Applied Thermal Engineering 196 (2021) 117320

Contents lists available at ScienceDirect

Applied Thermal Engineering

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

Solar-driven natural vacuum desalination system with inner condenser

Lu Wang, Hongfei Zheng, Yunsheng Zhao, Xinglong Ma *
School of Mechanical Engineering, Beijing Institute of Technology, Beijing 100081, China

A R T I C L E I N F O A B S T R A C T

Keywords: Desalination under low pressure is an effective way of increasing freshwater productivity. A natural vacuum
Solar energy desalination system with inner condenser driven by solar energy was proposed in this study. Through gravity and
Distiller local atmospheric pressure, a vacuum condition could be formed in distiller, which allows flowing seawater to
Natural vacuum desalination
absorb a small amount of solar heat to achieve phase change evaporation. In addition, the system can easily
Vacuum pressure
Performance ratio
remove accumulated non condensable gas through the water-filling and air-exhausting process. Transient
characteristics of the system under constant hot seawater temperature were presented. And performance of solar
energy-powered distillation was tested under actual weather using flat-plate solar collectors with a total area of
18 m2. The constant temperature experiment results show that freshwater yield and cycle time increase with the
increase of hot seawater temperature. Hourly water yield can reach 12.45 kg/h when the hot seawater tem­
perature is 60 ◦ C and the seawater mass flow rate is 0.1 kg/s. Daily water yield is 154.14 kg, and the corre­
sponding performance ratio is 1.36 under real weather with an average solar irradiance of 672 W/m2. Finally,
economic analysis results indicate that the pure water cost is $ 6.9 per ton.

problems of corrosion and pollution during long-term operation, thus,

operation life and reliability are difficult to guarantee [10]. Meanwhile,
1. Introduction
the vacuum enhancement method can be more facile than using special
materials. Vacuum enhancement mainly reduces vapor pressure of
Shortage of freshwater resources has become a key factor in
water surface by evacuating the evaporator to ensure that the water
restricting economic development and threatening human health with
inside the evaporator will rapidly evaporate and boil. This method
the rapid growth of the world population and the acceleration of ur­
presents the advantage of achieving rapid evaporation under low-
banization. An estimated 1.8 billion people will face water shortage by
temperature conditions with minimal energy consumption. Al-
2025 [1]. The daily production of around 63 × 106 m3 of freshwater
Hussaini and Smith [11] proved that using vacuum technology in a
requires approximately 3.78 × 108 kW⋅h/d of electricity [2]. The con­
solar still can increase the freshwater yield by nearly 100%.
sumption of a large amount of nonrenewable resources in conventional
The vacuum desalination system driven by solar energy has recently
seawater desalination techniques, such as reverse osmosis, thermal
attracted considerable research attention. Vacuum pumps are
distillation, electrodialysis, or their combinations, may have a negative
commonly used to extract gas from the evaporation chamber and ach­
effect on the environment.
ieve vacuum conditions. Xie et al. [12] theoretically and experimentally
Solar stills are low-cost desalination devices with a simple structure
investigated the performance of a tubular solar still under vacuum
that have been widely used in water shortage areas. However, its limi­
operation condition. The theoretical analysis showed that vacuum
tation of slow evaporation rate results in a typical efficiency of 30%–
operation pressure augments the concentration difference in humid
40% [3,4]. Hence, strengthening the evaporation process has become an
vapor around the trough and condensation surface and intensifies
important research topic. Studies on strengthening the evaporation
diffusion and natural convection of vapor in the cavity. Compared with
process can be classified into two categories, namely, functional mate­
the normal operating pressure condition, the energy efficiency of the
rial and vacuum enhancement. Functional material enhancement is the
system improved by more than 80%. Morad et al. [13] developed a
process of placing nano-photothermal materials with natural broad-light
solar-powered desalination system using a condenser integrated with a
spectrum range, such as graphenes, carbon nanotubes, polymers, and
flat-plate solar collector and vacuum pump for producing freshwater.
nanometals, into the distiller to improve solar absorption and evapora­
The result proved that the enhanced system can achieve higher
tion efficiency [5–9]. However, functional materials demonstrate

* Corresponding author.
E-mail address: [email protected] (X. Ma).

https://doi.org/10.1016/j.applthermaleng.2021.117320
Received 8 March 2021; Received in revised form 15 June 2021; Accepted 1 July 2021
Available online 7 July 2021
1359-4311/© 2021 Elsevier Ltd. All rights reserved.
L. Wang et al. Applied Thermal Engineering 196 (2021) 117320

Nomenclature Qc increased energy of cold seawater, W

Qf decreased energy due to freshwater outflow, W
Ac surface area of condenser, m2 Qh decreased energy of hot seawater, W
As area of flat-plate collector, m2 Qs solar energy absorbed by flat-plat solar collector, W
cp,s specific heat of seawater, J/kg⋅K tp runimg time of water pump, s
hf,g vaporization latent heat, kJ/K Tb,out brine outlet temperature, ◦ C
I radiation, W/m2 Tc,in cold seawater inlet temperature, ◦ C
j annual loss factor of the desalination system Tc,out cold seawater outlet temperature, ◦ C
K total heat transfer coefficient of the distiller, kW /m2◦ C Th,in hot seawater inlet temperature, ◦ C
mb mass flow rate of brine, kg/s Th,out hot seawater outlet temperature, ◦ C
me evaporation rate of seawater, kg/s Y service life of the device, yrs
mf mass flow rate of freshwater, kg/s Δtm logarithmic mean temperature difference of distiller, ◦ C
ms mass flow rate of seawater, kg/s ε uncertainty value
Mf,tot total freshwater production, kg ηs solar collector efficiency
Ms,tot total mass of seawater flowing through the distiller, kg

