Modelling and performance evaluation

of a low-temperature ammonia –water

absorption refrigeration system
..............................................................................................................................................................

Y.T. Ge1,*, S.A. Tassou1 and I. Chaer2

1
Mechanical Engineering, School of Engineering and Design, Brunel University, Uxbridge,
Middlesex UB8 3PH, UK
2
Faculty of Engineering, Science and the Built Environment, London South Bank
University (T617), Borough Road, London SE1 0AA, UK
.............................................................................................................................................
Abstract
This paper presents the simulation of a low-temperature gas-fired ammonia – water absorption chiller
for refrigeration applications. The model was developed as part of a research effort to investigate
microturbine-based tri-generation systems for application in the food retail industry. The absorption
chiller model was developed in the TRNSYS environment by integrating the main component models
in the system and will form part of an overall TRNSYS-based supermarket model. The chiller model
was validated against experimental results obtained on a 12 kW absorption chiller in the laboratory. The
model was subsequently used to investigate the influence of important design and operating parameters
on the performance of the chiller.

Keywords: ammonia – water absorption chiller; modelling and validation; low-temperature

refrigeration
*Corresponding author:
[email protected] Received 6 May 2009; revised 11 May 2009; accepted 12 May 2009
................................................................................................................................................................................

1 INTRODUCTION provides a refrigeration capacity of 12 kW at 2128C flow

temperature. A schematic diagram of this chiller is shown in
Gas-fired absorption chillers can offer advantages over electri- Figure 1. A similar type of chiller was modelled by Lazzarin
cally driven machines in applications where the local electrical et al. [3,4] with the integration of the main component models
supply grid has insufficient capacity and where noise is a con- in the system. Of the component models, the generator was
sideration. Other advantages of absorption chillers include the divided into two control volumes while other components had
use of environmental-friendly refrigerants and the possibility one control volume each. The system model was validated by
to use waste heat or the exhaust gases of local combined heat comparing the cooling capacities from manufacture at con-
and power systems to drive the generator. Most applications of ditions of different temperatures of air inlet and brine outlet.
absorption chillers have been in the air-conditioning field The model was then utilized to predict the variations of
where the flow temperature required is above 68C. These chil- solution concentrations, high and low solution pressures, com-
lers invariably use lithium bromide – water as fluid pair with ponent heat fluxes and cooling capacities with the tempera-
water being the refrigerant and lithium bromide the absorbent. tures of air inlet and brine outlet. Some constant parameters
There have also been applications of absorption chillers at were assumed for the component models such as the effective-
temperatures down to 2558C or even lower in the chemical ness for generator, condenser and evaporator, the temperature
and food manufacturing industries, but these have been of the differences between solution and air for the air-cooled heat
specially made type using ammonia– water as fluid pair with exchanger, and the temperature differences between vapour
ammonia being the refrigerant and water the absorbent [1,2]. and solution for the rectifier. These constant parameters which
More recently, small packaged ammonia –water chillers have are critical to the model accuracy can be determined by further
become available in the market and this has prompted their experiments of the chiller at different conditions.
evaluation for a wide range of applications at temperatures Experimental investigations on the similar type of chiller
below 08C. An example of this is the Robur unit which with different chilled solution and various structure of

International Journal of Low-Carbon Technologies 2009, 4, 68–77

# The Author 2009. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected]
doi:10.1093/ijlct/ctp015 Advance Access Publication 4 June 2009 68

Downloaded from https://academic.oup.com/ijlct/article-abstract/4/2/68/745728

by guest
on 08 May 2018
 Low-temperature ammonia – water absorption refrigeration system

Figure 1. Schematic diagram of the absorption refrigeration system.

