ARTICLE IN PRESS

Renewable Energy 32 (2007) 24612478 www.elsevier.com/locate/renene

Review

Ground heat exchangersA review of systems, models and applications

Georgios Florides, Soteris Kalogirou
Higher Technical Institute, P.O. Box 20423, Nicosia 2152, Cyprus Received 17 October 2006; accepted 23 December 2006 Available online 26 March 2007

Abstract The temperature at a certain depth in the ground remains nearly constant throughout the year and the ground capacitance is regarded as a passive means of heating and cooling of buildings. To exploit effectively the heat capacity of the ground, a heat-exchanger system has to be constructed. This is usually an array of buried pipes running along the length of a building, a nearby eld or buried vertically into the ground. A circulating medium (water or air) is used in summer to extract heat from the hot environment of the building and dump it to the ground and vice versa in winter. A heat pump may also be coupled to the ground heat exchanger to increase its efciency. In the literature, several calculation models are found for ground heat exchangers. The main input data are the geometrical characteristics of the system, the thermal characteristics of the ground, the thermal characteristics of the pipe and the undisturbed ground temperature during the operation of the system. During the rst stages of the geothermal systems study, one-dimensional models were devised which were replaced by two-dimensional models during the 1990s and three-dimensional systems during recent years. The present models are further rened and can accommodate for any type of grid geometry that may give greater detail of the temperature variation around the pipes and in the ground. Monitoring systems have been set up to test various prototype constructions with satisfactory results. r 2007 Elsevier Ltd. All rights reserved.
Keywords: Ground heat exchangers; Ground temperature; Heat exchanger models; Ground source heat pump

Corresponding author. Tel.: +357 22 406466; fax: +357 22 406480.

E-mail address: skalogir@spidernet.com.cy (S. Kalogirou). 0960-1481/$ - see front matter r 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.renene.2006.12.014

ARTICLE IN PRESS
2462 G. Florides, S. Kalogirou / Renewable Energy 32 (2007) 24612478

Contents 1. 2. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Types of ground heat exchangers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Open systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Closed systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Miscellaneous systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ground thermal behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Calculation models and evaluated performance of ground heat exchangers . . . . . . . . . Vertical U-tube ground heat exchangers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Deep geothermal systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2462 2463 2463 2464 2467 2467 2470 2473 2476 2476 2477

3. 4. 5. 6. 7.

1. Introduction Measurements show that the ground temperature below a certain depth remains relatively constant throughout the year. This is due to the fact that the temperature uctuations at the surface of the ground are diminished as the depth of the ground increases because of the high thermal inertia of the soil. Also, there is a time lag between the temperature uctuations at the surface and in the ground. Therefore, at a sufcient depth, the ground temperature is always higher than that of the outside air in winter and is lower in summer. The temperature variation of the ground at various depths in summer (August) and winter (January) is shown in Fig. 1. The graph shows actual ground temperatures as measured in a borehole drilled for this purpose in Nicosia, Cyprus. As can be seen, the temperature is nearly constant below a depth of 5 m for the year round. The difference in temperature between the outside air and the ground can be utilised as a preheating means in winter and pre-cooling in summer by operating a ground heat exchanger. Also, because of the higher efciency of a heat pump than conventional natural gas or oil heating systems, a heat pump may be used in winter to extract heat from the relatively warm ground and pump it into the conditioned space. In summer, the process

30

Temperature (C)

26 22 18 14 10 6 0 5 10 15 20 25 30 Depth in ground (m) 35 40 45 50 25-Jan-05 20-Aug-05

Fig. 1. Temperature variation with depth.

ARTICLE IN PRESS
G. Florides, S. Kalogirou / Renewable Energy 32 (2007) 24612478 2463

may be reversed and the heat pump may extract heat from the conditioned space and send it out to a ground heat exchanger that warms the relatively cool ground. Ground source heat pumps are receiving increasing interest in North America and Europe and the technology is now well established with over 550,000 units installed worldwide and with more than 66,000 units installed annually. About 80% of the units installed worldwide are domestic [1]. 2. Types of ground heat exchangers There are two general types of ground heat exchangers: open and closed [2]. In an open system, the ground may be used directly to heat or cool a medium that may itself be used for space heating or cooling. Also, the ground may be used indirectly with the aid of a heat carrier medium that is circulated in a closed system. The loop of the heat exchanger is made of a material that is extraordinarily durable but allows heat to pass through efciently. Loop manufacturers typically use high-density polyethylene which is a tough plastic, with heat fuse joints. This material is usually warranted for as much as 50 years. The uid in the loop is water or an environmentally safe antifreeze solution. Other types of heat exchangers used directly for heating and cooling utilise a copper piping placed underground. As refrigerant is pumped through the loop, heat is transferred directly through the copper to the earth. The length of the loop depends upon a number of factors such as the type of loop conguration used, the house heating and air conditioning load, soil conditions, local climate and many more. These types of systems are examined separately below. 2.1. Open systems In open systems, ambient air passes through tubes buried in the ground for preheating or pre-cooling and then the air is heated or cooled by a conventional air conditioning unit before entering the building (Fig. 2).

Fig. 2. Basic principle of ground preheating or pre-cooling of air in an open system.

