Sie sind auf Seite 1von 16

energies

Article
Closed Adsorption Heat Storage—A Life Cycle
Assessment on Material and Component Levels
Björn Nienborg 1, *, Tobias Helling 1 , Dominik Fröhlich 1 , Rafael Horn 2 , Gunther Munz 1
and Peter Schossig 1
1 Fraunhofer ISE Institute for Solar Energy Systems, Heidenhofstr. 2, 79110 Freiburg, Germany;
tobias.helling@ise.fraunhofer.de (T.H.); dominik.froehlich@ise.fraunhofer.de (D.F.);
gunther.munz@ise.fraunhofer.de (G.M.); peter.schossig@ise.fraunhofer.de (P.S.)
2 Institute for Acoustics and Building Physics, University of Stuttgart, 70569 Stuttgart, Germany;
rafael.horn@iabp.uni-stuttgart.de
* Correspondence: bjoern.nienborg@ise.fraunhofer.de; Tel.: +49-761-4588-5883

Received: 14 November 2018; Accepted: 3 December 2018; Published: 6 December 2018 

Abstract: Closed adsorption storages have been investigated in several projects for heat storage
in building applications with focus on energy density and performance. This study complements
this research with the assessment of the environmental impacts over the life cycle. Global warming
potential (GWP) was chosen as the assessment criterion. Selected sorption materials in combination
with water as the refrigerant were analyzed first by themselves and then embedded in a generic
storage configuration. Sensible storage in water served as the reference benchmark. Results on
material and component level showed that the relative storage capacity compared to water under
realistic operating conditions reached values of below 4 and 2.5, respectively, in the best cases.
Since the effort for producing the sorbents as well as the auxiliary material demand for assembling
storage components was significantly higher than in the reference case, the specific environmental
impact per storage capacity also turned out to be ~2.5 to ~100 times higher. We therefore suggest
focusing sorption storage research on applications that (a) maximize the utilization of the uptake of
sorbents, (b) do not compete with water storages, and (c) require minimal auxiliary parts.

Keywords: thermochemical storage; sorption storage; adsorption; storage capacity; life cycle
assessment; embedded energy; global warming potential

1. Introduction
The promotion and use of renewable energy sources is a key strategy for combating climate
change.
In order to match their fluctuating supply to the demand, energy storages play an essential
role [1]. In the field of heat storage, currently sensible hot water storages are widely used, especially
for short-term energy storage. In regard to their extended use in combination with renewables, storage
losses and space consumption are disadvantageous. Therefore, thermo-chemical materials (TCM) have
been widely studied in the past couple of years, as they potentially allow higher storage densities and
almost loss-free storage of thermal energy. The latter is the reason why they have been considered
very promising for seasonal storage in combination with solar thermal energy [2].
A common type of TCM studied is based on adsorption where a gas or vapor (sorbate) adheres to
a microporous sorbent material. During charging, which is an endothermic reaction, the sorbent is
heated and releases the adsorbed vapor. During discharging, the working pair is brought back together,
which releases heat in the exothermic adsorption process [3]. So-called open systems are connected
to the atmosphere, and here, the sorbate is released and extracted from air which also serves as heat

Energies 2018, 11, 3421; doi:10.3390/en11123421 www.mdpi.com/journal/energies


Energies 2018, 11, 3421 2 of 16

transfer medium. In contrast, closed systems consist of a hermetic sorption reactor with connected
evaporator or condenser for the sorbate and operate under a sorbate-specific pressure (e.g., a vacuum,
in the case of water or methanol, or overpressure for ammonia). Heat is supplied and extracted via
heat exchangers. Due to its favorable properties, many potential working pairs using water as the
sorbate have been investigated, e.g., silica gel [4], zeolites [5,6], aluminum phosphates (AlPOs) [7],
metal organic frameworks (MOFs) [8], activated carbon, and a variety of composites [4,7].
Jähnchen et al. reported theoretical energy storage densities for TCMs in closed storages of
93 kWh/m3 (silica gel/water) and 160 kWh/m3 for zeolite NaX/water systems [7]. Aydin et al.
quantified the storage densities of silica gel/water to be 182 kWh/m3 and 162 kWh/m3 for the
zeolite/water working pair [2]. By using salt-impregnated matrixes, high values also seem achievable:
For materials with vermiculite as the substrate, values of up to 253 kWh/m3 and even 330 kWh/m3
have been reported, although the latter analysis showed stability problems [9,10].
Even though these results indicate a great potential of solid sorption for thermal energy storage,
their performance is significantly lower considering actual operating conditions. Frazzica et al.
performed calculations assuming a solar thermal heat supply and low temperature radiators for
heat distribution (Tdesorpion = 120 ◦ C, Tadsorpion = 50 ◦ C, Tevaporator = 10 ◦ C; Tcondenser = 30 ◦ C, closed
system) and reported storage densities of 53 kWh/m3 with a water/zeolite 13X working pair and
38 kWh/m3 for water/silica gel [6]. Permyakova et al. determined a storage density of 51 kWh/m3 for
zeolite 13X based on thermogravimetric measurements of the material in combination with water [11].
Several pilot projects have studied the performance of solid-sorption-based storage components
in heating systems. In frame of the EU Project “Modestore”, the technical feasibility of a closed
thermochemical storage with water/silica gel storage was demonstrated. The realized unit reached a
storage density of 33 kWh/m3 at 85 ◦ C charging and 25 ◦ C discharging temperature. A laboratory
scale storage with Z13X/water yielded a specific capacity of 58 Wh/m3 with 180 ◦ C and 20 ◦ C
charging and discharging temperatures, respectively [12]. The University of Stuttgart demonstrated
in their SolSpaces project that their open TCS prototype (with zeolite) was able to meet the heating
requirements of a small house with 43 m2 [13].
As can be seen, a lot of research has been done on the performance of TCMs on different scales,
partly complemented with (usually rather disillusioning) cost estimations (e.g., References [14,15]).
In order to increase operating time and therefore economic viability, Vasta et al. therefore recommend
looking at multi-functional applications such as cooling and heat pumping [16].
Few publications include life cycle assessment (LCA) results on TCMs. Munz et al. performed
a generic potential analysis for open sorption using the cumulated energy demand for production
as the evaluation criterion. They found that the material impact must be below 4.5 kWh/kg but
ranges from 5 to 7 kWh/kg [15]. Horn et al. evaluated the energetic amortization of pure sorption
materials, ignoring any required auxiliary parts or energy, for an assumed solar thermal application
(100 ◦ C desorption, 35 ◦ C adsorption, and 10 ◦ C condensation/evaporation temperature, respectively).
Under these boundary conditions, silica gel and Z13X requires ~60 and ~260 cycles, respectively,
to supply the energy consumed during their production [17]. Bertsch compared the LCA of solar
thermal systems with water and open sorption storages. He concluded that the latter has a higher
thermal performance which is offset by a higher auxiliary energy demand [18].
Further studies on the LCA of (closed) sorption based thermal storages are not known. Evidently,
a holistic and comparative ecological evaluation including not only material but also component level is
still pending and therefore addressed in this work to address this issue for closed solid-sorption-based
thermochemical storage. This paper focuses on the storage and provision of heat in temperature ranges
suitable for building applications.

2. Materials and Methods


As sorption materials cannot be used in storage applications by themselves, this paper
contemplates the environmental impact not only at the material level but also in a generic sorption
Energies 2018,
Energies 2018, 11,
11, 3421
x FOR PEER REVIEW 33 of
of 16

comparative evaluation of the performance of the respective storage components in a heating system, lies
storage component,
outside the scope of thiswhich comprises
journal and willall
bethe required
provided in aparts
future(such as heat exchangers, containment,
publication.
and insulation). At this scale, an initial quantitative comparison with water storages as the reference
is possible.
2.1. Life CycleThe final step,
Assessment andthe comparative
Applied evaluation of the performance of the respective storage
Impact Categories
components in a heating system, lies outside the scope of this journal and will be provided in a
The method of life cycle assessment (LCA) refers to a systematic analysis of the environmental
future publication.
impacts of a product from cradle to grave. As shown in Figure 1, it includes all stages from resource
extraction
2.1. through
Life Cycle processing,
Assessment manufacturing,
and Applied transport, use, and including maintenance and end-
Impact Categories
of-life (reuse, recycling, or disposal). It is regulated by ISO standards [19,20] and further substantiated
Theconstruction
for the method of life cyclebyassessment
sector European (LCA)
Normsrefers to aAsystematic
[21,22]. analysis
large number of the environmental
of impact categories are
impacts
available for the assessment, which ranges from land use, water stress, and ecological stages
of a product from cradle to grave. As shown in Figure 1, it includes all from
and human
resource extraction through processing, manufacturing, transport, use, and
toxicity, to climate and energy related values such as global warming potential (GWP) or primaryincluding maintenance
and
energyend-of-life (reuse, recycling,
(PE) consumptions or disposal).
(comprising It is regulated
both energetic by ISO
and material standards [19,20] and further
use).
substantiated for the
In the present construction
study, sector by
we concentrated on European Norms
the latter two [21,22].
as they A large number
are recommended ofrelevant
in the impact
categories are available for the assessment, which ranges from land use, water
standards, are well known, and rest upon well-validated models [23]. The assessment was carried stress, and ecological
and human
out with thetoxicity,
software to climate
suite GaBi andand
energy related
relied values
on the such as global
corresponding warming
material andpotential (GWP) or
process databases
primary energy (PE) consumptions (comprising both energetic and material use).
(http://www.gabi-software.com).

