Building and Environment 98 (2016) 158e179

Contents lists available at ScienceDirect

Building and Environment

journal homepage: www.elsevier.com/locate/buildenv

A glazed transpired solar wall system for improving indoor

environment of rural buildings in northeast China
Tiantian Zhang*, Yufei Tan, Xuedan Zhang, Zhigao Li
School of Municipal and Environmental Engineering, Harbin Institute of Technology, Harbin 150090, China

a r t i c l e i n f o a b s t r a c t

Article history: In rural areas of northeast China, residents often use a traditional Chinese kang for domestic heating.
Received 19 October 2015 However, this heating method is associated with an uncomfortable thermal environment and poor in-
Received in revised form door air quality. Therefore, this paper proposes a glazed transpired collector (GTC) based solar wall
12 January 2016
system, as a solution. Similar to air conditioning systems, the GTC is able to generate various airflow
Accepted 13 January 2016
settings, which is one creative concept of the proposed system. Experimental and numerical studies were
Available online 14 January 2016
conducted to investigate the space heating and indoor air-quality improving performance of the system.
The experimental results show that the average indoor temperatures can be increased from 12.12
C to
Keywords:
Transpired solar collector
16.17e18.19
C, relevant to the various operating modes. The return-air mode reduces the concentrations
Operating modes of CO, CO2, PM2.5, and PM10 by 34.8%, 20.3%, 14.4%, and 11.6%, respectively. The reductions relevant to the
Indoor air temperature mixture mode are 69.6%, 28.0%, 45.0%, and 41.6%, respectively, and those to the fresh-air mode are 73.9%,
Indoor pollutants 42.7%, 56.2%, and 58.1%, respectively. By numerically simulating the distributions of indoor air temper-
Operating strategy ature and the concentrations of CO, CO2, PM2.5, and PM10, the GTC-based solar wall system is able to
Rural dwellings enhance the uniformity of temperature distribution and dilute indoor pollutant concentrations. To
conclude, a summary of the simple and easy operating manual of the solar wall system is presented in
this paper. The solar wall system can solve the problem of unacceptable indoor thermal comfort and
indoor environmental pollution in rural dwellings in cold climates.
© 2016 Elsevier Ltd. All rights reserved.

1. Introduction smoke produced by the stove, which burns biomass such as

In the cold regions of rural China, people often use a traditional
stove-kang (Chinese kang) system to heat their dwellings. The
stove-kang is an effective system that uses local biomass for heat-
ing, ameliorating the energy and environmental problems associ-
ated with rural residential buildings [1]. The system probably dates
back 2500 years, and is still widely used today [2]. A large-scale
field survey in 2006 indicated that this ancient system was still
used for domestic heating in nearly 85% of rural homes, by 175
million people [3]. A typical Chinese kang comprises a stove, the
kang body, and a chimney, and is often constructed from masonry,
adobe, and cement [4], as shown in Fig. 1. Various functions, such as
cooking, a place for sleeping, domestic heating, and ventilation are
integrated in this one system. In most of the kang applications, the
stove is placed in the kitchen, additionally serving as cooking place.
An inner chamber functions as a flue for the high-temperature
* Corresponding author.
E-mail address: x418298537@163.com (T. Zhang).
straw, corncobs, wood, and the like. Heat is transferred from the
smoke to the massive structure as the smoke flows through the
chamber to the chimney and subsequently outside. The massive
structure stores the heat and gradually releases it to the indoor
space to keep the room warm for several hours after the combus-
tion process.
The kang system can maintain the average indoor air temper-
ature at 13e16
C during the day [5]. Using straw, corncobs, and
wood as fuel helps to reduce fossil fuel consumption for cooking
and domestic heating. To some extent, the system solves the energy
and environmental problems related to rural residential buildings.
However, applying the system causes significant other problems. As
the main structures of the detached houses in the rural areas are
usually poorly insulated, the general indoor thermal condition re-
mains relatively unsatisfactory if the kang is the only heating
source. The kang system alone does not provide enough heat to
increase and maintain the indoor temperature to a comfortable
level on the coldest winter days. The average indoor temperature
often drops below 13
C [6], the acceptable minimum indoor