ARTICLE IN PRESS
2464 G. Florides, S. Kalogirou / Renewable Energy 32 (2007) 24612478

Fig. 3. Ground water heat pump.

In a similar way, the ground water of a water-bearing layer may be used as a cooling carrier medium, brought in direct contact with the heat pump coils. In most cases two wells are required, one for extracting the ground water and one for injecting it back into the water-bearing layer as indicated in Fig. 3. 2.2. Closed systems In this case heat exchangers are located underground, either in horizontal, vertical or oblique position, and a heat carrier medium is circulated within the heat exchanger, transferring the heat from the ground to a heat pump or vice versa. Fig. 4 indicates the horizontal type which has a number of pipes connected together either in series or in parallel. This conguration is usually the most cost-effective when adequate yard space is available and trenches are easy to dig. The trenchers have a depth of 12 m in the ground and usually a series of parallel plastic pipes is used. Fluid runs through the pipes in a closed system. A typical horizontal loop is 3560 m long per kW of heating or cooling capacity [3]. Horizontal ground loops are the easiest to install while a building is under construction. However, new types of digging equipment allow horizontal boring and thus it is possible to retrot such systems into existing houses with minimal disturbance of the topsoil and even allow loops to be installed under existing buildings or driveways. In USA, some special ground heat exchangers have been developed for heat pump systems, in which the pipe is curled into a slinky shape (Fig. 5). In this way, it is possible to place more pipes into shorter trenches in order to reduce the amount of land space needed [3]. These collectors are best suited for heating and cooling in places where natural temperature recharge of the ground is not vital. For all horizontal systems in heating-only mode, the main thermal recharge is provided by the solar radiation falling on the earth surface. Therefore, it is important not to cover the surface above the ground heat collector. Vertical ground heat exchangers or borehole heat exchangers, shown in Fig. 6, are widely used when there is a need to install sufcient heat exchange capacity under

ARTICLE IN PRESS
G. Florides, S. Kalogirou / Renewable Energy 32 (2007) 24612478 2465

Fig. 4. Horizontal-type ground heat exchangers (redrawn from Ref. [2]).

Fig. 5. Slinky-type ground heat exchanger.

a conned surface area such as when the earth is rocky close to the surface, or where minimum disruption of the landscape is desired. This is possible because the temperature below a certain depth remains constant over the year (see Fig. 1). In a standard borehole, which in typical applications is 50150 m deep, plastic pipes (polyethylene or polypropylene) are installed, and the space between the pipe and the hole is lled with an appropriate material to ensure good contact between the pipe and the undisturbed ground and reduce the thermal resistance [3].

ARTICLE IN PRESS
2466 G. Florides, S. Kalogirou / Renewable Energy 32 (2007) 24612478

Fig. 6. Vertical ground heat exchangers.

Fig. 7. Common vertical ground heat exchanger designs.

Vertical loops are generally more expensive to install, but require less piping than horizontal loops because the earth deeper down is cooler in summer and warmer in winter, compared to the ambient air temperature. Several types of borehole heat exchangers were tested and are widely used. These are classied in two basic categories as shown in Fig. 7: (a). U-pipes, consisting of a pair of straight pipes, connected with a U-turn at the bottom. Because of the low cost of the pipe material, two or even three of such U-pipes are usually installed in one hole.

ARTICLE IN PRESS
G. Florides, S. Kalogirou / Renewable Energy 32 (2007) 24612478 2467

Fig. 8. Standing column well.

(b). Concentric or coaxial pipes, joint either in a very simple way with one straight pipe inside a bigger diameter pipe or joint in complex congurations. 2.3. Miscellaneous systems A number of ground systems cannot be categorised either as open or as closed. Such a system is the standing column well shown in Fig. 8, where water is pumped from the bottom of the well to the heat pump. The exiting water is percolated through gravel in the annulus of the well in order to absorb heat. Standing wells are typically 15 cm in diameter and may be as deep as 500 m; therefore they are very expensive [3]. Other sources of heat are the use of water in mines and tunnels. This water has a steady temperature the whole year round and is easily accessible. 3. Ground thermal behaviour The use of direct or indirect earth-coupling techniques for buildings and agricultural greenhouses requires knowledge of the ground temperature prole at the surface and at various depths. The ambient climatic conditions affect the temperature prole below the ground surface (Fig. 9) and need to be considered when designing a heat exchanger. Actually the ground temperature distribution is affected by the structure and physical properties of the ground, the ground surface cover (e.g., bare ground, lawn, snow, etc.) and the climate interaction (i.e., boundary conditions) determined by air temperature, wind, solar radiation, air humidity and rainfall. The temperature distribution at any depth below the earth surface remains unchanged throughout the year with the temperature increasing with depth with an average gradient of about 30 1C/km. The geothermal

ARTICLE IN PRESS
2468 G. Florides, S. Kalogirou / Renewable Energy 32 (2007) 24612478

Fig. 9. Energy ows in ground.