evaporator were carried out by Horuz et al. [5]. Water-cooled system cycle and the thermal duty of each component. For the
absorber and condenser were introduced in addition to the solution to converge, it requires careful selection of input par-
original air-cooled ones so as to examine the system perform- ameters, such as the UA or effectiveness values of the heat
ance under the conditions of a wider range of high-side press- transfer components. This makes the model difficult to use by
ures. The experimental investigations were carried out under inexperienced users. To obtain a reasonable prediction of
six different operating conditions including variable chilled system performance by the model, the correlated value of UA
water inlet temperature and flow rate, changeable chilled water or effectiveness for each component from test results is still
level in the evaporator drum, varied generator heat input, and needed.
different cooling water flow rate and inlet temperature to the Apart from the above system models formed by the inte-
water-cooled condenser. It was found that the cooling capacity, gration of simplified component models, the detailed com-
component heat transfer rates, and cooling coefficient of per- ponent models are still significant to describe operationally
formance (COP) were all increased with higher chilled water and assist the optimal designs for the components. Chua et al.
inlet temperature. In the mean time, the above three perform- [11] developed a detailed absorption generator model in which
ance parameters were all increased with elevated energy input the multi-stream non-isothermal heat and mass transfers were
to generator. In addition, the cooling capacity and COP were recognized and considered. Goel et al. [12] developed an
both augmented with increased condenser cooling water flow analytical model to simulate the combined heat and mass
rate but the COP was decreased with higher condenser cooling transfer process in a counter-current ammonia– water-based
water temperature. The solution high and low pressures which absorber. The model requires detailed description of the geo-
were quite important to the system performance were however metric characteristics of the absorber. It uses empirical corre-
not presented. lations to predict the heat and mass transfer coefficients and
The models of different types of absorption refrigeration accounts for both liquid and vapour phase mass transfer resist-
systems were reported in series by Grossman et al. [6-10] and ances. Similar investigations on absorbers were carried out by
were finalized into a modular simulation tool—ABSIM for Kaynakli et al. [13] and Kwon et al. [14]. All these models
advanced absorption systems. The codes underneath the require detailed description of the components’ geometric
system models are based on unit subroutines containing the data, which inversely limits their applications to different types
governing equations for the system’s components. Each com- of system.
ponent is assumed as one control volume and the applied gov- Any simulation model requires a method for calculating
erning equations for the control volume are similar to those the thermophysical properties of the fluid pair in the
used by Lazzarin et al. [3,4]. Based on user-supplied cycle system. A simple, practical and fairly accurate method to do
diagrams, working fluid and given operating conditions, this for the ammonia– water fluid pair was proposed by
the model calculates the properties at each state point in the Conde [15].

International Journal of Low-Carbon Technologies 2009, 4, 68– 77 69

Downloaded from https://academic.oup.com/ijlct/article-abstract/4/2/68/745728

by guest
on 08 May 2018
 Y.T. Ge et al.

This paper presents a model of a low-temperature air- ‘1’. The weak solution is heated and distilled in the generator
cooled ammonia – water absorption chiller system. The system and as a result, a strong solution is formed at the bottom of
model is an integration of the main component sub-models. the generator and high concentration ammonia refrigerant at
To avoid the need to have detailed data on the geometric the top. The strong solution then exits the generator at ‘2’ and
characteristics of the components and ensure speed of solution, flows through an expansion device before entering the absorber
each component model was based on the lumped parameter where it is sprayed over the SHX. The refrigerant flows from
methodology. To facilitate the modelling processes, the genera- the top of the generator to the rectifier where more water in
tor and rectifier are combined into one control volume while the ammonia refrigerant condenses and returns back to the
the absorber is divided into two control volumes. In the mean generator. The nearly pure refrigerant then exits the rectifier at
time, extensive experiments were carried out including the ‘8’ and flows to the condenser for the repetition of the cycle.
variation of high and low solution pressures with the tempera- The model of each component was developed based on the
tures of air inlet and brine outlet. The test results are used to energy, mass and species conservation equations. For simpli-
validate the model and correct the heat transfer parameters city, each component was assumed as a single or two control
needed in the component models. The following sections volumes. The models of the more complicated components in
describe the modelling methodology adopted, validation the system are described as below.
against experimental measurements and model applications.
This is a first stage in the development of an integrated model
of a microturbine-based tri-generation system for low-
temperature food refrigeration applications. 2.1 Rectifier
The rectifier was assumed to be a single control volume. The
assumptions made in developing the model are:

2 MODELLING SCHEME WITH THE (1) Pure saturated refrigerant vapour at rectifier exit ( point 8
LUMPED METHOD in Figure 1).
(2) Saturated vapour solution at top of generator ( point 100 ).
A schematic diagram of the refrigeration system considered in (3) Saturated liquid solution from rectifier to generator ( point
this investigation is shown in Figure 1. Referring to the 10 ).
diagram, the system consists of eight main components: gen- (4) No heat loss in rectifier.
erator, rectifier, air-cooled condenser, sub-cooling heat exchan-
ger, liquid-cooling evaporator, absorber (solution-cooled), Conservation of mass for ammonia:
air-cooled heat exchanger and solution pump. After flowing ð1 þ RÞj100 ¼ j8 þ Rj10 ð1Þ
through the generator and rectifier, the nearly pure refrigerant
at ‘8’ enters the air-cooled condenser where it is condensed to Conservation of energy:
liquid at ‘9’. The refrigerant then passes through the connec- ð1 þ RÞh100 ¼ h8 þ Rh10 þ qr ð2Þ
tion pipes and arrives at ‘10’. From there the refrigerant flows
where
through a tube-in-tube heat exchanger where it exchanges heat
with the refrigerant vapour leaving the evaporator and is R¼m
_ 10 =m
_8
further subcooled at ‘11’. The refrigerant then flows through a
thermodynamic expansion valve at ‘12’ before entering the
evaporator coil where it evaporates while absorbing heat from
a brine solution. 2.2 Generator and rectifier
The refrigerant vapour exiting the evaporator at ‘13’ then The generator and rectifier are combined into one control
flows through the tube-in-tube heat exchanger where it is volume to facilitate the modelling. Assumptions for the gen-
superheated to ‘14’ before entering the absorber. The ‘strong’ erator are:
solution from the generator enters the absorber at ‘3’ where it
is first cooled by a solution heat exchanger (SHX) before (1) Saturated liquid for weak solution outlet of generator at
absorbing the refrigerant flowing from the evaporator. It ‘2’.
should be noted that in this paper ‘strong’ solution refers to a (2) No heat loss in generator.
solution that is rich in absorbent, whereas ‘weak’ solution
refers to a solution that is weak in absorbent and rich in Conservation of mass for ammonia:
refrigerant. The ‘weak’ solution formed in the absorber then
flows through the air-cooled heat exchanger where the heat f j1 ¼ ð f  1Þj2 þ j8 ð3Þ
generated by the absorption process is rejected to the ambient. Conservation of energy:
The solution then exits the heat exchanger at ‘5’ and is
pumped through the rectifier and absorber to the generator at qg ¼ h8  h2 þ qr þ f ðh2  h1 Þ ð4Þ

70 International Journal of Low-Carbon Technologies 2009, 4, 68– 77

Downloaded from https://academic.oup.com/ijlct/article-abstract/4/2/68/745728

by guest
on 08 May 2018
 Low-temperature ammonia – water absorption refrigeration system

where, in Equations (3) and (4) Conservation of energy:

m
_1 _ 3a h3a þ m
m _ 14 h14 ¼ m
_ 4 h4 ð10Þ
f ¼
m
_8
2.4 Air-cooled heat exchanger
2.3 Solution-cooled absorber The conservation equations for the air-cooled heat exchanger
To simplify the simulation, the solution-cooled absorber was are similar to those of equations in CV1 of solution-cooled
divided into two control volumes, CV1 and CV2, as shown in absorber except for the different endpoints of the two heat
Figure 2. The simulation assumes that the weak solution exchange sides. In this component model, the warm side is the
exiting the solution cooled absorber is saturated liquid. weak solution and the cold side is air.
For control volume CV1:Conservation of mass for solution:
2.5 Air-cooled condenser
_3¼m
m _ 3a ð5Þ Conservation of energy:
Conservation of mass for ammonia:
Q_ c ¼ m
_ 8 ðh8  h9 Þ ¼ m
_ a CPa ðTaout  Tain Þ ð11Þ
_ 3 j3 ¼ m
m _ 3a j3a ð6Þ
Heat transfer:
Conservation of energy:
Q_ c ¼ UA  LMTD ð12Þ
_a ¼m
Q _ 7 ðh1  h7 Þ ¼ m
_ 3 ðh3  h3a Þ ð7Þ
where
Heat transfer:
DTc;max  DTc;min
Q_ a ¼ UA  LMTD ð8Þ LMTD ¼

