Thermal Science and Engineering Progress 20 (2020) 100710

Contents lists available at ScienceDirect

Thermal Science and Engineering Progress

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

Investigation of steam jet flash evaporation with solar thermal collectors in T

water desalination systems
Akram W. Ezzata, , Eric Hub, Hussein M. Taqi Al-Najjarc, Zihui Zhaob, Xin Shub
⁎

a
Department of Mechanical Engineering, University of Baghdad, Iraq
b
School of Mechanical Engineering, The University of Adelaide, Adelaide, SA 5005, Australia
c
Department of Energy Engineering, University of Baghdad, Iraq

ARTICLE INFO ABSTRACT

Keywords: Experimental and theoretical studies were carried out to investigate the depressurization induced by steam jet
Flash evaporation using subsonic steam ejector with flat-plate solar thermal collector. Saline water flash evaporation could be
Desalination system realized by such depressurization in water desalination systems. The most critical component in such systems is
Subsonic ejector the steam ejector nozzle where the Mach number within the ejector is highly influenced by its geometry. The
Solar collector
main objective of the research is to evaluate the effect of different operating parameters on evaporation per-
Entrainment ratio
Experimental validation
formance which can be specified by two factors; first is the subsonic ejector efficiency governed by the steam
entrainment ratio, and second is the percentage gain of distilled water productivity. That goal served an in-
novative drive for the present work. The experimental test rig was designed and constructed with primary steam
pressure (1.25–2.5) bar and temperature (106–127) °C using a controlled boiler, while condenser pressure
ranged (0.974–1.0) bar. It was found that ejector efficiency increased up to 53% however the efficiency is
saturated beyond primary steam pressure of 2.0 bar due to the sonic velocity limitation. Also, it was noticed that
the percentage gain of distilled water productivity using steam ejectors with respect to that using conventional
evaporation ranged between 1.0%–5.5%. Based on current collector design considerations implemented in the
mathematical model, about 34% of the total thermal energy required for water heating and evaporation during
the desalination process can be covered by the solar collector. The error analysis for experimental validation
indicated an average value of 22.9%. Finally, it was concluded that the present research could be considered as a
good basis for further investigations of solar steam jet flash evaporation in desalination systems.

1. Introduction Fig. 1 shows the existing desalination methods according to the type
of energy used. Conventional energy sources such as non-thermal and
Freshwater became one of the world's biggest needs due to the in- electricity are being used to operate high capacity thermal desalination
crease in population. Water desalination contributed effectively in processes. The most technically applicable methods in water desalina-
solving this problem. Flash evaporation of seawater is one of the well- tion are RO, MVC, MSF, MED, TVC and ED/EDR [2]. However, the
known technologies which are greatly improved recently. Solar heating shortcomings of conventional desalination technologies are mainly laid
of saline water using Sun heat absorbed by solar thermal collectors is in the complexity of treatment process and the high energy consump-
one of the efficient ways. Normally, direct and indirect desalination tion. The combination of different technologies can integrate their ex-
processes are related to the energy transferred to freshwater. Direct pected advantages to decrease the energy consumption, the cost and to
solar desalination uses solar energy in solar stills, while indirect solar improve the desalination performance. In fact, analysis of energy con-
desalination systems utilize solar energy via solar collectors. Two al- sumption of different desalination technologies was taken much at-
ternatives are simply implemented in such systems; the first is using tention. At present, of the total production capacity, Reverse Osmosis is
thermal collectors for heat collection while the second depends on contributed up to 63%, multi-stage flash evaporation around 23% and
photovoltaic panels to convert the electromagnetic radiation into only 8% by MED [3]. Actually, newly developed desalination technol-
electricity [1]. Though, a number of desalination plants around the ogies are still in lab-scale and need further improvement.
world are still under continuous research and development. Different comparisons and analysis have been conducted for four

⁎
Corresponding author.
E-mail address: [email protected] (A.W. Ezzat).

https://doi.org/10.1016/j.tsep.2020.100710
Received 13 May 2020; Received in revised form 27 August 2020; Accepted 28 August 2020
Available online 06 September 2020
2451-9049/ © 2020 Elsevier Ltd. All rights reserved.
A.W. Ezzat, et al. Thermal Science and Engineering Progress 20 (2020) 100710

