Heat Transfer Engineering

ISSN: 0145-7632 (Print) 1521-0537 (Online) Journal homepage: https://www.tandfonline.com/loi/uhte20

Heat Transfer Research of a New Type of Radiant

Floor Heating System

Minzhang Liu, Huan Zhang, Wandong Zheng & Shijun You

To cite this article: Minzhang Liu, Huan Zhang, Wandong Zheng & Shijun You (2019): Heat
Transfer Research of a New Type of Radiant Floor Heating System, Heat Transfer Engineering,
DOI: 10.1080/01457632.2019.1661701

To link to this article: https://doi.org/10.1080/01457632.2019.1661701

Published online: 16 Sep 2019.

Submit your article to this journal

Article views: 7

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=uhte20
HEAT TRANSFER ENGINEERING
https://doi.org/10.1080/01457632.2019.1661701

Heat Transfer Research of a New Type of Radiant Floor Heating System

Minzhang Liua, Huan Zhanga,b, Wandong Zhenga,b, and Shijun Youa,b
a
School of Environmental Science and Engineering, Tianjin University, Tianjin, China; bKey Laboratory of Efficient Utilization of Low
and Medium Grade Energy (Tianjin University), Ministry of Education of China, Tianjin, China

ABSTRACT
In China, the energy policy of switching charcoal heating to electric heating is pushing
hard. Radiant floor heating system (RFHS) with air-source heat pump system is considered
to be a suitable scheme for energy retrofit. A new type of lightweight RFHS with the alumi-
num layer is put forward in this article, which has the benefit of lowering the supply water
temperature by enhancing the heat transfer performance and homogenizing the floor sur-
face temperature distribution. The theoretical model of the new lightweight RFHS is built
and solved by the Matlab program. Experiments are conducted to study the thermal per-
formance of the RFHS and verify the theoretical model. The effects of supply water tempera-
ture and water flow velocity on the thermal performances of the new RFHS are studied.
Then, the effects of aluminum layer on the heat transfer performance of the new system
are researched. Finally, comparative research of the new lightweight RFHS and traditional
heavy RFHS is carried out. The results show that the average floor surface temperature, heat
transfer rate, and annual exploitation cost of the new RFHS are all better than the traditional
one.

Introduction problem [5–7]. Furthermore, the RFHS was more

comfortable than the air conditioning system by
Background
reducing the draught risk [8]. However, there are also
In recent years, the energy shortages and environmen- disadvantages of the RFHS, such as inconvenient con-
tal degradation are getting worse. In China, the energy struction, high energy consumption, and high operat-
policy of switching charcoal heating to electric heating ing cost. For these reasons, numerous researchers
is put forward to reduce the coal consumption and studied the performance of the RFHS and tried to
protect environment. The low-temperature radiant improve the efficiency of this system.
floor heating system (RFHS) with air-source heat
pump system is considered to be a suitable scheme
for energy retrofit in the cold rural areas of China Research status of the RFHS
and related researches have been conducted [1]. The
RFHS has good application prospects for its energy The RFHS can be categorized into two types: the trad-
conserving and thermal performance [2]. Abdolzadeh itional heavy RFHS and the lightweight RFHS. In the
et al. [3] comparatively studied the RFHS and the traditional heavy system, the pipes are laid in the con-
skirting board heating system. The results showed that crete layer and the heat transfer resistance of the sys-
the RFHS followed more closely to the criteria of ther- tem is large. The lightweight system is characterized
mal comfort. The RFHS could use many low-tempera- by its less floor load, faster installation, and quicker
ture heat sources such as solar energy and geothermal heating. The thickness of the lightweight system is
energy. Zheng et al. [4] studied the use of geothermal also less than the traditional heavy system.
energy in the RFHS and provided an operation strat- There are numerous studies focused on the thermal
egy for the geothermal heating system. Compared performance and calculation methods of the trad-
with the air conditioning system, the RFHS was more itional heavy RFHS. Rhee and coworkers [9] and [10]
energy-saving, space-saving, and caused less noise summarized the research in recent 50 years on the

CONTACT Associate Professor Wandong Zheng [email protected] Building Environment & Energy Engineering, School of Environmental Science
and Engineering, Tianjin University, Beiyangyuanxiaoqu, Jinnan District, Tianjin 300350, China.
Color versions of one or more of the figures in the article can be found online at www.tandfonline.com/uhte.
ß 2019 Taylor & Francis Group, LLC
2 M. LIU ET AL.

Nomenclature
2
A area (m ) W pump shaft power (kW)
a heat conductivity coefficient (m2/s) x axial coordinate
C annual operating cost (USD/m2) X flow direction of the hot water
c specific heat of the fluid (kJ/(kg
C)) y life cycle (year)
d inner diameter of pipe (m) Z annual cost (USD/m2)
E experimental value (
C)
Gr Grashof number
Greek symbols
g gravitational acceleration (m/s2)
H pump head (m) k thermal conductivity (W/(mK))
I discount rate d thickness (m)
i node q density of the fluid (kg/m3)
L pipe length (m) a convective heat transfer coefficient (W/(m2K))
l length of the node (m) r Boltzmann constant (W/(m2K4))

