Energy and Buildings 109 (2015) 53–64

Contents lists available at ScienceDirect

Energy and Buildings

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

Performance study of counter-flow indirect evaporative air coolers

Demis Pandelidis a,∗ , Sergey Anisimov a , William M. Worek b
a
Department of Environmental Engineering, Wroclaw University of Technology, 27 Wyspiański st., 50-370 Wroclaw, Poland
b
Stony Brook University, College of Engineering and Applied Sciences, Department of Mechanical Engineering, 127 Engineering Bldg., Stony Brook,
NY 11794, USA

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

Article history: This paper numerically investigates the performance and heat and mass transfer processes that occur in
Received 4 May 2015 counter-flow indirect evaporative air coolers. Different types of counter-flow air coolers were compared:
Received in revised form 4 September 2015 a typical counter-flow unit, the same unit operating as a heat recovery exchanger, a regenerative unit and
Accepted 1 October 2015
a novel modified exchanger, proposed by authors. Additionally the paper investigated impact of location
Available online 9 October 2015
of the perforated holes in the channel plate of regenerative air cooler. The numerical model is based
on a one-dimensional heat and mass transfer analysis that was validated against existing experimental
Keywords:
data. The results obtained from the simulation reveal the high effectiveness of the presented units. It was
Indirect evaporative cooling
Mathematical model
established that selected arrangements of the presented exchangers are characterized by the different
Counter-flow heat and mass exchanger efficiency in different air-conditioning applications.
© 2015 Elsevier B.V. All rights reserved.

1. Introduction were the experiments took place. Heidarinejad and Moshari [5]
presented a new mathematical model of an indirect evaporative
The rise in living standards is leading to increased air- cooling system, which included the longitudinal heat conduction in
conditioning demand, especially in the summer time in high the channel plate and the effect of spray water temperature change
temperature areas. This high demand is also accompanied by the as well. The mathematical model was used for simulation of a com-
increasing electric energy consumption [1–3]. Consequently, it is bined system of indirect/direct evaporative cooling. The results
necessary to develop new sources of cooling, which could be less indicated that presented solution is able to achieve 50% higher wet-
dependent on electric energy. The use of indirect evaporative cool- bulb effectiveness (defined as the ratio of temperature difference
ing is a relatively new method of cooling air in air-conditioning between inlet and supply airflow to the difference between inlet air
systems which looks very promising, both in the area of reduced dry bulb and wet bulb temperature) than typical indirect evapora-
electricity consumption and in the area of using less ozone deplet- tive air coolers. Cui et al. [6] presented the performance analysis of
ing refrigerants, since its working medium is water. In typical indirect evaporative heat exchangers pre-cooling the air for typical
evaporative cooling systems, the supply airflow cannot be cooled mechanical compression cooling coils in hot and humid climate.
below its wet-bulb temperature. However, there are new methods The numerical study suggested that that the proposed exchanger
called sub-wet bulb evaporative cooling, which are able to pro- is able to meet about 47% of the cooling load from the ambient air,
duce very low outlet temperatures (theoretically, the limit for these therefore it is able to provide important energy savings. Gao et al.
cycles is the ambient air dew-point temperature). Sub wet-bulb [7] experimentally analyzed an integrated liquid-desiccant indi-
evaporative cooling is a new and innovative way to produce cooled rect evaporative air-cooling system using the Maisotsenko Cycle
air for air conditioning systems, without adding humidity to the air. (M-Cycle). The results showed that the dehumidification process
Recently, many researchers have analyzed these highly efficient in the first stage of the cycle has direct impact on the cooling
indirect evaporative air cooling cycles. Cruz and Krüger [4] analyzed capacity in the second stage, when the inlet parameters of the air-
passive indirect evaporative cooling in Brazilian climate conditions. flow or desiccant are changed. The energy balances obtained were
Different arrangements of passive air coolers were tested in their in the range of ±20% for all experimental conditions. Montazeri
experiments. The obtained results showed that the presented solu- et al. [8] simulated the impact of selected parameters on evapora-
tion can provide comfortable conditions in the most of the cities tive cooling efficiency using a computational fluid dynamics model.
The selected parameters were: air temperature, air humidity ratio,
airflow velocity, water temperature and droplet size distribution.
∗ Corresponding author. The results showed that the sensible cooling capacity of the sys-
E-mail address: [email protected] (D. Pandelidis). tem can be improve by more than 40%, for the given inlet water

http://dx.doi.org/10.1016/j.enbuild.2015.10.004
0378-7788/© 2015 Elsevier B.V. All rights reserved.
54 D. Pandelidis et al. / Energy and Buildings 109 (2015) 53–64

