M. Keshtkar, et al.

Fig. 18. Productivity of the basin solar still at different cover angles: (a) Hourly (b) Cumulative.

(29)
(32)
where is the ambient temperature (K).

2.3.3. Species boundary conditions

(33)

• The air is assumed to be saturated with water vapor (φ = 100%) on • Zero diffusive flux boundary condition is used for modeling im-
the evaporating and condensing surfaces (Rahbar and Esfahani,
2013). Accordingly, the following equations are used to calculate permeable walls.
the mass fraction of water vapor on these surfaces (Kreider et al.,
2009). (34)

(30)
(35)

(31) The initial temperature and velocity in the passive basin solar still
are set to 302 K and zero, respectively.
M. Keshtkar, et al.

Fig. 19. Temperature distribution in stepped solar stills at different water surface to glass cover distances at 3 PM.

Solar irradiance and ambient temperature shown in Fig. 3 are ap-

plied on the outer surface of the glass due to available experimental (40)
data (Feilizadeh et al., 2017).
Having defined all of the governing equations with the appropriate
boundary conditions and thermophysical properties, the proposed
2.4. Thermophysical and optical properties procedure for numerical solution of this problem is discussed in the
following section.
Molecular diffusion coefficient of water vapor and density of air and
water vapor are computed as functions of temperature for each cell by
Eqs. (8) and (36), respectively. Other properties are assumed to be 3. Solution procedure
constant as shown in Tables 2 and 3. Then the humid air properties are
obtained through Eqs. (37)–(40). All of the governing equations with their relevant boundary con-
ditions described in Section 2 are solved simultaneously in the present
method. Fig. 4 shows the flowchart of user defined functions (UDF) that
(36) are employed in order to implement the sink/source term in the energy
equation considering the instantaneous boundary conditions. At each
time step, instantaneous weather data including total solar irradiance
(37) and ambient temperature are applied as the only needed information.
Then, the molecular diffusion coefficient of water vapor into air and
density of air-water vapor mixture are calculated as functions of tem-
(38) perature by Eqs. (8) and (36). Simultaneously, mass fraction of water
vapor on the evaporation and condensation surfaces are computed ac-
cording to temperatures of evaporation and condensation surfaces
(39) through Eqs. (30)–(33). Next, species equation of water vapor (Eq. (11))
M. Keshtkar, et al.

Fig. 20. Productivity of stepped solar still at different water surface to glass cover distances: (a) Hourly (b) Cumulative.

is solved with its corresponding boundary conditions on the water and 29,673, 44,400 and 70,803, no significant difference is observed be-
glass surfaces. As a result, distribution of water vapor mass fraction in tween 44,400 and 70,803 grids, and therefore the mesh with 44,400
the solar still is obtained. After that, by calculating the gradient of grids is selected. Furthermore, a comparison of daily productivity for
water vapor mass fraction on the water and glass surfaces and in- time step sizes of 5, 10, and 20 s revealed that there is a negligible
tegrating Fick’s law on these surfaces, the total evaporation and con- difference between 5 and 10 s time steps and 10 s time step is chosen
densation rates are obtained through Eqs. (6) and (7). Each time, the consequently. Fig. 7 shows a schematic of the generated mesh used in
condensate water is reported as output and the amount of evaporated the present modeling.
water is deduced from the basin water. Afterwards, the latent heats are SIMPLE algorithm is used for pressure-velocity coupling. Besides,
calculated using evaporation and condensation rates and enthalpies of PRESTO scheme is suitable for rotational currents and used for pressure
evaporation and condensation at their relevant temperatures based on interpolations due to the natural convection vortices in the solar still.
Eq. (5). Finally, these latent heats are added to the energy equation as The momentum, energy, and species equations are discretized by
sink and source terms and temperature distribution in the solar still is second-order upwind scheme (ANSYS Fluent Documentation, 2016).
calculated. All of the mentioned processes are repeated every time step. The described procedure is now applied to the defined passive solar
The simulations are carried out in ANSYS Fluent 17.1 with finite basin and results are presented in the next section. Also, effects of
volume method. Grid and time step studies are performed to obtain different design and ambient parameters are going to be discussed.
independent results from grid and time step sizes as shown in Figs. 5
and 6. By comparing the daily productivity for grids numbers of 9500,
M. Keshtkar, et al.

Fig. 21. Temperature distribution in stepped solar stills at different number of steps at 3 PM.

4. Results and discussion surface to the ambient. Otherwise, increasing the wind velocity results
in a higher convection heat transfer coefficient. Therefore, the heat loss
4.1. Validation from the glass cover increases and the glass temperature decreases. The
higher temperature difference between water vapor and glass cover
Hourly productivity of the basin solar still is obtained by integrating leads to more productivity.
the instantaneous rate of condensation on the inner surface of glass for Fig. 10 presents the hourly and cumulative productivities of the
an hour and compared with the available experimental data of basin solar still at different wind velocities. It is revealed that total
Feilizadeh et al. (2017). As shown in Fig. 8a there is a very good productivity increases as the wind velocity increases from 1 m/s to 9 m/
agreement between the proposed numerical model and the experi- s. However, productivity does not change significantly from 6 m/s to
mental data. Besides, the cumulative productivity is computed by 9 m/s. It should be noted that productivity and wind velocity are
adding the hourly productivity of each hour to the previous accumu- proportional during the day but they are inversely proportional during
lated distilled water. The cumulative productivity is also compared with the night according to Fig. 10a.
experimental data of Feilizadeh et al. (2017) in Fig. 8b that verifies the
high agreement between present numerical solution and the experi-
mental measurements. The total yield obtained by the current numer- 4.3. Effect of water depth
ical modeling is 5.652 kg/m2 which differs about 0.73% from the ex-
perimental data. The present validated numerical method in the passive By maintaining the water surface to glass cover distance fixed at
basin solar still is now used for the parametric study in the following 8 cm and examining different water depths, it is found that the depth of
sub-sections. 2 cm results in the best performance. If the water depth increases, more
water under the same incoming solar irradiance should be heated and
4.2. Effect of wind velocity the water temperature and the production rate drop accordingly. The
cooling of the larger volume of water is also delayed which leads to an
Fig. 9 shows the temperature, velocity and water vapor mass frac- increase in the night production. On the other hand, by decreasing the
tion distributions at different wind velocities at 3 PM. The wind velocity water depth, the basin water warms up earlier due to its lower volu-
is assumed to be constant during the day for this study. As a result, metric heat capacity and increases the daily production. However, the
temperature in the solar still and water vapor mass fraction decrease by water cools earlier by decreasing the incoming solar irradiance in the
increasing velocity from 1 m/s. The temperature difference between afternoon and the production decreases. Fig. 11 shows that maximum
water and glass is lower in the case of V = 1 m/s and lower velocities water temperature occurs in the case of 2 cm water depth at 3 PM
are observed in the solar still consequently. In lower wind speeds, the which coincides with the maximum productivity in Fig. 12.
temperature rises in the solar still due to lower heat loss from the glass
M. Keshtkar, et al.

Fig. 22. Productivity of stepped solar still at different numbers of steps: (a) Hourly (b) Cumulative.

4.4. Effect of water surface to glass cover distance 4.6. Effect of cover angle

By maintaining the water depth at 2 cm and reducing the water to Different glass cover angles are investigated and the contours are
cover distance (WCD), the water surface becomes cooler due to the compared in Fig. 17. When the cover angle is 20°, two vortices are
smaller distance to the outside environment, resulting in lower pro- created in the still, but the second vortex disappears by increasing the
duction. According to Fig. 13, by increasing WCD, a larger vortex is angle. The temperature and the mass fraction are higher at 20°. The
created in the air-water vapor gap which results in lower heat transfer productivity at these three cover angles is compared in Fig. 18. Because
in the still and reduces the productivity. Therefore, there is an optimum of the higher temperature difference at 30°, the productivity is higher.
length in between where the productivity is the highest. Fig. 14 shows This result is again consistent with the experiment of Feilizadeh et al.
that the value of WCD = 8 cm is an appropriate choice which is in (2017) in which the same angle of 30° (which is equivalent to the la-
agreement with the optimization performed by Feilizadeh et al. (2017). titude of Shiraz) is recommended.

4.7. Stepped solar stills

4.5. Effect of glass thickness
Generally, the horizontal water surface receives less solar irradiance
Three different glass thicknesses are compared in Fig. 15. In thinner than a tilted plane. Due to the difficulty of keeping the water at a
glass covers, less radiation is absorbed by the glass cover and it is certain angle, the stepped solar still is a good alternative (Bundschuh
cooled more easily by the outside environment. The production with a and Hoinkis, 2012). In this section, the effect of converting a basin solar
4 mm glass cover can be increased by 3.5% if the thickness is reduced to still to a stepped solar still is investigated.
2 mm. According to Fig. 16, although the productivity increases at The volume of water considered in this type of solar still is equal to
lower glass thicknesses, this effect is not significant. the basin solar still and depth of water at each step is 2 cm. Initially, by
considering 5 equal steps, different distances from water level to glass
M. Keshtkar, et al.

Fig. 23. Contours of (a) Temperature (b) Water vapor mass fraction (c) Velocity (d) Velocity vector in six-stepped solar still at 3 PM.

cover are investigated to determine the optimum distance. Then, at the which allows the water vapor to condense on the colder inner surface of
optimum distance of water surface to glass cover, different number of glass. In the middle of the vortices, the properties are almost uniformly
steps are examined to obtain the optimal steps. In all of the studied distributed while a sudden change is observed on the water and glass
cases, the width and height of steps are taken in such a way that the surfaces. The sudden changes imply the evaporation and condensation
steps are parallel to the glass cover with a 30° angle. Due to the lower on these surfaces which are in agreement with the available results
distance between the water surface and the glass, the corresponding (Rashidi et al., 2017b).
Rayleigh number is low and therefore the flow is laminar as mentioned Productivity of the six-stepped solar still with WCD of 2 cm is
in Section 2. compared with the productivity of single basin solar still in Fig. 24.
After 10 PM, the productivities are almost the same and both of them
4.7.1. Effect of water surface to glass cover distance are approaching to zero. The total yield of the improved stepped solar
In Fig. 19, temperature contours of five-stepped solar still at dif- still is 6.633 kg/m2 which is about 17.4% higher than the single basin
ferent water surface to glass cover distances are shown. At 2 cm dis- solar still.
tance, a higher temperature in the solar still is observed. The pro- In this section, the numerical model was first validated for the
ductivity at these different distances is compared in Fig. 20. passive basin solar still. Afterwards, the numerical method was used to
Accordingly, the optimal distance is 2 cm, with the highest output. At investigate different parameters in order to enhance the performance of
lower distances, the water in the steps is near the glass and cools more solar still. The important conclusions of this study are highlighted in the
easily. next section.

4.7.2. Effect of number of steps 5. Conclusion

Temperature contours in stepped solar stills with different number
of steps are shown in Fig. 21 in case of WCD = 2 cm. The productivity The transient performance of solar stills is investigated through a
for various number of steps is also compared in Fig. 22. It can be general CFD modeling in the present study. As a result, the temperature,
concluded that the optimal number of steps is six with the most fresh velocity and water vapor mass fraction distributions are obtained. The
water production. absorption of solar irradiance in different components of solar still and
Contours of temperature, water vapor mass fraction, velocity, and radiative heat transfer between surfaces are considered by solving DO
velocity vector in the six-stepped solar still at 3 PM are presented in radiation equation. Instead of using time-consuming multi-phase
Fig. 23. At each step, two vortices are created that enhance the heat methods, latent heats are considered by coupling of energy and species
transfer between evaporation and condensation surfaces in comparison equations. By solving the species transport equation of water vapor and
to one vortex according to the research of Edalatpour et al. (2015). computing evaporation and condensation rates, latent heats of eva-
These vortices are formed by the buoyancy due to the temperature poration and condensation are applied to energy equation through user
difference between water surface and glass cover. The observed mass defined functions. Effects of climatic and design parameters on the
fraction difference is due to the corresponding temperature difference productivity of a single slope passive basin solar still are investigated
M. Keshtkar, et al.

Fig. 24. Comparison of productivity of the basin and six-stepped solar stills: (a) Hourly (b) Cumulative.

according to the instantaneous ambient conditions in a typical day. The • Productivity is higher at lower glass thicknesses. The reason is the
following conclusions are obtained by analyzing the results for a single colder glass temperature due to the increment in heat transfer with
slope passive basin solar still: the outside environment and decrease of radiation absorption in the
glass. The productivity in a system with 2 mm cover thickness is
• Higher wind speeds reduce the glass temperature by increasing the about 3.5% higher compared to the reference system with a 4 mm
heat transfer between the glass surface and the outside environment. thick glass cover.
The resulting increase in temperature difference between the water • Different cover angles are studied and it is revealed that cover angle
surface and the glass cover improves the productivity. Increasing the of 30° which is equivalent to the latitude in Shiraz has the highest
wind velocity from 1 m/s to 6 m/s, enhances the productivity about productivity. Altering this angle to 20° and 40° decreases the output
14.4%. But beyond the wind speed of 6 m/s, further increase in about 9% and 23%, respectively.
wind speed does not have a significant effect and there is only a 2% • The stepped solar still survey has been carried out in two sections.
improvement in the output. The optimum distance between the surface of the water and the
• Decreasing the water depth from 2 cm, increases the daily produc- glass and then the optimum number of steps are obtained. According
tion while decreases the output at night. On the other hand, in- to the results, the productivity of the improved stepped solar still
creasing the water depth decreases the daily production and in- with six steps and WCD of 2 cm increases about 17.4% compared to
creases night productivity. Overall, the depth of 2 cm is the most the basin solar still.
appropriate choice for the maximum productivity and its output is • Glass cover thickness has the least effect on the solar still pro-
15.8% higher than water depths of 1 and 4 cm. ductivity among the investigated parameters.
• Different water surface to glass cover distances are examined. It is • The highest productivity with the same basin area and an equal
found that the distance of 8 cm has the highest yield among the mass of water in the system is achieved by converting the solar still
investigated distances. Decreasing this distance to 4 cm and in- to an improved stepped solar still with six steps and WCD of 2 cm.
creasing it to 12 cm, decreases the output about 2.7% and 13.5%,
respectively.