Académique Documents
Professionnel Documents
Culture Documents
Doctor of Philosophy
in
Engineering
By
Jahar Sarkar
Under the Supervision of
CERTIFICATE
The thesis entitled Transcritical carbon dioxide heat pumps for simultaneous
cooling and heating submitted by Mr. Jahar Sarkar, for the award of the degree of
Doctor of Philosophy in Engineering to the Indian Institute of Technology Kharagpur is a
record of bonafide research work carried out by him under our guidance and supervision.
The results presented in this thesis have not been submitted to any other
university or institution for the award of any other degree of diploma.
Professor
Deptt. of Mechanical Engineering
Indian Institute of Technology Kharagpur
Associate Professor
Deptt. of Mechanical Engineering
Indian Institute of Technology Kharagpur
Acknowledgements
I also like to express special thanks to material and component suppliers for their
cooperation during my experimental work. I will never forget the cooperation of Mr A.
Nandy and other workshop staff in the fabrication work.
I will remember the contributions of my hall mates and other fellow research
scholars for their strong mental support during my entire period of research in IIT
Kharagpur. I will ever remember my hall of residence, Dr. B. C. Roy Hall of Residence.
Last, but not the least by any means, in this auspicious moment, I shall forever
remember the contributions of my mother, brother, sister and my other family members
for their constant inspiration during the entire period of the work with great patience and
understanding.
Date
IIT Kharagpur
Jahar Sarkar
ABSTRACT
ii
humidity, the effects of DE, RAR, ambient temperature and air mass flow rate are
significant on system behavior.
A fully instrumented prototype of CO2 heat pump for simultaneous water cooling
and heating has been developed and tested. The gas cooler pressure has been successfully
controlled by simultaneously controlling the total mass of the system and degree of
opening of expansion device. Uncertainty analyses show that the test data is fairly good.
Results of transient analysis and performance with different operating conditions (water
mass flow rates and inlet temperatures, pressures, valve opening, etc.) have been
presented. The valve opening has significant effect near the valve closing condition.
Effect of water mass flow rates is not significant for both evaporator and gas cooler,
whereas the effect of water inlet temperature to gas cooler on the system performance is
significant. Compressor performance results have been presented as well. Recently
available correlations for heat transfer and pressure drop, used in the theoretical analyses,
have been validated by test data obtained from the experiments and showed reasonable
agreement for both gas cooler and evaporator. Comparison between test results and
simulation model prediction has been presented as well and shows reasonable agreement
and the trends are fairly similar. Finally based on the theoretical and experimental work,
several conclusions have been drawn and recommendations have been made for future
studies.
iii
iv
TABLE OF CONTENTS
Certificate
Acknowledgements
Abstract
List of figures
List of tables
xiii
Nomenclatures
xiv
1. Introduction
1.1
Motivation
1.2
1.2.1
Background
1.2.2
1.2.3
1.3
1.4
9
11
2. Literature review
13
2.1
Introduction
13
2.2
13
17
2.3.1
17
2.3.2
17
2.3.3
Multistage cycle
18
2.3.4
18
2.3.5
Ejector-expansion cycle
19
2.4
19
2.5
20
2.6
22
2.7
25
2.8
2.9
2.7.1
26
2.7.2
27
Component design
28
2.8.1
Compressor
28
2.8.2
Gas cooler
30
2.8.3
32
2.8.4
Evaporator
33
2.8.5
Other components
34
Application areas
35
2.9.1
Automotive air-conditioning
36
2.9.2
Automotive heating
38
2.9.3
Residential cooling
39
2.9.4
Residential heating
39
2.9.5
Water heating
41
2.9.6
42
2.9.7
Refrigeration applications
43
2.9.8
44
2.9.9
45
2.10 Summary
45
47
3.1
Introduction
47
3.2
48
3.2.1
49
3.3
51
3.4
53
3.5
3.4.1
53
3.4.2
57
3.4.3
66
68
3.5.1
68
3.5.2
68
3.5.3
Multi-staging
69
vi
3.5.4
3.6
Ejector-expansion device
70
Summary
71
4. Simulation of transcritical CO2 heat pump for simultaneous cooling and heating
4.1 Introduction
73
75
4.2.1
Compressor
76
4.2.2
Gas cooler
76
4.2.3
Evaporator
78
4.2.4
80
4.2.5
Expansion device
81
4.2.6
Performance parameters
81
83
85
91
96
4.6.1
Compressor
97
4.6.2
97
4.6.3
Expansion device
98
4.7 Summary
98
101
102
104
105
5.4.1
105
5.4.2
106
5.4.3
107
107
5.5.1
108
5.5.2
112
5.6 Summary
115
vii
116
6.1 Introduction
116
117
6.2.1
117
6.2.2
Numerical procedure
120
6.2.3
120
123
124
6.4.1
Compressor model
124
6.4.2
125
6.4.3
Evaporator model
127
6.4.4
130
6.4.5
Dryer Model
130
6.4.6
Fan model
131
6.4.7
131
6.4.8
132
132
134
137
6.7.1
138
6.7.2
140
6.7.3
141
6.7.4
143
6.7.5
144
6.7.6
146
6.8 Summary
147
149
7.1 Introduction
149
149
7.2.1
Compressor
150
7.2.2
Expansion device
151
7.2.3
Evaporator
152
viii
7.2.4
Gas cooler
153
7.2.5
Separator
154
7.2.6
Receiver
156
7.2.7
Condensing unit
156
7.2.8
156
7.2.9
157
157
159
163
169
171
7.8 Summary
172
175
8.1 Conclusions
175
178
Appendix A
180
Appendix B
182
References
183
ix
List of Figures
Figure 1.1
Figure 2.1
Figure 3.1
Figure 3.2
Figure 3.3
Figure 3.4
Figure 3.5
Heat pump cycle on the P-h plane for various gas cooler pressures
Figure 3.6
Figure 3.7
Figure 3.8
Variation of maximum system COP and optimum discharge pressure with cooler
outlet temperature
Figure 3.9
Figure 3.10
Figure 3.11
Gas cooler inlet temperature (oC) at optimum discharge pressure contour (0.5
increment of iso-temperature lines)
Figure 3.12
Figure 3.13
Variation of second law efficiency with discharge pressure for different gas
cooler exit temperatures
Figure 3.14
Figure 3.15
Figure 3.16
Figure 3.17
Figure 3.18
Figure 4.1
Figure 4.2
Figure 4.3
Figure 4.4
Figure 4.5
Figure 4.6
Figure 4.7
Variation of optimum discharge pressure and mass flow rate with area ratio
Figure 4.8
Figure 4.9
Figure 4.10
Figure 4.11
Variation of optimum pressure and mass flow rate with water inlet temperature
Figure 4.12
Figure 4.13
Figure 4.14
Figure 4.15
Figure 4.16
Figure 4.17
Figure 4.18
Figure 4.19
Figure 4.20
Figure 4.21
Figure 5.1
Figure 5.2
Figure 5.3
Figure 5.4
Figure 5.5
Figure 5.6
Figure 5.7
Figure 5.8
Figure 6.1
Figure 6.2
Figure 6.3
Figure 6.4
Figure 6.5
Figure 6.6
Figure 6.7
Figure 6.8
Figure 6.9
Figure 6.10
Figure 6.11
Figure 6.12
Figure 6.13
xi
Figure 6.14
Effect of re-circulation air ratio on COP, MER and SMER; At RAR = 0, heating
COP = 4.3, MER = 7.558 kg/h, SMER = 2.905 kg/kWh
Figure 6.15
Figure 6.16
Figure 6.17
Figure 6.18
Figure 6.19
Figure 6.20
Figure 7.1
A fully instrumented CO2 heat pump test facility for simultaneous water cooling
and heating
Figure 7.2
Figure 7.3
Figure 7.4
Figure 7.5
Figure 7.6
Figure 7.7
Figure 7.8
Figure 7.9
Figure 7.10
Figure 7.11
Figure 7.12
Figure 7.13
Suction and discharge pressure and mass flow rate with varying expansion valve
opening
Figure 7.14
Variation of cooling and heating output with discharge pressure for suction
pressure of 35 bar
Figure 7.15
Variation of cooling and heating output with discharge pressure for suction
pressure of 40 bar
Figure 7.16
Figure 7.17
Figure 7.18
Figure 7.19
Figure 7.20
Figure 7.21
Figure 7.22
Figure 7.23
xii
List of Tables
Table 1.1
Table 2.1
Table 2.2
Table 6.1
Table 6.2
Component irreversibilities
Table 6.3
Table 7.1
Table 7.2
Test matrix
xiii
Nomenclatures
A
cp
dw
exergy (kW)
e*
f ,
friction factor
hfg
irreversibility (kW)
Ldc
number of pass
Nu
Nusselt number
pressure (bar)
Pr
Prandtl number
xiv
Re
Reynold number
RH
rp
sf
St
Stanton number
T,t
temperature (K or oC)
T0
Td
tf
trs
UA
Vc
volumetric capacity ( kJ m3 )
Vs
vapour quality
Greek symbols
density (kg/m3)
efficiency
surface effectiveness
volumetric efficiency
xv
is
isentropic efficiency
II
Subscripts
1
state points
1a-6a
air
ai
am
moist air
amb
ambient
as
asbestos
bulk properties
core
cr
critical
comp
compressor
dis
discharge
ef
external fluid
exp
expansion device
ev
evaporator
evr
refrigerant in evaporator
evw
water in evaporator
fin, foam
fan
fab
fabrication
gc
gas cooler
xvi
gcr
gcw
inner, inlet
ihx
ins
insulation
saturated liquid
measured, mechanical
man
related to manufacturing
nb
nucleate boiling
outer, outlet
oper
operation
opt
optimum
pc
pseudocritical
Pr
Prandtl number
r, ref
refrigerant
ri
rw
suc
suction
sys
tube
tp
two-phase
v,g
saturated vapour
wi
water inlet
wo
water outlet
Subscripts
i
computational segment
pressure drop
xvii
xviii
Chapter 1
INTRODUCTION
1.1 Motivation
Refrigeration, heat pump and air conditioning systems play an important role in
modern civilization. Environmental control is one of the major requirements of a healthy
and non-pollutant living condition. Hence the prudent strategy would be to use advanced
technologies that are eco-friendly. Over the last few decades, refrigeration, airconditioning and heat pump industries have seen major changes caused by restrictions on
specific refrigerant use due to their detrimental effects on our climate. Two successive
international agreements; Montreal Protocol and Kyoto Protocol were introduced to
combat the twin menace of ozone layer depletion and global warming. The Montreal
Protocol (MP) on substances that deplete the ozone layer was adopted in September 1987
to phase-out the use of Ozone Depleting Substances (ODSs) within a fixed time period.
Ozone depleting Potential (ODP), a comparative measuring index, is fraction of the
ozone depleting potency of a substance compared to that of R11 or R12. Kyoto Protocol
(KP) was adopted at the third conference of parties to the United Nations Framework
Convention on Climate Change (UNFCCC) in December 1997, which has imposed
restrictions on refrigerants on the basis of GWP. Global warming Potential (GWP) is an
index that relates the potency of green house gas to the CO2 emission over a 100-year
period. The CFC refrigerants, although once considered to be the best refrigerants, were
abandoned due to high ODP. The chlorine free synthetic refrigerants based on HFCs,
which were taken as permanent replacement of CFCs, are also in the list of regulated
substances due to their considerably high GWP.
In this situation, industries are obviously looking for long-term solution for
refrigerant related problems. This has triggered a large number of innovative studies to
develop new technologies. Instead of continuing search for new chemicals, there is an
increasing interest in technologies based on ecologically safe natural refrigerants, i.e. air,
water, noble gases, ammonia, carbon dioxide and hydrocarbons. In principle, it must be a
much better solution to use naturally occurring substances as refrigerants; these are
compounds already circulating in large quantity in the biosphere and which we know are
harmless. Because of non-toxicity and non-flammability, carbon dioxide offers both
environmental and personal safety.
Although carbon dioxide is an old refrigerant, it was abandoned earlier after the
discovery of synthetic refrigerants. Now due to harmful effects of the synthetic
refrigerants on the environment, it has been revived as a potential refrigerant by the
seminal work of Lorentzen [1-3] in 1994. This has inspired subsequent development of
transcritical carbon dioxide cycles where the condenser gets replaced by a gas cooler.
Due to the low critical temperature of CO2, the gas cooler is operated above the critical
pressure and the evaporator is operated below that; hence the cycle is called transcritical
cycle. These systems have strong potential in two sectors: i) automotive air-conditioning
and ii) heat pumps. It was found that the use of a gas cooler with heat rejection taking
place over an unusually large temperature glide offers several unique possibilities such as
simultaneous refrigeration and hot water heating/steam production, heat pump drying,
simpler control of capacity, etc. Along with eco-friendliness, CO2 based vapor
compression systems have various advantages over conventional systems such as,
compatibility with normal lubricants and common machine construction materials, nonflammability and non-toxicity, greatly reduced compression ratio, easy availability, high
volumetric refrigerant capacity, and excellent heat transfer properties. Several unique
application possibilities along with various advantages motivated the huge research work
and industrial innovation recently in this area.
Background
This section outlines the invention, decline, and reinvention of carbon dioxide as a
refrigerant. The role of carbon dioxide in refrigeration systems goes back well over a
century and possibly had the largest impact on early food refrigeration and human
occupied space air conditioning. Though the Evans-Perkins process, upon which modern
refrigerator and air conditioners are based, was developed in 1834, it was in 1866 when
Thaddeus S.C. Lowe first harnessed carbon dioxide for ice production in his British
Patent, although Alexander Twining first proposed CO2 as a refrigerant in his British
Patent [4] in 1850. Following a period of further development, Carl Linde build the first
CO2 machine in 1881 in Europe. W. Raydt first developed compression ice-making
system using carbon dioxide in 1884. The carbon dioxide compressor was built by
Windhausen of Germany in 1880 and was awarded a British patent. Other landmark uses
of carbon dioxide refrigeration was made by Britain company J & E Hall, who had
purchased the Patent right in 1887 and after further development of technology, they first
installed a carbon dioxide based marine plant in 1890 and the first continuous production
of carbon dioxide refrigerating equipment was started in USA in 1897. In the field of
marine applications, CO2 dominated as a refrigerant in the first half of 20th century. Not
only were carbon dioxide based machines were growing in numbers in the late 1800s, but
improvements were continually being made as well on the basic cycle. J & E Hall
demonstrated that the efficiency of the vapour compression process could be improved
through the use of two-stage compression and made the first two-stage CO2 machine in
1889. In 1905, Voorhees developed what is now known as the multiple effect cycle,
which involves a separation of liquid and vapour at an intermediate stage in the
expansion process.
The advent of refrigeration and air conditioning in the late 1800s had an
enormous impact on thousands of industries. In Europe CO2 machines were often the
only choice at that time. In United States, CO2 was used in refrigeration systems from
about 1890 and in comfort cooling from about 1900. Carbon dioxide was used with brine
distribution for most ship installations and also in many stationary ones. Until about
1940, along with carbon dioxide, ammonia was used for medium and large stationary
systems and also sometimes in ships, often with brine as a secondary refrigerant, but
increasingly with direct cooling. Sulphur dioxide was used as well for household
equipment and small commercial applications, but occasionally for capacities up to
several hundred kW. But use of ammonia and sulphur dioxide was being limited in
ozone layer has led to the Montreal Protocol and universal banning of most CFC and
HCFC compounds. This has forced refrigeration engineers to search for new refrigerants
such that there is no degradation of performance compared with that of proven CFC and
HCFC technology. HFCs are presently used in newly produced refrigeration and air
conditioning systems. HFCs are substances with zero ODP, and which exhibit thermal
and transport properties similar to CFCs and HCFCs. Hence this new class of fluids may
be used with machinery already designed for CFCs and HCFCs with minor
modifications. R134a is one of the HFC refrigerants extensively used in automobile air
conditioning. But R134a may be decomposed by sunlight in the troposphere resulting in
formation of acid and poisonous substances. So similar predicaments have occurred from
the release of many other new chemicals to the environment. In this situation it does not
seem very sensible to replace the CFC/HCFCs with a new family of related
hydrocarbons, equally foreign in nature, to be used in quantities of hundreds of thousands
of tons every year, and hence the whole industry was searching for viable refrigerant
alternatives.
In principle, it must be a much better solution to use naturally occurring
substances as refrigerants; these are compounds already circulating in large quantities in
the biosphere and which we know are harmless. Carbon dioxide is one such naturally
available old refrigerant, which has been completely abandoned for more than 40 years,
and is arguably the best future refrigerant both in air conditioning and heat pump
applications due to its environmentally benign nature. In the early 1990s, Norwegian
Professor Gustav Lorentzen proposed a transcritical carbon dioxide cycle where the high
side pressure can be controlled by throttling valve. Main difference between this cycle
and conventional cycle is that heat rejection occurs in supercritical regimes due to very
low critical temperature (31.2 oC) and the condenser gets replaced by a gas cooler.
Automobile air-conditioning, a sector that dominated the global CFC emissions, is one of
the intended applications of this system, along with applications where non-toxic and
non-flammable refrigerants are needed. The potential for more compact components due
to high pressure was also another interesting feature. In 1993, Lorenten and Petersen
published experimental data on a prototype CO2 system for automobile air-conditioning
and comparison was made with a state of the art R12 system with equal heat exchanger
dimensions and design point capacity [1]. They showed that the number of practical
factors made the efficiencies of two systems nearly equal, although simple cycle
calculation indicated lower COP for CO2 system. Based on their results the interest in
normal boiling point is low ( 78.4 oC), CO2 cannot be used down to that temperature due
to higher triple point temperature ( 56.6 oC). The main difference with other refrigerants
is the low critical temperature with comparatively higher critical pressure of CO2, which
is responsible for very high system pressure compared to others. However this gives an
advantage of very high volumetric capacity compared to others, which can lead to a
compact design. Due to a gliding temperature in the gas cooler, carbon dioxide can be
effectively used in heat pump applications. This offers very wide application ranges
compared to others as shown in Figure 1.1. Although the subcritical cycle of CO2 is
applicable to only low temperature applications, the transcritical CO2 cycle can cover
cooling up to 50 oC and heating up to 120 oC approximately.
Table 1.1 Comparison with other refrigerants [2, 6, 7]
Refrigerants
Chemical formula
Molecular. Weight
N.B.P., C
Triple point, C
Critical Pressure, bar
Critical Temp., C
Sat. Pr. (0C), bar
ODP
GWP* 100 years
20 years
Flammable or
Explosive?
Toxicity
Vol ref. capacity at
0C (kJ/m3)
R22
CHClF2
86.48
40.80
160.0
49.88
96.0
4.976
0.05
1500
4100
R134a
C2H2F4
102.03
26.15
96.6
40.56
101.1
2.929
0.0
1200
3100
R717
NH3
17.03
33.35
77.7
112.97
113.0
4.304
0.0
0.0
0.0
R290
C3H8
44.1
42.10
187.1
42.52
96.8
4.712
0.0
0.0
0.0
R600a
C4H10
58.0
11.60
159.6
36.40
134.7
1.564
0.0
0.0
0.0
R744
CO2
44.01
78.40
56.6
73.72
31.1
34.80
0.0
1(0)**
1(0)
No
No
Yes
Yes
Yes
No
Yes
Yes
Yes
No
No
No
4344
2860
4360
3870
1509
22600
* Global warming potential in relation to CO2 with 20 and 100 years integration time
(IPCC 1990, 1992)
** Abundant amounts of CO2 are recovered from waste gas. Thus the effective GWP of
commercial carbon dioxide, for instant use as refrigerant is 0.
140
120
Isobutane
Propane
20
Ammonia
40
R134
60
R22
Temperature, C
80
Carbon dioxide
100
0
-20
-40
-60
Refrigerant
8. All properties and characteristics of carbon dioxide are well known and
thoroughly documented. Further toxicity testing is not required.
9. CO2 has zero ozone depletion potential (ODP) and zero effective GWP, the latter
because more than sufficient quantities of CO2 are recovered from industrial
waste gas, e.g., in oil refineries or ammonia production plants. As a natural
constituent of the biosphere, CO2 will not have any unexpected long-term effects
on health or the environment.
10. Applicable temperature ranges are wider than other pure refrigerants.
1.2.3
commercialize it. Coca Cola and Kirin is in testing phase of CO2 based vending machine,
and planning commercial rollout in selected countries. Denso has developed first CO2 air
conditioning system. Visteon of Europe developed car air-conditioning system using CO2
as refrigerant. Nissan Motor Co., Ltd. is willing to launch the fuel cell vehicles (FCVs)
equipped with an CO2 air-conditioning system, jointly developed by Nissan and Calsonic
Kansei Corp. Danfoss A/S and Hydro Alunova have established a joint venture to
develop a new generation of aluminium tubes for use in automotive air conditioning
systems. Denso and DaimlerChrysler AG are working together to develop a carbon
dioxide air conditioning system for a Mercedes vehicle. Sanyo and Danfoss are main
suppliers of CO2 hermetic compressors. Tecumseh, USA has also developed its own CO2
compressor. Mitsubishi Heavy Industries Ltd developed CO2 scroll compressors for air
conditioning application. Dorin of Italy has been manufacturing semi-hermetic CO2
compressors lately [8].
optimization of cycle and components, and detailed theoretical and experiments studies
on simultaneous cooling and heating applications.
Simulation and optimization of CO2 heat pump for simultaneous heating and cooling
application, and system irreversibility analysis
CO2 heat pump dryer simulation, validation with experimental data and optimization
heat pumps for simultaneous cooling and heating. For theoretical analysis, a complete
property code has been developed for thermophysical and transport properties of CO2.
The detailed thermodynamic optimization of transcritical CO2 cycle has been carried out
and effect of cycle modifications on it has been studied. The CO2 heat pumps with
internal heat exchanger have been simulated to study its performance for simultaneous
water heating and cooling applications Energetic and exergeric optimization of such
systems have been done as well. A prototype heat pump including a well instrumented
test loop has been developed; performance tests have been conducted and the results from
mathematical models have been validated with these experimental results. Experimental
validation of available heat transfer and pressure drop correlations in both gas cooler and
evaporator have been carried out.
The exergetic optimization of evaporator and gas cooler is another aspect of this
research work. One of objectives is to model a CO2 heat pump dryer system, followed by
experimental validation and the study of performance characteristics. Since the
transcritical CO2 cycle is different from other refrigerant-based cycles, it is quite
interesting how the CO2 heat pump dryer behaves. Two process applications:
simultaneous water cooling and heating, and heat pump drying, are the areas where such
a research study could be beneficial to agro-based industries.
10
11
12
Chapter 2
LITERATURE REVIEW
2.1 Introduction
This chapter presents a survey of past transcritical CO2 cycle based refrigeration,
air-conditioning and heat pump investigations. Although CO2 was used as a refrigerant
in the early 1900s and some research activities such as modification of basic cycle and
compressors to improve the performance also were carried out; those technologies appear
rather ancient now. During several decades in the past, ground breaking changes have
been made on test facilities, instrumentation, manufacturing technique, etc. So after the
revival of CO2 as a refrigerant in transcritical cycle, researchers have started thinking
from the basics and even some old concepts and information have been employed. Hence
in this chapter, the detailed research developments of CO2 as a refrigerant with respect to
thermodynamic cycle, heat transfer, designs and application related issues will be
reviewed.
13
cooling of CO2 gas takes place is called a gas cooler, which replaces the condenser of
conventional systems.
100
Isotherm
90
Pressure (bar)
80
COPc =
70
q
w
60
50
40
30
20
specific enthalpy (kJ/kg)
( h1 h3 ) / ( h2 h1 ) ]
pressure yields a maximum COP for COPc / P = 0 at pressure (P) defined by [10]:
14
h3
h2
= COPc
P s
P T
(2.1)
This is the optimum pressure where marginal increase in capacity equals marginal
increase of work. Since this method is time consuming for actual cycles, Petersen et al.
[9] used simulation technique to optimize the high side pressure. Kauf [11] considered
component performance data and optimized the high side pressure in terms of ambient
temperature. Subsequently, Liao et al. [12] showed that the optimum pressure mainly
depends on outlet temperature of gas cooler, evaporation temperature and compressor
performance and they obtained a correlation for optimum heat rejection pressure in terms
of appropriate parameters for specific conditions, based on cycle simulation. Vaisman
[13] applied the modified cryogenic approach and showed that the maximum potential
COP is achieved when the it is equal to the product of three fundamental parameters,
density, Joule-Thomson coefficient and isobaric heat capacity determined at the
discharge pressure and ambient temperature, which is basically similar to the result
obtained from fundamental approach stated before. Srinivasan et al. [14] obtained
correlations for maximum COP and exergetic efficiency based on cycle analysis.
