Hybrid Trigeneration - Spiral
Hybrid Trigeneration
Thermally Activated Heat Pump Technologies
Luke Bannar-Martin
Design Engineering Group
Department of Mechanical Engineering
Imperial College London
A Thesis submitted to Imperial College London in partial fulfillment for the degree of
Doctor of Philosophy and Diploma of Imperial College
September 13, 2015
Declaration
I declare that the work reported in this Thesis was conducted by me at Imperial
College London and has not been submitted in part, or in whole, for any other
qualifications.
The copyright of this thesis rests with the author and is made available under
a Creative Commons Attribution Non-Commercial No Directives licence. Researchers are free to copy, distribute or transmit the thesis on the condition that
they attribute it, that they do not use it for commercial purposes and that they do
not alter, transform or build upon it. For any reuse or redistribution, researchers
must make clear to others the licence terms of this work.
i
Acknowledgements
I wish to record my thanks to Professor Peter R.N. Childs, Head of the School
of Design Engineering of Imperial College London, for his constant enthusiasm
and guidance as my supervisor. I also wish to express my sincere thanks to the
Royal Commission for the Exhibition of 1851 for their unwavering support of
this research. I would also like to extend my thanks the Climate Knowledge
and Innovation Community for their support of the research and the excellent
opportunities they afforded to me. Finally, I would like to thank all those who
have helped along the way; there are simply too many names to list.
ii
Abstract
This Thesis describes a theoretical approach to the design and analysis of concepts in the field of trigeneration, which is the combined generation of electricity,
hot water and chilled water from a single prime mover. The continuously increasing demand for cooling in modern buildings and cities due to economic expansion,
which is further compounded by the effects of climate change, means that there
is a tremendous opportunity for new concepts in the fields of trigeneration, refrigeration and heat pump cycles and technologies. It is this opportunity which
forms the context and primary aim for the research presented within this Thesis,
which is to identify high performance heat pump cycle concepts.
A thorough review of trigeneration systems, controls and operational strategies is presented, which demonstrates their energetical benefits in terms of Primary Energy Consumption (PEC) and Energy Utilisation Factor (EUF). This is
followed by a review of the state of the art in absorption chiller cycles, adsorption chiller cycles and R744 (carbon dioxide) Mechanical Vapour Compression
(MVC) heat pump cycles. To fulfil the main aim of this research, the focus of
the review is in multi - stage / multi - effect absorption and adsorption chillers, hybrid compression - absorption chillers and high temperature R744 heat pump and
supercritical Brayton cycles.
A number of novel absorption heat pump cycle concepts are modeled and analysed, the most advanced of which being: a double - stage / triple - effect ammonia water absorption heat pump cycle, and a triple - stage / triple - effect lithium bromide - water compression - absorption heat pump cycle. Under certain operating
iii
Hybrid Trigeneration - Thermally Activated Heat Pumps
parameters, these cycles are capable of achieving a Coefficient Of Performance
(COP) in cooling of upto 2.0. However, they are not believed to represent the
best opportunity for further research due to: the inability for simultaneous heat
and coolth production; potential corrosion problems resulting from the high temperatures; the extremely high volumetric flowrate of steam required through compressors; the significant body of research which has explored almost all avenues
for innovations in absorption heat pump cycles.
A novel thermally activated transcritical R744 heat pump cycle is developed,
which combines the principles of both the Brayton cycle and reverse Rankine
cycle. The cycle is designed to utilise relatively hot gas turbine exhaust (∼ 500◦ C)
gases to produce low temperature hot water and chilled water for domestic hot
water and space cooling, respectively. The cycle itself achieves a COP of 1.58
and 2.41 in cooling and heating respectively with a maximum inlet temperature
of 350 ◦ C. A Coefficient of Performance in cooling of 1.58 is comparable to a
triple - stage / triple - effect lithium bromide - water absorption heat pump cycle.
The principal advantage of the novel cycle is that it simultaneously produces hot
water (∼ 65◦ C), while absorption chillers are not capable of producing heat and
coolth simultaneously.
Certain components of the novel R744 heat pump cycle are analysed in detail,
including: R744 heat exchangers, R744 two - phase ejector, R744 compressor and
work expanders. A design and calculation methodology is presented with the purpose of forming the continuation of this research from a Technology Readiness
Level1 of 2 (technology concept and / or application formulated) to a Technology Readiness Level of 3 (analytical and experimental critical function and / or
characteristic proof - of - concept).
1
Technology Readiness Level is a widely used concept to estimate the maturity of specific
technologies
iv
Chapter 0
Acronyms
CCGT Combined Cycle Gas Turbine.
CCHP Combined Cooling, Heating & Power.
CCP Combined Cooling & Power.
CFC ChloroFluoroCarbon.
CHP Combined Heat & Power.
COP Coefficient Of Performance.
CSP Concentrated Solar Power.
DHW Domestic Hot Water.
DSG Direct Steam Generation.
EES Engineering Equation Solver.
EHR Exhaust Heat Recovery.
EHRSG Exhaust Heat Recovery Steam Generator.
ELF Electric Load Following.
EUF Energy Utilisation Factor.
FESR Fuel Energy Savings Ratio.
GAX Generator - Absorber Exchange.
v
Hybrid Trigeneration - Thermally Activated Heat Pumps
GAXAC Generator Absorber Exchange Absorption Compression.
GWP Global Warming Potential.
HC HydroCarbon.
HCFC HydroChloroFluoroCarbon.
HFC HydroFluoroCarbon.
HLF Hybrid Load Following.
HX Heat Exchanger.
LTHW Low Temperature Hot Water.
MTHW Medium Temperature Hot Water.
MVC Mechanical Vapour Compression.
ODP Ozone Depletion Potential.
ORC Organic Rankine Cycle.
PEC Primary Energy Consumption.
PGU Primary Generation Unit.
S-CO2 Supercritical Carbon Dioxide.
SCP Specific Cooling Production.
TLF Thermal Load Following.
VLTHW Very Low Temperature Hot Water.
VRA Vapour Recompression Absorber.
vi
Chapter 0
Subscripts
c
Cooling.
h
Heating.
m
Mechanical.
P EC
th
Primary Energy Consumption.
Thermal.
vii
Hybrid Trigeneration - Thermally Activated Heat Pumps
viii
Chapter 0
Refrigerants
CH3 OH - LiBr Methanol & Lithium Bromide.
CH3 OH - LiCl Methanol & Lithium Chloride.
CH3 OH - TEGDME Methanol / TetraEthyleneGlycol-DiMethylEther.
H2 O-LiBr:LiI:LiNO3 -LiCl Water / Lithium Bromide : Lithium Iodide : Lithium
Nitrate : Lithiurm Cloride (5 : 1 : 1 : 2).
H2 O - NH3 Water & Ammonia.
LiBr - H2 O Lithium Bromide Water.
R115 Chloropentafluoroethane.
R12 Dichlorodifluoromethane.
R125 Pentafluoroethane.
R134a 1,1,1,2-Tetrafluoroethane.
R143a 1,1,1-Trifluoroethane.
R22 Difluoromonochloromethane.
R22 - TEGDME R22 / TetraEthyleneGlycol-DiMethylEther.
R290a Propane.
R32 Difluoromethane.
ix
Hybrid Trigeneration - Thermally Activated Heat Pumps
R407c 23 % R32 / 25 % R125 / 52 % R134a.
R410a 50 % R125 / 50 % R32.
R502 51.2 % R115 / 48.8 % R22.
R507 50 % R125 / 50 % R143a.
R600a Isobutane.
R717 Ammonia.
R744 Carbon Dioxide.
R744a Nitrous Oxide.
TFE - TEGDME TriFluoroEthanol / TetraEthyleneGlycol-DiMethylEther.
x
Chapter 0
Symbols
).
N u = Nusselt Number ( hL
k
P r = Prandtl Number ( αν ).
).
Re = Reynolds Number ( vL
ν
Ω =
Q̇generator
Ẇpump
(i.e the ratio of external heat input to the generator and external
mechanical input to the solution pump).
Φ =
Q̇solution,recuperation
Q̇rectif ier,recuperation
(i.e. the ratio of recuperated heat from the solution heat
exchanger and the recuperated heat from the rectifier (relevant to H2 O NH3 absorption heat pumps)).
Ψ =
ṁref rigerant produced
ṁstrong solution
(i.e. the ratio of mass flow rate of refrigerant in the ab-
sorption chiller and the mass flow rate of solution exiting the absorber and
returning to the generator).
Q
gen.,in
(i.e. the proportion generator heat input which
Θ = 1− Qgen.,in +Qsolution
hx +Qrect.,out
is internally recuperated (relevant to H2 O - NH3 absorption heat pumps)).
f = Friction factor.
x =
mref rigerant
msolution
(i.e. the proportion of the fluid which passes through the con-
denser and evaporator in an ammonia - water absorption heat pump which
is the refrigerant, ammonia).
xi
Contents
1 Review of Trigeneration Systems
2
1.1
Context . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3
1.2
Trigeneration Systems . . . . . . . . . . . . . . . . . . . . . . . .
3
1.2.1
Trigeneration Subsystems . . . . . . . . . . . . . . . . . .
4
1.2.2
Trigeneration System Design . . . . . . . . . . . . . . . . .
12
1.2.3
Trigeneration Control Strategies . . . . . . . . . . . . . . .
15
1.3
Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
18
1.4
Aims and Objectives of Thesis . . . . . . . . . . . . . . . . . . . .
19
2 Review of Heat Pump Systems
21
2.1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
22
2.2
Refrigerants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
22
2.2.1
Single Component Refrigerants . . . . . . . . . . . . . . .
22
2.2.2
Multi Component Refrigerants . . . . . . . . . . . . . . . .
26
Absorption Heat Pumps . . . . . . . . . . . . . . . . . . . . . . .
28
2.3.1
Multi - Stage / Multi - Effect
. . . . . . . . . . . . . . . . .
29
2.3.2
Generator - Absorber Exchange Cycles . . . . . . . . . . .
32
2.3.3
Vapour Recompression Absorption Cycles . . . . . . . . .
35
Adsorption Pumps . . . . . . . . . . . . . . . . . . . . . . . . . .
35
2.4.1
Multi - bed adsorption chiller . . . . . . . . . . . . . . . . .
36
2.4.2
Multi - stage adsorption chiller . . . . . . . . . . . . . . . .
37
R744 MVC Heat Pumps . . . . . . . . . . . . . . . . . . . . . . .
37
2.3
2.4
2.5
xii
Hybrid Trigeneration - Thermally Activated Heat Pumps
2.6
Supercritical Carbon Dioxide Brayton Cycle . . . . . . . . . . . .
40
2.7
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
43
3 Absorption Heat Pump Cycle Concepts
3.1
46
Absorption Heat Pump Cycle Concepts . . . . . . . . . . . . . . .
47
3.1.1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . .
47
3.1.2
H2 O - NH3 Kangaroo Cycle . . . . . . . . . . . . . . . . . .
47
3.1.3
LiBr - H2 O Compression - Absorption Cycle . . . . . . . . .
61
3.1.4
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . .
74
4 R744 Heat Pump Cycle Concepts
76
4.1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
77
4.2
Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
78
4.2.1
Assumptions . . . . . . . . . . . . . . . . . . . . . . . . . .
79
Cycle 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
80
4.3.1
Description . . . . . . . . . . . . . . . . . . . . . . . . . .
80
4.3.2
Internal Heat Recuperation Heat Exchanger . . . . . . . .
82
4.3.3
Intercooling Pressure . . . . . . . . . . . . . . . . . . . . .
83
4.3.4
Results and Discussion . . . . . . . . . . . . . . . . . . . .
83
Cycle 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
89
4.4.1
Description . . . . . . . . . . . . . . . . . . . . . . . . . .
89
4.4.2
Results and Discussion . . . . . . . . . . . . . . . . . . . .
94
Cycle 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
97
4.5.1
Description . . . . . . . . . . . . . . . . . . . . . . . . . .
97
4.5.2
Results and Discussion . . . . . . . . . . . . . . . . . . . . 100
4.3
4.4
4.5
4.6
Cycle Selection and Discussion . . . . . . . . . . . . . . . . . . . . 102
4.7
Sensitivity Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 104
4.8
Comparison to a ‘Separate Cycle System’ . . . . . . . . . . . . . . 106
4.9
Integration into a Trigeneration System . . . . . . . . . . . . . . . 109
4.10 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
Chapter 0
xiii
Hybrid Trigeneration - Thermally Activated Heat Pumps
5 R744 Heat Pump Cycle Preliminary Component Analysis
115
5.1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
5.2
Heat Exchanger Modeling . . . . . . . . . . . . . . . . . . . . . . 116
5.3
5.4
5.5
5.2.1
Hot Water Generator . . . . . . . . . . . . . . . . . . . . . 121
5.2.2
Internal Heat Recuperator . . . . . . . . . . . . . . . . . . 124
Ejector Modelling . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
5.3.1
Review of previous work . . . . . . . . . . . . . . . . . . . 125
5.3.2
Ejector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Compressor / Work Expander Modeling . . . . . . . . . . . . . . . 131
5.4.1
Centrifugal Compressor . . . . . . . . . . . . . . . . . . . . 131
5.4.2
Gerotor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 144
6 Overall Conclusions
146
A R744 Heat Pump Cycle Early Concepts
165
B R744 Centrifugal Compressor Impeller Blade Geometry
169
xiv
Chapter 0
List of Figures
1.1
Relevance of demand side variability components to systems design
and control strategy; and the correlation of each to the amplitude,
frequency and predictability of the demand . . . . . . . . . . . . .
16
2.1
Classification of Heat Pump / Refrigeration System . . . . . . . .
22
2.2
Basic process flow diagram of a typical single-stage absorption chiller 29
2.3
Dühring plot - Kangaroo cycle [56] . . . . . . . . . . . . . . . . .
32
2.4
Basic Generator - Absorber Exchange (GAX) cycle . . . . . . . . .
33
2.5
Basic GAX cycle[59] . . . . . . . . . . . . . . . . . . . . . . . . .
34
2.6
Temperature - Entropy plot of a basic transcritical R744 heat pump
cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
38
2.7
Cp as a function of temperature for R744 at various pressures
39
3.1
1 - Stage / 1 - Effect H2 O - NH3 V.1 absorption heat pump cycle pro-
. .
cess flow diagram . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2
1 - stage / 1 - effect H2 O - NH3 V.2 absorption heat pump cycle process flow diagram . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.3
54
2 - stage / 3 - effect H2 O - NH3 V.2 absorption heat pump cycle process flow diagram . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.5
52
2 - stage / 3 - effect H2 O - NH3 V.1 absorption heat pump cycle process flow diagram . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.4
51
56
3 - stage / 5 - effect H2 O - NH3 V.1 absorption heat pump cycle process flow diagram . . . . . . . . . . . . . . . . . . . . . . . . . . .
xv
59
Hybrid Trigeneration - Thermally Activated Heat Pumps
3.6
3 - stage / 5 - effect H2 O - NH3 V.2 absorption heat pump cycle process flow diagram . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.7
1 - Stage / 1 - Effect LiBr - H2 O compression - absorption heat pump
cycle process flow diagram . . . . . . . . . . . . . . . . . . . . . .
3.8
64
1 - Stage / 1 - Effect LiBr - H2 O compression - absorption heat pump
cycle: COPc,th vs. rp . . . . . . . . . . . . . . . . . . . . . . . . .
3.9
60
64
1 - Stage / 1 - Effect LiBr - H2 O compression - absorption heat pump
cycle: COPc,th vs. Tgen . . . . . . . . . . . . . . . . . . . . . . . .
65
3.10 1 - Stage / 1 - Effect LiBr - H2 O compression - absorption heat pump
cycle: COPc,m vs. rp . . . . . . . . . . . . . . . . . . . . . . . . .
66
3.11 2 - Stage / 2 - Effect LiBr - H2 O compression - absorption heat pump
cycle process flow diagram . . . . . . . . . . . . . . . . . . . . . .
67
3.12 2 - Stage / 2 - Effect LiBr - H2 O compression - absorption heat pump
cycle: COPc,th vs. rp . . . . . . . . . . . . . . . . . . . . . . . . .
68
3.13 2 - Stage / 2 - Effect LiBr - H2 O compression - absorption heat pump
cycle: COPc,th vs. Tgen . . . . . . . . . . . . . . . . . . . . . . . .
69
3.14 3 - Stage / 3 - Effect LiBr - H2 O compression - absorption heat pump
cycle process flow diagram . . . . . . . . . . . . . . . . . . . . . .
70
3.15 3 - Stage / 3 - Effect LiBr - H2 O compression - absorption heat pump
cycle: COPc,th vs. rp . . . . . . . . . . . . . . . . . . . . . . . . .
71
3.16 3 - Stage / 3 - Effect LiBr - H2 O compression - absorption heat pump
cycle: COPc,th vs. Tgen . . . . . . . . . . . . . . . . . . . . . . . .
72
3.17 3 - Stage / 3 - Effect LiBr - H2 O compression - absorption heat pump
cycle: COPc,m vs. Tgen . . . . . . . . . . . . . . . . . . . . . . . .
73
4.1
Temperature - Entropy plot of R744 heat pump cycle 1 . . . . . .
82
4.2
Process flow diagram of R744 heat pump cycle 1 . . . . . . . . . .
82
4.3
Parametric analysis of the optimum intercooling pressure . . . . .
84
4.4
Plots of COP against ambient temperature for thermal and mechanical inputs for Cycle 1 . . . . . . . . . . . . . . . . . . . . . .
xvi
86
Chapter 0
Hybrid Trigeneration - Thermally Activated Heat Pumps
4.5
A plot of COPc against evaporation temperature for Cycle 1 . . .
87
4.6
A plot of COPh against evaporation temperature for Cycle 1 . . .
88
4.7
A plot of T7 and T2 against evaporation temperature for Cycle 1 .
89
4.8
Process flow diagram of R744 heat pump Cycle 2 . . . . . . . . .
91
4.9
Temperature - entropy plot of R744 heat pump Cycle 2 . . . . . .
92
4.10 A plot of COPC and COPh against first stage expander inlet temperature for Cycle 2 . . . . . . . . . . . . . . . . . . . . . . . . . .
95
4.11 A plot of fluid pressure at state point 1 (first stage expander outlet)
against temperature at state point 13 (first stage expander inlet
temperature) for Cycle 2 . . . . . . . . . . . . . . . . . . . . . . .
96
4.12 Process flow diagram of R744 heat pump Cycle 3 . . . . . . . . .
98
4.13 Temperature - entropy plot of R744 heat pump Cycle 3 . . . . . .
99
4.14 A plot of the relationship between COPc,P EC and COPh,P EC and
the split ratio, γ for Cycle 3 . . . . . . . . . . . . . . . . . . . . . 101
4.15 A plot of the relationship between COPc/h,h/m split ratio for Cycle 102
4.16 Sensitivity analysis of the COPc and COPh of Cycle 2 to the isentropic efficiency of the compressors . . . . . . . . . . . . . . . . . 104
4.17 Sensitivity analysis of the COPc and COPh of Cycle 2 to the isentropic efficiency of the expanders . . . . . . . . . . . . . . . . . . 105
4.18 Sensitivity analysis of the COPc and COPh of Cycle 2 to the isentropic efficiency the compressors and expanders . . . . . . . . . . 106
4.19 Supercritical CO2 Brayton cycle process flow diagram . . . . . . . 107
4.20 Basic transcritical R744 MVC cycle process flow diagram . . . . . 107
4.21 Supercritical CO2 Brayton cycle thermal to mechanical conversion
efficiencies for a range of inlet and outlet pressures. . . . . . . . . 108
4.22 A plot of the relationship between temperature at state point 12
(internal heat recuperator inlet) against temperature at state point
13 (first stage expander inlet temperature) for Cycle 2 . . . . . . . 111
Chapter 0
xvii
Hybrid Trigeneration - Thermally Activated Heat Pumps
4.23 A plot of the relationship between the exergetic efficiency of the
exhaust heat recuperation process and T13 . . . . . . . . . . . . . 111
4.24 A plot of the relationship between COPc,trigen.system and T13
. . . 112
4.25 A plot of the relationship between COPh,trigen.system and T13 . . . 112
4.26 A plot to compare the relationship between COPc,th,directf iring and
COPc,th,cycle against T13 . . . . . . . . . . . . . . . . . . . . . . . . 113
5.1
Diagram to indicate the numbering convention for counterflow heat
exchanger . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
5.2
Alrogithm used to model a counterflow heat exchanger . . . . . . 119
5.3
Diagrammatic representation of a cross section through the shell
and tube heat exchanger . . . . . . . . . . . . . . . . . . . . . . . 120
5.4
Schematic of the hot water generator . . . . . . . . . . . . . . . . 121
5.5
Hot water generator hot - side low pressure temperature profile . . 123
5.6
Hot water generator hot - side high pressure temperature profile . 124
5.7
Internal heat recuperator temperature profiles (first - stage work
expander inlet temperature of 350 ◦ C . . . . . . . . . . . . . . . . 125
5.8
Schematic diagram to illustrate the fluid inlets and outlets of an
R744 ejector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
5.9
Schematic diagram to illustrate the state points in the R744 ejector 128
5.10 Ejector motive flow nozzle shape and mean bulk fluid velocity along
the nozzle length . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
5.11 Ejector bulk fluid density along the nozzle length . . . . . . . . . 131
5.12 Ejector combined flow nozzle shape and mean bulk fluid velocity
along the nozzle length
. . . . . . . . . . . . . . . . . . . . . . . 132
5.13 Ejector bulk fluid density along the diffuser length
. . . . . . . . 132
5.14 Cross sectional view and notation of compressor impeller blade
profile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
5.15 Top view and notation of compressor impeller blade profile for the
intake area
xviii
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133
Chapter 0
Hybrid Trigeneration - Thermally Activated Heat Pumps
5.16 Geometry of a trochoid (peritrochoid) [118] . . . . . . . . . . . . . 137
5.17 Geometry of a trochoid envelope [118] . . . . . . . . . . . . . . . . 139
5.18 Coordinate systems and contact point for the outer rotor . . . . . 141
5.19 Illustration of a typical outer rotor profile
. . . . . . . . . . . . . 142
5.20 Illustration of a typical inner and outer rotor profile
5.21 Illustration a typical gerotor rotating through
2π
N
. . . . . . . 143
of a turn . . . . 144
A.1 Process flow diagram of R744 heat pump cycle concept ‘Cycle A’
166
A.2 Process flow diagram of R744 heat pump cycle concept ‘Cycle B’ . 166
A.3 Process flow diagram of R744 heat pump cycle concept ‘Cycle C’ . 167
A.4 Process flow diagram of R744 heat pump cycle concept ‘Cycle D’
167
A.5 Process flow diagram of R744 heat pump cycle concept ‘Cycle E’ . 168
B.1 Impeller outlet velocity triangle . . . . . . . . . . . . . . . . . . . 170
B.2 Impeller inlet velocity triangle
Chapter 0
. . . . . . . . . . . . . . . . . . . 170
1
Chapter 1
Review of Trigeneration Systems
2
Hybrid Trigeneration - Thermally Activated Heat Pumps
1.1
Context
It is estimated that residential, commercial and industrial air conditioning consumes approximately 1 × 1012 kWh of electricity annually worldwide [1]. The
effect of climate change is expected to increase worldwide air conditioning demand by 72% by the year 2100 [2]. This statistic is independent from increases in
air conditioning demand resulting from societal changes, economic changes and
changes in building design, use and thermal efficiency. The prospective economic
development of Asia and Africa, in particular, will have a profound effect on the
demand for air conditioning. Upper estimates predict a global increase in air
conditioning demand to rise to 1 × 1013 kWh [1]. It is well documented in the media of this day that China and India experience regular and significant shortfalls
in electricity supply during hot periods. This commonly results in brownouts
and blackouts, factory closures, disconnection of residential supples, and loss of
economic output.
Conversely, worldwide heating demand is expected to decrease by 34% over
the same period of time as a result of climate change. There is therefore significant scope and opportunity for new cooling technologies, and the integration
of heating and cooling technologies to deliver increased efficiencies and primary
energy savings; this is the basis for, and focus of, this Thesis.
1.2
Trigeneration Systems
The term trigeneration, or Combined Cooling, Heating & Power (CCHP), implies
the simultaneous production of electricity, heating and cooling from a single energy source. Trigeneration is a concept born from the ideal of fully utilising the
available energy from a power generation system, typically achieved through the
recuperation of exhaust heat from the Primary Generation Unit (PGU). Given the
relative difficulty of transmitting heating and cooling flows compared to electricity, trigeneration systems are solely used in distributed power generation systems.
Chapter 1
3
Hybrid Trigeneration - Thermally Activated Heat Pumps
Crucial in the design and optimsation of trigeneration systems is the matching of
economical, energetical and environmental factors; commonly referred to as the
‘3 - E Trilemma’ [3].
A review of trigeneration systems is given by Wu and Wang[4], which gives a
high level analysis of the performance and financial aspects of cogeneration and
trigeneration systems, PGUs, absorption and adsorption chiller working fluids. It
also provides a summary and commentary of the nature cogeneration and trigeneration systems all all scales across the USA, Europe and Asia. The trigeneration
system review presented within this Chapter includes an overview of Trigeneration system design, operation and control strategies, as well as a brief overview
of subsystems: Primary Generation Units (PGUs), thermal storage devices.
1.2.1
Trigeneration Subsystems
A typical trigeneration system will comprise of these basic subsystems; a Primary
Generation Unit (PGU), an Exhaust Heat Recovery (EHR) device and a chilling
device. Almost without exception further subsystems are necessary to ensure energetic feasibility, and these may include; an electrical storage device, a hot / cold
thermal storage device, an auxiliary boiler, a secondary heat pump (mechanically
activated to supplement thermally activated - See Chapter 3) and the wider electrical grid. When a trigeneration system has two subsystems for producing the
same energy flow, the system is referred to as a hybrid trigeneration system.
Primary Generation Unit
Suitable PGUs for trigeneration applications include: Rankine cycle devices, internal combustion engines (both liquid and gas fueled), open and closed cycle gas
turbines, fuel cells, concentrated solar driven Brayton and Rankine cycle devices.
Steam Turbine A principal design consideration when selecting steam turbines is whether the steam exit pressure is higher or lower than atmospheric
4
Chapter 1
Hybrid Trigeneration - Thermally Activated Heat Pumps
pressure, the former of which is commonly referred to as a backpressure turbine
and the latter a condensing turbine. A backpressure turbine enables steam at an
intermediate pressure to be produced for other uses, whilst a condensing turbine
allows electrical and thermal power outputs to be varied. Steam turbines are
very reliable and have an extremely long life if corrected maintained. However,
slow response time, low conversion efficiency at smaller scale and poor part load
performance severly limit their use in distributed trigeneration systems.
Internal Combustion Engines Commonplace in systems below 1 MWe , reciprocating engines are split into two categories: compression ignition and spark
ignition, which typically use diesel and gas. They are a very flexible power source,
with fast start up and excellent partial load efficiency. However, they are complex and require frequent maintenance, which means that their availability is
lower than the other PGUs mentioned. Further, it is difficult to fully utilise the
various heat sources and diverse temperature levels in the context of a trigeneration system.
Gas Turbines Whilst gas turbines are often used as the PGU in large scale
Combined Heat & Power (CHP), their use in systems below ∼5 MWe is uneconomical due to their low high cost per kWe . Gas turbines typically have lower
capital and maintenance costs than reciprocating engines, as well as more affordable NOx emissions control systems, which are vital for installation in emissions
controlled zones. With exhaust temperatures around 500 ◦ C, there is significant
scope for subsequent processes (e.g. high temperature Organic Rankine Cycle
(ORC) systems, multi - stage absorption chillers, phase change thermal storage).
Much like steam turbines, gas turbines also suffer from poor part load performance and slower response time than reciprocating engines, as well as significantly
diminished performance at higher ambient temperatures and lower ambient pressures.
Chapter 1
5
Hybrid Trigeneration - Thermally Activated Heat Pumps
Fuel Cells Fuel cells pose an interesting prospect for PGUs in a distributed
generation system, particularly in large cities where emissions and noise control
is vital. Fuel cells are quiet, compact and require very little maintenance. There
are five major fuel cell technologies [5], the most mature of which are based
on a phosphoric acid electrolyte; however, fuel cells based upon a solid oxide
electrolyte, typically a yttrium stabilised zirconia, show great promise in CHP
and CCHP applications due to their high efficiency and operational temperatures (600 - 1,000 ◦ C). Potential fuels for these types of fuel cell are numerous and
include hydrogen, natural or biogas and alcohols.
Chilling Device
This is presented in Chapter 2.
Thermal Storage Device
Sensible Sensible heat storage refers to thermal storage systems where the storage media does not undergo any chemical of phase related changes; energy stored
is simply the product of the mass, specific heat and the temperature difference
Z T2
(Qsens. =
m · cp (T ) · dT ). Suitable storage mediums are greatly influenced by
T1
a number of properties, including: thermal conductivity, thermal diffusivity, heat
loss coefficient, operational temperatures, vapour pressure, chemical compatibility and cost. Sensible heat storage media can be categorised as either solid or
liquid, and single or dual media [6].
In solid medium systems the media is usually packed in beds and a heat transfer fluid passes through the beds; this is known as a dual storage system. Layering the storage medium in strata (packed beds) promotes thermal stratification,
which is advantageous because it allows for a high level of thermal control. For
example, the hot strata is in discharging mode when cold fluid is passing through,
and the cold strata is in charging mode when hot fluid is passing through. One
major problem with dual storage stratified systems is the large pressure drop and
6
Chapter 1
Hybrid Trigeneration - Thermally Activated Heat Pumps
consequent required pumping power consumption, resulting from passing the heat
transfer fluid through the beds [7].
In liquid medium systems it is possible for the thermal stratification of the
storage media to be maintained passively . In order to utilize this characteristic,
the hot fluid must enter the storage vessel at the top (during charging), whilst the
cold fluid must exit the storage vessel at the bottom (during discharging). This
type of storage is known as thermocline and relies upon the density differences
between the hot and cold fluid, due to thermal expansion. The storage medium
itself has a large implication on the design of the storage system.
In single medium systems the storage medium itself circulates through the
charging heat exchanger directly. Single medium systems can operate with one
or two tanks. Dual tank systems utilise a cold tank for the cold fluid being
discharged and a hot tank for the hot fluid being charged. The flow rate of hot
and cold fluid is controlled, such that during periods of high demand and low heat
supply the cold fluid has a high flow rate and the hot fluid has a low flow rate
implying the storage system is in a pure discharging mode. Single tank systems
are known as thermocline storage systems and, as previously explained, operate
as a result of density variations in the hot and cold fluid. Extensive research into
thermocline storage has been conducted by Kandari et al.[8].
The principal characteristic of the dual media storage system is that the storage medium is stationary and does not circulate around the plant. The storage
medium can be a solid or liquid undergoing sensible heating, or a phase change
material. Typically, the heat transfer fluid passes directly through the storage
media, although for solid medium the heat is usually transferred by a heat exchanger and not direct contact. There are however some major drawbacks with
dual media systems [6]. Firstly, when the system is discharging the storage material naturally cools down which causes the temperature of the heat transfer
to decrease increasing the thermal inertia of the system, reducing the speed at
which the system can switch from discharging back to charging. Secondly, the
Chapter 1
7
Hybrid Trigeneration - Thermally Activated Heat Pumps
thermal conductivity capabilities of the storage media/HTF must be sufficient
to satisfy the internal heat transfer requirements, in order to charge or discharge
at an acceptable rate. A poor thermal conductive material can result in greatly
reduced performance and high thermal gradients across the media, a potentially
serious problem with some solid media.
Sensible liquid storage mediums in use today and under research are predominantly oils and salts [6, 7, 9–11]. Salts have the advantage of low cost, but most
exhibit a high freezing point, which means that parasitic heating may be required
to present them from freezing during periods of low storage demand. Typical salts
used for heat storage include nitrate, nitrite and carbonate salts. Synthetic and
silicone oils are also used, which these come at a higher cost. Synthetic oils are
also hazardous, which place restrictions on their use in certain environments.
