Desalination 435 (2018) 114–127

Contents lists available at ScienceDirect

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

Design strategies of conventional and modified closed-air open-water T

humidification dehumidification systems
⁎
Samih M. Elmutasim, M.A. Ahmed, Mohamed A. Antar, P. Gandhidasan, Syed M. Zubair
Mechanical Engineering Department, KFUPM Box # 1474, King Fahd University of Petroleum & Minerals, Dhahran 31261, Saudi Arabia

G RA P H I C A L AB S T R A C T

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

Keywords: Humidification dehumidification (HDH) systems are robust and known to withstand a wide range of saline water
Desalination without the need of complex maintenance. In this study, the closed-air open-water (CAOW) HDH arrangement is
Humidification-dehumidification examined, which can easily be integrated with solar or other renewable resources. Also, this system is modified
Extraction by incorporating heat recovery options for low performing units. The heat recovery process is executed through
Heat recovery
two approaches, (i) a mixing chamber and, (ii) by a heat exchanger. Thermal balancing through air extraction is
Enthalpy pinch
evaluated for the basic as well as the modified cycles, particularly for high performing units. Zero, single and
double extractions models are evaluated for the conventional CAOW water heated cycle. The performance of the
systems is characterized into three operating regions (high, moderate and low) based on the effectiveness of the
components. Based on the results of this evaluation, an operating scheme is then developed to decide where to
use the conventional or the modified systems with or without extraction.

1. Introduction mechanical and electrical techniques. One of the promising thermal

techniques is the humidification-dehumidification (HDH) desalination
There is about 40% of the world population that is suffering from system. The HDH desalination system is suitable for small-scale fresh-
the water shortage problems. It is expected to reach 60% by 2025 [1]. A water production. Thus, it is more suitable for villages and small
large portion of the world population lives within 70 km of seashores communities due to its simplicity. These systems are considered to have
[2], which qualifies for industrial desalination as a promising solution some advantages over other desalination technologies such as their
to this crisis. Industrial desalination is classified into thermal, capacity to operate over a wide range of untreated water quality with

⁎
Corresponding author.
E-mail address: smzubair@kfupm.edu.sa (S.M. Zubair).

https://doi.org/10.1016/j.desal.2017.11.011
Received 8 July 2017; Received in revised form 9 October 2017; Accepted 6 November 2017
Available online 28 November 2017
0011-9164/ © 2017 Elsevier B.V. All rights reserved.
S.M. Elmutasim et al. Desalination 435 (2018) 114–127

Nomenclature T temperature (°C)

Acronyms Greek

GOR gained output ratio Δ change

HDH humidification-dehumidification ε effectiveness (−)
HX heat exchanger Ψ enthalpy pinch (kJ/kg dry air)
MX mixing chamber ω absolute humidity (kg water vapor per kg dry air)
CAOW closed-air open-water φ relative humidity (−)

Symbols Subscripts

ṁ mass flow rate (kg/s) 0, 1, … state points

Mr water-to-air mass flow rate ratio (−) a air
Q̇ heat transfer rate (kW) b brine
h specific enthalpy (kJ/kg) w saline water
h∗ specific enthalpy (kJ/kg dry air) cw cooling water
hfg specific enthalpy of vaporization (kJ/kg) deh dehumidifier
RR recovery ratio (%) hum humidifier
m slope of humidifier and dehumidifier line (C·kg dry air/kJ) hx heat exchanger
C intersection of humidifier and dehumidifier line (°C) in entering
cp specific heat capacity at constant pressure (kJ/kg·K) pw pure water
tan correspondent point at humidifier line to the tangent point mix mixer
at air curve ext extraction point
tan′ tangent point at air curve b brine
X salinity (ppm)

minimum maintenance requirements [3]. The basic drawbacks of HDH consumption decreased to 800 kJ/kg. In their system, the air was ex-
systems remain that the total heat input is relatively high compared to tracted from two points in the humidifier and injected to the dehumi-
other conventional thermal desalination technologies; however, re- difier. They followed enthalpy-temperature diagrams, as used in several
newable energy as a source of heat input can be utilized in these sys- other publications [4,6,9,13], to illustrate the extraction impact on the
tems. design of HDH systems.
Müller–Holst [4,5] proposed to vary the water-to-air mass flow rate Thermal balancing by extracting air or water from the humidifier
ratio continuously in order to achieve thermal balancing of HDH sys- and injected it into the dehumidifier or vice versa has been investigated
tems. This variation will decrease stream-to-stream temperature dif- by Narayan et al. [14]. Mistry et al. [15] found that reducing the spe-
ference. He made use of natural convection to circulate the moist air cific entropy would result in minimizing the GOR. Miller et al. [16]
stream through ports in both the humidifier and dehumidifier. This studied the effects of extraction on balancing enthalpy rates in HDH
circulation resulted in variation of the water-to-air mass flow rate ratio. systems. They followed an effectiveness-based methodology. Their
After optimization, the system's specific energy consumption was main conclusion was that extractions are better for systems that have a
120 kWh/m3 (≈450 kJ/kg). Another novel approach is to vary the high effectiveness in both the humidifier and dehumidifier of an HDH
water-to-air mass flow rate ratio was introduced by Zamen et al. [6]. system.
They designed a multi-stage process, in which, HDH processes are The variation of temperature pinch effect on both the recovery ratio
executed in a sequence. The brine flow was common for all the stages, (RR) and gain output ratio (GOR) has been studied by McGovern et al.
while the air flow was separate for each stage. Schlickum [7] and Hou [13]. They showed an increase in GOR from 3.5 to 14 by incorporating
[8] reported a similar design. Zamen et al. [6] have defined the system single water extraction by assuming that effective heat and mass
by the temperature pinch approach, which is commonly used in process transfer area to be very large. That increase was achieved by using
industries. The total specific heat consumed by this system was about bottom- and top-cycle temperatures of 25 °C and 70 °C, respectively.
800 kJ/kg. The humidifier and dehumidifier both had a temperature Furthermore, for a single water extraction and under same operating
pinch of 4 °C, at a top- and bottom-cycle temperatures of 70 °C and conditions, they reported an increase in RR from 7% to 11%.
20 °C, respectively. Narayan et al. [17] defined a novel parameter called the enthalpy
Another novel HDH system driven by forced convection was in- pinch approach. They used this parameter to balance HME devices since
vented by Brendel [9,10]. Under the balanced temperature profiles, this parameter takes into account both heat and mass transfer processes
forced convection was used to extract water from the dehumidifier and that are occurring in HDH systems. Balanced systems that have zero
was injected to the humidifier. This extraction process was executed at extraction, one extraction and an infinite number of extractions were
several points in both the humidifier and dehumidifier. Thiel and studied using the enthalpy pinch approach. An increase in the GOR
Lienhard [11] have stated that the optimization of heat and mass from 2.6 to 4.0 for a system with single air extraction has been reported
transfer exchanger (HME) devices, thermodynamically requires con- in an experimental study by Narayan et al. [18]. In their experimental
sidering both the temperature and concentration profiles. They have study, the enthalpy pinch was 19 kJ/kg of dry air. The bottom- and top-
shown that balancing humidity profile have more significance in the cycle temperatures were 25 °C and 90 °C, respectively.
optimization of the system than balancing the temperature profile. Chehayeb et al. [19] in continuation of the previous work by
Forced convection driven HDH systems with air extraction and injec- Narayan et al. [17] have investigated the effect of extractions on the
tion have also been investigated by Younis et al. [12]. They have suc- GOR, RR, and the total heat input to the cycle. They examined a finite
ceeded to increase the system efficiency as the specific energy number of extractions and found that the smaller the enthalpy pinch,