l finalize design size (m) u radiative angle factor
MBE mean bias error (%) c bulk density of water (N/m3)
N group number of comparison g pump efficiency
n total number of nodes t kinematic viscosity (m2/s)
Na number of heating days
b coefficient of cubical expansion (1/K)
Nu Nusselt number
P pump power (kW) D difference value
P0 initial cost (USD/m2) Subscripts
Pr Prandtl number a aluminum layer
PE-RT polyethylene of raised temperature resistance f fluid
Q flow rate (m3/s) h concrete layer
Qa annual heating consumption (GJ) i design indoor temperature (
C)
Qh designing heating load (kW)
m cement mortar layer
Re Reynolds number
RFHS radiant floor heating system
n indoor air
S theoretical value (
C) o design outdoor temperature (
C)
t temperature (
C) p pipe wall

t average wall temperature in the laboratory (
C) v average outdoor temperature (
C)
u velocity of the fluid (m/s) w wooden floor layer
V control volume (m3)

radiant heating and cooling systems and found that the two-dimensional model for the RFHS to analyze its
thermal performance and energy efficiency of the radi- thermal performance and found that RFHS needed less
ant heating and cooling systems had been improved power to achieve the required condition and the tem-
greatly. Rahimi and Sabernaeemi [11] experimentally perature distribution was more uniform. There were
studied the contribution of radiation and free convec- also many researchers conducted other relative
tion heat transfer in radiant ceiling heating system. The research. Spiga and Vocale [15] focused on the influ-
heat transfer from the ceiling surface of a room to ence of floor geometry on the thermal processes in the
other internal surfaces has been investigated in this ground. The numerical results showed that floor geom-
study. The results showed that more than 90% of the etry has a significant influence on the steady-state
heat was transferred by radiation from the heated ceil- ground global heat transfer coefficient. Fontana [16]
ing to the other surfaces of enclosure. Cho and Zaheer- analyzed the thermal performance of radiant heating
uddin [12] conducted experiments to evaluate two floors in furnished enclosed spaces for different config-
flow-control schemes and found that the radiant heat urations and found that the furniture influence could
control under continuous and intermittent modes of not be neglected. Hu and Niu [17] reviewed the actual
operation was better. Besides the experimental method, operation of four projects in different climatic cities
many scholars adopted the simulation method to con- and found that most research at present focused on the
duct research. Wu et al. [13] proposed and established capacity of radiant system but there was little research
a new valid simplified model to calculate the heat concentrated on optimizing the overall system
transfer rate and surface temperature of the radiant performance.
floor heating and cooling systems, and the results With the development of RFHS, the lightweight
showed that the average water temperature and pipe RFHS is put forward. There were also many studies
space had large impact on the thermal performance of focused on the lightweight RFHS. Wang et al. [18] pro-
the radiant floor. Khorasanizadeh et al. [14] built a posed a new type of low temperature radiant floor panel
 HEAT TRANSFER ENGINEERING 3

Figure 1. The structure of traditional heavy RFHS.

Figure 2. Schematic of the new type lightweight RFHS.

system and carried out experiments to analyze its ther- Main study of this article
mal performance. The results indicated that the new
In this article, a new type of lightweight RFHS with
floor heating device could meet the heating demand
aluminum layer was put forward in order to match the
with lower water temperature when compared with
low-temperature heat sources, save energy, and
traditional radiant floor heating coil pipe. Yu and Yao
improve the system performance. A theoretical model
[19] made an experimental research under the steady
of the new RFHS was developed to analyze the influ-
state to study the heating and cooling performance of
ence of key parameters, i.e., supply water temperature
light radiant floor panels and found that the time to a
and water flow velocity, on the thermal performance of
steady state indoor environment for the light floor radi-
the new RFHS. Experiments were conducted to test the
ant panels was much smaller than that for the trad-
thermal performance of the new RFHS and verify the
itional concrete floor system. Rahimi and Sabernaeemi
accuracy of the model. The effects of aluminum layer
[20] conducted experiments in an enclosure with light-
on the thermal performance of the new lightweight
weight RFHS and found that 75–80% of the heat was
RFHS were also experimentally studied. Finally, com-
transferred through radiation from the heated floor to
parative studies between the new lightweight RFHS and
the other surfaces of the enclosure. Yu et al. [21]
the common traditional heavy RFHS were carried out
researched the thermal performance of precast light
with respect to both thermal and economic efficiency.
floor radiant panel. It was concluded that the distance
between heating pipes had obvious effects on the
equivalent thermal resistance and the thickness of alu- Theoretical model
minum foil had insignificant influence on the thermal
Configuration of the radiant floor heating system
performance of this system. Zhang et al. [22] built a
heat transfer model to analyze the lightweight RFHS. Figure 1 shows the structure of heavy RFHS. Heating
The model was verified by experiments and used to pipes are buried in the concrete layer, which is widely
analyze the thermal performance of this system. Based used in the traditional engineering. Compared with
on the above research, the form of lightweight RFHS the traditional heavy RFHS, the lightweight RFHS
needs to be further improved. does not have concrete layer. Therefore, it has the
4 M. LIU ET AL.

