University of Warwick institutional repository: http://go.warwick.ac.

uk/wrap

This paper is made available online in accordance with

publisher policies. Please scroll down to view the document
itself. Please refer to the repository record for this item and our
policy information available from the repository home page for
further information.
To see the final version of this paper please visit the publisher’s website.
Access to the published version may require a subscription.
Author(s): Y. Zhong, R.E. Critoph, R.N. Thorpe, Z. Tamainot-Telto
Article Title: Dynamics of BaCl2–NH3 adsorption pair
Year of publication: 2009
Link to published article:
http://dx.doi.org/10.1016/j.applthermaleng.2008.06.015
Publisher statement: “NOTICE: this is the author’s version of a work
that was accepted for publication in Applied Thermal Engineering.
Changes resulting from the publishing process, such as peer review,
editing, corrections, structural formatting, and other quality control
mechanisms may not be reflected in this document. Changes may
have been made to this work since it was submitted for publication. A
definitive version was subsequently published in Applied Thermal
Engineering, [VOL:29, ISSUE:5-6, April 2009, DOI:
10.1016/j.applthermaleng.2008.06.015”
 Dynamics of BaCl2 – NH3 Adsorption Pair

Y. Zhong *, R.E. Critoph, R.N. Thorpe, Z. Tamainot-Telto

School of Engineering, University of Warwick, CV4 7AL, UK

Abstract

To obtain accurate kinetics of adsorption pairs, dynamic tests of a composite

adsorbent material (BaCl2 impregnated into a vermiculite matrix) and ammonia were

performed on small samples under isothermal conditions and on a larger quantity in a

laboratory scale adsorption system. The experimental results show that the size of the

pressure swing plays an important role in the dynamics of adsorption pairs. This

driving pressure difference affects the mass transfer of NH3 through the pores of

adsorbent and therefore the performance of the complete adsorption system. A

modified LDF (Linear Driving Force) model was used to fit the experimental results.

It was shown that the model can predict the dynamics of both adsorption and

desorption fairly well and can be used for the modelling of the adsorption system and

adsorption cycles.

Keywords: Dynamics, Sorption rate, Modified LDF model

*
Corresponding author. Tel.: +44-2476-523137; Fax: +44-2476-418922
Email address [email protected] (Yong Zhong)

1
Introduction

Chemisorption could be useful in adsorption systems due to the large concentration

change compared with physisorption. The most commonly used adsorption pairs for

chemisorption are metallic salts and ammonia. However, the salt swelling, due to its

reaction with ammonia, can damage the matrix structure. It was found that composites

of metallic salts impregnated into various porous host matrices can accommodate the

swelling and abate the damage. An inorganic salt, expanded vermiculite , is one of the

promising host matrices. It allows insertion of a large amount of salt inside the pores,

reaching an ammonia uptake of 68.5 wt. % (for a composite “63.5 wt %

BaCl2/vermiculite”) [1]. And also, it can efficiently accommodate the salt swelling

due to its large pore volume.

As we know, for a real adsorption chiller, accurate isotherms, isosteric heat of

adsorption and the kinetic characteristics of adsorption and desorption are the key

factors to determine the performance of the system [2]. However, most of previous

studies on adsorption heat pumps have been performed using steady-state concepts,

while a real adsorption system can hardly reach the equilibrium state because of the

short cycle time.

Dawoud and Aristov carried out the kinetic study of water sorption on loose grains of

composite sorbent CaCl2 confined to meso-porous silica for various grain sizes

2
(between 0.34 and 3.2 mm) and various salt contents (12.6-33.7 wt.%) [3].