http://dx.doi.org/10.1016/j.buildenv.2016.01.011
0360-1323/© 2016 Elsevier Ltd. All rights reserved.
 T. Zhang et al. / Building and Environment 98 (2016) 158e179
Fig. 1. An illustration of the elevated kang system.
temperature according to the standard for rural residential build-
ings [7]. Furthermore, in winter, the stove is used twice daily,
namely, in the morning for cooking breakfast and late in the eve-
ning for cooking dinner. Sometimes, the kang is used again at
midday to cook lunch. Consequently, the indoor air temperature
can fluctuate wildly [8]. During the combustion period and for
several hours afterwards, the body of the kang remains warm and
transfers heat to the indoor air, raising the temperature to even
more than 20
C. However, eventually the air temperature drops,
sometimes to below 10
C, especially during the night [9]. The low
temperature during non-combustion periods and the temperature
fluctuations adversely affect the indoor thermal comfort. A survey
on the indoor thermal comfort in ten rural houses in two villages
near Harbin has indicated that the average indoor temperature was
12.3
C and the magnitude of temperature variation was 7.8
C.
Forty percent of the occupants indicated that they experienced cold
or intense cold on winter days and 60% were cold during winter
nights. In the daytime, the occupants usually wear thick clothes,
with an average clothing insulation of 1.47clo, to keep warm [10].
The poor indoor air quality of rural buildings, caused by fuel
combustion, is of serious concern [11]. The design and construction
of the traditional kang is mostly directed by intuition and the
practical knowledge of the craftsmen, but there is no scientific
159
theory or basis to ensure effective functioning. Consequently, the
heating efficiency of the traditional kang system is less than 20%
[12], and that of the improved elevated system is 45e60% [13]. The
low energy efficiency results in a significant waste of energy and
the system requires large amounts of fuel to maintain the indoor
thermal environment. The occupants often supplement the kang
system with coal-fired boilers or radiators to improve the indoor
temperature during the non-combustion periods [14]. The fuel
combustion process releases various types of pollutants to the in-
door space, including CO, particulate matters (PM2.5 and PM10), and
others. According to Zhang [15], a field test on the indoor envi-
ronment of 12 rural residential houses heated by the Chinese kang
has indicated that the average indoor temperature was 10e13
C,
i.e., far below the comfort level. Moreover, the main pollutants
found were CO2, particulate matters, NOx, and SO2, and their con-
centrations exceeded the acceptable standards. The average con-
centrations of CO2, PM2.5, NOx, and SO2 reached 1514 ppm,
0.425 mg/m3, 0.24 ppm, and 0.63 ppm, respectively. Wang et al.
[16] investigated the indoor air quality of rural dwellings in the
severe cold regions of China and concluded that the over-standard
ratios of PM2.5, PM10, SO2, NOx, and CO2 were 97%, 79%, 99%, 63%,
and 71%, respectively. According to another survey, more than 40%
of the farmers reported heavy indoor smoke during the cooking
160 T. Zhang et al. / Building and Environment 98 (2016) 158e179
periods. Numerous studies have shown that in the various prov-
inces of China the indoor particulate matter and gaseous pollutant
concentrations related to solid fuel combustion greatly exceeded
the national standards for indoor air quality [17]. Excessive
pollutant concentrations pose a high risk to the environment and
human health. Clear evidence exists for the association between
the air pollutants in dwellings, caused by solid fuel combustion, and
the national burden of diseases, such as respiratory infections,
respiratory tract cancers, lung cancer, and other chronic diseases
[18,19].
Using the kang for heating in rural dwellings causes mainly two
problems, namely, an uncomfortable thermal environment and
poor indoor air quality. Increasing the fuel amounts or the space
heating facilities should solve the first problem, but would lead to
an increase in pollutant emissions. Ventilation would reduce the
indoor pollutant concentrations by inducing the outdoor fresh air
to dilute the concentrations; in return, however, ventilation would
increase the demand for indoor heating. Consequently, in winter, a
conflict exists in rural dwellings between the demand for heating
and the need for ventilation. Increasing the amount of ventilation
would lead to a significant increase in the demand for heating, as
the outdoor air can be extremely cold. Resolving this conflict poses
a major challenge to efforts to improve indoor thermal comfort and
air quality. A potential solution is utilizing solar energy to pre-heat
the outdoor fresh air before it is sent indoors. In such instance, the
solar air collectors should be adapted to convert solar radiation to
thermal energy and to transfer the absorbed thermal energy to the
fresh air [20].
Solar air collectors can be used for various purposes, such as
drying textiles, marine and agricultural products, and heating in-
door spaces to obtain a comfortable environment, particularly in
extremely cold climates [21]. These collectors can be broadly clas-
sified as flat-plate and transpired collectors, with the transpired
collectors having been proven more efficient for converting solar
energy [22e25]. The transpired solar collectors can be further
classified into unglazed transpired collectors (UTC) and glazed
transpired collectors (GTC) [26,27]. UTC was introduced in the early
1990s to utilize solar energy for ventilation and hot-air supply [28].
The UTC system usually consists of a perforated and blackened
absorber plate, with an air cavity between the back wall and the
plate. A schematic diagram of a building-integrated UTC system is
presented in Fig. 2.
The transpired absorber used in the UTC system is often con-
structed of metal plate, with uniformly distributed perforations,
and with selective coatings to absorb and transform the solar ra-
diation. The plate is fixed approximately 100e300 mm from the
Fig. 2. Schematic diagram of building integrated UTC system.
south-facing external wall (the back sheet). Consequently, an air
channel is formed in the sealed space between the absorber surface
and the wall surface. The size of the perforations usually varies
between 0.5% and 2% of the total surface area [29]. The ambient air
flows through these small perforations, absorbing heat energy from
the metal plate and the heated wall surface. Subsequently, the
heated air is drawn by a fan through an outlet vent, located at the
top of the back sheet, and into the indoor space. In hot weather, the
heated air is exhausted directly to the atmosphere, or, making full
use of the system, it can be used for domestic water heating [30]. It
has been reported that approximately 62% of the heat gain of the air
is obtained from the absorber plate, and 28% and 10% from the
perforations and the back plate, respectively [31]. The performance
of a UTC is influenced by factors that include the porosity of the
metal plate, the size and shape of the perforations, the dimensions
of the air cavity, the material and absorption coefficient of the
collector, the suction velocity, the air velocity inside the cavity, the
crosswind velocity, the ambient conditions [32e35], and others.
A percentage of the solar energy that should have been absorbed
could be reflected or lost when a conventional flat-plate solar air
collector with no perforations is used. For instance, the glass cover
could reflect up to 15% of the total solar radiation back into space
[26]. Moreover, a crosswind would strengthen the convective heat
transfer between the ambient air and the glass cover. In compari-
son, a UTC does not have such weaknesses and is therefore more
efficient. The temperature of the metal plate of the UTC remains at a
lower level because of the heat transfer between the air sucked in
and the plate, which minimizes the loss of radiation heat. Accord-
ingly, more than 80% of the incoming solar radiation is absorbed. In
addition, the adverse effect of the crosswind could be eliminated
when the suction velocity is maintained at 0.03e0.05 m/s [36,37].
The research on and applications of UTC over the last 20 years
have been remarkable. A host of studies has investigated the heat
transfer characteristics of the system, the thermal efficiency,
airflow distribution, and the like [38e41]. Numerous UTC-based air
heating systems, crop drying systems, and cogeneration systems
have been set up and operated throughout the world [42e46].
Essentially, the feasibility of the UTC for such industrial applications
has been confirmed. However, this successful application of UTC
could not be replicated to provide space heating for buildings in
extremely cold climates [47]. The reason is that the main aim of
these industrial systems was to obtain higher airflow rates, with
only a reasonable temperature rise. However, the air temperature
obtained in this way is far colder than that required for space
heating. In relatively low outdoor air temperatures, lower suction
velocity is crucial to maintaining sufficient contact time between
the air and the absorber plate of the UTC system in order to achieve
a sufficient air temperature rise. Nevertheless, the low suction ve-
locity is unable to stop the adverse effect of significant heat loss
caused by convective heat transfer between the outside crosswind
and the surface of the metal plate. According to Pesaran [36], with a
suction velocity of 0.05 m/s, the temperature rise would be only
12e13
C above the ambient temperature. Cordeau and Barrington
[48] have pointed out that with a suction velocity between 0.012 m/
s and 0.016 m/s, the temperature of the supplied air is only between
10
C and 18
C during March. In addition, the efficiency of the UTC
system relies heavily on the outdoor wind velocity. In tests, this
efficiency was indicated as more than 65% when the outside wind
velocity was below 2 m/s. However, the efficiency dropped below
25% when the crosswind velocity exceeded 7 m/s [48]. Accordingly,
in an extremely cold climate, it is virtually impossible to apply UTC
in buildings, especially during the winters with typical strong
winds.
However, the GTC system provides a viable alternative to uti-
lizing transpired collectors in extremely cold and blustery weather
 T. Zhang et al. / Building and Environment 98 (2016) 158e179
conditions [49]. In the GTC system, compared with the UTC, a glass
cover with air vents is added. The configuration of the GTC is shown
in Fig. 3. The transpired absorber plate is mounted 100e200 mm
from the back wall surface, and the distance between the absorber
plate and the glass cover is approximately 100e200 mm. A fan
draws the outdoor air into the outer cavity between the glazing and
the absorber plate and subsequently through the transpired
absorber plate into the inner cavity, between the transpired
absorber plate and the back plate. Finally, the air is drawn through
air ducts into the indoor space that has to be heated. As solar energy
is absorbed by the metal plate, the air is heated when it is drawn
through the small holes on the absorber plate. At night, the air inlet
and the fan are closed to form an enclosed cavity in the solar wall
structure. The heat loss of the external wall is absorbed by the in-
ternal air of the enclosed cavity, which acts as an air insulation layer
for the external envelopes.
Even though the glass cover reflects part of the solar radiation,
nevertheless, the GTC system has several advantages. Firstly, the
glass cover separates the transpired collector from the ambient,
weakening the effect of the outdoor wind velocity. Therefore, GTC
can be utilized in blustery weather conditions. Secondly, the glass
cover prolongs the period the sucked air stays in the internal cavity,
thereby enabling a higher temperature rise and increasing the
space heating ability. Additionally, during the night, the GTC system
can be converted to the completely closed mode, achieving supe-
rior insulating performance compared with the UTC system. The
internal cavity of the UTC system cannot be closed completely
because of the numerous perforations in the absorber plate. Gao
[49] has conducted numerical research to compare the perfor-
mance of the GTC and UTC systems. The results indicated that in no-
crosswind conditions, the air temperature improvement achieved
by the GTC was approximately 3.4
C below that achieved by the
UTC, and the overall thermal efficiency was approximately 13.0%
lower. This finding can probably be ascribed to the glass cover,
which reduces the solar radiation. However, when the crosswind
velocity increased, the air temperature rise achieved by the UTC
decreased, whereas that of the GTC remained almost constant.
When the crosswind velocity exceeded 3.0 m/s, the exit tempera-
ture of the GTC was higher than that of the UTC. The author
concluded that in higher crosswind velocity conditions, the GTC
had superior thermal efficiency compared with the UTC. When the
indoor air was recirculated to the GTC system and the outdoor air
temperature was 20
C, the supply air temperature could be
maintained above 20
C, provided that the solar radiation was
above 150 W/m2. A case study at a five-story hotel in Harbin
revealed that the annual average solar fraction of a GTC-based solar
Fig. 3. Schematic diagram of building integrated GTC system.
161
air heating system was approximately 20%. This figure was twice
that of an air heating system based on a flat-plate collector, and
nearly nine times that of the UTC-based system. Consequently, in
cold climates, a large amount of energy can be saved by utilizing the
GTC system for indoor space heating.
Although a number of studies have been conducted on the space
heating performance of transpired solar collectors, these studies
mainly relied on laboratory tests or numerical simulation. Rela-
tively few studies have been carried out in actual buildings, with
actual outdoor conditions. In addition, the benefit of the transpired
collector in reducing thermal loads in buildings has been widely
accepted, but the possible positive effect on the indoor air quality
has rarely been evaluated. This paper therefore attempts to employ
GTC to solve the energy and the environmental problems in the
kang-heated rural dwellings of northeast China. In this climate
region, the outdoor air temperature is extremely low and the out-
door air velocity is often relatively high in winter. Utilizing the GTC
for these buildings is anticipated to increase the indoor tempera-
ture and dilute the indoor pollutant concentrations.
In this study, both experimental and numerical studies were
carried out on a specially designed GTC-based solar wall system for
kang-heated dwellings in northeast China. The system was
designed to act as the main heating source for the dwellings, as well
as an indoor air cleaning system. The major objective of this study is
to investigate the space heating capacity of this solar wall system in
dynamic outdoor conditions and its ability for improving the in-
door air quality. Experimental tests were carried out to investigate
the actual performance of the system in different operating modes,
and to provide the underlying data for numerical simulation and
validation. Numerical simulations were used to evaluate the tem-
perature distributions of the indoor air and the distribution of the
CO, CO2, PM2.5, and PM10 concentrations. In addition, the operating
strategy of this system is presented.
2. Experimental methodology
2.1. The rural building and the GTC-based solar wall system
The rural house with the specially designed solar wall system is
located in Wuchang city, Jilin Province, China (127.00
east longi-
tude and 44.90
north latitude). Wuchang is located in an
extremely cold region, with a temperate continental monsoon
climate. During the period when heating is needed (winter
months), the average outdoor air temperature is
approximately 10
C and the average outdoor wind speed is
approximately 3.2 m/s. The winter is typically cold and windy, with
high solar irradiation. The house is a stand-alone building, oriented
southenorth. The size of the indoor area is 74.5 m2, and the height
is 3 m. The external length in the eastewest direction is 12 m, and
the width in the northesouth direction is 8 m. The house has four
rooms, including two bedrooms, a kitchen, and a sitting room. Solid
bricks were used for the external wall, which has a cavity insulating
layer and two plaster layers. The total thickness of the wall is
500 mm, comprising a 20 mm thick plaster layer, 240 mm brick
wall layer, 100 mm internal insulation layer, 120 mm bricklayer, and
20 mm plaster layer. The external windows, 2.2 m  1.8 m in size,
are double-glazed. Various thicknesses of polystyrene board as
insulation layers were added to the roof and floor of the house (the
thickness of the polystyrene board for the roof is 100 mm and that
for the floor is 50 mm). The bedrooms are placed on the south side
and each bedroom is equipped with a kang (4.18 m long, 2.65 m
wide and 0.6 m high). The total surface heating area is approxi-
mately 9.66 m2.
A GTC-based solar wall system, similar to that presented in
Fig. 3, was designed to serve as a space-heating source and an
162 T. Zhang et al. / Building and Environment 98 (2016) 158e179
indoor air conditioning system for a rural house. It was installed on
the external surface of the south wall (see Fig. 4). The total area of
the collector is about 18 m2. The solar wall system provides hot air
for the bedrooms. The glass cover of the solar wall reduces the
energy exchange between the solar wall and the outdoor air by
weakening the natural convection and the radiation heat transfer.
As illustrated in Fig. 5, three fresh-air inlets (20 cm  8 cm) were
placed at the bottom of the wall cavity. An air supply outlet
(20 cm  20 cm) was placed on the ceiling of each bedroom, and a
return-air inlet (20 cm  20 cm) was placed on the external wall of
each bedroom. The hot air from the solar wall cavity is distributed
to the bedrooms by circular ducts (diameter 100 mm), aided by the
air fan in each bedroom.
The solar collector comprises transpired collector units
(55 cm  20 cm), made from black-painted galvanized steel sheets,
with uniformly distributed hemispherical concaves (see Fig. 6). The
perforations on the concaves are designed to improve the perme-
ation of fresh air. The diameter of the perforations is 1 mm, with the
horizontal pitch 25 mm and the vertical 23 mm. The transpired
absorber plate was mounted 200 mm from the back plate, while the
distance between the absorber plate and the glass cover was
100 mm. The assembled collector was fixed on a holder and the
entire system was fixed to the south wall of the building. The
necessary insulating measures were taken for the solar wall system.
A creative concept relevant to this solar wall system is that the
glass cover and the return-air inlets enable the GTC to generate
various airflow settings. The outdoor air, the indoor air, or a mixture
of indoor and outdoor air can serve as the air source for the solar
collector. In winter, with proper regulation of the air vents and air
dampers, four operating modes can be achieved by the GTC-based
solar wall system, including three hot-air supply modes and an off
mode, as shown in Fig. 7.
➢ Fresh-air mode: In this mode, all fresh air inlets are opened,
while all return air inlets are closed. The outdoor air is sent into
the indoor space after being heated by the transpired collector.
The ventilation capacity is the biggest in this mode, and it has
the most similar operating principle as the UTC system.
➢ Mixture mode: According to different ventilation demands, one
or more fresh air inlets are opened, while corresponding return
air inlets in the bedrooms are also opened. The mixture of the
fresh air and return air is sent into the indoor space after being
heated.
➢ Return-air mode: When the indoor air quality satisfies daily life
demands, the fresh air inlets are closed, while the return air
inlets are opened. The recirculated air is sent back to the room
Fig. 4. The appearance of the solar wall house.
space after being heated. In this mode, the heating effect is ex-
pected to be the largest. And the heated air temperature could
be raised above the indoor air temperature.
➢ Off mode: All air inlet and outlets are closed to form an enclosed
cavity in the solar wall structure. The air gap in the enclosed
system becomes an air insulation layer for the external
envelopes.
In all the hot-air supply modes, the internal air of the system is
heated up by the absorbed solar energy. A space heating effect is
achieved when the heated air is supplied to the indoor space.
Furthermore, in the fresh-air and the mixture modes, the air fan
draws fresh outdoor air into the indoor space, thereby producing a
positive internal pressure state. The pressure difference drives
surplus air out of the room through the reserved apertures in the
interior windows. As a result, a ventilation route is formed from the
south outdoor to the indoor space of the bedroom, further to the
indoor space of the kitchen, and finally to the north outdoors. The
airflow of the fresh-air and mixture modes can help to reduce the
indoor pollutant concentrations. In summer, the operation mode is
completely different, namely, all return-air inlets and all air fans are
closed, while all the fresh-air inlets and the outlets at the top of the
outer cavity are open. Air is exchanged between the outer cavity
and the outdoor environment because of the buoyancy effect. In the
summer mode, the heat gain of the solar wall system is released to
the ambient by the air movement. Consequently, a heat insulation
effect is achieved and the indoor air temperature is reduced.
2.2. Experimental procedure
The experiments were divided into two stages. The first stage
was conducted in Dec. 2014, in which the operating conditions of
the system in the various modes were tested. Four days with similar
weather conditions (including outdoor temperature and solar ra-
diation intensity) were chosen as the test period. The environ-
mental conditions during this test are listed in Table 1. The average,
maximum, and minimum outdoor temperatures, as well as the
average solar radiation intensity during the selected four days were
quite close. Consequently, it was possible to compare the space
heating performance and the indoor air quality improvement po-
tential of the system, relevant to the four operating modes. At this
stage, the airflow rates and the temperatures in the solar wall
system, the indoor air temperature, and the pollutant concentra-
tions were measured.
The second stage was processed in Feb. 2015. The purpose of this
stage was to investigate the dynamic performance and to
 T. Zhang et al. / Building and Environment 98 (2016) 158e179 163

