Solar Energy 166 (2018) 59–70

Contents lists available at ScienceDirect

Solar Energy
journal homepage: www.elsevier.com/locate/solener

Comprehensive study of wind supercharged solar chimney power plant T

combined with seawater desalination
⁎
Lu Zuo , Ling Ding, Jiajun Chen, Xiaotian Zhou, Bofeng Xu, Zihan Liu
College of Energy and Electrical Engineering, Hohai University, Nanjing 210098, China

A R T I C LE I N FO A B S T R A C T

Keywords: In this study, a model of wind supercharged solar chimney power plant combined with seawater desalination
Solar chimney (WSSCPPCSD) was proposed. The integrated system was divided into wind pressure ventilator and solar chimney
Wind pressure ventilator power generation system for numerical simulation. Three-dimensional unsteady state numerical investigation on
H-type vertical axis wind wheel blade wind pressure ventilator was carried out as well as steady state numerical study on the solar chimney power
Seawater desalination
generation system. The flow field characteristics and performance parameters of the two subsystems were
Numerical simulation
analyzed. Subsequently, the comparison between solar chimney power plant (SCPP), solar chimney power plant
combined with seawater desalination (SCPPCSD) and WSSCPPCSD was conducted. The results show that, al-
though SCPPCSD can greatly improve the solar energy comprehensive utilization, the power generation is re-
duced. The wind pressure ventilator proposed in this paper could not only significantly increase the power
generation but also increase the freshwater production by providing a negative pressure of 64.5 Pa at the
chimney outlet. It was found that at 800 W/m2, an increase in power output of about 14.7 kW and increase in
hourly freshwater production of 30 g/h were obtained.

1. Introduction (2008a) did a numerical simulation of solar chimney power plant

coupled with turbine, and studied the effect of turbine rotational speed
Solar chimney technology is one of the feasible ways to achieve on system performance. Afterwards, the heat storage was included in
large-scale utilization of solar energy, which integrates collector tech- the simulation to analyze the system flow and heat transfer character-
nology, chimney technology and wind turbine technology. The concept istics, together with the heat storage performance under different ir-
of solar chimney on power generation technology was first described by radiance (Ming et al. 2008b). Effects of heat storage thermal properties
Spanish Artillery Colonel, Isidoro Cabanyes (1903)which was later and soil compactness were analyzed by Hurtado et al. (2012) and
employed in the design and construction of the first practical solar Bernardes (2013). The effect of external crosswind on system was stu-
chimney power plant in Manzanares pioneered by Prof. Schlaich and died by Ming et al. (2012) through numerical simulation, and they
his colleagues (Robert 1981). This prototype plant of 50 kW operated reported that, the crosswind could be positive for the system when its
from 1982 to 1989 until being destroyed by a storm, which verified the velocity is above 15 m/s. This factor was also analyzed by Zhou et al.
technical feasibility of solar chimney power plant (SCPP). The experi- (2012), and the variation of enlargement degree on inflow with the
mental results of Spanish prototype were reported in Haaf (1984), chimney height and collector size was included. Considering the ra-
which also involved analysis of energy balance, collector efficiency, diation heat transfer in the collector, Guo et al. (2014) did a numerical
pressure losses in the turbine section, along with the power generation simulation which coupled with radiation, solar load and the fan model,
cost. then in Guo et al. (2015), they replaced the previous fan model with a
This attempt has greatly stimulated the research in this field. In real turbine model, and carried out a comparison between results of
order to evaluate the impact of various factors on the performance o these two methods. Recently, the effects of blade number, rotational
f the solar chimney power plant, an ideal air standard cycle analysis speed, collector diameter, and chimney height were reported in a nu-
with the main losses and the collector model was carried out by Gannon merical study by Kasaeian et al. (2017).
and von Backström (2000). Dai et al. (2003) proposed a simple model of In order to make it possible and economic for the commercial ap-
SCPP to analyze the influence of geometric sizes, environment condi- plication of large-scale Solar Chimney Power Plant, great efforts have
tion and some relevant coefficients on the output power. Ming et al. been devoted to finding a new design structure or to improving the

⁎
Corresponding author.
E-mail address: [email protected] (L. Zuo).

https://doi.org/10.1016/j.solener.2018.03.041
Received 17 September 2017; Received in revised form 13 March 2018; Accepted 14 March 2018
Available online 21 March 2018
0038-092X/ © 2018 Elsevier Ltd. All rights reserved.
L. Zuo et al. Solar Energy 166 (2018) 59–70

Fig. 1. Schematic diagram of WSSCPPCSD system structure.

