You are on page 1of 8

The importance of axial effects for borehole design of geothermal

heat-pump systems
D. Marcotte
a, b, c,
*
, P. Pasquier
a
, F. Sheriff
b
, M. Bernier
c
a
Golder Associates, 9200 lAcadie, Montreal, (Qc), H4N 2T2 Canada
b
CANMET Energy Technology Centre-Varennes, 1615 Lionel-Boulet Blvd., P.O. Box 4800, Varennes, (QC), J3X 1S6 Canada
c
Departement des genies civil, Geologique et des mines, E

cole Polytechnique de Montreal, C.P. 6079 Succ. Centre-ville, Montreal, (Qc), H3C 3A7 Canada
a r t i c l e i n f o
Article history:
Received 13 May 2008
Accepted 18 September 2009
Available online 23 October 2009
Keywords:
Innite line source
Finite line source
Ground loop heat exchangers
Hybrid systems
Underground water freezing
a b s t r a c t
This paper studies the effects of axial heat conduction in boreholes used in geothermal heat pump
systems. The axial effects are examined by comparing the results obtained using the nite and innite
line source methods. Using various practical design problems, it is shown that axial effects are relatively
important. Unsurprisingly, short boreholes and unbalanced yearly ground loads lead to stronger axial
effects. In one example considered, it is shown that the borehole length is 15% shorter when axial
conduction effects are considered. In another example dealing with underground water freezing, the
amount of energy that has to be removed to freeze the ground is three times higher when axial effects
are considered.
2009 Elsevier Ltd. All rights reserved.
1. Introduction
Geothermal systems using ground-coupled closed-loop heat
exchangers (GLHE) are becoming increasingly popular due to
growing energy costs. Such a system is presented in Fig. 1.
The operation of the system is relatively simple: a pump circu-
lates a heat transfer uid in a closed circuit fromthe GLHE to a heat
pump (or a series of heat pumps). Typically, GLHE consists
of boreholes that are 100150 m deep and have a diameter of
1015 cm. The number of boreholes in the boreeld can range from
one, for a residence, to several dozens, in commercial applications.
Furthermore, several borehole congurations (square, rectangular,
L-shaped) are possible. Typically, a borehole consists of two pipes
forming a U-tube (Fig. 1). The volume between these pipes and the
borehole wall is usually lled with grout to enhance heat transfer
from the uid to the ground. In some situations it is advantageous
to design so-called hybrid systems in which a supplementary heat
rejecter or extractor is used at peak conditions to reduce the length
of the ground heat exchanger.
Given the relatively high cost of GLHE, it is important to design
them properly. Among the number of parameters that can be
varied, the length and conguration of the boreeld are important.
There are basically two ways to design a boreeld. The rst method
involves using successive thermal pulses (typically 10-years 1
month 6 h) to determine the length based on a given congura-
tion and minimum/maximum heat pump entering water temper-
ature [8,3]. There are design software programs that perform these
calculations. Some use the concept of the g-functions developed by
Eskilson [5]. The g-functions are derived from a numerical model
that, by construction, includes the axial effects. The other approach
is to perform hourly simulation. This last approach is essential for
design of hybrid systems in which supplemental heat rejection/
injection is used. There are several software packages that can
perform hourly borehole simulations. For example, TRNSYS [9] and
EnergyPlus [4] use the DST [6] and the short-time step model [5],
respectively. Even though these packages account for axial effects,
they necessitate a high level of expertise. Furthermore, it is not
easily possible to obtain ground temperature distributions like the
ones shown later in this paper. In this paper hourly simulations are
performed using the so-called nite and innite line source
approximations where the borehole is approximated by a line with
a constant heat transfer rate per unit length. These approximations
present, in a convenient analytical form, the solution to the tran-
sient 2-D heat conduction problem. Despite their advantages,
hourly simulations based on the line source approximation are
* Corresponding author. De partement des ge nies civil, ge ologique et des mines,
E