115
S.M. Elmutasim et al. Desalination 435 (2018) 114–127

the larger would be the impact of balancing. That is, when the heat and 2. Performance and operating metrics
mass transfer areas decrease (large enthalpy pinch), the balancing loses
its significance. Furthermore, they found that balancing effect on water The gain output ratio (GOR), which is defined as the ratio of the
recovery is not significant compared to its effect on energy efficiency. latent heat of vaporization of fresh water to the amount of heat utilized
Chehayeb et al. [20] in another investigation studied the effect of ex- to produce it.
traction on a fixed-size HDH system with a two-stage HDH processes. A ṁ pw hfg
generalized energy effectiveness for HME devices was proposed in their GOR =
̇
Qin (1)
study. The model was constructed from a multi-tray bubble-column
dehumidifier and a packed-bed humidifier. The results pointed out that The effectiveness of dehumidifier, and humidifier, is expressed, re-
thermodynamic balancing maximizes both the GOR and water recovery spectively, as
while keeping entropy generation at minimum levels. Furthermore, it
Δh∗
was mentioned that the direction of extraction should always be from εdeh =
Δh∗ + Ψdeh (2)
the humidifier to dehumidifier to reach a balanced system.
It is important to note that the work in the literature regarding air
Δh∗
extraction has focused mainly on a water heated closed air open water εhum ≈
Δh∗ + Ψhum (3)
(CAOW) cycle as a basic configuration. This paper is intended to pro-
vide an analysis for this cycle combined with two modifications to close
the water loop. In this regard, the modified CAOW arrangement is 3. Conventional and modified cycles' description
proposed by using a heat recovery process. The heat recovery process
will be executed through two approaches, a mixing chamber, and a heat The HDH systems are generally simple and cheap. In addition, they
exchanger. Furthermore, the use of single- and double-air extractions may be fabricated by a low-skilled labor which may result in low ef-
and injections to thermodynamically balance the system following the fectiveness humidifier. Thus, the hot brine coming out of the humidifier
enthalpy pinch approach will be studied. (refer to Fig. 1) can be utilized to recover part of this heat instead of
throwing it away. This recovery can be attained either through mixing

Fig. 1. Water-heated CAOW HDH cycle with zero, single, and double extractions.

116
S.M. Elmutasim et al. Desalination 435 (2018) 114–127

Fig. 2. Water-heated CAOW HDH cycle after heat exchanger modification with zero, single, and double extractions.

part of the hot brine with the cooling water that leaves the dehumidifier 3.2. Modified closed-air open-water with heat recovery options
in a mixing chamber before the heater or, by installing a heat ex-
changer, as illustrated in Figs. 2 and 3. These modifications aim to re- The modification in this cycle includes adding a heat exchanger, as
duce the heat input and increase the GOR of the system. In this study, illustrated in Fig. 2 wherein the hot brine is piped to a heat exchanger
zero, single, and double air extractions will be evaluated for both the where it is used to heat the saline water coming from the dehumidifier
conventional and modified CAOW water heated cycles. before it enters the heater. The rest of the cycle remains the same as the
conventional cycle, which is explained earlier in the previous section.
In the other modified cycle as shown in Fig. 3, the heat recovery
3.1. Conventional closed-air open-water (CAOW), water heated cycle
process is carried out by mixing the hot brine with cooling water
coming out of the dehumidifier in a mixing chamber. To achieve this
In this arrangement as illustrated in Fig. 1 (with both valves closed)
and keep the salinity level steady, there is a need to have water from
saline water enters the dehumidifier as cooling water and warms up
two sources with different salinity levels. A low-salinity cooling water
before it is heated in a water heater or solar collector. The heated water
from the dehumidifier; that is, tap water (< 3000 ppm), and the saline-
is then sprayed in the humidifier which has a packing material to in-
water (40,000 ppm). Saline water is passed through the heater and then
crease the heat and mass transfer area that enhances the evaporation
sprayed in the humidifier where evaporation takes place. The re-
process, the remaining unevaporated brine is rejected. The air loop in
maining hot brine is then collected and part of it is circulated back to
this cycle is closed, it enters the humidifier at low humidity and leaves
the mixing chamber where it mixes with the make-up low-salinity
as hot and humid air. Thereafter, it enters the dehumidifier where it
cooling water to close the water loop. The mixed water is then passed
undergoes a cooling and dehumidifying process. The condensate water
through the heater under steady-state operating conditions. The air
is then collected as a fresh water at the bottom of the dehumidifier and
loop remains the same as the conventional CAOW; however, for single
air is circulated back to the humidifier to close the loop. In the case of
air extraction, the valve can be opened to allow humid air to flow into
single extraction, some of the air is extracted from the humidifier and
the dehumidifier. This cycle is shown schematically in Fig. 3 for zero,
injected into the dehumidifier by simply opening one of the valves, as
single and double extractions.
shown in Fig. 1, while the double extractions cycle is when the two
valves are opened.

