Category Archives: EUROSUN 2008

Air-sand heat exchanger

The heat exchanger should show the following properties:

• Good heat transfer, heat losses < 20 %, temperature difference between sand outlet and air inlet temperature < 10 %

• Low pressure loss ApAir < 5000 Pa

• Compact design

To meet these requirements a concept has been chosen, which is featured by:

• Cross flow arrangement of heat transfer media air and sand

• Low width design for reduction of pressure drop in the moving packed bed

The hot air enters through a porous wall, passes through the moving bed of sand, and flows out at lower temperature through another porous wall, see Fig. 3.

The requirements to be met by the porous walls include:

• Thermal strength

• Abrasion resistance

• Low pressure drop

• Mechanical strength

The Drake Landing Solar Community

The Drake Landing Solar Community (DLSC) consists of a small suburb of 52 homes, where at least 90% of the space heating load is to be provided from solar thermal energy within five years of its operation. A description of the DLSC’s operation is provided elsewhere [3]. In use at the DLSC are two short-term thermal stores (STTSs) which interact with the various thermal systems at the site. The STTSs are typical liquid thermal energy stores, albeit large. Their configuration is illustrated in Figure 1.



1st International Congress on Heating, Cooling, and Buildings, ■ 7th to 10th October, Lisbon — Portugal /

t ank Dimensions.

• ~11.5 m (length)

• ~3 m (diameter)

separating flow within each tank

Operating Flow Rates:

• 3.35 L/s to 14.9 L/s (Re ~ 100,000 ^ 450,000)

Connector Pipe:

• 0150 mm

Tank Insulation:

• R-20 along tank shell

• Adiabatic conditions assumed during charging/discharging

Three Operating Loops:

Solar Collector Loop (0100 mm) District Heating Loop (075 mm) Seasonal Storage Loop (050 mm)

Figure 1. The Short-Term Thermal Stores at the Drake Landing Solar Community

Prior Work: A CFD Model of the Drake Landing STTSs

In a previous study by the authors [4], an effective CFD model of the STTSs was developed using the FLUENT 6.3.26 commercial software package. The model was validated against real performance data recorded at the Drake Landing site in October 2007. The modelling procedure developed in the previous work is applied similarly in this study, and is listed below in Table 1. Further information is provided in the original paper.

♦ PISO Pressure-Velocity coupling with neighbour correction


♦ Double-precision, ♦

segregated 1st-order implicit unsteady solver

♦ k-epsilon Realizable turbulent model

♦ Default simulation convergence criteria

Polynomial correlation to define temperature dependent material properties (i. e., density). Thermodynamic data: [5]

Pressure under-relaxation of 0.9 Momentum under-relaxation of 0.2 Default values for all other components

Second Order Upwind ♦

discretization of momentum, turbulence, and energy. Body Force Weighted discretization of pressure

Подпись: Polynomial correlation to define temperature dependent material properties (i.e., density). Thermodynamic data: [5]

Table 1. Modelling parameters applied to the CFD simulation of the STTSs

Ice Store

For the purposes described above, the ice store has to comply with several requirements: manufactured from standard components, high capacity, low cost, high flexibility during charging and discharging performance and use of cheap, easy-to-handle and ecologically harmless phase change material. Based on these requirements a prototype was built at the beginning of 2007 and experimentally investigated. Possible heat exchanger materials are copper and polypropylene. These materials may offer satisfactory thermal conductivity. They also provide the possibility to operate the ice store as heat store, as well. This may be necessary if the chiller is used as a heat pump. The tank is a standard hot water heat store with a volume of approximately 500 litres. Especially simultaneous charging and discharging require external melting. For this purpose, the ice store is separated into two circuits. The first circuit is for the coolant and connected to the evaporator of the absorption chiller. The second circuit has two functions: 1) the water in the second circuit freezes at the surface of the heat exchanger and 2) water which does not freeze is used for air conditioning. Hence, the possibility of water flow into and through the tank has to be assured at any time. Therefore complete freezing of water in the tank is not possible.

Another possibility to discharge the ice store is internal melting. In this case the heat exchanger in the store is connected to the chilled ceiling circuit. The ice will melt from the inside to the outer boundaries of the ice layer. In this operational mode of the ice store simultaneous charging and discharging is not possible.


Dronninglund Fjernvarme is a district heating company producing 40.000 MWh/year with natural gas fuelled CHP and boilers. Dronninglund has got support to design an energy system with app. 35.000 m2 solar collectors, 50.000 m3 pit heat seasonal storage and 3 MW compression heatpump (thermal output) covering 50% of the yearly consumption. The heat pump uses CO2 as medium and can produce hot water at 80 oC as needed for forward temperature in the district heating sys­tem.