Latent The underlying concept of latent heat thermal storage is to utilise the
large amount of energy required in forcing a material to undergo a phase change
; ie. heat of vaporization, heat of fusion or heat of a solid-solid crystalline transformation. When latent heat themal storage is chosen over sensible heat storage
the result could be a significantly smaller storage system, without the negative
effects of temperature glide associated with sensible heat storage. The difficulties
with latent heat storage systems lie around the heat transfer mechanisms, transport mechanisms and the properties of the storage medium itself; research has
shown that the performance of medium degrades quite rapidly after only a small
number of freeze-melt cycles [6, 10]. A number of recent studies have focused
on methods of enhancing heat transfer through the addition of finds, multitubes,
insertion/dispersion of high thermal conductivity materials and micro/macro encapsulation [12].
A comprehensive list of potential Phase Change Materials (PCM) for use in
thermal storage systems is given by Agyenim et al. [12] and Sharma et al [13].
Considerable research into latent heat thermal storage has focused on applications
in domestic space heating and cooling, with common PCMs including: paraffin,
8
Chapter 1
Hybrid Trigeneration - Thermally Activated Heat Pumps
salts, wax, n-octadecane and stearic acid [10, 14]. Very few studies into medium
temperature latent heat thermal storage for absorption chiller space exist.
RedOx & Thermochemical RedOx, or reduction and oxidation, reactions
can be used as an extremely effective method of storing thermal energy, and have
therefore become the focus of significant amounts of research into thermal storage
techniques [3, 6, 13, 15]. During charging, thermal energy is essentially converted
into chemical bonds via a reduction reaction, whilst during discharging thermal
energy is released via an oxidation reaction. RedOx thermal storage has not been
demonstrated at a commercial level and only recently at a prototype level due to
the complex fundamental challenges which must be overcome.
There are an almost limitless number of materials which could potentially be
used in a RedOx thermal storage systems, including: hydroxides, carbonates and
metal oxides, and mixtures thereof. There are two fundamental design concepts
governing the operation of RedOx thermal storage reactors; direct and indirect
heat exchange. Direct heat exchange involves the flow of heat transfer fluid,
typically the gaseous reactant and/or an inert gas, through the storage media.
Indirect heat exchange means that the hot/cold fluid passes through an external
heat exchanger.
The performance of indirectly heated reactors is significantly hampered by
the poor thermal conductivity of the fixed bed, which in the case of calcium
hydroxide (Ca(OH)2 ) has been measured to be below 1 W/m.K [16]. This results
in poor conversion (oxidation) of the media during discharging. Hence, this is an
area which requires rigorous research and development to ensure the successful
deployment of indirectly heated RedOx thermal storage systems.
Directly heated reactors negate the effects of poor thermal conductivity associated with indirectly heated reactors [16], due to the direct mixing of the
hot/cold fluid with the media. However, this in turn creates other issues which
must be addressed; mass transfer, pressure gradients, reaction rates and thermal
capacity of the hot/cold feed stream. It is noted by Van Essen et al. [15] that
Chapter 1
9
Hybrid Trigeneration - Thermally Activated Heat Pumps
for an optimal heat storage media the rate of chemical reaction must be considerably higher than the rate of heat transport, a view echoed by Schaube [16].
Although an increase in reaction rate serves to improve the reactor conversion
performance, its effect is limited by the thermal capacitance of the feed stream.
To overcome this phenomenon the mass flow of feed stream has to be increased
in accordance with the heat generated by the reactor (provided the system is in
discharging / hydration / oxidation). The increased flow rate through the reactor
leads to an increase in pressure drop over the fixed bed reaction zone, which in
turn implies in a higher pumping power is required.
Metal oxide based heat storage has been highlighted by Wong and Brown
[17]; ten potential multivalent metal oxides as storage media are considered in
this study. The oxides were identified using both feasibility studies and thermodynamic modeling. It was discovered through non-isothermal redox measurements,
all ten candidate oxides underwent full thermal reduction under heating with air,
but only seven indicated a measurable level of oxidation during cooling. The
most promising of these was a mixture of 90 % cobalt oxide and 10 % iron oxide
(by mass).
Sorption Absorption thermal storage using NaOH and H2 O has been studied
in depth as a method of long term thermal storage. During charging, thermal
energy is produced (often by a solar collector) and drives the desorption process,
whilst the ground acts a heat sink. During discharging, the ground/ambient air
acts as a heat source, whilst the absorption process upgrades the heat source
to useful levels for space heating and hot water supply. The theoretical storage
density of NaOH is 250 kWh/m3 , and its cost is approximately e250 per m3 [3].
In order to improve the quality of heat produced during the discharging process, a
double-stage cycle has been conceptualised and is due to be created and analysed
by Weber and Dorer [18]. It is expected that the addition of the second stage
will decrease the storage density, resulting in a much larger and more complex
system.
10
Chapter 1
Hybrid Trigeneration - Thermally Activated Heat Pumps
One of the most promising thermochemical storage technologies for space
heating applications is the thermochemical accumulator . This is an absorption
process which uses a fluid pair operating in the solid, solution and vapour phases.
Sorption systems operate with either solution and vapour phases (absorption) or
solid and vapour phases (adsorption). TCA systems using LiCl salt and water
have achieved an experimental storage density of 253 kWh/m3 , and an experimental cooling COP
1
of 0.46 [19]. However, the very high cost of LiCl (e3500
per m3 ) is very prohibitive to the success of this technology.
Sorption storage can also be provided by adsorption chillers, via the desorption and adsorption of a fluid onto a solid surface. The quality of the heat stored
can vary dependent upon the number of stages within the cycle and the chosen
fluid-pair, but temperatures of below 80◦ C to above 300◦ C are possible [3]. The
prospect of seasonal sorption heat storage for space heating and cooling was studied by Mottillo, Zmeureanu and Beausoleil-Morrison [20]. The analysed system
consisted of a solid oxide fuel cell cogeneration system, coupled to a hot water
tank for short-term thermal storage and a sorption storage unit for long-term
thermal storage. The author concluded that the sorption storage system successfully provided heat during the winter months which would otherwise come from
a back-up boiler, but during the summer months the sorption unit did not store
sufficient amounts of thermal energy, as the temperature of the heat source (fuel
cell coolant) was not high enough to fully charge the unit.
There are two types of adsorption processes; physical adsorption (physisorption) and chemical adsorption (chemisorption), the former of which is due to Van
der Waals forces and the latter valency forces. Examples of physisorption solidfluid pairs are zeolite-water and silicagel-water. Zeolites are alumina silicates with
high micro-porosity, and are fully compliant with all environmental regulations.
Zeolite and water sorption storage systems using synthetic zeolites have reached
experimental storage densities of 124 kWh/m3 in heating and 100 kWh/m3 in
1
The Coefficient of Performance (COP) is the ratio of the external work input (mechanical
or thermal) into the system and the useful cooling or heating effect produced by the system.
Chapter 1
11
Hybrid Trigeneration - Thermally Activated Heat Pumps
cooling, with COPs of 0.9 and 0.86 respectively [3]. Silicagel-water has been
shown to reach an experimental storage density of 50 kWh/m3 , although theoretically 200-300 kWh/m3 . Examples of chemisorption pairs are sodium sulphidewater (Na2 S-H2O), magnesium sulphate heptahydrate (MgSO4 -7H2 O) and iron
hydroxide (Fe(OH)2 ). Sodium hydrate is a highly corrosive substance and operates under a vacuum, and has a measured storage density of 1980 kWh/m3 in
heating and 1300 kWh/m3 in cooling, with a COP of 0.84 and 0.57, respectively.
Magnesium sulphate heptahydrate is promising for solar energy storage, with a
storage density of 780 kWh/m3 and an inlet temperature of 122 ◦ C. MgSO4 -7H2 O
is non-corrosive and non-toxic but very expensive (e4870 per m3 ) [3]. Magnesium sulphate heptahydrate (MgSO4 7H2 O) and iron hydroxide (Fe(OH)2 ) appear
to be the most promising chemisoption pairs, with storage densities an order of
magnitude higher than that of conventional hot water storage.
Sorption storage technologies can be further defined in terms of open and
closed systems. Open systems operate at atmospheric pressure and the working
fluid vapour, typically water, is realised to the environment. In closed systems the
regeneration requires a higher temperature, which is useful if the desired storage
temperature is high.
1.2.2
Trigeneration System Design
Trigeneration system design is a term used to encompass all factors concerning
the basic design of the system: energy inputs, flows, outputs and wastes. This
is achieved via individual subsystem design and a top level system design. In
a separate system, energetic efficiency is the key performance driver behind the
system design, which can be complex in a trigeneration system design due to
the interactivity of the different subsystems. Subsystems will often be required
to operate at sub - optimal efficiencies, in order to maximise the EUF of the
Trigeneration system as a whole.
A theoretical study of a conceptual trigeneration system is documented by
12
Chapter 1
Hybrid Trigeneration - Thermally Activated Heat Pumps
Ameri et al. [21]. The system utilises a micro gas turbine coupled to a Exhaust
Heat Recovery Steam Generator (EHRSG) and three further subsystems which
achieve the follows tasks: domestic hot water production, winter heating production and summer cooling production. The proposed system has two ‘operational
modes’, in winter mode the system works purely as a CHP plant, whilst in the
summer the system works purely as a Combined Cooling & Power (CCP) plant.
In winter mode, the steam produced by the EHRSG directed to a heat exchanger where water is heated for building ambient space heating. In summer
mode, the steam produced by the EHRSG is directed to a steam ejector refrigeration cycle, which produces chilled water for building ambient space cooling.
In both modes, a proportion of the steam is directed to another heat exchanger
for the production of building hot water. Compared to a system consisting of a
grid connected centralised power station, a gas boiler and an electrically driven
MVC chilling device, the system proposed by Ameri et al. provides a Fuel Energy
Savings Ratio (FESR) of between 29 and 33 % and EUF of between 69 and 77 %
in the winter (CHP) mode. The corresponding values for FESR and EUF in the
summer (CCP) mode are between 18 and 21 %, and 53 and 58 %, respectively.
It can be noted that both the FESR and EUF are more favourable in the CHP
mode. However, the proposed system requires an auxiliary boiler to achieve the
required heating and cooling load in the winter and summer respectively.
A competing analysis of a similar system is put forward by Li et al. [22], with
very different results obtained. In typical cogeneration (dedicated CHP systems)
and trigeneration studies, the reference case is assumed to be a coal separate
production; this is a flaw which provides inaccurate results, due to the intrinsic
difference between coal and gas. In this study the reference case features a Combined Cycle Gas Turbine (CCGT) system as the generation method, and high
efficiency natural gas boilers as the heating method and high efficiency compression chillers as the cooling method. The trigeneration is as described by Ameri
et al. [21], although the ejector refrigeration cycle is replaced by an absorption
Chapter 1
13
Hybrid Trigeneration - Thermally Activated Heat Pumps
chiller (with an associated increase in efficiency). As a result of this viewpoint,
the results indicate that under certain conditions the trigeneration system actually has a negative FESR. In scenarios where there is a demand for cooling the
proposed systems exhibits a negative FESR almost without exception, whilst the
opposite is observed during periods of demand for heating. The values of FESR
for trigeneration systems which use gas engines as the PGU range from - 40 %
to 5 % in cooling mode and 15 % to 25 % in heating mode. Therefore, one can
conclude that the overall annual FESR of trigeneration systems strongly depends
on the ratio of heating hours to yearly operation hours, τ .
A range of algorithms used to obtain the optimal system design are well documented in the literature [23–26], including: linear sequential quadratic, non linear sequential quadratic, tri - commodity simplex, extended power simplex, genetic and Lagrangian relaxation.
A hybrid trigeneration system comprising a PGU, EHR device, auxiliary
boiler, absorption chiller and MVC chiller has been optimised using genetic algorithm by Wang. et al [23]. The analysis indicates that the proposed system
would, in the absence of the MVC chiller, produce a negative FESR when the
average temperature is above a certain threshold. It therefore follows that a system which features both absorption and MVC chillers would require significantly
less primary fuel to satisfy the cooling demand. An identical system is proposed
by Fumo et al. [27] and analysed for two cities in USA; the results indicating
that hybrid trigeneration is more likely to produce a positive FESR in colder climates, where the demand for cooling is less. Due to the low COP of absorption
chillers, the system thermal efficiency in electricity - cooling mode is lower than
in electricity - cooling mode.
A trigeneration system for total energy provision for a supermarket in the UK
is proposed and analysed by Tassou [28]; the system features an 80 kWe micro - gas
turbine with pre - heat, a EHR device and an absorption chiller. The absorption
chiller is a commercially available gas - fired ammonia - water chiller (ROBUR
14
Chapter 1
Hybrid Trigeneration - Thermally Activated Heat Pumps
ACR-60LB), with a cooling capacity of 12 kW and a measured COP of between
0.32 and 0.57, for brine outlet temperatures of −11 ◦ C and 3 ◦ C respectively.
In supermarkets, approximately 50 % of the electricity is used for refrigeration,
whilst 25 % is used for lighting and the remaining 25 % for cooking and ventilation
and computer systems [29, 30]. Assuming a conservative COP for the refrigeration
systems, this implies that the chilling demand in kWth is three times greater than
the total non - refrigeration based kWe electrical consumption. The price of gas
was found to have a notable effect on the system payback period; a change of
± 20 % from the base value alters the payback period by ± 40 %. Further, the
gas - to - electricity price ratio (‘spark spread’), ζ, has a profound effect on the
payback period; when ζ = 0.2 the payback period is 2.7 years, when ζ = 0.3 the
payback period is 4.5 years, when ζ = 0.4 the payback period is 13.7 years [28].
The high sensitivity of the economical feasibility of the system to gas prices is
most attributed to the very low COP of the absorption chiller, relative to the
COP for an equivalent MVC chiller.
1.2.3
Trigeneration Control Strategies
Whilst the selection, specification and sizing of plant is crucial to the energetic
and economical feasibility of trigeneration systems, the control strategy used to
monitor the interaction and energy flows between each of the subsystems is an
equally crucial factor. Given the variability of the demand side and the highly
complex and interdependent relationship between subsystems on the supply side,
managing and indeed correctly selecting an appropriate control strategy is difficult.
On the demand side, variability is comprised of four components, two external and two internal: seasonal fluctuations and climatic fluctuations representing
the external components; diurnal fluctuations and spontaneous fluctuations representing the internal components. Seasonal fluctuations are easily predictable,
and are likely to include demand side changes such as an increased heating or
Chapter 1
15
Hybrid Trigeneration - Thermally Activated Heat Pumps
cooling demand and a change in the diurnal pattern of electrical demand, resulting from changes in lighting hours due to longer days or nights. Climatic
fluctuations are less easily predictable and are dependent upon the variations
in local temperature, which can vary substantially over a period of a few days.
Both seasonal and climatic fluctuations are influenced by factors outside of the
end user control, these are therefore referred to as external components. Diurnal fluctuations are easily predictable and encompass the changes in demand
apparent due to behavioral changes in the users throughout the day. Volatile
fluctuations are unpredictable and apparent due to random changes in individual
user habits; if the number of users is large, the volatile fluctuations are under
most circumstances considerably less in magnitude than the others.
Figure 1.1: Relevance of demand side variability components to systems design
and control strategy; and the correlation of each to the amplitude, frequency and
predictability of the demand
Figure 1.1 gives an indication of the relevance of demand side variability
components to systems design and control strategy and the correlation of each
to the amplitude, frequency and predictability of the demand. Lower frequency
variations are more relevant to system design; the implication being that in order
to handle the effects of large seasonal variations in temperature (e.g. humid
continental climates, such as New England and Eastern Europe), appropriate
system design is crucial. High frequency variations are more relevant to the
control strategy; the implication being that in order to handle the effects of
anthropological cycles and unpredictabilities, an appropriate control strategy is
crucial [26, 31, 32].
16
Chapter 1
Hybrid Trigeneration - Thermally Activated Heat Pumps
A domestic - scale trigeneration system has been created and tested by Miguez
et al. [33]. The system comprises a small 2 - stroke petrol engine with a EHR
heat exchanger, and a reversible MVC heat pump. The heat pump is reversible
such that it can provide both summer cooling demands and compliment the
recuperated thermal energy during winter heating demands. There is also a large
thermal accumulator (a water tank) and deep cycle storage batteries to act as
buffers, to ensure consistent demand matching under a highly variable demand
scenario. When the demand side consists of a larger network of users, loads
and buildings; whose characteristics are more predictive and can be modeled and
controlled; it is possible to reduce the relative size of the buffers. Miguez et al.
describe five operational modes:
• Stand-by mode
• Electric generator mode
• Co-generation mode
• Heat pump winter mode
• Heat pump summer mode
Under heat pump summer mode the EUF of the system is poor because all the
useful thermal energy is rejected, unless there is thermal demand from either hot
water supply or to top - up the thermal accumulator. Further, if the demand for
electricity is high, the possibility that the system will become unable to meet the
combined electrical and cooling demand becomes a reality; the system becomes
energetically unfeasible. The addition of a thermally activated heat pump such
as a sorption chiller device would help negate this problem, which could indeed
be used in direct substitution of the MVC heat pump. This will be discussed
further in Chapter 3.
The most common control strategies are based upon operating the PGU in
accordance with the thermal demand (Thermal Load Following (TLF)), and in
Chapter 1
17
Hybrid Trigeneration - Thermally Activated Heat Pumps
accordance with the electrical demand (Electric Load Following (ELF)) [34]. The
philosophy behind TLF and ELF is such that the supply of either thermal or
electrical energy is aligned to the thermal or electrical demand, respectively. The
choice of TLF or ELF is usually affected by the operational characteristics of the
PGU, the presence and size of thermal storage and electrical storage systems, and
the ability to sell excess electricity to the grid. The cost of fuel and the cost of
grid electricity also has a huge implication upon the control strategy [35].
When a TLF strategy is adopted, both CHP and CCHP systems exhibit an
increase in EUF [36]. However, due to the low electricity demand to heat demand
ratio of most building sectors (typically 0.1 - 0.2 in the domestic sector and 0.4 0.6 in the tertiary sector [35]) this almost always results in excess electricity
being produced which must be either exported or stored for use at another time.
Therefore, when a ELF strategy is adopted, the thermal energy demands are
unlikely to be satisfied and supplementary boilers will be necessary.
An alternative operational strategy, a Hybrid Load Following (HLF), is proposed and analysed by Mago and Chamra [36]. Under this solution the ELF
strategy is employed under certain conditions, whilst under contrasting conditions the TLF strategy is employed. The philosophy of the HLF strategy is based
upon controlling the PGU such that no excess energy is produced, inevitably
creating a shortfall of one energy stream which is satisfied either by the grid
or a supplementary boiler. Although the model used by Mago and Chamra is
heavily simplified, particularly with respect to the relationship between power
output and recuperated heat of the PGU, the results indicate an increase EUF
and associated operational cost savings.
1.3
Conclusion
Worldwide demand for air conditioning is predicted to increase by 72 % by the
year 2100, primarily due to the increase in economic productivity of south east
Asia and Africa, which will result in an improvement in living standards and a
18
Chapter 1
Hybrid Trigeneration - Thermally Activated Heat Pumps
consequent increase in demand for air conditioning. It is predicted that this effect
will be compounded by climate change. Therefore, the efficient, low carbon and
low cost provision of chilled water for air conditioning systems is both a huge
challenge and opportunity. Trigeneration (the combined generation of electricity,
heating and cooling) is a progression of cogeneration (the combined generation
of electricity and heating), which will enable this goal to be reached.
This Chapter has given an overview of trigeneration system design, thermal
storage technologies and trigeneration control strategies. Trigeneration has been
shown to deliver an improved EUF over a system where electricity, heating and
cooling are produced separately [21, 28–30]. The magnitude of the improvement
in EUF is primarily dependent upon the performance characteristics of the PGU
and the device which converts waste heat into useable coolth. The control of
the system is an additional important factor in maximising the EUF [26, 31, 32],
and the integration of a thermal storage device can greatly improve the diurnal
and seasonal performance of the system. Trigeneration control and operation
strategies have been discussed and analysed at length in a number of studies
[23, 25, 26, 33, 36, 37], which conclude that the ideal control and operation
strategy is dependent upon climatic conditions and the performance parameters
of the trigeneration system.
1.4
Aims and Objectives of Thesis
A list of primary aims and objectives for the research documented in this Thesis
are given below.
1. Review the literature for interesting innovations in the field of mechanically
and thermally activated heat pump cycles. This will include academic journal and conference papers, PhD Theses, patents, commerically available
data.
Chapter 1
19
Hybrid Trigeneration - Thermally Activated Heat Pumps
2. Identify the innovations which exhibit a high level of perfomance or the
potential to achieve high levels of performance (n.b. COP is the primary
performance parameter of interest) given further development.
3. Design, model, analyse and optimise at least 3 thermally activated heat
pump cycles. The primary performance parameter is the cycle COP, but
consideration will also be given to the operating temperatures and pressures
as well as the practicability of the developing the cycle.
4. Develop the design of the most promising heat pump cycle, including performance optimsation and an analysis of the heat pump performance when
integrated within a trigeneration system.
5. Develop preliminary design data for specific components within the heat
pump cycle to serve as a basis for the continuation of research and development.
20
Chapter 1
Chapter 2
Review of Heat Pump Systems
21
Hybrid Trigeneration - Thermally Activated Heat Pumps
2.1
Introduction
Heat pumps encompass a wide variety of devices which transfer thermal energy
from a cold reservoir to a hot reservoir, as highlighted in Figure 2.1. In this
chapter the varying types of refrigerants are discussed, with particular focus given
to sorption heat pumps and R744 MVC heat pumps.
Figure 2.1: Classification of Heat Pump / Refrigeration System
2.2
2.2.1
Refrigerants
Single Component Refrigerants
Single component refrigerants are the most simple type of refrigerants available,
comprising a single substance. Refrigerants can be split into two categories:
synthetic and natural. Synthetic refrigerants are denoted by 000, 100, 200 or 300
22
Chapter 2
Hybrid Trigeneration - Thermally Activated Heat Pumps
series numbers under ASHRAE Std. 341 if they are methane based, ethane based
or propane based. Natural refrigerants are either inorganic compounds which
are denoted by 700 series numbers, or organic compounds which are denoted by
600 series numbers. Synthetic refrigerants include ChloroFluoroCarbons (CFCs),
HydroChloroFluoroCarbons (HCFCs) and HydroFluoroCarbons (HFCs), whilst
natural refrigerants include HydroCarbons (HCs), R717 and R744.
CFC
ChloroFluoroCarbons are synthetic man - made substances containing chlorine,
fluorine and carbon. Prior to the advent of CFCs, MVC heat pump systems
utilised more simple refrigerants such as ammonia, sulphur dioxide, carbon dioxide and chloromethane. It is well documented that CFCs contribute significantly
to the depletion of ozone in the upper atmosphere, which led to in the creation
of the Montreal Protocol in 1987 to phase out the use of CFC refrigerants. The
most common CFC refrigerant was R12, which has an Ozone Depletion Potential
(ODP) of 1 and a Global Warming Potential (GWP) of 2400 [38].
HCFC
HydroChloroFluoroCarbons are very similar in structure to CFCs, with the addition of one or more hydrogen atoms. The use of HCFC refrigerants became
widespread following the phasing out of CFCs in the 1980s, with refrigerants
such as R22 becoming ubiquitous in air conditioning units. HCFCs are less stable than CFCs and break down in the lower atmosphere, prior to reaching the
ozone layer. Most HCFC refrigerants have an ODP below 0.1 although GWP
levels remain very similar to CFCs. R22 has an ODP of 0.05 and a GWP of 1810
[38]. As of January 1 2010 the production of R22 for charging new systems was
be banned, and from January 1 2020 the production and import of all R22 will
be banned entirely. Global production and use of HCFC refrigerants, particu1
ASHRAE Standard 34:
www.ashrae.org
Chapter 2
Designation and Safety Classification of Refrigerants.
23
Hybrid Trigeneration - Thermally Activated Heat Pumps
larly R22, has been buoyant in recent years, owing to the rapid expansion of air
conditioning use in China and India [2].
HFC
HydroFluoroCarbons are a further evolution of HCFC refrigerants and do not
contain chlorine, which is the prime contributor to ODP in older refrigerants.
HFCs are non - toxic and have zero ODP, but do have a high GWP. One of the
most common HFC refrigerants is R134a, which is commonplace in domestic and
mobile air conditioning systems and has a GWP of 1430. Worldwide regulations
controlling the use of HFC refrigerants, such as the 1997 Kyoto Protocol [39] and
the 2006 EU MAC directive which controls the use of fluorinated gases with a
GWP greater than 150 in mobile air conditioning units [40] are driving innovation
in the refrigerant, heat pump and chiller market. It is clear that the age of
the chlorinated and fluorinated refrigerants is over, and research efforts must be
directed into systems which use alternative working fluids.
HC
Hydrocarbons have properties which make them ideal as refrigerants: they have
zero ODP, negligible GWP, low toxicity, are cheap to produce and have desirable
thermodynamic characteristics for achieving a high COP. The most commonly
used HC refrigerants are R600a and R290a, with GWPs of 4 and 20, respectively.
Since HCs are obviously highly flammable, there can be significant restrictions
regarding their use in certain applications.
Ammonia - R717
Ammonia is widely used as a refrigerant in large scale refrigeration facilities, and
has an ODP and GWP of 0. The thermodynamic properties of R717 are such that
it is often considered as a superior alternative to many older synthetic refrigerants
[41] because it has a higher heat transfer coefficient and a much higher volumetric
24
Chapter 2
Hybrid Trigeneration - Thermally Activated Heat Pumps
refrigeration capacity. R717 has a very distinctive smell, enabling leakages to be
easily detected by personnel and negating the possibility of asphyxiation, which
makes it an attractive refrigerant for large industrial facilities. However, R717
is highly toxic and combustible, and is also corrosive towards coppers and alloys
containing copper and zinc.
Carbon Dioxide - R744
Carbon dioxide is a non-toxic, incombustible substance with a ODP of 0 and a
GWP of 1, which is cheap, easy available, fully compatible with normal lubricants
and construction materials and is not required to be recovered at the end of
plant life. Due to the high volumetric capacity (approximately five times greater
than R22), plant size can be greatly reduced compared to common synthetic
refrigerant systems. R744 has a relatively low critical temperature of 31.1 ◦ C,
and a relatively high critical pressure of 73.77 bar. The optimum heat rejection
pressures are typically around 100 bar, which is significantly higher than for most
synthetic refrigerants, however the increased volumetric capacity implies that the
potential danger of pipe rupture is somewhat reduced (when considering both
pressure and volume as the prime contributers)
Carbon dioxide was first harnessed as a refrigerant by Thaddeus S.C Lowe in
1866 in the production of ice, and the first R744 compressor was designed and
built by Windhausen in 1880 [42]. R744 was used extensively in buildings where
humans could be exposed to the cooling systems: theatres, hospitals, restaurants.
However, the advent of synthetic refrigerants in the 1950s, which could provide a
greatly increased COP, resulted in the eventual phasing out of R744. In recent
years R744 has been making a comeback, owed in part to policy changes controlling the use of synthetic refrigerants (e.g. the Montreal Protocol), and also
to an appreciation of the numerous benefits of R744 heat pumps over synthetic
refrigerant heat pumps [see Section 1.4].
Chapter 2
25
Hybrid Trigeneration - Thermally Activated Heat Pumps
Nitrous Oxide - R744a
Nitrous oxide has a very similar molecular structure to carbon dioxide and an
almost identical molar mass, which results in both gases having very similar
thermodynamic properties and characteristics. It is acknowledged that R744a
have certain performance advantages over R744 heat pumps, although there are
limitations to these. R744a heat pumps exhibit a slightly higher COPc than R744
heat pump, require a lower pressure ratio and have a higher second law efficiency
[43, 44]. R744a also has a much lower triple point than R744 (−90.82 ◦ C vs
−56.56 ◦ C), which enables it to be suitable for use at very low temperatures.
However, R744a has a GWP or 298, has a lower volumetric heating and cooling
capacity, and begins to decompose at temperatures above 300 ◦ C.
2.2.2
Multi Component Refrigerants
Multi component refrigerants contain two or more refrigerants, and may be described as either azeotropic, zeotropic and near - azeotropic.
An azeotropic mixture is a combination of two or more refrigerants, which
all have the same concentration of liquid and vapour phase at a given pressure
and temperature. Azeotropic refrigerants are often suitable for low temperature
cooling, as they tend to have a lower triple point and vapouration temperature
then their constituents A common azeotropic refrigerant, and indeed exception to
the previous rule, is R507, which is a blend of R143a and R125 (50 : 50 by mass).
Azeotropic refrigerants are denoted by 500 series numbers under ASHRAE Std.
342 .
Very few azeotropic refrigerants exist, and in most instances a blended refrigerant is a near - azeotropic mixture. Common examples including R502 (a mix of
R115 and R22) and R410a (a mix of R125 and R32). Near azeotropic refrigerants
experience undesirable characteristics such as temperature glide and a very non2
ASHRAE Standard 34:
www.ashrae.org
26
Designation and Safety Classification of Refrigerants.
Chapter 2
Hybrid Trigeneration - Thermally Activated Heat Pumps
linear relationship between temperature and enthalpy. The former degrades the
overall COP of the by effecting the linearity of the evaporator temperature, whilst
the latter can result in undesirable temperature pinches in heat exchangers.
A zeotropic mixture is a combination of two or more refrigerants, whose liquid and vapour concentrations are never equal. Most blended refrigerants are
zeotropes to a varying degree, which results in systems using them experiencing
temperature glides in the evaporator and condenser to varying degrees. R410a (a
near - azeotropic refrigerant) typically experiences a temperature glide of 0.5 ◦ C,
whilst the glide for R407c (a blend of R32, R125 and R134a) is about 4 ◦ C. The
temperature glide is apparant due to differences between the bubble and dew
temperatures of the two fluids. Zeotropic and near-azeotropic refrigerants are
denoted by 400 series numbers under ASHRAE Std. 34.
Zeotropic mixtures are used in absorption chillers because the two fluids have
vastly different vapour pressures, which enables a desorption and resorption process of the refrigerant to the absorbent. Over 40 different refrigerant and 200
absorbents are discussed in the literature [4]. The two most common working
fluid pairs (LiBr - H2 O and H2 O - NH3 ) are highlighted below, although others
include: R22 - TEGDME, CH3 OH - LiBr ad CH3 OH - LiCl [45].
Lithium Bromide - Water
Lithium bromide and water (LiBr - H2 O) is one of the most common working fluid
pairs in absorption chillers, with lithium bromide serving as the absorbent and
water as the refrigerant. Absorption chillers which utilise LiBr - H2 O are capable
of operating at high temperatures, albeit with inhibitors to control corrosion in
the generator. LiBr - H2 O can be used in multi-stage absorption chillers, which are
capable of much of a higher COP [46–49]. In order to sufficiently cool the absorber
and condenser, a cooling tower is typically required. Due to crystallisation of
lithium bromide at relatively high temperatures, LiBr - H2 O chillers are not able
to provide low temperature cooling.
Chapter 2
27
Hybrid Trigeneration - Thermally Activated Heat Pumps
Water - Ammonia
Water and ammonia (H2 O - NH3 ) is another common working fluid pair in absorption chillers, with water serving as the absorbent and ammonia as the refrigerant. H2 O - NH3 is a less zeotropic mixture than H2 O - NH3 , which results in the
quality of ammonia refrigerant exiting the generator being low. The presence of
water in the ammonia circuit leads to a very significant temperature glide in the
condenser and evaporator. In order to counteract this a rectifier is located following the generator, which allows the entrained water vapour to condense untit
such a point that the ammonia quality is sufficient (usually greater than 0.995
[50, 51]). H2 O - NH3 chillers do not suffer from crystallisation problems in the absorber, and are therefore capable of providing cooling at very low temperatures,
therefore H2 O - NH3 are often used as the bottoming cycle to a LiBr - H2 O chiller
[52, 53].
2.3
Absorption Heat Pumps
Absorption heat pumps are commonly referred to as absorption chillers, due to
their predominant use as a way of converting low - to - mid grade waste heat into
cooling. Absorption chillers use a zeotropic mixture as the working fluid, where
one fluid is the refrigerant and the other the absorbent. In its most simplest form,
the absorption cycle process can be described as a desorption - condensation expansion - evaporation - absorption - compression process. A simplistic process
flow diagram of this process in given in Figure 2.2.
