Você está na página 1de 4

CFD modeling of ground source heat pump

exchanger and experimental verification

Pengyang Wang Ping Qin

School of Mechanical Engineering School of Mechanical Engineering
Southwest Jiaotong University Southwest Jiaotong University
Chengdu, China, 61003 Chengdu, China, 610031
Itch0118@gmail.com Qinping3159@sina.com

Abstract—Heat transfer around vertical ground Heat transfer (Different inlet temperature and velocity, different
around vertical ground heat exchangers is a common problem types of soil)
for the design and simulation of ground-coupled heat pump
(GCHP) systems. A serial of calculation models have been found II. METHOD
for ground heat exchangers to predict the temperature
distribution of superficial layers, but most of them still take the A. Simulation models
thermal characteristics of the ground as constant, which have
With the help of CFD-software Fluent [5], we can build and
great effort to the heat transfer efficiency. In this paper we will
propose a combination test and verify and thus provide a proper
simulate the steady-state models flexibly. The software Fluent
way to develop a simplified configuration in CFD software. is based on a finite volume method to convert the governing
equations to solve algebraic equations.
Keywords- Ground source heat pump; Ground heat exchangers; According to the study of Eui-Jong Kim et al [6], there were
simulation; u-pipe about 1,500,000 hexahedron and wedge-shaped volume
elements (fig.1). The bent section of the U-pipe was ignored
I. INTRODUCTION for some reasons below: the major purpose of this paper is the
The temperature at a certain depth underground almost study on the thermal interaction between the temperature
remain constant annually, so it provides a better way that distribution underground and the borehole exchangers, so
compared to the warm air in summer or cold air in winter to comparing to the total pipe, the bent section is too short to have
improve the air-conditioning efficiency. To extract or inject a notable amount in thermal transmissions. Furthermore,
from the superficial layer, an efficient heat-exchanger is because of the irregular volume around the bent pipe, the cells
needful and necessary. Most of heat pump units use vertical will increase dramatically and the whole quality of the mesh
ground heat exchangers which have less ground area went downward consequently.
requirement than horizontal arrangements in residential and
commercial buildings. As a result it is dominant to understand
the way of heat transfer between heat exchangers and the soil
Most borehole heat exchanger models recently developed
are based on line source theory or cylinder theory which the
thermal process are usually simulated in 1D or 2D. [1], [2], [3]
and [4] Therefore, the temperature field at certain depth of
underground are not clearly presented. This study concerns the
influence of temperature field at certain depth underground on
the process of heating transfer. The results of the simulation are
compared to laboratory measurements.
The main objectives of this field study are as follows:
x To select a reasonable configuration in Fluent to
predict the temperature field and thus to calculate the
coefficient of heat transmission effectiveness.
x To compare the laboratory measurement with
simulation results on the temperature field and verify
the accuracy of the configuration of simulation.
x Analyze the characters of temperature field at certain
depth of underground at all kinds of combinations. Figure 1

978-1-4244-8165-1/11/$26.00 ©2011 IEEE

The parameters and boundary conditions used for the
simulations are presented in Table i and Table ii respectively.

Figure 2

The base principles of setting the experimental boundary

conditions are presented below:
x By the way of calculation, we ensure the fluid in pipe
flows turbulently to make the heat exchanging fully
and completely.
There were also some simplifications in this model. x Mean annual temperature in Chengdu is 18oC, so the
initial soil and borehole temperatures are set to 18oC.
x Soil is considered as a solid material whose
thermal parameters keep consistent at all x Fig, 3 shows that the sections at the depth of 1m, 2m,
directions. 3m and 4m; each section has 4 sensors arranged as fig,
3. Data are recorded every 30s and usually the whole
x There is no change at thermal conditions (density, test lasts 8hrs to make the distribution steady.
heat conductivity and specific heat) in fluid.
x The fluid in U-pipe flows turbulently and instead
of the heat transfer at the direction of the fluid
flows, we only think about the heat transport
between u-pipe and borehole, and furthermore,
to the ground.
x We see there’s a tight contact between u-pipe
and borehole so the contact resistance could be

B. Laboratory measurement on a U-pipe borehole

A small-scale laboratory model of a heat-exchanger (fig.2)
was constructed at South-west Jiaotong University in 2010 so
the simulation above can be verified by the experimental
measurement. [7] The test equipment consisted of an outer 5m
high steel cylinder with a diameter of 0.8m. To test
conveniently the test equipment was set horizontally. An inner
steel cylinder with a diameter of 0.15m represented the
borehole wall. Furthermore the annular ground region between
the borehole and the steel cylinder was filled with wet sandy-
gravel soil and the thermodynamic characteristics of the
mixture have been surveyed previously. Two orthogonal U-
pipes whose diameters were separately 25mm and 32mm were
set in the borehole. To verify by the simulations, 32mm
diameter u-pipe was used in the following tests. Figure 3

A constant temperature is set at the outside of the pipe wall
which should be an acceptable approximation for a 5m long
borehole at steady-state conditions. Table iii shows all the
boundary conditions necessary in the experiments. To get a
notable temperature distribution, instead of 35oC that is regular
used in construction [8] as the inlet temperature, 40oC, 45oC,
50oC were selected in experiments. The statistical summary of
boundary conditions in experiments is shown in Table iii.

Figure 6
Inlet=50oC depth=4m
From fig.4, fig.5 and fig.6 the distribution present mostly
centrosymmetric. But as it goes closer to the central pipe, the
more asymmetrical the distribution presents. It can be inferred
that the branch of the u-pipe prevent the distribution from
presenting symmetrically. As a result of that the phenomenon
that it looks more circular when it goes further from the central



Figure 4
Inlet=40oC 4m depth Figure 7
Inlet temp=40oC the compare of test result and model

As what have been tested and measured earlier (section b),

we get a similar result in the tests. The whole thermal
distribution is not totally symmetrical, part of the reason of this
phenomenon is that the soil around the borehole was not
average, the heat conduction is affected, in addition, the inlet
pipe is slightly warmer than outlet pipe, which of two reasons
can cause the difference at symmetrical test points. This kind of
discrepancy presents obviously when the test points are near to
Figure 5 the exchangers while on the other hand as the distance between
Inlet= 45oC 4m depth central point and sensor points.

In the condition that the movement of fluid in porous of soil
is neglected, the absolute error between laboratorial
measurements and simulation results maintains less than 0.6oC.
The temperature distribution according to the tests and
simulations are roughly central symmetric, and the further
goes to the central, the colder the sensors test. With the rise of
u-pipe inlet temperature, the higher temperature the fixed
sensors will survey; the more overall heat-transmission will
inject into the ground the more time the ground temperature
recover to the oral state required.
On the basis of the conclusion above, we also can roughly
estimate of the porous-underground water situation. There’s
Figure 8 no doubt that the convection and conduction of underground-
Inlet temp=45oC the compare of test result and model water will improve the heat transmission and the temperature
simulation distribution will not exactly be symmetry, but we still need a
systematic, detailed and convictive models and simulations to
tell us in which condition we can neglect the underground
water or have no choice but to count the convective and
conductive heat transmission of porous water in the total
thermal calculation in.

[1] Jinxiang Wang and Sufen Li , “Model considering thermal conduction

and groundwater advection of soil viewed as porous medium and
numerical simulation on soil temperature field around underground heat
exchanger,” Dalian University of Technology, 2006
[2] Cane R L D, Forgas D A. Modeling of Ground-Source Heat Pump
Performance. ASHRAE Transactions, 1991, 97(1):909-925.
[3] Deerman J D, Kavanaugh S P. Simulation of Vertical U-Tube Ground-
Figure 9 Coupled Heat Pump Systems Using the Cylindrical Heat Source
Inlet temp=50oC the compare of test result and model Solution. ASHRAE Transactions, 1991, 97(1): 287-295.
simulation [4] Yavuzturk C, Spitler J D. A Short Time Step Response Factor Model for
Vertical Ground Loop Heat Exchangers. ASHRAE Transactions, 1999,
When the inlet temperature is 50oC, the temperature 105(2):475-485
distribution presents more symmetrical. The quantity of heat [5] http://my.fit.edu/itresources/manuals/
injection increase as the difference of soil-inlet temperature [6] Eui-Jong Kim , Jean-Jacques Roux , Gilles Rusaouen , Frédéric Kuznik,
grows. So it can be inferred that the same test point will grow “Numerical modelling of geothermal vertical heat exchangers for the
short time analysis using the state model size reduction technique”,
gradually following the rise of the inlet temperature, which also Applied Thermal Engineering 30 (2010) 706–714
cause the temperature gradient of soil increases. The simulation
[7] Yu Liu, Pin Qin,”Soil saturation transfer of heat penetration coupled
results fit the laboratory measurements well. numerical”, Southwest Jiaotong University,2007
[8] National Standards of Peoples Republic of China, “Technical code for
ground-source heat pump system”, GB 50366-2005, 2006.1.1