Technical Performance Comparison of Horizontal and Vertical Ground-Source Heat Pump Systems

,


Introduction
Climate change is an urgent global concern due to the large amount of greenhouse gases (GHG) emitted into the atmosphere [1].Fossil fuel consumption in sectors such as transport, business, and residential is largely responsible for the GHG emission.In the UK, for instance, it is reported that the residential sector emitted 68.1 MtCO 2 , accounting for 19.9% of all carbon dioxide emissions in 2021, and the main source is the use of natural gas to heat homes [2].With the increasing awareness of environmental protection and energy sustainability and security, shallow ground-source heat becomes a promising alternative to conventional fossil fuels to reduce the GHG emission for heating and cooling applications in residential, commercial, and public buildings [3][4][5][6][7].
Ground-source heat pump (GSHP) systems are usually installed to utilize the shallow ground-source heat [8,9].Depending on the configuration of ground heat exchangers (GHEs), GSHP systems can be classified as open or closedloop systems.Closed-loop GSHP systems can be further divided into two types based on the installation orientation of the GHEs [10].One is the GSHP system coupled with vertical boreholes, which usually reach depths of 15-200 m [11,12].The other is the GSHP system coupled with horizontal ground loops, which are installed in trenches at a depth of 1-3 m [10,13].These two types of GSHP systems are referred to as vertical and horizontal GSHP systems, respectively, in this study.
Numerical and experimental studies were conducted to investigate the technical and economic performance of horizontal GSHP systems under the various influential factors, including design configurations and climatic conditions.For example, Li et al. [10] developed a numerical model, which considered the geothermal gradient and varying ambient air temperature, to investigate the operating characteristics of horizontal spiral-coil GSHP system under the influences of soil thermal conductivity, buried depth, pipe spacing, and ambient air temperature.The study indicated that the soil thermal conductivity and pipe spacing are the main influential factors.Sedaghat et al. [14] proposed a ground thermal recovery system for horizontal GHEs in a hot climate, which supplied ambient-temperature air to the ground-air pipes installed between the GHEs to remove the accumulated heat from the ground.The effects of GHE length, pipe spacing, buried depth, and ground-air pipe diameter on the system performance were numerically studied and it was found that the annual coefficient of performance (COP) of the system was increased.Gao et al. [15] carried out a sandy soil container experiment to compare the thermal performance of a horizontal GHE with and without the supply of rainwater.The experiment showed the performance improvement of the system by rainwater harvest due to the increased thermal conductivity by increased moisture content.Kayaci and Demir [16] conduct an experiment to investigate the transient soil temperature profile (the horizontal and vertical temperature distribution in soil) of a horizontal GSHP system under the influences of real climatic conditions, and economic analyses considering the initial investment and operational costs were employed to study the effects of the increase rate in electricity prices, number of pipes, burial depth, pipe spacing, pipe diameter, and pipe length on the system.
The performance of vertical GSHP systems was also investigated numerically and experimentally.For example, Hein et al. [17] developed a numerical model, which included the groundwater flow and heat transport processes and the dynamics of heat pump efficiency, to study the sustainability and efficiency of vertical GSHP systems.It was found that groundwater flow and injection of excess heat would be beneficial to the energy recovery and efficiency of the heat pump.Li et al. [18] constructed a numerical model to analyse the influence of unsaturated soil properties and groundwater flow on the performance of vertical GSHP systems.The simulated results showed that neglecting variations in moisture content in unsaturated soil would underestimate the heat transfer capacity of the soil, and a rising groundwater table was beneficial to the heat transfer of the borehole and the operation of the vertical GSHP system.In addition, Soltani et al. [19] numerically investigated the capacity of various circulating fluids and their effects on energy consumption reduction of vertical GSHP systems.The simulations showed that utilizing varying levels of ethylene glycol, methanol, potassium acetate, sodium chloride, and Freezium as the heat carrier fluid would decrease the energy consumption significantly compared to the pure water.Boughanmi et al. [20] experimentally examine the performance of a vertical GSHP system coupled with a conic basket pipe for greenhouse cooling.Monitored results indicated that the air temperature inside the greenhouse decreased by 8-12°C with the operation of the cooling system.
Except for the individual study on the horizontal or vertical GSHP systems, research studies also compared the performance between the horizontal and vertical GSHP systems.To demonstrate the technical and design feasibility of GSHP systems in mild climate applications for greenhouse heating, Benli [21] conducted an experimental comparison using a heating system consisting of two different GHEs.Results showed that the heating COP of the overall system was 2.7-3.3 for the horizontal system and 2.9-3.5 for the vertical system.Lee et al. [22] carried out a field test to analyse the performance of GSHP systems when the coil-type GHE was installed horizontally or vertically.It was found that the amount of electric power consumed by the horizontal system was higher than that consumed by the vertical system, and the cooling COP was 3.9-4.3for the vertical system and 3.3-3.7 for the horizontal system.Yin et al. [23] compared the field performance of GSHP systems for residential space heating based on 32 residential houses with 16 vertical and 16 horizontal GHEs.Results showed the COP for horizontal systems ranged from 1.18 to 4.57 while the COP for vertical systems demonstrated less variation (1.96-3.8)due to more stable ground temperatures given the fact that vertical GHEs reached deeper into the ground than the horizontal GHEs.Furthermore, on the basis of a case study of a residential building with a fixed heating and cooling load under moderate climate conditions, Aresti et al. [24] performed a life cycle analysis (LCA) for a direct environmental impact comparison between different GHE configurations, including three types vertical GHEs and five types horizontal GHEs.It was found that the vertical coaxial GHE configuration led to the most negative environmental impact among all GHE configurations, and the horizontal GHEs outperformed the vertical GHEs in all impact categories.
Based on the literature review conducted, it becomes evident that the configuration of the GHEs is pivotal in determining the efficiency and sustainability of GSHP systems.However, a notable knowledge gap exists regarding the performance disparities between horizontal and vertical GSHP systems within the same project, particularly when confronted with varying heating and cooling demands.This knowledge gap consequently hinders the selection of appropriate GHEs to ensure the optimal performance of GSHP systems.
To address this research gap, this study focuses on a technical performance comparison between a horizontal GSHP system and a vertical GSHP system under the three distinct scenarios characterised by varying heating-to-cooling ratios.To conduct these comparisons, a coupled thermal-hydraulic (TH) model for unsaturated soils has been employed, which takes into account realistic ground surface boundaries, GHE boundary, and the dynamics of heat pump efficiency.Furthermore, the design of the GHEs utilised information gathered from an experimental site situated on the campus of a UK university.
This paper is structured as follows: first, a comprehensive description of the numerical model employed in this study is presented.After that, the model validation is provided.Next, the model is applied to the experimental site, and subsequently, results derived from the simulations and discussion are given.Finally, conclusions are drawn.