cole Polytechnique de Montre al, C.P. 6079 Succ. Centre-ville, Montre al, (Qc),
H3C 3A7 Canada. Tel.: 1 514 340 4711x4620; fax: 1 514 340 3970.
E-mail address: denis.marcotte@polymtl.ca (D. Marcotte).
Contents lists available at ScienceDirect
Renewable Energy
j ournal homepage: www. el sevi er. com/ l ocat e/ renene
0960-1481/$ see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.renene.2009.09.015
Renewable Energy 35 (2010) 763770
rarely used in routine design due to the perceived computational
burden.
The major difference between the nite and innite line source
lies in the treatment of axial conduction (at the bottom and top of
the borehole) which is only accounted for in the former. The
theoretical basis of the nite line source, although more involved
than for the innite line source, was rst established by Ingersoll
et al. [7]. It has been rediscovered recently by Zeng et al. [15] who
improved the model by imposing a constant temperature at the
ground surface. Lamarche and Beauchamp [11] have made a useful
contribution to speed up the computation of Zengs model. Finally,
Sheriff [13] extended Zengs model by permitting the borehole top
to be located at some distance below the ground surface. She also
did a detailed comparison of the nite and innite line source
responses, but did not examine the repercussion on boreeld
design.
At rst glance, the axial heat-diffusion is likely to decrease
(increase) the borehole wall temperature in cooling (heating)
modes respectively. Therefore, designing without considering
axial effects appears to provide a safety factor for the design. But,
is it really always the case? Moreover, are the borehole designs
incorporating axial effects signicantly different from those
neglecting it? Under which circumstances are we expected to
have signicant design differences? These are the main questions
we seek to answer. The main contribution of this research is to
describe, using synthetic case studies, the impact of considering
axial effects on the GLHE design. Our main nding is that for
many realistic circumstances the axial effects cannot be neglec-
ted. Therefore, design practices should be revised accordingly to
include the axial effects.
We rst review briey the theory for innite and nite line
source models. Then, we present three different design situations.
The rst two situations involve the sizing of geothermal systems
with and without the hybrid option, under three different hourly
ground load scenarios. The last design problem examines the
energy required and ground temperature evolution in the context
of ground freezing for environmental purposes.
2. Theoretical background
The basic building block of both innite and nite line source
models is the change in temperature felt at a given location and
time due to the effect of a constant point source releasing q
0
units of
heat per second [7]:
DTr; t
q
0
4pk
s
r
erfc
_
r
2

at
p
_
(1)
where erfc is the complementary error function, r the distance to
the point heat source, and a is the ground thermal diffusivity.
The line is then represented as a series of points equally spaced.
In the limit, when the distance between point sources goes to zero,
Fig. 1. Sketch of a GLHE system.
Nomenclature
a Thermal diffusivity (m
2
s
1
)
A, B, C, D Synthetic load model parameters (kW)
b r/H
C
s
Ground volumetric heat capacity (Jm
3
K
1
)
erfc (x) Complementary error function
(erfcx
1
2

p
p
_
N
x
e
t
2
dt
EWT Temperature of uid entering the heat pump (K or

C)
F
o
Fourier number, F
o
at/r
2
k
s
Volumetric ground thermal conductivity (Wm
1
K
1
)
H Borehole length (m)
HP Heat Pump
q
0
Radial heat transfer rate (W)
q Radial heat transfer rate per unit length (Wm
1
)
S Borehole spacing (m)
r Distance to borehole (m)
r
b
Borehole radius (m)
R
b
Borehole effective thermal resistance (KmW
1
)
t Time
DT (r, t) Ground temperature variation at time t and distance r
from the borehole (K or

C)
T
f
Fluid temperature (K or

C)
T
g
Undisturbed ground temperature (K or

C)
T
w
Temperature at borehole wall (K or

C)
u
H
2

at
p
x, y Spatial coordinates (m)
z Elevation (m)
D. Marcotte et al. / Renewable Energy 35 (2010) 763770 764
the combined effect felt at distance r fromthe source is obtained by
integration along the line.
2.1. Innite line source
In an innite medium, the line-integration gives the so-called
(innite) line source model [7]:
DTr; t
q
4pk
s
_
N
r
2
=4at
e
u
u
du (2)
2.2. Finite line source
In the case of a nite line source, the upper boundary is
considered at constant temperature, taken as the undisturbed
ground temperature [15]. This condition is represented by adding
a mirror image nite line source with the same load, but opposite
sign, as the real nite line. Then, integrating between the limits of
the real and image line, one obtains [15],[13]:
DTr; t; z
q
4pk
s
_
H
0
_
_
erfc
_
du
2