Nomenclature Subscripts

x Steam quality, kg/kg a Ambient

T Temperature, °C i Inlet
ΔT Temperature difference, °C n Nozzle
P Pressure, kpa d Diffuser
m Mass, kg f Liquid phase
M Mach number g Vapor phase
ṁ Steam mass flow rate, kg/s i Inlet
G Gain in water productivity, kg/kJ av Average
A Cross sectional area, m2 s Isentropic
Ac Collector area, m2 1,2,…,7,8 Section designation numbers in steam ejector
h Enthalpy, kJ/kg
hfg Latent heat of evaporation, kJ/kg Abbreviations
s Entropy, kJ/kg °K
UL Overall heat loss coefficient, W/m2.°C ABVC Absorption vapor compression
v Steam velocity, m/s ADVC Adsorption vapor compression
S Absorbed solar radiation, W/m2 MSF Multi stage flash
IT Incident solar radiation, W/m2 MED Multi effect desalination
Qu Thermal gain, W ED Electro-dialysis
FR Heat removal factor EDR electro-dialysis reversal
TVC Thermal vapor compression
Greek symbols RO Reverse Osmosis
MVC Mechanical vapor compression
α Effective absorptance of collector plate with glazing, 0.9 COP Coefficient of performance
β Water productivity, kg/ kJ MD Membrane distillation
η Efficiency H/D Humidification–Dehumidification
ρ Density, kg/m3
μ Entrainment ratio

different types of single-effect evaporator desalination systems [4]. The water rate.
systems were operated by vapour compression heat pumps including Another study proved that MED method does not require additional
thermal (TVC), mechanical (MVC), absorption (ABVC) and adsorption heat for evaporation at all levels and can be operated at lower tem-
(ADVC). The parameters used for the analysis are: specific power con- peratures (~70 °C) than MSF [6]. It has been demonstrated that a
sumption, performance ratio, specific cooling water flow rate and higher thermal efficiency could be obtained by increasing evaporator
specific heat transfer area. The performance ratio related to the thermal and generator temperatures or decreasing condenser saturated tem-
vapour compression system has been proved to be inversely propor- perature. Whereas, the exergy efficiency would be greater at higher
tional to boiling pressure and temperature. A suitable steady-state evaporator and generator temperatures as well as condenser saturated
mathematical model has been developed for a single-effect thermal temperatures. The performance of a steam ejector with the condensa-
vapour compression (TVC) desalination process [5]. The model was tion phenomenon has been investigated for desalination applications
used to study system performance due to variations of the physical [7]. It has been noticed that, in the chocking section, the liquid droplets
properties through the demister with fluid temperature and salinity, could be vaporized by increasing the superheat degree of the vapour to
pressure drop, boiling-point rise, specific heat transfer area and cooling 35 K in the inlet section. On the other hand, the results showed that the
irreversibility of a steam ejector could occur in the mixing process. The
irreversible loss in the mixing process has been estimated to be 73% of
the total irreversible losses of the ejector when the inlet vapours are
saturated. One dimensional model has been implemented to study the
efficiency of solar-assisted refrigeration systems using the supersonic
ejector technique [8]. Research outcomes proved that the generator
temperature should basically exceed 90 °C in order to obtain an ac-
ceptable performance coefficient. Evaporator temperatures have been
limited to values below 10 °C and condenser temperatures above 35 °C.
Constant pressure mixing ejector theory has been used to estimate the
ejector dimensions.
Flash evaporation of saline water has been used on a small-scale
system to conduct experiments for the purpose of investigating the ef-
fect of operating parameters on flash evaporation enhancement [9]. A
suitable vacuum pump has been implemented to create low pressure in
the evaporator to ensure flash evaporation. A freshwater production
rate of 4 l/h has been reached as the maximum value when 56 °C
temperature and 0.08 bar pressure have been ensured in the eva-
porator, while the flash rate of the feed water was limited to 3.6 l/min.
A testing rig has been designed and used for simulating water desali-
Fig. 1. Flow sheet of the existing desalination methods according to type of
energy used. nation systems to study the effect of seawater inlet temperature and

