.1016/j.ijheatmasstransfer.2010.03.001.

465 [18] Cui P, Li X, Fang Z. Heat transfer analysis of pile geothermal heat exchangers with spiral coils.

466 Applied Energy 2011;88(11):4113-9. http://dx.doi.org/10.1016/j.apenergy.2011.03.045.

467 [19] Man Y, Yang H, Diao N, Cui P, Lu L, Fang Z. Development of spiral heat source model for novel pile

468 ground heat exchangers. HVAC and R Research 2011;17(6):1075-88.

PT
469 http://dx.doi.org/10.1080/10789669.2011.610281.

470 [20] Park SK, Lee SR, Park HP, Yoon S, Chung J. Characteristics of an analytical solution for a spiral coil

RI
471 type ground heat exchanger. Computers and Geotechnics 2013;49:18-24.

472 http://dx.doi.org/10.1016/j.compgeo.2012.11.006.

SC
473 [21] Wang D, Lu L, Cui P. A new analytical solution for horizontal geothermal heat exchangers with vertical

474 spiral coils. International Journal of Heat and Mass Transfer 2016;100:111-20.

U
475 http://dx.doi.org/10.1016/j.ijheatmasstransfer.2016.04.001.
AN
476 [22] Duffy DG. Green’s Functions with Applications. Second ed. New York: CRC Press; 2015.

477 [23] Carslaw HS, Jaeger JC. Conduction of Heat in Solids. Second ed. New York: Oxford University Press;
M

478 2011.

479 [24] Bose JE, Parker JD, McQuiston FC. Design/Data Manual for Closed-Loop Ground-Coupled Heat
D

480 Pump Systems. Atlanta, GA: ASHRAE; 1985.

TE

481 [25] Fujii H, Nishi K, Komaniwa Y, Chou N. Numerical modeling of slinky-coil horizontal ground heat

482 exchangers. Geothermics 2012;41:55-62. http://dx.doi.org/10.1016/j.geothermics.2011.09.002.

483
EP

[26] Naylor S, Ellett KM, Gustin AR. Spatiotemporal variability of ground thermal properties in glacial

484 sediments and implications for horizontal ground heat exchanger design. Renewable Energy

485 2015;81:21-30. http://dx.doi.org/10.1016/j.renene.2015.03.006.

C

486 [27] Xing L. Estimations of Undisturbed Ground Temperatures Using Numerical and Analytical Modeling.
AC

487 PhD thesis. Oklahoma State University; 2014.

488 [28] Kusuda T, Achenbach RS. Earth temperatures and thermal diffusivity at selected stations in the United

489 States. ASHRAE Transactions 1965;71(1):61-74.

490 [29] Churchill SW. Friction factor equation spans all fluid-flow regimes. Chemical Engineering

491 1977;84(24):91-2.

492 [30] Gnielinski V. New Equations for Heat and Mass Transfer in Turbulent Pipe and Channel Flow.

18
 ACCEPTED MANUSCRIPT

493 International Chemical Engineering 1976;16:359-68.

494 [31] Kim MJ, Lee SR, Yoon S, Jeon JS. An applicable design method for horizontal spiral-coil-type ground

495 heat exchangers. Geothermics 2018;72:338-47. http://dx.doi.org/10.1016/j.geothermics.2017.12.010.

496 [32] Comsol Inc. Comsol Multiphysics Reference Manual Version. COMSOL 5.1., USA;2015.

497 [33] Abu-Hamdeh NH, Reeder RC. Soil thermal conductivity effects of density, moisture, salt concentration,

PT
498 and organic matter. Soil Science Society of America Journal 2000;64(4):1285-90.

499 http://dx.doi.org/10.2136/sssaj2000.6441285x.

RI
500 [34] Chalhoub M, Bernier M, Coquet Y, Philippe M. A simple heat and moisture transfer model to predict

501 ground temperature for shallow ground heat exchangers. Renewable Energy 2017;103:295-307.

SC
502 http://dx.doi.org/10.1016/j.renene.2016.11.027.