Experimental measurements were performed in a constant pressure unit based on a

microbalance under isothermal external conditions [4]. The results showed evidence

of a remarkable enhancement of the sorption rate and apparent diffusion constant with

the decrease in the particle size and salt content. The kinetic characteristic sorption

times were used to evaluate the specific power generated inside adsorbent grains

during the water sorption [4]. Aidoun and Ternan performed an experimental study on

the effect of the kinetic characteristics of CoCl2 – NH3 working pairs on the

performance of the chemical heat pump [5]. The experimental results showed that the

rate of the synthesis reaction in a chemical heat pump is controlled by the rate of

diffusion of ammonia gas into the chemical salt [6].

The mathematical models often used to describe the dynamics of adsorption system

are LDF (Linear Driving Force) and FD (Fickian Diffusion) model [7 – 11].

The Linear Driving Force (LDF) model for gas adsorption kinetics is frequently and

successfully used for analysis of adsorption column dynamic data and for adsorptive

process designs because it is very simple. This model was first introduced by

Gluekauf [7, 8]. According to the LDF model, the rate of adsorption of adsorbate into

adsorbent particles is given by [9]:

d x(t )
K L x* x(t ) (1)
dt

where, x (t ) is the average adsorbate concentration in the adsorbent particle at time t,

3
x * is the final equilibrium adsorbate concentration in the adsorbent particle, KL is

called the effective LDF mass transfer coefficient and usually given by [7],

15D0
KL (2)
r02
where D0 is intra-particle diffusivity of sorbate and r0 is radius of particle or crystal

Another well known model to describe the kinetics of adsorption system is the

isothermal Fickian diffusion (FD) model [9]. Compared to the LDF model, the FD

model imposes formidable mathematical complication in adsorptive process design

because the differential equations of the FD model have to be integrated at the

adsorbent particle level. Also, the above integration process must be repeated over

many cycles of operation in order to establish the final cyclic-steady-state separation

performance of the overall process. As a result, the FD model will generally require

large computational times for process simulation under realistic conditions.

On the other hand, the LDF model is much simpler and most importantly it eliminates

the integration step at the particle level so it can significantly reduce the

computational times required for realistic process simulations. This model with a

lumped mass transfer coefficient is very frequently used for practical analysis of

dynamic data and for adsorptive process design because it is simple, analytical, and

physically consistent [9]. The LDF model may not fit the experimental data very well

because its assumptions are not valid in some conditions in which heat transfer and

chemcial reaction have great effect on the sorption process. It is also worth pointing

4
out that in LDF and FD model, the chemical reaction of adsorbate and adsorbent are

not considered. Some mathematical models regarding the dynamics of chemical

reactions have been proposed recently. But these models are very complicated and the

characteristic coefficients in these models are determined by the adsorption pairs used

[12 – 14]. In this paper, in order to simplify the dynamic model, the LDF model is

modify by considering the chemical kinetics to predict the dynamics.

This paper presents experimental study on dynamics of one composite material (58.7

wt.% BaCl2 impregnated inside vermiculite matrix) for a small sample under

isothermal condition and for a larger sample in a laboratory scale adsorption system.

For isothermal condition, with the small sample, the reactor was weighed

continuously using a magnetic suspension balance. The object was to obtain an

instantaneous direct measurement of the salt conversion from the weight measurement.

A modified LDF model was proposed to analyze the dynamics of the system. For the

labortary scale adsorption system, the temperatures and pressure of the system were

measured, and the modified LDF model was applied to describe the dynamics of the

adsorption and desorption process. A one dimension mathematical model was used to

calculate adsorbent temperature and the concentration change during experiments.

The calculated results were compared with the experimental values to verify the

accuracy of the modified LDF model.