Numerical Model
2.1.Moisture and Heat Transfer in Unsaturated Soils.Shallow ground is generally unsaturated.The governing equation of moisture transfer within unsaturated soils can be given as follows [25]: As shown in Equation ( 1), moisture in unsaturated soils consists of liquid water and vapour.θl is the volumetric water content (-), θ a is the volumetric air content (-), t is the time (s), ∇ is the gradient operator, ρ l is the density of the water (kg/m 3 ), ρ v is the density of vapour (kg/m 3 ), v l is the velocity of water (m/s), and v v is the velocity of vapour (m/s).
Using Darcy's law and the equation proposed by Philip and Vries [26], the velocities of water and vapour are obtained, respectively: where u l is the pore-water pressure (Pa), γ l is the unit weight of water (N/m 3 ), y is the elevation (m), D atms is the molecular diffusivity of vapour through air, and D atms ¼ 5:893 × 10 −6 T 2:3 =u a with u a ¼ 1atm, v v is a mass flow factor (-), τ v is a tortuosity factor (-), and ∇ρ v is the spatial vapour density gradient.K l in Equation ( 2) is the unsaturated hydraulic conductivity, expressed by the Brooks and Corey [27] Model: in which K ls is the saturated hydraulic conductivity (m/s), θ ls is the saturated water content (-), and η is the shape parameter (-).The van Genuchten [28] model is used to characterize the soil water characteristic curve of soils: where h is the pressure head (m), h d is the scale parameter (m), and n is the shape parameter (-).The volumetric energy balance within unsaturated soils can be expressed [25]: where T is the temperature (K), L is the latent heat of vaporisation (J/kg), ϕ is the soil porosity (-), S a is the saturation degree of pore air (-), and H c is the unsaturated soil's heat capacity at a reference temperature T r , in J/m 3 /K, C pl and C pv are the specific heat capacities of water and vapour, respectively, in J/kg/K, and λ T is the thermal conductivity for the unsaturated soil (W/m/K).
The heat capacity of the unsaturated soil can be obtained as below: where C ps is the specific heat capacities of the solid (J/kg/K), S l is the degree of saturation of water (-), and ρ s is the density of the solid (kg/m 3 ).Moreover, the thermal conductivity for the unsaturated soil can be obtained based on the soil's components as follows [29]: where λ s , λ w , and λ a is the thermal conductivity corresponding to the solid, water, and air, respectively, in W/m/K, and χ s , χ w , and χ a is the volume fraction corresponding to the solid, water, and air, respectively, which can be calculated by the following equations: 2.2.Boundaries for GSHP Systems 2.2.1.Ground Surface Boundary.For a GSHP system coupled with horizontal ground loops, its performance is significantly affected by the heat and moisture exchanges at the ground surface [30].The energy balance equation at the ground surface is as follows [6]: where F NE is the net radiant energy flux absorbed or emitted at the ground surface (W/m 2 ), F absorbed SW is the absorbed shortwave radiation flux (W/m 2 ), F absorbed LW is the absorbed longwave radiation flux (W/m 2 ), F emitted LW is the emitted longwave radiation flux (W/m 2 ), F SEN is the sensible heat flux (W/m 2 ), and F LE is the latent heat flux (W/m 2 ).The sensible heat flux Journal of GeoEnergy is the transfer of heat caused by the difference in temperature between the ground surface and the air.The latent heat flux is the heat moved by the water evaporation.
The moisture balance at the ground surface is presented below [31]: in which M NM is the net moisture flux at the ground surface (kg/m 2 /s), P is the rainfall (kg/m 2 /s), E AE is the actual evaporation flux (kg/m 2 /s) and R RO is the run-off (kg/m 2 /s).The expression of each term in Equations ( 10) and ( 11) are detailed in [30].
Five climatic variables are needed to determine the ground surface boundary in terms of energy and moisture transfer of the coupled TH model, and they are ambient air temperature, shortwave solar radiation, air relative humidity, wind speed, and rainfall.