High-side pressure regulation can be applied to maintain the maximum COP and/or
to regulate the heating or cooling capacity. For the supercritical operation, high side
pressure is determined by the relationship between refrigerant charge (mass), inside
volume and temperature [5],
P = P ( v, T ) = P (V m , T )
(2.2)
Varying the refrigerant charge (m) in the high pressure side of the circuit,
The first technique is comparatively easy and common and can be achieved by
controlling the expansion valve opening. To avoid flooding or dry out of evaporator, a
different buffer system can be used for this case [5]. The inside volume of high pressure
side is varied by using a pressure vessel or cylinder of adjustable volume. The last
method is actually a passive scheme where refrigerant charge/volume conditions are
15
adopted to change the pressure when temperature varies. Casson et al. [15] proposed
another innovative throttling system, which consists of a differential valve, separator and
thermostatic expansion valve to control the high side pressure optimally as well as to
control the superheat. The proposed system showed an intrinsic self-adjusting capability
that led to COP values quite close to the maximum level when a fixed suitable value of
the differential pressure is chosen, even if the temperature of the secondary fluid varies to
a large extent.
It has been shown [2,5] that the thermodynamic loss in heat rejection is higher
compared to other refrigerant based cycles due to the gliding temperature heat rejection.
This loss will reduce in water heating (or coolant with single phase heat transfer) by
proper design of the counterflow heat exchanger. This feature can be utilized in heat
pumps for tap water heating or hydronic heating systems. In applications where the
rejected heat is not of interest, the gliding temperature is not an advantage since the
average temperature of heat rejection becomes higher than necessary. For heat pump
applications, the influence of evaporator temperature on the heating capacity and heating
COP is smaller compared to that of other refrigerants, which enables the CO2 system to
maintain a high heating capacity at low ambient temperature [16]. By raising the pressure
above optimum value, increase in capacity can be obtained further. However in actual
operation this depends on factors like maximum allowable pressure, maximum motor
load and compressor temperature limitations. Where rejected heat is not needed, the
thermodynamic losses due to heat transfer can be reduced by maintaining approach
temperature as low as possible; proper heat exchanger design can give temperature
approach of a few degrees. Owing to large throttling losses in CO2 system, the cooling
COP is more sensitive to gas cooler exit temperature compared to condenser exit
temperature in case of other refrigerants [16]. The close temperature approach, obtained
in CO2 gas coolers, therefore contributes significantly to practical COP improvement.
16
compressor work with use of internal heat exchange in the system, the influence of the
internal heat exchange on the system overall efficiency depends on the working fluids
and operating conditions. The effect of internal heat exchange on CO2 transcritical cycle
is found to be marginal. The internal heat exchange can increase the COP by a maximum
of 7% [17]. In case of high lift application, where internal heat exchange may produce
compressor discharge temperature high enough to damage the lubricant, the internal heat
exchanger may employ a parallel-flow configuration. Boewe et al. [18] showed that the
enhancement on cycle efficiency can be a substantial 25%, because of the relatively high
irreversibility at the expansion device in the standard transcritical cycle. The internal heat
exchanger brings lower and higher side pressures close together at optimal condition,
creating opportunities for using less precise and simpler control system and strategies [5].
2.3.2
device (expander), which has potential for COP improvement. Work recovery turbine
with isentropic efficiency of 60%, may reduce contribution to total irreversibility by
17
about 35% and cause an average increase of COP by 25% [17]. Heyl and Quack [19]
discussed various cycles with expanders and showed the design and results of a freepiston expander concept. Kim et al. [5] have discussed many subsequent developments
on expander devices. Huff et al. [20] analyzed three types of expanders and carried out a
comparative study with a baseline R22 system, Due to the high pressure difference, the
effect of expander efficiency on the system COP is very significant for a CO2 system
whereas it is negligible for the R22 system. Quack et al. [21] proposed the integration of
three-stage expander in CO2 refrigeration system with two-stage compression with intercooling. They installed a liquid-vapor separator between second and third stage of
expansion to get optimum performance.
2.3.3
Multistage cycle
Performance of the CO2 transcritical cycle can be improved by using multistage
investigations
on
the
multistage
transcritical
CO2
cycles
for
tubes in evaporator to handle high pressure without adding weight or bulk. But the
challenging problem is how to distribute the two-phase refrigerant uniformly after
expansion device into the many tube passes through the header. Under normal condition
the void fraction at the evaporator inlet exceeds 0.8 [5] and uniform distribution of liquid
and vapor is very difficult due to combined effect of surface and gravitational forces. One
option to deal with this problem is bypassing the vapor around the evaporator and
allowing only liquid to enter the multi-pass evaporator. An experimental comparison to a
18
conventional direct expansion system revealed that using the flash gas bypass, the
cooling capacity and the COP improves by up to 9 and 7% respectively [22]. At the same
time, the refrigerant side heat transfer coefficient improves and the refrigerant side
pressure drop reduces significantly.
2.3.5
Ejector-expansion cycle
In transcritical CO2 cycle, regenerating expansion energy and increasing refrigerant
19
20
Liao et al. [36] experimentally investigated heat transfer from supercritical CO2
flowing in horizontal mini/micro circular tubes cooled at constant temperature and
showed that the effect of buoyancy force on supercritical CO2 cooling became significant
as the tube diameter decreased. Experimental results also indicated that the previous
correlations for large tubes deviated significantly. Based on the experimental data authors
developed a correlation for the axially average Nusselt number for forced convection of
supercritical carbon dioxide cooled at constant temperature. Jiang et al. [37] also
observed significant influence of buoyancy on convective heat transfer of carbon dioxide
at supercritical pressures in vertical mini-tubes and in porous media.
Based on numerical and experimental study (counter-flow, CO2 inside tubes, water
in the outer annulus; stainless steel tube with OD of 6.35 mm and thickness of 0.815
mm), Pitla et al. [38] presented a new correlation based on mean Nusselt number, where
Nu at constant thermophysical property was evaluated from Gnielinski correlation.
Presence of lubricant reduced heat transfer and increased the pressure drop. Scalabrin et
al. [39] developed a neural network model to estimate the heat transfer coefficient of CO2
in supercritical cooling based on data available in the literature and showed that the
model very accurately predicts the heat transfer coefficient data presented by Olson
whereas the Pitla correlation under-predicts the data.
Yoon et al. [40] presented the experimental data for heat transfer and pressure drop
characteristics during supercritical cooling of CO2 in horizontal copper tubes with ID of
7.73 mm and with an inlet pressure range of 75 88 bar. Conventional single-phase
pressure drop correlation with friction factor calculated from Blasius equation accurately
predicted the measured pressure drop for CO2 cooling. However, the authors showed that
most of the existing correlations for supercritical heat transfer coefficient under-predicted
the measured data and hence a new empirical correlation was proposed for the near
critical heat transfer coefficients. Zhao et al. [41] experimentally investigated the heat
transfer characteristics of supercritical CO2 cooling in a microchannel heat exchanger
(tube inner diameter of approximately 1 mm). Authors showed that refrigerant mass flow
rate, pressure and temperature highly dominated the heat transfer characteristics.
21
Recently, Dang et al. [42] experimentally investigated the heat transfer coefficient
and pressure drop for supercritical carbon dioxide cooled in four horizontal circular tubes
with different inner diameters ranging from 1 to 6 mm and showed the effect of mass
flux, pressure, heat flux and tube diameter. Based on the test data, authors established the
modified Gnielinski correlation of heat transfer coefficient using the reference
temperature method for supercritical carbon dioxide cooling in horizontal tubes. Authors
also compared the heat transfer coefficients obtained from numerical simulation applying
four different turbulent models and showed that the JL model (low Reynolds number
k model by Jones and Launder) showed the best agreement with the experimental
data [43]. Table 2.1 shows some of the correlations presented in recent times for
estimating heat transfer coefficients in a gas cooler.
Table 2.1 Recent transcritical heat transfer correlations for CO2 cooling
Authors (Year)
Fang (1999)
Tube
material
NA
Diameter,
mm
NA
Flow rate,
kg/min
NA
Pressure range,
bar
NA
NA
0.8
NA
NA
SS
0.5 2.21
0.02 0.20
74 120
SS
6.35 OD
1.20 2.35
94 134
Cu
7.73 ID
0.63 1.27
75 88
Cu
16
0.04 0.68
80-100
* NA = Not available
22
indicated regimes of convective boiling at high mass flux and low evaporative
temperature, and nucleate boiling regimes at low mass flux and higher evaporating
temperature. At most conditions, the heat transfer coefficient increased up to vapor
quality of 0.9, but the behaviour is quite different above evaporating temperature of 5 oC,
with decreasing heat transfer coefficient at increasing vapor quality. Zhao et al. [45]
conducted tests on flow boiling of CO2 in microchannel (stainless steel tube) for mass
fluxes of 250-700 kg/m2s and heat fluxes of 8-25 kW/m2 and results indicated that under
identical testing conditions, the heat transfer coefficient of CO2 is found to be much
higher than that of R134a with much lower pressure drop. Zhao et al. [46] studied flow
boiling of CO2 with miscible oil in microchannels (0.86 mm diameter tube) and results
indicated that larger oil concentration degrades the heat transfer coefficient significantly,
while smaller oil concentrations (< 3%) at low vapor quality (x < 0.45) moderately
enhanced the heat transfer coefficient.
Yun et al. [47] experimentally investigated the boiling heat transfer and dry-out
phenomenon of CO2 in horizontal smooth tube (ID=6 mm) for saturation temperature of
5 to 10 oC and mass fluxes of 170 to 320 kg/m2s. They found that the heat transfer
coefficient of CO2 decreases with an increase of quality due to a lower dryout quality and
dominance of nucleate boiling compared to the conventional refrigerants. Authors
showed that Gungor and Winterton correlation [48] exhibits poor prediction at low mass
flux whereas good agreement at higher mass flux. Yun et al. [49] experimentally showed
that the dry-out phenomena of CO2 are similar to water in many respects while the effect
of mass flux on dry-out showed dissimilar behaviour.
Pettersen [50] conducted experiments on flow vaporization heat transfer coefficient
and pressure drop of carbon dioxide in extruded microchannel tubes with 25 flowchannels (ID=0.8 mm). Flow pattern maps were presented at an evaporation temperature
of 20 oC and the heat transfer coefficients were correlated using asymptotic models for
combinations of nucleate boiling, and convection evaporation, dryout inception and post
dryout heat transfer. Flow visualization showed the dominance of intermittent and
annular regimes, although the latter one became more significant at high mass flux.
Pettersen [51] separately visualized the flow pattern of CO2 flow boiling in microchanel
23
tube of 0.98 mm OD using a high-speed camera at temperatures 20oC and 0oC and for
mass flux ranging from 100 to 580 kg/m2s. The intermittent (slug) flow dominated at low
quality whereas the wavy annular flow with entrainment of droplets dominated at higher
quality. The aggravated dry-out problem reported from heat transfer experiments at high
mass flux could be explained by increased entrainment. The flow pattern observations did
not fit generalized maps or transition lines reported in the literature.
Yoon et al. [52] conducted experiments on evaporative heat transfer and pressure
drop of CO2 in a seamless SS tube with an inner diameter of 7.53 mm and length of 5 m,
for saturation temperatures of 4 to 20 oC, mass flues of 200 to 530 kg/m2s and heat
fluxes of 12 to 20 kW/m2. They showed that at low quality region, heat transfer
coefficient has a tendency to increase slightly as quality increases, with the increase in
saturated temperature heat transfer coefficient increases up to about 5 oC, after that heat
transfer coefficient decreases in low vapor quality region. A new correlation was
proposed to predict critical quality where liquid film breaks down and the heat transfer
coefficient for CO2, and also developed a correlation for frictional two-phase multiplier
for pressure drop calculation. Thome et al. [53] generalized the flow pattern map for flow
boiling of CO2 and developed a general correlation as an asymptotic model including
both nucleate and forced convection terms for flow boiling heat transfer coefficient for
CO2 based on data covering five tube diameters from 0.79 to 10.6 mm, mass velocities
from 85 to 1440 kg/m2, heat fluxes from 5 to 36 kW/m2 and saturation temperature from
25 oC to 25 oC.
Huai et al. [54] experimentally studied boiling heat transfer and pressure drop of
CO2 flowing in a multi-port extruded aluminum test section having 10 circular channels,
each with an inner diameter of 1.311 mm for the evaporation pressure ranging from 39.9
to 53.8 bar, inlet temperature of CO2 from 3.08 to 16.96 oC, heat flux from 10.1 to 20.1
kW/m2, mass velocities from 131.4 to 399 kg/m2s and vapour quality from 0.0 to 1.0.
Results showed that two-phase CO2 flow exhibited a higher heat transfer coefficient than
that of single-phase liquid or vapor flow and once the dry-out occurred the wall
temperature increased and the heat transfer decreased rapidly. Results also showed that
the mass velocity and the applied mass flux have significant effect on the flow boiling
24
characteristics. The measured heat transfer coefficients were found to deviate from those
obtained from correlations reported in the literature significantly. Table 2.2 shows a
summary of studies carried out recently on flow boiling aspect of CO2.
Table 2.2 Summary of experimental investigation on flow boiling of CO2
Investigator
Tube
Specifications
Mass flux
(kg/m2s)
Heat flux
(kW/m2)
Sat. temp.
(oC)
Studies
Bredesen et
al., 1997
Aluminum,
7 mm ID
200 - 400
3-9
-25 to 5
Local heat
transfer
Zhao et al.,
2000
Stainless Steel,
0.86 mm dia.
250 - 700
8 - 25
5 to 15
Heat transfer,
pressure drop
Zhao et al.,
2001
Stainless Steel,
0.86 mm dia.
250-700
8 - 25
5 to 15
Heat transfer
Yun et al.,
2003
Stainless Steel,
6.0 mm dia.
170 - 320
10 - 20
5 to 10
Heat transfer
Yun et al,
2003
7.2 - 48.1
0 to 10
Dry-out
Pettersen,
2004
Aluminum,
0.8 mm dia.
190 - 570
5 - 20
0 to 25
Heat transfer,
pressure drop
Yoon et al.,
2004
Stainless Steel,
7.53 mm dia.
200 - 530
12 - 20
-4 to 20
Heat transfer,
pressure drop
Huai et al.,
2004
Aluminum,
1.311 mm ID
-3 to 17
Heat transfer,
pressure drop
25
2.7.1
large pressure difference, but at modarate pressure ratio. Pettersen [16] showed that the
displacement of R134a machine is 6.5 times higher, and pressure ratio of CO2 is nearly
40% lower than R134a for equal cooling capacity at 0 oC. Re-expansion losses are much
smaller in CO2 system. Although a thicker wall is required due to high pressure, due to
the higher volumetric capacity, CO2 compressor will be smaller compared to
conventional refrigerant compressor for the same capacity [5]. Recent investigations
showed that internal leakage losses and the piston blow-by losses are much less (about
1%), which indicated that by proper design of the lubricated compressor the influence of
leakage can be neglected [5].
As the pressure level increases, allowable pressure drop becomes higher and this
gives rise to a possibility of improving heat transfer because of higher flow velocity
resulting in higher Reynolds number in heat exchanger. Both high pressure and proximity
to the critical point give considerable improvement of convective heat transfer.
Evaporator pressure drop leads to reduced temperature difference due to corresponding
drop in saturation temperature. Slope of saturation curve is very different for CO2
compared to the conventional refrigerants because of high pressure. Nucleate boiling heat
transfer is also affected to a large extent by pressure. The higher tolerable limit of
pressure drop leads to optimum design of compact heat exchanger with small diameter
tube and in many cases extruded multiport tubing with parallel flow of refrigerant in
several tubes and flow channels. Pettersen et al. [55] showed some compact heat
exchanger concepts for CO2 air conditioning systems having internal diameter of 2 mm.
Low side refrigerant line diameter are typically reduced to 60 70% compared to HFC
system due to higher density and flow velocity [5]. Higher side piping dimensions can
also be reduced. Assuming wall thickness that more or less same as in HFC piping of
equal capacity, pressure capability will be sufficient for CO2 due to reduced diameter.
Hence, the components size and weight reductions are possible due to reduced refrigerant
side volume and cross section.
26
2.7.2
27
supercritical fluid is rapidly depressurized, e.g. due to an initial crack or rupture. The
sudden depressurization gives a superheated liquid phase that is suddenly vaporized in an
explosive manner. This may give a transient overpressure peak inside the vessel, which
again may lead to a powerful burst of the whole vessel. However it has been observed
that maximum overpressure spike is not so high (about 7 % of initial) [5]. Thermal shock
effects on the pressure sensors were significant. In real systems the presence of
compressor
lubricant,
particles
and
contaminants,
as
well
as
unstable
pressure/temperature, would make homogeneous nucleation even less likely. Some rules
regarding pressure relief for safety standard for CO2 mobile air conditioning and heat
pump have been proposed [5].
Compressor
The vapor pressure inlet to the compressor for transcritical CO2 system is much
28
29
scroll compressor was the increase of mechanical loss at the thrust surface supporting the
orbiting scroll number, while it showed very smooth torque load variation.
Several manufacturers, mostly in Japan and Europe, have started manufacturing
CO2 compressors. Denso Corporation has developed a single stage semi-hermetic scroll
compressor [5]. Dorin developed the first high-pressure semi-hermetic CO2 compressor
series in the range of 1.710.7 m3/h swept volume [8]. The series comprises single- and
two-stage compressors with two cylinders, running at nominal speeds of 1450 and 900
rpm (50 Hz). Kim et al. [5] listed various CO2 compressor manufacturers in Japan. Some
important manufactures have been listed in chapter 1.
2.8.2
Gas cooler
Maximum research efforts have been spared on the cross-flow extended surface
gas cooler, where air is the secondary fluid, applicable to air conditioning and room/space
heating. The high working pressure and favourable heat transfer properties of CO2 enable
reduced tube diameters and small refrigerant-side surface areas. Since these reductions
may give room for more airside surface per unit core volume, the compactness can be
increased. Pettersen et al. [55] initially reported microchannel heat exchangers (both gas
cooler and evaporator) could be used for air-conditioning systems.
To handle the high pressures, many CO2 systems employ heat exchangers with
flat multiport (microchannel) tubes. Compared to conventional flat-fin/round-tube
designs, this technology provides additional benefits such as increase in refrigerant-side
area by about a factor of three, far less airside pressure drop due to the streamlined profile
presented by the tubes and flat tubes enable higher face velocities that increase the airside
heat transfer coefficient. Because of high investment costs, microchannel heat exchangers
for fluorocarbon refrigerants have appeared first in high-volume applications where
compactness is valued such as automotive applications. Some of the early theoretical
analyses of CO2 system performance assumed that both indoor and outdoor heat
exchangers would be of conventional flat-fin/round- tube design [5]. One issue in
compact gas cooler design is internal conduction due to large temperature differences
across small lengths. As pointed out by Pettersen et al. [55] internal conduction in fins,
30
tubes and manifolds may lead to performance reduction. Solutions to avoid these
problems include splitting of fins, use of several heat exchanger sections, and careful
design of manifold geometries.
Fang et al. [62] developed a simulation model of gas cooler for calculating heat
transfer and pressure drop at supercritical pressures in both transitional and fully
developed turbulent flow and verified with experimental data thereafter. Model
considered the louvered fin geometries of rectangular channels, triangular channels and
plate and tube types for airside heat transfer and pressure drop. Results showed that the
pass segmentation and uncertainty of pressure drop influenced the accuracy of capacity
calculation. Neglecting pressure drop had caused 10% uncertainty in capacity
calculation. The capacity increased with increase in tube depth, whereas fin height had
negligible effect on the capacity. A finite element model for a supercritical cross-flow gas
cooler model was developed and validated with data obtained in more than 350
experiments [63]. They proposed a multi-slab gas cooler design and reported that a newly
designed cross-counter flow gas cooler could improve the system capacity and COP by 34% and 5% respectively. Simulation of different arrangements of the gas cooler within
original packaged dimensions showed that the 3-pass gas cooler is the best single slab
design. However using multiple slabs seemed to be a more effective way to improve the
performance. In the transcritical CO2 cycle, system performance is very sensitive to gas
cooler design. A small change in refrigerant exit temperature can produce a large change
in gas cooler exit enthalpy (and evaporator inlet enthalpy) because specific heat becomes
infinite at the critical point and hence COP is very much sensitive to gas cooler exit
temperature [64]. A counterflow configuration is important for the gas cooler to exploit
the large refrigerant-side temperature glide. Moreover, the steep refrigerant temperature
glide allows for ideal cycle efficiency to be achieved at finite air flow rate, in contrast to
the infinite air flow required to achieve ideal efficiency in the subcritical cycle. Using a
multi-slab, counter-flow arrangement gas cooler design was proposed [64] and the
predicted approach temperature difference was reduced from 6.9 to 3.6 oC and the design
seemed to be very effective at lower airflow rate and high capacity conditions. The flat
tubes are vertical in this prototype, to facilitate condensate drainage and defrosting in
31
heating mode. The proposed design could give 11% higher gas cooler capacity than the
commonly used multi-pass design.
Garimella [65] has modeled a novel near counter-flow serpentine flow for the
tube side gas cooler for air conditioning system. The model used heat exchanger
geometry and inlet conditions as the inputs to predict the overall duty as well as the
temperature profiles of refrigerant and air. Model predicted that a gas cooler load of 6.97
kW could be transfered in very small envelope. The approach temperature difference was
predicted as 5.33 oC. However the optimum design of gas cooler such as optimizing
geometry parameters (e.g. core depth, face area, number of tube, number of serpentine
tube side passes, port diameter as well as fin pitch and thickness and louver geometry)
and refrigerant flow through port could give lower approach temperature difference.
Numerical investigation with validation with experimental data in open literature has
been done to investigate the effect of wall thermal conductivity on overall performance of
cross flow microchannel gas cooler [66]. Results showed that the longitudinal conduction
in fins, the transverse and the longitudinal conduction in tubes give negligible effects on
the total heat flow and on the temperature field. Neksa et al. [67] experimentally and
theoretically investigated the fan less gas cooler in CO2 system for heating purpose.
Along with the elimination of fan noise and fan power consumption, this concept reduces
the air draft in the room. The concept was found to be feasible also for high efficiency
heat pumps, as CO2 is partly characterized by a gas cooler giving a low cold end
approach temperature difference. It was therefore expected to be very well suited as an
indoor heat exchanger for an air-to-air heat pump for which fan noise may be an
important market-limiting factor.
2.8.3
32
2.0 m), having identical cross-sections, were used in the experiments in both parallel and
counter-flow configurations. The counter-flow arrangement of internal heat exchanger
was shown to be better than parallel arrangement. The longer internal heat exchanger
provides the greatest increase in COP and the greatest decrease in the corresponding
optimal discharge pressure. The experimental data were used to develop simulation
model, which was used to develop a optimal design for a COP maximizing of internal
heat exchanger. In automotive air-conditioning systems, internal heat exchange provides
the greatest capacity enhancement when it is needed most, that is during idling at high
ambient temperatures. Boewe et al. [68] showed how three microchannel tubes could be
stacked to provide many parallel ports to control pressure drop in the cold suction gas,
while forcing the supercritical fluid through smaller ports to maximize heat transfer
coefficients and areas upstream of the expansion device where larger pressure drop can
be tolerated. Compared to conventional concentric tube designs, the microchannel
configuration reduced material requirements by 50% while eliminating the need for long
suction and liquid lines, and increasing effectiveness by 10%.
2.8.4
Evaporator
Microchannel evaporators are currently the subject of research within the
33
correlations for the model. They reported that their model predicted the experimental data
with reasonable accuracy, and could be used for the performance analysis and designing
of a microchannel evaporator. Currently, all simulation models of microchannel heat
exchangers assume perfect distribution on the refrigerant side [69]. Their main focus is
on capturing accurately the important air-side phenomena such as the effects of
condensate and inclination angle.
Although several experimental studies have been reported on flow boiling heat
transfer in microchannel tubes, few experimental studies have been reported to measure
the effects of condensate retention on the performance of microchannel evaporator as
reported by Kim et al. [5]. The heat transfer coefficients and pressure drops for wet
conditions revealed the important of role of condensate drainage.
2.8.5
Other components
A variety of lubricants can be used for transcritical CO2 applications. In certain
systems synthetic hydrocarbons such as polyalpha olefins (PAOs) and alkyl benzenes
(ABs) can be used even though they have poor solubility. The poor solubility of the
synthetic hydrocarbons is compensated by their excellent low temperature flow properties
and can be improved still further by blending with more miscible lubricants (e.g.
polyalkylene glycols (PAGs), esters, etc.). A range of individual and blends of synthetic
lubricants are therefore being evaluated to find the more cost effective solution for a
particular application. Quite often lubricant selection will be based on logistic factors, i.e.
a lubricant that can work with a variety of refrigerants. Various synthetic lubricants
including mineral oil, Polyalkylene glycol (PAG), polyolester (POE), polycarbonate (PC)
and polyvinyl ether (PVE) were experimentally tested for CO2 applications [5].