Figure 3. Schematic of the new system control volumes.

advantages of convenient construction, light weight, The values are used to calculate the outlet water tem-
faster room warming, and little space occupation. The perature. The method of Gaussian elimination and
greatest advantage is that the thermal resistance of the backward substitution is adopted to solve this model.
lightweight RFHS is lower. Therefore, the lightweight The calculation process is specified in i node by the
RFHS is more suitable to be integrated with the low- following energy balance equations, i ¼ 0, 1, 2 … n.
temperature heat sources. While i ¼ 0, the temperature of fluid is the supply
The structure of the lightweight RFHS is shown in water temperature. For the control volume of pipe
Figure 2. In this system, the heating pipe is embedded wall (Vp), energy balance equation is
in the insulation materials and the metal thermal con- tp ð1Þtp ð0Þ Af , p
ductive layer (aluminum layer) is installed under the k p Ap þ af , p ½tf ð0Þ  tp ð0Þ
Dl 2
heating pipe. The heat exchange area with the heating Aa, p ta ð0Þtp ð0Þ (1)
pipe can be increased and the thermal resistance of þ dp
¼0
2 þ da
the system is reduced. Therefore, the thermal per- 2kp 2ka
formance of the system is improved. where the thermal conductivity of the pipe wall kp is
0.39 W/(m K). The outside diameter of the pipe is
Model assumptions 20 mm. The thickness of pipe wall dp is 2.3 mm. The
length of the pipe L is 55 m.
As to the complexity of the new lightweight RFHS
For the control volume of aluminum layer (Va),
heat transfer process, some assumptions are made for
energy balance equation is
quantitative analysis.
ta ð1Þta ð0Þ Aa, p tp ð0Þta ð0Þ
ka Aa þ
1. The heat transfer process of the lightweight RFHS Dl 2 dp
þ da
2kp 2ka
is in steady state. (2)
Aa, w tw ð0Þta ð0Þ
2. The conduction heat transfer between fluids in þ ¼0
2k þ 2k
2 dw da
the heating pipe is ignored. w a

3. The materials in each layer are homogeneous and where the thermal conductivity of the aluminum layer
the heat conductivity coefficient of each material ka is 237 W/(m K) and the thickness of the aluminum
is constant. layer da is 0.12 mm.
4. Each structure layer contacts closely and the con- For the control volume of wooden floor layer (Vw),
tact thermal resistance is ignored. the energy balance equation is
5. The heat flow in the downward direction
tw ð1Þtw ð0Þ Aa, w ta ð0Þtw ð0Þ
is ignored. kw Aw þ
Dl 2 dw da
þ
2kw 2ka
Theoretical model of the new lightweight system An , w An , w

þ an, w ½tn  tw ð0Þ þ ru ð
t þ 273Þ4
The model of new RFHS is constructed based on the 2 2
 4
energy balance equations. The model is divided into  tw ð0Þ þ 273  ¼ 0
four control volumes. The schematic of the control
volumes are shown in Figure 3. X direction is the (3)
flow direction of the hot water. The supply water tem- where r is the Boltzmann constant and equals to
perature and indoor air temperature are monitored. 5.67  108 W/(m2 K4); u is the radiative angle factor
 HEAT TRANSFER ENGINEERING 5

Figure 4. Schematic diagram of the experimental set-up.

between the floor and room, and the value is 0.87. For the control volume of pipe wall (Vp), energy
The material of the surface layer is wood and its ther- balance equation is
mal conductivity kw is 0.12 W/(mK). The thickness of tp ði  1Þ2tp ðiÞ þ tp ði þ 1Þ
the surface layer dw is 8 mm. The temperature of the kp Ap
Dl
heating room tn is constant at 18
C. Based on the ta ðiÞtp ðiÞ
theory of turbulent natural convection, the convective þ af , p Af , p ½tf ðiÞ  tp ðiÞ þ Aa, p dp
¼0
heat transfer coefficient an, w can be derived by the fol- 2kp þ 2k
da
a

lowing equations [23]. (12)

an, w ¼ Nun kn
l (4) For the control volume of aluminum layer (Va),
an, w 
l energy balance equation is
Nu ¼ (5)
k ta ði  1Þ2ta ðiÞ þ ta ði þ 1Þ tp ðiÞta ðiÞ

3 ka Aa þ Aa, p dp
g l bDt Dl þ da
Gr ¼ (6) 2kp 2ka
t2
t tcq tw ðiÞta ðiÞ
Pr ¼ ¼ (7) þ Aa , w ¼0
2kw þ 2ka
dw da
a k
While i ¼ 1, … , n  1, for the control volume of (13)
fluid (Vf), the energy balance equation is
  For the control volume of wooden floor layer (Vw),
tf ði  1Þtf ði þ 1Þ energy balance equation is
quAf c þ af , p Af , p ½tp ðiÞ  tf ðiÞ ¼ 0
2 tw ði  1Þ2tw ðiÞ þ tw ði þ 1Þ ta ðiÞtw ðiÞ
(8) kw Aw þ Aa, w
Dl dw da
þ
where the water density q is 1,000 kg/m3, the specific 2kw 2ka
heat c is 4.174 kJ/(kg K). For the forced convective þan, w An, w ½tn  tw ðiÞ
 4
heat transfer in the pipe, the flow regime of heat þruAn, w ½ð
t þ 273Þ4  tw ðiÞ þ 273  ¼ 0
transfer medium is turbulent and the heat transfer
coefficient af , p can be calculated by Eqs. (9–11). (14)

kf While i ¼ n, for the control volume of fluid (Vf),

af , p ¼ Nuf (9) the energy balance equation is
d  
Nuf ¼ 0:023Re0:8 0:3
f Prf (10) tf ðn  1Þtf ðnÞ Af , p
quAf c þ af , p ½tp ðnÞ  tf ðnÞ ¼ 0
ud 2 2
Ref ¼ (11)
 (15)
6 M. LIU ET AL.