Figure 1. Schematic visualization of the life cycle of a product.


product. Source: University of Stuttgart.

In the present
2.2. Description study, weMaterials
of Considered concentrated on the latter two as they are recommended in the relevant
standards, are well known, and rest upon well-validated models [23]. The assessment was carried
This study concentrated on five sorption materials, which are considered promising for building
out with the software suite GaBi and relied on the corresponding material and process databases
heating applications due to their thermodynamic properties. One representative of the well-known
(http://www.gabi-software.com).
material types, silica gel, silicoaluminumphosphates (SAPO), and zeolites, was each included. Two
newer
2.2. materialsofout
Description of the group
Considered of metal organic frameworks (MOFs) complemented the selection.
Materials
Finally one promising (because it is apparently stable) salt-composite material with “sorption-like”
This study concentrated on five sorption materials, which are considered promising for building
behavior was considered.
heating applications due to their thermodynamic properties. One representative of the well-known
material types, silica
2.2.1. Considered gel, silicoaluminumphosphates
Production Processes (SAPO), and zeolites, was each included.
Two newer materials out of the group of metal organic frameworks (MOFs) complemented the selection.
Essential
Finally for a substantiated
one promising evaluation
(because it is apparentlyof stable)
the materials is the realistic
salt-composite material implementation of the
with “sorption-like”
corresponding production
behavior was considered. processes. For the established materials, abundant information is available
on the synthesis routes (examples in Table 1) which have been scaled to industrial batch production
2.2.1.
levels.Considered
For the MOFsProduction
and the Processes
composite, information is scarcer and only refers to laboratory scales,
which, of course, is not comparable to industry scales in terms of the specific environmental impact
Essential for a substantiated evaluation of the materials is the realistic implementation of the
of the product. For this reason, their production has also been upscaled to industry scale batch
corresponding production processes. For the established materials, abundant information is available
processes using the same process models (with boundary conditions adapted to their synthesis
routes) as those for the remaining materials. All materials are assumed to be provided as pellets;
Energies 2018, 11, 3421 4 of 16

on the synthesis routes (examples in Table 1) which have been scaled to industrial batch production
levels. For the MOFs and the composite, information is scarcer and only refers to laboratory scales,
which, of course, is not comparable to industry scales in terms of the specific environmental impact of
the product. For this reason, their production has also been upscaled to industry scale batch processes
using the same process models (with boundary conditions adapted to their synthesis routes) as those
for the remaining materials. All materials are assumed to be provided as pellets; therefore, all but silica
gel and the composite contain ten mass-percent of a silicon-resin-based binder, which was chosen for
Energies 2018, 11, x FOR PEER REVIEW 4 of 16
its longevity and resistance.
therefore, all but silica gel and the composite contain ten mass-percent of a silicon-resin-based binder,
whichTable 1. Overview
was chosen for itson investigated
longevity sorption materials; MOF: Metal Organic Framework.
and resistance.

Table 1.Name
Overview on investigated sorption materials; Type
Reference for
MOF: Metal Organic Framework
Material Short Name
Production Process
Material Name Short Name Type Reference for Production Process
Silica gel (microporous) Silica gel Xerogel [24]
Silica gel (microporous) Silica gel Xerogel [24]
Silica-aluminophosphate 34 SAPO-34 Zeolite-like [25]
Silica-aluminophosphate 34 SAPO-34 Zeolite-like [25]
Zeolite type 13X Z13X Zeolite [26]
Zeolite type 13X Z13X Zeolite [26]
CAU-10-H CAU-10-H MOF [27]
CAU-10-H CAU-10-H MOF [27]
Aluminum-Fumarate Al-Fum. MOF [28]
Aluminum-Fumarate Al-Fum. MOF [28]
Vermiculite impregnated
Vermiculite impregnated LiCl-Ver Composite/sorption-like
Composite/ [9]
with Lithiumchloride LiCl-Ver [9]
with Lithiumchloride sorption-like

2.2.2. Determination of the


2.2.2. Determination Thermal
of the Storage
Thermal StorageCapacity
Capacity
In order
In order to determine
to determine the sorption-equilibria
the sorption-equilibria of theoffirst
thefive
firstmaterials
five materials
listed inlisted
Tablein1,Table 1,
volumetric
volumetric water sorption isotherms were measured at 25,◦ 40, and 60 °C on
water sorption isotherms were measured at 25, 40, and 60 C on a Quantachrome Vstar. Prior to the a Quantachrome Vstar.
Prior to the measurement, all samples were degassed under reduced pressure at 250 °C for 24 h.
measurement, all samples were degassed under reduced pressure at 250 ◦ C for 24 h.
For performance calculations, a number of functions have been published, such as multi-temperature
For performance calculations, a number of functions have been published, such as multi-
Toth [29,30], Sips [31], Dubinin [4], and others [32], which differ in complexity and suitability for different
temperature Toth [29,30], Sips [31], Dubinin [4], and others [32], which differ in complexity
materials. For the present study, an adapted Dubinin-Asthakov-Fit (see Equation (1)) was used as it
and suitability
showed decentfor results
different
acrossmaterials. For
all materials withthe present
limited study, The
complexity. an adopted
adaptedmodification
Dubinin-Asthakov-Fit
refers to
(see Equation (1))the
the fact that wasadsorbed
used as vapor
it showed
mass decent
instead results across allvolume
of the adsorbed materials withtolimited
is fitted complexity.
the adsorption
The adopted
potentialmodification refers
(see Equation (2)). As to
an the fact that
example, the adsorbed
the adapted vapor mass instead
Dubinin-Asthakov-Fit of the
for silica-gel adsorbed
is shown
volumein Figure 2 (for
is fitted to other fits, see Supplementary
the adsorption potential Materials).
(see EquationThe resulting
(2)). As coefficients
an example, can be the
found in
adapted
Table 2 for all five materials. The specific heat capacity specified in the table corresponds
Dubinin-Asthakov-Fit for silica-gel is shown in Figure 2 (for other fits, see Supplementary Materials). to the value
measuredcoefficients
The resulting at 60 °C with acan
MICROSC-4C
be found from Setaram.
in Table The
2 for allbulk
fivedensity was measured
materials. either directly
The specific (m)
heat capacity
or calculated from the measured density of the powder, assuming close-packing of equal spheres
specified in the table corresponds to the value measured at 60 ◦ C with a MICROSC-4C from Setaram.
with 74% of solid volume.
The bulk density was measured either directly (m) or calculated from the measured density of the
For determining the storage capacity of LiCl-Ver, no measured data was available at Fraunhofer
powder, assuming close-packing of equal spheres with 74% of solid volume.
ISE. Therefore, the equilibrium line and properties published by Grekova et al. [9] were adopted.

FigureFigure 2. of
2. Plot Plot of water
the the water sorption
sorption isothermsofofsilica
isotherms silica gel
gel at
at different
differenttemperatures
temperatures during ad- ad-
during and and
desorption. The red line depicts the adapted Dubinin-Asthakov-Fit; data points and
desorption. The red line depicts the adapted Dubinin-Asthakov-Fit; data points and fits for other fits for other
materials are provided in the Supplementary Materials.
materials are provided in the Supplementary Materials.

𝑋 = 𝑋 ∙ 𝑒𝑥𝑝⌊− 𝑏/𝐴 ⌋ (1)


𝑝
𝐴 =𝑅∙𝑇 ∙ ln (2)
𝑝
Energies 2018, 11, 3421 5 of 16

Table 2. Key information on the investigated materials; measured (m) and calculated (c) from the
measured density of the powder, assuming close-packing of equal spheres with 74% of solid volume.