The pit heat storage will be of same type as the 10.000 m3 storage in Marstal, but the floating cover construction will be changed to a solution where LECA is uses instead of mineral wool and EPS as insulation and where the cover can be parted in sections making it possible to construct large covers and reduce problems if the constructions is not tight

Fig. 2. Construction cross section, Marstal

Roof-foi hvt. Underlayment for roof-foi

SecuGrid 30 x 30 380-500 mm Leca

Geotextile kl. IV

2,5 mm HDPE Polymermenbrane BAM

3 Horizontal valleys and 4 toppoint

bottom of valley

Minimun 4 permille slope.

1 — 4 drain pipes



Water 2.5 mm HDPE



Geotexti e к. V

20000 85000.

Fig. 3. Contraction croos section, Dronninglund

The heat production price is calculated to 70 €/MWh (annuity 0,1). Investment costs are app. 11. mio. €. The Danish Energy Agency is applied for support.

Testing facility for Phase Change Slurries

Bjorn Nienborg[12]*, Stefan Gschwander1, Li Huang[13], Peter Schossig1
1 Fraunhofer Institute for Solar Energy Systems, Heidenhofstr. 2, 79110 Freiburg, GERMANY
2 Fraunhofer Institute for Environmental-, Safety — and Energy Technologies, Osterfelder Str. 3, 46047


* Bjoern. Nienborg@ise. fraunhofer. de

Phase change materials (PCM) offer a great potential for energy saving in heating and cooling applications as well as efficient energy storage. A series of solid materials has come into the market during the last years. Phase Change Slurries (PCS) are mixtures of a Phase Change Material and a carrier fluid so the material can be pumped. At Fraunhofer Institute for Solar Energy Systems (ISE) two types of PCS are investigated: emulsions and suspensions.

As PCS compete with water which is usually used as pumpable heat transfer medium, they need to meet a multitude of requirements in order to compensate the higher investment cost by lower operation expenses.

At our institute the materials can be analyzed on small scale in a laboratory. Once a promising product is detected, it is tested in a testing facility which reproduces reality-like operation. This work describes the testing facility and illustrates the parameters that can be measured. Subsequently results of an example measurement of a phase change emulsion for cooling applications are presented.

Keywords: phase change material, PCM, Phase Change Slurry, PCS, thermal energy storage

these ice-slurries a large amount of energy can be transferred at 0 °C. For many applications e. g. building climatization with comfort temperatures of 22 °C — 24 °C, working with temperatures around the freezing point has a great disadvantage as they unnecessarily reduce the efficiency of the chiller [5]. For this reason Fraunhofer ISE is working on PCS on basis of paraffin with the melting point close to the targeted working temperature.

There are two types of PCS being investigated:

— Emulsions with the dispersed paraffin mixed directly with water and emulsifiers preventing the accumulation of the paraffin drops.

— Suspensions of microencapsulated paraffin with the shell preventing an interaction of the paraffin with the water.

A testing facility for PCS has been set up at Fraunhofer ISE, which allows the analysis of their physical properties and the suitability for the designated use.

Experimental study of composite sorbents

2.1. Experimental setup

The chemical storage is designed to undergo the endothermic heat storage reaction during the summer, using heat obtained from tube solar collectors, and release the heat stored by using the exothermic reverse reaction during the winter. The hydration/dehydration cycle of magnesium sulphate has previously been identified as particularly promising [3]. An experimental study has been performed on this material to describe and understand its behaviour during its hydration by water vapour. The principal thermal properties of the MgSO4 / MgSO4.7H2O system are presented in Table 2.


MgSO4 + 7 H2O ^ MgSO4.7H2O hydration, exothermic — dehydration, endothermic

Enthalpy of hydration

— 410 kJ. mol-1

Temperature of dehydration


Density of the dense salt


2660 kg. m-3


1680 kg. m-3

Energy density of the dense salt


9.06 GJ. m-3 / 2.52 MWh. m-3


2.80 GJ. m-3 / 0.78 MWh. m-3

Table 2. Properties of the MgSO4 / MgSO4.7H2O system [4]

By dispersing the magnesium sulphate over a larger exchange surface, the thermal power released during hydration is greatly increased. Composite materials made of a sorption material and magnesium sulphate were prepared in such a way to preserve the porosity of the sorption material. This feature is essential in order to allow the water vapour to react with both the salt and the matrix. Sorption materials, such as zeolites and silica gels, were tested as host materials for the salt composites.

The samples were hydrated in a thermally insulated (non-isothermal) open reactor using a flow of moist air at a relative humidity close to 100% at ambient temperature.

2.2. Results

For all of the hydration tests, the rate of water uptake (grams of water per gram of material per hour) is calculated from the exit air humidity and temperature measurements. These values qualify the thermal characteristics of the system : the water uptake is proportional to the heat produced and the rate of water uptake is proportional to the power delivered by the reaction (Fig. 6).

