Applied Energy 261 (2020) 114361

Contents lists available at ScienceDirect

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

Underground solar energy storage via energy piles T

Qijie Ma , Peijun Wang
⁎

School of Civil Engineering, Shandong University, Jinan, Shandong, 250061, China

HIGHLIGHTS

• The thermal performance of energy piles for underground solar energy storage was investigated.
• AThelower flow rate of the circulating water was preferred.
• Thermal
maximum daily average rate of solar energy storage reached 150 W/m.
• interference induced a 10 W/m reduction in the daily average rate of solar energy storage.

ARTICLE INFO ABSTRACT

Keywords: Conventional piles embedded with geothermal loops, referred to as energy piles, have been successfully used as
Energy pile heat exchangers for the ground source heat pump system. For heating-dominated regions, it is crucial for the
Solar energy storage ground source heat pump system to keep the ground thermal balance in the long run. Solar energy is the most
Ground source heat pump system feasible source to charge the ground manually. In this study, thermal performance of an energy pile-solar col-
Thermal interference
lector coupled system for underground solar energy storage was investigated using numerical modeling. The
results suggested that a lower flow rate should be adopted for the energy pile-solar collector coupled system to
save the operational cost of the circulation pump. For the case with a pile length of 30 m, the decrease in the rate
of solar energy storage was about 2% when the mass flow rate was reduced from 0.3 to 0.05 kg/s. Throughout a
year, the maximum daily average rate of solar energy storage reached 150 W/m. It was also found that to
increase the length and the diameter of the pile improved the thermal performance of the system by keeping its
temperature relatively lower. In addition, the effects of the pile-pile thermal interference on reducing the rate of
solar energy storage after a one-year operation were quantified to be within 10 W/m for groups with the pile-pile
spacing of 3 times the pile diameter.

1. Introduction the ground enjoys the advantage of enormous quantity and being
widely accessible [4,5]. The conventional practice of underground
According to the International Energy Agency, buildings are re- thermal energy storage is burying heat exchange pipes into pre-drilled
sponsible for almost 40% of total final energy consumption in the vertical holes, referred to as the borehole thermal energy storage [6].
European Union, out of which 80% is due to the heat demand, ac- Heat transfer occurs by circulating heat carrier fluid through the pipes.
counting for 30% of the total CO2 emissions [1]. To reduce the carbon However, the cost of drilling deep holes can cause a breakdown of a
footprint and promote sustainable development, clean solar energy project [4]. In addition, with the quick development of urbanization,
offers excellent potential for heat production to meet the demands of the available free lands for drilling holes become increasingly more
space heating in winter and domestic hot water production. Never- scarce and costly in cities. Both factors impede the application of the
theless, solar radiation varies daily and seasonally and is not constantly borehole thermal energy storage technology.
present. To cover the intermittency of the solar radiation, thermal en- In recent years, energy piles have been attracting attention from the
ergy storage is necessary so that heat can be extracted when the solar academic field and getting more installations in engineering practice
radiation is not available [2,3]. Based on the medium adopted, thermal [7–9]. The energy piles combine the foundation piles with the heat
energy storage can be classified as sensible, latent, and chemical heat exchange pipes, the latter being attached to the steel cage and em-
storage. Of the common sensible mediums for thermal energy storage, bedded in the pile body, as illustrated in Fig. 1. In this way, the energy

⁎
Corresponding author.
E-mail address: qmaah@connect.ust.hk (Q. Ma).

https://doi.org/10.1016/j.apenergy.2019.114361
Received 4 September 2019; Received in revised form 28 November 2019; Accepted 10 December 2019
0306-2619/ © 2019 Elsevier Ltd. All rights reserved.
Q. Ma and P. Wang Applied Energy 261 (2020) 114361

Fig. 1. Schematic diagram of the GSHP system coupled with energy pile (Modified from Sani et al. [9]).

piles sustain the building load and hold the heat exchange pipes si- [25–28]. As a successful application, over 90% of the heat demand was
multaneously. Therefore, the underground space and the cost related to provided by the solar energy in the Drake Landing project, which em-
drilling additional boreholes can be saved. In addition, the concrete ploys 144 boreholes with a depth of 37 m [29].
energy piles enjoy relatively larger thermal conductivity and heat sto- The novelty of this study lies in the proposed energy pile-solar
rage capacity compared to the conventional boreholes [10]. The collector coupled system, the thermal performance of which for un-
thermal performance of energy piles has been tested with manually derground solar energy storage has not been studied yet. There are two
controlled inlet temperature or heat flux rate under different operation main different features of the energy pile-solar collector coupled system
modes (e.g., [11–13]). Zhao et al. [14] compared the effects of the pipe compared to the traditional borehole system for underground thermal
configuration and the spiral-shaped configuration was found to perform energy storage. First, the concrete pile material and the larger pile
better than others. Park et al. [15] found that as the number of heat diameter help to improve the thermal performance of energy piles.
exchange pipes increases, the contribution to improving the thermal Nevertheless, the energy piles are relatively shorter, usually less than
performance of energy piles gradually decreases due to the thermal 50 m. This leads to a smaller storage volume for each single energy pile.
interference between them. A novel truncated cone helix configuration It should also be always borne in mind that the primary function of the
was proposed and studied by Huang et al. [16] and Liu et al. [17] to energy piles is to support the structures built upon it. Therefore, the
reduce the thermal interference between pipes. Faizal et al. [18] gave a temperature changes of the energy piles should be kept within a safe
comprehensive review of the available ways to improve the thermal value to ensure that thermal effects on its geotechnical performance are
performance of energy piles. Li & Lai [19] reviewed the available acceptable [30,31].
analytical models for the thermal analysis of energy piles. Recently, Liu In this study, a mathematical model for the energy pile-solar col-
et al. [20] and Kong et al. [21] demonstrated the feasibility of energy lector coupled system was developed first and validated against field
piles for bridge snow melting. test results on energy piles. The model was built based on the compo-
The energy piles have been successfully used in the ground source nent approach, and the components were coupled through the heat
heat pump (GSHP) system to replace the traditional boreholes (see transfer between them. A systematic parametric study adopting steady-
Fig. 1). The GSHP system uses the ground as a heat source or a heat state analysis was then conducted to evaluate the effects of different
sink. Heat is injected into the ground in summer and extracted from it system parameters, including the flow conditions, the characteristics of
in winter. It takes advantage of the relatively constant ground tem- the solar collector, the intensity of solar radiation, the ambient air
perature throughout the year about 10 m below the ground surface. temperature, and the ground conditions. From the parametric study,
Therefore, it outperforms the air source heat pump system. A coefficient major factors affecting the thermal performance of the energy pile-solar
of performance close to 4.0 was reported for the GSHP system coupled collector coupled system for underground solar energy storage were
with energy piles [22]. For the GSHP system, however, it is critical to identified. In addition to the steady-state analysis, transient-state si-
keep the ground thermal balance to achieve a long-term high perfor- mulations were also performed to study the evolution with time of the
mance [21,24]. For heating-dominated areas, the natural recovery is thermal performance of the energy pile-solar collector coupled system.
usually not enough to meet the continuous heat extraction in winter, The focus was put on the rate of underground solar energy storage and
resulting in ground thermal imbalance and gradually decreasing heat the temperature change of the system. The effects of the pile length, the
exchange rate in the long run. The solar-assisted GSHP system has pile diameter, and the pile-pile thermal interaction were evaluated in
therefore been suggested to maintain the ground thermal balance the transient-state analysis.

