- Email: [email protected]

Applied Thermal Engineering 28 (2008) 2295–2304 www.elsevier.com/locate/apthermeng

Thermal performance and ground temperature of vertical pile-foundation heat exchangers: A case study Jun Gao, Xu Zhang *, Jun Liu, Kui Shan Li, Jie Yang Institute of HVAC&GAS Engineering, College of Mechanical Engineering, Tongji University, Shanghai 200092, China Received 21 August 2007; accepted 9 January 2008 Available online 26 January 2008

Abstract To assess the geothermal energy for a district heating and cooling system in Shanghai, China, a case study of ground heat exchangers for a ground-coupled heat pump (GCHP) system is presented in this study. Several types of vertical pile-foundation heat exchangers selected are intercompared to determine the most eﬃcient one. Based on a series of performance tests, experimental data of single pile-foundation heat exchanger are presented. Heat transfer performance is also evaluated by numerical method, which couples heat convection and conduction through water in pipeline, concrete pile and soil. It is further used to investigate ﬁve-year changes in the ground temperatures. Numerical results under two imbalance ratios between cooling and heating load are analyzed to evaluate the potential of geothermal energy in the present application. The present study is aimed to provide guidelines for better design of large-scale GCHP in a district heating and cooling system in Shanghai. Ó 2008 Elsevier Ltd. All rights reserved. Keywords: Geothermal energy; Ground-coupled heat pump; Ground temperature; Pile-foundation heat exchanger; Thermal imbalance ratio

1. Introduction Low-cost energy resources are more and more popular for energy-eﬃciency buildings. Sustainable geothermal energy technologies for heating and air-conditioning in buildings have been very attractive since signiﬁcant development of ground-source or ground-coupled heat pump (GSHP/GCHP) system was achieved in these years [1–6]. Bloomquist [2] indicated that the development of geothermal district heating has been one of the fastest growing segments of the geothermal space heating industry and now accounts for over 75% of all space heating provided from geothermal resources world wide. Ozgener et al. [6] presented some case studies to investigate the thermodynamic aspects in terms of energy and exergy and performance improvement opportunities of three geothermal district heating systems installed in Turkey. Their results are very beneﬁcial to the development of GSHP/GCHP system *

Corresponding author. Tel.: +86 21 65984243; fax: +86 21 65983605. E-mail address: [email protected] (X. Zhang).

1359-4311/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.applthermaleng.2008.01.013

for much improvement of the system design and project implementation can be achieved through the energetic and exergetic assessment. Recently, numerical methods in the GSHP/GCHP system, combined with the experiment in situ, have been widely applied and signiﬁcantly developed [5,7–11]. Esen et al. [11] experimentally studied a GCHP system with horizontal ground heat exchangers installed in a test room and its performance coeﬃcient COPsys based on the measured data. Further, a numerical model for heat transfer in the ground was also developed to determine the temperature distribution in the vicinity of the pipe and good agreement with the experimental results were obtained. As for the experiments for ground heat exchangers, much research [12–15] was found to focus on the thermal response and performance test based on the constant heat rejection rate. This kind of test was aimed to obtain such thermal parameters as the conductivity of whole heat exchanger system including backﬁlls, soil, and the heat exchangers in situ. Therefore, long time of experiment was necessary for the experiment system to achieve a steady-state heat transfer, i.e., the condition of constant