efficiency with a constant even reduced chimney height. Papageorgiou system, and the results showed that the solar energy utilization effi-
(2004) firstly proposed the concept of floating solar chimney tech- ciency was significantly improved in SCPPCSD. Niroomand and
nology, and it was proved to be more cost competitive compared with Amidpour (2013) proposed a new combination of solar chimney and
other renewable technology by Papageorgiou (2007) and Zhou et al. Humidification-Dehumidification desalination process, and discussed
(2008). Bilgen and Rheault (2005) proposed the design of sloped solar the effects of tube number, humidifier inlet water temperature and its
updraft power plant with the collector built on sloped surface or sui- mass flow rate on the performance of this process, the results also re-
table hills, which is applicable for mountainous region. They reported ported that with the increase of freshwater output, the power output
that the chimney height might be reduced by almost 90% and the would decrease.
thermal performance is slightly improved. Li et al., 2012 proposed a Obviously, the reduction in electricity generated by the coupled
solar and wind energy extraction system, which combined a tornado- system constitutes challenge to the commercialization of Solar Chimney
type wind tower with SCPP. They found that, the pressure deficit Power Plant. This is because the pressure difference in the integrated
generated by tornado-type wind tower could greatly improve the power system is affected as part of the energy gained at the collector was used
generation capacity with the utilization of high-elevation wind energy, for desalination, thus leading to slight decrease in system average
which means that the height can be greatly reduced compared with the temperature and consequently reduction in the pressure difference
same rated conventional SCPP. between the system air and the ambient air. This results in reduction to
In addition, limited by the relatively low overall efficiency of con- the power generation capacity of the coupled system in comparison to
ventional SCPP, the coupling of solar chimney technology and other the operation of conventional SCPP.
renewable technologies becomes a new competitive option, which can Considering the performance challenges associated with SCPPCSD,
improve the comprehensive utilization of solar energy and achieve a novel Wind Supercharged Solar Chimney Power Plant Combined with
multiple objectives simultaneously. Wang et al. (2006) proposed an Seawater Desalination (WSSCPPCSD) was proposed. The model has an
integrated system for output of electricity and freshwater, the climbing integration of supercharged wind pressure ventilator installed at the
saturated moist air got condensed in the heat exchanger installed at the chimney exit of SCPPCSD aimed at improving the performance. The
chimney outlet, and then the freshwater was used to drive the hydraulic numerical study on the proposed system was divided into two compo-
turbine to obtain power output. Based on this structure, some im- nents to study the operations of the solar chimney power generation
provements were made by Zhou et al. (2010), and then a comparison system and the wind pressure ventilator. The study of the flow field
between conventional SCPP and combined solar chimney system for characteristics and performance parameters of the system involves
power generation and seawater desalination (CSCSPD) was conducted. unsteady state numerical simulation of the wind pressure ventilator and
The research reported that the temperature, air velocity and power steady state numerical simulation of solar chimney power generation
generation in CSCSPD were both lower compared with conventional system. Besides, a comparison between SCPP, SCPPCSD and WSSCPP-
SCPP. Zuo et al. (2011) proposed a new solar chimney power system CSD was conducted to analyze the effects of the coupling of wind
with integration of sea water desalination (SCPPCSD), in which closed pressure ventilator.
seawater desalination zone (disk solar distiller) was adopted to replace
the heat storage layer, then the mathematical model was developed and
benefit analysis was conducted. In addition, Zuo et al. (2012) built up
an experimental system to verify the feasibility of this integrated

60
L. Zuo et al. Solar Energy 166 (2018) 59–70

Fig. 2. Energy flow through control volume of collector-desalination region.

2. Model description water vapor gets condensed into water droplets under the glass cover,
eventually a higher productivity of freshwater is achieved.
2.1. Physical model
2.2. Design parameters
The basic model of WSSCPPCSD is shown in Fig. 1. Most of the solar
radiation penetrating the collector is absorbed by seawater layer and Considering the Manzanares solar chimney power plant prototype
the bottom tank. The rise in the temperature of seawater results in adopted for this study, the cylindrical chimney is 194.6 m high with a
evaporation, heat is transferred from seawater to the glass cover by diameter of 10.16 m, the collector has a radius of 122 m and average
radiation and convection. The vapor gets condensed into water droplets height of roof above the ground is 1.85 m (Haaf, 1983, 1984). The flow
at the inner face of the glass cover, and then trickles along the glass passage adopts the design of bell-mouth shape (Fluri and Backström,
cover into the freshwater collecting trough under gravity. Glass cover 2008). The ambient temperature and solar radiation intensity con-
absorbs the long-wave radiation from seawater and the latent heat of sidered in this investigation are 298.15 K and 800 W/m2, respectively,
condensation of water vapor, which in turn heats the cold air flowing in and the external crosswind velocity at the top of the chimney is set as
the collector. Then the temperature of air rises, the density decreases, 15 m/s. In reference to the Glauert theory involved in blade aero-
and a density difference with external environment is formed, thus dynamic design of horizontal axis wind turbine, the single-stage axial
resulting in a pressure difference. The chimney, which acts as a nega- wind turbine blades are designed and generated by BladeGen, and then
tive pressure tube, increases the pressure difference between the system some optimizations have been done. The ventilator blades with CLARK
air and the ambient air, and thus creates a strong updraft that drives the Y airfoil are designed in the manner of variable circulation (Wu, 2007).
turbine installed in the base of the chimney. The turbine consumes most NACA 0012 airfoil is selected for the H-type vertical axis wind wheel
pressure potential and partial kinetic energy of the hot air, and the rest blade, 4° of installation angle, 10 m of wind wheel diameter (D). To
are used to drive the hot air to the outlet of chimney. ensure that the height to diameter ratio of the wind wheel is between
The wind pressure ventilator is installed at the chimney exit to 0.5 and 3, the height of the wind wheel is set to 15 m.
utilize the high altitude wind energy, which adopts a modified H-type
vertical axis wind wheel structure. Specifically, several ventilator 2.3. Mathematical model
blades are rigidly connected to the wind wheel blades and fixed on the
main shaft, and the main shaft is connected to the supporting plate The generalized governing equation of the whole system:
fixed on the chimney wall through bearings. When the wind drives H-
∂ (ρϕ)
type vertical axis wind wheel to rotate, the ventilator blades will rotate + ∇ ·(ρvϕ) = ∇ ·(Γϕ ∇ϕ) + Sϕ
∂t (1)
at the same angular speed. The negative pressure generated at the
bottom of rotating ventilator blades augments the system driving force, where ϕ is generalized dependent variable, can represent 1, v, T, k and ε
the pressure potential absorbed by the turbine increases, inducing a of continuity equation, N-S equation, energy equation, and k−ε equa-
higher output shaft power. Besides, when the velocity of the working tion, respectively; The universal diffusivity Γϕ , the universal source term
fluid in the system increases, it leads to higher convective heat transfer Sϕ and relevant coefficients are determined by the value of ϕ .
coefficient from glass cover to the buoyant airflow in a condition where The momentum equation for rotation coordinate system:
the cover temperature is higher than the air temperature, thus increases
∂ (ρ→vr )
the heat removal factor from the glass cover (Toure and Meukam, + ∇ ·(ρ→→
vr vr ) + ρ (2→
ω ×→
vr + →
ω ×→
ω ×→
r ) = −∇p + ∇ ·(μ∇→
vr )
∂t
1997). Due to the increasing temperature difference between seawater ⎯→
⎯
and the glass cover, as the evaporation of seawater increases, more + F (2)