Ikeda et al. [70] evaluated the chemical and physical properties (lubricity, chemical
solubility, miscibility, mixture viscosity and other properties) of possible base oils:
polyalkylene glycol (PAG), polyvinyl ether (PVE), polyol ether (POE) and polycarbonate
(PC) for a CO2 applications. It was found that PAG was the best lubricant for a CO2 heat
pump system due to its suitable miscibility, higher chemical stability and better lubricity
under high pressure. Although PAG was less miscible than other tested oils, oil return
34
problem was not occurred. POE showed poor lubricity and stability and PC was worse
than PAG and PVE in lubricity. Newly developed refrigeration lubricants using PAG
have been successfully introduced into the market for CO2 heat pump water heater
systems and CO2 automotive A/C systems.
Some issues are being studied with respect to elastomer materials for seals and hose
connections in CO2 systems Permeation rates are quite high, thus giving potential
problems regarding desired leakage rates in automobile air conditioning systems.
Explosive decompression may occur when CO2 systems or components are rapidly
depressurized, leading to fractured and ruptured sealing elements [71]. A fluorite
elastomer, FKM was regarded as promising due to its wide temperature range of
application and the negligible impact of explosive decompression. Other components,
including electronic expansion valves, accumulators, hoses, o-rings and fittings have
been developed for CO2 air-conditioning systems. Several manufacturers in Japan and
Europe are working on expansion valves and controls for CO2 systems [5].
35
Automotive air-conditioning
Mobile air conditioning applications were among the first to be considered for
application of the transcritical CO2 cycle due to various disadvantages with R22 and
R134a including high leakage rate through the flexible nylon or butyl rubber hoses
needed for vibration protection, and through the compressor shaft seal needed to avoid
the additional weight and conversion losses associated with the hermetic electric
compressors used in other applications as well as high GWP. In successive studies,
Lorentzen and Pettersen [1] and Pettersen [72] developed and tested a prototype of
transcritical CO2 automobile air conditioning system. They used, as a reference, a
commercially available R12 automotive air-conditioning system and built the prototype
of comparable cooling capacity. The CO2 system had a liquid line/suction line heat
exchanger to transfer heat between CO2 leaving the gas cooler and low pressure CO2
leaving the evaporator. To match the cooling capacity, the displacement volume and
speed of compressor were adjusted. The external dimensions of the air-to-refrigerant heat
exchangers were nearly identical for both systems; however the CO2 evaporator had 25%
larger airside surface than the R12 evaporator, and the CO2 gas cooler had 34% larger
airside surface than the R12 condenser. The authors demonstrated that the CO2 system
had comparable performance of the R12 system.
36
The ideal cycle comparisons of R-134a and CO2 for automotive air conditioning
carried out by Bhatti [73], rested on assumptions that failed to account properly for the
unique thermodynamic and transport properties of CO2 even though experimental data
for a prototype CO2 system had already shown competitive performance to a state-of-theart R-12 system. Not surprisingly the theoretical studies concluded that the total global
warming impact of CO2 systems would exceed that of R-134a, considering both direct
(leakage) and indirect (fuel combustion) emissions of greenhouse gases. Through
improvements in system operation and control [9], compressor performance, and heat
exchanger performance [55], the CO2 technology was able to compete even with the
improved R-134a systems that were introduced in the mid-1990s. McEnaney et al. [74]
presented experimental results for both steady state and cyclic operation for prototype
CO2 system and a commercially available 134a automotive air-conditioner. Both the gas
cooler and evaporator were microchannel heat exchangers. The external volumes of
evaporator were identical for both systems with the CO2 evaporator having 20% larger
airside surface area. The CO2 gas cooler had 23% lower external volume and 28% lower
air side surface than the R134a condenser. The test results indicated that the prototype
CO2 system provided a comparable performance to the current production R134a system
for both steady state and cyclic operation. Control strategy using clutch cycling, variable
displacement and variable speed compressor can be applied to adjust optimum high side
pressure for CO2 automotive air conditioning [75].
Brown et al. [76] evaluated merits of CO2 and R134a air conditioners using
simulation model and considering current production configuration of a R134a system
and a CO2 system, which was additionally equipped with liquid-line/suction line heat
exchanger. Results showed CO2 having an inferior COP to R134a. The COP disparity
depends on compressor speed (system capacity) and ambient temperature; the higher the
COP and discharge temperature, the lower was the COP difference. At the same speed
and lower ambient temperature, the COP disparity was lower; however at higher speed
and ambient temperatures, it was greater. Hence, better transport properties and
compressor isentropic efficiency did not compensate for its thermodynamic disadvantage
compared to R134a when equivalent heat exchangers were used for both refrigerants,
37
even if internal heat exchanger was used to reduce throttling irreversibilities. The entropy
generation calculation indicated that CO2 had somewhat better performance than R134a
in evaporator, however poorer performance in gas cooler than R134a in condenser.
Based on the results of an analysis of a large number of experiments and some
new concepts, next-generation prototype systems have been designed and are serving as
the focus for current research. Most are equipped with variable-displacement
compressors, and heat exchangers configured to exploit the unique transport and
thermodynamic properties of CO2 [5]. Liu et al. [77] evaluated the effect of CO2 mass
charge on the performance of an automotive R744 transcritical air conditioner operated
by manual expansion valve and showed that there was optimum mass charge, which
make the COP maximum; however no maximum cooling capacity was found to exist
with varying CO2 charge.
2.9.2
Automotive heating
Because of insufficient waste heat from efficient fuel-injection engines for heating
of the passenger compartment in the winter season and also unacceptable heating-up
period and slow defroster action in terms of both safety and comfort, supplementary
heating is necessary, and heat pump may be one of the attractive solutions. CO2 systems
have special benefits in heat pump mode, since high capacity and COP can be achieved
also at low ambient temperature and with high air supply temperature to the passenger
compartment. Giannavola et al. [78] first ran an auto air-conditioning prototype system in
reverse for automotive heating. However, the cross-counter-flow interior heat exchangers
were far from ideal. Results showed that the capacity is highest at startup, at least three
times higher than that of electric resistance or friction heater due to the high heat
pumping efficiency; and capacity and efficiency decline slowly at higher temperature lift
due to reduced volumetric and isentropic compressor efficiencies. These initial results
have proven quite valuable in guiding the design and development of improved
components for next-generation systems. Heat pumps are not currently employed in
automobile due to disadvantages of R134a as heat pump fluid. The technical problem is
accumulation of frost in heat exchanger and very little knowledge about the drainage
from ultra-compact microchannel heat exchangers. There is no such effective option to
38
avoid it. However, the ability of heat pump to provide instant heat and capability of CO2
heat pump to deliver that heat at high temperature while moving less air may give
potential advantage to CO2 systems [5].
2.9.3
Residential cooling
The first assessment of transcritical CO2 systems for residential air conditioning
Residential heating
Transcritical CO2 cycle can deliver air at 60 oC to achieve the same level of
comfort offered by a gas based furnace while quietly moving substantially less air than
conventional heat pumps Similarly, CO2 can provide heat via a hydronic secondary loop
without the energy penalty that would be incurred in a heat pump operating on a sub-
39
critical cycle. Several analytical studies of the global warming implications of alternative
heat pumping systems have concluded that the direct effect of refrigerant emissions is
relatively small, Hence, for almost all applications except automotive air-conditioning
and supermarket refrigeration, the highest R&D priority is to maximize the efficiency of
CO2 systems.
Initial theoretical study [81] on controlled ventilation air heating system with CO2
heat pump showed very promising results. The overall seasonal performance factor for a
Graz, Austria climate was calculated in the range of 6.15 6.5. Beaver et al. [82]
reported that the R-744 system operated with the same or slightly higher COP at lower
ambient temperatures (26/27 oC) and a slightly lower COP at higher ambient temperature
(26/35 oC indoor/outdoor) for tests performed at equal capacity. Although COP (cooling)
of CO2 system in cooling mode is lower than that of R22 system, it is slightly higher in
heating mode [83]. Experimental investigations of transcritical CO2 systems for
residential space heating were conducted on a prototype residential split system,
originally designed for cooling only, by simply reversing its operation to study heating
performance [84]. Since the original baseline system was obtained before the first R410A heat pump system became available, the package dimensions for the heating
comparisons were no longer equal: the baseline subcritical R-410A system had larger
heat
exchangers.
Nevertheless,
system
performance
was
compared
for
two
configurations: first when the CO2 prototype semi-hermetic compressor speed was set to
match heating capacity at 8.2 oC outdoors and 21.1 oC indoors; and second when the air
conditioning capacity was matched at 35 oC outdoors and 26.5 oC indoors. The CO2
system had comparable cycle COP in heating mode, but its higher capacity at lower
outdoor temperatures increased its heating performance factor by reducing the need for
supplemental heat, which illustrates the CO2 systems ability to select a compressor
discharge pressure that gives it extra capacity or efficiency when needed. Another
experimental investigation showed that the prototype CO2 unit had performed almost
equal as the R410A baseline unit in AC mode at 27.8 oC and 35 oC ambient temperature
and in heat pump mode it had performed around 30 to 40% better at 5 to 5 oC ambient
temperature [85].
40
For hydronic heating applications, CO2 heat pump systems showed favorable
seasonal performance compared to the system using R134a [5]. Theoretical study on CO2
heat pumps for retrofit in typical hydronic heating systems showed that the seasonal
performance was increased from 2.8 to 3.2. In addition, this system will be able to supply
hot tap water without any loss in energetic efficiency. Experimental results from two
prototype systems reported the same range of efficiency figures as those calculated [83].
2.9.5
Water heating
The first application of CO2 systems on the market is heat pump water heaters,
where the thermodynamic properties are very favorable. Due to gliding temperature in
gas cooler the temperature characteristics of the transcritical cycle matches the
temperature profiles of the heat source and heat sink, giving small heat transfer losses and
high efficiency. A pre-condition for high efficiency is a low water inlet temperature,
giving a low refrigerant inlet temperature to the throttling device. Thus, the design of the
hot water accumulating system for temperature stratification is essential in order to
achieve high heating COP. Experimental study on CO2 heat pump prototype showed very
promising performance in tap water heating [86]. A heating COP of 4.3 was achieved for
heating tap water from 9 oC to 60 oC at an evaporation temperature of 0 oC. The result led
to a seasonal performance of about 4 for an Oslo (Norway) climate using ambient air as a
heat source and the primary energy consumption could be reduced by more than 75%
compared to electrical or gas fired systems. Increasing the required hot water temperature
from 60 oC to 80 oC reduced the heating COP from 4.3 to 3.6 at an evaporation
temperature of 0 oC. A CO2 system can produce hot water with temperature of 90 oC
without operational difficulties, whereas conventional heat pump systems are often
restricted to hot water temperature lower than 55 oC.
Hwang and Radermacher [87,88] compared the water chilling and tap water
heating performance of CO2 system with R22 system through their theoretical and
experimental studies. Theoretical analysis showed that the reduction of heat exchanger
size and mass of CO2 is possible for tap water heating by appropriate design.
Experimental study showed that though the ideal cycle coefficient of performance (COP)
of CO2 is only 50% to 60% that of R22, the actual CO2 cycle performed similar to the
41
R22 cycle within 6% when the same outside volume of the heat exchanger was
applied for both refrigerants. This large deviation between ideal and actual cycles was
due to beneficial thermodynamic properties of CO2.
Important application areas for commercial size systems are in hotels, apartment
houses, hospital and food industries. Another experimental study on heat pump water
heater showed a heating capacity of 115 kW and heating COP of 3.4 for evaporation
temperature of +0.3 oC and hot water temperature of 77.5 oC [89]. Simulation model
based on equipment performance data from experiment showed that the hot water
temperature could be increased from 65 to 120 oC with relatively small reduction of
heating capacity and heating COP of 33% and 22%, respectively. Several manufacturers
introduced CO2 heat pump water heaters in the market during 2001-02 [5]. Stene [90]
developed a CO2 heat pump system for combined space heating and water heating and
tested for evaporation temperature of 5 oC to obtain hot water up to 80 oC. The test
results proved that an integrated brine-to-water CO2 heat pump may achieve the same or
higher seasonal performance factor than the most energy efficient state-of-theart brine
water heat pump systems.
2.9.6
modules, and vehicles have traditionally been met by procurement rather than R&D,
using custom-built units based on the same basic technology used for commercial
applications. The conventional Mil-Std (Military-Standard) ECU (Environmental Control
Unit) consists of a reciprocating compressor, copper tube and aluminum fin coils, swirl
cage fan assemblies, and a housing that has been hardened to meet the unique military
requirements. However, two recent developments have motivated the US military to
sponsor research on CO2 systems to meet its operational requirements for (1) lightweight
ultra-compact units for rapid deployment via air transport, and (2) a refrigerant that is
globally available and free of the diverse and extensive regulatory requirements and
logistical challenges associated with greenhouse gases. Although, one of the earliest
theoretical analyses [91] showed roughly equivalent performance under same conditions,
experimental study on microchannel based CO2 prototype showed that the CO2 based
42
ECU did not perform as well as the R22 ECU in terms of capacity and COP. The addition
of internal heat exchanger improved the capacity and COP but still fell short of the R22
baseline. However, capacity and COP could be further improved by using an appropriate
compressor, and a change in fan type [92]. Recent experimental study [93] indicated that
the CO2 based ECU demonstrates a higher cooling COP and a higher cooling capacity
compared to the R22 units.
2.9.7
Refrigeration applications
Research interest in CO2 has also been renewed in the area of transport
refrigeration for two reasons. The first relates to the relatively high density and capacity
of CO2 at low temperatures, compared to alternatives such as hydrocarbons or ammonia;
the advent of lightweight compact microchannel heat exchangers presents new
opportunities for system optimization. Second, the worldwide availability of CO2 and
freedom from HFC-related regulatory uncertainties fits well with the global nature of the
transport refrigeration industry. Studies showed that the performance of CO2 system
matched with equal size systems using R502 and 507 and very similar COP values in
freezing mode with R134a over a full range of ambient temperatures [5]. One problem
with CO2 may be its very high compressor discharge pressure. However, in shipping
application, this can be reduced due to lower ambient temperature in sea atmosphere.
Commercial refrigeration systems for shops, supermarkets, large kitchens, etc.
have large refrigerant emissions, and the energy use is in many cases high. Thus, there is
a need for efficient, safe and environment friendly refrigeration systems. New concepts
based on CO2 have been demonstrated for centralized systems using CO2 as a secondary
heat transfer fluid or in a low-temperature cascade stage, and recently decentralized
concepts with heat recovery have been shown. Schiesaro and Kruse [94] developed a
two-stage CO2 supermarket system and carried out experiments to optimize the operation
and performance. Results showed lower COP than theoretical value due to the effect of
large amount of oil, additional mass flow through venturi, internal leakage and pressure
drop. However, good performance could be achieved by using components specially
designed for this refrigerant. Use of CO2 as secondary fluid or lower side refrigerant in
43
and cooling is one of the most promising applications of transcritical CO2 cycle. A CO2
heat pump was constructed to enable the simultaneous production of refrigeration at less
than 0 oC and water heating to 90 oC for the New Zealand food processing industry [97].
The optimum heating COP of the prototype was about 3 in most trials, but increased to
3.2 when operating the heat pump compressor at part-load with no oil in the system. The
addition of oil caused fouling and reduced performance in the evaporator but had
minimal effect on the performance of the gas cooler. The addition of oil had little effect
on compressor isentropic efficiency when operating at maximum compressor speed but
caused a significant reduction in isentropic efficiency when operating the compressor at
reduced speed. Another experimental study on simultaneous air-conditioning and water
heating showed promising result as an energy recovery system [98]. Results showed that
the combined system is more effective compared to an air conditioning system without
heat recovery.
44
2.9.9
Schmidt et al. [99] theoretically showed possibility of energy saving due to better
temperature adaptation in the heat exchangers compared to sub-critical process and also
possibility to achieve higher temperature without loss in efficiency, thus better moisture
extraction rate. Results also showed equivalent or even better COP than the comparative
R134a system. Experimental results from Klcker et al. [100] reported heating COPs in
the range 5.5, and a 55% reduction in the energy consumption, including fan power,
compared to a traditional electrically heated dryer. The results were achieved after a first
optimization of the prototype system and it was hoped that further essential
improvements still could be reached.
2.10 Summary
This chapter outlined the recent developments in carbon dioxide as refrigerant in
transcritical cycle in various refrigeration, heat pump and air-conditioning applications. It
is observed that details of thermodynamic analysis and cycle optimization for a wide
temperature range are still insufficient. Although various methods have been suggested
to control the gas cooler pressure at optimum condition, still there is a need for some easy
method to control it. There is also need for rigorous theoretical and experimental
investigations on cycle modifications for optimum design, also in term of economics.
Because of high temperature lift (stated in chapter 1) for CO2, there is an opportunity for
investigation of multistage systems for heating or cooling at various temperatures. Also
investigations of heating or cooling systems with various mixtures of CO2 can be carried
out with a view to reduce the system pressure and take advantage of superior heat transfer
properties of CO2. Although various theoretical and experimental investigations on
supercritical heat transfer and pressure drop, boiling heat transfer and pressure drop, twophase flow have been reported, pseudocritical region of gas cooling is still one of the
interesting areas, where one can rigorously study heat transfer and fluid flow.
Relatively low isentropic efficiency of developed compressor showed that still there
is a need to improve the design to achieve higher performance. Research on CO2
45
compressors is required not only for improvement of performance, but also to reduce the
weight and cost. Although various types of heat exchangers have been theoretically and
experimentally investigated, detailed economic analysis of heat exchanger is missing,
which is essential in commercial applications. In the last few years, most of the research
was concentrated on mobile air conditioning and heat pump water heating, even though
transcritical heat pumps also have great potential in industrial applications where
simultaneous heating and cooling is required. However, the literature survey shows that
detailed theoretical and experimental studies on transcritical CO2 systems for
simultaneous cooling and heating applications are scarce. Detailed studies on CO2 heat
pump dryers have also been not done extensively. Hence in the present work, the
following aspects are studied:
i)
ii)
Simulation and optimization of CO2 heat pump for simultaneous heating and
cooling applications, and system irreversibility analysis
iii)
iv)
CO2 heat pump dryer simulation, validation with experimental data and
optimisation
v)
vi)
46
Chapter 3
47
One of the interesting features of the transcritical CO2 cycle is that the refrigerant
outlet temperature and the pressure in the gas cooler can be adjusted independently to get
optimum performance. The details of the cycle analyses with both energetic and exergetic
perspectives and optimization of compressor discharge pressure for a wide range of
temperature applications is demonstrated. Finally, effects of different cycle modifications
on the optimum discharge pressure have been presented.
48
critical points. For surface tension of CO2, REFPROP [104] source code was partly
incorporated in CO2PROP.
3.2.1
some distinct effect on component design. Since the transcritical CO2 cycle operates at a
very high pressure, the vapor density is high yielding 3-10 times greater volumetric
capacity compared to other refrigerants. Steeper vapor pressure curve near the critical
point gives a smaller temperature change for a given pressure change. Thus, the
temperature change associated with pressure drop through evaporator will become
smaller.
50
45
P = 40 bar
40
P = 75 bar
35
P = 80 bar
30
P = 90 bar
25
P = 100 bar
P = 120 bar
20
15
10
5
0
10
30
50
70
90
110
130
150
Temperature (oC)
49
temperature hence the pressure drops may be allowed to be higher. Near critical pressure,
more abrupt change in specific heat is observed as shown in Figure 3.2. Such large
variations in thermo-physical properties cause the heat transfer coefficient to be greatly
dependent on both the local temperature and the heat flux, and conventional design
techniques such as -NTU or LMTD method yield large error in the design of a gas
cooler. The pseudocritical temperature, Tpc of CO2 was calculated using the following
algebraic equation [36]:
Tpc = 122.6 + 6.124 P 0.1657 P 2 + 0.01773P 2.5 0.0005608P 3
(3.1)
50
P = 80 bar
P = 100 bar
Pr
P = 120 bar
6
4
2
0
0
20
40
60
80
100
120
140
Temperature ( C)
h2 h3
h2 h1
(3.2)
51
100
P2'
90
2'
P2
80
Pressure (bar)
h3
t3
3
70
h2
60
50
40
30
20
50
150
250
350
450
Figure 3.5 Heat pump cycle on the P-h plane for various gas cooler pressures
With increase in discharge pressure from P2 to P2 for a constant cooler exit temperature
of t 3 , the heating COP gets modified as:
52
COPheating
=
(h2 h3 ) + h2 + h3
(h2 h1 ) + h2
(3.3)
Due to the unique behavioral pattern of carbon dioxide properties around the
critical point and beyond, the slope of the isotherms is quite modest for a specific
pressure range; at other pressures above and below this range, the isotherms are quite
steep. As pressure increases, the quantity h3 is large compared to h2 , as is evident
from Figure 3.5, and this causes an increase in the modified COP value as can be
observed from Equation (3.3). At a particular pressure the COP attains a maximum value
and the corresponding pressure is termed the optimum pressure for the cycle. With
further increase in pressure, h3 does not produce the required gain over h2 and thus the
COP begins to fall. The pressure range where the isotherms are fairly flat and where this
beneficial gain in COP occurs varies considerably with cooler outlet temperature. Hence
the gas cooler outlet temperature plays an influential role in determining the optimum
operating conditions for the cycle.
53
process 2-3 represents the single phase heating of external fluid and dashed line above
evaporating process represents the single phase cooling of external fluid.
The entire system has been modeled based on the energy balance of individual
components of the system. Steady flow energy equations based on first law of
thermodynamics have been employed in each case and specific energy quantities are
used. The following assumptions have been made in the thermodynamic analysis:
1. Heat transfer with the ambient is negligible.
2. Only single-phase heat transfer occurs for the external fluid.
3. Compression process is adiabatic but non-isentropic.
4. Evaporation and gas cooling processes are isobaric.
380
2s
Temperature (K)
360
340
320
3
4
300
1
280
260
240
-2
-1.5
-1
-0.5
qev = h6 h5
(ii)
(3.4)
54
qgc = h2 h3
(3.5)
wcomp = h2 h1
(3.6)
h1 h6 = h3 h4
(v)
(3.7)
(3.8)
(vi) Energy balance for gas cooler with respect to external fluid being heated:
mgcef c p , gcef Tgcef = mr qgc
(3.9)
where
c p , gcef
is the average specific heat of external fluid being heated and Tp , gcef is
c p ,evef
(3.10)
evaporator and Tevef is its temperature drop across the evaporator. The COPs for
heating and cooling modes are given by:
COPheating =
qgc
wcomp
and
COPcooling =
qev
wcomp
(3.11)
(3.12)
(3.13)
It may be noted that is ,comp depends on the type of compressor, compressor design,
degree of superheat, etc. The isentropic efficiency of available CO2 compressors varies in
wide range. Hence is ,comp has been varied from 50 to 90% to observe its effect.
55
Exergy analysis: An exergy analysis has been performed for each component of the
system employing the fundamental steady state equation:
Net exergy transfer from the component = Exergy transfer due to heat transfer
+ Exergy transfer due to work transfer + Change in flow exergy.
i)
Compressor irreversibility,
icomp = To (s2 s1 )
ii)
(3.14)
iexp = To (s5 s4 )
iii)
(3.15)
(3.16)
ee v = To (s6 s5 ) qev
(3.17)
Neglecting irreversibility due to pressure drop, the evaporator irreversibility is given by:
To
(3.18)
T evef
To
(3.19)
T evef
egc
T9 T10
ln (T9 / T10 )
(3.20)
(3.21)
where, T gcef is the heating fluid entropic average temperature and is estimated the same
way as that for the evaporator.
Irreversibility due to pressure difference:
56
igc,P = RTo ln 1
Pgc
P2
(3.22)
(3.23)
(3.24)
(3.25)
Second law (exergetic) efficiency for the system is given by the ratio of net exergy output
and the work input to the compressor:
*
esys
II =
wcomp
(3.26)
57
COP
87.5
87
86.5
86
85.5
-10
-5
10
10
Evaporator temperature ( C)
Figure 3.7 Variation of maximum system COP and optimum discharge pressure
with evaporator temperature
58
120
COP
110
Pressure
100
6
90
5
80
70
30
35
40
45
50
Figure 3.8 Variation of maximum system COP and optimum discharge pressure with
cooler outlet temperature
In the present study, it is observed that the influence of internal heat exchanger
effectiveness on system COP and optimum pressure is marginal. With changes in
effectiveness for a cooler outlet temperature of 35 C and an evaporator temperature of 0
o
pressure. So, the performance of internal heat exchanger has a minor influence on system
optimization at low and moderate gas cooler exit temperatures. However, isentropic
efficiency of compressor has strong influence on system performance and design. For
evaporation temperature of 0 oC, refrigerant temperature at gas cooler exit of 40 oC and
59
4
4.5
6
7
8
10
12
60
120
115
110
105
100
95
90
85
80
75
C. Both the iso-optimum pressure lines and corresponding iso-cooler inlet temperature
lines are nearly parallel and vary the least at maximum cooler exit temperature and
minimum evaporator temperature as opposed to a maximum variation at minimum cooler
exit temperature and maximum evaporator temperature. So to obtain useful heating at
higher temperatures from the system, it has to be designed for high compressor discharge
pressure, which is corresponding to maximum cooler exit temperature and minimum
evaporator temperature, although the COP will be low under these conditions. However
to restrict the system to lower optimum cycle pressures, it has to be designed for the
minimum cooler exit temperature and the maximum evaporator temperature, yielding
high COP. With increase in cooler exit temperature or decrease in evaporator
temperature, the optimum discharge pressure increases. This implies that for high
temperature heating or low temperature cooling the system is not profitable in terms of
61
system COP as well as cost as it necessitates a high optimum discharge pressure. Keeping
the smallest possible refrigerant temperature difference between evaporator and cooler
outlet (yielding high COP), a system can be designed for a low optimum discharge
pressure (yielding lower pressure ratio) to obtain heating output at high temperature only
through high superheat.