Material X 0 (g/g) b (g/J) N cp,sorb,60 ◦ C (J/kg/K) Bulk Density 1 (kg/m3 )


Silica gel 0.31 232.8
1.673 1.04 545.6 (m)
Energies 2018, 11, x FOR PEER
SAPO-34 REVIEW415.9
0.315 2.9220 1.42 650.0 (c) 5 of 16
Z13X 0.321 1404.0 1.2350 1.68 703.5 (m)
Figure 3 illustrates
CAU-10-H 0.343the resulting
244.5 equilibrium
4.1510 curves
1.59and includes the466.0
limits
(c) of adsorption
Al-Fum. 0.496 195.6 2.3860 1.67 316.3 (m)
potential determined by the studied application scenarios (see Section 3). It shows that neither silica
LiCl-Ver
gel nor the -
MOFs are suitable for- adsorption
- (discharge)1.01 350.0
at 55 °C. Zeolite 13X, on the other hand,
1
Bulk density
cannot exploit much of the maximum uptakerefers
even to with
pelletsthe
(notmaximum
powder). desorption temperature of 120

°C. LiCl-Vermiculite is entirely regenerated at 90 °C and therefore the 120 °C results are not presented
For determining the storage capacity of LiCl-Ver, no measured data was available at Fraunhofer
in the subsequent sections.
ISE. Therefore, the equilibrium line and properties published by Grekova et al. [9] were adopted.
Table 2 Key information on the investigated materials; measured (m) and calculated (c) from the
X = X0 ·exp − (b/A)n (1)
measured density of the powder, assuming close-packing of equal spheres with 74% of solid volume.
 
Material X0 (g/g) b (g/J) A = R N· T · ln psat°C (J/kg/K) Bulk Density 1 (kg/m3)
cp,sorb,60
sorb (2)
Silica gel 0.31 232.8 1.673 p Tsorb
1.04 545.6 (m)
SAPO-34
Figure 3 illustrates0.315 415.9
the resulting 2.9220 curves and
equilibrium 1.42
includes the limits of650.0 (c)
adsorption potential
determinedZ13X 0.321 application
by the studied 1404.0 scenarios
1.2350 (see Section 1.68 3). It shows that703.5 (m)silica gel nor
neither
CAU-10-H
the MOFs are suitable0.343
for adsorption 4.1510 at 55 ◦ C.
244.5 (discharge) 1.59Zeolite 13X, on the466.0
other(c)hand, cannot
Al-Fum.
exploit much 0.496
of the maximum 195.6 even
uptake 2.3860
with the maximum 1.67 316.3 (m) of 120 ◦ C.
desorption temperature
LiCl-Ver is entirely
LiCl-Vermiculite - regenerated
- ◦
at 90- C and therefore ◦
1.01 the 120 C results are 350.0
not presented in
the subsequent sections. 1 Bulk density refers to pellets (not powder).

Figure
Figure 3.3. Plot of the
Plot of the adapted
adapted Dubinin-Astakhov-Regressions
Dubinin-Astakhov-Regressionsand andthe
theLiCl-Ver
LiCl-Ver equilibrium
equilibrium line
line used
used to
to determine the water uptake (and finally storage capacity) of the considered thermo-chemical
determine the water uptake (and finally storage capacity) of the considered thermo-chemical materials
materialsThe
(TCMs). (TCMs).
verticalThe
linesvertical
indicatelines indicate the
the applicable applicable
limits limitspotential
of adsorption of adsorption potentialscenarios
for the studied for the
(35 ◦
studied
C and ◦
scenarios (35 °C and
55 C adsorption and55 ◦
90 °CC and ◦
adsorption and 90 °Ctemperature,
120 C desorption and 120 °Crespectively;
desorption condensation
temperature,
respectively;
and evaporationcondensation
temperature andareevaporation ◦ C and 10 ◦ C,are
fixed at 20 temperature fixed at 20 °C and 10 °C, respectively).
respectively).

On a material level, the storage capacity is composed of evaporation enthalpy, the adsorption
energy, and
and the
the sensible
sensible heat
heat capacity.
capacity. Equation
Equation (3) (3) (c(cm,max
m,max)) refers
refers to the maximum mass-specific
capacity including the entire
entire sensible
sensible heat.
heat. Equation
Equation (4)(4) (c(cm,min
m,min)) describes
describes the
the capacity
capacity which remains
after the
after thematerial
materialhashas entirely
entirely cooled
cooled down down
to thetotemperature
the temperature of thearound
of the space space the
around the(assumed
storage storage
(assumed ◦
to be 20 °C)to bewhich
and 20 C) is and
then which is then
heated up to theheated up to(=the
adsorption adsorption
discharging) (=discharging)
temperature. temperature.
As a simplification,
a constant sensible heat capacity corresponding to the measured value at 60 °C was adopted.

𝑐 , = ∆𝑋 ∙ ℎ 𝑇 𝐴 𝑋 𝑑𝑋 𝑇 −𝑇 ∙𝑐 , , ° (3)

𝑐 , = ∆𝑋 ∙ ℎ 𝑇 𝐴 𝑋 𝑑𝑋 − 𝑇 −𝑇 ∙𝑐 , , ° (4)
Energies 2018, 11, 3421 6 of 16

As a simplification, a constant sensible heat capacity corresponding to the measured value at 60 ◦ C


was adopted.
Z X
Ads
cm,max = ∆X ·hevap ( Tads ) + A( X )dX + ( TDes − TAds )·c p,sorb,60 ◦ C (3)
XDes

Z X
Ads
cm,min = ∆X ·hevap ( Tads ) + A( X )dX − ( TAds − Tamb )·c p,sorb,60 ◦ C (4)
XDes
Energies 2018, 11, x FOR PEER REVIEW 6 of 16
For all materials but silica gel and LiCl-Ver, the ten mass-percent of the binder (cp = 1.45 kJ/kg/K,
rho =For1120 kg/m3 ) needed
all materials but silica
togel
be and LiCl-Ver,for
considered thethe
tendetermination
mass-percent ofofthe
thebinder (cp =capacity
storage 1.45 kJ/kg/K,
and
rho = 1120 kg/m
environmental 3 ) needed to be considered for the determination of the storage capacity and
impact.
environmental
Temperature impact.
gradients between the heat transfer fluid and the sorbent, as they occur in real
Temperature
components, gradients between the heat transfer fluid and the sorbent, as they occur in real
were ignored.
components, were ignored.
2.3. Definition of Generic Sorption Storage Configuration
2.3. Definition
In order toofevaluate
Generic Sorption Storage Configuration
the performance and life cycle impact of the sorbents related to an application,
a generic storage
In order component
to evaluate was defined.
the performance andInlife
this task,
cycle the experiences
impact fromrelated
of the sorbents past research projects
to an application,
and from storage
a generic the design of sorption
component waschillers and
defined. heattask,
In this pumps
the served as a good
experiences starting
from past point.projects
research The mainand
findings adopted, which led to a configuration as shown in Figure 4, are:
from the design of sorption chillers and heat pumps served as a good starting point. The main findings
adopted, which led to a configuration as shown in Figure 4, are:
• The storage is closed, operates under near vacuum, and water is the adsorbent;
•• AThe storage isevaporator/condenser
combined closed, operates underheat nearexchanger
vacuum, and watercan
(ECHX) is the
beadsorbent;
used since charging and
• discharging
A combinedoccur evaporator/condenser
at different times; heat exchanger (ECHX) can be used since charging and
• Itdischarging
is preferable occur at different
to combine thetimes;
sorption heat exchanger (SHX) and the evaporator/condenser
• heat
It is preferable
exchangerto combine
into the sorptiondue
one containment heattoexchanger (SHX) and the
easier manufacturing evaporator/condenser
and better vacuum stability heat
exchanger into one containment due to easier manufacturing
(compared to a solution with a separate component connected via vacuum pipework); and better vacuum stability
• A(compared to a solution
welded stainless with a separate
steel containment, component
supported connected
by the via vacuum
heat exchangers, pipework);
is favorable compared
• to A awelded
self-baring cylindrical vacuum vessels in terms of compactness and material mass;compared
stainless steel containment, supported by the heat exchangers, is favorable and
• Due self-baring
to a cylindrical
to the low thermal vacuum vessels
conductivity of thein terms ofacompactness
sorbents, good thermal and material mass;
connection to the and
fluid is
• obtained
Due to the low thermal conductivity of the sorbents, a good thermal
by a pellet filling in a finned tube heat exchanger (staggered tubes). connection to the fluid is
obtained by a pellet filling in a finned tube heat exchanger (staggered tubes).

Figure4.4. Schematic
Figure Schematic of
ofthe
theconsidered
consideredstorage
storageconfiguration.
configuration.