61
L. Zuo et al. Solar Energy 166 (2018) 59–70

Table 1 structure, the model is divided into solar chimney power generation
Grid partitioning scheme. system and wind pressure ventilator system respectively. The solar
chimney power generation system is composed of collector, chimney
System Domain Type Grids Minimum
(×104) mesh quality and turbine region, and the wind pressure ventilator system is divided
into outer domain, inner domain and rotation domain of H-type vertical
Solar chimney power Collector Structured 310.1 0.5 axis wind wheel.
generation Turbine Unstructured 221.8 0.35
Hybrid structured/unstructured grid method is adopted. For turbine
system region
Chimney Structured 151.3 0.7 blades and ventilator blades, the chord length and installation angle
both vary greatly from blade root to blade tip. Consequently, although
Wind pressure Inner Structured 113.6 0.7
ventilator domain
the Jacobian matrix quality could reach 0.5 when adopting the struc-
Rotation Unstructured 286.8 0.3 tured grid, the grid quality is only 0.1 according to the skew angle,
domain leading to the final quality of grid is only about 0.1. Therefore, un-
Outer Structured 235.2 0.7 structured grids are generated in these two rotation domains, and
domain
structured grids are employed in other regions due to the symmetrical
structure. The grids of rotation domain and the fixed domain are con-
Shaft power from the turbine can be obtained by: nected by sliding mesh interface, and the mesh size ratio across the
interface is between 0.5 and 2 in order to minimize the discretization
Ptur = TΩ (3) error. Besides, tests with different meshing schemes for each subsystem
The turbine efficiency can be obtained by: have been conducted to investigate the grid independence. Finally, for
solar chimney power generation system and wind pressure ventilator
ρPtur
ηtur = × 100% system, the partitioning schemes with total grid number of 683, 2000
ṁ Δp (4) cells and 635, 6000 cells were adopt, respectively. The specific scheme
By analyzing the energy transfer mechanism, a mathematical model is shown in Table 1 and the grid divisions of two subsystems are shown
of airflow and heat transfer to the control volume (see Fig. 2) from a in Figs. 3 and 4.
radius, r, to incremental radius of r+Δr is established with reference to
the procedure developed by Pasumarthi and Sherif (1998). 3.2. Boundary conditions and numerical solution
The heat balance equation of collector roof:
qr,g −c + α c G = qr,c −s + qc,c −a + qc,c −m (5) To obtain the boundary temperature distributions, the heat balance
equations are discretized and solved by MATLAB according to the
The heat balance equation of the airflow in the collector: procedure shown in Fig. 5. After multiple iterations, a series of tem-
cp,m ṁ ∂Tm perature scatter points are obtained, and the least squares method is
qc,c −m + qc,g −m = − · used to fit these discrete points with tenth-order polynomials. Then the
2πr ∂r (6)
polynomials are programmed in the method of cell-centered dis-
For the SCPP, there are only the heat balance equations of collector cretization and imported as Fluent boundary conditions by compiled
roof, airflow in the collector and the heat storage layer, and the sub- UDF (User defined function).
script g of glass cover in the formula (5) and (6) should be changed to The numerical simulation solver is Fluent. For the simulations of
the subscript p of heat storage layer. solar chimney power generation system in SCPP, SCPPCSD and WSS-
The heat balance equation of SCPP heat storage layer: CPPCSD, pressure inlet and pressure outlet are used for system inlet and
α p τc τg G = qr,p −c + qc,p −m + qh,soil (7) outlet, and temperatures of the collector inlet and the chimney outlet
are considered to be equal to the environment temperature. The gauge
The heat balance equation of glass cover: pressures at the collector inlet and the chimney outlet are both set at
qe,w + qc,w −g + qr,w −g + αg τc G = qc,g −m + qr,g −c 0 Pa in SCPP and SCPPCSD, while in WSSCPPCSD, the supercharging
(8)
system is expected to provide a certain negative pressure at the chimney
The heat balance equation of seawater layer: outlet. The turbine region adopts the multiple reference frame (MRF)
α w τc τg G + qc,b −w = qe,w + qr,w −g + qc,w −g model (Ming et al., 2008a), interface of rotational domain and fixed
(9)
domain are set as Interface, and no-slip wall boundary conditions are
The heat balance equation of pond bottom plate: adopted for the rest. In this high-Rayleigh buoyancy-driven subsystem,
the air density is mainly affected by temperature, so the Boussinesq
αb τg τc (1−α w ) G = qc,b −w + qh,ss (10)
The hourly productivity of freshwater can be written as:
qe,w
Mw = × 3600
hw (11)
The utilization efficiency of solar energy of the integrated system
can be achieved by:
0.25π(D2coll −D2ch) qe,w + Ptur
ηsys = × 100%
0.25π(D2coll −D2ch) G (12)