For the control volume of pipe wall (Vp), energy wood floor is used in this experiment as the surface.
balance equation is Meanwhile, the experimental results are used to verify
tp ðn  1Þtp ðnÞ Af , p the accuracy of the theoretical model. The compara-
kp Ap þ af , p ½tf ðnÞ  tp ðnÞ
Dl 2 tive experiments, with aluminum layer and without
Aa, p ta ðnÞtp ðnÞ (16) aluminum layer, are also conducted.
þ dp
¼ 0 The experiment of new RFHS is conducted in a
2 þ da
2kp 2ka
temperature-controlled chamber. The air temperature
For the control volume of aluminum layer (Va ), of the air jacket around the test chamber is main-
energy balance equation is tained by air conditioning system at 18
C and the air
ta ðn  1Þta ðnÞ Aa, p tp ðnÞta ðnÞ temperature in the chamber is maintained by the cold
k a Aa þ
Dl 2 dp air when the test starts. The heat moved away by the
2kp þ 2ka
da
(17) air conditioning system equals to the heat brought in
Aa, w tw ðnÞta ðnÞ by the new type of RFHS. Figure 4 shows the sche-
þ ¼0
2kw þ 2ka
2 dw da
matic diagram of the experimental system. Standard
For the control volume of wooden floor layer (Vw), experimental devices include the experimental cham-
energy balance equation is ber with floor heating system, circulation system of
tw ðn  1Þtw ðnÞ Aa, w ta ðnÞtw ðnÞ thermal fluid, ventilation cooling system, and param-
kw Aw þ eter control equipment. Water tank provides hot
Dl 2 dw da
þ water which is heated by electrical heating system.
2kw 2ka
An , w The heating wire placed in the low position water
þan, w ½tn  tw ðnÞ tank is used to adjust the water temperature prelimin-
2 
An , w  4  arily. The supply water can be adjusted accurately by
þru
4
ðt þ 273Þ  tw ðnÞ þ 273 ¼0
2 the heating wire located in the high position water
(18) tank. High position water tank is used to deliver the
Working fluid inlet temperature is the supply water hot water to the system and ensure stable flow rate.
temperature. Pipe wall, aluminum layer, and wooden The water flows from the low position water tank to
floor layer are adiabatic at the entrance. The control the high position water tank through circulating
volumes at the outlet are considered to be adiabatic. pump. And then, the hot water flows from high pos-
The boundary conditions are defined as follows: ition water tank into the heating pipes to heat the test
room and then flows through the rotameter. Finally,
dtp ð0Þ
While i ¼ 0, tf ð0Þ ¼ tinlet , j the water flows into the low position water tank or
dx x¼0
the electronic scale, which is used to measure the
dta ð0Þ dtw ð0Þ
¼ jx¼0 ¼ j ¼0 mass flow rate of water as the system achieves sta-
dx dx x¼0
ble state.
(19)
dtðnÞ To make the distribution of floor surface tempera-
While i ¼ n, j ¼0 (20) ture uniform, the pipe is arranged as shown in
dx x¼L
Figure 5. The pipe used in this experiment is PE-RT
To calculate the temperature of each node, a simula- (polyethylene of raised temperature resistance) pipe.
tion program is developed based on the Matlab program- The dimensions of the testing room are 4 m (length)-
ing language and solved using the iterative method. The 3 m (width)2.8 m (height). The inner diameter of
physical and operating parameters of the new lightweight the pipe is 20 mm and the pipe length is 55 m. The
RFHS are input, such as the supply water temperature distance between two pipes is 250 mm. Figure 6 shows
tinlet and water flow velocity u. The temperature of heat the structure of the heating pipe. The supply water
transfer units in each layer can be calculated with the temperature is varied from 32 to 40
C. The water vel-
energy balance equations above. The calculation results ocity is changed from 0.1 to 0.3 m/s and it remains
of tf, tp, ta and tw in each node are output. the same in each test. Ten groups of tests in different
working conditions are conducted. Each test lasts for
Experimental verification 1.5 h and there is half an hour interval during each
test to guarantee that the test room environment
Experimental system
returns to the original state. After five group tests of
The thermal performance experiments of the new different supply water temperatures, the water flow
lightweight RFHS are carried out. The compound velocity changes and the system is tested again.
 HEAT TRANSFER ENGINEERING 7

Figure 5. Heating pipe layout diagram.

Figure 6. Structure of the heating pipe.

The supply and return water temperature are meas-

ured by platinum resistance sensors. The flow rate
and supply water temperature are controlled by the
circulating pump and heating wire. Figure 7 shows
the distribution of test points. The ground is split into
nine equal parts. The test points from 1 to 9 are set
in the middle of each part on the floor to record the
dynamic change in temperature. The temperature of
each measuring point on the floor is measured by
copper-constantan thermocouple sensors. The infrared
spectrometer is used to measure the average floor sur-
face temperature and record the floor surface tem-
perature distribution. The monitored temperatures are
Figure 7. Sketch of test points distribution.
collected by the data acquisition instrument.

temperature. The supply water temperature varied

Model validation
from 32 to 40
C and the water velocity varied from
The calculation results and experimental data are 0.1 to 0.3 m/s.
compared to verify the accuracy of the theoretical In this article, the average deviation and Mean Bias
model. Figure 8 shows the comparison results of the Error (MBE) methods are adopted to process the data
theoretical calculations and experimental values on and evaluate the accuracy of the theoretical calculation
the return water temperature and average floor surface results, where MBE is defined as
8 M. LIU ET AL.