Fig. 5. System schematic of the solar wall system.

Fig. 6. Configurations of the transpired collector unit and its holder.

Fig. 7. Different operating modes of the solar wall system.

determine the operating strategy of the solar wall system. At this bedroom was tested successively to investigate the heat contribu-
stage, the indoor air temperature of both the east and the west tions of the solar wall system and the kang system. During the test,
164 T. Zhang et al. / Building and Environment 98 (2016) 158e179

Table 1
Outdoor weather conditions during the tests of different operating modes.

Date Operation mode Weather Average temperature Maximum Minimum Average solar intensitya




Dec.14th Off Sunny 16.73 C 8.04 C 26.85 C 242.01 W/m2




Dec.18th Full return air Sunny 16.81 C 8.48 C 26.36 C 239.31 W/m2




Dec.24th Mixture Sunny 16.36 C 8.16 C 27.64 C 232.22 W/m2




Dec.27th Full fresh air Sunny 16.19 C 7.92 C 28.64 C 236.21 W/m2
a
The average solar radiation intensity was calculated during the operating period (from 9:00AM to15:30PM).

the east bedroom was continuously heated by the solar wall and
intermittently heated by the kang, while the west bedroom was
heated only by the wall. In addition, the pollutant concentrations
relevant to the different operating modes were monitored to
determine the optimal running method of the system. At this stage,
the wall system had to be regulated between the different oper-
ating modes, and the air temperature of both bedrooms, as well as
the pollutant concentrations of the east bedroom had to be
measured.
During both test stages, the stove of the kang was used twice per
day for cooking and intermittent heating, namely, early in the
morning for breakfast at 8:30e9:30 AM and later in the day for
dinner at approximately 15:30e17:00 PM. The kang system was
therefore operated twice daily during the test as a secondary
heating device.
2.3. Test point layout and measuring instruments
The test parameters were mainly the operating parameters of
the solar wall system and the indoor air parameters. Based on the
test results, the heating capacity of the solar wall system and its
ability to improve the indoor air quality could be investigated. The
test parameters of the solar wall system included solar radiation
intensity, outdoor air temperature, cavity air temperature, the air
temperatures and velocities at the fresh-air inlets, return-air inlets,
and supply air outlets. A pyranometer, located on a support and
mounted on the south side of the roof, was used to measure the
global solar radiation intensity (see Fig. 8). Sixteen thermocouples
were used to test the temperatures at different parts of the solar
wall system. Two thermocouples were located next to the pyran-
ometer to measure the outdoor air temperature, two were posi-
tioned in the air cavity (as illustrated in Fig. 4), four were uniformly
distributed in the return-air inlets of the east bedroom, four were
uniformly distributed in the supply air outlets of the east bedroom
(see Fig. 9), two were placed in the return-air inlets of the west
bedroom, and two were placed in the supply air outlets of the west
bedroom. Eighteen air velocity test points were positioned in the
air inlets and outlets, including eight in the supply air outlets and
return-air inlets of the east bedroom, four in the supply air outlets
and return-air inlets of the west bedroom, and six in the three
fresh-air inlets. The average of the measured values was assumed
the actual air velocity at an air inlet or outlet.
The indoor air quality test was conducted in the east bedroom,
as it was the main living space in winter. During the day, the oc-
cupants would sit on the kang to rest or chat; therefore, the tem-
perature fields and the pollutant distribution in the space above the
kang were of concern. The kang and the space directly above it
(within a height of 0.6 me1.3 m) was considered the human ac-
tivity zone. The experimental tests on the indoor air quality were
processed mainly in this activity space of the east bedroom. The test
items included the temperatures and the concentrations of carbon
dioxide (CO2), carbon monoxide (CO), and inhalable particulate
matters (PM2.5 and PM10) in the residential activity space. Fig. 9
presents the measuring-point arrangement in the indoor space of
the east bedroom. The test points for the indoor air temperature
and indoor pollutant concentrations were all placed on the hori-
zontal surface at a height of 1.2 m from the floor. These test points
included twenty four for the temperature, eight for the CO and CO2
concentrations, and eight for the PM2.5 and PM10 concentrations. In
addition, other parameters were measured to obtain the initial and
boundary conditions for the subsequent numerical research. These
additional parameters included the surface temperature of the
kang, the pollutant concentrations in the vicinity of the kang sur-
face, the outdoor pollutant concentrations, and the air tempera-
tures in the adjacent spaces of the east bedroom. Eight test points
were located on the upper surface of the kang to test the surface
temperature and the concentrations of CO, CO2, PM2.5, and PM10.
Two air temperature test points were located in the west bedroom
and another two were located in the kitchen. The pollutant con-
centrations of the outdoor air were tested next to the fresh-air
inlets.
The measuring instruments and the associated amounts, types,
ranges, accuracies, and locations of the experimentation are indi-
cated in Table 2. In addition, the uncertainties relevant to the
measured variables are listed. Compliant with the test instruments,
the temperatures and the solar radiation intensity could be

Fig. 8. Photos of the pyranometer and the indoor air quality test instruments.
 T. Zhang et al. / Building and Environment 98 (2016) 158e179 165

Fig. 9. The test point layout of indoor air temperatures and indoor pollutant concentrations.