3. Numerical processing method

3.1. Grid division and independence verification

ICEM CFD is used for three-dimensional geometric modeling and

Fig. 3. Meshing of solar chimney power generation system.
meshing during pre-processing. Considering the huge size of the whole

62
L. Zuo et al. Solar Energy 166 (2018) 59–70

approximation is used to approximate the buoyancy forces, and the

thermal expansion coefficient takes a value of 0.00335. The standard
k−ε model coupled with the enhanced wall function is adopted to si-
mulate the flow in this part. SIMPLEC algorithm is applied for pressure-
velocity coupling, the PRESTO! discretization scheme is used for pres-
sure interpolation, and the QUICK discretization scheme is used for
other terms.
As for wind pressure ventilator system, the whole calculation do-
main is 4.5D × 2.5D × 2D (D is the wind wheel diameter). The hor-
izontal inlet is set as the speed-inlet boundary with a given speed of
15 m/s, the bottom of rotation domain is set as the pressure-inlet
boundary whose initial gauge pressure is 0 Pa, the horizontal outlet and
top outlet are considered as pressure-outlet boundaries, of which the
relative pressures are both 0 Pa. Frame motion and Mesh motion are
chosen for the rotation domain during the steady simulation and the
unsteady simulation of this subsystem, respectively. The bottom outer
domain adopts symmetry boundary conditions, the interface is set as
Interface condition, and no-slip-wall boundary conditions are adopted
for the rest. The RNG k−ε model with standard wall function is selected.
SIMPLE algorithm is applied for pressure-velocity coupling, the dis-
cretization scheme for terms of the governing equations are the same as
above.

4. Results and discussion

Fig. 4. Meshing of wind pressure ventilator system.
4.1. Validation of numerical simulation

The heat balance equations mentioned in 2.3 are discretized and

solved by MATLAB, and the boundary temperature distributions are
shown in Fig. 6. The temperatures of the SCPP collector roof and the
heat storage layer increase slowly from the inlet and drop sharply at the
collector outlet, the smooth increase can be explained by the small
airflow velocity in the collector except outlet area. When air flows
through the collector outlet, the contraction of the flow area results in a
higher velocity according to the mass flow conservation, and the con-
vective heat transfer coefficient also increases. Then more heat is taken
away by the hot air, leading to a sharp decrease in the temperatures of
the collector and the heat storage layer. For SCPPCSD, some of the heat
is absorbed by seawater, so the collector roof temperature is lower than
that in SCPP. In addition, the temperature difference between collector
roofs of SCPP and SCPPCSD is more obvious compared with the dif-
ference between SCPPCSD glass cover and SCPP heat storage layer. This
is because in SCPP, the heat absorbed by collector roof mainly comes
from the long-wave radiation of heat storage layer, which is sig-
nificantly reduced when using glass cover, resulting in a lower collector

Fig. 5. Flow chart of the solution procedure in Matlab.

Fig. 6. Boundary temperature distribution for SCPP and SCPPCSD.

63
L. Zuo et al. Solar Energy 166 (2018) 59–70

roof temperature in SCPPCSD. However, considering the long-wave region is formed from the rotating area of the wind wheel to the
radiation from seawater and latent heat of condensation absorbed by downstream area, and the cycle of the wake flow is 72°. The velocity at
glass cover, the temperature drop is relatively small. right downwind side of the wake region is smaller compared with that
According to Haaf (1984), with an insolation of 800 W/m2, when at left downwind side. This is because at right downwind side of the
the turbine rotational speed was 100 rpm, the maximum temperature wind wheel, the direction of rotational velocity component is opposite
difference between the inlet and outlet of the collector was 17.5 K, to the wind, so wind is blocked by the clockwise rotating wind wheel,
airflow velocity in the chimney was 9 m/s, turbine pressure drop was leading to an obvious decrease in velocity. As the wake flow moves
120 Pa, pressure drop factor was 0.67 and output power was 36 kW. As away from the wind wheel, the wake vortex falls off and the wake
to the simulation of the basis SCPP in this paper, under the same op- velocity gradually approaches the ambient wind velocity. The reason
erating condition, the maximum temperature difference of the airflow for this phenomenon is that there exists a velocity gradient between the
is 16.99 K, chimney outlet airflow velocity is 9.2 m/s, turbine pressure free flow and the wake, resulting in an additional shear turbulence,
drop is 116 Pa, pressure drop factor is 0.66 and output shaft power is which leads to the energy conversion and mixing between them. Finally
37.8 kW. The maximum tolerance of these mentioned parameters is 5% the velocity gradient is eventually eliminated as the diffusion of the
through comparison between the simulation results and the experi- mixing area.
mental results generated from the Manzanares pilot plant (Haaf, 1984), The pressure distribution at the bottom inlet of the rotation domain
so it is clear that good agreements are obtained, which proves the re- is shown in Fig. 8. The pressure on the windward surface is larger and
liability of numerical simulation method proposed in this paper. The there exists some positive pressure area, while the pressure on the
slightly higher simulation output power can be attributed to the bell- leeward surface is mainly negative. The airflow on the windward sur-
mouth shape flow passage design method adopted for transition sec- face is blocked by the wind wheel and gradually accumulates, which
tion, which results in a slight decrease in flow loss. results in positive pressure, however, the airflow on the leeward surface
is mainly influenced by ventilator blades, resulting in a wide range of
4.2. Simulation results of wind pressure ventilator negative pressure area, especially under the ventilator blades. Com-
pared with the right downwind side of the wind wheel, the pressure
The flow in the wind pressure ventilator changes notably with time, under the ventilator blades is larger on left downwind side, which even
so an unsteady numerical simulation is essential for detailed study. As equals the value of the surrounding airflow. The natural wind on left
the wind pressure ventilator proposed in this paper is a non-powered downwind side of the wind wheel has the same direction with the
device, the torque generated by H-type vertical axis wind wheel blades horizontal velocity component of rotating ventilator blades, so the re-
is totally used to drive the ventilator blades, it is necessary to carry out lative velocity between ventilator lower surface and airflow gradually
a steady simulation to determine the equilibrium condition before un- decreases, resulting in a higher pressure. As to the right downwind side,
steady simulation. the pressure under ventilator blades is relatively smaller inversely. In
After repeatedly adjusting the rotational speed of the wind wheel, the windward side of the H-type vertical axis wind wheel, the pressure
the torque balance is obtained at the rotational speed of 69.3 rpm with on the upper surface of the blade is larger, which pushes wind wheel to
the torques of 3648 N·m and −3627 N·m for wind wheel blades and rotate continually, however, for the leeward side, the pressure on the
ventilator blades, respectively, and the average pressure generated at lower surface is larger, which will prevent the rotation of wind wheel,
the bottom inlet of this subsystem is −64.5 Pa. Subsequently, based on so the rotation of blades in leeward side are driven by blades in
the results of the steady simulation, the unsteady simulation was car- windward side. The airflow near the blade leading edge is squeezed
ried out. The rotation period is 0.866 s, so the time step is set as 0.001 s. during rotation, so the pressure here is relatively large. Although there
Time period that the flow field is stable is intercept, and the time step is exits partial positive pressure area at the bottom inlet, the mean pres-
transformed into the change of azimuth angle of the blade for a more sure is still negative, and the negative pressure generated here can ef-
intuitive analysis. ficiently improve the performance of the solar chimney integrated
The velocity distribution in the cross section of wind pressure ven- system.
tilator at different azimuth angles is shown in Fig. 7. It can be seen that
the velocity gradient changes evidently before and after the wind wheel
and the flow velocity decreases gradually. A narrow and long wake