shading devices; indirect evaporative cooling, a combination of

Nomenclature shading devices and thermal mass and a combination of indirect
evaporative cooling, solar protection, thermal mass and nocturnal
cp specific heat capacity of moist air [J/(kg K)] radiative cooling. This work showed a high potential of using such
C counter-flow exchanger [–] systems in dry and warm climate conditions. From this cited work,
F surface area [m2 ] it is clear of the high potential of indirect evaporative cooling sys-
h height [m] tems to become a new, cheap and environmentally friendly source
G moist air mass flow rate [kg/s] of cooling energy. According to many studies [11–18], one of the
l streamwise length of cooler [m] most effective cooling solutions for indirect evaporative air-cooling
M modified counter-flow exchanger [–] is the counter-flow cycles (Fig. 1). Those air coolers can be arranged
q heat flux [W/m2 ] in a variety of configurations (Fig. 1(a)–(d)). The most “classical”
r0 specific heat of water evaporation [kJ/kg] solutions are a typical counter-flow exchanger and a regenerative
Q̂ cooling capacity Q̂ = W1 /W2 ·  · cp · (t1i − t1o ), counter-flow exchanger (Fig. 1(a) and (b): the psychrometric anal-
kW/(kg/s) [kW/m3 ] ysis of counter-flow and regenerative exchanger was presented
R regenerative exchanger [–] by authors in [15]). The regenerative heat exchanger can also use
RH relative humidity [%] different variants of perforation along its channel plate (Fig. 1(c),
s fin pitch [m] the psychrometric analysis of perforated regenerative air coolers
t temperature [◦ C] was presented by authors in [18]). The counter-flow exchanger
v air stream velocity [m/s] can operate as a heat recovery unit in conjunction with a standard
W heat capacity rate of the fluid [W/K] cooling coil from mechanical compression system (Fig. 1(d)). This
x humidity ratio [kg/kg] solution can be applied in any climate and it is less dependent
X coordinate along supply airflow direction [m] on the outdoor conditions. To make regenerative exchangers less
Y coordinate perpendicular to X coordinate [m] dependent on the ambient air parameters, a desiccant wheel may
Z coordinate along fins direction [m] be applied to the system (Fig. 1(e), the psychrometric analysis of
the desiccant system with regenerative exchanger was presented
Special characters
by Miyazaki et al. in [19]). The last system is a novel combination of
˛ convective heat transfer coefficient [W/(m2 K)]
heat recovery counter-flow exchanger operating with cooling coil
ˇ mass transfer coefficient [kg/(m2 s)]
and a regenerative heat exchanger (Fig. 1(f)). The novel solution is
ı thickness [m]
a compromise between a regenerative changer and a counter-flow
ε effectiveness [–]
exchanger operating with a cooling coil. Less air is returned to the
 density [kg/m3 ]
wet channel after passing through the exchanger, most of the air-
 surface wettability factor,  ∈ (0.0–1.0) [–]
flow is delivered to the cooling coil where it can be additionally
cooled and then to the conditioned space. After meeting the heat-
Non-dimensional coordinates
ing loads, the exhaust airflow is delivered to the working channels
Le Lewis factor Le = ˛/(ˇcp ) [–]
of the exchanger. Depending on the situation, more or less airflow
NTU number of transfer units NTU = ˛F/(Gcp ) [–]
can be returned directly after passing the exchanger to increase its
X̄ X̄ = X/lX − relative X coordinate [–]
temperature effectiveness. Therefore, during most time of its oper-
Ȳ Ȳ = Y/lY − relative Y coordinate [–]
ation the device should be able to achieve high cooling capacity and
Z̄ Z̄ = Z/hfin − relative Z coordinate [–]
low outlet temperatures.
This paper will focus on comparison of presented solutions
Subscripts
in order to establish which of them are more suitable for air-
1 main (supply) airflow
conditioning systems. It is obvious that above-mentioned systems
2 working airflow in the wet channels
are characterized by the different energy consumption: systems
E exhaust airflow
equipped only with evaporative air coolers consume significantly
fin referenced to the fins
less energy than desiccant systems which require energy for
i inlet
regeneration of the desiccant wheel and combined cooling coil-
l latent heat flow
evaporative air cooling systems. However, it should be stated that
met metal foil
it is impossible to precisely control the supply air temperature only
o outlet
with indirect evaporative air coolers. In case of the domestic appli-
p plate surface
cations this fact does not make a significant difference, because
s sensible heat flow
air conditioner is required to provide generally pleasant conditions
w water film
for the occupants. However, in many cases, even in office appli-
X air streamwise in the dry channel
cations, the precise level of supply air temperature is required to
Y air streamwise in the wet channel
 compensate the heat loads and obtain indoor conditions which
conditions at the air/water interface temperature
 guarantee the effective work of the employees or the safe oper-
referenced to the plate surface
ation of the equipment. In this case an additional device must be
added to the evaporative air cooler (desiccant wheel, cooling coil
or other). Such system consumes more energy, but it gives the ben-
temperature (35.2 ◦ C), if the temperature difference between the efits which a stand-alone evaporative air cooler cannot provide.
inlet air and the inlet water droplets increases from 0 ◦ C to 8 ◦ C. However, even though the systems with additional devices con-
Caliskan et al. [9] presented an energy and exergy analysis of one sume more energy, their energetic efficiency may be significantly
of the most effective indirect evaporative air cooling cycles: the increased using the highly effective indirect evaporative air coolers.
M-Cycle. The results indicated that maximum exergy efficiency is Therefore, this study focuses on analysis of the indirect evaporative
found to be 19.14% for a reference temperature of 23.88 ◦ C where air coolers in different arrangements in air conditioning systems.
the optimum operation takes place. Sosa and Gómez-Azpeitia [10] The paper does not analyze the energy consumed by the sys-
presented a field study of indirect evaporative air coolers and tem, because it may be significantly different, depending on the
 D. Pandelidis et al. / Energy and Buildings 109 (2015) 53–64 55

Fig. 1. Counter-flow indirect evaporative air coolers. (a) “Classical” counter-flow exchanger. (b) Regenerative exchanger. (c) Counter-flow exchanger with cooling coil. (d)
Perforated regenerative exchanger. (e) Regenerative exchanger with the desiccant wheel. (f) Novel modified counter-flow exchanger.