2296

J. Gao et al. / Applied Thermal Engineering 28 (2008) 2295–2304

heat ﬂux. In the cases when thermal performance of diﬀerent types of ground heat exchangers are to be compared, Li et al. [9] suggested a test method with constant inlet temperature and varied heat ﬂux using a large water tank of constant temperature. In such kind of experiment, test time was much reduced, but the thermal conductivity could not be correctly derived. Therefore, it is usually used to assess the thermal performance. As for the research method of ground temperatures, in the cases when long-period ground temperatures and the potential of geothermal energy in operation require to be investigated, numerical methods appear to be good alternatives to the experiments on the premise of being well veriﬁed and validated. Many recent studies have focused on the performance of diﬀerent ground heat exchangers and the system. Recently, utilization of the pile foundations of buildings as ground heat exchangers have attracted much attention for it reduces the cost [16–19]. It is often regarded as an energy pile or heat exchanger pile system in the literature and several types can be found in practical applications. Large amount of research on the ground heat exchangers and the heat pump systems can be found. However, there are few practical examples concerning the evaluation of pile-foundation heat exchangers and the underground ﬁeld performance, especially for the large-scale applications. In this present work, a case study is presented to assess the geothermal energy for a district heating and cooling system in Shanghai, China. First, the experimental setup and numerical method are introduced three-dimensional numerical simulations, coupling heat convection and conduction through water in pipeline, concrete pile and soil, are performed to determine the most eﬃcient type of pile-foundation heat exchangers. Experimental data are used to validate the numerical results. Second, numerical method is further applied to investigate ﬁve-year variations of the ground temperatures. The potential of geothermal energy and the operating performance of ground heat exchanger selected are analyzed using the modiﬁed energy output based on the practical ground temperatures in operation. Results of the present study are to be used by a practical application to a district heating and cooling system in Shanghai, China, based on which river water will be used as a supplementary source for an eventual GCHP system. 2. Selection on ground heat exchangers 2.1. Project introduction and pile-foundation heat exchangers An energy resource scheme for the district heating and cooling system and a report of engineering geological survey for the GCHP system were completed in 2006, based on which a group of 5500 pile foundations in a land parcel of 100 m 1000 m, cast-in situ and made from concrete, was used as energy piles. These energy piles will be operated in a GCHP system and are designed to take about 30% thermal load of the district heating and cooling sys-

tem. The potential of geothermal energy are to be discussed and supplementary source is to be provided by river water, which has been adequately calculated and tested last year. According to the climate data of Shanghai, space cooling season is designed from May to September and space heating season from December to next February. Depth of frozen earth in Shanghai is about 8 cm and soil temperature under 5 m depth stays almost constant. The pile-foundation heat exchangers, length 25 m, are then vertically laid at the depth of 5 m. Cast-in situ concrete bearing piles are used and heat exchange is performed by taking advantage of the inner portion of piles. Outer meter of the piles is 600 mm. The thermal medium is water, which ﬂows in the high-density polyethylene (HDPE) pipes cased in the piles. Four types of underground heat exchangers investigated and their sizes are shown in Fig. 1. In consideration of cost reduction of in situ experiment and avoiding the large pressure loss and water ﬂow rate, only the U and W types are applied. The water pipe arrangement is, however, limited by the structure design and construction work. It is necessary for the long pipeline to be ﬁrmly attached to the steel frame (see Fig. 2). Properties of concrete, HDPE and soil are listed in Table 1. 2.2. Experimental data Performance experiment of four types of ground heat exchangers have been conducted in situ. The view of the pile-foundation heat exchangers and experimental system are shown in Figs. 2 and 3, respectively. Water was supplied into the PE pipes and its temperature was stabilized at about 35 °C by a water tank of constant temperature and two electric heaters (see Fig. 3). Real supply temperature and return temperature were measured by platinum resistance thermometers with A-class PT100 sensor, whose precision is 0.15 °C and the errors during experiment are less than 1.0%. Volumetric ﬂow rate was measured by a turbine ﬂowmeter LWGY-10, whose precision is 0.005 m3/h and the errors during experiment are less than 0.5%. To observe the heat transfer performance of the four types of ground heat exchangers, dynamic measurements were carried out and results were obtained when all parameters came to be stable. Table 2 provides the stable results of water temperature and the performance of the four ground heat exchangers investigated. Water supply temperature is approximately 35 °C and the ﬂow rate is controlled at three levels: 0.342 m3/h, and its double, tripe. Energy output from the ground heat exchanger is calculated by the ﬂow rate and temperature diﬀerence, and the heat transfer coeﬃcient is derived by the energy output and the average temperature diﬀerence between water and soil. Soil temperature under 5 m depth stays almost constant and it is 18.2 °C according to the measurements. Experimental data provide the maximum value of energy output, and some percentages will be adopted as the design load. The results in Table 2 will be discussed in the latter numerical study. The most eﬃcient type of

600mm 210mm

300mm 100mm

210mm

100mm

100mm

210mm

600mm

100mm

600mm

2297

120°

J. Gao et al. / Applied Thermal Engineering 28 (2008) 2295–2304

concrete pile

soil

HDPE pipe

W-shaped type

Single U-shaped type

Double U-shaped type

Triple U-shaped type

Fig. 1. Four types of underground heat exchangers investigated (pipe inner diameter, d = 2 cm).