503 [35] Balland V, Arp PA. Modeling soil thermal conductivities over a wide range of conditions. Journal of

U
504 Environmental Engineering and Science 2005;4(6):549-58. http://dx.doi.org/10.1139/S05-007.
AN
505 [36] Lipiec J, Usowicz B, Ferrero A. Impact of soil compaction and wetness on thermal properties of

506 sloping vineyard soil. International Journal of Heat and Mass Transfer 2007;50(19-20):3837-47.
M

507 http://dx.doi.org/10.1016/j.ijheatmasstransfer.2007.02.008.

508 [37] Sohn BH, Choi HS. Thermal Diffusivity Evaluation of Backfilling Materials for Horizontal Ground
D

509 Heat Exchanger Using Single-Probe Method. Korean Journal of Air-Conditioning and Refrigeration
TE

510 Engineering 2011;23(5):356-64. http://dx.doi.org/10.6110/KJACR.2011.23.5.356.

511 [38] Naili N, Hazami M, Attar I, Farhat A. In-field performance analysis of ground source cooling system

512
EP

with horizontal ground heat exchanger in Tunisia. Energy 2013;61:319-31.

513 http://dx.doi.org/10.1016/j.energy.2013.08.054.
C

514 [39] Xiong Z, Fisher DE, Spitler JD. Development and validation of a Slinky™ ground heat exchanger
AC

515 model. Applied Energy 2015;141:57-69. http://dx.doi.org/10.1016/j.apenergy.2014.11.058.

516

517

518

519

520

19
 ACCEPTED MANUSCRIPT

521

522

523

524

525

PT
526

RI
527

528

SC
529 Nomenclature
530 hS spiral coil length ( m )
531 ql heat rate per unit length ( W / m )

U
532 d buried depth of coils ( m )
AN
533 pS pitch size ( m )
534 ro spiral coil radius ( m )
535 rin inner radius of pipe ( m )
M

536 rout outer radius of pipe ( m )

537 t time ( sec )
538 T temperature ( K )
D

539 u , u ' , u * ' position vector

TE

540 x , y , z Cartesian coordinate ( m )

541 x ' , y ' , z ' integral variable
542 Cp specific heat ( J / kg ⋅ K )
EP

543 ρ density ( kg / m3 )
544 k thermal conductivity ( W / m ⋅ K )
545 k wall thermal conductivity of pipe wall ( W / m ⋅ K )
C

546 α thermal diffusivity ( mm 2 / sec )

AC

547 As surface amplitude ( °C )

548 A p pipe cross section area
549 v velocity field ( m / sec )
550 p pressure of fluid ( Pa )
551 dh mean hydraulic diameter ( m )
552 e surface roughness ( m )
553 R volume force term
554 Q general heat source
555 S boundary of domain

20
 ACCEPTED MANUSCRIPT

556 Qwall external heat exchange through the pipe wall

557 Nu Nusselt number
558 Re Reynolds number
559 Pr Prandtl number
560 et unit tangent vector to the pipe axis
561 f D Darcy friction factor
562 hint heat transfer coefficient on the inside of pipe
563

PT
564 Greek symbols
565 Ω domain
566 θ temperature variable ( K )

RI
567 θs annual ground temperature ( °C )
568 ϕ phase change of the annual soil temperature cycle
τ

SC
569 time ( sec )
570
571 Subscripts

U
572 avg average

573
AN
o original

574 im image

575
M

g ground

576 f fluid
D

577 p pipe
TE

578 lam laminar flow

579 turb turbulent flow

EP

580

581
C

582
AC

583

584

585

586

587

588

21
 ACCEPTED MANUSCRIPT

589

590

591

592

593

PT
594

Table

RI
595
596
597

SC
598 Table 1. Input parameters for the three-dimensional numerical simulation
Parameters Unit Values
Density, ρ g kg / m 3
1663

Soil
U
Specific heat, Cp, g J / kg ⋅ K 1548
AN
Thermal conductivity, kg W / m⋅K 1.21
Thermal diffusivity, α 2
mm / sec 0.470