117
S.M. Elmutasim et al. Desalination 435 (2018) 114–127

Fig. 3. Water-heated CAOW HDH cycle after mixing chamber modification with zero, single, and double extractions.

4. Modeling flow rate ratio (Mr). This process can be achieved by extracting air from
the humidifier and injecting it into the dehumidifier either at one point
The present study is following the enthalpy pinch model for HME or two points (single or double extractions) as illustrated in cycles de-
devices as illustrated by Narayan et al. [17]. The enthalpy pinch pro- scription in Section 3. Zero extraction system refers to the case when
cedure is a novel procedure introduced to study the effect of mass the extraction valves in all cycles (conventional and modified) are
transfer as well as the heat transfer in the humidifiers and dehumidi- closed, i.e. no extraction. Fig. 4 represents the concept for zero, single
fiers. This method is applied to all the aforementioned systems. For and double extractions processes, respectively. Visualization and
example, a temperature-enthalpy diagram for HDH systems with zero modeling of these systems depend on this figure. The idea is to convert
extraction is exhibited in Fig. 4a. The curved line represents the air these figures into a mathematical model to evaluate the performance of
process through the humidifier and dehumidifier. The solid line re- zero, single, and double extractions systems, as shown in Fig. 5. The
presents the water process through the dehumidifier and the dotted line inputs of the system are as follows: the cooling water temperature, the
through the humidifier. The enthalpy pinch of the dehumidifier can be top brine temperature, the humidifier enthalpy pinch, and the dehu-
expressed as, midifier enthalpy pinch. The enthalpy pinch method is briefly discussed
in the literature [4,6,9,13]. These studies recommended using this
Ψdeh = (ha1 − ha) (4)
method instead of temperature pinch.
where ha is calculated at Two. As the inputs are the same for the conventional and modified cycles,
While the enthalpy pinch of the humidifier is, ̇ . The
the only difference is observed in the calculation of heat input, Qin
process of calculating Qiṅ for CAOW is summarized in the following
Ψhum = (htan ′ − htan ) (5)
equations:
here htan' represents the enthalpy calculated at the tangent point at the
̇ = ṁ cw cp (Tw2 − Tw1).
Qin (6)
air saturation curve, and htan represents the enthalpy obtained at the
tangent point at the humidifier line, as illustrated in Fig. 4a.
For the CAOW with a heat exchanger,
4.1. Zero extraction system ̇ = ṁ cw cp (Tw2 − Tw1, HX )
Qin (7)

Thermal balancing in HME devices is attained through varying mass and for the CAOW with a mixing chamber,

118
S.M. Elmutasim et al. Desalination 435 (2018) 114–127

ṁ cw, mix + ṁ b, mix = ṁ w (10)

m˙ cw, mix X cw + m˙ b, mix Xb = m˙ w Xw (11)

ṁ cw, mix h w1 + ṁ b, mix h w3 = ṁ w h w1, mix (12)

4.2. Single extraction system

The major variation in the single-extraction model is that thermal

balancing process is executed in two stages. The algorithm for the single
extraction model is illustrated in Fig. 5. Fig. 4b represents a tempera-
ture-enthalpy profile for all the systems with a single extraction. To
avoid entropy generation during the injection process, the air was ex-
tracted from state ‘ext’ at the humidifier and injected at the same state
‘ext’ at the dehumidifier. The entropy generation due to mixing results
in the increase in total entropy. For example, when several initially
separate fluids of different concentration and temperature, each in a
thermodynamic state of internal equilibrium, are mixed without che-
mical reaction followed by a time for the establishment of a new
thermodynamic state of internal equilibrium in the new mixture.
Therefore to avoid this increase in entropy, the constraint is placed so
that the state (temperature and humidity) of the injected air is the same
as the air it is injected into. Please note that “ext” state is the extraction
and injection state point. At this point, both the state of humidifier and
dehumidifier are in thermodynamic equilibrium. This model is applied
to the aforementioned systems. In an attempt to validate the results of
this work, terminal point temperatures of the current work for the
conventional cycle with zero and single extraction has been compared
to those reported by Narayan et al. [17] and presented in Table 1. The
deviation from Narayan work ranged between 0.3 and 2.2% for zero
extraction and between 0.3 and 1.2% for the single extraction case,
which represents a good agreement. It is important to emphasize that
Table 1 is a digitized version of Fig. 6 from [17]. This figure is gener-
ated at an enthalpy pinch of 20 kJ/kg dry air (that is, about 97%
component effectiveness). This is used merely to validate the authen-
ticity of a numerical code for the basic cycle with extraction. This ef-
fectiveness is very high which results in a reverse heat transfer when
heat recovery options are applied. However, in the low effectiveness
region, the water temperature at the humidifier outlet will be higher
than the water temperature at the dehumidifier outlet. Therefore, the
heat recovery process will be feasible, as shown in Table 2. For ex-
ample, at 77% component effectiveness, GOR of the proposed modified
systems are slightly higher than the conventional system. At 68% and
57% component effectiveness, GOR is increased by 59.6% and 178.6%,
respectively, for the heat exchanger option. However, for the mixing
chamber option, GOR increased by 65.4% and 207.1% at 68% and 57%
component effectiveness, respectively. This fact is further highlighted in
the Results and discussion section.