2.3.1. Sorption Heat Exchanger


2.3.1. Sorption Heat Exchanger
A parameter variation for the SHX was carried out in order to determine a configuration,
A parameter variation for the SHX was carried out in order to determine a configuration, which
which assures a rather uniform thermal conduction (meaning that the conductivity is not limited by a
assures a rather uniform thermal conduction (meaning that the conductivity is not limited by a single
single thermal resistance) between the sorbent pellets and the heat transfer fluid with low material
thermal resistance) between the sorbent pellets and the heat transfer fluid with low material usage.
usage. Thus, the available volume for sorbent shall be maximized and the additional LCI minimized.
Thus, the available volume for sorbent shall be maximized and the additional LCI minimized.
For the parameter variation, only a small but scalable representative section of the SHX—namely,
For the parameter variation, only a small but scalable representative section of the SHX—
a sixth of the part of a fin, which could be allocated to a single tube, including the respective sorbent
namely, a sixth of the part of a fin, which could be allocated to a single tube, including the respective
volume (see Figure 5)—was calculated [33]. The varied parameters and the respective value ranges are
sorbent volume (see Figure 5)—was calculated [33]. The varied parameters and the respective value
listed in Table 3, the underlying material conductivities and contact resistances are specified in Table 4.
ranges are listed in Table 2, the underlying material conductivities and contact resistances are
specified in Table 3.

Table 2. Value range for tube diameter and spacing as well as fin thickness and spacing covered
during parameter variation.
Energies 2018, 11, 3421 7 of 16
Energies 2018, 11, x FOR PEER REVIEW 7 of 16

Energies 2018, 11, x FOR PEER REVIEW 7 of 16

(a) (b)

Figure 5. (a)
Figure AA
5. (a) cross-cut
cross-cutalong
alongthe
theheat
heatexchanger
exchanger tubes; the dotted
tubes; the dottedline
linedelimits
delimitsthe
thesection
section calculated
calculated
as as
sorber. (b)(b)
sorber. AA plan
planview
viewofofone
onefin;
fin;the
theshaded
shaded area corresponds to the resulting hexagon when
(a) (b)area corresponds to the resulting hexagon when
dividing fins
dividing with
fins the
with sector
the sectormethod
method(Source:
(Source:Shat
Shat et
et al.).
al.).

Figure3.5.Value
Table (a) Arange for tube
cross-cut alongdiameter
the heatand spacingtubes;
exchanger as well asdotted
the fin thickness and spacing
line delimits covered
the section during
calculated
Table 3. Heat conductivity parameters of the heat exchanger.
parameter
as sorber. variation.
(b) A plan view of one fin; the shaded area corresponds to the resulting hexagon when
λ fin (W/m/K)
dividing fins with theλsector
tb (W/m/K) λ sorb (W/m/K)
method (Source: Shat et al.). r sorb,fin (W/m2/K) r fin,tb (W/m2/K)
Description d tb (m) d sp,tb (m) d fin (m) d sp,fin (m)
380 380 0.2 350 7000
Minimum 0.006
Table 3. Heat conductivity 0.001
parameters 10-4heat exchanger.
of the 0.0015
Maximum 0.016 0.005 2 × 10 -4 0.0215
Out of the results of the parameter variation, a pareto front could be derived, which gives
λ fin (W/m/K) λ tb (W/m/K) 0.002λ sorb (W/m/K)
Increment 0.0005 2.5 × 10-5
r sorb,fin (W/m2/K)
0.002 r fin,tb (W/m2/K)
achievable volumetric sorbent to metal ratio for any desired heat transfer coefficient (Figure 6).
Realized380
pilot storages have 380 shown that a heat 0.2 transfer coefficient 350above 1 kW/K/m37000 is essential,
Table 4. Heat conductivity parameters of the heat exchanger.
especially for discharging with a reasonably low driving temperature difference. In order to meet this
Outand
value of to
the resultsfor of the parameter variation,
in realasystems
pareto compared
front could be calculations,
derived, which gives
λ fin account
(W/m/K) possible imperfections
λ tb (W/m/K) λ sorb (W/m/K) r sorb,fin (W/m2 /K) to rthe 2
fin,tb (W/m /K)
a value
achievable volumetric sorbent to metal ratio for any desired
of 3 kW/K/m was adopted. Thus, the volumetric sorbent to metal ratio was 65.
3 heat transfer coefficient (Figure 6).
Realized pilot 380
storages 380
have shown that a heat0.2 transfer 350
coefficient above 7000
1 kW/K/m 3 is essential,
The analogous approach was followed for the determination of the sorbent share in the overall
especially
volume for discharging
of the SHX sectionwithwith
a reasonably
a result oflow driving
94%. temperature
The remaining difference.
space In order
corresponded totothe
meet
heatthis
value Out
andofto
transfer the
medium results
account forof
inside theSHX.
possible
the parameter variation,
imperfections
To account inthe
for anecessary
real pareto front
systems vaporcould
compared tobethe
volume, derived,
the whichawas
calculations,
SHX section gives
value
achievable
of enlarged
3 kW/K/m volumetric
by3 was sorbent
10% adopted.
of the to metal
Thus,
sorbent the ratio
volume. for anysorbent
volumetric desired toheat transfer
metal ratiocoefficient
was 65. (Figure 6). Realized
pilot storages have shown that was
a heat transferforcoefficient above 1 kW/K/m 3 is essential, especially for
The analogous approach followed the determination of the sorbent share in the overall
discharging
volume of the withSHX a reasonably
section withlow adriving
result temperature
of 94%. Thedifference.
remainingInspace ordercorresponded
to meet this value and
to the to
heat
account for possible imperfections in real systems compared to the calculations, a value of 3 kW/K/m 3
transfer medium inside the SHX. To account for the necessary vapor volume, the SHX section was
was adopted.
enlarged by 10%Thus, the sorbent
of the volumetric sorbent to metal ratio was 65.
volume.

Figure 6. Results of the parameter variation for the SHX; blue dots correspond to the identified Pareto
front.

2.3.2. Evaporator/Condenser Heat Exchanger


In adsorption chillers, cycles times are typically short (<30 min), and the required refrigerant
Figure
Figure
mass 6. Results
is limited, so of
Results ofthe
their parameter
the parametervariation forheat
variation
evaporator/condenser the the
for SHX; blue blue
SHX; dots
exchangers correspond to the to
dots correspond
(ECHX) are usually identified Pareto
the identified
designed for thin
front.
Pareto front.
film operation. In contrast, adsorption storages require high amounts of refrigerant and have longer
cycle durations. Therefore, a flooded ECHX is favorable in this application [34]. Since the
2.3.2.The analogous approachHeat
Evaporator/Condenser was followed
Exchanger for the determination of the sorbent share in the overall
condensation of refrigerant in general is a faster process than evaporation (because the entire HX
volume of the SHX section with a result of 94%. The remaining space corresponded to the heat transfer
surface is active), the
In adsorption design
chillers, is oriented
cycles times to thetypically
latter. short (<30 min), and the required refrigerant
medium inside the SHX. To account for theare
necessary vapor volume, the SHX section was enlarged by
mass is limited, so their
10% of the sorbent volume. evaporator/condenser heat exchangers (ECHX) are usually designed for thin
film operation. In contrast, adsorption storages require high amounts of refrigerant and have longer
cycle durations. Therefore, a flooded ECHX is favorable in this application [34]. Since the
condensation of refrigerant in general is a faster process than evaporation (because the entire HX
surface is active), the design is oriented to the latter.
Energies 2018, 11, 3421 8 of 16