Fig. 2. View of pile foundations before casted and buried.

Table 1 Physical properties of materials

Fig. 3. Experimental system.

Material

Thermal conductivity (W/m K)

Density (kg/m3)

Heat capacity (J/kg K)

Soil (sandy silt) Concrete HDPE

1.3 1.628 0.42

1847 2500 1100

1200 837 1465

pile-foundation heat exchanger is to be decided based on the experimental data and numerical results. 2.3. Numerical method As for numerical study on the solid–ﬂuid conjugated heat transfer encountered in the ground pile-foundation heat exchangers, the mass, momentum, turbulence and energy conservation of water ﬂow, and the energy conservation of soil and concrete piles are considered. The stan-

dard j–e model together with the standard wall functions [20] are employed for the water ﬂow and heat transfer trough the pipes. This traditional standard method should be applicable for the present engineering application of Re (8000–16,000), which can be recognized as fully turbulent pipe ﬂow. Conductive diﬀerential equation is used to model the heat transfer through the piles and soil. The pressure–velocity coupling in the ﬂow region is achieved by using the SIMPLEC algorithm [21]. The second-order upwind diﬀerencing scheme [22], which computes the quantities using a multidimensional linear reconstruction approach and achieves higher-order accuracy through a Taylor series expansion of the cell-centered solution about the cell centroid, is used to evaluate the advection terms for the Navier–Stokes equations, the energy equation and the turbulent transport equations. In Shanghai, groundwater

2298

J. Gao et al. / Applied Thermal Engineering 28 (2008) 2295–2304

Table 2 Performance of underground heat exchangers from experimental data Types

Supply temperature tin (°C)

Return temperature tout (°C)

Temperature diﬀerence Dt (°C)

Water ﬂow rate L (m3/h)

Energy output Q1 (W/m)

Heat transfer coeﬃcient K (W/m °C)

W-shaped W-shaped and double ﬂow rate Single U-shaped Double U-shaped Triple U-shaped

35.02 34.79 35.13 35.08 34.88

29.88 31.88 31.56 32.30 32.63

5.14 2.91 3.57 2.78 2.25

0.342 0.342 2 0.342 0.342 2 0.342 3

83.05 94.25 57.84 89.53 108.07

5.840 6.230 3.891 5.780 6.947

Mean ground temperature, 18.2 °C.

velocity in the soil is ranged from 3.65 m to 10.95 m/yr, so the eﬀect of groundwater advection is neglected. The ﬁnite volume method is applied to discretize the equations in the space. Sizes of a 3D soil zone used to study the heat transfer performance of diﬀerent ground heat exchangers are 5 m 5 m 35 m (depth). Unsteady simulation is performed to achieve the ﬁnal stable temperature variation along the water ﬂow and the return temperature. In accord with the experiments in situ, simulation is performed with a time period of 3 h. Stable results are obtained within the period. Two-order implicit scheme using the central diﬀerencing algorithm is applied for the unsteady simulation, and a varied time-step scheme of initial step 0.001 s, later 0.05 s and ﬁnally 0.2 s is used. In each time step, 20 iterations are speciﬁed. Boundary conditions are speciﬁed as the following. Inlet velocity: uin = 4L/pd2, i.e., 0.3 m/s for all the type cases in Table 2 except the W-shaped type with double ﬂow rate and 0.6 m/s for the W-shaped type with double ﬂow rate. Inlet temperature: tin, Refer to the values of inlet temperature measured in Table 2. Inlet turbulence: turbulent intensity are calculated through 0.16Re1/8, i.e., 4% and 5% for the W-shaped type with double ﬂow rate and the rest, respectively; hydraulic diameter is 0.02 m. Outlet: single direction outﬂow. Soil boundary: vertical surfaces and top surface of the soil zone are assumed adiabatic. Bottom surface is 18.2 °C. Initial temperature of the soil: 18.2 °C. Physical properties are provided in Table 1. Grid-independent solution is tested using three levels of non-structured grid sizes and the eventual grids are described in Fig. 4 (take the triple U-shaped type as an example). Meshes are gradually reﬁned from the outer soil to pile and intraductal ﬂow. In the grid-independent test, coarse grid case has 6 and 10 equal segments for the perimeters of HDPE and pile, respectively, moderate case 12 and 20 segments, and ﬁne case 16 and 30 segments. The soil boundary is divided into the same number of segments for the three sets of grid. Non-structured tetrahedral solid mesh with the increment less than 2.0 was used. Slight diﬀerences of the results are found in the three sets of grid size, so the moderate grid case (i.e., grid size of about 5 mm and 90 mm for the perimeters of