Fig. 7. Velocity distribution in the cross section of wind pressure ventilator.

64
L. Zuo et al. Solar Energy 166 (2018) 59–70

Fig. 8. Pressure distribution at the bottom inlet of wind pressure ventilator.

4.3. Blade surface pressure distribution and torque variation in wind The change period of H-type vertical axis wind wheel blades total
pressure ventilator torque is 72°. In the first cycle, the maximum torque appears at the
azimuth angle of 40°, and the minimum appears at the azimuth angle of
As is shown in Fig. 9, there exist great differences in wind wheel 3°. The total torque increases sharply at first and then has a significant
blade surface pressure distribution at different azimuth angles. When reduction with the increase of azimuth angle. The average total torque
the azimuth is between 90° and 270°, the pressure on the lower surface is similar with the value under equilibrium condition in steady simu-
of the blade is greater than the upper surface, which is exactly contrary lation, which also validates the unsteady numerical simulation method
to the distribution between 300° and 60°. In general, the pressure dis- used in this paper. The total torque of the ventilator blades fluctuates
tribution on the wind wheel blade varies periodically. The maximum slightly around the results of steady simulation, and the change of the
pressure difference between the upper and lower surface occurs near single blade torque is small, so it is not illustrated.
the blade leading edge, where generates great lift. The pressure dif-
ference is almost zero at trailing edge, this is mainly caused by the small 4.4. Results of steady simulation in solar chimney power generation system
trailing edge angle of the selected airfoil, resulting in an inconspicuous
pressure gradient. Fig. 11 shows the pressure distribution of system axial plane. It can
Fig. 10 shows the variation of blade torque in wind pressure ven- be seen that the internal pressure is symmetrically distributed about the
tilator with respect to azimuth angle. Specifically, the torque of a single central axis of the chimney, and the pressure variation is smooth and
wind wheel blade changes periodically with azimuth angle, and the uniform. The relative pressure at the collector inlet is 0, and the var-
maximum appears near the azimuth angle of 330° with a toque of iation of the pressure through the collector is much small, so it can be
11,360 N m, which is corresponding with the result shown in Fig. 9 l regarded as a constant-pressure endothermic process. At the transition
that the total pressure difference between the upper and lower surface section, airflow is squeezed and the speed increases accordingly, which
of the blade is maximum in 330°. It is actually the tremendous torque can explain the notable decrease of pressure in this area. There is a
generated by blade at azimuth angle of 330° that compensate for the slight increase in pressure at turbine inlet due to the obstruction, after
rather small or even negative torque of other blades to maintain the getting across the turbine, the pressure drops drastically to a minimum,
rotation of the wind wheel. The minimum torque appears near the and then slowly rebound until be same with the pressure at the chimney
azimuth angle of 150° with a value of −3634 N·m, which can also be outlet.
supported by the negative total pressure difference shown in Fig. 9f. It The characteristic of turbine area is analyzed according to the
should be noted that although the pressure difference at the blade pressure change due to its drastic variation trend, and several typical
leading edge in Fig. 9i is largest among all azimuth angles, the absolute streamlines have been selected. Fig. 12 shows the variation of pressure
value of total pressure difference is lower than that in Fig. 9f, which can on the streamline with respect to height in turbine area. At the inlet of
be attributed to the relatively smaller pressure difference from middle turbine area, the pressure increases slightly due to the resistance. After
to the trailing edge compared with Fig. 9f. Besides, the arm of the blade getting across the blade, the flow is not only affected by pumping force
lifting force for the rotation center at the azimuth angle of 150° is of the chimney but also the negative pressure surface of the rotating
greater than that of 240°, so the absolute value of blade torque is larger blade, so the negative pressure decreases rapidly to a few hundred
near the azimuth angle of 150°. Pascal, and it slowly picks up at the turbine outlet as the flow is only

65
L. Zuo et al. Solar Energy 166 (2018) 59–70

Fig. 9. Pressure distribution of H-type vertical axis wind wheel blade surface under different azimuth angles.