2.3.2. Evaporator/Condenser Heat Exchanger


In adsorption chillers, cycles times are typically short (<30 min), and the required refrigerant mass
is limited, so their evaporator/condenser heat exchangers (ECHX) are usually designed for thin film
operation. In contrast, adsorption storages require high amounts of refrigerant and have longer cycle
durations.
Energies 2018, Therefore,
11, x FOR PEER a flooded
REVIEW ECHX is favorable in this application [34]. Since the condensation of
8 of 16
refrigerant in general is a faster process than evaporation (because the entire HX surface is active),
In a flooded
the design ECHX,
is oriented to the
the evaporation
latter. mainly takes place at the contact line formed at the interface
betweenIn athe heat exchanger
flooded ECHX, thesurface and the
evaporation refrigerant.
mainly Therefore,
takes place at thethecontact
lengthlineof this line at
formed is the
the
decisive
interfacefactor
between for the
its heat
performance.
exchangerAn approach
surface and the for refrigerant.
an ECHX which is scalable
Therefore, withofthe
the length thisSHX
line
dimension
is the decisive was factor
derived for from the configuration
its performance. An approachdescribed for in
an[34]
ECHX(coated
whichcopper tube-fin
is scalable withheat
the
exchanger).
SHX dimension This component
was derived hasfrom
threethe
vertical tube rows,described
configuration and measurements in flooded
in [34] (coated operation
copper yielded
tube-fin heat
heat transfer coefficients
exchanger). This component in the has
range of 50
three to 20 W/K
vertical tube(decreasing with the filling level).
rows, and measurements The average
in flooded operation of
35 W/K was adopted. The construction volume of the ECHX section was derived
yielded heat transfer coefficients in the range of 50 to 20 W/K (decreasing with the filling level). from the finned tube
volume
The average(0.083ofm35 × 0.16
W/Kmwas × 0.12 m = 1.59The
adopted. × 10construction
−3 m3), and an additional 20% was assumed for headers
volume of the ECHX section was derived
and bends. Thus, the volume-specific UA-value
from the finned tube volume (0.083 m × 0.16 m × 0.12 m = 1.59 of the ECHX section
× 10−was
3 m318.2 kW/K/m
), and 3, excluding
an additional 20%
any
wasadditional
assumed for vapor volume.
headers and bends. Thus, the volume-specific UA-value of the ECHX section was
18.2 kW/K/m3 , excluding any additional vapor volume.
3. Results and Discussion
3. Results and Discussion
3.1. Results on the Material Level
3.1. Results on the Material Level
3.1.1. Environmental Impact of the Materials
3.1.1. Environmental Impact of the Materials
Figure 7 shows the environmental impact of the materials according to the two selected
FigureGWP
categories, 7 showsandthePE.environmental
Qualitativelyimpact of the
they show a materials according
similar trend. to the two
The absolute selected
impact categories,
of the binder
GWP and PE. Qualitatively they show a similar trend. The absolute impact
was constant among the materials due to the fixed mass-specific share. The relative share was of the binder was constant
below
among the materials due to the fixed mass-specific share. The relative share
3% for SAPO-34 and approximately one-third for the other binder-containing materials, independent was below 3% for SAPO-34
and
of theapproximately
impact category. one-third
Due to forthisthe other binder-containing
analogy, in the following sections, materials,
onlyindependent
the GWP categoryof the will
impact be
category. Due to this analogy, in the following sections, only the GWP category
evaluated. A tabular overview on all the values the result chapter relies on is available in the Appendix will be evaluated.
A tabular overview on all the values the result chapter relies on is available in the Appendix A.
A.

Figure
Figure 7.
7. Environmental
Environmentalimpact
impact of
of the
the storage
storage materials
materials per
per kg
kg of
of material
material (logarithmic
(logarithmic scales).
scales).

3.1.2. Material
3.1.2. Material Storage
Storage Capacity
Capacity
The volumetric
The volumetric storage
storage capacity
capacity was
was calculated
calculatedbased
basedononthe
thematerial
materialdata
dataprovided
providedin inTable
Table 12
and Equations
Equations (3)
(3) and
and (4).
(4). The
The results
results for
for the
the four
four considered
considered scenarios
scenarios (35 ◦
(35 °CC and
and 55 ◦
55 °CCadsorption,
adsorption,
and
corresponding to underfloor and radiator heating, and 90 ◦ C and 120 ◦ C desorption temperature,
corresponding to underfloor and radiator heating, and 90 °C and 120 °C desorption temperature,
respectively; condensation
condensation andand evaporation
evaporation temperature
temperature were
were fixed
fixed at
at 20 ◦ C and 10 ◦ C, respectively;
20 °C
respectively; and 10 °C, respectively;
and temperature
and temperature gradients
gradients in
in heat
heat exchangers
exchangers werewere neglected)
neglected) are
are depicted
depicted inin Figure
Figure 8.
8.
For ideal short-term storage (discharge begins immediately after charging, so no sensible heat is
lost), the sum of sorptive and maximum sensible capacity must be considered. This value can be
compared directly to the reference capacities of water (also loss-free, with a useful capacity between
ad- and de-sorption temperatures). Interestingly, the regular sorption materials rarely outperformed
water in terms of volumetric density. Especially for the 35 °C cases, SAPO-34 reached the storage
capacities noticeably above that of the reference. LiCl-Vermiculite significantly outperformed the
temperature, to the sorptive capacity. In this case, the effective capacity of the water storage will also
be lower and needs to be corrected by an appropriate efficiency [35]. Assuming, for example, 50%
sensible losses for the water storage and 100% sensible losses for the sorption alternatives due to less
insulation, some more sorption storages show benefits compared to water.
EnergiesThe fact
2018, that the MOFs did not show any sorptive capacity for the 55 °C cases can be explained
11, 3421 9 of 16
by Figure 2: At that temperature, they were still entirely desorbed.

Figure8.8. Volumetric
Figure Volumetricstorage
storagecapacity
capacityof
ofthe
theselected
selectedmaterials
materialsfor
fortwo
twodifferent
differentde- de-and
andadsorption
adsorption
temperatures(specified
temperatures (specifiedbelow
belowmaterial
materialnames
namesininthe
thefollowing
followingorder;
order;TTdes -Tads
des-T ads-T
-Tcond-T-T
cond evap). ).
evap

3.1.3.For idealper
Impact short-term storage (discharge
Storage Capacity beginsLevel
on the Material immediately after charging, so no sensible heat
is lost), the sum of sorptive and maximum sensible capacity must be considered. This value can be
After their separate analysis, the storage capacity and the environmental impact were combined
compared directly to the reference capacities of water (also loss-free, with a useful capacity between ad-
(Figure 9) in order to understand the specific impact of each sorbent. Due to its high GWP, SAPO-34
and de-sorption temperatures). Interestingly, the regular sorption materials rarely outperformed water
had the highest specific impact of several thousand times higher than that of water. This held true
in terms of volumetric density. Especially for the 35 ◦ C cases, SAPO-34 reached the storage capacities
even in the 35 °C cases, in which it appeared rather favorable in the preceding storage capacity section.
noticeably above that of the reference. LiCl-Vermiculite significantly outperformed the remaining
For silica gel and the MOFs in 35 °C cases, the multiplier lies in the range of various hundreds. The salt-
materials with a volumetric storage capacity up to four times that of water.
composite showed the best result in these scenarios, with an impact ~120 higher than that of water
Since 120 ◦ C as maximum temperature for water storages means additional effort for pressure-
(Tdes = 90 °C).
resistant systems, the 90 ◦ C bars can be adopted as references as well, which improves the perspectives
of sorption storage.
For longer term storage, the resulting storage capacity can be obtained by adding the (negative)
green bar, which represents the energy required to heat the sorbent from ambient to adsorption
temperature, to the sorptive capacity. In this case, the effective capacity of the water storage will
also be lower and needs to be corrected by an appropriate efficiency [35]. Assuming, for example,
50% sensible losses for the water storage and 100% sensible losses for the sorption alternatives due to
less insulation, some more sorption storages show benefits compared to water.
The fact that the MOFs did not show any sorptive capacity for the 55 ◦ C cases can be explained
by Figure 2: At that temperature, they were still entirely desorbed.

3.1.3. Impact per Storage Capacity on the Material Level


After their separate analysis, the storage capacity and the environmental impact were combined
(Figure 9) in order to understand the specific impact of each sorbent. Due to its high GWP, SAPO-34
had the highest specific impact of several thousand times higher than that of water. This held true
Figure 9. Global warming potential per storage capacity of selected materials (left axis: sorption materials;
even in the 35 ◦ C cases, in which it appeared rather favorable in the preceding storage capacity
right axis: water) for different boundary conditions (Tdes-Tads-Tcond-Tevap); for MOFs and Tads = 55 °C, c min
section. For silica gel and the MOFs in 35 ◦ C cases, the multiplier lies in the range of various hundreds.
is negative, so no values are shown.
The salt-composite showed the best result in these scenarios, with an impact ~120 higher than that of
water (Tdes = 90 ◦ C).
(Figure 9) in order to understand the specific impact of each sorbent. Due to its high GWP, SAPO-34
had the highest specific impact of several thousand times higher than that of water. This held true
even in the 35 °C cases, in which it appeared rather favorable in the preceding storage capacity section.
For silica gel and the MOFs in 35 °C cases, the multiplier lies in the range of various hundreds. The salt-
composite
Energies 2018,showed
11, 3421 the best result in these scenarios, with an impact ~120 higher than that of 10
water
of 16
(Tdes = 90 °C).