a

b

Z

X

Y

Fig. 4. Meshes used for the numerical simulation (88,000 nodes for the triple U-shaped): (a) view of the total mesh and (b) top view of the mesh in and around the pile.

HDPE and pile, respectively) is further used for all the cases in Table 2. In fact, compared to the solid grid size, the ﬁrst grid size near the inner wall of pipe can be more important for the numerical results. It is then well controlled in this study to ensure the ﬁrst grid point in the fully turbulent region for the applicability of the loglaw, i.e., the non-dimensional ﬁrst grid size y+ be located at 11.225–60. It is realized by doubling the mesh density with the staggered mesh technique near the interior surface when y+ is checked to be larger than 60 during calculation. Thus, y+ is not smoothly distributed along the surface and a singe grid size is diﬃcult to be presented. However, a total of 88,000 nodes are ﬁnally adopted for the triple U-shaped and the averaged y+ is 36.3. The averaged values of y+ are around 35–45 for other cases in Table 2. Stopping criteria for iterative calculations in each time step and the whole time-marching simulation of the discretized equations are as follows: 104 for the continuity equation, 107 for the energy equation, and 103 for the rest equations. If local convergence in each time step is not achieved, the total 20 iterations are performed. Actually, this convergence is obtained after certain number of steps. Convergence error in the present study is thus well controlled through the varied and small time step, and the criterion set for the continuity equation. Overall energy balance is also tracked to verify the numerical solution.

J. Gao et al. / Applied Thermal Engineering 28 (2008) 2295–2304

2.4. Numerical results and type selection

triple U-shaped type have double and triple ﬂow rate, respectively. However, their absolute energy output is only 28% and 56% larger than that of the W-shaped type with the reference ﬂow rate, respectively. Absolute energy output of the W-shaped type with double ﬂow rate is only 10% larger than with reference ﬂow rate, while, as for the W-shaped type with the reference ﬂow rate, energy output is 43% larger than that of the half ﬂow rate. Comparatively, it is decided that the W-shaped with the reference ﬂow rate is regarded as the most eﬃcient type and is employed by the project. In addition, the W-shaped type is preferred because it is easier to adjust the energy output to larger extent than the U-shaped type. As shown in Fig. 6, water temperature declines along the ﬂow direction. It seems that the change of water temperature is linear along the pipe. However, slight changes of the slope observed indicate that the temperature does not decline uniformly. The reason is that water temperature descending will reduce the heat transfer potential from water in pipe to the soil and thus lead to slower descending of water temperature. Therefore, water temperature descending is slightly decreased along the ﬂow direction. It is also seen from Fig. 6 that water return temperature simulated agrees well with the experimental data in Table 2.