Spiral coil-
Pitch size, pS m 0.05
M

type Radius of coil, ro m 0.15

HGHE Length, hS m 4.0
Install depth, d m 2.0
D

599
600
TE

601
602
EP

603
604 Table 2. Thermal properties of the soil at shallow depth for various sites in Korea
Thermal Thermal Volumetric
Water content
C

Site Longitude Latitude conductivity diffusivity heat capacity

(%)
(W / m ⋅ K ) ( mm 2 / sec ) ( J / m3 ⋅ kg )
AC

A1 126°56.547' 35°9.299'N 0.94 0.54 1725000 9.25

A2 126°42.876' 35°3.131'N 1.39 0.46 3253000 21.84
A3 126°24.280' 34°48.840'N 0.83 0.47 1770000 33.18
A4 126°44.763' 34°18.422'N 1.26 0.41 3097000 26.00
A5 126°53.460' 34°40.175'N 1.27 0.36 3511000 18.11
A6 127°19.748' 34°47.907'N 0.98 0.41 2371000 39.23
A7 127°37.267' 34°48.871'N 1.36 0.33 4069000 29.16
A8 127°15.062' 35°4.358'N 0.68 0.27 2542000 34.40

22
 ACCEPTED MANUSCRIPT

A9 127°22.721' 35°23.929'N 1.00 0.31 3174000 12.61

A10 127°20.562' 35°30.935'N 0.91 0.45 2032000 24.36
A11 127°31.442' 35°39.074'N 1.08 0.35 3069000 18.96
A12 127°10.357' 35°50.463'N 1.22 0.45 2686000 20.26
605
606

PT
607
608 Table 3. Input parameters for the simplified numerical simulation
609

RI
Parameters Unit Values
Density, ρ g kg / m 3
1500

SC
Specific heat, Cp, g J / kg ⋅ K 1613.33
Soil
Thermal conductivity, kg W / m⋅K 1.21
Thermal diffusivity, α 2
mm / sec 0.5

U
Spiral coil-
Pitch size, pS m 0.5
type Radius of coil, ro m 0.15
AN
HGHE Length, hS m 4.0
Injection heat, ql W /m 100
Elapsed time, t hours 6.25, 12.5, 62.5, 125, 625
M

610
611
D

612
TE

613

614
EP

615

616
C

617
AC

618

619

620

621

622

623

23
 ACCEPTED MANUSCRIPT

624

625

626

627

628 Figures

PT
629

RI
U SC
AN
M
D
TE
EP

630
631
C

632 Fig. 1. Concept of the ground source heat pump (GSHP) system for space heating and cooling
AC

633

634

635

636

637

638

639

24
 ACCEPTED MANUSCRIPT

640

641

642

643

644

PT
RI
U SC
AN
M
D

645
TE

646 Fig. 2. Schematic view of the various horizontal ground heat exchangers (HGHEs)
647

648
EP

649
C

650

651
AC

652

653

25
 ACCEPTED MANUSCRIPT

PT
RI
SC
654
655

U
656 Fig. 3. Schematic view of the spiral coil source for the VGHE
657
AN
658
659
660
M

661
662
D

663
664
TE

665
666
EP
C
AC

26
 ACCEPTED MANUSCRIPT

PT
RI
U SC
667
668 Fig. 4. Schematic view of the spiral coil source for the HGHE
AN
669
670
671
M

672
673
674
D

675
TE

676
677
678
EP

679
680
681
C

682
AC

683
684
685
686
687
688
689
690
691

27
 ACCEPTED MANUSCRIPT

692
693

PT
RI
694

SC
695 (a) Adiabatic surface condition (b) Isothermal ground surface condition

696 Fig. 5. Mirror image method

U
697
698
AN
699
700
701
M

702
703
D

704
705
TE

706
707
708
EP

709
710
C

711
712
AC

713
714
715
716
717
718
719
720

28
 ACCEPTED MANUSCRIPT

721

PT
RI
U SC
AN
722
723 (a) Three-dimensional numerical model
M

724
D
TE
EP
C
AC

725
726
727 (b) Mesh configuration
728 Fig. 6. Schematic view of the numerical simulation
729