Fig. 2. Initial conditions for considered exchangers. (a) Counter-flow exchanger. (b) Regenerative exchanger. (c) Perforated regenerative exchanger. (d) Novel counter-flow
exchanger.
56 D. Pandelidis et al. / Energy and Buildings 109 (2015) 53–64

Fig. 3. Comparison between the counter-flow and the regenerative exchanger. (a) Operation on ambient air-outlet air temperatures. (b) Operation on ambient air-obtained
cooling capacity. (c) Counter-flow exchanger operating on exhaust airflow (wet channels) and ambient air (dry channels), regenerative exchanger operating only on ambient
air-outlet air temperatures. (d) Counter-flow exchanger operating on exhaust airflow (wet channels) and ambient air (dry channels), regenerative exchanger operating only
on ambient air-obtained cooling capacity.

application. However, it is obvious that the more effective evap- is colder than the ambient air. The cooling coil can be supplied with
orative cooling process is, the less energy is consumed by the the traditional mechanical-compression system, but it can also be
system. supplied from the absorbent chiller running on the waste heat. The
It can be seen that flow arrangement of regenerative air cooler same situation occurs with the desiccant wheel: it may be regen-
(Fig. 1(b)) does not allow it to operate on the exhaust airflow. erated with the waste heat or heat obtained from the solar panels.
Therefore if this exchanger is placed in the system with cooling During certain parts of the year the evaporative air cooler is able to
coil it cannot benefit from using the colder exhaust air for the wet provide appropriate indoor conditions without switching the cool-
channels. On the other hand, the counter-flow unit can operate on ing coil on (in the desiccant systems the outdoor air is directed to
the exhaust airflow, but it cannot benefit from pre-cooling of the the bypass of the desiccant wheel under such conditions). In this
working airflow in the dry channels (Fig. 1(d)). The counter-flow case the energy consumption would be kept at the minimal level.
unit operating on the exhaust airflow is not the best solution for The key reason behind this study is to determine which of the
the desiccant system, because the exhaust air, which is later used exchangers is characterized by the highest efficiency in different
for the regeneration of the wheel, after passing the wet channels, systems, because this information is critical from the standpoint of
becomes very humid. This means that in order to regenerate the the total energy consumed by any air conditioning system equipped
wheel effectively it would require very high temperature. In case with evaporative air coolers.
of the regenerative exchanger, the exhaust air is dry and it does The analysis will use a experimentally validated numerical
not require high temperature to regenerate the desiccant wheel model, which distinguishes from other models existing in litera-
effectively. It is important to compare those exchangers to deter- ture by taking account the heat and mass transfer on the finned
mine which of these units guarantees more effective operation of surface of indirect evaporative air coolers and by considering dif-
different systems. ferent perforation arrangements in the plate of the air cooler. The
It should be mentioned that the name of the system “counter- model simulates the operation of all of indirect evaporative air cool-
flow unit with cooling coil” is used as a simplification. It only means ers presented in Fig. 1, by changing the operation arrangement and
that the counter flow unit is operating on the exhaust airflow, which initial conditions.
 D. Pandelidis et al. / Energy and Buildings 109 (2015) 53–64 57

Fig. 4. Comparison of the novel exchanger with counter-flow and regenerative unit. (a) Comparison with the counter-flow exchanger operating on exhaust airflow. (b)
Comparison with the regenerative air cooler operating on ambient air. (c) Comparison with the regenerative air cooler with the same amount of airflow returned to the wet
channel. (d) Cooling capacities obtained by the novel unit, counter-flow unit and a regenerative unit. (e) Switch points relative humidity at ambient air temperature equal
30 ◦ C. (f). Comparison with the regenerative air cooler operating with the desiccant wheel.

2. Methods important assumptions, were presented in [14–18]. The different

arrangements of indirect evaporative air coolers are characterized
The mathematical model developed to simulate the counter- by diffferent initial and operating condiitons (Fig. 2). In case of
flow heat and mass exchangers is based on the modified ε-NTU the perforated regenerative exchanger, three perforations were
method. The basis of the method, including main heat transfer and assumed. They can be distirbuted regularly or irregularly along the
58 D. Pandelidis et al. / Energy and Buildings 109 (2015) 53–64

plate (different values of X̄1 , X̄2 and X̄3 ). Different perfora- Table 1
Switch point relative humidity for different ambient and exhaust air parameters
tion distributions will be compared to establish their impact on
(20% of the supply airflow is returned to the wet channel).
the cooling performance (authors showed that different perforation
arrangement had a siginificant impact on the M-Cycle cross-flow tE = 24 ◦ C tE = 24 ◦ C tE = 25 ◦ C tE = 25 ◦ C tE = 26 ◦ C tE = 26 ◦ C
RHE = 50% RHE = 60% RHE = 50% RHE = 60% RHE = 50% RHE = 60%
heat exchanger [2,17]). The basic explenation of the assumptions
connected with analysis of the perforated air coolers was presented t1i = 25 65% 66% 65% 67% 67% 72%
t1i = 30 50% 56% 50% 57% 53% 57%
by authors in [15,18].
t1i = 35 42% 45% 42% 45% 43% 46%
The energy balance equations are the same for all of analyzed
indirect evaporative air coolers, the main difference lies in the cal-
culating algorithm, which is different for the assumed exchangers. Table 2
The model of perforated exchanger requires an additional algo- Assumed hole arrangement variants.

