261

Thermodynamic modelling of a single-effect

LiBr –H2O absorption refrigeration cycle
M A Mehrabian
and A E Shahbeik
Department of Mechanical Engineering, Shahid Bahonar University of Kerman, Kerman, Iran

The MS was received on 5 April 2004 and was accepted after revision for publication on 23 September 2004.

DOI: 10.1243/095440805X8656

Abstract: The objective of this paper is to develop a computer program for design and thermo-
dynamic analysis of a single effect absorption chiller using LiBr–H2O solution as working fluid. The
conditions of hot water entering and leaving the desorber, cooling water entering the absorber and
leaving the condenser, chilled water entering and leaving the evaporator, as well as the approach
temperatures in condenser, evaporator, desorber, and absorber, the effectiveness of solution heat
exchanger, the chiller refrigeration power, and the ambient temperature are used as input data.
The program then gives the thermodynamic properties of all state points, the design information
of all heat exchangers in the cycle and the overall cycle performance. The results deduced from the
computer program are used to study the effect of design parameters on cycle performance. For
example, increasing the evaporator and generator temperatures or decreasing the condenser
and desorber temperatures can improve the second-law efficiency of the cycle. It is also noticed
that the temperatures of hot water, cooling water, and chilled water, respectively, at the inlet
of the desorber, condenser, and evaporator have a great effect on cycle coefficient of performance.
The results of this program can be used either for sizing a new refrigeration cycle or rating an exist-
ing system. It can also be used for optimization purposes. The predictions of the present program
are compared with other simulating programs and qualitative agreement is achieved.

Keywords: thermodynamic modelling, absorption refrigeration system, second law efficiency,

coefficient of performance, lithium bromide/water solution

1 INTRODUCTION combinations, have been studied extensively from

both theoretical and experimental points of view.
Absorption refrigeration systems have been adopted Simulation programs have also been presented for
to produce cold since 1850. They are becoming more performance evaluation and calculations of design
attractive nowadays, especially because of their parameters. The accuracy of these models depends
environmental friendliness. There are, of course, strongly on the reliability of thermo-physical proper-
other advantages such as utilizing low-grade ener- ties of LiBr – H2O or NH3 – H2O solutions. Therefore,
gies like solar, geological, and waste heat discharged along with improvements in modelling techniques,
from various industrial processes. Absorption systems vapour –liquid equilibrium and calorimetric measure-
are still less efficient, bulkier, and more expensive ments of aqueous solutions have been carried out
than vapour compression systems. Absorption refri- covering the high temperature and concentration
geration cycles, including those using LiBr – H2O ranges to be incorporated in computer programs to
and NH3 – H2O solutions as refrigerant – absorbent design advanced absorption machines with optimal
and cost-efficient performance.
An experimental procedure conducted by Pichel


Corresponding author: Department of Mechanical Engineering, [1] on a 3517 kW capacity Li– Br absorption refriger-
Shahid Bahonar University of Kerman, PO Box 76175-133, ation machine showed that the machine can operate
Kerman, Iran. with hot water at 80 and 120 8C, and it was found that

E02604 # IMechE 2005 Proc. IMechE Vol. 219 Part E: J. Process Mechanical Engineering
262 M A Mehrabian and A E Shahbeik

the cooling water connected in parallel to both mass transfer correlations. A new model was intro-
absorber and condenser is more efficient than duced for representing the absorber. Simulta-
series connection. The coefficient of performance neous heat and mass transfer were considered in
was between 0.68 and 0.72. the absorber instead of heat transfer alone.
Eisa et al. [2] determined experimentally the flow The model described in Joudi and Lafta [8] was
ratio (the ratio of the mass flow rate of the solution applied to an actual commercial absorption refriger-
from the absorber to the mass flow rate of the ation plant manufactured by Mitsubishi-York [9],
refrigerant) and COP for all possible combinations model ES-2A4-MW working on LiBr –H2O and using
of operating temperatures of the evaporator, hot water as a heat source with a capacity of
condenser, absorber and desorber, and the concen- 211.1 kW refrigeration power.
tration in the absorber and the desorber up to the Heydari et al. [10] used entropy generation for-
crystallization limit. mulation to develop a thermal – hydraulic model for
Eisa and Holland [3] performed an experimental analysing the performance of a single effect absorp-
investigation to study the effect of changes in operat- tion chiller. They also presented a systematic
ing conditions on the performance of LiBr–H2O optimization technique based on second-law analy-
absorption cooler for optimization purposes. They sis of the absorption cycle.
found that the most significant parameter affecting Feuerecker et al. [11] developed a new method
COP is the desorber temperature. An increase in des- which was especially convenient to measure
orber temperature increases the COP. The flow ratio is vapour –liquid equilibrium and calorimetric proper-
also an important design and optimizing parameter. ties of concentrated aqueous LiBr solutions. They
Increasing flow ratio decreases the required desorber then deduced enthalpies and entropies in the con-
temperature at the expense of reduction in COP. centration range 40– 75% LiBr and temperatures up
Eisa et al. [4] conducted more experiments on to 190 8C. The data were presented as polynomial
the same system of Eisa and Holland [3] to study and diagrams including experimental accuracy.
the effect of operating temperatures of the absorber Mostafavi and Agnew [12] and [13] studied the
and condenser on performance. It was noticed effect of ambient temperature on component geo-
that as the temperature difference (Tc 2 Ta) is metry and performance of absorption chillers.
increased, the COP and the cooling capacity are The present study involves developing a com-
decreased. They also found that the COP is more puter program to evaluate single stage absorption
sensitive to the absorber temperature than to heat pump performance using water as the refri-
the condenser temperature. gerant. The program performs the cycle calculations
Waleed [5] developed a computer program to for single stage cycles based on the polynomial
design heat exchangers (desorber, condenser, and expressions developed to correlate experimental
evaporator) and to predict their performance for a vapour –liquid equilibrium and specific enthalpy –
7 kW capacity LiBr absorption machine under work- concentration data for LiBr – H2O solutions. This is
ing conditions different from the design conditions. the distinguished superiority of this program as com-
Mclinden and Klein [6] constructed a modular, pared with other simulations which suffer from lack
steady-state model for simulation of NH3 – H2O of accurate thermo-physical data.
absorption heat pumps. The model was based on
detailed mass and energy balances and heat and
mass transfer relationships for the components of
the cycle and was applied to a prototype unit and
2 MATHEMATICAL MODELLING
compared with experimental data.
Grossman and Michelson [7] developed a modular
The schematic diagram of a single effect LiBr – H2O
computer simulation program for absorption sys-
absorption chiller is shown in Fig. 1. The aim of the
tems, which makes it possible to simulate various
program is to calculate the UA values of the heat
cycle configurations. The program has been tested
exchangers and the thermodynamic properties of
on single and double stage absorption heat pumps
the working fluid at state points shown in Fig. 1.
and heat transformers with LiBr –H2O and NH3 –
The input data for the program are:
H2O solutions as the working fluid. The results have
been compared with experimental data from tests (a) temperatures of hot water entering and leaving
of a LiBr – H2O heat transformer and good agreement the desorber (T11 and T12);
has been achieved. (b) temperatures of cooling water entering the
Joudi and Lafta [8] dealt with a continuous absorp- absorber and leaving the condenser (T13 and
tion refrigeration LiBr – H2O system. They carried out T16);
a steady-state simulation based on mass and heat (c) temperatures of chilled water entering and
balance equations, as well as fluid flow, heat, and leaving the evaporator (T17 and T18);