gradient deviations from the average value are, in part, related to the type of rocks present in each section. Heat ow, which is a gauge of the amount of thermal energy coming out of the earth, is calculated by multiplying the geothermal gradient by the thermal conductivity of the ground. Each rock type has a different thermal conductivity, which is a measure of the ability of a material to conduct heat. Rocks that are rich in quartz, like sandstone, have a high thermal conductivity, indicating that heat readily passes through them. Rocks that are rich in clay or organic material, like shale and coal, have low thermal conductivity, meaning that heat passes slower through these layers. If the heat ow is constant throughout a drill hole (i.e., water is not owing up or down the hole), then it is obvious that low-conductivity shale layers will have a higher geothermal gradient compared to high-conductivity sandstone layers [4]. Mihalakakou et al. [5] present a complete model for the prediction of the daily and annual variation of ground surface temperature. The model uses a transient heat conduction differential equation and an energy balance equation at the ground surface to predict the ground surface temperature. The energy balance equation involves the convective energy exchange between air and soil, the solar radiation absorbed by the ground surface, the latent heat ux due to evaporation at the ground surface as well as the

ARTICLE IN PRESS
G. Florides, S. Kalogirou / Renewable Energy 32 (2007) 24612478 2469

long-wave radiation. Therefore, the ground surface temperature can be estimated from [6]: T sur t T m As Reeiwt , (1)

where Tm is the mean annual ground surface temperature, As is the amplitude of the temperature wave at the ground surface and w is the frequency of the temperature wave. In order to solve Eq. (1), the following energy balance equation at ground surface was used as a boundary condition equation at the ground surface [7]:  qT sur   K CE LR SR LE; (2) qy y0 where K is the thermal conductivity of the soil, CE is the convective energy exchanged between air and ground surface, LR is the long-wave radiation emitted from the ground surface, SR is the solar radiation absorbed from the ground surface and LE is the latent heat ux due to evaporation. The model is validated against 10 years of hourly measured temperatures for bare and short-grass covered soil in Athens and Dublin. The results are compared with the corresponding results of models using Fourier analysis. Furthermore, a sensitivity investigation is performed to investigate the inuence of various factors involved in the energy balance equation at the ground surface on the soil temperature prole. Popiel et al. [8] present the temperature distributions measured in the ground for the period between summer 1999 and spring 2001. The investigation was carried out in Poznan, Poland, for two differently covered ground surfaces, a bare surface and a surface covered with short grass. Temperatures were measured with thermocouples distributed in the ground at a depth from 0 to 7 m (bare surface) and from 0 to 17 m (short grass). It was found that the short-period temperature variations reached a depth of approximately 1 m. From July to the end of September, from the surface region at ground depth (below about 1.5 m), a heat ux of 3.6 W/m2 was transferred. Usually, the recommended depth for horizontal ground heat exchangers is from 1.5 to 2 m. The measurements also show that during the summer period the ground temperature under the bare surface below 1 m was about 4 1C higher in comparison to the temperature of the ground covered with short grass. Therefore, for the ground cold source, e.g., for the air conditioning application the surface covered with short grass is recommended. However, in winter, the temperature distributions were almost the same. A comparison of the Buggss formula for the ground temperature distribution adapted to the European region of Poznan shows a good agreement with the experimental data. From the point of view of the temperature distribution, Popiel et al. [8] distinguish three ground zones: 1. Surface zone reaching a depth of about 1 m, in which the ground temperature is very sensitive to short time changes of weather conditions. 2. Shallow zone extending from the depth of about 18 m (for dry light soils) or 20 m (for moist heavy sandy soils), where the ground temperature is almost constant and close to the average annual air temperature; in this zone the ground temperature distributions depend mainly on the seasonal cycle weather conditions. 3. Deep zone (below about 820 m), where the ground temperature is practically constant (and very slowly rising with depth according to the geothermal gradient).

ARTICLE IN PRESS
2470 G. Florides, S. Kalogirou / Renewable Energy 32 (2007) 24612478

