DESIGN CONSIDERATIONS FOR

REFRIGERATION CYCLES

by

MATTHIAS HARTMUT RAUCK

A thesis submitted in partial fulfillment of

the requirements for the degree of

,MASTER OF SCIENCE

(Chemical Engineering)

at the

UNIVERSITY OF WISCONSIN-MADISON
1992
 ABSTRACT

Refrigeration equipment consumes worldwide a major part of the annual electrical energy

production. The US domestic refrigerator industry is interested in methods which

improve refrigerator performance in order to meet US government energy standards. It is

necessary to find suitable replacements for refrigerants which are now known to deplete

the ozone layer or which contribute to the greenhouse effect. A possible solution may be

the use of non-azeotropic refrigerant mixtures (NARMs) instead of the common

chlorofluorcarbons (CFCs).
The isobaric phase change for a NARM does not occur isothermally as for pure
refrigerants, but over a temperature range. This behavior offers a potential improvement

of the coefficient of performance (COP) in a refrigeration cycle. Two models are

developed to investigate the potential improvements of the COP in an ideal refrigeration
cycle. These models allow the refrigerant to change temperature throughout the heat

transfer process. The models account for irreversibilities due to heat transfer rate
limitations from the refrigeration cycle to external fluids. The COP obtained offers a

realistic design goal for the refrigeration system. Design guidelines which lead to the
maximum possible COP for such a system are established.

ii
 A standard vapor compression cycle is then investigated with another simulation
model. NARMs consisting of real refrigerants are used to evaluate the performance of

the cycle and these results are compared to the results from the previous developed ideal

cycle.

iii
 ACKNOWLEDGMENT

Special thanks go to my academic advisors Professor Sanford A. Klein and Professor

William A. Beckman. I am thankful to Sandy for finding an interesting and suitable
research project for me as I decided, after my first semester, to go for a Master of Science
in Chemical Engineering. He and Bill believed in me, made the project possible and

supported me whenever I needed help or motivation. Osama A. Ibrahim deserves a

'special' thank you. He offered a great deal of help in tackling my first problems before

he left the Solar Energy Laboratory.

My time in the SEL was enjoyable and intense at all times. I am going to miss the

stimulating environment: the seminars, the discussions with other students all over the
globe, my own office on campus and all the social events. The Chemical Engineering
Department, which was very challenging for me, 'provided' me also with a lot of work,

but also with even more friends. When do I see Louis, Stephan, Jon, Christine, Brian,
Dan, Jdrg, Svein, Gerold, Oystein, Jtirgen, Nate, Jeff, Bob, Paul,.... and Bob, Frank,

Rahmat, Rajesh and the fun guys from the Nottingham Coop, again?

I want to thank my brother Thomas and his wife Cilly for visiting me, sharing some

american experiences and taking care of all these nasty things like going through all the
formalities with the University of Hannover and a lot more.

iv
My stay in Madison was made possible by the German Academic Exchange Service
(DAAD) and the Institut fUr Thermodynamik at the University of Hannover in Germany.

I am grateful to Professor H. D. Baehr and Dipl. Ing. Dipl. Ing.(MS) Frank-Detlef Drake
for putting their personnel effort into this program. Frank also spent 1 1/2 years in
Madison at the Solar Energy Laboratory. He knew how great the study in Madison is

and initiated the program. Thanks a lot, Frank, for making this great study abroad
experience possible.

Last, but not least, I want to thank Klaus. We became great buddies during our stay in
America. He was a wonderful friend and listener when the time in Madison, with all the

ups and downs, got a little bit rougher for me.

Finally I dedicate this 'work' to my father Wolfgang who always believed in me and who
is probably now looking down from far far above to me.

My time was a great experience and I will always keep good memories of Madison

and, of course, I will be back.

 TABLE OF CONTENTS

ABSTRACT ll

ACKNOWLEDGMENTS iv
TABLE OF CONTENTS vi
LIST OF TABLES x
LIST OF FIGURES x
NOMENCLATURE xv

CHAPTER 1 INTRODUCTION

1.1 Background and Scope of Study 1

1.2 Literature Search 3
1.3 Review of the Vapor Compression and the Carnot
Refrigeration Cycle 4
1.3.1 Vapor compression Refrigeration Cycle 5
1.3.2 Carnot Refrigeration Cycle 8

vi
1.4 Limitations of the Carnot COP as a Realistic Design Goal for
Refrigeration Systems 9
1.4.1 Background 9
1.4.2 External Boundary Conditions 11
1.4.3 Identification of the optimum Refrigeration Cycle
using a pure Refrigerant 12

CHAPTER 2 IDENTIFICATION OF THE

OPTIMUM REFRIGERATION CYCLE 17

2.1 Derivation of a Numerical Simulation Model 18

2.2 Results obtained from the Finite Difference Model 22
2.2.1 Necessary number of individual Carnot Cycle 22
2.2.2 Individual Cooling Load for each Carnot Cycle 25
2.2.3 Optimized Allocation of the Heat Exchanger Conductance 26
2.2.4 Optimized Allocation of the Heat Capacitance Rates 29
2.2.5 Optimum Refrigeration Cycle Shape 32
2.3 Chapter Summary 36

CHAPTER 3 ANALYTICAL SIMULATION MODEL 37

3.1 Derivation of an Analytical Model 38

3.1.1 General Expression for the COP 38
3.1.2 Partially Optimized COP 44
3.2 Results from the Analytical Model 45
3.2.1 Variation of the external inlet Temperatures
TL,mn and TH~m 45

vu
 3.2.2 Variation of the Gliding Temperatures Differences
GTDl,refr and the Cooling Load 48
3.2.2.1 Gliding Temperatures Differences GTDlrefr at constant
Heat Capacitance Rates 48
3.2.2.2 Influence of Cooling Load 50
3.2.2.3 Gliding Temperatures Differences GTDl,refr at
constant Cooling Load 52
3.2.3 Influence of total Heat Exchanger Conductance 54
3.2.4 Influence of total Heat Capacitance Rate 54
3.2.5 Optimum Refrigerant Heat Capacitance Rate 56
3.3 Comparison of the Numerical and Analytical Model 58
3.3.1 COP of the different Models 58
3.3.2 Accuracy of the Numerical Model 63
3.3.3 Allocation of Heat Exchanger Conductance and
Heat Capacitance Rate for an isothermal process 64
3.3.3.1 Heat Exchanger Conductance 64
3.3.3.2 Heat Capacitance Rates 66
3.4 Comparison of different Refrigeration Cycles 69
3.4.1 Carnot Refrigeration Cycle 69
3.4.3 Brayton Refrigeration Cycle 70
3.4.4 Optimum Cycle 70

3.5 Chapter Summary 71

CHAPTER 4 NARMs IN A VAPOR COMPRESSION

CYCLE 73
4.1 Description of NARM 74
4.1.1 Phase Change Process of a NARM 75

viii
 4.1.2 Components for a NARM 75
4.2 Description of the Simulation Model Cyclell 76
4.2.1 Heat Exchanger Model 76
4.2.2 Refrigerant Properties 77
4.3 Simulations with Cyclell 78
4.3.1 Base Case System 79
4.3.2 Optimum Composition of the NARM 80
4.3.3 Accompanying Superheat 81
4.3.4 Deviation of the optimum Temperature matching 83
4.3.5 Heat Exchanger Conductance Allocation 87
4.3.6 Comparison of Analytical Model and Results from
Cycle 11 89
4.4 Chapter Summary 92

CHAPTER 5 CONCLUSIONS & RECOMMENDATIONS 94

5.1 Conclusions 94
5.2 Recommendations for Future Work 97

APPENDIX Finite Difference Model (FDM) and

Analytical Solution Model (AM) 98
Bibliography 105

ix
 LIST OF TABLES

Table 4.1 Operating conditions for the standard vapor compression cycle
Table 4.2 COPs calculated from Cycle 11 and AM

LIST OF FIGURES

Figure 1.1 Vapor compression refrigeration cycle

Figure 1.2 Temperature-Entropy diagram for an ideal vapor compression cycle
Figure 1.3 Pressure-Enthalpy diagram for a vapor compression cycle
Figure 1.4 Temperature-Entropy diagram for an Carnot cycle
Figure 1.5 Refrigeration cycle with external streams
Figure 1.6 Schematic of a cooling engine
Figure 1.7 Temperature vs. Entropy for a Carnot cycle coupled to a heat sink and heat
source with infinite heat capacitance rates (left) or finite
heat capacitance rate (right)
Figure 2.1 Refrigeration cycle broken into several individual Carnot cycles in
sequence
Figure 2.2 Individual Carnot cycle coupled to external streams
Figure 2.3 COPn vs. number of individual Carnot cycles for different high external
stream inlet temperatures TH,in and low external stream inlet temperature of
TL,in = 273 K, UAL =UAH=4 kW/K, CLC=(H =1 kW/K, QL = 10 kW
Figure 2.4 COPn/COP 1 vs. number of individual Carnot cycles for different high
external stream inlet temperatures TH,in and low external stream inlet
temperature of TL,in = 273 K, UAL = UAH = 4 kW/K, CL = CH = 1 kW/K,
QL= 10kW

Figure 2.5 Optimized individual cooling load QL,i of each individual Camot cycle for
TL,in = 273 K, TH,in = 313 K, UAL = UAH = 4 kW/K,
CL = CH = 1 kW/K, QL = 10 kW

Figure 2.6 COP 10 vs. heat exchanger conductance UA fractions for different heat
capacitance ratios with UAtotal = 1 kW/K, C L = 1 kW/K, TL,in = 273 K,
THn = 313 K, QL = 10 kW

Figure 2.7 COP10 vs. heat exchanger conductance UA fractions for different heat
capacitance ratios with UAtota1 = 1 kW/K, Ctotal = 2 kW/K, TLin = 273 K,
THn = 313 K, QL,i

Figure 2.8 COP 10 vs. heat capacitance ( fractions for different heat exchanger
conductance ratios with UAtotai = 1 kW/K, Ctotal = 2 kW/K,
TL,in = 273 K, TH,in = 313 K, QL = 10 kW

Figure 2.9 COP 10 vs. heat capacitance rate C fractions for different heat exchanger
conductance ratios with UAL = 0.5 kW/K, Ctotal = 2 kW/K,
TL,in =273 K, TH,in = 313 K, QL = 10 kW

Figure 2.10 COP 1 o vs. heat capacitance C fractions for different heat exchanger
conductance ratios with UAtotal = 2 kW/K, Ctotal = 2 kW/K,
TL,in = 273 K, TH,in = 313 K, QL = 10 kW

Figure 2.11 COP 1Ovs. heat capacitance C fractions for different heat exchanger
conductance ratios with UAtota 1 = 2 kW/K, Ctotal = 0.5 kW/K,
TLin = 273 K, TH,in = 283 K, QL = 10 kW

Figure 2.12 Temperature vs. Entropy transfer rate with low and high external stream
inlet temperature of TL,in = 273 K, TH,in = 313 K respectively,
UAL = UAH = 1 kW/K, CL = CH='1 kW/K, QL = 10 kW

xi
Figure 2.13 Temperature vs. Entropy transfer rate with low and high external stream
inlet temperature of TL,in = 273 K, TH,in = 313 K respectively,
UAL = UAH = 1 kW/K, CL = (CH= 0.1 kW/K, OL = 10 kW
Figure 2.14 Temperature vs. entropy transfer rate diagram with low and high external
stream inlet temperature of TL,in = 273 K, TH,in = 313 K respectively,
UAL = UAH = 1 kW/K, (L = CH = 10 kW/K, QL = 10 kW
Figure 3.1 Schematic Temperature - Entropy diagram

Figure 3.2.a Potential COP improvement Q vs. temperature difference THin - TL,in for
different heat exchanger conductances, low external stream inlet
temperature of TL,in = 273 K, CL = CH = 1 kW/K, 0L = 10 kW

Figure 3.2.b Potential COP improvement Q vs. temperature difference TH,in - TL,in for
different heat exchanger conductances, low external stream inlet
temperature of TL,in = 273 K, CL = CH = 1 kW/K, (L = 10 kW

Figure 3.3 COPCO, COP 1 and potential COP improvement 92 vs. temperature difference
TH,in - TLin with external stream inlet temperature of TL,in = 273 K,
TH,in = 313 K and UAL = UAH = 2 kW/K, CL = CH = 1 kW/K

Figure 3.4 COPOO, COP 1 and potential COP improvement Q vs. temperature difference
THin - TL,in with external stream inlet temperature of TL,in = 273 K,
THin = 293 K and UAL = UAH = 2 kW/K, CL = CH = 1 kW/K

Figure 3.5 COP. vs. cooling capacity QL with external stream inlet temperatures of
TL,in = 273 K, THin = 313 K and UAL = UAH = 2 kW/K, CL = 1 kW/K

Figure 3.6 0 vs. cooling capacity OL with external stream inlet temperatures of
TL,in = 273 K, THin = 313 K , UAL = UAH = 2 kW/K, CL = 1 kW/K

Figure 3.7 COP., COP 1 and potential COP improvement Q vs. temperature
difference Tlin - TL,out with external stream inlet temperature of
TL,in = 273 K, TH,in = 313 K and UAL AH = 2 kW/K, QL = 10 kW

Figure 3.8 COPoo, COP 1 and potential COP improvement Q vs. temperature
difference Tl,in - TL,out with external stream inlet temperature of
TL,in = 273 K, TH,in = 293 K and UAL = UAH = 2 kW/K, QL = 10 kW

xii
Figure 3.9 COP,, and potential COP improvement Q vs. UAtotal with external stream
inlet temperature of TL,in = 273 K, TH,in = 293 K and UAL = UAH,
CL = CH = 1 kW/K, QL = 10 kW/K or QL =50kW/K

Figure 3.10 COP., and potential COP improvement 02 vs. Cexttotal with external
stream inlet temperature of TL,in = 273 K, THin = 293 K and
UAL = UAH = 2 kW/K, CL = CH, QL = 10 kW/K or QL = 50 kW/K
Figure 3.11 Optimum refrigerant heat capacitance rate Crefr,opt vs. external heat
capacitance rate ratio CH/CL for different fixed 1CL and UAL = UAH

Figure 3.12 COP vs. heat constant capacitance rate of the refrigerant Crefr for
TL,in = 273 K, THin = 313 K, UAL = UAH = 2 kW/K, CL = CH = 1 kW/K

Figure 3.13 COP vs. heat constant capacitance rate of the refrigerant Crefr for
TL,in = 273 K, TH,in = 313 K, UAL = UAH = 2 kW/K, CL = 1 kW/K and
CH =4 kW/K

Figure 3.14 COP vs. heat constant capacitance rate of the refrigerant Crefr for
TL,in = 273 K, THIn = 313 K, UAL = UAH = 2 kW/K, CL - 1 kW/K and
CH = 20 kW/K

Figure 3.15 COP vs. heat constant capacitance rate of the refrigerant Crefr for
TL,in = 273 K, TH,in = 313 K, UAL = UAH =2 kW/K, CL = 1 kW/K and
CH = 0.25 kW/K
Figure 3.16 COPn/COP 1 vs. different low external inlet temperatures TL,in for
different cooling loads Oj for TH,in = 313 K, UAH = 2 kW/K,
UAH = 4 kW/K, C.L = 1 kW/K and CH = 5 kW/K

Figure 3.17 COP., and COP 1 vs. heat exchanger conductances fractions UAL for
different external heat capacitance rates ratios with Cexttotai = 2 kW/K,
UAtotal = 4 kW/K, TH,in = 293 K, TLin = 273 K and QL = 20 kW

Figure 3.18 Optimum allocation of the total heat exchanger conductance UAtotal for a
single Carnot cycle system vs. external heat capacitance rates ratios CH/CL
with Cext,1totai = 2 kW/K, T,in = 293 K, TL,in = 273 K, QL =f20 kW and
varying UAtotai

xiii
Figure 3.19 COP. and COP 1 vs. heat capacitance rates fractions CL for different heat
exchanger conductance ratios UAH/UAL With Cexttotai = 2 kW/K,
UAtotai = 4 kW/K, TH,in = 293 K, TL,in = 273 K and QL = 20 kW

Figure 3.20 Optimum allocation of the total heat capacitance rates Cext,totai for a single
Carnot cycle system vs. external heat exchanger conductance ratios
UAHIUAL With UAtotal = 4 kW/K, THin = 293 K, TL,in = 273 K,
QL = 20 kW and varying Cext,total

Figure 4.1 Evaporation process of a NARM in a Temperature-Concentration diagram

Figure 4.2 COP vs. mole fraction fR22 for a refrigerant mixture of R22/R141b and
different polytropic compressor efficiencies
Fraction superheat fsh vs. mole fraction fR22 for a refrigerant mixture of
Figure 4.3
R22/ R14lb and different polytropic compressor efficiencies
Figure 4.4 Location of gliding temperature differences