rithm describing the airflow mixing process. The algorithm was V1 V2 V3

presented by authors in [17]. Due to the fact that authors described
X̄1 = 0.33 X̄1 = 0.35 X̄1 = 0.15
the process of conversion the basic equations describing heat and X̄2 = 0.33 X̄2 = 0.15 X̄2 = 0.15
mass transfer in indirect evaporative air coolers in detail in their X̄3 = 0.34 X̄3 = 0.15 X̄3 = 0.35
previous studies [14–18], this process will be omitted in this study
and only the final form of the equations will be presented.
The energy balance for the main airflow is given as: also supplemented by an empirical equation describing the rela-
   tion between the temperature of the water vapor and its saturation
dt1 ıfin1 2 pressure [15,20]. The model described above is implemented in a
= NTU1 (tp1

− t1 ) 1− + tanh (mfin1 hfin1 )
dX̄ sfin1 mfin1 sfin1 multi-module computer simulation program. A four-dimensional
(1) computational numerical code on the base of the modified
Runge–Kutta method was implemented using the Wolfram Math-
ematica environment. The validation of presented mathematical
The energy balance for the working airflow is given as:
  model is presented in Appendix A.
dt2 ıfin2
= −NTU2 1− 
(tp2 − t2 )
dX̄ sfin2 3. Results and discussion
 Z̄2 =1 
2hfin2  The exchangers will be characterized by following dimensions:
+ (tfin2 − t2 )dZ̄2 (2)
sfin2 Z̄2 =0
• Length: 0.7 m
The mass balance equation for the water vapor is given as: • Width: 0.5 m

dx2
 1   ıfin2
 • Channel height: 3.0 mm
= −NTU2 1− 
p2 (xp2 − x2 ) • Working to intake air ratio: 0.5 (regenerative exchangers), 1.0
dX̄ Le2 sfin2 (counter-flow exchangers)
 Z̄2 =1  • Returned supply airflow in modified counter-flow exchanger:
2hfin2 
+ fin2 (xfin2 − x2 )dZ̄2 (3) G1 /G1 = 0.2
sfin2 Z̄2 =0 • Primary airflow velocity: 3 m/s

The energy balance for the fins inside the dry channel is pre-
sented in form of the heat conduction equation considering only 3.1. Comparison of the presented units
sensible heat transfer.
∂2 tfin1 The comparison between the regenerative and the counter-flow
= −m2fin1 h2fin1 (t1 − tfin1 ) (4) units are presented in Fig. 3(a)–(d). Fig. 3(a) shows the compar-
∂Z̄12
ison when both units operate using only ambient air (in case of
where: counter flow unit inlet temperatures of main and working airflow
2˛1 are identical: t1i = t2i ). It is clearly shown that the regenerative air
m2fin1 = (5)
fin1 ıfin1 + 2(fin1 )met (ıfin1 )met cooler is characterized by higher temperature effectiveness than
the counter-flow unit. The differences are up to 1.5 ◦ C. This is caused
The energy balance describing finned structure inside the wet by the initial pre-cooling of the working airflow in the dry channel
channels includes combined sensible and latent heat transfer. of the regenerative exchanger: if one evaporative exchanger has
   
∂2 tfin2 
fin 
a hot air entering both channels, while the other has pre-cooled
= −m2fin2 h2fin2 (t2 − tfin2 )− (xfin2 0
− x2 )rfin2 (6) air with the same humidity ratio entering the wet channel, the
∂Z̄22 cp Le
2
second would obtain the lower supply air temperature. This phe-
where: nomenon was explained in detail by many authors, the additional
2˛2 information can be found in [3,11,14,21]. However, the regenera-
m2fin2 = (7)
fin2 ıfin2 tive air cooler obtains lower cooling capacity than the counter-flow
cooler (Fig. 3(b)). This follows from the flow arrangement inside
The overall balance relationship for the whole heat and mass the exchanger: even though it allows obtaining high temperature
exchanger structure (plate surfaces and fins in the dry and wet effectiveness by pre-cooling of the airflow, it requires a part of
channel) is given as: main airflow to be delivered to the wet channel. This results in
 W  dt  
dt2 r0 dx2 a lower airflow rate which is delivered to the conditioned space,
1 1
− − =0 (8) while in the counter-flow unit whole main airflow is delivered to
W2 dX̄ dX̄ cp dX̄
2 the system. For this reason the counter-flow exchanger obtains
To complete the set of differential equations (i.e., (1), (2), (3), a higher cooling capacity even when it is characterized by the
(8)) the initial conditions are needed (see Fig. 2). The model is lower temperature effectiveness. It can be seen that when those
 D. Pandelidis et al. / Energy and Buildings 109 (2015) 53–64 59

Fig. 5. Analysis of the perforation distribution inside regenerative heat and mass exchanger. (a) Obtained outlet air temperatures: W2 /W1 = 0.3. (b) Obtained outlet air tem-
peratures: W2 /W1 = 0.5. (c) Obtained outlet air temperatures: W2 /W1 = 0.7. (d) Temperature profiles: perforation V1. (e) Temperature profiles: perforation V2. (f) Temperature
profiles: perforation V3. (g) Humidity ratio profiles: perforation V1. (h) Humidity ratio profiles: perforation V2. (i) Humidity ratio profiles: perforation V3.
60 D. Pandelidis et al. / Energy and Buildings 109 (2015) 53–64

Fig. 6. Analysis of the novel counter-flow exchanger. (a) Obtained outlet temperatures: G1 = 0.1G1 . (b) Obtained outlet temperatures: G1 = 0.2G1 . (c) Obtained outlet
temperatures: G1 = 0.3G1 . (d) Analysis on the psychrometric chart: G1 = 0.1G1 .