4. Calculation models and evaluated performance of ground heat exchangers Several calculation models for ground-coupled heat exchangers are found in the literature. Early models generally used a one-dimensional description of the pipe to derive a relation between its inlet and outlet temperature. Tzaferis et al. [9] studied eight models to predict the performance of ground-to-air heat exchangers. The algorithms of the studied models either calculate the conductive heat transfer from the pipe to the ground mass or calculate the convective heat transfer from the circulating air to the pipe. Input data include the geometrical characteristics of the system, the thermal characteristics of the ground and the thermal characteristics of the pipe together with the undisturbed ground temperature during the operation of the system or only the temperature of the pipe surface. The algorithms of the eight models were introduced into computer programs to simulate the behaviour of the ground-to-air heat exchangers. Experimental results were also obtained for a PVC horizontal pipe, buried at a depth of 1.1 m and compared to the calculated values. Six of the eight models gave very close results to the actual values with an rms error in each case of about 3.5%. In the examined models, the thermal capacity of the ground is not considered and therefore the inuence of different pipes on each other and the temperature proles in the ground cannot be studied. Bi et al. [10] used a two-dimensional cylindrical coordinate system to model a vertical double spiral coil ground heat exchanger (GHx). This ground heat exchanger was designed by the authors for a ground source heat pump (GSHP) system. The underground temperature distribution of the coil was solved numerically and the results were compared to measured temperature data. They concluded that the temperature distribution is important to the performance improvement of the GSHP, and especially for the GHx and that the analytical and experimental results prove that the GHx design is reasonable. Mihalakakou et al. [11] present a model in which the ground surrounding the pipe and the pipe itself are described in polar co-ordinates. In this model the temperature and moisture proles of the ground are included in the equations. The inuence of the ground surface temperature is modelled by the superposition of the algebraic solution of the undisturbed temperature eld caused by the surface air temperature and the temperature eld caused by the pipe. The authors show the importance of including the moisture content in the soil. The model is solved in the TRNSYS (a modular energy system simulation program) environment and validated with good results. Mihalakakou et al. [12] investigated the heating potential of a single ground-to-air heat exchanger as well as the potential of a multiple parallel earth tube system. An accurate numerical model was used to investigate the dynamic thermal performance of the system during the winter period in Dublin. The model had been successfully validated against an extensive set of experimental data. The results showed that the heating potential of the system during winter is signicantly important. The obtained results showed that the effectiveness of the ground-to-air heat exchanger increases with an increase in the pipe length (checked range: 3070 m). Also, there is an increase in effectiveness when the pipe is buried in greater depths (3 m instead of 1.2 m). By increasing the pipe diameter from 100 to 150 mm, it was shown that the heating capacity of the system was reduced. This is due to a reduction in the convective heat transfer coefcient and an increase in the pipe surface, therefore, providing a lower air temperature at the pipe outlet. Finally, a higher air velocity in the pipe (checked range: 515 m/s) leads to a reduction of the systems heating capacity, mainly because of the increased mass ow rate inside the pipe.

ARTICLE IN PRESS
G. Florides, S. Kalogirou / Renewable Energy 32 (2007) 24612478 2471

Bojic et al. [13] developed a model in which the soil is divided into horizontal layers with uniform temperature. All the pipes are placed in one layer at the same depth and parallel to each other. The heat transported to the soil by convection from the air and the solar irradiation is calculated. Also an equation describing the heat ow between the airow in the pipe and the neighbouring soil layer is used. All equations used for the soil layers in each time step are steady-state energy equations. This model is a two-dimensional model therefore the inuence that pipes have on each other may not be evaluated. Gauthier et al. [14] describe a fully three-dimensional model. A simple Cartesian coordinate system is used and the round pipes are replaced with square pipes of equivalent areas. The thermophysical properties of the ground are considered constant and temperature independent, but actually the ground may not be homogenous. In this way, the inuence of different layers in the ground, concrete foundations and insulation can be evaluated. The heat transfer caused by moisture gradients in the ground is assumed to be negligible with respect to that caused by temperature gradients. Heat transfer in the pipes is dominated by convection in the axial direction but coupled with the temperature eld in the ground via the boundary condition on the pipe surface. The model is thoroughly validated with experimental data taken from a ground heat exchanger storage system installed in a commercial-type greenhouse. Finally, the various parameters that affect the behaviour of the ground heat exchanger storage system are examined. De Paepe and Willems [15] further rened the above Gauthier et al. [14] approach and the model was used to study the performance of a ground-coupled air heat exchanger in the Belgian climate. A three-dimensional unstructured nite volume model was derived and the FLUENT solver was used to obtain the numerical solutions. The model considers transient and fully three-dimensional conduction heat transfer in the soil and other materials. The heat transfer by moisture gradients in the soil is neglected and the heat transfer in the pipe is dominated by convection. The governing equation for the conduction in the soil may be stated as: qT lr T , (3) qt where r is the density (kg/m3), cp the heat capacity (J/kg K), l the thermal conductivity (W/m K), T the temperature (K) and t the time (s). The boundary conditions for the underground lateral external surfaces of the computational domain are assumed to be adiabatic, thus: rc p qT 0, (4) qn where n is the unit vector normal to the surface. A constant and uniform temperature for the horizontal plane deep underground is imposed. At the ground surface the heat ux from the ambient air to the surface is calculated from: qT hsurr T soil T surr , (5) qz where Tsurr is the temperature of the surrounding air and this can be a constant value or a time-dependent function, and hsurr is the convection coefcient. The results show that the inuence of the pipe on the temperature of the surrounding soil is limited to a distance of twice its diameter. To make optimal use of the thermal l

ARTICLE IN PRESS
2472 G. Florides, S. Kalogirou / Renewable Energy 32 (2007) 24612478