Proc. IMechE Vol. 219 Part E: J. Process Mechanical Engineering E02604 # IMechE 2005
 Single-effect LiBr –H2O refrigeration cycle 263

The refrigerant entering the evaporator at point 9 and

leaving the evaporator at point 10 are within the sat-
uration mixture zone at pressures P9 and P10,
respectively. Neglecting the pressure loss between
points 9 and 10 results in

T9 ¼ T10 (5)

The saturated liquid leaving the condenser at point 8

enters the expansion valve and undergoes a throt-
tling process in which enthalpy remains constant

h9 ¼ h8 (6)

In order to evaluate h9, we need to analyse the

condenser. The temperature at point 8 is evaluated
from the input data
Fig. 1 Schematic diagram of LiBr –H2O absorption
chiller T8 ¼ DT816 þ T16 (7)

Assuming saturated liquid leaving the condenser,

(d) the approach temperatures in the condenser the pressure in the condenser, as well as the
(DT8 – 16), evaporator (DT18 – 10), desorber (DT12 – 4), desorber (the high pressure line of the cycle), and
and absorber (DT1 – 14); the enthalpy at point 8 are read from the steam
(e) the effectiveness of solution heat exchanger tables
(Q/Qmax);
(f) the chiller refrigeration power (Qeva);
P8 ¼ P7 ¼ P3 ¼ P4 ¼ P2 ¼ P5 ¼ Psat (T8 ) (8)
(g) the ambient temperature (T0).
h8 ¼ h f (T8 ) (9)
The calculations start from a state point in the cycle
Enthalpy at point 9 is now evaluated from equation
and proceed to the next point. In order to analyse the
(6). As a result of pressure reduction in the expansion
whole cycle, it is required to conduct a fixed point
valve, a saturated mixture is formed. The quality of
iteration for T14 and x6. The final values for these
the mixture can now be calculated
properties are evaluated based on physical reasoning
and compared with the initial guessed values. If the
h9  hf (P9 )
final value is different from the initial value, the cal- q9 ¼ (10)
culations are repeated using the final value as the hg (P9 )  hf (P9 )
new guess. A typical loop of calculations may start
from point 10, whose temperature is evaluated The mass flow rate of working fluid in evaporator and
from the input data condenser is now evaluated

T10 ¼ T18  DT1810 (1) Qeva

m9 ¼ m10 ¼ m8 ¼ m7 ¼ (11)
h10  h9
Assuming dry saturated steam leaving the evapor-
ator, the pressure in evaporator, as well as in absor- In order to find the conditions at point 7, the absor-
ber (the low pressure line of the cycle) and the ber and desorber should be analysed. The tempera-
enthalpy at point 10 are read from the steam tables ture at point 1 can be evaluated from

P9 ¼ P10 ¼ P1 ¼ P6 ¼ Psat (T10 ) (2) T1 ¼ T14 þ DT114 (12)

h10 ¼ hg (T10 ) (3)
but T14 is not known. As the first guess, it is assumed
The mass flow rate of chilled water is evaluated from T13 þ T16
the input data T14 ¼ (13)
2
Qeva T1 is now evaluated from equation (12). P1 is already
m17 ¼ (4)
cp (T17  T18 ) known from equation (2). The program now reads