Fig. 6. Rate of water uptake during hydration

High rates of water uptake are obtained for a much longer period with the zeolite/MgSO4 composite ZM15 compared to the analogous silica composite SM16 and the reference pure materials (Fig. 6). This results in higher temperature lifts for the zeolite composite (Fig. 7). Maintaining the high rates of reaction over a long time will ensure that a maximum amount of usable high-grade heat is produced.

Fig. 7. Temperature lifts and peaks of power density during hydration

These peaks of temperature lift are associated with peaks of power of 28.4, 20.1, 20.7 and 15.6 mW per gram of material for the hydration of ZM15, zeolite, SM16 and silica gel respectively. The zeolite/ magnesium sulphate composite proves more favourable than the choice of a silica gel matrix. Other materials, such as polymer binders, are currently being investigated to create a porous expanded structure for pure magnesium sulphate.

4. Conclusion

A high performance long-term heat storage system is needed to meet the huge heat demand of the building sector and the seasonal variations in the availability of the solar resource. Thermochemical

energy storage proves to be a relevant solution for this purpose since it does not loose energy with time and can provide the high energy densities necessary for compact storage.

Modelling and experimental studies have been performed on the hydration/dehydration cycle of magnesium sulphate composites to design an innovative system of seasonal storage of thermal energy.

A reference solar combisystem, without any long-term heat storage, has been simulated. The size of the required heat storage material tank has been calculated for a 191 m2 single-family house, with a space heating demand of 37.2 kWh. m-2 (53.0 kWh. m"2 including domestic hot water), under the Parisian climate. The results predict a volume of 0.2 to 0.9 m3 of magnesium sulphate (i. e. 1.0 to 4.8 m3 per m2 of living space) to reach a solar fraction ranging from 50 to 57 %. For the same house located in Marseille, with a space heating demand of 15.4 kWh. m-2 (31.1 kWh. m-2 including domestic hot water), a 100% solar fraction is achievable with a volume of 0.7 m3 of magnesium sulphate.

The limited performances of pure magnesium sulphate powder in real conditions have led to the investigation of porous composite materials, such as zeolite/salt and silica gel/salt composites. Thus, for example, the design of a composite material with 50% of the theoretical energy density of the dense salt would be enough to reach an acceptable storage volume of 2 m3. Temperature lifts around 30°C and maximal power of 28 mW. g-1 have been obtained during the hydration of the zeolite/MgSO4 composite ; lower values were obtained with pure zeolite, silica, or silica/MgSO4 composite tested under the same conditions. Further experiments are in progress to test the stability of these materials after several cycles and to find other appropriate composites to improve the performances. Summer regeneration of the material along with complete modelling of the chemical heat store will also be investigated.


[1] « Factor 4 : Doubling wealth — halving resource use, A report to the Club of Rome », Earthscan Publications Ltd., London, 1997.

[2] TRNSYS : http://sel. me. wisc. edu/trnsys/

[3] K. Visscher and J. B.J. Veldhuis , “Comparison of candidate materials for seasonal storage of solar heat through dynamic simulation of building and renewable energy systems”, paper presented at the Buildings Simulations 2005, in Montreal, Canada, August 15-18, 2005.

[4] J. Van Berkel, “Storage of solar energy in chemical reactions”, Thermal energy storage for solar and low energy buildings, IEA SHC Task 32, JC Hadorn (Edt), ISBN 84-8409-877-X, 2005.

Economical Possibilities

High Storage Capacity

First pilot projects [3], [5] in Germany reached storage capacities of about 130 kWh/m3. That means such a TES can store about 2-3 times more thermal energy than hot water storages. If higher temperature around 200 °C would be available, Zeolite storages could reach capacities of about 250 kWh/m3. However it has to be stated that a sorption TES consist of more components that the storage container itself, e. g. heat exchangers. In closed systems the desorbed water has to be kept inside the system, which reduces the storage capacity as well [2], [3]. Open systems need in most cases a humidifier for discharging. These additional components are usually not taken into account, which makes it difficult to compare different TES technologies.

Heating and Air-Conditioning

Open and closed sorption storages are able to provide thermal energy for heating purposes as well as for air conditioning of buildings. For this application cold can be delivered from the evaporator of a closed system. Using an open system the air dried by adsorption will be humidified, which leads to low temperatures (“desiccant cooling”).

The achievable Coefficients of Performance COPth are usually between 0.3 and 0.8. The open Zeolite system in Munich [5] reached a value COPth = 0.87 at a charging temperature of 80 °C in the experiments. The storage capacity was in the range of 100 kWh/m3.

In the case of liquid absorption systems only air conditioning is possible. Such an open storage system for solar air conditioning was installed in Singapore by the company L-DCS [8]. In the current demonstration project L-DCS Technology supplied a liquid desiccant air de-humidification system (11,000 m3/h) for a factory unit in Singapore, owned by JTC Corporation. A 550 m2 flat plate solar collector array drives the desiccant regeneration and 12 m3 desiccant energy storage covers the difference between the energy need for air conditioning by absorption and the solar energy supply for regeneration.