exchangers are compared for the stand-alone system, the regen- cooler. It can be seen that when the system has to be equipped
erative air cooler obtains lower supply air temperatures. The with another cooling source (such as cooling coil supplied from
counter-flow unit obtains higher cooling capacity, but this can be the typical mechanical-compression system) it is more justified
compensated by using several regenerative air coolers instead of to use the counter-flow exchanger instead of the regenerative
one. Depending on the costs, both exchangers may be considered unit (especially in more humid climates). The main advantages
in a stand-alone system. of the system with the counter-flow exchanger and the cooling
Fig. 3(c) shows the comparison between the same exchangers, coil (presented in Fig. 1(d)) is the lower sensitivity on the outlet
but in this case the counter-flow unit uses the airflow exhausted conditions, possibility the precise regulation of the inlet tempera-
from the conditioned space. In such an arrangement, the evapo- ture and humidity (cooling coil may operate in drying mode) and
rative air cooler may operate as a pre-cooling unit for a standard recovery of the “free” cooling energy provided by the exhaust air.
cooling coil. This arrangement makes the system less sensitive to Furthermore, the indirect evaporative air cooler may also operate
outdoor conditions. It can be seen that in such system, for typ- as a typical heat recovery unit during the winter season. The main
ical indoor conditions, the counter-flow unit is able to achieve disadvantages are: using a mechanical vapor compression system
higher effectiveness than regenerative unit. For more humid out- as a backup cooling source (even if its energy consumption is sig-
door conditions (i.e., RH ≥ 50%), the counter-flow unit shows better nificantly lower) and sensitivity on the exhaust air parameters. In
performance for almost all of the typical exhaust air parameters. many cases, the ventilation ducts are placed in hot places, such as
Moreover, in this configuration, the counter-flow cooler obtains attics, where the airflow may be additionally heated, which may
significantly higher cooling capacity: the maximum difference affect the effectiveness of evaporative cooling process. Also, when
reaches 10 kW for every supply airflow considered. This shows the conditioned space is characterized by higher latent heat loads,
that the operation scheme, in an air-conditioning system, is essen- the exhaust airflow may be too humid to realize an evaporative
tial to obtain the highest efficiency of the indirect evaporative air cooling process with high effectiveness. In many cases, the most
 D. Pandelidis et al. / Energy and Buildings 109 (2015) 53–64 61