E02604 # IMechE 2005 Proc. IMechE Vol. 219 Part E: J. Process Mechanical Engineering
264 M A Mehrabian and A E Shahbeik

the solution concentration at point 1 using the P5 is already known (equation (8)); x5 can be
Duhring diagram [11] calculated from

x1 ¼ xsat (P1 ,T1 ) (14) x5 ¼ xsat (P5 ,T5 ) (22)

It should be noted that in the Duhring equilibrium Enthalpy at point 5 is either calculated from
equation, the solution concentration is an indepen-
dent variable. Numerical methods along with fixed h5 ¼ cp T5 (23a)
point iteration are used to evaluate concentration
when pressure and temperature are known. The or
program also calculates the enthalpy of LiBr – H2O
solution when temperature and concentration are h5 ¼ hsat (T5 ,x5 ) (23b)
known from the experimental charts incorporated
in the program A heat balance on the solution heat exchanger gives
the enthalpy at point 3
h1 ¼ hsat (T1 ,x1 ) (15)
Qshx ¼ (mcp )h (T4  T5 ) ¼ (mcp )c (T3  T2 )
The program now moves to point 4, whose pressure
is known from equation (8) and whose temperature ¼ m4 (h4  h5 ) ¼ m2 (h3  h2 ) (24a)
is worked out using the input data m4
h3 ¼ h2 þ (h4  h2 ) (24b)
m2
T4 ¼ T12  DT124 (16)
The ratio m4/m2 can be expressed in terms of
The program now reads the solution concentration solution circulation ratio which is defined as
and enthalpy at point 4 using the Duhring diagram
and enthalpy chart, respectively m3
f ¼ (25)
m7
x4 ¼ xsat (P4 ,T4 ) (17)
The solution enters the desorber with concentration
h4 ¼ hsat (x4 ,T4 ) (18)
x3. Heat transfer to the desorber makes the solution
boil and water evaporate. As a result of this, a
It is assumed that the solution concentration at point strong solution will leave the desorber whose
3 (desorber inlet) is equal to that at point 1 (absorber concentration is
outlet), meaning that
x 3 m3 x3 m3
x3 ¼ x1 (19) x4 ¼ ¼
m3  m7 m4
The solution pressure is increased in the pump, which leads to
while its temperature and concentration remain con-
stant. The work done on the working fluid in the x3 m4
pump is very small, making the enthalpy change in ¼ (26)
x4 m3
the pump negligible. The solution entering the
pump is in saturated liquid phase, while leaving the According to equation (21a), we have
pump is in compressed liquid phase. Assuming no
temperature rise during pumping process, we have m7 m4
¼1 (27)
m3 m3
T2 ¼ T1 x2 ¼ x1 h2 ¼ h1 (20)
Combining equations (25) –(27) results in
Temperature at point 5 is now calculated using the
effectiveness of solution heat exchanger. The hot x4
fluid has a lower mass flow rate, because a mass f ¼ (28)
x4  x3
balance on the desorber gives
Knowing that m2 ¼ m3, the ratio m4/m2 is now
m3 ¼ m7 þ m 4 ) m 4 , m 3 (21a)
expressed in terms of solution circulation ratio
Q (mcp )h (T4  T5 ) T4  T5
1¼ ¼ ¼ (21b)
Qmax (mcp )h (T4  T2 ) T4  T2 m4 1
¼1 (29)
T5 ¼ T4  1(T4  T2 ) (21c) m2 f

Proc. IMechE Vol. 219 Part E: J. Process Mechanical Engineering E02604 # IMechE 2005
 Single-effect LiBr –H2O refrigeration cycle 265

Equation (24b) now becomes As the liquid phase at point 5 changes to liquid –

 vapor mixture at point 6, we have
1
h3 ¼ h2 þ 1  (h4  h5 ) (30)
f m5 x 5 ¼ m f x 6 (38)
x4 and x3 are known from equations (17) and (19),
The quality of mixture at point 6 is
respectively; f is then calculated from equation (28).
In addition to this, the values of enthalpies at points mg
2, 4, and 5 are known (Equations (20), (18), and (23b)) q6 ¼ (39)
m6
therefore h3 is deduced from equation (30). T3 is then
calculated from
The enthalpy at point 6 is known, because the pro-
h3 cess in expansion valve is iso-enthalpic, and h5 is
T3 ¼ (31) given in equation (23a)
cp

The refrigerant mass flow rate in condenser and h6 ¼ h5 (40)

evaporator is the same as the steam mass flow rate
generated in the desorber Therefore, the quality at point 6 should satisfy the
following relationship
m 7 ¼ m8 (32)
h6  h f
The solution mass flow rate in desorber can be q6 ¼ (41)
deduced from equation (25) hg  h f

m3 ¼ fm7 (33) where hg and hf are the enthalpies of saturated vapor

and liquid at T6, where T6 can be found from the
The solution mass flow rate leaving the desorber is Duhring diagram

m 4 ¼ m3  m 7 (34) T6 ¼ Tsat (P6 ,x6 ) (42)