145
135
125
115
105
95
85
75
Figure 3.11 Gas cooler inlet temperature (oC) at optimum discharge pressure contour
(0.5 increment of iso-temperature lines)
Second law efficiency and percentages of irreversibility for different components
have been obtained for different operating conditions. For the gas cooler, mass flow rates
for both refrigerant and the fluid being heated are assumed to be the same (1 kg/s). In the
evaporator, the secondary fluid exit temperature is assumed to be 2 C above the
evaporator temperature and its mass flow rate has been calculated per unit refrigerant
mass flow rate basis. Inlet conditions for both secondary fluids have been taken as 10 oC
lower than the cooler exit temperature. The total pressure drop in the gas cooler is taken
as 2 bar. The minimum temperature difference required at cooler outlet to avoid pinch
problem in the gas cooler increases as compressor discharge pressure decreases. This
62
behaviour gets more complex in the neighbourhood of the critical point and the pinch
problem becomes quite significant due to the irregular constant pressure line in that zone.
To circumvent the pinch issue, we have taken an average temperature difference of 10 K
between CO2 gas and external fluid.
50
48
46
44
42
40
tev=-10
tev
= 10 oC
tevtev=-5
= 5 oC
tev=0
tev
= 0 oC
tev=5
tev
= 5 oC
tev=10
tev
= 10 oC
38
36
34
32
80
85
90
95
100
105
110
115
120
Figure 3.12 Variation of second law efficiency with discharge pressure for
different evaporator temperatures
Second law efficiency (combined) variations with compressor discharge pressure
for various evaporator temperatures at a cooler outlet temperature of 35 oC and internal
heat exchanger effectiveness of 60% are presented in Figure 3.12. Maximum values of
second law efficiency vary from 44.3 to 48.9% with corresponding discharge pressure
varying from 93 to 88 bar, respectively. Calculations show that unlike its effect on COP,
the heat exchanger effectiveness has some influence on the second law efficiency of the
system. With increase in heat exchanger effectiveness from 0.6 to 0.9, the maximum
value of second law efficiency increases by about 3%, although this effect is somewhat
more than that on the system COP (about 1%) due to assumptions made pertaining to the
temperature difference between refrigerant and secondary fluid in the heat exchanger.
63
Second law efficiency variation with compressor discharge pressure for different
gas cooler exit temperatures at an evaporator temperature of 0C and internal heat
exchanger effectiveness 60% is presented in Figure 3.13. With variation in gas cooler exit
temperatures, maximum values of second law efficiency are found to vary widely from
53.8 to 30.8% with corresponding discharge pressure varying from 75 to 125 bar,
respectively.
55
50
45
40
35
30
25
20
30 oC
40 oC
15
10
35 oC
45 oC
55 oC
50 oC
5
75
85
95
105
115
125
Figure 3.13 Variation of second law efficiency with discharge pressure for
different gas cooler exit temperatures
Figure 3.14 represents the variation of percentages of total irreversibility of
different components with discharge pressure at a condition specified by an evaporator
temperature of 0C, gas cooler inlet temperature of 35oC and internal heat exchanger
effectiveness of 60%. It is observed that the effect of heat exchanger effectiveness on
irreversibility is comparatively minor. It may be observed that the nature of the curves for
evaporator and gas cooler are different due to different assumptions for both. Near the
optimum discharge pressure, irreversibilities of compressor and evaporator are maximum
but the irreversibility of gas cooler is minimum. With increase in discharge pressure,
64
Compressor
Gas cooler
Evaporator
Expansion device
Internal HEX
Percentage of irreversibility
40
35
30
25
20
15
10
5
80
85
90
95
100
105
110
115
120
65
The energy and exergy flow per unit work input (or 100%) for an evaporator
temperature of 280 K, cooler exit temperature of 310 K, internal heat exchanger
effectiveness of 80%, optimum discharge pressure of 90 bar, and ambient temperature of
300 K, are presented in Figure 3.15. Energy flow diagram implies that system COP is
8.08. As discussed earlier, exergy loss through expansion device is comparatively large
due to large pressure difference between the two sides and also due to the distinct
properties of CO2; near the critical point the entropy change as abruptly as other
properties (as pressure drops from supercritical to subcritical). Exergy loss in internal
heat exchanger is about 4%. For the conditions stated above, irreversibility of evaporator
is greater compared to that in the gas cooler because of larger average temperature
difference across the evaporator compared to that in the gas cooler.
Heating output
Exergy input
100 %
454 %
Compressor
12.2%
Trans100 % critical
CO2
system
Compressor
input
18.7 %
Internal HEX
4.76%
Evaporator
irreversibility, 13%
Gas cooler
irreversibility, 5.2%
Cooling 354 %
output
Evaporator
output 12.3%
Expansion
device
Net = 46.3%
Gas cooler
output, 34%
66
(3.27)
The maximum system COP is given by, COPmax = f (tev , t3 ,is.comp , ) and the
In the
present study, it is observed that for the given input temperatures the internal heat
exchanger has a negligible effect on the system performance. Moreover isentropic
efficiency of the compressor is exclusively dependent on compressor design. Hence
fixing these two ( is ,comp and ) parameters at 70% and 60% respectively, the optimum
condition dependence reduces to its functional form of:
COPmax = f (tev , t3 ) ; p2,opt = f (tev , t3 )
(3.28)
C. Performing a regression analysis on the data, the following relations have been
COPsys ,max = 22.58 + 0.3575tev 0.6136t3 5.075 103 tevt3 + 1.8 103 tev 2 + 4.7 103 t32 (3.29)
p2,opt = 3.47 + 0.32 tev + 2.23 t3 0.0134 tev t3 + 3.7 103 t32
(3.30)
t (at , P
) = 10.65 + 3.78t 1.44t 0.0188t 2 + 0.009t 2
2
2, opt
3
ev
3
ev
(3.31)
It may be noted that in actual practice, the isentropic efficiency of a given compressor
decreases as the pressure ratio increases. However the correlation for optimum
compressor discharge pressure (Equation 3.30) can be widely used for any compressor
because of its near independence from isentropic efficiency.
67
Previous studies [17] show that the use of a work producing device (expander)
such as a turbine instead of an expansion valve can improve the COP for transcritical
CO2 systems. For the cycle with expander, system COP is evaluated by overall cooling
and heating output divided by net work input (compressor work turbine work). A
evaporation temperature of 0 oC and the refrigerant temperature at gas cooler exit of 40
o
70 %, the system COP improves by about 18% when an expander of 80% isentropic
efficiency is used. However the optimum compressor discharge pressure decreases by
only 2.5%. Hence the use of work producing device has marginal effect on the optimum
discharge pressure, whereas it has significant effect on system performance.
68
3.5.3
Multi-staging
Several types of configurations such as flash gas removal, flash gas intercooling,
compression intercooling for a multistage transcritical CO2 cycle can be adopted to
improve the system performance depending on the requirement. Figure 3.16 shows the
CO2 cycle of multistage compression with inter-cooling. Simulation result shows that for
evaporation temperature of 40 oC, refrigerant temperature at gas cooler outlet of 40 oC
and isentropic efficiencies of 70% for both compressors, the cooling COP (evaporative
cooling output divided by work of both compressors) improves by about 12% and the
optimum compressor discharge pressure reduces by 30% compared to single stage for a
degree of inter-cooling of 30 oC. The intermediate pressure has been taken as the
geometric mean of higher and lower side pressures. Hence the compressor discharge
temperature corresponding to optimal condition will reduce for multistage cycle. It is
noteworthy that all the three cycle modifications will improve the system performance
and reduce the optimum discharge pressure, which can be advantageous in high pressure
side component design.
Temperature
intercooling
Specific entropy
69
3.5.4
Ejector-expansion device
Recent studies [23,24] show that the use of an ejector-expansion device instead of an
expansion valve can improve the COP for transcritical CO2 cycles. For the ejectorexpansion cycle as given in Figures 3.17 and 3.18, for a evaporation temperature of 0 oC
and the refrigerant temperature at gas cooler exit of 40 oC, and pressures after nozzle and
diffuser of 0.8 lower and 5 bar higher than evaporation pressure, respectively, the system
COP improves by about 18.2% when expander, diffuser and compressor isentropic
efficiencies of 80%, 90% and 70 %, respectively are used. However the optimum
compressor discharge pressure decreases by only 2.4%. Result shows very similar effect
of ejector-expander as the work-producing device on the optimum discharge pressure and
system performance. All the improvements discussed above shows similar trend of
optimum discharge pressure, as it is lower than that of basic cycle with expansion valve.
70
100
Isotherm
90
Pgc
Pressure (bar)
80
70
60
50
40
7
30
8
4
10
Pev
20
specific enthalpy (kJ/kg)
Figure 3.18 P-h diagram of transcritical CO2 cycle with ejector-expansion device
3.6 Summary
A comprehensive property code for both thermophysical and transport property of
CO2 based on latest available correlations has been developed. A cycle model and
computer simulation of a transcritical carbon dioxide based simultaneous heating and
cooling system have been developed and the effect of various cycle modification
including internal heat exchanger, turbine and multi-staging have been studied. Based on
the results and optimization of the system, the following conclusions can be drawn.
1. Due to the near critical operation, CO2 exhibits some distinct properties (mostly
favourable to system design) such as high Prandtl number, homogeneous two-phase
flow compared to other conventional refrigerants.
2. The effects of evaporator temperature and gas cooler outlet temperature are more
predominant compared to internal heat exchanger effectiveness at optimized
conditions for the system. Although the internal heat exchanger effectiveness has
negligible effect on optimum condition for low and moderate gas cooler exit
71
temperature, its effect is more significant at high gas cooler exit temperature and
lower evaporator temperatures.
3. Analyses of the optimum condition indicate that a system meant for low or moderate
temperature heating is more economical not only due to high system COP but also
due to lower optimum discharge pressure (low operating pressure ratio). Such a
system will yield good performance at lower external fluid inlet temperatures.
However it is possible to obtain high temperature heating at the expense of COP.
Even though COP is lower, a system designed for such application is worthwhile
because conventional refrigeration systems do not offer this high temperature heating.
So there is some trade off among high COP, high outlet temperature and cost of
superheating.
4. Design of all heat exchangers must endeavour to involve lower temperature
differences between the two fluids to yield higher second law efficiency, although
that will also cause the heat exchangers to be bulkier, resulting in higher weight, cost
and pressure loss.
5. Expressions for optimum cycle parameters have been developed and these
correlations offer useful guidelines for optimal system design and for selecting
appropriate operating conditions.
6. Multi-staging has more significant effect than an internal heat exchanger, work
producing turbine and ejector-expander device on the optimum compressor discharge
pressure.
72
Chapter 4
73
exergy analysis has been done for the carbon dioxide based heat pump to cater for
heating and cooling services simultaneously. Irreversibilities of all components and the
second law efficiency of the system for different values of operating parameters have
been estimated. Finally, techniques to reduce the irreversibility for various components,
which leads to improved system exergetic efficiency, have been suggested.
74
75
4.2.1 Compressor
The refrigerant mass flow rate through the compressor is given by,
mref = 1v Vs
N
60
(4.1)
(4.2)
(4.3)
with compressor pressure ratio rp ( = Pdis / Psuc ) varying between 1.5 and 6.5.
The exergy input (same as power input) to the compressor is given by,
Ein = mref (h2 h1 )
(4.4)
(4.5)
Gas cooler
(4.6)
(4.7)
where Qgco i is heat loss to ambient. The heating effect, Qgcwi is given by:
76
(4.8)
Summation of heat transfer in all segments gives total heat transfer in gas cooler. The
overall heat transfer coefficient for the segment of gas cooler has been calculated using
the fundamental equation for overall heat transfer coefficient yielding:
ln ( d o / di )
1
1
1
=
+
+
i
UAgc r Ar
w Aw
2Lkt
(4.9)
Water flow (Annular side)
Tgcwi +1
mgcw
T gcw i
Tw , Pw
w
kw
Qgc i
mref
Trb , Prb
Tgcr i +1 , Pgcr i +1
di do
Di
T gcr i , Pgcr i
L
Refrigerant flow (Inner side)
Figure 4.3 A computational segment of the gas cooler
The resulting irreversibility is expressed as:
T
T
P
I gc = T0 m gcw c pw ln gco m ref (s 2 s3 ) + I gcr
+ I gcw
+ Qgc 0 1 0
Tgcw
Tgci
(4.10)
mr
gcr
Pgcr i +
mgcw
gcw
Pgcwi
(4.11)
Heat transfer and pressure drop correlations: To estimate heat transfer rates,
Gnielinski equation is not suitable for normal tubes (although it is valid for microchannels) due to large variation of fluid properties in the radial direction. To alleviate this
deficiency, Pitla et al. [38] proposed a modification for supercritical in-tube carbon
77
dioxide cooling, incorporating both bulk and wall properties. This correlation, used for
gas cooler model, is given by;
Nu + Nurb krw
Nur = rw
krb
r =
Nur
krb
di
(4.12)
Here, Nurb and Nurw are the Nusselt number at bulk and wall temperature respectively,
predicted by Gnielinski equation within the range 2300 < Re < 106 and 0.6 < Pr < 105 .
Nu =
( f / 8)(Re 1000) Pr
1.07 + 12.7( f / 8)1/ 2 (Pr 2 / 3 1)
(4.13)
( f / 8) Re Pr
1.07 + 12.7( f / 8)1/ 2 (Pr 2 / 3 1)
(4.14)
Ggcr 2
Lgc
+ 1.2
di
2 rb
(4.15)
rw rw
rb rb
(4.16)
Qgcr i
Ggcr
0.42
Evaporator
The evaporator also has been discretised lengthwise for computation similar to the
gas cooler. Employing LMTD expression, heat transfer in the i th segment is given by,
Qevr = (UA)ev
i
(T
evw
i +1
(4.17)
T i +1 T i +1
ln evw i evr i
Tevw Tevr
78
i +1
(4.18)
h )
i
where Qevo i is the heat gain with ambient and the cooling effect, Qevwi is given by:
Qevw = m evw c pw (Tevw
i
i +1
(4.19)
Tevw )
i
Summation of heat transfer in all segments gives total heat transfer in evaporator. The
overall heat transfer coefficient for each segment of the evaporator has been calculated in
the same way as for the gas cooler.
The irreversibility or exergy loss in the evaporator is expressed as:
T P
P
I ev = T0 m ref (s 6 s5 ) m evw c pw ln evi +Ievr
+ I evw
+ Qevo 0 1
Tevo
Tevw
(4.20)
where the temperature related terms on the right-hand side are due to the temperature
difference and heat interaction with ambient respectively. The pressure related terms are
due to pressure drop in refrigerant and water side respectively and are given by,
P
P
I evr
+ I evw
=
mref
evr
Pevr i +
mevw
evw
Pevwi
(4.21)
Heat transfer and pressure drop correlations: The convective heat transfer coefficient
has been estimated using the Wattelet-Carlo correlation, which was originally developed
for R-12, R-134a and a mixture of R22/R124/R1542a and later validated by Rieberer [18]
for R-744 evaporation in tubes. This is the preferred correlation for two-phase flow in the
evaporator and is expressed as:
r = F .1
1 = 0.023
F = 1 + 1.925 X tt 0.83
0.9
kl
Rel 0.8 Prl0.4
di
(Dittus-Boelter correlation)
(4.22)
1 x v l
X tt =
x l v
0.1
(4.23)
The refrigerant side pressure drop, Pevr , is given by (Lockhart and Martinelli equation):
Pevr i Pevr i +1 = 4
Lev f r
G2
(1 x) 2 evr l2
di 2
l
(4.24)
79
7.242
l = 1.376 + 1.655 where X tt is the Lockhart-Martinelli factor.
X tt
Waterside heat transfer and pressure drop: w is the waterside heat transfer
coefficient and has been evaluated by the conventional Dittus-Boelter equation for
annular flow in both evaporator and gas cooler. For water-side friction factor, the
classical Blasius correlation [107] is used:
f w = 0.0791Re w 0.25
(4.25)
All water properties are assumed to be temperature dependent only for which
polynomial expressions have been used. Viscosity (103 Ns/m2) and thermal conductivity
(W/mK) of water can be expressed as (Tw in K):
(4.26)
k w = 0.08331 + 0.00174 Tw
(4.27)
4.2.4
Discretisation and energy balance in the internal heat exchanger have been carried
out the same way as that in the gas cooler and similar correlations have been employed as
well for estimating heat transfer coefficients and pressure drop. It may be noted that
Reynolds number for annular flow is based on hydraulic diameter, ( = Di d o ) and
Nusselt number is based on equivalent diameter {= ( Di 2 d o 2 ) d o }. Internal heat
exchanger effectiveness () is expressed as:
=
h1 h6
h3 h6
(4.28)
(4.29)
where, Qihx 0 is the heat gain ( + sign) or loss ( sign) with the ambient.
The irreversibility in the internal heat exchanger is given by,
80
P
+ Qihx 0 0 1
I ihx = T0 mref [( s1 s6 ) ( s3 s4 )] + I ihx
T
ihx , r
(4.30)
It may be noted that the last term on the right hand side of this equation is always positive
whether Tihx , r > or < T0 , which indicates that exergy is always degrading whether heat
loss or gain with the component. Heat transfer with the ambient for all the components
has been estimated employing the conventional natural convection equations assuming
that no other heat transfer mode is existent.
4.2.5
Expansion device
h4 = h5
(4.31)
I exp = T0 mref ( s5 s4 )
4.2.6
(4.32)
Performance parameters
The system performance measures are based on the system COP and the exergetic
efficiency. The system COP (combined heating and cooling) is given by,
COPsys =
( Q
evw
+ Qgcwi
)W
(4.33)
comp
The exergetic efficiency is the percentage ratio of total exergy output to the
exergy input, where the output exergy can be found by subtracting the total system
irreversibility (summation of irreversibilities of all the components in the system) from
exergy input to the system and is given by:
II =
Ein I Components
Ein
I Components
100% = 1
100%
Ein
81
(4.34)
No
No
Update P1,
Superheat
Update P6
Yes
Maximum COP
No
Update Pdis
Yes
Output: state points, , COPmax, Pdis,opt
82
83
al. [38]. Figure 4.5 shows the variation of refrigerant side heat transfer coefficient and
mean values of Nusselt number in the gas cooler with bulk temperature for mass flow
rate of 0.029 kg/s, inlet pressure of 108 bar and water inlet and outlet temperatures of 30
o
particularly near critical region due to abrupt change in refrigerant properties; heat
transfer coefficient experiences a threefold change and Nusselt number varies from 720 at
gas cooler inlet to 550 at gas cooler exit with a peak value of 1180. This trend can be
attributed to the fact that, for certain tube diameter, though the Reynolds number of
refrigerant flow slightly decreases in the downstream direction due to increase in both
density and viscosity, the Prandtl number increases very significantly from 0.9 to 3.2 with
a peak value of 4.2 near the critical region due to sudden rise in isobaric heat capacity.
Similar theoretical and experimental results can be found in the literature [34,38] as well.
Absolute values of Nusselt number and heat transfer coefficients differ from
experimental values of Pitla et al. [38] due to different waterside temperatures resulting in
different tube wall temperatures. An inner diameter of 4.72 mm, used to plot Figure 4.5,
results in a total pressure drop of 1.2 bar; as expected, for an inner diameter of 7.875 mm
pressure drop becomes lower but the corresponding heat transfer coefficient gets halved.
The cooling COP as well as pressure loss increase with increase in heat exchanger length.
Here, results are presented for standard stainless steel inner tubes of 3/8 inch OD (9.525
mm, thickness of 0.815 mm, suitable to withstand high pressures) and outer tube of 5/8
inch OD (14.097 ID and 15.875 mm OD with 0.815 mm thickness) for evaporator,
internal heat exchanger and gas cooler. Internal heat exchanger length has been taken as 4
m, which yields an effectiveness of about 60 70%. A Dorin brand compressor (model
TCS113) with a rated speed of 2900 rpm and a swept volume of 11.7 cm3 has been
chosen [10]. Ambient temperature is assumed to be 30 oC (average in the Indian subcontinent) for the analysis.
Presented results show the effects of varying area ratio and operating conditions
for the same combined length (25 m) of evaporator and gas cooler. The parameters varied
are: compressor speed from 1500 to 3500 rpm, water inlet temperature from 20oC to 40oC
and gas cooler to evaporator heat transfer area ratio from 1.0 to 3.0 (same as the length
84
ratio of gas cooler and evaporator in this case because of equal diameters). Unless
otherwise specified, the mean values of these parameters are: compressor speed of 2900
rpm, water inlet temperature of 30oC and an area ratio of 1.8.
parameters is varied within the specified range stated above while the other two are kept
constant at the mean value to generate data for the plots. The compressor discharge
pressure has been optimized based on the maximum system COP using steady state
simulation and both energetic and exergetic results are presented corresponding to the
optimum value.
r ( kW / m 2 K )
Nu r
13
di = 4.72 mm
do = 6.35 mm
Lgc = 15 m
T2 = 398 K
11
1200
1100
1000
900
800
700
7
5
300
15
600
320
340
360
380
500
400
85
refrigerant flow rate. The decrease in mass flow rate is due to decrease in optimum
discharge pressure as decrease in suction density (Figure 4.7). Hence, the system COP
first increases and then decreases with a peak value of 3.95 at a gas cooler-to-evaporator
area ratio of 1.8 (Figure 4.6). Whereas the corresponding optimum gas cooler to
evaporator conductance (UA) ratio is 1.15 (close to equal allocation). This optimum area
ratio shifts to higher values with decrease in compressor discharge pressure as evident
from Figure 4.8. The maximum cooling COPs also decrease with decrease in discharge
pressure. It may be noted that the optimum value of heat exchanger area ratio also
depends on other factors, such as water inlet temperature, heat exchanger dimensions and
compressor specifications.
4
4.8
(kW)
3.92
3.8
3.88
Compressor work
Cooling output
System COP
3.3
System COP
3.96
4.3
3.84
2.8
3.8
1
1.5
2.5
86
pressure was found to remain almost constant varying between 107 and 109 bar as the
speed was modulated between 1500 and 3500 rpm.
110
mass flow rate
optimum pressure
108
0.026
106
104
0.024
Pdis,opt (bar)
0.028
102
0.022
100
1
1.5
2.5
Figure 4.7 Variation of optimum discharge pressure and mass flow rate with area ratio
4
System COP
3.8
3.6
3.4
90 bar
100 bar
3.2
110 bar
3
1
1.5
2.5
Figure 4.8 Variation of system performance with area ratio and discharge pressure
87
4.2
Compressor work
Cooling output
System COP
5
4.5
4.15
4.1
(kW)
4.05
3.5
3
System COP
5.5
2.5
3.95
2
1.5
1500
2000
2500
3000
3.9
3500
4.4
5.2
(kW)
4
4
3.8
Compressor work
Cooling output
3.4
System COP
4.2
4.6
3.6
System COP
2.8
3.4
20
25
30
35
40
88
0.025
optimum pressure
mass flow rate
0.0248
Pdis,opt (bar)
115
0.0246
110
0.0244
105
0.0242
100
120
0.024
20
25
30
35
40
Figure 4.11 Variation of optimum pressure and mass flow rate with water inlet
temperature
The effect of water inlet temperature at a compressor speed of 2900 rpm and area
ratio of 1.8 is shown in Figures 4.10 and 4.11. As the water inlet temperature increases,
the cooling COP decreases due to the combined effect of shifting of cooler exit
refrigerant temperature to higher value that cause the degradation of heat transfer
properties in gas cooler and decrease in refrigeration effect. Water inlet temperature has a
negligible effect on refrigerant mass flow rate as shown in Figure 4.11. However, the
optimum discharge pressure increases rapidly with increase in water inlet temperature
due to rapid change of refrigerant outlet temperature in the gas cooler.