capacity of the soil and to eliminate the inuence of the outside air, the tubes have to be buried below a depth of 2.5 m and the length of the tube can be optimised with the calculation model to obtain an efcient heat exchanger. Hollmuller and Lachal [16] examined the winter preheating and summer cooling potential of buried pipe systems under the central European climate. The simulation model of the air-to-ground heat exchanger used, accounts for sensible and latent heat transfer. The model additionally accounts for frictional losses and water inltration and ow along the tubes. It further allows for control of airow direction as well as for exible geometry (inhomogenous soils, diverse border conditions, use of symmetries or pattern repetitions for run-time economy) and is adapted to TRNSYS. The basic equations of the model describe the mass and energy exchanges between air and tube. The study concludes that in Central Europe there is a fundamental asymmetry between heating and cooling potentials with the ground used as a seasonal energy buffer. In winter, preheating of fresh air acts as a saving function on energy demand to which it is inherently linked by limitation of ow rate. In summer, inertial cooling (smoothening of ambient temperature below comfort threshold) increases along with ow rate and hence becomes an energy-producing service on its own. Air preheating with buried pipes is more expensive than that with fuel (about twice the cost per kWh), which it cannot substitute completely. On the contrary, buried pipe inertial cooling, together with an (avoided) air conditioning system, is competitive and allows simultaneous savings on electricity, capital costs and CFC gases. Buried pipe systems may be subject to water inltration, which can lower winter performance and enhance summer performance, but also raises sanitary problems related to stagnant water. These problems can be avoided by replacing buried pipes with a closed water underground circuit coupled to the fresh air system via a water/air heat exchanger. One of the economically important parameters to deal with is the pipe depth, which relates to surface temperature. Preliminary results in this climate show that for cooling purposes excavation should be kept to a minimum. Pfafferott [17] presents a paper dealing with the dynamic temperature behaviour and energy performance of three ground-to-air heat exchangers (GAHx) for mid European ofce buildings located in Germany. The rst GAHx is located at Hamm designed with pipe diameters of 200 and 300 mm, a total duct surface area of 1650 m2, depth of ducts 24 m around the foundation slab in dry rock, a mean airow of 10,300 m3/h and an air speed of 2.2 m/s. As it is mentioned, the main characteristic describing energy performance is the overall heat transfer coefcient h, which in this case is 5.5 W/m2 K. The second GAHx is located at Freiburg designed with a pipe diameter of 250 mm, a total duct surface area of 522 m2, depth of ducts 2 m partly below the foundation slab in dry gravel, a mean airow of 7000 m3/h and an air speed of 5.6 m/s. In this case, the overall heat transfer coefcient is 5.0 W/m2 K. The third GAHx is located at Weilheim designed with a pipe diameter of 350 mm, a total duct surface area of 198 m2, depth of ducts 2.3 m around the foundation slab in moist clay, a mean airow of 1100 m3/h and an air speed of 1.6 m/s. The overall heat transfer coefcient is 3.2 W/m2 K. The thermal performance of the GAHx is calculated using four different approaches, a dimensionless ratio of temperature variation RT, the mean heat transfer coefcient hmean, the temperature ratio Y and the coefcient of performance COP. It is concluded that the evaluation of a GAHx depends on project-specic criteria. Each of the evaluated GAHx is

ARTICLE IN PRESS
G. Florides, S. Kalogirou / Renewable Energy 32 (2007) 24612478 2473

shown to be the best from a certain point of view. The rst GAHx narrows the outlet air temperature close to the undisturbed earth temperature. The second GAHx supplies the highest specic energy gain based on the total surface area. And the third GAHx has the highest COP. Pipe lengths up to 100 m and pipe diameters around 250 mm are protable. If the GAHx aims at a high specic energy performance, a small specic surface area should be reached using fewer pipes. If the GAHx aims at a high temperature ratio, a high specic surface area should be reached using more pipes. Finally, it is mentioned that all three GAHx supply more heating and cooling energy than the primary energy they use for the fan input. De Paepe and Janssens [18] used a one-dimensional analytical method to examine the inuence of the design parameters of the heat exchanger on the thermo-hydraulic performance and devise an easy graphical design method which determines the characteristic dimensions of the ground-air heat exchanger in such a way that optimal thermal effectiveness is reached with acceptable pressure loss. Therefore, the choice of the characteristic dimensions is independent of the soil and climatological conditions. This allows designers to choose the ground-air heat exchanger conguration with the best performance. Their analysis considers the air mass ow rate, the inlet air temperature, the desired outlet air temperature, the ground temperature and the geometric sizing parameters which are the diameter of the tube, the length of the tube and the number of tubes in parallel in the heat exchanger. As they emphasise generally, lowering the diameter of the tube raises the effectiveness but on the contrary higher ow rates reduce the effectiveness. So it is better to have several tubes of small diameter over which the ow rate is divided. Long tubes with a small diameter are protable for the heat transfer but at the same time the pressure drop in the tubes is raised, resulting in high fan energy. On the other hand, having a small ow rate per tube and a large diameter gives the least pressure loss. This would mean that it is better to use many tubes, with a large diameter, which is in conict with the thermal demand of a small diameter. In both cases a large number of tubes is benecial. The tube length and diameter combination have to be optimised and this is achieved by a graphical method by reducing the inuencing parameters and introducing the specic pressure drop. The specic pressure drop is a measure for the pressure drop needed to achieve a given thermal performance. In this way, a maximal specic pressure drop can be calculated when a value for the effectiveness of the ground-air heat exchanger is chosen. The effectiveness is dictated by the design requirements and climatic conditions, but often an effectiveness of 80% is considered to be an optimum value for a ground-air heat exchanger [19]. A higher effectiveness is only achievable at the cost of a large increase in the tube length or in the number of tubes.

5. Vertical U-tube ground heat exchangers In a vertical U-tube ground heat exchanger, a water pump circulates uid through pipes inserted into a borehole in the ground. The borehole, after the insertion of the U-tube, is usually backlled with grout in order to ensure good thermal contact with the ground. The grout is often a bentonite clay mixture, with the possibility of having thermally enhanced additives in order to present a thermal conductivity signicantly lower than the surrounding ground. The circulating uid is usually water or a waterantifreeze mixture.