The vapour temperature leaving the desorber is As T6, x6, and q6 are unknowns, a fixed point iteration
approximately equal to the arithmetic mean of is required to determine them. In this procedure, x6 is
solution temperatures entering and leaving it [14] guessed, T6 is calculated from equation (42), hf from
the Duhring diagram, hg from the steam tables, and
T3 þ T4 q6 from equation (41). In order to check the accuracy
T7 ¼ (35)
2 of x6, the value of mf is evaluated first
The refrigerant vapour entering the condenser is
mf ¼ (1  q6 )m6
assumed to be saturated vapor at the condenser
temperature [8], whose enthalpy is the saturated
vapor enthalpy at the condenser’s temperature and then x6 is worked out from equation (38) for
the purpose of comparison. The calculations are
h7 ¼ hg (T8 ) (36) repeated until the final value is close enough to the
initial value.
We have now the properties of all the state points We realize that the whole process is based on the
except 6. The solution entering the absorber falls first guess for T14 (equation (13)). It is now time to
on the absorber tubes through the nozzles and check the accuracy of this initial value. This is con-
experiences a pressure drop. In practice, there is ducted in the following manner. The heat balance
not an expansion valve between the solution heat on the absorber and condenser gives
exchanger and the absorber and the pressure drop
between the desorber and absorber is provided in Qabs ¼ m10 h10 þ m6 h6  m1 h1 (43)
the nozzles at the absorber entrance. The expansion
Qcon ¼ m7 (h7  h8 ) (44)
valve in Fig. 1 is just for modelling the pressure drop
between the desorber and the absorber. As a result of
As the cooling water leaving the absorber enters the
pressure drop in the nozzles, the subcooled liquid
condenser, we have
entering the nozzles becomes saturated liquid and
may even change phase. In this case, we may write
Qabs þ Qcon
m13 ¼ (45)
m6 ¼ m f þ mg (37) cp (T16  T13 )

E02604 # IMechE 2005 Proc. IMechE Vol. 219 Part E: J. Process Mechanical Engineering
266 M A Mehrabian and A E Shahbeik

The temperature of cooling water at absorber outlet It should be noted that the vapor temperature at the
is then inlet of absorber has been ignored when calculating
LMTDabs. This is because its mass flow rate is
Qabs much lower (about 1/15) than the solution mass
T14 ¼ þ T13 (46)
m13 cp flow rate entering the absorber. In fact, the vapor is
not involved in heat transfer, while heat is transferred
This temperature is now compared with the value between the solution (shell side) and cooling water
guessed in equation (13). The calculations are repeated (tube side).
until the final value is close enough to the initial value. Qdes is obtained from a heat balance on the
desorber

3 DESIGN SPECIFICATIONS Qdes ¼ m4 h4 þ m7 h7  m3 h3

Once the calculation of thermodynamic properties at The LMTD and UA are as follows:
18 points of the cycle is carried out, the design specifi-
cations can easily be obtained. The heat load, logarith- (T11  T4 )  (T12  T7 )
LMTDdes ¼
mic mean temperature difference and conductance ln(T11  T4 )/(T12  T7 )
for the evaporator, condenser, solution heat exchan- Qdes
ger, absorber, and desorber are calculated according UAdes ¼
LMTDdes
to the following formulation. Each unit is assumed
to be a counter flow heat exchanger, so that the It is worth mentioning that, in calculating LMTDdes
logarithmic mean temperature difference is used. the solution temperature at the inlet of desorber
Qeva is known from the input data; the LMTD and has been ignored. This is common practice in kettle
UA are reboilers and provides a lower value for LMTD.
When the heat duty of desorber is fixed, a higher
(T17  T10 )  (T18  T9 )
LMTDeva ¼ value for UA is obtained and avoids underestimating
ln(T17  T10 )/(T18  T9 )
the desorber conductance.
Qeva
UAeva ¼
LMTDeva
4 CYCLE PERFORMANCE
Qcon is evaluated using equation (14), the LMTD and
UA are The cycle performance is designated by:
(T8  T16 )  (T8  T15 ) (a) the chiller refrigeration capacity;
LMTDcon ¼ (b) coefficient of performance;
ln(T8  T16 )/(T8  T15 )
Qcon (c) solution pumping power;
UAcon ¼ (d) total conductance of the component;
LMTDcon
(e) circulation ratio;
It should be explained that T7 is replaced with T8 (f) highest and lowest temperatures in the cycle;
when calculating LMTDcon. It has been seen that (g) highest and lowest pressures in the cycle;
LMTDcon based on T7 is much higher than the (h) highest and lowest concentrations in the cycle;
actual temperature difference and therefore results (i) ambient temperature.
in underestimating the condenser conductance.
The chiller refrigeration capacity is defined as the
Qshx is calculated using equation (24a); the LMTD
evaporator heat duty, which is given as an input data.
and UA are
The coefficient of performance for the refrigeration
(T4  T3 )  (T5  T2 ) cycle is
LMTDshx ¼
ln(T4  T3 )/(T5  T2 )
Qeva
Qshx COP ¼
UAshx ¼ Qdes
LMTDshx

Qabs is calculated using equation (43); the LMTD and The pump is assumed to perform reversibly. The
UA are pumping power is small compared with the heat
duties of the heat exchangers and can therefore
(T6  T14 )  (T1  T13 ) be ignored. It is, however, from the first law for
LMTDabs ¼
ln(T6  T14 )/(T1  T13 ) steady-state steady-flow process
Qabs
UAabs ¼
LMTDabs Wp ¼ m1 (h1  h2 )

Proc. IMechE Vol. 219 Part E: J. Process Mechanical Engineering E02604 # IMechE 2005
 Single-effect LiBr –H2O refrigeration cycle 267