4.3. Double extractions system

Fig. 4. Temperature-enthalpy profile of the system with: (a) zero extraction, (b) single As discussed in Section 4.1, the algorithm for the double extractions
extraction, and (c) double extractions.
model is also shown in Fig. 5. The modeling process of double extrac-
tion is performed in three stages rather than two. A temperature-en-
̇ = ṁ cw cp (Tw2 − Tw1, mix )
Qin (8) thalpy profile for all the systems with double extractions is illustrated in
Fig. 4c. As explained by the single extraction model, air is also extracted
In Eq. (6) the temperature of water at the outlet of the dehumidifier from states ‘ext1’and ‘ext2’ at the humidifier and injected to the same
(Tw1) is calculated using the mathematical approach as illustrated in states ‘ext1’and ‘ext2’ to the dehumidifier, this helps to avoid entropy
Fig. 5. While (Tw1, HX) and (Tw1, mix) can be found from mass and energy generation. These three models are then used to examine the perfor-
balance for the heat exchanger and mixing chamber, respectively, as mance of the conventional and modified cycles. For further validation,
shown in the following equations: the results of GOR for the conventional cycle using double extractions
ṁ cw cp, cw (Tw1, HX − Tw1 ) model are compared with Chehayeb et al. [19] and presented in
εHX = Table 3. The results show good agreement with the literature with a
ṁ b, HX cp, HX (Tw3 − Tw1 ) (9)
maximum deviation of 3.4%.

119
S.M. Elmutasim et al. Desalination 435 (2018) 114–127

Fig. 5. Zero, single, and double extractions algorithm for all cycles.

Table 1
Comparison between Narayan et al. [17] and the current work.

Terminal points Zero extraction Single extraction

Ref. [17] (°C) Current work (°C) Error (%) Ref. [17] (°C) Current work (°C) Error (%)

Dehumidifier outlet (Tw1) 62.6 64.0 2.2 70.2 69.9 0.4

Humidifier outlet (Tw3) 36.0 35.9 0.3 29.8 29.6 0.7
Bottom air temperature (Ta1) 25.5 25.2 1.2 25.5 25.2 1.2
Top air temperature (Ta2) 64.3 65.7 2.2 71.1 70.9 0.3
Dehumidifier extraction (Tdeh,ext) – – – 38.5 38.1 1.0
Air extraction (Ta,ext) – – – 41.9 41.6 1.0
Humidifier extraction (Thum,ext) – – – 48.2 47.7 1.0

120
S.M. Elmutasim et al. Desalination 435 (2018) 114–127

5. Results and discussion all cases. Since the bottom temperature is fixed for two cases at 20 °C
and 30 °C, respectively. It is found that the increase in temperature
In this section, the performance of the conventional CAOW water difference leads to increase in the total heat input. Thus, the GOR drops
heated cycle and the modified cycles (CAOW-HX, CAOW-MX) is eval- accordingly. In addition, it is to be noted that at 20 °C the GOR is higher
uated with zero, single and double extractions. than at 30 °C. This is because of the fact that the ability of cold water to
condensate more potable water.
5.1. Effect of the difference between top and bottom temperatures The reduction in entropy generation due to the thermal balance
associated with single extraction leads to tremendous increase in GOR,
5.1.1. Conventional CAOW water heated cycle which is almost tripled compared to zero extraction at 40 °C tempera-
Fig. 6a illustrates the effect of the temperature difference on water ture difference for bottom temperatures of 20 °C, and 30 °C, respec-
heated CAOW system performance with zero, single and double ex- tively. For the double extractions case, we notice the increase in GOR is
tractions at zero enthalpy pinch (that is, 100% component effective- even higher than five times when compared to zero extraction for the
ness). It is clear that as the difference increases the GOR decreases for same operating conditions. Fig. 6b illustrates the behavior of this

Fig. 6. Effect of the difference between the top and bottom temperatures on water heated CAOW performance: (a) Ψdeh = Ψhum = 0 kJ/kg dry air, (b) Ψdeh = Ψhum = 20 kJ/kg dry air,
(c) Ψdeh = Ψhum = 80 kJ/kg dry air, and (d) Ψdeh = Ψhum = 140 kJ/kg dry air.

121
S.M. Elmutasim et al. Desalination 435 (2018) 114–127

Table 2
Feasibility of heat recovery options in the low effectiveness region.

Enthalpy pinch (kJ/kg dry air) Tw1 (°C) Tw3 (°C) CAOW CAOW-HX CAOW-MX

GOR GOR Improvement in GOR (%) GOR Improvement in GOR (%)

105 (ε = 77%) 49.9 50.1 0.85 0.86 1.2 0.86 1.2

115 (ε = 75%) 47.6 52.3 0.74 0.85 14.9 0.86 16.2
125 (ε = 71%) 45.2 54.8 0.63 0.85 34.9 0.86 36.5
135 (ε = 68%) 42.6 57.4 0.52 0.83 59.6 0.86 65.4
145 (ε = 64%) 39.7 60.3 0.42 0.82 95.2 0.86 104.8
160 (ε = 57%) 34.7 65.3 0.28 0.78 178.6 0.86 207.1