Absorption chillers are a very well established technology, but compared to
MVC heat pumps have a relatively small market share. There are a number
of potential reasons for this: the relatively high COP of MVC heat pumps, the
simplicity of MVC heat pumps, the abundance and widespread availability of
cheap electricity, the increased size and weight of absorption heat pumps. In
distributed CCHP systems the PGU exhaust stream is a potential high grade
28
Chapter 2
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 2.2: Basic process flow diagram of a typical single-stage absorption chiller
heat source, which can be used to drive high efficiency multi - effect absorption
chillers.
2.3.1
Multi - Stage / Multi - Effect
Single - stage, single - effect absorption chillers tend to achieve a COPc of around
0.7, with a associated generation (desorption) temperature of around 80 ◦ C [45].
Double - stage, double - effect absorption chillers have corresponding values of
around 1.2 and 150 ◦ C, whilst triple - stage, triple - effect chillers achieve figures of
around 1.6 and > 200 ◦ C [48]. The limitation in the maximum number of stages
is the inhibition of corrosion in the high temperature generator as the generation
temperature and pressure increases. A further complication is the increased level
of complexity, which can make successfully controlling the system very difficult
[47].
In multi - stage absorption chillers the working fluids pass through a number
of generators, each operating at differing pressure levels, which greatly increases
the mass proportion of refrigerant produced relative to the mass flow of solution,
Ψ. An ‘effect’ refers to the process of transferring heat from one part of the cycle
to another; all multi - stage chillers are also multi - effect chillers. For LiBr - H2 O
chillers, multiple cycles are cascaded and share a single absorber, the thermal
Chapter 2
29
Hybrid Trigeneration - Thermally Activated Heat Pumps
energy from the hot refrigerant generated in the upper cycle is used to drive the
generator of the cycle below it; in some systems the refrigerant condenses during
this process. Monitoring and control of temperature pinch is key in ensuring
exergetic efficiency. There are three possible flow configurations in multi - stage
cascade chillers: series flow, reverse flow and parallel flow.
In series flow configuration, the strong3 refrigerant / absorbent solution exiting the absorber flows into the high temperature generator (through a series of
solution - to - solution Heat Exchangers (HXs)). The weaker solution exiting the
upper generator passes into the generator below, and so on, until the strong solution outlet of the low temperature generator which flows back into the absorber.
In reverse flow configuration, the wet solution passes to the lower generator first
and finishes at the high temperature generator. In parallel flow configuration, the
wet solution passes is split and passes through each generator independently, the
mass flow ratio of which is vital, and returns directly to the absorber. The performance characteristics of each configuration as a function of the absorber inlet
solution concentration has been discussed by Kaita [47]. Parallel flow configuration achieves the highest COPc throughout the range of absorber inlet conditions,
whilst series flow achieves the lowest COPc . The difference is more profound when
the absorber inlet solution is weaker, whilst when the solution is strong all configurations provide a similar and higher COPc , implying that if the generators
are efficient in producing refrigerant the choice of configuration is unimportant.
A similar study by Aghdam, Ranjbar and Mahmoudi [54] concluded that
series flow is capable of achieving a higher COPc than parallel flow, provided
that the evaporator pressure and generation temperature are favourable (both
being high). Series flow exhibits a less constant COPc than parallel flow over the
range of evaporator pressures and generation temperatures considered. In both
configurations, the generation temperature (ranging from 180 ◦ C to 210 ◦ C) has
much less profound effect on COPc than the evaporator pressure. The implication
3
‘Strong’ implies that the mass proportion of refrigerant in the solution is low, ie. it is strong
in absorbent
30
Chapter 2
Hybrid Trigeneration - Thermally Activated Heat Pumps
is that the minimising temperature pinch in the evaporator is more vital than in
the generator.
An analysis of a steam driven microscale double - stage LiBr - H2 O chiller, operating with parallel flow configuration has been presented by [55]. It was founds
that the high temperature generator produces approximately 60 % of the refrigerant, whilst the low temperature generator produces 40 %; this ratio remains
consistent regardless of the cooling output.
The number of effects in an absorption chiller can, and often does, exceed the
number of stages in the chiller, an example of which is the double - stage, triple effect H2 O - NH3 ‘Kangaroo’ cycle proposed by Shelton, Jacob and Schaefer [56].
A very similar concept was patented by DeVault in 1995 [53].
The kangaroo cycle features two independent cycles: a high pressure outer
cycle operating at 200 ◦ C, and a low pressure inner cycle driven by the heat
rejected by the heat of absorption and heat of condensation from the upper cycle.
Its name is derived from idea that the inner cycle is within the pouch of the
outer cycle, a phenomenon manifested through observation of the Duhring plot in
Figure 2.3. The pressure of the working fluid pair is plotted on the ordinate axis,
and the inverse of temperature is plotted on the abscissa. The cycle evaporators
are represented in the lower left portion of the plot, whilst the inner generator is
represented by the upper right portion of the red curve and the outer generator
by the upper right portion of the green curve. The left side of the plot represents
the part of the cycle which is predominantly ammonia refrigerant (i.e. x = 0.99),
whilst the right side of the plot is the sorbent rich mixed fluid (i.e. x = 0.2). This
cycle is capable of achieving a COPc of 1.50 when Tamb = 30 ◦ C, although the
performance is highly sensitive to the Tamb , with COPc dropping to 1.15 when
Tamb = 35 ◦ C.
The COPc of the double - stage triple - effect H2 O - NH3 ‘Kangaroo’ cycle is
much greater than the COPc of an equivalent double - stage double - effect LiBr H2 O cascaded cycle [55]. A similar system is proposed in a patent registered to
Chapter 2
31
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 2.3: Dühring plot - Kangaroo cycle [56]
DeVault [53]. The proposed system differs from the aforementioned ‘Kangaroo’
cycle in that the inner loop is effectively a double - stage cascade cycle (i.e. two
generators and condensers), which implies a dual stage condenser is required in
the outer loop
Numerous multi - stage and multi - effect absorption chiller cycles using non standard working fluids have been proposed in the literature, including: a double - stage absorption / compression cycle using CH3 OH - TEGDME and TFE TEGDME [57], CH3 OH - LiBr [45], CH3 OH - LiCl [45] and H2 O-LiBr:LiI:LiNO3 LiCl [58].
2.3.2
Generator - Absorber Exchange Cycles
GAX cycles recycle some of the low grade heat available in the absorber and use
it to drive a portion of generator [59]. This is made possible by a temperature
overlap between the hot side of the absorber and cold side of the generator. When
combined with solution HX transferring thermal energy from the strong solution
exiting the generator to the weak solution exiting the absorber, a high proportion
32
Chapter 2
Hybrid Trigeneration - Thermally Activated Heat Pumps
of the thermal energy supplied to the desorption process is internally recycled
heat. This provides a reduction in external heat requirements and increases the
COP of the cycle. The basic principal of the GAX cycle is highlighted on a
Dühring plot in Figure 2.4.
Figure 2.4: Basic GAX cycle
A computational analysis of four very similar H2 O - NH3 GAX cycles has been
documented by Park, Koo and Kang[59]. The cycles each feature three stages
within the absorber and three stages with the generator. In the generator the
high stage is the driven by external heat, the middle stage by the strong solution
exiting the high stage generator, and the low stage by the high stage absorber.
In the absorber, the high stage is cooled by the low stage generator, the middle
stage by the weak solution exiting the high stage absorber, and the high stage by
an external heat sink.
The difference between the four different cycles is the sequence and method
of heat rejection, and the parallelisation of the middle and low stage absorbers.
Three cycles are capable of providing simultaneous heating and cooling, whilst
one is capable of providing cooling only. The COPc as a function of hot water
outlet temperature is shown in Figure 2.5. This demonstrates that a COPc of
Chapter 2
33
Hybrid Trigeneration - Thermally Activated Heat Pumps
approximately 1 is achievable under the best operating conditions (lowest hot
water outlet temperature), but the cycle can only really be considered as a pure
cooling mode cycle under this scenario. For simultaneous heating and cooling
(i.e. hot water outlet temperature > 50 ◦ C) the best configuration achieves a
COPc of approximately 0.85.
Figure 2.5: Basic GAX cycle[59]
An advanced GAX cycle, developed by Kang et al., included the addition
of a fourth ‘waste heat’ stage within the generator [60]. It was found that the
generator outlet temperature could be reduced whilst simultaneously increasing
the COP for all scenarios. However, when considering the enthalpy of the waste
heat in the calculation of COP, the COP was lower than a standard GAX cycles
in almost all scenarios.
A further modification to the standard GAX cycle is the Generator Absorber
Exchange Absorption Compression (GAXAC) cycle [61]. Hybridised absorption
chiller cycles are well documented [62–66], but hybrid GAX absorption chillers
cycles are a relatively unknown concept. Kang, Hong and Park [67] analysed
four slightly different GAXAC cycles. The differences in the cycles was the of
the location of the compressor unit on the refrigerant side absorber inlet and
refrigerant side generator outlet. The aim of the former is to create a wide
34
Chapter 2
Hybrid Trigeneration - Thermally Activated Heat Pumps
temperature overlap between absorber and generator, and/or a lower evaporation
temperature. The aim of the latter is to reduce the generation temperature
(via a reduction in the generator pressure), and/or to increase the condenser
temperature for hot water production. The highest COPc calculated was 1.24
and was achieved in the former case, 85 % of energy input was thermal and 15 %
isentropic compression work. COPc increases as a function of absorber pressure,
by as much as 24 % over the standard GAX cycle (albeit with an assumed 100 %
compressor isentropic efficiency).
2.3.3
Vapour Recompression Absorption Cycles
The Vapour Recompression Absorber (VRA) is a noteworthy modification to the
basic absorption chiller cycle, which aims to increase the refrigerant production
rate and therefore the cycle COP [68, 69]. The VRA effectively adds a third stage
to a double stage cycle. It achieves two things: an increase in the strength of
the solution entering the absorber, and a decrease in the strength of the solution
entering the upper generator. When compared to a conventional triple effect
LiBr - H2 O absorption chiller cycle, the double effect LiBr - H2 O VRA absorpton
chiller cycle offers simplification, a reduction in upper generator temperature and
a potential decrease in the overall size of the system. The addition of the VRA
has been demonstrated to theoretically increase COPc by 8 % and reduces the
upper generator temperature by around 5 ◦ C. Whilst no laboratory study has
been identified within the literature, a patent for the concept does exist [70]
2.4
Adsorption Pumps
Adsorption heat pumps are devices which generally utilise a low grade heat source
to provide a cooling effect, with the potential for a subsequent heating effect.
Adsorption chillers operate using the same basic process as absorption chillers:
desorption, condensation, throttling, evaporation, adsorption, pumping. The maChapter 2
35
Hybrid Trigeneration - Thermally Activated Heat Pumps
jor difference is that the sorbent is in the solid phase and the sorbate / refrigerant
adheres to the surface, rather than permeating it. This consequence of this is
that in the adsorption / desorption process there is no change in the density of
the sorbent (in the absorption process there is an decrease in the density of the
sorbent as it absorbs the sorbate).
There are two types of adsorption processes: physical adsorption (physisorption) and chemical adsorption (chemisorption), the former of which is due to Van
der Waals forces and the latter valency forces [3]. Due to the fact that the sorbent is a solid and cannot be easily transported, the desorption and adsorption
processes both occur in the same location (i.e. the generator is the adsorber and
vice versa). This poses a problem in that multiple desorption / adsorption chambers / beds are required to maintain a quasi - continuous operation, with a control
system required to switch between the processes at the appropriate moment.
2.4.1
Multi - bed adsorption chiller
In order to maintain a quasi-continuous operation, a minimum of two sorption
beds are required, however, for performance reasons more than two beds are
often used. A theoretical study of a three - bed silica - gel adsorption chiller was
conducted, as reported in [28]. With a driving temperature of 80 ◦ C, a coolant
temperature of 30 ◦ C, and a chilled water temperature of 6 ◦ C, the adsorption
chiller achieves a COPc of 0.38. The three - bed cycle is shown to have more stable
performance than a two - bed cycle throughout an entire sorption / desorption
cycle, where a third ‘dormant’ bed can be used to provide a steady chilled water
output. The addition of a third bed represents an increase in capital cost, but will
extend the operational lifetime of the chiller. One could argue that the addition
of a fourth or fifth bed could provide a further increase in performance, but this
is a concept which has not been widely presented in the literature.
36
Chapter 2
Hybrid Trigeneration - Thermally Activated Heat Pumps
2.4.2
Multi - stage adsorption chiller
Of further interest are multi - stage adsorption chillers, which, like multi - bed adsorption chillers, feature more than the a single bed. However, unlike multi - bed
adsorption chillers, which have multiple beds arranged in a parallel formation,
multi - stage adsorption chillers have multiple beds arranged in a series formation. Multi - stage adsorption chillers allow for lower driving temperatures, as
demonstrated by Saha, Akisawa and Kashiwagi using a double - parallel and double - series silica gel and water chiller [71]. This can be advantageous in many
scenarios, most notably when the driving source is low temperature solar energy;
the system achieves a COPc of 0.36 with a driving temperature of just 55 ◦ C,
coolant temperature of 30 ◦ C and delivered chilled water temperature of around
7 ◦ C.
A triple - stage metal hydride adsorption chiller utilising different working fluids in each stage was analysed by Li et al. [72]. The sorbent of the high temperature bed was NiCL2 , and the sorbents of the medium and low temperature
bes were MnCL2 and SrCL2 , respectively; the sorbate in each bed was NH3 . The
system achieved a theoretical COPc of between 0.75 and 0.97, given a driving
temperature of 250 ◦ C. Compared to a cascade triple - stage absorption chiller,
the proposed adsorption chiller does not suffer from the corrosion or crystallisation problems but has a greatly diminished performance and potential operational
issues with the freezing of the metal hydrides. A thorough review of metal hydride
adsorption chiller systems is presented by Muthukumar and Groll [73].
2.5
R744 MVC Heat Pumps
The vast majority of heat pumps in use are electrically activated [2] and thus fall
into the MVC category. The MVC heat pump industry is continually in a state of
change due to the continually evolving political restraints regarding refrigerants.
In recent years, R744 has risen to the fore as a prime refrigerant [42, 74–77]. It
Chapter 2
37
Hybrid Trigeneration - Thermally Activated Heat Pumps
is also considered a prime candidate for future power conversion cycles [78–81],
in particular concentrated solar power applications [82–85]
The transcritical R744 heat pump, depicted in Figure 2.6, is ubiquitous due
to the requirement for a heat rejection temperature greater than 31 ◦ C in almost
all scenarios. The resulting high compressor outlet temperature gives rise to the
possibility for combined heating and cooling, and is explored by Sarkar, Bhattacharyya and Ram Gopal in the context of a dairy plant requiring chilling at
4 ◦ C and heating at 73 ◦ C [76, 86]. The analyses indicated that such a system is
well suited to high heating temperatures but modest cold temperatures [86].
Figure 2.6: Temperature - Entropy plot of a basic transcritical R744 heat pump
cycle
The replacement of an isothermal condenser with an anisothermal gas cooler
results in a significant temperature glide, which, combined with the close proximity to the critical point, gives rise to significant thermophysical fluid property
changes along the gas cooler. Therefore, simple heat exchanger models (e.g.
LMTD) cannot be used with R744. Figure 2.7 shows the variation of Cp as a
function of temperature for pressures close to the critical pressure, one notes the
pronounced peak in Cp when the when the critical pressure and temperature
intersect. At higher pressures, the peak is more subdued and is shifted towards
38
Chapter 2
Hybrid Trigeneration - Thermally Activated Heat Pumps
higher temperatures (the opposite being true for lower pressures). The significant
increase in Cp close to the critical point gives rise to pronounced pinch points in
heat exchangers.
Figure 2.7: Cp as a function of temperature for R744 at various pressures
Due to the high pressure of the cooled R744 exiting the gas cooler it is both
possible and desirable to extract work from the fluid using a work expansion
device or an ejector [42, 87, 88]. The addition of an ejector to the cycle increases
the COP by between 7 and 18 % relative to the basic cycle depicted in Figure
2.6. The ejector combines the high pressure supercritical / superheated vapour
R744 at the gas cooler outlet with the low pressure saturated vapour R744 at the
evaporator outlet , producing a fluid at an intermediate pressure in the saturated
region. The addition of the ejector reduces the exergetic losses associated with the
isenthalpic expansion of a valve, thus reducing the pressure ratio of the compressor
whilst not increasing the evaporation temperature, and increases the refrigerant
flow rate through the evaporator. Expanders can achieve a similar (∼15 % when
ηturb,is = 60% [42]), if not greater (∼40 % with a triple - stage expander and dual stage compressor [88]) boost in COP to ejectors.
Chapter 2
39
Hybrid Trigeneration - Thermally Activated Heat Pumps
2.6
Supercritical Carbon Dioxide Brayton Cycle
Although Supercritical Carbon Dioxide (S-CO2) was first investigated as a working fluid in the late 1960’s, Brayton cycles operating with S-CO2 is a relatively
new concept, with no commercial plant currently existing or currently being built.
Research into carbon dioxide Brayton cycles has experienced a significant increase
in recent years, due in part to a resurgence in the nuclear power industry. Carbon dioxide is a fluid which can be used to both cool the reactor core and as a
working fluid for the power cycle [79], thus potentially eliminating the need for
two separate fluid loops. When compared to the helium S-CO2 provides several
advantages over the the helium and steam indirect cycles. Thermoeconomic analyses have indicated that the S-CO2 cycle could provide cost savings of 30 % over
the steam indirect cycle and 15 % over the helium cycle [79].
The part-flow S-CO2 closed gas turbine cycle, proposed by [81], features an
additional compressor and heat recuperator as well as a part-flow valve to regulate
the flow to each compression stage. With a turbine inlet pressure and temperature
of 200 bar and 525 ◦ C respectively, a thermal to mechanical efficiency of 45 % is
achievable. It is estimated that due to the presence of a pinch point within the
recuperator, the conventional closed Brayton cycle suffers from a reduction in
conversion efficiency of around 7 % (absolute). The part flow cycle aims to reduce
the pinch effect, with 6 % (absolute) of the lost efficiency recoverable at optimal
conditions. Futher, it is suggested that the adiabatic efficiency of the compressor
and temperature effectiveness of the recuperator are not sensitive to efficiency
and, as such, the S-CO2 cycle is viable in distributed generation scenarios where
small capacities are required [81].
The recompression S-CO2 cycle with reheating proposed by [79] is a futher
evolution of the part-flow cycle, featuring the addition of a second reactor loop and
low pressure turbine. This cycle is theoretically capable of a maximum thermal to
40
Chapter 2
Hybrid Trigeneration - Thermally Activated Heat Pumps
mechanical conversion efficiency of 45.3 %, assuming a turbine inlet temperature
of 550 ◦ C (42.4 % for a turbine inlet temperature of 480 ◦ C) [79]. The addition
of reheating provides a maximum efficiency improvement of 3.5 % at optimum
conditions over the non-reheating case. Futher, it is noted that the “energetic
efficiency increases monotonically with increase in turbine inlet temperature” [79].
Whether the advantages proposed by the recompression S-CO2 cycle outweigh
the extra complexity and capital cost of the secondary turbine, compressor, heat
recuperator and reactor loop remain unknown.
A prototype test rig operating the S-CO2 simple Brayton cycle has been developed by Sandia National Laboratories [78]. The test rig is designed to produce
10 kW of mechanical power. The centrifugal compressor unit has a rotor of radius 19 mm, rotational speed of 75,000 rpm, pressure ratio of 1.8, mass flow rate
of 3.5 kg/s and exit blade height of 1.7 mm. Due to the extreme fluctuations in
thermophysical properties of carbon dioxide around the critical point, the density
of the fluid entering the compressor is between 65 - 80% the density of water in
liquid state. Coupled to the very low enthalpy change across the compressor, this
results in rather unique compressor design characteristics. The turbo - compressor
unit is noted as reaching energetic ‘break even’ when the turbine inlet temperature is 60 ◦ C [78]. Sandia National Laboratories are planning on subsequently
developing and testing a 10 MW system.
A prospect of utilising S-CO2 in Concentrated Solar Power (CSP) systems has
been considered in a number of different studies [82–84, 89]. The majority of operational parabolic trough CSP plants use synthetic oils as the heat transfer fluid,
which is used to produce steam either directly or indirectly (if thermal storage is
present) for the turbine. A smaller minority of parabolic trough CSP plants use
Direct Steam Generation (DSG), where the steam is produced and superheated
directly in the collector tubes. Both of these methods have a number of drawbacks, which can potentially be eradicated through the use of the S-CO2 Brayton
cycle. Synthetic oils are expensive, highly flammable, are limited to a maximum
Chapter 2
41
Hybrid Trigeneration - Thermally Activated Heat Pumps
temperature of around 400 ◦ C and over time degrade to their constituent parts
which reduces plant efficiency. DSG plants experience a phase change in the
collector tubes, which adds complexity to the design and control of the system.
Besides mitigating all of these drawbacks, a S-CO2 Brayton cycle has a much
smaller power block and can use forced air cooling far more effectively - condensing water requirements for standard CSP systems places significant restraints on
suitable sites. Indeed, the S-CO2 simple Brayton cycle has a theoretical efficiency
which surpasses that of the superheated steam Rankine cycle at turbine inlet temperatures above approximately 450 ◦ C and the supercritical steam Rankine cycle
at above approximately 550 ◦ C.
Possibly the greatest challenge facing S-CO2 parabolic trough CSP systems,
besides the closed Brayton cycle itself - which is being actively researched for
next generation nuclear plants, is the high pressure S-CO2 piping system running
through the collector fields or power tower receiver. The very high pressure fluid
combined with the distance that the fluid must travel brings technical challenges
with the collectors tubes themselves, which ordinarily carry high density oil and
low pressure (not low density fluid at very high pressure). The next biggest
challenge is the integration of thermal storage technologies, e.g. two-tank molten
salt systems versus packed-bed thermocline systems [82]. High pressure thermal
storage vessels require very thick walls and are limited in diameter, a low pressure
molton salt thermal storage vessel could be used but this would require a S-CO2 to - molten salt heat exchanger – these do not currently exist [82]. A prototype
low temperature evacuated tube S-CO2 solar collector system has been created
and tested [83], where comparisons between R744 and other refrigerants are made
in the context of CSP. The prototype evacuated tubes consist of an outer glass
envelope, and inner glass selectively coated tube, inside of which is an aluminium
fin housing two small steel tubes which carry the pressurised carbon dioxide at
120 bar. The collectors were capable of heating the CO2 to 165 ◦ C, which the
author predicts would correspond to a power generation efficiency of 25 % and
42
Chapter 2
Hybrid Trigeneration - Thermally Activated Heat Pumps
a heat recovery efficiency of 65 % - this appears optimistic, and will have to be
verified.
A S-CO2 CSP system using a central receiver tower has been proposed by
Chacartegui et al. [89]. Three cycles are proposed: a basic closed S-CO2 Brayton cycle, a part-flow recompression closed S-CO2 Brayton cycle, and a topping
basic closed S-CO2 Brayton cycle with a bottoming ORC. Central receiver tower
systems hold many advantages of parabolic trough systems when S-CO2 is the
working fluid. The high pressure associated with S-CO2 is best if kept refined to
a single location for cost, performance and safety reasons, as would be the case in
a central receiver tower. In a parabolic trough system the highly pressurised fluid
would be sent around the entire field, which would require no doubt expensive
receiver tubes, and result in notable pressure and heat losses around the field,
as well as present an increased safety hazard. Further, the concentration of the
incident radiation is considerably higher in central receiver towers than it is for
parabolic troughs - the implication being that central receiver towers are capable
of producing extremely high temperatures at the focal point. Higher temperatures allow for considerably higher thermal to mechanical conversion efficiencies
to be achieved. At a turbine inlet temperature of 450 ◦ C a supercritical steam
Rankine cycle is capable of a conversion efficiency of approximately 43 % whilst
a S-CO2 closed Brayton cycle is capable of a theoretical conversion efficiency
of approximately 40 %, at a turbine inlet temperature of 950 ◦ C a S-CO2 closed
Brayton cycle is capable of a theoretical conversion efficiency of approximately
55 % [82], which is well above the maximum conversion efficiency and turbine
inlet temperature possible in a supercritical steam Rankine cycle.
2.7
Conclusions
This Chapter provided a general review of literature relating to thermally activated heat pumps for integration into trigeneration systems. A review of refrigerants, single component and multi component, zeotropic and azeotropic is given.
Chapter 2
43
Hybrid Trigeneration - Thermally Activated Heat Pumps
Natural refrigerants such as R744 and R717 have been identified as featuring
heavily in research over the past decade, owing to a continued move towards low
GWP and low ODP refrigerants, driven in part by the Montreal protocol. The
use of highly zeotropic refrigerant mixtures, such as LiBr - H2 O and H2 O - NH3 ,
in thermally activated absorption heat pumps has been discussed, and a number
of concepts and recent advances to increase the COP have been presented. These
include: multi - stage / multi - effect cycles, generator - absorber exchanger cycles
and vapour recompression cycles. Triple - stage triple - effect absorption chillers
are capable of achieving a COPc of approximately 1.6, but suffer from corrosion
in the upper generator due to the high temperatures [48]. The COP is both a
function of the upper generation temperature, chilled water temperature, load
conditions and the flow arrangement. It is demonstrated by Kaita that the parallel flow configuration achieves the highest COP for all absorber and generator
conditions considered [47].
A brief review of multi - bed and multi - stage adsorption heat pumps, and sorbent and sorbate pairs demonstrates that COP of this technology is still far below
that of absorption heat pumps. The Chapter closes with a review of R744 Mechanical Vapour Compression (MVC) heat pumps along with a review advances
Supercritical Carbon Dioxide (S-CO2) power cycle concepts. It is evident from
the review of literature that there is a small scope for innovations in multi - stage
multi - effect absorption heat pump cycle design, and a significant scope for innovations in R744 heat pump cycle design. Therefore, the analysis of multi - stage
multi - effect absorption heat pump cycle and R744 heat pump cycle concepts
form the key contributions to research presented within this Thesis.
An overview of some common heat pump cycles found within the literature is
contained within Table 2.1.
44
Chapter 2
Chapter 2
Activated carbon - CO2
Zeolite - water
Silica gel - water
NiCl2 - NH3
MnCl2 - NH3
MnCl2 / NiCl2 - NH3
SrCl2 / FeCl2 - NH3
NiCl2 / SrCl2 - NH3
MnCl2 / SrCl2 - NH3
NiCl2 - MnCl2 - SrCl2 - NH3
H2 O-NH3
LiBr - H2 O
H2 O-NH3
LiBr - H2 O
H2 O - NH3
LiBr - H2 O
H2 O:NaOH - NH3
Adsorption
Adsorption
Adsorption
Adsorption
Adsorption
Adsorption
Adsorption
Adsorption
Adsorption
Adsorption
Absorption / Compression
Absorption / Ejector
Absorption
Absorption
Absorption
Absorption
Absorption
Three bed, non-regenerative. Hot water inlet T = 85 ◦ C,
coolant water inlet T = 30 ◦ C, chilled water inlet / outlet
T = 14 / 6 ◦ C
Double-bed with heat regeneration. Ambient
temperatures ranged from 27 ◦ C to 38 ◦ C, Specific Cooling
Production (SCP) was 36 W/kg and could be increased to
106 W/kg whilst sacrificing COP (0.7)
Triple-bed. Designed for low driving temperature
operation
Single-stage. Ideal COP
Single-stage. Ideal COP
Double-stage. Evaporation temperature of -10 ◦ C,
condensation temperature of 40 ◦ C
Double-stage. Evaporation temperature of -10 ◦ C,
condensation temperature of 40 ◦ C
Double-stage. Ideal COP
Doule-stage. Ideal COP
Triple-stage. Ideal COP
GAXAC cycle
Single stage absorption cycle utilising a steam ejector to
simulate a GAXAC-style cycle
Double stage triple effect ‘Kangaroo’ cycle. Tamb ranging
from 28 ◦ C to 35 ◦ C
Theoretical study of a 16 kW steam driven double effect
chiller. Performance is heavily dependant upon ambient
conditions
GAX cycle utilising four operational modes. COPc quoted
is for Tamb of 40 ◦ C
Three stage triple effect chiller in parallel or series flow
configuration. Generator temperature around 200 ◦ C
Single stage cycle. Addition of NaOH increases COPc by
around 15 %
Comment
Table 2.1: List of various absorption and adsorption cycles and relevant data
Silica gel - water
Working Fluids
Adsorption
Type of Technology
0.6 [94]
1.7 [54]
0.55, 0.85, 0.9 [59]
1.03 [55]
1.15 - 1.55 [56]
0.75 [72]
0.95 [72]
1.10 [72]
1.185 [61]
1.013 [93]
0.7 [92]
0.38 [72]
0.45 [72]
0.6 [92]
0.43 [90]
0.083 [91]
1.0 - 1.6 [92]
0.7 [90]
Typical range of COP
Hybrid Trigeneration - Thermally Activated Heat Pumps
45
Chapter 3
Absorption Heat Pump Cycle
Concepts
46
Hybrid Trigeneration - Thermally Activated Heat Pumps
3.1
3.1.1
Absorption Heat Pump Cycle Concepts
Introduction
Absorption heat pumps are typically used in CCHP systems (discussed in §2.2) as
chillers to produce chilled water for space cooling, dehumidification and process
chilled water. As mentioned previously, the technoeconomic feasibility of the such
systems is generally very dependent upon climatic conditions. The most common
type of absorption heat pumps, LiBr - H2 O and H2 O - NH3 , achieve COPc of ∼0.7
in a single - stage configuration, ∼1.2 in a double - stage configuration, and ∼1.7
in a triple - stage configuration (LiBr - H2 O only). In this Chapter, a number of
high performance and hybrid absorption heat pump cycles are conceptualised,
analysed and discussed.
These cycles are modeled and simulated using Engineering Equation Solver
(EES)1 . It is assumed that all processes are adiabatic and reversible, with the exception of the pump, which is modeled using an appropriate pumping efficiency.
Ambient temperature is specified as 30 ◦ C. Evaporator refrigerant inlet temperature is specified as 4 ◦ C, which, given a standard chilled water flow / return of
6 / 12 ◦ C, implies a ∆Tmin in the evaporator of 2 ◦ C. In all other HXs, ∆Tmin is
specified at 4 ◦ C.
3.1.2
H2 O - NH3 Kangaroo Cycle
The Kangaroo Cycle, as mentioned in §2.3.1, is a H2 O - NH3 multiple effect cycle,
where each stage is independent, but thermally coupled to the others. Six cycles
are presented in this chapter: single - stage / single - effect V.1 & V.2, double stage / triple - effect V.1 & V.2 and triple - stage / quintuple - effect V.1 & V.2.
Rectifier The purpose of the rectifier is to increase refrigerant mass fraction (x)
from that which leaves the generator to that which is deemed sufficient to negate
1
http://www.fchart.com/ees/
Chapter 3
47
Hybrid Trigeneration - Thermally Activated Heat Pumps
an unacceptable thermal gradient (temperature glide) in the evaporator. Typically a mass fraction (x) 0.995 kg/kg is deemed acceptable [51, 61, 95, 96], and
this research indicates that this value results in a temperature glide of between 5
& 6 ◦ C in the evaporator. Higher values of x require significantly increased plant
size, can lead to condensation of refrigerant in the rectifier (hence, reduction of
refrigerant mass flow), and provide minimal performance gain. Indeed, temperature glide in evaporators is rarely an issue, provided a direct exchange evaporator
is specified. Without rectification, the evaporator temperature glide is in excess
of 25 ◦ C, which is highly undesirable.