Ground Heat Exchanger Boundary.
With the circulation of fluid in the GHE, there is the following relationship between inlet fluid temperature and outlet fluid temperature: where T f ; i is the inlet fluid temperature (K), T f ; o is the outlet fluid temperature (K), C pf is the specific heat capacity of the fluid (J/kg/K), ρ f is the fluid density (kg/m 3 ), and r f is the fluid flow rate (m 3 /s), and Q GHE is the thermal load of the GHE, in W (positive in heating mode, negative in cooling mode).
The GHE in the model is discretized into a series of control volumes with a length of dL, and the fluid temperature in each control volume T f ; j is assumed to be constant.The heat flux at the GHE boundary F GHE can be obtained based on Fourier's law: where T g is the ground temperature adjacent to the control volume j in K, and R is the outer radius of pipes (m), and R res is the thermal resistance of pipes (K/m), which can be calculated as follows: where b is the thickness of pipes (m) and λ p is the thermal conductivity of pipes (W/m/K).

COP of Heat Pump.
The thermal load of the GHE in the heating and cooling modes, respectively, can be calculated based on the building thermal load Q building (W) and the heat pump's COP (-) as follows: The empirical model proposed by Staffell et al. [32] is employed to predict the COP of a heat pump.The model was obtained by fitting the temperature difference and the heat pump COP data taken from industrial surveys and field trials: where ΔT is the temperature difference (K), T hot is the temperature of the supplied hot water to the building (K), and T chilled is the temperature of the supplied chilled water to the building (K).In this study, the values of T hot and T chilled are set as 315.65 and 282.65 K, respectively (namely 42.5°C and 9.5°C, respectively) [33].Therefore, when the outlet fluid temperature increased above 315.65K (42.5°C) or dropped below 282.65 K (9.5°C), respectively, the free heating or free cooling operation would be available, i.e., heat pumps would stop, and the thermal load of the GHE equalled the building thermal load.