Table 2
List of testing and monitoring devices.

Device Specification Amount Measuring parameter Range Uncertainty Location

Pyranometer TBQ-2; 1 Solar radiation intensity 0e2000 W/m2 ±1 W/m2 On a support mounted on the roof
Thermocouple copper-constantan; 60 Temperature 250e300
C ±0.1
C Indoor, outdoor, air vents, cavity, surfaces
Data logger BES-Ca 2 Data storage and output e e In the room
Hot-wire anemometer Fluke923; 1 Air speed 0.20 m/se20.00 m/s ±0.01 m/s Air vents
Laser particle counter TSI8532; 1 Contents of PM2.5 and PM10 0.001e150 mg/m3 ±0.001 mg/m3 Activity space, kang surface
Gas detector GT-1000-CO; 1 CO concentration 0e1000 ppm ±0.1 ppm Activity space, kang surface
Gas detector GT-1000-CO2; 1 CO2 concentration 0e2000 ppm ±1 ppm Activity space, kang surface

measured continuously and recorded automatically, whereas the
pollutant concentrations and air velocities could only be tested
intermittently and recorded manually.
3. Simulation of indoor temperature field and indoor
pollutant distributions
To investigate the effect of the solar wall system on the indoor
temperature and air quality, the indoor temperatures and pollutant
distributions of the east bedroom were simulated with FLUENT
software. The physical model is shown in Fig. 10. The dimensions of
the room are 5.45 m  4.18 m  3.0 m (length  width  height).
The sizes and locations of the door, the windows, and the interior
furniture were specified according to the actual conditions. The
kang, with a geometrical dimension of 4.18 m  2.65 m  0.6 m, and
the space above it, within a height of 0.6 me1.3 m, is considered the
human activity zone. This activity space was assumed the CO2
source of human activity, and the near surface area of the kang was
assumed the source of indoor pollutants, including CO, CO2, PM2.5,
and PM10. A rectangular air outlet (1.2 m  0.05 m) was positioned
in the north window, and the air outlet and the return-air inlet of
the solar wall system were configured in accordance with the actual
conditions.
In reality, the heat and mass transfer processes in the solar wall
system and the bedroom were transient. The dynamic operating
parameters of the system and the indoor environment had already
been tested during experimentation and the main purpose of the
simulation study was to investigate the indoor temperature and the
pollutant distributions. Accordingly, we concentrated on modelling
the indoor temperature field and the pollutant concentration fields
in the different operating modes, specifically at 12:30 PM, when the
solar radiation intensity and the outdoor temperature were the
highest and the system performances were the best. Therefore, the
heat and mass transfer was simplified as a steady process.
Accordingly, a steady-state heat and mass transfer model was
established, with some necessary simplifications and hypotheses.
The interfacial thermal contact resistances between different sur-
faces were ignored; the physical properties of all materials, except
the air, were considered constants; the buoyancy force in the
airflow was assumed to satisfy the Boussinesq hypothesis; and the
air infiltrations through the solar wall and the windows were
ignored. Based on the above hypothesis, the governing equations of
the problem included the masseconservation equation, the mo-
mentum conservation equation, the energy conservation equation,
and the realizable keε equation model that included the turbulent
kinetic energy equation and the turbulent dissipation rate
166 T. Zhang et al. / Building and Environment 98 (2016) 158e179
Fig. 10. Mathematical model of east bedroom and its mesh.
equation. As there was no large pressure gradient, body force, or
high speed flow in the indoor airflow, the standard wall function
was used in the near wall region. The DO model of the FLUENT
software was adopted to evaluate the radiation heat transfer be-
tween the surfaces of the indoor space and to determine the effect
of solar radiation on the indoor environment. The location of the
solar calculator was set at 127.00
east longitude and 44.90
north
latitude, and the data were set according to the experimental
period, the time was set at 12:30 PM. The transmissivity, absorp-
tivity, and reflectivity of the exterior glaze of the window were set
at 0.77, 0.16, and 0.07, respectively. The coefficients of the interior
glaze were 0.61, 0.28, and 0.11, respectively. The emissivity and the
absorptivity of the interior surfaces of the room were set at 0.85.
To investigate the diffusion process of pollutants in the rural
building, the convectionediffusion equation was adopted. CO, CO2,
and particulate matters (PM) are considered the main contami-
nants in rural buildings in winter. The combustion of fuel for
cooking and space heating is the source of CO and particulate
matters, whereas combustion and human activities are the sources
of CO2. Accordingly, the fluid in the model was a mixture of indoor
air, CO, and CO2. In the Species Transport model of the FLUENT
software, the tabs for Inlet Diffusion, Diffusion Energy Source, and
Thermal Diffusion were activated to model the indoor CO and CO2
distributions.
To investigate the distribution of inhalable particles (particle
size less than or equal to 10 mm), the Eulerian model was employed
to simulate the multiphase flow of the mixture of indoor air and
inhalable particles. Both PM2.5 and PM10 were investigated in this
study. Accordingly, three phases were included in the model,
namely, the air, particulate matters with a diameter of 0e2.5 mm,
and particulate matters with a diameter of 2.5e10 mm. The density
of these particulates was 2500 kg/m3. The main acting forces be-
tween these particulate matters were drag force and lift force, and
the corresponding functions of these two forces were therefore
activated.
The heat transfer coefficients of the different components of the
Table 3
Thermal transmittances in the simulation.
Boundary type Door External window
Thermal transmittance (W/m2$K) 3 2.9
model are listed in Table 3. The surfaces of the television and the
indoor furniture were treated as the uniform heat flux boundary
(the second boundary condition). The surface heat flux of the
television was 86.96 W/m2 and that of the furniture was 0 W/m2.
The surface of the kang was treated as a constant wall temperature
condition (the first boundary condition).
In order to ensure the accuracy of the numerical results, a grid-
dependency study was conducted. Respective computations were
carried out with four gridline systems, with the grid numbers
440800, 661200, 881600, and 1102000. Based on the numerical
calculations, the computational grid supplying grid-independent
solutions was 881600, with a maximum deviation of 2% for the
average Nusselt number. Therefore, the 881600 grid was used in all
the instances considered here. The grid number appears to be too
large. However, considering that the computational domain is a 3-D
real space, which includes all indoor furniture and equipment, and
as it was necessary to adopt densified grids at the air vents and the
numerous interior surfaces, we consider this grid number accept-
able. The finite-volume solver using segregated methods was
employed to solve the problem. The SIMPLEC algorithm was used
for coupling the equations of pressure and velocity. The second-
order upwind scheme was adopted to discretize the governing
equations.
4. Results and discussions
4.1. Relationships between solar radiation intensity and cavity air
temperatures
Outdoor temperature and solar radiation intensity are the two
significant factors that influence the internal air temperature of the
cavity in the solar wall system. Fig. 11 demonstrates the test results
of the off mode on Dec. 14th, 2014. The variations of the solar ra-
diation intensity, the average cavity air temperature (two test
points), and the average outdoor temperature (two test points)
were plotted as curves to show the rising temperature process of
Inner window External wall Floor Roof Inner wall
3.2 0.38 0.22 0.41 1.72
 T. Zhang et al. / Building and Environment 98 (2016) 158e179 167

Fig. 11. Test results of solar radiation intensity, average cavity air temperature and average outdoor air temperatures on Dec.14th.

the wall system. In this mode, the air cavity served as an enclosed
air layer, within which natural convection was generated because of
the buoyancy effect. In addition, the solar energy absorbed by the
transpired metal plate was transferred into the inside air to prevent
the indoor heat transferring to the outdoor.
The average cavity air temperature showed the same variation
tendency as the solar radiation intensity. As shown in Fig. 8, the
tendencies were almost the same from 8:30 AM to 11:45 AM. The
internal air temperature responded quickly to the solar radiation,
which meant that the heat absorption and release process of the
transpired absorber plate was rapid. The peak value of the cavity air
temperature occurred from 12:00 to 13:15 PM, at about 6
C.
Subsequently, after a time delay, the air temperature dropped,
along with the solar radiation intensity. When the solar intensity
approached zero at about 15:30 PM, the cavity air temperature was
approximately 10
C. The air supply temperature of the solar wall
system should be higher than the designed indoor air temperature
(16
C); therefore, the system could operate from 8:30 AM to 15:30
PM.
4.2. Hot-air supply volumes in different modes
The measured average airflow rates in the different operating
modes are illustrated in Table 4.
During the experimental test, we adjusted the air fans and air
vents to keep the hot-air supply volume of the east bedroom con-
stant during the different operating modes. In the return-air mode,
the daily average hot-air volume during the operating period was
214.89 m3/h, i.e., equal to the average return-air volume. In the
mixture mode, the average hot-air volume was 212.71 m3/h, which
is the sum of the fresh-air volume and the return-air volume. The
fresh-air ratio was 0.75. In the fresh-air mode, the supply hot-air
volume was 199.84 m3/h, the same as the fresh-air volume.
According to the design manual for heating and ventilation, the
fresh-air volume for a rural house in winter should be larger than
42.5 m3/h [50]. Accordingly, the fresh-air mode and the mixture
mode could meet the fresh-air demand in a rural building. As the
hot-air supply flow rates in all three modes were similar, it was
reasonable and feasible to compare the thermal performances of
the different operating modes.
4.3. Indoor air temperature variations in different operating modes
The average indoor air temperature curves of the east bedroom
in the different operating modes are shown in Fig. 12. The average
temperature value was calculated from the measured values of the
24 test points located in horizontal plane at z ¼ 1.2 m.
This figure shows that in the off mode the indoor air tempera-
ture varied from 8.4
C to 15.12
C, whereas in the three hot-air
supply modes, the range was 14.64
Ce20.48
C. The daily
average temperatures were 12.12
C, 18.19
C, 16.72
C, and 16.17
C,
respectively, for the off mode, the return-air mode, the mixture
mode, and the fresh-air mode. The solar hot-air supply of the wall
system therefore has the advantage of improving the indoor air
temperature. As the design code states [50], the indoor air tem-
perature in winter should be maintained above 16
C in rural
buildings. Accordingly, with the kang as secondary heating source,
only the return-air mode could meet the indoor air temperature
requirement for the entire day, whereas the mixture and fresh-air
modes could meet the requirement during the day, whereas the
off mode was unable to do so. Furthermore, in the operating modes
of the solar wall system, the indoor air temperature fluctuation
range was 3.56
Ce3.96
C, whereas the fluctuation range of the off
mode was 6.72
C. Consequently, we concluded that the solar wall
system has the capacity to maintain indoor thermal stability.