at
p
_
du

erfc
_
d
0
u
2

at
p
_
d
0
u
_
_
du (3)
where du

r
2
z u
2
_
and d
0
u

r
2
z u
2
_
, z is the
elevation of the point where the computation is done. The left part
of the integrand in Equation (3) represents the contribution by the
real nite line, the right part, the contribution of the image line.
Fig. 2 shows the vertical temperature prole obtained with
Equation (3) at radial distance r 2 m, after 200 days, and at
r 1 m, after 2000 days of heat injection. The corresponding
innite lines-source temperature is indicated as a reference. In this
example, the borehole is 50 mlong, the ground thermal parameters
are k
s
2.1 Wm
1
K
1
and C
s
2e06 Jm
3
K
1
. The ground is inti-
tially at 10
o
C. The applied load is 60 W per m for a total heating
power of 3000 W. As expected, the importance of axial effects and
the discrepancy between innite and nite models increases with
the Fourier number (at/r
2
4.54 and 181.4 for these two cases).
In hourly simulations, the uid temperature (T
f
in Fig. 1) is
required. This necessitates knowledge of the borehole thermal
resistance R
b
(i.e. from the uid to the borehole wall), and of the
borehole wall temperature (T
w
in Fig. 1) [2]. The average borehole
wall temperature it obtained by integrating Equation (3) along z.
However, this is computationally intensive due to the double
integration. Lamarche and Beauchamp [11] have shown, using an
appropriate change of variables, how to simplify Equation (3) to
a single integration. Accounting for small typos in [11] and [15] as
noted by Sheriff [13], the average temperature difference, between
a point located at distance r fromthe borehole and the undisturbed
ground temperature, is given by:
DTr; t
q
2pk
s
_
_
_
_
_
_

b
2
1
_
b
erfcuz

z
2
b
2
_ dz D
A

b
2
4
_

b
2
1
_
erfcuz

z
2
b
2
_ dz D
B
_
_
_
_
_
(4)
where b r/H, r is the radial distance from the borehole center,
u
H
2

at
p
and D
A
, and D
B
are given by:
D
A

b
2
1
_
erfc
_
u

b
2
1
_ _
b erfcub

_
e
u
2
b
2
1
e
u
2
b
2
u

p
p
_
and
D
B

b
2
1
_
erfc
_
u

b
2
1
_ _
0:5
_
b erfcub

b
2
4
_
erfc
_
u

b
2
4
_ __

_
e
u
2
_
b
2
1
_
0:5
_
e
u
2
b
2
e
u
2
_
b
2
4
_
_
u

p
p
_
10 12 14 16 18 20 22 24
0
10
20
30
40
50
60
Temperature (
o
C)
D
e
p
t
h

(
m
)
Vertical temperature profile


Infline, r=2, t=200 d
Fline, r=2, t=200 d
Fline average, r=2, t=200 d
Infline, r=1, t=2000 d
Fline, r=1, t=2000 d
Fline average, r=1, t=2000 d
Fig. 2. Vertical ground temperature prole at radial distances r 1 m and r 2 m after
respectively 2000 days and 200 days, F
o
(r 1, t 2000) 181.4 and F
o
(r 2,
t 200) 4.54. Constant heat injectionof 3000 W. Thermal parameters: k
s
2.1 Wm
1
K
1
,
C
s
2e06 Jm
3
K
1
.
0 1000 2000 3000 4000 5000
0
2
4
6
8
10
12
Days


T

(
o
C
)


Infinite
Finite
FEM
Fig. 3. Comparison of Finite and Innite line source model with nite element model
(FEM) for a 30 m borehole. Average temperature variation computed at 0.5 m from the
borehole axis, over the borehole length. Constant heat transfer rate of 1000 W.
Thermal parameters: k
s
2.1 Wm
1
K
1
, C
s
2e06 Jm
3
K
1
.
D. Marcotte et al. / Renewable Energy 35 (2010) 763770 765
The particular case r r
b
in Equation (4) gives the borehole wall
temperature.
2.3. Numerical validation
Fig. 3 compares the variation in temperature over time
computed with nite and innite line source to the numerical
results of a nite element model (FEM) constructed within
COMSOL