ARTICLE IN PRESS
2474 G. Florides, S. Kalogirou / Renewable Energy 32 (2007) 24612478

A borehole heat exchanger is usually drilled to a depth between 20 and 300 m with a diameter of 1015 cm. A borehole system can be composed of a large number of individual boreholes. Several models for calculating the thermal properties of a ground heat exchanger are available. These models, which are based on Fouriers law of heat conduction, include the analytical line source model [20] and the cylindrical source model [21] and several numerical models [2225]. The most widely used method at this time is the line source model. The model is a simplication of the actual experiment, and data analysis is based on the theory describing the response of an innite line source. Mands and Sanner [26] carried out the thermal response test on a number of boreholes. For a thermal response test, basically a specied heat load is applied into the hole and the resulting temperature changes of the circulating uid are measured. In this way, the thermal conductivity of the borehole is measured on site allowing sizing of the boreholes based upon reliable underground data. The test data consist of one or more curves showing the uid temperature development against elapsed time. The easiest way to evaluate the test data makes use of the line source theory and the following formula is utilised to calculate the thermal conductivity: leff Q , 4pHk (6)

where leff is the effective thermal conductivity, including inuence of groundwater ow, borehole grouting, etc. (W/m K), Q is the heat injection/extraction (J), H is the length of borehole heat exchanger (m) and k is the inclination of the curve of temperature versus logarithmic time (K/s). A second value that can be determined by a response test is the borehole thermal resistance, Rb. This value gives the temperature drop between the natural ground and the uid in the pipes. It is also possible to calculate Rb from the dimensions and materials used. The authors present a table indicating the effective thermal conductivity and the thermal borehole resistance for various grounds in Germany (Table 1). Pahud and Matthey [27] in their paper explain how the thermal performance of a borehole heat exchanger can be assessed with a response test. The tested boreholes differ from one another by the lling material which may be a standard mixture of bentonite and cement, a standard mixture of bentonite and cement with the addition of quartz sand or
Table 1 Geology and results for some thermal response tests carried out in Germany since 1999 Geology Silt and clay (Quarternary/Tertiary) Mesozoic sediments Marl (Emschermergel, Cretaceous) Sand/silt, marl (Cretaceous) Sand and clay (Quarternary/Tertiary) Sand and clay (Quarternary/Tertiary) Marl, clayey Marl, sandstone, limestone (Mesozoic) Silt, sandy (Quarternary/Tertiary)
a

Thermal conductivity, leff (W/m K) 1.6 2.72.8 1.52.0 2.3 2.8 2.22.3 2.5 4.0 3.4

Resistance, Rb (K/(W/m)) 0.100.18 0.110.12 0.08a 0.11 0.070.08a 0.12 0.08a 0.06a

Filled with thermally enhanced grout (Stu watherm).

ARTICLE IN PRESS
G. Florides, S. Kalogirou / Renewable Energy 32 (2007) 24612478 2475

only quartz sand, and the use or omission of spacers to keep the plastic pipes apart from each other and close to the borehole wall. Using an average estimated ground thermal conductivity of 2.5 W/m K, the tests show that the thermal resistance can be decreased by 30% when quartz sand is used instead of bentonite and when spacers are used to keep the plastic pipes in contact with the borehole wall. With a common heat extraction rate of 50 W per metre of borehole length, the temperature gain in a heat pump evaporator is +2 K. Also, it is mentioned that for a typical residential house in Switzerland, a borehole heat exchanger of 100200 m is used with a diameter of 1015 cm, depending on the energy demand and the ground conditions. For typical ground conditions and a single borehole heat exchanger, the borehole length is sized for a heat extraction rate of 50 W per metre of borehole length. Roth et al. [28] describe their in situ experiment for the determination of the thermal properties of the soil (thermal conductivity, borehole thermal resistance and undisturbed soil temperature), which were needed for designing a ground heat exchanger for underground thermal energy storage (UTES). As they mention, about 10 countries in the world are dealing nowadays with this type of investigation including Germany, Sweden, Canada, USA, Norway, Netherlands, England and Turkey. Their test was carried out for over 9 days (24 June to 3 July 2003) while inlet and outlet uid temperatures of the borehole heat exchanger, the ambient temperature and various other necessary parameters were measured. A comparison of results between the conventional slope determination method, the geothermal properties measurement (GPM) data evaluation software method and the two-variable parameter tting method was performed in order to calculate the thermal conductivity and borehole thermal resistance. The determined value of thermal conductivity was 1.8 W/m K and the borehole thermal resistance was 0.3 m K/W. They also concluded the following: (a). The application of the classical slope determination and/or two-variable parameter tting can be used as a fast and reliable tool for data evaluation. (b). The accuracy of the evaluation depends on the care taken when performing the test. Important aspects are reliable temperature measurements, constant power supply, a proper determination of undisturbed underground temperature and weatherproong the system as much as possible. (c). The value of the thermal conductivity was quite insensitive to the operating hours before recording but exhibited an oscillatory behaviour regarding the duration of test condition. Large relative error for short tests, converging for test duration over 120 h was the trend of the thermal conductivity when compared to the 1.8 W/m K expected value. (d). The uctuations exhibited by the experimental curve are mainly governed by corresponding uctuations of ambient temperature and by uctuations of electric power supply especially during night hours. (e). An overall heat loss coefcient of 5.3 W/K was determined applying a model based on an energy balance between the hot parts of the system and the surrounding environment. Zeng et al. [29] in their paper present a new quasi-three-dimensional model for vertical ground heat exchangers (GHx) that accounts for the uid axial convective heat transfer and thermal short-circuiting among U-tube legs. Analytical expressions of the borehole