Fig. 10. The variation curves of blade torque with azimuth angle.

driven by pumping force of the chimney then. During the first rising
phase, pressure can reach hundreds of Pascal for minor blocked airflow
located near the positive pressure surface. As for the rest of airflow that Fig. 11. Pressure distribution of axial plane of solar chimney power generation system.
flows between adjacent blades, the pressure increases slightly as the
resistance is small, and the value is still negative.
Fig. 13 shows the velocity distribution of turbine inlet and outlet. It due to the relatively large obstruction of the blades, the velocity of the
can be seen that the airflow at the inlet and outlet of the turbine is airflow under the blade is smaller than the airflow around the blade. At
relatively uniform and there is no imbalance. At the inlet of the turbine, the outlet of the turbine, the static pressure drops sharply as the airflow
flowing through the turbine, and turns into dynamic pressure, so the

66
L. Zuo et al. Solar Energy 166 (2018) 59–70

Fig. 12. The variation curves of pressure with height on the streamline.

velocity of airflow increases. The shock of the airflow on blades not

only drives the blade rotates but also equips itself with rotational speed
at the same time, so the airflow at the turbine outlet is a vortex flow. In
addition, there is a small part of airflow rushing directly through the tip
clearance instead of flowing across the blades.
Fig. 14 shows the pressure distribution of the positive and negative
pressure surfaces of the turbine blades. The pressure difference between Fig. 14. Pressure distribution of positive and negative pressure surfaces of the turbine
two surfaces drives the rotation of the blade, and the pressure variation blade.
on the blade is gentle. As the leading edge is the part that first contacts
with the airflow when the blade rotates, the velocity gradient of the decrease, while the turbine pressure drop and the maximum tempera-
airflow changes remarkably, leading to a maximum pressure difference ture difference increase inversely. This is because the increase of the
at the leading edge and it gradually decreases towards the trailing edge. turbine rotational speed will lead to a larger resistance, which reduces
In addition, the pressure of blade tip is relatively small because part of the mass flow rate of the system and then the chimney outlet air ve-
the airflow rushes directly through the tip clearance. locity decreases accordingly. The airflow is thus heated for a longer
time in the collector, leading to a higher airflow temperature at the
5. Comparison between SCPP, SCPPCSD and WSSCPPSCSD chimney inlet. According to the Boussinesq approximation, the density
difference between inside and outside the chimney increases, so the
Figs. 15 and 16 show the effect of different turbine rotational speeds system draft increases, and then the pressure drop across the turbine
on maximum temperature difference, chimney outlet velocity, mass increases. Besides, the variation trend of pressure drop becomes gentle
flow rate and turbine pressure drop of SCPP, SCPPCSD and WSSCPP- when the speed surpasses a certain value.
CSD. Specifically, the chimney outlet pressure of WSSCPPCSD takes the For certain rotational speed, that relevant parameters mentioned
results obtained in 4.2, −64.5 Pa. With the increase of the turbine ro- above in SCPPCSD are smaller than SCPP system, this can be explained
tational speed, the system mass flow rate and chimney outlet velocity

Fig. 13. Air velocity distribution at turbine inlet and outlet.

67
L. Zuo et al. Solar Energy 166 (2018) 59–70

Fig. 15. The variation curves of maximum temperature difference and chimney outlet air Fig. 17. The variation curves of turbine shaft power and efficiency.
velocity.

Fig. 18. Variation curves of turbine shaft power and freshwater output with negative
Fig. 16. The variation curves of mass flow rate and turbine pressure drop. pressure at chimney outlet.