Figure 4.5 Deviation of optimum temperature matching ji vs. mole fraction fR22
for a refrigerant mixture of R22/R14lb and different polytropic compressor
efficiencies
Figure 4.6 Temperature-Entropy diagram for the system with a polytropic compressor
efficiency of rl = 0.5

Figure 4.7 COP vs. heat exchanger conductance fraction fUA,evap for a refrigerant
mixture of R22/R141b and different polytropic compressor efficiencies
Figure 4.8 Temperature-Entropy diagram for the system with a polytropic
compressor efficiency of 'q = 0.5 and optimized UA allocation

xiv
 NOMENCLATURE

ROMAN LETTER SYMBOLS

Symbol Definition SI-Units

C heat capacitance rate (mass flow rate - specific

heat product) [kW/K]

Cp specific heat capacity at constant pressure [kJ/kgK]

COP Coefficient of Performance

CSD Carnahan-Starling-DeSantis equation of state
EES Engineering Equation Solver

f fraction, mole fraction

GTD gliding temperature difference [K]
h specific enthalpy [kJ/kg]

rh mass flow rate [kg/s]

NTU number of transfer units

p pressure [N/m 2]

Q heat transfer rate [kW]

S entropy [kJ/K]

xv
 S entropy transfer rate [kW/K]

T temperature [K]

AT temperature difference defined in Equation (1.8) [K]

AT temperature difference defined in Equation (3.33) [K]

UA Heat Exchanger Conductance [kW/K]

W Power [W]
X quality
y parameter for the CSD equation of state defined in
Equation (4.6)

GREEK LETTER SYMBOLS

Symbol Definition

(3 parameter for the calculation of the COP defined in Equation (3.27)

A difference
heat exchanger effectiveness

polytropic compressor efficiency

X parameter for the calculation of the COP defined in Equation (3.29)

parameter for the calculation of the COP defined in Equation (3.35)

potential improvement of the COP defined in Equation (3.30)

xdeviation of the optimum temperature deviation defined in equation (4.7)

xvi
 SUBSCRIPTS

Symbol Definition

C compressor

cond condenser

evap evaporator

ext external stream

H high, referring to external stream condition in high temperature heat
exchange
h high, referring to refrigerant condition in high heat exchanger, liquid

hx heat exchanger

i cycle
in inlet
L low, referring to external stream condition in low temperature heat
exchanger
1 low, referring to refrigerant condition in low heat exchanger; liquid
Im logarithm mean

mean mean
min minimum

max maximum
n number of individual Carnot cycles

opt optimum

out outlet
refr refrigerant

xvii
sh superheat
T turbine

total total
v vapor

x local condition in the heat exchangers

1 one Carnot cycle
10 ten Carnot cycles in sequence

00 COP calculated from Analytical Simulation Model

SUPERSCRIPTS

Symbol Definition

partially optimized COP from analytical simulation model

xviii
 CHAPTER
ONE

INTRODUCTION

1.1 BACKGROUND AND SCOPE OF STUDY

Various economic and ecological reasons make it necessary to look more in detail at the
potential performance improvements of refrigeration systems. More energy efficient

systems have to be developed because of stricter energy standards issued all over the
world. American refrigerator manufacturers are being forced by the government and by
environmental and economic interests to produce more energy efficient systems and to

eliminate chlorofluorcarbons (CFCs) [1].

More energy efficient systems are needed to curb the greenhouse effect. The
greenhouse effect is caused by an increase of carbon dioxide (C02) and other gases in the

atmosphere in addition to refrigerants. These substances prevent radiation emitted or

reflected from the earth to leave the atmosphere and hence the atmospheric temperature
may increase. C02 is released in all energy producing processes based on fossil fuel.
 2

Consequently a decrease of the energy consumption and of the use of these refrigerants is
desired.
Substitutes for CFCs must be used. It is now well known that the ozone layer
protecting the earth from excessive ultraviolet radiation is already dangerously depleted

primarily by CFCs. CFCs, the main culprit that causes ozone destruction, work their

way slowly to the stratosphere, break apart, and release ozone destroying chlorine. In
many parts of the world, especially in the southern hemisphere, the uv radiation is

already above physical tolerable amounts. Especially the fully halogenated

chlorofluorcarbons like Ri1, R12, Ri13, Ri14 have a high ozone depleting potential.

Two of the most commonly used refrigerants, R11 and R12, are the major of ozone

destruction. These chemicals are used in small refrigeration systems, automobile air
conditioning and as a agent to blow plastic foam. These refrigerants have been used in
refrigeration systems for decades due to their desired characteristics like low toxicity,

non-flammability and their non-aggressive behavior against other design materials.

The purpose of this work is it to determine the optimum refrigeration cycle for a

vapor compression cycle. The maximum possible coefficient of performance (COP),

which may be used as a realistic design goal, will be identified. Design guidelines are

found for this cycle , which is first considered to be ideal with no irreversibilities. These
guidelines are then applied and checked for real vapor compression cycles using pure

refrigerants and mixtures.

The circumstances for which performance improvements may be obtained using

non-azeotropic refrigerant mixtures (NARMs) as the refrigerant will be identified. For a

NARM, the phase change does not occur isothermal as for pure refrigerants. This
temperature change is due to a composition change which accompanies the phase change
processes occurring in a refrigeration cycle. This property and other may be used to mix

suitable refrigerants for different applications.

1.2 LITERATURE SEARCH

Many studies on alternative refrigerants have been published in the last few years. The

use of NARMs as a refrigerant is considered by many authors, e.g., [8,9,10,11,12].

Atwood [8] reports that there is still insufficient experience with NARMs and that
refrigeration systems applying a NARM may require significant changes. It is most

likely that these changes result in refrigeration systems which are more costly than the

conventional in mass manufactured applications.

Kruse et al. [13] investigated a two-evaporator refrigeration system similar to a

household refrigerator. Theoretical and experimental investigations have been carried out

for the refrigerant mixtures R22/R114 and R13B1/R-114 and compared to R12. Kruse

calculated an increase of 18% and 20% of the COP of the mixtures R22/Rl14 and
R13Bl/R-114 respectively. The corresponding experiments leaded to improvements of
13% (R22/R114) and 15% (R13B1/R-114) compared to R12.
Domanski and McLinden [14] worked on refrigeration cycle simulations models

for the performance rating of refrigerants and their mixtures. Their model Cycle 11 is able
to predict the COP for refrigerants and their mixtures in a vapor compression cycle.
McLinden and Radermacher [15] describe methods for comparing the performance of
pure and mixed refrigerants in a vapor compression cycle. They computed the COPs for
the mixtures R22/R1l14 and R221R1 1 and reported that COP improvements may be
obtained for mixtures where the temperature change of the refrigerant mixture matches
with the temperature change of the heat sink or source.

Klein [16] pointed out that the Carnot COP for a refrigeration cycle does not
provide a realistic upper limit for design considerations. He derived another expression

for the COP of a simple internal reversible refrigeration cycle. This COP considers heat

transfer mechanism from and to the cycle which are a necessarily an irreversible process

and the analysis assumes that the heat transfer occurs isothermally.

1.3 REVIEW OF THE VAPOR COMPRESSION AND THE

CARNOT REFRIGERATION CYCLE

1.3.1 VAPOR COMPRESSION REFRIGERATION CYCLE

A review of the standard vapor compression cycle is presented [2,3]. Later in this work,

the use of a NARM will be investigated. The different components and the

thermodynamic basics are described and explained.

Figure 1.1 shows the schematic of a vapor compression refrigeration cycle

operating between a low temperature T1 and at a higher temperature Th. An accompanying

temperature-entropy and pressure-enthalpy diagram are presented in the Figures 1.2 and
1.3. The refrigerant is continuously circulating through this cycle and its different states

are described beginning at the inlet of the evaporator.

 tttt QHtttt

4444 QL4444f
Figure 1.1 Vapor compression refrigeration cycle

hL

Entropy
Figure 1.2 Temperature-Entropy diagram for a vapor compression cycle
The refrigerant enters the evaporator as a two phase liquid-vapor mixture at low quality at
state 4. It changes to vapor and leaves the evaporator at state 1 with a quality of x=1 or in

a slightly superheated state. Heat transfer in the evaporator occurs from an external cold

region at lower temperature as the refrigerant to the refrigerant. For a pure fluid, the
pressure and temperature remain approximately constant throughout the phase change,

but the temperature of the refrigerant rises the a superheated region. The refrigerant is

then compressed from state 1 to state 2, where it leaves the compressor as superheated
vapor at a higher pressure and temperature.

3 €2

Enthalpy

Figure 1.3 Pressure-Enthalpy diagram for a vapor compression cycle

The refrigerant passes from the compressor into the condenser, where it changes at
constant pressure to saturated liquid at a quality of x=1. The phase change in the

condenser occurs, as in the evaporator, isothermal and almost isobar. The refrigerant,
now at state 3, returns then into state 4 of the inlet of the evaporator by expanding
adiabatically through a expansion valve. This throttling process during which the
refrigerants expands from the condenser pressure to the evaporator pressure, takes place

at approximately constant enthalpy. In practice, irreversibilities occurs due to heat losses

during compression and expansion (non-adiabatic), pressure drops in the evaporator and

condenser, and additional heat- and pressure losses in the remaining lines.
As the refrigerant passes through the evaporator, heat transfer from the

refrigerated space into the cycle results in the vaporization of the refrigerant. This is the

refrigeration capacity or the cooling load and determined by the product of enthalpy

difference across the evaporator and the mass flow rate of the refrigerant.

QL = rfref" (hl- h4 ) (1.1)

The power requirement for the cycle is given by the product of the mass flow rate of the
refrigerant and the enthalpy difference between evaporator outlet and condenser inlet.

W=ithref•-(h 2 - hi) (1.2)

The rate of heat transfer from the condenser to the surroundings is given by

QH= rnrefl• (h2 - h 3 ) (1.3)

The throttling is assumed to be isenthalpic, but it is irreversible and there is an

accompanying entropy increase.
 h3 = h4 =(1.4)

The coefficient of performance (COP), is defined as the cooling load or cooling capacity

divided by the power supplied to the cycle. The COP may take values between zero and
infinite.

COP = QL _.hi- h4 (1.5)

-Vh 2 - hl

1.3.2 CARNOT REFRIGERATION CYCLE

The Carnot refrigeration cycle has no internal irreversibilities. Instead of the expansion

valve, as in the vapor compression cycle, an expander is applied to the system, so that the

energy which was dissipated in the expansion valve can now be gained as a power
output. Figures 1.4 illustrates the schematic of the Carnot cycle.

3 2

41

Entropy
Figure 1.4 Temperature-Entropy diagram for a Carnot cycle
The coefficient of performance for the Carnot cycle is determined by [2]

COP = T1 with T 4 = T 1 and T2 = T 3 (1.6)

T2 - T1

where T2 is the temperature at which heat is rejected from the cycle to an isothermal heat
sink and T1 is the temperature at which heat is added to the cycle from an isothermal heat

source.

1.4 LIMITATIONS OF THE CARNOT COP AS A REALISTIC

DESIGN GOAL FOR REFRIGERATION SYSTEMS

The Carnot COP does not consider heat transfer mechanisms, which are necessarily
irreversible. This reduces its usefulness as a realistic design goal for refrigeration

systems. A more realistic approach to determine the optimum COP of a refrigeration

system using a pure refrigerant is shown.

1.4.1 BACKGROUND
The design of vapor compression refrigeration cycle which attains the maximum possible

coefficient of performance (COP) for specified heat exchanger sizes, capacitance rates,

and external stream temperatures is of significant practical importance. The laws of

thermodynamics and heat transfer mechanisms are considered to identify a realistic design

goal for the COP of refrigeration cycles.

This COP differs from the Carnot COP, which is ordinarily used for the purpose

as a design goal for actual cycles. Carnot introduced the concept of reversibility and the
 10

principle that the thermal efficiency may be only expressed as a function of the heat

source and heat sink. Nevertheless, the Carnot COP, which assumes a thermodynamical
ideal cycle in which no irreversibilities exist and which yields the maximum COP, does

not provide a realistic upper limit for design considerations because of various reasons.
The reversible heat transfer processes, assumed in the Carnot analyses, do not

consider heat transfer mechanisms. Heat is added to the cycle in the evaporator by an

external stream, which is cooled as it flows through the evaporator; heat is rejected from

the cycle using an external stream, which is heated when it flows through the condenser.
Heat transfer at a finite rate is necessarily an irreversible process and unavoidable in a
refrigeration cycle. Further, heat exchangers represent a major size and cost constraint to

refrigeration cycles, but the Carnot cycle can not provide any useful information on heat

exchanger design.
The concept of constant temperature thermal reservoirs, as used in the Carnot

cycle analyses, has no direct parallel in practice. In reality, refrigeration cycles receive

and reject thermal energy from external streams. These streams, flowing at finite heat
capacitance rates and with known inlet temperatures, do not represent an isothermal heat

sink or heat source. This effect is not considered in the Carnot cycle analyses.
There is interest in non-azeotropic mixtures on refrigerants, for which the phase

change in the evaporator and condenser does not occur at a constant temperature.
Refrigerant mixtures offer the possibility of higher efficiencies and are possible

substitutes for CFCs. A simple determination of the maximum COP for these fluids is
desired and will be presented in the next chapter. This COP can not be calculated with a
simple Carnot cycle analysis because of the varying temperature of the refrigerant during
the phase changes and the heat exchanger limitations.
 11

1.4.2 EXTERNAL BOUNDARY CONDITIONS

The performance of the ideal vapor compression cycle, with no internal irreversibilities is

examined. The processes in the condenser and evaporator are assumed to be isobaric; the

compression process in the compressor and turbine is assumed to be isentropic. The

cycle is designed to provide a specified cooling load (Q to an external stream with the
heat capacity CL entering the evaporator at the temperature TL,in. Another stream with the

heat capacity CRH rejects the heat and enters the condenser and at the temperature TH,in.

Figure 1.5 shows the cycle examined.

WTIN, Turbine WC

Figure 1.5 Refrigeration cycle with external streams

 12

The evaluation of the heat transfer mechanism in the heat exchangers is based on the basic
heat transfer relations [4,5]. All the components of the refrigeration cycle are modeled

with the program Engineering Equation Solver (EES), [6] which is capable of solving
large systems of algebraic equations and has built-in mathematical and thermophysical

property functions.

1.4.3 IDENTIFICATION OF THE OPTIMUM REFRIGERATION CYCLE

USING A PURE REFRIGERANT
The Carnot cycle is used as a starting point for the determination of a realistic upper

bound on the COP

COP= QL - Tj(1.7)
WC-WVVT Th- T1

where QL is the cooling load (or cooling capacity), Vc and VT are the compressor and

turbine power respectively. T1 is the temperature at which the heat is transferred from the
refrigerated space through the low heat exchanger, the evaporator, to the cycle and Th is

temperature at which the heat is transferred from the cycle through the high heat

exchanger, the condenser, to the heat sink. The Carnot COP is the maximum COP
attainable for a refrigeration cycle with heat transfer at a constant temperature.
The determination of the Carnot COP is based only on the temperatures of the

working fluid (refrigerant) in the cycle. It does not take into account the cooling load or

any external conditions, like the external heat source (refrigerated space), external heat
sink and the heat transfer mechanism from the cycle to the surrounding.
 13

Th> TH OH from refrigeration cycle

Th temperature of condensing refrigerant

W supplied by auxiliar
Refrigeration
cycle

<TL
ATT,QL
temperature of evaporating refrigerant
from refrigerated space

ZL, temperature of heat source

Figure 1.6 Schematic of a cooling engine

In addition, it does not consider the heat capacitance rates CH and CL of the heat transfer

fluids in the condenser and evaporator, their overall heat transfer coefficients UA or the
effectiveness EL and EH of the heat exchangers. Klein [7] found a more realistic

expression for the Carnot COP in terms of the external conditions:

TL,in-AT (1.8)
COP=
TH,in - (TL,jn - AT)

where TL,in is the temperature at which the external fluid enters the low-temperature heat
exchanger and TH,in is the temperature at which the external fluid enters the high-
 14

temperature heat exchanger (Indices with capital letters refer to the external stream

conditions and heat exchangers, whereas lower cases refer to the refrigerant.).
The remaining term AT, a temperature difference, is a measure for the performance of the

cycle based on the above described external conditions and is defined as:

EL
AT = QL FHCH + CL (1.9)
£HCH L1CL

The heat exchanger effectiveness, with one fluid undergoing isothermal heat transfer, is

described by Kays and London [5] as

eL= 1-exp(.LUAL) (1.10)

CL

eH=Iexp -UAH)(.1
CH

where UAL and UAH are the heat exchanger conductances for the low- and high-
temperature heat exchanger respectively.
The COP defined in Equation (1.8) has its maximum for AT -4>0. This limit is possible

for a cooling capacity of QL = 0, but it is of no practical interest, because it would result

in a refrigeration cycle with zero cooling capacity. The heat capacitance rates have the
largest influence on the AT. If both are increased to infinite values, heat transfer with the

external streams becomes isothermal as assumed in the Carnot cycle analysis.