Energies 2018, 11, x FOR PEER REVIEW 10 of 16

3.2. Results on Component Level


Figure
Figure
The 9.9.Global
evaluation warming
Global warming
on potential
component per
potential storage
pertakes
level capacity
storage m3ofstorage
selected
a 1capacity materials
of selected (left
volumematerials axis:(left
(excluding sorption
axis:materials;
sorption
insulation) as the
right
reference. axis:
materials;
In thewater)
right for
axis:
case different
of water) boundary conditions
for different
the water storage,boundary (T -T -T
this wasconditions
des ads cond-T ); for MOFs
(Tdes -Tads -T
evap
a market-average and
cond -T
steel T = 55 °C,
evap ); forwith
ads
cylinder MOFs ca min
and
mass of
◦ C,
147.5is
Tkgnegative,
= 5510
adsand
socno
cmmin values are shown.
is negative,
mineral so no values[36].
wool insulation are shown.
The sorption storage was assumed to be a cube with
a3.2.
1 mResults
edge length and made
on Component of 1 mm stainless steel sheets plus 3.5 cm of insulation.
Level
3
3.2.1. The evaluation
Component on component
Storage Capacity level takes a 1 m storage volume (excluding insulation) as the
reference. In the case of the water storage, this was a market-average steel cylinder with a mass of
The storage capacity for the 1 m3 reference volume is presented in Figure 10. The values for water
147.5 kg and 10 cm mineral wool insulation [36]. The sorption storage was assumed to be a cube with
correspond to those in Figure 8, since a water storage requires no built-in components. The sorption
a 1 m edge length and made of 1 mm stainless steel sheets plus 3.5 cm of insulation.
storages show lower capacities because the entire volume was not available for sorbents. The ECHX
was
3.2.1.considered
Component asStorage
entirelyCapacity
inactive in terms of storage capacity while the SHX contributed to the
sensible share. Based on the sorbent-to-metal and sorbent-to-total-sorption-section volume ratios as
The storage capacity for the 1 m3 reference volume is presented in Figure 10. The values for water
well as the resulting geometry of an adapted ECHX, the available volume for sorbent pellets was
correspond to those in Figure 8, since a water storage requires no built-in components. The sorption
0.819 m3. The SHX had a material volume of 12.6 L and the ECHX had a material volume of 2.9 L and
storages show lower capacities because the entire volume was not available for sorbents. The ECHX
a construction volume of 43 L. Table 4 lists the corresponding sorbent pellet masses and the resulting
was considered as entirely inactive in terms of storage capacity while the SHX contributed to the
mass ratios. The vapor volume was assumed to constitute 10% of the volume not occupied by heat
sensible share. Based on the sorbent-to-metal and sorbent-to-total-sorption-section volume ratios as
exchangers and the sorbent.
well as the resulting geometry of an adapted ECHX, the available volume for sorbent pellets was
0.819 m3 . The SHX had
Table 4. aMain
material volume of
characteristics of 12.6 L and
sorption the ECHX
storage had a material
with considered volume of 2.9 L and
materials.
a construction volume of 43 L. Table 5 lists the corresponding sorbent pellet masses and the resulting
Description Silica Gel SAPO-34 Z13X CAU-10-H Al-Fum LiCl-Ver
mass ratios. The vapor volume was assumed to constitute 10% of the volume not occupied by heat
Mass (kg) 446.7 570.6 610.1 435.1 324.8 286.5
exchangers andtothe
Sorbent pellet SHX sorbent.
mass ratio (kg/kg) 4.0 5.1 5.4 3.9 2.9 2.5

Figure
Figure 10.
10. Storage
Storage capacity
capacity ofof 11 m
m3 storages
3
storages with
with the
the selected
selected materials
materials under
under different
different boundary
boundary
conditions (Tdes-T-Tads-T-T
conditions (T cond-Tevap). ).
-Tevap
des ads cond

Thus, considering idealized, loss-free short-term storage, among the typical sorbents, only the zeolite
SAPO-34 showed a higher storage density on component level compared to water. The outstanding
sorption-like LiCl-Vermiculite achieved a storage density ~3.5 higher than that of water. Assuming
long-term storage as above (50% sensible losses for the water storage and 100% sensible losses for the
Energies 2018, 11, 3421 11 of 16

Table 5. Main characteristics of sorption storage with considered materials.

Description Silica Gel SAPO-34 Z13X CAU-10-H Al-Fum LiCl-Ver


Mass (kg) 446.7 570.6 610.1 435.1 324.8 286.5
Sorbent pellet to SHX mass ratio (kg/kg) 4.0 5.1 5.4 3.9 2.9 2.5

Thus, considering idealized, loss-free short-term storage, among the typical sorbents, only the
zeolite SAPO-34 showed a higher storage density on component level compared to water.
The outstanding sorption-like LiCl-Vermiculite achieved a storage density ~3.5 higher than that
of water. Assuming long-term storage as above (50% sensible losses for the water storage and 100%
sensible losses for the sorption alternatives), some more sorption storages showed benefits compared
to water. Even then, their relative storage densities compared to water remained below two.

3.2.2. Environmental Impact of the Storage Component


Figure 11 breaks up the GWP of the resulting 1 m3 storages by the main parts. Interestingly,
the steel cylinder for the water storage with a mass of 147.5 kg and the stainless steel housing of
the sorption storages (47.4 kg) had a very similar impact due to the higher specific impact of the
latter metal. The thicker insulation of the water storage evidently resulted in a higher absolute11GWP.
Energies 2018, 11, x FOR PEER REVIEW of 16
The water itself had a negligible impact. The SHX contributed about half the GWP of the containment,
and the ECHX
contributed contributed
roughly anotherroughly
10%. Theanother 10%.accounted
sorbents The sorbents accounted
for the for the
major share ofmajor share GWP:
the overall of the
overall GWP: From 69%, in case of silica gel, up
From 69%, in case of silica gel, up to 98% for SAPO-34. to 98% for SAPO-34.

Figure 11. GWP of the single storage component elements (for 1 m33 storages).

3.2.3. Impact
3.2.3. Impact per
per Storage
Storage Capacity
Capacity on
on the
the Component
Component Level
Level
In order
In order toto determine
determine the the overall
overall environmental
environmental impact, the impact
impact, the factors of
impact factors of the
the standard
standard
materials (metals
materials (metals and
and insulation)
insulation) were
were extracted
extracted from
from the
the Gabi
Gabi database
database [36]
[36] and
and summed
summed to to the
the
sorbent values. Figure 12 links the storage capacity to the resulting overall global warming
sorbent values. Figure 12 links the storage capacity to the resulting overall global warming potential potential of
thethe
of sorption
sorptioncomponents,
components, normalized
normalized to the respective
to the results
respective forfor
results thethe
water storage.
water storage.
The impact of a storage with composite material was higher compared tothe
The impact of a storage with composite material was higher compared to thewater
waterreference
referencebybya
a factor of ~2.5, and classical sorption materials showed minimum factors around 10. Inworst
factor of ~2.5, and classical sorption materials showed minimum factors around 10. In the case
the worst
(SAPO-34), the factor ◦
case (SAPO-34), the went
factorupwent
to ~100. Assuming
up to a 90 C maximum
~100. Assuming a 90 °C reference
maximum storage temperature
reference storage
for the 120 ◦ C desorption temperature scenarios, 120 ◦ C results can be reduced to ~50% and ~65%,
temperature for the 120 °C desorption temperature scenarios, 120 °C results can be reduced to ~50%
respectively,
and for 55 ◦ C and
~65%, respectively, for35
◦ adsorption temperature ((90 ◦ C–55 or 35 ◦ C)/(120 ◦ C–55 or 35 ◦ C)).
55 C
°C and 35 °C adsorption temperature ((90 °C–55 or 35 °C)/(120 °C–55
or 35 °C)).
The impact of a storage with composite material was higher compared to the water reference by
a factor of ~2.5, and classical sorption materials showed minimum factors around 10. In the worst
case (SAPO-34), the factor went up to ~100. Assuming a 90 °C maximum reference storage
temperature for the 120 °C desorption temperature scenarios, 120 °C results can be reduced to ~50%
and ~65%, respectively, for 55 °C and 35 °C adsorption temperature ((90 °C–55 or 35 °C)/(120 °C–55
Energies 2018, 11, 3421 12 of 16
or 35 °C)).

Figure
Figure 12. GWPGWPper perstorage
storage capacity
capacity of of sorption
sorption storages
storages under
under different
different boundary
boundary conditions
conditions (Tdes-
(T -T
des-Tcond
Tads -T
ads-Tevap -T )
) compared
cond evap compared to
to water. water.
Energies 2018, 11, x FOR PEER REVIEW 12 of 16
3.2.4. Sensitivity Analysis
3.2.3. Sensitivity Analysis
lifetime of the storage and ignoring any additional effort for dismantling or reconditioning, the GWP
In order to get an impression on the sensitivity of the above results to the single inputs and
of the storages
In orderdropped
to get anbyimpression
between ~35% (silica
on the gel) andof48%
sensitivity the(SAPO-34).
above results to the single inputs and
assumptions, Figure 13 shows their sensitivity to certain variations. The reduction of the material
Replacing the
assumptions, energy-intensively
Figure produced silicon-based
13 shows their sensitivity binder by
to certain variations. Thecommon but
reduction of less
the water-
material
impact evidently had the largest impact. Assuming that 50% of it can be recycled after the useful
resistant
impact organic
evidentlyalternatives, methylcellulose
had the largest and polyvinyl
impact. Assuming that 50%alcohol, alsobe
of it can showed benefits
recycled of up
after the to
useful
lifetime of the storage and ignoring any additional effort for dismantling or reconditioning, the GWP
15% depending on its relative impact. Reducing the metal mass, be it in the SHX, ECHX, or the
of the storages dropped by between ~35% (silica gel) and 48% (SAPO-34).
containment, also yielded a GWP decrease.