3. Economical Limits

The invest costs of sorption storage systems are quiet high. Therefore the number of operating hours, hence storage cycles, and the amount of thermal energy provided from the storage per time has to be high as well. This makes the operation as a seasonal storage system, used for 1 cycle per year, from the economical point of view impossible.

However a higher number of storage cycles by applying the system for heating and cooling purposes can lead to economical advantages. Furthermore the higher prices for air conditioning compared to plain heating could shorten the pay back time for such installations. The Zeolite storage in Munich reached a 50 % reduction of the payback time from around 14 years (for heating only) to about 7 years (for heating in winter and air conditioning in summer) [5].

4. Conclusion

Sorption storages are due to their thermodynamic possibilities — high storage capacity and variable temperatures — very interesting for solar thermal applications. How ever for economical reasons seasonal storage of solar heat by such systems is not an option. Most of the installed demonstration plants are looking for many storage cycles per year. This can be achieved by using the system for heating and cooling or by using smaller sorption systems in heat pump applications


With respect to solar applications sorption systems for solar air conditioning are most suitable, because for a high solar fraction the integrated storage effect is crucial.

The paper concludes with the remark that even the high storage capacities and the possibility of providing heat and cold of sorption storages does not solve all solar thermal storage problems. It is still necessary to find an appropriate application and to carefully check the relevant boundary conditions.

5. Literature

[1] R. Sizmann, Speicherung thermischer Energie — Eine Ubersicht, BMFT Statusseminar “Thermische Energiespeicherung” Stuttgart, 1989.

[2] D. Jaehnig, Thermo-Chemical Storage for Solar Space Heating in a Single-Family House, Proceedings of the International Conference on Thermal Energy Storage, Ecostock 2006, Stockton, New Jersey, USA, May 31 — July 2 2006.

[3] Thomas Nunez, A Small Capacity Adsorption System in a Heating and Cooling Application: The German Field-Test in the MODESTORE Project, International Conference Solar Air-Conditioning, Kloster Banz, Bad Staffelstein, Germany, October 6th/7th, 2005

[4] ZeoTech GmbH, Internet: http://www. zeo-tech. de/

[5] A. Hauer, Thermal Energy Storage with Zeolite for Heating and Cooling, Proceedings of the 7th International Sorption Heat Pump Conference ISHPC ’02, Shanghai, China, 24.-27. September 2002.

[6] H. Kerskes, K. Sommer, H. Muller-Steinhagen, An Effective Application of an Open Adsorption, Process for Solar Thermal Heat Storage, Proceedings of the EuroSun 2006, Glasgow, UK, June 27-30, 2006. [21] [22]

Design Method

Fig. 4 and 6 present effectiveness curves in which the capacity of the TSU is constant. These figures can be readily used as design curves which characterise the performance of the TSU subject to a single parameter, phase change fraction. In any design, using Eqn. (2), the minimum effectiveness can be determined which will define the outlet temperature from the TSU which will remain within specifications. For a given set of parameters this effectiveness is now directly related to the capacity of the TSU.

Consequently the proportion of phase change can be found which will ensure that the outlet temperature during the charging/discharging process will meet design requirements. This proportion defines the redundant PCM which occurs due to the 2nd law losses described in [3]. Consequently, the size of the TSU can be determined for a given set of parameters which will reliably meet thermal performance criteria. For example, in the solar heating application the minimum outlet temperature during the freezing process is 30 oC. This equates to a minimum effectiveness of 0.43. If the TSU is defined by H=50 mm and a flow rate of 0.35 kg/s, then, referring to Fig. 6, the useful proportion of the TSU is 67%. Therefore, to achieve the required, capacity the TSU should have 50% more PCM.

3. Conclusions

Using the knowledge of the solid/liquid phase change profile, it is possible to develop a one dimensional equation of the effectiveness of a TSU for design purposes, applying the s-NTU approach. The effectiveness was defined in terms of the phase change fraction and is able to reflect two dimensional and one dimensional phase change within a PCM slab. The use of the phase change fraction enables the characterisation of the TSU into a single effectiveness chart. From a design perspective this presents a useful method for determining the size of the TSU, which is the principle unknown variable in the design of the thermal storage unit. This analysis process also provides a method for optimisation of a design by enabling direct comparison of the impact of different parameters.


[1] E. Halawa, F. Bruno and W. Saman, Energy Conversion and Management, 46 (2005) 2592-2604.

[2] D. Morrison and S. Abdel — Khalik, Solar Energy, 20 (1978) 57-67.

[3] H. El-Dessouky and F. Al-Juwayhel, Energy Conversion and Management, 38 (1997) 601-617.

[4] K. Ismail and M. Goncalves, Energy Conversion and Management, 40 (1999) 115-138.