Experimental and Results

5
Dynamic tests under isothermal condition

The simplest experimental method to determine the kinetics of sorption systems

involves the measurement of sorption curves, under isothermal conditions for a small

sample of the adsorbent subjected to a step change in the sorbate pressure. The

experimental apparatus consists of Rubotherm ISOSORP 2000 magnetic suspension

balance system, glycol water bath system (50 wt.% glycol to make sure that the

system can be cooled down to -30oC) and oil bath system，shown in Figure 1. In the

system, four thermocouples within the chamber were used to measure the adsorbate

temperature and the wall temperature. The system pressure was measured directly

with a calibrated Druck PDCR 920 transducer. The resolution of the pressure

transducer was 0.35bar/mV and the accuracy is 1.5%. The mass of the sample was

measured by Rubotherm balance system . The adsorbate concentration was calculated

by the mass change of the sample during the experiment. The resolution of the

equipment was 0.01 mg and the standard deviation of successive measurement was

less than 0.03 mg. The aluminium sample holder (Figure 2) acted as a thermal

ballast and its design ensured good thermal contact with the material under test. The

mass of the sample and sample holder were 0.2 and 11.2 g respectively. To make sure

that the adsorption system was under isothermal condition, the test cell was contained

within a jacket, through which heated oil was circulated in order to maintain the

temperatures of the vessel and sample constant. Thus the dynamics of the adsorption

6
was mainly controlled by mass transfer rather than heat transfer.

The sample was prepared by being held at 180oC for at least 10 hours at atmospheric

pressure and then kept for 2 hours under vacuum to make sure that all water vapor

was eliminated from the sample. The measuring cell was then connected to an

evaporator maintained at a temperature between -20 and 50oC until the system

reached an equilibrium state. At the beginning of the dynamic tests, the valve

separating the reactor and the reservoir was closed and the saturated temperature of

the reservoir was changed to a new value. After the adsorbent material and the

refrigerant in the reactor had attained thermal and chemical equilibrium, the valve

between the reactor and reservoir was opened. The pressure changed to the pressure

of the reservoir, causing a rapid rate of concentration change, because the adsorbent

bed was at conditions far from equilibrium. The values of temperature, pressure and

mass change during this process were recorded. After the kinetic curve had been

recorded, the evaporator was disconnected from the measuring cell and its

temperature was increased to fix the proper pressure for the next kinetic test.

Typical kinetic curves of ammonia sorption were shown in Figure 3. In Figure 3(a),

the test sample was at the temperature of 37oC and the pressure of the system dropped

from 5bar down to 2.15bar. It can be seen from Figure 3 that the concentration change

of the adsorbent was around 38% during a over a period of of 20 minutes. No more

ammonia was desorbed from the sample and the concentration remained at 2.5%,

7
which is considered as the contribution of vermiculite to the ammonia concentration.

Figure 3(b) showed the kinetic curve when the pressure of system dropped from 4.7

down to 4.5 bar at the same temperature. It took around 400 minutes to complete the

decomposition, which was around 20 times longer than that of large pressure

difference.

The results of the kinetic experiments on BaCl2 – NH3 showed that the size of the

pressure swing played an important role in the dynamics of adsorption pairs. This

driving pressure difference affected the mass transfer of NH3 through the pores of

adsorbent and therefore the performance of the complete adsorption system.

For chemisorption, the chemical reaction rate of adsorption pairs is as important as the

diffusion of refrigerant into the adsorbent pores. To better match the experimental data,

the LDF model was modified to predict the dynamics of BaCl2 and NH3 adsorption

pair, which was given by [15 - 17],

dx 1
(3)
dt C 1
p K x xeq
where C is a resistant coefficient due to chemical reaction between the adsorbate and

adsorbent, in Pa s and, K is the mass transfer diffusion coefficient in s-1. C and K can

be obtained by fitting the experimental data: for adsorption C=50 Pa s, K=-2500 s-1;

for desorption, C=30 Pa s,K=500 s-1. It is worth pointing out that the value of C and K

came from the data fitting of the experimental data.and that the form of equation 3 is

8
chosen as being empirically useful, without necessarily having established it as a true

model of all the physical processes involved.

xeq is the final equilibrium adsorbed phase concentration, p ptransition p , p is

the system pressure, ptransition is the onset pressure of the synthesis and

decomposition transitions and given by [15 - 18],

- 1000
ln ptransition 4.98 22.26 (4)
T
where, T is the temperature of reaction bed. The root mean square error

(corresponding to the Standard Estimated Error – SEE) is shown in Table 1

( p ptransition p ) . and It is defined as follow:

n
1 2
SEE= xi x cal (5)
n i 1

where n is number of experimental data, x is measured concentration, xcal is the

calculated concentration through the modified LDF model. The comparison of the

typical experimental data and mathematical model is shown in Figure 4 and Figure 5.