The rectifier can take a number of physical forms, but the main design choice
is how heat is removed from the low refrigerant mass fraction generator outlet
vapour, external heat rejection or internal heat rejection. With external heat
rejection, the rectifier simply rejects heat to an external sink, potentially via
dry cooling or wet cooling, the former of which would create a electrical load
on the system, and the latter of which would require a continuous water source
or, if recirculation is used, a coolant circuit and associated subsystems. Internal
heat rejection has none of the drawbacks of external, and actually has a positive
impact upon COP by reducing the generator’s requirement for heat. The mid
quality refrigerant / sorbent solution exiting the absorber is either passed through
the solution HX and then the rectifier (V.1) or passed through the rectifier and
then the solution HX (V.2). Recuperation via the rectifier accounts for around
13 % of internally recuperated heat.
Based upon Figure 3.1 the rectifier is modeled under the constraints of equations 3.1, 3.2 and 3.3.
48
m1 = m2 + m3
(3.1)
m1 x1 = m2 x2=xf + m3 x3=xg
(3.2)
Chapter 3
Hybrid Trigeneration - Thermally Activated Heat Pumps
2 O+N H3 mixture
2 O+N H3 mixture
h10 (3.3)
h9 − mH
m1 h1 = m2 h2=hf + m3 x3=hg + mH
10
9
Generator The generator produces refrigerant vapour and weak sorbent solution via an endothermic reaction, using an input of strong sorbent solution from
the absorber and weak sorbent solution returned from the rectifier. Depending
on the conditions, the refrigerant vapour has a value of x of around 0.98, with a
quality (q) slightly below the saturated vapour line. The weak sorbent solution
outlet is the highest temperature flow of the cycle, and has a typical mass fraction
of around 0.38.
In the single - stage cycle the generator, or in the case of multiple - stage cycles
the upper generator, is driven by an external heat source, which is likely to be a
heat transfer fluid which passes through the generator and is itself heated by an
external source. In the lower generators of multiple - stage cycles the generators
are driven either by the condenser or absorber from the upper cycle. In the case
of the former the condenser runs directly through the generator, whilst in the case
of the latter an intermediate heat transfer medium is required. Heat pipes could
be used as an effective way of transferring the heat from absorber to generator.
The minimum ∆T of upper condenser / absorber temperature to lower generator
is 10 ◦ C.
Based upon Figure 3.1 the generator is modeled under the constraints of
equations 3.7, 3.8 and 3.9.
Chapter 3
m1 = m2 + m10 − m11
(3.4)
m1 h1 = m2 h1 + m10 h10 − m11 h11 + m21 h21 − m22 h22
(3.5)
m1 x1 = m2 x2=xf + m10 h10 − m10 h10
(3.6)
49
Hybrid Trigeneration - Thermally Activated Heat Pumps
Absorber The absorber produces strong sorbent solution via an exothermic
reaction, using an input of weak sorbent solution from the generator and refrigerant vapour from the evaporator. The absorber is modeled to produce a ∆x
of 0.1, which typically equates to a refrigerant to solution mass flow rate ratio
(Ψ) of around 0.19. An increase in Ψ would decrease the pump work, but would
require a larger absorber; given that the pump work is less than 1 % of the energy
supplied to the cycle, increasing Ψ is typically not desirable.
m7 = m6 + m13
(3.7)
m7 h7 = m6 h6 + m13 h13
(3.8)
m7 x7 = m6 x6 + m6 h10 − m10 h10
(3.9)
Single - Stage / Single - Effect
The single - stage / single - effect H2 O - NH3 absorption heat V.1 and V.2 are depicted in Figures 3.1 and 3.2; Tables 3.1 and 3.2 show corresponding data for
the state points under the operational criteria specified prior. The generation
temperature is allowed to ‘float’ in order to obtain the highest possible value for
COP.
V.1 This cycle, with the solution HX situated at the solution pump outlet,
achieves a theoretical maximum COPc 2 of 0.725, given a generation temperature
of 105 ◦ C. This implies a heating fluid outlet temperature of approximately 115 ◦ C;
the inlet temperature depending upon the Cp and ṁ of the fluid. The ratio of the
mass flow rate of refrigerant in the evaporator and the mass flow rate of solution
exiting the absorber (Ψ) is 0.191 kg/kg. The ratio of generator heat input to
2
COPc is defined as the ratio of thermal energy input into the cycle to the production of
useful coolth within the evaporator(s)
50
Chapter 3
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 3.1: 1 - Stage / 1 - Effect H2 O - NH3 V.1 absorption heat pump cycle process
flow diagram
pump work (Ω) is 99.4.
State Point
h (kJ/kg)
P (bar)
m (kg/s)
T ( ◦ C)
s (kJ/kgK)
x (kg/kg)
1
2
3
4
5
6
7
8
9
10
11
12
13
1446
147.1
1299
185.7
185.7
1281
-38.94
-36.51
176.0
207.1
249.7
-3.259
-3.259
15.13
15.13
15.13
15.13
5.155
5.155
5.155
15.13
15.13
15.13
15.13
15.13
5.155
1
0.03494
0.9651
0.9651
0.9651
0.9651
6.031
6.031
6.031
6.031
5.066
5.066
5.066
85.3
85.3
44.8
39.0
5.00
10.5
44.8
45.1
86.9
88.6
105.3
49.1
49.3
4.64
1.06
4.20
0.642
0.679
4.60
0.508
0.512
1.14
1.22
1.32
0.595
0.599
0.9812
0.4745
0.9995
0.9995
0.9995
0.9995
0.4745
0.4745
0.4745
0.4745
0.3745
0.3745
0.3745
Table 3.1: State properties - 1 - Stage / 1 - Effect H2 O - NH3 V.1 cycle
V.2 This cycle, with the rectifier is situated at the solution pump outlet and
solution HX is situated following the rectifier, achieves a theoretical maximum
COPc of 0.680, given a generation temperature of 98 ◦ C; implying a heating fluid
outlet temperature of 108 ◦ C. Ψ is 0.205 kg/kg and Ω is 111.4.
There is a marginal increase in both Ψ and Ω when progressing from V.1 to
Chapter 3
51
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 3.2: 1 - stage / 1 - effect H2 O - NH3 V.2 absorption heat pump cycle process
flow diagram
V.2. This indicates that V.2 produces more refrigerant vapour in the generator
than V.1, although the COPc is marginally lower. The generator temperature
in V.2 is lower than V.1, which is the cause of the marginal decrease in COPc
despite the higher Ψ.
The ratio of recuperated heat from the solution HX to the rectifer (Φ) is 6.83
for V.1, whilst for V.2 it is 6.76. The similar values for Φ indicate that the order
of recuperation bears little effect upon the relative source of the recuperated heat;
the predominant factor being the heat available, and the refrigerant mass fraction,
x, of the fluid entering the rectifier.
The proportion of generator heat input which is internally recuperated (Θ)
is 0.502 for V.1 and 0.432 for V.2. The substantially lower Φ in V.2 is the
predominant cause of the reduction in COPc ; a reduction in internally recuperated
heat, due to the exergy of the weak sorbent solution entering the solution HX
52
Chapter 3
Hybrid Trigeneration - Thermally Activated Heat Pumps
State Point
h (kJ/kg)
P (bar)
m (kg/s)
T ( ◦ C)
s (kJ/kgK)
x (kg/kg)
1
2
3
4
5
6
7
8
9
10
11
12
13
1425
119.3
1299
185.7
185.7
1281
-64.64
-62.18
-5.36
146.2
207.9
-10.83
-10.83
15.13
15.13
15.13
15.13
5.155
5.155
5.155
15.13
15.13
15.13
15.13
15.13
5.155
1.000
0.02348
0.9765
0.9765
0.9765
0.9765
5.745
5.745
5.745
5.745
4.768
4.768
4.768
78.9
78.9
44.8
39.0
5.00
10.5
39.0
39.4
45.3
80.2
97.6
49.3
49.5
4.58
0.973
4.20
0.642
0.679
4.60
0.423
0.427
0.512
1.05
1.22
0.585
0.588
0.9880
0.5112
0.9995
0.9995
0.9995
0.9995
0.5112
0.5112
0.5112
0.5112
0.4112
0.4112
0.4112
Table 3.2: State properties - 1 - Stage / 1 - Effect H2 O - NH3 V.2 cycle
hot side, means that more heat must be supplied from external sources to the
generator. The order of rectifier and solution HX in V.2 results in a decreased
2nd Law efficiency and unnecessary exergy destruction. It also leads to a hitting
of the minimum ∆Tmin 3 at the cold end of the solution HX, indicating that the
potential level of heat recuperation is restricted; whilst in V.1 the equivalent
temperature pinch is 4.19 ◦ C, the cycle is not restricted by the ∆Tmin .
Double - Stage / Triple - Effect
The double - stage / triple - effect H2 O - NH3 absorption heat pump V.1 and V.2
are depicted in Figures 3.3 and 3.4; Tables 3.3 and 3.4 show corresponding data
for the state points under the operational criteria specified prior.
V.1 This cycle features two separate but thermally coupled single - stage cycles,
depicted in Figure 3.3. The upper cycle (Figure 3.3a), is thermally activated by
an external source and much the same as the single - stage V1 cycle depicted in
Figure 3.1. The primary difference being an increase in temperature of the external heat source (Point 51), resulting in increased temperatures and pressures
throughout the cycle. The upper condenser and upper absorber are both cooled
by the two generators in the lower cycle (Figure 3.3b). The specific means by
3
∆Tmin refers to the minimum pinch temperature, or temperature difference between the
two fluid flows in a heat exchanger
Chapter 3
53
Hybrid Trigeneration - Thermally Activated Heat Pumps
which heat is transferred is not included in this analysis; direct heat exchange is
possible, although indirect exchange via an active system such as a circulated heat
transfer fluid, or a passive system such as heat pipes would simplify system design
and control. In order to accommodate any heat transfer method, a minimum temperature pinch between the lower generator(s) and upper absorber / condenser of
10 ◦ C has been specified. The fluid entering both the upper and lowers evaporator
is throttled to the same pressure, thus both produce a cooling effect of the same
temperature, but not necessarily the same magnitude.
(a) Upper Cycle
(b) Lower Cycle
Figure 3.3: 2 - stage / 3 - effect H2 O - NH3 V.1 absorption heat pump cycle process
flow diagram
This cycle achieves a COPc of 1.765, given an upper generation temperature
54
Chapter 3
Hybrid Trigeneration - Thermally Activated Heat Pumps
of 224 ◦ C; thus implying a heating fluid outlet temperature of approximately
234 ◦ C. The upper cycle has a Ψ of 0.131 kg/kg, and the lower cycle has a Ψ of
0.217 kg/kg.
State Point
h (kJ/kg)
P (bar)
m (kg/s)
T ( ◦ C)
s (kJ/kgK)
x (kg/kg)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
1957
755
1247
434
434
1280
228
238
764
939
908
313
313
1399
74
1298
161
161
1280
-85
-83
94
116
152
-64
-64
47.52
47.52
47.52
47.52
4.97
4.97
4.97
47.52
47.52
47.52
47.52
47.52
4.97
13.12
13.12
13.12
13.12
4.97
4.97
4.97
13.12
13.12
13.12
13.12
13.12
4.97
0.602
0.239
0.363
0.363
0.363
0.363
3.127
3.127
3.127
3.127
2.763
2.763
2.763
0.630
0.011
0.619
0.619
0.619
0.619
3.478
3.478
3.478
3.478
2.859
2.859
2.859
200.5
200.5
87.3
86.0
4.0
9.6
90.1
91.8
201.0
210.2
224.4
95.74
96.4
68.7
68.7
40.9
34.0
4.0
9.6
34.0
34.3
69.6
70.6
86.1
38.3
38.5
5.241
2.428
3.675
1.363
1.571
4.610
1.182
1.195
2.449
2.811
2.661
1.285
1.297
4.563
0.838
4.563
0.564
0.592
4.610
0.350
0.353
0.896
0.960
1.047
0.423
0.426
0.6975
0.2386
0.9995
0.9995
0.9995
0.9995
0.2386
0.2386
0.2386
0.2386
0.1386
0.1386
0.1386
0.9915
0.5377
0.9995
0.9995
0.9995
0.9995
0.5377
0.5377
0.5377
0.5377
0.4377
0.4377
0.4377
Table 3.3: State properties - 2 - Stage / 3 - Effect H2 O - NH3 V.1 cycle
V.2 This cycle is similar to V.1, but the position of the rectifier and solution
HX in both the upper and lower cycles has been altered. The rectifiers are now
situated following the absorber strong sorbent solution outlet, whilst the solution
heat exchanger follows on from the rectifier. A schematic of the double - effect
triple -effect cycle is shown in Figure 3.4.
This cycle achieves a COPc of 1.422, given an upper generator temperature
of 187 ◦ C; thus implying a heating fluid outlet temperature of approximately
197 ◦ C. The upper cycle has a Ψ of 0.147 kg/kg, and the lower cycle has a Ψ of
0.217 kg/kg. When compared to the values for Ψ achieved by V.1, this represents
a small increase in generation efficiency in the upper cycle, whilst the generation
Chapter 3
55
Hybrid Trigeneration - Thermally Activated Heat Pumps
(a) Upper Cycle
(b) Lower Cycle
Figure 3.4: 2 - stage / 3 - effect H2 O - NH3 V.2 absorption heat pump cycle process
flow diagram
efficiency in the lower cycle remains constant.
In V.1, the evaporator in the upper cycle produces 30.7 % of the total chilling
effect, whilst in V.2 the evaporator produces 28.4 % of the total cooling effect;
56
Chapter 3
Hybrid Trigeneration - Thermally Activated Heat Pumps
State Point
h (kJ/kg)
P (bar)
m (kg/s)
T ( ◦ C)
s (kJ/kgK)
x (kg/kg)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
1738
547
1274
360
360
1280
117
124
229
580
693
290
290
1402
74
1298
161
161
1280
-85
-83
-61
98
152
-42
-42
35.37
35.37
35.37
35.37
4.97
4.97
4.97
35.37
35.37
35.37
35.37
35.37
4.97
13.12
13.12
13.12
13.12
4.97
4.97
4.97
13.12
13.12
13.12
13.12
13.12
4.97
0.403
0.091
0.309
0.309
0.309
0.309
2.411
2.411
2.411
2.411
2.102
2.102
2.102
0.650
0.010
0.640
0.640
0.640
0.640
3.594
3.594
3.594
3.594
2.955
2.955
2.955
163.7
163.7
74.5
72.6
4.0
9.6
72.6
73.8
97.3
165.8
187.4
101.3
96.6
68.6
68.6
40.9
34.0
4.0
9.6
34.0
34.4
39.2
69.7
86.1
43.2
43.3
4.960
2.019
3.828
1.161
1.307
4.610
0.929
0.939
1.231
2.096
2.286
1.319
1.329
4.571
0.838
4.254
0.564
0.592
4.610
0.350
0.353
0.425
0.909
1.067
0.492
0.495
0.8440
0.3189
0.9995
0.9995
0.9995
0.9995
0.3189
0.3189
0.3189
0.3189
0.2189
0.2189
0.2189
0.9924
0.5377
0.9995
0.9995
0.9995
0.9995
0.5377
0.5377
0.5377
0.5377
0.4377
0.4377
0.4377
Table 3.4: State properties - 2 - Stage / 3 - Effect H2 O - NH3 V.2 cycle
the lower evaporators produce 69.3 % and 71.6 %, respectively. This difference
in cooling effect contributions between V.1 and V.2 is contrary to the difference
in generation efficiency. The mass flow rates of refrigerant, sorbent and refrigerant / sorbent solution in the upper cycle are between 20 and 30 % higher in V.1,
relative to V.2. The mass flow rates in the lower cycle in both V.1 and V.2 are
very similar. The higher generation efficiency and reduced chilling effect in V.2
are likely the reason for the reduced mass flow rates.
The values for Φ for V.1 are 0.786 and 0.433 in the upper and lower cycles,
respectively. The same values for V.2 are 0.609 and 0.390. The difference in Φ
between V.1 and V.2 is apparent due to the reduction in exergetic efficiency resulting from the ordering of the rectifier and solution HX in the heat recuperation
process. The difference between the temperature of the strong solution entering
the generator and the generation temperature (T11 − T10 and T24 − T23 in Table
Chapter 3
57
Hybrid Trigeneration - Thermally Activated Heat Pumps
3.3 and 3.4), is considerably less in V.1 than V.2 in both the upper and lower
cycles. This also occurs as a direct result of the reduction in exergetic efficiency,
the result of which is a requirement for more externally supplied thermal energy
(observed by a decrease in Φ); this reduces the COP.
Overall, the ‘Double - Stage / Triple - Effect’ V.1 cycle achieves a significantly
higher COPc than V.2, with a 24 % increase resulting from the change in the
positions of the solution HX and rectifier.
Triple - Stage / Quintuple - Effect
The triple - stage / quintuple - effect absorption heat pump V.1 and V.2 are depected in Figures 3.5 and 3.6, respectively. These cycles are an evolution of the
double - stage / triple - effect absorption heat pumps V.1 and V.2 described previously. An additional H2 O - NH3 absorption heat pump cycle loop is added around
the ‘outside’ of the previous upper loop; such that the upper loop becomes the
middle loop.
Now, the upper loop activates the middle loop by passing thermal energy
from the upper condenser and upper absorber to the middle generator. The
middle loop activates the lower loop by passing thermal energy from the middle
condenser and middle absorber to the lower generator.
The addition of a outer loop to the single - stage / single - effect H2 O - NH3 absorption chiller increased the COPc by between 0.8 and 1.0 to 1.422 and 1.765
for V.2 and V.1, respectively. It could therefore be expected that the addition of
a further outer loop would yield another significant boost in performance. The
double double - stage / triple - effect V.1 cycle has an upper generator temperature
of 224 ◦ C, whilst the V.2. cycle has an upper generator temperature of 187 ◦ C;
this is already quite high for an ammonia mixture. Neither simulation models
of the triple - stage / quintuple - effect heat pumps depicted in Figures 3.5 and 3.6
produce a solution; the thermodynamic properties of the non-azeotropic water
and ammonia mixture is outside of solvable range in the upper generator. Low58
Chapter 3
Hybrid Trigeneration - Thermally Activated Heat Pumps
(a) Upper Cycle
(b) Lower Cycle
(c) Lower Cycle
Figure 3.5: 3 - stage / 5 - effect H2 O - NH3 V.1 absorption heat pump cycle process
flow diagram
ering the upper generator temperature by ‘compressing’ the temperature range
across the loops (i.e. reducing pinch temperatures, altering pressures, isentropic
efficiencies) does not yield a solution. Therefore, the triple - stage / quintuple effect H2 O - NH3 ‘Kangaroo’ absorption heat pump is not a worthy candidate for
further investigation.
Chapter 3
59
Hybrid Trigeneration - Thermally Activated Heat Pumps
(a) Upper Cycle
(b) Lower Cycle
(c) Lower Cycle
Figure 3.6: 3 - stage / 5 - effect H2 O - NH3 V.2 absorption heat pump cycle process
flow diagram
60
Chapter 3
Hybrid Trigeneration - Thermally Activated Heat Pumps
3.1.3
LiBr - H2 O Compression - Absorption Cycle
The LiBr - H2 O compression / absorption heat pump cycle is a variant of the conventional LiBr - H2 O absorption heat pump cycle, and features a low pressure
steam compressor. The compression / absorption heat pump cycle is a hybrid
heat pump cycle; both thermal and mechanical energy inputs are required. The
proportion of energy input which is thermal and mechanical has a profound effect
on the performance characteristics of the cycle. The addition of a compressor into
the cycle can produce a range enhancements over the standard absorption cycle,
including:
• Decreasing the generation temperature
• Decreasing the evaporator temperature
• Increasing the absorber pressure
• Increasing the COP
In general, only one of these enhancements can be realised through the addition of the compressor; this is dependent upon the location of the compressor
in the cycle and the pressures selected at the relevant points. It is possible to
produce a blend of two of these enhancements, but this is not typically desirable.
A brief introduction to each enhancement is given below, which this is followed
by an analysis of compression - absorption heat pump cycles and a discussion of
their technical feasibility.
Decreasing the generation temperature Lowering the generation temperature is achieved by lowering the fluid pressure within the generator. This may be
desirable if the primary heat source is low temperature or has an efficiency which
increases at lower temperatures. Consequently, the refrigerant vapour exiting the
generator and entering the condenser is at a lower temperature, and the condenser
may be unable to reject heat to atmosphere; i.e condenser exergy is reduced to
Chapter 3
61
Hybrid Trigeneration - Thermally Activated Heat Pumps
zero. The knock - on effect through the rest of the cycle is a drastic lowering
of COP. However, by slightly re-pressurising the refrigerant vapour exiting the
generator prior to it entering the condenser, the potential for heat rejection in
the condenser can be maintained; thus the COP can remain constant. As will be
shown later, the mechanical work required to re-pressurise the refrigerant vapour
is relatively small.
Decreasing the evaporator temperature Lowering the evaporation pressure will reduce the evaporation temperature, thus enabling the cycle to produce
chilled water at a lower temperature (n.b. this may be for process chilled water,
rather than building space cooling chilled water). Ordinarily, lowering the evaporation pressure, and thus absorber pressure, would decrease the COPc of the
cycle. However, if the water vapour exiting the evaporator can be re-compressed
to a higher pressure, the absorber can operate at a higher pressure, thus enabling
the COPc of the cycle to remain at consistent at lower chilled water temperatures.
Increasing the absorber pressure Raising the absorber pressure can be desirable for a number of reasons:
• To increase the absorption temperature. In general the waste heat from the
absorber is of a temperature unusable for processes and hot water. Increasing the pressure in the absorber allows a higher operating temperature to
be achieved, whilst maintaining a similar operating characteristics. Raising
the operating temperature in the absorber can enable it to produce Very
Low Temperature Hot Water (VLTHW) for space heating and Domestic
Hot Water (DHW) in buildings, as well as low temperature industrial processes. This also reduces the requirement for, and possibly eliminates, the
need for a cooling tower; dry air coolers may be sufficient for cooling the
condenser.
62
Chapter 3
Hybrid Trigeneration - Thermally Activated Heat Pumps
• To decrease the size of the absorber A higher pressure within the absorber
results in a more efficient absorption process; the rate of refrigerant absorption by the sorbent is a function of pressure and temperature, and increases
at higher pressures and lower temperatures. As one of the largest components of an absorption chiller, reducing the size of the absorber may results
in a more compact system.
Increasing the COP If a compressor is added at any point in the cycle, and
neither of the previous three enhancements are realised, it is possible to increase
the cycle COP. The increase in COP is dependant upon a number of factors,
which will be discussed in the analysis.
When discussing the COP of a compression - absorption heat pump cycle, it
is necessary to distinguish between the COP relative to the thermal energy input
(COPth ), and the COP relative to the mechanical energy input (COPm ), such
that:
COPth =
Qheating/cooling,out
Qgenerator
(3.10)
COPm =
Qheating/cooling,out
Wcompressor
(3.11)
Single - Stage / Single - Effect
A single stage compression - absorption LiBr - H2 O heat pump with the steam
compressor located at the evaporator outlet is shown in Figure 3.7. As with
previous analyses, Engineering Equation Solver has been used to simulate this
cycle under a range of performance criteria.
The cycle is analysed using the same ambient criteria as used in the H2 O - NH3
Kangaroo cycle detailed in the previous Section.
The effect of the compressor pressure ratio (rp ) on the COPc,th (i.e. coefficient of performance relative to the heating work in) at four different generator
Chapter 3
63
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 3.7: 1 - Stage / 1 - Effect LiBr - H2 O compression - absorption heat pump
cycle process flow diagram
Figure 3.8: 1 - Stage / 1 - Effect LiBr - H2 O compression - absorption heat pump
cycle: COPc,th vs. rp
temperatures is shown in Figure 3.8. It can be observed that when the compression ratio is increased, the COPc,th increases, and this effect becomes far more
profound at lower generation temperatures. With a generation temperature of
80 ◦ C, the COPc,th decreases sharply when the compression ratio is below 2.0,
64
Chapter 3
Hybrid Trigeneration - Thermally Activated Heat Pumps
and tends towards zero when the compression ratio is around 1.2. As generation
temperature increases, the relationship between COPc,th and compression ratio
tends to a linear relationship. Indeed, with a generation temperature of 90 ◦ C, a
near linear relationship is observed. Of further note is that at higher generation
temperatures, the gradient of the line reduces (i.e. increasing compression ratio
has less of an effect on COPc,th ).
Figure 3.9: 1 - Stage / 1 - Effect LiBr - H2 O compression - absorption heat pump
cycle: COPc,th vs. Tgen
A similar graph comparing COPc,th to generation temperature is shown in
Figure 3.9. The five curves denote five compression ratios under consideration,
the compression ratio of 1.00 indicating a conventional absorption chiller cycle.
As the compression ratio is increased, the curves appear to shift to the left and
up; implying that the minimum generation temperature at which the cycle works
decreases, and the COPc,th increases for a given generation temperature. This is
essentially the same effect observed and noted from Figure 3.8.
A graph of the cooling output relative to the mechanical work input,COPc,m ,
is given in Figure 3.10. The relationship between COPc,m and rp is such that
Chapter 3
65
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 3.10: 1 - Stage / 1 - Effect LiBr - H2 O compression - absorption heat pump
cycle: COPc,m vs. rp
COPc,m tends towards a very high value when rp is below 1.4. The minimum
value for COPc,m under the specified rp range is approximately 10. The COPc,m
for a MVC is typically around 5, whilst the COPc,m for a conventional absorption
chiller (i.e. the pump) may be around 100.
Double - Stage / Double - Effect
A double stage double effect compression - absorption LiBr - H2 O heat pump with
the steam compressor located at the evaporator outlet is shown in Figure 3.11.
In this cycle, heat is passed from the upper condenser directly to the lower generator via condensation of the refrigerant vapour (state point 17) within the lower
generator; indicated by the dotted line. Besides the the addition of the compressor, this cycle is the same as a conventional double stage absorption chiller, used
prevalently in commercial and industrial chilling processes.
The effect of rp on the COPc,th at four different generator temperatures is
shown in Figure 3.12. The generation temperatures simulated range between
66
Chapter 3
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 3.11: 2 - Stage / 2 - Effect LiBr - H2 O compression - absorption heat pump
cycle process flow diagram
100 ◦ C and 160 ◦ C (20 - 50 ◦ C greater than the previous single stage cycle). In
general, it can be observed that when the rp is increased, the COPc,th increases,
and this effect becomes far more profound at lower generation temperatures (i.e.
this characteristic is as observed in the single stage compression - absorption cycle). The relationship between COPc,th and rp becomes less profound at higher
generation temperatures. Further, it can be noted that this relationship tends to
become more linear at higher generation temperatures, whilst at lower generation
temperatures the effect is more parabolic.
It is evident from Figure 3.12 that when rp is between 1.0 and 1.2 the highest
COPc,th is achieved with a generation temperature of 160 ◦ C. When rp is between
1.2 and 1.5, the highest COPc,th is achieved with a generation temperature of
140 ◦ C. When rp is between 1.5 and 2.1, the highest COPc,th is achieved with a
generation temperature of 120 ◦ C . When rp is between 2.1 and 3.0, the highest
COPc,th is achieved with a generation temperature of 100 ◦ C. This effect is conChapter 3
67
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 3.12: 2 - Stage / 2 - Effect LiBr - H2 O compression - absorption heat pump
cycle: COPc,th vs. rp
trary to that experienced with conventional absorption chiller cycles (Chapter
2).
It is believed that this effect is due to a ‘compacting’ of the different stages
with increasing rp . In other words, the generation temperature required in the
upper cycle necessary to produce sufficient exergy in the upper condenser relative
to the lower generator diminishes with greater absorber pressures. This effect
will be analysed and discussed further with the 3 - stage / 3 - effect compression
absorption cycle. The effect of rp upon the generation temperature crossover
point between the number of stages (Nstage ) relative to conventional absorption
cycles will also be discussed.
A graph indicating the relationship between COPc,th and Tgen at different rp
is shown in Figure 3.13. The five different curves indicate the five values of rp
simulated (n.b. when rp = 0, this implies a conventional absorption chiller cycle).
The shape and alignment of the curves is very similar to that discussed for the
single - stage cycle in Figure 3.9.
68
Chapter 3
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 3.13: 2 - Stage / 2 - Effect LiBr - H2 O compression - absorption heat pump
cycle: COPc,th vs. Tgen
There is however a noticeable difference between in the graphs; the curves are
more compact in the single - stage cycle than in the double - stage cycle. In the
single - stage cycle, assuming that the target COPc,th is 0.7 and the compression
ratio can float between 1.0 and 2.0, the generation temperature must be between
56 ◦ C and 77 ◦ C; a ∆Tgen of 21 ◦ C. In the double - stage cycle, assuming that the
target COPc,th is 1.2 and the compression ratio can float between 1.0 and 2.0, the
generation temperature must be between 83 ◦ C and 124 ◦ C; a ∆Tgen of 41 ◦ C. This
represents an almost doubling in the possible range of Tgen required to sustain
the target COPc for a conventional absorption chiller cycle.
This is particularly important because reducing the generation temperature
from over 120 ◦ C to below 90 ◦ C means that a far wider range of heat sources and
low grade waste heat sources can be utilised. For example, Low Temperature Hot
Water (LTHW) recovered from the cooling jacket of a CHP gas - fired engine, and
first and / or second stage intercooler(s). Ordinarily, only Medium Temperature
Hot Water (MTHW) or steam have the required temperatures necessary to drive a
Chapter 3
69
Hybrid Trigeneration - Thermally Activated Heat Pumps
double - stage absorption chiller cycle. By using a small amount of mechanical (i.e.
electrical) work it is possible to utilise a much greater range of low temperature
heat sources.
Triple - Stage / Triple - Effect
Figure 3.14: 3 - Stage / 3 - Effect LiBr - H2 O compression - absorption heat pump
cycle process flow diagram
A triple - stage triple - effect compression absorption LiBr - H2 O heat pump
cycle with steam compressor located at the evaporator outlet is shown in Figure
3.14. This cycle is similar to the double - stage double - effect cycle; heat is passed
from the upper and middle condensers (state points 27 and 17) directly to the
middle and lower generators respectively. Besides the addition of the compressor,
70
Chapter 3
Hybrid Trigeneration - Thermally Activated Heat Pumps
this cycle is the same as a conventional triple stage absorption chiller.
Figure 3.15: 3 - Stage / 3 - Effect LiBr - H2 O compression - absorption heat pump
cycle: COPc,th vs. rp
The effect of rp on the COPc,th at four different generation temperatures is
shown in Figure 3.15. The generation temperatures simulated range between
140 ◦ C and 200 ◦ C (40 ◦ C greater than the previous double - stage cycle, but with
a 20 ◦ C temperature overlap to enable a comparison to be made).
In general, the relationship between rp and COPc,th is very similar to that
observed in the single - stage and double - stage compression - absorption cycles.
Although there were notable differences in the shape of the curves and hence
nature of the relationship between the single - effect and the double - effect cycles,
the differences between the double - effect and triple - effect cycles appear to be
in magnitude only. The addition of the extra stage and a corresponding increase
in upper generation temperature of approximately 40 ◦ C tends to produce an
increase in COPc,th of between 20 and 30 % over the double - stage cycle.
A graph indicating the relationship between COPc,th and Tgen at different rp
is shown in Figure 3.16. The shape and alignment of the curves is quite different
Chapter 3
71
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 3.16: 3 - Stage / 3 - Effect LiBr - H2 O compression - absorption heat pump
cycle: COPc,th vs. Tgen
to that observed for the single - stage and double - stage cycles in Figures 3.9 and
3.13.
In the single - stage and double - stage cycles, the COPc,th increases initially
very rapidly as Tgen increases, and then appears to reach an asymptote. At higher
rp , changes in the magnitude of Tgen becomes almost like an on / off switch for the
cycle; below a certain Tgen the COPc,th is almost zero, and above the appropriate
Tgen the COPc,th is almost constant. There is also a very even spacing between
the curves at different rp , which suggests that COPc,th exhibits an almost linear
system response to an increase in rp .
However, in the triple - stage cycle both of these effects are no longer observed.
The relationship between COPc,th and Tgen is highly dependent upon rp . When rp
is 1.25, COPc,th changes dramatically with different values of Tgen , and increases
steadily before reaching an asymptote of around 1.7 when Tgen > 200 ◦ C; the
effect of increasing Tgen is less profound on COPc,th in the triple stage cycle than
the single or double stage cycles. When rp is 2.0, COPc,th is almost constant (at
72
Chapter 3
Hybrid Trigeneration - Thermally Activated Heat Pumps
approximately 2.0) through the range of Tgen simulated.