[5] E. Halawa (2006) Thermal Performance Analysis of a Roof Integrated Solar Heating System Incorporating Phase Change Thermal Storage, PhD Thesis, University of South Australia.

[6] A. Sari and K. Kaygusuz, Energy Conversion and Management, 43 (2002) 863-876.

[7] J. P. Holman (1992) Heat Transfer, 7th edn, Mc Graw Hill, London.


A, m2

Heat transfer area


Reynolds Number

Cp, kJ/kgK

Specific heat of fluid


Total thermal resistance

h, W/m2K

Convection heat transfer coefficient T1, oC

Inlet fluid temperature

H, m

Half slab thickness




Outlet fluid temperature

k, W/mK

Thermal conductivity

T, oC

PCM melting point

L, m

Length of slab

U, W/m2K

Overall heat transfer coefficient

m, kg/s

Fluid mass flow rate

W, m

PCM slab width


Number of transfer units

x, m

Direction in flow path


Nusselt Number

y, m

Direction perpendicular to flow path


Prandtl Number


Heat exchange effectiveness


Phase change fraction

Materials for thermochemical storage: characterization of magnesium sulfate

V. M. van Essen1*, H. A. Zondag1, R. Schuitema1, W. G.J. van Helden1 and C. C.M. Rindt2

1Energy research Centre of the Netherlands ECN, P. O. Box 1, 1755 ZG Petten, The Netherlands
2Department of Mechanical Engineering, Eindhoven University of Technology (TU/e), 5600 MB Eindhoven,

The Netherlands

* Corresponding Author, v. vanessen@ecn. nl

Magnesium sulfate hepta hydrate (MgSO4.7H2O) was studied as possible thermochemical material for seasonal storage of solar heat. Both hydration and dehydration were investigated and it was found that the material was able to take up and release almost 10 times more energy than water of the same volume. The amount of water taken up and energy released by the material turned out to be strongly dependent on the water vapor pressure and temperature. Unfortunately, the material was not able to release all the stored heat under practical conditions. Despite this problem, valuable information on the dehydration and hydration behavior of MgSO4.7H2O was acquired and the characterization procedure will also be used for future characterization of other salt hydrates for thermochemical materials.

Keywords: compact thermochemical seasonal heat storage, characterization, magnesium sulfate, salt hydrate

1. Introduction

Households in the Netherlands use about 15% of the total energy consumption for space heating and domestic hot water. The energy consumption in the built environment can be reduced by energy saving measures (improved insulation, heat recovery, etc.). A substantial part of the remaining energy demand can be fulfilled by using renewable energy sources such as solar energy.

The heat demand in summer can be completely fulfilled using solar heat, but in winter the heat demand exceeds the solar supply. To accommodate this difference in time between the solar energy surplus in summer and the energy demand in winter, a seasonal thermal storage is needed.

Traditionally water is used for storing solar heat (e. g. solar boiler) for short time periods, however, long-term heat storage will require a large water tank (>50 m3) that is often too large to be placed inside a building. As an alternative, it is possible to store energy by means of chemical processes, making use of the reversible reactions: C + heat «• A + B

During summer, the thermochemical material (TCM) dissociates under influence of solar heat into components A and B, which are stored separately. In the discharging mode, the two components (A and B) react to form the original TCM while releasing the stored solar heat. No reactions occur as long as the two components are stored separately, which means that thermo chemical storage can be used for loss-free storage and transportation of heat.

Interesting TCMs are cheap, non-toxic, non-corrosive, have sufficient energy storage density, and have reaction temperatures in the proper range. These requirements are fulfilled by a number of salt hydrates. A previous theoretical study [1] identified magnesium sulfate hepta hydrate (MgSO4.7H2O) as a promising material for long-term heat storage, by means of the following reaction:

MgSO4.7H2O(s) + heat «• MgSO^s) + 7H2O (g)

In this work, measurements on magnesium sulfate are presented that give information on the suitability of magnesium sulfate as thermochemical material for seasonal heat storage. Additionally, information is given on parameters important for the design of thermochemical seasonal heat storage using magnesium sulfate.

Temperature and pressure development in the fixed bed — temperature in the water tank

In Fig. 4 the normalized temperature Tn=T(t)/Tmax development in function of time t in a fixed bed of spherical zeolite 13X particles of average diameter dp=1.51mm is shown. The direct correlation of T and p in the fixed bed sorption process was shown in Fig. 3. An increased time delay At can be seen in which two succeeding temperature sensors in the equally spaced array of distance l=5cm are measuring the same temperature values. For example the time delay At=[t(T8)-t(T7)]=250s or At=[t(T10)-t(T9)]=500s can be measured for the normalized temperature of Tn=0.5 at the sensor T7 (z=33cm) and T8 (z=38cm) or T9 (z=43cm) and T10 (z=48 cm). These time delays are strongly dependent on the average particle diameter dp and the vapour pressure p(T) at the entrance to the sorption material fixed bed. For comparison in Fig. 5 the data for a pressure of p=29.7mbar i. e. the temperature T=25°C in the sorbate water tank, at the entrance to the same fixed bed are shown and a reduction of the time delay for Tn=0.5 to At=[t(T8)-t(T7)]=60s and At=[t(T10)-t(T9)]=100s can be seen.

time t [s]

Fig. 4: Normalized Temperature Tn in function of time t and position z in a fixed bed of spherical zeolite 13X particles of average diameter dSp=1.51mm. The water vapour pressure at the entrance to the fixed bed was p=16.3mbar at the start of the measurement. The delay time At(z) of two succeeding temperature sensors measuring the same temperature depends of the position z in the fixed bed.