2
A.W. Ezzat, et al. Thermal Science and Engineering Progress 20 (2020) 100710

mass flow rate on system performance [10]. It has been concluded that primary steam generation in the mathematical model. Saline water is
the performance falls with the inlet temperature of the seawater. added to the feed tank for heating purpose by the solar thermal col-
An experimental study has been realized using a steam ejector in a lector. The pre-heated water is then heated up to the saturation tem-
desalination system with thermal vapour compression (TVC) approach perature in the stem boiler to generate primary steam at design pressure
[11]. The system showed that steam ejector can work effectively when and temperature. Portion of the pre-heated water in solar thermal
operated below 100 °C. It has been also proved that coefficient of collector is fed to the evaporator. The pre-heated water in the eva-
performance (COP) of the steam ejector is proportional to secondary porator is heated to evaporator temperature using immersed electrical
steam temperature and inversely proportional to primary steam tem- heater. The primary and secondary steam are condensed in the con-
perature. Moreover, steam ejectors are found to work at critical con- denser, and then collected in distilled water tank.
densation temperatures higher than that based on typical operating A single nozzle steam ejector is used in the present research as
conditions. Ejector assisted passive solar desalination system has been shown in Fig. 3. The ejector is the most critical component in this
investigated using detailed CFD analysis complemented by experiments system because it realizes system function during desalination process
for both open and closed ejectors [12]. The research concentrated on [13]. The shape of the ejector nozzle, convergent or convergent-di-
the capability of the new design to reduce the consumed energy of the vergent, directly influences the Mach number within the ejector. In the
process. The configuration related to closed ejectors showed reliable current research, the Mach number is controlled to be less than 1
performance. The results proved the possibility of energy reduction by (subsonic) by selecting the convergent shape for ejector nozzle. The
reducing the feed water water flow rate. high pressure and temperature of the saturated steam produced by the
The prime motive behind the present research was designated as to: boiler are allowed to flow through the ejector nozzle to the subsequent
1) Improve knowledge of the fluid flow and thermodynamics inside and parts. The primary steam flow into the convergent nozzle of the ejector
around the subsonic steam ejector. 2) Validate theoretical results based depressurizes the evaporator entrance area due to the conversion of
on an experimental approach. 3) Investigate subsonic ejector efficiency primary steam potential energy into kinetic energy and ejects the steam
according to operating parameters. 4) Estimate the enhancement in the evaporated from the evaporator to the mixing area of the ejector. This
distilled water productivity by both prime and induced evaporation steam flow from the evaporator is nominated as secondary flow. The
using subsonic ejector in comparison with that realized due to eva- ratio of secondary steam flow rate to that of primary could be con-
poration at atmospheric pressure. 5) Calculate the effectiveness of in- trolled through the design of the ejector and choosing the operating
corporating solar thermal collector within the system. parameters such as pressure and temperature in its primary, secondary
In the experimental part of the present research, the solar collector and outlet ports. Mathematical modelling of the system includes the
is simulated by a variable power heat source to control primary steam solar thermal collector and flash evaporation process.
pressure and temperature. While both primary and secondary steam
condensation processes are realized by open-loop condensation in 2.2. Thermal design of flat-plate solar collector
which the condensed steam is collected at the outlet of the condenser.
The water inside the evaporator is preheated using an electrical heater A solar thermal collector is incorporated within the mathematical
during the experiments. modelling of the system to back up both boiler and evaporator with
appropriate portion of thermal energy at certain prerequisite tem-
2. Theoretical approach perature, see Fig. 2. A flat-plate collector is designed for that purpose
following the method of [14] using four specific collector factors with
2.1. Physical modelling of the system the local solar radiation and the ambient temperature of Baghdad, Iraq.
The basic equation of flat-plate collector is given by [14] as:
The block diagram of Fig. 2 shows the contribution of both solar Qu = Ac FR [S UL (Ti Ta )] (1)
thermal collector and steam boiler to the total energy required for

Fig. 2. Block diagram of the desalination system used in the present research.

3
A.W. Ezzat, et al. Thermal Science and Engineering Progress 20 (2020) 100710

Fig. 3. (A) 3D geometry of the ejector. (B) Ejector sections.

The heat removal factor FR includes parameters of collector piping The pressure of steam into the suction chamber P3 is the same at
design, fluid properties and its heat transfer coefficient, and mass flow nozzle outlet (P2 = P 3); therefore, s2f = s 3f and s2g = s 3g. Assuming
rate. isentropic process between states at the sections 1 and 2, then s1 = s2s.
The absorbed solar radiation S is found by [14] as: Therefore, the steam quality at position 2 could be calculated as:
S = IT (2) x2 = (s1 s3f )/(s3g s3f ) (4)
Collector design should provide the required fluid temperature with The isentropic enthalpy at state 2, h2s could be calculated based on
reasonable partial energy for water heating and evaporation during the relationship between the relative vapor quality and the nozzle ef-
desalination process. The incident solar radiation IT was calculated ficiency ηn = 0.7.
according to the isotropic model using beam, diffuse, and reflected
components for tilted collector due south [14–16]. The annual average n = (h1 h2)/(h1 h2s ) (5)
hourly solar radiation was found to be 480 W/m2 for a good atmo- Accordingly, steam density and enthalpy at state 2 could be calcu-
spheric condition at the optimum collector tilt angle [16]. The annual lated. Based on above assumptions, the expanding and accelerating
average daily ambient high temperature is 30 °C [16]. process of the primary fluid in the primary nozzle should meet energy
The design of the thermal collector for the present study is accom- and mass conservation. The mass conservation equation between sec-
plished using a computer program that is specially developed for that tion 1 and section 2 as ṁ 1= ṁ2, could be written as:
purpose. The main input and output parameters of the program, as
hourly average values, are shown in Table 1. 1 A1 v1 = 2 A2 v 2 (6)
While the energy equation between the same sections could be
2.3. Mathematical modelling of flash evaporation written as:

The design pressure and temperature of primary steam are ranged h1 + v12/2 = h2 + v22/2 (7)
from (1.25–2.5) bar and (106–127) °C respectively, while those related The velocity at state 1 and 2 can be calculated by solving Eqs. (6)
to condenser pressure and temperature ranged (0.974–1.0) bar and and (7) together.
(97–100) °C respectively. The estimated thermal energy for water Moreover, the mass flow rate at state 1 and 2 can be calculated by
heating and evaporation is 4 kW for one hour. The simplified diagram Eqs. (8) and (9):
of the process flow of the studied case and its related p-h diagram are
shown in Fig. 4. The following assumptions are considered in the m1 = 1 A1 v1 (8)
mathematical model: The pressure losses in the condenser, evaporator
m2 = 2 A2 v 2 (9)
and connection pipeline of system components are ignored, no heat
exchange between system parts and the environment, an isenthalpic Process 7–8:
process in the throttling process: state 1 to 2, state 3 to 4 and state 7 to There is an isentropic process between state 7 to state 8, therefore,
8, the fluid in the ejector is one-dimensional homogeneous flow, steam s8 = s7s = s7g. The steam quality at state 8 is found as:
flow from sections 1, 7 and 4 are considered as a saturated vapour,
primary and secondary steam is assumed to be entirely mixed and
x 8 = (s8 s8f )/(s8g s8f ) = (s7g s3f )/(s3g s3f ) (10)
further compressed in the subsonic diffuser and then discharged to the Therefore, the relative property of the steam in state 8, such as ρ8
condenser, the pressure of both primary and secondary steam into the
mixing chamber at section 3 is the same. The pressure in sections 1 and Table 1
4 corresponds to design pressure, while the pressure in section 3 is a Thermal collector design parameters.
pre-assumed value. This value is cross-checked and corrected using part
no parameter value no parameter value
2 of process 2 + 7–4. The mathematical model for flash evaporation is
based on simple thermodynamic equations that track water evaporation 1 tilt angle 30° 6 heat transfer 270 W/m2.
and condensation processes at different sections [17–20]. The percen- coefficient °C
2 inlet temperature 30 °C 7 outlet temperature 75 °C
tage salinity for water physical properties in the evaporator is ignored
3 overall heat loss 7 W/m2. 8 collector 60 °C
due to its minor effect [21]. coefficient °C temperature
The steam quality at state 2 is found as: 4 mass flow rate 7.2 g/s 9 thermal efficiency 52%
5 collector area 6 m2 10 thermal gain 1.35 kW
x2 = (s2s s2f )/(s2g s2f ) (3)

4
A.W. Ezzat, et al. Thermal Science and Engineering Progress 20 (2020) 100710

Fig. 4. (A) Process flow of the studied case. (B) P-h diagram of the system.

and h8 can be calculated based on x8. 1. The mixing process of the primary and secondary steam streams
Process 3–4: in the mixing chamber satisfies both energy and mass conservation
The process from 3 to 4 is similar to the process from 1 to 2. The equations respectively as follows:
vapor quality in state 3 needs to be known by:
m4 (h4 + v42/2) = m2 (h2 + v22/2) + m8 (h8 + v82 /2) (13)
x3 = (s3s s3f )/(s3g s3f ) (11)
4 A 4 v4 = 2 A2 v 2 + 8 A8 v 8 (14)
where s3s = s4s = s4g, since there is an isentropic process from state 3 to
whereṁ 4 is the sum of the primary and secondary mass flow rates.
4. Therefore, Eq. (11) could be transformed as x3 = (s4g − s3f)/(s
Eqs. (13) and (14) could be solved together in order to find the
3g − s3f), ρ3 and h3 could be calculated based on the relative vapor
velocity at state 4 and state 8 (ν4 and ν 8).
quality x3.
The mass flow rates at section 8 and section 4 could be calculated by
The diffuser efficiency ηd = 0.8, is represented as follows:
Eqs. (15) and (16) respectively as follows:
= (h4s h3)/(h4 h3) (12)
d m8 = 8 A8 v 8 (15)
where h3 is enthalpy at state 3 and h4s is isentropic enthalpy at state 4.
m4 = 4 A 4 v4 (16)
Then h4 could be calculated by Eq. (12).
Process 2 + 7–4: 2. Check the assumed value of P3

Fig. 5. The conceptual design of experimental test rig.

5
A.W. Ezzat, et al. Thermal Science and Engineering Progress 20 (2020) 100710

%G = [1/hfg, T 7 1/(hfg ,100 + Cp (100 T7 )]/[1/(hfg,100 + Cp (100 T7)]

(22)

3. Experimental approach

3.1. Experimental test rig

Flash evaporation of water in sub-cooled phase depends on the

technique used for depressurization. Some of these techniques are
based on static depressurization of the upward flowing fluid in the test
section [22,23]. Other techniques depend on the depressurization in-
duced by steam jet using steam ejectors.
Testing rig is designed and constructed to validate the mathematical
model according to the operating pressures, temperatures and thermal
power consumed during desalination process using flash evaporation
induced by steam jet in subsonic ejector. The conceptual design of the
testing rig simulated the solar collector by variable power boiler which
is used to heat up the water to produce primary steam pressure of
Plate 1. Experimental test rig. (1.25–2.5) bar and temperature of (106–127) °C. The configured test rig
ensures water evaporation in the evaporator using an electric heater
and open-loop condenser cooling that covers the operational condi-
The mixing process of the two fluids in the mixing chamber satisfies
tions. Fig. 5 illustrates the schematic diagram of the test rig. The
momentum conservation equation:
pressure control valve 1 is used to regulate steam pressure and flow rate
m4 v4 = P3 A3 + m2 v2 + m8 v8 P4 A 4 (17) according to the design limits. The primary steam for the subsonic
ejector actuates the secondary flow of steam from the evaporation tank
where ν4 and ν7 refer to the velocity at state 4 and 7; P4 and P2 refer to
due to the low-pressure condition created in the mixing area at the
the pressure at state 4 and 2, and A3 is the area directly after mixing
upstream of nozzle outlet which also ensures water evaporation at
primary and secondary flow. Eq. (17) can be arranged to find P3 as:
temperatures ranging (75–95) °C. The initial water temperature in the
P3 = (m4 v4 m2 v2 m8 v8 + P4 A 4 )/ A3 (18) evaporator is controlled by the electrical power of the preheating
heater. The mixed primary and secondary steam flow are drafted
Eq. (18) can be solved using iteration procedure by replacingṁ 8 through ejector divergent area to the condenser. The mixed steam
with ṁ7. flowing inside the condenser is condensed after cooling by the water
The entrainment ratio of the ejector μ is defined as: coil fixed inside. A properly designed cooling coil is used in the con-
µ = m7 / m1 (19) denser to remove heat from collected steam and condense it to the
operating temperature of (25–50) °C. Valve 4 is used to drain water
The coefficient of performance COP for subsonic ejector could be from the evaporator at the end of the experiments. The following design
represented by steam entrainment μ as the enthalpy difference along criteria are considered for the test rig:
primary steam path is almost equal to enthalpy difference along sec-
ondary steam path, as given by: 1- Sizing of the subsonic steam ejector is complied with the design
COP = m7 (h7 h 4)/m1 (h1 h 4) = µ (20) parameters of the system by selecting a typical ejector that ensures
nozzle replacement capability.
Distilled water productivity β is estimated based on the steam mass 2- Condenser design ensures the range of the ejector back pressure
generated during evaporation process per thermal energy unit con- implemented during the experiments by using a proper cooling coil
sumed during this process, according to the following equation: for the steam condensation process.
3- The evaporation tank has the capability to keep the water tem-
= 1/ hfg (21)
perature within specified limits by pre-heating due to solar collector
The percentage gain %G of distilled water productivity using flash design.
evaporation at the evaporator temperature T7 corresponding to the 4- Use of distilled water instead of saline water due to the minor effect
evaporator pressure P7 with respect to that produced by the evapora- of water salinity on the physical properties [21].
tion process at the atmospheric pressure Pa is estimated according to
following equation: Plates 1 and 2 show the test rig and the steam ejector used in the

Plate 2. Manufacturing process for two ejector models.

6
A.W. Ezzat, et al. Thermal Science and Engineering Progress 20 (2020) 100710

experiments. Table 3
Experimental results.
4. Results and discussion Steam line Condenser Evaporator ṁ1 (g/s) ṁ7 (g/s) μ exp μ av
pressure, pressure, P4 pressure, p7
P1 (bar) (bar) (bar)
Table 2 illustrates the results obtained using the mathematical
model while Table 3 shows the experimental results, where ṁ1 is the 1.250 1.0 0.940 0.533 0.185 0.347 0.247
primary steam mass flow rate and ṁ7 is the secondary steam mass flow 1.500 0.998 0.928 0.90 0.197 0.219 0.205
rate. 1.750 0.980 0.908 1.05 0.212 0.202 0.279
2.000 0.976 0.886 1.20 0.275 0.229 0.38
The mathematical model used for the present research assumes that
the steam mass flow rate at position 7, ṁ7 is equal to that at position 8,
ṁ8 ignoring the connection area between these positions. This as-
sumption is justified based on the continuity equation between these
two points as there is no mass addition or subtraction among them. The
theoretical values of steam velocity at the outlet of the ejector nozzle, ν2
is cross-checked based on the chocking condition at the nozzle outlet
when this velocity approaches sound velocity, M2 = 1. Primary steam
mass flow rate ṁ1 behavior versus variable primary steam pressure at
position 1, P1 is shown in Fig. 6. The theoretical results illustrate the
saturation of primary flow rate due to the sonic condition in the nozzle
outlet. However, there is a significant gap between the model results
and the experimental results which reach its maximum value at primary
steam pressure of 2 bar. This gap is related to the collection of distilled
water produced from the primary steam condensation process. The
underestimation in theoretical values is justified due to neglecting the
vapor compressibility in the mathematical model for its dependency on
the relative vapor speed to sound speed along the vapor path within
ejector sections. The analytical solution of the mathematical model can
only verify vapor speed at specified areas while ignoring this effect
along the vapor path.
The secondary steam mass flow rate showed an increase in its trend
versus primary steam pressure, see Fig. 7. Steam velocity saturation in
the nozzle outlet affects evaporator depressurization and then subse-
quently saturates the mass flow rate of secondary steam. However,
there is an underestimate in the model results with respect to experi-
mental results at low primary steam pressure. While, overestimation is
noticed as primary steam pressure increased. These differences are due
to neglecting of irreversibility. The experimental validation for the
mathematical model is executed using error analysis between theore-
tical and experimental values of entrainment ratio. The error in en-
trainment ratio has been chosen as it represents the effect of the most
operating conditions in both mathematical model and experimental
approach. The average value of this error was 22.9% due to the effect of
theoretical assumptions on the mathematical results. This error could
be narrowed using CFD model for the theoretical calculations. More-
over, ejector and test rig design could be modified such that to elim-
inate the condensed droplets in the evaporator which affects the ex-
perimental results of entrainment ratio.
Steam entrainment ratios versus steam pressures at nozzle inlet,
evaporator and condenser are plotted using their average values μav of
the experimental and theoretical results; see Table 3. Entrainment ratio
Fig. 6. Primary steam mass flow rate versus primary steam pressure.
trend increases with primary steam pressure P1, as shown in Fig. 8. The
average to mainstream value of the secondary steam mass flow rate
affects this increase. However, this increase is restricted as notified Fig. 9 illustrates the influence of the evaporator pressure P7 on the
early due to the limitation of the primary steam velocity to subsonic average entrainment ratio. The evaporator pressure P7 is proportional
values, as noticed from Table 2. to the steam pressure at nozzle outlet P2, which in turn inversely affects

Table 2
Mathematical Modelling results.
Steam line pressure, P1 (bar) Condenser pressure, P4 (bar) Evaporator pressure, p7 (bar) ṁ1 (g/s) ṁ 7 (g/s) μ th v2 (m/s)

1.250 1.0 0.940 0.485 0.071 0.146 311.7

1.500 0.998 0.928 0.596 0.114 0.191 390.6
1.750 0.980 0.908 0.659 0.235 0.357 437.8
2.0 0.976 0.886 0.718 0.381 0.531 476.8
2.25 0.974 0.878 0.743 0.394 0.533 493.2
2.5 0.974 0.878 0.746 0.396 0.531 495.4

7
A.W. Ezzat, et al. Thermal Science and Engineering Progress 20 (2020) 100710

proportional to the entrainment ratio, as both of them are influenced by

the additional evaporation induced by depressurization in the eva-
porator. This additional steam using subsonic ejector increases the
steam mixture mass flow rate to the condenser.
Fig. 12 shows the comparison between average entrainment ratios
obtained from present research in comparison to that obtained by Liu
research [17] versus primary steam superheat degree. The comparison
proves reliable agreement despite the different range in primary steam
superheat degree and primary steam temperature. The latter is ranged
(106–127) °C for the present research and 140 °C for Liu research [17].
Fig. 13 illustrates the pressure and velocity profiles along the ejector
sections based on the results obtained from the mathematical model.
The figure shows that the highest velocity occurs at the outlet of ejector
nozzle, Section 2, which is close to the sonic velocity but never exceeds
Fig. 7. Secondary steam mass flow rate versus primary steam pressure.
it. The lowest pressure occurs at the mixing chamber, Section 3 so that
the secondary flow could be sucked in. The model proved that velocity
and pressure profiles do not change significantly in the diffuser.
Therefore, in the subsonic velocity ejector, the diffuser could be de-
signed as a constant cross-section with a minor effect on pressure and
velocity profile downstream the nozzle section of the ejector.

5. Conclusions

Experimental and mathematical studies were performed on steam

jet flash evaporation for desalination system using subsonic ejector and
incorporating solar thermal collector. The effects of primary steam
pressure, evaporator and condenser pressures on system performance
were investigated at the primary steam pressure ranging from 1.25 bar
Fig. 8. Average entrainment ratio versus primary steam pressure. to 2.5 bar and condenser pressures ranging from 0.974 bar to 1.0 bar. A
mathematical model was developed and validated against the experi-
mental data generated by the steam ejector desalination system ex-
periments conducted on a well-designed testing rig. The model is then
applied to estimate the effects of operating parameters on secondary
steam entrainment ratio and the percentage gain of water productivity.
The following outcomes are concluded from the present research:

1) Steam ejector efficiency represented by entrainment ratio showed

an increase of 53% when the primary steam pressure changed from
1.25 bar to 2.0 bar. However, beyond 2 bar this ratio saturates due
to the limitation of the primary steam velocity to subsonic values.
2) The entrainment ratio is inversely proportional to evaporator pres-
sure ranged from 0.898 bar to 0.94 bar.
Fig. 9. Average entrainment ratio versus evaporator pressure. 3) The tendency of ejector efficiency decreases as the condenser pres-
sure increases from 0.974 bar to 1.0 bar.
4) The percentage gain of distilled water productivity at selected eva-
the secondary steam mass flow rate. This inverse proportionality jus-
porator temperature with respect to that produced during the clas-
tifies the trend of entrainment ratio with evaporator pressure. In con-
sical evaporation process ranged (1.0–5.5)%.
clusion, evaporator pressure P7 is another important parameter that
5) Based on a sensitivity study, the mathematical model could be ap-
influences ejector efficiency.
plied to correlate between ejector geometry, operating parameters
Additionally, Fig. 10 proves that the average entrainment ratio is
and ejector performance. The best combination could be selected for
inversely proportional to condenser pressure P4 and that the trend of
the highest ejector efficiency at different conditions.
proportionality saturates due to the saturation of condenser pressure P4
6) According to the present solar collector design, 34% of the total
at 0.974 bar. This relationship demonstrates that condenser pressure
has another role in controlling the ejector and the efficiency of the
whole process. However, there is a critical condenser pressure for each
ejector [24]. The influence of condenser pressure on the entrainment
ratio is restricted to such nominated critical pressure, which means that
condenser pressure P4 should not be set smaller than the critical pres-
sure to avoid the reduction of ejector efficiency. Steam ejector effi-
ciency governed by the entrainment ratio showed an average increase
of 53% when the primary steam pressure changed from 1.25 bar to
2.0 bar.
Fig. 11 shows that the percentage gain of distilled water pro-
ductivity at selected evaporator temperature with respect to that pro-
duced during the classical evaporation process, T7 = 100 °C, ranged
(1.0–5.5)%. The figure proves that water productivity is directly Fig. 10. Average entrainment ratio versus condenser pressure.

8
A.W. Ezzat, et al. Thermal Science and Engineering Progress 20 (2020) 100710

condensation process for water pre-heating in the evaporator. The au-

thors also recommend implementing solar energy in the experimental
part. The heat gain from the collector could be shared for heating up the
water inside evaporator and for ensuring the primary steam production
rate.

CRediT authorship contribution statement

Akram W. Ezzat: Conceptualization, Data curation, Formal ana-

lysis, Investigation, Methodology, Project administration, Software,
Supervision, Validation, Visualization, Writing - original draft, Writing -
review & editing. Eric Hu: Conceptualization, Formal analysis, Funding
acquisition, Investigation, Methodology, Project administration,
Fig. 11. Percentage gain in distilled water productivity versus evaporator
Resources, Supervision. Hussein M. Taqi Al-Najjar: Data curation,
temperature.
Formal analysis, Investigation, Methodology, Software, Writing - re-
view & editing. Zihui Zhao: Data curation, Formal analysis, Resources,
Investigation, Software, Visualization. Xin Shu: Data curation, Formal
analysis, Resources, Investigation, Validation, Visualization.

Declaration of Competing Interest

The authors declare that they have no known competing financial

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

Acknowledgments

The authors acknowledge the support of the Adelaide University

Fig. 12. Average entrainment ratio versus primary steam superheat degree.
workshop and RAB engineering services in Adelaide- Australia for their
support to fabricate the ejectors used in the test rig to validate the
mathematical model. This acknowledgment is extended to appreciate
the support of Chemical Engineering School in Adelaide University
–Australia to utilize the steam boiler in the distillation Lab. during
conduction the validation experiments.

References

[1] A. Abutayeh, et al. Solar Desalination. NextEra Energy Resources, Juno Beach,
Florida, USA, 2014, pp. 531–551.
[2] V. Belessiotis, et al., Thermal Solar Desalination Methods and Systems, Science
direct. (2016) 1–382.
[3] Z. Zheng, et al., State-of-the-art of R&D on seawater desalination technology, Chin.
Sci. Bull. 61 (2016) 2344–2426.
[4] Al-Juwayhel, et al., Analysis of single-effect evaporator desalination systems com-
bined with vapor compression heat pumps, Desalination 114 (1997) 253–275.
[5] H. El-Dessouky, et al., Single-effect thermal vapor-compression desalination pro-
Fig. 13. Normalized steam pressure and velocity profile along ejector sections cess: thermal analysis: thermal analysis, Heat Transf. Eng. 20 (1999) 52–68.
with respect to condenser pressure and sound velocity. [6] H. Rostamzadeh, et al., Energetic and exergetic analyses of modified combined
power and ejector refrigeration cycles, Therm. Sci. Eng. Progress 2 (2017) 119–139.
[7] J. Liu, et al., Thermodynamic analysis of steam ejector for desalination applications,
thermal energy required for water heating and evaporation during Appl. Therm. Eng. 159 (2019) 1–11.
[8] S. Varga, et al., Analysis of a solar-assisted ejector cooling system for air con-
desalination process is covered by the 6-m2 flat-plate collector. ditioning, Int. J. Low-Carbon Tech. 4 (2009) 2–8.
However, during specific months of higher incident solar radiation [9] G. Wessley, et al., Investigations on low temperature flash evaporation desalination
and/or larger collector area greater coverage fraction can be for small-scale applications, Int. J. Eng. Tech. 5 (5) (2013) 4109–4116.
[10] J. Ji, et al., Simulation and analysis of a single-effect thermal vapor-compression
achieved. desalination system at variable operation conditions, Chem. Eng. Technol. 30
7) Experimental validation showed an average error of 22.9% in the (2007) 1633–1638.
steam entrainment ratio. [11] J. Dong, et al., Experimental investigation of the steam ejector in a single-effect
thermal vapor compression desalination system driven by a low-temperature heat
8) The present research could be considered as a great basis for further
source, Energies 11 (2018) 1–13.
investigations of solar power-sharing between main steam produc- [12] M. Ahmadi, et al., Experimental and numerical investigation of a solar educator-
tion and saline water pre-heating in the evaporator. assisted low-pressure water desalination system, Eng. Sci., Chinese Sci. Bull. 61
(2016) 12.
[13] Croll Reynold Steam Ejectors, 2018, https://croll.com/vacuum-systems/applica-
It should be noted that due to large differences in operating con- tions/chemical-processing/24-2/.
ditions, working fluid and/or steam ejector geometry, significant [14] J.A. Duffie, et al., Solar Engineering Thermal Processes, John Wiley & Sons Inc,
comparisons of the present research with other related works can only 2006.
[15] H.M.T. Al-Najjar, et al., Experimental prediction of hourly diffuse solar radiation
be attained for the entrainment ratio at various ranges of primary steam with clearness index in Baghdad (Iraq), Int. J. Computer Appl. 158 (7) (2017)
temperatures and superheat degrees. 20–28.
For future work, the authors recommend investigating the effect of [16] H.M.T. Al-Najjar, Electric, heating and cooling yields of solar collectors for different
atmospheric conditions and tilt angles, Int. J. Computer Appl. 141 (10) (2016)
operating parameters on ejector thermal efficiency using the steam 1–10.

9
A.W. Ezzat, et al. Thermal Science and Engineering Progress 20 (2020) 100710

[17] S. Liu, et al., Thermodynamic analysis of steam ejector refrigeration cycle, Int. [21] A. Kendoush, et al. Investigation the effect of water salinity on flash evaporation,
Refrigeration Air Conditioning Conf. (2014) 2306–2307. ASME International Mechanical Engineering Congress and Exposition, 2007.
[18] R. Kelso, Applied aerodynamics: compressible flow, PowerPoint Presentation [22] A. Izzat, et al., Flash evaporation enhancement by electrolysis of saturated water
(2018). flowing upwards in vertical channel, J. Eng. 16 (3) (2010) 5620–5713.
[19] S. Han, et al., One-dimensional numerical study of compressible flow ejector, AIAA [23] A. Ezzat, et al., Investigation of flash evaporation in solar water desalination sys-
J. 40 (7) (2002) 1469–1473. tems, J. Mech. Eng. Res. Devel. 42 (4) (2019) 135–140.
[20] Antonio, et al., Thermodynamic modelling of an ejector with compressible flow by a [24] Yu y, et al. Analysis of operating parameters of steam injector for refrigeration.
one-dimensional approach, Entropy 14 (4) (2012) 599–614. Research and discussion 2018, 1–3.

10