A nomogram has been developed based on the results of steady state simulation
model (Figure 4.12). This graphical aide to system design has been developed to obtain a
design through optimal area ratio, discharge pressure for achieving maximum cooling
COP. The parameters varied are compressor speed, water inlet temperature and area ratio.
Such a nomogram helps the design engineer to ascertain expected output parameters once
the input design and operating parameters are fixed. For example, when the water inlet
temperature is 30C, compressor speed is 2500 rpm, and consequently to attain maximum
89
system COP of 4.01, optimum area ratio and discharge pressure have to be 1.76 and 107
respectively as is illustrated in the nomogram, and the system is expected to yield a
cooling output of 4 kW.
Figure 4.12 Design nomogram for a transcritical carbon dioxide heat pump
The optimization of a carbon dioxide based cooling-heating system is quite
complex as it depends on several parameters such as compressor speed and efficiency,
water inlet temperature and flow rates, heat exchanger dimensions, etc. However,
constraining some of the parameters such as required cooling/heating output, specific
90
compressor, area ratio, etc. based either on the requirements or other limitations makes
the task somewhat simpler. For the system studied here, the output cooling capacity and
water outlet temperatures are fixed by the application; in addition a standard Dorin
compressor with fixed displacement volume and speed is selected and the area ratio is
taken as 1.8. Under these conditions, the maximum COP (at optimum discharge pressure)
and the optimum discharge pressure itself become a function of the water inlet
temperature only. For the above input conditions, the following equations can be
regressed from the generated data with correlation coefficients (R2 values) of 0.999 and
0.995 respectively:
COPsys ,max = 5.12 0.039 twi and Pdis ,opt = 85.45 + 0.774 twi
(4.35)
where water inlet temperature (twi) ranges between 20 and 40C. COPsys ,max is the system
COP at optimum discharge pressure.
91
be attributed to pressure drop in evaporator being more than that in gas cooler and
internal heat exchanger due to both frictional and momentum effects.
24
3.95
System COP
23
3.85
22.5
22
3.8
21.5
3.75
21
Exergetic efficiency
3.7
23.5
3.9
20.5
System COP
3.65
20
1
1.5
2.5
Figure 4.13 System performances with varying heat exchanger area ratio
35
Compressor
30
Internal HEX
25
Exp. valve
Evaporator
20
Gas cooler
15
10
5
0
1
1.5
2.5
Figure 4.14 Variation of component irreversibility with heat exchanger area ratio
92
0.75
0.65
0.55
Evaporator
0.45
Internal HEX
0.35
Gas cooler
0.25
0.15
0.05
1
1.5
2.5
Figure 4.15 Influence of heat exchanger area ratio on irreversibility due to pressure drop
5
25
System COP
4.5
23
22
4
21
3.5
20
System COP
Exergetic efficiency
24
19
Exergetic efficiency
3
18
20
25
30
35
40
93
work and also due to decrease in cooling output. Water inlet temperature has a negligible
effect on refrigerant mass flow rate. However, the optimum discharge pressure increases
rapidly with increase in water inlet temperature due to rapid change of refrigerant outlet
temperature in the gas cooler. Hence the exergetic efficiency of the system deteriorates
with rise in water inlet temperature as the exergy losses in the evaporator and in the
internal heat exchanger increase rapidly due to the increase in heat exchanger temperature
differences. Although the irreversibility in gas cooler remains fairly constant with
changes in water inlet temperature, the exergy loss in the expansion valve increases.
The effect of compressor speed on system performance at a heat exchanger area
ratio of 1.8 and a water inlet temperature of 30oC is presented in Figure 4.18. It is
observed that the system COP at optimum discharge pressure decreases as both
compressor work and cooling output increase with compressor speed. Maximum
increment of irreversibility occurred in the gas cooler. This may be attributed to an
increase in temperature difference in the gas cooler as well as the increase in pressure
loss due to rapid increase in flow velocity. Thus increase in compressor speed yields
higher capacity and higher irreversibility as well due to higher mass flow rate and higher
frictional pressure loss.
35
Compressor
30
Internal HEX
25
Exp. device
20
Evaporator
Gas cooler
15
10
5
0
20
25
30
35
40
94
4.1
System COP
24
4
23
22
System COP
3.9
21
25
Exergetic efficiency
3.8
2000
2500
20
3500
3000
Compressor
input,
100%
Output
23.1 %
Internal
HEX,
2.7 % Gas
cooler,
Compressor,
17.0 %
33.5 %
Expansion
device,
10.6 %
Evaporator,
13.1 %
95
exchanger is negligible. Due to the high pressure drop occurring in the system being
studied, the expansion device contributes a much larger fraction of irreversibility
compared to conventional systems.
System COP
System COP
38
Exergetic efficiency
36
34
32
30
4.5
28
26
24
40
5.5
22
3.5
20
50
55
60
65
70
75
80
96
System COP
4.4
25
24
23
22
21
20
4
3.6
System COP
3.2
19
18
17
16
Exergetic efficiency
2.8
15
20
25
30
28
27
26
4.8
35
Figure 4.21 System performances with varying total heat exchanger length
4.6.1 Compressor
Irreversibilities in the evaporator and the gas cooler occur due to the temperature
difference existing between two heat exchanger fluids, pressure loss, flow imbalance and
heat transfer with the ambient. Results show that almost 90% of the irreversibility occurs
due to fluid temperature difference and 10% due to the rest. At the mean conditions, the
average fluid temperature differences in evaporator and gas cooler are about 22oC and
40oC respectively, whereas the pressure losses are 2.5 bar and 0.9 bar respectively. The
97
effective fluid temperature difference can be reduced by increasing heat transfer area,
either by increasing the heat exchanger length or by incorporating fins; however, both
will result in higher pressure drop. System COP and exergetic efficiency of the system
increase, first rapidly and then slowly, because of increase in irreversibility due to
pressure loss (Figure 4.21). Heat transfer with the ambient is inconsequential for the
system performance; if better insulation reduces net outer wall conductivity from 20 to 1
W/mK, exergetic efficiency will increase by a mere 0.2 %.
4.6.3 Expansion device
4.7 Summary
The steady state performances for both energetic and exergetic points of view of a
carbon dioxide based transcritical heat pump to cater for heating and cooling
simultaneously have been presented in this chapter. The results are obtained for a system
that can be used, for example, in dairy applications of chilling and pasteurization of milk.
Unlike previous studies reported in literature, realistic heat transfer and fluid flow effects
have been included in this comprehensive analysis. The highly variable heat transfer
characteristics of the refrigerant in the gas cooler have been included in the analysis for
better accuracy. Spatial discretisation of all the heat exchangers has been carried out as
well to yield better precision where fluid properties change rapidly. A computer model
98
has been developed to first simulate the system at steady state for different operating
conditions and then evaluate the system performance based on COP as well as exergetic
efficiency. Additionally, component level irreversibility analyses have been performed.
Results are obtained by varying important operating and design parameters such as heat
exchanger area ratio, compressor speed and water inlet temperature over a given range.
An enumerated summary of the results is as follows:
1.
Optimum heat exchanger area ratio ranges between 1.7 and 1.9 for maximum
system COP as well as maximum exergetic efficiency at optimum discharge
pressure, although optimum value for later is slightly more.
2.
3.
4.
The temperature difference in heat exchangers contributes more than 90% of the
irreversibility, whereas the rest occurs due to pressure drop and heat transfer with
surroundings in the heat exchangers.
5.
99
6.
The compressor, evaporator, gas cooler and expansion device contribute to system
irreversibility to a larger extent while the internal heat exchanger has negligible
effect. The expansion valve contributes a significant amount of exergy loss here
whereas it is negligible for a conventional system.
7.
8.
Replacement of expansion valve with a turbine will increase the COP as well as the
exergetic efficiency significantly, but it will also raise issues related to cost, design
and dynamic balancing of the system. It is advisable to employ a turbine for large
systems, such as a large dairy plant or other large system where simultaneous
cooling and heating is useful.
100
Chapter 5
5.1 Introduction
For effective sizing of heat exchangers, detailed knowledge of total irreversibility
is essential. For example, for the same capacity, length can be reduced by decreasing
diameter, so that irreversibility due to material reduces, but the irreversibly due to
pressure drop will increase rapidly. To circumvent this problem, a multi-pass
arrangement can be used with very small tubes (microchannel heat exchanger), but this
will give rise to higher manufacturing irreversibility. So it is a fairly complex task to
choose an effective set of diameter, length and number of passes for heat exchangers.
Irreversibility occurs in fluid flow system through three mechanisms of entropy
generation: molecular thermal dissipation, viscous dissipation and chemical dissipation.
The first mechanism leads to irreversibility due to temperature difference and the second
one leads to irreversibility due to pressure drop, while the third one is not generally
experienced in vapour compression heat pumps due to chemical stability of the
refrigerant. Different approaches such as entropy generation minimization, life cycle
analysis, exergoeconomic or thermoeconomic analysis have been employed in recent
years for optimization of heat exchangers mostly with constant fluid properties. High
pressure, distinct dry out phenomenon and near critical operation in evaporator, and
supercritical operation and abrupt variation of thermophysical and transport properties of
CO2 in gas cooler make this type of analysis very interesting in transcritical CO2 heat
pump systems.
In this chapter, detailed irreversibility analysis of both the evaporator and gas
cooler is presented for a CO2 based heat pump system of 1 ton cooling capacity with
101
water as the secondary fluid for both heat exchangers. This analysis includes both
operational and material irreversibility. The present study includes the numerical
optimization of multi-pass counter-flow double-tube evaporator and gas cooler. Effect of
three parameters: tube diameter, length and number of passes have been studied in detail.
Results reveal the right combination of optimum diameter and length, which depends on
number of passes, capacity and operating parameters.
(5.1)
where, I T and I P indicate the irreversibilities due to thermal and viscous dissipation
respectively. The parameters kT and k P depend on the nature of applications of thermal
system, i.e., whether the heat exchanger is used in a power plant, heat pump,
refrigeration, etc. and its working characteristics. The irreversibility associated with use
of material includes total life cycle of material and the effect on environment. However
neglecting recycling and the environmental effect, the irreversibility associated with use
of material can be simply written as:
I man =
1
(M i , p + M o , p )C p + ( M i ,man + M o ,man )C man + LC fab + M ins C ins
t n
(5.2)
where, the first term denotes irreversibility due to raw material for inner and outer tubes,
the second term denotes that for inner and outer tube manufacturing processes, the third
102
and fourth terms for heat exchanger fabrication and insulation respectively and t
indicates the total cycle time. So the total irreversibility is given by,
(5.3)
For certain applications and working fluids, if we try to reduce the I T , I man and
I P will increase. Likewise, with the same diameters, we can reduce I T by increasing
length, but I man and I P will increase and hence the total irreversibility will be minimum
for a particular optimum length. Analytical formulation to obtain optimum length based
on minimum total irreversibility is impossible for a real heat exchanger due to
dimensional variation of properties. Bejan [110] has optimized the single tube aspect
ratio based on combined thermal and frictional irreversibility for a balanced heat
exchanger with an ideal fluid and derived the following equation for a simplified case of
constant Stanton number (constant Reynolds number):
4L
=
1/ 2
D opt G ( R / c p ) f St
(5.4)
where, St = f (Re, Pr) . This optimization analysis can be extended by adding I man . For a
single tube, equation (5.2) can be rewritten as I man L or I man = CL for constant
diameter. The total irreversibility can thus be written as:
I tot = kT T0
C p D 2
4 LSt
+ kP
DG 3 f L
+ CL
2 2
(5.5)
Lopt =
kT T0 C p D / 4St
(5.6)
C + k P DG 3 f / 2 2
It is clear from equations (5.4) and (5.6) that the optimum length will reduce if the
irreversibility associated with use of material is included.
103
I T
I P
I T
I
(5.7)
(5.8)
104
total irreversibility of evaporator and gas cooler has been evaluated by summing up the
irreversibility in each segment and pass.
Specifications of a Dorin compressor (Model TCS113) have been used in the
simulation and the following correlation have been used for volumetric and isentropic
efficiency respectively, based on regression of experimental data [112]:
(5.9)
T
I evT = T0 mref (s 6 s5 ) mevw c pwln evi
Tevo
pass
(5.10)
(5.11)
Similarly, the irreversibility due to thermal dissipation in gas cooler is expressed as:
T
I gcT = T0 m gcw c pw ln gco m ref (s 2 s3 )
Tgci
pass
(5.12)
and the irreversibility due to pressure drop in the gas cooler is given by:
(5.13)
supercritical pressures where the thermo physical properties of carbon dioxide change
drastically. The great variations in the thermo physical properties cause the heat transfer
coefficient to be greatly dependent on both the local temperature and the heat flux. The
variation includes two aspects: changes along and perpendicular to the direction of fluid
105
flow. Longitudinal discretisation accommodates the former effect. To account for the
variation in the perpendicular direction, Pitla et al. [38] correlation has been used in this
study (Equations 4.12-4.14). Pressure drop has been estimated following Equations 4.154.16.
5.4.2 Evaporator (refrigerant side): High pressure, very low viscosity and surface
tension, and near critical operation make the flow boiling heat transfer and pressure drop
phenomenon of carbon dioxide distinct from conventional refrigerants. Distinct film
breakdown and dry-out phenomena make most of the general correlation unusable. In this
analysis, the recently developed Yoon et al. [52] correlation has been used for boiling
heat transfer coefficient. The following correlation was proposed to predict critical
quality:
1.64
Bd 4.7
(5.14)
2
2
htp = ( S .hnb ) + ( E.hl )
1/ 2
(5.15)
0.12
( log ( P / P ) )
10
cr
1.11 1
htp =
0.55
M 0.5 q 0.67
(5.16)
0.11
(5.17)
(5.18)
hwet is the heat transfer coefficient on the wetted portion of the tube, given by,
hwet = Ehl ,
E = 1 + 3000 Bo
0.86
x
+ 1.12
1 x
0.75
0.41
(5.19)
where, dry is angle of dry portion, which is closely related with the flow pattern, given
by:
106
dry
1
= 36.23Re3.47 Bo 4.84 Bd 0.27
2
X tt
2.6
(5.20)
hl and hg are the heat transfer coefficients corresponding to saturated liquid and vapour,
given by Dittus-Boelter equation.
Pressure drop has been calculated from modified Martinelli correlation
pTP =
2 f fo G 2 L 1 x 2
TP dx ,
d x 0
(5.21)
0.25
( P / Pcr )
1.64
(5.22)
5.4.3 Waterside heat transfer and pressure drop: Waterside heat transfer coefficient
has been evaluated by the conventional Dittus-Boelter equation for annular flow in both
evaporator and gas cooler.
Nuw = 0.023Re0.8 Pr n
(5.23)
(5.24)
To simplify the analysis, the area ratio between annulus and inner cross section is
assumed to be constant (=2.0). This is a more reasonable assumption than the case of
constant diameter ratio because of the fact that the relative mass velocity and Re will
107
Tube material for both evaporator and gas cooler is stainless steel. The information on the
exergy losses for the material uses is limited. The exergy loss in steel production depends
on the manufacturing processes and is found to vary from 4 to 18 MJ/kg [114,115]. Here,
an average value of 11 MJ/kg has been taken for exergy losses for primary steel. The
exergy losses due to tube manufacturing, fabrication and insulation (density: 30 kg/m3)
have been taken as 5.7 MJ/kg, 0.26 MJ per metre length and 71 MJ/kg respectively [114].
The following operating parameters have been assumed for both evaporator and
gas cooler. For evaporator, inlet and outlet temperatures of water are 30 oC and 4 oC
respectively, typically used for dairy application. For gas cooler, inlet temperature and
mass flow rate of water are taken as 30 oC and 2 kg/min respectively, which will yield an
outlet water temperature of about 70-80 oC, typically required in dairy applications. The
heating capacity and the corresponding outlet temperature may vary.
The ambient
temperature has been taken as 30 oC. Calculations are based on 1 ton of cooling capacity;
compressor work (exergy input) and heating capacity may vary based on the simulation.
5.5.1 Exergetic optimization of gas cooler
The effect of a tube diameter and length of a 2-pass gas cooler on the total
irreversibility is shown in Figure 5.1. The minimum irreversibility is seen to lie in the
range of 6 mm diameter and 25 m length. Decreasing the diameter beyond about 4 mm
yields a rapid increase in irreversibility due to a rapid increase in pressure drop. This can
be avoided by increasing the number of passes. For a 5-pass unit, the optimum diameter
and length are found to be 4.5 mm and 22 m respectively as shown in Figure 5.2. For
larger diameter and smaller length, the irreversibility due to thermal dissipation is very
high because of lower heat transfer coefficients; increasing the length can reduce this
irreversibility, but that will also give rise to higher I man and I P . However, the effect of
108
I P (< 0.1 %) and I man (< 4 %) at larger diameter is fairly small as shown in Figures 5.3
and 5.4. On the other hand, for a very small diameter, the fluid velocity rises (25 m/s for
d = 2mm and n = 5) leading to high Re (1.6 105 approx.) and a larger pressure drop as
shown in Figure 5.3, although the heat transfer coefficient will also increase in this case.
For smaller diameter and long heat exchangers, two problems will arise: i) Due to
the increase in heat transfer coefficient the temperature approach becomes zero, which is
practically not feasible and hence there is no further decrease in thermal dissipative
irreversibility, ii) A very rapid increase in pressure drop related irreversibility occurs. If
we recall the pressure drop equation, for certain size and operating condition, G n 1 ,
Re n 1 and thus the pressure drop is given by, P n x . For water, x = 1.75 and for
CO2, x > 1.5 . So, the irreversibility due to pressure drop I P n1 x (according to
Equation 5.13). In the present case, for a 4 mm diameter and 30 m long heat exchanger,
I P = 21 W with n = 2 and I P = 10 W with n = 5, so the value of x will be
approximately 1.8. This small deviation is due to the change in other properties. I man will
increase approximate linearly with n. It is very clear from the above discussion that with
increase in n, I P will decrease and I man increases, so the optimum diameter and length
will both decrease, but it is not an easy task to find the relationship. With increase in
number of passes from 2 to 5, the optimum diameter decreases as a function of n 0.4
approximately, although the value 0.4 will reduce with increase in n due to increase in
I man , which will give some rough idea about the minimum number of passes required in a
microchannel heat exchanger to attain minimum irreversibility. For a 5-pass system with
inner diameter of 4.5 mm, the Equation 5.6 predicts the optimum length of about 8 m, for
C = 0 that will be about 13 m. However, for the water side, Equation 5.6 estimates a very
large optimum length (about 40 m) due to negligible pressure drop. Since Equation 5.6 is
based on several assumptions such as balanced heat exchanger, constant properties and
Reynolds number, the deviation in the predicted results from the numerical results is
quite intuitive.
109
110
111
The effects of diameter and length of evaporator with 2 and 5 passes on the total
irreversibility are shown in Figures 5.5 and 5.6, respectively. Unlike the gas cooler, there
does not exist any optimum length for the evaporator, although some optimum diameter
exists having values of 7.5 mm and 6 mm for 2 and 5 passes respectively. Reducing the
diameters (below 4 mm for 2 passes and 3 mm for 5 passes approximately) leads to very
rapid increase in irreversibility due to a fast rise in pressure drop. This can be avoided by
increasing the number of passes. As can be observed in the figure, the portion of the
contour at lower diameter and higher length is incomplete. This can be attributed to the
simulation being not feasible due to temperature approach tending to zero because of
higher pressure drop. So, beyond certain value of length the temperature approach
becomes negative, which is an unacceptably trivial situation as the heat transfer direction
reverses. So, the minimum temperature difference approaches zero before attaining the
optimum length.
112
113
114
5.6 Summary
Exergetic analyses of both evaporator and gas cooler of a CO2 based heat pump
system have been presented in this chapter. Water has been employed as the secondary
fluid in both heat exchangers. Typical operating conditions, required in dairy plants, have
been chosen for the analysis. Results clearly show that higher heat transfer coefficient can
be achieved by decreasing the diameter only to a limited extent due to rapid increase in
pressure drop thereafter. With increase in length, thermal irreversibility decreases but
irreversibilities due to both pressure drop and material become larger. So, for certain
operating conditions and capacity of gas cooler, a set of optimum diameter and length is
possible for each set of passes and the optimum diameter and length will decrease with
increase in number of passes. For the evaporator, although an optimum diameter has been
obtained, optimum length could not be found since as temperature approach becomes
zero before attaining the optimal length. The effect of material use on the irreversibility is
negligible. Although the effect of pressure drop on the irreversibility can be neglected for
higher diameter, it is quite significant for smaller diameter tubes. The effect of pressure
drop depends on the number of passes and mass velocity of the fluid. Irreversibility due
to pressure drop is higher for the evaporator compared to that in the gas cooler.
Outcome of such exergetic optimization exercise is expected to help design the
optimal heat exchanger (in terms of diameter, length and number of passes) for a given
capacity and the operating parameters.
115
Chapter 6
6.1 Introduction
Drying is an energy intensive process and is very common in many chemical,
food and process industries such as milk powder production, wood processing kilns. In
the conventional dryer, the exhaust air from dryer is vented to atmosphere and the useful
part of its energy is lost resulting in lower performance and lower specific moisture
extraction rate, SMER (about 0.2-0.6 kg/kWh). Using a heat pump dryer (HPD), which is
a combination of heat pump and dryer unit, both the latent heat and sensible heat can be
recovered from exhaust air thus improving the overall thermal performance and yielding
effective control of air condition at the inlet to the dryer. Studies show that SMER of a
heat pump dryer is in the range of 1.0-4.0 kg/kWh. Another advantage of HPD is that it
can yield dry (very low humidity) air of low temperature, which is essential in some
applications, such as drying of medicines. After the first installation of HPDs mainly for
timber drying in 1960s, many synthetic refrigerants such as R114, R22, R134a have been
used in HPDs. However, due to the adverse environmental effect of synthetic
refrigerants, natural refrigerants are being promoted lately. Carbon dioxide (R744) is one
such natural refrigerant, which has high potential for this application because of available
effective heating in the gas cooler due to gliding refrigerant temperature. CO2 also offers
various advantages such as high volumetric capacity, superior environmental properties
and favorable heat transfer properties.
To fully examine the operating characteristics, computer simulation of HPDs has
received attention in recent years. The simulation results explicitly explain the HPD
characteristics with respect to various operating parameters and results also help to
establish a fundamental guideline for the HPD operation.
However, experimental
116
In the present work, a mathematical model and simulation code has been
developed to investigate the performance of a transcritical CO2 heat pump dryer. The
model takes into account the momentum, heat and mass transfer phenomena taking place
in all components in the system. To take care of the variable heat transfer properties, the
heat exchanger components were divided into several small segments to examine the
state, heat and mass balance and pressure drop for both refrigerant and air, and hence
accurate results are expected from the present study. The simulation model has been first
validated with experimental data available in open literature and then the model is used to
investigate effects of important operating parameters on the system performances such as
heating COP, moisture extraction rate and specific moisture extraction rate. Simulation
results show the effect of key operating parameters such as by-pass air ratio, recirculation air ratio, dryer efficiency, ambient condition (temperature and relative
humidity) and air mass flow rate. Anomalies in the behavioral trends of the transcritical
CO2 heat pump are explained through comparison with such data for other refrigerant
based systems reported in the literature.
117
(6.1)
(6.2)
(6.3)
(6.4)
118
Figure 6.2: T-s diagram of R22 & R134a heat pump dryer cycle
(6.5)
(6.6)
119
A computer code was developed based on energy and exergy balance equations
incorporating the developed property codes CO2PROP for R744 and REFPROP [104]
for R22 and R134a. Although the AT is attained at point 4 in the evaporator
(Ta3 T4 = AT ), in the condenser the AT may be attained at saturated vapor point ( 2 ) or
state 2, so the temperature approach condition for R22/R134a HPD can be
mathematically represented as:
h h
min (t2 ta4 ), t2 ta3 + 2 3 (ta4 ta3 ) = AT
(6.7)
h
h
2
3
For CO2 systems, the entire temperature range is divided into elemental sub-ranges to
find the minimum temperature difference between refrigerant and air (AT), and the gas
cooler pressure is set as the optimum compressor discharge pressure (Equation 3.30).
With the given input parameters: ta1, a1 , ta 4 , TR , ma , other state points are
calculated by employing an effective iterative technique incorporating property data and
then performance parameters COPc, MER, SMER and irreversibilities have been
estimated.
6.2.3 Result and discussion
Results are presented in Table 6.1 for unit mass flow rate of air (1 kg/s) and for
the following input parameters: ta1 = 30 oC , ta 4 = 80 oC , inlet moisture content of
evaporator a1 = 0.02 kg/kg of dry air and reference temperature of 30 oC. Evaporator
outlet saturated air temperatures are 11.5 oC, 13.0 oC and 12.4 oC for R22, R134a and
R744 respectively. Results clearly show that due to the high system pressure the
volumetric capacity of R744 is very large compared to others. Another advantage for
R744 is the smaller pressure ratio compared to others, although the high system pressure
can pose some difficulty in design of components. COP of R744 based heat pump dryer
is 10% higher than that of R134a units, whereas it is 7% lower than that of R22 units. In
terms of moisture extraction rate as well R22 yields the best performance. SMER of
R744 system is 11% higher than that of R134a and 4% lower than that of R22 systems.
