Вы находитесь на странице: 1из 9

HEAT AND MASS TRANSFER MODEL OF A GROUND HEAT EXCHANGER

Mirek Piechowski MP Energy Consulting PTY. LTD. 3 Kent Court, Glen Waverley 3150, Austarlia Tel/Fax +61-03-9884 5221

ABSTRACT A new approach to the simulation of a horizontal type Ground Heat Exchanger is proposed resulting in a better accuracy and at the same time a reduced computational effort. These results come from the concentration of the computational effort at the locations with the largest temperature and moisture gradients, i.e. the pipe-soil interface. The model takes into account heat and moisture transfer in the soil allowing for more accurate predictions of the soil thermal response to the heat fluxes induced by the GHE operation. This in turn allows for a more accurate prediction of the soil temperature field and the circulating fluid temperature profile.

1.

INTRODUCTION

A Ground Heat Exchanger (GHE) links a heat pump unit with the soil. There are two basic GHE configurations; vertical and horizontal. For any GHE configuration, the aim of the design process is to arrive at a length of pipe, which will be capable of absorbing or rejecting the required amount of thermal energy. The difficult part in the process is to predict the thermal response of the soil surrounding the pipe through which either hot or cold water is circulated. A more specific question is: what would be the soil resistance to the heat flow and how would it change during operation of the system? The answer to this question is neither simple nor universal. It is not simple because soil transport phenomena are very complex and usually require an advanced mathematical model for an accurate description. This however tends to lessen the applicability of such a model in every day engineering practice.

Any such model is not universal because it depends on many factors such as the operating strategy of the system, the GHE geometry, and local soil and climatic conditions, which by their nature are very site specific. Various models of a GHE have been developed, based on a line source theory and including the following simplifying assumptions (Bose , 1985): i. the soil properties are uniform and constant, ii. the heat flow rate per unit of pipe length is constant over a period of time and, iii. the heat sink/source is considered to be a line source, i.e. very long and of extremely small diameter. In order to use this model we have to make some assumptions regarding soil, or field, resistance to the heat flow. Typically, the soil resistance is taken to be equal to that which results from a long term continuous operation of a heat exchanger. In practice this assumption may lead to an oversized GHE. However it needs to be said that this approach allows for a very quick and easy way of sizing of a GHE, and various modifications are currently widely used by the industry. In this work a mathematical model of a horizontal type GHE is developed, which uses elements of the model proposed by Mei (Mei, 1986). However, some major modifications are made in order to include heat and mass transfer in the soil as well as to enhance the accuracy of the model. This is achieved by concentrating the computational effort in the vicinity of the pipe, where the most important heat and mass transfer processes are taking place. The heat and mass transfer in the soil is calculated using the model developed by Philip and de Vries (Philip and de Vries, 1957)

2.

ASSUMPTIONS FOR MODEL DEVELOPMENT

A GHE model has been developed in order to simulate the operation of a GCHP system, mainly in the cooling mode. Thus the model does not take into account the soil freezing around the pipe as a result of a prolonged heat extraction regime. Heat and mass transfer phenomena in the soil are very complex and difficult to model in all their detail. However, by making some assumptions, it is possible to develop a GHE model, which is capable of an accurate prediction of the 'Entering Water Temperature' to a heat pump unit. We have made the following simplifying assumptions in the analysis: (a) the soil is homogeneous and the soil type does not change along the pipe (b) the soil temperature at a certain distance from the pipe is assumed to fluctuate, only with diurnal and seasonal variation, and does not depend on the GHE operation (c) the heat transfer in the soil in the direction parallel to the pipe is negligible (d) the air-soil surface boundary is assumed to be of the convective type (e) the temperature and velocity of the circulating fluid are assumed constant at any cross section of the pipe

(f) the influence of gravity on the soil moisture transfer in the unsaturated soil is assumed to be negligible. For an accurate modeling of the soil transport phenomena, one would need to know the exact soil composition and physical properties at the location of a GHE. This would require an extensive geological surveying of the site, which would considerably increase the installation cost. Even if such detailed data were available, it would be very difficult to accommodate them in a general model, which is supposed to be relatively simple in terms of the input data and versatile in its applications. Based on the local site geological data, we can arrive at some average soil properties based on the prevailing soil type and conditions. These arguments lead us to assumption (a) above. Assumption (b) results from the translation of a real situation to the numerical reality of a computer model. In practice we are dealing with an infinite soil region. In numerical reality however, we have to draw a boundary for our system, see Figure 1. There are a number of ways of specifying those boundaries, either numerical or experimental.