2
Q. Ma and P. Wang Applied Energy 261 (2020) 114361

phases of the soil were assumed being at local equilibrium, implying the
same temperature for them. Accordingly, the following three-dimen-
sional heat diffusion equation was solved for the soil and also the
concrete energy pile, assuming they share the same temperature at the
pile-soil interface:
cp ( T / t ) ·(k T ) q=0 (2)
where T is the temperature; t is the time; q is the heat source term; , k ,
and cp are the density, thermal conductivity, and specific heat capacity
of the materials, respectively.
Density, thermal conductivity, and specific heat capacity of a soil
are functions of its void ratio (e) and degree of saturation (Sr).
According to Johansen [35] and Brandl [7], they can be related through
the following equations:

soil = (Gs + Sr e )· water /(1 + e) (3)

ksoil = (k i ) x i (4)

(cp )soil = ci x i (5)

where Gs is the specific weight of the soil particles with a typical value
of 2.65; x i is the specific volume of each phase of the soil: x w = nSr for
Fig. 2. Schematic diagram of the energy pile-solar collector coupled system for the water phase, x a = n (1 Sr ) for the air phase, and xs = 1 n for the
underground solar energy storage. solid phase, where n = e/(1 + e ) is the soil porosity. The adopted
thermal properties of the water, air, and soil particles are listed in
2. Methodology Table 1 [7,36]. The typical thermal properties of the concrete pile are
based on the European design code [37].
2.1. Mathematical formulation The water circulation through the pipes was modeled using the
following one-dimensional convection-diffusion heat transfer equation:
Fig. 2 shows a schematic diagram of the energy pile-solar collector T T T
coupled system to be modeled in this study. It consists of a flat-plate cp + v· · k· q=0
t s s s (6)
solar collector, an energy pile with the surrounding soil, and the water-
bearing pipes. The building upon the pile foundation is also shown in where v is the velocity of the circulating water; s is the distance along
the figure, but not modeled. The effect of the building on the thermal the flow path. When the thermal conduction term is dropped (k = 0 ),
performance of the system was accounted for by setting a constant- Eq. (6) is close to the further simplification adopted by Park et al. [38].
temperature boundary condition at the top ground surface of the model To ignore the thermal conduction term is reasonable when the heat
domain. Both single energy piles and energy pile groups considering the transport by convection dominates over that by conduction for turbu-
potential pile-pile thermal interference were modeled in this study. For lent conditions. In this study, the complete form of Eq. (6) was solved,
the energy pile groups, the same independent flat-plate solar collector considering that a wide range of flow velocity was simulated.
was connected to each energy pile of the group. Eq. (1) serves to determine the heat source term (q) of Eq. (6) when
According to Duffie & Beckman [32], the flat-plate solar collector the water passes through the solar collector. The coupling between Eq.
was modeled using the following equation: (2) and Eq. (6) is also realized via the heat source term (q) when the
water passes through the pipes embedded in the energy pile. The
qu' = w·F '·[( · GT ) U ·(Tf Ta)] (1) amount of heat transfer (q) between the circulating water and the en-
ergy pile is calculated by:
where qu' is the useful energy gain of the circulating water per unit
length of the solar collector tube; w is the tube spacing of the solar
q = hf ·(Tf Tp) (7)
collector; F' is the collector efficiency factor; is the transmittance- where Tf and Tp are the temperature of the circulating water and the
absorptance product; GT is the total available solar irradiation; U is the adjacent energy pile, respectively. hf is the convective heat transfer
overall heat loss coefficient to the ambient; Tf and Ta are the fluid coefficient between them. It was estimated using the Dittus-Boelter
temperature and the ambient air temperature, respectively. For all the correlation:
analyses conducted in this study, typical values of the tube spacing
kf
(0.15 m), the collector efficiency factor (0.9), and the transmittance- hf = 0.023R e 0.8Pr n
D (8)
absorptance product (0.75) were adopted.
Heat transfer through the ground (soil in this study) can include all where kf , R e , and Pr are the thermal conductivity, the Reynolds number,
the three modes conduction, convection, and radiation, with the con- and the Prandtl number of the circulating fluid, respectively. D is the
duction dominating most cases when the groundwater flow is negligible
[7]. It should be noted that the convective heat transfer due to Table 1
groundwater flow can cause opposite effects on the thermal perfor- Thermal properties of different materials adopted in this study.
mance of the underground solar energy storage system and the GSHP Material Density (kg/ Thermal conductivity (W/ Heat capacity (J/
system. For the underground solar energy storage system, the ground- m3) m/°C) kg/°C)
water flow can increase the heat loss due to self-discharge [33], while
for the latter it facilitates the heat restoration [34]. In this study, only Water 1000 0.57 4200
Air 1.25 0.026 1000
the thermal conduction was considered for the heat transfer through the
Soil particles 2650 4.0 1000
soil. This assumption is reasonable considering that the dry soil con- Concrete 2500 1.6 900
dition was adopted for most analyses conducted in this study. The three

3
Q. Ma and P. Wang Applied Energy 261 (2020) 114361

Fig. 3. Dimensions and mesh of the model domain for a single energy pile-solar collector coupled system.

inside diameter of the pipe. n = 0.4 for heating (Tf > Tp ) and n = 0.3 for an inside diameter of 0.02 m. The distance from the center of the pipe
cooling (Tf < Tp ). Although the water-bearing pipe was not modeled to the periphery of the pile was taken to be 0.05 m. The concrete pile
directly in this study, its effects on the thermal performance of the and the ground were modeled using three-dimensional diffusive heat
system were evaluated as to be discussed later in Section 3.1. transfer elements DC3D8. The 3U-in-series shaped water-bearing pipes
embedded within the pile body was modeled using the one-dimensional
2.2. Numerical implementation forced convection heat transfer elements DCC1D2. The mesh of the pile
and the adjacent ground region was particularly refined to capture their
The above developed mathematical model was solved using the fi- relatively large temperature gradients. The mesh of the water-bearing
nite element software Abaqus/Standard [39]. A typical finite element pipes was manually numbered in an ascending order along the flow
mesh of the model domain for a single energy pile-solar collector direction. The inlet and outlet of the water circulation denoted in the
coupled system is shown in Fig. 3. Typical dimensions were chosen for figure are defined with respect to the energy pile. The heat transfer
the simulated energy pile and the embedded water-bearing pipes ac- between the circulating water and the surrounding concrete pile was
cording to Bozis et al. [40]. The water-bearing pipes were taken to have modeled using thermal contact with a flow velocity-dependent heat

Fig. 4. Effects of the radius of the model domain on: (a) temperature distribution along the radial direction; (b) inlet, outlet temperature and their difference.