29
 ACCEPTED MANUSCRIPT

730
731
732
733

PT
RI
U SC
AN
734
735
M

736 Fig. 7. Spiral coil-type horizontal ground heat exchanger

737
D

738
739
TE

740
741
EP

742
743
744
C

745
746
AC

747
748
749
750
751
752
753
754

30
 ACCEPTED MANUSCRIPT

PT
RI
SC
755

U
756
757 Fig. 8. Schematic diagram of the on-site thermal response test
AN
758
759
M

760
761
762
D

763
764
TE

765
766
EP

767
768
769
C

770
771
AC

772
773
774
775
776
777
778

31
 ACCEPTED MANUSCRIPT

PT
RI
U SC
779
780 (a) Volumetric inlet flow rate
AN
M
D
TE
EP
C
AC

781
782 (b) Inlet fluid temperature

783 Fig. 9. Measured data during the on-site thermal response test
784
785

32
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
786
787
788 Fig. 10. Measured ground temperature using RTDs
M

789
790
D

791
TE
EP
C
AC

792
793
794 Fig. 11. Locations of RTDs after the on-site thermal response test
795

33
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
796
797 Fig. 12. Results of the numerical simulation and on-site experiment
798
M

799
800
D

801
802
TE

803
804
805
EP

806
807
808
C

809
AC

810
811
812
813
814
815
816
817
818
34
 ACCEPTED MANUSCRIPT

819

PT
RI
U SC
AN
820
821 Fig. 13. Calculated injection heat and step-wise injection heat
M

822
823
D

824
825
TE

826
827
828
EP

829
830
C

831
832
AC

833
834
835
836
837
838
839
840

35
 ACCEPTED MANUSCRIPT

841
842

PT
RI
U SC
AN
843
M

844 Fig. 14. Results of the mathematical model and on-site experiment
845
846
D

847
TE

848
849
850
EP

851
852
853
C

854
855
AC

856
857
858
859
860
861
862
863

36
 ACCEPTED MANUSCRIPT

864

PT
RI
U SC
AN
865
866 Fig. 15. Comparison of the temperature distributions on the positive and negative side
M

867
868
D

869
870
TE

871
872
873
EP

874
875
C

876
877
AC

878
879
880
881
882
883
884
885

37
 ACCEPTED MANUSCRIPT

886

PT
RI
U SC
AN
887
888
889 Fig. 16. Results of the proposed mathematical model and numerical simulation
M

890 for long-term operation (solid line: mathematical model, circle: numerical solution)
891
D

892
893
TE

894
895
896
EP

897
898
899
C

900
AC

901
902
903
904
905
906
907
908

38
 ACCEPTED MANUSCRIPT

PT
RI
909

SC
910 (a) Illustration of two source models for spiral coil-type HGHEs

U
AN
911
912 (b) Top view of two source models for spiral coil-type HGHEs
M

913 Fig. 17. Proposed spiral coil source and multiple ring-coil source
914
D

915
TE

916
917
918
EP

919
920
921
C

922
AC

923
924
925
926
927
928
929
930
931

39
 ACCEPTED MANUSCRIPT

PT
RI
U SC
AN
932
M

933 Fig. 18. Comparison of results from the proposed mathematical model and multiple ring-coil
934 source model (solid line: proposed mathematical model, dotted line: multiple ring-coil source
D

935 model)
936
TE

937
938
939
EP

940
941
C

942
943
AC

944
945
946
947
948
949
950
951

40
 ACCEPTED MANUSCRIPT

952

PT
RI
U SC
AN
953
954 Fig. 19. Ground temperature variation in a heat extraction condition
M

955 (solid lines: mathematical model results, circles: numerical solution solutions)
956
D
TE
C EP
AC

41
 ACCEPTED MANUSCRIPT

Highlights

Mathematical model for spiral coil-type HGHEs was presented.

Green s function, mirror image and superposition method were used.

For validation, the on-site experiment and numerical simulation were employed.

PT

The proposed mathematical model provides more accurate results.

RI
U SC
AN
M
D
TE
C EP
AC