Once the conductances of heat exchangers are calcu- DE2con is the exergy loss due to heat transfer
lated, the total conductance is determined as a per- irreversibility of cold stream,
formance criterion. The circulation ratio is calculated
from equation (25). Highest and lowest temperatures, DE2con ¼ m15 (e15  e16 )  Qcon (1  T0 /Tcon )
pressures, and concentrations in the cycle are part of
the response of the program when calculating the ther- and DE3con the exergy loss due to temperature rise of
modynamic properties of the state points. The ambient the cold stream; this is, on the other hand, the exergy
temperature is given as input data. gain of the ambient

DE3con ¼ m15 (e16  e15 )

5 EXERGY ANALYSIS

The flow availability or exergy of a substance in any

state as it enters or leaves the control volume in a 5.3 Solution heat exchanger
steady-state steady-flow process is defined as [15] The total exergy change in solution heat exchanger
is due only to heat transfer irreversibility which
e ¼ (h  T0 s þ V 2 /2 þ gz)  (ho  T0 so þ gzo ) consists of two terms

The program calculates the exergy at each state point DEshe ¼ DE2shx ¼ DEhs þ DEcs
of the absorption cycle according to the above
equation. Based on such information, the individual where DEhs is the exergy change due to heat transfer
components of the cycle can be analysed from the irreversibility of hot stream
second law point of view.
DEhs ¼ m4 (e4  e5 )  Qshx (1  T0 /Tshx )
5.1 Evaporator
and DEcs is the exergy change due to heat transfer
The total exergy change in evaporator consists of irreversibility of cold stream
three terms
DEcs ¼ m2 (e2  e3 ) þ Qshx (1  T0 /Tshx )
DEeva ¼ DE1eva þ DE2eva þ DE3eva
Summing up the above equations gives
where DE1eva is the exergy loss due to internal
irreversibility of cold stream DE2shx ¼ m4 (e4  e5 ) þ m2 (e2  e3 )

DE1eva ¼ m9 (e9  e10 ) þ Qeva (1  T0 /Teva )

5.4 Absorber
DE2eva is the exergy loss due to heat transfer
irreversibility of hot stream The total exergy loss in absorber consists of three
terms
DE2eva ¼ m17 (e17  e18 )  Qeva (1  T0 /Teva )
DEabs ¼ DE1abs þ DE2abs þ DE3abs
and DE3eva is the exergy gain due to temperature drop
of the hot stream; this is, on the other hand, the where DE1abs is the exergy loss due to internal
exergy loss of the ambient irreversibility of LiBr –H2O solution

DE3eva ¼ m17 (e18  e17 ) DE1abs ¼ m6 e6 þ m10 e10  m1 e1  Qabs (1  T0 /Tabs )

5.2 Condenser DE2abs is the exergy loss due to heat transfer

irreversibility of cold stream (water)
The total exergy change in the condenser consists of
three terms DE2abs ¼ m13 (e13  e14 ) þ Qabs (1  T0 /Tabs )
DEcon ¼ DE1con þ DE2con þ DE3con
and DE3abs is the exergy loss due to temperature rise
where DE1con is the exergy loss due to internal of the hot stream; this is, on the other hand, the
irreversibility of hot stream exergy gain of the ambient

DE1con ¼ m7 (e7  e8 )  Qcon (1  T0 /Tcon ) DE3abs ¼ m13 (e14  e13 )

E02604 # IMechE 2005 Proc. IMechE Vol. 219 Part E: J. Process Mechanical Engineering
268 M A Mehrabian and A E Shahbeik

5.5 Desorber where DEloss is the sum of exergy loss due to internal
irreversibility, exergy loss due to heat transfer
The total exergy change in desorber consists of three
irreversibility, and exergy loss in the expansion valve
terms
X X
DEdes ¼ DE1des þ DE2des þ DE3des DEloss ¼ DE1 þ DE2 þ DEsev þ DEeev

where DE1des is the exergy gain due to internal The cold water leaving the evaporator is usually used
irreversibility of LiBr –H2O solution to cool the water leaving the condenser. Therefore,
exergy given to the chiller in the evaporator, is
DE1des ¼ m3 e3  m4 e4  m7 e7 þ Qdes (1  T0 /Tdes ) returned to the water in cooling tower and is reduced
from DEloss .
DE2des is the exergy loss due to heat transfer DEgiven is the sum of exergy given to the chiller in
irreversibility of LiBr –H2O solution the evaporator, desorber, and pump

DE2des ¼ m11 (e11  e12 ) þ Qdes (1  T0 /Tdes ) DEgiven ¼ DE3eva þ DE3des þ DEp

and DE3des is the exergy gain due to temperature rise putting all these together, the second law efficiency
of the hot stream; this is, on the other hand, the reads
exergy gain of the ambient P P
DE1 þ DE2 þ DEsev þ DEeev  DE3eva
h¼1
DE3eva þ DE3des þ DEp
_ 11 (e12  e11 )
DE3des ¼ m