Figure 1:

Schematic of the soil domain.

The industry standard (NRECA/IGHSPA, 1988) for the spacing of loops of a GHE is about 34m, which gives about 1.5-2.0m on each side of the loop. From our experiments, in which the hot water was circulated continuously for fourteen weeks, the soil temperature at 0.6m from the pipe rose by about 7C. Some of this temperature rise was also caused by the seasonal temperature variation, since the experiment was conducted at the beginning of Spring. But clearly the distance from the pipe surface to the boundary should not be smaller than 1.0m. On the other hand, the distance of about 1.5-2.0m, for a horizontal GHE, seems to be a reasonable assumption for the far field boundary. Although generally the bigger the distance from the pipe the more realistic results one is likely to obtain. Limitations faced with at this point come from the memory capacity of the computer and numerical algorithms employed to solve the resulting set of equations. From an analysis of the experimental results we see that the steepest temperature gradient in the soil region occurs at the pipe-soil interface. We also observe that a few centimeters away from the pipe surface the soil temperature is reduced by up to 30-40%. On the other hand, inspection of the longitudinal distribution of the water temperature reveals that a much

smaller temperature reduction takes place over a much longer distance. These observations lead us to make assumption (c) above. Considering a multiple pipe configuration of a GHE loop, we notice that the distance from the uppermost pipe to the soil surface is smaller than the far field radius. In this case we cannot assume the air-soil surface boundary temperature to be totally dependent on the diurnal and seasonal temperature variation. In order to make the model relatively easy to use in terms of the input data required, we assume that the net effect of a GHE operation on the soil surface phenomena can be modeled using a convective boundary approach, assumption (d). The convective heat transfer coefficient can be calculated using average wind velocity data and knowledge of the type of soil surface. The circulating fluid flow rate and the pipe diameter are chosen to maintain the flow in the turbulent region. This fact, and the pipe length to pipe diameter ratio, leads us to assumption (e) in the above listing. Assumption (e) holds for the turbulent regime flow with a big pipe diameter to pipe length ratio. Assumption (f) holds for the soil which moisture content is below the saturation level (unsaturated soil) and in the presence of the induced temperature gradient. In these circumstances the thermal and isothermal moisture diffusion becomes the dominant force in the moisture transfer (Sophocleous, 1979).

3.

THEORETICAL DEVELOPMENT

The model consists of three equations: one is related to the energy balance of the circulating fluid and the two other equations describe transient, simultaneous heat and mass transfer in the soil region. The energy balance for a differential section of a GHE can be expressed in the following form (Piechowski, 1996): v T f l + 2U i Tr 2rp , o f c f r =
rp , o

T f t

(1)

where U i is the overall heat transfer coefficient through the pipe wall. The temperature distribution in the pipe is not explicitly calculated here. It is assumed that the thermal capacity of the pipe material is sufficiently small compared with that of the circulating fluid and the soil. For this reason we do not include the heat diffusion equation for the pipe wall. It is represented by the overall heat transfer coefficient. Transient, simultaneous heat and mass transfer in the soil region is described by the following equations: C T K h = (KT ) + (D l ) + Ll t y K h = (DT T ) + (D l ) + t y (2)

(3)

or on the basis of assumption (f): C T K h = (KT ) + (D l ) + Ll t y K h = (DT T ) + (D l ) + y t (4)

(5)

The boundary conditions for the pipe-soil interface consist of conditions for the heat and moisture transfer. The heat flux at the pipe surface is given by the following equation: q = K sT + Ds Substituting for q yields: Ui rp ,i rp , o = K s T + D (7) (6)

where T f is the circulating fluid temperature and rp ,i and rp , o is the pipe inside and outside radius, respectively. Since the pipe surface is impermeable, the moisture flux at the pipe surface is given by the following equation: 0 = DT T + DT (8)