From Figure 4 and Figure 5 , we can conclude that the mathematical model described

in equation (3) can used to predict the dynamics of BaCl2 and NH3 under isothermal

conditions. However, during the experiments, the reactor was kept isothermal, which

is not consistent with real adsorption systems. The accuracy of this model is still

worth testing in conditions where both pressure and temperature change.

Dynamic test in a laboratory scale adsorption system

9
The experimental rig (as described in Figure 6) consists of a reactor connected with a

receiver and three water thermal baths, which were maintained at around 20oC

(receiver water bath), 100oC (desorption water bath) and 20oC (adsorption water bath).

Four K-type thermocouples (two on the generator, which were placed in the centre

and on the tube outer wall respectively; another one was placed on receiver outer wall,

the other one was placed in the hot water bath) and a pressure transducer which was

used to monitor the pressure within the generator were linked to a PC via an interface

(Data Shuttle DA-16-TC-AO). A program written with Workbench 2.3 was used to

monitor and record the temperatures and the pressure.

In order to measure the amount of liquid collected during the dynamic test, a glass

tubing receiver rated to 40 bar at 20oC (20mm OD and 12mm ID) was connected to

the generator. The schematic diagram of the receiver is shown in Figure 7. The mass

of ammonia collected in the receiver can be calculated by the following expression:

6
m 36 NH 3 10 h (6)

Where, m is the mass of ammonia collected in the receiver (g); NH 3 is the density of

ammonia liquid, h is the height of ammonia liquid level (mm).

Initially, the test generator was at ambient temperature (typically 20oC). After

typically 16 hours, the system reached equilibrium state. Then the reactor was

submerged suddenly into the boiling water bath (at about 100oC). The pressure and

temperatures were recorded for the desorption procedure. After around 1 hour almost

all of the ammonia in the reactor was desorbed and condensed in the receiver. The

10
reactor was moved suddenly from the hot water bath to the cold water bath (at around

20oC) and the adsorption dynamic test was started. During the dynamic tests, the

receiver was kept at around 20oC, controlled by the cold water bath.

A one-dimensional transient heat conduction generator model was used to simulate

the dynamics of the set-up. The thermal conductivity of the bed and the contact heat

transfer coefficient between the bed and the tube wall were set to be 0.2 W m-1 K-1

and 150 W m-2 K-1 [16, 19]. The explicit scheme of finite difference was used in the

modelling. To simplify the simulation, the pressure in a single control volume was

assumed to be uniform. Therefore, the bed pressure was uniform and the resistance of

mass diffusion through the sample pore in the reactor was neglected. The equations

used to monitor the transient sorption characteristics were heat and mass conservation

and the sorption dynamic equation.

The transient heat and mass transfer is given by the following equation [16, 19]:

T ks T x
s C ps C pa x g C pg r H (7)
t r r r t

where:

s is the sample density, kg/m3

Cps is the specific heat of the sample, J/(kg K)

Cpa is the specific heat of adsorbed ammonia, J/(kg K)

T is the sample temperature, K

t is the time, s

g is the density of free gaseous ammonia,(kg)

11
ks is the thermal conductivity of the sample, W/(m K)

A is the area of conduction for sample = 2 (r + r /2) z , m2

H is the heat of sorption which can obtained form the slope of isotherms [16, 17], J/kg

x is the ammonia concentration, kg NH3/ kg sample

The mass of adsorbed ammonia ma is a function of the ammonia concentration and is

given by:
ma
Ms (8)
x
where: Ms is the total mass of the sample, kg
dma dx
Then, Ms (9)
dt dt
Equation (9) is used to describe the concentration change of adsorption pair.