120
For optimum evaporator inlet moisture content of 0.0195 kg/kg, the system is found to
yield a maximum cooling COP. This type of behavior is attributed to the nature of
saturated line in psychometric chart.
Table 6.1 Performance Comparison for R22, R134a and R744 HPD
Refrigerant
a1
a4
rp
COPc
R22
0.02
0.0082
4.605
R134a
0.02
0.0090
R744
0.02
0.0087
2.170
Vc
kJ m3
3090.6
MER
kg/min
0.708
SMER
kg/kWh
1.840
5.838
1.812
1775.6
0.660
1.599
2.523
2.020
23172.2
0.6765
1.7779
Icomp (%)
Iex (%)
Iev (%)
Ic/gc (%)
R22
10.96
10.45
15.67
10.88
47.96
R134a
12.55
12.47
12.94
13.35
51.06
R744
13.20
15.75
13.15
7.75
49.85
121
Figure 6.4 Schematic diagram of a CO2 based heat pump dryer system
2c
2
Pressure, P
1
4
1c
Enthalpy, h
122
0.04
0.03
6a
2a'
0.025
0.02
1a
0.015
0.01
3a
5a
4a
0.005
0.035
0
-0.005
10
20
30
40
50
60
70
80
123
cooling and dehumidification (2a-2a) takes place. After the evaporator, both air streams
are mixed and the resulting dry and cold air (state 3a) is then heated sensibly in the gas
cooler (3a-4a) and through the fan (4a-5a). Then it passes through the dryer (5a-6a) to dry
the product. The amount of fresh air introduced in the system is equal to the exhaust air
from dryer vented to the atmosphere. Hence by adjusting re-circulation and bypass,
condition of air at inlet to dryer can be controlled.
The refrigerant mass flow rate through the compressor is given by,
mr = 1v Vs
N
60
(6.8)
Detailed modeling of the compressor is quite complex. The volumetric and isentropic
efficiencies and pressure drop in inlet and outlet valves depend on the compressor design,
inlet conditions, wall temperature, properties of coolant and fluid flow. To avoid the
124
complexity, the pressure drop in both inlet and outlet are taken as 0.1379 bar [117] and
specifications of a Dorin R744 compressor have been used for simulation and
correlations have been used for volumetric and isentropic efficiency, respectively, based
on regression of manufacturers performance data (Equation 5.9).
Compressor work is given by:
Wcomp = mr ( h2 h1 )
(6.9)
Gas cooler model is comparatively simpler due to absence of mass transfer. Gas
cooler considered here is a plate fin and tube cross-flow heat exchanger, consisting of
several rows and passes, It is assumed that both air and refrigerant mass flow is equally
divided in each pass. Considering the large property change of supercritical carbon
dioxide, the gas cooler is divided into number of sections to calculate the heat transfer
rate. Each section has been individually treated as a cross flow finned tube heat
exchanger and log mean temperature difference method is used in each section. The heat
transfer properties of both air and refrigerant are evaluated based on the mean
temperature and the pressure of respective section. The energy balance in each section is
given by (Figure 6.7):
(UA) ( LMTD ) = mri ( hr i hr i +1 ) = mai c pam (Ta i +1 Ta i )
(6.10)
For multi row plate fin and tube (without fouling), the overall heat transfer coefficient is
given by:
ln(d o / di )
1
1
1
=
+
+
UA r Ai
as Ao
2Lkt
(6.11)
s = 1 (1 f )
Af
Ao
f =
tanh ( m d o / 2 )
m d o / 2
(6.12)
and m and are determined following McQuiston et al. [118]. Summing up of heat
transfer in all the sections yields the total heating capacity of gas cooler.
125
The heat transfer coefficient on the refrigerant side is calculated using Pitla et al.
[38] correlation based on mean Nusselt number and is given in Chapter 4 (Equations
4.12-4.14). Pressure drop has been estimated in the same way as given in Equations 4.154.16 in chapter 4.
The airside forced convection heat transfer coefficient for a finned coil is given by
Rich [119]:
Ga trs
Pra =
(6.13)
a c pam
ka
This correlation is suitable for fin pitch within the range: 118 < N f < 787 .
Air side pressure drop in each segment is given by:
f
P = Ga 2 a
2 a
Ao 1
1
+ i +1 i
a
Ac a
(6.14)
126
f ad
A
= 0.589 o
A
p
0.28
Re a 0.27 ,
Re a =
Ga ( 4 Ac Ldc Ao )
(6.15)
As the evaporator of a HPD is used for cooling and dehumidifying of air, it plays
a very important role in moisture removing capability of the dryer system. The model of
evaporator is very complex due to simultaneous occurrence of heat and mass transfer.
The evaporator considered in the present study is a plate fin and tube type cross
countercurrent flow heat exchanger. Usually the working fluid, which flows inside the
tube, enters as a two-phase mixture and evaporates and thereafter may get superheated.
Hence in the evaporator model, refrigerant side consists of two regions: two-phase and
superheat, whereas outside the tubes, air is first cooled to dew point temperature and then
dehumidified by removing the moisture (both sensible and latent heat transfers occur
simultaneously). Hence the airside also consists of both dry region and wet region as
shown in Figure 6.8. The boundary of these two regions is set by dew point of the moist
air. To take into account the property variation, the evaporator is analyzed by dividing it
into number of sections and the heat transfer properties in each section are based on mean
temperature, pressure and/or humidity ratio of corresponding section.
127
Dry region: The dry region is treated similar to the gas cooler as only sensible heat
transfer takes place in this region. So heat balance in each computational section is
written as:
mri ( hr i +1 hr i ) = (UA )i (Tp Tr ) = (UA )o (Ta Tp ) = mai c pam (Ta i Ta i +1 )
(6.16)
where,
ln(d o / di )
1
1
=
+
,
(UA)i r Ai 2 kt L
(UA)o = as Ao
(6.17)
Water film thickness on both tube and fin surface are assumed to be equally distributed.
Tp and Tw are taken as mean surface temperatures of tube and water, respectively. So,
using the energy balance in each section (Figure 6.9), the following equations are
obtained:
q = mri ( hr i +1 hr i )
(6.18)
(6.19)
q = (UA )i (Tw Tr )
(6.20)
128
(6.21)
(UA)o = as Ao
(6.22)
(6.23)
where, s is the saturated specific humidity of air corresponding to the water surface
temperature ( Tw ). Lewis correlation is given by:
m = a c pam
(6.24)
Using the Lewis correlation and replacing the mean value , yields:
i +1 = i
2
1 + 2mai c pam
(
(UA)
s )
(6.25)
Simultaneously solving of equations (6.18-6.21) and (6.25), the air outlet temperature and
humidity, surface temperature, refrigerant outlet enthalpy and heat transfer rate can be
determined. Water film thickness is taken as uniform and equal to 0.1016 mm [121].
Determination of wet region
To distinguish the dry and wet region the air temperature at which condensation
first occurs should be determined. This occurs when the airside surface temperature
reaches dew point of air. The following correlation is employed:
Tad = Td +
r Ai (Td Tr )
as Ao
(6.26)
For dry region: Tp > Td or Ta > Tad ; for wet region: Tw Td or Ta Tad
Heat transfer equation
To estimate single-phase heat transfer for the refrigerant, the gas cooler equations
have been employed once more. The recently developed Yoon et al. [52] correlation has
been used for boiling heat transfer coefficient, as given in chapter 5 (Equations 5.145.20). For airside heat transfer coefficient, same equations have been used as for the gas
cooler.
129
For refrigerant side, two-phase pressure drop has been calculated from the
modified Martinelli correlation, presented in chapter 5. For single-phase refrigerant side
and airside dry region pressure drop, the gas cooler correlations have been used.
The following correlation [120] for friction factor was employed to estimate the
airside pressure drop in wet region:
f w = 0.318 f d
0.04
sf
tf
0.4
0.42
Re a
(6.27)
It may be noted that momentum effect reduces the overall pressure drop of air in the
evaporator as the second term of the pressure drop equation (6.14) is negative.
6.4.4. Expansion device model
The expansion process in CO2 heat pump dryer system is crucial as it is required
to control both the amount of superheat at the evaporator outlet and optimum compressor
discharge pressure to get maximum COP. There are various ways to control these as
discussed in chapter 2. In the present model, the expansion process is assumed to be
isenthalpic, i.e.
h3 = h4
(6.28)
The capacity of expansion device is assumed to be large enough so that the refrigerant
mass flow passing through the expansion device is the same as that through the
compressor.
6.4.5. Dryer Model
130
in the psychrometric chart. In this model, the dryer is treated as a black box; such an
approach was successfully used previously [117,122], where a contact factor (CF) or
dryer efficiency (DE) is introduced. Using the psychrometric relations, the outlet
condition of air in the dryer can be calculated by DE, which is defined as [117]:
DE =
T5 a T6 a 6 a 5 a
=
T5 a Ts
s 5a
(Figure 6.6)
(6.29)
To avoid actual simulation of the dryer the pressure drop in the dryer is evaluated using
the following correlation [100]:
P =
ma
am
(6.30)
SRC
The system resistance constant (SRC) of the dryer depends on various parameters such as
flow condition and product. In this model SRC is taken as an input parameter.
3.4.6. Fan model
The fan is used in a heat pump dryer system to generate a pressure head sufficient
to maintain flow of air in the system. Neglecting pressure drop in the duct, total air
pressure drop in the heat pump dryer system is given by:
Ps = Pev + Pgc + Pd
(6.31)
The power input to the fan, which can produce pressure rise of Ps is given by
WF =
1 ma
F am
Ps 100
(6.32)
The power input to the fan will get transferred as sensible heat to air and is expressed as:
WF = ma c pam (T5 a T4 a )
(6.33)
In the present model, only the temperature dependent air properties have been
used. The following polynomial correlations found by regression analysis based on
available data have been used for density, isobaric heat capacity, viscosity and thermal
conductivity of air:
131
(6.34)
(6.35)
(6.36)
(6.37)
where, temperature t is in oC. For the psychrometric properties of moist air, the
correlations given in ASHRAE Handbook [123] are used in this model. For
thermophysical and transport properties of CO2, the exclusive property subroutine
CO2PROP, the development of which was presented in chapter 3, has been employed
in this model.
6.4.8 Performance criteria of heat pump dryers
(6.38)
(6.39)
where, Wcomp and WF are the compressor power and fan power respectively.
MER = ma ( d ,o d ,i )
(6.40)
where ma , d ,i and d ,o are mass flow rate of air through the dryer, inlet and outlet
humidity ratios of air, respectively. The MER is vital for the drying process as it
indirectly indicates the dry product throughput rate. SMER is the most commonly used
performance criteria for heat pump dryers.
132
discharge pressure and using compressor specifications, refrigerant mass flow rate,
compressor work and state point 2 (inlet to gas cooler) are calculated from the
compressor model. After that, evaporator outlet condition of air (sate 2a and 3a) and inlet
condition of refrigerant (P4, h4) are determined by the evaporator model, then outlet
condition of refrigerant in gas cooler (P3, h3) and outlet condition of air in gas cooler
(state 4a) are determined by the gas cooler model. In both the gas cooler and evaporator
model, capacity, other state points, pressure drop for both refrigerant and air side are
determined using the heat exchanger configurations, input state points, and mass flow
rate of refrigerant and air. Particular attention is given to each computational section,
where the heat and mass balance, heat transfer and pressure drop correlations are used
and the Newton-Raphson method is used for iteration. Point 1 is iterated using Equation
(6.28) unless enthalpy of state 3 agreed with that of state 4 within a specified tolerance,
iteration was continued. In each iteration step, the discharge pressure of compressor is
updated by the following optimum correlation (Chapter 3):
P2 = 3.47 + 0.32 ( t1 t ) + 2.23 t3 0.0134 ( t1 t ) t3 + 3.7 103 t32
(6.41)
In the air circuit simulation, outlet condition of dryer (state 6a) was assumed to be the
same as ambient condition. Using RAR, state 1a is determined, then by refrigerant circuit
simulation, state 4a and heat pump performance is determined. Then by fan model, state
5a is obtained and by dryer model state 6a is determined. Unless the new condition of
state 6a agreed with the initially assumed state 6a within a specified tolerance, iteration
(fixed iteration) was continued using new values until convergence was obtained. The
flow chart of this simulation procedure is shown in Figure 6.10. Input parameters in the
simulation model are as follows:
1. Ambient condition ( tamb , RH)
2. Re-circulation air ratio (RAR)
3. Bypass air ratio in the evaporator (BAR)
4. Air mass flow rate ( ma )
5. Amount of superheat at evaporator outlet ( t ).
6. Compressor specifications
7. Evaporator and gas cooler geometries
133
Figure 6.10: Flowchart of the entire air and refrigerant loop simulation
134
required information reported, a few approximations have been made while using the
available data. Heat exchanger specifications are as follows. Gas cooler: inner and outer
diameters of the refrigerant tube are 8 mm and 9.6 mm respectively, aluminum fin of 0.3
mm thickness, tube spacing is 22 mm 25.4 mm and total air side surface area is 118 m2;
evaporator: same as gas cooler except that total air side surface area is 30.1 m2. Tube
array, pitch and length are not given; however, through visual inspection of the gas cooler
presented, an array of 8 20 tubes have been taken and length and pitch have been
adjusted in a way that the air side heat transfer occurs over a surface area of 118 m2.
Similarly, for evaporator, 30.1 m2 has been attained by taking an array of 6 10 tubes.
However, the tube arrangement is found to have little influence compared to other
parameters. Because of the reported transient characteristics, system behavior at 50 min is
considered for validation of steady state. The compressor inlet and outlet pressures are 46
bar and 80 bar, respectively. Basic strategy of this validation exercise is that first input
parameters of the simulation model are calculated from state points given in the results
using energy balance for both refrigerant and air, and then validate the numerical results
obtained from simulation. Due to lack of information on compressor characteristics,
refrigerant mass flow rate is taken as input parameter. For a given heating capacity of 12
kW, mass flow rate is estimated as 0.0665 kg/s. Kiln mass transfer coefficient is given as
0.75, which is higher than DE. DE and air mass flow rate are obtained graphically
through the psychrometric chart using available temperature data (Figure 6.11). Point 4
on the saturated line represents the minimum refrigerant temperature in the evaporator
and 4-KM is the tangent of saturated curve at that point. Correlations of saturated curve
and the tangent of the curve are given by [124]:
(6.42)
d s
= 0.0002 + 0.000036 ts
dts
(6.43)
Value of dts / d s at point 4 (11 oC) is 1677. Air temperature at cooler inlet (16 oC) lies
on the tangent and 3a to 5a (52 oC) is the constant humidity line. So, specific humidity in
gas cooler is 0.01113 kg/kg d.a. Energy balance on air in the gas cooler yields an air mass
135
flow rate of 0.334 kg/s. By taking the constant wet bulb line for dryer, the dryer
efficiency (DE) is given by:
5a
DE = KM KM
s 5a
(6.44)
Experi-
Numerical
Results
mental
Experi-
Numerical
mental
Cooling load, kW
10.15
11.56
COP
6.5
7.1
Heating load, kW
12 k
13.45
MER, kgw/hr
6.02
Work, kW
1.85
1.89
SMER, kgw/kWh
2.05
2.45
136
55
50
5a
Constant wet
bulb line
45
40
dt
= 1677
d
6a
35
30
KM
25
20
Saturation line
3a
3a
15
10
5
0
0.01
0.02
0.03
0.04
Figure 6.11: Air cycle with the results obtained by Klocker et al. [100]
137
It is assumed that water vapor condenses without freezing. The fan efficiency is
taken as 35%. The variation band of the following operating parameters is as follows:
RAR from 0 (open system) to 1 (closed system), BAR from 0 to 0.5, dryer efficiency
(DE) from 50% to 90%, ambient temperature from 20oC to 40oC, ambient relative
humidity from 30% to 70% and air mass flow rate from 0.2 to 0.5 kg/s. Unless otherwise
specified, the mean values of these parameters are: RAR of 0.5, BAR of 0.2, DE of 70%,
ambient temperature of 30oC, relative humidity of 50% and air mass flow rate of 0.25
kg/s. To generate plots, one of these parameters is varied within the specified range stated
above while the others are kept constant at the mean value.
6.7.1. Effect of bypass air ratio (BAR)
The effect of evaporator by-pass air ratio on the heat pump performance, MER
and SMER is shown in Figure 6.12. It is obvious that the cooling load on the evaporator
decreases and the heating load follows suit whereas the ratio of heating load to cooling
load increases with BAR. Due to bypassing of air, cooling and dehumidification rate of
evaporator will improve because of decrease in evaporation temperature and increase in
humidity ratio at the inlet to the evaporator; however, the temperature rise between
evaporator and gas cooler increases, and hence the compressor work increases due to the
increase in pressure ratio. Consequently the COP decreases (with increase of BAR by
0.5, 18% drop in COP occurs) with increase in BAR. However, on the other hand, the
influence of BAR on MER and SMER is not very significant. The increase in
dehumidification rate and decrease in COP points to the existence of an optimum BAR.
Although the MER increases continuously with BAR, the SMER attains a maximum
value, which depends on refrigerant properties and other operating parameters. With
increase in BAR by 0.5, MER increases by only 3.5%, and the maximum value of SMER
at BAR of 0.3 is only 4% more than that at BAR of 0. Prasertsen et al. [125] reported
similar effect of BAR (less than 2%) but did not get the optimal BAR for R134, whereas
Jia et al. [126] showed significant effect (20%) and dependency of optimum BAR on air
mass flow rate. These trends clearly indicate that bypassing of air is not a very effective
mode of system control.
138
3.02
7.5
6.5
2.98
Heating COP
5.5
2.96
MER
SMER
4.5
2.94
3.5
2.92
0
0.1
0.2
0.3
0.4
SMER (kg/kWh)
8.5
0.5
10
4
COP
MER
3.6
SMER
8
3.2
7
2.8
6
SMER (kg/kWh)
2.4
5
4
2
50
60
70
80
90
139
25
heating COP
20
MER
SMER
15
10
5
0
-5
-10
0
0.2
0.4
0.6
0.8
Figure 6.14 Effect of re-circulation air ratio on COP, MER and SMER;
At RAR = 0, heating COP = 4.3, MER = 7.558 kg/h, SMER = 2.905 kg/kWh
6.7.2. Effect of dryer efficiency (DE)
The influence of dryer efficiency (ranging from 50% to 90%) on heat pump COP,
MER and SMER is depicted in Figure 6.13 for an air mass flow rate of 0.25 kg/s. The
dryer efficiency represents the ratio of actual moisture extraction from the product to
maximum possible moisture extraction when drying takes place up to the saturation
condition. In the open system (RAR = 0), the heat pump COP is independent of DE and
the inlet condition of dryer is invariant with DE; hence MER and SMER are linear
functions of moisture extraction and vary linearly with dryer efficiency, whereas heating
COP is held constant. On the other hand, for RAR > 0 (Figure 6.13), the heat pump
performance is influenced by DE, although the effect is marginal. This influence is
dependent on the RAR and the relative condition of air at dryer exit and ambient. In this
simulation, the ambient condition is fixed. So, with increase in dryer efficiency, the
evaporator temperature should drop due to the decrease in air temperature at inlet to the
evaporator. Conversely, condensation in the evaporator increases due to increase in
140
humidity ratio at evaporator inlet, and hence the evaporator temperature tends to increase.
Due to these two contrasting effects, the evaporator temperature barely decreases (1% as
DE rises from 50% to 90%). Due to the decrease in air temperature at evaporator outlet,
the optimum gas cooler pressure and temperature decrease with increase in DE. So, both
lower side and higher side temperatures decrease, whereas the difference decreases
slightly. It is observed that the heating COP increases by only 6.5% when DE varies from
50% to 90% and the optimum compressor discharge pressure drops from 82 bar to 75.5
bar. There is little change in air condition at inlet to the dryer, whereas the MER and
SMER increase by 80% and 75%, respectively, as DE varies from 50% to 90% and the
variation is nearly linear.
70
10
Temperature
RH
62
58
54
50
temperature ( C)
66
5
0
0.2
0.4
0.6
0.8
Figure 6.14 shows the relative variation of heating COP, MER and SMER with
re-circulation air ratio compared to those for open system (RAR = 0). With increase in
RAR, less warm air is vented to the atmosphere and hence more energy gets accumulated
within the system, which increases the operating temperatures of both refrigerant and air.
Hence the inlet air to the dryer becomes warmer. As RAR rises from 0 to 1, air
141
temperature at inlet to dryer rises from 54 to 69oC whereas relative humidity increases
from 6.5 to 8.5% as shown in Figure 5.15. This will increase the MER and SMER, but it
also increases the temperature difference between evaporator and the gas cooler as well
as the individual temperatures of both and hence COP of the heat pump will reduce.
Because the optimum discharge pressure is more dependent on the refrigerant outlet
temperature in gas cooler than evaporation temperature, the optimum compressor
discharge pressure increases sharply with RAR. As shown in Figure 6.14, with increase
in RAR from 0 (open system) to 1 (closed system), heating COP decreases by 10%,
whereas the MER and SMER increase by 20% and 12%, respectively. The optimum
discharge pressure is found to increase by 50%. Due to the increase in compressor work,
the increase of SMER is much less than MER. The results yield an optimum value of
RAR for maximum MER as well as maximum SMER similar to that for refrigerant R22
[125]. However, in actual systems many other operating parameters influence this
optimal behavior of RAR.
11
3.3
3.2
3.1
9
MER (kg/h)
SMER (kg/kWh)
10
2.9
8
SMER
MER
2.8
2.7
7
20
25
30
35
40
142
Although the closed system (RAR = 1) is better in terms of MER and SMER, it is
required to ventilate some warm and moist air to the atmosphere (partially open: RAR <
1) to maintain the maximum refrigerant temperature and pressure within limit and also to
control the air condition at inlet to the dryer. In the end, the choice of open or closed
system is dependent on the user requirement and many operating parameters such as
ambient condition, dryer type and dryer efficiency as both the dryer exit and ambient
conditions affect the humidity of the air entering the evaporator. Clements et al. [127]
suggested that a heat pump dryer for the constant rate drying should be designed to
operate in two modes depending on the seasons (temperature). In winter, the dryer should
operate as a closed system and in summer the open system was suggested. Although in
summer, if relative humidity of ambient air is very high, then the open system may not be
desirable as it will increase the system load pulling the heat pump performance down.
6.7.4. Effect of ambient temperature
The performance based on the SMER and MER, for an ambient temperature
variation of 20-40oC for a relative humidity of 50%, is illustrated in Figure 6.16. With
increase in ambient temperature, inlet air temperature for the evaporator increases and
hence the working fluid temperature at all the points increases, which gives warm air at
inlet to the dryer. On the other hand, the specific humidity also increases with the
ambient temperature. With increase in ambient temperature from 20 to 40oC, the air inlet
temperature to the dryer increases from 55 to 73oC (about 33%) as shown in Figure 6.17,
whereas the inlet relative humidity to the dryer increases slightly. As a result, both MER
and SMER increase. With increase in ambient temperature from 20 to 40oC, MER
increases by 40% and the SMER increases by 18%. The effect of ambient temperature on
optimum gas cooler pressure is predominant because of significant increase in air inlet
temperature to the gas cooler. As both the evaporator temperature and the gas cooler
temperature increase proportionately, the temperature difference remains approximately
the same and hence the compressor work remains same. So the MER and SMER show
similar trends. On the other hand, due to the deterioration of heat transfer properties of
CO2 in the gas cooler, heating output decreases and hence the heating COP decreases. At
ambient temperature lower than 25oC, the optimum gas cooler pressure falls below the
143
critical pressure. Hence a constant gas cooler pressure of 74 bar has been considered in
the simulation, as a result COP is nearly the same for this region as shown in Figure 6.17.
Although Prasertsan et al. [125] showed that MER and SMER are not strong functions of
ambient temperature; the result obtained here shows that the ambient temperature
strongly influences both the CO2 heat pump and dryer performance.