Figure 1.7 shows the two possibilities for the temperature distribution of the
external streams for a coupling to a heat source and sink with infinite or finite heat

capacitance rates.
 15

3. 2 3
-Th -Th
-TH R- ooo -TH,out 4
-0

-TH,in -

V -TL,in

-TL 4-
-TL,out

-T1 -T1 1 2
4

Entropy Entropy

Figure 1.7 Temperature vs. Entropy for a Carnot cycle coupled to a heat sink and
heat source with infinite heat capacitance rates (left) or finite heat
capacitance rate (right)

In reality, the heat transfer to and from the external streams is not isothermal. Klein [7]
shows that the optimum COP in that case is attained for equal products of the heat
exchanger effectiveness and their heat capacitance rates

£HCH = £LCL (1.12)

The temperature difference AT reduces then to

AT - 2EC
0L for EC = EHCH = ELCL (1.13)
 16

For equal heat capacitance rates of the external streams CH = CL becomes Equation (1.12)

with Equations (1.9) and (1.10)

1-exp(-UAL)= 1- exp(-UAH) (1.14)

This equation indicates, that the total heat exchanger conductance UAtotal = UAL + UAH

should be split up equally UAL = UAH for equal heat capacitance rates.
 17

CHAPTER
TWO

IDENTIFICATION OF THE OPTIMUM

REFRIGERATION CYCLE

In this chapter, the optimum refrigeration cycle-is determined using a finite difference
approach. The finite difference model is based on the Carnot cycle, but it considers the
external boundary conditions e.g., the heat exchanger sizes, the inlet temperatures of the

external stream at the heat exchangers and their heat capacitance rates. The model is able
to predict the maximum obtainable COP. First design guidelines are established and the

optimum refrigeration cycle shape is determined. The potential improvements of the COP
in a refrigeration system using a NARM, compared to a pure refrigerant are shown.
 18

2.1 DERIVATION OF A NUMERICAL SIMULATION

MODEL

The Carnot cycle is used as starting point for the determination of a realistic upper limit

on the COP. To simulate a non-isothermal phase, it is possible to break the one Carnot
cycle into several smaller ones which together provide the same total cooling load. The

shape of such a cycle is shown in Figure 2.1. The COP obtained with this finite

difference model tends towards to the thermodynamical optimum COP (COP is

maximum) as the number of segments is increased. The Carnot COP is the maximal

thermodynamically possible COP and is simulated in each temperature range.

high external stream

(condenser)

individual Carnot cycles

low external stream

(evaporator)

Entropy

Figure 2.1 Refrigeration cycle broken into several individual Carnot cycles
 19

The temperature of the refrigerant during the phase change processes in both heat

exchangers of every single cycle remains constant, but it is different for each individual

cycle. Hence, a discrete representation of a variable temperature distribution of the

refrigerant throughout the phase changes is modeled. The temperature distribution of the

external stream is necessarily changing along the heat exchangers and in general leads the

trend of the refrigerant.

QH,i+I

QH,i
Th+1
Qi- 1

][ Thi

Th,i-I

external stream with THi

heat capacitance rate
HI

external stream with

Tl,i+l

Ti,i- 1 QL,i-1

OL,i+l

Figure 2.2 Individual Carnot cycle coupled to external streams

The first and second laws of thermodynamics for each reversible Carnot cycle may be

expressed as

ri- =QHJi- QLJi (2.1)

QL~i
 20

and
- (2.2)
Th,i Tli

Th,i is the temperature of the refrigerant in the high-temperature heat exchanger and Tli is

the temperature of the refrigerant in the low-temperature heat exchanger. The COP of this
refrigeration cycle with n individual Carnot cycles is then obtained with
n n

ILi XQL,i
COP = n n i=1 n (2.3)
XQH,i- QL,i wi
i=1 i=1 i=l

where the heat rates and temperatures of all streams may vary for each individual cycle.

In order to determine the COP for this modified cycle, is it necessary to

investigate each Carnot cycle. The examined system is shown in Figure 2.2. Energy

balances and rate equations are to set up.

The energy balances for the external streams in the high and low-temperature heat

exchanger respectively are

QH,i = CH, (TH,i - TH,i-1) (2.4)

(J= CL,i (TL,i - TL,i-1) (2.5)

The rate equations between the refrigerant and the external streams in the high and low-

temperature heat exchanger respectively become

 21

QH,i = EH,i CH,i (TH,i - TH,i-1) (2.6)

Q-L,i = L,i CL,i (TL,i - Tlj)(2.7)

where the heat exchanger's effectiveness are [5]

EL=- exp ( UAL) (2.8)

EH=1- exp(-UAHi) (2.9)

The heat capacitance rates of the external streams are assumed to be constant throughout

the heat transfer process. This is a good assumption considering that fluids such as air or

water will be used for these purpose and that the temperature change in each heat

exchanger is usually relatively small.

The individual cooling load QL,i is specified for every single cycle. It is initially

assumed, that the cooling load provided by each cycle is equal and that the sum of all
individual cooling loads is equal to the total cooling load. The individual cooling load

may be then expressed as:

QL
QL,i- n(2.10)
 22

In section 2.2.2 it is shown, that a significant increase of the COP does not occur if one
tries to optimize the allocation of the heat transferred from the refrigerated space to the
refrigerant and the heat exchanger conductance for every individual cycle.

Hence, the overall heat transfer coefficient area product of the heat exchangers is

determined as

UAi- UA (2.11)
n

For the determination of the COP in terms of external conditions, it is necessary to

specify the following boundary conditions: the heat capacitance rates of the external
streams CH and CL, the overall heat transfer coefficients area products of the high and
low heat exchangers UAH and UAL, the desired total cooling load QL and the inlet
temperatures THin and TLjn of the external fluids in both heat exchangers. With this

information it is possible to calculate the COP with an arbitrary number of individual

Carnot cycles.

2.2 RESULTS OBTAINED FROM THE FINITE

DIFFERENCE MODEL
2.2.1 NECESSARY NUMBER OF INDIVIDUAL CARNOT CYCLES
The program FDM (Finite Difference Method, Appendix), written in EES is, is able to

calculate the desired COP. All external parameters and the number of Carnot cycles may
be varied. The COPn, obtained for constant boundary conditions, is shown versus the
number of individual Carnot cycles in sequence in Figure 2.3.
 23

COPn
7.5

6.5
Hii T 293K - -T~~
323
K

FET 33K -- 33

5.5
. ...........
i ............
!, !..............
o i .............
ii - i..........
.............. ..............
4.5
' 0 40 -o --- 4 - 1- -- -

3.5

2.5

1 2 3 4 5 6 7 8 9 10

n, number of Carnot Cycles

Figure 2.3 COPn vs. number of individual Carnot cycles for different high
external stream inlet temperatures TH,in and low external stream inlet
temperature of T L,in = 273 K, U AL = UAH = 4 kW/K,
CL=CH=lkW/K, QL=10 kW

The inlet temperatures of the external fluid flowing into the high-temperature heat
exchanger are varied for the different curves in the figure. It is first assumed that the heat
capacitance rates (C = 1 kW/K) and heat exchanger conductances (UA = 4 kW/K) are the

same in both heat exchangers.

The COPn, determined with n individual Carnot cycles in sequence, increases in an

asymptotic manner with increasing number of Carnot cycles, as shown in Figure 2.3,
leveling off to a maximum which is independent of n.
 24

1.20

1.15

COPn

1.10

2
1.05= 93 K 3

THi -0K
o--TH -343K

--- TH9* =313K

1.00
1 2 3 4 5 6 7 8 9 10

n, number of Carnot Cycles

Figure 2.4 COPn/COP1 vs. number of individual Carnot cycles for different high
external stream inlet temperatures TH,in and low external stream inlet
temperature of TL,in = 273 K, UAL = UAH = 4 kW/K,
CL=CH=l1 kW/K, QL=10 kW

Figure 2.4 shows the potential COP improvement of COPn/COP 1 versus the number of

Carnot cycles in sequence. A COPn/COP 1 of increase of more than 17% may be

obtained. It depends on the operating conditions of the refrigeration cycle, a small TH,in

leads to large potential performance increases. The lower the inlet temperature of TH,in,

the higher is the calculated COP, but also the higher the potential COP improvement of

the COP.

The required number of Carnot cycles in sequence that will lead closely enough to

the optimum COP of the refrigeration cycle is about 6 cycles. However, for all following
simulations is the refrigeration cycle broken into 10 individual Carnot cycles.
 25

2.2.2 INDIVIDUAL COOLING LOAD OF EACH CARNOT CYCLE

In order to derive this numerical model, the assumption of constant heat flow and equal
heatexchanger conductances for every individual cycle was made. The effect of this

assumption is shown in Figure 2.5. The individual cooling capacity and heat exchanger

conductivity was optimized for each of the 8 individual Carnot cycles. The cycle is first

broken into two sub-cycles in sequence. The COP of the total cycle is optimized with
respect to the allocation of the cooling load and heat exchanger conductance on the two

sub-cycles. Each sub-cycle is then broken into two other sub-cycles and the optimization
is repeated for these cycles. At this state consists the refrigeration cycle of 4 individual

sub-cycles. Every sub-cycle is broken into two further sub-cycles and the optimization
procedure is repeated.

1.2750

U optimum individual Carnot Cycle Cooling Load

OL kW]
constant individual Carnot Cycle Cooling Load
1.2 62 5 ...........................................................................................
------ .....

COP8 ,QJindiv onst = 1.000013

COP 8 , i.,nciv=const

1.25 00 .. .. ......... '.".-.'

..." .. . "...

1.2375 . .. ......... . ... ...

1.2250
1 2 3 4 5 6 7 8

i, individual Carnot Cycle number

Figure 2.5 Optimized individual cooling load Q L,i of each individual Carnot cycle
for TL,in = 273 K, TH,in = 313 K, U AL - UAH, = 4 kW/K,
CL=CH--1kW/K, Q L=10kW
 26

The optimized cooling load is for the Carnot cycles operating at relatively low
temperatures smaller and at relative high operating temperatures larger than the equal

average load. The optimized individual cooling load increases continuously throughout

the heat exchanger, but the influence on the over all COP is negligible. The optimized
heat exchanger conductances turn out to be nearly constant for all individual Carnot

cycles. The potential improvement of the COP optimizing the allocation of the individual
cooling loads QLi to constant QL,i is less than 0.001%. The assumption of constant heat

exchanger conductance and constant cooling load for every individual Carnot cycle is

very good.
The optimum distribution of the heat exchanger conductances and heat capacitance

rates in both heat exchangers is investigated next.

2.2.3 OPTIMIZED ALLOCATION OF THE HEAT EXCHANGER

CONDUCTANCE
The influence of the distribution of the heat exchanger conductance UAtotai with varying

heat capacitance ratios (H/C L of the external streams on the COP is shown in Figure 2.6.
The inlet temperatures of the external fluids are fixed at TL,in = 273 K and TH,in = 313 K

and the low heat capacitance is fixed at CL = 1 kW/ K. The sum of the overall heat
transfer coefficient area products UAH and UAL in the condenser and evaporator is held

constant at UAtotal = 1 kW/K. The allocation of the UA's is shifted between the two heat

exchangers. Figure 2.6 indicates that the maximum COP is always obtained for
UAL = UAH (the deviation of UAH to UAH is less than 0.5 percent). The distribution of
the COP is symmetrical about the even allocation line and exhibits a flat peak.
 27

2.80

2.40o
05
CO
!...
......................
.............
... .
..
.......
........ UAL2 10
10

2 .0 0 ................ :................ i................. ................ ................. ................ ................. ................

0 .2 ' '' "

10 ........ ................. ................. ................. I................... ................ ................ ...............

1.20
30 35 40 45 50 55 60 65 70
UAL •100
UAL + UAH

Figure 2.6 COPio vs. heat exchanger conductance UA fractions for different heat
capacitance ratios with UAtotal = 1 kW/K, CL = 1 kW/K,
TL,in = 273 K, TH,in = 313 K, QL=10 kW

A little deviation from the optimum allocation would still be very close to the maximum

attainable COP. The heat capacitance rate in the evaporator C L was fixed and the total
heat capacitance rate Cext = CH + CL changes for the different curves in Figure 2.6. The

optimum allocation of the overall heat transfer coefficient is independent on the heat
capacitance ratio CH/CL.

An increase of the high heat capacitance rate CH strongly increases the COP for

CH < C L. A further increase of the high heat capacitance rate for CH> CL does not result

in such a significant improvement of the COP. Hence there is a trade off, depending on

the particular operating conditions, between the increase of CH and the increase of the
COP.
 28

2.501 --
COP 10

302 35 40 45 50 .55...........
0 ..........
2.00 ..........

:..............
17------- i.....-----
.......
....
...... 0.1

1.50 L . .. . . . . .. . .. . . .. . . . . . . . . . . . ..
30 35 40 45 50 55 60 65 70
UAL •10
T---23K T 1 1 = 31 UAL
K, Q1 kW
+ UAH.10

Figure 2.7 COPi0 vs. heat exchanger conductance UA fractions for different heat
capacitance ratios with U Atotal = 1 kW/K, Ctotal-" 2 kW/K,
TL,in = 273 K, TH,in = 313 K, QL=10 kW

The same boundary conditions as these applied in Figure 2.6 are used to create

Figure 2.7, only this time the total heat capacitance rate and the total heat exchanger

conductance are held constant. The total heat capacitance rate is fixed at Ctotal = 2 kW/K,
so that the maximum curve in Figure 2.7 matches the curve of Figure 2.6 for equal heat
capacitance rates. The maximum COP is obtained as before for equal heat exchanger
conductances. In addition, it is best to allocate the total heat capacitance rate evenly on

both external streams. The ratio of the two external heat capacitance rates is important,
but in this analysis it does not matter which one of the heat capacitance rates larger.
 29

2.2.4 OPTIMIZED ALLOCATION OF THE HEAT CAPACITANCE

RATES
A result similar to that found in section 2.2.3 is obtained if the external heat capacitance

rates in the evaporator and condenser are varied instead of the heat exchanger

conductances. The same boundary conditions as these applied for Figure 2.7 are used to

obtain the results shown in Figure 2.8. The only difference is the fixed total heat

capacitance rate of Ctotal = 2 kW/K, which is now allocated between the two external
streams and that the overall heat transfer coefficient area product ratio UAH /UAL is

varied, where still UAtotal = 1 kW/K. The different curves refer to different ratios of the

heat exchanger conductances.

2.50 , , ' i

COP1-
1UAo2 1or 1/2
..
---
. ---
.. --.............
. - - .....
......
2.00 ! ............ ....... 3 orl/3

- 1o0
or -
' IJI.
UAL" o 150 ............
"......!... ....... ........ s........ ....... :

0.5

FUAH 10

30 35 40 45 50 55 60 65 70
CL •100
1L+ H

Figure 2.8 COP1o vs. heat capacitance C fractions for different heat exchanger
conductance ratios with UAtotal = 1 kW/K, Ctotai = 2 kW/K,
TL,in = 273 K, T H,in = 313 K, QL=10 kW
 30

The Optimum COP for the specified boundary conditions is, as expected, always obtained

for equal heat capacitance rates CH = CL and for equal heat exchanger conductances

UAH = UAL. The peak again is flat and symmetrical to the even allocation line. A

deviation of the optimum allocation is not severe for this special application. The
calculated COP for constant UAtotal is the same when UAH / UAL or UAL / UAH is used

to calculate it. Hence, it would be possible to exchange the heat exchangers without

changing the COP even if the heat exchanger sizes were different.

3.5
' ' ,, i ' 10
.................... ...........
-----------------
3.0 2
................. ................... ..................
. . . . ..... . . . . . .................... •..................!..................
COP10
2.5
.................
..................
A...........
.....
t..................
..................
..................
..................
..................

2.0
-- ' 0o.5''

.................-------------------.-
...- ..................
;.-------------- ................. ................. . .................. ------------------.
1.5

1.0

0.5

n n' .... .. L......

'.V

30 35 40 45 50 55 60 65 70

eL
CL+ CHi10

Figure 2.9 COPlo vs. heat capacitance rate C fractions for different heat
exchanger conductance ratios with UAL = 0.5 kW/K, C total = 2 kW/K,
TL,in = 273 K, TH,in = 313 K, QL= 10 kW

The total heat exchanger conductance is varied in Figure 2.9. An increase in the COP for
larger ratios of UAH/UAL is obtained for constant UAL. The total heat exchanger
 31

conductance varies for the different curves and an increase of the COP for increasing

ratios of UAH/UAL is obtained for constant UAL.