Numerical Solutions.
The proposed coupled model is implemented into a thermal-hydraulic-chemical-mechanical (THCM) modelling platform-COMPASS (code of modelling prtially saturated soils) [34].The governing equations can be expressed in terms of pore-water pressure and temperature as follows: where C ll , C lT , C TT , C Tl , K ll , K lT , K TT , K Tl , J l and J T are detailed by Gao et al. [25].
The Galerkin finite-element method [35] is employed to spatially discretize the governing equations of the coupled model, and an implicit mid-interval backward difference time-stepping algorithm is employed for the temporal discretisation.The discretised system of linear equations is solved iteratively using a predictor-corrector algorithm [36] to obtain the ground temperature and pore-water pressure 4 Journal of GeoEnergy distributions.A local time-step is prescribed between two consecutive global time-steps, and the fluid temperature profile is calculated.The fluid circulation can be modelled by repeating the multiple time-steps procedure.Meanwhile, the updated fluid temperature profile is used to update the heat flux at the GHE boundary to obtain the ground thermal behaviour.

Model Validation
The TH model presented in this study has been previously thoroughly validated against the experimental data, including an evaporation monitoring study in the Southern France [37,38], an on-site heating experiment utilising a vertical borehole installed in a three-layered ground [39], and a laboratory heating experiment involving a horizontal GHE in a two-layered ground [40], as well as established numerical solution [39].The numerical results predicted by this model showed great agreement with the experimental data and numerical results.Due to the length constraints, readers interested in further details are referred [30,41].

Model Application
4.1.Experimental Site.The technical performance comparisons are conducted based on an experimental site located in a UK university [25,30].As shown in Figure 1, a GSHP system, which can be coupled with two typical types of GHEs, i.e., horizontal ground loops and vertical boreholes, was constructed to provide heating and cooling for campus buildings.The GHEs were designed to connect in parallel to reduce the thermal interferences between each other when heat was extracted from or injected into the ground.Two buildings (A and B), a student accommodation and a data centre, with different heat demands were each equipped with a water source heat pump and connected to the respective GSHP systems.
4.2.Thermal Load of Buildings.Figures 2(a) and 2(b) illustrate the total heating load from Building A and the total cooling load from Building B, respectively, which are provided monthly.The heating load data were collected and provided by the Estate Office of the university, and the cooling load data were estimated by the consultation and experience.It should be noted that three scenarios of total cooling load were planned, which corresponded to a heating-to-cooling ratio (H : C) of 6 : 1, 2.4 : 1, and 1 : 1, respectively, to investigate the effects of heating-to-cooling ratios on the thermal behaviour of the GSHP system.In this study, the connections between the heat pumps and the network were not modelled.The total thermal load applied to the GHEs was the sum of the regulated building thermal loads from these two buildings based on Equations ( 15) and ( 16).

D Domains
4.3.1.Horizontal Ground Loops.Based on the available land area near the two buildings, as well as to minimize the thermal interferences between the GHEs, the configurations of the GHEs in the GSHP system were designed.For the GSHP system coupled with horizontal ground loops, it was comprised of 200 U-tube ground loops buried at 3.0 m below the ground surface (H = 3.0 m).A representative unit of the horizontal ground loops is illustrated in Figure 3.The length (L) and the spacing (S) of each leg of the U-tube ground loops were 200.0 and 1.0 m, respectively.The outer radius (R) of  Journal of GeoEnergy the ground loop pipes was 0.02 m and the pipe thickness (b) was 0.003 m.The thermal conductivity of pipe material (λ p ) was 0.45-W/m/K.Along the depth of the 3D domain (D = 4.0 m), three different soil layers were identified based on the borehole logs from the British Geological Survey [42].The corresponding soil parameters are given in Table 1.
To enhance computational efficiency while maintaining accuracy, ground surface temperature, and pore-water pressure 6 Journal of GeoEnergy have been prescribed at the domain surface, which were determined by the local climatic conditions in the year of 2019, including ambient air temperature, shortwave solar, air relative humidity, and wind speed [25].It should be noted that the variations in the climatic conditions are not considered for the 5-year-long simulations in the current study.At the bottom of the domain, a fixed ground temperature (T g ) of 12.0°C was applied based on the literature [43] and a saturation of 0.75 were assumed considering the local shallow groundwater level.Pure water with an initial temperature (T f ) of 12.0°C was adopted as the heat carrier fluid, which entered from the right inlet and exited from the left outlet.A constant difference of 4°C between the outlet and inlet fluid temperatures was set.In addition, the specific heat capacity (C pf ) and the density ( ρ f ) of pure water were taken as 4180.0J/kg/K and 1000.0 kg/m 3 , respectively.