Numerical results are compared with the experimental data and are used to evaluate to heat transfer performance of pile-foundation heat exchangers investigated. Energy output and heat transfer coeﬃcient are calculated by the following correlations, respectively. Q ¼ cqLðtin tout Þ=l Q K¼ tin þtout 0:02plp 2 18:2

2299

ð1Þ ð2Þ

where, l is the depth of pipeline and lp is the total length of the pipeline. As shown in Fig. 5, good agreement of energy output and heat transfer coeﬃcient between the numerical and experimental results is achieved. To investigate the eﬀect of water ﬂow rate on the heat transfer, an extra case, W-shaped type with half ﬂow rate 0.342 m3/h 0.5, is also calculated. From the numerical results and experimental data, absolute energy output of the triple U-shaped type is the largest among all heat exchangers, followed by the double U-shaped, W-shaped with double ﬂow rate, W-shaped with reference ﬂow rate, U-shaped and W-shaped with half ﬂow rate. Based on the reference of 0.342 m3/h, the double and

8

140

Experimental data 7

Numerical results

120 6

2

Q1 (W/m)

5

K (W/m K)

100 80

4

60

3

not measured

40 20 0

2 1 0

W-shaped

W-shaped and double flow rate

W-shaped and SingleU-shaped Double U-shaped Triple U-shaped half flow rate

Fig. 5. Numerical results of heat transfer performance and the comparisons with experiments.

(a) W-shaped type double flow rate

35

35

34

34

34

33

33

33

32

32

32

31

31

31

30

30

30

29

29

29

28

28

28

36

single U

double U

Triple U

35 34

o

35

t ( C)

o

t ( C)

half flow rate

(b) U-shaped type reference flow rate

33

0 5 10 15 20 25 5 10 15 20 25 5 10 15 20 25

z (m)

32 0 5 10 15 20 25 5 10 15 20 25 5 10 15 20 25

z (m)

Fig. 6. Axial temperature variation along water ﬂow in pipes.

2300

J. Gao et al. / Applied Thermal Engineering 28 (2008) 2295–2304 Table 3 Design load and energy output for a single pile (W-shaped, reference ﬂow rate)

3. Simulation of ground temperatures 3.1. Geometry and boundary conditions A geometrically periodic unit of the pile group discussed in Section 2.1 is presented in Fig. 7, which consists of eight engineering piles (W-shaped with the reference ﬂow rate) and four lattice piles. In the total area of 100 m 1000 m for the pile group, a corner of 10 unit 10 unit is used to ﬁgure out the ground temperature distribution and variation. It is implemented through using the simpliﬁed symmetrical boundary conditions for two interior faces of the corner and using soil’s natural temperature as the boundary conditions for two exterior faces. To incorporate the eﬀect of surrounding soil, this corner, 55 m 55 m, is enlarged to an area of 75 m 75 m and the two exterior faces are therefore moved outward 20 m (see Fig. 7). As for the long-period simulation of ground temperatures by the group of pile-foundation heat exchangers, a twodimensional volume-control method is used. Therefore, the vertical heat transfer is neglected. A consequence of this assumption is that the detailed ground temperatures at different depth can not be obtained. Owing to the small diﬀerence in the mean water temperature at diﬀerent depth (water temperature varied approximately linear along the pipeline and water ﬂow direction is recirculated), small discrepancy between the assumed and real cases exists. At least, this assumption provides the averaged results of the real 3D case. In addition to the boundary conditions discussed above, speciﬁed heat ﬂux is fed to the exterior surface of the water pipes without interior water ﬂow. The values of heat ﬂux are determined by experimental data of energy output in Table 2 which are regarded as the maximum energy output. Design load is decided by a percentage of the maximum value. Table 3 presents two design loads (65% and 55% of the maximum value, respectively) and the corresponding total energy output under two different imbalance ratios between cooling and heating season.

Thermal imbalance ratio

Case 1: 10% Case 2: 3%

Total energy output (MJ)

Design load (W/m)

Total energy output (MJ)