4
Q. Ma and P. Wang Applied Energy 261 (2020) 114361

Fig. 5. Effects of the clearance distance from the pile base to the bottom of the model domain on: (a) temperature distribution along the radial direction; (b) inlet,
outlet temperature and their difference.

transfer coefficient. The pile-soil interface was modeled using common The initial temperature of the whole domain was assumed to be at
nodes, assuming they share the same temperature. The heat transfer 15 °C, as adopted by Bourne-Webb et al. [41]. Constant-temperature
with the ambient air when the water passes through the solar collector boundary conditions were set at the outer lateral and the bottom sur-
was realized by setting two thermal boundary conditions. One is a face of the model domain. To fulfill this assumption, the model domain
concentrated heat flux boundary condition corresponding to the first should be large enough so that the assumed boundary conditions have
term of Eq. (1). The other is a thermal contact boundary condition negligible effects on the results. This was achieved by carrying out a
corresponding to the second term of Eq. (1). sensitivity analysis to determine the dimensions of the model domain.

Fig. 6. Validation of the developed model: (a) inlet temperature; (b) inlet-outlet temperature difference; (c) pipe temperature at different locations along the flow
path; (d) concrete temperature.

5
Q. Ma and P. Wang Applied Energy 261 (2020) 114361

Based on the results from the sensitivity analysis (see Figs. 4 & 5), the
Pile diameter
radius and the clearance distance from the pile base to the bottom of the

From 0.5 to
model domain were determined to be 100 m and 10 m, respectively.
From Fig. 4(a) it can be seen that the soil temperature corresponding to
(m)

1.0

1.5
1.0
the steady-state is not affected when the radial distance reaches about
100 m. Fig. 4(b) shows that the radius of the model domain has almost

From 10 to
Pile length

no effects on the steady-state inlet-outlet temperature difference, which

represents the rate of solar energy storage. Fig. 5 demonstrates that the
(m)

30

50
30
effects of the clearance distance from the pile base to the bottom of the
model domain are negligible after it exceeds 10 m. A constant-tem-
saturation of soil

From 0 to 100%

perature condition was also set at the top surface of the model domain.
This is based on the consideration that the top surface is covered by the
Degree of

building, and its basement temperature can be assumed to be main-

tained constant throughout the year [41].
(%)

0
Void ratio

2.3. Model validation

From 0.6
of soil

to 0.9
0.7

0.7

The above-developed model was validated against the field test on a

single energy pile conducted by Park et al. [38]. For the field test, there
collector tube (m)

was no solar collector, and the input heat flux was controlled manually.
Length of solar

From 10 to 50

In addition, the top ground surface was insulated so that a no-heat-flux

boundary condition, instead of a constant-temperature boundary con-
dition, was adopted in the model validation. The geometrical dimen-
20

20

sions and material properties used in the model validation followed the
details provided by Park et al. [38]. The measured variations of the inlet
From 100 to 650
Solar irradiance

temperature (see Fig. 6(a)) were curve-fitted and used as the input in
the model validation. It should be noted that the solar collector was not
(W/m2)

considered in the model validation. Nevertheless, the mathematical

Time variations (Fig. 13)
600

600

equation adopted to describe the performance of the solar collector was

well developed and widely used [32]. The practical issue lies in de-
termining the parameters characterizing the specific solar collector of
Air temperature

From 15 to 35

interest.
The simulated outlet temperature, pipe temperature at different
locations along the flow path, and the concrete temperature are com-
(°C)

15

15

pared to the measured values in Fig. 6(b–d). It should be noted that

grout was used in the original publication. Concrete is used here to be
coefficient (W/m2/

consistent with other sections. Overall, they show a reasonably good

agreement with each other. Regarding the inlet-outlet temperature
From 1 to 10
Overall loss

difference, the simulated results show a slightly higher value after it

turns to be almost constant, as shown in Fig. 6(b). The pipe temperature
decreases gradually in the flow direction (from P1 to P3) due to con-
°C)

tinuous heat loss to the surrounding concrete. The numerical simulation

Convective heat transfer

reproduces the same trend. Quantitatively, it overestimates the pipe

coefficient (W/m2/°C)

temperature, particularly for the third point P3, with a maximum dif-
From 100 to 1000

ference of about 4 °C. It should be noted that the measured temperature

drop from P2 to P3 is more significant compared to that from P1 to P2.
This is not as anticipated considering that the distance between them is
Eq. (8)

the same. The temperature drop from P2 to P3 should be smaller due to

the continuously decreased temperature gradient between the circu-
lating water and the surrounding concrete. In reality, it is likely that the
rate (kg/s)

From 0.05
Mass flow

non-uniform soil properties and other unidentified factors lead to the

to 0.3

observed larger temperature drop from P2 to P3. Fig. 6(d) indicates that
0.05

the simulated evolution of the concrete temperature correlates well

Transient-state

with the measured results with a maximum difference of about 2 °C.

Analysis type

Steady-state

2.4. Numerical program

Details of the parametric study.

Both steady-state and transient-state heat transfer analyses were

2×2&3×3

conducted in this study. The steady-state analysis was to evaluate the

energy piles

effects of different parameters on the thermal performance of the

Number of

Single pile

Pile group

system. For each steady-state simulation, the solar irradiation and the
ambient air temperature were kept constant. The transient-state ana-
lysis was to study the evolution of the thermal performance of the
Series No.

system with time-varying solar irradiation and ambient air temperature

Table 2

through a year. The parameters studied are summarized in Table 2. For

1

2
3
4

most cases, the dry soil condition was used. This is based on the

6
Q. Ma and P. Wang Applied Energy 261 (2020) 114361

Fig. 7. Effects of the convective heat transfer coefficient and the mass flow rate on: (a) inlet-outlet temperature difference; (b) rate of solar energy storage; (c) system
efficiency.