important factor characterizing indirect evaporative air coolers is and exhaust air parameters were calculated and are shown in
the possibility of achieving the lowest outlet temperatures [14,18]. Table 1.
From Fig. 3 it can be observed that, depending on the ambient and Another solution, which needs to be compared with the pre-
exhaust air parameters, both regenerative and counter-flow units sented unit is a regenerative heat exchanger coupled with a
may be alternately more effective. For this reasons the authors have desiccant wheel (Fig. 4(f)). In such system, the airflow is dried (and
proposed the novel solution: modified counter-flow exchanger also heated) by the desiccant rotor before it is delivered to the indi-
(Fig. 1(f)). This device has all the advantages of the presented rect evaporative cooler. Dryer air is produced by the evaporative
exchangers: depending on the indoor and outdoor conditions and cooling process with higher effectiveness. Such system is also less
it may operate more as a counter-flow or regenerative air cooler. sensitive to the ambient air conditions. However, it can be observed
The comparison of presented unit with other systems is shown in from Fig. 4(f), the ambient air has to be dried to a very low humid-
Fig. 4. ity ratio level (which corresponds to the relative humidity of the
Fig. 4(a) shows the comparison of the outlet supply air tem- processed air equal to 10%) to be able to achieve higher effective-
peratures obtained by the novel unit and a typical counter-flow ness than the novel unit. Such low humidity ratio levels require
exchanger using exhaust air. It can be seen that in all cases, the mod- additional power for regeneration of the sorbent in the desiccant
ified exchanger is able to achieve lower outlet temperatures, with wheel. Some novel systems have tried to cover the required power
differences are up to 1.2 ◦ C. The comparison of the outlet temper- with solar energy [22]. However solar systems are also dependent
atures obtained by the novel device with regenerative exchanger on the ambient conditions and usually the system requires an addi-
is shown in Fig. 4(b). It can be observed that even for unfavorable tional heater [23]. Moreover, the SDEC (i.e., solar descant evaporative
indoor conditions (i.e., higher temperature and humidity levels) the cooling) system is more expensive and also more complicated due
modified unit is able to achieve similar or higher effectiveness than to the additional solar system and complicated airflow scheme. For
the air-cooled exchanger. This improved efficiency is obtained by a humidity ratio equal to 15% and the higher the regenerative air
returning only 20% of the main airflow to the wet channel. In com- cooler achieves higher outlet temperatures than the novel device.
parison, in the regenerative unit 50% of the main airflow is returned Also, the high efficiency of the regenerative air cooler is obtained
to the working air channel. If this value is reduced to 20% the tem- when 50% of the airflow is returned to the wet channel, which
perature efficiency of the regenerative unit significantly decreases. results in lower cooling capacity, as described earlier.
Fig. 3(c) shows comparison of novel unit with the regenerative It can be seen that presented solution is able to achieve a high
exchanger with the same value of returned airflow (i.e., 20% of the temperature effectiveness and cooling capacity and its sensitivity
main air stream). It can be seen that under such operating con- to ambient conditions is low. It is easy to implement in classical air
ditions, the modified counter-flow exchanger obtains significantly handling units, due to the fact that it can operate in supply-exhaust
lower outlet temperatures than the regenerative unit, with differ- airflow scheme. The main disadvantage of presented exchanger is
ences are up to 5 ◦ C. This follows from the very low working air the fact that it still requires traditional mechanical vapor compres-
mass flow rate in the regenerative air cooler under such conditions. sion systems as supplementary cooling source. However, due to its
In case of the novel exchanger the working to main air ratio (G2 /G1 ) adjustability it is able to significantly reduce the vapor compression
is always equal 1 (the proportion of the returned primary airflow system operation time. This shows that when object is not equipped
and the exhaust airflow in the working air stream may vary, but with a source of the waste heat or does not have the space for solar
the working to main air ratio remains unchanged). In this case, the panels, which can provide enough power for the regeneration of
working to main air ratio in the regenerative air cooler is equal 0.2. the desiccant wheel should rather consider systems with the novel
Sensible heat flux from the product channel is equal G1 cp (t1o − t1i ) cooler or counter-flow unit and cooling coil. If the object has the
and it is completely delivered to the wet channel. Total heat flux in source of waste heat it may also considered presented solution,
the wet channel is equal G2 (i2o − ī2i ). Because G2 value is smaller because it can be used to supply the absorbent chiller instead of
than G1 , specific enthalpy drop (and therefore temperature drop) the mechanical compression unit, which can minimize the energy
in the dry channel must be smaller, to fulfill the balance equation. consumed by the system and make it much more effective and keep
This shows the high potential of the novel exchanger, where it is it simple and inexpensive in terms of investment.
able to achieve higher thermal effectiveness due to the favorable
combination of keeping the higher heat capacity of the working 3.2. Selected operational and technological aspects
airflow and reduction of exhaust air humidity ratio by mixing it
with dryer supply air. Due to the fact that only 20% of the airflow is 3.2.1. Regenerative exchanger
returned to the wet channel, the unit is able to achieve much higher The authors have presented the detail analysis of regenerative
cooling capacities than the regenerative exchanger (see Fig. 4(d)). heat and mass exchangers in Refs. [14,18]. This section will focus
It can be seen that novel exchanger obtains a little lower cooling on an aspect of the exchanger that has not yet been analyzed, that
capacity than the counter-flow unit (the differences are up to 3 kW is the location of perforated holes along the channel plate. Three
per every kg/s of main airflow), however it still is able to obtain perforation arrangements will be studied (see Table 2). These are:
over two times higher cooling capacity than the regenerative unit. holes distributed evenly along the plate (V1), holes located closer to
Moreover, depending on the outlet conditions, the value of air- the end of the dry channel (V2) and holes closer to the dry channel
flow returned to the wet passages in the novel unit is variable. inlet (V3).
During operation, the control system is able to change to airflow The results of the simulation are presented in Fig. 5. It
rate, which is delivered to the wet channel with the adjustable can be observed that depending on the working to intake air
damper. This allows for higher energy savings with a simple tech- ratio (W2 /W1 ), different arrangements show higher efficiency
nical solution. The analysis of “switch” points, when the modified (Fig. 5(a)–(c)). For the lower working to intake air ratio, the
exchanger should operate as a typical counter-flow unit (no air- exchanger with second hole arrangement is characterized by
flow returned to the wet channel) is shown in Fig. 4(e): for hot and lower outlet temperatures than other units (up to 0.5 ◦ C differ-
humid exhaust airflow (tE = 26 ◦ C; RHE = 60%) the “switch point” for ence). For higher working to intake air ratios (W2 /W1 = 0.5 and
an ambient air temperature of 30 ◦ C is a relative humidity of 57%. For 0.7) the exchanger with an even hole distribution achieves lower
colder and dryer exhaust airflow conditions (tE = 25 ◦ C; RHE = 55%) supply air temperatures (up to 0.5 ◦ C). As it was earlier established
for the same ambient air temperature, the switch point is at rela- by the authors [14,18], the perforated regenerative heat exchang-
tive humidity equal 50%. Other switch points for different ambient ers are very sensitive to the perforation arrangement at different
62 D. Pandelidis et al. / Energy and Buildings 109 (2015) 53–64

working to intake air ratios. This is connected to the perforation is always characterized by the lowest efficiency. This follows from
impact on the indirect evaporative air-cooling process. The positive the very low heat capacity of the working airflow through most of
effects are the reduction of the working airflow humidity ratio, the channel. The airflow with such low capacity becomes saturated
increasing airflow heat capacity along the channel and allowing to very quickly (Fig. 5(f)). The positive trend, associated with the mix-
keep higher NTU values along the main air channel as compared to ing of the airflows, occurs only in very short section of the channel
a typical regenerative air cooler. The negative trend following from and cannot overcome the negative impact of other effects. It can be
the perforation is increasing the working airflow temperature at seen that primary airflow is cooled with a high effectiveness only
the mixing points. at beginning of the primary air channel and then its temperature
The location of the holes close to the working air channel inlet change becomes minimal (Fig. 5(i)). For the above mentioned rea-
allows for keeping higher heat capacity of the working airflow sons the perforated regenerative exchangers should be equipped
through most of the wet channel. When working to intake air ratio with holes located closer to the beginning of the working air chan-
is low, keeping the highest possible heat capacity of the working nel at lower working to main air ratios and distributed more evenly
airflow becomes critical which overcomes the other trends con- at higher working to intake air ratios.
nected with the mixing of the air streams at the beginning of the
wet channel. As it can be seen in Fig. 5(d)–(f), the most even supply 3.2.2. Novel counter flow exchanger
air temperature drop can be observed in exchanger with holes per- Another aspect, which needs to be analyzed, is operation of the
forated close to the beginning of the working air channel (Fig. 5(e)). operation of the novel counter-flow exchanger for different values
The air humidity ratio also increases very evenly along the wet of the main airflow returned to the wet channel. The results of the
channel as shown in Fig. 5(h). For higher W2 /W1 ratios, the neg- numerical simulation for different G1 values (0.1G1 ; 0.2G1 ; 0.3G1 )
ative trends connected with keeping a lower NTU value through are presented in Fig. 6. It can be observed that increasing the airflow
most of the primary air channel overcomes the positive trend of rate which is returned to the wet channel may improve or reduce
higher heat capacity and the second variation of hole arrangement the temperature effectiveness. When the exhaust air temperature
becomes less effective than the first one (i.e., an even hole distri- and humidity are low, an exchanger with G1 = 0.1 achieves the
bution). It can be seen that even though the primary airflow is not highest efficiency (Fig. 6(a)). For hotter and more humid exhaust
cooled perfectly evenly under such hole arrangement (Fig. 5(d)), airflows exchanger a value of G1 = 0.2 and 0.3 achieves lower
the humidity ratio at final points of the channel is lower than in outlet temperatures (Fig. 6(b) and (c)). This is caused by the
case of the second hole arrangement due to the mixing process fact that mixing of the airflows causes two main changes in the
(Fig. 5(g)). It can be also observed that the third hole arrangement working airflow. First it decreases its temperature and it changes