Figure 2.10 shows results with the total heat exchanger conductance enlarged by the

factor of 4. Larger values for the COP are obtained and the peak for the optimum COP
sharpens with increasing UAtotal. The larger UAtotal, the more sensitive is the allocation

of Ctotai on the two external streams.

3.00

2.75 2 or 1/2
COP10 3 or 1/3
2.50
4 or 1/4

2.25
..............
...............
..............
---------I--.
----..............
............
.....
.........
.....
or 1/5

2.00
7 r1 4

1.75
.... 1UA
M MrIoo
7 A=
-10 t . .... .. ......
.......
............. ..i........".......
1.50
30 35 40 45 50 55 60 65 70

CL
-•. 100
C, +C14

Figure 2.10 COP1o vs. heat capacitance C fractions for different heat exchanger
conductance ratios with UAtotal = 2 kW/K, Ctotal = 2 kW/K,
TL,in = 273 K, TH,in = 313 K, QL=10 kW

For Figure 2.11 is the total heat capacitance rate decreased by the factor of 4. The COP
increases and again the peak for the optimum COP sharpens with decreasing Ctotai. The

smaller Ctotal, the more sensitive is the allocation of Ctotal on the two external streams.
 32

3.50

3 .00 ...................... ... ;a0W =q .................. ................... i............................

COP1 0
2 52 or0 5..... i...... .......... --

2.0 = rO2
"5.................. ............ ................. ................. .

2.00 ............. ........ . . . . . .. . . . ..............................

---------- ----- ................. .......... ................- --

UA = 10 or - 1
UAL 10
1.00

25 35 45 55 65 75

CL 100

Figure 2.11 COP10 vs. heat capacitance C fractions for different heat exchanger
conductance ratios with UAtotal = 2 kW/K, Ctotail=-0.5 kW/K,
TL,in = 273 K, TH,in = 283 K, QL=10 kW

2.2.5 OPTIMUM REFRIGERATION CYCLE SHAPE

Figures 2.12, 2.13 and 2.14 show the shape of the optimum refrigeration cycle in a

temperature-entropy transfer rate diagram for different heat capacitance ratios. The shape
of the optimum refrigeration cycle as shown in Figure 2.12 assumes a constant total heat
capacitance rate and total heat exchanger conductance which may be allocated between the

two heat exchangers. The temperature difference between the heat transferring fluids in
each heat exchanger is constant. For the cycle having the maximum COP, the heat
exchange process between the refrigerant and the external stream occurs throughout the
 33

340

320
3202refrigerant in condenser

ATl = const

5_..l
0
WS300 external streams

280

260 ATim-const
refrigerant in evaporator

240 I I I I
-0.005 0.005 0.015 0.025 0.035 0.045

Entropy transfer rate [kW/K]

Figure 2.12 Temperature vs. entropy transfer rate with low and high external
stream inlet temperature of TLin = 273 K, TH,in = 313 K respectively,
UAL = UAH = 1 kW/K, CL=CH= 1 kW/K, QL=10 kW

whole heat transfer process at a constant temperature difference. Hence the temperature

change of the external stream should match the temperature change of the refrigerant.
The temperature profiles of the external stream and the refrigerant are parallel to each
other in each heat exchanger as seen in Figure 2.12. It turned out that the COP i for every

individual cycle is constant and consequently equal to the total COP. A change of the

heat capacitance rates to CH <<CL or CH >> CL does not change this behavior

significantly. The temperature difference in the evaporator remains constant, due to the
equal heat rate in each single Carnot cycle in the evaporator. The temperature difference
in the condenser is still fairly constant.
 34

330
I V -I! I
ATim slightly decreasng

I I1lob.
310 "

6..
\ s
external streams
290 •
3-

270 -
..I.
&TI =co I

1)I I
Liu.-
005 0.005 0.015 0.025 0.035 0.045

Entropy transfer rate [kWIK]

Figure 2.13 Temperature vs. entropy transfer rate diagram with low and high
external stream inlet temperature of TL,in = 273 K, TH,in = 313 K
respectively, UAL = UAH = 1 kW/K, CL= 1 kW/K, CH= 10 kW/K,
QL=1O kW

Only for very small heat capacitance rates CH (C << CL) increases the temperature

difference, between the external stream and refrigerant in the condenser. Nevertheless,
the increase in the temperature difference is fairly small compared to the absolute increase
of the fluid temperatures. The temperature change of the external stream should match

the temperature change of the refrigerant also for refrigeration system with unequal heat
capacitance rates of the external stream. The optimum refrigeration cycle shape will be

investigated more thoroughly in the next chapter.

 35

500 3

450

3.. 400

CU ATIm slightly increasing 1

350
external streams

300

. . . . . .
250 ....... I

-0.005
0.005 0.015 0.025 0.035 0.045

Entropy transfer rate [kW/K]

Figure 2.14 Temperature vs. entropy transfer rate diagram with low and high
external stream inlet temperature of TL,in=273 K, TH,in = 313 K
respectively, UAL = UAH = 1 kW/K, C L= 1 kW/K,CH=O.1 kW/K,
QL= 10 kW
 36

2.3 CHAPTER SUMMARY

It was shown that a relatively small number of individual Carnot cycles is sufficient to

obtain a COP very close to the maximum COP.

First design guidelines could be established:

For specified total heat exchanger conductance and specified total heat capacitance rates

Equal allocation of the total heat exchanger conductance on the high- and low-

temperature heat exchanger, independent on the heat capacitance ratio (H/CL and

the inlet temperatures of the external streams.

Equal allocation of the total heat capacitance rates on the external streams,

independent on the heat exchanger conductance ratio UAH/UAL and the inlet

temperatures of the external streams.

For a varying total heat exchanger conductance and a varying total heat capacitance rates

An increase of the total heat exchanger conductance increases the COP and

increases the potential COP improvement.

An increase of the total external heat capacitance rate increases the COP, but

decreases the potential COP improvement.

For the temperature profiles of the external streams and the refrigerant
40 The slope of the temperature change of the refrigerant in the heat exchanger

should be parallel to that of the external stream flowing through that heat

exchanger.
 37

CHAPTER
THREE

ANALYTICAL SIMULATION MODEL

The calculation of the COP with the numerical model described in Chapter 2 leads to a

large set of non-linear equations. The solution of the equations requires extensive

calculations. So it is necessary to look for an analytic possibility to evaluate the optimum

COP in terms of the external conditions.

In this chapter, another approach is presented for the determination of the

optimum refrigeration cycle. This Analytical Simulation Model (AM, Appendix) does not
provide information about the temperature distribution of the heat transferring fluids in

the heat exchangers, but it is possible to specify internal heat capacitance rates of the
refrigerant. A general and simple expression for the COP of an arbitrary system is

found. The numerical and analytical model are then compared with each other.
 38

3.1 DERIVATION OF AN ANALYTICAL MODEL

3.1.1 GENERAL EXPRESSION FOR THE COP

A refrigeration cycle broken into n individual Carnot cycles where n - >00 is assumed.

Such a procedure leads to the optimum COP. The temperature of the external fluids and

the refrigerant flowing through the heat exchangers change continuously. The general

definition of the COP applied to this cycle is again

COP= QL(1.5)

In order to determine the COP, it is necessary to set up energy balances and rate

equations for the high temperature heat exchanger (the condenser). The heat transfer
does not occur isothermally for the refrigerant, as in each individual Carnot cycle, and the

energy balance for the refrigerant in the condenser yields

QH = Ch (Th,in - Th,out) (3.1)

where 'Ch is the heat capacitance rate of the refrigerant in the condenser and Th,in and

Th,out are the temperatures of the refrigerant at the condenser inlet and outlet,

respectively. An energy balance of the external stream yields

QH =CH (TH,out - TH,in) (3.2)

 39

The energy balance for the refrigerant in the low temperature heat exchanger, the

evaporator, is similar obtained as

QL = I1(Ti,out - Tl,in) (3.3)

where C1 is the heat capacitance rate of the refrigerant in the evaporator and Tl,0 ut and

T 1,in are the temperatures of the refrigerant at the condenser outlet and inlet, respectively.

The energy balance for the external stream may be expressed as

QL = CL (TL,in - TL,out) (3.4)

The rate equations between the refrigerant and the external streams for the condenser are

given as

QH = CH CH,min (Th,in - TH,in) (3.5)

and for the evaporator as

QL = CL CL,min (TL,in - Tl,in) (3.6)

The heat exchanger effectiveness factor is defined as [5]

 40

CH1mi
1- exp -NTUH 1 - CH,min
CH ____
~ ~ [
H,maxj(37
=. xl(3.7)
1-(ICHmin')exp 1-NTUH1
- H'min
(CH,max I L H,max /

for the high temperature heat exchanger and

= 1-exp [-NTUL(1C-Lm2:)1L(3.8)
e-L L a(3.8)

1 CLmin exp [-NUL(1 - rL'min)]

CLmax I LCL,maxj

for the low temperature heat exchanger, where NTUL and NTUH are defined as

NTUH=.UAH (3.9)
CH,min

NT UAL (3.10)
CL,min

The heat capacitance rates appearing in Equations (3.5)-(3. 10) are obtained with

CH,max = max (CH,Ch) (3.11)

CH,min = min (CH,Ch) (3.12)

CL,max = max (CL,C1) (3.13)

CLmin = mi ((L,41) (3.14)

 41

If the heat capacitance rates ratio is CH,min/CH,max = 1, then the heat exchanger

effectiveness for the condenser in Equation (3.7) reduces to [5]

NTUH for CH,min/CHmax = 1 (3.15)

SH = NTUH + 1

The evaporator heat exchanger effectiveness in Equation (3.8) for CL,min/CL,max-= 1

reduces to

= NTUL for CL,min/CL,max = 1 (3.16)

NTUL + 1

The application of the second law of thermodynamics for this refrigeration cycle is more

complex then for the Carnot cycle. The temperatures of the refrigerant in the low and
high temperature heat exchanger are not constant. Suitable mean temperatures for non-
isothermal heat transfer must be found so that the entropy balance for a reversible cycle

becomes

= H(3.17)
T1,mean Th,mean

The heat and entropy transfer must be examined carefully to determine these

thermodynamic mean temperatures.

The entropy transfer rate due to the heat flow may be expressed as

SQJ TTm (3.18)

 42

where the second equation defines the thermodynamic mean temperature for a non-

isothermal heat transfer process.

Tm = ---T2dQS Q (3.19)

fTj

In order to calculate the mean temperature Equation (3.19) is written as

TM =AQ _ h2 - hi (3.20)
Tm Ss2-S1

where 1 and 2 refer to inlet and outlet conditions respectively.

For a fluid with constant specific heat cp, isobar enthalpy and entropy change and

negligible kinetic and potential energy becomes the enthalpy

= ho + cp ( Tx-To )) + vo (Px- Po) (3.21)

and the entropy

sx=so + Cp In -Tx - R In Po (3.22)

To PO

combining Equations (3.19) to (3.22) yields

Tm = ( h0 + CT 2 )-( h0 + CpT1) - T2 - T 1 (3.23)

To)2 5 + c lnnT2
1)
 43

where Tm is the thermodynamic mean temperature for non-isothermal heat transfer. The
mean temperatures which are to be used in Equation (3.17) are

Th,mean - Th,out - Th,in (3.24)

InTh,ou~t
ITh in

T 1,out - T lin
( . 5
Ti,mean = n (3.25)
Tl,in

It is now possible to evaluate the coefficient of performance only in terms of external

conditions. Solving the Equations (1.5), (3.1) to (3.17), (3.24) and (3.25) for the COP
leads to the following general expression.

QL1+ [ CH. -1

COPL= ICh(3.26)

TH,inCHC-H,min (131) - Q 1+ (EH CRm-1

where 3 is defined by

l ( nL,min
QLEL +1 (3.27)
LC1 (TLjnFl L1CL,mnin - QL)J
 44

3.1.2 PARTIALLY OPTIMIZED COP

The optimum refrigeration cycle shape was shown in a temperature-entropy transfer rate

diagram in Figure 2.12. Parallel temperature curves of the fluids in each heat exchanger
have been obtained for that cycle. That indicates, that the heat capacitance rates of the

refrigerant and the external fluid flowing through the same heat exchangers
(CL,ext = Ci,ext & CH,ext = Ch,refr) should be equal in order to obtain an optimized COP.

Equation (3.26) reduces then to

COP = QL [1+ - (3.28)

TH,inSCH H(X- 1)- QLji +X(SH-1)]

where Xis defined by

X[=(TL,n CL )+1 (3.29)

Equations (3.28) and (3.29) do not require equal heat capacitance rates of the external
streams flowing through the different heat exchangers (CL,ext CH,ext). The COP in

Equation (3.28) is only optimized with respect to the ratio of external to refrigerant heat

capacitance rate (CLext/C1,refr & CH,ext/Ch,refr), hence it is called a partially optimized

COP. A verification of that result is fairly simple. COP's obtained with the expression

in Equation (3.28) must be equal or slightly larger then the COP calculated with the Finite

Difference Method for 10 cycle. The two models are compared in section 3.3.
 45

3.2 RESULTS FROM THE ANALYTICAL MODEL

3.2.1 VARIATION OF THE EXTERNAL INLET TEMPERATURES

TL,in AND TH,in

The influence of different external inlet temperatures TH,in and TL,in will be investigated

along with different temperature drops TLin - TL,out of the external stream in the low-

temperature heat exchanger.

The same boundary conditions applied as these for the numerical investigation in
the previous chapter with the additional assumption of equal heat exchanger

conductances, UAH = UAL and equal heat capacitance rate in each heat exchanger,

C H = Ch & CL = C1. A new parameter, the potential improvement of the COP named Q,
is defined as

COP=00
f2 o
(3.30)
COP 1

COPOO represents the COP obtained from the Analytical Simulation Model and COP 1 is

the COP obtained from the Numerical Simulation Model with n=l (one Carnot cycle).

Figure 3.1 shows a schematic temperature - entropy diagram. The external inlet

temperatures and a temperature difference, the gliding temperature difference GTD are
labeled in the figure. The GTD refers to the total temperature change which a fluid

acquires during the heat transfer process. These and other parameters will be
investigated. Q decreases as the difference between TH,in and TL,in increases. Figure

3.2.a indicates also, that 2 increases with increasing heat exchanger conductance. This
behavior is especially significant for small temperature differences and becomes less
important for higher incoming temperatures TH,in (TL, in is fixed at TL,in = 273 K).
 46

5..

Cu
5..

GTD1 refr

Entropy

Figure 3.1 Schematic Temperature - Entropy diagram

4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.50

0.0
-10 0 10 20 30 40 50

TH,in- TL,in [K]

Figure 3.2.a Potential COP improvement Q vs. temperature difference TH,in-TL,in

for different heat exchanger conductances, low external stream inlet
temperature of TLjn= 273 K, C L = C H= 1 kW/K, QL=10 kW
 47

The range of larger temperature differences represents the realistic operating conditions of
a refrigeration system. TH,in will ordinarily be well above TL,in. It is unlikely that the

fluid which has to be cooled enters the low-temperature heat exchanger at the same or

littlehigher temperature than the temperature of the high external stream, which enters the
high-temperature heat exchanger usually at ambient temperature TH,in. The section of

interest is enlarged and shown in Figure 3.2.b. The potential improvement 92 can be as

high as 50 % for infinite heat exchanger size.

The smaller the inlet temperature difference of the external streams THin -TL,in,

the larger the potential COP improvement Q. The larger the total heat exchanger
conductance UAtotai, the larger the potential COP improvement Q.

1.6 i : '
. -- UAL=UAH=0.25 kW/K

1.51.
... increasing
i....
....... A......................
U........ UAL=UAH= 1kW/K
UAL=UAWf4kW/K

... i. ........
.*.
. .... ... ............
1 ... ...
.......... UAL=UAH=I16 kW/K

.. . . .. .. . . . .. . . .. . .. . .
1.3 -- ------- -- ---.... . .. . ..
-.- ------- ---.
.-..-- . . .. . ,.
~ ... U..
i .... n.t
....--...

Immmmm mmmmmm

1.2

1.1

1.0
10 15 20 25 30 35 40
[
TH,in - TL,in K]

Figure 3.2.b Potential COP improvement 02 vs. temperature difference TH,in - TL,in
for different heat exchanger conductances, low external stream inlet
temperature of TLjn= 273 K, CL = C H= 1 kW/K, QL=10 kW
 48

3.2.2 VARIATION OF THE GLIDING TEMPERATURE

DIFFERENCES GTDl,refr AND THE COOLING LOAD QL

3.2.2.1 Gliding Temperature Difference GTDI,refr at constant Heat

Capacitance Rates
The temperature drop of the low external stream represents the ability of the refrigeration

system to cool down a fluid to a certain temperature. The temperature drop of the system
investigated in chapter 3.2.1 was fixed at TLin - TL,out = 10 K. What is the influence of

other temperature differences?