system under the same operating conditions but at 20 kJ/kg dry air increase sharply from 0.035 to slightly > 0.45. This GOR is low mainly
enthalpy pinch (that is, about 97% component effectiveness). The GOR due to the low effectiveness of the components. After 50 °C, the increase
for the 20 °C bottom temperature case increases for all the layouts (with in temperature difference starts to impact the performance as the in-
and without extraction) as temperature difference increases. This in- crease in top cycle temperature results in an adequate amount of eva-
crease is associated with the increase in the top cycle temperature poration. This means more condensation at the dehumidifier, which
which will result in a better evaporation process in the humidifier, thus results in an increase in GOR. For the case of 30 °C bottom temperature,
more distilled water in the dehumidifier. The single- and double-ex- the GOR increases as the temperature difference increases. This in-
traction curves for this case coincide, while the GOR almost doubled crease of condensate flow rate is associated with the increase in top
compared to zero extraction at 60 °C temperature difference which is cycle temperature. From this figure, the effect of a smaller amount of
mainly due to the thermal balancing. total heat input required can be observed since the case when Tmin is
For the case where the bottom temperature is 30 °C, the GOR for 30 °C has a higher GOR than that of the 20 °C case as the temperature
zero extraction decreases as the temperature difference increases due to difference increases. The GOR at 60 °C temperature difference for the
the increase in the amount of total heat input. Thermal balancing 30 °C bottom temperature case is almost doubled when compared to
through single and double extractions for this case increases the GOR Tmin of 20 °C case.
until temperature difference approaches 55 °C, it then starts to drop.
This behavior is associated with the increase in the top cycle tem-
perature which means better evaporation until a certain point at which 5.1.2. Modified (CAOW)-HX cycle
the effect of an increase in total heat input surpasses the effect of vapor Fig. 7 illustrates the effect of temperature difference between the
generated through increasing the top cycle temperature, thus the GOR top and bottom cycle temperatures on the GOR for the modified cycle
drops. By using extraction, GOR increases by more than double for the with a heat exchanger for a two fixed bottom temperatures cases and at
temperature difference of 55 °C due to the reduction in entropy gen- different enthalpy pinch values. Fig. 7a shows that at zero enthalpy
eration associated with the extraction process. The GOR for 30 °C pinch (that is 100% component effectiveness) for both cases where
bottom temperature case is higher when compared to the 20 °C case for there is no effect of extraction on the GOR. This is because of the fact
zero, single, and double extraction cases. This is mainly because of the that thermal balancing, on one hand, increases the effectiveness of the
less amount of heat input required for 30 °C case. heat transfer process through the humidifier which will lead to a low
The GOR for single and double extractions for the case when the temperature of the brine that should be utilized in the heat recovery
bottom temperature is 30 °C is doubled at the temperature difference of process. On the other hand, it will increase the temperature of water
40 °C. Then, this elevation in GOR decreases until it reaches a difference coming out of the dehumidifier which then exchanges heat in the
of almost 0.75 units at temperature difference of 60 °C. This decrease in heating exchanger with the brine. This means that the effect of in-
the difference may be associated with the decrease in the total heat creasing temperature of the cooling water will be canceled by the effect
input required compared to the effect of cooler water's ability to con- of cooling down the brine temperature which will be the case in both
densate more potable water. After reaching a temperature difference of CAOW-HX and CAOW-MX modified systems, as illustrated in Figs. 7
59 °C, it is noticed that the GOR for the 20 °C case with zero extraction through 10.
slightly exceeds the 30 °C case. At this point, the effect of the cooler The GOR of the modified cycles is low when compared to the CAOW
water to condensate the water exceeds that of the total heat input re- conventional cycle. This is mainly due to the use of 100% component
quired. effectiveness for the humidifier which will result in a low brine tem-
Figs. 6c and 6d show the impact of running the system with the perature. A high effectiveness humidifier means the hot water that
above operating conditions at an enthalpy pinch of 80 and 140 kJ/kg enters will be cooled efficiently which will result in a low temperature
dry air, respectively (that is, 85% and 66% component effectiveness). of the rejected brine. This brine will exchange heat in the heat ex-
From both the figures, it is clear that there is no effect of extraction on changer with water coming out from the dehumidifier before it is he-
the GOR with increasing the temperature difference as the thermal ated in the heater. This process results in high heat input requirements,
balancing effect appears clearly for high effectiveness components. thus a lower GOR. From this figure and for the 20 °C bottom tem-
Fig. 6c illustrates that increase in the temperature difference increases perature case, the increase in temperature difference increases the GOR
the GOR for all the cases. This increase in GOR is due to the amount of slightly for this system by almost 3.9% at its peak when there is a 58 °C
evaporation in the humidifier due to the increase of the top cycle temperature difference.
temperature which will result in a higher amount of potable water and When the bottom temperature is fixed at 20 °C, increasing the
thus higher GOR. The GOR for 30 °C bottom temperature case is higher temperature difference means increasing the top cycle temperature
than that of the 20 °C case. This due to the less amount of heat input which results in a relatively higher brine temperature for the heat re-
required for the 30 °C bottom temperature case. covery process; thus, a higher GOR. In contrast to the above case, the
Fig. 6d shows the increase in temperature difference for the case 30 °C case increases slightly at the beginning from a temperature dif-
when Tmin is 20 °C, GOR decreases slightly. It is very low until the ference of 40 to 42 °C but then starts to decrease by about 5% as the
temperature difference reaches 50 °C. After this point, the GOR starts to temperature difference increases. The increase, in the beginning, is due
to the increase in brine temperature and evaporation in the humidifier.