Fig. A1. Scheme of the testing bench used in [11].

 D. Pandelidis et al. / Energy and Buildings 109 (2015) 53–64 63

Fig. A2. Validation of the ε-NTU method results against experimental measurements data from [2]: (a) outlet product air temperature at variable velocity v1 and constant inlet
air parameters (x1i = 11.2 g/kg), (b) outlet product air temperature at variable velocity v1 and constant inlet air parameters (x1i = 19.0 g/kg), (c) outlet product air temperature
at variable inlet air temperature t1i and constant velocity v1 (x1i = 6.9 g/kg), (d) outlet product air temperature at variable inlet air temperature t1i and constant velocity v1
(x1i = 11.2 g/kg), (e) outlet product air temperature at variable inlet air temperature t1i and constant velocity v1 (x1i = 20.0 g/kg), and (f) outlet product air temperature at
variable inlet air temperature t1i and constant velocity v1 (x1i = 26.4 g/kg).

its humidity ratio (Fig. 6(d)). For a more humid exhaust airflow Results of numerical simulations allowed for a detail compar-
humidity ratio of the inlet working airflow is reduced (process ison of presented unit. It was established that depending on the
11o − E 1 ), but for the more dry exhaust airflows it may be increased application in air-conditioning systems all of studied exchangers
(process 12o − E 2 ). Depending on the value of mixed air and its result in different performance. The regenerative unit is able to
temperature this effect may improve or reduce the effectiveness of achieve the lowest outlet temperatures, however it is character-
the unit. The main advantage of the novel device is the operation ized with the lowest cooling capacity. The counter-flow unit may
with adjustable damper. This control system is able to change achieve similar efficiency when it is operating as a heat recovery
the returned airflow rate or eventually stop returning the supply unit in supply-exhaust system with cooling coil. The evaporative
airflow to the working air channel (at points shown in Table 1). cooling process runs with higher effectiveness due to the low tem-
perature of the exhaust airflow.
4. Conclusions The novel counter-flow unit is a compromise between a regen-
erative exchanger and a counter-flow unit. It is able to achieve
A comparison of the different counter-flow indirect evapora- very low outlet temperatures, similar to the regenerative unit while
tive air coolers: “classical” counter-flow exchanger, a regenerative keeping relatively high cooling capacity, like typical counter-flow
exchanger and the novel modified counter-flow exchanger is heat and mass exchanger. Its main advantage is its ability to be
presented and analyzed. Additionally, selected operational and adjusted: the control system in air handling unit is able to change
technological aspects of regenerative exchanger and the novel the value of returned airflow with damper. The unit may operate as
modified unit are investigated. The study used an experimentally a typical counter-flow or more like regenerative exchanger depend-
verified numerical model based on previous work of the investiga- ing on the outdoor conditions. This allows reduction in the energy
tors. The validation of the model describing indirect evaporative air consumed by the air conditioning system equipped with an indi-
cooler is presented in Appendix A. rect evaporative air cooler and classical cooling coil to the lowest
64 D. Pandelidis et al. / Energy and Buildings 109 (2015) 53–64