Figure 13.13.
Figure Influence ofof
Influence different measures
different measuresononGWP
GWPofofthe
thestudied
studiedstorage
storageconfigurations.
configurations.

4. Conclusions
Replacing the energy-intensively produced silicon-based binder by common but less
water-resistant organic alternatives,
The storage capacity methylcellulose
and environmental impact (in and polyvinyl
terms of GWP)alcohol, also showed
of five classical benefits
sorbents and
of up to 15% depending on its relative impact. Reducing the metal mass, be it in the
one sorbent-like material regarded promising for application in heat storages for building SHX, ECHX,
or the containment,
applications also yielded a GWP decrease.
were studied.
On the material level, it turned out that the storage capacity of classical sorbents under four
4. Conclusions
application scenarios was lower compared to water than that commonly reported in literature, even
thoughThe storage capacity
temperature and environmental
gradients impact
occurring in real (in termshave
applications of GWP)
beenofneglected.
five classical sorbents
Among and
them,
one sorbent-like
SAPO-34 showed material
the best regarded promising
performance for application
and reached capacityinfactors
heat storages
betweenfor2 building applications
and 4 compared to
were studied.
water, depending on the assumed boundary conditions values. Silica gel and the MOFs turned out
to be unsuitable for 55 °C heating applications, as they were not able to adsorb significant amounts
of water vapor at this temperature. The sorbent-like salt-composite (LiCl-Vermiculte) showed the
highest volumetric storage capacity, roughly twice that of SAPO-34.
At the same time, the manufacturing of all materials leads to environmental impacts much
higher than that of water. Again, SAPO-34 stood out of the mass, this time because of its very
significant impact, approximately 10 times higher than that of the zeolite, the MOFs, and the
Energies 2018, 11, 3421 13 of 16

On the material level, it turned out that the storage capacity of classical sorbents under four
application scenarios was lower compared to water than that commonly reported in literature,
even though temperature gradients occurring in real applications have been neglected. Among
them, SAPO-34 showed the best performance and reached capacity factors between 2 and 4 compared
to water, depending on the assumed boundary conditions values. Silica gel and the MOFs turned out
to be unsuitable for 55 ◦ C heating applications, as they were not able to adsorb significant amounts of
water vapor at this temperature. The sorbent-like salt-composite (LiCl-Vermiculte) showed the highest
volumetric storage capacity, roughly twice that of SAPO-34.
At the same time, the manufacturing of all materials leads to environmental impacts much higher
than that of water. Again, SAPO-34 stood out of the mass, this time because of its very significant
impact, approximately 10 times higher than that of the zeolite, the MOFs, and the composite, and almost
30 times higher than that of silica gel. The specific impact per storage capacity of the storage materials
was above around one hundred (LiCl-Vermiculite) to several thousands (SAPO-34) times higher than
that of water.
The subsequent evaluation on the component level, which includes auxiliary parts required
for the exploitation of the storage potentials, such as containment, insulation, and heat exchangers,
brings the results closer to an actual application. A 1 m3 containment volume was chosen as the
reference. The volumetric storage density of the sorption variants was further decreased by these
additional components. It still reached a factor of ~7 for the salt-composite but was only 2.5 times
higher than that of water in the best case of SAPO-34. The environmental impact of all storage types
suffered from the incorporation of the additional components. The specific impact of sorption storages
was about 2.5 to 100 times higher than that of water. Only with very optimistic (and not necessarily
realistic) assumptions such as recycling of the sorbent pellets, highly efficient and light-weight auxiliary
components, and long term storage which makes water storages a very lossy alternative, can an
environmental benefit compared to this reference seem possible.
The presented findings suggest that the chances that closed adsorption storages in building
applications can provide an environmental benefit are small. When a significant increase of storage
capacity compared to water is required, salt-composites are an option under the preconditions
that, for the entire aspired useful life time, salt leakage (also due to condensation) and corrosion
can be eliminated. For the classical sorbents, more promising applications should fulfill the
following conditions:

• Better exploitation of the potential uptake of the sorbent;


• Seek applications where (non-pressurized) water storage cannot serve as benchmark, e.g., due to
temperature limitations; and
• Minimize the amount and mass of auxiliary equipment.

For new material developments, the presented methodology may provide guidance during
the selection and optimization of raw materials and production routes in order to minimize the
environmental impact. Further thermochemical storage materials such as liquid sorbents still need to
be evaluated, yet it is likely that they face similar challenges to solid sorbents.

Author Contributions: D.F. and T.H. prepared the material data for the LCIA performed by R.H. and the
performance calculations conducted by B.N.; B.N. conceived the study under guidance of P.S. and G.M. and
drafted the manuscript with input from the above named authors to the respective sections. All authors read and
approved the final manuscript.
Funding: This research was funded by the Germany Federal Ministry of Economic Affairs and Energy and
supported by Project Management Jülich under grant number 03ET1333.
Conflicts of Interest: The authors declare no conflict of interest. The funders had no role in the design of the
study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to
publish the results.
Energies 2018, 11, 3421 14 of 16

Nomenclature
Symbol Description Unit
A Adsorption Potential J/g
b Characteristic Dubinin-Astakhov Parameter kg/J
c m,max maximum specific heat capacity (loss-free) J/kg/K
c m,min minimum specific heat capacity (100% sensible losses) J/kg/K
cp specific isobaric heat capacity J/kg/K
d fin fin thickness m
d sp,fin fin spacing m
d sp,tb tube spacing m
d tb tube diameter m
h evap evaporation enthalpy kJ/kg
n exponent -
psat saturation pressure Pa
p Tsorb actual pressure at sorbens temperature Pa
R Ideal gas constant of water vapor J/kg/K
r fin,tb Resistance at W/m2 /K
r sorb,fin W/m2 /K
T Ads Adsorption temperature K
T amb Ambient temperature K
T Des Desorption temperature K
T sorb Sorbens temperature K
X Sorbent uptake g/g
X uptake g/g
X0 Maximum sorbent uptake g/g
λ fin Fin thermal conductivity W/m/K
λ sorb Sorbent thermal conductivity W/m/K
λ tb Tube thermal conductivity W/m/K

Appendix A

Table A1. Overview on material properties and life cycle impact values used for calculations.

Density Sensible Heat Capacity GWP PE


Material
(kg/m3 ) (kJ/kg/K) (kg CO2 eq/kg) (MJ/kg)
Silica gel 546 1.0 2.3 19.3
SAPO-34 650 1.4 50.2 745.3
Z13X 704 1.7 5.1 62.0
CAU-10-H 466 1.6 3.5 58.1
Al-Fum 316 1.7 4.8 70.9
LiCl-Ver 350 1.0 8.5 85.4
Water 1000 4.2 0.0 0.06
Silicon-based binder 1120 1.5 11.1 191.9
Methylcellulose 1260 1.5 1.4 -
Polyvinyl alcohol 1190 1.5 3.6 -
Copper 8940 - 1.1 -
Mineral wool 80 - 1.4 -
Steel 7860 - 1.8 -
Stainless steel 8940 - 5.9 -
Energies 2018, 11, 3421 15 of 16