Table 4
Tested average air flow rates in different operating modes.

Mode Test date Air flowrate East bedroom West bedroom Total
3 3
Return-air Dec.18th Supply 214.89 m /h 185.56 m /h 400.45 m3/h
Fresh air e e e
Return air 214.89 m3/h 185.56 m3/h 400.45 m3/h
Mixture Dec.24th Supply 212.71 m3/h 212.71 m3/h 425.42 m3/h
Fresh air 159.06 m3/h 159.06 m3/h 318.12 m3/h
Return air 53.65 m3/h 53.65 m3/h 107.30 m3/h
Fresh-air Dec.27th Supply 199.84 m3/h 109.71 m3/h 309.55 m3/h
Fresh air 199.84 m3/h 109.71 m3/h 309.55 m3/h
Return air e e e
168 T. Zhang et al. / Building and Environment 98 (2016) 158e179
Fig. 12. Average indoor air temperature curves in different operating modes.
4.4. Relationship between supply air temperature and indoor
temperature
The major function of the solar wall system is to increase the
temperature of the indoor air by supplying heated air to the indoor
space; accordingly, it is essential to evaluate the relationship be-
tween the supply air temperature and the indoor air temperature.
Figs. 13e15 present the average value curves of the two tempera-
tures, as well as the outdoor temperature in the return-air mode,
the mixture mode, and the fresh-air mode. The average value of the
supply air temperature was obtained from the test results at the
four test points in the supply air outlet, whereas the average value
of the indoor temperature was calculated from the test results of
the twenty four indoor test points.
In the return-air mode during the operating period, the average
supply temperature increased from 19.16
C at 9:00 AM, when the
system operation started, to 36.81
C at 12:30 PM. The average
indoor temperature varied from 18.20
C to 20.48
C during this
period, whereas the outdoor temperature varied
between 16.28
C and 8.48
C.
In the mixture mode, the supply air temperature increased from
16.98
C to 30.53
C, the indoor temperature varied from 15.84
C to
19.12
C, and the outdoor temperature varied between 16.32
C
and 8.16
C.
In the fresh-air mode, the supply air temperature increased from
15.81
C to 26.42
C, the indoor temperature varied from 15.40
C to
Fig. 13. Measured results of average indoor air temperature, average supply air temperature and average outdoor air temperature in the return-air mode on Dec.18th.
18.20
C, and the outdoor temperature varied between 15.80
C
and 7.92
C.
In all three hot-air supply modes, the variation tendencies of the
supply air temperature, the indoor air temperature, and the out-
door air temperature were similar, indicating that the supply air
temperature has a greater influence on the indoor temperature. The
indoor air temperature increased in the morning and dropped in
the afternoon, along with the supply air temperature. The peak
values of these two temperatures in the different modes occurred
at approximately 12:30 PM, and the minimum values appeared at
approximately 9:00 AM, when the system started to operate. In all
three modes, the indoor air temperatures showed extremely small
fluctuations. We can therefore conclude that the indoor thermal
environment can be kept stable by a combination of the solar wall
system and the kang system. In contrast, the test results indicated
that the return-air mode had the highest supply air temperature
and maintained the indoor temperature at the highest level. The
heating capacity of the mixture mode was located at mid-level,
whereas, compared with the other two modes, the fresh-air
mode had the smallest heating capacity.
4.5. Indoor air quality assessment
The indoor air quality measurement of each operating mode was
conducted at 12:30 PM on the test day, when the solar wall system
operated under relatively stable and efficient working conditions.
 T. Zhang et al. / Building and Environment 98 (2016) 158e179 169

Fig. 14. Measured results of average indoor air temperature, average supply air temperature and average outdoor air temperature in the mixture mode on Dec.24th.

Fig. 15. Measured results of average indoor air temperature, average supply air temperature and average outdoor air temperature in the fresh-air mode on Dec.27th.

The average concentrations of CO, CO2, PM2.5, and PM10 are listed in
Table 5. Each of the above items was tested at the eight points
located in the activity space of the east bedroom. The values in the
table show the average of all the test points.
From Table 5, we can observe that all the hot-air supply modes
had the capacity to reduce the concentrations of indoor pollutants.
Compared with the off mode, the return-air mode reduced the CO
concentration by 34.8%, and the reduction proportions relevant to
CO2, PM2.5, and PM10 were 20.3%, 14.4%, and 11.6%, respectively.
Although no fresh air was supplied to the indoor space, the air
movement helped to even out the pollutant distributions, thereby
reducing the pollutant concentrations in the activity space. In the
mixture mode, the reduction proportions of CO, CO2, PM2.5, and
PM10 were 69.6%, 28.0%, 45.0%, and 41.6%, respectively. In the fresh-
air mode, these proportions were 73.9%, 42.7%, 56.2%, and 58.1%,
respectively. These two modes significantly reduced the pollutant
concentrations, as the outdoor fresh air was introduced into the
room space and ventilation was generated between the indoor and
outdoor environment. Compared with the standard values indi-
cated in the table, the CO concentrations in all the modes were
lower than the standard value. The fresh-air supply in the mixture
and fresh-air modes contributed significantly to reducing the CO
concentration. The fresh-air component modes were able to reduce
the CO2 content from an over-weight range to the standard range.
The PM2.5 and PM10 concentrations were reduced dramatically by
the fresh-air supply in the mixture and fresh-air modes, although
the concentrations still exceeded the standard range. The concen-
tration values in the different modes showed that the fresh-air
mode performed the best at diluting the indoor pollutants, fol-
lowed by the mixture mode. The return-air mode was unable to
reduce the total indoor pollutant amount, but was able to even out
the pollutant distributions, slightly reducing the pollutant

Table 5
Measured average indoor pollutant concentrations and the standard values [50].

Type of pollutant Test date CO (ppm) CO2 (ppm) PM2.5 (mg/m3) PM10 (mg/m3)

