adsorption
description
Transcript of adsorption
Technology development in the solar adsorption
refrigeration systems
K. Sumathy*, K.H. Yeung, Li Yong
Department of Mechanical Engineering, University of Hong Kong, Hong Kong, People’s Republic of China
Received 9 May 2002; accepted 30 April 2003
Abstract
Solar adsorption refrigeration devices are of significance to meet the needs for cooling requirements such as air-conditioning,
ice-making and medical or food preservation in remote areas. They are also noiseless, non-corrosive and environmentally
friendly. Various solar powered cooling systems have been tested extensively; however, these systems are not yet ready to
compete with the well-known vapor compression system. For these reasons, research activities in this sector are still increasing
to solve the technical, economical and environmental problems. The objective of this paper is to provide fundamental
knowledge on the adsorption systems and present a detailed review on the past efforts in the field of solar refrigeration systems.
A number of attempts have been made by researchers to improve the performance of the solar powered adsorption subsystems.
It is seen that, for successful operation of such systems, a careful selection of the adsorbent-adsorbate pair is essential apart from
the collector choice, system design and arrangement of subsystems.
q 2003 Elsevier Ltd. All rights reserved.
Keywords: Adsorbate; Adsorbent; Adsorption; Performance; Refrigeration; Solar energy
Contents
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
2. Principle of adsorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
2.1. Adsorption equilibria . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
2.2. Refrigerants and adsorbents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
2.3. Heat of adsorption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306
3. Thermodynamics of adsorption cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
3.1. Basic adsorption cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
3.2. Heat recovery adsorption refrigeration cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
3.3. Mass recovery adsorption refrigeration cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
3.4. Thermal wave cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
3.5. Convective thermal wave cycle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
3.6. Multi-stage and cascading cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
4. Solar cooling technologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
4.1. Intermittent adsorption systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312
4.1.1. Silica-gel/water and silica-gel methanol systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
4.1.2. Zeolite–water systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
4.1.3. Activated carbon–methanol systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
4.1.4. Activated carbon–ammonia systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 314
0360-1285/03/$ - see front matter q 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0360-1285(03)00028-5
Progress in Energy and Combustion Science 29 (2003) 301–327
www.elsevier.com/locate/pecs
* Corresponding author. Tel.: þ852-2859-2632; fax: þ852-2858-5415.
E-mail address: [email protected] (K. Sumathy).
4.2. Continuous adsorption systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
4.2.1. Multi-stage and cascading systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315
4.2.2. Thermal wave adsorption systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
4.2.3. Convective thermal wave adsorption systems. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
4.2.4. Hybrid system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
4.3. Solar desiccant cooling and air-conditioning systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 321
5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 324
1. Introduction
The use of solar energy for environmental control is
receiving much attention as a result of the projected world
energy shortage. Refrigeration is particularly attractive as
a solar energy application because of the near coincidence
of peak cooling loads with the available solar power.
Solar refrigeration has the potential to improve the quality
of life of people who live in areas with electricity
insufficient. It is usually used for storage of agricultural
products, food and medicines (e.g. vaccines) in remote
areas. Solar cooling to produce ice accumulates latent
heat, thus leading to smaller volume of ice-makers. The
adsorption system is one of the promising solar thermal
refrigeration methods, and it is environmentally friendly
along with low cost and low maintenance requirements
[1]. Adsorptive processes have been applied extensively
for gas separation and catalysis, but it is only recently that
adsorptive processes have been widely studied for
refrigeration and heat pumps.
Despite a large potential market, existing solar refriger-
ation systems are not competitive with electricity-driven
refrigeration systems because of their high capital costs.
Improvements such as reduced collector area, improved
system performance, and reduced collector cost will lower
the cost of solar components. Several solar refrigeration
systems have been proposed and are under development
such as sorption systems including liquid/vapor, solid/
vapor absorption, adsorption, vapor compression and
photovoltaic-vapor/compression systems. Most of the
above mentioned systems have not been economically
justified.
Solar refrigeration is highly dependent upon environ-
mental factors such as cooling water temperature, air
temperature and solar radiation. The energetic conversion
efficiency is low, and solar cooling and refrigeration are not
yet competitive economically with the conventional
systems. This article details the various research aspects
of adsorption refrigeration, which includes adsorption
mechanism, the criteria to choose an appropriate working
pair, thermodynamic analysis of several refrigeration
cycles, adsorbent properties and various solar powered
adsorption refrigeration systems based on various cooling
technologies.
2. Principle of adsorption
Adsorption occurs at the surface interface of two phases,
in which cohesive forces including electrostatic forces and
hydrogen bonding, act between the molecules of all
substances irrespective of their state of aggregation [2].
Unbalanced surface forces at the phase boundary cause
changes in the concentration of molecules at the solid/fluid
interface. The process of adsorption involves separation of a
substance from one phase accompanied by its accumulation
or concentration at the surface of another. The adsorbing
phase is the adsorbent, and the material concentrated or
adsorbed at the surface of that phase is the adsorbate.
Adsorption processes can be classified as either physical
or chemical, depending on the forces causing the adsorp-
tion process. Physical adsorption (physisorption) occurs
when Van der Waals forces bind the adsorbing molecule to
the solid phase, these intermolecular forces are as same as
ones that bond molecules to the surface of a liquid.
Molecules that are physically adsorbed to a solid can be
released by applying heat; therefore, the process is
reversible. Chemical adsorption (chemisorption) occurs
when covalent or ionic bonds are formed between the
adsorbing molecules and the solid substance. The bonding
forces of chemical adsorption are much greater than that of
physical adsorption. Thus, more heat is liberated. This
bonding leads to change in the chemical form of the
adsorbed compounds and hence, it is irreversible. For this
particular reason, most of the adsorption processes
applicable to the thermal system or cooling machine
mainly involve physical adsorption.
Adsorption is an exothermic process accompanied by
evolution of heat, the quantity of heat release depends upon
the magnitude of the electrostatic forces involved, latent
heat, electrostatic and chemical bond energies. The heat of
adsorption is usually 30–100% higher than the heat of
condensation of the adsorbate. In general adsorption is
stronger than condensation to liquid phase. Hence, if a fresh
adsorbent and adsorbate in liquid form coexist separately in
a closed vessel, transport of adsorbate from the liquid phase
to the adsorbent occurs in the form of vapor. The liquid
temperature becomes lower while the adsorbent temperature
rises. Air-conditioning and refrigeration utilize this
phenomenon to obtain a cooling effect [3].
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327302
Heat of adsorption is either derived from adsorption
isotherms, generally referred to as either the isosteric heat
(the energy released in the adsorption process), or, as the
differential heat of adsorption determined experimentally
using a calorimetric method. Differential heat of adsorption
for some adsorbent/adsorbate pairs are given in Table 1. The
performance of adsorbents used in physisorption is governed
by surface properties, such as surface area, micro-pores and
macro-pores, size of granules in powders, crystals or in
pellets. Adsorbents having special affinity with polar
substances like water are termed ‘hydrophilic’. These include
silica gel, zeolites and porous or active alumina. Non-polar
adsorbents, termed ‘hydrophobic’, have more affinity for oils
and gases than for water. These substances include activated
carbons, polymer adsorbents and silicalites.
The general term ‘sorption’ is used when both adsorption
and absorption occurs simultaneously. ‘Desiccants’ are a
type of adsorbent having special affinity for water and have
been used extensively for dehumidification or drying in air
processing applications.
Adsorbents are characterized by surface properties such
as surface area and polarity. A large specific surface area is
preferable for providing large adsorption capacity, but the
creation of a large internal surface area in a limited volume
inevitably gives rise to large numbers of small sized pores
between adsorption surfaces. The pore size distribution of
micropores which determines the accessibility of adsorbate
molecules to the internal adsorption surface, is important for
characterizing adsorptivity of adsorbents. Materials such as
zeolite and carbon molecular sieves can be engineered
specifically for precise pore size distributions and hence
‘tuned’ for a particular separation.
2.1. Adsorption equilibria
To understand the adsorption process, the adsorption
equilibria are introduced to describe the adsorption process
and several state equations known as isotherms of adsorp-
tion are proposed [2,4]. Several basic theories have been
proposed and ussed to define the main isotherms of an
adsorption process and are listed below:
Henry’s law. For an adsorption process on a uniform
surface at sufficiently low concentrations (such that all
molecules are isolated from their nearest neighbors), where,
Nomenclature
A adsorption potential (J mol21)
Cm adsorbed concentration (mol kg21)
Cms saturation adsorbed concentration (mol kg21)
E interaction energy between solid and adsorbing
molecule (J mol21)
E0 characteristic energy used in Dubinin equation
(J mol21)
H enthalpy (J mol21)
k characteristic parameter of adsorption pair
K adsorption parameter for D-A equation
l micropore half width (nm)
l0 half width of a slit sharped micropore (nm)
n adsorption parameter of adsorption pair
P pressure (Pa)
PC condensing pressure (Pa)
PE evaporation pressure (Pa)
Ps saturated vapor pressure (Pa)
Rg gas constant (J mol21 K21)
T temperature (8C, K)
TA ambient temperature (8C)
TC condensing temperature (8C)
TE evaporation temperature (8C)
TR regeneration temperature (8C)
Ts saturated refrigerant liquid temperature (8C)
W volume adsorption capacity (l kg21)
W0 maximum volume adsorption capacity (l kg21)
W00 total micropore volume (l kg21)
Greek symbol
b affinity coefficient
d variance
u fractional loading
Table 1
Heat of adsorption of some adsorbent/adsorbate pairs
Adsorbent Adsorbate Heat of
adsorption
(kJ kg21)
Remarks
Activated
alumina
Water 3000 Water is applicable
except for very low
operating pressures
Zeolite
(various
grades)
Water 3300–4200 Natural zeolites have
lower values than
synthetic zeolites
Ammonia 4000–6000
Carbon
dioxide
800–1000
Methanol 2300–2600
Silica gel Methyl
alcohol
1000–1500 Not suitable above
200 8C
Water 2800 Used mostly for
desiccant cooling
Charcoal C2H4 1000–1200 Reacts at approximate
100 8C. Ammonia and
methanol are not
compatible with
copper at high
temperature
Ammonia 2000–2700
Water 2300–2600
Methanol 1800–2000
C2H5OH 1200–1400
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327 303
the equilibrium relationship between fluid phase and
adsorbed phase concentrations will always be linear [5].
Langmuir’s approach. In order to understand the
monolayer surface adsorption on an ideal surface, Lang-
muir’s approach is used and this approach is based on
kinetic equilibrium; that is, the rate of adsorption of the
molecules is assumed to be equal to the rate of desorption
from the surface [6].
Gibbs’ theory. This is based on the ideal gas law, in
which the adsorbate is treated in microscopic and bi-
dimensional form and provides a general relation between
spreading pressure and adsorbed phase concentration. In this
concept, volume in the bulk phase is replaced by the area
and the pressure is replaced by the so-called ‘spreading
pressure’. By assuming some forms of thermal equation of
state relating the number of moles of adsorbate, the area and
the spreading pressure and using them in the Gibbs equation,
a number of fundamental equations can be derived, such as
the linear isotherm, the Volmer isotherm, etc.
Adsorption potential theory. It is purely a thermodyn-
amic approach, suitable for adsorption in microporous
materials which was based on a model originally proposed
by Polanyi and further developed by Dubinin by the end of
the 1920s.
Recent equations for ideal solids are introduced by Nitta
and his co-workers [4] are based on statistical thermo-
dynamics with features similar to the Langmuir theory and
encompasses the Langmuir equation as a special case.
Due to the complexity in structure of practical solids, the
above fundamental adsorption isotherm cannot completely
describe the process. Hence, the analysis of ideal solids are
initially used as a basis for the development of equilibria
theory for practical solids, and a number of empirical as well
as semi-empirical equations to describe adsorption equili-
bria have been developed.
In order to describe adsorption of gases and vapors below
the capillary condensation region, equations such as
Freundlich, Langmuir–Freundlich (Sips), Toth, Unilan,
and Dubinin–Radushkevich (DR) have been used. Simi-
larly, to describe equilibrium data in the region of multi-
layering adsorption, the classical equation of Brunauer,
Emmett and Teller (BET) is used. To account for the various
features inherent with real solids, BET equation is further
modified and is listed in the reference [4]. The above
discussed empirical relations are applicable to both
subcritical as well as the wider range of supercritical vapors.
To analyse more critically the adsorption process due to
the presence of subcritical vapors in the micropores solids,
the semi-empirical DR equation’ was developed by Dubinin
et al. [7].