level. The additional advantage of the presented unit is the abil- [2] D. Pandelidis, S. Anisimov, Numerical analysis of the selected operational and
ity to operate as a heat recovery exchanger during winter season. geometrical aspects of the M-Cycle heat and mass exchanger, Energy Build.
87 (2015) 413–424.
Analysis using the numerical model allowed establishing the [3] M. Jradi, S. Riffat, Experimental and numerical investigation of a dew-point
“switch points”, where no supply airflow should be returned to the cooling system for thermal comfort in buildings, Appl. Energy 132 (2014)
wet channel for different sets of ambient and exhaust air parame- 524–535.
[4] E.G. Cruz, E. Krüger, Evaluating the potential of an indirect evaporative
ters. The study shows the high potential for the application of the passive cooling system for Brazilian dwellings, Build. Environ. 87 (2015)
novel device to air-conditioning systems. 265–273.
[5] G. Heidarinejad, S. Moshari, Novel modeling of an indirect evaporative cooling
system with cross-flow configuration, Energy Build. 92 (2015) 351–362.
Appendix A. Validation of the mathematical model [6] X. Cui, K.J. Chua, M.R. Islam, K.C. Ng, Performance evaluation of an indirect
pre-cooling evaporative heat exchanger operating in hot and humid climate,
The mathematical model presented in this article was validated Energy Convers. Manage. 102 (2015) 140–150.
[7] W.Z. Gao, Y.P. Cheng, A.G. Jiang, T. Liu, K. Anderson, Experimental
against existing experimental data published by Riangvilaikul and
investigation on integrated liquid desiccant – indirect evaporative air cooling
Kumar [11]. The testing bench used in this study was equipped system utilizing the Maisotsenko-cycle, Appl. Therm. Eng. 88 (2015) 288–296.
with the air handling unit, the evaporative air cooler and the mea- [8] H. Montazeri, B. Blocken, J.L.M. Hensen, CFD analysis of the impact of physical
parameters on evaporative cooling by a mist spray system, Appl. Therm. Eng.
surement devices (Fig. A1). The parameters measured included: the
75 (2015) 608–622.
volumetric airflow rate and the dry and wet-bulb temperatures. [9] H. Caliskan, A. Hepbasli, I. Dincer, V. Maisotsenko, Thermodynamic
The parameters of the regenerative exchanger are listed below: performance assessment of a novel air cooling cycle: Maisotsenko Cycle, Int. J.
Refrig. 34 (2011) 980–990.
[10] L.C. Sosa, G. Gómez-Azpeitia, Cooling average potential of evaporative cooling
• Length: 1.2 m, system in dry warm climate, Energy Procedia 57 (2014) 2554–2563.
• Plate thickness: 0.5 mm, [11] B. Riangvilaikul, S. Kumar, An experimental study of a novel dew point
• Channel height: 5 mm, evaporative cooling system, Energy Build. 42 (2010) 637–644.
[12] Z. Duan, Z. Changhong, X. Zhang, M. Mustafa, B. Alimohammadisagvand, A.
• Channel width: 80 mm, Hasan, X. Zhao, Indirect evaporative cooling: past, present and future
potentials, Renew. Sustain. Energy Rev. 16 (2012) 6823–6850.
The model was set to the same operating conditions as for the [13] A. Hasan, Indirect evaporative cooling of air to a sub-wet bulb temperature,
Appl. Therm. Eng. 30 (2010) 2460–2468.
experimental cases, including the exchanger geometry and inlet [14] S. Anisimov, D. Pandelidis, J. Danielewicz, Numerical analysis of selected
air-flow conditions. The validation of the mathematical model is evaporative exchangers with the Maisotsenko Cycle, Energy Convers.
presented in Fig. A2(a) and (b) (constant inlet temperature and Manage. 88 (2014) 426–441.
[15] S. Anisimov, D. Pandelidis, Theoretical study of the basic cycles for indirect
humidity ratio, variable airflow velocity) and in Fig. A2(c)–(f) (con-
evaporative air cooling, Int. J. Heat Mass Transf. 84 (2015) 974–989.
stant airflow velocity, variable inlet temperature and humidity [16] D. Pandelidis, S. Anisimov, W.M. Worek, Performance study of the
ratio). For the first case (Fig. A2(a)), the differences in obtained Maisotsenko Cycle heat exchangers in different air-conditioning applications,
Int. J. Heat Mass Transf. 81 (2015) 207–221.
supply air temperature were at most 4%, the correlation between
[17] D. Pandelidis, S. Anisimov, Numerical analysis of the heat and mass transfer
experiment and model was 0.998. For the second case, discrepan- processes in selected M-Cycle heat exchangers for the dew point evaporative
cies between the model and experiment are at most 2% and the cooling, Energy Convers. Manage. 90 (2015) 62–83.
correlation is equal 0.999. It can be seen that the experimental and [18] D. Pandelidis, S. Anisimov, W.M. Worek, Comparison study of the
counter-flow regenerative evaporative heat exchangers with numerical
simulation results show satisfactory agreement, thus the model can methods, Appl. Therm. Eng. 84 (2015) 211–224.
be successfully used to predict performance of the counter-flow [19] T. Miyazaki, I. Nikai, A. Akisawa, Simulation analysis of open cycle adsorption
regenerative indirect evaporative air cooler. The accuracy of the air conditioning system – numeral modeling of a fixed bed dehumidification
unit and the Maisotsenko Cycle cooling unit, Int. J. Energy Clean Environ. 12
mathematical model predictions are considered to be similar for (2011) 341–354.
other exchangers analyzed in this paper, since the numerical model [20] M.N. Golubovic, H.D. Hettiarachchi, W. Belding, W.M. Worek, A new method
describing heat and mass transfer in the investigated exchanger is for the experimental determination of Lewis’ relation, Int. Commun. Heat
Mass Transf. 33 (2006) 929–935.
identical to one describing other counter-flow units. [21] L. Gillan, Maisotsenko Cycle for cooling process, Int. J. Energy Clean Environ. 9
(1–3) (2008) 47–64, http://dx.doi.org/10.1615/InterJEnerCleanEnv.v9.i1-3.50.
[22] S. Alizadeh, Performance of a solar liquid desiccant air conditioner – an
References experimental and theoretical approach, Sol. Energy 82 (2008)
563–572.
[1] E.D. Rogdakis, I.P. Koronaki, D.N. Tertipis, Experimental and computational [23] T. Miyazaki, A. Akisawa, I. Nikai, The cooling performance of a building
evaluation of a Maisotsenko evaporative cooler at Greek climate, Energy integrated evaporative cooling system driven by solar energy, Energy Build.
Build. 70 (2014) 497–506. 43 (2011) 2211–2218.