consideration that due to its relatively low thermal conductivity and system to different parameters. The rate of solar energy storage corre-
small heat capacity, results corresponding to the dry soil condition sponding to the steady-state is supposed to be at the lowest level due to
should represent the worst case in terms of the thermal performance of the gradual increase in the ground temperature. In addition, corre-
the system. In addition, during underground solar energy storage, sponding to the steady-state, there is no change in the temperature of
temperature increase drives the moisture away from the storage zone the system, indicating that the model domain has been fully charged.
[42]. Correspondingly, a dry soil condition is usually assumed in the Therefore, the rate of solar energy storage corresponding to the steady-
pure conduction analysis [4]. state also equals that of the energy loss to the building and the ground
outside the model domain.
3. Steady-state analysis
3.1. Effects of the convective heat transfer coefficient and the mass flow rate
In the following interpretation of the numerical results, the focus is
put on the inlet-outlet temperature difference (Tin Tout ), the rate of The convective heat transfer coefficient between the circulating
solar energy storage per unit of pile length (qsto /L ), and the efficiency of water and the surrounding concrete pile depends on the flow rate of the
the system ( ). The rate of solar energy storage qsto and the system ef- water and the characteristics of the water-bearing pipes. To model it
ficiency are calculated as follows: accurately, an equivalent heat transfer coefficient should be adopted,
accounting for the thickness and the thermal conductivity of the pipe
qsto = mc w (Tin Tout ) (9)
[38]. In this study, the effects of the water-bearing pipes were evaluated
by keeping the mass flow rate constant and changing the convective
= qsto /(F '· GT ·wl) (10)
heat transfer coefficient. Fig. 7 shows its effects on the thermal per-
where m is the mass flow rate; c w is the specific heat capacity of water; formance of the system for 5 different mass flow rates. It is expected
Tin and Tout are the inlet and outlet temperature of the circulating water that the temperature drop of the circulating water from the inlet to the
(see Fig. 3), respectively. F ' GT is the absorbed solar energy by the outlet increases with the convective heat transfer coefficient. This is
circulating water per area of the solar collector. w and l are the spacing observable as the convective heat transfer coefficient increases from
and the length of the solar collector tube, respectively. Essentially, the 100 to about 600 W/m2/°C, particularly for smaller mass flow rates, as
efficiency defined here characterizes the relative proportion between shown in Fig. 7(a). With a further increase in the convective heat
the solar energy injected into the ground and that lost to the air as the transfer coefficient, its contribution is almost negligible. Compared to
water passes through the solar collector. the convective heat transfer coefficient, the temperature drop is more
The steady-state analysis was conducted to study the sensitivity of sensitive to the mass flow rate. Fig. 7(a) shows that it increases from
the thermal performance of the energy pile-solar collector coupled about 1 to 6 °C as the mass flow rate decreases from 0.3 to 0.05 kg/s.

7
Q. Ma and P. Wang Applied Energy 261 (2020) 114361

Fig. 8. Effects of the air temperature and the overall loss coefficient on: (a) inlet-outlet temperature difference; (b) rate of solar energy storage; (c) system efficiency.

This is because a smaller mass flow rate allows more time for the heat 3.2. Effects of the air temperature and the overall loss coefficient of the solar
transfer between the circulating water and the surrounding pile. The collector
same trend was also observed by Gao et al. [8] and Başer et al. [43].
Both a larger convective heat transfer coefficient and a higher mass As expressed in Eq. (1), the air temperature and the overall loss
flow rate contribute to increasing the rate of solar energy storage, as coefficient affect the heat transfer between the circulating water and
shown in Fig. 7(b), being consistent with the experimental findings the air during its flow through the solar collector. The amount of heat
from Jalaluddin et al. [44]. While the increase is quite small, less than transfer depends linearly on the temperature difference between the
3 W/m, as the convective heat transfer coefficient increases to 1000 W/ circulating water and the ambient air. Therefore, it is expected that
m2/°C. When the energy pile serves as a heat exchanger for the GSHP higher air temperature can promote solar energy storage by reducing
system, the mass flow rate is usually about 0.25 kg/s to create a tur- the heat loss to the air, as demonstrated in Fig. 8(a). This is more evi-
bulent flow condition and increase the convective heat transfer coeffi- dent for cases with larger overall loss coefficients. When the air tem-
cient [45]. When used for underground solar energy storage, the results perature increases from 15 to 35 °C, the increase in the inlet-outlet
suggest that the mass flow rate should be reduced to save the opera- temperature difference reaches about 2.5 °C for the cases with the
tional cost of the circulation pump. This only causes a slight sacrifice of overall loss coefficient U = 10 W/m2/°C. Correspondingly, the rate of
the rate of solar energy storage, less than 2%, as the mass flow rate is solar energy storage increases by about 15 W/m, as shown in Fig. 8(b).
reduced from 0.3 to 0.05 kg/s. Therefore, a mass flow rate of 0.05 kg/s The rate of solar energy storage decreases with the increase of the
was used for the remaining analyses of this study. This is close to the overall loss coefficient due to more heat loss to the air. It drops more
mass flow rate (0.09 kg/s) for the heat injection test adopted by Zarrella quickly when the air temperature is low. The minimum value is slightly
et al. [46]. For the remaining analyses, the convective heat transfer more than 35 W/m. The system efficiency follows a similar pattern to
coefficient was calculated using Eq. (8) directly without considering the that of the solar energy storage. When the overall loss coefficient is
effects of the water-bearing pipes as they are relatively small. Regarding small, the system efficiency reaches 95%. For the worst condition, it
the system efficiency, it is supposed to follow a similar pattern to the drops to about 60%.
rate of solar energy storage as the intensity of solar radiation and the
area of the solar collector are the same for the cases shown in Fig. 7. 3.3. Effects of the solar irradiance and the length of the solar collector tube
Overall, the system efficiency corresponding to the steady-state is more
than 70%, as shown in Fig. 7(c). This means less than 30% of the solar The solar irradiance varies over time and locations, and it depends
energy absorbed by the circulating water was lost to the ambient air on the incidence angle and the climate conditions [32]. As the solar
during its pass-through the solar collector. irradiance increases, a larger inlet-outlet temperature difference is

8
Q. Ma and P. Wang Applied Energy 261 (2020) 114361

Fig. 9. Effects of the solar irradiance and the length of solar collector tube on: (a) inlet-outlet temperature difference; (b) rate of solar energy storage; (c) system
efficiency.

expected since the circulating water absorbs more heat during its pass- is even more significant for the case with l = 50 m. It increases from
through the solar collector. This is confirmed by the numerical results about 15 to almost 90 W/m.
shown in Fig. 9(a). A larger inlet-outlet temperature difference corre- Although an increase in the solar irradiance improves the rate of
sponds to a higher rate of solar energy storage. Fig. 9(b) shows that for solar energy storage, it has almost no effect on the efficiency of the
the case with the length of the solar tube l = 10 m, the rate of solar system, as shown in Fig. 9(c). This is because as the circulating water
energy storage increases from slightly more than 0 to about 20 W/m absorbs more heat, its temperature increase turns to be higher. There-
when the solar irradiance increases from 100 to 650 W/m2. This trend fore, it loses more heat to the air for the simulated condition with

Fig. 10. Distribution of the circulating water temperature along its flow path for cases with different lengths of the solar collector tube.