6 RESULTS AND DISCUSSION

5.6 Expansion valves and pump
The exergy change in the solution expansion valve This paper presents the results of a simulation
(DEsev), evaporator expansion valve (DEeev) and program to study the performance and design charac-
pump (DEp) are negligibly small. They are, however, teristics of single effect LiBr–H2O absorption refriger-
respectively ation cycles. A typical (351.7 kW) absorption chiller is
studied based on the developed model. A set of input
DEsev ¼ m5 (e5  e6 ) data is listed in Table 1. The corresponding output
DEsev ¼ m8 (e8  e9 ) data are shown in Tables 2–5. The program is run
when one item in Table 1 is varied, while the other
DEp ¼ m1 (e1  e2 )
items remain constant. In other words, the effect of
individual input data on design specifications, chiller
5.7 Second-law efficiency performance, and second-law efficiency are studied.
Such an investigation is performed to examine the
Second-law efficiency is a measure of reversible
effect of input data on design specifications, coeffi-
performance of the cycle. It is defined in terms of
cient of performance, and second law efficiency (h).
exergy changes in the components of the cycle, gen-
The following results are obtained.
erally speaking it is defined as the ratio of total exergy
taken from the system to the total exergy given to
1. The water temperature entering the desorber
the system
(T11) has a dramatic effect on desorber conduc-
tance. The second-law efficiency is also affected
DEtaken
h¼ by T11.
DEgiven 2. The water temperature leaving the condenser
(T16) has a tremendous effect on solution heat
or exchanger conductance. The absorber conduc-
tance is also significantly affected by T16.
DEloss  DE3eva 3. The water temperature leaving the evaporator
h¼1
DEgiven (T18) has an important effect on evaporator

Table 1 Typical set of input data (SI units)

T11 T12 T13 T16 T17 T18 Qeva DT8 – 16 DT18 – 10 DT12 – 4 DT1 – 14 1 T0

88 83 31 36.5 13 8 351.7 2.9 1.2 1.6 2.7 0.52 20

Proc. IMechE Vol. 219 Part E: J. Process Mechanical Engineering E02604 # IMechE 2005
 Single-effect LiBr –H2O refrigeration cycle 269

Table 2 Thermodynamic properties (SI units) of state points corresponding to input data in Table 1
h  1023 m P  1023 q T x s  1023 e  1023

1 87.4 2.4886 0.988 0 36.8 55.1 0.236 21.4

2 87.4 2.4886 7.15 0 36.8 55.1 0.236 21.4
3 130.3 2.4886 7.15 0 58.6 55.8 0.373 23.97
4 190.1 2.3388 7.15 0 81.4 59.63 0.477 53.31
5 144.1 2.3388 7.15 0 58.2 58.63 0.345 46.29
6 144.5 2.3388 0.988 96 44.8 58.63 0.275 46.29
7 2639.2 0.1498 7.15 100 70 0 0.268 208.42
8 165 0.1498 7.15 0 39.4 0 0.565 2.55
9 165 0.1498 0.988 5.49 6.8 0 0.591 25.15
10 2512.9 0.1498 0.988 100 6.8 0 8.98 2116.51
11 368.2 24.5139 – – 88 – 1.17 28.18
12 347.3 24.5139 – – 83 – 1.112 24.41
13 129.9 37.6706 – – 31 – 0.451 0.87
14 143.1 37.6706 – – 34.1 – 0.494 1.4
15 143.1 37.6706 – – 34.1 – 0.494 1.4
16 152.9 37.6706 – – 36.5 – 0.526 1.87
17 54 16.7564 – – 13 – 0.195 20.38
18 33.5 16.7564 – – 8 – 0.121 21.09

Table 3 Design specifications (SI units) corresponding to input data in Table 1

Evaporator Absorber SHX Desorber Condenser

Heat duty 351 685 496 650 106 670 514 068 368 000
LMTD 3.04 8 22.07 9.44 3.96
UA 115 552 62 119 4835 54 511 93 252

Table 4 Chiller performance (SI units) corresponding to input data in Table 1

P
Qeva COP Wp UA f Tmax Tmin Pmax Pmin xmax xmin T0

351 685 0.684 0 330 269 15.88 81.4 6.8 7150 988 58.63 55.1 20

Table 5 Exergy analysis (SI units) of the cycle corresponding to input data in Table 1
EVA ABS SHX DES CON SEV EEV Pump

DE1 0.1 4.65 – 211.71 8 – – –

DE2 28.47 12.94 10.02 7.88 5.13 – – –
DE3 211.89 19.96 – 292.42 17.71 – – –
DE – – – – – 0 1.15 0

conductance. The solution heat exchanger con- 7. When the approach temperature in the evapor-
ductance is most seriously affected by T18. ator (DT18 – 10) varies, the biggest changes
4. The absorber and solution heat exchanger geo- belong, respectively, to evaporator and solution
metries are seriously affected by the water temp- heat exchanger geometries.
erature at absorber inlet (T13). 8. When the approach temperature in desorber
5. When the solution heat exchanger effective- (DT12 – 4) varies, the biggest changes belong,
ness (1) varies, the solution heat exchanger respectively, to the solution heat exchanger and
conductance experiences the biggest variations. desorber geometries.
The desorber conductance is also slightly 9. When the approach temperature in absorber
affected. (DT1 – 14) varies, the biggest changes belong,
6. When the approach temperature in the conden- respectively, to solution heat exchanger and
ser (DT8 – 16) varies, the condenser conductance absorber geometries.
is mostly affected. The solution heat exchange 10. The ambient temperature (T0) has a significant
conductance also varies considerably. effect on the second-law efficiency of the cycle.