. The nite element model is 2-D with axial symmetry


around the borehole axis. The ground is represented by a 50 mlong
and 50 m radius cylinder. The borehole is represented by a 30 m
long and 0.075 m radius cylinder delivering 1000 W. The axis of
revolution is located at the borehole center and constitutes
a thermal insulation boundary whereas all external boundaries are
set to the undisturbed ground temperature. Over 6000 triangular
elements equipped with quadratic interpolating functions are used
to discretize the model. The agreement between the FEM model
and the nite line source is almost perfect, the maximum absolute
difference in temperature over the 5000 days period being only
0.019
o
C.
Fig. 4 compares the temperature obtained with the innite
and nite line source models, at r 1 m and r 0.075 m (a
typical value for r
b
), with the thermal parameters specied
above. A 1
o
C temperature difference between the innite and
nite models is obtained after 2.5 y and 2 y, at 1 m and 0.075 m
respectively. Note that the temperature reaches a plateau for the
nite line source model indicating that a steady-state condition
has been reached. In contrast, the innite line source model
exhibits a linear behavior.
Fig. 5 shows the ground temperature, computed at a distance of
1 m fromthe borehole, for increasing values of the borehole length.
As expected, the nite line source solution reaches the innite line
source solution for long boreholes.
0.001 0.01 0.1 1 10 100 1000
10
20
30
40
50
60
70
80
r
=
0
.
0
7
5

m
r
=
1

m
Ground temperature
Time (y)
T
e
m
p
e
r
a
t
u
r
e

(
o
C
)
Fig. 4. Comparison of Finite (solid) and Innite (broken) line source model, computed
at distance 1 m and 0.075 m from the borehole. Constant heat transfer rate per unit
length of 100 W/m. Thermal parameters: k
s
2.1 Wm
1
K
1
, C
s
2e06 Jm
3
K
1
.
0 100 200 300 400 500 600 700 800 900 1000
12
12.5
13
13.5
Borehole length (m)
T
e
m
p
e
r
a
t
u
r
e

(
o
C
)
Average temperature vs borehole length


Infinite linesource
Finite linesource
Fig. 5. Innite vs nite line source average temperature along a vertical prole. The load
is 20 W/m, thermal parameters: k
s
2.1 Wm
1
K
1
, C
s
2e06 Jm
3
K
1
. Temperature
computed after one year at r 1 m from the borehole.
1 2 3 4 5 6
100
0
100
Cooling (+) Heating () load
Time (h)
L
o
a
d

(
k
w
)
1 2 3 4 5 6
200
100
0
100
Load decomposition
L
o
a
d

(
k
w
)
Fig. 6. Principle of temporal superposition for variable loads.
0 5 10 15 20 25 30 35
3
3.5
4
4.5
5
5.5
6
6.5
7
7.5
COP vs EWT
EWT
C
O
P


Cooling
Heating
Fig. 7. COP as a function of EWT.
D. Marcotte et al. / Renewable Energy 35 (2010) 763770 766
3. Design of complete geothermal systems
In this section we compare the design length of boreelds
obtained with the nite and innite line source models for given
hourly ground load scenarios. These calculations imply that single
borehole solutions will need to be superimposed spatially. We have
already seen an instance of this principle of superposition while
computing the line source solution from a series of constant point
sources along a line [7], see Equations (1 and 2). The additivity of
effects (variation in temperature) stems from the linear relation
between q and DT, and the fact that energy is an extensive and
additive variable. The temporal superposition also follows the same
general principle of addition of effects as described by Yavuzturk
and Spitler [14] and illustrated by Fig. 6. When the load is varying
hourly, a new pulse is applied each hour. It is simply the difference
between the load for two consecutive hours. More formally, for the
innite line source as an example, with a single borehole, we have:
DTr; t

i; t
i
t
q

i
4pk
_
N
r
2
=4att
i

e
u
u
du (5)
where: q*
1
q
1
, and q*
i
q
i
q
i1
, i 2.I, t
I
t, is the incremental
load between two successive hours. With multiple boreholes,
DTx
0
; t