9
Q. Ma and P. Wang Applied Energy 261 (2020) 114361

Fig. 11. Effects of the void ratio and the degree of saturation on soil properties: (a) density; (b) thermal conductivity; (c) specific heat capacity.

constant air temperature. Consequently, the system efficiency remains 3.4. Effects of the ground conditions
almost unchanged as it characterizes the relative proportion between
the stored solar energy over that lost to the air. If the variations of the Both the void ratio and the degree of saturation have some effects
air temperature with the solar irradiation were considered, some non- on the soil thermal properties, including the density, the thermal con-
linearity should then be observed. Nevertheless, the system efficiency ductivity, and the specific heat capacity [4]. Therefore, they can affect
decreases from about 80% to about 50% as the length of the solar the thermal performance of the system. As demonstrated in Fig. 11,
collector tube increases from 10 to 50 m. This is because the tem- where the values are calculated using Eqs. (3)–(5), all three properties
perature of the circulating water continues to increase as it flows increase with the degree of saturation. Both the density and the thermal
through the solar collector. Therefore, as the length of the solar col- conductivity increase as the void ratio decreases. This is because the
lector tube increases, the circulating water loses increasingly more heat soil particles have larger values of density and thermal conductivity
to the air, and the efficiency of the system decreases accordingly. compared to the other two components (see Table 1). Nevertheless, the
The temperature distribution of the circulating water along its flow specific heat capacity decreases with the void ratio due to the smaller
path is shown in Fig. 10 for cases with different lengths of the solar heat capacity of soil particles compared to that of water. This is more
collector tube. It shows that during pass-through the solar collector, the evident for cases with higher degrees of saturation, as shown in
temperature gradient of the water decreases as the length of the solar Fig. 11(c). The thermal properties of the soil studied here are within
collector tube increases. This indicates more heat loss to the air, and their typical ranges, as reported by Bozis et al. [40]. The density, the
thus a lower system efficiency. It can also be observed from Fig. 10 that thermal conductivity, and the specific heat capacity range from 1400 to
the gradient of temperature drop during the three loops gradually de- 2100 kg/m3, from 0.3 to 2.0 W/m/°C, and from 1000 to 2500 J/kg/°C,
creases from the 1st loop to the 3rd loop. This is due to the continuous respectively.
decrease of the water temperature. The part between the 1st and the The resultant effects of the void ratio and the degree of saturation
2nd U loop and that between the 2nd and the 3rd U loop are at the top on the thermal performance of the energy pile-solar collector coupled
of the energy pile. They are in direct contact with the basement of the system are shown in Fig. 12. Overall, as the degree of saturation in-
building, the temperature of which was assumed to be a constant value creases, the thermal performance of the system improves remarkably.
of 15 °C. Therefore, the gradient of temperature drop along these two When the soil condition turns from being completely dry to being fully
parts is relatively more significant. saturated, the rate of solar energy storage increases by about 10 W/m

10
Q. Ma and P. Wang Applied Energy 261 (2020) 114361

Fig. 12. Effects of the void ratio and the degree of saturation of soil on: (a) inlet-outlet temperature difference; (b) rate of solar energy storage; (c) system efficiency.

(20%), and the system efficiency also increases by about 10%. They are temperature was used for simplicity, as shown in Fig. 13(d). It increases
consistent with the results on the thermal performance of energy piles, from the minimum value at sunrise to the maximum value at 2 pm and
as reported by Park et al. [45] and Akrouch et al. [47]. The reason lies drops linearly to the minimum value at sunrise of the next day. The
in the degree-of-saturation-driven increase in the density, the thermal system was supposed to work when the absorbed solar irradiance ex-
conductivity, and the specific heat capacity of the soil, as illustrated in ceeds 50 W/m2. In the numerical simulation, this was achieved by
Fig. 11. Compared to the degree of saturation, the void ratio has rela- controlling the mass flow rate of the circulating water. The adopted
tively smaller effects on the thermal performance of the system due to time increment for the analysis was 5 mins.
its limited effects on the soil thermal properties.
4.1. Transient performance of a typical case
4. Transient-state analysis
Shown in Fig. 14 are the obtained results for the case of a single
A total of seven cases were analyzed to study the evolution of the energy pile with L = 30 m and Dp = 1.0 m. To make it clear, only
thermal performance of the system with varying solar irradiance and results for a few days at intervals of 30 days are shown in the figure. For
ambient air temperature, as shown in Fig. 13. A time period of each day, in accordance with the hourly variations of the solar radiation
12 months from Jan 01 to Dec 30 was simulated to consider the sea- and the ambient air temperature (see Fig. 13), the inlet and outlet
sonal solar energy storage throughout a year. The adopted daily and temperature increase first, reach a peak value and then decrease. The
hourly variations of the solar irradiance and the ambient air tempera- maximum inlet temperature gradually increases with time and reaches
ture were based on the climate conditions of Jinan, China (North La- slightly more than 30 °C after about 210 days of operation. This is due
titude 36˚40′). The solar irradiance was calculated based on the stan- to the continuous solar energy storage in the ground and the resultant
dard clear-sky condition [48]. The maximum absorbed daily solar gradual increase of the ground temperature, as shown in Fig. 14(b). The
irradiance is about 650 W/m2 in summer (see Fig. 13(a)). Regarding its curve of the ground temperature distribution along path 1 (see Fig. 3)
hourly variations, it increases from 0 at sunrise to the maximum value continues to expand outwards with time. This means a wider zone being
at solar noon and then drops to 0 at sunset, as shown in Fig. 13(b). In thermally influenced and a gradual increase in the ground temperature.
terms of the ambient air temperature, its daily variations are in ac- After 210 days of solar energy storage, the temperature of the energy
cordance with those of the solar irradiance. The difference between the pile reaches the maximum value of about 24 °C. The corresponding
maximum and the minimum daily ambient air temperature is 15 °C (see temperature increase of the pile is about 9 °C, which is within the
Fig. 13(c)). Over a typical day, a linear variation of the ambient air normal operating temperature range of energy piles ( T 20 °C ) when

11
Q. Ma and P. Wang Applied Energy 261 (2020) 114361

Fig. 13. Adopted daily and hourly variations of the absorbed solar irradiance and the ambient air temperature for the transient-state analysis: (a) maximum daily
variations of the absorbed solar irradiance throughout a year; (b) hourly variations of the absorbed solar irradiance over a typical day; (c) maximum and minimum
daily variations of the ambient air temperature throughout a year; (d) hourly variations of the ambient air temperature over a typical day.