W ¼ W0 exp 2A
E
� �2" #
ð1Þ
In the above equation, A is the adsorption potential and is
defined as A ¼ RgT lnðP0=PÞ; E; is the interaction energy
between solid and adsorbing molecule. The DR equation has
been widely used to describe adsorption isotherms of sub-
critical vapors in microporous solids such as activated
carbon and zeolite.
The DR equation describes fairly well many carbon-
aceous solids with low degree of burned-out. For carbon-
aceous solids resulting from a high degree of burn-out
during activation, the degree of heterogeneity increases
because of a wider pore size distribution, and for such cases
the DR equation does not describe well the equilibrium data.
To take into account of this surface heterogeneity,
Dubinin and Astakhov [4] proposed another equation (DA)
of the following form:
W ¼ W0 exp 2A
E
� �n� �ð2Þ
where, the exponent n describes the surface heterogeneity.
Rewriting the above equation in terms of the characteristic
energy of the reference vapor, the expression becomes,
W ¼ W0 exp 2A
bE0
� �n� �ð3Þ
where, the parameter b represents the affinity coefficient,
which is the ratio of the liquid molar volume to that of the
reference vapor. With the additional exponent n in the
adsorption isotherm equation, the DA equation provides
flexibility in the description of adsorption data of many
microporous solids ranging from a narrow to wide
micropore size distribution.
Adsorption isotherms of many microporous solids do not
usually conform to the simple DR equation. Even with the
inclusion of adjustable exponent n in the DA equation, it
cannot reflect experimental data corresponding to different
situations. This inability to fit the data is attributed to the
heterogeneity of the system, in which the characteristic
energy varies with the different regions in the solid.
Since adsorption of many adsorbates in micropores of
carbonaceous solids is due to the dispersion force, the
micropore size is therefore playing a major role in the
attraction of adsorbate molecules. In this sense, distribution
of the micropore size dictates the heterogeneity of the solid
surface than the distribution in characteristic energy. For a
micropore volume, W00 f ðxÞ; (f ðxÞ represents the micropore
size distribution), having micropore size between l and l þ
dl; the volume of micropore occupied by adsorbate at a
given adsorption potential A is:
W ¼ W00
ðlmax
lmin
uðA=bE0; nÞf ðlÞdl ð4Þ
where u is the fraction of the micropore volume occupied by
the adsorbate, which is a function of ðA=bE0; nÞ [4]. l is the
micropore half width, with lmax and lmax are its minimum
and maximum, respectively. However, for mathematical
convenience, those limits are usually replaced by 0 and 1,
respectively, to facilitate the analytical integration of the
above integral.
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327304
To evaluate the integral Eq. (4), the parameters of the
local isotherm must be expressed in terms of the micropore
half width, l:The parameter n is usually regarded as a
constant. The parameter E0; is the characteristic energy of
the reference vapor, which follows the following relation-
ship with the micropore half width, as:
E0 ¼K
lð5Þ
The micropore size distribution can then be assumed to take
the following Gaussian distribution:
dW0
dx¼
W00
dffiffiffiffi2p
p exp 2ðl 2 l0Þ
2
2d2
" #¼ W0
0 f ðlÞ ð6Þ
where, W00 ; is the total micropore volume, l0 is the half-
width of a slit shaped micropore which corresponds to the
maximum of the distribution curve, and d is the variance.
Using the relationship between the characteristic energy and
the micropore half width, the DR equation can be written in
terms of this half width l as:
W ¼ W0 exp½2ml2A2� ð7Þ
where m ¼ 1=ðbkÞ2
In order to obtain an equation for a heterogeneous solid,
the above adsorption equation is differentiated to obtain
micropore volume element dW :
dW ¼ dW0 expð2ml2A2Þ ð8Þ
Combining Eqs. (6) and (8), yields the volume occupied by
the adsorbate ðWÞ :
W ¼W0
0
dffiffiffiffi2p
pð1
0exp 2
ðl 2 l0Þ2
2d2
" #expð2ml2A2Þdx ð9Þ
Evaluation of the above integral results:
W ¼W0
0
2ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi1 þ 2md2A2
p exp 2ml20A2
1 þ 2md2A2
!
� 1 þ erfl0
dffiffi2
p ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi1 þ 2md2A2
p !" #
ð10Þ
This equation was first obtained by Dubinin and Stoeckli
[8], and hence referred as DS equation. In this equation, ‘erf’
represents the error function. Using the above equation to fit
experimental data, three parameters can be extracted from
this fitting process, namely W00 ; l0; d: Knowing these
parameters, the micropore size distribution in terms of
volume can then be calculated from Eq. (6).
Thermodynamic analysis suggests that an adsorption
isotherm must exhibit the Henry law behavior when
pressure is very low. Unfortunately, the DA as well as the
DR equations do not follow the perfect Henry law when the
pressure is approaching zero. The above DR, DA, DS
equations, however, suffer from the disadvantage of zero
Henry constant at zero pressure. This difficulty in describing
the adsorption behavior at low pressure has been addressed
by a number of workers [9].
2.2. Refrigerants and adsorbents
There are several working pairs for solid adsorption. For
the successful operation of a solid adsorption system,
careful selection of the working medium is essential. It is
because, the performance of the system varies over a wide
range using different working pairs at different tempera-
tures. The advantages and disadvantages of different
working media and their properties are listed and discussed
in this section. For any refrigerating application, the
adsorbent must have high adsorptive capacity at ambient
temperatures and low pressures but less adsorptive capacity
at high temperatures and high pressures. Thus, adsorbents
are first characterized by surface properties such as surface
area and polarity. A large specific surface area is preferable
for providing large adsorption capacity, and hence an
increase in internal surface area in a limited volume
inevitably gives rise to large number of small sized pores
between adsorption surfaces. The size of the micropores
determines the effectiveness of adsorptivity and therefore
distribution of micropores is yet another important property
for characterizing adsorptivity of adsorbents.
Based on the above discussion, the choice of the
adsorbent will depend mainly on the following factors:
† high adsorption and desorption capacity, to attain high
cooling effect;
† good thermal conductivity, in order to shorten the cycle
time;
† low specific heat capacity;
† chemically compatible with the chosen refrigerant;
† low cost and widely available.
The selected adsorbate (working fluid) must have most
of the following desirable thermodynamics and heat transfer
properties:
† high latent heat per unit volume;
† molecular dimensions should be small enough to allow
easy adsorption;
† high thermal conductivity;
† good thermal stability;
† low viscosity;
† low specific heat;
† non-toxic, non-inflammable, non-corrosive; and
† chemically stable in the working temperature range.
Based on the above criteria, some of the appropriate
working pairs are zeolite–water, zeolite–organic refriger-
ants, silica gel–water, zeolite–water and activated carbon–
methanol in solid adsorption systems. Several refrigeration
applications have been studied using various adsorbent and
adsorbate pairs. Initially, zeolite and activated carbon were
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327 305
the only adsorbent most popularly used in many systems. In
the early 1980s, Tchernev [10] carried out an investigation
of adsorption refrigeration with zeolite–water pair. Also,
Pons et al. [11] worked on the adsorption pair zeolite and
water, to produce refrigerating effect achieving a coefficient
of performance (COP) of only about 0.1. Later, in 1987,
Pons and Crenier demonstated [12,13] that activated carbon
and methanol can serve as a suitable pair for a solar
powered, solid-adsorption ice-maker. Critoph [14,15] had
studied the performance limitations of adsorption cycles
with different adsorbates, for solar cooling and concluded
that, in general, activated carbon–methanol combination
was preferable for solar cooling, giving the best COP
achievable in a single-stage cycle. In order to increase the
desorption temperatures and have a good cooling effect
during the adsorption period at night, Headley et al. [16]
constructed a charcoal–methanol adsorption refrigerator
powered by CPC concentrating collectors, but the solar COP
achieved was very low, of about 0.02.
The recent new development of activated carbon fibre—
ACF, shows the possibility for applications in adsorption
refrigeration. One good example is the development in
Byelorussia of a refrigerator prototype using ACF–ethanol
and ACF–acetone pairs has been reported [17]. New
experiments have also been shown to use a heat pipe for
heating/cooling ACF adsorbers for the ACF–NH3 pair [18].
In China, several studies had been carried out on solar
powered refrigerators using different adsorption pairs such
as zeolite–water and activated carbon-methanol. In search
of new adsorbent materials, Wang et al. [19] have reported
that, for specially treated ACF, the measured adsorption
capacity of methanol is two to three times greater than that
of normal activated carbon, and the estimated adsorption
time is only about 1/5 to 1/10 of that of normal activated
carbon [20].
Meunier et al. [20,21] have compared synthetic zeolite–
water, synthetic zeolite–methanol, and reported that
charcoal–methanol resulted in a better system COP
provided that the night time temperature were low. When
the night time ambient temperature to evaporating tempera-
ture difference is particularly high, zeolite–water combi-
nation showed better performance. However, this will also
require a higher generating temperature from the solar
collector during the day.
Critoph and Vogel [22] compared zeolites and activated
carbon with refrigerants R11, R12, R22 and R114, and in all
cases found activated carbon a preferable adsorbent for solar
cooling. A theoretical analysis was carried out by Oertel and
Fischer [23] for a prediction of the achievable COP using
methanol/silica gel. Since water was used as refrigerant, the
adsorption chiller was restricted to cooling temperatures
above the freezing point of water at 0 8C. It was shown that
by changing to the working pair silica gel–methanol it
might be possible to reach cold brine temperatures of 25 8C
with a COP ¼ 0.25. A similar study on the dynamic
performance of a closed-cycle, solar-adsorption refrigerator
was carried out by Hajji et al. [24] numerically using zeolite
13X-water and chabazite–methanol working pairs for a
chosen set of weather conditions.
Recently, Tamainot and Critoph [25] investigated the
thermophysical properties of two types of monolithic
activated carbons with an intention to design and fabricate
a high performance generator for sorption refrigeration
systems and heat pumps using ammonia as refrigerant. It
was found that, reduction in volume from granular bed to
monolithic bed was up to 50% which could lead to a
substantial economic gain.
2.3. Heat of adsorption
All adsorption processes are accompanied by heat
release, i.e. they are, exothermic processes. Since in
adsorbed phase, adsorbate molecules are in more ordered
configuration, entropy decreases; and from DG ¼ DH 2 T
DS; exothermic character of the process is obvious. The
three terms which are often used in adsorption process are:
(i) Integral heat of adsorption—which is the total heat
released from initial state to final state of adsorbate loading,
at constant temperature; (ii) Differential heat of adsorp-
tion—is the change in integral heat of adsorption with
change in loading; and (iii) Isosteric heat of adsorption is
defined by using adsorption isosters and Clausius Clapryron
relationship.
In practice, the difference between differential heat of
adsorption and isosteric heat of adsorption is so small that it
can easily be neglected, as well as, these two can be
considered as identical. At constant loading ðdCm ¼ 0Þ;
using the DA Eq. (3) and the following Clapeyron equation,
DH
RgT2¼ 2
› ln P
›T
� �Cm
ð11Þ
an expression for the isosteric heat can be obtained.
2DH ¼ A þ DHvap þðbE0Þ
ndT
nAn21ð12Þ
Thus, the isosteric heat is a summation of three terms. The
first term is due to the adsorption potential, the second is
the heat of vaporization and the third corresponds due to the
change in maximum capacity with temperature.
To express the isosteric heat of adsorption in terms of
adsorption rate, the above expression can be rewritten as:
2DH ¼ DHvap þ bE0 ln1
u
� �þ
ðbE0ÞdT
nln
1
u
� �2ðn21Þ=n
ð13Þ
where the fractional loading u can be expressed as Cm=Cms:
The above expressions (11)–(13) are often used in the
study of adsorption refrigeration system and the detail of
adsorption equilibria, working pair and heat of adsorption is
summarized in Table 2.
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327306
Table 2
A review of adsorption equilibrium equations used in different models.
References Equilibrium equation Adsorbent-adsorbate pair Adsorption heat
Hajji, 1995, 1996
Zhang, 1999, 2000 [61,93]
lnðpÞ ¼ aðwÞ þ bðwÞ=T aðwÞ ¼ a0 þ a1w þ a2w2 þ a3w3;
bðwÞ ¼ b0 þ b1w þ b2w2 þ b3w3 ¼ DHðwÞ=R
4A zeolite–water, 13X zeolite–water,
AC35 activated carbon–methanol
Enthalpy of adsorption:
bðwÞ ¼ b0 þ b1w þ b2w2 þ b3w3 ¼ DHðwÞ=R
Cho, 1992 Simplified Freundlich’s equation: simplified Freundlich’s
equation: q ¼ qsatðP=PsatÞ1=n; qsat ¼ 0:522 kg kg21, n ¼ 1:6
Fuji RD-type silica gel–Water
[14] D-A equation: q ¼ W0rðTÞ exp½2ðA=bE0Þn� Adsorbent (charcoal): AC35, DEG,
LH, BPL, NORIT RB, PKST,
207E and C; adsorbate: NO2,
methanol, SO2, amine, methyl,
ammonia, acetonitrle and
formaldehyde.