Interestingly, COPc,th appears to reduce very slightly from 2.0 to 1.95 when
Tgen > 170 ◦ C. This is due to a decrease in exergetic efficiencies of the internal
heat recuperation process (from condenser to generator). This reduction in cycle
exergetic efficiency (COP) implies that an extra stage is likely to be required. In
other words, at generation temperatures above 160 ◦ C (i.e. driving temperatures
over ∼170 ◦ C) and with rp > 2.0, a fourth stage may be required.
Figure 3.17: 3 - Stage / 3 - Effect LiBr - H2 O compression - absorption heat pump
cycle: COPc,m vs. Tgen
Figure 3.17 shows the relationship between COPc,m and Tgen for four different
values of COPc,th . When COPc,th is 1.9, COPc,m ranges between 12 and 18,
reaching a maximum when Tgen is 175 ◦ C. This suggest that in order to achieve
a COPc,th of 1.9, the optimum generation temperature to minimise mechanical
work input (i.e. maximise COPc,m ) is 175 ◦ C.
Conversely, when COPc,th is 1.6, COPc,m ranges from 24 to >70. Given that a
standard triple - stage LiBr - H2 O absorption chiller will produce a COPc of around
1.6 given a generation temperature of around 200 ◦ C, it therefore follows that the
Chapter 3
73
Hybrid Trigeneration - Thermally Activated Heat Pumps
COPc,m for the triple - stage compression - absorption should tend towards infinity
when Tgen = 200 ◦ C and COPc,m =1.6, as approximately observed in Figure 3.17.
Whilst the 3 - stage / 3 - effect LiBr - H2 O compression - absorption heat pump
cycle has a significantly better performance that a 3 - stage / 3 - effect LiBr - H2 O
absorption heat pump cycle, there is a major technical hurdle which may prevent
it from becoming a viable product. The very low pressure (∼ 10 mbar) and very
low density (0.03745 kg/m3 ) of the steam exiting the evaporator, means that the
steam compressor must be unfeasibly large. For a specified coolth output of
100 kW, the volumetric flowrate at the compressor inlet is 6.1 m3 /s. This would
be beyond the flow rate capabilities of a centrifugal compressor, and the pressure
ratio (∼ 5) would mean a very large and complex axial compressor will multiple
stages. Therefore, unfortunately the LiBr - H2 O compression - absorption heat
pump cycle is not considered to be a practicable solution.
3.1.4
Conclusions
Two types of thermally activated heat pumps have been explored in this Chapter: multi - effect / multi - stage H2 O - NH3 absorption heat pumps and multi effect / multi - stage LiBr - H2 O compression - absorption heat pumps. It has been
demonstrated that the double - stage / triple - effect H2 O - NH3 absorption heat
pump cycle can achieve a COPc equivalent to a triple - stage / triple - effect LiBr H2 O absorption heat pump, but with one fewer stage.
The sequence of internal heat recuperation devices within the cycle has been
shown to be vital to the performance. The strong solution which exits the absorber should pass through the solution HX first and then through the rectifier in
order to achieve the highest COPc . The double - stage / triple - effect H2 O - NH3
absorption heat pump cycle was able to achieve a theoretical COPc of 1.765.
A triple - stage / quintuple - effect cycle was conceptualised, but temperature required in the upper generator was too high for the thermophysical properties of
the ammonia and water mixture to be resolved.
74
Chapter 3
Hybrid Trigeneration - Thermally Activated Heat Pumps
Whilst the double - stage / triple - effect H2 O - NH3 absorption heat pump cycle
appears to be promising, a number of patents for similar concepts already exist
or have existed [52, 53]. Therefore this cycle is not considered suitable for further
research and development.
Singe - stage / single - effect, double - stage / double - effect and triple - stage / triple effect LiBr - H2 O compression - absorption heat pump cycles have been demonstrated to achieve a significant improvement in COPc over their conventional
absorption heat pump cycle counterparts. The triple - stage / triple - effect LiBr H2 O compression - absorption heat pump has been shown to achieve a COPc,th
of 2.0, when the steam compressor has a pressure ratio of 2.0. The net effect of
the addition of the compressor to the evaporator outlet is not only an increase
in COPc,th , but also a reduction in generation temperature. The corresponding
COPc,th has been demonstrated to be over 16.0 in all scenarios analysed.
Whilst the triple - stage / triple - effect LiBr - H2 O absorption - compression heat
pump cycle appears to be a promising area for future research, the extremely low
pressure and density of the steam exiting the evaporator means that the very
high volumetric flowrate through the compressor would result in it being almost
impracticably large. (6.1 m3 /s for 100 kW of cooling output). Therefore the absorption - compression cycle is not recommended as a topic of further research.
Chapter 3
75
Chapter 4
R744 Heat Pump Cycle Concepts
76
Hybrid Trigeneration - Thermally Activated Heat Pumps
4.1
Introduction
The very high pressure of the R744 and presence of three phases (liquid, vapour,
supercritical fluid) presents a number of challenges in the design of the cycle processes and in the design of specific components. Processes in which the operating
fluid passes close to the critical point are to be avoided wherever possible, due to
the severe changes in thermophysical properties which occur in this region.
A range of heat pump / chiller cycle concepts are conceptualised, developed
and analysed in this Chapter. Potential applications of theses cycles include:
• Desalination - low temperature evaporative desalination, such as multieffect evaporation or freeze desalination. Typically, such processes would
require a heat source of over ∼70 ◦ C or a coolth source of below ∼5 ◦ C
• District heating and cooling - district heating and cooling systems provide heat or coolth (hot or chilled water) to a network at a temperature
which is appropriate for building space temperature control; typically at
>70 ◦ C flow for heat and <12 ◦ C for cooling
• Low temperature processes - required by many processess, in particular
the food production industry.
• Solar driven cooling - using solar concentrating collectors to provide high
temperatures to a cycle which produces chilled water for cooling in buildings
located in arid and semiarid climates
The cycles which follow are thermally assisted transcritical MVC and thermally activated heat pump cycles. They combine the principles of both the reverse
Rankine cycle and Brayton cycles: isothermal heat input and isobaric heat rejection, and; isobaric heat input and heat rejection, respectively. The cycles use a
naturally occurring refrigerant (R744), rather than the currently commonly used
synthetic HFC refrigerants. They are capable of accommodating a range of mechanical and thermal energy inputs. Thermal inputs with temperatures between
Chapter 4
77
Hybrid Trigeneration - Thermally Activated Heat Pumps
150 ◦ C and 500 ◦ C are researched; based on an understanding of thermodynamic
principles, a higher input temperature can result in a lower mechanical energy
input requirement.
A thermally and mechanically activated R744 heat pump has been conceptulaised and analysed by Deng et al. [97]. This cycle utilises a solar - thermal
activated absorption chiller to assist a R744 MVC heat pump cycle to ‘super cool’ the refrigerant exiting the gas cooler. This lowers the vapour fraction, x,
of the refrigerant as it enters the evaporator, thereby increasing the COP. With
an evaporator outlet temperature of 7 ◦ C and a gas cooler outlet temperature of
38 ◦ C, the cycle achieved a COPc,m of 3.8. The cycles presented below differ from
the cycle proposed by Deng et al. in that they do not use an absorption chiller.
A considerable number of alternative cycles have been conceptualised, modelled and analysed, but are not presented within this Thesis. These are variants
of Cycle 1, Cycle 2 and Cycle 3 presented below which were either unsuccessful
or were part of the conceptualisation and development process.
4.2
Methodology
The cycles presented below are modelled and analysed for a ‘standard case’. In
these cases, the exergy of the coolth output of the cycle remains constant at
1 MW. The maximum available temperature is assumed to be 350 ◦ C; this is a
conservative estimate based upon known exhaust gas temperatures of commercially available small to medium scale gas turbines (5 MW to 20 MW) of between
450 ◦ C and 550 ◦ C.
The minimum heat rejection temperature was based either upon the ambient
dry bulb temperature (if dry cooling is used) or the ambient wet bulb temperature
(if adiabatic cooling or a cooling tower is used). This temperature is assumed to
be 30 ◦ C.
Engineering Equation Solver (EES) is used to model and simulate the cycles.
EES is a commercially available equation solving programme which is capable
78
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
of numerically solving coupled non-linear algebraic and differential equations,
conducting optimisation and sensitivity analyses. It uses an updated variation
of the Peng-Robinson equation of state model [98] to generate thermodynamic
and transport properties of single substances and mixtures. Approximately 1,000
equations are required to accurately define and model the various processes of
the cycles.
Details of the modelling and simulation tools for specific components within
the cycles is given in Chapter 5. An analysis and discussion of potential preliminary design methodologies for these components is also included.
4.2.1
Assumptions
The models use rely upon a number of assumptions, which are detailed below.
• The total thermal energy input is 1 MW.
• The maximum temperature at the work - expander inlet is 215 ◦ C.
• The minimum temperature achievable in the heat rejection unit is 35 ◦ C.
• The minimum pinch temperature in working fluid to water heat exchangers
is 2 ◦ C.
• The minimum pinch temperature in working fluid to working fluid heat
exchangers is 5 ◦ C (n.b. this is to prevent them from being impracticably
large).
• Pressure losses only occur in heat exchangers (i.e. not in pipework).
• The expansion valve - orifice plate is isenthalpic.
• All process are treated as adiabatic (unless there is an intentional transfer
of heat).
• Compressor and expander devices have an isentropic efficiency of 80%.
Chapter 4
79
Hybrid Trigeneration - Thermally Activated Heat Pumps
• Ejector nozzle, mixing section and diffuser are adiabatic and have an isentropic efficiency of 80%.
A method of considering Coefficient Of Performance (COP) based upon the
Primary Energy Consumption (PEC) is synthesized, such that the heat pump
performance can be analysed against its environmental impact in terms of primary energy. This value is termed COPc/hP EC , and gives an indication of the
true primary fuel -to - heat / coolth conversion efficiency of the cycle. In order
to calculate the glscopc/hP EC , a value for the thermal - to - mechanical work conversion efficiency of electricity delivered to point of use through the grid (ηgrid )
must be assumed. For this analysis, it is assumed that ηgrid is 30%. This value
encompasses an average primary fuel - to - electrical conversion efficiency in the
generation of electricity from fossil fuels (35%), an electrical grid transmission
and distribution efficiency (90%), and a electrical - to - mechanical efficiency of the
motor which drives the heat pump compressor. (95%) This assumes centralised
electricity generation. Equation 4.1 below indicates the process for calculating
COPc/hP EC , which is used in determining the performance of Cycle 3.
COPc/h,P EC =
4.3
4.3.1
1
1
+
COPc/h,t ηgrid × COPc/h,m
−1
(4.1)
Cycle 1
Description
In Cycle 1, a low quality heat source with input temperature of approximately
220 ◦ C (selected as a comparison to the maximum driving temperature used in
LiBr - H2 O absorption heat pumps) is required to provide the thermal input to the
Brayton portion of the cycle. The components which enable the processes of the
cycle to occur are: a transcritical superheated vapour compressor, a supercritical
fluid compressor, one supercritical turbine and a further supercritical turbine, a
recuperating heat exchanger, two gas/supercritical fluid coolers, an evaporator
80
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
and an expansion valve. The processes associated with this cycle are described
below:
• 1 - 2: internal heat recuperation (hot side)
• 2 - 3: supercritical fluid cooler (providing heat for external use)
• 3 - 4: second stage expansion turbine
• 4 - 5: thermal expansion valve
• 5 - 6: evaporator (providing coolth for external use)
• 6 - 7: first stage compressor
• 7 - 8: supercritical fluid intercooler (providing heat for external use)
• 8 - 9: second stage compressor
• 9 - 10: internal heat recuperation (cold side)
• 10 - 11: heat addition (from external source)
• 11 - 1: first stage expension turbine
The working fluid operates at four pressure levels as it passes through the
cycle. The minimum pressure is 43 bar, which occurs in the evaporator and is
results in an evaporation temperature of 8.3 ◦ C. The maximum pressure is tentatively set at approximately 250 bar, and occurs in the cold side of the supercritical
fluid intercooler and the heat addition heat exchanger. Whereas the minimum
pressure is dictated by the desired chilled water temperature, the maximum pressure is a specified design parameter which affects the performance characteristics
of the cycle.
The cycle produces a chilled water at 10 ◦ C, which is suitable for a chilled
water flow and return temperature in modern air conditioning systems of 10 ◦ C
and 15 ◦ C, respectively (older systems typically operate with a flow and return of
6 ◦ C and 12 ◦ C).
Chapter 4
81
Hybrid Trigeneration - Thermally Activated Heat Pumps
A temperature - entropy plot of the cycle is given in Figure 4.1. A single line
progress flow diagram of the cycle is shown in Figure 4.2.
Figure 4.1: Temperature - Entropy plot of R744 heat pump cycle 1
Figure 4.2: Process flow diagram of R744 heat pump cycle 1
4.3.2
Internal Heat Recuperation Heat Exchanger
An accurate pinch analysis of the internal heat recuperation heat exchanger is
crucial in determining how much exergy can be contained and re-used within the
cycle. Due to the characteristics of R744 and the close proximity to the critical
point in the hot side of the recuperator, the thermophysical properties of the fluid
82
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
are highly variable depending upon small changes in temperature and pressure.
A temperature pinch occurs due to the difference in Cp between the hot and
cold flows (implied by the ‘flattening’ of isobars close to the critical point on a
Temperature - Entropy plot.
A one-dimensional analysis heat transfer characteristics of the internal heat
recuperator is presented in Chapter 5.
4.3.3
Intercooling Pressure
Optimising the intercooling pressure minimises compression work, thereby increasing the COP relative to mechanical work input. In order to determine the
optimum intercooling pressure (i.e. pressure at the first - stage compressor outlet), a parametric analysis is conducted. However, optimising the first - stage
compressor outlet pressure may lower the intercooling temperature, which could
reduce the amount of useful heat available in the intercooler. The analysis takes
account of this by to ensuring that the intercooler inlet temperature (state point
7) is closely matched to the gas cooler inlet temperature (state point 2).
Figure 4.3 shows the relationship between compression work and the intercooler pressure. The optimum intercooler pressure, as defined by the minimum
compression work required, is identified as 97 bar.
4.3.4
Results and Discussion
The fluid properties at each state point are given in Table 4.1. For the maximum at the work - expander inlet of 215 ◦ C, the cycle is operating under the joint
thermally and mechanically activated regime.
For the mechanical work input, the cycle achieves a COPc,m of 7.0 and a
COPh,m of 12.4. For the thermal work input, the cycle achieves a COPc,th of 1.6
and a COPh,th of 2.8.
The COP values for mechanical work input conversion to coolth and heating
are on par with, if not marginally higher than the best R744 based MVC heat
Chapter 4
83
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 4.3: Parametric analysis of the optimum intercooling pressure
pumps in the literature [99] and indeed commercially available HFC based MVC
heat pumps.
The COP values for thermal input conversion to coolth and heating are
equal to a modern triple - effect LiBr - H2 O absorption chiller. Neither MVC heat
pumps, nor absorption heat pumps are capable of providing simultaneous heating
and cooling at the temperatures achieved by this cycle.
State Point
T ( ◦ C)
P (bar)
h (kJ/kg)
s (kJ/kgK)
x (-)
ρ (kg/m3 )
1
2
3
4
5
6
7
8
9
10
11
168.9
81.0
40
28.8
8.4
8.4
73.0
40.0
73.8
150.0
215.0
150
150
150
80
43.3
43.3
97
97
250
250
250
573.3
41.6
285.5
277.9
277.9
424.6
461.0
318.0
345.6
503.3
604.5
2.039
1.636
1.246
1.252
1.274
1.795
1.814
1.374
1.388
1.801
2.024
0.278
1.000
-
214.1
420.7
780.2
724.9
227.1
603.9
717.7
415.5
308.1
Table 4.1: Cycle 1 fluid properties at each state point
The minimum heat rejection temperature (state points 3 and 8) has a strong
effect on the COP. Figure 4.4 shows the relationship between COPc/h,th/m and
84
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
the ambient temperature.
As the ambient temperature increases, COPc reduces, both in terms of the
mechanical and thermal work input. This is due to the increased enthalpy at state
point 3 and consequent vapour mass fraction, x, entering the evaporator at state
point 5; this is a known performance characteristic of a heat pump. However, the
close proximity of the fluid to the critical point through both the gas cooler and
intercooler means that changes in ambient temperature become more influential
on the cycle performance.
Upon observing Figure 4.1, it can be seen that isobars ‘flatten’ close to the
critical point. Therefore, a small increase in temperature results in a large shift
to the right on a Temperature - Entropy plot, resulting in a large increase in both
entropy and vapour mass fraction (assuming that the fluid is to consequently be
isenthalpically throttled). In the same manner, a small decrease in temperature
results in a large shift to the left on the Temperature - Entropy plot, resulting in
a large decrease in entropy and vapour mass fraction. The former results in a
decrease in COPc , whilst the latter results in an increase in COPc , as observed
in Figure 4.4a.
It is evident from Figure 4.4a that the changes in ambient temperature affect
the COPc,m considerably more than the COPc,th . Throughout the range of ambient temperatures considered, COPc,th changes by 30%, whilst COPc,m changes by
57%. The COPc,m is more dependent upon ambient temperature due to the dependency of the intercooler on the ambient temperature and consequent changes
the second stage compression work.
The relationship between ambient temperature and COPh , shown in Figure
4.4b, is very similar to that between the ambient temperature and COPh . The
changes in ambient temperature affect the COPh,m considerably more than the
COPh,th . Throughout the range of ambient temperatures considered, COPh,th
changes by 13%, whilst COPh,m changes by 46%.
The change in COPh is less than that shown by COPc . An increase in ambiChapter 4
85
Hybrid Trigeneration - Thermally Activated Heat Pumps
(a) COPc
(b) COPh
Figure 4.4: Plots of COP against ambient temperature for thermal and mechanical inputs for Cycle 1
ent temperature primarily affects the evaporator inlet conditions and intercooler
outlet conditions. Intercooler pressure affects the compression work, and therefore both COPc,m and COPh,m show high dependency upon ambient temperature
(57% and 48%). The evaporator inlet conditions affect the COPc but not does
not affect COPh , and this explains why COPc,th is much more dependent upon
ambient temperature than COPh,m (30% vs 13%).
86
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 4.5: A plot of COPc against evaporation temperature for Cycle 1
Figure 4.5 shows the relationship between the evaporation temperature (i.e.
pressure), and COPc,m and COPc,th . The COPc,m decreases sharply with a reduction in evaporator temperature, dropping from 12.2 at an evaporator temperature of 14 ◦ C down to 4.5 at an evaporation temperature of -4 ◦ C. Conversely,
the COPc,th increases very marginally with a reduction in evaporation temperature, increasing from 1.7 at an evaporator temperature of 14 ◦ C to 1.9 at an
evaporation temperature of -4 ◦ C.
When the evaporation temperature is reduced (i.e. evaporation pressure is
reduced), the first stage compressor pressure ratio increases, which results in an
increase in mechanical work input. This causes a decrease in the COPc,m . However, because R744 is a wet fluid, as the evaporation pressure is reduced the
enthalpy of evaporation increases; therefore the coolth produced by the evaporator increases. This slows down the decrease in COPc,m as the evaporation
temperature is reduced.
The increase in the enthalpy of evaporation at lower pressures explains the
marginal increase in COPc,th when the evaporation temperature is reduced. There
Chapter 4
87
Hybrid Trigeneration - Thermally Activated Heat Pumps
is no further relationship between evaporation temperature and thermal input.
Figure 4.6: A plot of COPh against evaporation temperature for Cycle 1
Figure 4.6 shows the relationship between the evaporation temperature and
COPh,m and COPh,th . The relationship shows an almost identical trend to that
shown in 4.5 between evaporation temperature and COPc . When evaporation
temperature is reduced, COPh,m decreases; this is due to the same reasons as
COPc,m . However, the small increase in COPh,th that occurs when evaporation
temperature is reduced is due purely to the increase in enthalpy of evaporation
(which increases enthalpy at state point 6). The thermophysical properties of
R744 (i.e. the curvature of isobars and isenthalps on a Temperature - Entropy
plot) are such that when the pressure of dry saturated vapour (or, in reality,
slightly superheated vapour) entering the first stage compressor is reduced, the
compressor outlet conditions (state point 7) are at a much higher temperature
and enthalpy. This not only increases the useful heat available (i.e. COPh ), but
also the maximum temperature for external use.
Figure 4.7 shows the relationship between the temperature at the first stage
compressor outlet (state point 7) and the evaporation temperature. The internal
88
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 4.7: A plot of T7 and T2 against evaporation temperature for Cycle 1
heat recuperator hot side outlet (state point 2) is also plotted for reference. In
order to simplify the hot water circuit, it is desirable to match the upper temperatures of the two hot streams (T2 and T3 ). With this in mind, the ideal
evaporation temperature is between 2 and 3 ◦ C. The target evaporator temperature of approximately 8 ◦ C results in a temperature difference between T2 and
T3 of 7 ◦ C.
4.4
4.4.1
Cycle 2
Description
Cycle 2 represents a minor modification of the Cycle 1 discussed previously. The
primary difference is the addition of an ejector following the second stage expander. The purpose of the ejector is to decrease the compression work required
by increasing the pressure of the dry saturated vapour (or, in reality, slightly
superheated vapour entering the first stage compressor, whilst not altering the
Chapter 4
89
Hybrid Trigeneration - Thermally Activated Heat Pumps
evaporation temperature. The ejector also decreases the enthalpy/ vapour mass
fraction of the working fluid entering the evaporator. This increases the exergy
of the fluid, and therefore reduces the mass flow through the evaporator.
A review of literature and intellectual property regarding the design, integration and performance of ejectors within R744 heat pump cycles is provided in
Chapter 5. A preliminary design method for a ejector for this cycle (and the
subsequent Cycle 3) is also provided, along with a discussion of the expected
performance characteristics.
As a result of the inclusion of an ejector, it is necessary to have a vapour liquid separator to split the two - phase flow at the outlet of the ejector expansion
diffuser. The saturated liquid is goes into the evaporator, whilst the dry saturated
vapour goes into the first stage compressor. This is discussed in more detail in
Chapter 5.
The addition of the ejector allows the reverse Rankine aspect of the cycle
to be powered purely from the Brayton aspect of the cycle, such that the only
energy stream required to activate the cycle is a mid - grade temperature source.
As before, the cycle can provide simultaneous heating and cooling effects. As
an evolution of the analysis of Cycle 1, the performance characteristics of the
supercritical fluid cooler has been conducted. This includes and optimisation
of the hot water circuit configuration, volume flow rate and flow and return
temperatures.
The cycle is designed to operate in a trigeneration system, where the prime
mover is a gas turbine. It produces hot water at both 65 ◦ C and 40 ◦ C, and chilled
water at 10 ◦ C. Therefore the maximum expander inlet temperature is specified
such that the heat input can be provided exhaust gases of the gas turbine but a
recuperator. The amount and quality of thermal energy available for recuperation
(i.e. exergy) bears significance upon the cycle performance, specifically COP, as
will be demonstrated.
A single line progress flow diagram of Cycle 2 is shown in Figure 4.8, and a
90
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
temperature - entropy plot of the cycle is given in Figure 4.9.
Figure 4.8: Process flow diagram of R744 heat pump Cycle 2
Chapter 4
91
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 4.9: Temperature - entropy plot of R744 heat pump Cycle 2
The processes associated with this cycle are described below:
• 1 - 2: internal heat recuperation (hot side)
• 2 - 3: supercritical fluid cooler (providing heat for external use) air cooler
• 3 - 4: second stage expander
• 4 - 5: nozzle stage of ejector (pressure reduction of ejector motive flow)
• 5 - 6x: mixing and diffuser stage of ejector
• 6x - 6l: separation of two - phase flow into dry saturated vapour
• 6l - 7: thermal expansion valve
• 7 - 8: evaporator (providing coolth for external use)
• 8 - 5: nozzle stage of ejector (entrained flow pulled in to negatively pressurised motive flow)
• 6x - 6: separation of two - phase flow into saturated liquid
• 6 - 9: first stage compressor
92
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
• 9 - 10: supercritical fluid intercooler (providing heat for external use) and
air cooler
• 10 - 11: second stage compressor
• 11 - 12: internal heat recuperation (cold side)
• 12 - 13: heat addition (from external source)
• 13 - 1: first stage expander
The maximum working fluid pressure, which occurs at the outlet of the second
compressor, cold side of the internal heat recuperator and at the expander inlet,
is 250 bar. The minimum working fluid pressure is 43 bar, which occurs in the
refrigerant side of the evaporator heat exchanger. There are four intermediate
pressure which the working fluid operates at.
The pressure at first stage compressor outlet and intercooler hot side has been
previously optimised to 97 bar. The pressure at first stage expander outlet varies
depending upon the first expander inlet temperature; if the inlet temperature
increases, the mechanical work output increases for a constant pressure ratio and
isentropic efficiency.
In order to ensure that the net mechanical work output of the expanders to
the compressors of zero, the first stage expander pressure ratio must decrease
as the inlet temperature increases. Decreasing the first stage expander pressure
ratio increases the pressure in the hot side of the internal heat recuperator, and
consequently increases the exergy at state point 2. This subsequently reduces
the enthalpy at state point 3 (inlet to the second stage expander), which in turn
reduces the vapour fraction of the working fluid in the mixing section and diffuser
section of the ejector (state points 5 and 6x, respectively). The overall result in
an increase in the liquid refrigerant produced in the vapour - liquid separator,
and therefore an increase in the COPc . The nature of this relationship will be
discussed in the results section.
Chapter 4
93
Hybrid Trigeneration - Thermally Activated Heat Pumps
The second expander outlet pressure (state point 4) is a fixed parameter, specified above the critical pressure at 80 bar to ensure that there is no mixed phase
flow within the expander. The ejector outlet pressure is a calculated parameter,
which is dependent upon the motive and entrained fluid inlet pressure and the
isentropic efficiency of the nozzle, mixing and diffuser sections (this is discussed
in more detail in Chapter 5).
4.4.2
Results and Discussion
Cycle 2 has been simulated simulated across a range of maximum first stage
expander inlet temperatures. A first stage expander inlet temperature of 350 ◦ C
is denoted as the ‘base case’.
The fluid properties at each state point for the ‘base case’ are given in Table
4.2. Under base case conditions, the cycle achieves a COPc,th of 1.58, which
is directly comparable to a triple - stage LiBr - H2 O absorption chiller [54] [47]
[48]. The corresponding COPh,th is 2.41, which is achievable in a double - stage
or triple - stage LiBr - H2 O absorption heat pump, provided that the heat source
(evaporation temperature) is sufficiently high. Both COPc,m and COPh,m are
zero.
State Point
T ( ◦ C)
P (bar)
h (kJ/kg)
s (kJ/kgK)
x (-)
ρ (kg/m3 )
1
2
3
4
5
6x
6l
6
7
8
9
10
11
12
13
305.1
67.6
33.0
21.4
5.3
11.6
11.6
11.6
8.4
8.4
68.3
30.0
51.3
266
350
160
160
160
80
40
46.8
46.8
46.8
43.3
43.3
97
97
250
250
250
736.6
364.1
258.8
251.0
328.7
332.0
230.1
421.1
230.1
424.5
454.5
273.1
296.6
674.1
781.9
2.361
1.482
1.156
1.161
1.461
1.461
1.103
1.774
1.105
1.795
1.791
1.229
1.242
2.159
2.345
0.539
0.533
0
1.000
0.042
1.000
-
151.5
569.5
857.1
815.3
193.2
233.1
849.2
142.6
700.7
128.3
26.0
761.9
827.5
261.6
213.6
Table 4.2: Cycle 2 fluid properties at each state point
As per Cycle 1, Cycle 2 provides coolth at approximately 8 ◦ C, which implies
94
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
that chilled water can be produced at approximately 10 ◦ C. State points 2 and 9
indicate the maximum temperature available at the hot side of the heat pump (in
the intercooler and supercritical fluid cooler, respectively), which can be recovered
for use in the production of hot water. The overall temperature lift is 60 ◦ C, and
herein lies the significant advantage of Cycle 2 over multi - stage absorption chiller;
the ability to provide chilled water and hot water simultaneously.
The relationship between COPc,th and COPh,th and the first stage expander
inlet temperature (state point 13) is shown in Figure 4.10. It is evident that as the
inlet temperature is reduced, both COPc,th and COPh,th decrease, and that this
effect is non - linear. The lowest temperature considered in this analysis of 200 ◦ C
produces a COPc,th of 0.78 and a COPh,th of 1.62. The highest temperature
analysed is 450 ◦ C and produces a COPc,th of 1.86 and a COPh,th of 2.80.
Figure 4.10: A plot of COPC and COPh against first stage expander inlet temperature for Cycle 2
As the first stage expander inlet temperature is reduced, the first stage expander pressure ratio must increase to ensure the same mechanical work output.
Figure 4.11 shows the relationship between the inlet temperature and outlet presChapter 4
95
Hybrid Trigeneration - Thermally Activated Heat Pumps
sure for the first stage expander (n.b. expander inlet pressure is a fixed parameter,
therefore changes in outlet pressure are as a direct result of changes in pressure
ratio). This correlation is very similar in shape to the relationship to that observed in Figure 4.10. This can be said to imply that the first stage expander
bears significant influence upon both COPc,th and COPh,th .
Figure 4.11: A plot of fluid pressure at state point 1 (first stage expander outlet)
against temperature at state point 13 (first stage expander inlet temperature) for
Cycle 2
At expander inlet temperatures below 200 ◦ C the COP drops sharply, and the
model indicates that with a temperature below approximately 190 ◦ C the cycle
becomes unfeasible; that is to say the expanders do not produce enough mechanical work to drive the compressors, and the cycle requires a net mechanical work
input. This can be explained in conjunction with Figure 4.11, which shows that
at an expander inlet temperature of 200 ◦ C, the fluid pressure in the supercritical
fluid cooler is 90 bar. In order to prevent condensation in the supercritical fluid
cooler, the fluid pressure must be above the critical pressure (73.8 bar). As the
expander inlet temperature drops below approximately 190 ◦ C, the fluid pressure
in the supercritical fluid cooler drops to below the critical pressure.
96
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
4.5
Cycle 3
4.5.1
Description
Cycle 3 is an evolution of Cycle 2, and includes a modification which could potentially allow for the continuous variation of mechanical and thermal energy inputs
to the cycle. The second - stage compressor has a split - flow arrangement, with
two outlets at different pressures. The lower pressure outlet is at the same pressure as the second stage expander inlet, whilst the second stage outlet is the first
stage expander inlet pressure.
A single line progress flow diagram of Cycle 3 is shown in Figure 4.12, and a
temperature - entropy plot of the cycle is given in Figure 4.13.
The processes associated with this cycle are described below:
• 1 - 2: internal heat recuperation (hot side)
• 2 - 3: supercritical fluid cooler (providing heat for external use) air cooler
• 3 - 4: second stage expander
• 15 - 4: second stage expander
• 4 - 5: nozzle stage of ejector (pressure reduction of ejector motive flow)
• 5 - 6x: mixing and diffuser stage of ejector
• 6x - 6l: separation of two - phase flow into dry saturated vapour
• 6l - 7: thermal expansion valve
• 7 - 8: evaporator (providing coolth for external use)
• 8 - 5: nozzle stage of ejector (entrained flow pulled in to negatively pressurised motive flow)
• 6x - 6: separation of two - phase flow into saturated liquid
• 6 - 9: first stage compressor
Chapter 4
97
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 4.12: Process flow diagram of R744 heat pump Cycle 3
98
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 4.13: Temperature - entropy plot of R744 heat pump Cycle 3
• 9 - 10: supercritical fluid intercooler (providing heat for external use) and
air cooler
• 10 - 14: second stage compressor (first split)
• 10 - 11: second stage compressor (second split)
• 11 - 12: internal heat recuperation (cold side)
• 12 - 13: heat addition (from external source)
• 13 - 1: first stage expander
• 14 - 15: external heat rejection
By increasing the mass flow leaving the split in the second stage compressor (state point 14), it is possible to increase the mechanical work input and
decrease the thermal work input whilst maintaining a constant cooling work output. Therefore, depending upon the availability of thermal energy (if recuperated
exhaust heat) and / or the price of electricity, the mix can be adjusted accordingly.