used for the GSHP system. Afterward, the temperature of the energy 4.2. Effects of the pile length
pile and the adjacent ground decreases with time, while that of the far-
away ground increases. This indicates a transfer of stored solar energy Fig. 15 compares the thermal performance of the single energy pile-
from the ground adjacent to the energy pile to the far-field. This solar collector coupled system with three different pile lengths
transfer is induced by the gradual decrease in the rate of solar energy (L = 10 m, 30 m, and 50 m). It can be seen from Fig. 15(a) that as the
storage after about 210 days (see Fig. 14(c)). By the end of the one-year pile length increases, the daily average inlet temperature decreases. The
operation, the temperature of the energy pile drops to about 20 °C with maximum daily average inlet temperature for the case with L = 10 m is
an increment of about 5 °C. During the one-year operation, the max- about 10 °C higher than that for the case with L = 50 m. This is due to
imum temperature of the storage system is below 30 °C. Therefore, it the higher temperature increase of the pile and the surrounding ground
can be characterized as a low-temperature system [49], which is ad- for the shorter energy piles, as shown in Fig. 15(b). As the pile length
vantageous to reduce heat loss due to self-discharge. decreases, the volume of ground for solar energy storage decreases. For
The gradual increase of the ground temperature leads to a decrease the same amount of solar energy storage, the temperature increase is
in the rate of solar energy storage. A higher ground temperature makes higher for a smaller storage volume. For the case with L = 10 m, the
it more difficult to further inject heat into the ground due to the re- maximum temperature increase of the pile is about 16 °C after
duced temperature gradient between the circulating water and the 210 days, which is about 10 °C higher than that for the case with
surrounding ground. As demonstrated in Fig. 14(c), by the end of the L = 50 m. Although the temperature increase is higher for cases with
one-year operation, both the peak and the average rate of solar energy shorter energy piles, the zone being thermally influenced is roughly the
storage are slightly smaller than the corresponding values at the be- same.
ginning. The radius of the ground being thermally affected reaches As the pile length increases, it allows more time for heat transfer
about 10 m (see Fig. 14(b)), which is 10 times the pile diameter. between the circulating water and the energy pile. This results in a
Usually, the pile-pile spacing is about 3–5 times the pile diameter. larger inlet-outlet temperature difference for the longer energy piles, as
Therefore, thermal interference between energy piles will occur if in- shown in Fig. 15(c). The maximum daily average inlet-outlet tem-
stalled in a pile group. The rate of solar thermal energy storage is thus perature difference for the case with L = 50 m is about 2 °C higher than
expected to decrease more due to an even higher increase in ground that for the case with L = 10 m. This means a higher total solar energy
temperature for an energy pile group. storage rate for the longer energy piles. If the total solar energy storage

12
Q. Ma and P. Wang Applied Energy 261 (2020) 114361

Fig. 14. Evolution with time of: (a) inlet and outlet temperature; (b) distribution of ground temperature along path 1 (vertically halfway the pile length); (c) rate of
solar energy storage.

rate is divided by the pile length, however, the shorter energy piles are deformation and bearing capacity. From an economic point of view,
superior over the longer energy piles (see Fig. 15(d)). The maximum Fig. 15(d) can serve to determine the operation of the system while
daily average rate of solar energy storage decreases from as high as taking into consideration the cost of the circulation pump. A minimum
150 W/m for the case with L = 10 m to about 35 W/m as the pile value of solar energy storage rate can be calculated, below which the
length increases to 50 m. The maximum daily average rate of solar operation of solar energy storage should be stopped.
energy storage for the case with L = 30 m is slightly over 50 W/m. To
improve its thermal performance, solar collectors with a larger area 4.3. Effects of the pile diameter
should be adopted for the longer energy piles. Further study of optimal
design is necessary to maximize its thermal performance of under- The effects of the pile diameter on the thermal performance of the
ground solar energy storage while ensuring its safety in terms of both energy pile-solar collector coupled system are shown in Fig. 16. It can

13
Q. Ma and P. Wang Applied Energy 261 (2020) 114361

Fig. 15. Comparison of the thermal performance of the system with different pile lengths: (a) daily average inlet temperature during operation period; (b) tem-
perature distribution along path 1 after 210 days (vertically halfway the pile length); (c) daily average inlet-outlet temperature difference during operation period;
(d) daily average rate of solar energy storage during operation period.

be seen from Fig. 16(a–b) that both the inlet temperature and the pile rate of solar energy storage for each energy pile of the groups to that of
temperature are lower for cases with larger pile diameters. The pile the single energy pile with the same pile length and pile diameter. It
temperature for the case with Dp = 1.5 m is about 5 °C smaller than shows that initially there is no thermal interference between piles until
that with Dp = 0.5 m, which has a maximum pile temperature increase after about 90 days of solar energy storage. This is because the thermal
of about 13 °C after 210 days of operation. The density and thermal front generated by each energy pile did not meet each other initially.
conductivity of concrete are larger than those of the soil. As the pile Afterward, the effects of the pile-pile thermal interference on reducing
diameter increases, there has a relatively larger volume of concrete for the rate of solar energy storage appear and gradually increase. As ex-
solar energy storage, leading to a lower pile temperature. As a result of pected, the middle pile of the 3 × 3 group is to suffer most from the
its lower temperature, a higher rate of solar energy storage is observed pile-pile thermal interference. By the end of the one-year operation, its
for cases with larger pile diameters. As shown in Fig. 16(d), the max- daily average rate of solar energy storage is about 10 W/m less than
imum daily average rate of solar energy storage for the case with that of the single energy pile. In addition, the daily average rate of solar
Dp = 1.5 m is about 5 W/m (10%) larger than that with Dp = 0.5 m. In energy storage of the corner pile of the 3 × 3 group is also less than that
combination with the previous section, it can be concluded that to in- of the corner pile of the 2 × 2 group. This is due to that the former is
crease either the pile length or the pile diameter can contribute to thermally affected by more energy piles than the latter, leading to a
maintaining the temperature of the system relatively lower and thus higher increase in the ground temperature as discussed below.
improving the rate of solar energy storage. Shown in Fig. 18 is the temperature distribution across the region
surrounding the energy piles with a diameter of about 9 m at three
4.4. Effects of the pile-pile thermal interference different time points. The case of the single energy pile has been ana-
lyzed before (see Fig. 14(b)). For the 2 × 2 and the 3 × 3 group,
As shown above, after one year of solar energy storage, the ground temperature changes of each energy pile and its adjacent ground are
being thermally affected had a radius of up to 10 m. Therefore, when roughly the same at the beginning. This condition exists during the first
installed in a group pile-pile thermal interference is expected. To 90 days of operation and indicates no pile-pile thermal interference, as
quantify the effects of the pile-pile thermal interference on the rate of also confirmed by the results of the solar energy storage rate shown in
solar energy storage, one 2 × 2 and one 3 × 3 energy pile group were Fig. 17. As the process of solar energy storage continues, the pile-pile
simulated (see Table 2 series 7). The pile-pile spacing was 3 times the thermal interference occurs, resulting in that the temperature at the
pile diameter, which represents the minimum pile spacing for most center of the pile group gradually increases. By the end of the one-year
engineering applications. Therefore, the pile-pile thermal interference solar energy storage, the contour of the temperature distribution shows
is expected to be most noticeable. Fig. 17 compares the daily average concentric circles centered at the middle point of the group for both the