Standard values e 8 1000 0.075 0.15

Off mode Dec.14th 2.3 1633 0.313 0.534
Return-air mode Dec.18th 1.5 1301 0.268 0.472
Mixture mode Dec.24th 0.7 1176 0.172 0.312
Fresh-air mode Dec.27th 0.6 935 0.137 0.224
170 T. Zhang et al. / Building and Environment 98 (2016) 158e179
concentrations in the activity space.
4.6. Indoor air temperature distributions and indoor pollutant
distributions in different operating modes
4.6.1. Simulation inputs and model validation
The simulation was carried out to investigate the indoor tem-
perature and pollutant distributions in the various operating modes
at 12:30 PM. The outdoor temperature was 8
C at that time. The
solar radiation intensity was obtained from the solar calculator of
the FLUENT software. The temperatures of the two adjacent bed-
rooms were 10
C and 6
C, respectively. The upper surface tem-
perature of the kang was 22
C. The measured average supply air
temperatures, and the velocities and cavity air temperatures of the
solar wall are indicated in Table 6.
To simulate the indoor pollutant distributions, the generation
rates of CO, CO2, PM2.5, and PM10 at the surface of the kang were
obtained from the field test. The generation rate of CO2 in this ac-
tivity space was obtained by the following assumption, namely,
four adults are in the room, including two in the daily activity
condition and two in the rest condition. The CO2 generating rates
were 1 L/min and 0.25 L/min respectively for these two conditions.
The volume of the activity space is 14.37 m3. Consequently, the CO2
generation rate of the activity space is calculated as 2.27  106 kg/
(s$m3). The generation rates of the different types of pollutants at
the kang surface (with a volume of 0.068 m3) are indicated in
Table 7. The pollutant concentrations of the outdoor air are indi-
cated in Table 8.
Fig. 16 displays a comparison between the tested and the
simulated results of the indoor air temperatures along the three
test lines (y ¼ 2.5 m, z ¼ 1.2 m; y ¼ 3.5 m, z ¼ 1.2 m; and y ¼ 4.5 m,
z ¼ 1.2 m) in the mixture mode. Taking everything into account, we
found that the simulated temperatures were 0.6
C higher than the
tested results, which could be ascribed to the assumption of the
steady-state heat and mass transfer. Fortunately, the difference
between the tested and the simulated results was acceptable, and
the variation tendencies were almost the same, showing that the
established simulation model has high accuracy and reliability in
predicting the indoor temperature distribution. Figs. 17 and 18
show comparisons between the test results and the simulation
results for the CO, CO2, PM2.5, and PM10 concentrations relevant to
the mixture mode. The simulated results were close to the test
results, and the changing trends were similar. We can therefore
conclude that the established model was able to predict the indoor
pollutant distributions. Our conclusions were similar, regarding the
test and simulation results of the other operating modes.
4.6.2. Horizontal temperature distributions in different operating
modes
Fig. 19 presents the simulated indoor temperature fields at a
height of 1.2 m (z ¼ 1.2 m) in the different operating modes of the
solar wall system.
In the off mode, the indoor air temperatures in the horizontal
plane were not evenly distributed. The temperatures in the east
side were clearly higher than in the west side of the room. The
northeast corner and the southwest corner of the room had higher
Table 6
Model input of the cavity air temperature, the supply air temperatures and velocities in different modes.
Mode Supply air temperature
Off e 59.90
Return-air 36.81
C
Mixture 30.53
C
Fresh-air 26.42
C
Table 7
Measured pollutant generation rates of the kang
CO2 CO PM2.5 PM2.5~10
(kg/(s$m3)) (kg/(s$m3)) (kg/(s$m3)) (kg/(s$m3))
3.39  105 5.29  107 2.10  107 4.35  108
Table 8
Test pollutant concentrations of outdoor air.
Type CO(ppm) CO2(ppm) PM2.5(mg/m3) PM10 (mg/m3)
Content 0 400 0.059 0.066
temperatures than the other regions, as the kang and the television
were the heat sources in the room. In the hot-air supply modes, the
dissymmetrical distribution of air temperatures in an eastewest
direction still existed, but the magnitudes of the imbalances were
considerably smaller. The temperatures of the local scope below the
air inlet were higher than those in the other regions. The operation
of the solar wall system helped to improve the air temperature in
the human activity space of the room and to distribute the tem-
peratures more uniformly. Taking into account the temperature
fields, the heating effectiveness of the different operating modes
can also be investigated. The overall temperature of the return-air
mode was the highest, followed by the mixture mode, and the
fresh-air mode.
4.6.3. Vertical temperature distributions in different operating
modes
To investigate the vertical temperature distributions, we used
the average temperature curves along the x direction at heights of
0.1 m, 0.8 m, 1.2 m, and 1.7 m, as shown in Fig. 20.
In the off mode, the temperature difference in the vertical di-
rection was obvious, as cold air deposits at the bottom of the room
space because of indoor natural convection. At a height of 0.1 m, the
average temperature was approximately 9
C. From x ¼ 3.5 to
x ¼ 4.2, the exothermic television raised the localized temperature.
At a height exceeding 0.8 m, the temperatures rose higher than
13.59
C. The temperature differences between the different
heights were small when the solar wall system was in operation.
Obviously, the mechanical air supply produced a stable flow field in
the room, which helped to create a more uniform indoor temper-
ature field. In all hot-air supply modes, the air temperature reached
a maximum at the location of the air inlet (x ¼ 3.0). Taking into
account the temperature ranges of the different modes, we can
conclude that the solar wall system not only improved the indoor
air temperature but also enhanced the uniformity of temperature
distribution.
4.6.4. Distributions of indoor carbon oxides
Figs. 21e24 illustrate the distributions of CO and CO2 at a height
of 1.2 m relevant to the different operating modes.
Although the distributions of CO and CO2 were not completely
consistent in the different modes, the maximum point, minimum
point, and the variation tendencies were almost identical. In the off
Cavity air temperature Supply air velocity


C e
32.00

C 3.34 m/s
25.67

C 3.58 m/s
22.51

C 3.36 m/s
 T. Zhang et al. / Building and Environment 98 (2016) 158e179 171

Fig. 16. Comparison between the measured and calculated temperature values along the lines of y ¼ 2.5 m, 3.5 m, 4.5 m and z ¼ 1.2 m.

Fig. 17. Comparison between the measured and calculated CO and CO2 concentrations along the line of y ¼ 3.5 m and z ¼ 1.2 m.

Fig. 18. Comparison between the measured and calculated PM2.5 and PM10 concentrations along the line of y ¼ 4 m and z ¼ 1.2 m.

mode, the only path for indoor pollutant discharging was the air
outlet, positioned in the north window. Consequently, the distri-
butions of both the CO and the CO2 contents were characterized by
these contents being higher in the north and lower in the south.
The minimum concentration point was located next to the air outlet
in the north window, and the maximum point was located at the
activity space above the kang. In the hot-air supply modes (the
return-air mode, the mixture mode, and the fresh-air mode), the
concentrations of CO and CO2 were clearly lower than those in the
off mode. These distributions showed a completely different trend,
namely, that the content was lower in the southwest and higher in
the northeast. In all three modes, the minimum concentration point
was below the air inlet. The maximum points were all located in the
activity region above the kang, but a difference in the specific
locations was indicated. The maximum point of the return-air
mode was located at the centre of the east wall, while those of
the mixture mode and the fresh-air mode were positioned at the
northeast corner. The distribution indicated that a focus region of
CO and CO2 was generated in the activity space above the northeast
corner, whilst the hot-air supply of the solar wall system aiding the
dilution of the CO and CO2 concentrations of the room, which could
prevent the pollutants from affecting the health of the occupants.
Fig. 25 illustrates the vertical distributions of CO in the vertical
plane of x ¼ 2.68 m in the mixture mode and the return-air mode. It
is clear that the greater the distance from the air supply outlet was
the higher was the concentration. From an overall point of view, the
concentrations in the upper space were lower than those in the
nether space. The concentration layering was more obvious at a
172 T. Zhang et al. / Building and Environment 98 (2016) 158e179
Fig. 19. Simulated indoor temperature distributions in different operating modes: a-the return-air mode; b-the mixture mode; c-the fresh-air mode; d-the off mode.
lower height. A focus region was found on the kang surface in each
mode.
4.6.5. Distributions of inhalable particles
Fig. 26 shows the distributions of the inhalable particles, with
different particle sizes (smaller than 2.5 mm, and from 2.5 mm to
10 mm), at a height of 1.2 m, in the mixture mode. The distributions
were similar, as the high concentration zones and low concentra-
tion zones were almost identical. Consequently, in the following
discussion, the distributions of PM2.5 are taken as representative to
investigate the inhalable particle distributions.
Fig. 27 illustrates the concentration distributions of PM2.5 at a
height of 1.2 m in the different operating modes. Unlike the dis-
tributions of CO and CO2, that of PM2.5 in each operating mode
appeared scattered and chaotic, this was consistent with the
principle of the multiple-phase flow model. The distributions in all
modes generally showed a trend of the concentrations being high
in the east and low in the west, since the kang was the only source
of PM2.5.
In the off mode of the solar wall system, the distribution was
even, as there was no indoor disturbance, while in the hot-air
supply modes, influenced by the inlet airflow, the particles accu-
mulated in the localized regions. In the return-air mode, the con-
centration was high along the south and east walls, with the
accumulation of pollutants being the densest in the northeast
corner. The concentrations in the south side were lower, as the air
circulated through the ceiling air inlet and the return-air vent is
situated there. In the mixture mode and the fresh-air mode, the
supply air was discharged from the air outlet in the north window.
Air travelled through the north and south of the room in these two
modes, rendering the PM2.5 distributions uniform.
Fig. 28 illustrates the vertical distributions of PM2.5 at x ¼ 2.68 m
in the mixture mode and the fresh-air mode. From an overall point
of view, the concentration was high in the east side and low in the
west side. The inlet airflow produced a ribbon region, in which the
CO content was relatively low. Compared with the CO and CO2
distributions, the influence of the inlet airflow on the PM2.5 dis-
tribution was weaker.
4.7. The heat contributions of the solar wall and the kang
In the previous sections, in both the experimental and the nu-
merical investigations, the solar wall and the kang were treated as a
combined space heating system. In all the test and simulated
conditions, the kang and the GTC operated together to improve the
indoor environment. To determine the load contributions of the
solar wall and the kang for domestic heating, a comparison test
between the indoor air temperatures of the east and west
 T. Zhang et al. / Building and Environment 98 (2016) 158e179
Fig. 20. Simulated indoor temperature variations at different heights: a- the off mode; b- the fresh-air mode; c- the mixture mode; d- the return-air mode.
bedrooms was conducted from Feb. 18th to Feb. 23rd of 2015.
During the test, the solar wall operated in the fresh-air mode from
8:30 AM to 3:30 PM. The kang in the east bedroom was heated
twice daily, namely, early in the morning at 8:30 to 9:30 AM and
later in the evening at approximately 15:30 to 17:00 PM. The kang
in the west bedroom was not used during the test period. Fig. 29
presents the test results of the average indoor temperatures of
the two rooms, as well as the outdoor temperature.
The figure indicates that by using the solar wall system the in-
door temperature of the west bedroom (without kang heating)
could be maintained between 3
C and 12
C, whereas that of the
east bedroom (with kang heating) could be maintained between
10
C and 16
C. Taking into account the operating period (between
8:30 AM and 5:00 PM), the temperature difference between the
two rooms varied from 2.56
C to 9.24
C, depending on the time of
day. The average indoor temperature in the east room was 6.62
C
higher than that in the west room. During the test period, the
average outdoor air temperature was 4.33
C, the average west
room temperature was 7.17
C, and the average east room
173
temperature was 13.76
C. Considering the total areas and the
building materials, the interior decorations of the two rooms were
almost the same. Consequently, the heating contributions of the
kang and the solar wall could be simplified, namely, that the solar
wall system improved the indoor temperature from 4.33
C
(outdoor temperature) to 7.17
C, with the kang contributing to
raise the indoor temperature to 13.76
C.
The heat contributions of the solar wall and the kang in domestic
heating are therefore 63.6% and 36.4%, respectively. The ratios were
calculated from the fresh-air mode; however, a higher duty ratio
could be achieved in the mixture mode or the return-air mode. On
sunny days, the running time of the solar wall system is about 7 h,
much longer than that of the kang that operated intermittently for
only 2.5 h daily. The solar wall benefits from the hot-air supply and
from the heat transfer from the high cavity temperature to the in-
door environment to improve the indoor temperature. Therefore, in
the heating of rural houses, the solar wall system appears to carry
more than 63.6% of the heating load and the kang provides the rest of
the thermal energy for the indoor environment.
174 T. Zhang et al. / Building and Environment 98 (2016) 158e179

Fig. 21. Simulated CO and CO2 concentration distributions in the off mode: a- the CO distribution; b- the CO2 distribution.