This is, in principle, a fully hybrid heat pump cycle.
Chapter 4
99
Hybrid Trigeneration - Thermally Activated Heat Pumps
Although in theory, this appears to be an extremely advantageous feature over
the previous Cycle 2, in reality there would be significant control complexities and
design challenges to be overcome. Firstly is the design of a split flow supercritical
R744 compressor with an inlet pressure of ∼95 bar, and intermediate pressure of
∼115 bar, and a final pressure of ∼250bar, with a variable mass flow ratio at the
two outlets. Secondly, is the control of the second stage expander inlet conditions
for the two streams (state point 3 and state point 15). Thirdly, the design of the
heat addition and heat recuperation heat exchangers, such that they are able to
operate with a variable mass flow rate of working fluid, whilst maintaining the
desired inlet and outlet conditions. It is also necessary to include an air cooled
heat exchanger to reject heat from the flow exiting the split flow compressor at
an intermediate pressure (state point 14).
The split ratio, γ, is defined as the mass flow ratio of working fluid at state
point 11 relative to the mass flow of working fluid at state point 10 (γ =
ṁ11
).
ṁ10
In pure thermal work input γ = 1, whilst in pure mechanical work input γ = 0.
4.5.2
Results and Discussion
When operating in a pure thermal work input mode, Cycle 3 achieves the same
values for COPc,th and COPh,th as Cycle 2 (1.58 and 2.41, respectively). When
operating in a pure mechanical work input, Cycle 3 achieves a COPc,m of 5.12
and a COPh,m of 5.80. These values are are higher than many conventional MVC
R744 heat pumps [76] [100], although a patent was filled in 2010 for a R744 heat
pump cycle which can achieve a COPc of 7.47 under the ideal operating conditions
[99].
Figure 4.14 demonstrates the relationship between COPc,P EC and COPh,P EC
and the split ratio, γ. The data presented in figure 4.14 is based upon a first
stage expander inlet temperature (state point 13) of 250 ◦ C. It can be observed
that an increase in γ produces contrasting behaviour in COPc,P EC to COPh,P EC .
When the cycle is operating in pure thermal input mode (γ = 1) COPc,P EC
100
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 4.14: A plot of the relationship between COPc,P EC and COPh,P EC and
the split ratio, γ for Cycle 3
is 1.03 and COPh,P EC is 2.02. As γ decreases, and the proportion of mechanical
work input increases, COPc,P EC increases and COPh,P EC decreases. The increase
in COPc,P EC is due to the difference in COPc when operating in pure thermal
work input mode (1.02) and COPc when operating in pure mechanical input mode
(5.12). The decrease in COPh,P EC can be explained via an understanding of the
cycles process. When the γ is increased, the mass flow of working fluid passing
through the Brayton portion of the cycle is reduced, and subsequently the mass
flow of working fluid passing through the supercritical fluid cooler (state points 2
to 3) is reduction. Consequently, the heat production of cycle is reduced. When
the cycle is operating in pure mechanical input mode the COPc,P EC is 1.54 and
the COPh,P EC is 1.74.
Figure 4.15 shows the relationship between COPc/h,m/th across the entire
range of split ratio, γ. Observation of Figures 4.15a and 4.15b shows that the
coolth output from mechanical work input behaves in a manner which is the opposite to the cooling work output from the thermal work input, as γ changes.
Chapter 4
101
Hybrid Trigeneration - Thermally Activated Heat Pumps
(a) COPc,m
(b) COPc,th
(c) COPc,m
(d) COPc,th
Figure 4.15: A plot of the relationship between COPc/h,h/m split ratio for Cycle
The COPc,m tends to infinity as γ approaches 1, and COPc,th tends to infinity
as γ approaches 0. In reality, this implies that the mechanical and thermal work
inputs are approaching 0. Exactly the same relationship is observed in Figures
4.15c and 4.15d for COPh,m and COPh,m .
4.6
Cycle Selection and Discussion
Three cycles have been presented and discussed, although many more were conceptualised, modelled and analysed; a basic block schematic of other cycles is
given in the Appendix.
Cycle 1 is an earlier concept, which was not capable of operating on thermal
work input alone at the temperatures discussed. The input of both mechanical
and thermal work provides an interesting range of possible operating modes for
hybridised trigeneration systems with gas - fired reciprocating engines or gas turbine. Conventional thermal load or electrical load following control algorithms
102
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
are unlikely to be appropriate. A hybrid control algorithm which considers both
supply and demand factors would be required; the complex performance characteristics of the heat pump, particularly with regard to ambient temperature and
evaporator pressure, mean that this is necessary.
In order for Cycle 1 to operate under a pure thermal regime, it is necessary to
have an expander inlet temperature of approximately 400 ◦ C. However, the COP
achieved is relatively low when compared to Cycle 2. Therefore, due to potential
control issues and poor performance, Cycle 1 is not considered to be a viable
option for further consideration.
Cycle 2 is a evolution of Cycle 1, which aimed to overcome the potential
control complexities resulting from two types of work inputs, as well as poor performance. The principal desire was to increase the COP to be equivalent to a
triple - stage LiBr - H2 O absorption chiller, but enabling the simultaneous production of heat and coolth and with significantly fewer processes (heat exchangers,
vessels, pumps, etc.). The addition of a transcritical ejector to the cycle enable
this to be possible; a COPc,th of 1.58 and a COPh,th of 2.41 have been demonstrated to be theoretically possible with a first stage expander inlet temperature
of 350 ◦ C to the cycle. As with Cycle 1, the performance of Cycle 2 is highly
dependent upon the first stage expander inlet temperature and the evaporator
temperature. However, the cycle is able to function at first stage expander inlet
temperatures at low as 190 ◦ C; although COPc,th is only marginally better than
a single - stage LiBr - H2 O absorption chiller. The maximum COPc,th of 1.86 and
COPh,th of 2.691 (at first - stage expander inlet temperatures of 450 ◦ C) mean
that Cycle 2 has potential for further investigation.
Cycle 3 is an evolution of Cycle 2, which incorporates the potential for a
hybrid mix of continuously variable thermal and mechanical work inputs. The
performance of Cycle 3 under pure thermal work input is identical to Cycle 2.
However, increasing the mix of mechanical work input increases the COPc in
terms of Primary Energy Consumption (PEC), but decreases the COPh in terms
Chapter 4
103
Hybrid Trigeneration - Thermally Activated Heat Pumps
of Primary Energy Consumption (PEC). This cycle has great potential in a hybrid
trigeneration scheme, due to the flexibility of the work inputs and the opportunity
this affords to alter the mix of energy flows to best match demands. However, the
complexity of the split flow compressor means that this cycle will be considered
no further.
Therefore Cycle 2 is believed to be the best candidate for further research and
development.
4.7
Sensitivity Analysis
The COP of Cycle 2 is particularly a function of the isentropic efficiencies of
the two compressors and the two expansion turbines. A sensitivity analysis has
been conducted using the EES software, to determine the effect of the isentropic
efficiencies on the COPc and COPh of Cycle 2.
Figure 4.16: Sensitivity analysis of the COPc and COPh of Cycle 2 to the isentropic efficiency of the compressors
Figure 4.16 highlights the relationship between the isentropic efficiency of the
compressors and the COP. The relationship demonstrates that the COP reduces
as the compressor isentropic efficiencies are reduced, in an approximately linear
fashion. The gradient of the curve for COPh is marginally greater than for COPc ,
104
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 4.17: Sensitivity analysis of the COPc and COPh of Cycle 2 to the isentropic efficiency of the expanders
indicated that the isentropic efficiency of the compressors has a bigger effect on
COPh .
Figure 4.17 highlights the relationship between the isentropic efficiency of
the expanders and the COP. The relationship demonstrates that the COP reduces as expander isentropic efficiencies are reduced. The gradient of the curves
increases at higher isentropic efficiencies and reduces at lower isentropic efficiencies. As with the relationship observed in Figure 4.16, the COPh appears to vary
marginally more over the range of expander isentropic efficiencies considered.
Figure 4.18 shows the effect of varying the isentropic efficiencies of both the
two compressors and the two expanders on the COP. This demonstrates that
there is a very non linear relationship between the isentropic efficiencies and
the COP. A COPc of 2 and COPh of 3 is achieved with isentropic efficiencies
of approximately 87 %. If the isentropic efficiencies reduce to 77 % the COPc
reduces to approximately 1.3 and the COPc reduces to approximately 2.1, a 35 %
and 30 % reduction, respectively.
Chapter 4
105
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 4.18: Sensitivity analysis of the COPc and COPh of Cycle 2 to the isentropic efficiency the compressors and expanders
4.8
Comparison to a ‘Separate Cycle System’
It must be demonstrated that Cycle 2 gives a significant performance benefit over
a ‘separates’ system in order to justify its complexity. Therefore, the performance
of Cycle 2 is analysed relative to a low temperature recuperated exhaust heat
supercritical CO2 Brayton cycle, linked to a basic transcritical R744 MVC heat
pump.
The R744 heat pump cycle generates hot water using a supercritical fluid
cooler, which generates hot water at 40 ◦ C and 60 ◦ C. In order to reduce the
temperature of the working fluid as close as possible to ambient, the supercritical
fluid cooler is followed by an air cooler.
The thermal to mechanical work conversion efficiency of the Brayton cycle is
dependent upon the expander inlet and outlet conditions, in particular temperature and pressure. Figure 4.21 shows the relationship between expander inlet
pressure and outlet pressure, assuming an expander inlet temperature of 250 ◦ C.
The maximum thermal to mechanical conversion efficiency the cycle can achieve
is 19.24%. The data in Figure 4.21 shows that cycle efficiency is greater at lower
expander outlet pressures (Pmin ) and higher inlet pressures (Pmax ), as would typ106
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 4.19: Supercritical CO2 Brayton cycle process flow diagram
Figure 4.20: Basic transcritical R744 MVC cycle process flow diagram
ically be expected of a Brayton cycle. Increasing the pressure ratio across the
expander is desirable from an efficiency perspective, but high pressure ratios can
lead to reduced isentropic efficiencies across compression and expander devices
(this analysis assumes the same isentropic efficiency for all pressure ratios). A further comment of caution with a high pressure ratio is the decrease in effectiveness
Chapter 4
107
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 4.21: Supercritical CO2 Brayton cycle thermal to mechanical conversion
efficiencies for a range of inlet and outlet pressures.
of the internal heat recuperator (for a given heat exchanger surface area), due
to the difference in Cp between the high pressure cold side and low pressure hot
side; this problem becomes much worse when the expander outlet temperature is
close to the critical pressure. The decrease in recuperator effectiveness occurs as
a result of temperature pinch towards the cold end of the heat exchanger.
The R744 MVC heat pump cycle achieves a simultaneous COPc and COPh
of 3.52 and 4.12, respectively. Coupling the two cycles, and not accounting for
mechanical to electrical losses, if any, the implied COPc,th and COPh,th are defined
by Equation 4.2.
COPc/h,th = ηBrayton × COPc/h
(4.2)
Therefore, COPc,th is 0.68 and COPc,th is 0.80. The corresponding values
COPc,th and COPh,th achieved by Cycle 2 with an first stage expander inlet
temperature of 250 ◦ C are 1.12 and 1.96, respectively - this is a 65% and 145%
increase in COP. This is a significant improvement in performance which makes
a very strong argument for the integration of a CO2 mechanical work cycle and
a heat pump cycle into a single loop. The thermodynamic benefit derived from
the sharing of compression work, the internal recuperated of exergy, the thermo108
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
physical properties of R744 and the two - phase ejector yields this performance
gain.
The ejector increases the inlet pressure of first stage compressor, which reduces
the work required to compress the gas, which subsequently allows the first stage
expander pressure ratio to be reduced which consequently increases the cooling
work potential of the evaporator. The advantage gained by sharing compression
work is best understood with reference to the overlap in pressures of the two
parts of the cycle: the Brayton cycle and the reverse Rankine cycle. This can be
observed on Figure 4.8 by the difference in pressure between P3 and P9 , and the
fact that P3 > P9 .
4.9
Integration into a Trigeneration System
To assess the suitability of Cycle 2 for integration within a trigeneration system which has a gas turbine as the PGU, a high level assessment of the total
efficiency of the system is required. In order to do this, the exergetic efficiency
of the exhaust heat recuperation process must be ascertained. This is dependent upon the conditions of the exhaust gases after they exit the gas turbine,
the conditions of the exhaust gases after the recuperator and the ambient conditions. The conditions of the exhaust gases is fixed parameters and are dependent
upon the specific performance of the gas turbine, which are a commonly available
performance parameter from suppliers and manufacturers. For the sake of this
high level analysis, it is assumed that the exhaust gases can be treated as air at
500 ◦ C and 0.5 barg. Assuming that the heat recovery process is fully reversible,
the exergetic efficiency (ηex ) of the exhaust heat recuperation process can simply
defined by Equation 4.3.
ηex =
∆hrecuperator
BG.T.exhaust
(4.3)
BG.T.exhaust represents the exergy in the exhaust stream of the gas turbine.
Chapter 4
109
Hybrid Trigeneration - Thermally Activated Heat Pumps
Thus, assuming that Cp is constant with temperature, Equation 4.3 can be expressed as shown in Equation 4.4.
ηex =
TG.T.exhaust − T12
TG.T exhaust − Tamb
(4.4)
Figure 4.22 shows the relationship between T12 and T13 (i.e. the internal heat
recuperator code side outlet temperature and first stage work expander inlet /
exhaust heat recuperator working fluid side outlet temperatures, respectively).
The relationship between T12 and T13 is almost linear, and it can be observed
that T13 − T12 ≈ 100◦ C. This implies that the cp of the two fluid streams (air
and R744) are approximately constant, and therefore the enthalpy transfer in the
exhaust heat recuperator is linear..
Using Equation 4.4, Figure 4.22 can be converted to demonstrate the relationship between the exergetic efficiency of the exhaust heat recuperation process
and the first stage expander inlet temperature, T13 . This relationship is shown in
Figure 4.23. It is assumed that the ambient temperature is 30 ◦ C.
The exergetic efficiency is a direct measure of the enthalpy of the gas turbine
exhaust which can be recovered by Cycle 2, enabling it to ultimately produce
chilled water and hot water for building and process consumption. Ascertaining the optimum first stage expander temperature (T13 ) to maximise the overall
production of chilled water and hot water is vital in optimising the performance
for the trigeneration system. The COP of Cycle 2 when integrated within a
trigeneration system is defined in Equation 4.5.
COPc,trigen.system = COPc,th × ηex
(4.5)
Figure 4.24 shows the relationship between COPc,trigen.system and T13 . This
indicates the optimum first stage work expander inlet temperature to maximise
the coolth production potential of Cycle 2 integrated within a gas turbine trigeneration system; it is evident that this temperature is approximately 275 ◦ .
110
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 4.22: A plot of the relationship between temperature at state point 12
(internal heat recuperator inlet) against temperature at state point 13 (first stage
expander inlet temperature) for Cycle 2
Figure 4.23: A plot of the relationship between the exergetic efficiency of the
exhaust heat recuperation process and T13
Whereas a higher T1 3 gives a higher COPc for the cycle alone, this is clearly not
the optimum solution for the cycle when integrated within a trigeneration system.
Figure 4.24 shows the relationship between shows the relationship between
Chapter 4
111
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 4.24: A plot of the relationship between COPc,trigen.system and T13
Figure 4.25: A plot of the relationship between COPh,trigen.system and T13
COPh,trigen.system and T13 . This indicates the optimum first stage work expander
inlet temperature to maximise the heat production potential of Cycle 2 integrated
within a gas turbine trigeneration system; it is evident that this temperature is
approximately 240 ◦ C. This is slightly lower than the optimum T13 for coolth
production, which could allow for system fine tuning.
112
Chapter 4
Hybrid Trigeneration - Thermally Activated Heat Pumps
An plot illustrating the potential COPc in a ‘direct firing’ scenario (i.e. where
natural gas is combusted to directly provide thermal energy to the cycle via the
exhaust heat recuperator, or via an intermediate heat transfer fluid) is given in
Figure 4.26. This is compared to the COPc,th for Cycle 2 itself. The difference
between the two curves is an illustration of the exergetic efficiency of the heat
addition process. Unlike with heat addition from exhaust, direct firing of natural
gas results in a much T13 giving the highest performance (i.e. it gives a higher
exergetic efficiency).
Figure 4.26: A plot to compare the relationship between COPc,th,directf iring and
COPc,th,cycle against T13
4.10
Conclusions
Three novel thermally and mechanically or pure thermally activated transcritical
R744 heat pump cycles have been conceptualised, modelled and analysed. Each
of these cycles are capable of providing simultaneous heating and cooling, and
are designed to use the thermal energy in the exhaust gases from a Primary
Generation Unit (PGU), in this instance a gas turbine. Cycle 2 was identifed as
the preferred heat pump cycle as it provides a higher COPc and COPh than Cycle
Chapter 4
113
Hybrid Trigeneration - Thermally Activated Heat Pumps
1. Cycle 3 is a minor evolution of Cycle 2 and is capable of providing mechanical
work output as well as producing hot water and chilled water, however the small
magnitude of the mechical work output and complexity of a split flow compressor
meant that Cycle 3 was not pursued further. Cycle 2 is capable of providing
coolth (10 ◦ C) at a COPc,th of 1.58 and hot water (65 ◦ C) at a COPh,th 2.41, with
a first stage expander inlet temperature of 350 ◦ C.
These performance figures for COPc are very similar to those theoretically
achieved by triple - stage LiBr - H2 O absorption chillers, as demonstrated earlier
(n.b. typically COPc = 1.6). However, Cycle 2 is capable also simultaneously
producing hot water at 65 ◦ C. Cycle 2 requires half the number of processes of a
triple - stage LiBr - H2 O absorption chiller and uses a single working fluid, further
simplifying the system and the control. Chapter 5 provides an overview of an
initial component analysis and design for key components of Cycle 2, such as: the
internal heat recuperator, ejector, work expander and supercritical fluid cooler.
It was then shown that Cycle 2 performs considerably better than a ‘separates
system’ of a supercritical CO2 Brayton cycle linked to a R744 MVC heat pump
cycle, achieving a 65% and 145% increase in COPc,th and COPh,th respectively,
with an expander inlet temperature of 250 ◦ C.
Lastly, it has been demonstrated that when integrated within a trigeneration
system with a gas turbine as the PGU, the optimum first stage work expander
inlet temperature is between 240◦ C and 275 ◦ C. The cycle was able to achieve a
maximum COPc of 0.84, and maximum COPc of 1.45 in this scenario (i.e. for
every 1 kWh of exhaust heat, 0.84 kWh of coolth and 1.45 kWh of hot water are
produced).
A considerable number of alternative cycles have been conceptualised, modelled and analysed, but are not presented within this Thesis. These are viewed as
stepping stones towards the cycles that have been presented within this Chapter,
and form the basis of the trial - and - error style design process.
114
Chapter 4
Chapter 5
R744 Heat Pump Cycle
Preliminary Component Analysis
115
Hybrid Trigeneration - Thermally Activated Heat Pumps
5.1
Introduction
This Chapter provides an overview of preliminary component analysis for the
novel thermally activated R744 simultaneous heating and cooling heat pump cycle
‘Cycle 2’ discussed in Chapter 4. Components analysed include:
• The supercritical fluid cooler / hot water generator, which cools the working
fluid following the first stage compressor and the hot side of the internal
heat recuperator.
• The internal heat recuperator, which preheats the supercritical fluid exiting
the second stage compressor prior to the external heat addition.
• The ejector.
• Concepts for the compressor and work expander.
5.2
Heat Exchanger Modeling
Temperature pinch in heat exchangers is modeled using a discretised − N T U
method. The − N T U method can be used to calculate the rate of heat transfer
in heat exchanger when the Log Mean Temperature Difference method is not
suitable, for example when cp and the heat transfer coefficient, U , are not constant. The Number of Transfer Units, N T U , is defined by Equation 5.1, where
A is the heat transfer area.
NT U =
UA
Cp,min
(5.1)
The effectiveness, , is defined by Equation 5.2.
=
Where Cr =
116
1 − e(−N T U (1−Cr ))
1 − Cr e(−N T U (1−Cr ))
(5.2)
Cp,max
.
Cp,min
Chapter 5
Hybrid Trigeneration - Thermally Activated Heat Pumps
The heat exchanger is discretised into > 250 blocks (n number), with fluid
thermophysical properties recalculated for each block.
The numbering convention used to describe the counterflow heat exchanger is
shown in Figure 5.1.
Figure 5.1: Diagram to indicate the numbering convention for counterflow heat
exchanger
The − N T U method requires the hot and cold side inlet temperatures to
be known for every block; this is not possible with a discretised counterflow
heat exchanger. Whilst iterative finite volume techniques (such as are used in
computational fluid dynamics software) would yield a solution, such software is
not available and creating an custom code script would take considerable time and
may require significant computational time. Therefore, a simple finite difference
method is proposed, which relies on an assumption. The assumption produces
an error (e), such that limn→∞ e = 0.
The temperature and thermophysical properties at the hot side outlet (i.e.
expander outlet) and cold side inlet to the heat exchanger side are both known;
this means that T1,i=1 and T2,i=n are known. In order to resolve the thermophysical properties in, say, block i = 1 using the − N T U method, T2,i=1 must be
known; however this is an output of the model, not an input. If T2,i=1 is assumed,
T3,i=1 and T4,i=1 can be calculated using an appropriate empirical heat transfer
Chapter 5
117
Hybrid Trigeneration - Thermally Activated Heat Pumps
relationship. The calculated values for T2,i=1 and T3,i=1 are then assigned to T4,i=2
and T1,i=2 , respectively. However, the same problem occurs again; T2,i=2 but now
be assumed.
Given that the hot side inlet ,T1,i=1 , is known (defined by the expander outlet
conditions), a sensible estimate for the cold side outlet, T4,i=1 , can be made (with
a value much less than T1,i=1 ). If the temperature drop of each side across the
block is very small compared to the temperature difference between the hot and
colde side then T4,i=1 can be assumed to be equal to T2,i=1 (which is known), such
that:
if
T1,i − T3,i T1,i − T4,i
(5.3)
T2,i ' T4,i
(5.4)
then
Each block can now be solved linearly and the temperature profile of the two
streams can be ascertained, yielding the pinch point temperature and location
along the heat exchanger. The cold side outlet temperature, T4,i=1 , is then increased and the calculations re-run until. This is repeated until the desired pinch
point temperature is achieved. This is a semi-iterative scheme, as demonstrated
in the algorithm which follows in Figure 5.2.
The development and validation of a three - pass R744 air cooled gas cooler
is discussed by Yin, Bullard and Hrnjak [101]. The model was based upon the
Gnielinski empirical relationship for heat transfer in a micro - tubes, and was
found to be within 2% of the results of the physical test. Therefore, in this
instance the empirical Gnielinski-Petukhov relationship for turbulent flow within
an annulus [102] is used to calculate the overall heat transfer coefficient, U , for
each block of the shell and tube heat exchanger. The cold side is the flow within
118
Chapter 5
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 5.2: Alrogithm used to model a counterflow heat exchanger
the annulus, and the hot side is the flow around the annulus.
N ucold,i =
· Recold,i · P rcold,i · (1 + Dinner /L)0.666
0.5
f
0.666
1.07 + 12.7 · cold,i
·
P
r
−
1
cold,i
8
fcold,i
8
N uhot,i = N ucold,i · 0.86 · (Douter /Dinner )0.16
Chapter 5
(5.5)
(5.6)
119
Hybrid Trigeneration - Thermally Activated Heat Pumps
The f , used in the calculations for N u number, is calculated using the Filonenko (1954) relationship defined by Equation 5.7.
fcold,i = (0.79 · ln (Recold,i ) − 1.64)−2
(5.7)
The Prandtl Number, P r, represents the ratio of momentum diffusivity to
thermal diffusivity, and the Reynolds Number, Re, represents the ratio of inertial
forces to viscous forces.
By assuming that each annulus represents an individual tube of a single pass
countrerflow shell and tube heat exchanger. Assuming that each tube and surrounding fluid is adiabatic (i.e. the temperature profile on the shell (hot) side is
uniform in a cross sectional slice of the heat exchanger), the entire shell and tube
heat exchanger can be approximated using a series of parallel annuli.
Figure 5.3: Diagrammatic representation of a cross section through the shell and
tube heat exchanger
A diagrammatic representation of a cross section through the shell and tube
120
Chapter 5
Hybrid Trigeneration - Thermally Activated Heat Pumps
heat exchanger is given in Figure 5.3. The inner diameter of the annular flow, Di ,
is the inner diameter of the tube. The outer diamter of the annular flow, Do , is
the distance between interference zones in the shell, which is represented by the
dotted lines on Figure 5.3. The tubes are relatively thin (1.5 mm), therefore it
is assumed that they have infinite heat transfer conductance (zero heat transfer
resistance).
5.2.1
Hot Water Generator
Optimisation of the supercritical fluid cooler / hot water generator is vital in ensuring the maximum COP is achieved. [75]. The design and analysis of R744 hot
gas coolers is well documented [76, 101, 103]. US patent #2010/0313582 A1 [99]
is for a high efficiency R744 heat pump cycle, and specifically identifies the use
of a internal recuperator to cool the working fluid prior to the expansion value,
which is preceded by a gas cooler; this is similar in concept to Cycle 2, which has
a double - stage gas cooler preceded by an internal heat recuperator.
The hot water generator is modeled as a twin - parallel, two - stage, counter flow, single - pass shell and tube type heat exchanger. It cools the refrigerant at
two points within the cycle and provides hot water for external use. The hot
side flow inlets are the compressed flow exiting the first stage compressor (the
‘hot - side low pressure flow’) and the outlet of the hot side of the internal heat
recuperator. (the ‘hot - side high pressure flow’). The fluid entering at both of the
cold side inlets is water at ambient temperature. A schematic of the hot water
generator with approximate inlet and outlet temperatures is given in Figure 5.4.
Figure 5.4: Schematic of the hot water generator
Chapter 5
121
Hybrid Trigeneration - Thermally Activated Heat Pumps
The two hot - side flows are at different pressures: 97 bar for the hot - side low
pressure flow and ∼150 bar for the hot - side high pressure flow. The difference
in pressures means that there thermophysical properties of the working fluid as
it moves along the tubes and cools vary differently; the Cp is very non - linear
in the hot - side low pressure flow. However, both flows can be made to exhibit
a very similar temperature glide throughout the heat exchanger by altering the
flow speeds and ratio of heat transfer area (i.e. number of tubes). Therefore, it
is possibly for the two flows to be cooled within a single heat exchanger, thus
simplifying the system complexity, cost and size.
The hot - side low pressure flow and hot - side high pressure flow pass through
tubes of inner diameter 3 mm and outer diameter 4.5 mm (safety factor of > 3
for mild steel). In total there are 1,000 tubes; 525 on the hot - side low pressure
flow and 475 on the hot - side high pressure flow. These criteria ensure that the
pressure drop through the tubes is below 1 bar.
The calculated temperature profile of the hot - side low pressure flow and associated cold - side is shown in Figure 5.5. The cooling water flow is provided in
two stages which both have inlet temperature of 25 ◦ C and outlet temperatures
of 60 ◦ C and 40 ◦ C. The mass flowrate of cooling water in the first - stage of the
hot water generator is the equal to the mass flow rate of refrigerant. However,
the mass flowrate of cooling water in the second - stage of the hot water generator
is 230% greater than the mass flow rate of refrigerant. This large increase is due
to the increase in Cp in working fluid as it passes the second - stage of the hot
water generator and is close to the critical point. A slight ‘bulge’ in the temperature profiles can be observed in the second - stage, as the working fluid is around
40 ◦ C. The lengths of the first - stage and second - stage are 1.34 m and 1.62 m,
respectively.
The calculated temperature profile of the hot - side high pressure flow and
associated hot - side is shown in Figure 5.6. As before, the cooling water flow is
provided in two stages which both have inlet temperature of 25 ◦ C and outlet
122
Chapter 5
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 5.5: Hot water generator hot - side low pressure temperature profile
temperatures of 60 ◦ C and 40 ◦ C. As before, the mass flowrate of cooling water
in the first - stage of the hot water generator is equal to the mass flow rate of the
refrigerant. The mass flowrate of cooling water in the second - stage of the hot
water generator is 33% greater than the mass flow rate of refrigerant. This is a
significant reduction from the hot - side high pressure flow second - stage, which is
due to the reduction in Cp . The lengths of the first - stage and second - stage are
1.19 m and 1.50 m, respectively.
This analysis demonstrates that the two parallel hot - side flows (high pressure
and low pressure) exhibit very similar temperature profiles. Therefore, both flows
can be integrated within a single heat exchanger, provided that number and
length of tubes is selected appropriately. The approximate specification given
in this analysis (for a 1 MWc heat pump) indicate that when compared to the
complexity and number the heat exchangers required in a triple - stage LiBr H2 O absorption heat pump, the hot water generator proposed here is relatively
technically simple, but requires a high number of small diameter tubes. A more
detailed analysis would provide possible methods of conducting a detailed design
Chapter 5
123
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 5.6: Hot water generator hot - side high pressure temperature profile
of a heat exchanger which fulfills the requirements of the hot water generator.
5.2.2
Internal Heat Recuperator
The internal heat recuperator is modeled as a single - pass shell and tube heat
exchanger. It allows the cycle to re-use the relatively high grade thermal energy
at the first - stage work expander outlet, and transfer it to the high pressure
low leaving the second stage compressor. This increases the cycle performance
by reducing the heat input requirement from external sources and increases the
exergetic efficiency of the cycle.
Chapter 4 highlighted the effect of the first - stage work expander inlet temperature upon the COP, and eluded to the fact that it is partly due to an increasing
‘pinch effect’ in the internal heat recuperator when the inlet temperature, and
consequent outlet pressure, are reduced.
The higher pressure cold side passes through the tubes, whilst the lower hot
side passes through the shell. There are 3,000 mild steel tubes of inner diameter 4 mm and outer diameter 5 mm. A minimum pinch temperature of 8◦ C is
124
Chapter 5
Hybrid Trigeneration - Thermally Activated Heat Pumps
specified; based on simulations, this is believed to be a value which prevents the
overall length of the heat exchanger becomes too large.
The temperature profile of the internal heat recuperator for the case where
the first - stage work expander inlet temperature is 350◦ C is given in Figure 5.7.
Figure 5.7: Internal heat recuperator temperature profiles (first - stage work expander inlet temperature of 350 ◦ C
For the case shown in Figure 5.7, the overall bulk heat transfer value, U , is
586 W/m2 .K. The local heat transfer conductance varies along the heat exchanger
between 552 W/m2 .K and 620 W/m2 .K. The total length of the tubes is 6.4 m,
which indicates that the the internal heat recuperator will be large and may be
expensive.
5.3
5.3.1
Ejector Modelling
Review of previous work
Nakagawa et al. [104] conducted a physical test to analyse the effect of internal
heat exchanger (from gas cooler outlet to evaporator outlet) on the performance
Chapter 5
125
Hybrid Trigeneration - Thermally Activated Heat Pumps
of a two - phase ejector R744 refrigeration cycle. The results suggested an improvement in COPc of upto 27% can be achieved with the addition of an ejector
and maximum heat recuperation. It is suggested that a higher COPc could be
attained through the recovery of more heat, although this may cause overheating
of the compressor.
The results also show that the COPc is more sensitive to gas cooler pressure
in the ejector cycle than in a basic R744 refrigeration cycle; the motive fluid inlet
total pressure should be at least 100 bar in order to obtain a reasonable performance gain. It is noted that the “ejector system without [internal heat exchanger]
had lower COP than similar conventional systems in most of the conditions used,
except at high Pc ” [104].
The study shows that for the four configurations of basic R744 refrigeration
cycles analysed (with internal heat exchanger, without internal heat exchanger,
with ejector and without ejector), an increase in gas cooler temperature decreases
COP. It also shows that an increase in evaporator temperature increases COP.