122
S.M. Elmutasim et al. Desalination 435 (2018) 114–127

Table 3 The decrease after that is associated with the fact that as we increase
Comparison between Chehayeb work and the current work for GOR at different enthalpy the top cycle temperature, the total heat input required will increase.
pinch for the conventional cycle with double extractions.
This will result in a decrease in GOR. The 30 °C case has a higher GOR
Enthalpy pinch (kJ/kg dry air) Chehayeb et al. [19] Current work Error (%) than the 20 °C case until 54.5 °C temperature difference, after this point
the later exceeds the 30 °C case. Before this point, the impact of less
0 17.4 17.5 0.6 heat input required has a higher effect on GOR than the effect of in-
10 8.2 8.2 0
creasing both the inlet and brine temperatures. After this point, the
20 4.3 4.2 2.3
30 2.9 2.8 3.4 effect of having better evaporation and heat recovery processes takes
40 1.9 1.9 0 over which explains the better GOR for the 20 °C case.
Fig. 7b is produced at 20 kJ/kg dry air enthalpy pinch (that is, 97%
component effectiveness). It shows similar behavior as explained above
for Fig. 7a. The GOR drops slightly for both cases at the start when

Fig. 7. Effect of the difference between top and bottom temperatures on water heated CAOW-HX performance: (a) Ψdeh = Ψhum = 0 kJ/kg dry air, (b) Ψdeh = Ψhum = 20 kJ/kg dry air,
(c) Ψdeh = Ψhum = 80 kJ/kg dry air, and (d) Ψdeh = Ψhum = 140 kJ/kg dry air.

123
S.M. Elmutasim et al. Desalination 435 (2018) 114–127

Fig. 8. Effect of the difference between top and bottom temperatures on water heated CAOW-MX performance: (a) Ψdeh = Ψhum = 0 kJ/kg dry air, (b) Ψdeh = Ψhum = 20 kJ/kg dry air,
(c) Ψdeh = Ψhum = 80 kJ/kg dry air, and (d) Ψdeh = Ψhum = 140 kJ/kg dry air.

compared to the previous figure because of using less effective com- is found that there is no product until 48 °C for the 20 °C case, which is
ponents. However, it increased at 60 °C temperature difference as it due to the low effectiveness of the components. After 48 °C, the increase
suggests that the effect of the higher top cycle and brine temperatures in temperature difference starts to impact the performance as the in-
are more influential here. The point after which the 20 °C case exceeds crease in top cycle temperature enhances the evaporation process that
the 30 °C is the temperature difference is about less than 55 °C. Fig. 7c means more condensate at the dehumidifier, thus higher GOR. The
examines this system at an enthalpy pinch of 80 kJ/kg dry air (85% 20 °C case does not exceed the 30 °C case in terms of GOR. The GOR for
component effectiveness). The GOR for 30 °C case reaches a peak of the 20 °C starts from slightly above 0.73 and then increases as the
0.885 at 44 °C temperature difference. It then starts to decrease as the temperature difference increases to about 0.825 at a temperature dif-
temperature difference increases which is also similar to what is ex- ference of 58 °C. The effect of temperature difference increase on the
plained earlier in the case for Fig. 7a. The GOR is slightly less than that GOR of the 30 °C case is marginal unlike the significant increase in the
of Fig. 7b, because of the lower component effectiveness. The point at 20 °C case. This implies that the 20 °C is more sensitive to the compo-
which the 20 °C case surpasses the 30 °C one at a temperature difference nent effectiveness and the top cycle temperature. From all the previous
of about 58 °C. figures, we notice that the point at which the 20 °C case exceeds the
This system is examined then at enthalpy pinch of 140 kJ/kg dry air 30 °C one is shifting from 54 to 58 °C until it doesn't surpass the 30 °C
(that is 66% component effectiveness), as shown in Fig. 7d. Similarly, it case for the last case (refer to Fig. 7d). This indicates that the impact of

124
S.M. Elmutasim et al. Desalination 435 (2018) 114–127

top cycle temperature and brine temperature is more dominant at the 5.2. Effect of the enthalpy pinch on modified systems performance
high effectiveness regions than the impact of total heat input required.
Fig. 10a and b exhibit the effect of enthalpy pinch on the perfor-
mance of the CAOW-HX and CAOW-MX systems, respectively with zero,
5.1.3. Modified (CAOW)-MX cycle single and double extractions. As illustrated in Fig. 10a the GOR in-
Fig. 8 shows the effect of the temperature difference between the creases as the enthalpy pinch increases up to 50 kJ/kg dry air and then
top and bottom cycle temperatures on GOR for the modified cycle with it starts to decrease noticeably to a GOR just below 0.78 at 160 kJ/kg
the mixing chamber for two fixed bottom temperatures cases and at dry air enthalpy pinch. The increase in GOR in the first portion is due to
different enthalpy pinch values. The GOR for this system is slightly less decrease in humidifier effectiveness, which will result in a higher brine
than that of the previous one at the same operating conditions. This is temperature for the heat recovery process. In the second portion, the
because part of the recirculated brine is injected in the mixer resulting continuous decrease in GOR is attributed to the decrease in effective-
in less heat recovery and hence, less GOR. ness of the dehumidifier. Thus, the temperature of water that enters the
The effect of difference between the top and bottom temperatures heat exchanger will decrease accordingly and more heat will be re-
on heat input required and fresh water production for the modified quired at the heater; i.e., less GOR.
options is shown in Fig. 9. This is carried out for two fixed bottom As shown in Fig. 10b the GOR starts to increase as enthalpy pinch
temperature cases at constant enthalpy pinch value equal to 20 kJ/kg increases until it reaches its peak at GOR slightly above 0.86 at 120 kJ/
dry air. It shows that both the heat input required and fresh water kg dry air enthalpy pinch. It then drops to a GOR of slightly < 0.86 at
production increases as the top temperature increases. This is because 160 kJ/kg dry air. This increase is explained previously as it indicates
as top temperature increases, more energy is required in the heater to that this system is more sensitive to the brine temperature effect. The
achieve this temperature. Also, the increase in top temperature in- slight drop after 120 enthalpy pinch indicates the impact of low ef-
creases the evaporation in the humidifier and makes the condensation fectiveness in the dehumidifier starts to outplay the benefit of low ef-
process more effective. The seawater that enters the dehumidifier is fectiveness in the humidifier.
used as a coolant for the condensation process. As this temperature The yield of heat recovery process in both the heat exchanger and
increases, the condensation process decreases and hence, the decrease mixing chamber is dependent on the energy associated with the cir-
in freshwater produced, as shown in Fig. 9. culated brine and cooling water at the outlet of the dehumidifier. When
single and double extractions are used, the brine is coming out of the
humidifier with less energy and the cooling water coming out of the
dehumidifier with a higher energy content. This loss in energy across
the humidifier is compensated by the increase in energy of the cooling
water. In conclusion, the resultant of heat recovery process will remain
the same for zero, single and double extractions. This explains the
overlapping between zero, single and double extractions lines for
modified systems as shown in Figs. 7 through 11.