n
j 1

i; t
i
t
q
0
i
4pk
_
N
kx
j
x0k
2
=4att
i

e
u
u
du (6)
where: n is the number of boreholes, x
j
and x
0
are the coordinate
vectors of borehole j and point where temperature is computed,
respectively. Note that for long simulation periods, the computa-
tional burden becomes important.
In the test cases that follow we assume that all of the building
heating and cooling loads are to be provided by the GLHE system,
i.e. there is no supplementary heat rejection/injection. Synthetic
building loads are used to enhance the reproducibility of our
results. These building loads are simulated using:
Qt A B cos
_
t
8760
2p
_
C cos
_
t
24
2p
_
D cos
_
t
24
2p
_
cos
_
2t
8760
2p
_
(7)
In Equation (7), t is in hours, A controls the annual load
unbalance, B the half-amplitude of annual load variation, C and D
40 30 20 10 0 10 20 30 40
40
30
20
10
0
10
20
30
40
1
2
3 4
5
6 7
8 9
10
11 12
13
14 15
16 17
18 19
20 21
22 23
24 25
26
27 28
29
30 31
32 33
34 35
36 37
38 39
40 41
42 43
44 45
46
47 48
49
50 51
52 53
54 55
56 57
58 59
60 61
62 63
64 65
66 67
68 69
70
71 72
73 74
75 76
77 78
79 80
81
82 83
84 85
86 87
88 89
90 91
92 93
94 95
96 97
98 99
100 101
102 103
104 105
106 107
108 109
110
111 112
113
114 115
116 117
118 119
120 121
122 123
124 125
126 127
128 129
130 131
132 133
134 135
136 137
138 139
140 141
142 143
144 145
146
147 148
149
150 151
152 153
154 155
156 157
158 159
160 161
162 163
164 165
166 167
168 169
170 171
172 173
174 175
176 177
178 179
180 181
182 183
184 185
186 187
188 189
190 191
192 193
194 195
196 197
198 199
200 201
202 203
204 205
206 207
208 209
210 211
212 213
214 215
216 217
218 219
220 221
222 223
224 225
Borehole location and priority number
Coord. x (m)
C
o
o
r
d
.

y

(
m
)
Fig. 8. Borehole grid and priority number. Number indicates order of inclusion in the design when required.
Table 1
Number of boreholes required, complete geothermal system. Constant T assumes
a constant ground surface temperature of 10
o
C, Periodic T assumes a periodic
ground surface temperature with an amplitude of 20
o
C in phase with the heat
load.
Scenario Borehole length Innite line Finite line
Constant T Periodic T
Balanced (A 17) 100 m 33 33 34
Balanced 50 m 76 74 80
Cooling dominant (A 17) 100 m 39 36 37
Cooling dominant 50 m 93 79 81
Heating dominant
(A 30)
100 m 57 53 56
Heating dominant 50 m 134 115 124
Table 2
Number of boreholes required, hybrid system. HP capacity represents 40% of
maximum building load. The last two column represent the percentage of the
building load supplied by the HP for each mode.
Scenario Borehole
length
Number of
boreholes
% Energy
Innite Finite Cooling Inf.
(Fin.)
Heating Inf.
(Fin.)
Balanced 100 m 19 19 69 (69) 77 (78)
Balanced 50 m 37 37 67 (67) 72 (73)
Cooling dominant 100 m 24 24 69 (69) 86 (86)
Cooling dominant 50 m 41 39 70 (69) 90 (88)
Heating dominant 100 m 37 37 67 (67) 83 (86)
Heating dominant 50 m 55 53 72 (70) 70 (73)
D. Marcotte et al. / Renewable Energy 35 (2010) 763770 767
the half-amplitude of daily load uctuations. D/C controls the
relative importance of the damped component used to simulate
larger daily uctuations in winter and summer. Coefcients A to D
are in kW.
We consider three different load scenarios, each with B 100,
C 50, and D25. One is approximately balanced (A17), one is
a cooling dominated load (A17) and the other is a heating domi-
natedload(A30). Conversionof buildingloads togroundloads is
done with: q
ground
q
building
(1 1/COP). The heat pump COP varies
as a function of entering water temperature as depicted in Fig. 7.
For all scenarios, we consider a unique set of possible locations
for the boreholes. The locations are at the nodes of a regular grid of
mesh S 6 m. The boreholes are assigned a priority number (lower
number / highest priority), moving excentrically from the grid
center to the fringes (see Fig. 8). The system is simulated with
100 m and then, 50 m long boreholes. The borehole heads are
located at the ground surface.
For each scenario and given number of boreholes, we simulate
the hourly uid temperature for a 10-year period. We repeat this
simulation for different number of boreholes and nally keep the
lowest number of boreholes ensuring that the uid temperature at
the HP entrance never exceeds the HP limit specications. We do
this computation using both the innite and nite line source
models. Thermal parameters of the ground are: thermal conductivity
k
s
2 Wm
1
K
1
, volumetric heat capacity C
s
3.4e06 Jm
3
K
1
and borehole thermal resistance R
b
0.1 mKW
1
. The ground is
initially at 10
o
C and all the boreholes have a radius of r
b
0.075 m.
Table 1 gives the number of boreholes required for a specic
scenario and model. The choice of model does not have a strong
impact on the balanced load scenario. However, for cooling and
heating dominant scenarios, the nite line source model indicates
a reduction in the number of boreholes of approximately 7% and
15% for the 100 m and 50 m borehole length. As expected, the
shorter the borehole length, the greater are the discrepancies
between both models.
3.1. Seasonal effects
The nite line source model assumes a constant temperature at
the ground surface equal to the undisturbed ground temperature.
This hypothesis would more or less correspond to the case of
geothermal boreholes located under a building slab. However, in
many cases, the boreholes are located outside the building area
where the ground surface temperature varies in phase with the
heat load. One can expect the axial effect at the ground surface will
be less than under the constant temperature assumption. The
inuence of ground surface temperature variation on the ground
temperature, at any depth and time, can be computed for a periodic
signal [7]. Using the principle of superposition, the inuence on the
design can be assessed.
Last column of Table 1 gives the number of boreholes required
when the ground surface temperature is periodic, in phase with the
heat load, and shows a yearly variation of 20
o
C. The borehole
heads are located at the ground surface, a feature that maximizes
the seasonal effects on the design. As expected, the designs with
periodic ground surface temperature require more boreholes than
with constant ground surface temperature. For unbalanced
scenarios, the periodic ground surface temperature solutions
obtained are intermediate between the innite and the nite line
source designs.
0 5 10 15 20 25 30
6
4
2
0
2
4
6
8
Time (year)
o
T
e
m
p
e
r
a
t
u
r
e