14
Q. Ma and P. Wang Applied Energy 261 (2020) 114361

Fig. 16. Comparison of the thermal performance of the system with different pile diameters: (a) daily average inlet temperature during operation period; (b)
temperature distribution along path 1 after 210 days (vertically halfway the pile length); (c) daily average inlet-outlet temperature difference during operation
period; (d) daily average rate of solar energy storage during operation period.

2 × 2 and the 3 × 3 energy pile group. This indicates a concentration turned from being completely dry to being fully saturated, the rate
of the stored solar energy within the region of the pile group, being of solar energy storage increased by about 20%.
consistent with the analysis results from Başer et al. [43]. By the end of 3) Throughout a year, the rate of solar energy storage changed in ac-
the one-year operation, the maximum temperature for the 2 × 2 and cordance with that of the solar irradiance and the ambient air
the 3 × 3 energy pile group is about 30 °C and 35 °C, respectively. Both temperature. Under the specific thermal boundary conditions
are higher than that of the single energy pile, about 20 °C. This is the adopted, the maximum daily average rate of solar energy storage
reason behind the relatively higher rate of solar energy storage gener- reached 150 W/m for the 10 m-long energy pile. It decreased to
ated by the single energy pile (see Fig. 17). about 35 W/m as the pile length increased to 50 m. In addition, due
to the gradual build-up of the ground temperature, the rate of solar
5. Conclusions energy storage by the end of the one-year operation was smaller
than that at the beginning.
This study evaluated the thermal performance of an energy pile- 4) It was found that a larger pile size in terms of both the pile diameter
solar collector coupled system for underground solar energy storage. and the pile length was favorable to keep system temperature re-
Both steady-state and transient-state analyses were conducted to latively lower, and thus improved the total rate of solar energy
quantify the effects of relevant design parameters and the evolution storage. This is because the concrete pile material has a relatively
with time of its thermal performance throughout a year. Based on the larger density and thermal conductivity compared to the soil. This
results and discussions presented, the authors attempt to draw the also implies that to increase the density, thermal conductivity, and
following conclusions: heat capacity of the pile material can further improve the thermal
performance of the energy pile-solar collector coupled system for
1) For the energy pile-solar collector coupled system to store solar underground solar energy storage.
energy underground, lower flow rates of the circulating water were 5) For the simulated cases, the maximum temperature increase ex-
preferred to save the operational cost of the circulation pump at perienced by the energy pile was about 16 °C after a 210-day op-
only a slight sacrifice of the rate of solar energy storage. For the eration. By the end of the one-year operation, the ground being
specific case simulated, the decrease in the rate of solar energy thermally affected had a radius of about 10 m, indicating potential
storage was about 2% as the mass flow rate was reduced from 0.3 to pile-pile thermal interference. Quantitatively, the effects of the pile-
0.05 kg/s. pile thermal interference on reducing the rate of solar energy sto-
2) The thermal performance of the system was quite sensitive to the rage were less than 10 W/m for an energy pile group with pile
water content of the ground (soil in this study). When the ground spacing of 3 times the pile diameter after a one-year operation.

15
Q. Ma and P. Wang Applied Energy 261 (2020) 114361

Fig. 17. Effects of the pile-pile thermal interference on the daily average rate of solar energy storage.

Fig. 18. Evolution of temperature distribution with time for different cases (vertically halfway the pile length): (a) single energy pile; (b) 2 × 2 energy pile group; (c)
3 × 3 energy pile group.