E02604 # IMechE 2005 Proc. IMechE Vol. 219 Part E: J. Process Mechanical Engineering
270 M A Mehrabian and A E Shahbeik

A diagrammatic representation of the effect of input

data on second law efficiency, coefficient of perform-
ance, and geometry of heat exchangers is shown in
Figs 2 – 14. Figures 2 – 6 represent the effect of temp-
eratures (T11, T16, Tl8, T13, and T0) on second-law
efficiency. Figures 7 – 11 indicate the effect of
temperature approaches (DT8 – 16, DT18 – 10, DT12 – 4,
DT1 – 14) and solution heat exchanger effectiveness
(1) on the coefficient of performance. Figures 12– 14
present the effect of solution heat exchanger
effectiveness (1), water temperature leaving the con- Fig. 5 Effect of water temperature at the inlet of
denser (T16), and temperature approach in evapor- absorber on second-law efficiency
ator (DT18 – 10) on geometry of heat exchangers.
Figures 2 – 5 present the variations of second-law
efficiency with respect to desorber, condenser,
evaporator, and absorber temperatures, respectively.

Fig. 6 Effect of ambient temperature on second-law

efficiency

Fig. 2 Effect of water temperature at the inlet of

desorber on second-law efficiency

Fig. 7 Effect of approach temperature in condenser on

Fig. 3 Effect of water temperature at the outlet of coefficient of performance
condenser on second-law efficiency

Fig. 4 Effect of water temperature at the outlet of Fig. 8 Effect of approach temperature in evaporator
evaporator on second-law efficiency on coefficient of performance

Proc. IMechE Vol. 219 Part E: J. Process Mechanical Engineering E02604 # IMechE 2005
 Single-effect LiBr –H2O refrigeration cycle 271

Fig. 9 Effect of approach temperature in desorber on

Fig. 13 Effect of water temperature at the outlet of
coefficient of performance
condenser on geometries of heat exchangers

Fig. 10 Effect of approach temperature in absorber on

coefficient of performance Fig. 14 Effect of approach temperature in evaporator
on geometries of heat exchangers

component at which the second-law efficiency is

maximum. Further study of these predictions helps
the design engineer to select the conditions under
which the maximum performance is achieved.

7 COMPARISON

Fig. 11 Effect of solution heat exchanger effectiveness

Heydari et al. [10] have performed a basic analysis to
on coefficient of performance
study the amount of entropy generation due to irre-
versibility in the heat transfer, mixing, and fluid
flow processes in various components of the chiller
It can be seen that, changing the working temp- based on geometric and manufacturing constraints.
erature of constitutive components, which is the They have used a very basic mathematical equation
result of variations in physical characteristics, makes solver to solve the non-linear set of simultaneous
the second-law efficiency go through significant equations describing the performance of an absorp-
changes. There is a working temperature for each tion cihiller. In part of their analysis, they have
plotted the second-law efficiency with respect to
desorber, absorber, condenser, and evaporator
temperatures. However, the ranges of efficiencies
and temperatures used in their investigation are
quite different from those used in this paper; quali-
tative agreement is achieved when comparing their
plotted results with those presented in Figs 2 –5.

8 CONCLUSION

Fig. 12 Effect of solution heat exchanger effectiveness A computer program has been developed to predict
on geometries of heat exchangers the geometry, performance, and second law

E02604 # IMechE 2005 Proc. IMechE Vol. 219 Part E: J. Process Mechanical Engineering
272 M A Mehrabian and A E Shahbeik

efficiency of a LiBr – H2O absorption refrigeration 6. The evaporator conductance is not affected by
cycle. The conditions of hot water entering and leav- the approach temperature in the condenser.
ing the desorber, cooling water entering the absorber Increasing this temperature difference increases
and leaving the condenser, chilled water entering the absorber, solution heat exchanger, and
and leaving the evaporator, as well as the approach desorber conductances. Increasing this temp-
temperatures in condenser, evaporator, desorber, erature difference decreases the condenser
and absorber, the effectiveness of solution heat conductance, the coefficient of performance,
exchanger, the chiller refrigeration power, and the and the second law efficiency.
ambient temperature are used as input data. The 7. The evaporator conductance, the coefficient of
program then gives the thermodynamic properties performance, and the second law efficiency are
of all state points in the cycle, the design information decreased when the approach temperature in
of all heat exchangers in the cycle, the overall per- the evaporator is increased. The absorber, sol-
formance, and the second-law efficiency of the ution heat exchanger, desorber, and condenser
cycle. The results deduced from the computer pro- conductances are increased when the approach
gram are used to study the effect of input parameters temperature in the evaporator is increased.
on geometry, cycle performance, and second-law 8. The evaporator conductance remains constant
efficiency. The program can be used either for when the approach temperature in the desorber
sizing a new refrigeration cycle or rating an existing is varied. The absorber, solution heat exchanger,
system. It can also be used for optimization pur- and condenser conductances are increased when
poses. Some concluding remarks from the analysis the approach temperature in desorber is
performed in this paper reveals the following. increased. The desorber conductance, coefficient
of performance, and second-law efficiency are
1. Water temperature entering the desorber does not reduced when the approach temperature in the
have any effect on evaporator, absorber, solution desorber is increased.
heat exchanger, and condenser geometries. It 9. The evaporator conductance remains
does not affect the coefficient of performance. unchanged when the approach temperature in
Increasing this temperature decreases the deso- the absorber varies. The absorber conductance,
rber conductance and also decreases the the coefficient of performance, and the second-
second-law efficiency. law efficiency are decreased when the approach
2. Water temperature exiting the condenser does temperature in the absorber is increased. The
not have any effect on evaporator geometry. solution heat exchanger, desorber, and conden-
Increasing this temperature increases the ser conductances are increased when the
absorber, solution heat exchanger, and desorber approach temperature in absorber is increased.
conductances and decreases the condenser con- 10. The ambient temperature does not have any
ductance. Increasing this temperature decreases effect on geometry of heat exchangers. The
the coefficient or performance, but may increase coefficient of performance is not affected by
or decrease the second law efficiency. the ambient temperature, but the second-law
3. Increasing the water temperature exiting evapor- efficiency is reduced when the ambient temp-
ator increases the evaporator conductance, the erature is increased.
coefficient of performance, and the second-law
efficiency. Increasing this temperature reduces
the absorber, solution heat exchanger, desorber, REFERENCES
and condenser conductances.
4. Water temperature entering absorber does not 1 Pichel, W. Development of large capacity lithium
affect the evaporator conductance. Increasing bromide absorption refrigeration machine in USSR.
this temperature, increases the conductance of ASHRAE J., 1996, 85(8).
all other heat exchangers. Increasing this tempera- 2 Eisa, M. A. R., Devotta, S., and Holland, F. A. Thermo-
ture reduces the coefficient of performance but dynamic design data for absorption heat pump systems
increases the second-law efficiency. operating on water lithium bromide: Part I: cooling.
5. The evaporator conductance is not affected by Applied Energy, 1986, 24, 287 – 301.
3 Eisa, M. A. R. and Holland, F. A. A study of the opera-
effectiveness of solution heat exchanger. Increas-
ting parameters in a water lithium bromide absorption
ing the effectiveness reduces the conductances
cooler. Energy Res., 1986, 10(2), 137 – 144.
of absorber and condenser, and increases the 4 Eisa, M. A. R., Diggory, P. J., and Holland, F. A. Experi-
solution heat exchanger and desorber mental studies to determine the effect of differences in
conductances. Increasing the effectiveness absorber and condenser temperatures on the perfor-
improves both the coefficient of performance mance of water lithium bromide absorption cooler.
and second-law efficiency. Energy Convers. Mgmt, 1987, 27(2), 253 – 259.