The temperature difference of the low external stream is now called external
gliding temperatures difference GTDL,ext and defined as

GTDL,ext = TL,in - TL,out (3.29)

The heat capacitance rates of the streams in each heat exchanger are assumed equal and

consequently the temperature drop of the external stream is equal to the temperature drop

of the refrigerant.

GTD,ext = TIout-- Tin = GTDL,ext (3.30)

A variation of the GTDi,ext is attained by varying the cooling capacity Qi. or by varying

the heat capacitance rates. An increase of GTDi,ext increases the potential improvement

Q. It is possible to double the COP for very large gliding temperatures GTDi,ext in the

low-temperature heat exchanger, but the refrigeration system at this operating conditions
would operate at a very low COP. The COP and the potential improvement are strongly
dependent on the operating conditions. In Figure 3.3 for example is a CO = 3 and a
 49

= 1.10 (10%) obtained for a system at GTDl,refr = 20 K. The external high inlet

temperature is THn = 313 K = 40 'C.

The COP of the system decreases with increasing GTDl,refr. This behavior is

non-intuitive and the reader might ask why it is useful to apply a refrigerant mixture when

the COP decreases. Actually the cooling capacity is increased in order to vary the
GTD,refr. The refrigeration system itself is not changed and the same system has to cope

with a higher cooling load which explains the reduced COP. Nevertheless it is seen, that

a system operating at large GTD and using a suitable refrigerant mixture would perform

better than a pure refrigerant.

COP
6 1.75

1.25

0 20 40 60 80 100
GTDI,ref [K]

Figure 3.3 COPo., COPI and potential COP improvement 2 vs. temperature
difference Tlin- Ti,out with external stream inlet temperature of
TL,in=273 K, TH,in=313 K and UAL = UAH = 2 kW/K, CL=CH= 1 kW/K
 50

TH,in in Figure 3.3 is fairly high for ambient heat rejection. If the temperature TH,in is

decreased from TH,in = 313 K to TH,in = 293 K = 20 'C, which is more realistic, then the
COP increases to COPO = 4, and Q increases even to 0 = 1.13 (13%), as shown in

Figure 3.4. Consequently there is a trade off between the attainable COP, the potential

improvement Q, and the operating conditions.

10 3.5

COP
.............
.........
":
............................
i...........................
..!.....
..
....
....
..
...............
.........
2
6
2.

2 1.5

0
20 40 60 80 100
GTDlref [K]

Figure 3.4 COP., COPI and potential COP improvement Q vs. temperature
difference TIin- Ti,out with external stream inlet temperature of
TL,in = 273 K, TH,in=293 K and UAL = UAH = 2 kW/K, CL=dH= 1 kW/K

3.2.2.2 Influence of Cooling Load

The variation of GTDlrefr at constant heat capacitance rates causes also a variation of the

cooling load, GTDLrefr ~ QL. Hence refrigeration systems of very different sizes have

been compared in section 3.2.2.1. The cooling capacity changed from

 51

QL > 0 to QL = 100 kW. In Figure 3.5 is the influence of QL on the COP for a
refrigeration cycle operating at different external high heat capacitance rates CH shown.

The COP decreases very quickly with an increasing cooling load. An increase of CH at

constant CL certainly helps the system, but does not change the tendency.

cop 0

10 100

QL [KW]
Figure 3.5 COP. vs. cooling capacity QL with external stream inlet temperatures
of TLin=273 K, TI~in= 313 K and UAL = UAH = 2 kW/K, CL = 1 kW/K

An accompanying plot for the potential performance increase 92 is given in Figure 3.6. It
indicates that 0 increases very rapidly for large cooling loads QL. On the one hand, large
0 may be obtained for large QL, but on the other hand would such a refrigeration system

run at a low COP.

 52

4.0

3.5

C-- =0.25kW
4 ---
3.0
..i....
.....
...
....
..
...........
..........
..........
...
.........
'...............
.................
...
...
.. ..
......
4--- ......
.. --
----- -- ..
... ---....
.. ....
... ........
.....

2.5
-
...
........
..........
...
..
.....
..
...
...
..
...
..
..
.....
.. increasing. CH
0.5

2.0
....... .....
.......
.....
.....................
.....................
1.5
......
i....
.......
....
........
i.....
.............. ...... 2
............ i
8 4
1.0
10 100

QL [KW]

Figure 3.6 Q vs. cooling capacity QL with external stream inlet temperatures of
TL,in=273 K, TH,in=313 K , UAL = UAH = 2 kW/K, CL = 1 kW/K

3.2.2.3 Gliding Temperature Difference GTDI,refr at constant

Cooling Load
The GTDl,refr can be varied at a constant QL by a variation of the heat capacitance rate

CL. Figure 3.7 shows the same tendency as these indicated by Figure 3.3, but the
potential improvement Q and the COP. are larger. For a gliding temperature difference

of GTDI,refr = 20 K is a COP of COP.0 = 3.7 and a Q of Q = 1.18 (18%) obtained. Q

increases more rapidly and the COP levels off slower than in the case of non-constant

cooling capacity.
 53

10 3.5

COP
6 . .2.5

01
0 20 40 60 80 100
GTDI, ref [K]

Figure 3.7 COPo., COPI and potential COP improvement Q vs. temperature
difference Ti n- Ti,out with external stream inlet temperature of
TL,in=273 K, TH,in=313 K and UAL = UAH = 2 kW/K, QL=10 kW,
CL = 1 kW/K

A decrease of TH,in from THjn = 313 K to TH,in = 293 K yields to the same tendency as

before. The COP and Q increases become even more significant as seen in Figure 3.8.
At a GTDl,refr of GTDl,refr = 20 K is now a COP of COP.o= 4.9 and an Q of

Q = 1.28 (28%) obtained.

The potential performance improvement Q becomes more significant with
increasing gliding temperature differences GTDl,refr, but the COP itself decreases. Q

alone does not provide sufficient information about the performance of the system.
 54

10 3.5

8
......
.... ..............
.....................
y'-.... 3
COP I

COP
6
%T.......
.......
................ .............
.......................... D2 2.5
...................
2.
...................

..........................
.. .....
.. ....
. .......
....... -..
... .....
.........
....... 1.5
2

0
0 20 40 60 80 100
GTDIref [K]

Figure 3.8 COP., COPI and potential COP improvement Q vs. temperature
difference TIin- TI,out with external stream inlet temperature of
TL,1n = 273 K, TH,in= 293 K and U AL = UAH = 2 kW/K, Q L = 10 kW,
CL = 1 kW/K

3.2.3 INFLUENCE OF TOTAL HEAT EXCHANGER CONDUCTANCES

The next investigation is devoted to the question how significant the potential
performance increase 0 is for different total heat exchanger conductances. The

performance of a refrigeration system which has to meet two different loads, QL = 10 kW

and QL = 50 kW is shown in Figure 3.9. The COP increases with increasing UAtotai

and also does the potential performance improvement Q. The refrigeration system which
has to meet the smaller cooling load QL achieves the larger COP and also the smaller Q.
The larger the cooling load QL and the larger the total heat exchanger size UAtotai, the
larger the potential performance improvement Q .
 55

10 2.5

COP
00 ...
....... ......................
.......
.............
1 k .
2.-.-.---------...... ---------- : : L: . . . ..i...
........................ ............. 1.3
"
40
-"i"" nQL = 0kw

1.0
0 10 20 30 40 50
UAtotal [-K

Figure 3.9 COPo and potential COP improvement Q vs. UAtotai with external
stream inlet temperature of TL,i.=273 K, TH,in=293 K and UAL = UAH,
= =50 kW
CL C H= 1 kW/K, QL =10 kW and QL

3.2.4 INFLUENCE OF TOTAL HEAT CAPACITANCE RATES

It is interesting to consider the influence of the total external heat capacitance rate
Cexttotal-=CL,ext + CH,ext on the COP and Q. An increase of Cext,total results in a

smaller GTDI,refr, which reduces the potential COP improvement but increases the COP.
The heat transfer approaches an isothermal process for Cexttota l -4 oo and hence there are

no further improvements possible in operating the cycle with a working fluid which
undergoes non-isothermal heat transfer (COP 0) instead of a working fluid undergoing

isothermal heat transfer.

 56

10 : :: i :: : : : 1.5
:':2 i@QL=5OkW:"

).....
,...... .........
8. . . . . -.......... ....... .....
i ... 1.4
..............
4.....
.--
COP 0 41COP 2 @Q ! 10k COP@Q
@QL OL10kW
50k 1 2i
6 ............. .'........ ..... i... ., , ------ ...
----.. . ...f --- '- ---... 1.3
..... .. ---...
. 4...... .. ...---..-----------...
..--

UA- = UA - 2 kW/K-, L QL =10 kW" or QL =50 kWk-1

01 1 100 100
: ii~ Ce, oti [1W.] i i!i~ i
.........
chpt..ha.tehet
Upton......wisitasumd... .apcianeraesof.... t..-- t
Fiur
310 COO. an . COP iroemen
ponial i2 Cetti
VS ith xtena
0 i : a d i Id :'I ,.!!I i

o.1 1 10 100
Cexttotal [Kl-

Figure 3.10 COP., and potential COP improvement K2 vs. Cext, total with external
caaiac:aestream
of th inlet temperature
externa exhngr K and have notbee
TH,in=293
l streams ofi TL,in=273 K, iffrntha
U AL = UAH = 2 kW/K, ( L =( H, ( =10 kW or & = =50 kW

3.2.5 OPTIMUM REFRIGERANT HEAT- CAPACITANCE RATE

Up to now is it assumed in this chapter that the heat capacitance rates of the streams in

each heat exchanger (CL,ext = Cl,,refr and C H,ext = C h,refr) are equal, whereas the heat

capacitance rates of the external streams in the different heat exchangers have not been

necessarily the same (CL CH). Consequently as a result of this assumption are the

refrigerant heat capacitance rates in the different heat exchangers different. It is surely not

easy to find a fluid with these required properties. It is now time to look at a refrigerant
 57

with constant heat capacitance rate Crefr in order to find the corresponding optimum heat

capacitance rate of the refrigerant for arbitrary given external heat capacitance rates. The
optimum heat capacitance rate of the refrigerant Crefr,opt is shown in Figure 3.11 and

plotted versus the external heat capacitance ratios CR/CL.

10.0

C~refr,opt [K]

1.0

0.1
0.1 1 CH 10 100

eL

Figure 3.11 Optimum refrigerant heat capacitance rate Crefr,opt vs. external heat
capacitance rate ratio CH/CL for different fixed CL and UAL = UAH

The low heat capacitance rate CL is fixed at different values for each curve in Figure 3.11.

The heat exchanger conductance of the low- and high-temperature heat exchanger are
equal, UAL = UAH. All curves level off to a constant Crefr,opt, which is always twice as

much as the constant CL. These values represent the optimum (refr for an refrigeration
cycle with isothermal heat rejection.
 58

The optimum Crefr,opt does not change for different inlet temperatures of the

external stream, different total heat exchanger conductances or different cooling loads.
The optimum heat capacitance rate of the refrigerant Crefr,opt is solely dependent on the

two heat capacitance rates of the external streams. The Crefr,opt for the optimum heat

exchanger conductance allocation UAL = UAH may be expressed as

Crefr,opt 2(L+3

Crefr,opt = 2 ( +CH (3.31.b)

3.3 COMPARISON OF THE NUMERICAL AND

ANALYTICAL MODEL
3.3.1 COP OF THE DIFFERENT MODELS
The coefficients of performances calculated with the different models are compared. The

four most interesting COP's are:

* The COP 1 calculated from the numerical model with one single Carnot Cycle

The COP tO calculated from the numerical model with 10 Carnot cycle in sequence

The COP. calculated from the analytical model with fluids of the same heat
capacitance rates in each heat exchanger (CL,ext = 11,refr & CHext = Ch,refr)

The COP * calculated from the analytical model with a constant heat
capacitance rate Crefr of the refrigerant
 59

In Figures 3.12 to 3.15, plots of COP0 . vs. a constant heat capacitance rate of the

refrigerant Cefr are shown. The heat capacitance rates of the external streams are varied

for the different figures; all other boundary conditions are constant. In order to note the
similarities and the differences, the corresponding COP 1 , COP 10 and COP. are plotted

as horizontal lines. The value of the optimized Crefr is indicated in each figure. This

value is calculated with the optimization process in EES and gives the same result as

Equation (3.31.a).

4.30
COP*@ refr,opt - 1.00 kW/K
CoP COP
. ..........
...... ........ .................. ......... .................. ... .....~ ° ...........................
4.20 -. . Co 10

4.10

4.00

,'i' I COP I
CL- CH 1kW/K
3.90
10
C refr Ik

Figure 3.12 COP vs. heat constant capacitance rate of the refrigerant Crefr for
TL,in=273 K, TH,in=313 K, UAL=UAH=2 kWK, CL=CH= 1kW/K

COP* levels off to the COP 1 for increasing heat capacitance rate Creft- The heat transfer

is isothermal for Crefr->coo and is identical to a Carnot cycle process. The heat
 60

capacitance rate Crefr should be at least Creft -CL- Otherwise the COP. may drop

rapidly.

An increase of the high external heat capacitance to CH =4 kW/K results in a

higher COP, but COP * can not be higher than the value of the COP... The optimum COP

would be obtained for CH = Ch,refr = 4 kW/K and CL = C1,refr = 1 kW/KL which would

lead to the optimum temperature matching where the gliding temperature differences are

equal, GTDext = GTDrefr. It is not possible to meet both conditions for a refrigerant with
a constant heat capacitance rate throughout the whole refrigeration process.

4.60

COP

4.55

4.50

4.45

4.40
0 2 4 6 8 10
Crf kW

Figure 3.13 COP vs. heat constant capacitance rate of the refrigerant Crefr for
TL,in=273 K, TH,in=313 K, UAL=UAH=2 kW/K CL= 1 kW/K and
Cu=4 kW/K
 61

The corresponding Crefr,opt is somewhat between the external heat capacitance rates CL

and CR and indicated in Figure 3.13. The value is obtained by optimizing the COP with
respect to Crfr. This number is identical to the Crefr,opt obtained with Equation (3.31.a).
The difference between the COP and COP* becomes even larger for a further

increase of the heat capacitance rate to CH =20 kW/K, as Figure 3.14 shows. The heat
rejection becomes virtually isothermal and the optimum C refr,opt approaches the number

of Crefr "" 2 CL = 2 kW/K.

4.70

Cop COP

4.65
... . . . . . . .... .......
... . . . .... . . . . . . . . . . . ....
....
.. . . . . . . . . . . . . . ...
...
po.
CO0P,-@ Crefr opt 1.90 kW/K
C re r°P
................ .....................
.... :..... .......
. .......... .... ...................... .
4.60

4.55

, , kW
,1 . . . .oi . . . . .=... k . . .
4.50
10 15 20 25
Crefr

Figure 3.14 COP vs. heat constant capacitance rate of the refrigerant Crefr for
TL,in = 273 K, TH,in= 313 K, UAL=UAH= 2 kW/K CL= 1 kW/K and
CH= 20 kW/K
 62

The same behavior is obtained for a very small heat capacitance rate of CH = 0.25 kW/K.
The optimized Crefr,opt is again somewhat between the external heat capacitance rates CL

and 'CR as indicated in Figure 3.15. A deviation of the Creft from the Crefr,opt results in a

significant decrease. The heat capacitance rate of the refrigerant should be at least
Crefr min (ICL,'CH).

3.25

COP

3.00

2.75

2.50

2.25
0.5 1 1.5 2 2.5
Crefr [kw]

Figure 3.15 COP vs. heat constant capacitance rate of the refrigerant Crefr for
TL,in=273 K, TH,in=313 K, UAL=UAH=2 kW/K CL= 1 kW/K and
CH= 0.25 kW/K
 63

3.3.2 ACCURACY OF THE NUMERICAL MODEL

In this section is first shown, that the solutions of the numerical and analytical model are
nearly identical. Then is the distribution of the total heat exchanger conductance UAtotai

on the two heat exchangers presented.

Some arbitrary conditions have been picked to compare the models in Figure

3.16. The cooling load QL is changed for the different curves in the Figure and the ratio
of COPn/COP. vs. the incoming external temperature TL,in is shown.

1.000 ....... ...... ......

0.999
COPn ..........
............

coPoo
0.998 -.O.,,[email protected].... ....