Fig. 22. Simulated CO and CO2 concentration distributions in the return-air mode: a- the CO distribution; b- the CO2 distribution.

Fig. 23. Simulated CO and CO2 concentration distributions in the mixture mode: a- the CO distribution; b- the CO2 distribution.
 T. Zhang et al. / Building and Environment 98 (2016) 158e179
Fig. 24. Simulated CO and CO2 concentration distributions in the fresh-air mode: a-the CO distribution; b- the CO2 distribution.
Fig. 25. Simulated CO concentration distributions in the mixture and return-air
modes: a- the mixture mode; b- the return-air mode.
4.8. Adjustment and operating strategy of the solar wall system
In daily operations, the operating modes of the system should be
properly adjusted, in accordance with the indoor and outdoor
conditions. At night or on overcast days (no sunshine), the system
should operate in the off mode; however, on sunny days, the system
should operate in the three hot-air supply modes, depending on the
175
outdoor and indoor conditions. The previous test results indicated
that on sunny winter days, the solar wall system could function
from 8:30 AM to 15:30 PM, therefore, on sunny days, the ON/OFF
switch should be flicked at approximately 8:30 AM and 15:30 PM.
The ON/OFF switch changes all the air vents and air fans between
the open and the closed statuses. There are two main purposes of
using the solar wall system in the hot-air supply modes, namely,
space heating and fresh-air supply. As was previously indicated, the
return-air mode attained the best heating performance, with the
least capacity to dilute the indoor pollutants. The performance of
the fresh-air mode was quite the opposite, whereas the mixture
mode had an integrated performance. Consequently, it was
necessary to find an optimum operating strategy for the proposed
system to improve the heating capacity, while ensuring that the
indoor air quality was enhanced.
Determining the operating strategy should be based on the
dynamic changes in the indoor pollutant concentrations. Fig. 30
shows the continuous changes of the average CO and CO2 con-
centrations on Feb 25th, 2015, in the return-air mode. Note that the
measured values for the CO2 concentration in this figure were
lower than the average measured values shown in Table 5. This is
ascribed to the combustion processes releasing less CO2 at this time
than in December when it is colder. As the outdoor temperature
rises and heat demand decreases considerably, a smaller amount of
straw is required for fuel to enhance the indoor thermal
environment.
This figure indicates that the CO2 concentration fluctuates
wildly, while the CO concentration shows a steady change. This
difference is caused by CO and CO2 deriving from different pollution
sources. The CO content derives only from the combustion process,
while the CO2 content is affected by human activity and the com-
bustion process. During the periods 8:30 to 11:10 AM and 15:20 to
17:50 PM, the concentrations of CO and CO2 were relatively high
and significantly exceeded the standard range [50] (8 ppm and
1000 ppm, respectively, for CO and CO2, as illustrated in Table 5).
The peak region agrees well with the time periods when fuel was
combusted for cooking and heating with the kang (8:30 to 9:30 AM
and 3:30 to 5:00 PM). At other times of the day, the concentrations
remained below the standard value, indicating that during non-
combustion periods, the return-air mode was able to meet the
air-conditioning standard. During combustion periods, the system
should be switched to the mixture mode or the fresh-air mode to
eliminate the pollutants released during the combustion.
We tested the CO and CO2 concentrations during the
176 T. Zhang et al. / Building and Environment 98 (2016) 158e179

Fig. 26. Simulated PM2.5 and PM2.5e10 concentration distributions in the mixture mode: a- PM2.5 distribution; b-PM2.5e10 distribution.

Fig. 27. Simulated PM2.5 concentration distributions in different operating modes: a- the off mode; b- the fresh-air mode; c- the mixture mode; d- the return-air mode.

combustion periods in the mixture mode on Feb. 26th, 2015. The
results are displayed in Fig. 31.
Fig. 31 a compares the CO2 concentrations in the mixture mode
and the return-air mode during the combustion period in the
morning. In addition, a standard curve with the concentration value
of 1000 ppm is included. In the return-air mode, during
combustion, the CO2 concentration exceeded 1000 ppm, but star-
ted to drop slowly to below this level after combustion had ended.
In the mixture mode, the concentration stayed below 1000 ppm
during the entire combustion process, and the fluctuation was less
than that of the return-air mode. Fig. 25 b shows a comparison of
the CO concentrations. In the return-air mode, the concentration
 T. Zhang et al. / Building and Environment 98 (2016) 158e179
Fig. 28. Simulated PM2.5 concentration distributions in mixture and fresh-air modes:
a- the mixture mode; b- the fresh air mode.
Fig. 29. Comparison between the test results of average indoor air temperatures in the east bedroom and west bedroom from Feb.18th to Feb.23rd.
increased rapidly to more than 8 ppm and then decreased slowly
after the combustion had ended; however, it was still more than
8 ppm. In the mixture mode, the concentration remained below the
standard, with slight rises and drops. The tests during the afternoon
showed similar results. Therefore, we can conclude that the
mixture mode ensures that the indoor CO and CO2 contents stay
below the standard during the combustion process.
In this test, the outdoor temperature was not extremely low;
therefore, a combination of the mixture mode (during cooking and
177
heating the kang) and the return-air mode at other times of the day
would be sufficient to maintain the indoor air quality and achieve
an optimal space heating capacity. However, during the colder
months (Dec. and Jan.), when the heating load is higher, the com-
bustion process releases more indoor pollutants, the fresh-air mode
could be necessary during non-combustion times.
In conclusion, based on the above test and simulation results, a
simple and easy operating manual for the solar wall system can be
summarized for users, as follows:
➢ In overcast weather (non-sunshine days), the system should
operate in the off mode for the entire day, i.e., all the air inlets
and outlets, and all the air fans are closed.
➢ On sunny days during Dec. and Jan., when the occupants rise
and cooking starts, the system is switched on by starting the air
fan and opening the fresh-air inlets to operate in the fresh-air
mode. After cooking, the fresh-air inlets should be closed and
the return-air inlets opened, the system is operating in the
return-air mode. In the afternoon, when the cooking of dinner
starts, the process is repeated, namely, the return-air inlets are
closed, the fresh-air inlets are opened, and the system operates
in the fresh-air mode. After cooking, the air fans and the fresh-
air inlets are closed, and the system operates in the off mode
during the night.
➢ On sunny days in Nov. and Feb., when morning cooking starts,
the system is activated by starting the air fan and opening the
fresh-air inlets and return-air inlets, and the system operates in
the mixture mode. When cooking ends, the fresh-air inlets are
closed, and the system operates in the return-air mode. Again,
when the cooking of dinner starts, the fresh-air inlets are
opened, the system operates in the mixture mode. When
cooking ends, the air fans and the fresh and return-air inlets are
closed, and the system operates in the off mode for the night.
The users of the system would find it easy to follow this oper-
ation manual. When the system is regulated according to the
manual guidelines, the solar wall system provides a balance be-
tween the purposes of heating the indoor space and diluting the
indoor pollutants.
5. Conclusions
The combustion of fuel by the kang system, used for cooking and
heating in rural dwellings in northeast China, causes uncomfortable
178 T. Zhang et al. / Building and Environment 98 (2016) 158e179

Fig. 30. Continuous changes of average CO and CO2 concentrations on Feb. 25th.