5.3. Effect of enthalpy pinch on water heated cycles

Fig. 11 represents the effect of enthalpy pinch on the performance

for all the cycles (investigated in this paper) with zero, single and
double extractions. It can be seen from this figure that the GOR de-
creases as the enthalpy pinch increases for the CAOW system with zero,
single, and double extractions. The GOR behavior for both CAOW-HX
and CAOW-MX is explained earlier in Fig. 10a and b. Region ‘A’ is
considered the high effectiveness region with enthalpy pinch ranges
from 0 to 40 kJ/kg dry air, which is equivalent to effectiveness varying
from 100% to 93% for both the humidifier and dehumidifier. This re-
gion is further divided into two regions ‘A1’ and ‘A2’ with enthalpy
pinch ranges from 0 to 18 kJ/kg dry air and 18 to 40 kJ/kg dry air (that
is, equivalent to 100%–98% and 98%–93% component effectiveness).
The double extractions have the highest GOR in region ‘A1’ for the
conventional CAOW system. It starts with a GOR of almost 17.5 and
then decreases to GOR slightly above 5.
The reduction in entropy generation due to thermal balancing as-
sociated with single and double extractions leads to a tremendous in-
crease in GOR for the conventional CAOW while for the modified
system there is no effect of extraction as explained previously. In region
‘A2’ the effect of double extractions coincides with the single extraction
line for the conventional CAOW system as they start from slightly above
5 to a value of 2, as the component effectiveness drops from 98% to
93%. Both modified cycles with zero, single and double extractions
have the lowest GOR in region ‘A’. The conventional CAOW without
extraction exhibits a moderate GOR that ranges from almost 3 to 2,
which lies between the modified cycles and the conventional one with
single and double extractions across region ‘A’.
Fig. 9. Effect of the difference between the top and bottom temperatures on water heated In region ‘B’ GOR of the conventional CAOW with zero, single and
(a) CAOW-HX, and (b) CAOW-MX heat input rate and fresh water production, double extractions coincides, the values are slightly higher at the start
Ψdeh = Ψhum = 20 kJ/kg dry air.
than the modified cycles. For this region, the enthalpy pinch starts from

125
S.M. Elmutasim et al. Desalination 435 (2018) 114–127

40 to 105 kJ/kg dry air, which is equivalent to effectiveness values from

93% to 77%. While in region ‘C’, the modified cycles have the highest
GOR. Region C ranges from 105 to 160 kJ/kg dry air (that is, 77 to 57%
component effectiveness). The right extreme end of region ‘B’ exhibits
the point where all systems have the same GOR. As the enthalpy pinch
exceeds this point (i.e. 105 kJ/kg dry air) the effectiveness starts to
drop noticeably, which justifies the use of heat recovery option (i.e. the
modified systems). From Figs. 7 to 11, it can be seen that the GOR for
the modified systems is always < 1 no matter what the operating
conditions are. However, these values are the highest in region ‘C’
which justifies using them only at this low energy effectiveness region.
The results of this figure are summarized in Table 4 to indicate the best
configuration for each effectiveness region.

6. Concluding remarks

Zero, single and double extractions effect on the performance of the

conventional water heated closed-air open-water cycle and both the
modified cycles have been examined. Based on the analysis and dis-
cussion presented in the paper, the following are the main findings:

• There are three regions with different performance behavior and are
classified into high effectiveness region A (100–93% effectiveness),
moderate effectiveness region B (93–77% effectiveness), and low
effectiveness region C (77–57% effectiveness).
• In the high effectiveness region, which is further divided into two
regions, the conventional cycle with extraction has the highest
performance. The double extraction systems are better in the upper
portion A1 (100–98% effectiveness) of this region, while the single
extraction is recommended for the lower portion A2 (98–93% ef-
fectiveness) as it coincides with the double extractions one.
• In the moderate effectiveness region, the conventional cycle with
zero, single and double extractions are identical and have the
highest performance. It is recommended to use the conventional
Fig. 10. Effect of the enthalpy pinch on water heated (a) CAOW-HX, and (b) CAOW-MX cycle with zero extraction for this region.
performance: Tw0 = 20 °C; Tw2 = 80 °C.
• In the low effectiveness region, the water temperature at the

Fig. 11. Operating scheme for water heated systems (CAOW, CAOW-HX and CAOW-MX) with and without extraction: Tw0 = 20 °C; and Tw2 = 80 °C.

126
S.M. Elmutasim et al. Desalination 435 (2018) 114–127

Table 4 (2013) 286–305, http://dx.doi.org/10.4236/nr.2013.43036.