(
C
)
Rock temperature at x=(13,5,12.0)


Infinite linesource
Finite linesource
15 10 5 0 5 10 15
15
10
5
0
5
10
15
Control point
x (m)
y

(
m
)
Fig. 9. Vertical average rock temperature at point x (13.5, 12.0). Thermal parameters: k
s
2 Wm
1
K
1
, C
s
3.4e06 Jm
3
K
1
.
0 5 10 15 20 25 30
1000
900
800
700
600
500
400
300
200
100
0
Time (year)
L
o
a
d

(
W
)
Ground cooling load / borehole


Infinite linesource
Finite linesource
Fig. 10. Evolution of average load per borehole.
D. Marcotte et al. / Renewable Energy 35 (2010) 763770 768
4. Design of hybrid geothermal systems
The design situation borrows its essential features (borehole
grid, thermal parameters, heat load scenarios) from the previous
complete geothermal system example. The main difference is that
the HP and auxiliary system power are respectively selected at 40%
and 60% of the maximum building load. The simulation works the
same way as previously with the following modication. Each hour,
the uid temperature at the HP entrance (EWT) is computed. If
EWT, at any hour, exceeds the HP specication limit, it is assumed
that the HP cycles to keep the EWT at its limit value, the excess load
being taken by the auxiliary system. We keep track of all the loads
provided by the auxiliary system. At the end of the 10-year simu-
lation, we verify that the load to the auxiliary systemnever exceeds
its capacity. If exceeded, we add more boreholes. The nal design is
obtained with the smallest number of boreholes compatible with
the auxiliary system capacity.
Table 2 shows the number of boreholes for a specic scenario
and computation model. The proportion of energy provided by the
geothermal system, relative to the total building energy required
for the 10-year period, is also given. Unbalanced scenarios with
short boreholes display reduction of only 6% for the number of
boreholes required. However, comparison of Tables 1 and 2 reveals
striking reduction in number of boreholes. The hybrid systems need
only between 46% and 70% of the number of boreholes for the
complete geothermal system. Yet, the geothermal components of
hybrid systems are able to supply between 67% and 70% of the
cooling load and between 73% and 88% of the heating load for the
various scenarios examined. As the borehole construction cost
represents the main investment in a geothermal system, it indi-
cates that hybrid systems can be economically advantageous.
Even though hourly simulations are computationally more
involved, efcient methods exist to speed up the computations
[1,10,12]. For example, a hybrid system 10-years hourly simulation
for a 45 boreholes eld is computed in less than 2 min on a stan-
dard laptop.
5. Comparing performances for a rock freezing problem
In some environmental problems, like those involving DNAPL
(dense non-aqueous phase liquid), the contaminants reach the
bedrock and then can propagate through the rock fractures. In such
X coordinate (m)
Y