References
1. Arce, P.; Medrano, M.; Gil, A.; Oró, E.; Cabeza, L.F. Overview of thermal energy storage (TES) potential
energy savings and climate change mitigation in Spain and Europe. Appl. Energy 2011, 88, 2764–2774.
[CrossRef]
2. Aydin, D.; Casey, S.P.; Riffat, S. The latest advancements on thermochemical heat storage systems.
Renew. Sustain. Energy Rev. 2015, 41, 356–367. [CrossRef]
3. Scapino, L.; Zondag, H.A.; van Bael, J.; Diriken, J.; Rindt, C.C.M. Sorption heat storage for long-term
low-temperature applications: A review on the advancements at material and prototype scale. Appl. Energy
2017, 190, 920–948. [CrossRef]
4. Stach, H.; Mugele, J.; Jänchen, J.; Weiler, E. Influence of Cycle Temperatures on the Thermochemical Heat
Storage Densities in the Systems Water/Microporous and Water/Mesoporous Adsorbents. Adsorption 2005,
11, 393–404. [CrossRef]
5. Zettl, B.; Englmair, G.; Somitsch, W. An Open Sorption Heat Storage Concept and Materials for Building
Heat Supply. Energy Procedia 2015, 73, 297–304. [CrossRef]
6. Frazzica, A.; Freni, A. Adsorbent working pairs for solar thermal energy storage in buildings. Renew. Energy
2017, 110, 87–94. [CrossRef]
7. Jänchen, J.; Ackermann, D.; Stach, H.; Brösicke, W. Studies of the water adsorption on Zeolites and modified
mesoporous materials for seasonal storage of solar heat. Sol. Energy 2004, 76, 339–344. [CrossRef]
8. Jeremias, F.; Lozan, V.; Henninger, S.K.; Janiak, C. Programming MOFs for water sorption: Amino-
functionalized MIL-125 and UiO-66 for heat transformation and heat storage applications. Dalton Trans.
2013, 42, 15967–15973. [CrossRef]
9. Grekova, A.D.; Gordeeva, L.G.; Aristov, Y.I. Composite “LiCl/vermiculite” as advanced water sorbent for
thermal energy storage. Appl. Therm. Eng. 2017, 124, 1401–1408. [CrossRef]
10. Gaeini, M.; Rouws, A.L.; Salari, J.W.O.; Zondag, H.A.; Rindt, C.C.M. Characterization of microencapsulated
and impregnated porous host materials based on calcium chloride for thermochemical energy storage.
Appl. Energy 2018, 212, 1165–1177. [CrossRef]
11. Permyakova, A.; Skrylnyk, O.; Courbon, E.; Affram, M.; Wang, S.; Lee, U.-H.; Valekar, A.H.; Nouar, F.;
Mouchaham, G.; Devic, T.; et al. Synthesis Optimization, Shaping, and Heat Reallocation Evaluation of
the Hydrophilic Metal-Organic Framework MIL-160(Al). ChemSusChem 2017, 10, 1419–1426. [CrossRef]
[PubMed]
12. Bales, C.; Gantenbein, P.; Jaenig, D.; Kerskes, H.; van Essen, M.; Weber, R. Final Report of IEA SHC Task 32
Subtask B—“Chemical and Sorption Storage”; IEA Solar Heating and Cooling Programme: Paris, France, 2008.
13. SolSpaces—Entwicklung und Erprobung Einer Autarken Solaren Wärmeversorgung für Energieeffiziente
Kompaktgebäude. Available online: http://www.itw.uni-stuttgart.de/forschung/projekte/abgeschlossen/
solspaces.html (accessed on 29 August 2018).
14. Scapino, L.; Zondag, H.A.; van Bael, J.; Diriken, J.; Rindt, C.C.M. Energy density and storage capacity cost
comparison of conceptual solid and liquid sorption seasonal heat storage systems for low-temperature space
heating. Renew. Sustain. Energy Rev. 2017, 76, 1314–1331. [CrossRef]
15. Munz, G.; Jeremias, F.; Krämer, B.; Wolff, T.; Henninger, S.K. Adsorbents for Seasonal Heat Storage—Economical
and Energetic Assessment; Deutsche Zeolith Tagung: Gießen, Germany, 2016.
16. Vasta, S.; Brancato, V.; La Rosa, D.; Palomba, V.; Restuccia, G.; Sapienza, A.; Frazzica, A. Adsorption Heat
Storage: State-of-the-Art and Future Perspectives. Nanomaterials 2018, 8, 522. [CrossRef] [PubMed]
17. Horn, R.; Burr, M.; Fröhlich, D.; Gschwander, S.; Held, M.; Lindner, J.P.; Munz, G.; Nienborg, B.; Schossig, P.
Life Cycle Assessment of Innovative Materials for Thermal Energy Storage in Buildings. Procedia CIRP 2018,
69, 206–211. [CrossRef]
18. Bertsch, F. Umsetzung und Bewertung der Thermochemischen Energiespeicherung für die Solare
Gebäudebeheizung. Ph.D. Thesis, University of Stuttgart, Stuttgart, Germany, 2014.
19. Deutsches Institut für Normung. DIN EN ISO 14044:2018-05 Environmental Management—Life Cycle
Assessment—Requirements and Guidelines; Beuth Verlag GmbH: Berlin, Germany, 2018.
20. Deutsches Institut für Normung. DIN EN ISO 14040:2009-11 Environmental Management—Life Cycle
Assessment—Principles and Framework; Beuth Verlag GmbH: Berlin, Germany, 2009.
Energies 2018, 11, 3421 16 of 16

21. Deutsches Institut für Normung. DIN EN 15978:2012-10 Sustainability of Construction Works—Assessment of
Environmental Performance of Buildings—Calculation Method; Beuth Verlag GmbH: Berlin, Germany, 2012.
22. Deutsches Institut für Normung. DIN EN 15804:2014-07 Sustainability of Construction Works—Environmental
Product Declarations—Core Rules for the Product Category of Construction Products; Beuth Verlag GmbH: Berlin,
Germany, 2014.
23. Fantke, P. UNEP/SETAC Scientific Consensus Model for Characterizing Human Toxicological and Ecotoxicological
Impacts of Chemical Emissions in Life Cycle Assessment; USEtox®International Center Hosted at the Technical
University of Denmark: Lyngby, Demark, 2017.
24. Newalkar, B.L.; Komarneni, S. Synthesis and Characterization of Microporous Silica Prepared with Sodium
Silicate and Organosilane Compounds. J. Sol-Gel Sci. Technol. 2000, 18, 191–198. [CrossRef]
25. Liu, G.; Tian, P.; Li, J.; Zhang, D.; Zhou, F.; Liu, Z. Synthesis, characterization and catalytic properties of
SAPO-34 synthesized using diethylamine as a template. Microporous Mesoporous Mater. 2008, 111, 143–149.
[CrossRef]
26. De Lucas, A.; Uguina, M.A.; Covian, I.; Rodriguez, L. Synthesis of 13X zeolite from calcined kaolins and
sodium silicate for use in detergents. Ind. Eng. Chem. Res. 1992, 31, 2134–2140. [CrossRef]
27. Lenzen, D.; Bendix, P.; Reinsch, H.; Fröhlich, D.; Kummer, H.; Möllers, M.; Hügenell, P.P.C.; Gläser, R.;
Henninger, S.; Stock, N. Scalable Green Synthesis and Full-Scale Test of the Metal-Organic Framework
CAU-10-H for Use in Adsorption-Driven Chillers. Adv. Mater. 2018, 30. [CrossRef]
28. Leung, E.; Müller, U.; Trukhan, N.; Mattenheimer, H.; Cox, G.; Blei, S. Process for Preparing Porous
Metal-Organic Frameworks Based on Aluminum Fumarate. US Patent 8,524,932 B2, 30 September 2011.
29. Wang, Y.; LeVan, M.D. Adsorption Equilibrium of Carbon Dioxide and Water Vapor on Zeolites 5A and 13X
and Silica Gel: Pure Components. J. Chem. Eng. Data 2009, 54, 2839–2844. [CrossRef]
30. Onyango, M.S.; Kojima, Y.; Aoyi, O.; Bernardo, E.C.; Matsuda, H. Adsorption equilibrium modeling and
solution chemistry dependence of fluoride removal from water by trivalent-cation-exchanged zeolite F-9.
J. Colloid Interface Sci. 2004, 279, 341–350. [CrossRef]
31. Gorbach, A.; Stegmaier, M.; Eigenberger, G. Measurement and Modeling of Water Vapor Adsorption on
Zeolite 4A—Equilibria and Kinetics. Adsorption 2004, 10, 29–46. [CrossRef]
32. Foo, K.Y.; Hameed, B.H. Insights into the modeling of adsorption isotherm systems. Chem. Eng. J. 2010, 156,
2–10. [CrossRef]
33. Shah, R.K.; Sekulić, D.P. Fundamentals of Heat Exchanger Design; Wiley-Interscience: Hoboken, NJ, USA, 2003.
34. Volmer, R.; Eckert, J.; Füldner, G.; Schnabel, L. Evaporator development for adsorption heat transformation
devices—Influencing factors on non-stationary evaporation with tube-fin heat exchangers at sub-atmospheric
pressure. Renew. Energy 2017, 110, 141–153. [CrossRef]
35. Data Measurement and Analysis of a Solar Heating System with Seasonal Storage. Energy Procedia 2015, 70,
241–248. [CrossRef]
36. Gabi Product Databases; Thinkstep AG: Stuttgart, Germany, 2018.

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).