A more detailed discussion on the boundary conditions and the diffusion coefficients can be found in (Piechowski, 1997). Initial conditions at the start-up of the system ( at time t=0) are: (r , l ) = s,ini ( y ) T f ,ini (l ) = Ts , ini ( y ) where Ts ,ini is an initial soil temperature distribution. For ON/OFF operation of the system, the initial fluid temperature and soil moisture content for the next ON period are given as: (r , l ) = (r , l )OFF T f ,ini (l ) = Ts (rp , o , l )OFF (11) (12) (9) (10)

where (r, l )OFF and Ts (rpo , l )OFF are the soil moisture content and soil temperature distribution after the preceding OFF period of the system.

Both the fluid and the soil initial temperatures can be either calculated from the empirical formulas, or determined by the field temperature measurement. It is assumed that, at the startup of the system, the circulating fluid temperature is equal to the undisturbed soil temperature at a given depth. In order to simulate the GHE performance we need to solve Eq.1 for the circulating fluid temperature. To do so we need to know the soil temperature distribution around the GHE pipe. In order to calculate the soil temperature distribution, we need to solve the set of Eq.4 and Eq.5. This procedure results in the set of three simultaneous heat and mass transfer equations that need to be solved for the entire soil domain, what requires a considerable computing effort for a full scale installation. However, this procedure can be further simplified without a significant loss of accuracy. It can be demonstrated (Piechowski, 1996) that the heat transfer rate to or from the pipe is dictated by a relatively small region of soil around the pipe, where the biggest temperature gradient occurs. Taking advantage of the experimental observation that the steepest temperature gradient exists within the region of approximately 0.15m away from the pipe (Piechowski, 1996) we assume that the most significant moisture content changes will be limited to this region. It is therefore proposed to limit the heat and mass transfer analysis to this region only, thus significantly reducing the computational burden without compromising the accuracy while still preserving the essential meaning of the physical processes taking place. The two governing equations, Eq.4 and Eq.5 for heat and mass transfer in the soil region can be discretised in either cylindrical or Cartesian coordinates. Discretisation of the governing equations in cylindrical coordinates appears to be a more natural choice. However, if we take multiple pipe configuration into account this approach loses its attractiveness. Discretisation in Cartesian coordinates becomes now more attractive. There is though one inherent difficulty with representing the pipe temperature. The pipe-soil interface temperature is represented as an average temperature for a rectangular node. Thus in order to accurately evaluate the pipe temperature, and consequently the heat rejection or extraction rate, the node region should be sufficiently small. This, in turn, increases demand on computing power, since the number of nodes for a given soil region is going to increase with the reduction of the node size. In order to calculate soil temperature and moisture distribution in the vicinity of the pipe surface, it is proposed to single out a cylindrical region around the pipe from the rest of the node. Further, we assume that the moisture content at the outermost radius of this region remains constant and equal to the far field moisture content for a given site. This may not be always the case, but for the soil which is not very dry, the thermal conductivity changes little with changing soil moisture content. The error then, associated with evaluating the soil thermal conductivity, is not significant in comparison with the uncertainty of the available soil thermal conductivity data itself. By doing so, we make the node temperature equal to the outermost radius of the radial region. This temperature is calculated as a balance temperature between the radial region and the rest

of the pipe node. This way we 'stitch' the radial region around the pipe together with the rest of the soil domain. The pipe-soil interface temperature is still calculated as an average temperature of the node, but this time the node is much smaller in size, and the assumption of the node temperature representing a true temperature of the region is much more accurate. Temperatures of all the remaining nodes in the soil domain are calculated based on the classical finite difference formulation for a rectangular mesh. The advantage of such an approach is two-fold. Firstly, the temperature of the soil region with a large temperature gradients are calculated more precisely. Secondly, in the regions of small temperature gradients we can afford a larger size of the mesh what leads to reduced computer time and memory requirements. This is particularly significant in real GCHP applications, where the soil domain is large and long simulation periods are usually required. The validity of the concept of using a radial region around the pipe can be demonstrated on the following example. Figure 2 represents transient temperature profile of the pipe-soil interface calculated with and without radial region around the pipe.

pipe/soil interface temperature, [C]

42 38 34 30 26 22 18 0 60 time, [min] 120 180


Tp_dS=0.05,cartesian Tp_dS=0.3,cartesian Tp_dS=0.3,radial

Figure 2:

Transient temperature profile around the GHE as calculated by the model, with andwithout the radial region.