ARTICLE IN PRESS
2476 G. Florides, S. Kalogirou / Renewable Energy 32 (2007) 24612478

resistance have been derived for different congurations of single and double U-tube boreholes and analytical solutions of the uid temperature proles along the borehole depth have been obtained. As they mention, borehole depths usually range from 40 to 200 m with diameters of 75150 mm. Analyses have shown that the single U-tube boreholes yield considerably higher borehole resistance than double U-tube boreholes. The double U-tubes in parallel conguration provide better thermal performance than those in series. Calculations show that the double U-tube boreholes are superior to those of the single U-tube with reduction in borehole resistance of 3090%. Calculations on typical GHx boreholes indicate that the U-tube shank spacing and the thermal conductivity of the grout are the prevailing factors in all the congurations considered in determining the borehole thermal resistance. Finally, discussions in this paper are limited to the thermal resistance inside the boreholes with U-tubes even though the thermal conduction outside the boreholes often plays an even more important role in the GHx heat transfer process. In addition, many factors such as the capital cost of borehole elds and circulating pump energy consumption must also be taken into consideration when merits and weaknesses of different borehole congurations are to be assessed. 6. Deep geothermal systems Kujawa et al. [30] studied the case of deep geothermal heat plants. These plants operate with one or two-hole systems. A computational model is presented which estimates the temperature of the geothermal water extracted to the earths surface as well as the temperature of the water injected into a deposit level. The predicted characteristics do not take into account specic working conditions of the systems. It is mentioned that the high expenditure incurred in drilling holes deters one from using this method in gaining thermal energy. The one-hole injection system or the use of existing single holes, made during crude-oil and or natural-gas exploration, reduces the capital cost. In one-hole systems, the hole is adapted to locate in it a vertical exchanger with a double-pipe heat exchanger in which the geothermal water is extracted via the inside pipe. Published characteristics allow one to estimate the gained geothermal heat-energy ux as a function of the difference of temperatures of extracted as well as injected water at different volume uxes of the geothermal water. In general, the two-layer systems and two-hole systems are more advantageous than the one-hole system. 7. Conclusions In this paper, various types of ground heat exchangers are described. Ground heat exchangers are used to exploit effectively the heat capacity of the soil and commonly they are coupled to heat pumps for increasing their efciency. One, two and three-dimensional models can be found in the literature that simulate the heat transfer process. Simulation models may be used successfully for sizing and predicting the thermal performance of ground heat exchangers. These exchangers usually supply more heating and cooling energy than the primary energy they use for power input for the fan or pump. As it can be concluded from the studies presented in this paper at a sufcient depth, the ground temperature is always higher than that of the outside air in winter and is lower in summer. This difference in temperature can be utilised as a preheating means in winter and pre-cooling in summer by operating a ground heat exchanger. Usually the recommended

ARTICLE IN PRESS
G. Florides, S. Kalogirou / Renewable Energy 32 (2007) 24612478 2477

depth for horizontal ground heat exchangers is from 1.5 to 2 m. In order to minimise interference between multiple pipes, a separation distance of 30 cm between pipes is recommended and trenches should be at least 2 m apart. A typical horizontal loop is 3560 m long per kW of heating or cooling capacity. A vertical borehole heat exchanger is usually drilled to a depth of 20300 m with a diameter of 1015 cm. A borehole system can be composed of a large number of individual boreholes. For a typical residential house in Switzerland, a borehole heat exchanger of 100200 m is required with a diameter of 1015 cm, depending on the energy demand and the ground conditions. For typical ground conditions and a single borehole heat exchanger, the borehole length is sized for a heat extraction rate of 50 W/m of borehole length. Performance degradation can occur if adjacent boreholes are too close, and a separation distance of 5 m is usually considered adequate. Calculations on typical GHx boreholes indicate that the U-tube shank spacing and the thermal conductivity of the grout are the prevailing factors in all the congurations considered in determining the borehole thermal resistance. Generally, the effectiveness of ground-to-air heat exchangers increases with an increase in the pipe length. Also, there is an increase in effectiveness when the pipe is buried in greater depths. The heating capacity of the system was reduced by increasing the pipe diameter. Finally, a higher air velocity in the pipe leads to a reduction of the systems heating capacity, mainly because of the increased mass ow rate inside the pipe. The line source model is an easy method of evaluating the characteristics of the borehole and does not need expensive equipment. By evaluating the borehole characteristics, no oversized or undersized equipment are installed; therefore, there is lower investment cost and the equipment function with the required output and performance.