During the experiments, the system is closed and the total mass of ammonia mt

remains constant and is the sum of the mass of liquid ammonia in receiver, mr, the

free mass of ammonia gas, mg, and the adsorbed ammonia in the sample porosity, ma:

mt=ma+mr+mg (10)

The ammonia gas is assumed as perfect gas so that the free mass of ammonia gas can

be calculated from the ideal gas law by,

pv
mg (11)
RT
where, p is the pressure of the system, Pa; v is the void volume of the system, m3; R is

the gas constant, J/(K kg); T is the temperature of the ammonia gas.

The mass of liquid ammonia in receiver, mr can be calculated by equation (6).

Therefore, the rate of change in adsorbed ammonia in the sample porosity can be

12
presented as:

ma=mt-mr-mg (12)

The calculated mass of adsorbent ammonia is compared with the simulated results.

The simulation uses the measured tube wall temperature (boundary condition) and

ammonia pressure as input. The heat transfer within the bed is based on the

conduction through the bed and with the mass transfer. The initial temperature of the

generator is assumed uniform. The simulated centre temperature of the reactor and the

concentration of the sample are compared with the experimental results.

Figure 8 and 9 showed the comparison of the experimental data and simulation. For

adsorption, except the first one or two minutes, the model can predict the centre

temperature and the concentration fairly well. The maximum error is around 10%.

For the desorption dynamic test, the mathematical model can predict the centre

temperature well, while the model underestimated the concentration especially after

30 minutes. The probable reason is that the ammonia gas might have condensed in the

horizontal connecting tube or other void volume. Therefore the measured ammonia

level in the receiver would be lower than that expected. In other words, the calculated

concentration of the sample from experimental data will be greater than ideal

situation.

Conclusion

13
The kinetics of ammonia sorption under isothermal conditions and in a laboratory scale

system was studied at T = 20 to 100 C and p = 0.2 to 20 bar. The pressure difference

between the reactor and condenser plays an important role in the dynamics of

adsorption pairs. The pressure difference affects the mass transfer of NH3 through the

pores of adsorbent and therefore the performance of the complete adsorption system.

A modified LDF model is used to predict the dynamics of the adsorption system using

BaCl2 – NH3. Through the comparison of the experimental data and simulation, the

model has been shown to predict the dynamics of both adsorption and desorption

fairly well and so can be used for the modelling of adsorption system and adsorption

cycles.

14
References:

[1] V.E. Sharonov, J.V. Veselovskaya, Yu.I. Aristov, Ammonia sorption on

composites "CaCl2 in inorganic host matrix": isosteric chart and its

performance, Int. J. Low Carbon Thecn., vol.1(3), pp. 191-200, 2006

[2] C.H. Liaw, J.S.P. Wang, R.H. Greenkorn, and K.C. Chao, Kinetics of fixed

bed adsorption: a new solution, AIChE Journal, vol.25, p.376, 1979

[3] B. Dawoud, Yu.I. Aristov, Experimental study on the kinetics of water vapor

sorption on selective water sorbents, silica gel and alumina under typical

operating conditions of sorption heat pumps, International Journal of Heat and

Mass Transfer, vol.46, pp.273–281, 2003

[4] Yu.I. Aristov, I.S. Glaznev, A. Freni , G. Restuccia, Kinetics ofwater sorption

on SWS-1L (calcium chloride confined to mesoporous silica gel): Influence

of grain size and temperature, Chemical Engineering Science, vol.61, pp.1453

– 1458, 2006

[5] Z. Aidoun, M. Ternan, The synthesis reaction in a chemical heat pump reactor

filled with chloride salt impregnated carbon fibres: the NH3–CoCl2 system,

Applied Thermal Engineering, vol.22, pp.1943–1954, 2002

[6] Z. Aidoun, M. Ternan, The unsteady state overall heat transfer coefficient in a