Initially the interface temperature was calculated with the mesh size equal to S = 0.05m . In the next calculation the mesh size was changed to S = 0.3m , while the remaining setting was unaltered. As we can see there is a dramatic difference in the transient temperature profile of the pipe soil interface. The third temperature profile was obtained for a mesh size equal to S = 0.3m but this time the pipe-soil temperature interface was calculated using the radial region singled out from the pipe node. The outer radius of the radial region is equal to Rmax = 0.5S . The difference between the case with the mesh size equal to S = 0.05m is only about 0.3C.

Figure 2 and Figure 3 present soil temperature distribution around the pipe after 5 and 60 minutes of operation, respectively.

38.0 34.0
Tp_dS=0.05,cartesian

soil temperature, [K]

30.0 26.0 22.0 18.0 0

Tp_dS=0.3,cartesian Tp_dS=0.3,radial

0.3 0.6 distance from the pipe, [m]

0.9

Figure 3:

Soil temperature distribution around the GHE pipe after 5min operation of the GHE.

38.0 34.0
Tp_dS=0.05,cartesian

soil temperature, [K]

30.0 26.0 22.0 18.0 0

Tp_dS=0.3,cartesian Tp_dS=0.3,radial

0.3 0.6 distance from the pipe, [m]

0.9

Figure 4: the GHE.

Soil temperature distribution around the GHE pipe after 60min operation of

As we can see there is no significant loss of accuracy in the soil temperature calculations compared with results obtained for a finer mesh. This allows for a fast and accurate calculation of the heat transfer in the soil without a significant loss of accuracy. The model described so far refers to the ON operation of a heat pump unit. When the heat pump is not operating we are simply solving for the transient temperature and moisture distribution in the soil for a known initial and boundary conditions. The only difference is that in this case we neglect effects due to the circulating fluid since its heat capacity is negligible compared with that of the soil region.

A more detailed discussion of the proposed model and its validation can be found in (Piechowski, 1997).

CONCLUSIONS A new approach to modeling of a GHE was presented. The new approach results in an increased accuracy of the model and at the same time offers a considerable reduction in the simulation time. This was achieved by concentrating the computational effort at the pipe-soil interface where the changes in both the temperature and moisture gradients are the most significant.

ACKNOWLEDGEMENTS Presented work constitutes part of a PhD research supervised by Prof. Bill Charters whose support is gratefully acknowledged. The author would also like to acknowledge support of The Department of Mechanical and Manufacturing Engineering at the University of Melbourne for providing all necessary facilities. The financial support for this research, project No.2215, came from Energy Research and Development Corporation.

REFERENCES Bose, J.E., et.al. (1985). Design/data manual for closed-loop ground-coupled heat pump system. Oklahoma State University. Mei, V.C. (1986), Horizontal ground-coupled heat exchanger, theoretical and experimental analysis., Technical Report ORNL/CON-193, Oak Ridge National Laboratory. NRECA/IGHSPA (1988), Closed-loop ground-source heat pump systems., Oklahoma State University. Philip, J.R., de Vries, D.A. (1957), Moisture movement in porous materials under temperature gradients., American Geophysical Union, Transactions}, 38, 222-232. Piechowski, M. (1996), A Ground Coupled Heat Pump system with Energy Storage., PhD Thesis, Melbourne University. Piechowski, M. (1997), Heat and Mass Transfer Model of a Ground Heat Exchanger. Theoretical Development. Int.J.Energy Research, (submitted 10.1997). Sophocleous, M. (1979), Analysis of water and heat flow in unsaturated-saturated porous media., Water Resources Research, 15(5), 61-75.

Вам также может понравиться