ln DTc;max =DTc;min
where DTa;max ¼ maxðT8  Taout ; T9  Tain Þ

DTa;max  DTa;min DTa;min ¼ minðT8  Taout ; T9  Tain Þ

LMTD ¼

ln DTa;max =DTa;min
DTa;max ¼ maxðT3  T1 ; T3a  T7 Þ 2.6 Liquid-cooling evaporator
DTa;min ¼ minðT3  T1 ; T3a  T7 Þ Conservation of energy:

For control volume CV2: Q_ e ¼ m

_ 12 ðh13  h12 Þ ¼ m
_ cf CPcf ðTcf in  Tcf out Þ ð13Þ
Conservation of mass for ammonia: where

_ 3a j3a þ m
m _ 14 ¼ m
_ 4 j4 ð9Þ _ 12 ¼ m
m _8
Heat transfer:

Q_ e ¼ UA  LMTD ð14Þ
where

DTe;max  DTe;min
LMTD ¼

ln DTe;max =DTe;min
DTa;max ¼ maxðTcf in  T13 ; Tcf out  T12 Þ
DTa;min ¼ minðTcf in  T13 ; Tcf out  T12 Þ

As a result, system COP can be calculated as:

Qe
COP ¼
Qg þ Qp

2.7 System model

The component models were integrated to form a system model
within the TRNSYS [16] environment as shown in Figure 3.
TRNSYS is a transient system simulation programme with a
modular structure. It recognizes a system description language
Figure 2. Two control volumes of solution-cooled absorber. in which the user specifies the components that constitute the

International Journal of Low-Carbon Technologies 2009, 4, 68– 77 71

Downloaded from https://academic.oup.com/ijlct/article-abstract/4/2/68/745728

by guest
on 08 May 2018
 Y.T. Ge et al.