Clausius–Clapeyron equation:
DH ¼ R lnðp2=p1Þ=ðð1=T1Þ2 ð1=T2ÞÞ
[43] Modified DR equation Carbon–ammonia Constant
Gui, 2001, 2002,
Chahbani, 2002 [38,75]
Modified D-A equation: q ¼ q0exp½2kðT=Tsat 2 1Þn� Active carbon–methanol [17,18,28];
monolithic carbon–ammonia [35]
Constant
Guilleminot, 1987,
Boubakri, 2000, Passos, 1989,
Mhimid, 1998, [41]
q ¼ W0r exp 2D T lnPsatðTÞ
PsatðTcondÞ
� �n� �Activated carbon AC-35–methanol;
activated carbon AS–methanol;
carbon AX21–ammonia; zeolite–water
Clapeyron equation for ideal gas:
DH ¼ RT2ð› lnp=›TÞ
Llobet, 2000 D-A equation Adsorbent (activated carbon):
PX21, KL93, KF1500, TA90, TA60,
BPL; adsorbate: ammonia
QstðT ;PevapÞ ¼ RT2½› lnðpÞ=›T�
¼ DHNH3þ A þ 2:5 £ 1023TbnEn
0A12n=n
Luo, 2000 Modified DR equation: q ¼ q0exp½2kðT=Tsat 2 1Þn� Activated carbon AC-35–methanol,
Activated carbon AC-35–ethanol
DH ¼ RT þ L þ ð2303b=ffiffik
pÞ £ ½ln1=2ð1=uÞ
þðaT=2Þln21=2ð1=uÞ� k;b are coefficients,
a is the thermal expansion coefficient of
liquid, and u ¼ q=q0
Saha, 1995, Chua, 1999 q ¼ AðTadÞ½PsatðTrÞ=PsatðTadÞ�BðTadÞ; AðTadÞ ¼ A0 þ A1Tad þ
A2T2ad þ A3T3
ad; BðTadÞ ¼ B0 þ B1Tad þ B2T2ad þ B3T3
ad ; A:
26.5314, 0.72452 £ 1021, 20.23951 £ 1023, 0.25493 £ 1026; B:
215.587, 0.15915, 20.50612 £ 1023, 0.53290 £ 1026
Silica gel–water Constant
Sakoda, 1984,1986,
Saha, 1995, Alam, 2000
Simplified Freundlich’s equation: q ¼ qsatðP=PsatÞ1=n; qsat ¼
0:346 kg kg21; n ¼ 1:6
Fuji-gel A type silica gel–water Constant
Sami, 1996 lnðp=psatÞ ¼P
i Aici þ ð1=TaÞP
i Bici Carbon–methanol, AC-HCFC123/
HCFC124
Clapeyron equation for ideal gas
Shelton, 1990, Fuller, 1994 q ¼ a1 þ a2ðTz=TsatÞ þ a3ðTz=TsatÞ2 þ a3ðTz=TsatÞ
3 Zeolite–ammonia Activated
carbon–ammonia
Constant
Sward, 2000 [41,52] qv ¼qs1b1p
1 þ b1pþ
qs2b2p
1 þ b2pþ
qs3b3p
1 þ b3p
qs;k and bk ðk ¼ 1; 2; 3Þ are function of temperature
Zeolite 13X–water Zeolite
13X–ammonia
Constant
[40] q ¼ 221:2 exp½21:916 £ 1027ðT lnðPsðTÞ=PÞÞn� Zeolite NaX–ammonia Constant
[74] Q ¼ f ðTad; pÞ Activated carbon–ammonia Constant
K.
Su
ma
thy
eta
l./
Pro
gress
inE
nerg
ya
nd
Co
mb
ustio
nS
cience
29
(20
03
)3
01
–3
27
30
7
3. Thermodynamics of adsorption cycles
Adsorption refrigeration with different cycles have been
studied extensively [26], and the typical adsorption
refrigeration cycles are: basic cycle, continuous heat
recovery cycle, mass recovery cycle, thermal wave cycle,
convective thermal wave cycle, cascade multi effect cycle,
hybrid heating and cooling cycle etc.
3.1. Basic adsorption cycle
A basic adsorption cycle consists of four thermodynamic
steps, which can be well represented with the aid of the
Clapeyron diagram, as shown in Fig. 1. The idealized cycle
begins at a point (point A in Fig. 1) where the adsorbent is at
low temperature TA and at low-pressure PE (evaporating
pressure). A–B represents the heating of adsorbent, along
with adsorbate. The collector is connected with the
condenser and the progressive heating of the adsorbent
from B to D causes some adsorbate to be desorbed and its
vapor to be condensed (point C). When the adsorbent
reaches its maximum temperature TD; desorption ceases.
Then the liquid adsorbate is transferred into the evaporator
from C to E and the collector is closed and cooled. The
decrease in temperature D to F induces the decrease in
pressure from PC to PE: Then the collector is connected to
the evaporator and adsorption and evaporation occur while
the adsorbent is cooled from F to A. During this cooling
period heat is withdrawn to decrease the temperature of the
adsorbent.
3.2. Heat recovery adsorption refrigeration cycle
The semi-continuous heat recovery cycle is usually
operated with two adsorption beds. The adsorber to be
cooled will transfer its heat to the adsorber to be heated,
which includes sensible heat as well as heat of adsorption.
This heat recovery process will lead to a higher system COP.
Multi-beds could be also adopted to get more heat recovery
and thereby to attain higher COP, but the operation of
a practical system will be complicated. A quasi-continuous
adsorption refrigeration system with heat recovery was
investigated by Wang et al. [26,27] and the flow path is
shown in Fig. 2. While adsorber 1 is cooled and connected to
the evaporator to realize adsorption refrigeration in
evaporator, the adsorber 2 connected to the condenser is
heated to obtain heating-desorption-condensation. The
condensed refrigerant liquid flows into evaporator via a
flow control valve. The operation phase can be changed, and
the go-between will be a short time heat recovery process.
Two pumps are used to drive the thermal fluid in the circuit
between two adsorbers (the connection to the heater and
cooler are blocked during this process).
Miles [28] obtained an experimental COP of 0.8 for the
activated carbon/ammonia system with an evaporation
temperature and a condensation temperature of about 5
and 35 8C, respectively. At the same time, Istria et al. [29]
achieved approximately the same theoretical value for these
temperatures with the multi-salt system. Pons et al. [30]
investigated the zeolite þ water adsorption system under
the fixed test temperature and offers the best possible
theoretical maximum COP of 1.5.
Jones [31] suggested an improvement to the process by
installing more than two adsorbers into the system. The
operating principle of the cycle remains the same, relying on
heat transfer flowing between the adsorbers and the
desorbers. As compared to the basic cycle, heat recovery
in this process is only effective if the heat transfer fluid
temperature leaving the adsorbers is sufficiently high.
Simulation results have shown that the maximum value of
the COP depends on the number of adsorbers and desorbers
installed. The analysis was further extended to a system
containing six adsorbers and six desorbers at the same test
temperature conditions (evaporation at 5 8C and conden-
sation at 35 8C) and it was possible to obtain COPs in the
range of 1.16 [32].Fig. 1. Clapeyron diagram (ln P vs.21=T) of ideal adsorption cycle.
Fig. 2. Schematics of heat recovery two-beds adsorption refriger-
ation system.
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327308
In order to simplify the management of the cycles and to
achieve continuous operation avoiding any control during
the cycle as well as to obtain good heat regeneration
between adsorption and desorption, systems with several
elementary adsorbers coupled with evaporators/condensers
and rotated about a central axis, have been proposed [33,34].
Maier-Laxhuber [35] proposed a rotary system which, in
principle, can be used for continuous production of cold
based on a single effect adsorption cycle. Erickson [36]
adapted the principle of the rotary system to a double effect
chemical reaction cycle operating at three pressure levels.
Although there is a problem for the practical management of
the air flows, i.e. the heat transfer fluid used in rotary
systems, these processes seem highly attractive as continu-
ous cooling effect as well as high COP, in the range of 1, can
be obtained.
Based on above reasons, Ebbeson [37] proposed a rotary
process which is designed for continuous operation with the
concept of a heat regeneration developed for solid sorption
cold production systems. The theoretical COP was about
0.9, which is comparatively low, due to the consideration of
the thermal masses and thermal pinches necessary for heat
exchange between the air and the elementary modules.
In a recent work, Critoph [38] describes the operation of
a continuous multiple-bed regenerative adsorption system.
The complete system, consisting of 32 modules was
modeled in detail. The design has not yet been optimized
but a parametric study has revealed the way in which key
parameters affect COP and SCP. The effect of five major
parameters such as thermal capacity ratio, number of
modules, temperature of air leaving the heater, generator
heat transfer coefficient and the evaporator air inlet
temperature were reported.
3.3. Mass recovery adsorption refrigeration cycle
Apart from the above discussed heat recovery operation,
it had been proved that mass recovery is also very effective
for heat recovery adsorption heat pump operation. In this
process, at the end of each half cycle, one adsorber is cold
and the other one is hot. Meanwhile, the former one which is
at low pressure ðPCÞ must be pressurized up to the conden-
ser pressure, and similarly, the other one which is at high
pressure must be depressurized down to the evaporator
pressure. With just one tube between the adsorbers and a
vapor valve, part of this pressurization–depressurization
can be achieved by transferring vapor from the latter
adsorber to the former one. This process can also be called
as an ‘internal vapor recovery process’, and is reported to
enhance the cooling power of the unit without reducing the
COP by more than 10%.
The above explained process involves only mass transfer
and hence the process is rapid. To obtain a ‘double effect’,
mass recovery could be initiated followed by heat recovery.
An ideal heat and mass recovery cycle is shown in Fig. 3, in
which the heat recovery state for a two bed system is shown
by the state points e 2 e0:
The mass recovery cycle ða2 2 a3 2 g01 2 g1 2 g2 2
g3 2 a01‘ 2 a1 2 a2Þ is an extended form of a two bed
basic cycle or two bed heat recovery cycle
(a2 2 g1 2 g2 2 a1 2 a2 shown in Fig. 3), and the cycled
mass is increased from Dx to Dx þ dx; which causes the
refrigeration effect to increase. The principle of these cycles
can be described using Fig. 3. The very first part of each half
cycle is the mass recovery process (path g2 2 g3 and
a2 2 a3). Then the heat recovery process proceeds: heat is
transferred from the hot adsorber to the cold one (path
g3 2 e0). As a consequence, the hot adsorber is first
depressurized (path g03 2 a01), it then adsorbs vapor from
the evaporator (path ). Meanwhile, the cold adsorber is first
pressurized (path a3 2 g01), and then vapor that is desorbed
passes into the condenser (path g01 2 e). Theoretically, the
heat recovery process develops until the adsorbers reach the
same temperature. Actually, there still remains a tempera-
ture difference between the adsorbers at the end of this
period. Then, for closing each half cycle, the adsorbers are,
respectively, connected to the heat source and heat sink
(path e 2 g2 and e0 2 a2). The second half-cycle is
performed the same way except that the adsorbers now
exchange their roles. Due to this process, about 35% of the
total energy transmitted to each adsorber can be internally
recovered, including part of the latent heat of sorption.
3.4. Thermal wave cycle
To further improve the heat regenerative ratio, Shelton
[39] had proposed an attractive cycle called ‘thermal wave
cycle’. In this process, it is assumed that a large temperature
gradient exists along an adsorption bed.
Heating and cooling of the adsorbent beds is achieved
via a heat transfer fluid such as high temperature oil. The
system consists of two adsorber beds and two heat
Fig. 3. Diagram of heat and mass recovery cycle.
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327 309
exchangers connected in series (Fig. 4) to effect semi-
continuous process. The function of the bed and heat
exchanger is to combine a large area of heat transfer surface
with a low oil flow rate.
A typical thermal wave cycle is shown in Fig. 4. The
cycle consists of two phases: In the first phase, the oil
recovers heat from bed 2 (hot), has a further heat addition
from the heat exchanger and then proceeds to heat bed 1
(cold). As the heating of the bed proceeds, bed 1 desorbs
refrigerant which passes to the condenser (giving a useful
heat output in the case of a heat pump) and bed 2 adsorbs gas
from the evaporator which provides cooling. In the
following phase (second phase) of the cycle the pump is
reversed, and hence, bed 1 is cooled (adsorbing) and bed 2 is
heated (desorbing) in a similar fashion until the original
conditions are reached and the pump can again be reversed.