16
Q. Ma and P. Wang Applied Energy 261 (2020) 114361

CRediT authorship contribution statement bridge decks using geothermal energy piles integrated with heat pump in Canada.
Renewable Energy 2019;136:1266–80.
[21] Kong G, Wu D, Liu H, Laloui L, Cheng X, Zhu X. Performance of a geothermal energy
Qijie Ma: Conceptualization, Methodology, Data curation, deicing system for bridge deck using a pile heat exchanger. Int J Energy Res
Validation, Writing - original draft, Writing - review & editing. Peijun 2019;43:596–603.
Wang: Funding acquisition, Resources, Software, Writing - review & [22] Wood CJ, Liu H, Riffat SB. An investigation of the heat pump performance and
ground temperature of a piled foundation heat exchanger system for a residential
editing. building. Energy 2010;35(12):4932–40.
[23] You T, Li X, Cao S, Yang H. Soil thermal imbalance of ground source heat pump
Declaration of Competing Interest systems with spiral-coil energy pile groups under seepage conditions and various
influential factors. Energy Convers Manage 2018;178:123–36.
[24] Akrouch GA, Sánchez M, Briaud JL. Thermal performance and economic study of an
The authors declared that there is no conflict of interest. energy piles system under cooling dominated conditions. Renewable Energy 2018.
[25] Shah SK, Aye L, Rismanchi B. Seasonal thermal energy storage system for cold
climate zones: A review of recent developments. Renew Sustain Energy Rev
Acknowledgements
2018;97:38–49.
[26] Welsch B, Göllner-Völker L, Schulte DO, Bär K, Sass I, Schebek L. Environmental
This work was supported by the National Natural Science and economic assessment of borehole thermal energy storage in district heating
Foundation of China (grant number 51908328). systems. Appl Energy 2018;216:73–90.
[27] Yang H, Cui P, Fang Z. Vertical-borehole ground-coupled heat pumps: a review of
models and systems. Appl Energy 2010;87(1):16–27.
References [28] Rad FM, Fung AS, Leong WH. Feasibility of combined solar thermal and ground
source heat pump systems in cold climate, Canada. Energy Build 2013;61:224–32.
[29] Mesquita L, McClenahan D, Thornton J, Carriere J, Wong B Drake. Landing Solar
[1] International Energy Agency. World Energy Outlook 2018. Paris: OECD/IEA; 2018. Community: 10 years of operation. ISES Solar World Congress 2017.
[2] Yan J, Yang X. Thermal energy storage. Appl Energy 2019;240:A1–6. [30] Ng CWW, Ma QJ, Gunawan A. Horizontal stress change of energy piles subjected to
[3] Guelpa E, Verda V. Thermal energy storage in district heating and cooling systems: a thermal cycles in sand. Comput Geotech 2016;78:54–61.
review. Appl Energy 2019;252:113474. [31] Ng CWW, Ma QJ. Energy pile group subjected to non-symmetrical cyclic thermal
[4] Pinel P, Cruickshank CA, Beausoleil-Morrison I, Wills A. A review of available loading in centrifuge. Géotechnique Letters 2019;3:1–5.
methods for seasonal storage of solar thermal energy in residential applications. [32] Duffie JA, Beckman WA. Solar engineering of thermal processes. 2nd ed. New York:
Renew Sustain Energy Rev 2011;15(7):3341–59. Wiley Interscience; 1991.
[5] Xu J, Wang RZ, Li Y. A review of available technologies for seasonal thermal energy [33] McDaniel B, Kosanovic D. Modeling of combined heat and power plant performance
storage. Sol Energy 2014;103:610–38. with seasonal thermal energy storage. J Storage Mater 2016;7:13–23.
[6] Gao L, Zhao J, Tang Z. A review on borehole seasonal solar thermal energy storage. [34] Li C, Mao J, Xing Z, Zhou J, Li Y. Analysis of geo-temperature restoration perfor-
Energy Procedia 2015;70:209–18. mance under intermittent operation of borehole heat exchanger fields.
[7] Brandl H. Energy foundations and other thermo-active ground structures. Sustainability 2016;8(1):35.
Géotechnique 2006;56(2):81–122. [35] Johansen O. Thermal conductivity of soils. PhD thesis, University of Trondheim,
[8] Gao J, Zhang X, Liu J, Li K, Yang J. Numerical and experimental assessment of CRREL Draft English Translation 637; 1977.
thermal performance of vertical energy piles: an application. Appl Energy [36] Choi JC, Lee SR, Lee DS. Numerical simulation of vertical ground heat exchangers:
2008;85(10):901–10. intermittent operation in unsaturated soil conditions. Comput Geotech
[9] Sani AK, Singh RM, Amis T, Cavarretta I. A review on the performance of geo- 2011;38(8):949–58.
thermal energy pile foundation, its design process and applications. Renew Sustain [37] EN 1992-1-2 (2004) European Committee for Standardization (CEN). Design of
Energy Rev 2019;106:54–78. concrete structures—Part 1-2: General rules—Structural fire design. EN 1992
[10] Batini N, Loria AFR, Conti P, Testi D, Grassi W, Laloui L. Energy and geotechnical Eurocode 2; 2004.
behaviour of energy piles for different design solutions. Appl Therm Eng [38] Park H, Lee SR, Yoon S, Choi JC. Evaluation of thermal response and performance
2015;86:199–213. of PHC energy pile: Field experiments and numerical simulation. Appl Energy
[11] Go GH, Lee SR, Yoon S, Kang HB. Design of spiral coil PHC energy pile considering 2013;103:12–24.
effective borehole thermal resistance and groundwater advection effects. Appl [39] ABAQUS Inc. ABAQUS user’s manual Ver. 6.13. Rhode Island 2013.
Energy 2014;125:165–78. [40] Bozis D, Papakostas K, Kyriakis N. On the evaluation of design parameters effects on
[12] Park S, Lee D, Lee S, Chauchois A, Choi H. Experimental and numerical analysis on the heat transfer efficiency of energy piles. Energy Build 2011;43(4):1020–9.
thermal performance of large-diameter cast-in-place energy pile constructed in soft [41] Bourne-Webb PJ, Bodas Freitas TM, Freitas Assunção RM. Soil–pile thermal inter-
ground. Energy 2017;118:297–311. actions in energy foundations. Géotechnique 2015;66(2):167–71.
[13] Guo Y, Zhang G, Liu S. Investigation on the thermal response of full-scale PHC [42] Sani AK, Singh RM. Response of unsaturated soils to heating of geothermal energy
energy pile and ground temperature in multi-layer strata. Appl Therm Eng pile, Renewable Energy, https://doi.org/10.1016/j.renene.2018.11.032.
2018;143:836–48. [43] Başer T, Lu N, McCartney JS. Operational response of a soil-borehole thermal en-
[14] Zhao Q, Chen B, Liu F. Study on the thermal performance of several types of energy ergy storage system. J Geotech Geoenviron Eng 2015;142(4):04015097.
pile ground heat exchangers: U-shaped, W-shaped and spiral-shaped. Energy Build [44] Miyara A, Tsubaki K, Inoue S, Yoshida K. Experimental study of several types of
2016;133:335–44. ground heat exchanger using a steel pile foundation. Renewable Energy
[15] Park S, Lee S, Lee D, Ahn D, Choi H. Effect of thermal interference on energy piles 2011;36(2):764–71.
considering various configurations of heat exchangers. Energy Build [45] Park S, Lee S, Oh K, Kim D, Choi H. Engineering chart for thermal performance of
2019;199:381–401. cast-in-place energy pile considering thermal resistance. Appl Therm Eng
[16] Huang G, Yang X, Liu Y, Zhuang C, Zhang H, Lu J. A novel truncated cone helix 2018;130:899–921.
energy pile: modelling and investigations of thermal performance. Energy Build [46] Zarrella A, De Carli M, Galgaro A. Thermal performance of two types of energy
2018;158:1241–56. foundation pile: helical pipe and triple U-tube. Appl Therm Eng 2013;61(2):301–10.
[17] Liu Y, Huang G, Lu J, Yang X, Zhuang C. A novel truncated cone helix energy pile: [47] Akrouch GA, Sánchez M, Briaud JL. An experimental, analytical and numerical
numerical and laboratory investigations of thermal performance. Int J Heat Mass study on the thermal efficiency of energy piles in unsaturated soils. Comput Geotech
Transf 2018;126:681–99. 2016;71:207–20.
[18] Faizal M, Bouazza A, Singh RM. Heat transfer enhancement of geothermal energy [48] Hottel HC. A simple model for estimating the transmittance of direct solar radiation
piles. Renew Sustain Energy Rev 2016;57:16–33. through clear atmospheres. Sol Energy 1976;18(2):129–34.
[19] Li M, Lai ACK. Review of analytical models for heat transfer by vertical ground heat [49] Hesaraki A, Holmberg S, Haghighat F. Seasonal thermal energy storage with heat
exchangers (GHEs): a perspective of time and space scales. Appl Energy pumps and low temperatures in building projects—A comparative review. Renew
2015;151:178–91. Sustain Energy Rev 2015;43:1199–213.
[20] Liu H, Maghoul P, Bahari A, Kavgic M. Feasibility study of snow melting system for

17