Recommended configuration for each effectiveness region. [3] F.A. Al-Sulaiman, M.I. Zubair, M. Atif, P. Gandhidasan, S.A. Al-Dini, M.A. Antar,
Humidification dehumidification desalination system using parabolic trough solar
Region Enthalpy pinch Effectiveness (%) Recommended configuration air collector, Appl. Therm. Eng. 75 (2015) 809–816, http://dx.doi.org/10.1016/j.
(kJ/kg dry air) applthermaleng.2014.10.072.
[4] H. Müller-Holst, Solar thermal desalination using the multiple effect humidification
A1 0–18 100–98 Double Extraction (CAOW) meh-method, Solar Desalination for the 21st Century, 2007, pp. 215–225.
[5] H. Müller-Holst, Mehrfacheffekt-Feuchtluftdestillation bei Umgebungsdruck –
A2 18–40 98–93 Single Extraction (CAOW)
Verfahrensoptimierung und Anwendungen (PhD thesis), Technische Universität
B 40–105 93–77 Zero Extraction (CAOW)
Berlin, 2002.
C 105–160 77–57 Zero Extraction (CAOW-MX),
[6] M. Zamen, S.M. Soufari, M. Amidpour, Improvement of solar humidification– de-
Zero Extraction (CAOW-HX) humidification desalination using multi-stage process, Chem. Eng. Trans. 25 (2011)
1091–1096.
[7] T. Schlickum, “Device for separating a liquid from its dissolved matters”, European
humidifier outlet will be higher than the water temperature at the Patent, EP 1770068 A2, 2007.
[8] S. Hou, Two-stage solar multi-effect humidification dehumidification desalination
dehumidifier outlet. Therefore, the heat recovery process will be process plotted from pinch analysis, Desalination 222 (2008) 572–578, http://dx.
feasible. doi.org/10.1016/j.desal.2007.01.127.
• The GOR for both zero, single, and double extraction systems in- [9] T. Brendel, Solare Meewasserental sungsanlagen mit mehrstufiger verdungtung
(PhD thesis), Ruhr University Bochum, 2003.
creases as the top temperature increases up to a certain point
[10] T. Brendel, “Process to distil and desalinate water in contra-flow evaporation hu-
wherein it reaches its optimum value and starts to decrease after it. midifier unit with progressive removal of evaporated fluid”, 2003. German Patent
GOR increases as bottom temperature increases because of the less #DE10215079 (A1).
amount of heat input rate required until a certain point at which the [11] G.P. Thiel, J.H. Lienhard V, Entropy generation in condensation in the presence of
high concentrations of noncondensable gases, Int. J. Heat Mass Transf. 55 (2012)
effect of coolant seawater to condensate the air exceeds that of the 5133–5147, http://dx.doi.org/10.1016/j.ijheatmasstransfer.2012.05.014.
total heat input rate required. [12] M.A. Younis, M.A. Darwish, F. Juwayhel, Experimental and theoretical study of a

• The performance of the closed-air open-water with a heat exchanger humidification–dehumidification desalting system, Desalination 94 (1993) 11–24.
[13] R.K. McGovern, G.P. Thiel, G. Prakash Narayan, S.M. Zubair, J.H. Lienhard V,
and the closed-air open-water with a mixing chamber have been Performance limits of zero and single extraction humidification-dehumidification
studied. desalination systems, Appl. Energy 102 (2013) 1081–1090, http://dx.doi.org/10.
• These modified cycles have the best performance for the low ef- 1016/j.apenergy.2012.06.025.
[14] G. Prakash Narayan, J.H. Lienhard V, S.M. Zubair, Entropy generation minimization
fectiveness region. of combined heat and mass transfer devices, Int. J. Therm. Sci. 49 (2010)
• The extraction or balancing lost its significance in the high effec- 2057–2066, http://dx.doi.org/10.1016/j.ijthermalsci.2010.04.024.
[15] K.H. Mistry, J.H. Lienhard V, S.M. Zubair, Effect of entropy generation on the
tiveness region for the conventional cycle. Furthermore, it has no
performance of humidification- dehumidification desalination cycles, Int. J. Therm.
useful impact for the modified cycles in the three regions.
•
Sci. 49 (2010) 1837–1847, http://dx.doi.org/10.1016/j.ijthermalsci.2010.05.002.
The heat exchanger and mixing chamber modifications show almost [16] J.A. Miller, J.H. Lienhard V, Impact of extraction on a humidification-dehumidifi-
identical performance as heat recovery solutions. cation desalination system, Desalination 313 (2013) 87–96, http://dx.doi.org/10.

• The mixing chamber modification is simpler, cheaper and easy to 1016/j.desal.2012.12.005.

[17] G.P. Narayan, K.M. Chehayeb, R.K. McGovern, G.P. Thiel, S.M. Zubair,
maintain when compared to the other heat recovery option, thus it J.H. Lienhard V, Thermodynamic balancing of the humidification dehumidification
is recommended to be used in the low effectiveness region. desalination system by mass extraction and injection, Int. J. Heat Mass Transf. 57
(2013) 756–770, http://dx.doi.org/10.1016/j.ijheatmasstransfer.2012.10.068.
[18] G. Prakash Narayan, M.G. St. John, S.M. Zubair, J.H. Lienhard V, Thermal design of
Acknowledgement the humidification dehumidification desalination system: an experimental in-
vestigation, Int. J. Heat Mass Transf. 58 (2013) 740–748, http://dx.doi.org/10.
The authors acknowledge the support provided by King Fahd 1016/j.ijheatmasstransfer.2012.11.035.
[19] K.M. Chehayeb, G. Prakash Narayan, S.M. Zubair, J.H. Lienhard V, Use of multiple
University of Petroleum & Minerals through the project # IN151001. extractions and injections to thermodynamically balance the humidification dehu-
midification desalination system, Int. J. Heat Mass Transf. 68 (2014) 422–434,
References http://dx.doi.org/10.1016/j.ijheatmasstransfer.2013.09.025.
[20] K.M. Chehayeb, G.P. Narayan, S.M. Zubair, J.H. Lienhard V, Thermodynamic bal-
ancing of a fixed-size two-stage humidification dehumidification desalination
[1] H.M. Ettouney, H.T. El-Dessouky, Fundamentals of Salt Water Desalination, Elsevier system, Desalination 369 (2015) 125–139, http://dx.doi.org/10.1016/j.desal.2015.
Science Publishers, Amsterdam, The Netherlands, 2002. 04.021.
[2] A.E. Kabeel, M.H. Hamed, Z.M. Omara, S.W. Sharshir, Water desalination using a
humidification-dehumidification technique—a detailed review, J. Nat. Resour. 4

127