chemical heat pump reactor: the NH3–CoCl2 system, Chemical Engineering

Science, vol.59, pp.4023 – 4031, 2004

[7] E. Gluekauf, Theory of chromatography, Part 10, Formula for diffusion into

15
 spheres and their application to chromatography, Transactions of the Faraday

Society, vol.51, pp. 1540–1551, 1955

[8] J.Kaerger, D.M.Ruthven, Diffusion in zeolites and other microporous solids,

J.Willy, N.Y., 1992.

[9] S. Sircar, J.R. Hufton, Why does the linear driving force model for adsorption

kinetics work?, Adsorption, vol.6, pp.137–147, 2000

[10] J. Karger, D.M. Ruthven, Diffusion in zeolites and other microporous solids,

John Wiley & Sons, INC., USA, 1991

[11] Chi Tien, Adsorption calculations and modeling, Butterworth-Heinemann,

London, 1994

[12] V. Goetz V, A. Marty, A model for reversible solid-gas reactions submitted to

temperature and pressure constraints: simulation of the rate of reaction in

solid-gas reactor used as chemical heat pump, Chemical Engineering Science,

47(17-18):4445-4454, 1992

[13] M. Lebrun, B. Spinner, Models of heat and mass transfers in solid-gas

reactors used as chemical heat pumps. Chemical Engineering Science,

45(7):1743-1753, 1990

[14] H.B. Lu, N. Mazet, B. Spinner, Modeling of gas-solid reaction – coupling of

heat and mass transfer with chemical reaction, Chemical Engineering Science,

51(15): 3829 – 3845, 1996

[15] V.E. Sharonov, J.V. Veselovskaya, Yu.I. Aristov, Y. Zhong, R.E. Critoph,

New composite adsorbent of ammonia “BaCl2 in vermiculate” for adsorptive

16
 cooling, CHP (2006), Proc. HPC06, Newcastle, 2006

[16] Y. Zhong, Studies on equilibrium and dynamic characteristics of new

adsorption pairs, Ph.D. Dissertation, University of Warwick, 2006

[17] R.E. Critoph, Y. Zhong, S.J. Metcalf, Dynamic modelling of solar thermal

cooling using the BaCl2 – ammonia pair, proc. World Renewable Energy

Congress, Florence, August 2006

[18] Y. Zhong, R.E. Critoph, R.N. Thorpe, Z. Tamainot-Telto, Yu.I. Aristov,

Isothermal sorption characteristics of the BaCl2 – NH3 pair in a vermiculite

host matrix, Applied Thermal Engineering, ATE-2206-628R1, Accepted, 2007

[19] Z. Tamainot-Telto, R.E. Critoph, Cost effective carbon-ammonia generators,

Report No ENK5-CT2002-00632-UW1, October 2003, School of

Engineering, University of Warwick

17
Nomenclature

A Area of conduction for sample, m2

C Resistant coefficient due to chemical reaction, Pa s

Cp Specific heat, J kg-1 K-1

D Intra-particle diffusivity, m2s-1H Heat of sorption, J kg-1

h Height of ammonia liquid level, m

-1
K Effective LDF mass transfer coefficient, s

Mass transfer diffusion coefficient, s-1

k Thermal conductivity , W m-1 K-1

m Mass, kg

p Pressure, Pa

R Gas constant, J kg-1 K-1

r Radius or Position of the sample cell from the centre, m

r Increment in radius, m

T Temperature, K

t Time, s

v Volume, m3

x Mass concentration. kg ammonia kg-1 adsorbent

z Total height, m

18
Subscripts

a ammonia

eq equilibrium

g gas

NH3 ammonia liquid

r receiver

s solid

transition Synthesis or Decomposition reaction

Greek letter

Density, kg/m3

Angle, rad

Increment in angle, rad

19
Figure captions

Table 1 Root mean square error

Figure 1 Schematic diagram of porosity test rig

Figure 2 Diagram of the sample holder

Figure 3 Concentration change against time at 37 oC

Figure 4 Comparison of desorption experimental data and simulation at T=44.8 °C

Figure 5 Comparison of adsorption experimental data and simulation at T=45.5 °C

Figure 6 Schematic diagram of dynamic test rig

Figure 7 Schematic diagram of the glass tubing ammonia receiver

Figure 8 Comparison of adsorption experimental results and mathematical model (the