54.0 45.7

8748 7395

81.0 73.8

7873 7173

Simulations for the two cases of diﬀerent thermal imbalance ratio are carried out under the design loads and conditions above. Based on the initial value of 18.2 °C, ﬁve-year ground temperature distributions are solved numerically. Mean ground temperature of every month and its variations in the ﬁrst year for the two cases are shown in Fig. 8. It is observed that mean temperature rise due to the pile-foundation heat exchanger is mainly determined by the thermal imbalance ratio and the highest temperature in the year by the design load of cooling season. For case 1-thermal imbalance ratio 10%, ground temperature ﬁelds are presented by Figs. 9 and 10 for cooling and heating seasons, respectively. Fig. 9 shows the temperature increasing during the cooling season from May to September. Obviously, high temperature occurs in the vicinity of energy piles. Averaged temperature rises up to 28.7 °C at the end of cooling season, which means the eﬃcient energy output from the energy piles are gradually reduced during the cooling operation of the GCHP system. This eﬀect is somewhat recovered in the subsequent transitional season (October and November) and is, however, very beneﬁcial to the succeeding heating season. It exhibits the energy shift for use from summer to winter. Fig. 10 presents the initial ﬁeld at the beginning of heating season and the temperature decreasing during the heating season from December

lattice pile

engineering pile

75 55

2750

Design load (W/m)

engineering pile

11

2750

89 38

11000 2750 2750

Heating season (3 months, December–February)

3.2. Results and analysis

11

lattic pile

Cooling season (5 months, May–September)

2750

2750 2750 11000

2750 55

unit (mm)

75

Fig. 7. View of a periodic unit of pile group and a corner of 10 10 units used.

unit (m)

J. Gao et al. / Applied Thermal Engineering 28 (2008) 2295–2304

23

o

t ( C)

22

cooling season

20

heating season

21

cooling season

22

21 20 19

19

heating season

23

transition season

24

transition season

24

transition season

25

transition season

(b) Case2

25

o

t ( C)

(a) Case1

2301

18

18 5

6

7

8

9

10 11 12 1

2

3

4

5

month

5

6

7

8

9

10 11 12

1

2

3

4

5

month

Fig. 8. Variations of mean ground temperature in the ﬁrst year of running.

Fig. 9. Ground temperature ﬁelds during cooling season at thermal imbalance 10%: (i) at the end of May; (ii) at the end of June; (iii) at the end of July; (iv) at the end of August; (v) at the end of September.

to next February. Soil temperature in the vicinity of energy piles descends signiﬁcantly during the three months of heating operation and total averaged temperature drops down to 19.1 °C, which means an elevated mean temperature of 0.9 °C in the ﬁrst year of operation at thermal imbalance 10%, thus the cooling capacity of the pile-foundation heat exchangers willed be reduced about 6%. More information of the ground temperatures and the reduced energy output are obtained from the ﬁve-year simulations. Using the design load for the two cases, total elevated mean temperature before cooling season is 2.77 °C and 0.81 °C for the two cases, respectively, and total elevated mean temperature after cooling season is 8.39 °C

and 6.02 °C, respectively. Fig. 11 presents the annual mean ground temperature before and after the cooling season, from which the potential of geothermal energy is very encouraging. Furthermore, the highest temperature in the ﬁeld for both cases is found much less than 32 °C, so the so-called thermal screen does not occur after ﬁve years and the GSHP/GCHP system can proceed with its stable running. Due to the ground temperature rise, the design energy output of heat exchangers may be reduced. It is difﬁcult to precisely derive the real energy output from the soil. However, the mean temperatures throughout the whole cooling or heating season can be used to evaluate the potential of geothermal energy.

2302

J. Gao et al. / Applied Thermal Engineering 28 (2008) 2295–2304

Fig. 10. Ground temperature ﬁelds during heating season at thermal imbalance 10%: (i) initial ﬁeld at the beginning of December; (ii) at the end of December; (iii) at the end of January; (iv) at the end of February.

Case 1

Case 2

Case 1

28

Case 2

1.05 heating season

1.00

After cooling season 26

Modified load

0.95

o

t ( C)

24

22

Before cooling season

0.90 0.85 0.80

20 0.75 cooling season

18

0.70 1

2

3

4

5

year

1

2

3

4

5

year

Fig. 11. Annual mean ground temperature.

Fig. 12. Modiﬁed load using the averaged temperatures.