the model. The chiller has a nominal refrigeration capacity of

12 kW at conditions of ambient air temperature of 358C and
chilled water flow temperature of 258C.
The unit was installed and commissioned in the laboratory
to facilitate the experimental investigations. The test facility is
shown schematically in Figure 5. It consists of the air-cooled
absorption chiller and an air-cooling loading circuit which is a
cooling coil to heat up the low-temperature secondary refriger-
ant from the chiller and provide the refrigeration load to the
chiller. The refrigeration load can be adjusted by modulating
Figure 3. Model of absorption chiller within the TRNSYS environment.
the fans’ frequency to facilitate testing at a fluid temperature at
the evaporator outlet down to 2128C. The fluid in the closed
system and the manner in which they are connected. The circuit was a mixture of 40% mass propylene glycol/water. The
outputs of each component model can be inputs to another test facility was instrumented with pressure transducers of
component according to the actual component layout in the model OMGA PX771A-100DI with +3 Pa accuracy to
system. The accuracy of the system simulation is dependent on measure the pressures in the low and high pressure sides of the
the correctness of the component models and their integration. system; thermocouples with +0.28C accuracy to measure
Inputs to the system model include mass flow rate and temperatures at various points in the refrigeration cycle, before
temperature of chilled fluid at inlet to the evaporator and air at and after each major component; thermocouples also with
inlet to the condenser, heat input to generator, UA values for +0.2 C accuracy to measure the temperature of the secondary
the heat exchange components and degrees of superheating fluid at inlet and outlet of the evaporator coil; a flow meter of
and subcooling of the refrigerant at evaporator and condenser model KROHNE IFC 010D with an accuracy 0.2% to measure
outlet, respectively. As a first step, assumptions are made for the flow rate of the secondary fluid through the evaporator
the pressures on both the high and low pressure sides of the coil.
system, and the mass flow rate and mass concentration of The experiments were carried out over a range of ambient
ammonia in the weak solution. Based on these initial assump- temperatures between 16 + 0.58C and 17 + 0.58C and chilled
tions, the equations for each component model on the high water flow temperatures between 228C and 2128C at a con-
pressure side are solved through an iterative process around the stant secondary fluid flow rate of 0.67 kg/s. Figure 6 shows the
degree of subcooling in the condenser as shown in Figure 4. temperature test results for the ammonia– water solution at the
Through each iteration the pressure in the high pressure side inlets and outlets of generator and absorber. In the figure, T1,
of the system is adjusted until the actual and desired values of T2, T4 and T7 are referred to the solution temperatures at the
subcooling agree to the required degree of accuracy. A similar generator inlet, generator outlet, absorber outlet and absorber
procedure is then followed for the evaporator model where the cooler inlet, respectively.
pressure of the low pressure side of the system is adjusted until Parameters used for model validation and component UA
the actual and set value of superheat agree to the desired value adjustment are pressures in the low and high pressure
degree of accuracy. Once this is achieved, the model simulates sides of the system, cooling capacity and cooling COP. Figure 7
the processes in the absorber and determines the concentration shows a comparison between the experimental and simulation
and enthalpy of the weak solution at the absorber outlet. If the results for the pressure in the high pressure (generator) side of
enthalpy does not correspond to the saturation enthalpy at the the system for a range of ambient and chilled secondary fluid
absorber pressure, the concentration of the weak solution is outlet temperatures. It can be seen that the pressure in the gen-
adjusted until the weak solution at the absorber outlet is satu- erator side of the system increases with increasing chilled fluid
rated liquid. Once this is achieved, the solution is further sub- outlet temperature. Higher ambient temperatures also lead to
cooled by the air-cooled heat exchanger. The pump model slightly higher pressures on the high pressure side of the system.
determines the input power, enthalpy and temperature of the Simulation accuracy for the high pressure side is quite good.
weak solution entering the generator. If the concentration, Similar characteristics were obtained for the pressure in the
temperature and mass flow rate of the weak solution entering low pressure side of the system. As can be seen from Figure 8,
the generator at point 1 are not close to the assumed or calcu- higher chilled fluid outlet temperatures lead to higher press-
lated values from the last iteration, then the model iterates on ures. It can also be seen that ambient temperature has less
these values until convergence is achieved. effect on the low pressure side of the system. At low secondary
fluid flow temperatures, below 258C, the simulation accuracy
is very good. At higher fluid flow temperatures, experimental
3 MODEL VALIDATION and simulation results begin to diverge. This may be owing to
the fixed UA values used in the simulation.
The gas-fired ammonia– water absorption chiller, shown sche- Figures 9 and 10 show the comparisons between the cooling
matically in Figure 1, was used to obtain test results to validate capacity and cooling COP, respectively, determined from tests

72 International Journal of Low-Carbon Technologies 2009, 4, 68– 77

Downloaded from https://academic.oup.com/ijlct/article-abstract/4/2/68/745728

by guest
on 08 May 2018
 Low-temperature ammonia – water absorption refrigeration system

Figure 4. Flowchart of system simulation.