Though the procedure is simple, significant heat recovery
can be achieved. Further, the system would achieve much
better performance due to the combination of the special
nature of the internal bed heat exchangers and the low flow
rate.
Although many researchers have studied the cycle, up to
now, there is no report of a successful prototype adopting
thermal wave cycle. Also, some experimental reports had
shown that the performance of the thermal wave cycle is not
very good. The efficiency of the thermal wave regenerative
system depends on a relatively large number of parameters:
for example, rates of various heat transfer processes, the
flow rate of the circulating fluid, the cycle time, the adsorber
configuration, etc. A numerical analysis of adsorptive heat
pumps with thermal wave heat regeneration had been
presented by Sun et al. [40]. They had derived two time
constants which can be used directly to quantify the relative
importance of the two heat transfer processes. This allows
ready determination of which of the two processes is rate
limiting and needs to be improved. The work has also
confirmed that the performance of an adsorptive heat pump
system using a traditional packed-bed would be too low,
even with a heat regeneration, and therefore a significant
enhancement of heat transfer properties inside the adsorber
is necessary.
Similar to the above numerical study, the effect of
various operating parameters on the performance of an
adsorptive thermal wave regenerative heat pump had been
studied by Ben Amar et al. [41], theoretically. They had
developed a two-dimensional model which simultaneously
considers heat and mass transfer in the bed. The results have
shown that under ideal conditions, the performance of a
thermal wave regenerative heat pump is considerably better
than that of a basic ‘uniform temperature’ heat pump. The
study showed that a COP greater than 1 and a power of cold
production of near 200 W per kg of adsorbent could be
obtained. For air-conditioning applications, these figures are
slightly higher than those obtained with other single-stage
solid-gas systems as chemical heat pumps [42].
3.5. Convective thermal wave cycle
Thermal wave cycles normally suffer from low power
density because of poor heat transfer through the adsorbent
bed. Rather than attempting to heat the bed directly, it is
possible to heat the refrigerant gas outside the bed and to
circulate it through the bed in order to heat the sorbent. The
high surface area of the grains leads to very effective heat
transfer with only low levels of parasitic power needed for
pumping. Hence, Critoph [43,44] has presented a modified
version of a thermal wave cycle, known as ‘convective
thermal wave cycle’. The concept is the same as thermal
wave cycle, however, the thermal fluid for heating and
cooling to the beds is initiated by the refrigerant itself, thus
the heat transfer between thermal fluid and adsorption bed is
a direct contact heat transfer, which is incorporated with
mass transfer in the system.
A practical schematic of the proposed system is shown in
Fig. 5. The two ‘active’ beds are packed with activated
Fig. 4. Thermal wave cycle.
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327310
carbon and the two ‘inert’ beds are packed with non-reactive
particles such as steel balls. The diagram shows the first half
of the cycle, during which Active bed 1 is heated and
desorbs ammonia while Active bed 2 is cooled, adsorbing
ammonia.
In the fluid circulation loop shown on the left, a low
power pump circulates ammonia steam through inert bed 1
which is initially hot. The gas stream is heated by the bed
and a ‘cold’ wave passes through the bed from right to left.
Having been preheated by the inert bed, the ammonia stream
is heated to the maximum cycle temperature (150–200 8C)
in a heat exchanger. The ammonia gas then passes to active
bed 1 where it heats the carbon. A ‘hot’ thermal wave passes
from left to right through the active bed. As the temperature
of the active bed rises it desorbs ammonia which first
increases the pressure in the left hand loop and then
condenses in the condenser, rejecting heat to the environ-
ment. The mass flow rate of circulating ammonia is typically
ten times that of the condensing stream of ammonia and it
may take about ten minutes for the two thermal waves to
travel the length of their respective beds. In similar fashion
to the left hand loop the circulating flow might be ten times
the adsorption flow from the evaporator.
The advantages of this system are:
† The four packed beds are in effect heat exchangers of
very high surface areas but at minimal cost. They are not
only cheap but very compact.
† There are only four conventional heat exchangers and
this is the minimum number allowed by thermodyn-
amics. These are the evaporator and condenser, a gas
heater whereby high-grade heat is input and a gas cooler
whereby the low grade heat of adsorption is rejected to
the environment.
† The cycle is highly regenerative since the packed beds
act like large counterflow heat exchangers. This results
in good energy efficiency (i.e. high COP).
3.6. Multi-stage and cascading cycle
The adsorption cycles discussed in previous sections are
applicable only to a single stage cycle. The single stage
cycle systems have certain limitations, that is, they cannot
effectively utilize high temperature heat source, as well as
do not perform well at very low temperatures. Hence, to
improve the system performance under such situations,
adsorptive processes may be adapted for advanced cycles,
such as, multi-stage and cascading cycle.
The basic idea of a multi-stage cycle is to perform the
desorption –condensation processes and evaporation–
adsorption processes at different temperature/pressure levels
by using the ‘same working pair’. The internal re-use of heat
of condensation or adsorption can increase the system
performance significantly. Another practical cycle that can
make good use of high temperature heat source is the
‘cascading cycle’, which operates with ‘different working
pairs’ (either liquid/liquid or solid/liquid), such as zeolite–
water/activated carbon–methanol, or zeolite–water/silica
gel–water, etc. These cascading cycles are applied to
situations especially, when there exists a large temperature
difference between the heat source/ambient and the
temperature in the evaporator/refrigeration space. For such
situations, it may not be practical to use single stage cycle.
Hence, one way of dealing with such situations is to perform
the evaporation/refrigeration process in stages, that is, to
have two or more cycles that operate in series at different
temperature levels (cascading). A high temperature heat
source (e.g. boiler) is used to drive the high temperature
Fig. 5. Adsorption refrigeration system with convective thermal wave cycle.
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327 311
stage adsorption refrigeration cycle. The low temperature
stage adsorption refrigeration is driven by sensible heat and
heat of adsorption obtained from high temperature stage. To
minimize the contribution of sensible heat, special care has
been attached to the heat management of the adsorbers;
n-adsorber cycles operating with a single evaporator and a
single condenser have been proposed with sequences of heat
recovery between adsorbers. Such cycles offer some
advantages: for example, a single condenser is used and
pressure in the n-adsorber unit is not higher than that in the
unit operating an intermittent cycle; moreover, adsorption
heat at high temperature is used as desorption heat at low
temperature. Counteracting heat transfer fluid circuits
between adsorbers reduces entropy generation in compari-
son with what happens in intermittent cycles. The driving
heat supplied to the cycle is at high temperature level (Fig. 6)
so that the entropy generation—due to the inadaptation
between the temperature levels of the source and of the
adsorber is much less in an n-adsorber cycle than in an
intermittent cycle. The same thing happens for the rejected
heat: the rejection temperature is much closer from the
utility temperature with an n-adsorber cycle than with an
intermittent cycle.
Very similar conclusions to that drawn by Scharfe et al.
[45] have been presented by Meunier [46] in the case of heat
recovery between adsorbers. In a particular case, Meunier
has shown that using an infinite number of adsorbers with
ideal heat recovery between adsorbers, the maximum
achievable—with given conditions of operating tempera-
tures—would be a cooling COP equal to 1.85 corresponding
to 68% of ideal Carnot COP. An other cascading cycle
which includes a triple effect machine operating a cascade
between a water zeolite heat pump and a single stage LiBr–
H2O refrigerator has been tested in Munich [47]. A lot of
theoretical studies exists on the possibilities of cascading
cycles while few experimental data are available.
4. Solar cooling technologies
In order to evaluate the potential of the different solar
cooling systems, a classification has been made by Best
and Ortega [48]. The relevant cooling technologies are:
† intermittent adsorption;
† continuous adsorption;
† diffusion; and
† absorption systems;
Adsorption cycles which were detailed in Section 3.6,
can be used in heat-driven refrigerators, air conditioners or
heat pumps in which the energy source can be solar energy
or waste heat. Examples include the use of waste heat to
provide air-conditioning, heating, industrial refrigeration,
and domestic refrigeration in parts of the developing world.
There is a well-documented need for domestic refrigeration
in areas that do not have access to grid electricity. Spoilage
of many products, particularly fish, can be as high as 50%.
This crisis has led to various international agreements
including the Kyoto ‘Declaration and Plan of Action on the
Sustainable Contribution of Fisheries to Food Security’
(NRI, 2000) [49].
Renewed interest in adsorption refrigeration is based on
the advantages of (i) being CFC-free; (ii) cost effective;
(iii) simple construction; (iv) no need for solution pumps,
and (v) they can be driven by low grade energy heat.
Nineteen different possible technologies have been ident-
ified [48] for solar refrigeration. Photovoltaic/vapor com-
pression and photovoltaic/thermoelectric systems have been
predominant in the application of small refrigerators for
vaccine storage in isolated areas where high cost is justified.
In this section, brief details on various solar refrigeration
systems have been listed and their unique features are also
discussed. Absorption and diffusion systems are not
included in this study as the main focus is on adsorption
systems.
4.1. Intermittent adsorption systems
Because of the intermittent nature of solar energy,
intermittent adsorption refrigeration cycles have long been
considered as logical approaches to solar cooling systems.
Adsorbent–refrigerant combinations which have been
examined for solar application include silica gel–water,
Fig. 6. n-Adsorber cascading cycle.
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327312
zeolite–water, activated carbon–methanol and activated
carbon–ammonia. The theoretical principles of basic
intermittent adsorption cycles have been described in detail
in Section 3.
4.1.1. Silica-gel/water and silica-gel methanol systems
Since early 1980s, the work on silica-gel/water systems
have been popular and lot of work was carried out mainly in
Japan. In an effort to utilize solar heat, Sakoda and Suzuki
[50] achieved a solar COP of about 0.2 with a solar collector
500 £ 500 £ 50 mm3 depth, packed with 1 kg of silica-gel
particles and with 1.5 kg of distilled water in the evaporator.
On a clear day with total solar insolation of 19.3 MJ m22
day21, it was estimated that a COP of about 0.4 can be
possible just with 0.4 m2 solar collector.
Hisaki et al. [51], utilizing the benefits of cogeneration,
manufactured a 34.5 kW (10 TR or tonne of refrigeration)
adsorption chiller and ice thermal storage system, compris-
ing two adsorption units, an evaporator and a condenser, as
shown in Fig. 7. Both adsorbers were installed in an
evacuated hermetic enclosure, and each adsorber was
alternatively heated and cooled in a cycle time of 7 min.
The COP of the system was found to be 0.4. They had
reported that a 100 TR system can obtain a COP of about
0.6. Based on the above approach, Saha et al. [52], also had
worked on a 10 kW system using silica gel as the working
media. The silica gel was packed around finned tubes and
the experiments had shown that each adsorption/desorption
cycle takes nearly 3–13 min and heat-exchange cycle is
about 30 s. The system could achieve a COP of 0.48
and was found that the COP tends to rise with cycle time
above 5 min and with the flow rate of chilled water.
A solar powered refrigeration system with a 0.25 m2 flat
plate collector has been developed [53]. This system would
realize an evaporating temperature of 5 8C when the
condensation temperature was around 35 8C and the attained
regeneration temperature was 100 8C. Ortel and Fischer [54]
used methanol with silica gel instead of water, so that the
system could operate at an evaporating temperature below
0 8C. It was found that a two-chamber silica-gel/water
system could be operated by methanol also, but because of
methanol’s inferior thermodynamic properties the COP of
the system was considerably reduced by about 30%.
4.1.2. Zeolite–water systems
Since the late 1970s, significant efforts were made by
Tchernev [55] to develop a solar-powered natural-zeolite/
water system. Natural zeolites, with a heat of adsorption of
2.8 MJ kg21, were found to be more suitable for cooling
systems than synthetic zeolites having a heat of adsorption
of 4.2 MJ kg21. This is due to the fact that with lower heat
of adsorption the adsorbent bed can be cooled more
effectively, which in turn results in better adsorption of
refrigerant. This air-conditioning system with the exposed
collector area of 1 m2 could produce ice at a rate of
6.8 kg day21. Tchernev [55] also designed and tested a solar
adsorption machine with different combinations of working
media. Flat plate collector having an exposed area of
2.44 m2 was used to produce 100 kg of ice per day. An
additional 1.5 m2 flat plate collector was used along with the
system to refrigerate milk. However, its performance
Fig. 7. Adsorption chiller.
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327 313
was found to be unsatisfactory, which was due to the
presence of inert gas inside zeolite bed chamber.
Dupont et al. [56] developed a 13X-zeolite-water cold-
storage system, which had an ice-making capacity of
10 kg day21. The system could attain a solar COP of 0.15.