Fig. 31. Variation tendencies of CO2 and CO during combustion period on Feb 26th: a- the CO2 concentration; b- the CO concentration.

thermal conditions and poor air quality. In this paper, a GTC-based
solar wall system is proposed as solution to these problems.
Employing both experimental and numerical methods, the appli-
cability and capability of the system were investigated. The results
showed that the proposed solar wall system was effective for
improving the indoor thermal comfort and the indoor air quality.
We offer a simple and easy operation manual for the occupants to
operate the system. This specially designed system offers the po-
tential of managing the ubiquitous energy and environmental
problems in the cold and severely cold rural areas of China. It
should be noted that all analyses in this research are based on field
test and steady-state simulation; therefore, the applicability of the
proposed system in other climates, as well as and the dynamic
modelling of the system, must be evaluated in future study.
Acknowledgements
This research was supported by the National Natural Science
 T. Zhang et al. / Building and Environment 98 (2016) 158e179
Foundation of China (Grant No. 51578177) and by the Hong Kong
Polytechnic University in G-UB65.
References
[1] Z. Zhuang, Y. Li, B. Chen, J. Guo, Chinese kang as a domestic heating system in
rural northern China-A review, Energy Build. 41 (2009) 111e119.
[2] Zhai Z, P.Y. Andrew, L. Duanmu, Z. Wang, An evaluation and model of the
Chinese Kang system to improve indoor thermal comfort in northeast rural
China-part-1: model development, Renew. Energy 8 (2015) 3e11.
[3] Tsinghua University Team of Northern Rural Energy and Environment Survey,
Report on Survey and Investigation of Rural Energy and Environment in
Northern China in 2006, Tsinghua University, 2006 (in Chinese).
[4] Z. Zhuang, Y. Li, B. Chen, Thermal storage performance analysis on Chinese
kangs, Energy Build. 41 (2009) 452e459.
[5] Z. Wang, X. Sheng, J. Ren, Y. Ji, Field survey on indoor thermal comfort in rural
houses around Harbin in winter, HV&AC 44 (12) (2014) 71e75 (in Chinese).
[6] J. Liu, The current situation analysis of rural residence in northern of China,
Build. Technol. 2 (2011) 20e23 (in Chinese).
[7] Ministry of Housing and Urban-Rural development of China, Design Standard
for Energy Efficiency of Rural Residential Buildings (GB/T50284-2013), 2012.
Chp.3.
[8] G. Cao, J. Jokisalo, G. Feng, et al., Simulation of the heating performance of the
Kang system in one Chinese detached house using biomass, Energy Build. 43
(2011) 189e199.
[9] Z. Wang, X. Sheng, J. Ren, D. Xie, Field study on thermal environment and
thermal comfort at rural houses in severe cold areas, 8th Int. Symp. Heat. Vent.
Air Cond. ISHVAC (2013) 403e409.
[10] X. Sheng, Field Test and Analysis of Thermal Comfort at Rural Houses and
Urban Residential Buildings in Severe Cold Region, Dissertation for the Master
Degree, Harbin Institute of Technology, 2013 (in Chinese).
[11] M. Shan, P. Wang, J. Li, et al., Energy and environment in Chinese rural
buildings: Situations, challenges, and intervention strategies, Build. Environ.
91 (2015) 271e282.
[12] D. Wang, Popularize energy-efficient Kang and stove and realize reasonable
utilization of straw, Renew. Energy 3 (2003) 52e53 (in Chinese).
[13] B. Chen, Z. Zhuang, X. Chen, X. Jia, Field survey on indoor thermal environ-
ment of rural residences with coupled Chinese kang and passive solar col-
lecting wall heating in Northeast China, Sol. Energy 81 (2007) 781e790.
[14] Q. Guo, The Chinese domestic architectural heating system (kang): origins,
applications and techniques, Archit. Hist. 45 (2002) 32e48.
[15] T. Zhang, Y. Tan, X. Zhang, C. Niu, Current situation of energy and indoor
environment in rural buildings of northeast China, Appl. Mech. Mater. 522
(2013) 1756e1759.
[16] Z. Wang, D. Xie, R. Tang, X. Sheng, Investigation on indoor air quality at rural
houses in winter in severe cold region, 8th Int. Symp. Heat. Vent. Air Cond.
ISHVAC (2013) 69e75.
[17] R. Edwards, Y. Liu, G. He, Z. Yin, J. Sinton, J. Peabody, et al., Household CO and
PM measured as part of a review of China's national improved stove program,
Indoor Air 17 (3) (2007) 189e203.
[18] D. Gu, Life Cycle Assessment of Building Environmental Load, Dissertation for
the Doctoral Degree, Tsinghua University, 2006 (in Chinese).
[19] G. Yang, Y. Wang, Y. Zeng, G. Gao, X. Liang, M. Zhou, et al., Rapid health
transition in China, 1990-2010: findings from the global Burden of disease
Study 2010, Lancet 381 (9882) (2013) 1987e2015.
[20] C.M. Lai, S. Hokoi, Solar façades: A review, Build. Environ. 91 (2015) 152e165.
[21] A.K. Mishra, M. Ramgopal, Field studies on human thermal comfort -an
overview, Build. Environ. 64 (2013) 94e106.
[22] S. Abhishek, Varun, A.A. El-Sebaii, A thermodynamic review of solar air
heaters, Renew. Sustain. Energy Rev. 43 (2015) 863e890.
[23] S.B. Mahmut, B.R. Saffa, Building integrated solar thermal collectors-A review,
Renew. Sustain. Energy Rev. 51 (2015) 327e346.
[24] H.A. Zondag, Flat-plate PV-thermal collectors and systems: a review, Renew.
Sustain. Energy Rev. 12 (2008) 891e959.
179
[25] M.A. Mahmud, K. Sopian, M.A. Alghoul, et al., Review of solar air collectors
with thermal storage units, Renew. Sustain. Energy Rev. 15 (2011)
1476e1490.
[26] A. Shukla, N.N. Dan, Y.J. Cho, A state of art review on the performance of
transpired solar collector, Renew. Sustain. Energy Rev. 16 (2012) 3975e3985.
[27] S. Francesca, M. Alessio, M. Placido, Life cycle assessment approach for the
optimization of sustainable building envelopes: an application on solar wall
systems, Build. Environ. 58 (2012) 278e288.
[28] B. Laura, M. Concetta, M. Francesco, An overview on solar shading systems for
buildings, Energy Procedia 62 (2014) 309e317.
[29] C. Dymond, C.F. Kutscher, Development of a flow distribution and design
model for transpired solar collectors, Sol. Energy 60 (5) (1997) 291e300.
[30] B. Messaoud, R.R. Daniel, H. Stephane, Q. Guillermo, Experimental and nu-
merical simulation of a two-dimensional unglazed transpired solar air col-
lector, Sol. Energy 93 (2013) 209e219.
[31] D. Van, K. Hollands, A.P. Brunger, Heat-exchange relations for unglazed
transpired solar collectors with circular holes on a square or triangular pitch,
Sol. Energy 71 (2001) 33e45.
[32] C.F. Kutscher, Heat exchanger effectiveness and pressure drop for air flow
through perforated plates with and without crosswind, J. Heat Transf. 116
(1994) 391e399.
[33] D. Van, Asymptotic Thermal Effectiveness of Unglazed Transpired Plate Solar
Air Heaters, M.Sc. Thesis, Department of Mechanical Engineering University of
Waterloo, Canada, 1996.
[34] L. Zalewski, S. Lassue, B. Duthoit, M. Butez, Study of solar walls-validating a
simulation model, Build. Environ. 37 (2002) 109e121.
[35] J. Zhu, B. Chen, Experimental study on thermal response of passive solar house
with color changed, Renew. Energy 73 (2015) 55e61.
[36] A.A. Pesaran, K.B. Wipke, Use of unglazed transpired solar collectors for
desiccant cooling, Sol. Energy 52 (1994) 419e427.
[37] C.F. Kutscher, C.B. Christensen, G.M. Barker, Unglazed transpired solar col-
lectors: heat loss theory, ASME J. Sol. Energy Eng. 115 (1993) 182e188.
[38] R.C. Michael, A. Hani, An evaluation of heat transfer and effectiveness for
unglazed transpired solar air heaters, Sol. Energy 99 (2014) 231e245.
[39] V. Roozbeh, M. Ilkan, F. Egelioglu, Experimental performance of perforated
glazed solar air heaters and unglazed transpired solar air heater, Sol. Energy
119 (2015) 251e260.
[40] L. Gao, H. Bai, X. Fang, T. Wang, Experimental study of a building-integrated
solar air heating system in cold climate of China, Energy Build. 65 (2013)
359e367.
[41] B. Li, S. You, T. Ye, Mathematical modeling and experimental verification of
vacuum glazed transpired solar collector with slit-like perforations, Renew.
Energy 69 (2014) 43e49.
[42] L. Gao, H. Bai, X. Wu, Numerical analysis of heat transfer in unglazed tran-
spired collectors based on field synergy principle, Sol. Energy 95 (2013)
336e344.
[43] K.A. Andreas, B. James, O. Brendan, A prototype photovoltaic/thermal system
integrated with transpired collector, Sol. Energy 85 (2011) 139e153.
[44] V. Yasin, F.O. Hakan, A comparative numerical study on natural convection in
inclined wavy and flat-plate solar collectors, Build. Environ. 43 (2008)
1535e1544.
[45] Z.D. Chen, P. Bandopadhayay, J. Halldorsson, An experimental investigation of
a solar chimney model with uniform wall heat flux, Build. Environ. 38 (2003)
893e906.
[46] V. Yasin, F.O. Hakan, Buoyancy induced heat transfer and fluid flow inside a
tilted wavy solar collector, Build. Environ. 42 (2007) 2062e2071.
[47] D.L. Zhao, Y. Li, Y.J. Dai, R.Z. Wang, Optimal study of a solar air heating system
with pebble bed energy storage, Energy Convers. Manag. 52 (2011)
2392e2400.
[48] S. Cordeau, S. Barrington, Performance of unglazed solar ventilation air pre-
heaters for broiler barns, Sol. Energy 85 (2011) 1418e1429.
[49] L. Gao, H. Bai, S. Mao, Potential application of glazed transpired collectors to
space heating in cold climates, Energy Convers. Manag. 77 (2014) 690e699.
[50] National standard of the People's Republic of China. Code for Design of
Heating Ventilation and Air Conditioning. GB 50736e2012.