4.4
4.3
heating COP
70
75
4.2
65
4.1
60
4
55
3.9
Heating COP
Dryer inlet temperature
3.8
50
20
25
30
35
40
Ambient temperature ( C)
Figure 6.17 Effect of ambient temperature on dryer inlet temperature and COP
6.7.5. Effect of ambient relative humidity
The dryer performance, MER and SMER are not strong functions of ambient
relative humidity. In general, the MER and SMER are expected to decrease as ambient
relative humidity increases in the open system. But in the closed system, evaporator inlet
temperature increases due to the increase in dryer outlet temperature and hence the
operating temperature of both CO2 and air increase. The air inlet temperature to dryer
increases by 17% as ambient relative humidity increases from 30 to 70%. As a result, the
MER and SMER increase slightly. With increase in ambient relative humidity from 30 to
70 %, MER and SMER increase by 9% and 2% respectively (Figure 6.18). Because of
the increase in temperature difference between evaporator and gas cooler, the compressor
144
work increases and on the other hand both cooling and heating outputs increase and
hence COP decreases slightly. As shown in Figure 6.19, with increase of ambient relative
humidity from 30 to 70 %, the heating COP decreases by 2.5%.
3.2
9.5
SMER
MER
3.1
8.5
3.05
MER (kg/h)
SMER (kg/kWh)
3.15
7.5
30
40
50
60
70
68
Heating COP
66
4.34
64
4.32
62
4.3
60
4.28
58
4.26
56
4.24
54
30
40
50
60
heating COP
4.36
70
Figure 6.19 Effect of ambient relative humidity on dryer inlet temperature and COP
145
3.3
4.9
4.7
4.5
3.1
4.3
3
4.1
2.9
SMER
3.9
COP
3.7
2.8
heating COP
SMER (kg/kWh)
3.2
3.5
2.7
3.3
0.2
0.25
0.3
0.35
0.4
0.45
0.5
Figure 6.20 Effect of air mass flow rate on SMER and COP
6.7.6. Effect of air mass flow rate
An increase in total mass flow rate of air causes the temperature difference
between the evaporator and the gas cooler to reduce, thus improving the heating COP and
SMER. On the other hand, the temperature at all points decreases and the refrigerant
mass flow rate increases due to increase in load. Influence of other parameters on
pressure drop is marginal (refrigerant side pressure drop in gas cooler and evaporator are
marginal at 0.5% and 2.5%, respectively) whereas mass flow rate influences pressure
drop significantly. Pressure drop for both CO2 and air increases with mass flow rate,
which in turn increases the compressor pressure ratio and hence the compressor work.
These two opposing effects combine to yield an optimum mass flow rate for maximum
COP and SMER, not necessarily at the same value. Although other operating parameters
have little influence on fan power (within 23% of compressor power), effect of mass
flow rate is quite significant. With an increase in air mass flow rate from 0.2 to 0.5 kg/s,
the fan power increases by 400% (4-5% of compressor power). Consequently, fan power
becomes significant at higher mass flow rates (8% of compressor power at 0.5 kg/s). This
causes the maximum SMER to occur earlier than maximum COP. As shown in Figure
146
6.20, the maximum SMER occurs well before a mass flow rate of 0.5 kg/s, whereas the
maximum COP occurs beyond that, although this optimization is not very definite as fan
power is a small fraction of compressor power.
The simulation results reported here show that the effect of ambient conditions
(mainly temperature) and dryer efficiency are significant on the performance of the CO2
heat pump dryer. Degree of RAR (open, RAR = 0; close, RAR = 1; or partially close, 0 <
RAR < 1) can be decided based on the ambient conditions and dryer efficiency to obtain
optimum performance. Control of RAR is fairly vital for some applications, where the air
temperature at inlet to the dryer has to be controlled.
6.8 Conclusions
The thermodynamic comparison of R744 with two other conventional refrigerants
R22 and R134a for heat pump drying in terms of both 1st and 2nd laws showed that R744
yields better performance than R134a whereas its performance is poorer when compared
to R22. Irreversibility of expansion device for R744 is higher whereas that of gas cooler
is smaller as compared to R22 and R134a systems.
A mathematical model and simulation code of a CO2 heat pump dryer has been
developed. Particular attention has been given to developing a detailed evaporator model.
Highly variable heat transfer properties of CO2 were also carefully considered. The heat
exchanger components were divided into several small segments in which the state, heat
and mass balance and pressure drop for both refrigerant and air were calculated. By this
approach, accurate results are expected. The numerical model has been first validated
with experimental data available in the literature and then used to investigate effect of
different operating parameters on important performance parameters such as COP, MER
and SMER of dryer system.
Individual validation of evaporator and gas cooler models shows that the gas
cooler model accurately predicts the performance whereas the evaporator model
overpredicts the performance. The assumption of an adiabatic model for the compressor
and negligible heat interaction with the ambient could be the other reasons behind the
slight overprediction of overall system performance. The present simulation model is
147
suitable for a continuous dryer where the dryer efficiency is constant and the operation is
fixed for a particular ambient condition. Results show that unlike BAR and ambient
relative humidity, the effects of DE, RAR, ambient temperature and air mass flow rate
are significant on system behavior. The performance of the dryer system increases
linearly with increase in DE. Although the SMER increases with ambient temperature,
COP deteriorates, whereas both decrease with ambient relative humidity. Both BAR and
RAR yield some optimum values for maximum SMER, but the COP decreases in both
cases. An optimization of air mass flow rate is possible to maximize both COP and
SMER although optimum values for maximum COP and SMER are not necessarily
identical.
148
Chapter 7
149
Figure 7.1: A fully instrumented CO2 heat pump test facility for simultaneous water
cooling and heating
7.2.1
Compressor
TCS 113
Bore/ stroke
22 mm/ 17 mm
Displacement
2.2 m3/h
Input voltage
3-phase, 380-420 V @ 50 Hz
Max Pressure
150 bar
Oil
Dorin CL80
Capacity
A Dorin CO2 compressor (model TCS113) was chosen for the experimental
investigation. The compressor is a single stage two-cylinder model. The detailed
specifications for TCS113 are given in Table 7.1. The motor is provided with Thermik
150
thermal protection. The inlet port of compressor was designed for a minimum suction
pressure of 10 bar. The compressor contains a safety valve for a maximum pressure 168
bar. Expected compression ratio was 2 to 4 and the mass flow rate range for the CO2 was
0.025-0.45 kg/s.
7.2.2
Expansion device
Numerical results in chapter 4 showed that the optimum discharge pressure varies
between 100 to 116 bar for the given range of water inlet and outlet temperatures. Hence
a maximum limit of 120 bar was considered to select the expansion device, although the
minimum discharge pressure was considered as 80 bar. Since it is recommended by the
compressor manufacturer to ensure a minimum evaporation temperature of 10 oC
(saturation pressure is about 26.3 bar), the minimum suction pressure limit was taken as
26 bar whereas 50 bar was set as the maximum limit. Hence the minimum and maximum
expansion ratios for the expansion valve should be 80/50 and 120/26 bar respectively. A
Swagelok integral bonnet needle valve (model SS-1RS4), shown in figure 7.2, was
chosen for this application. Detailed specifications are: stainless steel body, regulating
stem, 1/4" inch Swagelok tube fitting for both inlet and outlet, 0.37 Cv (flow coefficient),
pressure rating of 345 bar at 38 oC and temperature range of 54 oC to 232 oC.
151
7.2.3
Evaporator
Heat transfer and pressure drop analysis of the evaporator showed that a decrease
in tube diameter could improve the heat transfer although at the cost of pressure drop
increasing at a very faster rate (as discussed in chapter 5). For the present system design
condition, 3/8 inch (9.5 mm) OD tube will give rise to considerable pressure drop
(expected pressure drop will be up to 2.5 bar), whereas a lower standard diameter 1/4
inch (6.35 mm) OD will lead to very high pressure drop. Hence, 3/8 inch OD was chosen
for the evaporator. The tube thickness was taken as 1 mm (yielding an ID of 7.5 mm),
which is sufficient to sustain the expected refrigerant pressure. For water side (annulus),
standard stainless tube of 5/8 inch (16 mm approximately) OD with 1 mm thickness
(hence ID = 14 mm) was taken, although that was slightly higher than the design value.
Design mass flow rate of water was 1.2-3.2 L/min. The cross-sectional ratio of water to
refrigerant was 1.88. Total design length for the evaporator tubes was taken as 7.2 m.
Figure 7.3 shows a layout of the tube-in-tube counter-flow evaporator, which was
designed and developed for the test facility. Refrigerant flows through inner tube whereas
the water flows through the annulus. For the sake of simplicity of fabrication, only one
row was considered. The evaporator contains 9 parallel segments, each having 0.8 m
length (total 7.2 m), where the refrigerant tubes are connected by 180o circular bends
having the same diameter and the water tubes are connected by 90o straight tube of
inch OD. Sufficient gap between the two parallel segments was maintained in fabrication
for proper insulation and handling. The fabrication of the evaporator was done by brass
and silver brazing within the laboratory. Special care was taken during fabrication to
maintain uniform gap between the two tubes to obtain uniform peripheral heat transfer.
After the fabrication, the evaporator was tested up to a pressure of 75 bar for leak
detection and pressure sustainability. The fabricated evaporator has an effective length
0.825 m (total length is 7.4 m considering the 9 segments). The evaporator was properly
insulated using glassfibre insulation to reduce the heat transfer with the ambient.
Thermocouples were connected to each segment for detailed study of heat transfer
through the evaporator.
152
Gas cooler
Due to single-phase flow in the gas cooler, pressure drop is very low compared to
that of evaporator for the same conditions and the same diameter (as discussed in chapter
5). For 1/4 inch OD, the maximum pressure drop is around 1 bar accompanied by
excellent heat transfer rate as is evident from the simulation. Hence for the present
system design condition, 1/4 inch (6.35 mm) OD was chosen for the gas cooler. The tube
thickness was taken as 0.8 mm (leading to 4.75 mm ID), which is sufficient to withstand
the expected refrigerant pressure in the gas cooler. For the water side (tube annulus),
standard stainless tube of 12 mm OD with 1 mm thickness (10 mm ID) was taken. Mass
flow rate of water was in the range of 1.2-3.2 L/min for design conditions. The crosssectional area ratio of water to refrigerant was 2.64. Total designed heat transfer length
for the gas cooler was taken as 14 m. Figure 7.4 shows the layout of tube-in-tube counterflow gas cooler, which was designed for the experiment. Refrigerant flows through inner
tube whereas the water flows through the annulus. For the sake of simplicity of
fabrication, only two rows were considered. Gas cooler contains 14 parallel segments in
153
two rows, each 1 m long, where the refrigerant tubes are connected by 180o circular
bends having the same diameter and the water tubes are connected by 90o straight tubes
of 9.5 mm OD. Sufficient gap between the two parallel segments was maintained in
fabrication for proper insulation and handling. The gas cooler fabrication closely
followed that of the evaporator and similar practices were implemented in both the
fabrication processes. After the fabrication, the gas cooler was tested up to a pressure of
120 bar for leak detection and pressure sustainability. The fabricated gas cooler, as shown
in Figure 7.5, has an effective total heat transfer length of about 13.6 m. The gas cooler
was properly insulated by glassfibre insulation to reduce the heat transfer with the
ambient. The thermocouples were connected to each segment for detailed study of heat
transfer through the gas cooler.
Separator
The separator was designed for a total refrigerant capacity of 8 L. Figure 7.6
shows the separator, which was fabricated by rolling of stainless steel sheet metal
followed by tungsten inert gas welding. Mean diameter and height of separator are 175
mm and 350 mm, respectively. A wall thickness of 6 mm was taken for both side-wall
154
and upper and lower plates. Sufficient insulation was added to reduce the heat transfer
with the ambient. The separator has two ports; the inlet port is at the lower end and the
outlet port is at the upper end. Due to atomization of liquid refrigerant, it is possible that
carry-over of some liquid occurs which along with the vapor goes to the compressor. To
prevent this, two parallel horizontal thin plates covering 2/3 area, were installed inside
the cylinder for separation of the liquid fraction.
155
7.2.6
Receiver
It was expected that the receiver would contain mostly liquid (high density).
Hence, it was designed for a total refrigerant capacity of 2 L (lower than that of
separator). Figure 7.6 shows the receiver, which was fabricated similar to the separator.
Inner diameter and height of receiver are 90 mm and 300 mm, respectively. A thickness
of 6 mm was taken for both side-wall and upper and lower plates. Receiver was properly
insulated as well. Among the two ports, inlet port is at the higher side and the outlet port
is at the lower side. A pressure gauge was fitted on the upper side.
7.2.7
Condensing unit
An arrangement similar to the gas cooler was incorporated with the evaporator to
supply water at constant temperature and flow rate. Water re-circulation loop for
evaporator contains heater, insulated tank, centrifugal pump, flow control valve and mass
flow meter. Heating arrangement comprises number of immersion heaters of total
capacity 3.5 kW. By on-off control of each heater, the input heater load (equal to cooling
load of CO2 heat pump) was controlled. The pumping power was calculated following
the procedure of the condensing unit. By attenuating the heater and the valve, inlet
temperature and flow rate to the evaporator was controlled.
156
7.2.9
All the connections for the refrigerant circuit was made by SS-316 tube of 1/4
inch OD (6.35 mm) of thickness 0.8 mm (ID = 4.75 mm). Tubing was designed for
maximum pressure of 150 bar. Total tube length was estimated as 8 m. All the tube
connections were made by SS-316 ferrule adaptors. All the tubes were insulated by
asbestos rope covered by foam.
157
liquid or control the pressures. A Coriolis mass flow meter (item 3) is installed between
the separator and the compressor to measure the mass flow rate and temperature of CO2
vapour entering the compressor. Two Swagelok safety valves (item 2) are used in both
low and high pressure sides to control the higher limit of pressure. Four pressure gauges
(item 5) have been used in different locations. One differential pressure gauge (item 8)
has been used to measure the refrigerant side pressure drop in the gas cooler. CO2
cylinder (item 1) is used for external charging of CO2. A W-bend is provided before
compressor to provide superheating of CO2 vapour, if required. A temperature-controlled
bath is used in which W-bend is immersed. Although this was never used, because the
outlet of evaporator was already sufficiently superheated at operating conditions.
However, this provision for heating the refrigerant could be requisitioned during cooler
water inlet ambient temperatures in winter. A separate air-cooled condensing unit has
been used to supply water at required temperature and flow rate to the gas cooler. For
water inlet to evaporator at required flow rate and temperature, a separate heating unit
was used (not shown in figure). Thermocouples were fitted in different sections to
measure various temperatures; the sensors are subsequently connected to the data
acquisition system (DAS), which is interfaced with a computer as shown in figure 7.8.
Before test run, the system was purged by nitrogen gas and then evacuated by a
vacuum pump. Then CO2 gas was charged from the cylinder. Before setting, all the
measuring devices were calibrated properly (Appendix A). Reading of differential
pressure gauge and mass flow meters are set to zero. Then external heating and
condensing unit are started and set to certain inlet temperature and mass flow rate of
water. After recording initial reading and starting the data scan of the DAS, the
compressor was switched on. To measure the power input to the compressor, a 3-phase
energy meter was employed. Controlling the discharge pressure at a required level is
important for a transcritical CO2 system as it needs to operate at optimum discharge
pressure. This was achieved by simultaneous control of the total mass of the system and
degree of opening of the expansion device. The operating parameters were varied
following a test matrix as listed in Table 7.2.
158
Ranges
Parameters
Ranges
Psuc
26-45 bar
Tcwi
30-40 oC
Pdis
75-110 bar
mew
1.0-2.5 kg/min
Tewi
25-35 oC
mcw
0.7-1.5 kg/min
159
ql =
Tm Tamb
R
where, R =
ln( D1 / d o ) ln( D2 / D1 )
1
+
+
2k as
2k f
D2 f o
D2
D1
do
di
Tb
Tm
Asbestos Foam
Tube
wall
(7.1)
Tamb
ql
(7.2)
i has been found by Dittus-Bolter equation for water and Gnielinski correlation for
single phase refrigerant. For Tm < Tamb (evaporator side), heat transfer is from ambient
and same procedure has been used; with opposite sign. i for two-phase refrigerant has
been found by Yoon et al. correlation.
Volumetric efficiency of compressor has been determined by,
v =
mr
, where, 1 = f (P1 ,T1 )
1Vs N
(7.3)
160
is =
h2 s h1
where, h1 = f (P1 ,T1 ) , h2 = f (P2 ,T2 ) and h2 s = f (P1 , T1 , P2 )
h2 h1
(7.4)
(7.5)
(7.6)
Evaporating capacity and actual cooling effect of water are given by,
Qevr = mr ( hevr ,o hevr ,i ) and
(7.7)
Heat rejection of refrigerant and heat gain by water are given by,
Qgcr = mr ( hgcr ,i hgcr ,o ) and Qgcw = mgcw c pw (Tgcw,o Tgcw,i )
(7.8)
(7.9)
To find the heat transfer coefficient, gas cooler and evaporator was divided into 14
sections (1 m each) and 9 sections (0.85 m each), respectively. By measuring heat
transfer qr between CO2 refrigerant and water, and log mean temperature difference
(LMTD), UA can be calculated as: UA = q r LMTD (Note: qr , also heat gain/rejection by
CO2 = heat rejection/gain by water + heat gain/loss with ambient in one segment). Now
for segment of length L ,
ln(d o / d i )
1
1
1
=
+
+
UA r Ai
w Ao
2k t L
(7.10)
LMTD ln(d o / d i )
1
w Ao
2k t L
qr
(7.11)
161
Water side heat transfer coefficient ( w ) was estimated through Gnielinski correlation
[30].
1.8
test 1
Cooling COP
1.7
test 2
1.6
1.5
1.4
1.3
1.2
1.1
75
80
85
90
95
Cooling COP
1.8
1.6
1.4
1.2
test 1
test 2
0.8
72
82
92
102
112
162
120
85
100
65
Suction pressure
55
Discharge pressure
80
Discharge temperature
45
t gcw,o
tevw ,o
35
60
Temperature (oC)
Pressure (bar)
75
40
25
15
20
0
10
12
14
16
Time (min)
163
1.4
80
Pressure (bar)
1.3
70
1.2
60
1.1
1
50
0.9
Suction pressure
Discharge pressure
Mass flow rate
40
30
10
190
370
0.8
1.5
90
0.7
730
550
Figure 7.13 Suction and discharge pressure and mass flow rate with varying
expansion valve opening
3.2
Qevw,Qgcw (kW)
2.8
2.4
2
1.6
Cooling
Heating
1.2
75
80
85
90
95
Figure 7.14 Variation of cooling and heating output with discharge pressure for suction
pressure of 35 bar
164
3.8
3.4
3
2.6
2.2
1.8
Cooling
Heating
1.4
1
72
82
92
102
112
Figure 7.15 Variation of cooling and heating output with discharge pressure for suction
pressure of 40 bar
Variation of suction and discharge pressures, and refrigerant mass flow rate for
various valve openings (degree of opening measured from completely closed condition)
of the needle valve are shown in Figure 7.13. For this study, test was started by initial
opening of 720o and different angles of opening were maintained by gradual closing of
valve, measurements were recorded at steady-state for each position. When the valve is
gradually closed, pressure difference rises rapidly due to instant accumulation of mass in
gas cooler and reduction of mass in the evaporator, and then reduces gradually to steady
state. Valve opening exhibits a very significant effect near the full valve closing condition
as shown in figure.
Cooling and heating outputs of the system are plotted with discharge pressure in
Figures 7.14 and 7.15 for suction pressure of 35 and 40 bar, respectively. Both cooling
and heating outputs increase with discharge pressure due to increase in mass flow rate of
refrigerant. As the water mass flow rate in gas cooler and evaporator are kept constant at
1 kg/min and 1.5 kg/min, respectively, water outlet temperature increases in gas cooler
while it decreases in the evaporator with increase in discharge pressure due to increase in
both cooling and heating capacities.
165
50
48
46
44
42
Pev = 35 bar
Pev = 40 bar
40
2
2.2
2.4
2.6
2.8
Pressure ratio
166
4.2
3.5
4.1
2.5
System COP
Outputs (kW)
System COP
4.3
Cooling output
3.9
Heating output
3.8
1.5
0.7
0.9
1.1
1.3
1.5
Figure 7.17 Variation of system performance with water flow rate in gas cooler
4.4
4.35
3.5
3
4.25
2.5
4.2
4.15
Outputs (kW)
COPsys
4.3
1.5
System COP
Cooling output
Heating output
4.1
4.05
0.9
1.2
1.5
1.8
2.1
2.4
1
0.5
2.7
Figure 7.18: Variation of system performance with water flow rate in evaporator
Conversely, the effect of water temperature at inlet to gas cooler on the system
performance is fairly significant. Figure 7.19 exhibits the experimental data for a suction
pressure of 40 bar and a discharge pressure of 90 bar, mass flow rates of water in gas
167
cooler and evaporator of 1 kg/min and 1.5 kg/min respectively, and water inlet
temperature to evaporator of 31 oC. With almost 10 oC increase in water inlet temperature
the system COP decreases by about 13% due to deterioration of heat transfer properties
of CO2 as discussed in chapter 4.
4.8
4.7
tevw,i = 31 oC
System COP
4.6
4.5
4.4
4.3
4.2
4.1
30
32
34
36
38
40
tgcw,i (oC)
Figure 7.19: Variation of cooling COP with water inlet temperature in gas cooler
Effects of some important operating parameters on the system performance have
been discussed above. Maximum cooling and heating output from the system have been
recorded as 3 kW and 5 kW respectively. Maximum heat gain or loss from evaporator
and gas cooler were 20% and 30%, respectively. It can be noted that calculated system
COP is less than the ratio of (heating output + cooling output) and (heating output
cooling output) due to heat gain or loss. Maximum volumetric and isentropic efficiencies
of compressor have been recorded as 60% and 50% respectively.
168
169
7
6.5
Pitla et al.
correlation
Experimental
6
5.5
5
4.5
4
3.5
3
2.5
2
42
51
60
69
78
87
96
105
Bulk temperature ( C)
1.4
(100, 0.91)
1.2
(115, 0.83)
(105,
1
0.8
(90,
0.6
(Pressure in
bar, mass flow
rate in kg/min)
0.4
(82,
0.2
0.2
0.4
0.6
0.8
1.2
1.4
Figure 7.21 Predicted versus measured pressure drop of CO2 in gas cooler
For the evaporator, two different correlations were compared against the test data,
which were used in the analyses presented in previous chapters and varying degrees of
match were attained. The Yoon et al. [52] correlation showed very good comparison
with the test data following the trend throughout the range of tests carried out with a
maximum difference of 50% between them; however, the modified Wattlet correlation
170
[18] exhibited a distinctly opposite trend against the test data towards higher vapour
quality with increasing difference from the measured data, although it matches very well
up to a vapour quality of 0.75 compared to Yoon et al. correlation as shown in Figure
7.22. Comparison of pressure drop data in evaporator predicted from correlation with
measured data showed that the used correlation underpredicted the measured data by a
maximum of 45%. This may be attributed to similar reasons as was cited for the gas
cooler. We may conclude that our test data has performed reasonably well against some
25
20
15
10
5
0
0.4
0.5
0.6
0.7
0.8
0.9
Vapour quality
Figure 7.22: Boiling heat transfer coefficient of CO2 versus vapour quality
171
accordingly. One such comparison for water mass flow rates of 1.5 kg/min and 1 kg/min,
and water inlet temperatures of 30 oC and 30.5 oC in evaporator and gas cooler,
respectively and evaporator pressure of 40 bar is shown in Figure 7.23. Comparison
between the test results and the model prediction shows a modest agreement with a
maximum deviation of 20% and the trends are fairly similar. Comparison for other
operating parameters also shows fairly similar deviation between the test results and the
model prediction.
4.5
4.3
System COP
4.1
3.9
3.7
3.5
3.3
Experimental
numerical
3.1
2.9
70
80
90
100
110
7.8 Summary
This chapter described the experimental testing of a prototype CO2 heat pump
system for simultaneous water heating and cooling, and presented the experimental
results. An experimental facility was developed and detailed tests were carried out to
study the performance of system and components at different operating conditions.
Validation of heat transfer and pressure drop correlations for both gas cooler and
evaporator, and validation of system simulation model are also presented. The system has
been designed for a cooling capacity of 1 ton (3.56 kW). Due to the presence of very
high-pressure in the system and some of the special characteristics, special attention was
given in designing all the components. The gas cooler pressure was attenuated by
172
simultaneously controlling the total mass of the system and degree of opening of
expansion device.
The system performance for various operating parameters was studied closely.
Uncertainty analyses show fairly acceptable results. Lack of precise control and presence
of measurement errors cause a few data points to lie beyond the uncertainty bar.
Transient study shows that the system takes about 15 minutes to attain steady state. Study
on the system behaviour with different expansion valve opening shows that the valve
opening has significant effect near the full valve closing condition. Performance study
with different compressor discharge pressures shows that both the cooling and heating
outputs increase with discharge pressure due to increase in mass flow rate of refrigerant.
Water outlet temperature increases in gas cooler whereas, it decreases in evaporator with
increase in discharge pressure. Effect of water mass flow rates is not so significant for
both evaporator and gas cooler, whereas the effect of water inlet temperature to gas
cooler on the system performance is significant. Maximum cooling and heating output
from the system have been recorded as 3 kW and 5 kW respectively. Maximum
percentage heat gain/loss in evaporator and gas cooler were 20% and 30% respectively.