Fig. 5: Normalized Temperature Tn in function of time t and position z in a fixed bed of spherical zeolite 13X particles of average diameter dSp=1.51mm. The water vapour pressure at the entrance to the fixed bed was p=29.7mbar at the start of the measurement. The delay time At(z) of two succeeding temperature sensors measuring the same temperature depends of the position z in the fixed bed.

Подпись: Fig. 5: Normalized Temperature Tn in function of time t and position z in a fixed bed of spherical zeolite 13X particles of average diameter dSp=1.51mm. The water vapour pressure at the entrance to the fixed bed was p=29.7mbar at the start of the measurement. The delay time At(z) of two succeeding temperature sensors measuring the same temperature depends of the position z in the fixed bed.

In Fig. 6 the time dependent development of the water temperature in the evaporator heat exchanger inlet T(HX in) and outlet T(HX out) are shown for a zeolite 13X fixed bed of dp=1.51mm a) and dp=2.63mm b) and a water inlet temperature of T(HX in)=25°C. After a time of t=200s to t=250s the minimum T(HX out) is reached and the temperature difference AT= T(HX in)-T(HX out)=3.5°C and AT=5.0°C are reached for the two particle size distributions. A linear temperature increase T(HX out) in the sorbate water tank is observed for adsorption process times longer than about t=300s.






temperature T(HX out) temperature T(HX in)


AT= 5.0 °C




……………… «***

0 50 100 150 200 250 300 350 400 450 500 550 600

time t [s]


Подпись: 0 Подпись: 300
Подпись: тппптплпппппп □ □ о □ с О temperature T(HX out) temperature T(HX in) о AT= 5.0 °C «о**30 о o< «*** 0 50 100 150 200 250 300 350 400 450 500 550 600 time t [s]

In Fig. 7 the pressure p measurement in function of the position z and the time t is shown for the pressure sensors installed at the first z=28cm of the fixed bed. A linear dependence of the water vapour pressure p from the position z can be assigned. While the process time t is increasing the pressure gradient dp/dz in the fixed bed is decreasing. This effect is illustrated with the two lines a) and b). Two effects are contributing to this behaviour. The adsorption which in this zeolite 13X is governed through the Langmuir isotherm [13] is dominating as long the equilibrium concentration c(p, T) is not reached.

a) b)

Fig.6: water temperature T as a function of time t at the outlet T(HX out) and the inlet T(HX in) of the evaporator heat exchanger at the start temperature T(HX in)=25°C for a fixed bed with dp=1.51mm a) and a fixed bed with dp=2.63mm b). After a time of approximately t=200s the minimum of the outlet temperature T(HX out) in is reached and from there on a linear increase of the temperature T(HX out) can be observed.

Fig. 7: Pressure p development in function of the
position z in the fixed bed of spherical zeolite 13X
particles of average diameter dSp=1.51mm and water
vapour pressure p(T)=16.3mbar at the entrance to the
fixed bed. The curves with the parameter time t are
showing a linear dependence of the pressure p(z). A
transition from a higher pressure gradient a) to a lower
one can be seen b).

8 18 23

position z in the fixed bed [cm]

Подпись:After the particles have reached the equilibrium concentration the water vapour flowing through

this region of the fixed bed undergoes the typical pressure drop a gas is subject to [14]. Calis [15] showed this in simulations for a validation of experimental results in a simplified model. The normalized temperature Tn in function of the position z is shown in Fig. 8 for a water vapour pressure of p(T)=16.3mbar at the entrance to the fixed bed and the average particle diameter dp=1.51mm and the parameter time t. The time delay At to measure the same normalized temperature Tn in two sensors at different positions z is increased for example from At=t(Tn, z=23cm)-t(Tn, z=13cm)=320s to At=t(Tn, z=43cm)-t(Tn, z=33cm)=660s. Thus the speed uT of the fixed bed traversing moving temperature front is decreased by the increasing length z of the fixed bed. From curve 1 and 2 in Fig. 8 the mass transfer zone length LMtZ can be measured to LMTZ=0.18m.