0.997
..............................
{...............................
............................--
...............--............--......
COP../COP 2 o @ QL= 10 kW/K
COP./COPl 0 @ QL= 10 kW/K
0.996
COP ICOP2 o @ QL = 50 kW/K
COP./COPlo @QL = 50 kW/K
0.995
200 220 240 260 280 300

TL,in [K]

Figure 3.16 COP/COPn vs. different low external inlet temperatures TL,in for
=
different cooling loads QL for THn 313 K, U AL = 2 kW/K,
=
UAH=4 kW/K, CL=l1kW/K and CH 5 kW/K

The COP's are always nearly identical. A model of 10 Carnot cycles in sequence

calculates the COP accurately. The error is less than 0.5% for the system with a cooling
 64

load of QL = 50 kW/K. For large QL is the agreement less than for smaller ones. Large

QL leads to a stronger temperature change of the fluids, so that the alignment of several

Carnot cycles in sequence to the continuous temperature change of the external stream

becomes more difficult. The deviation is still less than 0.5 % for relatively large QL.
The numerical model has the advantage of being able to solve for the temperature

distribution of the fluids throughout the heat exchangers and it represents close enough

the maximum possible COP. On the other side, it is not possible to specify the heat

capacitance rates of the refrigerant. This is the great benefit of the analytical model.
Furthermore it offers a single algebraic equation only in terms of external boundary
conditions for the determination of the COP.

3.3.3 ALLOCATION OF HEAT EXCHANGER CONDUCTANCE AND

HEAT CAPACITANCE RATE FOR AN ISOTHERMAL PROCESS

3.3.3.1 Heat Exchanger Conductance

It is known from previous investigations with the numerical model that the total heat

exchanger conductance UAtotal and total heat capacitance rate should be allocated equally
for constant Cext,totai and UAtotal. The analytical model leads to the same result as

Figure 3.17 indicates. The corresponding COP 1 for an isothermal heat transfer process

is also presented. The optimum allocation of the UAtotai is not equal for the different

simulations. As expected, the analytical simulation shows an even allocation of the total
heat exchanger conductances, whereas this is not true for the simulation with one Carnot

cycle. For this system, an even allocation is only recommended for equal heat
capacitance rates of the external stream. Otherwise it is useful to reallocate the UA for a

given UAtotai between the heat exchangers.

 65

As shown in Figure 3.18, the optimum allocation of the UAtotal depends on the

heat capacitance ratio CH/CL. For ratios of CH/CL < 1 is the optimum allocation shifted to
larger UAL, whereas it is shifted to larger UAH for CH/CL> 1.

4.0

3.5
COP
3.0

2.5

2.0

1.5

1.0

0.5

0.0
0 10 20 30 40 50 60 70 80 90 100

100 UAL
UAL + UAH

Figure 3.17 COP. and COPi vs. heat exchanger conductances fractions UAL for
different external heat capacitance rates ratios with Cext, total== 2 kW/K,
UAtotal=4 kW/K, TH,in=293 K, TL,in=273 K and QL= 2 0 kW

The different curves in the figure refer to different total heat exchanger conductances and

it is seen that the larger the total heat exchanger conductance is, the more sensitive is the
allocation of the UAtotai. Nevertheless the optimum COP 1 is always obtained at the same

operating point, which are equal heat exchanger conductance and heat capacitance rates.
 66

Hence Figure 3.18 is a little misleading for heat capacitance rates CH/CL far off unity,

because such a system would probably not be designed.

1.00
.. fOALopt fUAtotal=16 kW/K

. fUA,total=4 kW/K

0,.7..............7:.. . ......................... .......... ........... . . , ..... L..........

0.75

0.50 . ............ _.

conditions of corresponding COP 1,max

-------------------.. ..... . .... .... ...--
0 .25 -------------- i--------

0.00 I A

0.1 1 10
cH
CT.

Figure 3.18 Optimum allocation of the total heat exchanger conductance UAtotai for
a single Carnot cycle system vs. external heat capacitance rates ratios
CH/CL with Cext, total= 2 kW/K, THin = 293 K, TL,in = 273 K, QL = 20 kW
and varying UAtotai

3.3.3.2 Heat Capacitance Rates

An investigation of the allocation of the total heat capacitance rate C exttotal on the two,

external streams leads to very similar results to these of section 3.3.3.1. The analytical
 67

simulation results in a even distribution of Cexttotai on the external streams, whereas this

is not always the case for the refrigeration cycle applying a pure refrigerant.

4.0

3.5
COP
3.0

2.5

2.0

1.5

1.0

0.5

0.0
0 10 20 30 40 50 60 70 80 90 100

100 (L
CL + CH

Figure 3.19 COP.. and COP1 vs. heat capacitance rates fractions CL for different
heat exchanger conductance ratios UAHI/UAL with C ext, total=2 kW/K,
= =
UAtotai=4 kW/K, TH,in 293 K, TL,in 273 K and OL=20 kW

An even allocation of heat capacitance rates is optimum only if the UAtotai is distributed

equal, but the cycle achieves the maximum COP 1 under these operating conditions. For

ratios of UAH/UAL < 1 is the optimum allocation shifted to larger ICL, whereas it is
shifted to larger CH for UAHUAL > 1 . The allocation of the Cext,totaI for the chosen

operating conditions was most sensitive for a total external heat capacitance rate of
Cexttota 1 = 2.7 kW/K. Nearly 60 % of the total heat capacitance rate should have been
 68

passed through the low temperature heat exchanger at a ratio of UAH/UAL = 0.1 kW/K.
It is less sensitive for other ratios.

0.60

flL,opt

0.55

0.50

0.45

0.40 L
0.1 1 10

UAH
UAL

Figure 3.20 Optimum allocation of the total heat capacitance rates Cext,total for a
single Carnot cycle system vs. external heat exchanger conductance
ratios UAH/UAL with UAtotal=4 kW/K, TH,in=293 K, TL,in=273 K,
QL =20 kW and varying Cext,total

It has been demonstrated, that the optimum allocation of Cext,total and UAtotaI for a

refrigeration system applying a pure refrigerant should be both equal in order to obtain

the optimum COP 1 . If one of these parameters is not allocated evenly, then the remaining

parameter is also not equally allocated for the corresponding optimum COP 1 .
 69

3.4 COMPARISON OF DIFFERENT REFRIGERATION

CYCLES
The COP for an arbitrary system in terms of external boundary conditions was developed

in chapter 3.1. First are the general Equations (3.28) and (3.29) recalled.

CCmin
TH,inFHCH,min( -1) - L 1 + HH -1

with

eL (TL,in + 1j.
EL CL,mnin - L)+11h
[Ci.~rnin ~QLSFL (3.27)

3.4.1 CARNOT REFRIGERATION CYCLE

The Carnot refrigeration cycle assumes a isothermal heat transfer in the evaporator and

condenser. Consequently the heat capacitance rates of the refrigerant are infinite, it is
CL -> oo and Ch -> oo. Klein [7] shows that

COP = TL,in - AT (1.7)

TH,in - (TL,in - AT)

where

AT = Q L EHCH + SLCL (1.8)

SHCH SLICL
 70

3.4.3 BRAYTON REFRIGERATION CYCLE

A constant heat capacitance rate Crefr = Ch = i for the refrigerant is assumed. The

Equations (3.28) and (3.29) reduce then to

COP= TLn - AT (3.32)

TH,in - (TL,in - AT)

where

AT'= 0 HCH,min Crefr + ELCL,min Crefr - ELCL,min EHCH,min (3.33)

EHCH,min ELCL,min Crefr

This equation holds for the ideal Brayton refrigeration cycle. The heat exchanger

effectiveness factors are defined by the Equations (3.15) and (3.16). The temperature
difference AT is a measurement of the performance of the system. The smaller AT, the
larger the COP.

3.4.4 OPTIMUM CYCLE

A expression for a partially optimized COP with equal heat capacitance rates in each heat

exchanger was given in the Equations (4.30) and (4.31).

It was shown, that the optimum cycle is obtained for equal allocation of the total
heat exchanger conductance, equal allocation of the external heat capacitance rates, and

equal heat capacitance rate of the refrigerant and external stream in each heat exchanger.
The COP for this cycle may be expressed as
 71

COP =HL TH,in, I (1 +

('1-Q[+
(jl+](3.34)
%(-)

where X is defined as

= oF
QL +11 (3.35)

The heat exchanger effectiveness and heat capacitance rates in the Equations (3.34) and

(3.35) are e =CL = 8 H (defined by Equation (3.15)) and C =CL = CH, respectively.

3.5 CHAPTER SUMMARY

A analytical simulation model was developed in this chapter. It is possible to specify
different heat capacitance rates of the refrigerant in the low and high-temperature heat

exchanger. However, the analytical model does not provide the temperature profile of the
refrigerant, as for the numerical model.

Important aspects and design guidelines for the optimum cycle are:

It was shown, that very high Q may be obtained, but the corresponding COP
was always very small for unreasonable high 92.

The potential improvement of the performance 92 increases significantly with

increasing gliding temperature differences GTDrefr.
* The £ increases with increasing total heat exchanger conductance UAtotai.
* The i decreases with increasing total heat capacitance rates Cext,total.
 72

The UAtotal should be allocated evenly on the two heat exchangers,

independent on C H, CL, THin, TL,in, QL

The Cext,totalshould be allocated evenly on the two external heat capacitance

rates independent on UAH, UAL, TH,in, TL,in, QL-

The COP does not vary significantly for moderate deviations of the optimum
distribution of UAtotal and Cexttotal-

The heat capacitance rate of a refrigerant with constant heat capacitance rate
Crefr = Ch = C, should be at least Crefr > min (CL, CH).

Design guideline for a refrigeration system undergoing iso-thermal heat transfer.

• The UAtotal and Cext,total should be also for this system allocated evenly for the

two heat exchangers.

The total heat exchanger conductance should not be allocated evenly if

UAH # 1 for CH CL
UAL
The total heat capacitance rate should not be allocated evenly if
CH #1 for UAH UAL
CL
 73

CHAPTER
FOUR

NARMs IN A STANDARD
VAPOR COMPRESSION CYCLE

In this chapter, a realistic simulation of a standard vapor compression cycle is reported.

A simulation program, Cycle 11 [14,16,17,18] is used to investigate the performance of a

specified refrigeration system. The use of a non-azeotropic mixture (NARM), as an

option to improve the performance of the system is investigated and compared to a pure
refrigerant. The optimum possible COP.. for this system is determined with the

Analytical Model from Chapter 3 and compared to the COP obtained with the simulation
from Cyclel 1.
 74

4.1 DESCRIPTION OF NARM

4.1.1. PHASE CHANGE PROCESS OF A NARM

A non-azeotropic refrigerant mixture (NARM) behaves differently from a pure refrigerant

or an azeotropic mixture. It may be described most easily as a fluid which changes its
or condensation.
composition in the liquid and vapor phases during boiling

NARMs have different vapor and liquid lines as shown in Figure 4.1. This

results in a different composition of the refrigerant in the boiling or condensing liquid

phase and the corresponding vapor phase. For example, a binary NARM with a mole
fraction of fAo and a temperature of T1 is heated isobarically to the liquid line. The

refrigerant with the lower boiling point will preferentially evaporate first at the
temperature T2 .

II
[- .+

f3,A,l f Ao

mole fraction fA

Figure 4.1 Evaporation process of a binary NARM in a Temperature-Concentration

diagram
 75

The composition of the vapor phase in equilibrium with the liquid is f2,A,v and the

composition of the remaining liquid is fA. The mole fraction fA of the less volatile

component A is larger than the initial mole fraction fAo. Consequently the boiling

temperature of the remaining liquid shifts to the component with the higher boiling
temperature for instance to the temperature T3 . The composition of the liquid and vapor

phase at that state is f3,A,1 and f3,A,v, respectively. At the temperature T4 the NARM
consists only of saturated vapor with the same composition fA0 as at the temperature T 1.

Hence, the isobaric evaporation and condensation occur over a temperature range.

This temperature range is strongly dependent on the mixture components and the

composition. The choice of the mixture composition offers an additional degree of

freedom which may be used as a design parameter. For instance, a NARM consisting of

two components of very different boiling temperature and with a suitable initial
composition will result in a mixture with a large temperature range.

4.1.2 COMPONENTS FOR A NARM

The use of any fully halogenated CFCs is to avoided. Refrigerants like R11, R12, R113,
R1 14 and Rl15 are in the Montreal Protocol [23] classified as group 1 chemicals, which

have to be replaced. Other aspects like low toxicity, non-flammability, chemical and

thermal stability, non-aggressive behavior towards design materials and commercial

availability should also be taken into account.

 76

4.2 DESCRIPTION OF THE SIMULATION MODEL

CYCLEll

The CYCLE11 model was developed by Domanski and McLinden [14,16,17,18] at the

National Institute of Standards and Technology. It is able to evaluate the performance of

a theoretical standard vapor compression refrigeration cycle. The investigated

refrigeration cycle works like the cycle described in section 1.3.1, but with an polytropic
compressor efficiency il smaller than unity and it accounts for pressure drops in the

condenser and evaporator.

4.2.1 HEAT EXCHANGER MODEL

Counterflow heat exchangers are assumed in the evaporator and condenser. The
performance of the heat exchangers is specified in terms of an average effective

temperature difference A Thx between the refrigerant and the external streams and

pressure drops in the condenser and evaporator.

Domanski [14] assumes for his model the same overall heat transfer coefficients,

U = const, of the heat exchanger and a linear pressure drop A p in the heat exchangers
with respect to the length L, dp/dL = constant. The heat exchanger average temperature

difference is defined as

-x
A Thx- UAhx (4.1)

Individual temperature differences are considered in different sections of the heat

exchangers. The two phase region in the condenser is broken into several individual
 77

sections, because the temperature profile of the refrigerant may be non-linear. Cycle 11

splits the two-phase portion of the heat exchangers into a number of sub-sections,

computes the log mean temperature of each of them and then evaluates the heat exchanger

average effective temperature difference consistently with Equation (4.4). The individual

rate equation between the refrigerant and the external stream is

Qi = UAi A Ti (4.2)

where Ai and A Ti are the individual heat transfer area and temperature difference in each

section. The.total heat transferred in the heat exchanger is

Qhx X Qi (4.3)

Domanski expressed the average effective temperature difference as a mean weighted

with the fraction of heat transferred in the individual sections of the heat exchanger.

_1 Q(4.4)
Thx Qhx A Ti

4.2.2 REFRIGERANT PROPERTIES

Morrison and McLinden [16] developed a routine for the calculation of the refrigerant

properties. In that are the refrigerant thermodynamic properties represented by the

Carnahan-Starling-DeSantis (CSD) equation of state [17].
 78

pVl 1 + y + y2 -y3 a (45)

RT (l-y)3 RT (V + b)

where R is the universal gas constant, T is the fluid temperature, V is the molar volume

of the mixture and p is the pressure. The values of the parameter a and b are strong

functions of the mixture composition and the temperature. The parameter y is defined by

b (4.6)
4V

McLinden [15, 24] showed that the CSD equation of state accurately represents the vapor

and liquid phase for CFC refrigerants and their mixtures. This routine was built in

Cycle 11 by Domanski.

4.3 SIMULATIONS WITH CYCLEll

A vapor compression cycle is considered with the following simplifications. No pressure

losses occur in the evaporator and condenser and no heat and pressure losses occur in the
manifold. The electric efficiency of the compressor is assumed to be unity.

The performance of the system is investigated following this procedure:

1. Specify the initial operating conditions, assuming arbitrary heat exchanger

conductance UA allocation on the condenser and evaporator

2. Optimize the COP with respect to the mole fraction fR2
 79

3. Match the gliding temperature differences GTDevap to GTDL,ext and

GTDcond to GTDH,ext in the evaporator and condenser, respectively

4. Optimize the COP with respect to the UA allocation

4.3.1 BASE CASE SYSTEM

The initial operating conditions are given in table 4.1. The chosen numbers are similar to

a small refrigeration application like a household refrigerator.

Electric motor efficiency 1

Polytropic efficiency 0.510.7/1.0
Compressor swept volume [m3 ] 0.170
Compressor RPM 1800
External fluid entering condenser [K] 293.1
External fluid exiting condenser [K] 323.1
GTD of the external stream in the condenser [K] 30
UA for condenser [kW/K] 0.35
External fluid entering evaporator [K] 283.1
External fluid exiting evaporator [K[ 268.1
GTD of the external stream in the evaporator [K] 15
UA for evaporator [kW/K] 0.25

Table 4.1 Operating conditions for the standard vapor compression cycle

Irreversibilities such as pressure drops in the heat exchangers are not taken into account,
because, by far, the largest irreversibilities, besides these generated by the heat transfer,
occur in the compressor. The potential improvement of the COP increases with larger

GTDext if the GTDrefr of the refrigerant mixture matches the temperature change of the
 80

external streams. The gliding temperature differences of the incoming and outgoing
external streams in the heat exchangers GTDL,ext and GTDH,ext are 15 and 30 K,

respectively. These GTD's will be varied in section 4.3.5.