Li and Groll [87] provide a further analysis of an ejector - expansion device
within a transcritical R744 refrigeration cycle when used for air conditioning
purposes. The effect odf the entrainment ratio and pressure drop in the receiving
section of the ejector on the relative performance of the cycle was investigated
using a theoretical model. It is found that the ejector provides a 7 - 18% increase
in COP over the basic transcritical R744 cycle.
An experimental evaluation of a prototype R744 ejector was conducted by
Elbel and Hrnjak [105], with the aim of reducing throttling losses in transcritical
R744 refrigeration cycle operation. The study was based upon a full cycle simulation defined within EES, where 2,500 non - linear time - dependent equations.
Experimentally obtained correlations for volumetric, compression and isentropic
efficiency are used within the compressor module. Assumptions include:
• Fluid properties are uniform across the cross - sectional diameters
• Thermodynamic non - equilibrium effects are not considered
126
Chapter 5
Hybrid Trigeneration - Thermally Activated Heat Pumps
• The ejector is to be treated as adiabatic
• Inlet and outlet flow kinetic energies are to be treated as negligible
• Ejector geometries are not accounted for
A needle extending into the throat of the motive nozzle allows for the control
of the high - side pressure; this represents an additional degree of freedom in the
system. The study indicated that the value of mixing pressure is irrelevant, so
long as it is lower than the entrained inlet pressure.
5.3.2
Ejector
The operating principle of a two - phase ejector is based upon the conversion
of static pressure into kinetic energy and vice versa, as defined by Bernoulli’s
Principle.
The high velocity low static pressure refrigerant (‘motive flow’) entering the
ejector is accelerated in the nozzle. The motive flow is throttled to such an extent
that it has a lower static pressure that the low pressure dry saturated refrigerant
vapour exiting the evaporator (‘entrained flow’), but a much higher velocity. This
allows the motive flow to transfer momentum to the entrained flow, this process
occurs in the mixing chamber.
Following the mixing chamber, the two - phase high speed mixed flow is expanded in the diffuser, where it is slowed down and regains static pressure. The
two - phase flow exiting the diffuser has a resultant pressure which is higher than
the entrained flow entering the ejector. Figure 5.8 shows a basic schematic of an
ejector.
Modelling
In order to model the ejector a number of state points must first be assigned,
these are illustrated on Figure 5.9.
Chapter 5
127
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 5.8: Schematic diagram to illustrate the fluid inlets and outlets of an R744
ejector
Figure 5.9: Schematic diagram to illustrate the state points in the R744 ejector
A series of simple equations enable the input and output states of the ejector
to be modeled. This relies upon a number of assumptions regarding isentropic
efficiencies, conservation of mass, momentum and energy.
The ratio of the entrained mass flowrate to the motive mass flowrate as the
termed the entrainment ratio, µ, and is calculated using Equation 5.8:
µ=
ṁ8
ṁ4
(5.8)
Assuming an isentropic efficiency for the motive flow nozzle, η4n , the actual
value for the enthalpy at state point 4i (ie. the motive flow at the motive nozzle
exit) can be calculated using Equation 5.9:
h4i = h4 −
η4n
h4 − h4s
(5.9)
By applying the conservation of momentum and energy, the velocity of the
motive fluid as it exits the motive nozzle, u4 i, can be calculated from Equation
5.10:
u4i =
128
p
2 · (h4 − h4i )
(5.10)
Chapter 5
Hybrid Trigeneration - Thermally Activated Heat Pumps
Assuming the flow is fully mixed at the outlet of the mixing section, the
velocity of the combined flow at the outlet of the mixing section can be calculated
using Equation 5.11:
u5 = ρ 5 ·
1
ρ4i u4i (1 + µ)
+
µ
ρ8i u8i (1 + µ)
−1
(5.11)
, where u8i is approximated by Equation 5.12:
u8i =
p
2 · (h8 − h8i )
(5.12)
, and densities are calculated from fluid thermophysical properties. The enthalpy of the entrained flow nozzle outlet,h8i , is calculated using assumed isentropic efficiency for the entrained nozzle, η8n , and Equation 5.13:
h8i = h8 −
η8n
h8 − h8s
(5.13)
The enthalpy of the combined flow, h6x , can be defined by Equation 5.14
(assuming that the process is adiabatic):
h6x =
h4
1+µ
+ h8 ·
µ
1+µ
(5.14)
Similarly, the enthalpy of the combined flow, h6x , can be also be calculated
assuming an isentropic efficiency of the diffuser, ηd , using Equation 5.15:
h6x =
h5s − h5
+ h5
ηd
(5.15)
The enthalpy of the fluid at the end of the mixing section can be calculated
using the conservation of energy and mass, as shown in Equation 5.16:
h5 =
h4 + µh8 1 2
− u5
1+µ
2
(5.16)
The pressure of the combined flow at the outlet of the ejector can be calculated
via assuming a value for the pressure at the end of the mixing section, P5 , with
Chapter 5
129
Hybrid Trigeneration - Thermally Activated Heat Pumps
an iterated procedure and Equation 5.17:
1
µ
P6x ·
+
+ u5 =
u4i · ρ4i · (1 + µ) u8i · ρ8i · (1 + µ)
µ
1
µ
1
+
+
· u4i +
· u8i
P5
u4i · ρ4i · (1 + µ) u8i · ρ8i · (1 + µ)
1+µ
1+µ
(5.17)
An approximation of the motive flow nozzle shape and combined flow diffuser shape is calculated based upon a linear increase / decrease in fluid velocity,
respectively. This is calculcated using a simple 1 dimensional analysis, and is
expected to serve as a concept design, which will be further optimised with 2 or
3 dimensional computational fluid dynamics packages and / or physical testing.
Figure 5.10 shows the calculated profile for the motive flow nozzle and the
corresponding mean bulk fluid velocity. It can be noted that the nozzle is of a
particularly elongated shape; this is believed to be due to the drastic drop in
bulk fluid density once the R744 drops below the saturation curve as the static
pressure drops (at ∼ 60 bar), as illustrated in Figure 5.11. As the fluid density
drops, the consequent reduction in cross - sectional area to maintain a constant
increase in velocity reduces.
The length of the nozzle section is shown as dimensionless on Figure 5.10, and
will require further analysis to ascertain the optimum length. If the nozzle is too
long, the boundary layer friction losses will be too great and isentropic efficiency
will reduce; if the nozzle is too short there will be undesirable flow characteristics
and the isentropic efficiency will reduce.
Figure 5.12 shows the calculated profile for the combined flow diffuser and
the corresponding mean bulk fluid velocity. As was noted with the nozzle, the
diffuser is also elongated in shape; this is also due to the non-linearity of bulk
fluid density as the fluid velocity reduces, as illustrated in Figure 5.13.
130
Chapter 5
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 5.10: Ejector motive flow nozzle shape and mean bulk fluid velocity along
the nozzle length
Figure 5.11: Ejector bulk fluid density along the nozzle length
5.4
5.4.1
Compressor / Work Expander Modeling
Centrifugal Compressor
Cycle 2 has two compressors, and a high level preliminary design of the blade
channels is given here.
Chapter 5
131
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 5.12: Ejector combined flow nozzle shape and mean bulk fluid velocity
along the nozzle length
Figure 5.13: Ejector bulk fluid density along the diffuser length
A notation for the cross section view and top view of the impeller blades is
defined in Figures 5.14 and 5.15.
The area of the impeller channel inlet can be defined in accordance with Figure
132
Chapter 5
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 5.14: Cross sectional view and notation of compressor impeller blade profile
Figure 5.15: Top view and notation of compressor impeller blade profile for the
intake area
5.15 and Equation 5.18.
Chapter 5
133
Hybrid Trigeneration - Thermally Activated Heat Pumps
Ap = π(rb2 − ra2 )
θ
θ
= ((rb2 − ra2 )
2π
2
(5.18)
Assuming that the lower curve on Figure 5.14 is circular (i.e. H = R1 ), and
therefore the channel has an arc shape, the cross sectional area of the channel
can be defined as:
θ
1
Ap = ((rb2 − ra2 )
,
2
cosλ
∀θ, θ 6= 0, θ 6=
π
2
(5.19)
Given that ra = R1 − c, it can be deduced that:
1
c = 2R1 sin
2
2(π − λ)
2
= R1 cosλ
(5.20)
Now, ra can be defined as R1 (1 − cosλ), and therefore:
A−p=
θ 2
rb − (R1 (1 − cosλ))2
2cosλ
(5.21)
This can be re-arranged to enable rb to be calculated, as follows:
2cosλ
rb = Ap
+ (R1 (1 − cosλ))2
θ
21
(5.22)
The impeller blade inlet and outlet geometry has been calculated as is included
in the Appendix. For the sake of brevity, it has not been included in the main
body of the thesis, given the established nature of research in this area. However,
of particular note is the very compact size of the impeller, due to the high density
134
Chapter 5
Hybrid Trigeneration - Thermally Activated Heat Pumps
of supercritical R744.
5.4.2
Gerotor
A gerotor is a type of positive displacement device, which is a type of trochoidal
device (along with the Wankel engine). Gerotors are suitable as work expansion
and compression devices where the working fluid is of a relatively high density
(i.e. similar magnitude to that of water at Normal Conditions). Gerotors are
commonly used as pumps, but there has been recent interest in exploring their
use in refrigeration cycles, Organic Rankine Cycles and low temperature Brayton cycles [106–110]. Various methods for the generation of trochoidal device
geometries are discussed in the literature [111–113], along with analyses of their
performance and flow characteristics and optimisation [114–117]. An alternative
method for the generation of a gerotor profile is presented later in this section.
Trochoidal devices are part of the family of planetary rotation machines and
offer advantages over other types of machines; eg. reliability, simplicity, possibility of higher speeds and a wide variety of applications (pumps, compressors, expanders, blowers, engines) [118]. There are two major components in a trochoidal
device; the rotor and the chamber (piston and cylinder), which are referred to as
the trochoid and envelope, and vice versa. In a trochoidal type pump, such as
a gerotor, the rotor is a trochoid and the chamber is an envelope, however in a
Wankel engine the rotor is an envelope and the chamber is a trochoid.
The theory of trochoidal devices in summarised by Shung and Pennock. [118].
A trochoid is defined by the path created by a fixed point on a circle of radius rt,m
with distance rc to the cycle centre, which rolls on a fixed circle of radius rt,f ; the
moving and fixed pitch circles, respectively [118]. If the moving pitch circle has a
greater radius than the fixed pitch circle (ie. rt,m > rt,f ), the trochoid is called an
epitrochoid. If the moving pitch circle has a smaller radius than the fixed pitch
circle (ie. rt,m < rt,f ), the trochoid is called a hypotrochoid. If the moving pitch
circle has the same radius as the fixed pitch circle (ie. rt,m = rt,f ), the trochoid
Chapter 5
135
Hybrid Trigeneration - Thermally Activated Heat Pumps
is called a cycloid. Trochoids are further categorised into prolate, ordinary and
curtate if they are smooth, have a cusp, or cross over themselves, respectively.
Ordinary hypotrochoids and epitrochoids are also classified as epicycloids and
hypocyloids, respectively. Observation of gerotors indicates that they are often
generated using prolate hypotrochoids, or some variation thereof.
A notation system has been introduced to describe the geometry of the trochoids analytically [118]. The centers of the fixed and moving pitch circles are
denoted Ot,f and Ot,m , respectively. The reference line is defined as the line con−−−−−→
necting the two centres, Ot,f Ot,m . The reference line passes through the point
of contact of the two circles, point P . The radii of the fixed and moving pitch
circles are rt,f and rt,m , respectively; the distance between the centres, referred
to as the eccentricity, is denoted by et =| rt,f − rt,m |. The Cartesian reference
−−−−→
frame attached to the fixed pitch circle is denoted by the vectors Ot,f Xt and
−−−−→
Ot,f Yt , representing the Xt and Yt axis respectively, and is referred to as the trochoid frame. The cartesian reference frame attached to the moving pitch circle
−−−−→
−−−−→
is denoted by the vectors Ot,m X0t and Ot,m Yt0 , representing the Xt0 and Yt0 axis
respectively, and is refered to as the reference frame. When the reference line
−−−−−→
(Ot,f Ot,m ) rotates at an angle α to the Xt axis, the generating frame turns at an
angle
α
kt
relative to the same axis, where kt is referred to as the angular velocity
ratio for the trochoid. A positive value of kt implies an epitrochoid, whilst a
negative value of kt implies a hypotrochoid.
The trochoid generation point, C, is located on the Xt0 axis at a radius of rc
from the moving pitch circle centre Ot,f . The coordinates of the generation point
C is defined by equations 5.23 5.24. Figure 5.16 gives a visual representation of the
geometry and notation used in the generation of a trochoid, using a peritrochoid
as an example.
The generation point of the trochoid is defined by Equations 5.23 and 5.24,
−−−−→
−−−−→
which are expressed in relative to the trochoid frame (Ot,f Xt and Ot,f Yt ).
136
Chapter 5
Hybrid Trigeneration - Thermally Activated Heat Pumps
Xt = et cos α + rc cos
α
kt
(5.23)
Yt = et sin α + rc sin
α
kt
(5.24)
Figure 5.16: Geometry of a trochoid (peritrochoid) [118]
There are two types of conjugate envelopes that can be generated using a
trochoidal profile; an inner envelope and and outer envelope, which are located
inside and outside the trochoid profile, respectively. The conjugate envelope is
defined by the maximum inner or minimum outer envelope for which there is no
interference throughout one entire motion; the contact force between the trochoid
Chapter 5
137
Hybrid Trigeneration - Thermally Activated Heat Pumps
and the envelope is zero when they are static. Further, and most crucially for a
positive displacement device, the compartment sealing is inherently optimised by
the geometry of the two parts.
The generation of the conjugate envelope can be best visualised by imagining
that the trochoid is initially surrounded by lots of tiny beads, and as it undergoes
planetary motion (as guided by the moving and fixed pitch circles) the beads are
pushed outwards by the trochoidal lobes, resulting in the identification of the
outer conjugate envelope profile. The same principle can be used to explain the
generation an inner conjugate envelope, except the beads are inside the trochoid
and the moving and fixed pitch circles are reversed.
The geometry used to describe the conjugate envelope is much the same as the
geometry used to describe the trochoid. The centers of the fixed and moving pitch
circles are denoted Oe,f and Oe,m , respectively. The reference line is defined as
−−−−−−→
the line connecting the two centres, Oe,f Oe,m . The reference line passes through
the point of contact of the two circles, point P . The radii of the fixed and moving
pitch circles are re,f and re,m , respectively; the distance between the centres,
referred to as the eccentricity, is denoted by ee =| re,f − re,m |. The Cartesian
−−−−→
reference frame attached to the fixed pitch circle is denoted by the vectors Oe,f Xe
−−−−→
and Oe,f Ye , representing the Xe and Ye axis respectively, and is referred to as
the trochoid frame. The Cartesian reference frame attached to the moving pitch
−−−−→
−−−−→
circle is denoted by the vectors Oe,m Xt and Oe,m Yt , representing the Xt and Yt
axis respectively, and is referred to as the reference frame. When the reference
−−−−−−→
line (Oe,f Oe,m ) rotates at an angle α to the Xe axis, the generating frame turns
at an angle
β
ke
relative to the same axis, where kt is referred to as the angular
velocity ratio for the envelope. A positive value of kt implies an that the angular
−−−−−−→
displacement of the reference line Oe,f Oe,m is in the same direction as the angular
displacement of the trochoid frame. Figure 5.17 gives a visual representation of
the geometry and notation used in the generation of the envelope.
The rotor profile can be defined according to Equations 5.25 and 5.25, which
138
Chapter 5
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 5.17: Geometry of a trochoid envelope [118]
calculate the coordinates of any point on the rotor.
Xe = ee cos β + Xt cos
Ye = ee sin β + Xt sin
β
ke
β
ke
− Yt cos
− Yt sin
β
ke
β
ke
(5.25)
(5.26)
Substituting Equation 5.23 into 5.25 and Equation 5.24 into 5.26, the coordinates of the generating point are now:
β
Xe = ee cos β + et cos α +
ke
Chapter 5
− rc cos
α
β
+
kt ke
(5.27)
139
Hybrid Trigeneration - Thermally Activated Heat Pumps
β
α
β
Ye = ee sin β + et sin α +
− rc sin
+
ke
kt ke
(5.28)
According to Shund and Pennock [118], the condition for a Type 2 conjugate
envelope defined by Equations 5.29, 5.30, 5.31, 5.32 & 5.33:
ke = kt − 1
(5.29)
rc
2
2µ
Xt = −2 cos 1 −
µ cos µ + rc cos
kt
kt
kt
(5.30)
2
2µ
rc
µ sin µ + rc sin
Yt = −2 cos 1 −
kt
kt
kt
(5.31)
1
kt
µ=
α+
β
2
kt − 1
(5.32)
1
2 − kt
ν=
α+
β
2
kt − 1
(5.33)
An alternative method of generating a trochoid inner rotor and outer rotor is
presented below, beginning with the outer rotor. The number of lobes that the
inner rotor has is denoted, N , such that the angle of rotation of the inner rotor,
θi is defined by:
θi = θo
N
N +1
(5.34)
where θo is the angle of rotation of the outer rotor. Therefore, we can define
the rotation of the contact point frame, θc , is N θo .
Denoting r as the fixed pitch circle diameter and E as the distance between
the fixed and the moving pitch circles, it can be deduced that the radius of the
fixed pitch circle (outer rotor), r3 , is defined by rE. Likewise, the radius of the
moving pitch circle (inner rotor), r2 , is defined by r2 = E(N + 1), according
140
Chapter 5
Hybrid Trigeneration - Thermally Activated Heat Pumps
to Kennedy’s theorem. Figure 5.18 shows the geometry and notation for the
coordinate systems, and contact point, C.
Figure 5.18: Coordinate systems and contact point for the outer rotor
It is possible to define ω on Figure 5.18 using Equation 5.35:
ω = tan−1
sinθ
R
− cosθ
rE
!
(5.35)
The x and y Cartesian coordinates of the inner rotor are defined by Equations
5.36 and 5.37:
Xo = Rcosθ + Rc cos(θ + ω) + Ecos(N θ)
(5.36)
Yo = Rsinθ + Rc sin(θ + ω) + Esin(N θ)
(5.37)
Figure 5.19 illustrates the profile of a typical outer rotor, calculated using this
method.
Now the inner rotor can be defined, beginning with Equation 5.39 for the
coordinate system, Xi,c and Yi,c of the inner rotor for when αi < κ < αi+1 , ∀i, i =
(1, 2, 3, ..., N ) ·
Chapter 5
2π
,
N
where:
141
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure 5.19: Illustration of a typical outer rotor profile
−1
κ = tan
Yi,c
Xi,c
(5.38)
and:
Xi,c = Xo − E
(5.39)
Yi,c = Yo
It can now be derived that the radius of the inner rotor can be described using
the following Equations:
r = Rcos(κ − αi ) +
p
Rr2 − R2 + R2 cos2 (κ − αi )
(5.40)
and
142
Chapter 5
Hybrid Trigeneration - Thermally Activated Heat Pumps
π
r = δcos(κ − (αi + )) −
N
r
π 2
2
2
2
Rb − lf + lf cos (κ − (αi + ))
N
(5.41)
r
π
Rcos2 − R2 + (Rb + Rr )2 )
N
(5.42)
where
π
δ = Rcos +
N
and
r = Rcos(κ − αi+1 ) +
p
Rr2 − R2 + R2 cos2 (κ − αi+1 )
(5.43)
Rb is a user definable quantity, which adjust the profile of the inner rotor
between the lobes, thereby affecting the volume of the enclosed control volume
between the outer and inner rotors.
Equations 5.40, 5.41 and 5.43 are calculated N number of times (i.e. the
number of lobes on the inner rotor), where αi =
2π
N
· i, ∀i, i = (1, 2, 3, ..., N )
Figure 5.20 illustrates the profile of a typical inner rotor, along with its respective outer rotor calculated using this method.
Figure 5.20: Illustration of a typical inner and outer rotor profile
Chapter 5
143
Hybrid Trigeneration - Thermally Activated Heat Pumps
An illustration of the gerotor rotating through
2π
N
is shown in Figure 5.21.
Figure 5.21: Illustration a typical gerotor rotating through
5.5
2π
N
of a turn
Conclusions
This Chapter has set out a high level preliminary design of a number of components within the novel thermally activated R744 simultaneous heating and cooling
144
Chapter 5
Hybrid Trigeneration - Thermally Activated Heat Pumps
heat pump conceptualised in Chapter 4. A design methodology for R744 heat
exchangers (both R744 - R744 and R744 - water) have been provided. This was
used to give an estimation of the size, number and length of tubes required in
a shell and tube heat exchanger, for both the internal heat recuperator and the
twin - parallel, two - stage hot water generator.
The two - phase ejector has been modelled, and this model was primarily used
in the simulation of Cycle 2 in Chapter 4. The methodology and calculation
process behind the ejector model has been explained, along with a preliminary
assessment of the ideal motive flow nozzle and combined flow diffuser profiles for
ensuring linear acceleration / deceleration of the fluid. The ideal motive flow nozzle was found to be particularly elongated in shape, which is due to the substantial
decrease in density once the fluid drops into the saturation region.
It was intended to continue analysing the optimum shape of the ejector using
a 2 or 3 dimensional fluid dynamics package. A database of R744 thermophysical properties was created, which was formatted into a manner compatible with
ANSYS CFX. However, the author was unable to make the software accept the
database in the time available, despite repeated attempts. Therefore, 2 or 3
dimensional analysis was not possible.
Potential design methodologies for the compressors and work expanders have
been explored, including: a centrifugal compressor and a trochoidal gerotor pump.
A centrifugal compressor impeller design methodology has been presented, and
this is included with Appendix B. The basic theory trochoidal devices has been
discussed, and a methodology for the generation of an outer and inner rotor profile
is given. The author’s intention was to continue this area of research by analysing
centrifugal compressor concepts and gerotor concepts in a fluid dynamics package,
but this was not possible.
Chapter 5
145
Chapter 6
Overall Conclusions
Worldwide demand for air conditioning is predicted to increase by 72 % by the
year 2100, primarily due to the increase in economic productivity of south east
Asia and Africa, which will result in an improvement in living standards and a
consequent increase in demand for air conditioning. It is predicted that this effect
will be compounded by climate change. Therefore, the efficient, low carbon and
low cost provision of chilled water for air conditioning systems is both a huge
challenge and opportunity. Trigeneration (the combined generation of electricity,
heating and cooling) is a progression of cogeneration (the combined generation
of electricity and heating), which will enable this goal to be reached.
This Thesis begins by providing an overview of trigeneration system design,
thermal storage technologies and trigeneration control strategies. Trigeneration
is shown to deliver an improved EUF over a system where electricity, heating and
cooling are produced separately [21, 28–30]. Trigeneration control and operation
strategies are discussed and analysed at length in a number of studies [23, 25,
26, 33, 36, 37], which conclude that the ideal control and operation strategy
is dependent upon climatic conditions and the performance parameters of the
trigeneration system.
Thermally activated heat pumps which can be integrated into trigeneration
systems are identified as a key research topic. Chapter 2 begins with a review
of single component and multi component, zeotropic and azeotropic refrigerants.
146
Hybrid Trigeneration - Thermally Activated Heat Pumps
Natural refrigerants such as R744 and R717 are identified as featuring heavily
in research over the past decade, owing to a continued move towards low GWP
and low ODP refrigerants, driven in part by the Montreal protocol. The use of
highly zeotropic refrigerant mixtures, such as LiBr - H2 O and H2 O - NH3 , in thermally activated absorption heat pumps is discussed, and a number of concepts
and recent advances to increase the COP are presented. These include: multi stage / multi - effect cycles, generator - absorber exchanger cycles and vapour recompression cycles. Triple - stage triple - effect absorption chillers are capable of
achieving a COPc of approximately 1.6, but suffer from corrosion in the upper
generator due to the high temperatures [48]. The COP is both a function of the
upper generation temperature, chilled water temperature, load conditions and the
flow arrangement. A brief review of multi - bed and multi - stage adsorption heat
pumps, and sorbent and sorbate pairs demonstrates that COP of this technology
is still far below that of absorption heat pumps. A review of R744 Mechanical
Vapour Compression (MVC) heat pumps along with a review advances Supercritical Carbon Dioxide (S-CO2) power cycle concepts is provided. It is evident from
the review of literature that there is a small scope for innovations in multi - stage
multi - effect absorption heat pump cycle design, and a significant scope for innovations in R744 heat pump cycle design. Therefore, the analysis of multi - stage
multi - effect absorption heat pump cycle and R744 heat pump cycle concepts
form the key contributions to research presented within this Thesis.
Chapter 3 explores two types of thermally activated absorption heat pumps:
multi - effect / multi - stage H2 O - NH3 absorption heat pumps and multi - effect / multi stage LiBr - H2 O compression - absorption heat pumps. It is demonstrated that
the double - stage / triple - effect H2 O - NH3 absorption heat pump cycle can achieve
a COPc equivalent to a triple - stage / triple - effect LiBr - H2 O absorption heat
pump, but with one fewer stage.
The sequence of internal heat recuperation devices within the cycle is shown
to be vital to the performance. The double - stage / triple - effect H2 O - NH3 abChapter 6
147
Hybrid Trigeneration - Thermally Activated Heat Pumps
sorption heat pump cycle is able to achieve a theoretical COPc of 1.765. A triple stage / quintuple - effect cycle is conceptualised, but temperature required in the
upper generator is too high for the thermophysical properties of the ammonia
and water mixture to be resolved.
Whilst the double - stage / triple - effect H2 O - NH3 absorption heat pump cycle
appears to be promising, a number of patents for similar concepts already exist
or have existed [52, 53]. Therefore this cycle is not considered suitable for further
research and development.
Singe - stage / single - effect, double - stage / double - effect and triple - stage / triple effect LiBr - H2 O compression - absorption heat pump cycles are demonstrated to
achieve a significant improvement in COPc over their conventional absorption
heat pump cycle counterparts. The triple - stage / triple - effect LiBr - H2 O compression - absorption heat pump is shown to achieve a COPc,th of 2.0, when the
steam compressor has a pressure ratio of 2.0. The net effect of the addition
of the compressor to the evaporator outlet is not only an increase in COPc,th ,
but also a reduction in generation temperature. The corresponding COPc,th is
demonstrated to be over 16.0 in all scenarios analysed.
Whilst the triple - stage / triple - effect LiBr - H2 O absorption - compression heat
pump cycle appears to be a promising area for future research, the extremely low
pressure and density of the steam exiting the evaporator means that the very
high volumetric flowrate through the compressor would result in it being almost
impracticably large. (6.1 m3 /s for 100 kW of cooling output). Therefore the absorption - compression cycle is not considered as a topic of further research.
Chapter 4 identifies three novel thermally and mechanically or pure thermally
activated transcritical R744 heat pump cycles which are conceptualised, modelled
and analysed. Each of these cycles are capable of providing simultaneous heating
and cooling, and are designed to use the thermal energy in the exhaust gases
from a Primary Generation Unit (PGU), in this instance a gas turbine. Cycle 2
is identifed as the preferred heat pump cycle as it provides a higher COPc and
148
Chapter 6
Hybrid Trigeneration - Thermally Activated Heat Pumps
COPh than Cycle 1. Cycle 3 is a minor evolution of Cycle 2 and is capable of
providing mechanical work output as well as producing hot water and chilled
water, however the small magnitude of the mechical work output and complexity
of a split flow compressor meant that Cycle 3 is not pursued further. Cycle 2 is
capable of providing coolth (10 ◦ C) at a COPc,th of 1.58 and hot water (65 ◦ C) at
a COPh,th 2.41, with a first stage expander inlet temperature of 350 ◦ C.
These performance figures for COPc are very similar to those theoretically
achieved by triple - stage LiBr - H2 O absorption chillers. However, Cycle 2 is capable also simultaneously producing hot water at 65 ◦ C. Cycle 2 requires half the
number of processes of a triple - stage LiBr - H2 O absorption chiller and uses a
single working fluid, further simplifying the system and the control.
It is demonstrated that Cycle 2 performs considerably better than a ‘separates
system’ of a supercritical CO2 Brayton cycle linked to a R744 MVC heat pump
cycle, achieving a 65% and 145% increase in COPc,th and COPh,th respectively,
with an expander inlet temperature of 250 ◦ C. It is also demonstrated that when
integrated within a trigeneration system with a gas turbine as the PGU, the optimum first stage work expander inlet temperature is between 240◦ C and 275 ◦ C.
The cycle is able achieve a maximum COPc of 0.84, and maximum COPc of 1.45
in this scenario (i.e. for every 1 kWh of exhaust heat, 0.84 kWh of coolth and
1.45 kWh of hot water are produced).
Three novel thermally and mechanically or pure thermally activated transcritical R744 heat pump cycles have been conceptualised, modelled and analysed.
Each of these cycles are capable of providing simultaneous heating and cooling,
and is designed to use the thermal energy in the exhaust gases from a Primary
Generation Unit (PGU), in this instance a gas turbine. Cycle 2 was identifed as
the preferred heat pump cycle as it provides a higher COPc and COPh than Cycle
1. Cycle 3 is a minor evolution of Cycle 2 and is capable of providing mechanical
work output as well as producing hot water and chilled water, however the small
magnitude of the mechical work output and complexity of a split flow compressor
Chapter 6
149
Hybrid Trigeneration - Thermally Activated Heat Pumps
meant that Cycle 3 was not pursued further. Cycle 2 is capable of providing
coolth (10 ◦ C) at a COPc,th of 1.58 and hot water (65 ◦ C) at a COPh,th 2.41, with
a first stage expander inlet temperature of 350 ◦ C.
These performance figures for COPc are very similar to those theoretically
achieved by triple - stage LiBr - H2 O absorption chillers, as demonstrated earlier
(n.b. typically COPc = 1.6). However, Cycle 2 is capable also simultaneously
producing hot water at 65 ◦ C. Cycle 2 requires half the number of processes of a
triple - stage LiBr - H2 O absorption chiller and uses a single working fluid, further
simplifying the system and the control. Chapter 5 provides an overview of an
initial component analysis and design for key components of Cycle 2, such as: the
internal heat recuperator, ejector, work expander and supercritical fluid cooler.
It was then shown that Cycle 2 performance considerably better than a ‘separates system’ of a supercritical CO2 Brayton cycle linked to a R744 MVC heat
pump cycle, achieving a 65% and 145% increase in COPc,th and COPh,th respectively, with an expander inlet temperature of 250 ◦ C.
Lastly, it has been demonstrated that when integrated within a trigeneration
system with a gas turbine as the PGU, the optimum first stage work expander
inlet temperature is between 240◦ C and 275 ◦ C. The cycle is capable achieve a
maximum COPc of 0.84, and maximum COPc of 1.45 in this scenario (i.e. for
every 1 kWh of exhaust heat, 0.84 kWh of coolth and 1.45 kWh of hot water are
produced).
Chapter 5 documents and initial component analysis and design for key components of Cycle 2, including: the internal heat recuperator, ejector, work expander and supercritical fluid cooler. A design and calculation methodology is
presented with the purpose of forming the continuation of this research from a
Technology Readiness Level1 of 2 (technology concept and / or application formulated) to a Technology Readiness Level of 3 (analytical and experimental critical
function and / or characteristic proof - of - concept).
1
Technology Readiness Level is a widely used concept to estimate the maturity of specific
technologies
150
Chapter 6
References
[1] Stan Cox.
Loosing Our Cool: Uncomfortable Truths About Our Air-
Conditioned World. New Press, 2010.
[2] Morna Isaac and Detlef P. van Vuuren. Modeling global residential sector
energy demand for heating and air conditioning in the context of climate
change. Energy Policy, 37(2):507–521, February 2009.
[3] K. Edem NTsoukpoe, Hui Liu, Nolwenn Le Pierrès, and Lingai Luo. A review on long-term sorption solar energy storage. Renewable and Sustainable
Energy Reviews, 13(9):2385–2396, December 2009.