Proc. IMechE Vol. 219 Part E: J. Process Mechanical Engineering E02604 # IMechE 2005
 Single-effect LiBr –H2O refrigeration cycle 273

5 Waleed, A. F. Study of the effect of design parameters

e exergy per unit mass, J kg21
on a two ton lithium bromide absorption unit. MSc
Thesis, University of Technology, Baghdad, Iraq, 1983. E exergy, J
6 Mclinden, M. O. and Klein S. A. Steady-state modelling f solution circulation ratio
of absorption heat pumps with a comparison to experi- g gravitational acceleration, m s22
ments. ASHRAE Trans. Part 2B, 1985, 91, 793 – 806. h enthalpy per unit mass, J kg21
7 Grossman. G. and Micheslon E. A. A modular compu- m mass flow rate, kg s21
ter simulation of absorption systems. ASHRAE Trans. P pressure, Pa
Part 2B, 1985, 91, 1808– 826. q quality of saturated mixture
8 Joudi, K. A. and Lafta, A. H. Simulation of a simple Q heat transfer rate, W
absorption refrigeration system. Energy Convers. s entropy per unit mass, J kg21 K21
Mgmt, 2001, 42, 1575 – 1605.
T0 ambient temperature, 8C
9 Catalogue for LiBr Absorption Chiller Model ES-
T temperature, 8C
2A4.MW, 60 TR Capacity (Mitsubishi Heavy Indus-
tries Ltd, Energy and Environment Research Center, DT temperature approach, 8C
Baghdad). U overall heat transfer
10 Heydari, A., Babadi, F., Jani, S., and Bafkar R. Second coefficient, W m22 K21
law based thermodynamic analysis of a lithium bro- UA conductance, W K21
mide absorption chiller. In 20th International Congress V velocity, m s21
of Refrigeration IIR/IIF, Sydney, 1999. x concentration of LiBr in
11 Feuerecker, J., Scharfe, J., Greiter, I., Frank, C., and LiBr – H2O solution
Alefeld, G. Measurement of thermophysical properties z height, m
of aqueous LiBr solutions at high temperatures and 1 effectiveness
concentrations. In International Absorption Heat
h second law efficiency
Pump Conference, 1993 (ASME, New York).
12 Mostafavi, M. and Agnew B. The effect of ambient
temperature on the surface area of components of an
air-cooled lithium bromide/water absorption unit.
Appl. Therm. Engng, 1996, 16, 515 – 522.
13 Mostafavi, M. and Agnew B. The impact of ambient Subscripts
temperature on lithium bromide/water absorption
ABS absorber
machine performance. Appl. Therm. Engng, 1996, 16,
CON condenser
313 – 319.
14 Grossman, G. and Perez-Blanco, H. Conceptual design COP coefficient of performance
and performance analysis of absorption heat pumps for cs cold stream
waste heat utilization. ASHRAE Trans., 1993, 2691, DES desorber
451 – 460. EVA evaporator
15 Sonntag, R. E., Borgnakke, C., and Van Wylen, G. J. EEV evaporator expansion value
Fundamentals of Thermodynamics, 1988 (Wiley, f saturated liquid
Chichester). g saturated vapour
hs hot stream
APPENDIX LMTD logarithmic mean temperature
difference, 8C
max maximum
Notation
p pump
A heat transfer area, m2 sat saturated
cp specific heat at constant SEV solution expansion valve
pressure, J kg21 K21 SHX solution heat exchanger

E02604 # IMechE 2005 Proc. IMechE Vol. 219 Part E: J. Process Mechanical Engineering