A refrigerant mixture of R22/R141b is used to investigate this cycle. The

components R22 and R141b are possible replacements for the CFCs.

4.3.2 OPTIMUM COMPOSITION OF THE NARM

Figure 4.3 shows the COP vs. the mole fraction fR22 for a refrigeration system operating

at the conditions specified in Table 4.1. The different curves refer to different polytropic
compressor efficiencies, rj = 1.0, ril = 0.7 and ril = 0.5. The maximum COP is always

obtained at a mole fractions of about fR22 = 0.1. All three curves demonstrate the same

behavior. The COP obtained for a system employing the pure refrigerant R141b is in
every case about twice as high as the COP obtained when pure R22 is employed.

The COP of the NARM might be considered in a first approach as a weighted

average of the COPs of the pure components. A composition with a large mole fraction

fR22, the refrigerant which leads to a low COP, does not show any improvement of the
mixture COP relative to the COP of pure R141 b. Otherwise is for a mixture with a small

mole fraction fR22 a significant improvement of the COP obtained. The COP obtained
with such a mixture would always out performance the COP obtained with pure R141 b.
The increase of the COP is about 16% for the isentropic compressor and about 13% for

the systems with the lower polytropic compressor efficiencies.

 81

6.0

1.0
COPn=o
COP 5
4.0 . . 09..=0.5
. . .. . . .. . . . . . . .. ..-----
..--- --------- --
.--- ----. . .
--- .- .. . . . .. . . . . . .. . . . .. . . . .. .

. .............
.... A......
.....
................ ..... ...
..........................
2 .0. 08........
................

....
2. ............... .. " " "............ ............................. .............................

10.L______L . L_..
0.0

0 0.2 0.4 0.6 0.8 1

mole fraction fR22 (mixture R22/ R141b)

Figure 4.2 COP vs. mole fraction fR22 for a refrigerant mixture of R22/R141b
and different polytropic compressor efficiencies

It is of interest to know how large the heat transferred in the superheated region in the

condenser Qsh is and how well the GTDrefr of the refrigerant matches the GTDext of the

external streams for different mole fractions.

4.3.3 ACCOMPANYING SUPERHEAT

In Figure 4.4 is the mole fraction fR22 vs. the fraction of heat transferred in the condenser

in the superheated region fsh shown. The parameter fsh is defined as

 82

Qsh
fsh = Qcond - Qmain
Ocond Qcond (4.1)

where Qcond is the total heat transferred in the condenser, Qain is the heat transferred in

the two phase vapor liquid region in the condenser and Qsh is the heat transferred in the
superheated region in the condenser.

The fsh is small for NARMs at lower mole fractions fR22. The minimum occurs
about fsh = 0.3. Only about 5% or less of the heat is transferred in the superheated
region. Whereas the superheat is large at higher mole fractions fR22. A system using

pure R22 as the refrigerant instead of a mixture would run with a fsh of 20% or more.

25

20

. 15

Cm 10

3'-

0.2 0.4 0.6 0.8

mole fraction R22 (mixture R22/R141b)

Figure 4.3 Fraction superheat fsh vs. mole fraction fR22 for a refrigerant mixture
of R22/ R141b and different polytropic compressor efficiencies
 83

At these operating conditions is it due to the large superheat not possible that the
temperatures of the two streams in the condenser are well matched. A larger condenser

superheat is corresponding to a worse temperature matching and should be avoided.

4.3.4 DEVIATION OF THE OPTIMUM TEMPERATURE MATCHING

As described in Chapter 2, the optimum temperature matching is obtained for equal GTDs
in the evaporator (GTDL,ext = GTDevap) and for equal GTDs in the condenser

(GTDH,ext = GTDcond).

GTDcond~n GTDH ext fsh

I-GTDH,ext (sh)
KHin ---------- f

------- TL,in

Entropy

Figure 4.4 Location of gliding temperature differences

 84

A measurement for the deviation of the optimum temperature matching is introduced as V:

iV= A GTDevap + A GTDcond (4.7)

where the A GTDevap is the absolute difference of the GTDs of the refrigerant

GTDevap,refr and the external stream GTDL,ext in the evaporator.

A GTDevap = II GTDevap,refr I - I GTDLext II (4.8)

The absolute difference of the GTDs from the refrigerant and the external stream in the
condenser A GTDcond is broken into two regions. These regions, the liquid vapor phase

and superheated region, are weighted by the fraction of heat transferred in each region.

A GTDcond = A GTDcondmai n (1- fsh) + A GTDsh fsh (4.9)

where A GTDcond,main is the difference of the GTD of the refrigerant and the external

stream in the two phase region in the condenser and A GTDsh is the difference of the
GTD in the superheated region respectively. The A GTDcond,main is defined as

A GTDcond,main = II GTDH,ext (1-fsh) I - I GTDcond,refr,main II (4.10)

where GTDcond,refr,main is the GTD of the refrigerant in the two phase region and

GTDH,ext (1-fsh) is the GTD of the external stream in the two phase region.
 85

The deviation of the GTDs in the superheated region is determined by

A GTDsh = II GTDrefr,sh I - I GTDH,ext fshII (4.11)

where GTDrefr,sh is the GTD of the refrigerant in the superheated region and GTDH,ext fsh

is the GTD of the external stream in the superheated region.

50

40

30

20

10

0
0.2 0.4 0.6 0.8
mole fraction fR22 (mixture R221Rl41b)

Figure 4.5 Deviation of optimum temperature matching NVvs. mole fraction fR 22

for a refrigerant mixture of R22/R141b and different polytropic
compressor efficiencies

In Figure 4.5, the deviation of the optimum temperature matching V, is plotted as a

function of the mole fraction fR22 shown. The curves are very similar to each other and
two minima at the same mole fractions of about fR22 = 0.12 and fR22 = 0.8 are obtained
 86

for each of the systems with the different polytropic compressor efficiencies r1. V
increases with decreasing 11, due to the larger superheat occurring in the condenser. The

smaller minimum of V is obtained for mixtures at a mole fractions of about fR22 = 0.12.

This mole fraction corresponds to the mole fraction which leads to the maximum COP, as

indicated in Figure 4.3.

It could be expected that another peak for the COP in Figure 4.3 occurs at a mole fraction
of about fR22 = 0.8. At this conditions is Vi also relatively small and the temperature

matching of such a mixtures is relative good. The problem is, that the COP of the pure

refrigerant R22 is compared to the R141b very small so that the benefit of the relative

well matched GTDs is not sufficient to cause an increase of the COP at that large mole

fraction fR22.

An accompanying scaled temperature-entropy diagram of the system at minimum Nf of

= 7 K is shown in Figure 4.6 for a polytropic compressor efficiency of 1 = 0.5. The

in- and outlet temperatures of all fluids in the heat exchangers, the saturated vapor

temperature of the refrigerant in the condenser and the corresponding entropies are used
to present the diagram. The Cycle 11 program does not provide information concerning
the temperature profiles of the fluids in the heat exchangers so the temperature profile is

approximated as straight lines.

The temperature matching of the fluids in each temperature heat exchanger at this

operating condition (Table 4.1, fR22 = 0.12) indicates nearly parallel temperature profiles
in the evaporator and in the condenser. The temperature profiles of the streams are nearly
ideal matched, except in the superheated region. The amount of heat transferred in the
 87

superheated region is less than 5% of the total energy and the mismatch of the GTDs in

that region is negligible.

400

375

350
3-

325

300

275

I I I I I I I I I
I jvk

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.1

Entropy [kJ/kgK]

Figure 4.6 Temperature-Entropy diagram for the system with a polytropic

compressor efficiency of rl = 0.5

4.3.5 HEAT EXCHANGER CONDUCTANCE ALLOCATION

The GTD of the external streams may be altered until V -> min. Such a system transfers

the heat in the evaporator and condenser at the best possible temperature matching. The
outlet temperatures of the external streams are changed until V reaches a minimum. The

change of the GTD causes a change of the heat capacitance rates of the external streams,
because Cext = 0/ GTD and Q remains nearly constant. The refrigeration systems differ
 88

now not only in different compressor polytropic efficiencies 71, but also in different

external heat capacitance rates (Table 4.2). The GTDs of both external streams were
reduced, which results in larger heat capacitance rates of the external streams and hence in

a higher COP. This alteration increased the COP compared to the COP obtained with the
optimum mole fraction (section 4.2.2). The improvement is for ' =-0.5, i =0.7 and
= 1.0 in average about 9%, 2.5% and 5%, respectively.

7.0

6.0o
..............
00

5 .0 .........
........... .........................-----
-- ...............
.....
J..... ........
..............

1] 0.7

3.0
........... 1.=.0.5 . ................... .----
--- -_..... ........................
....... i...
....................
3 . ---- . 5. -------
........ . . . . . .

2. 0

0.2 0.3 0.4 0.5 0.6 0.7 0.8

f
UA,evap

Figure 4.7 COP vs. heat exchanger conductance fraction fUA,evap for a
refrigerant mixture of R22/ R141b and different polytropic compressor
efficiencies

The distribution of the total heat exchanger conductances on the evaporator and condenser
is now optimized. The heat exchanger conductance of the evaporator and condenser have
been initially fixed at UAevap = 0.25 kW/K and UAcond = 0.35 kW/K. The total heat
 89

exchanger conductance UAtotal = 0.60 kW/K remains constant, but the allocation is

varied. Figure 4.7 indicates, that the total heat exchanger size should be allocated nearly

evenly for the two heat exchangers, independent on the compressor polytropic efficiency
fl. The optimum distribution for the three different systems occurs if 48% of the total

heat exchanger conductance UAtotal are allocated to the evaporator. The optimum
distribution is always about UAcond = 0.315 kW/K and UAevap = 0.285 kW/K. A
moderate deviation of the optimum distribution does not decrease the COP significantly.

This indicates, that the UAtotal should be allocated even for that system.

4.3.6 COMPARISON OF ANALYTICAL MODEL AND RESULTS

FROM CYCLEll
The COPO for the system with ri = 1.0 obtained with the analytical ideal model from

Chapter 3 is also shown in Figure 4.7. In the plot the upper line represents the upper
limit of the COP. This COP is determined with considerations of heat transfer
mechanism. The COP of the refrigeration system with the isentropic compressor is very
close to that limit and reaches nearly the optimum COPO. The difference between the
COPs of the systems with the non-isentropic compressor and the optimum COP. is of

course larger.
The external heat capacitance rates have been changed to obtain the best temperature
matching (section 4.3.5). For this reason, the optimum COPO for the systems with the

different polytropic compressor efficiency r" varies. Table 4.2 shows the different

operating conditions and the corresponding COPs.

 90

polytropic low ext.heat high ext. heat COP COPCO COP

compressor capacit. rate capacit. rate (Cyclel 1) (AM) COPO
efficiency CL,ext [k] CH,ext []kWI

1.0 123.8 10- 3 87.6 10-3 5.25 6.00 0.88

0.7 135.9 10-3 94.6 10- 3 3.73 7.02 0.53
0.5 161.4 10-3 125.3 10- 3 2.91 7.76 0.38
Table 4.2 COPs calculated from Cyclell and AM

The external heat capacitance rates are calculated as the heat transferred in each heat

exchanger over the corresponding gliding temperature difference.

CLext -(Qevap (4.12)

GTDLext

Hext cond(4.13)
GTDH, ext

The COP of the cycle with rj = 1.0 falls only 12% short of the optimum COP. = 6.0

determined from the analytical model. Vapor-slip and reduced heat transfer coefficients
in the heat exchanger leads to smaller COPs than predicted. For the system with lower

polytropic compressor efficencies is the difference to the optimum COP larger. A parallel
temperature profile as obtained for xj -0 corresponds to identical heat capacitance rates

of the external stream and the refrigerarant. An energy balance for the refrigerant in the

condenser leads to

Qcond,refr - Ccond,refr GTDcond,refr (4.14)

 91

and for the external stream flowing through the condenser

QH,ext - CH,ext GTDext (4.15)

The heat exchanger as a closed system yields Qcond,refr = QH,ext and for equal GTDs

(GTDH,ext = GTDcond,refr) must be CH,ext = Ccond,refr.

The same result is for the evaporator attained, so that the heat capacitance rates of
the external stream and the refrigerant are also equal, CL,ext = Cevap,refr.

375

350

325

P-9
300

275

I I I I I I I I I
L.-)v

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1

Entropy [kJ/kgK]

Figure 4.8 Temperature-Entropy diagram for the system with a polytropic

compressor efficiency of 11 = 0.5 and optimized UA allocation

A accompanying T-S diagram for optimized fR22 = 0.08, fUAevap = 0.48 and 11 = 0.5 is
presented in Figure 4.8. The change of the initial UA allocation (for which the GTDs
 92

have been matched) to the optimum UA allocation alters the temperature profile a little.
Nevertheless are the temperature profiles nearly parallel lines.

4.3 CHAPTER SUMMARY

In this chapter was a binary NARM in a standard vapor compression refrigeration cycle
investigated and compared to the COP. determined from the Analytical Model developed

in Chapter 3.

The COP reaches a maximum for a refrigerant compositions which lead to the

smallest deviation of parallel temperature profiles of the external stream and the
refrigerant in each heat exchanger.
COP --->COPmax for A -0

A parallel temperature profile corresponds to identical heat capacitance rates of

the external stream and the refrigerant. The heat capacitance rate of the
external stream should match the heat capacitance in the heat exchanger.
Cext - Chx,refr

Matching the gliding temperature difference in the evaporator and condenser to

the corresponding gliding temperature difference of the external stream is not

sufficient to obtain an increase of the COP.

 93

The COPs obtained with a refrigeration system employing the pure refrigerants of

the mixture should be reasonable large, otherwise the benefit of the temperature
matching is undermined by the low COP of one of the components.

The mixture should consist of a less and a more volatile component,

depending on the desired gliding temperature difference.

The COP experiences an optimum for equal allocation of the total heat

exchanger conductance if the gliding temperature differences are matched to

each other.

The analytical COP defined by Equation (3.26) represents an upper limit for the
Carnot COP considering heat transfer mechanism and may be used as a design

goal.
 94

CHAPTER
FIVE

CONCLUSIONS AND RECOMMENDATIONS

5.1 CONCLUSIONS

The Carnot cycle places an upper limit on the COP of a refrigeration cycle. This Carnot

COP assumes a thermodynamic cycle with no internal irreversibilities. It does not

consider heat transfer mechanism, which are necessarily a irreversible process. The

Carnot COP may only be obtained for a refrigeration cycle with zero cooling capacity.
Furthermore the Carnot analysis assumes that the heat transfer from and to the cycle

occurs at a constant temperature. All this reduces its usefulness as a realistic design goal

for a refrigeration system.

The maximum possible COP for a standard vapor compression cycle, which takes

specified external boundary conditions into account and does not require an isothermal
heat transfer for the refrigerant and the heat transferring external fluid, is given by
Equations (3.28) and (3.29). The COP is determined by the cooling load, the heat
 95

exchanger conductances of the low- and high temperature heat exchangers, the inlet
temperatures of the external heat transferring fluids in each heat exchanger and their heat

capacitance rates. A numerical and an analytical simulation model for the refrigeration

cycle has been developed. The investigation of the cycle leads to several design

guidelines for the cycle obtaining the maximum COP. The most important are:
The slope of the temperature change (or gliding temperature difference (GTD)) of

the refrigerant in the heat exchangers should be parallel to that of the external
stream flowing through that heat exchanger. This result corresponds to equal
heat capacitance rates of the refrigerant and the external stream flowing through

the heat exchanger.

The total heat exchanger conductance should be evenly allocated between the two

heat exchangers independent on the cooling load, the inlet temperatures and heat

capacitance rates of the external streams.

The total external heat capacitance rate should be evenly allocated between the two

external heat capacitance rates independent on the inlet temperature of the external
stream, the cooling load and the heat exchanger conductances.

The potential performance improvement of the COP, Q, is defined as the COP of a

system employing a refrigerant undergoing non-isothermal heat transfer to a system

employing a refrigerant undergoing isothermal heat transfer. 0 is strongly dependent on

the gliding temperature difference GTD (temperature difference between the in- and outlet
temperature of the refrigerant or external stream in the heat exchanger) of the refrigerant
and the external streams. Assuming, that the GTD of the refrigerant matches the GTD of
the external system:
 96

The larger the GTD of the refrigerant, the larger the potential COP improvements

of the refrigeration system.

A change of the refrigerant GTD as the GTD of the external streams remains

constant results in a smaller COP.