by the lower maximum temperature difference of airflow as part of performance with different negative pressure at chimney outlet. It can
solar radiation is absorbed for desalination. By comparing the curves of be seen from Fig. 18 that the turbine shaft power increases almost
SCPPCSD and WSSCPPCSD, the chimney outlet velocity, mass flow rate linearly with the increase of negative pressure. With a higher negative
and turbine pressure drop in WSSCPPCSD are both higher than pressure, the driving force increases, which leads to the increases in
SCPPCSD except the maximum temperature difference. This is because turbine shaft power and mass flow rate, then the heat convection be-
with the supplement of negative pressure generated by wind pressure tween airflow and glass cover is enhanced, the temperature of glass
ventilator, the pressure difference is greatly enhanced, pressure drop cover consequently decreases. A larger temperature difference between
absorbed by turbine and the velocity therefore increase, resulting a seawater and glass cover improves the evaporation of seawater and
larger mass flow rate. Then, heating time during air flowing through the eventually the hourly freshwater productivity. In fact, there is a limit to
collector is reduced due to the higher velocity, so the maximum tem- the rate of seawater evaporation, which means that with the increase of
perature difference is relatively smaller. chimney outlet negative pressure, the hourly freshwater productivity
Fig. 17 shows the variations of shaft power and turbine efficiency will finally reach a certain value.
with rotational speed in SCPP, SCPPCSD and WSSCPPCSD. The peak More specifically, on the basis of simulation results of wind pressure
shaft power and efficiency both occur at the rotational speed of ventilator, if the chimney outlet pressure of SCPPCSD is −64.5 Pa, the
100 rpm. Because of the additional freshwater output in SCPPCSD, the turbine shaft power can increase by 14.7 kW, and hourly freshwater
shaft power is always lower than that of SCPP at the same rotational productivity can simultaneously increase by 30 g/h. The system com-
speed, resulting in a 15% loss of electricity generation, which is con- prehensive utilization of solar energy will rise by 30% and 70% com-
sistent with the results of Figs. 15 and 16. In addition, the shaft power pared with SCPPCSD and SCPP respectively. Admittedly, the negative
of WSSCPPCSD is 45% higher than SCPPCSD. The efficiency of pressure of the chimney outlet is lower than 64.5 Pa when the wind
SCPPCSD is highest when turbine rotational speed is relatively low, but pressure ventilator proposed is directly coupled with SCPPCSD, which
lowest at high rotational speed contrarily. Under the same low rota- can be regarded as a SCPPCSD under the load condition of the wind
tional speed, the heat flow power is small in both cases. Due to the slow pressure ventilator, but this value can be reached by increasing the
and smooth airflow in SCPPCSD, aerodynamic loss of this system is height of the wind pressure ventilator. As the height increases, the
much smaller, so its turbine efficiency is higher. On the contrary, when torque of H-type vertical axis wind wheel blades increases, leading to a
it comes to a high rotational speed, the heat flow power is large, so the higher rational speed of ventilator, and the driving force is improved
proportion of aerodynamic loss is small, efficiency is mainly dominated consequently.
by shaft power, so the efficiency of WSSCPPCSD is highest then.
Fig. 18 shows the variation curve of WSSCPPCSD overall