and simulation over a range of chilled fluid outlet temperatures experiments, the ambient temperature which is quite impor-
and ambient temperatures. The scatters that can be observed in tant to the system performance varied between 15.58C and
the experimental results are mainly because of the variation in 17.58C. To investigate the performance of the system over a
ambient temperature during the tests and the slow response of wider range of ambient temperatures likely to be encountered
the system to variations in both the ambient temperature and in practice, simulations were carried out with ambient temp-
chilled fluid temperature. Despite the scatters, however, there eratures between 108C and 308C and chilled fluid flow temp-
are clear trends that show both cooling capacity and coling eratures between 2308C and 58C.
COP of the system to increase with increasing chilled fluid Figure 11 shows the variation of ammonia concentration in
outlet tempertaure at specific ambient temperature. This is the weak and strong solution with chilled fluid flow tempera-
mainly because of the higher evaporating temperature, which ture and ambient temperature. It can be seen that both the
leads to a higher refrigerant density and hence higher refriger- strong and weak solution concentrations increase with chilled
ant flow rate in the system. fluid temperature.
As the chilled fluid tmperature increases, the presures in
both the low and high pressure sides of the system increase as
4 MODEL APPLICATION shown in Figure 12. In addition, it shows that an increase in
the ambient temperature leads to an increase in the pressure in
The validated model can be used to explore the performance the condenser and the generator but has little effect on the
of the absorption unit at various operating conditions. In the pressure in the evaporator and absorber.

International Journal of Low-Carbon Technologies 2009, 4, 68– 77 73

Downloaded from https://academic.oup.com/ijlct/article-abstract/4/2/68/745728

by guest
on 08 May 2018
 Y.T. Ge et al.

Figure 5. Schematic diagram of the experimental facility.

Figure 6. Tested ammonia water solution temperatures at generator and Figure 8. Comparison between simulated and experimental pressures in the
absorber inlets and outlets. low pressure side of the system.

Figure 7. Comparison between experimental and simulation results for the Figure 9. Comparison between experimental and simulation results for
high pressure side of the system. cooling capacity.

74 International Journal of Low-Carbon Technologies 2009, 4, 68– 77

Downloaded from https://academic.oup.com/ijlct/article-abstract/4/2/68/745728

by guest
on 08 May 2018
 Low-temperature ammonia – water absorption refrigeration system

Figure 10. Comparison between test and simulation results for coefficient of Figure 13. Influence of chilled fluid flow temperature and ambient
performance. temperature on refrigerant and strong solution flow rates.

Figure 11. Variation of ammonia concentration in strong and weak solution

with chilled fluid and ambient temperature. Figure 14. Influence of chilled fluid flow temperature and ambient
temperature on cooling capacity.

flow rate of the strong solution decreases. An increase in the

ambient temperature, at constant generator heat input, causes
a reduction in the refrigerant mass flow rate, because of its
effect on condensing pressure and temperature and the
refrigerant enthalpy entering the evaporator.
The increase of refrigerant flow rate with chilled fluid flow
temperature leads to an increase in the cooling capacity as
shown in Figure 14. It can also be seen that an increase in the
ambient temperature causes a reduction in the cooling capacity
because of its influence on the condensing pressure and temp-
erature and the refrigerant enthalpy entering the evaporator. At
constant energy input to the generator, the increase in the
cooling capacity of the system leads to an increase in the COP
Figure 12. Variation of pressure in the high and low pressure sides of the
as shown in Figure 15.
system with chilled fluid outlet temperature and ambient temperature. The chilled water outlet temperature of the chiller is con-
trolled through the on – off control of the gas supply and heat
input to the generator. To investigate the effect of heat input
Figure 13 shows the variation of mass flow rate of refriger- on the performance characteristics of the system, the model
ant and strong solution with chilled fluid flow temperature was run for a range of heat inputs and constant chilled fluid
and ambient temperature. It can be seen that as the chilled and air flow rates and temperatures at inlet to the evaporator
fluid and hence evaporating temperature increase, the refriger- and condenser, respectively. The results for a chilled fluid inlet
ant mass flow rate increases and as a consequence, the mass temperature of 08C, constant chilled water mass flow rate of

International Journal of Low-Carbon Technologies 2009, 4, 68– 77 75

Downloaded from https://academic.oup.com/ijlct/article-abstract/4/2/68/745728

by guest
on 08 May 2018
 Y.T. Ge et al.