Vertical
Boreholes.The GSHP system coupled with 200 vertical boreholes was also designed.These vertical boreholes were evenly distributed in the experimental site with a spacing of 8.0 m between each other and were arranged in a regular hexagon layout.the length (L) of each vertical borehole was 210.0 m.Consequently, the total length of the two types of GHEs is comparable in the two systems.These vertical boreholes passed through three soil/ rock layers, which can be considered saturated without no groundwater flow based on the local borehole logs.Table 2 gives the material parameters for geologic materials and the grout.Moreover, the same size U-tube pipe as in the horizontal ground loops was installed in each vertical borehole with a diameter (P) of 0.14 m, and the spacing between the two legs of the U-tube pipe (S) was 0.06 m.It should be pointed out that these vertical boreholes were buried 1.2 m below the ground surface to avoid the influence of varied ground surface temperature due to the local climatic conditions, therefore, a fixed temperature (T) of 10.62°C corresponding to the undisturbed average ground temperature was prescribed on the top of the domain.Owing to the depth of the vertical boreholes, a temperature gradient (∇T) of 25.0°C/1,000 m was considered based on the literature [44], and hence, a constant ground temperature of 15.87°C was determined at the bottom of the domain.Pure water also circulated through each vertical borehole.It was assumed, a constant difference of 4°C between the outlet and inlet, the same as in the horizontal ground loops, ensuring the temperature differences in the warm and cold pipes in both GSHP systems are the same.

Results and Discussion
5.1.Horizontal GSHP System.Simulations of the GSHP system coupled with horizontal ground loops under various heating-tocooling ratios were conducted for 5 years, which is long enough to observe the long-term patterns.As shown in Figures 5 and 6, the results of fluid temperatures and each heat pump's COP This observed cyclic behaviour was expected owing to the annually prescribed temperature and pore-water pressure on the ground surface and the fixed ground temperature and saturation at the bottom of the domain.10