freshwater productivity of 3.67 L/d with lower cost than the system then proposed a two-stage NVSD system with a specific energy con­
without a vacuum pump. Wessley and Mathews [14] designed a small- sumption of 1500 kJ/kg [26]. Ambarita [27] studied the work charac­
scale low-temperature flash desalination system using a flat collector. teristics and performance of a NVSD system using low grade heat source.
The system can produce 3 L of freshwater per day to meet a single-family The simulation showed that the proposed system with evaporator sur­
need. Besides, in order to develop a desalination system suitable for the face area of 0.2 m2, operation time of 9 h and a source temperature of
Mediterranean region, Andrés-Mañas et al. [15] evaluated the perfor­ 80 ◦ C would produce 6.63 L freshwater. The corresponding thermal
mance of a vacuum multi-effect membrane distillation modules at pilot efficiency was 70.4%. To use the input energy more effectively, Maroo
scale. They could get results that maximum distillate flux reached 8.5 L and Groswami [28] proposed a two-stage flash evaporation desalination
h− 1m− 2 for the hot feed temperature of 75 ◦ C and flow rate of 150 L h− 1. system with natural vacuum. When coupled with a solar collector of 1
The gained-output ratio was about 3.19 and specific thermal energy m2 area, a single-stage system produced 5.54 kg water in 7.83 h, while
consumption was around 200 kWhthm− 3. Ahmed et al. [16] put forward the two-stage system produced 8.66 kg water in 7.7 h. Abbaspour et al.
the concept of multi-stage vacuum solar distiller. The preliminary [29] studied the NVSD system with vacuum tube collector through ex­
experimental results showed that the maximum daily water output was periments. The vacuum in the evaporation chamber was created with a
about three times than the basin type solar distiller. vacuum pump at the system startup and maintained by natural vacuum
These desalination systems use vacuum pumps to maintain the vac­ during the day. The maximum water yield of the system in a day was
uum state in the evaporation chamber. However, vacuum pumps require 8.065 kg/m2. In addition, Choi [30] presented a NVSD system with a
a considerable amount of power and the mixture of lubricating oil in passive vacuum pipe based on the hydrostatic head. Theoretical calcu­
pump body with liquid water during the pumping process will reduce lations showed that the freshwater production rate was over 7% of the
the service life and heat dissipation capacity of the lubricating oil or supplied seawater when the height of a vacuum pipe was 9.8 m and the
even damage the vacuum pump. Therefore, using a vacuum pump is seawater temperature was heated to 80 ℃.
unsuitable for small systems and also causes high operation cost. In view Non condensable gases (NCG), such as nitrogen, oxygen, and carbon
of this, other scholars have proposed the use of a natural vacuum to dioxide, dissolved in seawater are released during the evaporation
achieve vacuum conditions in the evaporation chamber. The basic process. Although NCG account for only 1% of vapor, they can form a
principle of natural vacuum solar desalination (NVSD) is to use the at­ “gas barrier” between evaporation and condensation surfaces to reduce
mospheric pressure and gravity of the earth to form a low-pressure space the heat transfer rate seriously while increasing the pressure of the
above the 10.33 m-high water column, which can be used as an evap­ distiller and reducing the evaporation rate [31,32]. Studies typically use
oration chamber, and easily evaporate the feed seawater heated by solar a vacuum pump to extract NCG from the evaporator but require high
energy [17,18]. Ayhan and Madani [19] proved that the NVSD system power consumption and maintenance cost.
requires less electrical energy than a conventional desalination plant Therefore, an active NVSD system coupled with flat-plate solar col­
with similar capacity for freshwater production. Al-Kharabsheh and lectors was developed in this study to reduce the operation energy
Goswami [20] theoretically calculated effects of heat source tempera­ consumption and water production cost. Compared with previous
ture, water body depth, and withdrawal rate on the system performance. studies, the novel NVSD system demonstrates the following character­
Eames et al. [21] explored a solar-powered NVSD system through istics: (1) The system only needs a water pump to fill seawater into the
theoretical analysis and experimental testing in 2007. The researchers distiller for extruding NCG and then form the natural vacuum in the
reported that the system can provide 30 L of freshwater per day with an distiller after seawater draining. Thus, the desalination system requires
evaporation area of 4.727 m2. Setyawan et al. [22] introduced NVSD low energy consumption to recover the vacuum state. (2) The condenser
system coupling with a hybrid solar collector. The maximum water yield is installed inside the evaporation chamber to improve the compactness
of the entire day was only 220 ml due to the insufficient heat transfer of the system and increase the freshwater production effectively by
between the collector and the evaporator. Zhang et al. [23,24] presented decreasing the heat loss. (3) The cold seawater in the condenser can
a passive NVSD system based on open loop heat pipes that uses siphon recover the latent heat of vapor to improve water yield and efficiency
effect to evaporate the seawater continuously. The system can obtain a before heating via solar energy. (4) The system demonstrates wide­
water productivity of 20.5 L/h under a heat vapor temperature of 55 ◦ C. spread application and can be combined with not only solar energy but
A combined natural vacuum desalination and solar-assisted air- also other industrial waste heat because the low pressure condition in
conditioning system was presented in reference [25]. The heat rejected the distiller reduces the heat energy for evaporation.
by the condenser of a modified absorption refrigeration system was
evaluated as a possible source to drive the desalination process. The
combined system can produce 4.5 kg of freshwater per hour. The author

2
L. Wang et al. Applied Thermal Engineering 196 (2021) 117320

2. Description of the operation principle and structure 2.2. Experimental set-up

2.1. Working principle and system proposal The NVSD system presents strict requirements on the installation site
because the evaporation chamber must be higher than 10.33 m above
The diagram of the designed NVSD system is shown in Fig. 1. The the ground. The experimental prototype in this study was installed on
system includes flat-plate solar collectors, a condenser, a freshwater top of a three-story building in Nanning City (longitude: 108◦ and lati­
collection groove, two water pumps, a freshwater tank, and some valves. tude: 23◦ ) in China, as shown in Fig. 2. Ten flat-plate solar collectors
The experimental system exploits the “Torricelli” phenomenon to form a with copper wing plate are used to collect irradiation and purchased
low-pressure space to reduce the evaporation temperature of water in from Guangdong Five Star Solar Energy Company. The aperture area for
the distiller. The specific experimental process is presented as follows: each collector is 1.8 m2, and the installation angle relative to the ground
Open valve 4 and water pump switch to fill the 10 m-high distiller with is 20◦ . Light transmittance of the glass cover plate on the heat collector is
seawater and expel NCG in the evaporation chamber from the discharge 0.9, and the absorptivity of the selective absorption coating on the heat
valve at this time. Open valve 2 to drain the water when the seawater in absorption wing plate is 0.93. Each plate contains 12 fluid pipes, and a 3
the distiller is full. A vacuum state will be formed in the distiller after cm-thick glass fiber cotton is pasted on the back as the insulation layer.
draining. Open switches of the water pump and valves 1, 3, 5, and 6. The The distiller is a horizontal tank-shaped container with a volume of 215
cold seawater initially flows through the condenser for cooling by L and length of 2 m and its internal structure is shown in Fig. 3. The
recovering the latent heat of vapor condensation. The preheated exterior of the distiller is wrapped with insulation rock wool with a
seawater then flow through the solar collector for further heating and thickness of 3 cm to prevent heat loss during the experiment. The in­
finally enter the distiller. After condensation, the freshwater drops on ternal condenser is an 11 m-long stainless steel hose arranged evenly in
the freshwater groove, which is set to a definite inclination angle, to the distiller according to the layout method of serpentine coil. The
ensure that freshwater can passively flow into the freshwater storage condenser pipe extends from the inlet end after winding to ensure that
tank through the pipeline. only one head is available in the distiller for convenience in machining.
NCG dissolved in seawater will be released from the seawater along The opening angle of the freshwater collection groove is 130◦ , and four
with the water vapor after a period of working time. Due to the hin­ support plates are fixed at the bottom. The inclination angle of the
drance to the vapor diffusion process, the heat transfer coefficient be­ collection groove is set to 5◦ with respect to the horizontal plane to
tween the condenser and the water vapor is reduced. Therefore, refilling improve the collection of freshwater. A submersible water pump with a
the distiller with seawater is necessary for exhausting the NCG to head of 15 m and rated power of 60 W is selected to feed water.
recover the natural vacuum. The specific operation is presented as fol­
lows: Open valve 4 and discharge valve. Close other valves to ensure that 3. Heat and mass transfer analysis and performance evaluation
the seawater can gradually fill the evaporation chamber. Then open indexes
valve 2 to discharge the seawater again when NCG is discharged
completely to allow the evaporation chamber to recover the low pres­ It is assumed that the distiller with insulation layer is adiabatic, and
sure condition. the heat capacity loss of each component is ignored. The heat and mass
transfer relationship of each part is as follows:
The solar energy absorbed by flat-plat solar collectors is given as:

Fig. 1. Diagram of the NVSD system.

3
L. Wang et al. Applied Thermal Engineering 196 (2021) 117320

Fig. 2. Photograph of the experimental set-up with insulation.

Fig. 3. Perspective view of three dimensional structure of distiller.

Qs = ms cp,s (Th,in − Tc,out ) (1) where mf is the mass flow rate of freshwater, Tf,out is the freshwater
outlet temperature.
where ms is the mass flow rate of seawater, cp,w is specific heat of
Compared with the mass flow of the inlet hot seawater, the mass flow
seawater, Th,in is the hot seawater inlet temperature, Tc,out is the cold
of brine is lower, and the difference is equal to the freshwater yield.
seawater outlet temperature.
Therefore, the relevant mass transfer equilibrium equation can be
The efficiency of the flat-plat solar collector is given as [33]:
expressed as follows:
Qs
ηs = (2) mf = ms -mb (7)
As I
According to the relation of countercurrent heat transfer, the total
where As is the surface area of the flat plat solar collector, I is the heat transfer coefficient of the distiller is calculated as [26]:
radiation.
Qc
The seawater heated by the solar collector evaporates in the distiller, K= (8)
so the mass and heat are reduced continuously. Taking the distiller as the Ac Δtm
research object, according to the first law of thermodynamics, the
where Ac is the surface area of the condenser, Δtm is the logarithmic
reduced energy of hot seawater is equal to the sum of the increased
mean temperature difference.
energy of cold seawater and the reduced energy of freshwater outflow,
Solar desalination system performance basically depends on solar
which can be expressed as following:
radiation and operation condition. The performance or energy efficiency
Qh = Qc + Qf (3) of designed NVSD system is defined as the performance ratio (PR). A
larger value of PR in a desalination system means high energy efficiency
Among them, the reduced energy of hot seawater is given as:
and better distillate production performance. Based on the relationship
Qh = ms cp,s T h,in -mb cp,s T b,out (4) of the input and output energy, the PR is given as follows [34]:
Mf,tot hfg
where mb is the mass flow rate of brine, Tb,out is the hot seawater outlet PR = ⋅1000 (9)
As Itot + tp Pp
temperature.
The increased energy of cold seawater can be written as:
where Mf,tot is total freshwater production, hfg is latent heat of water, Itot
Qc = ms cp,s (Tc,out − Tc,in ) (5) is the total radiation absorbed by the collector in a day, tp is running time
of the water pump, Pp is energy consumed by water pump.
where Tc,out is the cold seawater outlet temperature. The recovery ratio (RR) is another important performance indicator
The reduced energy due to freshwater outflow is given as: of the NVSD system. RR is the ratio of the rate of freshwater production
to the rate of feed seawater which is shown in Eq. (10) [15].
Qf = mf cp,s T f,out (6)

4
L. Wang et al. Applied Thermal Engineering 196 (2021) 117320

RR =
mf,tot
(10) uncertainly of total water yield and PR are 1% and 6%, respectively.
ms,tot

where Ms,tot is the total mass of seawater flowing through the distiller in 4.2. Experiment of constant inlet seawater temperature with simulated
a day. heat source

4. Experimental results and discussion Maintaining the inlet seawater temperature at a constant value is
necessary to analyze the transient heat and mass transfer process inside
In the experiment, salt was dissolved in tap water to make up the the distiller accurately. An electric heating rod is used to replace the
solution with 3.5% salinity instead of seawater. The experiment includes solar energy as the heat source of the experiment in this section.
the following parts: (1) constant inlet seawater temperature of the
simulated heat source and (2) experiment under actual solar radiation. 4.2.1. Transient temperature and pressure
A complete working cycle includes water filling and discharging
4.1. Measurement system and uncertainty analysis process as well as seawater desalination process. The water filling and
discharging time lasts for approximately 20 min due to the limitation of
A TBQ-2 radiometer was used to measure the total irradiance during pipe diameter and pump power. Fig. 4 shows the pressure of the evap­
the experiment. K-type thermocouples were used to measure inlet hot oration chamber and variation of each point temperature when the inlet
seawater temperature Th,in, inlet cold seawater temperature Tc,in, outlet hot seawater temperature is 60 ◦ C and the seawater mass flow rate is
brine temperature Tb,out, outlet cold seawater temperature Tc,out, and 0.05 kg/s. The initial pressure in the distiller can reach around 8 kPa
outlet freshwater temperature Tf,out. All temperature statistics were after drainage and then gradually increase over time. Notably, the
transferred to the data acquisition instrument JLS-XMT. The internal pressure of the distiller consists of NCG and vapor pressures. NCG dis­
pressure of the distiller was measured using a pressure vacuum gauge. solved in seawater will be released continuously during the evaporation
An electronic scale was placed below the freshwater tank to weigh the process. In this case, an “NCG barrier” will build up around the
quality of freshwater. Feed seawater mass flow rate was measured by a condensation surface to present the evaporated vapor condensing in
turbine flow meter set in the seawater circulation loop. The measure­ time. The gathered vapor increases its partial pressure, and further leads
ment devices and corresponding accuracy are shown in Table 1. to the decrease of evaporating rate. In addition, it can be found from the
All sensors connected to the data logger were calibrated to determine figure that the brine temperature Tb,out increases rapidly in the first five
their sensitivity before measuring. During the test period, temperature minutes. After that, due to the decrease of evaporation, the latent heat
was automatically recorded every 1 min, while water production and taken away by vapor keeps decreasing, which leads to the nearly linear
pressure were manually recorded every 5 min. increase of temperature Tb,out. At the end of the cycle, Tb,out reaches
In this study, the uncertainty analysis of the measured experimental about 56.4 ◦ C. It is also worth to note that the cold seawater outlet
data was estimated by using the Holman method [35]. PR and fresh­ temperature Tc,out rise rapidly at the beginning of desalination. The
water yield are two main parameters in the experiment. According to Eq. reason is that the high concentration of water vapor promotes the
(9), PR is mainly related to freshwater yield and solar irradiance. condenser to absorb a large amount of condensation latent heat. Then
Therefore, the uncertainty value of PR is calculated by the following due to the increase of NCG, Tc,out decreases gradually. Moreover, the
equation: curve of the freshwater outlet temperature Tf,out is similar to that of Tc,
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ out, because the condensing temperature of vapor is mainly affected by
(
∂PR 2 (
)
)2
(
∂PR 2
) the temperature of the cooling pipe.
εPR = εmf,tot + (εI tot )2 (11)
∂mf,tot ∂I tot
4.2.2. Transient heat transfer coefficient and water production rate
The uncertainty value of cumulative freshwater yield and solar Curves of heat transfer coefficient K and freshwater yield rate
irradiance are expressed as follows: decrease nonlinearly with running time, as shown in Fig. 5. Each group
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ of measurement data presents a corresponding fitting curve. At the
n ( )
∑ ∂Mf ,tot 2 beginning of desalination, there is almost no NCG in distiller, so the heat
εMf,tot = (εf ,i )2 (12)
i=1
∂Mf ,i transfer coefficient K can reach a high value of 18 kW /m2◦ C, and the
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ water production rate can reach 2.13 kg/5min. However, the continuous
∑ n
∂Itot 2 release of NCG hinders the condensation and heat transfer, resulting in
εItot = ( ) (εIi )2 (13)
i=1
∂ I i

In the above equations, εf,i is the error of each measurement of water

yield, that is, the error of electronic scale. εIi is the error of each mea­
surement of radiometer, as shown in Table 1. After the calculation, the

Table 1
Specification of the different measuring devices.
Test instruments Accuracy Range Measuring Error
(%)

Radiometer/ TBQ-2 ±1 W/m2 0− 2000 W/ 5

m2
Pressure vacuum gauge ±0.1 kPa − 0.1 MPa ~ 0 1.6
K-type thermocouple / WRNT- ±0.1 ◦ C 0–300 ◦ C 0.5
01
Liquid turbine flow meter ±0.01 0.05–5 m3/h 1
m 3/h
Electronic scale ±0.01 kg 0–80 kg 1
32 channel digital data- ±0.1 ◦ C − 200–600 ◦ C 1
recording /JLS-XMT
Fig. 4. Variation of temperature and pressure with running time.

5
L. Wang et al. Applied Thermal Engineering 196 (2021) 117320

4.2.4. Effect of seawater flow rate on hourly yield

The effect of different seawater flowrates ms and hot seawater tem­
peratures Th,in on hourly yield are shown in Fig. 7. Each parameter was
recorded four times and the average value was taken as the test result. It
can be seen from the figure that the hourly yield increases with the in­
crease of Th,in and ms, the maximum value can reach 12.45 kg/h, and the
corresponding working condition is Th,in = 60 ◦ C and ms = 0.1 kg/s.
Although these results depend on the experimental conditions applied in
each study, the trend is justified. These results are also applicable to
small and medium-sized desalination systems that use waste heat such
as household air conditioners or diesel engines as heat sources.
Fig. 8 shows the freshwater salinity under different parameters. The
average salinity of freshwater was around 390 ppm, which is lower than
that of drinking water recommended by the World Health Organization
(≤1000 ppm) [36]. The freshwater quality did not change with oper­
ating conditions such as water temperature and flowrate. This is because
in the designed desalination system, evaporation area and condensation
Fig. 5. Variation of heat transfer coefficient K and yield rate with running time.
area are completely separated. Such an observation is the same as a
previous study [26]. The residual salt in desalinated water is mainly
the decrease of heat transfer coefficient and freshwater yield. The heat caused by the seawater passing through the freshwater groove during
transfer coefficient is 4 kW/m2 ◦ C and the water production rate is the water filling process.
reduced to 0.12 kg/5 min when the operation time is 95 min, thereby
indicating that the desalination ability of the system is significantly
4.3. Experimental research under actual weather
weak. Closing valve 2 on the brine pipeline and filling the distiller to
discharge the accumulated NCG and recover the natural vacuum are
To further study the water production performance of the system, the
necessary at this time.
instantaneous freshwater yield characteristics and seawater mass flow
rate on the system daily yield performance were carried out under real
4.2.3. Effect of operating temperature on cycle time and water production
weather condition.
A working cycle begins when the filling and exhausting of the
distiller starts and then ends when the yield rate is lower than 0.15 kg/5
4.3.1. Instantaneous freshwater yield characteristics under real weather
min. Fig. 6 shows the variation of accumulative water yield in an
condition
operation cycle of the system with running time under different tem­
An experiment under real weather condition was performed on
peratures of cold and hot seawater. The tested hot seawater temperature
September 30, 2019, which was a sunny and windless weather. The test
Th,in was in the range of 45 ◦ C–60 ◦ C. As shown in Fig. 6, a high tem­
time was from 9:00 to 17:30 and the ambient temperature was about
perature of hot seawater corresponds to a high yield of accumulated
34 ◦ C. Fig. 9(a) shows the measured solar irradiance and accumulated
water and a long cycle time. Notably, the water yield in one cycle is 1.01
solar irradiance during the test day. It is seen that the total solar intensity
kg and the corresponding cycle time is only 40 min when Th,in is 45 ◦ C,
reaches summit at about 13:00 with the peak value of 1001 W/m2, and
while the water yield in one cycle is 12.91 kg and the corresponding
the accumulated irradiance is 22.1 MW/m2 in the experiment.
cycle time is 95 min when Th,in is 60 ◦ C. This phenomenon is due to the
The distiller started filling with seawater at 9:00 a.m. After the
increased evaporation heat from the high seawater temperature.
natural vacuum was formed in the distiller, the feed seawater with a
Furthermore, higher temperature of input cold seawater Tc,in implies a
mass flow rate of 0.15 kg/s was supplied by the water pump. The
lower yield and a longer cycle time. This translates to lower temperature
desalination process was stopped when the water production rate was
difference of heat exchange between vapor and condenser.
less than 0.5 kg/5min, then the distiller was filled again to continue the
next cycle. The curve of the pressure in the distiller during the desali­
nation process is illustrated in Fig. 9(b). The figure shows that the

Fig. 6. Variation of accumulated water yield with running time for various Th,in
and Tc,in. Fig. 7. Variation of the hourly yield under the different Th,in and ms.

6
L. Wang et al. Applied Thermal Engineering 196 (2021) 117320

seawater flow rate is 0.15 kg/s on September 30. The freshwater pro­
duction rate of the system remarkably increases as the mass flow rate of
seawater increases. On the one hand, increasing the mass flow rate al­
lows the seawater to obtain additional solar energy absorbed by the
collector and result in increased evaporation. On the other hand,
improving the seawater mass flow rate can increase the temperature
difference between the vapor and condenser. Furthermore, the results
demonstrate that the number of work cycles increases with the increase
of mass flow rates. During the 3 test days, the numbers of work cycles are
5, 7 and 11 respectively. Because the increasing evaporation rate im­
proves the content of NCG in the distiller, which leads to less time per
cycle.
Table 2 lists the performance evaluation indexes of the NVSD system
in different seawater flow rates. The results demonstrate that RR de­
creases with the increase of seawater mass flow rate ms. The RR of the
system can reach a maximum value of 10.9 when the seawater flow rate
is 0.05 kg/s. The low seawater flow rate can prolong the evaporation
time of hot seawater when flowing in the distiller to ensure that the unit
Fig. 8. Salinity of the freshwater under the different Th,in and ms. mass of seawater can evaporate more vapor and lead to high RR. In
addition, the recovery process of latent heat is enhanced when the
pressure changes periodically, because the system needs periodic water seawater mass flow rate increases, results in a higher PR. According to
filling to discharge NCG. As shown in Fig. 9(b), the system demonstrates Formula (9), the maximum PR can reach 1.36. If multistage distillers are
11 working cycles in a day, and the absolute pressure in the distiller can used to cascade the sensible heat carried by brine, then PR will be
reach a maximum value of 14.54 kPa. improved further.
Fig. 9(c) presents the temperature change that corresponds to the
pressure in Fig. 9(b). The temperature of inlet hot seawater Th,in is 5. Economic analysis of the system
clearly high at the beginning of the work cycle and then decreases
gradually, with the maximum temperature reaching 87.2 ◦ C during the The economy of the solar desalination device is one of the most vital
test. This phenomenon is primarily caused by the absence of seawater factors which determines whether the device could be commercialized
flow in flat-plate solar collectors in the stage of seawater filling and and widely applied in the market or not. So, it is of great significance to
discharging. The heat absorbing plate in the collector will absorb the analyze the economy of the device. Here, we refer to the method pro­
solar radiation and enable seawater temperature to rise continuously. posed in reference [37], as expressed in Eq. (14) and (15), the total costs
When the system begins to desalinate, the feed seawater will continue to (TC) consist of fixed costs (FC) and variable costs (VC). FC refers to the
absorb the solar heat accumulated in the collectors and cause the cost of purchasing each kind of equipment and installation cost, while
gradual decrease of the Th,in. Notably, the temperature of the outlet VC refers to the maintenance cost caused by system loss. TC and VC are
brine Tb,out gradually increases owing to the reduction of evaporation expressed as follows:
rate. TC = FC + VC (14)
Fig. 9(d) shows the efficiency of the flat-plate solar collectors under
different working cycles in a day. The efficiency of the solar collector is VC = j × y × FC (15)
low in the forenoon due to thermal inertia of the heat absorption plate.
The test results show that the daily average efficiency of the entire day where j is annual loss factor; Y is the service life of the device.
can reach 0.68. Table 3 lists the construction cost for each of the system components,
The freshwater yield rate of the NVSD system is depicted in Fig. 9(e). the total FC of the entire system is $2210 after verifying each part of the
The comparison of Fig. 9(e) and 9(c) demonstrated that the change cost of the apparatus. Since the system uses fewer components and no
tendency of freshwater yield rate under real weather condition is similar high-precision components, we assumed that the j = 0.08.
to that of the Th,in. The maximum water yield rate during the experiment If we conservatively assume that the service life of the system is Y =
reached 5.28 kg/5 min. In addition, the freshwater yield of the sixth 15 years, then TC is $4862. Nanning can present 300 sunny days each
cycle is the maximum, with a total yield of 24.3 kg in 55 min. This year [38]; hence, the total amount of freshwater that the system can
finding is due to the occurrence of the maximum solar irradiance in this produce is 697500 kg during a service life of 15 years. The calculated
period. average production cost of freshwater is 0.0069 $/kg for the Chinese
Fig. 9(f) illustrates the PR of each working cycle. It is apparent that market.
PR is lower under stronger solar radiation. This phenomenon is mainly Compared with the method of using vacuum pump to achieve vac­
due to higher temperature of brine outflowing from the distiller under uum in the evaporation chamber, the natural vacuum method is more
higher heating power. In other words, the hot seawater flowing out of low-carbon and energy-saving. Due to the mature process and fairly
the distiller without effective evaporation leads to the waste of solar heat cheap of the flat-plat solar collector, using solar energy as the driving
at noon stage. energy has a lower cost. When the system works, it requires lower en­
ergy consumption and maintenance costs. Because of these character­
4.3.2. Influence of seawater mass flow rate on freshwater production istics, the system is improved in competitiveness and becomes
The effect of different seawater mass flow rates on the system environmentally friendly. In addition, compared with other vacuum
freshwater yield was investigated from the 28th to 30th of September, desalination systems in Table 4, the system in this study demonstrates
and their daily cumulative water yield curves are presented in Fig. 10. better water production performance. This may be explained that the
The average solar irradiances for the three test days are 693, 682, and designed NVSD system with an internal condenser can reduce the heat
672 W/m2, respectively. The results show that the daily water produc­ loss and enhance the mass transfer of the vapor.
tion is 112.21 kg when the seawater flow rate is 0.05 kg/s on September
28. However, the daily water production can reach 154.4 kg when the

7
L. Wang et al. Applied Thermal Engineering 196 (2021) 117320

Fig. 9. (a) Solar irradiance change versus local time. (b) Curves of the internal pressure of the distiller. (c) Curves of temperature measuring points. (d) Efficiency of
the solar collector at different cycles. (e) Curves of the freshwater yield rate. (f) PR of the NVSD system at different cycles.

6. Conclusions When the average of daily irradiance is 672 W/m2 and ms is 0.15 kg/s,
the water production per unit collector could reach 8.61 kg/m2⋅day, and
A novel NVSD system that contains an internal condenser is inves­ the corresponding RR and PR are 8.3% and 1.36, respectively. Finally,
tigated in this study. The system uses water filling and emptying process the economic analysis based on the designed prototype showed that the
to remove NCG and recover the vacuum state. The NVSD system dem­ water production cost is 6.9 $/ton, thereby indicating a broader market
onstrates a high water yield rate because the latent heat of vapor can be prospects than other vacuum desalination system.
recovered through the internal condenser during desalination. An In the future research, the brine removed NCG can be collected and
experimental prototype is built and the effect of key parameters on desalinated again to simplify the operation process of NVSD system. In
system performance is investigated. addition, a multi-stage NVSD system that can use the sensible heat of
Under the fixed feed seawater temperature, the system can produce brine in cascade will be explored to improve the freshwater yield and PR
12.91 kg freshwater in 95 min when Th,in = 60 ◦ C and ms = 0.05 kg/s. In further.
the actual weather, the experimental results proved that the freshwater
yield and PR could be improved significantly with the increase of ms.

8
L. Wang et al. Applied Thermal Engineering 196 (2021) 117320

Declaration of Competing Interest

The authors declare that they have no known competing financial

interests or personal relationships that could have appeared to influence
the work reported in this paper.

Acknowledgements

This work was supported by the National Natural Science Foundation

of China (No. 51976013), and Beijing Natural Science Foundation
(No.3192031).

References:

[1] A.B. Pouyfaucon, L. García-Rodríguez, Solar thermal-powered desalination: A

viable solution for a potential market, DESALINATION 435 (2018) 60–69.
[2] N. Zehtabiyan-Rezaie, N. Alvandifar, F. Saffaraval, M. Makkiabadi, N. Rahmati,
M. Saffar-Avval, A solar-powered solution for water shortage problem in arid and
semi-arid regions in coastal countries, Sustainable Energy Technologies and
Fig. 10. Variation of the accumulated yield under the different seawater mass Assessments 35 (2019) 1–11.
flow rate. [3] S.W. Sharshir, N. Yang, G. Peng, A.E. Kabeel, Factors affecting solar stills
productivity and improvement techniques: A detailed review, APPL. THERM. ENG.
100 (2016) 267–284.
[4] J. Hou, J. Yang, Z. Chang, H. Zheng, Y. Su, The mass transfer coefficient assessment
Table 2
and productivity enhancement of a vertical tubular solar brackish water still, APPL.
The system performance under different seawater flow rate. THERM. ENG. 128 (2018) 1446–1455.
Date September September September [5] M.M. Ghafurian, H. Niazmand, E. Ebrahimnia-Bajestan, R.A. Taylor, Wood surface
treatment techniques for enhanced solar steam generation, RENEW. ENERG. 146
28 29 30
(2020) 2308–2315.
ms (kg/s) 0.05 0.1 0.15 [6] A. Rafiei, R. Loni, S.B. Mahadzir, G. Najafi, S. Pavlovic, E. Bellos, Solar desalination
Average solar irradiance (W/m2) 693 682 672 system with a focal point concentrator using different nanofluids, APPL. THERM.
Number of cycles 5 7 11 ENG. 174 (2020), 115058.
Total yield (kg) 112.21 130.4 154.14 [7] B. Hou, D. Kong, Z. Chen, Z. Shi, H. Cheng, D.D. Guo, X. Wang, Flexible graphene
Yield per unit collector area (kg/ 6.23 7.24 8.61 oxide/mixed cellulose ester films for electricity generation and solar desalination,
APPL. THERM. ENG. 163 (2019), 114322.
m2⋅day)
[8] C. Li, D. Jiang, B. Huo, M. Ding, C. Huang, D. Jia, H. Li, C. Liu, J. Liu, Scalable and
RR (%) 10.9 8.98 8.3
robust bilayer polymer foams for highly efficient and stable solar desalination,
PR 0.83 1.08 1.36
NANO ENERGY 60 (2019) 841–849.
[9] J. Xu, F. Xu, M. Qian, Z. Li, P. Sun, Z. Hong, F. Huang, Copper nanodot-embedded
graphene urchins of nearly full-spectrum solar absorption and extraordinary solar
desalination, NANO ENERGY 53 (2018) 425–431.
Table 3 [10] Y. Xu, J. Ma, Y. Han, H. Xu, Y. Wang, D. Qi, W. Wang, A simple and universal
Fixed costs related to the proposed system. strategy to deposit Ag/polypyrrole on various substrates for enhanced interfacial
solar evaporation and antibacterial activity, CHEM. ENG. J. 384 (2020), 123379.
Components Fixed costs/$
[11] H. Al-Hussaini, I.K. Smith, Enhancing of solar still productivity using vacuum
Distiller shell 560 technology, ENERG. CONVERS. MANAGE.36.11 (1995) 1047–1051.
Serpentine cooling pipe 320 [12] G. Xie, L. Sun, T. Yan, J. Tang, J. Bao, M. Du, Model development and experimental
Freshwater groove 60 verification for tubular solar still operating under vacuum condition, ENERGY 157
(2018) 115–130.
Plate solar collector 10 × 101
[13] M.M. Morad, H.A.M. El-Maghawry, K.I. Wasfy, A developed solar-powered
Water pump 15
desalination system for enhancing freshwater productivity, SOL. ENERGY 146
Valves and pipelines 35 (2017) 20–29.
Installation cost 210 [14] G.J.J. Wessley, P.K. Mathews, Investigations on solar desalination based on
Total costs 2210 vacuum evaporation for small scale applications[J], Procedia Eng 38 (2012)
2413–2419.
[15] J.A. Andrés-Mañas, A. Ruiz-Aguirre, F.G. Acién, G. Zaragoza, Assessment of a pilot
system for seawater desalination based on vacuum multi-effect membrane
Table 4 distillation with enhanced heat recovery, DESALINATION 443 (2018) 110–121.
Comparison of the water production and cost in different vacuum desalination [16] M.I. Ahmed, M. Hrairi, A.F. Ismail, On the characteristics of multistage evacuated
systems. solar distillation, RENEW. ENERG. 34 (2009) 1471–1478.
[17] S. Al-Kharabsheh, D. Yogi Goswami, Analysis of an innovative water desalination
Literature Method of Concentrator Daily yield per Water system using low-grade solar heat, DESALINATION 156 (2003) 323–332.
making type area (kg/ cost [18] S. Al-Kharabsheh, D.Y. Goswami, Experimental study of an innovative solar water
vacuum m2⋅day) ($/ton) desalination system utilizing a passive vacuum technique, SOL. ENERGY 75 (2003)
395–401.
Morad [13] Vacuum pump Flat-plat 5.38–8.1 35–37 [19] T. Ayhan, H. Al Madani, Feasibilty study of renewable energy powered seawater
Gude [26] Natural Flat-plat 11.11 6.7 desalination technology using natural vacuum technique, RENEW. ENERG. 35
vacuum (2010) 506–514.
Abbaspour Natural Evacuated 8.07 94 [20] S. Al-Kharabsheh, D.Y. Goswami, 2004. Theoretical analysis of a water
[29] vacuum and tubular desalination system using low grade solar heat. J. Sol. Energy Eng, 126(2004) 774-
vacuum pump 780. (https://doi.org/10.1115/1.1669450).
Bilgil [39] Vacuum pump / / 32 [21] I.W. Eames, G.G. Maidment, A.K. Lalzad, A theoretical and experimental
Gude [40] Natural Flat-plat 5.56 11.7 investigation of a small-scale solar-powered barometric desalination system,
APPLIED THERMAL ENGINEERING 27 (2007) 1951–1959.
vacuum
[22] E.Y. Setyawan, R.A.M. Napitupulu, P. Siagian, H. Ambarita, Field tests of a natural
Current Natural Flat-plat 8.61 6.9
vacuum solar desalination system using hybrid solar collector, IOP conference
study vacuum
series, Materials Science and Engineering 237 (2017) 12012.
[23] X. Zhang, W. Kan, H. Jiang, Y. Chen, T. Cheng, H. Jiang, X. Hu, Capillary-driven
low grade heat desalination, DESALINATION 410 (2017) 10–18.
[24] X. Zhang, H. Wang, H. Jiang, C. Li, X. Hu, Open loop heat pipes for high-efficiency
desalination plant, APPLIED THERMAL ENGINEERING 193 (2021), 117027.

9
L. Wang et al. Applied Thermal Engineering 196 (2021) 117320

[25] V.G. Gude, N. Nirmalakhandan, Combined desalination and solar-assisted air- [33] C. Kutlu, J. Li, Y. Su, Y. Wang, G. Pei, S. Riffat, Investigation of an innovative PV/T-
conditioning system, ENERG. CONVERS. MANAGE. 49 (2008) 3326–3330. ORC system using amorphous silicon cells and evacuated flat plate solar collectors,
[26] V.G. Gude, N. Nirmalakhandan, S. Deng, A. Maganti, Feasibility study of a new ENERGY 203 (2020), 117873.
two-stage low temperature desalination process, ENERG. CONVERS. MANAGE. 56 [34] W. Gang, Y. Qichang, Z. Hongfei, Z. Yi, F. Hui, J. Rihui, Direct utilization of solar
(2012) 192–198. linear Fresnel reflector on multi-effect eccentric horizontal tubular still with falling
[27] H. Ambarita, Study on the performance of natural vacuum desalination system film, ENERGY 170 (2019) 170–184.
using low grade heat source, Case Stud. Therm. Eng. 8 (2016) 346–358. [35] Z. Rahimi-Ahar, M.S. Hatamipour, Y. Ghalavand, A. Palizvan, Comprehensive
[28] S.C. Maroo, D.Y. Goswami, Theoretical analysis of a single-stage and two-stage study on vacuum humidification-dehumidification (VHDH) desalination, APPL.
solar driven flash desalination system based on passive vacuum generation, THERM. ENG. 169 (2020), 114944.
DESALINATION 249 (2009) 635–646. [36] W.H. Organization, Safe drinking-water from desalination, Environ. Sci. Technol.
[29] M.J. Abbaspour, M. Faegh, M.B. Shafii, Experimental examination of a natural 27 (2011) 2295–2297.
vacuum desalination system integrated with evacuated tube collectors, [37] Y. Zheng, K.B. Hatzell, Technoeconomic analysis of solar thermal desalination,
DESALINATION 467 (2019) 79–85. DESALINATION 474 (2020), 114168.
[30] S. Choi, Thermal type seawater desalination with barometric vacuum and solar [38] Z. Zhu, H. Zheng, Q. Wang, M. Chen, Z. Li, B. Zhang, The study of a novel light
energy, ENERGY 141 (2017) 1332–1349. concentration and direct heating solar distillation device embedded underground,
[31] H. Glade, A.E. Al-Rawajfeh, Modeling of CO2 release and the carbonate system in DESALINATION 447 (2018) 102–119.
multiple-effect distillers, DESALINATION 222 (2008) 605–625. [39] A. Bilgil, B. Hırlakoğlu, An experimental study on desalination at vacuum
[32] A.E. Al-Rawajfeh, H. Glade, J. Ulrich, CO2 release in multiple-effect distillers environment under low pressure and low condensation temperatures,
controlled by mass transfer with chemical reaction, DESALINATION 156 (2003) DESALINATION 411 (2017) 9–18.
109–123. [40] V.G. Gude, N. Nirmalakhandan, S. Deng, A. Maganti, Low temperature desalination
using solar collectors augmented by thermal energy storage, APPL. ENERG. 91
(2012) 466–474.

10