References
[1] Lund J. Ground heat worldwide utilization of geothermal energy. Renew Energy World 2005;8(4):25460. [2] Mands E, Sanner B. Shallow geothermal energy, /http://www.ubeg.de/Downloads/ShallowGeothEngl.pdfS; 2005. [3] Geothermal Heat Pump Consortium. /http://www.geoexchange.orgS; 2005. [4] Kelley S. Temperatures in the Kiowa drill hole. Denver basin projectThe Denver Museum of Nature Science (DMNS). /http://www.dmnh.org/denverbasin/r_temp.htmlS; 2005. [5] Mihalakakou G, Santamouris M, Lewis O, Asimakopoulos D. On the application of the energy balance equation to predict ground temperature proles. Sol Energy 1997;60(3/4):18190. [6] Carlslaw HS, Jaeger JC. Conduction of heat in solids, 2nd ed. Oxford: Oxford Science Publishers; 1980. [7] Krarti M, Lopez-Alonso C, Claridge DE, Kreider JF. Analytical model to predict annual soil surface temperature variation. Int J Energy Eng 1995;117:919. [8] Popiel C, Wojtkowiak J, Biernacka B. Measurements of temperature distribution in ground. Exp Therm Fluid Sci 2001;25:3019. [9] Tzaferis A, Liparakis D, Santamouris M, Argiriou A. Analysis of the accuracy and sensitivity of eight models to predict the performance of earth-to-air heat exchangers. Energy Buildings 1992;18:3543. [10] Bi Y, Chen L, Wu C. Ground heat exchanger temperature distribution analysis and experimental verication. Appl Therm Eng 2002;22:1839. [11] Mihalakakou G, Santamouris M, Asimakopoulos D. Modelling the thermal performance of earth-to-air heat exchangers. Sol Energy 1994;53(3):3015. [12] Mihalakakou G, Lewis J, Santamouris M. On the heating potential of buried pipes techniques-application in Ireland. Energy Buildings 1996;24:1925. [13] Bojic M, Trifunovic N, Papadakis G, Kyritsis S. Numerical simulation, technical and economic evaluation of air-to-earth heat exchanger coupled to a building. Energy 1997;22(12):11518.

ARTICLE IN PRESS
2478 G. Florides, S. Kalogirou / Renewable Energy 32 (2007) 24612478 [14] Gauthier C, Lacroix M, Bernier H. Numerical simulation of soil heat exchanger-storage systems for greenhouses. Sol Energy 1997;60(6):33346. [15] De Paepe M, Willems N. 3D unstructured modelling technique for ground-coupled air heat exchanger. In: Clima 2000/Napoli World Congress, Napoli, Italy; 2001. p. 158. [16] Hollmuller P, Lachal B. Cooling and preheating with buried pipe systems: monitoring, simulation and economic aspects. Energy Buildings 2001;33:50918. [17] Pfafferott J. Evaluation of earth-to-air heat exchangers with a standardized method to calculate energy efciency. Energy Buildings 2003;35:97183. [18] De Paepe M, Janssens A. Thermo-hydraulic design of earth-air heat exchangers. Energy Buildings 2003;35:38997. [19] DBA Annex 28. Early design guidance for low energy cooling technologies. Energy conservation in buildings and community systems programme, Annex 28: Low energy cooling, International Energy Agency, Subtask 2, Report 2; 1999. [20] Ingersoll LR, Zobel OJ, Ingersoll AC. Heat conduction with engineering, geological and other applications. New York: McGraw-Hill; 1954. [21] Carslaw HS, Jaeger JC. Conduction of heat in solids, 2nd ed. Oxford: Clarendon Press; 1993. [22] Eskilson P. Thermal analysis of heat extraction boreholes. Ph.D. thesis. Lund, Sweden: Lund University, Department of Mathematical Physics; 1987. [23] Muraya NK. Numerical modeling of the transient thermal interference of vertical U-tube heat exchangers. Ph.D. thesis. College Station, TX: Texas A&M University; 1995. [24] Shonder JA, Beck JV. Determining effective soil formation thermal properties from eld data using a parameter estimation technique. ASHRAE Trans 1999;105(1):45866. [25] Kavanaugh SP, Rafferty K. Ground source heat pumps: design of geothermal systems for commercial and institutional buildings. Atlanta: American Society of Heating, Refrigeration and Air-Conditioning Engineers; 1997. [26] Mands E, Sanner B. In-situ-determination of underground thermal parameters. /http://www.geothermie.de/ egec-geothernet/igd2k1/In-situ-determination%20of%20underground%20thermal%20parameters.pdfS; 2005. [27] Pahud D, Matthey B. Comparison of the thermal performance of double U-pipe borehole heat exchangers measured in situ. Energy Buildings 2001;33:5037. [28] Roth P, Georgiev A, Busso A, Barraza E. First in situ determination of ground and borehole thermal properties in Latin America. Renew Energy 2004;29:194763. [29] Zeng H, Diao N, Fang Z. Heat transfer analysis of boreholes in vertical ground heat exchangers. Int J Heat Mass Transf 2003;46:446781. [30] Kujawa T, Nowak W, Stachel A. Heat-ow characteristics of one-hole and two-hole systems for winning geothermal heat. Appl Energy 2003;74:2131.