For the different inlet temperature levels of T(HX in)=15°C to 25°C and the average particle diameter dp=1.51mm and dp=2.63mm the temperature T(HX out) decrease reaches the minimum after

10 15 20 25 30 35 40

position z in the fixed bed [cm]

Подпись:approximately t=200s to t=300s followed by a linear increase. The temperature difference AT development as a function of time indicates that the (cooling) power in the water tank reaches a maximum in the above mentioned time t.

Fig. 8: Normalized temperature Tn development as a
function of the position z in the fixed bed of spherical
zeolite 13X particles of average diameter dSp=1.51mm
and water vapour pressure p(T)=16.3mbar at the
entrance to the fixed bed. The curves with the parameter
time t are showing a moving temperature front through
the fixed bed and the speed uT of the front is depending
of the position z (see for example At=320s and At=660s).

To reach a high power of the closed sorption system an optimum of the fixed bed geometry related to the MTZ length can be determined for the sorbent — sorbate combination of a granular zeolite 13X sorbent material and a vapour phase sorbate. But this geometry limits the total energy output Qth of the system. In addition a power density of Pth/Afb =33kW/m2 was determined where the area A represents the cross section of the fixed bed. In the closed adsorption system with spherical zeolite 13X particles the length of the MTZ is LMTZ(dp=1.51mm)=0.18m and LMTZ(dp=2.63mm)=0.38m, respectively. And, in a closed sorption system with zeolite 13X — water as the sorbent — sorbate combination the cycle time of discharging and charging is in the range of 5 Min to 8 Min [5]. As a conclusion to the findings the closed fixed bed
storage system has a limited energy and power output determined by the sorbent — sorbate material combination and the granular structure of the sorbent i. e. the diameter dp of the particles.

Table 2. Summary of the experimental results.

Temperature T(HX in) [°C] Pressure p(T) (sorbate tank) [mbar]







Particle diameter dp (av.) [mm]







hydraulic diameter dh [mm]







Time At to T10(t)>T10(t=0) [s]







Average speed uT [mm/s]







Length of MTZ LMTZ [m]







Cooling power Pth/A [kW/m2]









(1 — A





Подпись: B
Подпись: Ap P Подпись: (3)

While the thermal power Pth of an adsorption system depends of the vapour mass m adsorbed in time multiplied by the heat of adsorption AhA(T, c), the vapour flow through a granular fixed bed undergoes a pressure drop Ap depending on the inverse of the hydraulic diameter dh. Ergun [14] proposed an equation for the pressure drop in a fixed bed. In equation (3) the normalized pressure drop Ap/p depending on the vapour flow velocity u in a fixed bed of length L depending of the particle Reynolds number Rep and pressure drop factors A and B is given [11]:

According to equation (3) the pressure drop is a linear function of the length L. Because of the direct correlation of the vapour pressure p to the mass m in a given volume V through the equation of the ideal gas a linear power drop in the sorption system will occur when vapour flows through a section of saturated particles in the fixed bed.

The energy output Qth of the closed sorption system is the integral of m (sorbent) * AhA(T, c) over the cycle time t, sorbate content c and temperature T. The diameter dp of the particle also determines the length of MTZ and the total reaction (macroscopic) surface on which the reaction (adsorption) takes place. The total adsorption surface AMTZ in the MTZ is the product of the number n of particles and their individual surface Ap=n*dp2. The power Pth of the sorption system is a function of the reaction surface Pth(AMTZ = LAp). So, by measuring the temperature T and pressure p in function of time t and position z in the fixed bed the dynamics of a closed sorption system can be better understood.

4. Conclusion

The length of the mass transfer zone LMTZ and its moving speed uT through the fixed bed depends of the pressure p(T) at the entrance to the fixed bed and the average particle diameter dp which determines the hydraulic diameter dh. The higher the pressure p(T) — the driving force — and the larger the hydraulic diameter the longer LMTZ and the higher uT. The sorption behaviour of the fixed bed can be explained by a quasi continuum particle model.

In a closed solid sorbent — vapour sorbate sorption system no pump or fan is transporting the vapour into the fixed bed. The external temperature level T(HX in) to the sorbate tank determines the vapour pressure p(T) in the tank and the average particle diameter dp determines the hydraulic diameter dh of the fixed bed. Therefore, beside the heat transfer from the solid sorbent to a heat exchanger, after the two components sorbent — sorbate and the temperature level i. e. the pressure level is determined, the thermal power Pth and the energy Qth output of a closed solid sorption system is limited through the fixed bed hydraulic diameter dh. And because the single particle is saturated in a few minutes the size
of a solid sorption heat storage system is limited by the sorbent fixed bed hydraulic diameter dh i. e the particle diameter dp.