Fig. 12 shows the ratio of modiﬁed load to the design load according to the averaged mean temperatures which are obtained by averaging the mean ground temperature in cooling and heating season, respectively. The modiﬁed load is deﬁned using the averaged mean ground temperature, mean water temperature (7.5 °C in heating season) and heat transfer coeﬃcient (K assumed not varied). From Fig. 12, it is observed that the real energy output from the soil is signiﬁcantly inﬂuenced by the ground temperature variation. Owing to the thermal imbalance, averaged ground temperature of cooling, transition and heating sea-

sons keeps increasing during the ﬁve years. Therefore, the modiﬁed load in cooling season tends to decrease, while that in heating season tends to increase. The GSHP/GCHP system related to present ground heat exchangers can apparently disregard the potential of geothermal energy in heating season. As for the potential of geothermal energy in cooling season, larger thermal imbalance leads to faster decreasing in the energy output from the soil. In the present case study, design load is really reduced by 20–30% in the ﬁve years, i.e., 38–43 W/m output for case 1 and 35–37 W/m for case 2. However, as a conservative

J. Gao et al. / Applied Thermal Engineering 28 (2008) 2295–2304

parameter, the modiﬁed load is not to directly indicate a sharp drop of the potential of geothermal energy in cooling season. As a matter of fact, the moderate value of elevated mean temperature conﬁrms that larger design load in cooling load can be applied on the basis of small thermal imbalance ratio. Another reason is that the ground temperature is derived based on the design load and it is then higher than under the modiﬁed load. Therefore, the current modiﬁed load is a conservative parameter and the real potential of geothermal energy is certainly higher than that indicated in Fig. 12. 4. Conclusions Geothermal energy utilization by GSHP/GCHP system for a district heating and cooling system in Shanghai is introduced in this study. Pile-foundation ground heat exchangers are investigated and its most eﬃcient type is decided based on the 3D ﬂuid–solid coupled numerical simulations and corresponding experimental data. Five-year simulations of the ground temperatures for the practical pile group are performed to evaluate the potential of geothermal energy in the present application. Some concluding remarks are extracted from this study as follows: (1) In the present practical application, both experimental data and numerical results of heat transfer performance demonstrate that the W-shaped type of pile-foundation ground heat exchanger with moderate medium ﬂow rate appears to be most eﬃcient and is ﬁnally applied. (2) Water temperature descending in the pipe reduces the heat transfer potential from water in pipe to the soil and it tends to slightly decreased with the ﬂow direction. (3) Under thermal imbalance ratio of 10% and 3%, total elevated mean ground temperature of ﬁve-year running is 2.77 °C and 0.81 °C, respectively. Design load employed, 65% and 55% of the maximum energy output from experiment, will not produce thermal screen. It conﬁrms that the GSHP/GCHP system can proceed with its stable running. (4) According to the averaged mean ground temperature in cooling and heating season, modiﬁed load is deﬁned to investigate the impact of ground temperature variation on the energy output of ground heat exchangers. As a conservative parameter, the reduced load, 70–80% of the design load, is proved to underestimate the real potential of geothermal energy. Therefore, higher energy output of the pile-foundation heat exchangers can be expected for the practical engineering. As far as the fully turbulent pipe ﬂow is concerned in this work, the most traditional standard k–e model with the standard wall functions is applicable. Certainly, more advanced turbulence model and modern methodology such