Similarly, in France, Grenier et al. [57] had developed a
system using a simple solar collector with an area of 20 m2
containing 360 kg of NaX zeolite for a cold storage plant.
The cold production during evaporation was 1.93 MJ m22
and net production was only about 1.88 MJ m22 when the
incident radiation was 17.8 MJ m22. The operation con-
ditions were set to a condensation temperature of 32 8C,
evaporating temperature at 1 8C and the regeneration
temperature at 118 8C. The system could attain a net solar
COP of 0.105, while its cycle COP was 0.38. Tather and
Erdem-Senatalar [58] also utilized zeolite–water as work-
ing pair and reported that, the system could operate
successfully if the ambient temperature is less than 20 8C.
In the literature [53] it is mentioned that Guilleminot et al.
tried to use zeolite composites to replace zeolite as an
adsorbent and two different composites with different ratios
was used in his study. (i) 65% of zeolite þ 35% of metallic
foam; (ii) 70% of zeolite þ 30% of natural expanded
graphite. It was expected that, the latter combination
resulted in a better performance. A similar performance
was noticed by Poyelle et al. [59] and found that, if the
evaporation temperature is set to 24 8C, the required
condition temperature should not exceed 30 8C.
Meunier [60] had been a pioneer in the field of
adsorption. Initially, in 1979, his group theoretically
analyzed the zeolite and water pair and found that the
COP depends largely on the condensation temperature,
followed by evaporation temperature and then regeneration
temperature. In 1984, they developed an adsorption machine
which was operated at different temperature conditions.
When TC ¼ 30 8C; TE ¼ 5 8C and TR ¼ 70 8C, the system
could attain a COP of 0.14, while when the TC and TE was
set to 48 and 121 8C, respectively, the COP reduced
drastically to 0.04. Further work was continued and in
1986, they had attempted to use various working media like
Z13X and Z4AX-water in order to improve the performance
of the system.
A detailed thermodynamic model had been developed by
Cacciola and Restuccia [61] to obtain all the data needed
during the design step of an adsorption machine. Using this
model, it is possible to calculate the operative conditions,
the energy balance in each component of the machine and
the COP.
From the comparison of performance reported for
different adsorbent/adsorbate pairs, it appears that zeolite/
water is the most suitable pair to realize machines to be used
in domestic applications in southern European countries,
where the market demand for devices that are capable to
produce heating during winter and cooling during summer is
increasing considerably.
4.1.3. Activated carbon–methanol systems
A solar-powered adsorption system using activated
carbon–methanol pair was investigated by Delgado et al.
[62] in France. With an exposed area of 4 m2 solar collector
area, it would produce nearly 25 kg ice in a day. While the
thermodynamic COP was 0.4, the overall system COP was
only 0.15. Li and Sumathy [63] had presented the
description and operation of a solar-powered ice-maker
with the solid adsorption pair of activated carbon and
methanol using basic adsorption cycle. A domestic type of
charcoal was chosen as the adsorbent, and a simple flat-plate
collector with an exposed area of 0.92 m2 was employed to
produce ice about 4–5 kg per day at an evaporator
temperature of about 26 8C. The above system could
achieve solar refrigeration COP of about 0.1–0.12. A
similar experiment was performed by Pons et al. [11] with
activated carbon and methanol as the working pair and the
condensers used were of air-cooled type. The system was
loaded with 130 kg of activated carbon, which produced
around 30–35 kg of ice on sunny days. The system
employed flat-plate collector having an exposed area of
6 m2 and was packed with charcoal at a ratio of 20 kg m22.
It was reported that the system COP was about 0.12 and was
comparable with Delgado’s results.
They further extended their work at Agadic, Morocco
and demonstrated the operation of a mini-ice maker which
could produce about 4 kg of ice a day. The limitation of the
performance was mainly due to the ambient temperature,
and an increase in the ambient temperature beyond 23 8C
reduced the system performance significantly.
Since 1986, at AIT Bangkok, Exell et al. [64] have
worked extensively on the development of a charcoal–
methanol system. With just 1 m2 collector area, the system
could produce 4 kg of ice a day. The COP of the system was
about 0.123 which is marginally better than the systems
reported by other researchers. With a similar collector of
area (0.92 m2), Meunier et al. [60] as well as Cacciola and
Restuccia [61] also carried out many experimental investi-
gations on development of activated carbon–methanol
adsorption machines. Their observations were similar to
that reported by Pons and Guilleminot.
4.1.4. Activated carbon–ammonia systems
Investigations in the use of charcoal–ammonia are
apparently more recent—mainly during the 1990s. Jones
[65] used a novel carbon-moulding technique and incorpor-
ated a thermal wave-regeneration concept employed in the
drying of gas streams; a small unit consuming 0.51 kg of
charcoal produced 293 W cooling with an adsorbent heating
and cooling cycle of 6 minutes, with ammonia as the
adsorbate. With R22 and R134a as adsorbates, cooling rates
of 113 and 99 W were reported. Jones [65] also reported that
larger multi-bed systems were under development which
could have a cooling COP of 1.0. Miles and Shelton [66]
were involved in a similar development work, they claimed
to have achieved a heating COP of 1.21 and a cooling COP
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327314
of 0.76, using a two-bed system with half-cycle time of
2.6 min.
In order to improve the heat transfer effectiveness, finned
tubes are used instead of a simple flat-bed [53], which had an
outer diameter of 53 m and relatively length (L ¼ 0:2 m).
The system was tested at various operating conditions, and
was shown to be optimum, when the condensation
temperature of and regeneration temperature were 20 and
25 8C, respectively.
4.2. Continuous adsorption systems
Continuous solar adsorption refrigeration systems are
being reported widely because of their higher system
performance over intermittent alternatives and for their
timely coincidence with the requirement of the cooling and
refrigeration demand. Continuous adsorption can be rea-
lized through various cycles which are detailed in (Sections
3.2–3.6). Hence, various continuous adsorption cycles
depending on their technologies, can be grouped as:
† multi-stage and cascading systems;
† thermal wave adsorption systems;
† convective thermal wave adsorption systems; and
† hybrid systems.
In general, thermal wave and convective thermal wave
adsorption systems are also called as ‘heat regenerative
systems’ owing to its heat recovery properties.
4.2.1. Multi-stage and cascading systems
A lot of theoretical studies exist on the possibilities of
cascading cycles while very few experimental data are
available. Douss and Meunier [67] had reported the
experiments on a cascading adsorptive heat pump. The
cycle comprised of a two-adsorber zeolite–water system at
high temperature stage and an intermittent active carbon–
methanol system at low temperature stage. Driving heat was
supplied by a boiler to zeolite adsorbers while active carbon
adsorber was heated by heat recovered from zeolite adsorber
under adsorption. Evaporators from both basic cycles
operated at the same temperature and contributed to the
evaporating load. Experimental cooling COP was found to
be 1.06, which is much higher than the COP of an
intermittent cycle (<0.5) and that of a two-adsorber zeolite
water cycle (<0.75). An analysis of the results showed that
the components which limit the power of the unit were the
evaporators and basically it was the water evaporator.
The COP of the cascading cycle was very sensitive to the
evaporating temperature lift. Also, it was reported that
the cycle had adapted well to air conditioning as long as the
evaporation temperature lift was less than 45 8C. A similar
study on combined cycle comprising of an adsorption–
absorption cascading multi-effect refrigeration cycle was
carried out by Shu et al. [68]. The system consists of a high
temperature stage of solid adsorption unit working with
zeolite water and a low temperature stage of double effect
absorption unit working with LiBr–water The working
principle of this multi effect cascading cycle can be
explained using Clapeyron diagram shown in Fig. 8. The
single stage cycle with heat source Qh represents the
adsorption cycle using zeolite þ water as working medium.
The double stage cycle obtains the absorption heat of the
single stage cycle as its driving heat ðQadsÞ for its high
temperature stage. The heat obtained from the condenser of
the single stage cycle is used as the heat source for its low
temperature stage. The study has shown that the COP of the
cycle can be greatly improved by efficient utilization and
recovery of energy within the system. Also, they had
reported that there exists no corrosive problem though the
temperature of the working pair is higher than 200 8C in the
adsorption unit.
Saha et al. [69] has proposed a two-stage non-
regenerative adsorption chiller design and experimental
prototype using silica gel–water as the adsorbent refrigerant
pair. The main advantage of the two-stage adsorption chiller
is its ability to utilize low temperature solar/waste heat
(40-75 8C) as the driving heat source in combination with a
coolant at 30 8C. With a 55 8C driving source in combi-
nation with a heat sink at 30 8C, the COP of the two-stage
chiller is 0.36. Wang [26,70] carried a study on a four-bed
adsorption refrigeration system and the multi-stage system
Fig. 8. Adsorption–absorption cascading multi-effect refrigeration
cycle: (1) adsorption cycle; (2) high pressure absorption cycle and
(3) low pressure absorption cycle.
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327 315
configuration is shown in Fig. 9. The system consisted of
four adsorbers ðA1;A2;B1;B2Þ; one condenser (C), and one
evaporator (E). The working principle can be explained as
follows according to the Clapeyron diagram as shown in
Fig. 10: To begin with, A1 and A2 are in the high temperature
stage, for which generation (temperature about 200 8C) is
initiated by heat input. The desorbed water vapor will go
through the adsorber B1 or B2 at the low temperature stage to
release heat for desorption. The adsorption pressure of the
two stages are the same because only one evaporator is used
in the system. During phase 1, the desorption process in
desorber B1 is furnished by the heat from desorber A1 and
adsorber A2 in the high temperature stage. The coordination
of adsorption and desorption of the two stages is very
important to operate the system properly. By the way, the
desorption pressure in desorber A1 must be higher than
the desorption pressure of desorber B1 in order to facilitate
the vapor desorbed from A1 to pass through desorber B1
effectively. In this multi-stage system, water is the
refrigerant and the desorbed vapor flows through the adsor-
bent bed in desorber B1 to get a triple-effect installation.
Here, a double effect is accomplished with sensible and
adsorption heat recovery to generate low temperature stage
in addition to the heat input to the high temperature stage, a
triple effect is accomplished with the heat recovery of the
desorbed vapor in the high temperature stage in addition to
Fig. 9. The system configuration of four-bed cascading adsorption refrigeration cycle.
Fig. 10. Clapeyron diagram of four-bed cascading adsorption refrigeration cycle.
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327316
the heat used in a double effect arrangement to generate low
temperature stage.
The maximum energy recovery from high temperature
stage can be arranged by a triple effect system, in which the
sensible heat of adsorbent bed, heat of adsorption and the
latent heat of refrigerant vapor are fully used to drive the low
temperature stage system. The simulation work has shown
that, the performance of a cascading cycle interve of COP
can be triple (COP ¼ 1.56), compared to two independent
stage system (COP ¼ 0.6). This concept could be used for
multi-effect adsorption systems, as well as extended to
adsorption–absorption cascading systems [70,71].
4.2.2. Thermal wave adsorption systems
Thermal wave adsorption is the most commonly
attempted technology. In a two-bed adsorption system
shown in Fig. 4, substantial part of heat from the adsorbent
bed is rejected at a temperature greater than the temperature
of the bed being heated. Therefore, it was soon recognized
that part of this rejected heat could be used for heating the
desorbing bed, thus providing gains in efficiency. Shelton
et al. [72] and Fuller et al. [73] were among the pioneers in
devising analytical models for this technique, by means of a
single heat transfer fluid loop passing through adsorbent
beds, a fluid cooler, a fluid heater and a reversible
circulation pump. For a two-bed system, high temperature
thermal fluid flows into the adsorber, exchanges heat with
the bed, and the temperature goes down along the bed
rapidly, thus the outlet temperature will be close to ambient.
After being cooled by ambient surroundings, the fluid flows
into another adsorption bed, absorbs heat from the bed, and
the temperature of the fluid goes up. At the exit of this bed,
the thermal fluid temperature will be very close to the
temperature of heat source. In this case, only less heat is
added to the system, and less heat released to the
environment. Thus, heat recovery ratio is high and the
COP is increased significantly.
Analytical models of square and ramp waveforms
predicted that heating COP of up to 1.6 is possible. In a
similar effort, Zheng et al. [74], using a charcoal–ammonia
pair, found that system performance improves significantly
with a reduction of ambient temperature, and is also affected
by cycling speed and regeneration temperature.
Wang et al. [75] had developed two heat regenerative
adsorption systems, with two different types of heat
exchangers: (i) heat regenerative adsorption refrigerator
with a spiral plate heat exchangers adsorbers; (ii) plate fin
heat exchangers or plate fin shell heat exchangers adsorbers.