Journal of GeoEnergy
As demonstrated in Figure 5, the amplitude of the simulated fluid temperature increased with the decrease in the heating-to-cooling ratio, and a fixed difference of 4°C between the inlet and outlet fluid temperatures can be observed.For the scenarios of a heating-to-cooling ratio of 6 : 1 or 2.4 : 1, the inlet fluid temperature would quickly decrease below 0°C, namely the freezing point of pure water, on the 19 th day and 24 th day of the 1st year, respectively; and starting from the 2nd year, the inlet fluid temperature would always be lower than 0°C in the first 3.5 and 4 months of every year, respectively, implying the malfunction of the horizontal GSHP system.In comparison, when more heat was injected into the ground, i.e., the scenario of a heatingto-cooling ratio was 1 : 1, the lowest fluid temperature occurred at the inlet and increased to −0.83°C, and the situation when the fluid temperature was lower than 0°C would last for 20 days in February of each year.Meanwhile, since the heat would be easily dissipated in the shallow ground or fast replenished by the adjacent ground, the fluid temperature would remain dynamically stable on a yearly basis.
Figure 6 shows the COPs of two heat pumps for the heating and cooling modes, respectively.It should be noted that the heat pump's COP depends on the temperature difference between the outlet fluid temperature and the temperature of the supplied hot/chilled water to the building.The heat pump under heating mode would operate for 5 years due to the exclusion of free heating, the range of its COP varied from 3.62 and 6.53 under three heating-to-cooling ratios (Figure 6(a)).As the outlet fluid temperature was always below the temperature of the supplied hot water (42.5°C) as shown in Figure 5(b), the lower the heating-tocooling ratios, the smaller the temperature difference, and the higher the COP under the heating mode.
In contrast, as the free cooling could happen in the three heating-to-cooling ratio scenarios (Figure 5(b)), a discontinuity in the heat pump's COP can be found in Figure 6(b).The COP for the heat pump running in cooling mode changed from 5.46 to 7.77.A higher outlet fluid temperature than the temperature of the supplied chilled water (9.5°C) would lead to a lower COP under the cooling mode.Overall, the decreasing heating-to-cooling ratios (more heat injected into the shallow ground) would improve the COP of the heat pump under heating mode, while having opposite effects on the COP of the heat pump under cooling mode.5.2.Vertical GSHP System.Simulations of the GSHP system coupled with vertical boreholes under various heating-tocooling ratios were carried out for 6 years to clearly exhibit the changing patterns.The results of fluid temperatures and two heat pumps' COPs under the three heating-to-cooling ratios are plotted in Figures 7 and 8, respectively.
As shown in Figure 7, a fast decline in the fluid temperatures can be observed in the 1st week of the 1st year regardless of the heating-to-cooling ratios due to the high-heating demands in January.The status of the inlet fluid temperature that was lower than 0°C would remain for 2-3 months in the 1st year in different heating-to-cooling ratio scenarios, while the fluid temperature would gradually increase above the freezing point of water mainly due to the increased cooling demand.Therefore, choosing a suitable heat carrier fluid with a low-freezing point was of importance to this system, especially to ensure its performance in the 1st year.In the long-term, a general downward trend in the fluid temperatures can be clearly noted when the heating-to-cooling ratio equalled 6 : 1 or 2.4 : 1 in the vertical system rather than a steady annual cycle as noted in the horizontal system.In comparison, when the heating-to-cooling ratio decreased to 1 : 1, i.e., more heat was injected into the ground each year through vertical boreholes, and an overall upward trend in the fluid temperature was generated.
Figures 8(a) and 8(b) illustrate the COP of the heat pump in heating mode and cooling mode, respectively.For the heat pump operating in the heating mode (Figure 8(a)), when the heating-to-cooling ratio was 6 : 1 or 2.4 : 1, the COP decreased year by year owing to the increasing difference between the outlet fluid temperature and the supplied hot water temperature (42.5°C), as shown in Figure 7(b).For example, the highest COP for the case of a heating-to-cooling ratio of 6 : 1 was 4.88 in the 1st year but decreased to 3.71 in the 6th year.In contrast with that, the COP for the case of a heating-to-cooling ratio of 1 : 1 increased year by year, for instance, the lowest COP was 3.71 in the 1st year but rose to 4.69 in the 6th year.
As shown in Figure 7(b), the situation that the outlet fluid temperature was lower than the supplied chilled water temperature (9.5°C) can be observed in all scenarios, indicating the occurrence of free cooling and the discontinuity in the heat pump's COP (Figure 8(b)).In the cases when the heating-to-cooling ratio was 6 : 1 or 2.4 : 1, the duration when the outlet fluid temperature was higher than 9.5°C lasted for a few months.Hence, in these two cases (6 : 1 and 2.4 : 1) the heat pump operating in cooling mode was off most of the time, and a high COP (>7.25) was achieved when the heat pump was running.In contrast, in the heating-to-cooling ratio of 1 : 1 scenario, the increasing temperature difference between the outlet fluid and the supplied chilled water on one hand increases the cooling operational period of the heat pump annually, on the other hand, led to a general downward trend in COP in cooling mode.