2303

as large eddy simulation (LES) can be bale to predict more correct results, if well implemented. A very traditional methodology, but not the best numerical approach have been applied at present. It should be pointed out that the assumption of fully developed turbulence, negligible molecular viscosity, constant-shear and local equilibrium wall treatment could be inapplicable when the ﬂow exhibits in or near the laminar, transitional, low-turbulent region. More advanced numerical method should be considered if the ﬂow pattern appears to be in such region, where the present numerical method may overestimate the water-to-pipe convection and lead to unreasonable assessment of the heat transfer performance of the pile-foundation heat exchangers. Acknowledgements The authors gratefully acknowledgement the ﬁnancial support from the Ministry of Science and Technology of the People’s Republic of China through the Science Research for the 11th Five-year Plan with Grants No. 2006BAJ01B05 and 2006BAJ01A05. This work was also partially supported by the National Natural Science Foundation of China under Grant No. 50578113. The authors are also grateful to the valuable advice of the reviewers on this article and the suggestion of using advanced methodology for the modern research work. References [1] S. Kavanaugh, Ground source heat pumps simple eﬃcient reliable, ASHRAE J. 40 (1998) 31–36. [2] R.G. Bloomquist, Geothermal space heating, Geothermics 32 (2003) 513–526. [3] L. Ozgener, A. Hepbasli, I. Dincer, Performance investigation of two geothermal district heating systems for building applications: energy analysis, Energ. Build. 38 (2005) 286–292. [4] J. Spitler, Ground-source heat pump system research – past present and future, HVAC&R Res. 11 (2005) 165–167. [5] W.B. Yang, M.H. Shi, H. Dong, Numerical simulation of the performance of a solar–earth source heat pump system, Appl. Therm. Eng. 26 (2006) 2367–2376. [6] L. Ozgener, A. Hepbasli, I. Dincer, A key review on performance improvement aspects of geothermal district heating systems and applications, Renew. Sust. Energ. Rev. 11 (2007) 1675–1697. [7] K. Den Braven, E. Nilson, Performance prediction of a sub-slab heat exchanger for geothermal heat pumps, ASME Trans. J. Sol. Eng. 120 (1998) 282–288. [8] C. Gauthier, M. Lacroix, H. Bernier, Numerical simulation of soil exchanger-storage systems for greenhouses, Sol. Energ. 60 (1997) 333–346. [9] X. Li, Y. Chen, Z. Chen, et al., Thermal performances of diﬀerent types of underground heat exchangers, Energ. Build. 38 (2006) 543–547. [10] O. Ozgener, A. Hepbasli, Modeling and performance evaluation of ground source (geothermal) heat pump systems, Energ. Build. 39 (2007) 66–75. [11] H. Esen, M. Inalli, M. Esen, Numerical and experimental analysis of a horizontal ground-coupled heat pump system, Build. Environ. 42 (2007) 1126–1134. [12] C. Eklof, S. Gehlin, TED – a mobile equipment for thermal response test, Master’s Thesis, 198E, Lulea˚ University of Technology, Sweden, 1996.

2304

J. Gao et al. / Applied Thermal Engineering 28 (2008) 2295–2304

[13] B. Sanner, M. Reuss, E. Mands, J. Mu¨ller, Thermal response test – experiences in Germany, in: Proceedings Terrastock 2000, Stuttgart, Germany, pp. 177–182. [14] S. Signorelli, S. Bassetti, D. Pahud, T. Kohl, Numerical evaluation of thermal response tests, Geothermics 36 (2007) 141–166. [15] K. Lim, S. Lee, C. Lee, An experimental study on the thermal performance of ground heat exchanger, Exp. Therm. Fluid Sci. 31 (2007) 985–990. [16] K. Morino, T. Oka, Study on heat exchanged in soil by circulating water in a steel pile, Energ. Build. 21 (1994) 65–78. [17] L. Laloui, M. Moreni, L. Vulliet, Behavior of a dual-purpose pile as foundation and heat exchangers, Canadian Geotech. J. 40 (2003) 388– 402.

[18] L. Laloui, M. Nuth, L. Vulliet, Experimental and numerical investigations of the behavior of a heat exchanger pile, Int. J. Numer. Anal. Meth. Geomech. 30 (2006) 763–781. [19] Y. Hamada, H. Saitoh, M. Nakamura, et al., Field performance of an energy pile system for space heating, Energ. Build. 39 (2007) 517–524. [20] B.E. Launder, D.B. Spalding, The numerical computation of turbulent ﬂows, Comput. Meth. Appl. Mech. Eng. 3 (1974) 269–289. [21] J.P. Van Doormaal, G.D. Raithby, Enhancement of the SIMPLE method for predicting incompressible ﬂuid ﬂows, Numer. Heat Transf. 7 (1984) 147–163. [22] T.J. Barth, D. Jespersen, The design and application of upwind schemes on unstructured meshes, Technical Report AIAA-89-0366, AIAA 27th Aerospace Sciences Meeting, Reno, Nevada, USA, 1989.