system. Some meaningful results can be obtained from the

simulation with the validated model, which can help to exten-
sively understand the performance of the absorption system:

† The concentrations of both strong and weak solutions

increase with higher chilled fluid outlet temperature at con-
stant ambient air temperature but decrease with higher
ambient air temperature at fixed chilled fluid outlet temp-
erature. When ambient air temperature varies from 108C to
308C and the chilled fluid outlet temperature changes from
2308C to 58C, the concentrations of strong and weak sol-
utions fluctuate from 0.1% to 27.4% and from 22.4% to
58.9%, respectively.
Figure 15. Variation of cooling coefficient of performance with ambient and
† The ambient air temperature has little effect on pressures in
chilled fluid outlet temperature. the evaporator and absorber.
† At constant ambient air temperature, the higher chilled fluid
outlet temperature leads to increased refrigerant mass flow
Table 1. Effect of variation of heat input to the generator on system rate but decreased strong solution mass flow rate. While at
performance constant chilled fluid outlet temperature, the higher ambient
Qgen Pg Pa _ r (kg/h)
m js jw Qe COP temperature causes decreased refrigerant mass flow rate but
(kW) (bar) (bar) (%) (%) (kW) increased strong solution mass flow rate. When ambient air
temperature varies from 108C to 308C and the chilled fluid
20 11.25 2.87 36.07 25.69 50.54 11.26 0.56
22 11.89 2.78 38.71 21.51 49.62 12.15 0.55 outlet temperature changes from 2308C to 58C, the mass
24 12.45 2.69 41.32 17.03 48.74 12.96 0.54 flow rates of refrigerant and strong solutions fluctuate from
26 12.94 2.62 43.90 12.26 47.89 13.68 0.53 26.3 kg/h to 47.8 kg/h and from 81.7 kg/h to 60.4 kg/h,
28 13.36 2.56 46.44 7.19 47.07 14.31 0.51 respectively.
30 13.70 2.51 48.94 1.82 46.28 14.85 0.49
† The increase of generator heat input from 20 kW to 30 kW
COP, coefficient of performance. greatly decreases the concentration of strong solution (from
25.69% to 1.82%) but has little effect on the concentration
0.7 kg/s, constant ambient temperature of 108C and variation of weak solution (from 50.54% to 46.28%) and eventually
of heat input to the generator from 20 kW to 30 kW are shown increases the cooling capacity from 11.26 kW to 14.85 kW
in Table 1. but slightly decreases the COP from 0.56 to 0.49.
It can be seen that increasing the heat input to the generator
will cause an increase in the generator pressure and a reduction
in the absorber pressure. This is a result of the increase in the
refrigerant flow rate in the system which will also result in a NOTATION
significant reduction in the concentration of refrigerant in the
strong solution and a slight reduction in the concentration of CHP combined heat and power
refrigerant in the weak solution. The higher heat input and COP coefficient of performance
increased refrigerant flow rate will cause an increase in the Dtsc refrigerant subcooling of condenser outlet (K)
cooling capacity of the system and a reduction in the Dtsc specified refrigerant subcooling of condenser outlet
chilled fluid flow temperature. The lower evaporating tempe- (K)
rature and pressure will also lead to a reduction in the COP of Dtsh refrigerant superheating of evaporator outlet (K)
the system. Dtshs specified refrigerant superheating of evaporator outlet
(K)
f ratio of cycle
h enthalpy (J/kg)
5 CONCLUSIONS m_ mass flow rate (kg/s)
LMTD logarithmic mean temperature difference (K)
A model of an ammonia – water absorption refrigeration Mref mass flow rate of refrigerant (kg/h)
system has been developed using the TRNSYS simulation Mfs mass flow rate of strong solution (kg/h)
environment. The model was validated against test results over Pa solution pressure inside absorber (bar)
a range of chilled fluid flow temperatures and ambient temp- Pg solution pressure inside generator (bar)
eratures. These, together with the heat input to the generator Q heat transfer rate (W)
are the main parameters that influence the performance of the q heat transfer effect (J/kg)