Comparisons between the Horizontal and Vertical Systems.
Table 3 compares GSHP systems coupled with horizontal ground loops and vertical boreholes from a technical point of view.As can be seen from the table, the land area of horizontal ground loops was approximately 9.6 times the floor area of the vertical boreholes.Owing to the differences in the soil profiles and temperature boundaries as listed in the table, the fluid temperatures and two heat pumps' COPs under three heating-to-cooling ratios exhibited disparate trends and results.
As shown in Figure 5, an annual periodic pattern in the fluid temperatures and heat pump's COPs regardless of the heating-to-cooling ratios indicated a high-natural heat regeneration potential yet unideal heat storage capacity for the horizontal system.As shown in Figure 7, for the vertical system, a downward trend in the fluid temperatures when the Journal of GeoEnergy heating-to-cooling ratios were 6 : 1 and 2.4 : 1 indicated a lowheat recovery potential of the system, while an upward trend in the fluid temperature when the heating-to-cooling ratio was decreased to 1 : 1 indicated a good heat storage ability of the system.When considering the COPs of the heat pump as shown in Figures 6 and 8, although there are differences in the values of COPs, for both systems, reducing the heating-tocooling ratio would increase the COP of the heat pump in heating mode but would decrease the COP of the heat pump in cooling mode.For example, in the horizontal system, during the heating mode, the lowest COP values were 3.62, 3.75, and 4.00 under heating-to-cooling ratios of 6 : 1, 2.4 : 1, and 1 : 1, respectively.That implies that when the heating-tocooling ratio was decreased from 6 : 1 to 2.4 : 1 and 1 : 1, the lowest COP values in the heating mode were increased by approximately 3.6% and 10.5%, respectively.In contrast, the lowest COP values during the cooling mode were 7.54, 7.13, and 5.46 under heating-to-cooling ratios of 6 : 1, 2.4 : 1, and 1 : 1, respectively.Consequently, the lowest COP values in the cooling mode were reduced by 5.4% and 27.6%, respectively, as the heating-to-cooling ratio was decreased from 6 : 1 to 2.4 : 1 and 1 : 1.The simulated results revealed the significance of the heat injection process on the performance of GSHP systems.

Journal of GeoEnergy
Beyond the current study, it is worth comparing both systems from the economic and environmental perspective, and the life-cycle assessment will be the further research.

Conclusions
This study presents a quantitative investigation comparing the technical performance of GSHP systems coupled with horizontal ground loops and vertical boreholes, respectively, considering various heating-to-cooling ratios (6 : 1, 2.4 : 1, and 1 : 1).Through extensive simulations spanning multiple years, valuable insights into the system dynamics and efficiency have been obtained.
The simulated results reveal significant differences between the horizontal and vertical GSHP systems in terms of fluid temperatures and heat pump COPs.In the horizontal system, the fluid temperatures and COPs reached a steady annual cycle after 2 years, regardless of the heating-to-cooling ratios.However, in the vertical system, there was a general downward trend in the fluid temperatures for heating when the ratios were 6 : 1 or 2.4 : 1, while an overall upward trend was observed for a ratio of 1 : 1.
Quantitatively, the analysis demonstrates that the horizontal system exhibited a better heat recovery potential, with a COP range of 3.62-6.53for heating under three heating-tocooling ratios.In contrast, the vertical system displayed a superior heat storage ability, with increasing fluid temperatures and COPs for heating as the heating-to-cooling ratio decreased.Notably, the COPs for cooling mode in both systems were affected by the heating-to-cooling ratio, with higher ratios leading to longer cooling operational periods and lower COPs.
This study highlights the significance of design considerations and the impact of heating-to-cooling ratios on the performance of GSHP systems.The quantitative analysis provides valuable insights for the system optimisation and decision-making processes in the selection of horizontal or vertical configurations.Future research in this area can build upon these findings to further enhance the efficiency and sustainability of GSHP systems for heating and cooling applications.

FIGURE 1 :
FIGURE 1: A schematic diagram illustrating the main components of the GSHP system (HP means heat pump).

Figure 4
shows the 3D domain for a representative unit of the vertical boreholes.As can be seen from the figure, only half of the domain with a width (W) of 4.0 m is needed for the simulation due to the symmetry of the vertical boreholes.Similar to the horizontal ground loops,

FIGURE 3 :
FIGURE 3: Schematic shows the initial and boundary conditions and 3D discretized mesh for a representative unit of the horizontal ground loops [25].

FIGURE 4 :FIGURE 5 :
FIGURE 4: Schematic shows the initial and boundary conditions and 3D discretized mesh for a representative unit of the vertical boreholes.

1 ðbÞFIGURE 6 :
FIGURE 6: Variations in (a) COP of heat pump for heating load and (b) COP of heat pump for cooling load in the horizontal GSHP system.