68
L. Zuo et al. Solar Energy 166 (2018) 59–70

6. Conclusion Cabanyes, I., 1903. Las Chimeneassolares (Solar Chimneys). La Ernergiaeléctr.

Dai, Y., Huang, H., Wang, R., 2003. Case study of solar chimney power plants in
Northwestern regions of China. Renew. Energy 28, 1295–1304.
To offset the shortcoming of the SCPPCSD, this paper proposed a Fluri, T.P., Backström, T.W.V., 2008. Performance analysis of the power conversion unit
new integrated system, wind supercharged solar chimney power plant of a solar chimney power plant. Sol. Energy 82 (11), 999–1008.
combined with seawater desalination (WSSCPPCSD). During the nu- Gannon, A.J., von Backström, T.W., 2000. Solar chimney cycle analysis with system loss
and solar collector performance. J. Solar Energy Eng. – Trans. ASME 122 (3),
merical simulation, the WSSCPPCSD was divided into wind pressure 133–137.
ventilator and solar chimney power generation system. Unsteady si- Guo, P., Li, J., Wang, Y., 2014. Numerical simulations of solar chimney power plant with
mulation for wind pressure ventilator was carried out to study its flow radiation model. Renew. Energy 62 (3), 24–30.
Guo, P., Li, J., Wang, Y., Wang, Y., 2015. Numerical study on the performance of a solar
field characteristic, and the analysis on surface pressure and torque chimney power plant. Energy Convers. Manage. 105 (WOM), 197–205.
variation of H-type wind wheel blade and ventilator blade were con- Haaf, W., 1984. Solar chimneys part II: preliminary test results from the Manzanares pilot
ducted. Afterwards, steady simulations of solar chimney power gen- plant. Int. J. Solar Energy 2, 141–161.
Haaf, W., Friedrich, K., Mayr, G., Schlaich, J., 1983. Solar chimneys part I: principle and
eration system in SCPP, SCPPCSD and WSSCPPCSD were carried out,
construction of the pilot plant in Manzanares. Int. J. Solar Energy 2, 3–20.
the flow field of this subsystem in WSCPPCSD was analyzed, especially Hurtado, F.J., Kaiser, A.S., Zamora, B., 2012. Evaluation of the influence of soil thermal
the turbine region. Then the performance characteristics were obtained inertia on the performance of a solar chimney power plant. Energy 47 (1), 213–224.
through comparisons between the three. Finally, in order to explain the Kasaeian, A., Mahmoudi, A.R., Astaraei, F.R., Hejab, A., 2017. 3d simulation of solar
chimney power plant considering turbine blades. Energy Convers. Manage. 147,
superiority of the supercharger system more specifically, the influence 55–65.
of chimney outlet negative pressure on the performance of WSSCPPCSD Li, J., Guo, P., Wang, Y., 2012. Preliminary investigation of a novel solar and wind energy
was investigated. The main conclusions can be summarized as follows: extraction system. Procee. Institut. Mech. Eng. Part A J. Power Energy 226 (Energy),
73–85.
Ming, T., Liu, W., Xu, G., Xiong, Y., Guan, X., Pan, Y., 2008a. Numerical simulation of the
(1) The flow field of wind pressure ventilator is relatively complicated, solar chimney power plant systems coupled with turbine. Renew. Energy 33 (5),
although there exists partial positive pressure at the bottom inlet, 897–905.
Ming, T., Liu, W., Pan, Y., Xu, G., 2008b. Numerical analysis of flow and heat transfer
the mean pressure is always negative. The blade torque of wind characteristics in solar chimney power plants with energy storage layer. Energy
pressure ventilator varies periodically. The torque of single H-type Convers. Manage. 49 (10), 2872–2879.
vertical axis wind wheel blade fluctuates strongly with the change Ming, T., Wang, X., Richter, R.K.D., Liu, W., Wu, T., Pan, Y., 2012. Numerical analysis on
the influence of ambient crosswind on the performance of solar updraft power plant
of azimuth angle, most of the torque on the windward side is po- system. Renew. Sustain. Energy Rev. 16 (8), 5567–5583.
sitive, and most is negative on the leeward side. The total torque of Niroomand, N., Amidpour, M., 2013. New combination of solar chimney for power
H-type vertical axis wind wheel blades substantially increases at generation and seawater desalination. Desalin. Water Treat. 51 (40–42), 7401–7411.
Papageorgiou C., Floating solar chimney. WO2004/085846 A1/07-10-2004.
first and then has a significant reduction in a cycle, and the total
Papageorgiou, C., 2007. Floating solar chimney versus concrete solar chimney power
torque of the ventilator blades is almost constant. plants. Int. Conf. Clean Electric. Power IEEE 760–765.
(2) For the solar chimney power generation system, the pressure is Pasumarthi, N., Sherif, S.A., 1998. Experimental and theoretical performance of a de-
symmetrically distributed with a smooth variation trend. The monstration solar chimney model—Part I: mathematical model development. Int. J.
Energy Res. 22 (3), 277–288.
pressure increases slightly through the collector, and then drops Robert, R., 1981. Spanish solar chimney nears completion. MPS Rev. 6, 21–23.
sharply at the transition section. There exists a slight increase be- Toure, S., Meukam, P., 1997. A numerical model and experimental investigation for a
fore the turbine, the minimum is reached after the airflow getting solar still in climatic conditions in abidjan (côte d'ivoire). Renew. Energy 11 (3),
319–330.
across the turbine, and then slowly rebound until be same with the Wang, Y., Fang, Z., Li, Z., Yang, Z., Wang, J., Han, L., 2006. Study on the integrated
pressure at the chimney outlet. The airflow at the inlet and outlet of utilization of seawater by solar chimney. Acta Energiae Solaris Sinica 27 (4), 382–387
the turbine region is relatively uniform and there is no imbalance, (in Chinese).
Wu, B., 2007. Engineering design of axial fan. Jilin: Jilin University Press, pp. 201–203
and there exists a vortex flow at the turbine outlet. In addition, the (in Chinese).
pressure difference between turbine blade surfaces is largest at the Zhou, X., Yang, J., Wang, F., Xiao, B., 2008. Economic analysis of power generation from
leading edge and gradually decreases towards the trailing edge. floating solar chimney power plant. Renew. Sustain. Energy Rev. 13, 736–749.
Zhou, X., Xiao, B., Liu, W., Guo, X., Yang, J., 2010. Comparison of classical solar chimney
(3) With the increase of the turbine rotational speed, the system mass
power system and combined solar chimney system for power generation and sea-
flow rate and the airflow velocity at chimney outlet decrease, the water desalination. Desalination 250, 249–256.
maximum temperature difference and the turbine pressure drop Zhou, X., Bernardes, M.A.D.S., Ochieng, R.M., 2012. Influence of atmospheric cross flow
on solar updraft tower inflow. Energy 42 (1), 393–400.
increase contrarily, and the increasing trends become gentle.
Zuo, L., Zheng, Y., Li, Z., Sha, Y., 2011. Solar chimneys integrated with sea water desa-
Besides, the turbine shaft power and efficiency have a peak at lination. Desalination 276 (1–3), 207–213.
certain speed. Although the SCPPCSD greatly improves the com- Zuo, L., Yuan, Y., Li, Z., Zheng, Y., 2012. Experimental research on solar chimneys in-
prehensive utilization rate of solar energy compared with SCPP, the tegrated with seawater desalination under practical weather condition. Desalination
298 (16), 22–33.
generation capacity is reduced. At the rotational speed of 100 rpm,
the shaft power of WSSCPPCSD is 45% higher compared with
Glossary
SCPPCSD. With the supplement of negative pressure generated at
the chimney outlet, the output power of the system increases line-
arly and the freshwater production increases accordingly. Symbol
Generally, the comprehensive utilization rate of solar energy is ef-
ficiently improved by the wind pressure ventilator. cp : specific heat capacity, J/(kg·K−1)
D : diameter, m
F : mass force, m/s2
Acknowledgements G : solar radiation intensity, W/m2
h : latent heat of unit area, J/(kg·m2)
k : kinetic energy of turbulent fluctuation, J
This research is financially supported by National Key Research and M : hourly freshwater productivity, kg/h
Development Program of China (No. 2016YFB0900103). ṁ : mass flow, kg/s
P : power, W
p : pressure, Pa
References q : heat transfer rate, W/m2
r : radius, m
Bernardes, M.A.D.S., 2013. On the heat storage in solar updraft tower collectors – in- T : temperature, K or torque, N·m
fluence of soil thermal properties. Sol. Energy 98 (3), 49–57. v : velocity, m/s
Bilgen, E., Rheault, J., 2005. Solar chimney power plants for high latitudes. Sol. Energy
79 (5), 449–458.

69
L. Zuo et al. Solar Energy 166 (2018) 59–70

Subscript sys: system

tur: wind turbine
w: seawater
a: environment or air
b: pond bottom plate
c: collector roof or heat convection Greek letter
ch: chimney
coll: collector α : absorptivity
e: evaporation ω : rotation angular velocity vector
g: glass cover η : efficiency
h: heat conduction μ : dynamic viscosity, Pa·s
m: hot airflow ρ : density, kg/m3
p: heat storage layer ϕ : universal variable
r: radiation ε : the turbulent dissipation rate of turbulent kinetic energy
s: sky τ : transmissivity
soil: soil Ω : turbine rotational speed, rad/s
ss: bottom and sides of distillation pond

70