76 International Journal of Low-Carbon Technologies 2009, 4, 68– 77

Downloaded from https://academic.oup.com/ijlct/article-abstract/4/2/68/745728

by guest
on 08 May 2018
 Low-temperature ammonia – water absorption refrigeration system

R ratio of return flow [2] Herold KE, Radermacher R, Klein S. Absorption Chillers and Heat Pumps.
T, t temperature (8C, K) CRC Press, 1996.
Tamb ambient air temperature (8C) [3] Lazzarin RM, Gasparella A, Longo GA. Ammonia-water absorption
machines for refrigeration: theoretical and real performances.
DT temperature difference (K)
Int. J. Refrigeration 1996;19:239– 46.
UA overall conductance (W/K)
[4] Lazzarin RM, Gasparella A, Romagnoni P. Experimental report on the
j concentration (% by mass) reliability of ammonia-water absorption chillers. Int. J. Refrigeration
1996;19:247– 56.
SUBSCRIPTS [5] Horuz I, Callander TMS. Experimental investigation of a vapor absorption
refrigeration system. Int. J. Refrigeration 2004;27:10– 16.
4l saturated liquid at point ‘4’ in Figure 1 [6] Grossman G, Perez-Blanco H. Conceptual design and performance analy-
a absorber, air sis of absorption heat pumps for waste heat utilization. Int. J. Refrigeration
cf chilled fluid 1982;5:361– 70.
[7] Grossman G, Gommed K, Gadoth D. A computer model for simulation of
e evaporator
absorption systems in flexible and modular form of absorption systems in
g generator
flexible and modular form (ORNL/Sub/90 ¼ -89673). Oak Ridge National
in inlet Laboratory, Oak Ridge, TN, 1991.
out outlet [8] Grossman G. Modular and flexible simulation of advanced absorption
r rectifier systems. In: Proceedings of the International Absorption Heat Pump
s strong solution (rich in absorbent) Conference, Vol. 31, AES, New Orleans, LA, 1994, 345– 52.
p pump [9] Grossman G, Wilk M. Advanced modular simulation of absorption
w weak solution (weak in absorbent) systems. Int. J. Refrigeration 1994;17:231– 44.
[10] Grossman G, Zaltash A. ABSIM-modular simulation of advanced absorp-
tion systems. Int. J. Refrigeration 2001;24:531 – 43.
[11] Chua HT, Toh HK, Malek A, Ng KC, Srinivasan K. A general thermodyn-
ACKNOWLEDGEMENTS amic framework for understanding the behaviour of absorption chillers.
Int. J. Refrigeration 2000;23:491– 507.
The authors wish to acknowledge the Food Technology Unit of [12] Goel N, Goswami DY. Analysis of a counter-current vapour flow absorber.
DEFRA for their financial support for this project and the con- Int. J. Heat Mass Transfer 2005;48:1283– 92.
tribution of the industrial collaborators, Bond Retail Services [13] Kaynakli O, Horuz I. Comparison of parallel and counter flow coil absor-
Ltd, Apex Air Conditioning, Doug Marriott Associates and ber performance. Int. J. Heat Mass Transfer 2006;33:211 –23.
Bowman Power. [14] Kwon K, Jeong S. Effect of vapor flow on the falling-film heat and mass
transfer of the ammonia/water absorber. Int. J. Refrigeration 2004;27:
955 – 64.
REFERENCES [15] Conde M. Thermophysical properties of fNH3 þ H2Og solution for the
industrial design of absorption refrigeration equipment. M.Conde
Engineering, 2006.
[1] Faramarzi PT, Walker DH. Investigation of secondary loop supermarket
[16] TRNSYS. A Transient System Simulation program. TRNSYS 16, 2005.
refrigeration systems. Consultant Report 500-04-013, March 2004.

International Journal of Low-Carbon Technologies 2009, 4, 68– 77 77

Downloaded from https://academic.oup.com/ijlct/article-abstract/4/2/68/745728

by guest
on 08 May 2018