The activated carbon–methanol adsorption pair was used
for the above two adsorption systems. The two adsorbers
were independently operated for heating and cooling, along
with the intermediate heat recovery process. Fig. 4 shows
the schematic of the unit. A receiver was installed for easy
observation of refrigerant flow in the system. An ice box
was used which was cooled by a salt water cooling heat
exchanger. With a heat source temperature of 100 8C,
the refrigerator achieved a refrigeration power density of
more than 2.6 kg per day of ice per kg of activated carbon
with a COP of 0.13, and the heat pump achieved
150 W kg21 activated carbon for air conditioning with a
COP of about 0.4.
4.2.3. Convective thermal wave adsorption systems
Critoph et al. [43,44] suggested the use of a forced
convection effect by direct use of refrigerant vapors. During
desorption, the refrigerant is heated externally and passes
through the adsorber bed, where it heats the adsorbent
rapidly because of good heat transfer coefficient and the
large surface area of the adsorbent bed. The desorbed vapor
is condensed in a condenser in the usual way. During the
adsorption process, the flow is reversed and heat of
adsorption is removed by the refrigerant, which is cooled
externally. This technique enhances performance by virtue
of the combined effects of forced convection and thermal
wave. Flow of hot gas through the adsorbent gives rise to a
thermal wave traveling through the bed during desorption,
and vice versa during adsorption process. Based on this
theory, a new continuous adsorption refrigeration/heat
pump system using a number of simple tubular adsorption
modules was developed. Each single module comprised of a
generator and a receiver/condenser/evaporator. A complete
module was tested in a simple rig, which was subjected to
alternating hot and cold air streams, desorbing and
adsorbing ammonia.
The performance of a temperature wave regenerative
heat pump has been analyzed using a two-dimensional
model by Ben Amar et al. [41]. The model simultaneously
considers heat and mass transfers in the adsorbent bed. The
mathematical model has been solved numerically by use of
the ADI finite different method. The consideration of both
heat and mass transfer in the model has made it possible to
determine the impact of the convective transport of the
gaseous adsorbate in the adsorber on the global perform-
ance, in addition to that of the heat transfer. It has been
shown that with a traditional packed bed, which generally
has a low thermal conductivity (,0.2 W m21K21), the
performance of the resulting heat pump is very poor. In
contrast, with a consolidated adsorbent material, which can
have a much improved thermal conductivity but relatively
poor mass transfer properties, the heat pump performance
could be poor as well, due to the large mass transfer
resistances encountered by low pressure adsorbates like
water. A proper design of the adsorber should therefore
ensure good properties for both heat and mass transfer, in
particular when the adsorbate used has low working
pressures.
4.2.4. Hybrid system
Although solid adsorption refrigeration has received
much attention from most of the refrigeration companies
and laboratories around the world, some of the main
disadvantages such as long adsorption/desorption time
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327 317
and small refrigeration capacity per unit mass of adsorbent,
have become obstacles for the real mass production of the
system.
Better design of the adsorbers and improvements in the
thermal conductivity of the adsorption bed are the main
tools to reduce cycle time, while treatment of adsorbent to
increase its adsorption capacity on a refrigerant could
increase the refrigeration capacity of adsorbent per unit
mass and also the COP of the cycle. Besides, several
investigations had attempted to further improve (hybrid) the
performance of the existing solar refrigerators, (i) by
combining the basic adsorption cycle with other refriger-
ation cycles, as well as, (ii) in designing a dual system. The
features of the different hybrid systems are discussed in
Section 4.3.
In an attempt to overcome the intermittence of adsorp-
tion refrigeration, and realize the continuity, a novel
combined cycle of a solar-powered adsorption–ejection
refrigeration system has been studied by Li et al. [76]. As
shown in Fig. 11, it includes two subsystems: the adsorption
sub-system, which refrigerates at night time and the ejection
subsystem, which refrigerates during the day.
During the daytime, with valves v1; v4 kept open while v2
is closed, the vapor at high temperature and high pressure
produced in generator ðgÞ and auxiliary heater ðbÞ enters the
ejector and forms supersonic primary fluid at the exit of the
nozzle. This enables the entrainment of the secondary fluid
from the evaporator. The secondary fluid subsequently
mixes with the primary fluid in the mixing chamber of the
ejector, and when flowing into the condenser, the mixed
fluid condenses into liquid. This liquid is finally divided into
two parts: one part goes into the evaporator as in other
cycles; the other is pumped back to the generator [77]. At
the same time, the adsorber gets heated up by solar energy.
When the pressure in the adsorber rises to a certain value Pb
(correspondent with Tb), the valve v5 opens, the desorbed
vapor then enters the generator as the primary fluid for the
ejection cycle. The valve v5 should be closed and v3 should
be open, as the temperature within the adsorber reaches Tg2.
During night time, valves v1; v3; v4 and v5 are closed, and
when the temperature within the adsorber decreases to Ta2;
valve v2 opens. As a result, the adsorbent begins to adsorb
the refrigerant. In order to recover the sensible heat and the
heat of adsorption, a heat pipe is used. The recovered heat is
applied to heat the refrigerant or thermal storage material in
the generator for the ejection cycle the next day. When
thermal equilibrium is reached, the heat pipe will stop
working, then a heat exchanger ðx2Þ is needed to cool the
adsorber further.
The main advantage of this system is to overcome the
intermittence problem in solar-powered adsorption refriger-
ation, together with increasing the performance of the
ejection system. It is reported that increasing the tempera-
ture or reducing pressure within the adsorbent bed, the COP
of the ejection subsystem improve very slightly, but if much
more adsorbent is used, a better result might be obtained.
Though solar ice making is attractive by an adsorption
system, it needs both good solar collecting (high desorption
temperature) and heat release characteristics for the
adsorber (low adsorption temperature), which seems to be
Fig. 11. Schematic of the ejection refrigeration and adsorption refrigeration systems.
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327318
contradictory, specially for a single bed system. Because, in
order to obtain a desorption temperature, the losses from the
bed should be as low as possible. Whereas, during
adsorption period, the same bed should attain a low
adsorption temperature, by releasing heat to the surround-
ing. It should be noted that, heat release means energy
losses; therefore, in order to make the system more efficient
and useful in practice, heat recovery must be considered. A
hybrid system of solar-powered water heater and ice-maker
(Fig. 12) has been developed by Wang et al. [78]. The
system consists of a solar collector, water tank adsorber/
generator, condenser, evaporator, receiver, ice-box, etc. The
working principle is based on the combination of a solar
water heater and an adsorption refrigerator. In the morning,
the solar collector heats the water tank and along with the
increase of water temperature, the temperature in the
adsorbent bed rises. In an ideal process, the adsorbent
temperature could reach a level very close to the water
temperature in the tank.
When the temperature in the adsorbent rises up to a
temperature which causes the vapor pressure of the
desorbed refrigerant up to the condensing pressure,
desorbed vapor would be condensed in the condenser and
collected in the receiver. This liquid refrigerant then flows
to the evaporator via a flow rate regulating valve. On the
other hand the temperature of the water and the adsorbent
bed continues to rise to a maximum temperature of about
80–100 8C. The high temperature water is finally drained
out into a separate tank, be used more flexibly for any
domestic purposes.
The features of the hybrid system include (1) water
heating and refrigeration with one solar collector, which is
suitable for household applications; (2) adsorber/generator
is separated from collector, thus high efficiency vacuum
collector can be used for water heating, thereby heating
the adsorber at same time. The high efficiency heating does
not mean a bad cooling of the adsorber through the night, as
by draining the hot water from the tank, cold water is refilled
to the tank, thus the adsorber is cooled and refrigeration will
take place.
Recently, Dai et al. [79] have investigated the perfor-
mance of a hybrid solar cooling system (Fig. 13), which
combines the technologies of rotary desiccant dehumidi-
fication and solid adsorption refrigeration, for cooling grain.
The key components of the system are a rotary desiccant
wheel and a solar adsorption collector. The former is used
for dehumidification and the later acts as both an adsorption
unit and a solar collector. The heating load from sunshine
can thus be reduced to a greater extent since the solar
adsorption collector is placed on the roof of the grain depot.
Compared with the solid adsorption refrigeration system
alone, the new hybrid system performs better. Under typical
conditions, the COP of the system was round to be 0.4 when
the outlet temperature was ,20 8C. It is believed that the
system can be used widely in the regions with abundant
solar resources due to such advantages as environmental
protection, energy saving and low operation costs.
Yet another application of natural ventilation using the
adsorption refrigeration cycle has been studied by Dai et al.
[80]. The study presents a parameter analytical study of
a solar house (Fig. 14) with enhanced natural ventilation
derived by a solar chimney and a solar adsorption cooling
cavity.
The proposed solar house consists of two subsystems, i.e.
the solar chimney and the solar adsorption cooling cavity.
Both subsystems contribute to inducing outdoor air into the
room space. The components of a solar chimney include
glazing, air channel, selective coating, and a thermal storage
wall. It works on the principle of thermal buoyancy induced
by solar heating. Similarly, the solar adsorption cooling
system is configured by a flat plate adsorber, a condenser,
and an evaporator, etc. The adsorber is regenerated during
daytime when solar irradiation heats the adsorption bed and
desorption process begins; during night time cooling effect
is produced when the adsorber cools down and adsorption
process is initiated. A water tank is provided to store the
chilled water from the evaporator, which is circulated
(thermosyphoning) through a heat exchanger in the cooling
cavity. During this period, the ambient air is captured
through the cavity inlet (0), gets cooled by the chilled water
through the heat exchanger, and further descends down-
wards into the room. The air motion is mainly due to the
density difference of the air at the inlet and outlet of the
system.
Hence, during daytime, both solar chimney and cooling
cavity induce ventilation, while at night, ventilation is
induced mainly by the heat storage in solar chimney, the
heat released during adsorption process, and by a chilled
water circulation in the cooling cavity. In order to enhance
Fig. 12. Schematic of the hybrid system: (1) solar collector; (2)
water tank; (3) adsorber; (4) condenser; (5) receiver; (6) valve; (7)
evaporator; (8) refrigerator.
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327 319
Fig. 14. Schematic of the solar house with enhanced natural ventilation driven by solar chimney and adsorption cooling cavity [80].
Fig. 13. Schematic of the solar cooling system with rotary desiccant dehumidification and solar adsorption refrigeration [78].
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327320
the adsorption process, two dampers that enclose the space
between the glass cover and the adsorption bed are released
apart which facilitates the adsorption bed to be cooled more
efficiently. The chimney effect can thus be realized in the air
channel between the glass cover and the adsorption bed.
4.3. Solar desiccant cooling and air-conditioning systems
Desiccant technology has become a valuable tool in the
industry’s arsenal of space-conditioning options. In certain
cooling applications, desiccant cooling units provide
advantages over the more common vapor-compression and
absorption units. For example, desiccant units do not require
ozone-depleting refrigerants, and they can use natural gas,
solar thermal energy, or waste heat as heat input, thus
lowering peak electric demand.
An open-cycle desiccant cooling system operating in a
recirculation mode is shown in Fig. 15. The room air is
recirculated and ambient air is used only for regeneration.
Room air is dehumidified and heated by the desiccant wheel,
regeneratively cooled, and then evaporatively cooled prior
to re-entering the room. Ambient air is evaporatively
cooled, regeneratively heated, and then heated by an energy
supply. The heated ambient air passes through the
dehumidifier and regenerates the desiccant. For this system,
the sensible heat regenerator and dehumidifier are assumed
to be rotary elements, but direct transfer exchangers and
fixed beds could equally well be used.
A 10.5 kW open-cycle liquid desiccant dehumidification
system was designed and operated by Patnaik et al. [81].
Packed bed units were used to dry the air in the dehumidifier
and to concentrate the desiccant in the regenerator. Liquid
distribution in the regenerator was studied for two systems:
a gravity tray distributor, and spray nozzle system. Higher
capacities (40–50% increase) and lower pressure drop
(30–40% reduction) for the air flow were observed with the
spray system. Cooling capacities of 3.5–14.0 kW were
achieved for both the regenerator and dehumidifier.
Functional relationships correlating the independent vari-
ables to the rate of vaporization in the regenerator and rate
of condensation in the dehumidifier were obtained by
statistical analysis of experimental data. These studies thus
provide data and correlations useful for design guidance and
performance analysis of similar open-cycle liquid desiccant
systems.
Recent research in the development of a solar desiccant
cooling system has focused on the development of advanced
desiccant materials that give improved sorption capacity,
favorable equilibrium isotherms, and better moisture and
heat rates. Improved performance of these systems will
lower their initial costs and make these systems a more
attractive alternative to existing vapor compression systems.
An improved system with Type 1M isotherm shape
(modified Langmuir Type 1) has been proposed by Collier
et al. [82]. Further development of such a system were
continued by Belding et al. [83] and Worek et al. [84]. The
suitability of different technologies for binding silica gel
particles has been investigated [85].