[4] D Wu and R Wang. Combined cooling, heating and power: A review.
Progress in Energy and Combustion Science, 32(5-6):459–495, September
2006.
[5] Frank de Bruijn. The current status of fuel cell technology for mobile and
stationary applications. Green Chemistry, 7(3):132, 2005.
[6] Antoni Gil, Marc Medrano, Ingrid Martorell, Ana Lázaro, Pablo Dolado,
Belén Zalba, and Luisa F. Cabeza. State of the art on high temperature
thermal energy storage for power generation. Part 1Concepts, materials
and modellization. Renewable and Sustainable Energy Reviews, 14(1):31–
55, January 2010.
[7] H Michels and R Pitzpaal. Cascaded latent heat storage for parabolic trough
solar power plants. Solar Energy, 81(6):829–837, June 2007.
151
Hybrid Trigeneration - Thermally Activated Heat Pumps
[8] a Kandari. Thermal stratification in hot storage-tanks. Applied Energy,
35(4):299–315, 1990.
[9] U Herrmann. Two-tank molten salt storage for parabolic trough solar power
plants. Energy, 29(5-6):883–893, May 2004.
[10] B Zalba. Review on thermal energy storage with phase change: materials, heat transfer analysis and applications. Applied Thermal Engineering,
23(3):251–283, February 2003.
[11] Doug Brosseau, Mike Edgar, John W. Kelton, Kye Chisman, Daniel Ray,
and Blaine Emms. Testing of Thermocline Filler Materials and MoltenSalt Heat Transfer Fluids for Thermal Energy Storage Systems in Parabolic
Trough Power Plants. Solar Energy, (July):587–595, 2004.
[12] Francis Agyenim, Neil Hewitt, Philip Eames, and Mervyn Smyth. A review
of materials, heat transfer and phase change problem formulation for latent
heat thermal energy storage systems (LHTESS). Renewable and Sustainable
Energy Reviews, 14(2):615–628, February 2010.
[13] a Sharma, V Tyagi, C Chen, and D Buddhi. Review on thermal energy
storage with phase change materials and applications. Renewable and Sustainable Energy Reviews, 13(2):318–345, February 2009.
[14] E. Halawa and W. Saman. Thermal performance analysis of a phase change
thermal storage unit for space heating. Renewable Energy, 36(1):259–264,
January 2011.
[15] V. M. van Essen, H. a. Zondag, J. Cot Gores, L. P. J. Bleijendaal,
M. Bakker, R. Schuitema, W. G. J. van Helden, Z. He, and C. C. M.
Rindt. Characterization of MgSO4 Hydrate for Thermochemical Seasonal
Heat Storage. Journal of Solar Energy Engineering, 131(4):041014–1 –
041014–7, 2009.
152
Chapter 6
Hybrid Trigeneration - Thermally Activated Heat Pumps
[16] Franziska Schaube, Antje Wörner, and Rainer Tamme. High temperature
thermo-chemical heat storage for CSP using gas-solid reactions. In System.
[17] Bunsen Wong, Lloyd Brown, Franziska Schaube, Rainer Tamme, and Christian Sattler. Oxide based thermochemical heat storage. In Lloydia (Cincinnati), pages 1–8.
[18] R Weber and V Dorer. Long-term heat storage with NaOH. Vacuum,
82(7):708–716, March 2008.
[19] Chris Bales and Svante Nordlander. TCA Evaluation - Lab Measurements,
Modelling and Systems Simulations. Technical Report December, 2005.
[20] Maria Mottillo. Modelling an Adsorption System for the Seasonal Storage
of Heat in Residential Buildings. Technical report, Department of Building, Civil and Environmental Engineering, Concordia University, Montréal,
Canada, 2002.
[21] Mohammad Ameri, Ali Behbahaninia, and Amir Abbas Tanha. Thermodynamic analysis of a tri-generation system based on micro-gas turbine with
a steam ejector refrigeration system. Energy, 35(5):2203–2209, May 2010.
[22] H Li, L Fu, K Geng, and Y Jiang. Energy utilization evaluation of CCHP
systems. Energy and Buildings, 38(3):253–257, March 2006.
[23] Jiang-Jiang Wang, You-Yin Jing, and Chun-Fa Zhang. Optimization of
capacity and operation for CCHP system by genetic algorithm. Applied
Energy, 87(4):1325–1335, April 2010.
[24] Aiying Rong, Henri Hakonen, and Risto Lahdelma. An efficient linear
model and optimisation algorithm for multi-site combined heat and power
production. European Journal of Operational Research, 168(2):612–632,
January 2006.
Chapter 6
153
Hybrid Trigeneration - Thermally Activated Heat Pumps
[25] a Rong, R Lahdelma, and P Luh. Lagrangian relaxation based algorithm
for trigeneration planning with storages. European Journal of Operational
Research, 188(1):240–257, July 2008.
[26] K.C. Kavvadias, a.P. Tosios, and Z.B. Maroulis. Design of a combined
heating, cooling and power system: Sizing, operation strategy selection and
parametric analysis. Energy Conversion and Management, 51(4):833–845,
April 2010.
[27] N Fumo, P J Mago, and L M Chamra. Hybrid-cooling, combined cooling,
heating, and power systems. Proceedings of the Institution of Mechanical
Engineers, Part A: Journal of Power and Energy, 223(5):487–495, August
2009.
[28] S Tassou, I Chaer, N Sugiartha, Y Ge, and D Marriott. Application of
tri-generation systems to the food retail industry. Energy Conversion and
Management, 48(11):2988–2995, November 2007.
[29] S Acha, G Mavromatidis, V Caritte, and N Shah. Effective low-cost energy
saving strategies in supermarkets: An uk case study. Proceedings of the
26th International Conference on Efficiency, Cost, Optimization, Simulation and Environmental Impact of Energy Systems, ECOS 2013, 2013.
[30] G Mavromatidis, S Acha, and N Shah. Diagnostic tools of energy performance for supermarkets using artificial neural network algorithms. ENERGY AND BUILDINGS, 62:304–314, 2013.
[31] Sau Man Lai and Chi Wai Hui. Feasibility and flexibility for a trigeneration
system. Energy, 34(10):1693–1704, October 2009.
[32] Sau Man Lai and Chi Wai Hui. Integration of trigeneration system and
thermal storage under demand uncertainties. Applied Energy, 87(9):2868–
2880, September 2010.
154
Chapter 6
Hybrid Trigeneration - Thermally Activated Heat Pumps
[33] J Miguez. Feasibility of a new domestic CHP trigeneration with heat pump:
I. Design and development. Applied Thermal Engineering, 24(10):1409–
1419, July 2004.
[34] E. Cardona and a. Piacentino. A methodology for sizing a trigeneration
plant in mediterranean areas. Applied Thermal Engineering, 23(13):1665–
1680, September 2003.
[35] E Cardona, a Piacentino, and F Cardona. Matching economical, energetic
and environmental benefits: An analysis for hybrid CHCP-heat pump systems. Energy Conversion and Management, 47(20):3530–3542, December
2006.
[36] P.J. Mago and L.M. Chamra. Analysis and optimization of CCHP systems
based on energy, economical, and environmental considerations. Energy
and Buildings, 41(10):1099–1106, October 2009.
[37] P.J. Mago, L.M. Chamra, and J. Ramsay. Micro-combined cooling, heating
and power systems hybrid electric-thermal load following operation. Applied
Thermal Engineering, 30(8-9):800–806, June 2010.
[38] Piers Forster and Venkatachalam Ramaswamy. Changes in Atmospheric
Constituents and in Radiative Forcing. Technical report.
[39] Kyoto Protocol to the United Nations Framework Convention on Climate
Change, 1998.
[40] Directive 2006/40/EC of the European Parliament and of the Council, 2006.
[41] L Kairouani and E Nehdi. Cooling performance and energy saving of a
compressionabsorption refrigeration system assisted by geothermal energy.
Applied Thermal Engineering, 26(2-3):288–294, February 2006.
[42] Douglas M. Robinson and Eckhard a. Groll. Efficiencies of transcritical
Chapter 6
155
Hybrid Trigeneration - Thermally Activated Heat Pumps
CO2 cycles with and without an expansion turbine. International Journal
of Refrigeration, 21(7):577–589, November 1998.
[43] Jahar Sarkar and Souvik Bhattacharyya. Thermodynamic analyses and optimization of a transcritical N2O refrigeration cycle. International Journal
of Refrigeration, 33(1):33–40, January 2010.
[44] H Kruse and H Russmann. The natural fluid nitrous oxidean option as
substitute for low temperature synthetic refrigerants. International Journal
of Refrigeration, 29(5):799–806, August 2006.
[45] P Srikhirin. A review of absorption refrigeration technologies. Renewable
and Sustainable Energy Reviews, 5(4):343–372, December 2001.
[46] DeVault. Triple effect absorption chiller utilising two refrigeration circuits,
1993.
[47] Y Kaita. Simulation results of triple-effect absorption cycles. International
Journal of Refrigeration, 25(7):999–1007, November 2002.
[48] Leilei Wang, Shijun You, Huan Zhang, and Xianli Li. Simulation of gasfired triple-effect LiBr/water absorption cooling system with exhaust heat
recovery generator. Transactions of Tianjin University, 16(3):187–193, May
2010.
[49] G Grossman and R DeVault. Simulation and Performance Analysis of a
4-Effect Lithium Bromide-Water Absorption Chiller under Contract No .
DE-AC05-840R21400 Prepared by the Dl SCLAl MER Portions of this document may be illegible in electronic image products . Images are produced
fro. In Simulation, 2008.
[50] Yousef S H Najjar. Pergamon. Applied Thermal EngineeringThermal Engineering, 16(2), 1996.
156
Chapter 6
Hybrid Trigeneration - Thermally Activated Heat Pumps
[51] S Adewusi. Second law based thermodynamic analysis of ammonia?water
absorption systems. Energy Conversion and Management, 45(15-16):2355–
2369, September 2004.
[52] R DeVault. Seven-effect absorption refrigeation, 1989.
[53] R DeVault. Dual-circuit, multiple-effect refrigeration system and method,
1995.
[54] Hajizadeh Aghdam, F. Ranjbar, and S.M. Seyed Mahmoudi. Performance
Comparison of Triple-Effect Parallel Flow and Series Flow Absorption Refrigeration Systems. Journal of Applied Sciences, 8(16):2913–2918, December 2008.
[55] Hongxi Yin, Ming Qu, and David H. Archer. Model based experimental performance analysis of a microscale LiBrH2O steam-driven doubleeffect absorption Chiller. Applied Thermal Engineering, 30(13):1741–1750,
September 2010.
[56] Sam V Shelton, Central Park West, and Laura A Schaefer. Modelling and
Analysis of the Air Cooled Ammonia-Water Triple Effect Cycle. Technical
report, 1995.
[57] D Boer, M Valles, and A Coronas. Performance of double effect absorption compression cycles for air-conditioning using methanolTEGDME and
TFETEGDME systems as working pairs. International Journal of Refrigeration, 21(7):542–555, November 1998.
[58] H Lee, K Koo, S Jeong, J Kim, and Y Oh. Thermodynamic design data and
performance evaluation of the water + lithium bromide + lithium iodide
+ lithium nitrate + lithium chloride system for absorption chiller. Applied
Thermal Engineering, 20:707–720, 2000.
[59] C Park, J Koo, and Y Kang. Performance analysis of ammonia absorption
Chapter 6
157
Hybrid Trigeneration - Thermally Activated Heat Pumps
GAX cycle for combined cooling and hot water supply modes. International
Journal of Refrigeration, 31(4):727–733, June 2008.
[60] Y Kang. An advanced GAX cycle for waste heat recovery: WGAX cycle.
Applied Thermal Engineering, 19(9):933–947, September 1999.
[61] a. Rameshkumar, M. Udayakumar, and R. Saravanan. Heat transfer studies on a GAXAC (generator-absorber-exchange absorption compression)
cooler. Applied Energy, 86(10):2056–2064, October 2009.
[62] J Kim. Simulation of the compressor-assisted triple-effect H2O/LiBr absorption cooling cycles.
Applied Thermal Engineering, 22(3):295–308,
March 2002.
[63] Keith E Herold, Lawrence A Howe, and Reinhard Radermacher. Analysis
of a hybrid compression-absorption cycle using lithium bromide and water
as the working fluid. International Journal of Refrigeration, 14:264–272,
1991.
[64] R. Ventas, a. Lecuona, a. Zacarı́as, and M. Venegas. Ammonia-lithium nitrate absorption chiller with an integrated low-pressure compression booster
cycle for low driving temperatures. Applied Thermal Engineering, 30(1112):1351–1359, August 2010.
[65] R. Ayala, C.L. Heard, and F.a. Holland. Ammonia/lithium nitrate absorption/compression refrigeration cycle. Part II. experimental. Applied
Thermal Engineering, 18(8):661–670, August 1998.
[66] R. Ayala, C.L. Heard, and F.a. Holland. Ammonia/lithium nitrate absorption/compression refrigeration cycle. Part I. Simulation. Applied Thermal
Engineering, 17(3):223–233, March 1997.
[67] Y Kang. Recherche avancée : analyse de la performance des cycles hybrides
GAX (HGAX). International Journal of Refrigeration, 27(4):442–448, June
2004.
158
Chapter 6
Hybrid Trigeneration - Thermally Activated Heat Pumps
[68] W Worek. Enhancement of a double-effect absorption cooling system using
a vapor recompression absorber. Energy, 28(12):1151–1163, October 2003.
[69] Daniele Ludovisi, William M Worek, and Milton Meckler. Simulation of
a Double-Effect Absorber Cooling System Operating at Elevated Vapor
Recompression Levels Simulation of a Double-Effect Absorber Vapor Recompression Levels. HVAC&R Research, 12(3):533–547, 2006.
[70] Milton Meckler. Enhanced lithium bromide absorption cycle water vapor
recompression absorber.pdf, 1998.
[71] B Saha. Solar/waste heat driven two-stage adsorption chiller: the prototype. Renewable Energy, 23(1):93–101, May 2001.
[72] T.X. Li, R.Z. Wang, J.K. Kiplagat, L.W. Wang, and R.G. Oliveira. A conceptual design and performance analysis of a triple-effect solidgas thermochemical sorption refrigeration system with internal heat recovery. Chemical
Engineering Science, 64(14):3376–3384, July 2009.
[73] P. Muthukumar and M. Groll. Metal hydride based heating and cooling
systems: A review. International Journal of Hydrogen Energy, 35(8):3817–
3831, April 2010.
[74] Gustav Lorentzen. Revival of carbon dioxide as a refrigerant. International
Journal of Refrigeration, 17(January):292–302, 1994.
[75] Jahar Sarkar. Review on Cycle Modifications of Transcritical CO 2 Refrigeration and Heat Pump Systems. Advanced Research in Mechanical
Engineering, 1:22–29, 2010.
[76] J Sarkar, S Bhattacharyya, and M Gopal. Simulation of a transcritical
CO2 heat pump cycle for simultaneous cooling and heating applications.
International Journal of Refrigeration, 29(5):735–743, August 2006.
Chapter 6
159
Hybrid Trigeneration - Thermally Activated Heat Pumps
[77] Yang Chen. Thermodynamic Cycles using Carbon Dioxide as Working
Fluid. PhD thesis, 2011.
[78] Steven A Wright, Ross F Radel, Milton E Vernon, Gary E Rochau, and
Paul S Pickard. Operation and Analysis of a Supercritical CO2 Brayton
Cycle. Technical Report September, Sandia National Laboratories, 2010.
[79] J. Sarkar and Souvik Bhattacharyya. Optimization of recompression SCO2 power cycle with reheating. Energy Conversion and Management,
50(8):1939–1945, August 2009.
[80] Masanori Aritomi, Takao Ishizuka, Yasushi Muto, and Nobuyoshi Tsuzuki.
Performance Test Results of a Supercritical CO2 Compressor Used in a New
Gas Turbine Generating System. Journal of Power and Energy Systems,
5(1):45–59, 2011.
[81] Motoaki Utamura. Thermodynamic Analysis of Part-Flow Cycle Supercritical CO2 Gas Turbines. Journal of Engineering for Gas Turbines and
Power, 132(11):111701, 2010.
[82] Craig S Turchi. Supercritical CO 2 for Application in Concentrating Solar
Power Systems. In Supercritical CO2 Power Cycle Symposium, number
Figure 1, pages 1–5, 2009.
[83] H Yamaguchi, X Zhang, K Fujima, M Enomoto, and N Sawada. Solar
energy powered Rankine cycle using supercritical CO2. Applied Thermal
Engineering, 26(17-18):2345–2354, December 2006.
[84] Zhiwen Ma and Craig S Turchi. Advanced Supercritical Carbon Dioxide
Power Cycle Configurations for Use in Concentrating Solar Power Systems
Preprint. In Supercritical CO2 Power Cycle Symposium, number March,
2011.
[85] Li Jing, Pei Gang, and Ji Jie. Optimization of low temperature solar thermal
160
Chapter 6
Hybrid Trigeneration - Thermally Activated Heat Pumps
electric generation with Organic Rankine Cycle in different areas. Applied
Energy, 87(11):3355–3365, November 2010.
[86] J Sarkar. Optimization of a transcritical CO2 heat pump cycle for simultaneous cooling and heating applications. International Journal of Refrigeration, 27(8):830–838, December 2004.
[87] D Li and E Groll.
Transcritical CO2 refrigeration cycle with ejector-
expansion device. International Journal of Refrigeration, 28(5):766–773,
August 2005.
[88] J Nickl, G Will, H Quack, and W Kraus. Integration of a three-stage
expander into a CO2 refrigeration system. International Journal of Refrigeration, 28(8):1219–1224, December 2005.
[89] R. Chacartegui, J.M. Muñoz de Escalona, D. Sánchez, B. Monje, and
T. Sánchez. Alternative cycles based on carbon dioxide for central receiver solar power plants. Applied Thermal Engineering, 31(5):872–879,
April 2011.
[90] B Saha. Performance evaluation of a low-temperature waste heat driven
multi-bed adsorption chiller. International Journal of Multiphase Flow,
29(8):1249–1263, August 2003.
[91] Skander Jribi, Anutosh Chakraborty, Ibrahim I El-sharkawy, Bidyut Baran
Saha, and Shigeru Koyama. Thermodynamic Analysis of Activated CarbonCO 2 based Adsorption Cooling Cycles. Technical report, 2009.
[92] R Wang and R Oliveira. Adsorption refrigerationAn efficient way to make
good use of waste heat and solar energy. Progress in Energy and Combustion
Science, 32(4):424–458, 2006.
[93] By Shenyi Wu. INVESTIGATION OF EJECTOR RE-COMPRESSION.
PhD thesis, University of Nottingham, 1999.
Chapter 6
161
Hybrid Trigeneration - Thermally Activated Heat Pumps
[94] Simona Steiu, Daniel Salavera, Joan Carles Bruno, and Alberto Coronas. A
basis for the development of new ammoniawatersodium hydroxide absorption chillers. International Journal of Refrigeration, 32(4):577–587, June
2009.
[95] J Fernandezseara. Heat and mass transfer analysis of a helical coil rectifier
in an ammonia?water absorption system. International Journal of Thermal
Sciences, 42(8):783–794, August 2003.
[96] J Fernández-Seara. Study and control of the optimal generation temperature in NH3H2O absorption refrigeration systems. Applied Thermal Engineering, 21(3):343–357, February 2001.
[97] S. Deng, Y.J. Dai, R.Z. Wang, T. Matsuura, and Y. Yasui. Comparison
study on performance of a hybrid solar-assisted CO2 heat pump. Applied
Thermal Engineering, February 2011.
[98] Ding-Yu Peng and Donald B. Robinson. A New Two-Constant Equation
of State. Industrial & Engineering Chemistry Fundamentals, 15(1):59–64,
February 1976.
[99] JongSik Oh, Giridhari L. Agrawal, and Charles William BUckley. High
efficiency R744 refrigeration system and cycle, 2010.
[100] M Kim. Fundamental process and system design issues in CO2 vapor compression systems. Progress in Energy and Combustion Science, 30(2):119–
174, 2004.
[101] Jian Min Yin, Clark W Bullard, and Predrag S Hrnjak. R-744 gas cooler
model development and validation. International Journal of Refrigeration,
24:692–701, 2001.
[102] B.S. Petukhov. Heat Transfer and Friction in Turbulent Pipe Flow with
Variable Physical Properties. Advances in Heat Transfer, 6:503–564, 1970.
162
Chapter 6
Hybrid Trigeneration - Thermally Activated Heat Pumps
[103] P Simoes, J Fernandes, and J Mota. Dynamic model of a supercritical carbon dioxide heat exchanger. The Journal of Supercritical Fluids, 35(2):167–
173, September 2005.
[104] Masafumi Nakagawa, Ariel R. Marasigan, and Takanori Matsukawa. Experimental analysis on the effect of internal heat exchanger in transcritical
CO2 refrigeration cycle with two-phase ejector. International Journal of
Refrigeration, pages 1–10, March 2010.
[105] S Elbel and P Hrnjak. Experimental validation of a prototype ejector designed to reduce throttling losses encountered in transcritical R744 system
operation. International Journal of Refrigeration, 31:411–422, August 2007.
[106] Mark Holtzapple and Andrew Rabroker. Gerotor Apparatus for a QuasiIsothermal Brayton Cycle Engine, 2010.
[107] Mark Holtzapple and George Rabroker. Gerotor apparatus for a quasiisothermal Brayton cycle engine, 2010.
[108] James a. Mathias, Jon R. Johnston, Jiming Cao, Douglas K. Priedeman,
and Richard N. Christensen. Experimental Testing of Gerotor and Scroll
Expanders Used in, and Energetic and Exergetic Modeling of, an Organic
Rankine Cycle. Journal of Energy Resources Technology, 131(1):012201–
012210, 2009.
[109] Mark Holtzapple, George Rabroker, Michael Ross, Steven Atmur, and Gary
Noyes. Gerotor apparatus for a quasi-isothermal Brayton cycle engine, 2010.
[110] Huijuan Chen, D. Yogi Goswami, and Elias K. Stefanakos. A review of
thermodynamic cycles and working fluids for the conversion of low-grade
heat. Technical report, July 2010.
[111] Jia Yan, Daniel C. H. Yang, and Shih-Hsi Tong. A New Gerotor Design
Method With Switch Angle Assignability. Journal of Mechanical Design,
131(1):011006, 2009.
Chapter 6
163
Hybrid Trigeneration - Thermally Activated Heat Pumps
[112] Giovanni Mimmi, Università Pavia, Meccanica Strutturale, and Via Ferrata. On Design and Performance of Internal Epitrochoidal Pumps. In
ASME International Design Engineering Technical Conferences, pages 1–8,
2006.
[113] Yii-Wen Hwang and Chiu-Fan Hsieh. Geometric Design Using Hypotrochoid and Nonundercutting Conditions for an Internal Cycloidal Gear.
Journal of Mechanical Design, 129(4):413, 2007.
[114] G Bonandrini, G Mimmi, and C Rottenbacher. Theoretical analysis of internal epitrochoidal and hypotrochoidal machines. In Proceedings of the
Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science, volume 223, pages 1469–1480, June 2009.
[115] C-F Hsieh. Non-undercutting region and property evaluation of epitrochoidal gerotor geometry. Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science, 224(2):473–481,
January 2010.
[116] D C H Yang, J Yan, and S-H Tong. Flowrate formulation and displacement analyses for deviation function-based gerotor pumps. Proceedings of
the Institution of Mechanical Engineers, Part C: Journal of Mechanical
Engineering Science, 225:480–487, January 2010.
[117] Chiu-Fan Hsieh. Influence of Gerotor Performance in Varied Geometrical
Design Parameters. Journal of Mechanical Design, 131(12), 2009.
[118] Great Bnutin and Mechanical Engineering. Geometry for Trochoidal-Type
Machines with Conjugate Envelopes. Mechanism and Machine Theory,
29(I):25–42, 1994.
164
Chapter
Appendix A
R744 Heat Pump Cycle Early
Concepts
165
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure A.1: Process flow diagram of R744 heat pump cycle concept ‘Cycle A’
Figure A.2: Process flow diagram of R744 heat pump cycle concept ‘Cycle B’
166
Chapter A
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure A.3: Process flow diagram of R744 heat pump cycle concept ‘Cycle C’
Figure A.4: Process flow diagram of R744 heat pump cycle concept ‘Cycle D’
Chapter A
167
Hybrid Trigeneration - Thermally Activated Heat Pumps
Figure A.5: Process flow diagram of R744 heat pump cycle concept ‘Cycle E’
168
Chapter A
Appendix B
R744 Centrifugal Compressor
Impeller Blade Geometry
169
Hybrid Trigeneration - Thermally Activated Heat Pumps
Velocity triangles are given below.
Figure B.1: Impeller outlet velocity triangle
Figure B.2: Impeller inlet velocity triangle
The output of the computational code used to calculate the velocity triangles
is given below.
170
R$ = ‘R744’
(B.1)
Nblade = 6
(B.2)
Nsplitter = 6
(B.3)
N = 50000
(B.4)
b1 = 0.01
(B.5)
R1 = 0.015
(B.6)
β2,b = 60
(B.7)
η = 0.8
(B.8)
Chapter B
Hybrid Trigeneration - Thermally Activated Heat Pumps
m = 10
(B.9)
P01 = 150
(B.10)
P03 = 250
(B.11)
T01 = 350
(B.12)
ρ01 = ρ (R$, T = T01 , P = P01 )
(B.13)
S01 = s (R$, T = T01 , P = P01 )
(B.14)
H01 = h (R$, T = T01 , P = P01 )
(B.15)
S03,is = S01
(B.16)
H03,is = h (R$, S = S03,is , P = P03 )
(B.17)
ρ03 = ρ (R$, T = T03 , P = P03 )
(B.18)
S03 = s (R$, T = T03 , P = P03 )
(B.19)
H03 = H01 +
H03,is − H01
η
(B.20)
T03 = T (R$, H = H03 , P = P03 )
(B.21)
W ork in = (H03 − H01 ) · 1000
(B.22)
W ork in
U2 =
Chapter B
1.04·σ
Cw,2
(B.23)
R2 = U2 /ω
(B.24)
U1 = ω · R1
(B.25)
Q1 = m/ρ01
(B.26)
ω = 2 · 3.1415 · N/60
(B.27)
171
Hybrid Trigeneration - Thermally Activated Heat Pumps
(R1 )2 − (R1 − b1 )2
A1,annulus = 3.1415 ·
A1,volute = 3.1415 ·
(R1 )2 − (R1 − b1 )2 · (sin (α1 ))
(B.28)
(B.29)
A1,volute,check = 3.1415 · b1 · (sin (α1 )) · (Cos(3.1415/N )) · ((2 · R1 ) − b1 ) (B.30)
A2 = 2 · 3.1415 · R2 · b2
(B.31)
A2,volute = 2 · 3.1415 · R2 · b2 · (Cos(β2 ))
(B.32)
∆P = P03 − P01
(B.33)
∆T = T03 − T01
(B.34)
∆H = H03 − H01
(B.35)
Head = δP · 9.81
(B.36)
(m/ρ01 )0.5
N s = (2 · 3.1415 · N/60) ·
(∆H · 1000)0.75
(B.37)
T1,1 = T01
(B.38)
P1,1 = P01
(B.39)
ρ1,1 = ρ01
(B.40)
Ca,1,1 =
m
ρ1,1 · A1,annulus
Cp1,2 = cp (R$, T = T1,1 , P = P1,1 )
(B.42)
Cv 1,2 = cv (R$, T = T1,1 , P = P1,1 )
(B.43)
γ1,2 =
T1,2 = T01 −
P1,2 = P01 ·
Cp1,2
Cv1,2
(B.44)
2
Ca,1,1
2 · Cp1,2 · 1000
172
(B.41)
T1,2
T1,1
(B.45)
M A12 !
GAM
γ
−1
1,2
(B.46)
Chapter B
Hybrid Trigeneration - Thermally Activated Heat Pumps
ρ1,2 = ρ (R$, T = T1,2 , P = P1,2 )
Ca,1,2 =
m
ρ1,2 · A1,annulus
Ca,1 = Ca,1,2
q
(B.48)
(B.49)
2
+ (U12 )
Ca,1
(B.50)
α1 = arctan (Ca,1 /U1 )
(B.51)
Cr,2 = Ca,1
(B.52)
Cw,2,b = U2 − Cr,2 · tan (β2,b )
!
p
Cos(β2,b )
σ =1−
(Nsplitter + Nblade )0.7
(B.53)
Cw,2 = σ · Cw,2,b
(B.55)
V1 =
C2 =
q
2
2
+ Cr,2
Cw,2
q
2
(U2 − Cw,2 )2 + Cr,2
U2 − Cw,2
β2 = arctan
Cr,2
V2 =
ηc = 0.8
ηi =
Chapter B
(B.47)
√
ηc
(B.54)
(B.56)
(B.57)
(B.58)
(B.59)
(B.60)
Cp2 = cp (R$, P = P02 , T = T02 )
(B.61)
Cv 2 = cv (R$, P = P02 , T = T02 )
(B.62)
γ2 = Cp2 /Cv 2
(B.63)
∆H imp,is = ηi · ∆H
(B.64)
H02,is = H01 + ∆H imp,is
(B.65)
173
Hybrid Trigeneration - Thermally Activated Heat Pumps
S02,is = S01
(B.66)
P02 = P (R$, H = H02,is , S = S02,is )
(B.67)
T02 = T03
C22
T2 = T02 −
2 · Cp2 · 1000
γ2 P2 = P02 · (T2 /T02 ) γ2 −1
(B.68)
(B.69)
ρ2 = ρ (R$, T = T2 , P = P2 )
(B.71)
A2 =
m
(ρ2 · Cr,2 )
Calculate b2!
(B.70)
(B.72)
Solution
α1 = 24.24 [deg]
A1,annulus = 0.0006283 [m2 ]
A1,volute = 0.0002579 [m2 ]
A1,volute,check = 0.0002579 [m2 ]
A2 = 0.0004941 [m2 ]
A2,volute = 0.0002004 [m2 ]
β2 = 66.07 [deg]
β2,b = 60
b1 = 0.01 [m]
b2 = 0.001963 [m]
Cp2 = 2.087 [kJ/kgK]
Cv 2 = 0.9024 [kJ/kgK]
C2 = 134.8 [m/s]
Ca,1 = 35.36 [m/s]
Cr,2 = 35.36 [m/s]
Cw,2 = 130.1 [m/s]
Cw,2,b = 148.6 [m/s]
∆H = 24.87 [kJ/kg]
∆Himp,is = 22.24 [kJ/kgK]
∆P = 100 [bar]
∆T = 34.2 [K]
η = 0.8
ηc = 0.8
ηi = 0.8944
γ2 = 2.312 [-]
Head = 981 [m]
H01 = 403.4 [kJ/kg]
H02,is = 425.6 [kJ/kg]
H03 = 428.2 [kJ/kg]
H03,is = 423.3 [kJ/kg]
174
Chapter B
Hybrid Trigeneration - Thermally Activated Heat Pumps
m = 10 [kg/s]
N = 50000
N s = 0.3945 [kg0.75 -m1.5 /s0.5 -kJ0.75 ] Nblade = 6
Nsplitter = 6
ω = 5236
P01 = 150 [bar]
P02 = 263 [bar]
P03 = 250 [bar]
P2 = 257.8 [bar]
Q1 = 0.02226 [m3 /s]
R$ = ‘R744’
ρ01 = 449.2 [kg/m3 ]
ρ03 = 540.4 [kg/m3 ]
ρ2 = 572.4 [kg/m3 ]
R1 = 0.015 [m]
R2 = 0.04007 [m]
σ = 0.8758
S01 = 1.601 [kJ/kgK]
S02,is = 1.601 [kJ/kgK]
S03 = 1.614 [kJ/kgK]
S03,is = 1.601 [kJ/kgK]
T01 = 350 [K]
T02 = 384.2 [K]
T03 = 384.2 [K]
T2 = 379.8 [K]
U1 = 78.54 [m/s]
U2 = 209.8 [m/s]
V1 = 86.13 [m/s]
V2 = 87.18 [m/s]
W ork in = 24866 [W]
Chapter B
175
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement