Desalination 351 (2014) 202–212

Contents lists available at ScienceDirect

Desalination
journal homepage: www.elsevier.com/locate/desal

A novel process for low grade heat driven desalination

Bijan Rahimi a,b, Alexander Christ a,b, Klaus Regenauer-Lieb b, Hui Tong Chua a,c,⁎
a
School of Mechanical and Chemical Engineering, The University of Western Australia, 35 Stirling Hwy, Perth, WA 6009, Australia
b
School of Earth and Environment, The University of Western Australia, 35 Stirling Hwy, Perth, WA 6009, Australia
c
School of Environmental Science and Engineering, Taiyuan University of Technology, Taiyuan, Shanxi Province, China

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• A flash-boosted multi-effect distillation A novel waste heat powered flash boosted MED system delivering 50% more freshwater.
process has been developed.
• The process maximally exploits the
potential of waste sensible heat source.
• 40–50% more freshwater compared with
conventional multi-effect distillation
process.
• Auxiliary power consumption only
increases modestly by 22–34%.
• 4% to 6% decrease in the specific capital
cost

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

Article history: Low grade process heat and geothermal energy with temperatures typically below 100 °C are significant un-
Received 17 April 2014 tapped environmentally friendly resources for desalination. This article reports on a novel multi-effect distillation
Received in revised form 11 July 2014 (MED) desalination process that is boosted by a multi-stage flashing process. Specifically the low grade heat first
Accepted 16 July 2014
heats up the multi-effect distillation plant and is then maximally exploited through a multi-stage flashing
Available online 15 August 2014
process, with the produced steam of which being judiciously introduced into the multi-effect distillation plant
Keywords:
for enhanced freshwater production. Compared with optimized conventional MED processes, the performance
MED improvement is up to around 50% better in terms of freshwater production, with a modest increase in the
Multi effect distillation pumping power consumption and 4% to 6% decrease in the specific capital cost.
MSF © 2014 Elsevier B.V. All rights reserved.
Multi stage flash
Flashing
Waste heat

Nomenclature CP specific heat capacity (kJ/kg.°C)

A area (m2) Dt total plant production rate (m3/day)
a, α constant DFC generated vapor mass flow of the flashing chamber section
β constant (m3/day)
BPE boiling point elevation (°C) Δhref specific reference enthalpy of the distillate (kJ/kg)
Δhavail. maximum exploitable energy of the heat source (kJ/kg)
⁎ Corresponding author. Δhref specific reference enthalpy of the distillate (kJ/kg)
E-mail address: huitong.chua@uwa.edu.au (H.T. Chua). Δ Ḣ rate of energy (enthalpy) transfer (MW)

http://dx.doi.org/10.1016/j.desal.2014.07.021
0011-9164/© 2014 Elsevier B.V. All rights reserved.
 B. Rahimi et al. / Desalination 351 (2014) 202–212
ΔP pressure difference (kPa)
ΔTlm log mean temperature difference (°C)
ΔTst stage temperature decrement (°C)
hf enthalpy of saturated water (kJ/kg)
hfg latent heat (kJ/kg)
hg enthalpy of saturated steam (kJ/kg)
H brine level (m)
HEX heat exchanger
L flash chamber length (m)
Liq. liquid
M ̇ total total freshwater mass flow rate (kg/s)
ṁ mass flow rate (kg/s)
NCG non-condensable gas
NEA non-equilibrium allowance for flashing stage (°C)
NEA10 non-equilibrium allowance for a 10-ft flashing stage (°C)
PR performance ratio
T temperature (°C)
Q̇ heat transfer rate (kW)
U overall heat transfer coefficient (kW/m2.°C)
V̇ volumetric flow rate (m3/s)
W brine flow rate per unit length of chamber width (kg/(m.s))
X salinity
η efficiency
Ψ cost function for MED plants
Φ cost function for MSF plants
1. Introduction
Nearly 71% of the surface of the earth (510 × 106 km2) is covered by
the oceans and the remaining 29% is occupied by the lands [1]. There is
certainly bountiful water available on earth, but only 3% of which is drink-
able and 97% is saltwater [2]. The demand for freshwater is set to increase
significantly due to a combination of population growth, acceleration of
industrialization in developing countries and increased agriculture
activity throughout the world [3]. Water scarcity is the mismatch of the
demand and availability of freshwater in a particular location. As
mentioned above, traditional sources of available freshwater such as un-
derground aquifers and surface water constitute a limited quantity
worldwide. Furthermore, depletion and degradation of these sources
are increasing at accelerating rates [4]. As a result, seawater desalination
has become an essential option to augment freshwater resources, espe-
cially in developing countries and many arid zones. In 2011, approxi-
mately 150 countries worldwide used around 15,988 desalination
plants (these include online, under construction and contracted) to pro-
duce desalinated waters [5]. The total global capacity of all online plants
was 70.8 Mm3/day [6] in 2011, which had a 10% increase in comparison
with that in 2010. Based on the recent information from the International
Desalination Association [6], 632 new plants have been added from mid-
2011 to Aug-2012, which has increased the installed capacity to
74.8 Mm3/day. These data indicate the potential of desalination market.
In general all applicable desalination processes can be divided into
two main categories based on the phase change of saline feedwater.
• Desalination with phase change: This category includes all heat
driven processes where freshwater is produced by evaporation and
condensation phenomena. The principle examples are MED (Multi-
Effect Distillation), MSF (Multi-Stage Flash), TVC (Thermal Vapor
Compression), HD (Humidification-Dehumidification) and MD
(Membrane Distillation). In Aug-2012, the share of MED and MSF
technologies of the total installed capacity for all feedwater types
was around 31% [6].
• Desalination without phase change: In this category, separation is
achieved by passing saline water through membranes without involv-
ing phase change and RO (Reverse Osmosis) is the iconic example.
RO's share of installed capacities was 63% in 2012 [6].
203
Desalination is an energy intensive process. For example the RO pro-
cess has a high overall efficiency at the expense of consuming a large
amount of electricity and for thermal processes such as MSF and MED
they relate to large thermal energy consumption. Therefore in respect
of production rate, economic feasibility and environmental friendliness,
the optimization of desalination methods should be considered in the
context of minimizing energy consumption. In the face of pressing con-
cerns regarding anthropogenic carbon emissions and global warming,
meeting an increasing worldwide demand for freshwater is not a simple
matter of increasing desalination capacity. Sustainable and renewable/
waste energy based approaches must also be considered.
Waste thermal energy has always been an important issue in the
process industries. Management of waste heat resources is one of the
important subjects in process plants. For this purpose waste heat
streams is being divided into two major categories: high and low grades.
High grade refers to those heat sources that are typically recovered by
the plant processes, while low grade refers to those which are not eco-
nomically viable to recover and rejected to the environment [7]. One op-
tion to utilize the low grade heat sources of a process plant is to produce
freshwater for internal or external usage (e.g. [7–13]). Furthermore, low
grade renewable energy resources such as geothermal with a wellhead
temperature lower than 100 °C [14] can be considered for the desalina-
tion purposes (e.g. [14–22]).
One of the main advantages of low-grade heat sources is related to
carbon dioxide emission and global warming issue. If the required energy
hails from fossil fuel source then the freshwater production will contrib-
ute to carbon dioxide emission and consequently global warming. Low
grade heat sources such as waste heat from process plants and geother-
mal energy generate minimal greenhouse gasses. This article introduces
a novel MED process that is coupled with low grade sensible (b 100 °C)
heat sources.
2. Low grade heat driven multi effect distillation
In the temperature range of low grade sensible heat sources, MED is
ideal as its top brine temperature varies between 60 °C to 75 °C. Fig. 1
shows a conventional MED system. In this system the feedwater is dis-
tributed onto the heat exchanger surfaces of the first effect. The heat
source fluid that flows through the heat exchanger releases its energy
to the distributed feedwater and evaporates a portion of the feedwater.
The produced vapor then condenses in the heat exchanger of the second
effect to evaporate more feedwater in that effect. The brine from
the first effect is then purged. At the second effect, the evaporated
feedwater goes on to power the third effect with the resulting brine
being drained from the bottom of that effect. This process continues to
the last effect (the fifth effect in Fig. 1) with the corresponding produced
vapor entering the condenser section and condensed by the incoming
saline water acting as a coolant. Part of the preheated saline water is
then sent to the various effects as feedwater. In the context of geother-
mal and sensible waste heat sources, this process is inefficient as the
outgoing heat source temperature is still sufficiently high [13].
Recently a novel boosted MED system tailored for waste heat streams
has been reported [11,12,23,24]. In this method and as shown in Fig. 2, a
steam booster unit is installed to better exploit the waste heat stream so
as to increase the freshwater yield. This booster unit (or an evaporator) is
powered by the outgoing waste heat source of the primary MED plant.
The generated vapor from the booster unit is then introduced into an ap-
propriate effect of the primary MED plant. This scheme substantially
heightens the production rate, but the extent of improvement is limited
by the temperature drop across the booster unit.
3. Description of the novel system — flash boosted MED (FB-MED)
To further exploit waste heat, an improved system has been devel-
oped, as shown in Fig. 3 [25]. The improvement is derived from the abil-
ity of the system to extract the maximal energy from the waste heat and
204 B. Rahimi et al. / Desalination 351 (2014) 202–212

Fig. 1. Optimized conventional MED system (parallel feed configuration), typical temperature distribution as per the simulation detailed below.

transform it into heightened freshwater production in the primary MED
plant. To this end, multiple flashing chambers are installed to efficiently
transform the waste energy into valuable steam. Referring to Fig. 3, the
outgoing heat source from the primary MED plant goes on to heat up the
feedwater via a liquid-to-liquid heat exchanger, which has been mar-
ginally preheated by the outgoing brine from the last flashing chamber.
The heightening of the feedwater temperature via this approach is only
limited by the practical temperature of approach of the liquid-to-liquid
heat exchanger and top brine temperature and this scheme ensures that
it comes sufficiently close to the outlet temperature of the heat source
leaving the primary MED plant. The heated feedwater then goes
through a series of flashing chambers. The generated vapor from each
stage of flashing is then injected into the judicious effect of the primary
MED plant for further boosting as in the previous steam boosted
scheme.
It is discernible that this scheme permits the waste heat source
outlet temperature to approach the mixing point temperature between
the flashing chamber outlet and the condenser outlet temperatures as
close as is practicable. This process therefore engenders a maximally ef-
ficient scheme to exploit the potential of waste heat, far more than the
conventional and boosted MED technologies.
The recycling of brine emanating from the series of flashing cham-
bers, other than the nominal energetic advantage as noted above, also
effectively suppresses the amount of released non-condensable gasses
and thereby further improves the heat transfer efficacy and attenuates
the venting and vacuuming power consumptions. To prevent cavitation,
the make-up water added to the brine recirculation stream is also
deaerated.
4. Numerical analysis and validation
A simulated model has been developed to quantify the efficiency of
the flash boosted MED system. It is a modified version of the earlier
model [11,12], and considers the effect of multiple flashing chambers.

Fig. 2. Schematics of the improved low grade heat MED plant (boosted MED System), typical temperature distribution as per the simulation detailed below.
 B. Rahimi et al. / Desalination 351 (2014) 202–212 205

Fig. 3. Schematic design of the novel system. (Flash Boosted MED System) The process temperatures indicated come from the simulation as detailed below.

The basic simulation has been validated against market available data BPE is the boiling point elevation that can be calculated from
from a reputable manufacturer [11,26]. The energy balance in the first Eq. (B.1) [27] for both Eqs. (3) and (5). In all the above mentioned equa-
effect is expressed as tions, the feedwater inlet temperature to each MED effect (Tf,k) is the


 same as the condenser outlet temperature (Tcond,out) as a portion of
Q̇ k ¼ ṁhs  C p;hs  T hs;in −T hs;out ¼ ṁ f ;k  C p; f  T b;k −T f ;k þ ṁv;k  hfgðT Þ this coolant is used as the feedwater.
vs;k
The energy balance for the condenser is presented as
¼ ðUAÞk  ΔT lm;k ; k ¼ 1: ð1Þ


hfg ðT vs Þ is the latent heat of evaporation and can be quantified from Q̇ cond ¼ ṁv;n;mix  hfgðT Þ ¼ ṁcond  C p;cond  T cond;out −T cond;in
vs;n;mix

Eq. (A.1). ¼ ðUAÞcond  ΔT lm;cond ð4Þ

According to which, within the first effect the sensible heat source
increases the temperature of feedwater to the boiling point and enjoins
partial evaporation. The energy balance for the kth effect is expressed as
in Eq. (2). This equation is applicable to those effects which have not re-
ceived vapors from the flashing chambers, such as the second and third
effects as in Fig. 3.


Q̇ k ¼ ṁv;k−1  hfgðT Þ ¼ ṁf ;k  C p; f  T b;k −T f ;k þ ṁv;k  hfgðT vs;k Þ
vs;k−1
¼ ðUAÞk  ΔT lm;k ; k∈f2; 3g:
For the remaining effects, save the condenser, that are empowered
by the additional vapor injected from the flashing chambers, the energy
balance is described by Eq. (3) below. In this equation, Tvs,k − 1,mix, is
interpreted as the mixing temperature of the injected vapor and the
vapor from the immediate upstream MED effect (viz. the vapor that
comes from the (k-1)th effect).


Q̇ k ¼ ṁv;k−1;mix  hfgðT
Þ ¼ ṁ f ;k  C p; f  T b;k −T f ;k þ ṁv;k  hfgðT vs;k Þ
vs;k−1;mix
¼ ðUAÞk  ΔT lm;k ; k∈f4; ::; ng
ṁv;k−1;mix ¼ ṁv;k−1 þ ṁv;FC;i ; i∈f1; ::; jg :
In the above equations, ‘i’ is the flashing chamber stage number and
‘j’ is the total number of flashing stages; mv;FC;i is the amount of vapor
generated from each flashing stage which will be accounted for in the
ṁv;n;mix ¼ ṁv;n þ ṁv;FC; j : ð40 Þ
The amount of feed seawater introduced into the series of flashing
chambers is predicated on the amount of energy obtained through the
liquid-to-liquid heat exchanger and the corresponding cold side tem-
ð2Þ perature difference. This feed seawater is primarily comprised of the
recycled brine that emanates from the last flashing chamber, with the
balance being made up of preheated cooling seawater. This scheme
will expectedly decrease the consumption of chemical additive, the
size of pretreatment facilities, and the amount of non-condensable
gas. The ratio between the make-up water and the brine recirculation
mass flow rate is set based on its impact on the BPE as in the first flash-
ing chamber, which is considered to be typically no more than 0.65 °C in
the present simulation, and the mixing temperature at the inlet to the
liquid-to-liquid heat exchanger. A lower ratio decreases the amount of
non-condensable gas, but the resultant higher BPE can give rise to unde-
sirable pressure (or saturated temperature) gradient in the downstream
ð3Þ flashing chambers, which leads to lower energy recovery in the primary
MED plant. In all the simulations this ratio is varied between 3 and 10%.
ð30 Þ The amount of vapor so generated from each stage of flashing can be
calculated from the equation below


ṁrb;i;in h f ;rbðT Þ −h f ;rbðT
b;i;out Þ
ṁv;FC;i ¼
b;i;in
 ð5Þ
hgðT Þ −h f ;rbðT
vs;FC;i Þ b;i;out

following Eq. (5); and ‘n’ is the total number of MED effects. Tvs is the
saturated vapor temperature in MED effects and a function of pressure that
and is less than the boiling temperature (Tb) as per Eq. (3')

T vs;k ¼ T b;k −BPE: ð300 Þ T vs;FC;i ¼ T b;i;out −ðBPE þ NEAÞ: ð50 Þ

206 B. Rahimi et al. / Desalination 351 (2014) 202–212

hf,rb is the saturated liquid enthalpy of recycled brine that comes Table 2
from Eq. (A.2). NEA or non-equilibrium allowance for flashing chambers Production rate for different heat source inlet temperatures.

is a function of flashing temperature range, saturation temperature, Heat source inlet temp. (°C) 65 70 75 80 85 90
mass flow rate of brine per unit of chamber width, brine level inside Conventional MEDa freshwater yield 510 723 696 914 1450 1984
the flashing chamber, and the design of the flash chamber such as cham- (m3/day)
ber length, width and the orifice type of the brine transfer device from Number of effects (max.) 5 6 8 9 9 9
chamber-to-chamber (Appendix B) [28]. Under typical operating condi- Optimized conventional MED 622 910 1222 1578 1961 2377
tions, NEA varies between 0.03 °C (as in the first few stages) and 0.8 °C freshwater yield (m3/day)
(as in the last low-temperature stages) in an MSF desalination plant Number of effects 3 4 5 5 6 6
[29]; in the present simulations, NEA is varied from 0.3 °C (as in the Boosted MED freshwater yield (m3/day) 812 1150 1528 1948 2402 2861
first chamber) to 0.6 °C (as in the last chamber), but the results that Number of effects/injected effect's 5/5 6/6 6/6b 7/6 8/7 8/7
we present here pertains to a uniform 0.6 °C NEA for all flashing number

chambers so as to be conservative. Flash boosted MED freshwater yield 874 1289 1826 2339 2810 3284
The energy balance for the liquid-to-liquid heat exchanger is written (m3/day)
c
Number of MED effects/flashing stages 5/3 6/4 8/6 9/7 9/7 9/7
as
a
This refers to a conventional MED system with the maximum nunber of effects but not


optimized.
Q̇ hex ¼ ṁh hex  C p;h hex  T h hex;in −T h hex;out b

 This particular configuration is represented in Fig. 2, so that the steam from the booster
¼ ṁc hex  C p;c hex  T c hex;out −T c hex;in ¼ ðUAÞhex  ΔT lm;hex : ð6Þ unit is injected into the sixth effect of the primary MED plant.
c
This particular configuration is represented in Fig. 3, so that the six flashing stages are
matched with the fourth effect, fifth effect right down to the condenser, respectively of the
The total amount of freshwater generated is the sum total of the primary MED plant.
condensate from each of the MED effects:
has been compared. The first item in this table represents the maximum
Xn Xj
Ṁtotal ¼ ṁ þ
k¼1 v;k
ṁ
i¼1 v;FC;i
: ð7Þ number of effects that a conventional MED system can have, based on
its heat source inlet temperature, and subject to the assumptions in
Table 1. This maximum number of effects also sets the limit for the num-
The GRC method (generalized reduced gradient) [13,30] is used to
ber of effects in the primary MED plant for both the boosted and flash
solve Eqs. (1) to (7) by imposing the boundary conditions from the
boosted MED systems. The number of effects for the last three schemes
law of thermodynamics and appropriate operational, technical and eco-
as in Table 2 (namely optimized conventional, boosted MED and flash
nomic constraints as shown in Table 1.
boosted MED) generally increases with temperature, so as to maximize
freshwater production. Subject to the same heat source inlet tempera-
5. Results and discussions
ture, the heat source temperature drop across the primary MED plant,
for the three schemes presently considered, is generally different so
5.1. Production rate and thermal energy consumption
that each technology maximizes its respective freshwater production.
Specifically, the heat source temperature drop across the primary
The present flash boosted MED scheme is benchmarked against the
plant is generally the smallest for the flash boosted MED scheme, as
standard conventional MED and the boosted MED processes over a wide
the remaining heat source enthalpy can be better exploited at the flash-
range of heat source inlet temperatures germane to low grade process
ing chambers. This is then followed by the boosted MED process, where-
heat.
by the steam booster unit can still further process the remaining waste
Table 1 delineates all the assumptions inherent in the present simu-
heat. The largest heat source temperature drop is verily found across the
lation. The heat source medium is water and its flow rate has been fixed
optimized conventional MED plant, as this is the only avenue to utilize
at 100 kg/s in all simulations. The inlet seawater temperature at the
the waste heat. While more effects can certainly be built into the con-
condenser is 28 °C and the salinity of the seawater is 35,000 ppm. The
ventional MED system at any particular heat source inlet temperature,
top brine temperature is considered to be 70 °C, so that the top brine
as delineated in Table 2, this is done at the expense of freshwater pro-
temperature is capped at 70 °C. For the liquid-to-liquid heat exchanger
duction, as less heat via a temperature drop can be exploited [13]. Con-
the minimum temperature of approach is taken to be 3 °C. For vapor
sequently both the boosted and flash boosted MED systems can realize a
injection to happen from the flashing chamber to the judicious effect
successively larger number of effects than the conventional MED system
in the primary MED plant, a pressure difference equivalent to 3 °C satu-
for the same heat source inlet temperature. For clarity, in Fig. 1 for the
rated temperature difference is considered to be sufficient.
conventional MED plant, the important process temperatures stemming
Table 2 considers six different heat source inlet temperatures from
from the simulation are included. The same is done to both the boosted
65 °C to 90 °C. The performance of the flash boosted MED scheme
MED scheme as in Fig. 2, and flash boosted MED scheme as in Fig. 3.
with those of the optimized conventional and boosted MED processes
Fig. 4 presents the temperature–energy profiles of the heat source
medium of both conventional and Flash Boosted MED systems. For the
Table 1 conventional MED the heat source outlet temperature is around 10 °C
Assumptions for all simulations. higher than that of the FB-MED system, although the temperature
drop over the 1st effect is higher. For both systems, the heat source tem-
Maximum top brine temperature (°C) 70
Heat source flow rate, ṁhs (kg/s) 100 perature decreases due to the sensible heat transfer. However for the
Heat source temperature (°C) 65, 70, 75, 80, 85, 90 feed, the heat transfer includes both sensible (increasing the feed inlet
Feed to vapor ratio 2.2 temperature from 38 °C (Table 1) to the relevant boiling point temper-
Cp,hs (kJ/kg ∙ K) 4.187 ature) and latent heat transfers (that relates to the evaporation at the
Tcond,in (°C)/Tcond,out (°C) 28/38
vapor temperature that is slightly lower than the boiling point temper-
ΔTinj a (°C) 3
Liquid-to-Liquid heat exchanger approach temperature (°C) 3 ature due to the BPE).
Xf (ppm), MED effects inlet 35,000 As shown in both Figs. 3 and 4, the first effect's higher heat source
Xb (ppm), MED effects outlet 64,200 outlet temperature of the FB-MED allows it to maintain more effects.
a
Saturated temperature difference between the flashed vapor and the relevant MED ef- The liquid–liquid heat exchanger can then exploit more energy from
fect (for steam injection purpose). the heat source and hence the outlet heat source temperature of the
 B. Rahimi et al. / Desalination 351 (2014) 202–212 207

80 80
∆Ḣt = 7.62 MW ∆Ḣt = 11.60 MW

70 70

Temperature (°C)

Temperature (°C)
60 60 Vapor Temperature
Profile
Vapor Temperature Profile

50 50 ∆Ḣs-l = 4.11 MW
∆Ḣs-l = 7.18 MW

Feed Temperature
40 Profile 40 ∆Ḣs-s = 7.12 MW

"Liq.-Liq. HEX" "Primary MED Secon"

30 30
Energy (∆Ḣ) Transferred, (MW) Energy (∆Ḣ) Transferred, (MW)

Fig. 4. Temperature profile for the 1st effect of the conventional MED (left) and both the 1st effect of the primary MED section and the liquid–liquid heat exchanger of the FB-MED system
(right) for 75 °C heat source inlet temperature configuration.

FB-MED is around 10 °C less than conventional MED. This exploited en- steam injection and this trend continues to the end of the process. As
ergy is transferred to the flashing chamber feed stream and gradually shown in Fig. 5 the total amount of released energy in the primary
released over the train of flashing chambers and injected into the proper MED section of the FB-MED system is more than that of the convention-
effects of the primary MED section of the FB-MED system and thereby al MED system and as mentioned before, this extra energy is realized
increasing the freshwater generation. through the flashing chambers that effectively process more energy
The trend of the energy transferred over all sections of both systems from the heat source by using the liquid–liquid heat exchanger (first
is shown in Fig. 5 (dashed line separates the two systems). Because of column as shown in Fig. 5 for FB-MED).
the higher temperature drop over the 1st effect of the conventional Fig. 5 also shows the UA value distribution for both conventional and
MED heat source (Fig. 4), the amount of energy released in the first flash boosted MED systems. For the FB-MED's liquid–liquid heat ex-
effect of the conventional MED is around 70% more than that at the changer, the heat transfer rate is around 7.12 MW (Figs. 4 and 5), coupled
1st effect of the FB-MED system at 75 °C configuration. The amount of with a constant approach temperature of 3 °C for both sides of the heat
released energy decreases from effect to effect for conventional exchanger (Table 1), 2.37 MW/°C UA value is required. Since the major
MED and for FB-MED till the 3rd effect (i.e. before injection occurs). part of heat transfer occurs in both the 1st effect and in the condenser
For FB-MED, starting from the 4th effect, the energy increases due to pertains to sensible–latent heat transfer instead of latent–latent heat

10 2.7
Optimized conventional MED Flash Boosted MED (FB-MED)
9 2.4

8 2.1
Energy transferred (MW)

7
1.8 UA Value (MW/ °C)
6
1.5
5
1.2
4
0.9
3
0.6
2

1 0.3

0 0.0

Effects number
Added energy by vapor injecon from the flash chambers to the primary MED effects (FB-MED)
Energy Transferred
UA Value

Fig. 5. Energy released and UA value of each effect of the conventional MED and the primary MED section of the Flash Boosted MED systems for 75 °C heat source inlet temperature
configuration.
208 B. Rahimi et al. / Desalination 351 (2014) 202–212

3.5 38%

43% 20%

Production Rate (1000 m3/day)

3.0

48% 22%
2.5

49% 23%
2.0
25%
1.5 42%
26%
40%
1.0 30%

0.5

0.0
65 70 75 80 85 90
Heat Source Inlet Temperature ( °C)

Opmized convenonal MED Boosted MED Flash Boosted MED

Fig. 6. Freshwater production rate comparison among the three systems for an assortment of heat source temperatures. The percentages on top of both the boosted and flash boosted MED
columns represent the increment of freshwater production over and above the conventional optimized MED process.

transfer as in other effects this results in a higher UA values for those ef- ongoing cost of thermal energy used. A waste–heat performance ratio
fects compared with that of the 1st effect and condenser. [13], or PRWH, as in Eq. (8), is used instead which encourages the maxi-
Fig. 6 highlights the characteristics of the flash boosted MED scheme mal use of the enthalpy of the low grade heat source relative to the heat-
in relation to the conventional and steam boosted MED processes. Over sink
the range of temperatures from 65 °C to 90 °C, the flash boosted scheme
is 10% to 20% better in terms of freshwater production than the boosted ṁd  Δhref ṁ  Δhref
PRWH ¼ ¼
d : ð8Þ
MED process. This potential is better discerned where the production ṁhs  Δhavail: ṁ  h
hs f ;hs;in −h f ;cond;in
increment rates of both the flash boosted MED and boosted MED
schemes are compared against the conventional MED system. The pro-
duction rate increment decreases gradually for the boosted MED pro- Δhref is the specific reference enthalpy of the distillate that is equal to
cess from around 30% at 65 °C heat source inlet temperature to close 2336 kJ/kg as an industrial benchmark [31]. Δhavail. represents the max-
to 20% at 90 °C. For the flash boosted MED scheme, on the other hand, imum exploitable energy of the heat source relative to the lowest avail-
the production rate increment begins at around 40% at 65 °C and able temperature which in this case is the condenser inlet temperature.
peaks at about 49% at 75 °C, and then pares back to around 38% at Fig. 7 shows the waste–heat performance ratio for all three systems.
90°C. The reduction is enjoined by the top brine temperature constraint From this figure, the waste–heat performance ratio of the flash boosted
of 70 °C at the first effect of the primary MED plant and forces the num- MED experiences a substantial jump at 75 °C. This mirrors the boost in
ber of primary MED effects to remain the same for heat source temper- production rate at 75 °C as shown in Fig. 6.
atures higher than 80 °C. Insofar as low grade heat applications are
ṁ d Δhref 5.2. Pumping power consumption
concerned, the conventional performance ratio or ṁhs ðh f ;hs;in −h f ;hs;out Þ
which holds that heat comes with a premium as it is consumed does All the above mentioned comparisons were related to the produc-
not capture the essence of a desalination system driven by such heat tion rate and thermal energy consumption, another important aspect
sources, which only attracts a one-off investment cost, instead of an is the electrical power consumption that stems primarily from pumping

3.5 38%
Waste Heat Performance Ratio (PR WH)

43%
20%
3.0 48%
22%
49% 23%
2.5
42% 25%
2.0 26%
40%
1.5 30%

1.0

0.5

0.0
65 70 75 80 85 90
Heat Source Inlet Temperature ( °C)

Opmized convenonal MED Boosted MED Flash Boosted MED

Fig. 7. Waste Heat Performance Ratio (PRWH) comparison among the three systems and the percentage of increment of boosted and flash boosted MED's performance ratios compare to
the optimized conventional MED for an assortment of heat source temperatures.
 B. Rahimi et al. / Desalination 351 (2014) 202–212
power consumption in desalination plants. Pumping power is calculat-
ed according to the equation below:


ΔP ðkPaÞ V̇ m3 =s
Pumping Power ðkWÞ ¼ : ð9Þ
ηpump  ηmotor
ηpump and ηmotor are the pump and motor efficiencies and equate to
0.7 and 0.9, respectively in the present simulation. As schematically
shown in Figs. 1 and 3, in conventional and flash boosted MEDs, the
pumping power accounts for the following:
(1) Saline water pump: This pump conveys saline water through the
condenser of the primary MED and feeds the MED effects. The
pressure drop across the condenser is considered to be 0.5 bar.
The overall pressure difference in the saline water pump is
taken to be 1.5 bar (0.5 bar for the condenser and 1.0 bar to access
the atmosphere).
(2) Brine recirculation pump: This pump is used in the flash boosted
MED system to recirculate the brine through the brine heater
(liquid–liquid heat exchanger) and the flashing chambers.
Sommariva et al. [32] showed the brine recirculation pump head
contributions in the Al-Taweelah B plant that are related to the
tube bundles, water boxes, brine heater, spray pipe, control valves,
pipeline, geodetical and pressure heads. All the pressure drops,
save those associated with the tube bundles and water boxes
which are absent in the present flashing chamber section, and
the reported pressure head, are adopted from [32]; this according-
ly amount to 1.7 bar [32]. The germane pressure head is calculated
based on the flashing chamber inlet and outlet saturated pressure
difference.
(3) Distillate extraction and brine blowdown pumps: These pumps are
used to extract the distillate (freshwater) and brine from the pri-
mary MED plant. For these pumps a 2 bar differential pressure is
considered sufficient.
(4) Heat source medium pump: This pump is used for pumping the
heat source liquid through the evaporator (first effect) of the pri-
mary MED plant and the liquid–liquid heat exchanger (for the
flash boosted MED). A 0.5 bar drop in the evaporator and a
0.5 bar in the liquid–liquid heat exchanger have been considered.
The total pressure drop for the conventional MED is accordingly
0.5 bar that is related to the evaporator's pressure drop.
(5) Drain pump: The drain pump is used for brine blowdown from the
brine recirculation stream of the flash boosted MED. The pumping
power of this stream is negligible because of the small flow rate.
However a 2 bar pressure differential has been considered for
this pump for pumping from vacuum to the atmospheric pressure.
(6) Make-up water pumps: The make-up water pump is used to pro-
vide sufficient pressure to the make-up water that exits the deaer-
ator. A 2 bar differential pressure has been considered for this
pump.
(7) NCG extraction vacuum pump: One of the important issues for low
grade heat driven desalination processes is related to the venting
process, which serves to remove the generated non-condensable
gasses (NCG) and the air leakage into the system. In the steam
heat driven desalination process, such as MED/TVC or MSF plants,
steam ejectors are used for vacuuming and non-condensable gas
removal; but in low grade heat driven application there is no pres-
surized steam for driving ejectors, and so two options are available.
The first option is to use water eductors or air ejectors, and the sec-
ond is to use water ring vacuum pumps. In the present simulation,
the second option is adopted and the corresponding pumping
power consumption is evaluated according to the manufacturer's
catalogs for water-ring vacuum pumps [33]. The amount of non-
condensable gas released is a function of feedwater composition,
evaporator design and working conditions. Literatures expounding
the methods for estimating the amount of non-condensable gasses
209
abound [34–36], but in terms of design, it is typical to consider an
over designed venting system for handling the air leakage into the
system [37]. As a matter of comparing between the conventional
MED and flash boosted MED, a cascaded venting scheme is consid-
ered for both systems which means that the non-condensable gas-
ses are extracted from each effect and cascaded downstream and
eventually into the condenser where all accumulated NCGs are ex-
tracted by the vacuum pump. On account of the low top brine tem-
perature and the make-up/brine recirculation ratio of the present
flashing chambers and also the use of deaerated make-up water,
the amount of generated NCGs in the present flashing chambers
as compared to the released NCGs in the primary MED effects is
relatively low. In the present simulation the amount of NCG re-
leased in the flashing chambers is considered based on the ratio
between the make-up water of the flashing chambers and the
feedwater of the primary MED plant that is varied between 15
and 25%. Therefore a venting system is considered for handling
all NCGs released from both the MED plant and flashing chambers.
Following routine considerations the venting system is designed to
remove 1% of the vapor in the condenser as NCGs [38].
The normalized pumping power consumptions of the flash boosted
MED as compared to conventional MED are shown in Fig. 8. It demon-
strates that the auxiliary power consumption of flash boosted MED is
well within the range of 22 to 34% more than that of the optimized con-
ventional MED system.
5.3. Capital cost analysis
Besides the operating cost as represented by the pumping power,
the overall capital cost of the flash boosted MED is a key factor for any
financial decision-making process. Since obviously the actual capital
costs depend on various factors, including material and equipment
specifications, the focus herein is to analyze the relative capital cost in-
crease of flash boosted MED against conventional MED plants and its
impact on the specific plant costs.
For conventional MED plants, the capital cost can be approximated
based on the GWI/IDA database [39]. For the capacity range of 100 to
10,000 m3/day, this results in the following cost function
0:9795
Capital CostMED ¼ ΨDt ¼ 3018:8  Dt ð10Þ
where Dt is the total production rate of a conventional MED plant in
m3/day.
The capital cost increase associated with a flash boosted system can
be evaluated based on the additional expenditures for the added units,
namely:
• Flashing chambers.
The flashing chambers resemble a major part of an MSF process.
Adopting the costing trend as provided by the GWI/IDA database
[39] for MSF plants in the range of 100 to 1000 m3/day, yields the fol-
lowing function:
0:9878
Capital Cost MSF ¼ ΦDFC ¼ 2867:7  DFC ð11Þ
with DFC being the equivalent plant capacity, determined according to
j
the total produced vapor from the flashing chambers, ∑mv;FC;i. Howev-
i¼1
er, in contrast to MSF plants, no condenser tubing and no heat rejection
section is required for the flashing chambers, which are major cost
items in thermal desalination plants [40]. Sommariva, et al. [41,42]
subdivided the capital costs for MSF and MED systems into five main
sectors: Capital cost of the evaporator (40% of the total capital cost),
equipment piping (29%), erection (14%), engineering and commission-
ing (10%), and electrical, instrumentation and control (7%). Following
210 B. Rahimi et al. / Desalination 351 (2014) 202–212

3.5 3.28

FB-MED
Normalized Pumping Power (kWh/m 3)
3.0
2.54 2.56

FB-MED
2.5
2.05
2.0 1.91

FB-MED
1.78
1.61

FB-MED
1.53 1.46 1.48

FB-MED
1.5

FB-MED
1.22 1.19

Convenonal MED

Convenonal MED

Convenonal MED

Convenonal MED

Convenonal MED
Convenonal MED
1.0

0.5

0.0
65 70 75 80 85 90
Heat Source Inlet Temperature ( °C)

Saline Water Intake Pump Heat Source Pump Brine Blowdown Pump Disllate Extracon Pump
Vacuum Pump Drain Pump Make-up Water Pump Brine Recirculaon Pump

Fig. 8. Normalized pumping power (kWh/m3) vs. heat source inlet temperature (°C) of the optimized conventional MED and the flash boosted MED processes.

the further cost breakdown for the evaporator section of MSF plants as Following this correlation, the flash boosted MED scheme gives rise to
explained in [40], approximately 30% of the evaporator costs are associ- between 32% and 41% higher total capital costs as compared with con-
ated with the flashing chambers. The capital cost for the flashing cham- ventional MED units (Fig. 9). This is however outweighed by the re-
bers can therefore be estimated with: sultant significantly higher production capacity, so that the specific
capital cost ratio (Eq. (15)) is actually reduced by around 4 to 6%
Capital Cost FC ¼ ð0:30  0:40Þ  ΦDFC þ 0:29  ΦDFC ¼ 0:41  ΦDFC : ð12Þ over the application range, indicating a favorable influence over the
final water cost by utilizing the heat source down to a much lower
• Increase in MED plant size due to steam injection. thermodynamic potential
Besides the flashing chambers, another major cost component
comes from the required up scaling of the primary MED plant so Capital Cost Ratio
Specific Capital Cost Ratio ¼ : ð15Þ
as to accommodate the additional vapor injection and allow for en- Production Rate Ratio
ergy recovery in the downstream effects from the resultant boost.
This can be analyzed with the full cost function of an MED plant ac-
cording to Eq. (10) by considering the capacity of the flash boosted 6. Conclusion
MED system, save the vapor generated in the flashing chambers.
For the vapor injected from the flashing chambers, only condensa- A novel flash boosted multi-effect distillation desalination process
tion takes place in the primary MED plant. No additional evapora- has been developed for the maximal exploitation of the potential of
tion surface needs to be considered for the vapor generation in low grade heat. Its performance has been further benchmarked with re-
the primary MED plant, insofar as capital cost analysis is con- spect to conventional and the recently reported boosted multi-effect
cerned. By assuming similar overall heat transfer coefficients for
50% Capital Cost Increment
both evaporation and condensation processes [43], only half of
the typical heat transfer area to process the additional vapor injec- 45% Specific Capital Cost Decrement
tion is therefore needed, which is then translated to a 50% discount 40%
of the earlier introduced cost breakdown for the evaporator
Percentage (Increment/Decrement)

35%
(namely 50% of 40% for the cost of evaporator). This overall factor
is then applied to the capital cost differential between an MED 30%
plant with a production capacity of an FB-MED plant and an MED
25%
plant with a production capacity of an FB-MED plant less the total
vapor injection rate from the flashing chambers 20%

15%
Capital Cost PrimaryMED ¼ ΨDt;FBMED −DFC þ ð0:50  0:40Þ

 10%
 ΨDt;FBMED −ΨDt;FBMED −DFC : ð13Þ 5%

0%
With other cost functions remaining essentially unaffected from the
process modification, combining Eqs. (12) and (13), the total capital -5%
cost of the flash boosted MED can therefore be approximated with -10%
65 70 75 80 85 90
h
i
Heat Source Inlet Temperature ( °C)
Capital Cost FBMED ¼ ΨDt;FBMED −DFC þ 0:20  ΨDt;FBMED −ΨDt;FBMED −DFC
Fig. 9. Capital cost increment and specific capital cost ($/(m3/day)) reduction of the flash
þ 0:41  ΦDFC : ð14Þ
boosted MED process as compared to a conventional MED scheme.
 B. Rahimi et al. / Desalination 351 (2014) 202–212 211

distillation processes [11,13], in terms of freshwater production and Which is valid for 10≤Tb ≤ 120°C and 0 ≤ X ≤ 0.12kg/kg range with
waste–heat performance ratio [13]. The flash boosted MED scheme an accuracy of ±0.5%.hfw is the saturated enthalpy of pure water:
has been demonstrated to be around 40% to 50% better than the opti-
2
mized conventional MED process over the entire practical range of h f ;w ¼ 141:355 þ 4202:07  T b þ 0:004
waste heat temperatures and the relevant assumptions as in Table 1. A 3 ∘
 T b ; 5≤T b ≤200 C and an accuracy of  0:02%:
capital cost analysis indicates a 4 to 6% decrement for the specific capital
cost of this system over the conventional systems. The normalized electri-
And the constants are as below.
cal power consumption of a flash boosted MED is 22% to 34% more than
that of the optimized conventional MED process. The flash boosted 4 5 6
a1 ¼ −2:348  10 ; a2 ¼ 3:152  10 ; a3 ¼ 2:803  10 ;
MED process is able to exploit waste heat significantly beyond the prom-
7 3 1
ise of conventional and even boosted MED systems. a4 ¼ −1:446  10 ; a5 ¼ 7:826  10 a6 ¼ −4:417  10 ;
−1 4 4 1
a7 ¼ 2:139  10 ; a8 ¼ −1:991  10 ; a9 ¼ 2:778  10 ; a10 ¼ 9:728  10

Subscripts
b brine
Appendix B. BPE and NEA
c_hex cold side of the liquid–liquid heat exchanger
cond condenser
Boiling Point Elevation [27]:
d distillate
f feedwater 2
FBMED flash boosted MED system BPE ¼ a  X þ β  X
−4 2 −1
FC flashing chamber a ¼ −4:584  10  T b þ 2:82:3  10  T b þ 17:95
−4 2 −2
hex liquid-to-liquid heat exchanger β ¼ 1:536  10  T b þ 5:267  10  T b þ 6:56
hs heat source
h_hex hot side of the liquid–liquid heat exchanger The vadlidity of equation (B.1) is for 0 ≤Tb ≤ 0.12 kg/kg range with
i flashing chamber number an accuracy of ±0.018K.
in inlet Non-Equilibrium Allowance (NEA) [28]:
inj injection
T H −6
j total number of flashing stages NEA10 ¼ ð0:9784Þ vs  ð15:7378Þ  ð1:3777ÞW  3600  10
k MED effect number W ¼ 105:6 þ 18:06  DFC
MED conventional MED process Xj
ðFor the flashing chamber section : DFC ¼ ṁ Þ
i¼1
MSF conventional MSF process v;FC;i
 0:3281L
mix mixed streams NEA10
n the total number of MED effects NEA ¼ ð0:5 ΔT st þ NEA10 Þ 
0:5  ΔT st þ NEA10
out outlet
PrimaryMED primary MED section of the flash boosted MED system
rb recycled brine (through the flashing chambers) References
s-l sensible–latent
s-s sensible–sensible [1] I.A. Shiklomanov, J.C. Rodda, World water resources at the beginning of the twenty-
first century, International Hydrology Series 2003, Cambridge University Press,
t total 2003.
v vapor [2] H. Cooley, P.H. Gleick, G. Wolf, Desalination, With a Grain of Salt, Oakland, CA 94612,
vs saturated vapor 2006.
[3] M.A. El-Tawil, Z. Zhengming, L. Yuan, A review of renewable energy technologies
w water (pure water) integrated with desalination systems, Renew. Sust. Energ. Rev. 13 (2009)
WH waste heat 2245–2262.
[4] A.D. Khawaji, I.K. Kutubkhanah, J.M. Wie, Advances in seawater desalination tech-
nologies, Desalination 142 (2008) 47–69.
[5] L. Henthorne, T. Pankratz, S. Murphy, The state of desalination 2011, IDA World
Acknowledgment Congress on Desalination and Water Reuse, Desalination: Sustainable Solutions for
a Thirsty Planet, 2011.
[6] T. Pankratz, IDA Desalination Yearbook 2012–2013, Media Analytics Ltd., Oxford,
We gratefully acknowledge the financial support of the Western 2013.
Australian Geothermal Centre of Excellence and the National Centre [7] Y. Ammar, S. Joyce, R. Norman, Y. Wang, A.P. Roskilly, Low grade thermal energy
of Excellence in Desalination Australia (NCEDA) which is funded by sources and uses from the process industry in the UK, Appl. Therm. Eng. 89 (1)
(2012) 3–20.
the Australian Government through the National Urban Water and
[8] Y. Ammar, H. Li, C. Walsh, P. Thornley, V. Sharifi, A.P. Roskilly, Desalination using low
Desalination Plan. grade heat in the process industry: challenges and perspectives, Appl. Therm. Eng.
48 (2012) 446–457.
[9] H. Shih, Evaluating the technologies of thermal desalination using low-grade heat,
Appendix A. Enthalpies
Desalination 182 (2005) 461–469.
[10] A. Ophir, F. Lokiec, Advanced MED process for most economical sea water desalina-
Latent heat of evaporation [44]: tion, Desalination 182 (2005) 187–198.
[11] X. Wang, A. Christ, K. Regenauer-Lieb, K. Hooman, H.T. Chua, Low grade heat driven
multi-effect distillation technology, Int. J. Heat Mass Transf. 54 (25–26) (Dec. 2011)
−3 2 5497–5503.
hfgðTvsÞ ¼ 2499:5698−2:204864  T vs −1:596  10  T vs ðA:1Þ [12] A. Christ, X. Wang, K. Regenauer-Lieb, H.T. Chua, Low grade waste heat driven desa-
lination technology, Int. J. Simul. Multidiscip. Des. Optim. 5 (2014) A02.
[13] A. Christ, K. Regenauer-Lieb, H.T. Chua, Thermodynamic optimisation of multi-
effect-distillation driven by sensible heat sources, Desalination 336 (2014) 160–167.
Seawater enthalpy [27] [14] A.M. El-Nashar, Desalination with renewable energy — A review, Desalination and
Water Resources (DESWARE), Renewable Energy Systems and Desalination, Vol. 1,
2 3 2 EOLSS Publisher, 2010, pp. 88–160.
h f ;rb ¼ h f ;w −X:ða1 þ a2 :X þ a3 :X a4 :X þ a5 :T b þ a6 :T b ðA:2Þ [15] V. Belessiotis, E. Delyannis, Renewable energy resources, Desalination and Water
3 2 2 Resources (DESWARE), Renewable Energy Systems and Desalination, Vol. 1,
þa7 :T b þ as :X:T b þ a9 :X :T b þ a10 :X:T b Þ EOLSS Publisher, 2010, pp. 49–87.
212 B. Rahimi et al. / Desalination 351 (2014) 202–212
[16] A.M. El-Nashar, Why use renewable energy for desalination, Desalination and Water
Resources (DESWARE), Renewable Energy Systems and Desalination, Vol. 1, EOLSS
Publisher, 2010, pp. 202–215.
[17] E. Barbier, Geothermal energy technology and current status: an overview, Renew.
Sust. Energ. Rev. 6 (2002) 3–65.
[18] K. Karytsas, V. Alexandrou, I. Boukis, The Kimolos geothermal desalination project,
International Workshop on Possibilities of Geothermal Energy Development in the
Aegean Islands Region, Milos Island, Greece, 2002. 206–219.
[19] L. Garcia-Rodriguez, Seawater desalination driven by renewable energies: a review,
Desalination 143 (2002) 103–113.
[20] A.M. El-Nashar, Economics of small solar-assisted multiple-effect stack distillation
plants, Desalination 130 (2000) 201–215.
[21] E. Mathioulakis, V. Belessiotis, E. Delyannis, Desalination by using alternative
energy: review and state-of-the-art, Desalination 203 (2007) 346–365.
[22] A. Husain, Renewable energy and desalination systems, Desalination and Water
Resources (DESWARE), Renewable Energy Systems and Desalination, Vol. 1,
EOLSS Publisher, 2010, pp. 161–201.
[23] H.T. Chua, X. Wang, K. Regenauer-Lieb, A desalination plant, World Intellectual
Property Organization (WIPO), International Publication Number: WO 2012/
003525 A1, 2012.
[24] A. Christ, K. Regenauer-Lieb, H.T. Chua, Development of an advanced low-grade heat
driven multi effect distillation technology, IDA World Congress on Desalination and
Water Reuse, Tianjin, China, 2013.
[25] B. Rahimi, A. Christ, H.T. Chua, System and method for desalination, Australian Pro-
visional Patent Application AU 2014901335, 2014.
[26] Single Effect Freshwater Generator, Model JWP-16/26-C Series, Alfa Laval marine &
diesel product catalogue, Alfa Laval Corporate AB, 2003.
[27] M.H. Sharqawy, J.H. Lienhard, V, S.M. Zubair, Thermophysical properties of seawa-
ter: a review of existing correlations and data, Desalin. Water Treat. vol. 16 (no.
10) (2010) 354–380.
[28] H. El-Dessouky, H.I. Shaban, H. Al-Ramadan, Steady-state analysis of multi-stage
flash desalination process, Desalination 103 (1995) 271–287.
[29] P. Fiorini, E. Sciubba, C. Sommariva, A new formulation for the non-equilibrium al-
lowance in MSF processes, Desalination 136 (2001) 177–188.
[30] L.S. Lasdon, A.D. Waren, A. Jain, M. Ratner, J. Rice, Design and testing of a generalized
reduced gradient code for nonlinear programming, ACM Trans. Math. Softw.
(TOMS) 4 (1) (March 1978) 34–50.
[31] L. Awerbuch, Understanding of Thermal Distillation Desalination Processes, IDA
Academy, Singapore, 2012.
[32] C. Sommariva, R. Borsani, M.I. Butt, A.H. Sultan, Reduction of power requirements for
MSF desalination plants: the example of Al Taweelah B, Desalination 108 (1996)
37–42.
[33] Robuschi Liquid Ring Vacuum Pumps Catalogue, GARDNER DENVER S.r.l., Robuschi
Pumps and Blower Manufacture, Italy, www.robuschi.com 2013.
[34] K. Genthner, A. Seifert, A calculation method for condenser in multi-stage evapora-
tors with non-condensable gases, Desalination 81 (1991) 349–366.
[35] H. Glade, J. Meyer, S. Will, The release of CO2 in MSF and ME distillers and its use
for the recarbonation of the distillate: a comparison, Desalination 182 (2005)
99–110.
[36] A.E. Al-Rawajfeh, H. Glade, H.M. Qiblawey, J. Ulrich, Simulation of CO2 release in
multiple-effect distillers, Desalination 166 (2004) 41–52.
[37] M.A. Darwish, M.M. El-Refaee, M. Abdel-Jawad, Developments in the multi-stage
flash desalting system, Desalination 100 (1995) 35–64.
[38] A. Seifert, K. Genthner, A model for stagewise calculation of non-condensable gases
in multi-stage evaporators, Desalination 81 (1991) 333–347.
[39] Desaldata.com, IDA Desalting Plants Inventory, Media Analytics Ltd., 2011.
[40] C. Sommariva, H. Hogg, K. Callister, Cost reduction and design lifetime increase
in thermal desalination plants: thermodynamic and corrosion resistance com-
bined analysis for heat exchange tubes material selection, Desalination 158
(2003) 17–21.
[41] C. Sommariva, H. Hogg, K. Callister, Maximum economic design life for desalination
plant: the role of auxiliary equipment materials selection and specification in plant
reliability, Desalination 153 (2002) 199–205.
[42] C. Sommariva, Water Management and Economics, IDA Academy, Singapore, 2012.
[43] H.T. El-Dessouky, H.M. Ettouney, Fundamentals of Salt Water Desalination, 1st ed.
Elsevier Science B.V., 2002.
[44] H.T. El-Dessouky, I. Alatiqi, S. Bingulac, H.M. Ettouney, Steady-state analysis of the
multi effect evaporation desalination process, Chem. Eng. Technol. 21 (1998) 437–451.