Performance of various types of commercially available
desiccant wheels have been experimentally and analytically
evaluated by Collier et al. [86]. This study included the use
of lithium chloride, molecular sieve, silica gel and activated
carbon in the manufacture of wheels. Observations revealed
that the rate of moisture removal was roughly the same for
core materials of all types, which implies that cooling
system’s performance is similar for desiccant wheels of all
types. The important merits of solid-desiccant cooling
systems include: simplicity in construction and maintenance
Fig. 15. An open solid desiccant cycle.
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327 321
Table 3
An overview of developments in adsorption technologies
System Heat source Collector
area (m2)
Operating temperature (8C) Performancea Working pairs Investigators Remarks
TC TE TR TA COP SCP
(W kg21)
Ice
(kg)
Intermittent Solar heat 1 – – – 21–29 0.12 – – Activated carbon/methanol Boubakri [94] Collector-
condenser
technology
Experiment
Solar heat 2 – – – – 0.09–0.13 – – Activated carbon/methanol Buchter
et al. [95]
Experiment
Measured solar
irradiation
– 33 20.9 105 20.6–27.3 0.24 – 7.3 Activated carbon/methanol Leite and
Daguenet [96]
Simulation
Solar heat 0.92 – 26 70–78 6–13 0.1–0.12 – 4–5 Activated carbon/methanol Li and
Sumathy [63]
Experiment
Quartz lamp 1 28–35 3 – – 0.11 – 5–6 Activated carbon/methanol Li et al. [97] Experiment
– 1 40 0 – – 0.15 – – Zeolite 13-X/water Tchernev [55]
Heat
recovery
– – 35 – 120 – 0.75–0.85 – – Activated carbon/methanol Cacciola and
Restuccia [61]
Simulation
þZeolite/water Single stage
system
Boiler – – – – – 0.21–0.48 – – Silica gel/water Saha et al. [52] Experiment
– – – – – – 0.3–0.5 – – Silica gel/water Saha et al. [52] Simulation
Heater – 24–29.6 6–9.9 96.8–106 – 0.32–0.4 133–151 – Activated carbon/methanol Wang [27] Experiment
Mass
recovery
– – 30 5 80–120 – 0.6–0.8 – – Activated carbon/methanol Wang [27] Simulation
Mass
recovery
Generator – 38 8.9 – – 0.60 142 – Monolithic active carbon/
ammonia
Critoph [38] Experimental
simulation
Multi-bed
system
Cascading Thermostated
bath
– 35 25 – – 1.06 – – Zeolite/water þ activated
carbon/methanol
Douss and
Meunier [67]
Experiment
Cascading – – – – – – – – – Zeolite/water þ LiBr–water Shu et al. [68] Simulation
of COP
Multi effect
Thermal
wave
– – 35 2 – – – – – Ditto Miles and
Shelton [66]
Heater – 40 5 – – – – – Zeolite NaX/ammonia Sun et al. [40] Simulation
K.
Su
ma
thy
eta
l./
Pro
gress
inE
nerg
ya
nd
Co
mb
ustio
nS
cience
29
(20
03
)3
01
–3
27
32
2
Convective
thermal
wave
Heater – 42 0 – – 0.95 – – Carbon/methanol Critoph [43,44] Experiment
Hybrid – – 20 22.5 – – 0.064–0.067 – 10–10.5 Activated carbon/methanol Wang et al. [78] Experiment
Heating þ
ice making
Assumed solar
irradiation
2 15–35 210 84.9–100 – 0.039–0.046 – 3.05–8.7 Activated carbon/methanol Wang et al. [78] Simulation
Heating þ
ice making
– 0.4 – – 5 – 0.18 87.8 – Activated carbon/methanol Zhang and
Wang [98]
Simulation
Hybrid
collector
– 2 30 5 – – 0.33 5.67 – Zeolite 13-X/water Zhang and
Wang [99]
Simulation
Adsorption-
ejection
technology
Generator – 40 10 120 – 0.4 – – Zeolite 13-X/water Li et al. [76] Experiment
Adsorption-
ejection
technology
Binary
coupling
Electric heater – – – 130 – – – – Zeolite/Ammonia/water Wang and
Zhu [100]
Adsorption þ
absorption
Experiment
a The cooling performance of the system.
K.
Su
ma
thy
eta
l./
Pro
gress
inE
nerg
ya
nd
Co
mb
ustio
nS
cience
29
(20
03
)3
01
–3
27
32
3
(as only air and water are needed); low pressure operation;
improved indoor air quality; potential for high energy
savings; and operation with potential use of diverse energy
sources such as solar energy, natural gas and bio-gas. Their
major limitation is that their performance may be influenced
by variations in weather conditions. Their COP ranges from
0.5 to 1.0, although analytical models for modified systems
have predicted that COP values greater than 2.0 are possible.
An experimental investigation with a lithium-chloride-
impregnated wheel regenerated, by solar energy (using a
commercially available evacuated tube collector) showed
that 0.5 units of cooling can be obtained per unit of solar
energy incident on solar collectors. In summer, an average
cooling of 10 MJ m21 per day could be achieved [87]. The
use of desiccant wheels for humidity control has been tried
successfully in applications such as the prevention of metal
corrosion in storage, and in photolithography, by Harriman
[88,89]; their potential has been evaluated by Banks [90,91]
for use in applications including ice rinks and supermarkets.
Hybrid systems, which integrate desiccant dehumidifiers
with conventional cooling systems are proven to provide
substantial energy savings. The required energy can be
supplied by low temperature heat sources for the regener-
ation if the rotary dehumidifier matrix is properly con-
structed and appropriate selections are made for the type and
amount of the desiccant. The thermal energy can be
provided by a combination of an array of solar flat plate
collectors and natural gas [92].
5. Conclusions
The development of adsorption system for refrigeration
is promising. An overall thermodynamics-based comparison
of sorption systems shows that the performance of
adsorption systems depend highly on both the adsorption
pairs and processes [93]. The technology continues to
develop, and the cost of producing power with solar thermal
adsorption refrigeration is falling. If the costs of fossil fuels,
transportation, energy conversion, electricity transmission
and system maintenance are taken into account, the cost of
energy produced by solar thermal adsorption systems would
be much lower than that for conventional refrigeration
systems. This paper presents an overall review on the
fundamental understanding on the various adsorption
refrigeration cycles and the applicability of solar adsorption
both in air conditioning and refrigeration, with the
improvement of the COP.
Of the several kinds of adsorption systems analyzed in
this paper, the intermittent system has been extensively
studied both theoretically and experimentally, owing to its
simplicity and cost effectiveness. However, the main
disadvantages such as long adsorption/desorption time
have become obstacles for commercial production of the
system. Hence, to compete with conventional absorption
and vapor-compression technologies, more efforts should be
made in enhancing the COP and SCP (A summarization on
the different systems along with their respective COPcooling
and SCPs is listed in Table 3). Heat recovery, mass recovery,
multi-bed and multi-stage technologies are promising
technologies in improving the COP and SCP. Also, recent
developments [94,95] of activated carbon series—activated
carbon fibre (ACF)—has called attention to the possibility
for application in adsorption refrigeration. For instance,
with a specially treated ACF, the measured adsorption
capacity on methanol is 2–3 times that of normal activated
carbon, and the estimated adsorption time of ACF on
methanol is about 1/5 to 1/10 of that of normal activated
carbon. Besides, activated carbon can be made with
properties to suit particular applications by varying the
activation time and temperature. Hence, with such technol-
ogies the specific refrigeration capacity as well as the COP
of the adsorption refrigeration cycle can be improved.
Adsorption technology combined with other tech-
nologies for multi-purpose application seems to be a new
trend in the research. This will widen the area of
applications of adsorption technologies and make the
adsorption refrigeration more cost effective. Any method
that improves the efficiency even marginally would improve
the economic viablilty of operating such devices. Thus,
further studies need to be carried out to validate the potential
for possible application in household refrigerators.
References
[1] ASHRAE. Absorption air-conditioning and refrigeration
equipment. ASHRAE guide and data book, equipment.
New York: ASHRAE; 1972. Chapter 14.
[2] Ruthven DM. Principles of adsorption and adsorption
processes. New York: Wiley; 1984.
[3] Suzuki M. Adsorption for energy transport. Adsorption
engineering, Japan; 1980.
[4] Do DD. Adsorption analysis: equilibria and kinetics. London:
Imperial College Press; 1998.
[5] Ponec V, Knor Z, Cerny S. Adsorption on solids. London:
Butterworths Group; 1974.
[6] Oscik J, Cooper IL. Adsorption. Chichester: Ellis Horwood
Limited; 1982.
[7] Dubinin MM, Radushkevich LV. Dokl Akad Nauk SSSR
1947;55:327.
[8] Dubinin MM, Stoeckli HF. Homogeneous and heterogeneous
micropore structures in carbonaceous adsorbents. J Colloid
Interface Sci 1980;75:34.
[9] Yang RT. Gas separation by adsorption processes. London:
Imperial College Press; 1997.
[10] Tchernev. Closed cycle zeolite regenerative heat pump. Heat
transfer enchancement and energy conservation, Hemisphere
Publ. Co; 1988. p. 747–55.
[11] Pons M, Guilleminot JJ. Design of an experimental solar
powered solid-adsorption ice maker. ASME J Solar Energy
Engng 1986;108:332–7.
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327324
[12] Pons M, Grenier PH. Experimental data on a solar-powered
ice maker using activated carbon and methanol adsorption
pair. ASME J Solar Energy Engng 1987;108:303–10.
[13] Pons M, Grenier PH. A phenomenological adsorption
equilibrium law extracted from experimental and theoretical
considerations applied to the activated carbon þ methanol
pair. Carbon 1987;24(5):615–25.
[14] Critoph RE. Limitations of adsorption cycles for solar
cooling. Solar Energy 1988;41(1):23–31.
[15] Critoph RE. Rapid cycling solar/biomass powered adsorption
refrigeration system. Proceedings of the World Renewable
Energy Congress, Brighton, UK, vol. 2.; 2000. p. 959–64.
[16] Headley OS, Kothdiwala AF, Mcdoom IA. Charcoal–
methanol adsorption refrigerator powered by a compound
parabolic concentrating solar collector. Solar Energy 1994;
53:191–7.
[17] Vasiliev LL, Gulko NV, Khaustov VM. Solid adsorption
refrigerators with active carbon–acetone and carbon ethanol
pairs. Proceedings of the International Sorption Heat Pump
Conference, Montreal, Quebec, Canada, vol. 1.; 1996. p. 3–6.
[18] Vasiliev LL, Antux AA, Kulaov AG, Amishkinis D. Heat
pipe cooled and heated solid adsorption refrigerator.
Proceedings of the 19th International Congress on Refriger-
ation, vol. 3.; 1995. p. 200–8.
[19] Wang RZ, Jia JP, Zhu YH, Teng Y, Wu JY, Cheng J. Study
on a new solid adsorption refrigeration pair: active carbon
fiber–methanol. ASME J Solar Energy Engng 1997;119:
214–9.
[20] Meunier F, Grenier PH, Guilleminot JJ, Pons M. Solar
powered refrigeration using intermittent solid adsorption
cycles. Internal paper, Campus Universitaire, Orsay, France;
February 1986
[21] Meunier F. Etude des Coupies Adsorbant-Solide–Refriger-
ant Liquide. EEC report EUR 7708 FR
[22] Critoph RE, Vogel R. Possible adsorption pairs for use in
solar cooling. Int J Ambient Energy 1986;7(4):183–90.
[23] Oertel K, Fischer M. Adsorption cooling system for cold
storage using methanol/silicagel. Appl Thermal Engng 1998;
18:773–86.
[24] Hajji A, Worek WM, Lavan Z. Dynamic analysis of a closed-
cycle solar adsorption refrigerator using two adsorbent–
adsorbate pairs. ASME J Solar Energy Engng 1991;113:73–9.
[25] Tamaniot-Telto Z, Critoph RE. Thermophysical properties of
monolithic carbon. Int J Heat Mass Transfer 2000;43:
2053–8.
[26] Wang RZ. Adsorption refrigeration research in Shanghai Jiao
Tong University. Renewable Sustainable Energy Rev 2001;
5(1):1–37.
[27] Wang RZ. Performance improvement of adsorption cooling
by heat and mass recovery operation. Int J Refrigeration
2001;24:602–11.
[28] Miles DJ, Sanborn DM, Nowakowski GAS, Shelton V. Gas-
fired sorption heat-pump development. Heat Recovery Syst
CHP 1993;13(4):347–51.
[29] Istria S, Castaing J, Neveu P. Energetic analysis, application
field and performance of a new thermochemical sorption
cycle: the multisalt system. Appl Thermal Engng 1996;16(1):
875–89.