A simulation of a standard vapor compression cycle with internal irreversibilities and real

non-azeotropic refrigerant mixtures (NARMs) developed by Domanski showed that the

COP of a specified system experiences a maximum for the best possible GTD matching

(GTD of the refrigerant in the heat exchanger nearly identical to the GTD of the external
stream flowing through that heat exchanger). In addition it was found, that the total heat

exchanger conductance for the two heat exchanger should be evenly allocated, as shown
for the ideal model.
The COP determined by Equation (3.26) and (3.27) offers a realistic upper limit

of the COP of a vapor compression refrigeration cycle employing a NARM. Contrary to

Equations (3.28) and (3.29) is it possible to specify the heat capacitance rates of the

refrigerant. This becomes important for refrigeration systems where the GTDs of the
refrigerant and external stream in each heat exchanger are not matched and Equations

(3.28) and (3.29) do not apply.

NARMs offer the potential for significant improvements of the COP in refrigeration

systems. These potential gains are strongly dependent on the fluid properties. A mixing

of two refrigerants often yields lower COPs than the COPs which would be obtained if
the pure refrigerants were used in the same system. The choice of the components is
difficult because several properties like ozone depletion potential, toxicity and commercial
 97

availability exclude many refrigerants. The refrigerant mixture should consist out of
refrigerants with different boiling point temperatures depending on the desired gliding

temperature difference of the refrigerant.

5.1 RECOMMENDATIONS FOR FUTURE WORK

In this work simulation models are used to investigate the performance of a refrigeration
cycle and to establish design guidelines. A verification of the design guidelines using

measurements of a real refrigeration cycle employing NARMs should be performed.

Data for the temperature profile of all fluids in the heat exchangers, the corresponding

heat capacitance rates, cooling load and COP are most important. The potential

performance improvements are significant, but experimental results have shown, that
they might be quite smaller than expected.
A refrigeration system using dedicated mechanical subcooling is of interest. In

such a refrigeration system, a second smaller vapor compression is applied, which task it
is to subcool the refrigerant flowing out of the main cycle's condenser. Studies have

shown [22], that large energy savings are possible for systems employing pure
refrigerants. A investigation with a NARM instead of a pure refrigerant might lead to

even larger energy savings. Potential performance improvements should be evaluated

and design guidelines should be established.
 98

APPENDIX

{ *************************************************************
Matthias Rauck November 1992
Calculation for the Coefficient of Performance (COP)
for an ideal Refrigeration cycle
Modeled with •
I) single Carnot cycle
II) Numerical Solution Model - Finite Difference Method (FDM) with n
Carnot Cycle
III) Analytical Solution Model (AM)

INPUT DATA:

Specification of the operating conditions of the refrigeration cycle:

{Number of individual Carnot Cycles in sequence, use replace command to change n

change 10{n} to #{n} }
n=10{n}
{cooling load QLjtotal }
QLtotal=10 {kW}
{Low and high temperature heat exchanger conductance}
UAL=5 {kW/K}
UAH=5
{Low and high external heat capacitance rates)
CextL=2 {kW/K}
CextH=3
 99

{Inlet temperature low (cold) external fluid in low temperature heat exchanger}
DUPLICATE a=n,n
TLext[i10{ n]=273 {K}
END

{inlet tenperature high (warm) external stream in high temperature heat exchanger)
THext[O]=313 {K)

{ I & IIl) Refrigeration cycle modeled with one and n Carnot Cycle
NOTE: Infinite heat capacitance rates of the working fluid for Carnot analysis)

{Define low and high inlet temperatures for single Carnot cycle)
DUPLICATE a=n,n
TLextin=TLext[10j{n]
END
THextin=THext[O]

{III) Analytical Solution Model (AM)

NOTE: Specify heat capacitance rates of the refrigerant and external stream)

{Defime low and high inlet temperatures of the external stream for AM)
THextanalyin=THext[O]
TLextanalyin=TLext[ 10(n)]

(NOTE: Put non corresponding paragraph in brackets (A or B), depending on

which definition of the heat capacitance rates is required
{
A) for optimum cycle (COP @ maximum): equal heat capacitance rates of
external stream and refrigerant (wf, working fluid))
CwfL=CextL
CwfH=CextH
(corresponding heat exchanger effectiveness factors for the
low and high temperature heat exchangers @ CwfL=CextL & CwfH=CextH!!
EpsanalyH=NTUanalyH/(NTUanalyH+ 1)
EpsanalyL=NTUanalyL/(NTUanalyL+1)
 100

{B) for arbitrary cycle (COP not @ maximum): unequal heat capacitance rates of external
stream and refrigerant (wf, working fluid) }
{specify heat capacitance rates of the refrigerant }
CwfL=l.5 {kW/K}
CwfH=2.5
{corresponding heat exchanger effectiveness factors for the low and high temperature
heat exchangers @ CwfL unequal CextL & CwfH unequal CextH !!}
EpsanalyH=(l-exp(-NTUanalyH*(l-A)))/(1-A*exp(-NTUanalyH*(1-A)))
A=min(CextH/CwfH, CwfH/CextH)
EpsanalyL=(1-exp(-NTUanalyL*(1-B)))/(l-B*exp(-NTUanalyL*(1-B)))
B=min(CextL/CwfL, CwfL/CextL)

CALCULATIONS:

Refrigeration cycle modeled with one Carnot Cycle as reference:

{Determination of Number of Transfer Units NTU for the low and high hx }
NTUCalH=UAH/CextH
NTUCa1L=UAL/CextL
{heat exchanger effectiveness factors for the low and high temperature hx }
EpsCa1H=1-exp(-NTUCa1H)
EpsCalL=I-exp(-NTUCalL)
{Energy balances and rate equations: (wf=refrigerant (workingjfluid)=Carnot cycle) }
QHCa1=CextH*(THextout-THextin)
QL_total=CextL*(TLextin-TLextout)
QHCal=EpsCa1H*CextH*(TwfH Cal-THextin)
QL total=EpsCalL*CextL*(TLextin-TwfLCal)
{Entropy transfer rate balance)I
QLjtotal/TwfLCa1=QHCa l/TwfHCa1

{compressor work)
WorkCal=QHCal-QL_total
{COP definition: (reversibel))
COPCal1 =QLtota/WorkCal1
 101

{Refrigeration cycle modeled with 1O{n} .Carnot Cycle:

{Creating the required equation set for n=10{ n } Carnot cycles:)I

DUPLICATE i=1,10{n }

{single cooling load per cycle:)

QL[i]=QLtotal/10{ n}
{heat exchanger effectiveness factors for the low and high temperature hx
for each individual Carnot cycle)I
EpsCanH[i]=1-exp(-NTUCa1H/10{ n))
EpsCanL[i]=1-exp(-NTUCa1L/10{ n})
{Energy balances and rate equations: wf=refrigerant (workingjfluid)=Camot cycle))
QH[i] =CextH* (THext[i]-THext[i- 1])
QL[i] =CextL* (TLext[i]-TLext[i- 1])
QH[i]=EpsCanH[i] *CextH*(Tcon[i]-THext[i- 1])
QL[i]=EpsCanL[i]*CextL*(TLext[i]-Tevap[i])

{Entropy transfer rate balance)I

QL[i]/Tevap[i]=QH[i]/Tcon[i]

{Entropy transferred with heat transfer in high hx)

{ a) for refrigerant)I
Scon[i]=QH[i]/Tcon[i]
{ b) for high external stream)
SH[i]=QH[i]/((THext[i- 1]-THext[i])/(ln(THext[i- 1]/THext[i])))
{Entropy transferred with heat transfer in low hx)
{ a) for refrigerant)I
Sevap[i]=QL[i]/Tevap[i]

I b) for low external stream)

SL[i]=QL[i]/((TLext[i- 1]-TLext[i])/(In(TLext[i- 1]/ILext[i])))
{Determination of the logarithm mean temperature differences:)
{logarithm mean temperature difference in low hx)
dtL[i]=((TLext[i- 1]-Tevap[i])-(TLext[i]-Tevap[i]))/lookl [i]
look1 [i]=ln((TLext [i- 1]-Tevap[i])/(TLext[i]-Tevap [i]))

{temperature difference between external streams)

dtextf[i] =THext[i]-TLext[i]
 102

{logarithm mean temperature difference in high hx)I

dtH[i]=-((THext[i- 1]-Tcon[i])-(THext[i]-Tcon[i]))/look2[i]
look2[i]=ln((THext[i- 1]-Tcon[i])/(THext[i]-Tcon[i]))

END

{compressor work)
W=sum(QH [i] ,i=1 ,10 {n })-QLtotal

{total rejected heat)I

QHCan=sum(QH[i],i= 1,10 {n })

{total entropy generated by heat transfer)

{in high temperature hx }
EntropyH=sum(SH[i]-Scon[i],i=1, 10{ n })

{in low temperature hx }

EntropyL=sum(Sevap[i]-SL[i],i=1,10 {n))

{COP for n Carnot cycles)

COPCan=QLtotal/(W)

{potential COP improvement)

COPimprCanvsCal=COPCan/COPCal {performance increase vs one cycle)
{Average temperature differences for one complete circulation (n cycles)
between the refigerant and external stream in the each hx)
dtevapLextav=sum(dtL[i] ,i=1,10{ n ))/10{n)
dtHextcon-av=sum(dtH[i] ,i= 1,10 {n ))/10{ n)
dtHextLext=sum(dtext-fl[i],i= 1,10{ n })/10{ n)

f Standard deviation for the temperature difference between

the external stream and working fluid)I
stdv-evapLext=(sum((dtevapLextav)A2,i=l,l0 {n )))A0.5/(10 {n )-1)
stdvHext_con=(sum((dtHexLcon-av)A2,i=1,10 {n )))AO.5/( 10 {n) -1)

{Analytical Model (AM):

{NOTE: We have different heat exchanger effectiveness factors depending on the

ratio of the heat capacitance rates of the external fluid and refrigerant.
 103

Possible: a) CwfH=CwfL and CextL=CextH than Eps=NTU/(I+NTU)

b) CwfH=CextH=CwfL=CextL than Eps=NTU/(1 +NTU)
c) CwfH=CwfL unequal to CextL=CextH than Eps=Epsanaly (earlier defined)

{Determination of Number of Transfer Units NTU for the low and high hx)
NTUanalyH=UAH/CminH
CminH=min(CextH, CwfH)
NTUanalyL=UAL/CminL
CminL=min(CextL, CwfL)

(Energy balances and rate equations:)I

QHanaly=CextH*(THextanalyout-THextanalyin)
QHanaly=CwfH* (Twf analyH in-TwfanalyH-out)
QHanaly=EpsanalyH*CminH*(Twf analyH-in-THextanalyin)
QL total=CextL*(TLextanalyin-TLextanalyout)
QL total=CwfL*(Twf analyL._out-Twf analyLin)
QL total=EpsanalyL*CminL*(TLextanalyin-Twf analyLin)

(Entropy transfer rate balance)I

QHanaly/TmeanwfH=QLtotalffmeanwfL
TmeanwfHl=(Twf analyH-in-Twf analyH-out)/n(Twf analyH-in/Twf analyH-out)
TmeanwfL=(Twf analyLout-Twf analyLin)/ln(Twf analyLout/TwfanalyLJn)

(COP defmition:)I
COPanaly=QL-total/(QHanaly-QL-total)

[potential COP improvement)

COPimpr-analyvsCan=COPanaly/COPCan

{Additional}
(equations (3.26) and (3.27)
(Unequal heat capacitance rates of external stream and refrigerant)I
Cop-general=QLtotal*alpha/(THextanalyin*(betha- l)*gamma-QLjtotal*alpha)

(where}
alpha=CwfH+betha*(EpsanalyH*CminH-CwfH)
betha=((QLjtotal*EpsanalyL*CminL)/(CwfL*(TLextanalyin*EpsanalyL*CminL-
QL_total))+)"delta
gamma=EpsanalyH*CminH*CwfH
delta=CwfL/CwfH
 104

{equations (3.28) and (3.29))

{ {Equal heat apacitance rates of external stream and refrigerant)I
{CwfH=CwfL; Cwfl=cwf}
COPpart-op_=(TLextanalyin-omega)/(THextanalyin-(TLextanalyin-omega))
omega=QLjtotal*(1/(EpsanalyH*CminH)+ 1/(EpsanalyL*CminL)- 1/(Cwfpartop_))
Cwfpart-op_=CwfH)}

{Deviation of Numerical and Analytical Model)

Accuracy=COPCan/COPanaly
)
 105

REFERENCES

1. Holusha, J., The Refrigerator of the Future,for Better and Worse , New York
Times, August 30, 1992

2. Moran, M. and Shapiro, H. N., Fundamentals of Engineering Thermodynamics,

John Wiley & Sons, New York (1988)

3. Baehr, H. D., Thermodynamik, 7th Edn., Springer-Verlag Berlin/Heidelberg

(1989)

4. Incropera, F. P. and DeWitt D. P., Fundamentalsof Heat and Mass Transfer,

3rd Edn., John Wiley & Sons, New York (1990)

5. Kays, W. M. and London, A. L., Compact Heat Exchanger, 2nd Edn,

McGraw-Hill, New York (1964)

6. Klein, S.A. and Alvardo, F.L., EES: Engineering Equation Solver F-Chart
Software, 4406 Fox Bluff Road Middleton, WI 53562 (1990)

7. Klein, S. A., Design Considerationsfor RefrigerationCycles, accepted for the

publication in the International Journal of Refrigeration, (November 1991)

8. Atwood, T., NARBs (Non-azeotropicRefrigerantBlends) The Promise and the

Problems, ASHRAE Paper No. 86-WA/HT-61, presented at the 1986 Winter
Annual Meeting, Dec 7-12, 1986

9. Vineyard, E.A., Sand, J.R. ,Statt, T.G., Selection of Ozone-Safe, Nonazeotropic

RefrigerantMixturesfor CapacityModulation in ResidentialHeat Pumps,
ASHRAE Transactions 1989, Vol. 95, Pt. 1, No. 3199
 106

10. Spauschus, H.A., Refrigerant mixtures challenges and opportunities,ASHRAE

Journal, pp. 28-41, November 1989

11. Atwood, T., (The ABCs of NARBs (NonazeotropicRefrigerant Blends),

ASHRAE Transactions, Vol. 91, Pt. 2, Paper HI-85-18 No. 1, 1985

12. Kruse, H., Hesse, U., Possible substitutesfor fully halogenated chlorofluor-
carbons using fluids already marketed, International Journal of Refrigeration,
Vol. 11, pp. 276-283, July 1988

13. Kruse, H., Schroeder, M., Kuever, M., Upmeier, B., Quast, U., Theoretical and
Experimental Investigations of Advantageous RefrigerantMixture Applications,
ASHRAE Transactions, Vol. 91, Pt. 2, Paper No. HI-85-27 No.4, 1985

14. Domanski, P.A. , McLinden, M.O., A Simplified Cycle Simulation Model for the
PerformanceRating of Refrigerants and RefrigerantMixtures, International
Journal of Refrigeration, Vol. 15, No. 2, pp. 81-88, 1991

15. McLinden, M.O. and Rademacher, R., Methods for Comparing the Performanceof
Pure and Mixed Refrigerants in the Vapor Compression Cycle, International
Journal of Refrigeration, Vol. 10, pp. 318-326, November 1987

16 Morrison, G., McLinden M., Two Refrigerant Mixtures and the Hard Sphere
Fluid, ASHRAE Transactions, Vol. 91, Pt. 2, Paper HI-85-18 No. 3, 1985

17. Morrison, G., McLinden M., Azeotropy in Refrigerant Mixtures, NBS Technical
Note, National Bureau of Standards and Technology, Gaithersburg, Md 20899,
USA

18. Morrison, G. , Gallagher, J.S., REPROP: A Thermodynamic PropertiesSoftware

Programfor Refrigerants and their Mixtures, Thermophysic Division at the
National Institute of Standards and Technology, Gaithersburg, Md. 20899,
USA
 107

19. Jung, D.S., Rademacher, R., Performance simulation of single-evaporator

domestic refrigeratorschargedwith pure and mixed refrigerants,International
Journal of Refrigeration, Vol.14, pp. 223-232, July 1991

20. Herold, H. E., PerformanceLimits for Thermodynamic Cycles, ASME Winter

Annual Meeting, San Francisco, December 1989

21. Ibrahim, O.M., Klein, S. A., and Mitchell, J.M., Optimum Heat Power Cycles for
Specified Boundary Conditions, ASME Journal for Gas Turbines and Power,
Vol 113, No. 4, pp. 514-521, 1991

22. Thornton, J.W., Supermarket Refrigeration Options, MS Thesis, Solar Energy

Laboratory, Dept. of Mechanical Engineering, University of Wisconsin -
Madison, 1991

23. Montreal Protocol on Substances that Deplete the Ozone Layer, Final Act, United
Nations Environment Program, Nairobi (1987)

24. Morrison, G., McLinden M.O., Application of a Hard Sphere Equation of State to
Refrigerants and RefrigerantMixtures, NBS Technical Note 1226, National
Bureau of Standards, Gaithersburg, Md 20899, USA