c
o
o
r
d
i
n
a
t
e

(
m
)
Infinite line: Temperature after 5 years


15 10 5 0 5 10 15
15
10
5
0
5
10
15
12
10
8
6
4
2
0
2
X coordinate (m)
Y

c
o
o
r
d
i
n
a
t
e

(
m
)
Finite line: Temperature after 5 years


15 10 5 0 5 10 15
15
10
5
0
5
10
15
12
10
8
6
4
2
0
2
X coordinate (m)
Y

c
o
o
r
d
i
n
a
t
e

(
m
)
Infinite line: Temperature after 20 years
15
10
5
0
5
10
15
12
10
8
6
4
2
0
2
X coordinate (m)
Y

c
o
o
r
d
i
n
a
t
e

(
m
)
Finite line: Temperature after 20 years


15 10 5 0 5 10 15 15 10 5 0 5 10 15
15
10
5
0
5
10
15
12
10
8
6
4
2
0
2
Fig. 11. Ground temperature after 5 years (top) and 20 years (bottom). Innite line source model (left) and nite line source model (right).
D. Marcotte et al. / Renewable Energy 35 (2010) 763770 769
circumstances, remediation can be extremely difcult and expen-
sive. One solution is to catch and treat all the contaminants owing
through the fractures. This requires designing a series of wells
forming an hydraulic trap. The pumping and treatment is expensive
and it has to be maintained typically for hundreds of years. More-
over, the ow of water to treat could be relatively important. An
alternative is to permanently freeze the liquid within the fractures
to avoid any movement of the contaminant.
In the following example, we consider the problem of freezing
water inside rock fractures, in a domain 30 m by 30 m by 30 m, for
the purpose of long term connement of toxic DNAPLs. A 10 10
boreeld is considered (see Fig. 9). Thermal parameters of the rock
are: thermal conductivity k
s
2 Wm
1
K
1
and volumetric heat
capacity C
s
3.4e06 Jm
3
K
1
. The borehole thermal resistance is
R
b
0.1 mKW
1
. The rock is initially at 10
o
C. The rock porosity is
small enough so as to neglect the latent heat of water and the
increase of rock thermal conductivity occurring when water
freezes. The total HP capacity is 100 kW (one kW per borehole)
and the EWT limit is 8
o
C. The mass owof the circulating uid is
19 l/s (or 0.19 l/s for each borehole). The simulation is run for 30
years. The HP capacity is kept at its maximum until the EWT has
reached the limiting value of 8
o
C. Then, the HP cycles to
maintain the EWT at this temperature. The performance of the
system is monitored by computing rock temperature at a control
point located at the boreeld periphery at (x, y) (13.5, 12) m (see
Fig. 9).
Fig. 9 shows the rock temperature obtained at this control point
using the innite and nite line source models. Fig. 10 shows the
corresponding average hourly load per borehole. In both gures,
the rugged nature of the load curves is due to intermittent opera-
tion of the HP. As shown in Fig. 9, the rock becomes permanently
frozen after 10 months and 2.2 years for the innite and nite
models, respectively. The temperature difference between both
models increases steadily with time, to reach approximately 2
o
C
after 30 years. The nite line source model predicts that about
150 W per borehole are required compared to only 50 W per
borehole for the innite line source model (Fig. 10). Thus, despite
a higher heat transfer rate per borehole, the nite line source model
predicts a higher rock temperature at the control point.
In this next example, the same 10 10 boreeld is cooled at
a rate of 10 W/m (300 W/borehole) for the rst year and then
3.33 W/m (100 W/borehole) permanently. The temperature distri-
bution in the ground is computed using the innite and nite line
source models after 5 and 20 years. Results are presented in Fig. 11.
The proportion of the domain with temperature below freezing is
79% and 100% after 5 and 20 years for the innite line source
calculation whereas it is only 10% and 52% with the nite line
source model. Therefore, a design based on the use of the innite
line source would have wrongly predicted ground freezing. Thus,
contrary to the two previous examples, the innite model does not
provide a conservative design.
6. Conclusion
This paper studies the effect of axial heat conduction in bore-
holes used in geothermal heat pump systems. The axial effects are
examined by comparing the results obtained using the nite and
innite line source methods. Using various practical design prob-
lems it is shown that axial effects are relatively important. Even
with a borehole spacing / borehole length (S/H) as lowas 0.06, the
nite line source model indicated that 7% less boreholes are
required, for a complete geothermal system, when the yearly load is
unbalanced either in cooling or in heating. The reduction in the
number of boreholes due to axial effects increases to 15% for 50 m
long boreholes (ratio S/H0.12). Differences between the nite
and innite line source models were noticeably smaller for the
hybrid system but still indicate that less boreholes are generally
required with the nite line source model. As sizing of geothermal
systemis a determining factor for the protability of a project, axial
effects should certainly be considered in similar circumstances.
Also, the value of using hybrid systems to substantially reduce the
number of boreholes, yet preserving most energy savings, was
conrmed.
In another example dealing with underground water freezing,
the amount of energy that has to be removed to freeze the ground is
three times higher when axial effects are considered. Furthermore,
for this particular example, a design based on the innite line
source would not achieve the ground freezing goal. Indeed, while
the innite line source calculations show that 100% of the ground
would be frozen, the nite line source model predicts that only 52%
of the ground would be frozen.
Together, these examples clearly indicate that axial effects are
problem dependent. The magnitude of these effects is difcult to
anticipate and cannot generally be neglected. Moreover, neglecting
the axial effects (i.e. using the innite line source model) does not
always provide a conservative design. Since the computing cost of
nite line source model, when properly implemented, is only
slightly superior to the computing cost of the innite line source
model, there is no reason to overlook axial effects in any boreeld
design.
Acknowledgments
Part of this research was nanced by the National Science and
Engineering Research Council of Canada. We thank Jeremie
Gaucher for English editing. Helpful suggestions from anonymous
reviewers are acknowledged.
References
[1] Bernier M, Pinel P, Labib R, Paillot R. A multiple load aggregation algorithm for
annual hourly simulations of gchp systems. HVAC&R Research
2004;10(4):47187.
[2] Bernier M. Ground-coupled heat pump system simulation. ASHRAE Trans-
actions 2001;107(1):60516.
[3] Bernier M. Closed-loop ground-coupled heat pumps systems. ASHRAE Journal
2006;48(9):129.
[4] Crawley D, Lawrie L, Pedersen C, Strand R, Liesen R, Winkelmann F, et al.
Energyplus: creating a new-generation building energy simulation program.
Energy and Buildings 2001;33(4):31931.
[5] Fisher D, Rees S, Padhrmanabhan S, Murugappan A. Implementation and
validation of ground-source heat pump system models in an integrated
building and system simulation environment. IJHVAC &R Research
2006;13(3a):693710.
[6] Hellstrom G. Ground heat storage. Thermal analysis of duct storage systems:
Part i, theory, Ph.D. dissertation, University of Lund, Sweden, 1991.
[7] Ingersoll L, Zobel O, Ingersoll A. Heat conduction with engineering, geological
and other applications. New York: McGraw-Hill; 1954.
[8] S. Kavanaugh and K. Rafferty. Ground-source heat pumps. Design of
geothermal systems for commercial and institutional buildings. ASHRAE,
editor. ASHRAE, 1997.
[9] Klein S, Beckman W, Mitchell J, Dufe J, Dufe N, Freeman T, et al. In: S.E.
Laboratory, editor. TRNSYS 16 A transient system simulation program, user
manual. Madison: University of Wisconsin; 2004.
[10] Lamarche L. A fast algorithm for the hourly simulations of ground-source heat
pumps using arbitrary response factors. Renewable Energy 2009;10:22528.
[11] Lamarche L, Beauchamp B. A new contribution to the nite line-source model
for geothermal boreholes. Energy and Buildings 2007;39:18898.
[12] Marcotte D, Pasquier P. Fast uid and ground temperature computation for
geothermal ground-loop heat exchanger systems. Geothermics 2008;37:
65165.
[13] F. Sheriff. "Ge ne ration de facteurs de re ponse pour champs de puits ge o-
thermiques verticaux." Masters thesis, E

cole Polytechnique de Montre al.


2007.
[14] Yavuzturk C, Spitler J. A short time step response factor model for vertical
ground loop heat exchangers. ASHRAE Transactions 1999;105:47585.
[15] Zeng H, Diao N, Fang Z. A nite line-source model for boreholes in geothermal
heat exchangers,". Heat Transfer - Asian Research 2002;31:55867.
D. Marcotte et al. / Renewable Energy 35 (2010) 763770 770

You might also like