pressure changed from 11.78 to 6.47 bar)

Figure 9 Comparison of desorption experimental results and mathematical model (the

pressure changed from 7.99 to 12.78 bar)

20
Temperature, oC p , bar SEE, %

Desorption
37.7 2.47 0.35
38.2 0.1 1.87
44.8 3.78 0.13
44.6 0.17 0.91
53.4 5.4 0.59
53.7 0.21 1.16
Adsorption
38.9 -5.88 0.17
37.9 -6.12 0.33
45.5 -6.35 0.29

Table 1. Standard Estimated Error (SEE)

21
 Magnetic Suspension
Valve
Balance
Rubotherm ISOSORP 2000
Reactor

Oil Bath

16 cm
Sample
System
Liquid NH3

Water Bath
Thermocouples
Glycol Water Bath System 10.5cm
12.75cm

Figure 1 Schematic diagram of porosity test rig

22
 Moving
part
35mm

Φ22mm
cm

Fixed part
Φ25mm
cm

Figure 2 Photograph of the sample holder

23
 40

35

30
Concentration, %

25

20

15

10

0
0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000

Time, seconds
(a) Large pressure difference (pressure drop 2.75bar)

40

35

30

25
Concentration

20

15

10

0
-0.5 0 0.5 1 1.5 2 2.5 3 3.5
Time, seconds 4
x 10

(b) Small pressure difference (pressure drop 0.2bar)

Figure 3 Concentration change against time at 37oC

24
 Concentration, %

Experimental results
Simulation

Time, seconds

Figure 4 Comparison of desorption experimental data and simulation at T=44.8°C

25
 Simulation
Concentration, %

Experimental results

Time, seconds

Figure 5 Comparison of adsorption experimental data and simulation at T=45.5°C

26
 PC + DataShuttle

Pressure Transducer

Adsorption bed

Thermocouples
170mm

25 mm Receiver

Hot water bath (100oC) Cold water bath (20oC)

Figure 6 Schematic diagram of the test rig

27
 Evaporator/Condenser

200 mm

h
(Ammonia liquid
collected)

20 mm

Figure 7 Schematic diagram of the glass tubing ammonia receiver

28
 100

90

80
Temperature, CC

70
o
Temperature,

60

50 Simulation of adsorbent temperature

40
Temperature of
Measured adsorbent temperature
30 the wall

20
0 500 1000 1500 2000 2500 3000
Time,Second
Time, seconds

(a) Temperature vs. time

40
35
30
25
x, %

20
15
10
5
0
0 500 1000 1500 2000 2500 3000 3500
Time, seconds

Experimental Results Simulation

(b) Concentration vs. time

Figure 8 Comparison of adsorption experimental results and mathematical model (the

pressure changed from 11.78 to 6.47 bar)

29
 60

55

50 Simulation of adsorbent temperature

45
Measured adsorbent temperature
Temperature,CC
o

40
Temperature,

35

30

25

20

15
0 500 1000 1500 2000 2500 3000 3500 4000
Time,Seconds
Time, seconds

(a) Temperature vs. time

50
40
30
x, %

20
10
0
0 500 1000 1500 2000 2500 3000 3500 4000
Time, Seconds

Experimental data Simulation

(b) Concentration vs. time

Figure 9 Comparison of desorption experimental results and mathematical model (the

pressure changed from 7.99 to 12.78 bar)

30
31