Figure 8. The comparison results of experimental and theoretical values: (a) return water temperature and (b) average floor sur-
face temperature.

Table 1. Average deviations and MBEs.

Return water temperature Average floor surface temperature


Water velocity (m/s) Average deviation ( C) MBE (%) Average deviation (
C) MBE (%)
0.1 0.68 2.16 0.47 1.76
0.3 0.80 2.37 0.53 2.17

X
n
jSðiÞ  EðiÞj MBEs of return water temperatures are 2.16% and
MBE ¼ ð1=NÞ  100% (21) 2.37%. On the other hand, the average deviations of
i¼1
EðiÞ
average floor surface temperatures are 0.47 and
The average deviations and MBEs are shown in 0.53
C, and the MBEs between theoretical calculation
Table 1. The average deviations of return water tem- results and experimental values of the average floor
peratures between the theoretical calculation results surface temperatures are 1.76% and 2.17%. All the
and experimental values are 0.68 and 0.80
C. The MBEs are <3%, which is acceptable in this research.
 HEAT TRANSFER ENGINEERING 9

Figure 9. Average floor surface temperature for different supply water temperatures.

Figure 10. Heat transfer rate for different supply water temperatures.

The deviation is mainly caused by the unstable experi- As shown in the figures, the average floor surface
mental conditions and the errors of the experimental temperature and heat transfer rate rise with the
equipment. Therefore, the theoretical model is reliable increase of supply water temperature, and the tem-
and accurate enough to predict the thermal perform- perature difference between the floor surface and
ance of the new RFHS for further studies. indoor air increases. Therefore, the heat transfer
between floor and indoor air is promoted. As shown
in Figure 9, the average floor surface temperature
Results increases by almost 0.32
C with the supply water
temperature increasing by 1
C. The floor surface tem-
Effects of supply water temperature
perature is 27.8
C with the water flow velocity of
The effects of supply water temperature on the ther- 0.2 m/s and supply water temperature of 46
C. All the
mal performance are studied through the theoretical temperatures are within the prescribed scope and the
model. The temperature is varied from 32 to 46
C. supply water temperature should not exceed 46
C
The results are shown in Figures 9 and 10. when the water flow velocity is 0.2 m/s.
10 M. LIU ET AL.

Figure 11. Average floor surface temperature for different water flow velocities.

Figure 10 shows the correlation between the heat thermal performance of the proposed heating system
transfer rate and the supply water temperature. The is improved with the increase of water flow velocity,
heat transfer rate increases linearly with the increase but the growth rate is deteriorated.
of supply water temperature. Taking the flow velocity
of 0.2 m/s as an example, when the supply water tem-
Effects of aluminum layer
perature rises from 32 to 46
C, the heat transfer rate
increases from 39.8 to 72.2 W/m2. On the average, the To study the effects of the aluminum layer on the sys-
heat transfer rate increases by 2.3 W/m2 when the tem thermal performance, the lightweight RFHS with
supply water temperature increases by 1
C. aluminum layer and without aluminum layer were
tested, respectively. The air temperature of the test
chamber was set at 18
C and the water velocity was
Effects of water flow velocity
0.2 m/s. The experimental results are shown in Figures
The velocity of supply water is another important par- 13 and 14.
ameter that can be controlled in the system. It also It can be seen from Figures 13 and 14, the average
has a great impact on the thermal performance of the floor surface temperature and heat transfer rate of the
system. The water flow velocity is varied from 0.1 to RFHS with aluminum layer are obviously higher than
0.3 m/s. Through the calculation of theoretical model, that without aluminum layer. And as the supply water
the results are given below. temperature increases, the growth rate of the system
As shown in Figures 11 and 12, the average floor with aluminum layer is higher. The average floor sur-
surface temperature and heat transfer rate increase face temperature of the proposed system with alumi-
with the water flow velocity when the supply water num layer is improved by 8.2% and the heat
temperature is constant. While the increasing trend transfer rate is improved by 37.4% while the supply
gradually slows down. On the condition that the sup- water temperature is constant. When the average floor
ply water temperature is 32
C, the average floor sur- surface temperature and the heat transfer rate of the
face temperature increases by 0.49
C and the heat two systems achieve the same values, the supply water
transfer rate increases by 4.84 W/m2 as the water flow temperature of the system with aluminum layer is
velocity increases from 0.1 to 0.15 m/s. However, the lower than the system without aluminum layer. The
average floor surface temperature and the heat trans- supply water temperature can be reduced by 6–8
C in
fer rate only increase by 0.34
C and 1.51 W/m2, the system with aluminum layer.
respectively, as the water flow velocity increases from The infrared spectrometer records the heat transfer
0.25 to 0.3 m/s. The growth rate of the average floor rate intuitively. Figure 15 shows the floor temperature
surface temperature is reduced from 2.3% to 1.5% and distribution of the system with and without aluminum
it is reduced from 15.8% to 3.6% for the growth rate layer as the supply water temperature is 38
C.
of heat transfer rate. It can be concluded that the Obviously, the temperature distribution on the RFHS
 HEAT TRANSFER ENGINEERING 11

Figure 12. Heat transfer rate for different water flow velocities.

Figure 13. Comparison of average floor surface temperature with and without the aluminum layer.

Figure 14. Comparison of heat transfer rate provided by the system with and without the aluminum layer.

with aluminum layer is more uniform and the average grows from 1.4 to 2.2
C while they are 3.5–6.4
C for
temperature is higher than that without aluminum the system without aluminum layer.
layer. As the supply water temperature increases from To sum up, the aluminum layer can reduce the
32 to 40
C, the floor surface temperature difference supply water temperature and uniform the floor sur-
in distribution of the system with aluminum layer face temperature distribution, which can not only save
12 M. LIU ET AL.

Figure 15. Floor surface temperature distribution of the system with and without aluminum layer.

Figure 16. Schematic of traditional system control volumes.

the heating energy, integrate the low-temperature heat For the control volume of concrete layer (Vh ), the
source, but also eliminate the defect of uneven surface energy balance equation is
temperature distribution. Therefore, the aluminum
th ði  1Þ2th ðiÞ þ th ði þ 1Þ tp ðiÞth ðiÞ
layer has great influence on the thermal performance k h Ah þ Ah, p dp
of the new lightweight RFHS. Dl þ dh
2kp 2kh
tm ðiÞth ðiÞ
þ Am, h ¼0
2km þ 2kh
dm dh
Discussion
Theoretical model of traditional heavy RFHS (22)

Comparative research between the new lightweight where the thermal conductivity of the concrete kh is
RFHS and the traditional RFHS are conducted. Firstly, 1.51 W/(m K), and the thickness dh is 60 mm.
the model of traditional heavy RFHS is built based on For the control volume of cement mortar layer
the energy balance equations. The traditional system (Vm ), the energy balance equation is
is divided into five control volumes. The schematic of
tm ði  1Þ2tm ðiÞ þ tm ði þ 1Þ th ðiÞtm ðiÞ
control volumes are shown in Figure 16. The model km Am þ Am, h d
Dl 2k þ 2k
m dh
construction method is the same with the new light- m h

weight RFHS. The calculation process is also specified tw ðiÞtm ðiÞ

þ Aw , m d w ¼0
in i node, i ¼ 0, 1, 2, … , n. Therefore, the equations 2kw þ 2km
dm

are not introduced in detail, an example of i ¼ 1, … , (23)

n  1 is described.
While i ¼ 1, … , n  1, the energy balance equation where the thermal conductivity of the cement mortar
for the control volume of fluid (Vf ) is the same as layer km is 0.93 W/(m K), and the thickness dm
Eq. 8. is 20 mm.
For the control volume of pipe wall (Vp ), energy For the control volume of wooden floor layer (Vw ),
balance equation is the same as Eq. 12. the energy balance equation is the same as Eq. 14.
 HEAT TRANSFER ENGINEERING 13

Figure 17. Comparison of average floor surface temperature of the two systems.

Figure 18. Comparison of heat transfer rate of the two systems.

Other physical parameters are the same with the geothermal energy. Figure 18 shows that the heat
proposed lightweight system. The Matlab software and transfer rate of the new system is 9.6% higher than
the iterative method were adopted to solve this model. that of the traditional system, which shows an obvious
energy-saving potential. Therefore, the proposed sys-
Thermal performance comparison tem can lower the supply water temperature and
achieve greater heat transfer rate due to its outstand-
The thermal performance of the proposed RFHS and ing thermal performance.
the traditional heavy RFHS is compared with the veri-
fied theoretical model. The distances between two
pipes in the two systems are both 250 mm. The indoor
Economic performance comparison
air temperatures are both set at 18
C. Figures 17 and
18 show the comparison of floor surface temperature Economic performance is another important index to
and heat transfer rate between the proposed system evaluate the system. Table 2 shows the comparison of
and the traditional system. As illustrated in Figure 17, the initial cost between the proposed RFHS and the
the average floor surface temperature of the proposed traditional one. As can be seen from the comparison
system is almost 6.3% higher than that of the trad- results, the initial cost of the proposed system is 49%
itional system, which means that the supply water cheaper than that of the traditional one.
temperature can be reduced by 5
C in the proposed The city of Tianjin, China is used as an example to
system under the same floor surface temperature. calculate the operating costs in this section. The num-
Therefore, the proposed system is more promising to ber of heating days in Tianjin is 118 and the heating
be integrated with the low-temperature heat source, load of unit area is calculated as 45 W/m2. According
such as solar energy, air-source heat pump, and to the local charge standard, the heating price in
14 M. LIU ET AL.

Table 2. Comparison of initial cost.

The new type of lightweight system Traditional heavy system
Specification Price Specification Price
Items thickness, mm USD/m2 of installation Items thickness, mm USD/m2 of installation
PE-RT heating pipe 20 1.36 PE-RT heating pipe 20 1.36
Heat-insulating layer 40 2.89 Heat-insulating layer 40 2.89
Aluminum layer 0.12 0.89 Concrete 60 4.44
Plastics clamps 0.24 Cement mortar 20 1.63
Plastics clamps 0.24
Total price 5.38 Total price 10.56

heating system. The heat transfer rate of the proposed

Table 3. Economy comparison.
The new type of system Traditional system
system is higher than the traditional one. In other
Initial cost (USD/m2) 5.38 10.56 words, the proposed system can adjust indoor tem-
Operating cost (USD/m2) 2.21 2.41 perature more quickly, so as to reach the required
Annual cost (USD/m2) 2.84 3.65
indoor environment. In addition, the initial cost and
operating cost are both cheaper than the traditional
Tianjin is 0.0187 USD/kWh and the electricity price one. Therefore, the new lightweight RFHS has a better
is 0.0705 USD/kWh. prospect of application.
The annual heating consumption of the room can
be calculated by Eq. 24 [24]:
ti tv Conclusions
Qa ¼ 0:0864  Na  Qh  (24)
ti  to This article puts forward a new type of lightweight
The system is powered by the circulating pump RFHS with aluminum layer. The thermal performance
which was shown in Figure 4. The circulating pump of this system is discussed through theoretical and
power can be calculated by Eq. 25: experimental methods. The following conclusions
1:25cQH are drawn:
P ¼ 1:25W ¼ (25)
1000g
1. A simplified theoretical model of the new light-
The operating cost of the system can be calculated weight RFHS is established and validated by
by Eq. 26: experiments. Return water temperature and aver-
Qa  103 age floor surface temperature are compared
C¼  0:0187 þ P  Na  24  0:0705 between theoretical and experimental values. The
3600
(26) Mean Bias Errors are all <3%.
2. The average floor surface temperature and heat
The annual cost of the system can be calculated by transfer rate increases linearly with the increase of
Eq. 27: supply water temperature. With the increase of
!
Ið1 þ IÞy water flow velocity, the growth rate of the average
Z ¼ P0  þC (27) floor surface temperature and heat transfer rate
ð1 þ IÞy  1
gradually slow down.
In this article, the life cycle y is calculated as 3. The aluminum layer has great influence on the
20 years and the discount rate I is 10%. The economy thermal performance of the proposed lightweight
comparison of the two systems is shown in Table 3. RFHS. Compared with the system without alumi-
The annual cost of the new system is only 77.8% of num layer, the temperature of supply water can
the traditional one. be reduced by 6–8
C for the system with alumi-
Through the comparison research, it can be seen num layer. The average floor surface temperature
that the new lightweight RFHS has advantages in both can be improved by 8.2% and the heat transfer
thermal performance and economy. Lower the supply rate can be improved by 37.4% for the system
water temperature can effectively save heating energy. with aluminum layer.
The low grade heat sources such as solar energy, air- 4. The average floor surface temperature and heat
source heat pump, and geothermal energy can be uti- transfer rate of the new lightweight RFHS are
lized in the proposed system. It can provide great 6.3% and 9.6% higher than those of the trad-
convenience especially for the village without central itional RFHS, respectively. And the annual cost of
 HEAT TRANSFER ENGINEERING 15

the new system is only 77.8% of the trad- [2] B. Lin, Z. Wang, H. Sun, Y. Zhu, and Q. Ouyang,
itional one. “Evaluation and comparison of thermal comfort of
convective and radiant heating terminals in office
buildings,” Build. Environ., vol. 106, pp. 91–102, Sep.
Notes on contributors 2016. DOI: 10.1016/j.buildenv.2016.06.015.
[3] M. Abdolzadeh, M. S. Sargazizadeh, and M. H.
Dehghan, “Quantitative and qualitative energy
Minzhang Liu is a PhD candidate in the assessments of floor heating and skirting board heat-
School of Environmental Science and ing systems in a room,” J. Energy Eng., vol. 143, no.
Engineering, Tianjin University, Tianjin, 1, pp. 04016032, Feb. 2017. DOI: 10.1061/(ASCE)EY.
China. She is pursuing the doctorate under 1943-7897.0000385.
the supervision of Dr Huan Zhang. Her [4] G. Zheng et al., “Operation strategy analysis of a
current research area involves radiant floor geothermal step utilization heating system,” Energy,
heating system, heat and mass transfer vol. 44, no. 1, pp. 458–468, Aug. 2012. DOI: 10.
mechanism, and tunnel ventilation system. 1016/j.energy.2012.06.006.
She has published three international journal papers and [5] O. B. Kazanci, M. Shukuya, and B. W. Olesen,
one conference paper. “Exergy performance of different space heating sys-
tems: A theoretical study,” Build. Environ., vol. 99,
pp. 119–129, Apr. 2016. DOI: 10.1016/j.buildenv.
Huan Zhang is a Professor and the 2016.01.025.
[6] M. S. Shin, K. N. Rhee, S. R. Ryu, M. S. Yeo, and
Department Head of Building
K. W. Kim, “Design of radiant floor heating panel in
Environment & Energy Engineering in the
view of floor surface temperatures,” Build. Environ.,
School of Environmental Science and
vol. 92, pp. 559–577, Oct. 2015. DOI: 10.1016/j.buil-
Engineering, Tianjin University, Tianjin,
denv.2015.05.006.
China. Her main research area is heat and
[7] B. C. Ahn and J. Y. Song, “Control characteristics and
mass transfer mechanism, performance
heating performance analysis of automatic thermo-
improvements of heat pump system, build-
static valves for radiant slab heating system in residen-
ing energy conservation, and heat utilization of renewable
tial apartments,” Energy, vol. 35, no. 4, pp. 1615–1624,
energy. She is the Director of Tianjin Association of
Apr. 2010. DOI: 10.1016/j.energy.2009.11.007.
Refrigeration Air Conditioning Heat Pump Committee.
[8] F. Causone, F. Baldin, B. W. Olesen, and S. P.
Corgnati, “Floor heating and cooling combined with
displacement ventilation: Possibilities and limi-
Wandong Zheng is an Associate Professor tations,” Energy Build., vol. 42, no. 12, pp.
in the School of Environmental Science 2338–2352, Dec. 2010. DOI: 10.1016/j.enbuild.2010.
and Engineering, Tianjin University, 08.001.
Tianjin, China. His main research area is [9] K. N. Rhee, and K. W. Kim, “A 50 year review of
heat pump system, solar thermal utiliza- basic and applied research in radiant heating and
tion, and heat and mass transfer mechan- cooling systems for the built environment,” Build.
ism. He has more than 30 international Environ., vol. 91, pp. 166–190, Sep. 2015. DOI: 10.
journal publications. 1016/j.buildenv.2015.03.040.
[10] K. N. Rhee, B. W. Olesen, and K. W. Kim, “Ten
questions about radiant heating and cooling sys-
Shijun You is a Professor in the School of tems,” Build. Environ., vol. 112, pp. 367–381, Feb.
Environmental Science and Engineering, 2017. DOI: 10.1016/j.buildenv.2016.11.030.
Tianjin University, Tianjin, China. His [11] M. Rahimi and A. Sabernaeemi, “Experimental study
main research area is heat and mass trans- of radiation and free convection in an enclosure
fer mechanism, air-conditioning system, with a radiant ceiling heating system,” Energy Build.,
and heat pump system. He is the director vol. 42, no. 11, pp. 2077–2082, Nov. 2010. DOI: 10.
of Tianjin Association of Architecture 1016/j.enbuild.2010.06.017.
HVAC Committee, the deputy director of [12] S. H. Cho and M. Zaheer-Uddin, “An experimental
National Association of HVAC Heating Committee, and the study of multiple parameter switching control for
member of Chinese Association of Refrigeration Air condi- radiant floor heating systems,” Energy, vol. 24, no. 5,
tioning Heat Pump Committee. pp. 433–444, May. 1999. DOI: 10.1016/S0360-
5442(98)00101-7.
References [13] X. Wu, J. Zhao, B. W. Olesen, L. Fang, and F. Wang,
“A new simplified model to calculate surface tempera-
[1] J. Dong, L. Zhang, S. Deng, B. Yang, and S. Huang, ture and heat transfer of radiant floor heating and
“An experimental study on a novel radiant-convect- cooling systems,” Energy Build., vol. 105, pp. 285–293,
ive heating system based on air source heat pump,” Oct. 2015. DOI: 10.1016/j.enbuild.2015.07.056.
Energy Build., vol. 158, pp. 812–821, Jan. 2018. DOI: [14] H. Khorasanizadeh, G. A. Sheikhzadeh, A. A.
10.1016/j.enbuild.2017.10.065. Azemati, and B. S. Hadavand, “Numerical study of
16 M. LIU ET AL.