[30] Pons M, Feng Y. Characteristic parameters of adsorptive
refrigeration cycles with thermal regeneration. Appl Thermal
Engng 1997;17(3):289–98.
[31] Jones JA. Regenerative adsorbent heat pump. US Patent
5046319; 1991.
[32] Jones JA. Sorption refrigeration research at JPL/NASA. In:
Proceedings of the Solid Sorption Refrigeration, Paris; 1992.
p. 126–35.
[33] Lotz H. Continuously operating and adsorption refrigeration
plant, especially for operation by waste heat of combustion
engines or the like. US Patent 4478057; 1984
[34] Duran M. Chemisorption air conditioner. US Patent
4574874; 1986.
[35] Maier-Laxhuber P. Continuously acting adsorption devices
and process for their operation, US Patent 4660629; 1987
[36] Erickson DC. Rotary trisorption heat pump, US Patent
5279359; 1994
[37] Ebbeson B. Sorption device. US Patent 5431716; 1995
[38] Critoph RE. Simulation of a continuous multiple-bed
regenerative adsorption cycle. Int J Refrigeration 2001;24:
428–37.
[39] Shelton SV, Wepfer WJ. Solid–vapor heat pump technology.
In: Proceedings of the IEA Heat Pump Conference, Japan:
Tokyo; 1990. p. 525–35.
[40] Sun LM, Feng Y, Pons M. Numerical investigation of
adsorptive heat pump systems with thermal wave heat
regeneration under uniform-pressure conditions. Int J Heat
mass Transfer 1997;40(2):281–93.
[41] Ben Amar N, Sun LM, Meunier F. Numerical analysis of
adsorptive temperature wave regenerative heat pump. Appl
Thermal Engng 1996;16(5):405–18.
[42] Spinner B. Ammonia-based thermochemical transformers.
Heat Recovery Syst CHP 1993;13:301–8.
[43] Critoph RE. Forced convection adsorption cycles. Appl
Thermal Engng 1998;18:799–807.
[44] Critoph RE. Forced convection adsorption cycle with packed
bed heat regeneration. Int J Refrigeration 1999;22:38–46.
[45] Scharfe J, Ziegler F, Radermacher R. Analysis of advantages
and limitations of absorber-generator heat exchange. Int J
Refrigeration 1986;9.
[46] Meunier F. Second law analysis of a solid adsorption heat
pump operating on reversible cycles: application to the
zeolite–water pair. J Heat Recovery Syst 1985;4:133–41.
[47] Ziegler F, Brandl F, Volkel J, Alefeld G. A cascading two-
stage sorption chiller system consisting of water–zeolite high
temperature stage and a water–LiBr low temperature stage.
Absorp Heat Pump Cong Paris 1985;231–8.
[48] Best R, Ortega N. Solar refrigeration and cooling. Renewable
Energy 1999;16:685–90.
[49] Natural Resources International Ltd. http://www.
nrinternational.co.uk/fishpi1.htm; 2000
[50] Sakoda A, Suzki M. Simultaneous transport of heat and
adsorbate in closed type adsorption cooling system utilizing
solar heat. ASME J Solar Energy Engng 1986;108:239–45.
[51] Hisaki H, Kobayashi N, Yonezawa Y, Morikawa A.
Development of ice-thermal storage system using an
adsorption chiller. Int Absorption Heat Transfer Conf,
ASME-AES 1993;31:439–44.
[52] Saha BB, Boleman EC, Kashiwagi T. Experiemntal invest-
gatiion of silica gel–water adsorption refrigeration cycle—
the influence of operating conditions on cooling output and
COP. ASHRAE Trans 1995;101:425–31.
[53] Dieng AO, Wang RZ. Literature review on solar adsorption
technologies for ice-making and air-conditioning purposes
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327 325
and recent developments in solar technology. Renewable
Sustainable Energy Rev 2001;5(4):313–42.
[54] Ortel K, Fischer M. Adsorption cooling system for cold
storage using methanol and silica gel. Proceedings of the
International Heat Powered Cycles Conference, University of
Nottingham, UK; 1997. p. 15–17.
[55] Tchernev DI. Solar air conditioning and refrigeration systems
utilizing zeolites. Proc. Meetings of IIR Commissions E1-E2,
Jerusalem 1982;209–215.
[56] Dupont M, Guilleminot JJ, Meunier F, Nguyen P. Etude de
glaciers solaires utilisant le cycle intermittent jour-nuit
zeolithe 13x-eau en climat tempere et en climat tropical.
Proceedings of the meeting of IIR Commissions E1-E2,
Jerusalem 1982;14–15:193–208.
[57] Grenier PH, Guilleminot JJ, Meunier F, Pons M. Solar
powered solid adsorption cold store. ASME J Solar Energy
Engng 1998;110:192–7.
[58] Tather M, Erdem-Senatalar A. The effects of thermal
gradients in a solar adsorption heat pump utilizing the
zeolite – water pair. Appl Thermal Engng 1999;19:
1157–72.
[59] Poyelle F, Guilleminot JJ, Meunier F. Experimental tests and
predictive model of an adsorptive air conditioning unit. Ind
Engng Chem Res 1999;38:298–309.
[60] Meunier F. Solid sorption heat powered cycles for cooling
and heat pumping applications. Appl Thermal Engng 1998;
18:715–29.
[61] Cacciola G, Restuccia G. Reversible adsorption heat pump: a
thermodynamic model. Int J Refrigeration 1995;18(2):
100–6.
[62] Delgado R, Choisier A, Grenier P, Ismail I, Meunier F, Pons
M. Etude de cycle intermittent charbon actifmethanol en vue
de la realisatioon d’une machine a fabriquer de la glace
fonctionnant a l’energie solaire. Proc Meetings IIR Commis-
sions E1–E2, Jerusalem 1982;14–15:185–91.
[63] Li ZF, Sumathy K. A solar powered ice maker with the solid
adsorption pair of activated carbon and methanol. Int J
Energy Res 1999;23:517–27.
[64] Exell RHB, Bhattacharya SC, Upadhyaya YR. Research and
development of solar powered desiccant refrigeration for
cold storage applications. Asian Institute of Technology,
Bangkok, AIT Research Report No. 265; 1993
[65] Jones JA. Carbon/ammonia regenerative adsorption heat
pump. Int Absorption Heat Pump Conf, ASME-AES 1993;
449–55.
[66] Miles DJ, Shelton SV. Design and testing of a solid
sorption heat pump system. Appl Thermal Engng 1996;16:
389–94.
[67] Douss N, Meunier F. Experimental study of
cascading adsorption cycles. Chem Engng Sci 1989;
44(2):225–35.
[68] Shu JZ, Wang RZ, Lu YZ, Shao Y. Research on an
adsorption–absorption cascading multi-effect refrigeration
cycle. J Engng Thermophys 2001;5:546–8. in Chinese.
[69] Saha BB, Akisawa A, Kashiwagi T. Solar/waste heat driven
two-stage adsorption chiller: the prototype. Renewable
Energy 2001;23:93–101.
[70] Wang RZ. Study on a four-beds cascade adsorption
refrigeration cycle capable of COP over 1.1. In: Proceedings
of the 20th International Congress of Refrigeration. Aus-
tralia: Sydney; 1999.
[71] Tan ZC, Wang RZ. Research on double effect adsorption
refrigeration cycle. Acta Energiae Solaris Sinica 1998;19(2):
156–60.
[72] Shelton SV, Wepfer WJ, Miles DJ. Ramp wave analysis of
solid/vapor heat pump. ASME J Energy Resour Technol
1990;112:69–78.
[73] Fuller TA, Wepfer WJ, Shelton SV, Ellis M. A two-
temperature model of the regenerative solid–vapor heat
pump. ASME J Energy Resour Technol 1994;116:297–312.
[74] Zheng W, Worek MW, Nowakowski G. Effect of operating
conditions on the performance of two-bed closed cycle soli-
sorption heat pump systems. ASME J Solar Energy Engng
1995;181–6.
[75] Wang RZ, Wu JY, Xu YX, Wang W. Performance researches
and improvements on heat regenerative adsorption refriger-
ator and heat pump. Energy Conversion Mgmt 2000;42(2):
233–49.
[76] Li CH, Wang RZ, Lu YZ. Investigation of a novel combined
cycle of solar powered adsorption–ejection refrigeration
system. Renewable Energy 2002;26:611–22.
[77] Sun DW. Evaluation of a combined ejector-vapour-
compression refrigeration system. Int J Energy Res 1998;
22(4):333.
[78] Wang RZ, Li M, Xu YX, Wu JY. A new hybrid system of
solar powered water heater and adsorption ice maker. Solar
Energy 2000;68(2):189–95.
[79] Dai YJ, Wang RZ, Xu YX. Study of a solar powered solid
adsorption desiccant cooling system used for grain storage.
Renewable Energy 2002;25:417–30.
[80] Dai YJ, Sumathy K, Wang RZ, Li YG. Enhancement of
natural ventilation in a solar house with a solar chimney and a
solid adsorption cooling cavity. Solar Energy 2003. in press.
[81] Patnaik S, Lenz TG, Lof GOG. Performance studies for
an experimental solar open-cycle liquid desiccant air
dehumidification system. Solar Energy 1990;44(3):
123–35.
[82] Collier Jr.RK. Desiccant cooling systems, inert heat capacity;
regeneration air stream. New York: ASME; 1986.
[83] Belding WA, Worek WM, Novosel D, Holeman WD.
Development for gas-fired desiccant cooling systems.
ASHRAE Trans Tech Papers 1991;587–94.
[84] Worek WM, Zheng W, San JY. Thermodynamic properties
of adsorbed water on silica gel. Energy losses in adiabatic
sorption processes. J Thermophys Heat Transfer 1991;5(3):
435–40.
[85] Majumdar P. Heat and mass transfer in composite desiccant
pore structures for dehumidification. Solar Energy 1998;
62(1):1–10.
[86] Collier RK, Cohen BM, Slosberg RB. Desiccant properties
and their effect on the performance of desiccant cooling
systems. ASHRAE Desiccant Cooling Dehumidification
1992;75–81.
[87] Srivastava NC, Eames IW. A review of developments in
solid–vapour adsorption-refrigeration and heat-pump sys-
tems. J Inst Energy 1997;70:116–27.
[88] Harriman IIILG. Prevention of metallic corrosion in storage
areas. ASHRAE, Desiccant Cooling Dehumidification 1992;
58–68.
[89] Harriman IIILG. Moisture control in photolithography areas.
ASHRAE, Desiccant Cooling Dehumidification 1992;
58–68.
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327326
[90] Banks NJ. Desiccant dehumidifiers in ice arenas. ASHRAE,
Desiccant Cooling Dehumidification 1992;69–71.
[91] Banks NJ. Utilization of condenser heat for desiccant
dehumidifiers in supermarket applications. ASHRAE, Desic-
cant Cooling Dehumidification 1992;21–8.
[92] Davanagere BS. A feasibility study of a solar desiccant air-
conditioning system—part I: psychrometrics and analysis of
the conditioned zone. Int J Energy Res 1999;23:7–21.
[93] Pons M, Meunier F, Cacciola G, Critoph RE, Groll M,
Puigjaner L, Spinner B, Ziegler F. Thermodynamic based
comparison of sorption systems for cooling and heat
pumping. Int J Refrigeration 1999;22:5–17.
[94] Boubakri A. A new conception of an adsorptive solar-
powered ice maker. Renewable Energy 2003;28:831–42.
[95] Buchter F, Dind PH, Pons M. An experimental solar-powered
adsorptive refrigeration tested in Burkina–Faso. Int J
Refrigeration 2003;26:79–86.
[96] Leite APF, Daguenet M. Performance of a new solid
adsorption ice maker with solar energy regeneration. Energy
Conversion Mgmt 2000;41:1625–47.
[97] Li M, Wang RZ, Luo HL, Wnag LL, Huang HB.
Experiments of a solar at plate hybrid system with
heating and cooling. Appl Thermal Engng 2002;22:
1445–54.
[98] Zhang XJ, Wang RZ. Design and performance simulation
of a new solar continuous solid adsorption refrigeration
and heating hybrid system. Renewable Energy 2002;27:
401–15.
[99] Zhang XJ, Wang RZ. A new combined adsorption–ejector
refrigeration and heating hybrid system powered by solar
energy. Appl Thermal Engng 2002;22:1245–58.
[100] Wang S, Zhu D. A novel type pf coupling cycle for
adsorption heat pumps. Appl Thermal Engng 2002;22:
1083–8.
K. Sumathy et al. / Progress in Energy and Combustion Science 29 (2003) 301–327 327