The actual interpretation of the experimental results in the closed sorption system take also account of the findings presented in literature about open sorption or catalytic systems in which for example a gas is the heat and mass transporting fluid. And the setup of the presented sorption system is in principle a heat pump and so the results can be used for the development of hest pumps. The measurements are showing that a solid sorbent — gas/liquid sorption system is more suitable for heat pump or cooling machine applications rather than thermal storage systems. But further experiments for a better understanding of the dynamic behaviour of the closed sorption system fixed bed will be needed.





length of the fixed bed, m




position in the fixed bed, m




particle diameter, mm


maximum (Tmax)


hydraulic diameter, mm


mass transfer zone


fixed bed diameter, mm


normalized (Tn=T(t)/Tmax)


speed (dz/dt) of the temperature front in the



fixed bed, mm/s


temperature (uT)


volume, m3




particle shape factor (sphere f=1)




fixed bed area Afb=n*D2/4, m2


particle surface Ap=n*dp2, m2

Greek letters


adsorption surface in MTZ AMTZ=n*n*dp2,







pressure, mbar


difference (heat of adsorption)


temperature, °C


concentration, kg/kg


mass, kg


water vapour mass flow, kg/s


water vapour velocity, m/s


power, thermal, kW


energy, thermal, kJ


heat of adsorption, kJ/kg


Reynold number

A, B

coefficients in Eq. (3)


time, s


Financial support of the Swiss Federal Office of Energy is gratefully acknowledged. The work was done under the IEA Task 32 Program and so special acknowledgment goes to the members of the IEA Task 32 group for very fruit full discussions. The author also would like to acknowledge F. Flueckiger who designed the data acquisition concept and W. Camenisch who did the welding of the laboratory vacuum equipment.


[1] Mittelbach, W. et al (2000), Solid sorption thermal energy storage for solar heating systems. TERRASTOCK 2000, Proceedings of the 8th International Conference on Thermal Energy Storage, Stuttgart Germany, Aug. 2000, 415-420.

[2] Purkarthofer, G., Sorptionsspeicher — Langzeitspeicherung von Warme mit hohen Energiedichten. SOLAR 2002, 6. Internationales Symposium fur Sonnenenergienutzung, Oktober 2002, 85-90.

[3] Stach, J. Mugele, J. Jaenchen, E. Weiler, Influence of the cycle temperature on the thermocemical heat storage densities in the systems water/microporous and water/mesoporous adsorbents. Adsorption 11: 393-404, 2005.

[4] Jaenchen J., D. Ackermann, H. Stach, W. Broesicke, Studies of water adsorption on zeolite and modified mesoporous materials for seasonal storage of solar heat. Solar Energy 76 (2004) 339-344.

[5] Gantenbein P., Adsorption speed and mass transfer zone analysis of water vapour on the solid sorbent materials zeolite and Silicagel with the focus on the heat exchanger design. ECOSTOCK 2006.

[6] Saha, B. B.; Boelman, E. C., and Kashiwagi, T. (1995), Computer simulation of a silica gel-water adsoption refrigeretion cycle — the influence of operating conditions on cooling output and COP. Technical and symposium papers. Annual meeting of the American Society of Heating, Refrigeration and Air-Conditioning Engineers, Inc. (ASHRAE Transactions); 24 June 1995 -28 June 1995. San Diego, CA. Atlanta, GA: American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.; Vol. 101, Part 2, pp. 348-357.

[7] Boelman E. C., Saha, B. B.; Kashiwagi, T. (1995), Experimental investigation of a silica gel — water adsorption refrigeration cycle — the influence of operating conditions on cooling output and COP. ASHRAE Transactions, Vol. 101, part 2, 1995, p. 358.

[8] Boelman E. C., Saha, B. B.; Kashiwagi, T. (1997), Parametric study of a silica gel — water adsorption refrigeration cycle — the influence of thermal capacitance and heat exchanger UA-values on cooling capacity, power density, and COP. ASHRAE Transactions, Vol. 103, part 1, 1997, p. 139.

[9] Zeochem / Uetikon, private communication.

[10] VDI-Waermeatlas (1997), Recherchieren — Berechnen — Konstruieren. Herausgeber: Verein Deutscher Ingenieure, Springer Verlag Berlin Heidelberg.

[11] Keule F., Das Trocknen, Grundlagen der chemischen Technik, Verlag Sauerlander & Co. Aarau und Frankfurt am Main.

[12] Freund H., Zeiser T., Huber F., Klemm E., Brenner G., Durst F., Emig G., Numerical simulations of single phase reacting flows in randomly packed fixed-bed reactors and experimental validation. Chemical Engineering Science 58 (2003) 903 — 910.

[13] http://en. wikipedia. org/wiki/Adsorption & references therein.

[14] Ergun S., Fluid flow through packed columns, Chemical Engineering Progress Vol. 48 No. 2, 1952, 89­94.

[15] Calis H. P. A., Nijenhuis J., Paikert B. C., Dautzenberg F. M., Van den Bleek C. M., CFD modelling and experimental validation of pressure drop and flow profile in a novel structured catalytic reactor packing. Chemical Engineering Science 56 (2001) 1713 — 1720.