Maximum volumetric and isentropic efficiencies of compressor have been recorded as
60% and 50% respectively. The Pitla correlation for gas cooler heat transfer coefficient
showed a reasonably close match with the test results (within 5% error), whereas the
validation of pressure correlation shows that the measured values are significantly higher
(up to 55% deviation) than the model predicted values. For the evaporator, the Yoon et al.
correlation showed good comparison with the test data following the trend throughout the
range of tests carried out with a maximum difference of 50% between them; however, the
modified Wattlet correlation exhibited a distinctly opposite trend against the test data
toward higher vapour quality with increasing difference in coefficient prediction from the
measured data. However, at lower vapour quality, the Wattlet correlation predicted the
test results even better than the Yoon et al. correlation. Comparison of pressure drop data
in the evaporator predicted from correlation with measured data showed that the
correlations underpredicted the measured data by a maximum of 45%. This is attributed
to pipe roughness which is not included in the estimation calculation. Comparison
173
between the test results and the simulation model prediction shows reasonably good
agreement and the trends are fairly similar.
174
Chapter 8
summarized below.
Due to the near critical operation, CO2 exhibits some distinct properties (mostly
favorable to system design) compared to other conventional refrigerants. The effects of
evaporator temperature and gas cooler outlet temperature are more predominant
compared to internal heat exchanger effectiveness at optimized conditions for the cycle.
However, the effect of internal heat exchanger effectiveness is more significant at higher
gas cooler exit temperature and lower evaporator temperature. Analyses for the optimum
condition indicate that a system meant for low or moderate temperature heating is more
economical not only due to high system COP but also due to lower optimum discharge
pressure (low operating pressure ratio) as well. Expressions for optimum cycle
parameters have been developed and these correlations offer useful guidelines for optimal
system design and for selecting appropriate operating conditions. Multi-staging is found
to have more significant effect compared to the use of internal heat exchanger, expansion
turbine and ejector-expander device on the optimum condition.
175
The simulation of CO2 heat pump for simultaneous water cooling and heating
applications shows that the optimum gas cooler to evaporator heat transfer area ratio
ranges between 1.7 and 1.9 for maximum system COP as well as maximum exergetic
efficiency at optimum discharge pressure, although optimum value for the latter is
slightly more. Favourable heat transfer properties of carbon dioxide in both two-phase
and supercritical region and an efficient compression process contribute significantly
toward high system COPs and exergetic efficiency values. A nomogram, applicable for
optimum design, with compressor speed and water inlet temperature as independent
parameters and optimum discharge pressure, optimum area ratio and maximum COP as
output parameters has been presented. This is expected to be of simple help to designers
of such systems. The temperature difference in heat exchangers is found to contribute
more irreversibility than pressure drop in the heat exchangers. The compressor,
evaporator, gas cooler and expansion device contribute to system irreversibility to a
larger extent, while the internal heat exchanger has negligible effect. The expansion valve
contributes a significant amount of exergy loss in case of CO2 systems whereas it is
negligible for a conventional system. Replacement of the expansion valve by an
expansion turbine will increase the COP as well as the exergetic efficiency significantly,
but it will also raise counter-issues related to cost, design and dynamic balancing of the
system.
Exergetic analysis of the gas cooler in a CO2 heat pump shows that for given
operating conditions and capacity, a set of optimum diameter and length is possible for
certain set of tube passes to get minimum total irreversibility associated with thermal,
pressure drop and material. The optimum diameter and length will decrease with increase
in number of passes. For the evaporator, although an optimum diameter has been
obtained, optimum length could not be found as the temperature approach becomes zero
before attaining the optimal length. The effect of material use on the irreversibility is
found to be negligible. Although the effect of pressure drop on the irreversibility can be
neglected for higher diameter, it is quite significant for smaller diameter tubes.
Irreversibility due to pressure drop is higher for the evaporator compared to that in the
gas cooler. Such exergetic optimization exercise is expected to help design the optimal
176
heat exchanger (in terms of diameter, length and number of passes) for a given capacity
and the operating parameters.
The thermodynamic comparison showed that R744 yields better performance than
R134a whereas it performs poorly compared to R22 for a heat pump drying applications.
Experimental validation of CO2 heat pump dryer simulation model, to investigate effect
of different operating parameters on important performance parameters such as COP,
MER and SMER of dryer system, shows that the gas cooler model accurately predicts the
performance whereas the evaporator model overpredicts the performance. The
assumption of an adiabatic model for the compressor and negligible heat interaction with
the ambient could be the reasons behind the slight overprediction of overall system
performance. Simulation results show that unlike bypass air ratio and ambient relative
humidity, the effects of dryer efficiency, recirculation air ratio, ambient temperature and
air mass flow rate are significant on system behavior. The performance of the dryer
system increases linearly with increase in dryer efficiency.
increases with ambient temperature, COP deteriorates, whereas both decrease with
ambient relative humidity. Both bypass air ratio and recirculation air ratio yield some
optimum values for maximum SMER, but the COP decreases in both cases. An
optimization of air mass flow rate is possible to maximize both COP and SMER although
optimum values for maximum COP and SMER are not necessarily identical.
Experimental studies show that the gas cooler pressure can be successfully
controlled by simultaneously controlling the total mass of the system and degree of
opening of expansion device. Repeatability and uncertainty analyses have been presented
and they exhibit reasonably acceptable trends. Transient study shows that the system
takes about 15 minutes to attain steady state. Study on the system behavior with different
degree of expansion valve opening shows that it has very significant effect near the fully
closed condition. Performance study with different compressor discharge pressures shows
that both cooling and heating outputs increase with discharge pressure due to increase in
mass flow rate of refrigerant. Effect of water mass flow rates is not so significant for both
evaporator and gas cooler, whereas the effect of water inlet temperature to gas cooler on
the system performance is significant. Maximum cooling and heating output from the
177
Also
investigations on heating or cooling systems with various mixtures of CO2 can be done
with a view to reduce the system pressure and take advantage of superior heat transfer
properties of CO2. Although various theoretical and experimental investigations on
supercritical heat transfer and pressure drop, boiling heat transfer and pressure drop, twophase flow have been done, pseudocritical region of gas cooling is still one of the
interesting areas, where one can rigorously study heat transfer and fluid flow. Relatively
low isentropic efficiency of developed CO2 compressors show that still there is a need to
improve the design to get higher performance. Research on compressor is required not
178
only for improvement of performance, but also to reduce the weight and cost.
Transcritical CO2 heat pump has great potential in process heat industries. Prototype
developments and experiment CO2 heat pump for dairy and drying applications in food
industry are some of the very promising applications that can be pursued.
179
Appendix A
(ii)
(iii)
(iv)
(v)
been used, whose range is 0-160 bar with an accuracy of 1.5% of full scale reading;
i.e. a maximum error of 2.4 bar can occur. For low pressures, locally procured pressure
gauges were used, with a range of 0-70 bar with much less accuracy. A Coriolis mass
flow meter (Rheonik brand of Rockwin Flowmeter) was installed for refrigerant mass
flow measurement with a range of 0.210 kg/min with fairly high accuracy (0.2% of
span) i.e. a maximum error of 0.02 kg/min can occur. For water side mass flow meters,
due to budgetary restrictions inexpensive models were used having lower accuracy. For
all pressure gauges, mass flow meters and energy meter, calibration charts supplied by
manufactures have been used.
For calibration of thermocouple a temperature controlled thermostatic bath was
used. The correction factors have been evaluated for different temperatures between 0 to
100 oC and plotted as shown in Figures A1 and A2 for two types of thermocouples used
in experiment. Calibration chart shows that the variation of correction factor is not linear.
Temperatures can be evaluated by using calibration chart or correlation of correction
factor fitted by calibration data. It may be noted that temperatures above 100 oC have
been evaluated by extrapolating the curve. The corrected temperature is evaluated by:
corrected temperature = measured temperature + correction factor
180
2
1.6
Correction factor (oC)
1.2
0.8
0.4
0
-0.4
-0.8
-1.2
-1.6
-2
0
10 20 30 40 50 60 70 80 90 100
Temperatre (oC)
6
5
Correction factor (oC)
4
3
2
1
0
-1
-2
-3
-4
0
10 20 30 40 50 60 70 80 90 100
Temperatre (oC)
181
Appendix B
Uncertainty Analysis
Functional form of the various performance parameters are given by
v = f ( mr , P1 , T1 )
is = f ( P1 , P2 , T1 , T2 )
Qevw = f ( mevw , Tevw,i , Tevw,o )
Qgcw = f ( mgcw , Tgcw,o , Tgcw,i )
COPsys = f ( mr , P1 , T1 , P2 , T2 , .................)
r = f ( mr , mw , Tr i , Tr i +1 , Twi , Twi +1 , d , D )
where Tr i , Tr i +1 , Twi , Twi +1 are the inlet and outlet temperatures or both water and
refrigerant in measured section.
Uncertainty analyses have been performed using the square root formula [128].
i.e. if y = f ( x1n1 , x2 n 2 , x3n 3 ,..................) , then
2
x x x
y
= n1 1 + n2 2 + n3 3 + ...............
y
x2
x3
x1
The multiplying factors (ni) have been taken as 1 for all measuring parameter except for
diameter that is 2.
182
REFERENCES
1.
2.
3.
4.
William S, Bodinus PE. The rise and fall of carbon dioxide systems, ASHRAE
Journal, 1999; 37-42.
5.
Kim MH., Pettersen J, Bullard CW. Fundamental process and system design issues
in CO2 vapor compression systems, Prog Energy Comb Sc, 2004; 30(2); 119-174.
6.
Riffat SB, Afonso CF, Aliveira AC, Reay DA. Natural refrigerants for refrigeration
and air conditioning systems, App Thermal Engg, 1997, 17(1); 33-42.
7.
Pearson S. F. Natural refrigerants for heat pumps, IEA heat pump center newsletter
article, 22 (1/2004), 2004.
8.
Casini D. CO2 compressors and equipment, use and availability, EURO cooling and
heating, Centro Studi Galileo; www.eurocooling.com.
9.
183
13. Vaisman IB. COP analysis of carbon dioxide cycles, ASHRAE Trans, 2002; 108;
252-262.
14.
15.
22. Elbel S, Hrnjak P. Flash gas bypass for improving the performance of transcritical
R744 systems that use microchannel evaporators, Int J Refrigeration, 2004; 27(7);
724-735.
184
185
33. Pitla SS, Groll EA, Ramadhyani S. Convective heat transfer from in-tube flow of
turbulent supercritical carbon dioxide Part 2 Experimental data and numerical
prediction, HVAC & R Research, 2001; 7(4); 367-382.
34. Fang, X, Bullard, CW, Hrnjak BS. Heat Transfer and Pressure drop of gas coolers,
ASHRAE Trans, 2001; 107; 255-266.
35. Wang J, Hihara E. Study on carbon dioxide gas cooler heat transfer process under
supercritical pressures, Int J Energy Research, 2002; 26(14); 1237-1251.
36. Liao SM, Zhao TS. Measurement of heat transfer coefficients from supercritical
carbon dioxide flowing in horizontal mini/micro channels, ASME J Heat Transfer,
2002; 124(1); 413-420.
37. Jiang PX, Xu YJ, Lv J, Shi RF, He S, Jackson JD. Experimental investigation of
convective heat transfer of CO2 at super-critical pressures in vertical mini-tubes and
in porous media, App Thermal Engg, 2004; 24(8-9); 1255-1270.
38. Pitla SS, Groll EA, Ramadhyani S. New correlation to predict the heat transfer
coefficient during in-tube cooling of turbulent supercritical CO2; Int J Refrigeration,
2002; 25(7); 887-895.
39. Scalabrin G, Piazza L, Condosta M. Convective cooling of supercritical carbon
dioxide inside tubes: heat transfer analysis through neural networks, Int J Heat Mass
Transfer, 2003; 46(23); 4413-4425.
40. Yoon SH, Kim JH, Hwang YW, Kim MS, Min K, Kim Y. Heat transfer and
pressure drop characteristics during the in-tube cooling process of carbon dioxide in
the supercritical region, Int J Refrigeration, 2003; 26(8); 857-864.
41. Zhao Y, Ohadi MM. Experimental study of supercritical CO2 gas cooling in a
microchannel gas cooler, ASHRAE Trans, 2004; 110; 291-300.
42. Dang C, Hihara E. In-tube cooling heat transfer of supercritical carbon dioxide. Part
1. Experimental measurement, Int J Refrigeration, 2004; 27(7); 736-747.
43. Dang C, Hihara E. In-tube cooling heat transfer of supercritical carbon dioxide. Part
2. Comparison of numerical calculation with different turbulence models, Int J
Refrigeration, 2004; 27(7); 748-760.
186
44. Bredesen AM, Hafner A, Pettersen J, Neksa P, Aflekt K. Heat transfer and pressure
drop for in-tube Evaporation of CO2, Proceedings International Conference on Heat
Transfer Issues in Natural Refrigerants, Nov. 6-7, 1997, College Park, pp. 1-15.
45. Zhao Y, Molki M, Ohadi MM, Dessiatoun SV. Flow boiling of CO2 in
microchannels, ASHRAE Trans, 2000; 106; 437-446.
46. Zhao Y, Molki M, Ohadi MM, Franca FHR, Radermacher R. Flow boiling of CO2
with miscible oil in microchannels, ASHRAE Trans, 2001; 107; 135-143.
47. Yun R, Kim Y, Kim MS, Choi Y. Boiling heat transfer and dryout phenomenon of
CO2 in a horizontal smooth tube, Int J Heat Mass Transfer, 2003; 46(13); 23532361.
48. Gungor, KE, Winterton RHS. A general correlation for flow boiling in tubes and
annuli, Int J Heat Mass Transfer, 1986; 29(3); 351-358.
49. Yun R, Kim Y. Critical quality prediction for saturated flow boiling of CO2 in
horizontal small diameter tubes, Int J Heat Mass Transfer, 2003; 46(14); 25272535.
50. Pettersen J. Flow vaporization of CO2 in microchannel tubes, Exp Thermal Fluid
Science, 2004; 28(2-3); 111-121.
51. Pettersen J. Two-Phase Flow Patterns in Microchannel Vaporization of CO2 at
Near-Critical Pressure, Heat Transfer Engg, 2004; 25(3); 5260.
52. Yoon SH, Cho ES, Hwang YW, Kim MS, Min K, Kim Y. Characteristics of
evaporative heat transfer and pressure drop of carbon dioxide and correlation
development, Int J Refrigeration, 2004; 27(2); 111-119.
53. Thome JR, Hajal JE. Flow boiling heat transfer to carbon dioxide: general
prediction method, Int J Refrigeration, 2004; 27(3); 294-301.
54. Huai X, Koyama S, Zhao TS, Shinmura E, Hidehiko K, Masaki M. An
experimental study of flow boiling characteristics of carbon dioxide in multiport
mini channels, App Thermal Engg, 2004; 24(10); 1443-1463.
187
188
68. Boewe D, Yin J, Park Y, Bullard C, Hrnjak P. The role of suction line heat
exchanger in transcritical R744 mobile A/C systems, SAE Paper No. 1999-01-0583;
1999.
69. Kim MH, Bullard CW. Development of microchannel evaporator model for a CO2
air-conditioning system, Energy, 2001; 26(10); 931-948.
70. Ikeda H, Yagi J, Kawaguchi Y. Evaluation of various compressor lubricants for a
carbon dioxide heat pump system, 6th IIR Gustav Lorentzen Natural Working Fluid
Conference, Glasgow UK; 2004.
71. Leisenheimer B, Fritz T. Interaction between CO2 and elastomers with respect to
permeation and explosive decompression, The Fourth IIR-Gustav Lorentzen
Conference on Natural Working Fluids, West Lafayette, IN. 2000; 201208.
72. Pettersen J. An efficient new automobile air-conditioning system based on CO2
vapour compression, ASHRAE Trans, 1994; 100; 657-665.
73. Bhatti M. A critical look at R-744 and R-134a mobile air conditioning systems,
SAE paper No. 970527; 1997.
74. McEnaney R, Park Y, Yin J, Bullard C, Hrnjak P. Performance of the prototype of a
transcritical R744 mobile A/C system. SAE Paper No. 99PC-6-7; 1999.
75. McEnaney RP, Hrnjak PS. Control strategies for transcritical R744 systems, SAE,
Paper No. 2000-0-1272, 2000; 1711-1720.
76. Brown JS, Yana-Motta SF, Domanski PA. Comparative analysis of an automotive
air conditioning systems operating with CO2 and R134a, International Journal of
refrigeration, 2002; 25 (1); 19-32.
189
77. Liu H, Chen J, Chen Z. Experimental evaluation of the optimal refrigerant mass
charge in a CO2 automotive air conditioner, 6th IIR Gustav Lorentzen Natural
Working Fluid Conference, Glasgow UK; 2004; paper 7/A/2.20.
78. Giannavola M, Murphy R, Yin J, Kim MH, Bullard C, Hrnjak P. Experimental
investigations of an automotive heat pump prototype for military, SUV and
compact cars, The Fourth IIR-Gustav Lorentzen Conference on Natural Working
Fluids, West Lafayette, IN. 2000; 115122.
79. Rozhentsev A, Wang C-C. Some design features of a CO2 air conditioner, Applied
Thermal Engineering, 2001; 21(8); 871-880.
80. Brown JS, Kim Y, Domanski PA. Evaluation of carbon dioxide as R-22 substitute
for residential air-conditioning, ASHRAE Transactions, 2002; 108(2); 954-963.
81. Rieberer R, Halozan H. CO2 heat pumps in controlled ventilation systems.
Proceedings of the Third IIR-Gustav Lorentzen Conference on Natural Working
Fluids, Oslo, Norway; 1998; 212222.
82. Beaver A, Yin J, Bullard CW, Hrnjak PS. Experimental and model study of the heat
pump/air conditioning systems based on transcritical cycle with R744, 20th
International Congress of Refrigeration, IIR/IIF, Sydney, 1999; Paper 434.
83. Neks P. CO2 heat pump systems, Int J Refrigeration, 2002; 25(4); 421-427.
84. Richter MR, Song SM, Yin JM, Kim MH, Bullard CW, Hrnjak PS. Experimental
results of transcritical CO2 heat pump for residential application, Energy, 2003;
28(10); 1005-1019.
85. Jakobsen A, Skaugen G, Skiple TV, Neksa P, Andresen T. Development and
evaluation of a reversible CO2 residential air conditioning system compared to a
state-of-the art R410A unit, 6th IIR Gustav Lorentzen Natural Working Fluid
Conference, Glasgow UK; 2004; paper 5/A/10.10.
86. Neksa P, Rekstad H, Zakeri GR, Schiefloe PA. CO2 heat pump water heater:
characteristics, system design & experimental results, Int J Refrigeration, 1998;
21(3); 172-179.
190
191
98. Adriansyah W. Combined air conditioning and tap water heating plant using CO2 as
refrigerant, Energy and Buildings, 2004; 36(7); 690-695.
99. Schmidt EL, Klcker K, Flacke N, Steimle F. Applying the transcritical CO2
process to a drying heat pump, Int J Refrigeration, 1998; 21(3); 202-211.
100. Klcker K, Schmidt EL, Steimle F. Carbon dioxide as a working fluid in drying
heat pumps: Int J Refrigeration, 2001; 24(1); 100-107.
101. Span R, Wagner W. A new equation of state for Carbon dioxide covering the fluid
region from triple point temperature to 1100 K at pressure up to 800 MPa, J. Phys.
Chem. Ref. Data 1996; 25; 1509-1596.
102. Vesovic V., Wakeham W.A., Olchowy G.A., Sengers J.V., Watson J.T.R, Millat J.
The transport properties of carbon dioxide, J. Phys. Chem. Ref. Data, 1990; 19;
763-808.
103. Fenghour A, Wakeham WA, Vesovic V. The viscosity of carbon dioxide, J. Phys.
refrigerant in air-to-air air conditioners and heat pumps: system modelling and
analysis, ARTI final report, no. 21CR/610-10030, 2003.
107. Kakac S, Liu H. Heat Exchangers: Selection, rating and thermal design, CRC Press,
Landon; 2002.
108. Aprea C, de Rossi F, Greco A, Renno C. Refrigeration plant exergetic analysis
varying the compressor capacity. Int J Energy Research 2003; 27: 653-669.
109. Stegou-Sagia A, Paignigiannis N. Exergy losses in refrigeration systems. A study
192
111. Yilmaz M, Sara ON, Karsli S. Performance evaluation criteria for heat exchanger
193
124. Shiming D. A dynamic mathematical model of a direct expansion (DX) watercooled air-conditioning plant, Building & Environment, 2000, 35: 603-613.
125. Prasertsan S, Saen-Saby P, Ngamsritrakul P, Prateepchaikul G. Heat pump dryer
Part 2: Result of the simulation, Int J Energy Research, 1997; 21(1): 1-20.
126. Jia X., Jolly P, Claments S. Heat pump assisted continuous drying - Part 2:
Simulation results, Int J Energy Research, 1990; 14: 771-782.
127. Clements S, Jia X, Jolly P. Experimental verification of a heat pump assisted
continuous dryer simulation model, Int J Energy Research, 1993; 17: 19-28.
128. Beckwith TG, Buck NL, Marangoni RD. Mechanical measurement, 3rd Ed, Narosa
Publishing House, New Delhi, 1994.
194
Sarkar J, Bhattacharyya S, Ramgopal M. Optimization of a transcritical CO2 heat pump cycle for simultaneous
cooling and heating applications. International Journal of Refrigeration 2004; 27(8): 830-838.
2.
Sarkar J, Bhattacharyya S, Ramgopal M. Transcritical CO2 heat pump systems: Exergy analysis including heat
transfer and fluid flow effects, Energy Conversion and Management, 2005; 46(13-14): 2053-2067.
3.
Sarkar J, Bhattacharyya S, Ramgopal M. Simulation of a transcritical CO2 heat pump cycle for simultaneous cooling
and heating applications, International Journal of Refrigeration, 2006; 29(5): 735-743.
4.
Sarkar J, Bhattacharyya S, Ramgopal M. CO2 heat pump dryer: Part 1. Mathematical model and simulation, Drying
Technology, 2006; 24(12): 1583-1591.
5.
Sarkar J, Bhattacharyya S, Ramgopal M. CO2 heat pump dryer: Part 2. Validation and simulation results, Drying
Technology, 2006; 24(12): 1593-1600.
6.
Sarkar J, Bhattacharyya S, Ramgopal M. Irreversibility minimization of heat exchangers for transcritical CO2 systems,
International Journal of Thermal Sciences, 2009; 48(1): 146-153.
7.
Sarkar J, Bhattacharyya S, Ramgopal M. A transcritical CO2 heat pump for simultaneous water cooling and heating:
Test results and model validation, International Journal of Energy Research, 2009; 33(1): 100-109.
8.
Sarkar J. Transcritical CO2 heat pump simulation model and validation for simultaneous cooling and heating,
International Journal of Mathematical, Physical and Engineering Sciences, 2009; 3(4): 199-204.
9.
Sarkar J, Bhattacharyya S, Ramgopal M. Performance of a transcritical CO2 heat pump for simultaneous water
cooling and heating, International J Applied Science, Engineering and Technology, 2010; 6(1): 57-63.
10. Sarkar J, Bhattacharyya S, Ramgopal M. Experimental investigation of transcritical CO2 heat pump for simultaneous
water cooling and heating, Thermal Science, 2010; 14(1): 57-64.
Conference Papers published from Thesis
1.
Sarkar J, Bhattacharyya S, Ram Gopal M. Carbon dioxide based heat pump dryers in food industry, Int conf Emerging
technologies in Agri food engg, IIT Kharagpur, Dec 14-17, 2004.
2.
Sarkar J, Bhattacharyya S, Ramgopal M, Gautam S. Comparison and validation of heat transfer correlations for intube cooling of supercritical CO2, 18th national & 7th ISHMT-ASME heat & mass transfer conf, IIT Guwahati, Jan 4-6,
2006.
3.
Sarkar J, Bhattacharyya S, Ramgopal M. Transcritical CO2 heat pump prototype development for simultaneous water
cooling and heating, 22nd IIR Int Congress of Refrigeration, Beijing, China, 2007; ICR07-E2-548.
4.
Sarkar J, Bhattacharyya S, Ramgopal M. Pressure drop for in-tube supercritical co2 cooling: comparison of
correlations and validation, 19th national & 8th ISHMT-ASME heat & mass transfer conf, JNTU India, Jan 3-5, 2008.
5.
Sarkar J, Bhattacharyya S, Ramgopal M. Comparison and validation of in-tube CO2 boiling heat transfer correlations
(Paper 58), 20th national & 9th ISHMT-ASME heat & mass transfer conf, IIT Bombay, India, Jan 4-6, 2010.
195