air flow and heat transfer in a two-dimensional radiant panels,” presented at the 9th Int. Symposium
enclosure with floor heating,” Energy Build., vol. 78, on Heating, Ventilation and Air Conditioning, vol.
no. 78, pp. 98–104, Aug. 2014. DOI: 10.1016/j. 121, pp. 1349–1355, Jul. 12–25, 2015. DOI: 10.1016/j.
enbuild.2014.04.007. proeng.2015.09.018.
[15] M. Spiga and P. Vocale, “Effect of floor geometry on [20] M. Rahimi and A. Sabernaeemi, “Experimental study
building heat loss via the ground,” Heat Transf. Eng., of radiation and free convection in an enclosure
vol. 35, no. 18, pp. 1520–1527, Dec. 2014. DOI: 10. with under-floor heating system,” Energy Convers.
1080/01457632.2014.897560. Manage., vol. 52, no. 7, pp. 2752–2757, Jul. 2011.
[16] L. Fontana, “Thermal performance of radiant heating DOI: 10.1016/j.enconman.2011.02.020.
floors in furnished enclosed spaces,” Appl. Therm. [21] G. Yu, W. Guo, and H. Chen, “Thermal performance
Eng., vol. 31, no. 10, pp. 1547–1555, Jul. 2011. DOI: of precast light floor radiant panel,” presented at the
10.1016/j.applthermaleng.2010.12.014. 10th Int. Symposium on Heating, Ventilation and
[17] R. Hu and J. L. Niu, “A review of the application of Air Conditioning, vol. 205, pp. 198–204, Oct. 19–22,
radiant cooling & heating systems in Mainland 2017. DOI: 10.1016/j.proeng.2017.09.953.
China,” Energy Build., vol. 52, no. 52, pp. 11–19, [22] D. Zhang, N. Cai, and Z. Wang, “Experimental and
Sep. 2012. DOI: 10.1016/j.enbuild.2012.05.030. numerical analysis of lightweight radiant floor heat-
[18] Y. Wang, Z. Zhang, F. He, and X. Liu, ing system,” Energy Build., vol. 61, pp. 260–266, Jun.
“Experimental investigations on thermal performance 2013. DOI: 10.1016/j.enbuild.2013.02.016.
of a new floor heating device,” presented at the 9th [23] A. Bejan and A. D. Kraus, Heat Transfer Handbook.
Int. Symposium on Heating, Ventilation and Air Hoboken, NJ: John Wiley & Sons, Inc., 2003.
Conditioning, vol. 121, pp. 2132–2140, Jul. 12–25, [24] “Industrial Standard of the People’s Republic of
2015. DOI: 10.1016/j.proeng.2015.09.084. China, CJJ 34-2010,” Design Code for City Heating
[19] G. Yu and Y. Yao, “The experimental research on Network. Beijing, CN: China Architecture & Building
the heating and cooling performance of light floor Press, 2010.